ArticlePDF Available

Abstract and Figures

Vertebrate populations at the periphery of their range can show pronounced genetic drift and isolation, and therefore offer unique challenges for conservation and management. These populations are often candidates for management actions such as translocations that are designed to improve demographic and genetic integrity. This is particularly true of coldwater species like brook trout (Salvelinus fontinalis), whose numbers have declined greatly across its historic range. At the southern margin, remnant wild populations persist in isolated headwater streams, and many have a history of receiving translocated individuals through either stocking of hatchery reared fish, relocation of wild fish, or both during restoration attempts. To determine current genetic integrity and resolve the genetic effects of past management actions for brook trout populations in SC, USA, we genetically assessed all 18 documented remaining brook trout populations along with individuals acquired from six hatcheries with recorded stocking events in SC. Our results indicated that six of the 18 streams showed signs of hatchery admixture (range 57–97%) and restored patches retained genetic signatures from multiple source populations. Populations had among the lowest genetic diversity (min average HE = 0.147) and effective number of breeders (mean Nb = 31.2) estimates observed throughout the native brook trout range. Populations were highly differentiated (mean pair-wise FST = 0.396), and substantial genetic divergence was evident across major river drainages (max pair-wise FST = 0.773). The lowest local genetic diversity and highest genetic differentiation ever reported for this species make its conservation a challenging task, particularly when combined with other threats such as climate change and non-native species. We offer recommendations on managing peripheral populations with depleted genetic characteristics and provide a reference for determining which existing populations will best serve as sources for future translocation efforts aimed at enhancing or restoring wild brook trout genetic integrity.
This content is subject to copyright. Terms and conditions apply.
1 3
Conservation Genetics
Characterizing genetic integrity ofrear-edge trout populations
inthesouthern Appalachians
KaseyC.Pregler1,2· YoichiroKanno1,2· DanielRankin3· JasonA.Coombs4,5· AndrewR.Whiteley6
Received: 22 September 2017 / Accepted: 12 October 2018
© Springer Nature B.V. 2018
Vertebrate populations at the periphery of their range can show pronounced genetic drift and isolation, and therefore offer
unique challenges for conservation and management. These populations are often candidates for management actions such
as translocations that are designed to improve demographic and genetic integrity. This is particularly true of coldwater spe-
cies like brook trout (Salvelinus fontinalis), whose numbers have declined greatly across its historic range. At the southern
margin, remnant wild populations persist in isolated headwater streams, and many have a history of receiving translocated
individuals through either stocking of hatchery reared fish, relocation of wild fish, or both during restoration attempts. To
determine current genetic integrity and resolve the genetic effects of past management actions for brook trout populations in
SC, USA, we genetically assessed all 18 documented remaining brook trout populations along with individuals acquired from
six hatcheries with recorded stocking events in SC. Our results indicated that six of the 18 streams showed signs of hatchery
admixture (range 57–97%) and restored patches retained genetic signatures from multiple source populations. Populations
had among the lowest genetic diversity (minaverage HE = 0.147) and effective number of breeders (mean Nb = 31.2) esti-
mates observed throughout the native brook trout range. Populations were highly differentiated (mean pair-wise FST = 0.396),
and substantial genetic divergence was evident across major river drainages (max pair-wise FST = 0.773). The lowest local
genetic diversity and highest genetic differentiation ever reported for this species make its conservation a challenging task,
particularly when combined with other threats such as climate change and non-native species. We offer recommendations
on managing peripheral populations with depleted genetic characteristics and provide a reference for determining which
existing populations will best serve as sources for future translocation efforts aimed at enhancing or restoring wild brook
trout genetic integrity.
Keywords Appalachian mountains· Effective population size· Genetic drift· Admixture· Microsatellite· Translocation
Widespread evidence indicates that the modern rates
of extinction in plants and animals exceed background
rates in the fossil record (Burkhead 2012). In the north-
ern hemisphere, southern populations of vertebrates have
Electronic supplementary material The online version of this
article (https :// 2-018-1116-1) contains
supplementary material, which is available to authorized users.
* Kasey C. Pregler
1 Department ofForestry andEnvironmental Conservation,
Clemson University, Clemson, SC29631, USA
2 Department ofFish, Wildlife, andConservation Biology,
Colorado State University, FortCollins, CO80523-1474,
3 South Carolina Department ofNatural Resources, Clemson,
SC29631, USA
4 Department ofEnvironmental Conservation, University
ofMassachusetts Amherst, Amherst, MA01003, USA
5 USDA Forest Service, Northern Research Station, University
ofMassachusetts, Amherst, MA01003, USA
6 Wildlife Biology Program, Department ofEcosystem
andConservation Sciences, University ofMontana,
Missoula, MT59812, USA
Conservation Genetics
1 3
suffered disproportionately (Hoffmann etal. 2010; Evans
etal. 2016) and this trend is expected to accelerate due
to future environmental changes (Currie 2001; Davis and
Shaw 2001). Declining populations at the southern edge
of their range are often relegated to small, isolated habi-
tat, which can have important demographic and genetic
consequences. Cessation of gene flow following isolation
can lead to a loss of genetic variation and pronounced
genetic divergence through genetic drift (Allendorf 1986;
Lesica and Allendorf 1995; Hampe and Petit 2005). Thus,
rear-edge populations have often received intensive man-
agement actions (Evans etal. 2016). For example, trans-
locations of individuals are used to mitigate the impacts
of decline (Griffith etal. 1989; Fischer and Lindenmayer
2000), where individuals are moved to establish repre-
sentative, replicate, and resilient populations throughout
the species’ historical range (IUCN 1987; Armstrong and
Seddon 2007).
While translocations can be particularly important for
species conservation, they can have variable success (Fis-
cher and Lindenmayer 2000; Weeks etal. 2011; Redford
etal. 2011). Translocations typically move individuals from
one or more source populations into existing populations or
vacant habitat (Weeks etal. 2011). It is important to bal-
ance the potential benefits of such translocations against
trade-offs and risks that could counteract conservation goals
(Weeks etal. 2011). Despite extensive research on transloca-
tions, the majority of these studies have focused on birds and
mammals (Brichieri-Colombi and Moehrenschlager 2016;
Hampe and Petit 2005; Fischer and Lindenmayer 2000).
Translocations are common in fishes, but freshwater habitats
have been far less studied in this regard (however see Huff
etal. 2010, 2011; Robinson etal. 2017).
Freshwater habitats occupy < 1% of the Earth’s surface,
yet are hotspots that support ~ 10% of all known species and
1/3 of vertebrate species (Dudgeon etal. 2006). Further-
more, in addition to translocations for conservation, fishes
have been extensively stocked with hatchery-raised conspe-
cifics for recreational angling (Laikre etal. 2010; Gozlan
etal. 2010). For instance, stocking is common for cold-water
species like members of the Salmonidae family (Haak etal.
2010). Salmonids represent an important natural resource
globally, and are one of the most angled fish species in the
United States (Maillett and Aiken 2015). In particular, rain-
bow trout (Oncorhyncus mykiss), brown trout (Salmo trutta)
and brook trout (Salvelinus fontinalis) have expanded their
distribution due to stocking (Gozlan etal. 2010). Histori-
cally, these hatchery fish were stocked with little regard for
the genetic make-up of or potential interactions with wild
populations. Hybridization with hatchery fish may result in
outbreeding depression and loss of local adaptation in wild
populations (Rhymer and Simberloff 1996; Currens etal.
1997; Allendorf etal. 2001; Araki etal. 2007), increasing
the likelihood of decreased fitness.
Brook trout are an iconic native salmonid in eastern North
America occupying habitat from Canada to the southern
Appalachians. Populations have declined throughout its
native range, particularly at the southern range (Hudy etal.
2008) due to competition with invasive species (Larson and
Moore 1985; Dewald and Wilzbach 1992), habitat loss, and
climate change (Meisner 1990; Curry and MacNeill 2004;
Wenger etal. 2011). Brook trout are typically restricted
to small headwater streamsat the southern range, which
makes their isolated populations challenging to conserve
(Hudy etal. 2008). In order to compensate for the decline
in wild fish populations, hatchery reared brook trout have
been stocked since the late 1800s (Krueger and Menzel
1979). Hatchery produced brook trout that are derived from
northeastern US stocks are used in the majority of stock-
ing efforts because southern brook trout were challenging
to raise in hatcheries (Lennon 1967). Evidence of genetic
swamping by hatchery fish in wild populations has been
documented throughout the brook trout range (Krueger and
Menzel 1979; McCracken etal. 1993; Hayes etal. 1996;
Marie etal. 2010; Lamaze etal. 2012), and conserving and
restoring populations with native genotypes is a manage-
ment priority in southern populations (Habera and Moore
2005). These anthropogenic alterations can have deleteri-
ous effects on a species’ genetic integrity. Here integrity is
defined asconditions that have little to no anthropogenic
influence, such thatnatural evolutionary and biogeographic
processes are allowed to occur (Angermeier and Karr
1994).While factors that influence genetic integrity have
been studied in parts of the northern range (Castric etal.
2001; Marie etal. 2010; Kanno etal. 2011; Annett etal.
2012; Fraser etal. 2014; Kelson etal. 2015), few studies
exist for the southern-most range of brook trout (Stoneking
etal. 1981; Hayes etal. 1996; Kazyak etal. 2018). This is
of concern because the southern portion of the range is the
most anthropogenically influenced as evidenced by a 50%
reduction in current distribution compared to historic occur-
rence (Hudy etal. 2008; Whiteley etal. 2014b). In addition
to hatchery stocking, translocations of wild individuals have
occurred as well. These wild translocations have included
efforts to improve genetic integrity of extant populations
(Robinson etal. 2017), and to reestablish populations in his-
torical habitats (Kanno etal. 2016).
Here we present a case study of rear-edge brook trout
populations at the southern-most limit in SC, USA, that have
been subject to both historical stocking events for recrea-
tional fishery and contemporary translocation actions for
conservation. Whether translocations or stocking, records
of fish movement have not always been kept and their demo-
graphic and genetic consequences are rarely monitored
(however see Robinson etal. 2017), presenting challenges
Conservation Genetics
1 3
for conserving small, isolated populations. This has created
a need for an updated assessment of genetic diversity and
integrity for the documented remaining wild populations in
SC. We addressed the following questions: (1) what were
the genetic effects of past management practices such as
translocations and hatchery stocking events? (2) are southern
genotype signatures still present in these populations, and
how is that variation partitioned within and among remain-
ing populations? Answers to these questions were then used
to guide future conservation efforts at the southern edge of
the range for brook trout.
Study area
This study was conducted in mountain streams of SC, USA,
located at the southern distribution of brook trout, the only
salmonid native to this region (Fig.1). In North America,
introduced rainbow trout have displaced many populations
of native brook trout (Habera and Moore 2005; Kanno etal.
2016). Introduced rainbow trout typically occupy habitat
downstream, whereas brook trout occur upstream, often
only above physical barriers that block upstream migration
for rainbow trout (Larson and Moore 1985). Natural popu-
lations of brook trout in the state are currently restricted
to the 18 documented streams used in this study. Through-
out the paper, we refer to study streams as patches. A patch
represents a spatially continuous headwater network, within
which movement of individuals was assumed in the absence
of physical barriers, and follows the definition of the Eastern
Brook Trout Joint Venture (Whiteley etal. 2014b; EBTJV
2016). Historic brook trout stocking in SC occurred from
the 1800s to approximately the 1970s. In the 1990s, stocking
resumed in response to a demand by anglers and has con-
tinued in areas where stocked fish cannot access the extant
brook trout streams due to physical barriers to movement.
In the 1990s, the SC Department of Natural Resources
(SC-DNR) initiated an allozyme analysis to determine if
hatchery admixture was present in remnant brook trout
populations. The allozyme study assessed 11 out of the 18
patches (Bad, Crane, Emory, Falls, Headforemost, Indian
Camp, Ira, Jacks, Matthews, Pig Pen, and Slicking) included
in the present study. They identified evidence of two patches
replaced by hatchery genotypes, and two southern genotype
Fig. 1 Map of eastern United States with current brook trout range
shaded in gray and locations of hatchery samples in stars (a). Previ-
ous translocation efforts in SC brook trout patches (b). Blue lines rep-
resent translocations of southern fish, dashed blue lines represent fish
that were moved after the source patch was restored. Red lines rep-
resent translocations of fish with identified hatchery admixture based
on the 1990s allozyme results. (Color figure online)
Conservation Genetics
1 3
patches. The remaining patches were an admixture of hatch-
ery and southern genotypes (Table1) (Guffey 1993). Man-
agement actions followed to conserve and restore southern
brook trout populations, including removal of hybridized
fish and subsequent translocation events. Most translocation
efforts aimed at moving individuals from southern popula-
tions to restore and enhance brook trout populations at other
locations (i.e., conservation focus), but some moved hybrid-
ized fish to other sites primarily for providing recreational
angling opportunities (e.g., translocations of known hybrid-
ized fish to Tammassee and Carrick Creeks; Fig.1). While
the allozyme study informed management actions, the des-
ignation of southern (native) versus hatchery samples relied
exclusively on a single diagnostic southern allele at the CK-
A2* locus (Guffey 1993). This single-locus analysis lacked
statistical power and, more importantly, used a locus that
might not have been diagnostic across the southern range
(Hayes etal. 1996). Accordingly, the present multi-locus
microsatellite study was intended to more accurately assess
the current genetic characteristics of brook trout populations
at the southern distributional margin.
Field sampling
Genetic samples were collected between July 2014 and
July 2016. Using backpack electrofishing, a total of 1111
young-of-the-year (YOY) brook trout were captured from
18 patches in upstate SC in the Santee and Savannah River
drainages (Fig.1). Our goals were to determine spatial
population structure, degree of hatchery admixture, and to
estimate single-cohort effective number of breeders (Nb)
per patch. We sampled YOY (defined as less than 100mm
in total length) to conduct all of these analyses following
recommendations in Whiteley etal. 2013. During summer,
YOY can be identified based on length frequency histo-
grams. Since sampling strategy (the number of individuals
and locations where they are collected) can influence esti-
mates of Nb through family representation effects (Whiteley
etal. 2012), each patch was divided into a lower, middle,
and upper section with a goal of 25 YOY sampled from
each section (i.e., 75 YOY per patch). Smaller patches only
contained two sections whereas one large patch (Matthews
Creek) contained five sections.
Hatchery samples (ranging from 30 to 46 individuals per
hatchery) were obtained from six hatcheries (Berlin Hatch-
ery, New Hampshire; Burton State Fish Hatchery, Georgia;
Table 1 Allozyme results from the unpublished SC-DNR hatchery admixture study (Guffey 1993) and the subsequent management actions and
translocation events that took place from 2005 to 2010 during the restoration of Crane, King, Laurel Fork, and Moody Creeks
NA values for allozyme results indicates that a given patch was not included in the study conducted in the 1990s
Allozyme results Hatchery admixed fish and invasives removed Received translocations from Year of
Savannah drainage
Bad Creek Hybrid
Carrick Creek NA Received hatchery admixed fish from Crane Restored King and Crane 2006
Crane Hybrid Hatchery admixed fish removed Jacks, Slicking, and 3 creeks in Georgia 2006
Indian Camp Hybrid
Ira Branch Hybrid
Jacks Branch Native
King NA Brown trout removed Jacks, and 1 creek in Georgia and North Caro-
Laurel Fork NA Restored King 2010
Moody NA Restored King 2008
Pig Pen Hybrid
Tammassee NA Received hatchery admixed fish from Crane
Santee drainage
Emory Creek Hybrid Restored Crane and King 2012
Falls Creek Hybrid
Headforemost Hybrid
Laurel Creek NA
Matthews Hatchery in origin
Slicking Native
South Saluda NA
Conservation Genetics
1 3
Dry Mill Hatchery, Maine; Governor Hill Hatchery, Maine;
Walhalla Hatchery, SC; Wytheville Hatchery, Virginia)
along the East Coast. These hatcheries were chosen based on
historic stocking records from D.C. Booth Historic National
Fish Hatchery and Archives (Spearfish, South Dakota) and
anecdotal records in SC (D. Rankin, SC-DNR, unpublished
data). Stocking has not occurred in these stream patches
since the 1970s. Our inference is based on the assumption
that our hatchery samples accurately represent genetic com-
position of the hatchery populations at the time of stocking.
Individuals were measured for total length and weight,
and anal or caudal fin tissue was collected non-lethally as a
DNA source. Samples were genotyped using 12 microsatel-
lite markers: SfoC113, SfoC88, SfoD100, SfoD75, SfoC24,
SfoC115, SfoC129, SfoB52, SfoC86, SfoD91a, SfoC38 (King
etal. 2003) and SsaD237 (King etal. 2005), following pro-
tocols for DNA extraction and amplification detailed in King
etal. (2005). Loci were electrophoresed on an ABI Prism
3130xl genetic analyzer (Applied Biosystems Inc., Foster
City, CA), and alleles were hand-scored using Geneious ver-
sion 7.0.6 (Kearse etal. 2012).
Statistical analysis
Single cohort samples with large numbers of siblings from
the same family can cause deviations from Hardy–Weinberg
(HW) expectations, elevated linkage disequilibrium (LD),
and bias in genetic structure analyses (Allendorf and Phelps
1981; Anderson and Dunham 2008; Rodriguez-Ramilo and
Wang 2012; Whiteley etal. 2013; Waples and Anderson
2017). We performed all analyses, with the exception of
Nb (for which all individuals were always included), with
the entire dataset and with a ‘sib-purged’ dataset. The sib-
purged data set was a truncated dataset where we randomly
sampled full-siblings within families (Waples and Anderson
2017). We followed the ‘yank-2’ procedure of Waples and
Anderson (2017), where we included all individuals from
estimated full-sib families of size one and two, and randomly
selected two individuals from families of size three or larger.
This approach is expected to provide a balance between
sib over-representation and the effects of reduced sample
size from more severe ‘sib-purging’ (Waples and Ander-
son 2017). Full-sibling family structure was determined for
each patch using COLONY V2 (Wang 2004). Settings for
COLONY analyses included the assumption of male and
female monogamy, outbreeding model, very long run length
with the full-likelihood model, sibship prior, and no allele
frequency updates. Finally, we used a paired t-test to evalu-
ate if summary statistic results were significantly different
between the full and sib-purged datasets.
Since our stream patches were determined a-priori with-
out any genetic information, and barriers to gene flow may
be present in these patches, we assessed genetic structuring
within patches. If within-structure was found, patches were
split. To investigate within patch structure, first we ran
genetic clustering analyses using STRU CTU RE version 2.3.4
(Pritchard etal. 2000), and then calculated within patch genic
differentiation tests. STRU CTU RE runs were conducted hier-
archically, with a first run to identify hatchery admixture by
including all 18 SC brook trout patches and six hatcheries.
A second STRU CTU RE run was performed on only the18
SC patches to identify genetic clustering among wild popula-
tions. To verify spatial structuring of southern genotype fish,
a third STRU CTU RE run was performed where we omitted
any patches with evidence of hatchery admixture greater than
20 percent, as identified after the first and second runs. All
STRU CTU RE runs were performed using 20,000 burn-in
and 100,000 iterations, with five replicates for each value
of K. We used the admixture model, with correlated allele
frequencies, and no location prior. We tested K = 1–24 for the
patch set that included hatcheries and we tested K = 1–18 for
the patch set that excluded hatcheries. The number of clusters
were determined by visually inspecting the likelihood plots.
STRU CTU RE results were visualized by creating bar plots in
the program STRU CTU RE PLOT (Ramasamy etal. 2014).
We also used a discriminate analysis of principle compo-
nents (DAPC) using the adegenet package (Jombart 2008) in
program R (R Development Core Team 2008). DAPC was
performed where clusters were determined using the find.
cluster function for k ranging from 1 to 24 (SC patches and
hatcheries dataset), and 1–18 (SC patches only). Clusters
were evaluated using Bayesian Information Criterion (BIC)
to determine the appropriate k value, where the k with the
lowest BIC value is typically the optimal number of clus-
ters. However, BIC values may keep decreasing after the true
k value (Jombart etal. 2010), so we visually examined the
rate of decrease in BIC to identify values of k, after which
BIC values decreased only slightly (Jombart etal. 2010). A
DAPC analysis was performed for each grouping using the
dapc function for the best k identified as described above.
We retained all axes of the principal component analysis to
explain the variation within the data, and created an ordi-
nation plot with the first and second axes to visualize the
clusters. To further investigate within patch structure, we cal-
culated pairwise FST values among sections within patches,
and performed genic tests using Fisher’s exacts tests for all
pairs of sections in GENEPOP. We applied the B-Y FDR
(Benjamini and Yekutieli 2001) correction method to genic
tests following Narum (2006).
After we redefined patches, genotype data were analyzed
using GENEPOP (Raymond and Rousset 1995) to gener-
ate summary statistics for genetic diversity, specifically the
number of alleles and the proportion of observed (HO) and
expected (HE) heterozygotes. We investigated non-random
mating in our patches through assessing the conformity of
loci to HW proportions, and linkage disequilibrium (LD)
Conservation Genetics
1 3
to assess non-random association of alleles among loci.
Given the small and isolated study patches, we estimated
FIS for each patch to assess the departure of observed from
expected heterozygosity (Keller and Waller 2002). We
estimated genetic differentiation with pairwise FST values
among patches, and performed genic tests using Fisher’s
exacts tests for all pairs of patches in GENEPOP and applied
the B-Y FDR correction. Lastly, we performed a Mantel test
(Mantel 1967) in GENALEX (Peakall and Smouse 2005)
to investigate associations between FST and geographic
distances across all patches. Statistical significance for the
Mantel test was obtained by using 9999 permutations and a
p-value of 0.05.
The effective number of breeders (Nb) was estimated
for each stream patch using the linkage disequilibrium
method in program NeEstimator 2.01 (Do etal. 2013) using
a monogamous mating system and a minimum allele fre-
quency of 0.02, following the method of Whiteley etal.
(2013). We used the monogamy mating model because
brook trout appear to conform more closely to monogamy
than random mating (Coombs 2010). Nb is calculated as an
equivalent to effective population size (Ne) when working
with samples from a single cohort for species with genera-
tional overlap and estimates the effective number of breeders
that gave rise to that cohort (Waples 2005; Waples and Do
2010). All individuals were used to calculate Nb since family
structure forms the basis for this estimation, and sib-purg-
ing causes upward bias in estimates (Waples and Anderson
2017). Since studies have observed positive relationships
among patch size, genetic diversity and effective population
size (Whiteley etal. 2010; Peacock and Dochtermann 2012)
we used a simple linear regression model to examine the
effect of patch size on estimates of genetic variation within
patches and Nb. We also analyzed the relationship between
Nb and genetic variation within patches, and used a Pearson’s
correlation to examine if our sample sizes were correlated
with Nb.
COLONY identified 485 families across the 18 SC patches,
and 51% were single fish families (mean family size = 2.28).
Letcher etal. (2011) demonstrated high accuracies of sibship
reconstruction in COLONY based on the same 12 micros-
atellite loci used in the present paper. Reconstructed full-
sibling families composed of at least two individuals had a
rate of correct family inference of 91.2% (0.7% SE) and for
full-sibling families of at least five individuals the correct
family inference was 97.7% (0.4% SE) (Letcher etal. 2011).
Furthermore, it is common to have close to 50% singleton
families for brook trout (Whiteley etal. 2014a), especially
when sampling the population to minimize full-sibling
overrepresentation (Whiteley etal. 2012). Sibship removal
reduced our mean patch sample size from ~ 62 for all indi-
viduals to ~ 39 for the siblings purged dataset. We present
genetic clustering results with the all individuals dataset due
to the trade-offs between accounting for family effects and
having large enough sample size for genetic structuring anal-
yses (Patterson etal. 2006; Anderson and Dunham 2008).
However results between the full and sibling-purged dataset
were similar and we reached the same conclusions with both.
Our first STRU CTU RE run, which included genotype
data from all 18 SC patches as well as the six hatcheries,
revealed four genetic clusters. One of these clusters had six
out of the 18 SC patches assigned with the hatchery sam-
ples which presented the first evidence that these patches
may have hatchery admixture (Fig. S1). The second STRU
CTU RE run excluding hatchery samples showed that the
18 stream patches made up five genetic clusters, exhibiting
further subdivision in comparison to the first STRU CTU
RE run. We again had one cluster characterized with those
six patches that may have hatchery admixture (Fig.2), and
the four remaining clusters which may consist of southern
genotype fish. The third STRU CTU RE run, which removed
those patches with evidence of hatchery admixture, corrobo-
rated the genetic clustering assignment of the second run and
showed four genetic clusters of southern fish.
STRU CTU RE and DAPC results were concordant, and
the k-means clustering BIC and DAPC also found four clus-
ters when 18 SC patches and six hatcheries were analyzed
(Fig. S2). Once hatcheries were removed, BIC and DAPC
results provided further evidence for five distinct clusters
(Fig.3) and DAPC cluster assignment (TableS1) was similar
to the STRU CTU RE results. While hatchery admixture was
present in six out of 18 patches, three of those six patches
appeared to consist almost entirely of hatchery-descendant
fish (Bad, Matthews, and Tammassee) (range 0.891–0.963),
whereas the remaining three patches (Carrick, Emory, and
Indian Camp) had varying proportions of admixture (range
0.255–0.542) (TableS2 and S3). Several patches (i.e. Car-
rick, Emory, and Indian Camp) showed genetic assignment
to multiple clusters and even to clusters from different river
drainages (Fig.4). The lower and middle sections of Car-
rick could be entire replacement by hatchery descendants.
Emory and Indian Camp exhibited hatchery admixture
across all sections sampled within each of these patches.
The DAPC plot (Fig.3) showed evidence of a river drainage
effect where axis 1 aligned wild brook trout clusters from
an east–west gradient. Clusters belonging to Savannah (blue
and green) and Santee (yellow and purple) river drainages
separated from one another. Further structuring occurred
within each of these drainages. Within the Santee there were
clusters representing the middle and south Saluda basins,
while the Savannah River drainage was structured into two
clusters (blue and purple).
Conservation Genetics
1 3
We detected evidence for admixture among wild popula-
tions that was likely associated with previous management
actions (translocations) conducted in the early 2000s. In
the STRU CTU RE bar plots (Figs. S1 and S2), Crane and
Carrick both retained genetic signatures from previous
translocations of the source populations such as the Santee
drainage (yellow) fish that were moved into Crane, and the
Savannah drainage (blue) fish that were moved into Carrick.
Emory exhibited Savannah drainage (blue), hatchery (red),
and Santee drainage (yellow) genetic signatures. The San-
tee genetic signature present in Emory was likely whatever
remnant population remained prior to receiving translocated
fish from King and Crane.
There was little to no fine scale genetic structuring
within the majority of SC brook trout patches, with the
exception of Carrick and Laurel Creeks. FST values were
greater than 0.2 in sections within Carrick Creek and
Laurel Creek (Fig. S3). Using Fisher’s method for com-
bined p-values across all loci for genic differentiation tests
(B-Y FDR correction, p < 0.0119), we found lower Carrick
was significantly different (p < 0.0001) from the middle
(FST = 0.11) and upper (FST = 0.33) sections, but the mid-
dle and upper were not significantly different from each
other (p = 0.0768, FST = 0.20). Individuals from upper Car-
rick were assigned to patches with the apparent southern
genotype signature in the STRU CTU RE plots, with several
possible downstream migrants in middle Carrick (Figs.
S1 and S2). STRU CTU RE q-score estimates revealed that
eight individuals in middle Carrick were assigned to the
blue Savannah drainage cluster where individual q-scores
ranged from 0.611 to 0.943 (mean = 0.782). Documenta-
tion of past management actions included the transloca-
tion of individuals from restored King and Crane Creeks
into upper Carrick. During sample collection we noted
Fig. 2 STRU CTU RE likelihood plot (a) and bar plot (b) for data including only the 18 South Carolina brook trout patches. Bar plot for K = 5.
Solid black lines separate patches, dashed black lines represent approximate within patch sample divisions ordered from downstream to upstream
Conservation Genetics
1 3
the presence of a waterfall barrier between middle and
upper Carrick sections. These two points taken together
could explain the genetic differentiation observed with
putative downstream migrants likely responsible for the
non-significant results between upper and middle Carrick.
In Laurel Creek, the largest FST values were observed
when comparing the lower section of Laurel Creek to the
middle (FST = 0.265) and upper sections (FST = 0.245).
Based on Fisher’s method for combined genic tests (B-Y
FDR correction, p < 0.0117), we found that the lower sec-
tion was significantly different from the middle and upper
sections (p < 0.001), but the middle and upper sections
were not significantly different from each other (p = 0.020;
FST = 0.043). Laurel Creek has multiple waterfalls between
the lower and middle sections, as well as a steep bedrock
slide between the middle and upper sections that may be dif-
ficult for fish to ascend. All Laurel Creek fish clustered with
fish of southern genotype from the Saluda drainage (yellow)
in the STRU CTU RE analysis. Based on these results, we
assigned individuals in Carrick and Laurel Creeks to within
patch origin prior to estimation of within-population genetic
diversity summary statistics and Nb.
Within-population genetic diversity was low at many
study patches. Expected heterozygosity (HE) ranged from
0.147 (Slicking Creek) to 0.667 (Emory Creek) (Table2).
Mean allelic richness ranged from 1 (Headforemost and
Jacks Branch) to 7 (lower and middle Carrick Creek)
(Table2). Across patches, mean FIS was 0.022 and ranged
from − 0.141 to 0.325 (Table2). High positive FIS values
were present in Slicking (0.325) and Falls Creek (0.141).
Positive FIS values can be due to family structure or a Wahl-
und effect caused by population substructure (Wright 1951).
However there doesn’t appear to be a clear biological cause
of the variation in FIS values, particularly because sib-purged
samples didn’t change FIS much which can suggest that fam-
ily structure isn’t responsible. Following sequential Bonfer-
roni correction, tests of deviation from HW proportions
were significant in 3% of cases (6 of 240 tests, p < 0.0002)
(Table2). Prior to correction for multiple tests, deviations
from HW proportions were significant (p < 0.05) in 24% (58
of 240) of cases (Table2). Following sequential Bonferroni
correction, tests for linkage disequilibrium were significant
in 9% of cases (113 of 1320 tests, p < 0.00004) (Table2).
Prior to correction for multiple tests, 28% (367 of 1320) of
tests for linkage disequilibrium were significant (p < 0.05).
We found that by-locus summary statistics calculated with
the all individuals dataset (TableS4) and the siblings purged
dataset (TableS5) did not statistically differ (p > 0.05), with
the exception of HE (p = 0.005) and LD (p = 0.007). Sib-
purging decreased evidence of LD in these patches.
Genetic differentiation varied substantially among
patches (Table3). Pairwise FST values ranged from 0.020
Fig. 3 Bayesian information criterion (BIC) plot (a) and discriminant
analysis of principal components (DAPC) plot (b) for K = 5 clusters.
Cluster 1 (green) represents falls, and headforemost. Cluster 2 (blue)
represents Crane, King, Laurel Fork, and Moody. Cluster 3 (red) rep-
resents the SC brook trout patches that have hatchery admixture (Bad,
Carrick, Emory, Indian Camp, Matthews, and Tammassee). Cluster 4
(purple) represents Ira Branch, and Pig Pen. Cluster 5 (yellow) repre-
sents Laurel Creek, Slicking, and South Saluda. (Color figure online)
Conservation Genetics
1 3
(Laurel Fork and Moody) to 0.773 (Ira Branch and Slick-
ing), with a mean pairwise FST value of 0.396 across all
comparisons. Large pairwise FST values, such as observed
in Ira Branch and Slicking, could be due to spatial struc-
ture by major river basins. We followed up this obser-
vation with a hierarchical AMOVA that grouped identi-
fied southern genotype patches (n = 12) into two groups
representing each major basin (Savannah and Santee).
AMOVA results showed evidence for spatial structure by
basin in which among basin groups accounted for 20%
of the variance, among patches within groups accounted
for 30%, and within patches accounted for the remaining
50%. When comparing patches with evidence of hatchery
admixture, FST values ranged from 0.083 (Matthews and
lower and middle Carrick) to 0.309 (Bad and Tammassee)
with a mean of 0.2202. Samples from hatcheries had rela-
tivelylow pairwise FST estimates, and ranged from 0.051
(Berlin and Walhalla) to 0.303 (Wytheville and Dry Mill)
with a mean of 0.193. Hatcheries have an anecdotal history
of swapping strains among other hatcheries so it is not a
surprise that the hatchery samples aremore genetically
similar. However, there are no known published records
of how much trading has occurred among hatcheries (but
see Kazyak etal. 2018). Genic differentiation tests showed
that all pair-wise patch comparisons were significant using
Fisher’s method for combined p-values across all loci (B-Y
FDR correction, p < 0.0057). Mantel test results revealed a
correlation of r = 0.251 and p-value = 0.0004 across popu-
lations illustrating a relationship between differentiation
and geographic distance.
Point estimates of effective number of breeders (Nb)
varied from 5.0 (upper Carrick Creek) to 116.1 (Matthews
Creek) and patch size ranged from 84ha (lower and mid-
dle Carrick Creek) to 1413ha (Matthews Creek) (Table2).
Confidence intervals for Nb were very wide for patches
with low genetic diversity (i.e. Ira, Jack, Slicking, South
Saluda) (TableS6). There was a significant (p < 0.001)
positive relationship (effect size = 0.077) between patch
size and Nb (Fig.5). However, this was driven by Mat-
thews Creek, and when it was omitted there was not a
significant relationship (p = 0.054). There also was not a
significant relationship between genetic diversity (HE) and
patch size (p = 0.357) (Fig. S4), nor between HE and Nb
(p = 0.885) (Fig. S5). Pearson’s correlation between patch
sample size and Nb revealed a correlation coefficient of
0.42 (p = 0.071), which again appears to be driven by
Fig. 4 STRU CTU RE results that illustrate five clusters of brook trout
patches in SC. Cluster 1 (green) represents Falls, and Headforemost.
Cluster 2 (blue) represents Crane, King, Laurel Fork, and Moody.
Cluster 3 (red) represents the SC Brook trout patches that have hatch-
ery admixture (Bad, Carrick, Emory, Indian Camp, Matthews, and
Tammassee). Cluster 4 (purple) represents Ira Branch, and Pig Pen.
Cluster 5 (yellow) represents Laurel Creek, Slicking, and South
Saluda. (Color figure online)
Conservation Genetics
1 3
Matthews Creek. After Matthews is omitted the correla-
tion coefficient is 0.064 (p = 0.798).
Genetic characteristics ofsouthern populations
Our brook trout populations at the southern margin con-
tained a mix of southern and northern-admixed genetic
signatures. Six out of the 18 study patches exhibited signs
of admixture from past stocking events. Three of those
six patches showed high amounts of hatchery admixture
and could be entirely hatchery descendants (Bad Creek,
Matthews, and Tammassee; range 0.891–0.963), and 12
patches did not show signs of hatchery influences. We can-
not definitively exclude the possibility of missing a hatch-
ery source that was used for stocking. While over time
genetic drift may have made these populations more dif-
ferentiated from the individuals that were initially stocked,
our genetic structuring results still showed evidence of
admixture. Our review of historic records and anecdotal
evidence was exhaustive and the six hatcheries used in
this study should represent historical sources of hatchery
samples. Thus, these 12 populations are likely of southern
descent and have persisted at the southern edge despite
small habitat size and historical stocking efforts.
Genetic diversity in the southern patches was low rela-
tive to other studies of brook trout. In fact, these patches
have some of the lowest genetic diversity recorded
(mean HE= 0.442, range 0.147–0.712) when compared to
existing microsatellite-based studies from other parts of
the brook trout range (across studies mean = 0.617; range
0.190–0.797, see TableS7 for study-specificestimates)
(Castric etal. 2001; Kanno etal. 2011; Annett etal. 2012;
Whiteley etal. 2013; Hoxmeier etal. 2015; Kelson etal.
2015). Low genetic diversity is a common trait of rear edge
populations (Davis and Shaw 2001; Hampe and Petit 2005).
We also observed very high pairwise FST values (mean =
0.396, range 0.020–0.773) among patches, suggesting a
great deal of genetic drift has occurred in these isolated
patches. Studies of similar spatial scale of < 100km report
mean FST = 0.124 (range 0.096–0.159) among brook trout
sites (Whiteley etal. 2013). Rear edge populations can
exhibit disproportionately high levels of genetic differentia-
tion among populations, even between nearby ones (Hampe
and Petit 2005).
Table 2 Summary statistics for SC brook trout patches averaged across all 12 loci
Table represents number of individuals (n) from each patch, and statistics include proportion of expected (HE) and observed (HO) heterozygotes,
allelic richness (A), inbreeding coefficient (FIS), number of significant (p < 0.0002) deviations from Hardy–Weinberg (HW) following sequential
Bonferroni correction, and significant (p < 0.00004) instances of linkage disequilibrium (LD) following sequential Bonferroni correction, and
effective number of breeders (Nb). Results are presented for all individuals, and for thedataset with siblings removed (noted with a ) with the
exception of Nb which was only calculated for all individuals
Bad Creek (n = 55, n = 36) 0.529 0.533 0.534 0.495 3 3 0.043 0.111 0 0 4 0 22.3
Carrick Creek (n = 43, n = 22) (lower and middle) 0.639 0.686 0.636 0.621 7 7 0.075 0.108 0 0 13 4 6.5
Carrick Creek (n = 23, n = 4) (upper) 0.634 0.617 0.702 0.687 4 3 − 0.110 − 0.152 0 0 2 0 5
Crane Creek (n = 94, n = 56) 0.587 0.580 0.588 0.588 5 5 0.009 0.014 1 0 22 8 22
Emory Creek (n = 84, n = 49) 0.667 0.679 0.672 0.683 6 6 − 0.005 − 0.005 0 0 32 9 16.8
Falls Creek(n = 75, n = 47) 0.332 0.368 0.299 0.340 3 3 0.121 0.101 0 0 3 3 19.1
Headforemost (n = 82, n = 64) 0.205 0.203 0.209 0.206 1 1 − 0.024 − 0.016 0 0 0 0 19
Indian Camp (n = 46, n = 28) 0.613 0.623 0.596 0.613 5 5 0.029 0.025 0 0 7 0 34.5
Ira Branch (n = 46, n = 28) 0.252 0.255 0.297 0.294 2 2 − 0.148 − 0.137 0 0 0 0 38.1
Jacks Branch (n = 30, n = 19) 0.208 0.220 0.247 0.258 1 1 − 0.141 − 0.145 0 0 0 0 18.7
King (n = 85, n = 54) 0.553 0.563 0.514 0.526 5 5 0.073 0.070 0 0 14 2 23.8
Laurel Creek (n = 30, n = 18, 20) (lower) 0.255 0.270 0.238 0.259 1 2 0.038 0.022 0 0 0 0 Infinite
Laurel Creek (n = 37, n = 23) (middle and upper) 0.272 0.292 0.270 0.278 2 2 0.006 0.037 0 0 0 0 11.7
Laurel Fork (n = 37, n = 20) 0.498 0.506 0.527 0.520 4 4 − 0.058 − 0.030 1 0 1 0 12.1
Matthew (n = 104, n = 91) 0.614 0.623 0.627 0.636 6 7 − 0.010 − 0.018 0 0 3 1 116.1
Moody (n = 35, n = 22) 0.411 0.458 0.422 0.465 3 3 − 0.033 − 0.031 0 0 2 0 14.1
Pig Pen (n = 63, n = 38) 0.500 0.518 0.519 0.532 4 4 − 0.047 − 0.029 1 0 7 1 53.6
Slicking (n = 46, n = 33) 0.147 0.148 0.121 0.123 1 1 0.325 0.312 1 1 0 0 56.9
South Saluda (n = 39, n = 30) 0.282 0.287 0.252 0.263 2 2 0.068 0.045 0 0 0 0 50.6
Tammassee (n = 57, n = 36) 0.505 0.503 0.539 0.525 3 3 − 0.069 − 0.042 0 0 2 1 25.9
Conservation Genetics
1 3
Table 3 Pairwise FST (below diagonal) values, and number of significant exact tests for genic differentiation (above diagonal) following B-Y FDR correction for multiple tests (p = 0.047) for 18
brook trout patches in SC, USA, and six hatcheries
Patch Bad Carrick
(lower and
Crane Emory Falls Headforemost Indian Ira Jacks King Laurel
Laurel Creek
(middle and
Bad 12 12 12 12 12 12 12 12 12 12 12 12
(lower and
0.243 12 12 12 12 12 12 12 12 12 12 12
0.338 0.211 8 10 12 11 10 11 11 10 11 12
Crane 0.358 0.251 0.057 11 12 11 12 11 11 11 11 11
Emory 0.266 0.141 0.101 0.143 12 11 11 11 11 11 11 12
Falls 0.495 0.412 0.436 0.445 0.347 12 12 12 12 12 11 12
Headforemost 0.591 0.508 0.498 0.480 0.406 0.473 12 10 10 11 11 10
Indian 0.314 0.179 0.177 0.207 0.137 0.347 0.449 12 12 12 12 11
Ira 0.535 0.466 0.469 0.397 0.391 0.657 0.701 0.417 11 11 11 11
Jacks 0.555 0.423 0.425 0.383 0.341 0.593 0.642 0.332 0.702 11 11 11
King 0.369 0.243 0.136 0.145 0.167 0.417 0.420 0.188 0.441 0.173 11 11
Laurel Creek
0.500 0.459 0.451 0.453 0.321 0.588 0.700 0.418 0.720 0.694 0.472 9
Laurel Creek
(middle and
0.524 0.437 0.440 0.447 0.284 0.547 0.646 0.410 0.710 0.660 0.441 0.244
Laurel Fork 0.416 0.277 0.131 0.118 0.187 0.492 0.542 0.217 0.516 0.287 0.045 0.548 0.530
Matthews 0.233 0.083 0.262 0.281 0.168 0.436 0.491 0.217 0.445 0.438 0.291 0.465 0.447
Moody 0.435 0.302 0.173 0.151 0.211 0.516 0.576 0.242 0.558 0.308 0.052 0.581 0.559
Pig Pen 0.407 0.256 0.273 0.258 0.222 0.461 0.542 0.214 0.410 0.408 0.258 0.537 0.517
Slicking 0.569 0.487 0.491 0.494 0.303 0.573 0.670 0.479 0.773 0.735 0.490 0.508 0.452
South Saluda 0.513 0.410 0.409 0.443 0.281 0.487 0.594 0.375 0.706 0.636 0.422 0.463 0.388
Tammassee 0.309 0.198 0.348 0.363 0.286 0.444 0.571 0.250 0.534 0.512 0.356 0.523 0.534
Berlin H. 0.225 0.168 0.360 0.376 0.260 0.532 0.640 0.298 0.568 0.593 0.390 0.588 0.584
Burton H. 0.275 0.141 0.216 0.246 0.143 0.429 0.499 0.197 0.432 0.432 0.257 0.464 0.439
Dry Mill H. 0.294 0.242 0.283 0.306 0.218 0.482 0.588 0.267 0.518 0.523 0.341 0.469 0.480
Governor Hill
0.287 0.203 0.381 0.397 0.299 0.519 0.623 0.322 0.579 0.589 0.411 0.579 0.578
Walhalla H. 0.257 0.145 0.350 0.365 0.261 0.512 0.618 0.288 0.519 0.576 0.383 0.566 0.567
Wytheville H. 0.292 0.210 0.410 0.425 0.314 0.558 0.643 0.353 0.600 0.638 0.441 0.618 0.609
Conservation Genetics
1 3
Table 3 (continued)
Patch Laurel Fork Matthews Moody Pig Pen Slicking South Saluda Tammassee Berlin H. Burton H. Dry Mill H. Governor
Hill H.
Walhalla H. Wytheville H.
Bad 12 12 12 12 12 12 12 12 12 12 12 10 11
(lower and
12 12 12 12 12 12 12 11 12 12 12 12 12
11 12 12 11 11 11 12 12 12 12 12 12 12
Crane 11 12 11 12 11 12 12 12 12 12 12 12 12
Emory 10 12 11 12 11 12 12 12 11 12 12 12 12
Falls 12 12 12 12 11 11 11 12 12 12 12 12 12
Headforemost 10 12 11 11 9 11 12 12 12 12 12 12 12
Indian 11 12 11 12 12 12 12 12 12 12 12 12 12
Ira 10 12 11 11 11 11 12 12 12 12 12 12 12
Jacks 10 12 10 11 11 10 12 12 12 12 12 12 12
King 9 12 8 12 11 12 12 12 12 12 12 12 12
Laurel Creek
11 12 10 12 6 7 12 12 12 12 12 12 12
Laurel Creek
(middle and
11 12 11 12 9 8 12 12 12 12 12 12 12
Laurel Fork 12 5 10 11 11 12 12 11 12 12 12 12
Matthews 0.315 11 12 12 12 12 10 12 12 11 12 12
Moody 0.020 0.347 10 11 11 12 12 11 12 11 12 12
Pig Pen 0.282 0.290 0.312 12 11 12 12 12 12 12 12 12
Slicking 0.596 0.480 0.631 0.589 7 12 12 12 12 12 12 12
South Saluda 0.516 0.427 0.550 0.509 0.404 12 12 12 12 12 12 12
Tammassee 0.402 0.280 0.422 0.344 0.583 0.520 12 12 11 12 12 12
Berlin H. 0.425 0.099 0.463 0.398 0.647 0.568 0.300 12 12 9 10 12
Burton H. 0.285 0.130 0.322 0.280 0.502 0.430 0.288 0.183 12 12 12 12
Dry Mill H. 0.349 0.224 0.381 0.370 0.545 0.478 0.313 0.231 0.188 12 10 11
Governor Hill
0.444 0.175 0.478 0.437 0.628 0.551 0.313 0.078 0.220 0.257 11 10
Walhalla H. 0.414 0.114 0.448 0.388 0.624 0.551 0.303 0.051 0.201 0.258 0.080 9
Wytheville H. 0.490 0.171 0.527 0.450 0.661 0.585 0.372 0.117 0.251 0.303 0.155 0.143
Conservation Genetics
1 3
Estimates of Nb in study patches were low, indicative of
small adult population size and/or limited spawning and
rearing habitat (Whiteley etal. 2012; Fraser etal. 2014).
Estimates of Nb were < 100 in all patches except Matthews
Creek (Nb = 116), which was the largest patch in the study
area. Variation in Nb estimates across sites has been linked to
differences in habitat variability and quality (Belmar-Lucero
etal. 2012; Whiteley etal. 2013; Ruzzante etal. 2016). In
patches with low Nb estimates relative to their patch size
(e.g., Laurel Fork; Nb = 12.1, patch size = 482ha), conser-
vation actions could be pursued to increase Nb by habitat
improvement and/or removal of downstream interspecific
competitors like non-native rainbow trout to increase habitat
size. Low Nb might also suggest that these populations have
reduced adaptive potential (Lande and Barrowclough 1987;
Weeks etal. 2011; Fraser etal. 2014).
Despite extensive stocking and presumably an effect of
genetic drift, there was evidence of a river drainage effect
with distinct clusters in the Savannah and Santee river
basins. Major river drainages in SC have had long-term tem-
poral isolation due to lack of glacial modification (Rohde
etal. 2009). As a result, populations can be highly differen-
tiated between major drainages. In contrast, northern brook
trout have relatively lower genetic differentiation among
populations (Davis and Shaw 2001) given those geographic
regions were recolonized by brook trout following glacial
occupation (Danzmann etal. 1998). Similar north–south pat-
terns of genetic diversity have been documented in other fish
species (Bernatchez and Wilson 1998). Given the amount of
genetic divergence we observed among patches, there is the
potential that themanagement actions of translocating fish
across major drainage boundaries could negativelyimpact
the recipient population. Our results (high divergence and
genetic isolation) provide an opportunity to evaluate the
risks and trade-offs with past conservation actions (Weeks
etal. 2011).
Conservation andmanagement
Common conservation practices may not apply to rear edge
populations in part due to the combination of being severely
impoverished in genetic diversity and lack of gene flow
among populations in comparison to core-range populations
(Hampe and Petit 2005). As such, conservation strategies
need to be designed that consider unique aspects such as lim-
ited habitat and low productivity, especially since conserv-
ing the genetic integrity of rear edge populations requires
strategies that maintain the greatest possible number of local
populations, and connectivity among them. However, this
can be of little use at rear edges particularly when inva-
sive species are present. For instance, improving trout patch
size and connectivity is typically important to increase Nb
(Whiteley etal. 2013) and population size. Given that many
brook trout patches in SC are located above a barrier with
invasive rainbow trout below, it is very difficult to expand
and connect previous habitat due to tradeoffs between iso-
lation and competition with invasives (Fausch etal. 2009).
Despite the risk of local extirpation if left isolated, isolation
may be a necessity to preserve thesebrook trout populations.
For situations where habitat expansion through increased
Fig. 5 Relationship between
effective number of breeders
(Nb) and patch area (hectares)
(p < 0.001) for brook trout
patches in SC. Gray shading
represents 95% confidence
Conservation Genetics
1 3
connectivity are not feasible, more aggressive management
actions such as translocations become important.
Translocations are apotential management tool to facili-
tate genetic rescue or restoration for southern brook trout.
Genetic rescue has recently gained traction in conservation
of imperiled species like the Florida panther (Hedrick 1995)
due to its ability to restore genetic diversity and reduce
extinction risk by increasing a population’s absolute fit-
ness through an increase in population size or growth rate
(Whiteley etal. 2015). A recent study provides evidence
for a positive effect of genetic rescue on genetic diversity,
body size, and population growth rate through the first gen-
eration in Virginia brook trout populations (Robinson etal.
2017). However, given that the streams we examined are
very unproductive systems due to underlying geology and
high acid deposition rates (Cada etal. 1987; Kulp and Moore
2005), population carrying capacity is inherently limited and
potential for increased population size and growth rate may
not be an expected outcome. However, even if an increase in
absolute fitness is unlikely for the populations we examined,
an increase in relative fitness and genetic diversity (genetic
restoration; Hedrick 2005) by translocating just a few indi-
viduals per generation into these low diversity patches could
greatly benefit them by increasing evolutionary potential and
ability to adapt to future environmental changes (Whiteley
etal. 2015; Nathan etal. 2017).
Historical conservation approaches involved translocating
individuals from nearby sources due to the risk that more
distant populations may be locally adapted to their envi-
ronment and translocations can potentially reduce fitness
through outbreeding depression (Tallmon etal. 2004). In
this situation, we recommend the following approach for SC
brook trout patches. In our DAPC analysis, because Savan-
nah drainage (blue) and Santee drainage (yellow) clusters
are distinct from hatchery strains (red), they may be good
candidates for restoration, whereas green and purple are
closer spatially on the DAPC plot to the red cluster. Yel-
low patches may be better sources when restoring patches
in Santee River drainage, and blue patches are better suited
for restoring patches in the Savannah River drainage. This
should be taken with the caveat that drift could have resulted
in the green and purple clusters being closer to the hatchery
cluster by chance, and could still be their own distinct groups
given that sub-basin groups in the southern range may be
long isolated. However, it is important to balance and con-
sider the risks of outbreeding depression against the risks
that low genetic diversity and inbreeding depression pose
to the long-term persistence of a population (Weeks etal.
2011). Little research has directly tested how geographically
or genetically distant source populations can be from receiv-
ing populations without negatively impacting fitness.
Depending on the number of fish removed, translocations
also have potential to negatively impact source populations
(Armstrong and Seddon 2007), which are typically small in
size in the case of southern brook trout. Reciprocal trans-
locations have been demonstrated as an option to improve
isolated and bottlenecked populations (Heber etal. 2013;
Pavlova etal. 2017). An alternative approach could be the
use of multiple source populations to minimize the negative
demographic and genetic impacts on source populations,
particularly when translocating to restored habitats with
extirpated populations. Benefits of using multiple source
populations have been documented in species of mammals
and plants (Bodkin etal. 1999; Kirchner etal. 2006) such
as increased population growth and genetic diversity. Addi-
tionally, we may need to identify some source populations
outside of SC. Patches like King and Crane Creeks, which
received translocations from multiple source populations
including streams from outside of SC, seem to have ben-
efitted from this type of restoration. We lack a baseline for
comparison of initial genetic diversity prior to restoration,
but these two patches have maintained the highest genetic
diversity relative to the other wild brook trout patches in SC.
However we have not measured the effects of these translo-
cations on fitness, so while these populations exhibited high
genetic diversity, outbreeding depression may still have been
a factor (Frankham etal. 2011). Given the current status of
brook trout populations in SC in terms of number, census
size, and genetic diversity levels, a more regional effort may
be needed to restore these populations at the southern-most
aspect of the brook trout range.
Future climate change scenarios predict increased popula-
tion fragmentation and isolation, and further reductions in
patch size (Hudy etal. 2008; Wenger etal. 2011). Under-
standing mechanisms of population persistence for small
populations, such as headwater brook trout, in an uncertain
and changing climate is vital and necessitates a greater com-
prehension of how translocations can be most effectively
used to maintain genetic diversity. Maintenance of genetic
diversity in these extant populations is critical to their future
potential for adaptive response to environmental changes
because records of extirpations and extinctions suggest that
limits to adaptation are greatest during periods of rapid cli-
mate change (Davis and Shaw 2001).
Overall this study provides insight into the genetic
integrity of brook trout populations at the periphery of
their range. Persistence of small populations with south-
ern genetic signature is encouraging, but their low genetic
diversity and lack of gene flow call for active management
actions such as translocations. The use of translocations to
mitigate anthropogenic impacts on biodiversity is increasing
(Brichieri-Colombi and Moehrenschlager 2016), but there
is little effort devoted to genetic monitoring post-release
(Laikre etal. 2010) (however see Robinson etal. 2017).
Translocation success should be measured based on ben-
efits on receiving populations and negative effects on source
Conservation Genetics
1 3
populations. Ideally, knowledge generated from monitoring
can further feed into conservation actions to maintain local
genetic diversity in small, isolated populations for their long-
term persistence.
Acknowledgements This study was financially supported by the South-
east Aquatic Resources Partnership, Trout Unlimited, Duke Energy,
and the South Carolina Department of Natural Resources (SC-DNR).
We thank a number of SC-DNR fisheries biologists and volunteers who
conducted field sampling, as well as the Greenville Water Company for
access to field sites. Two anonymous reviewers provided constructive
comments that improved an earlier version of this manuscript.
Allendorf FW (1986) Genetic drift and the loss of alleles versus het-
erozygosity. Zoobiology 5:181–190
Allendorf FW, Phelps SR (1981) Use of allelic frequencies to describe
population structure. Can J Fish Aquat Sci 38:1507–1514
Allendorf FW, Leary RF, Spruell P, Wenburg JK (2001) The problems
with hybrids: setting conservation guidelines. Trends Ecol Evol
Anderson EC, Dunham KK (2008) The influence of family groups
on inferences made with the program STRU CTU RE. Mol Ecol
Resour 8:1219–1229
Angermeier PL, Karr JR (1994) Biological integrity versus biologi-
cal diversity as policy directives: protecting biotic resources. In:
Ecosystem Management. Springer, New York, NY, pp. 264–275
Annett B, Gerlach G, King TL, Whiteley AR (2012) Conservation
genetics of remnant coastal brook trout populations at the south-
ern limit of their distribution: population structure and effects of
stocking. Trans Am Fish Soc 141:1399–1410
Araki H, Cooper B, Blouin MS (2007) Genetic effects of captive breed-
ing cause a rapid cumulative fitness decline in the wild. Science
Armstrong DP, Seddon PJ (2007) Directions in reintroduction biology.
Trends Ecol Evol 23:20–25
Belmar-Lucero S, Wood SLA, Scott S, Harbicht AB, Hutchings JA,
Fraser DJ (2012) Concurrent habitat and life history influences
on effective/census population size ratios in stream-dwelling
brook trout. Ecol Evol 2:562–573
Benjamini Y, Yekutieli D (2001) The control of the false discovery rate
in multiple testing under dependency. Ann Stat 29:1165–1188
Bernatchez L, Wilson CC (1998) Comparative phylogeography of
Nearctic and Palearctic fishes. Mol Ecol 7:431–452
Bodkin JL, Ballachey BE, Cronin MA, Scribner KT (1999) Population
demographics and genetic diversity in remnant and translocated
populations of sea otters. Conserv Biol 13:1378–1385
Brichieri-Colombi TA, Moehrenschlager A (2016) Alignment of threat,
effort, and perceived success in North American conservation
translocations. Conserv Biol 30:1159–1172
Burkhead NM (2012) Extinction rates in North America freshwater
fishes, 1900–2010. Bioscience 62:798–808
Cada GF, Loar JM, Sale MJ (1987) Evidence of food limitation of rain-
bow and brown trout in southern Appalachian soft-water streams.
Trans Am Fish Soc 116:692–702
Castric V, Bonney F, Bernatchez L (2001) Landscape structure and
hierarchical genetic diversity in the brook charr, Salvelinus fon-
tinalis. Evolution 55:1016–1028
Coombs JA (2010) Reproduction in the wild: the effect of individual
life history strategies on population dynamics and persistence.
University of Massachusetts Amherst, Dissertation
Currens KP, Hemmingsen AR, French RA, Buchanan DV, Schreck
CB, Li HW (1997) Introgression and susceptibility to disease
in a wild population of rainbow trout. N Am J Fish Manag
Currie DJ (2001) Projected effects of climate change on patterns of
vertebrate and tree species richness in the coterminous United
States. Ecosystems 4:216–225
Curry RA, MacNeill WS (2004) Population-level responses to sediment
during early life in brook trout. J N Benthol Soc 23:140–150
Danzmann RG, Morgan IIRP, Jones MW, Bernatchez L, Ihssen PE
(1998) A major sextet of mitochondrial DNA phylogenetic
assemblages extant in eastern North American brook trout
(Salvelinus fontinalis): distribution and postglacial dispersal
patterns. Can J Zool 76:1300–1318
Davis MB, Shaw RG (2001) Range shifts and adaptive responses to
quaternary climate change. Science 292:673–679
Dewald L, Wilzbach MA (1992) Interactions between native brook
trout and hatchery brown trout: effects on habitat use, feeding,
and growth. Trans Am Fish Soc 121:287–296
Do C, Waples RS, Peel D, Macbeth GM, Tillet BJ, Ovenden JR (2013)
NeEstimator V2: re-implementation of software for the estima-
tion of contemporary effective population size (Ne) from genetic
data. Mol Ecol Resour 14:209–214
Dudgeon D, Arthington AH, Gessner MO, Kawabata ZI, Knowler DJ,
Lévêque C, Naiman RJ, Prieur-Richard AH, Soto D, Stiassny
ML, Sullivan CA (2006) Freshwater biodiversity: importance,
threats, status and conservation challenges. Biol Rev 81:163–182
Eastern Brook Trout Joint Venture (EBTJV) (2016) Range-wide assess-
ment of brook trout at the catchment scale: a summary of find-
ings. https ://www.easte rnbro oktro Accessed Jan 2017
Evans DM, Che-Castaldo JP, Crouse D, Davis FW, Epachin-Niell R,
Flather CH, Frohlich RK, Goble DD, Li YW, Male TD, Master
LL, Moskwik MP, Neel MC, Noon BR, Parmesan C, Schwartz
MW, Scott JM, Williams BK (2016) Species recovery in the
United States: increasing the effectiveness of the endangered
species act. Issues Ecol 20:1–28
Fausch KD, Rieman BE, Dunham JB, Young MK, Peterson DP
(2009) Invasion versus isolation: trade-offs in managing native
salmonids with barriers to upstream movement. Conserv Biol
Fischer J, Lindenmayer DB (2000) An assessment of the published
results of animal translocations. Biol Conserv 96:1–11
Frankham R, Ballou JD, Eldridge MDB, Lacy RC, Ralls K, Dudash
MR, Fenster CB (2011) Predicting the probability of outbreeding
depression. Conserv Biol 25:465–475
Fraser DJ, Debes PV, Bernatchez L, Hutchings JA, Fraser DJ (2014)
Population size, habitat fragmentation, and the nature of adaptive
variation in a stream fish. Proc R Soc B 281:1–8
Gozlan RE, Britton JR, Cowx I, Copp GH (2010) Current knowl-
edge on non-native freshwater fish introductions. J Fish Biol
Griffith B, Scott MJ, Carpenter JW, Reed C (1989) Translocation
as a species conservation tool: status and strategy. Science
Guffey SZ (1993) Allozyme genetics of South Carolina brook trout.
South Carolina Department of Natural Resources, Columbia
Haak AL, Williams JE, Neville HM, Dauwalter DC, Colyer WT (2010)
Conserving peripheral trout populations: the values and risks of
life on the edge. Fisheries 35:530–549
Habera J. Moore S (2005) Managing southern Appalachian brook trout:
a position statement. Fisheries 30(7):10–20
Hampe A, Petit RJ (2005) Conserving biodiversity under climate
change: the rear edge matters. Ecol Lett 8:461–467
Hayes JP, Guffey SZ, Kriegler FJ, McCracken GF, Parker CR (1996)
The genetic diversity of native, stocked, and hybrid populations
Conservation Genetics
1 3
of brook trout in the southern Appalachians. Conserv Biol
Heber S, Varsani A, Kuhn S, Girg A, Kempenaers B, Briskie J (2013)
The genetic rescue of two bottlenecked South Island robin pop-
ulations using translocations of inbred donors. Proc R Soc B
Hedrick PW (1995) Gene flow and genetic restoration: the Florida
panther as a case study. Conserv Biol 9:996–1007
Hedrick PW (2005) “Genetic restoration”: a more comprehensive per-
spective than “genetic rescue”. Trends Ecol Evol 20:109
Hoffmann M, Brooks TM, Butchart SHM, Carpenter KE, Chanson
J etal (2010) The impact of conservation on the status of the
world’s vertebrates. Science 330:1503–1509
Hoxmeier RJH, Dieterman DJ, Miller LM (2015) Brook trout distri-
bution, genetics and population characteristics in the driftless
area of Minnesota. N Am J Fish Manag 35:632–648
Hudy M, Thieling TM, Gillespie N, Smith EP (2008) Distribution,
status, and land use characteristics of subwatersheds within the
native range of brook trout in the eastern United States. N Am
J Fish Manag 28:1069–1085
Huff DD, Miller LM, Vondracek B (2010) Patterns of ances-
try and genetic diversity in reintroduced populations of the
slimy sculpin: implications for conservation. Conserv Genet
Huff DD, Miller LM, Chizinski CJ, Vondracek B (2011) Mixed-source
reintroductions lead to outbreeding depression in second-gen-
eration descendants of a native North American fish. Mol Ecol
IUCN (1987) IUCN position statement on translocation of living
organisms: introductions, re-introductions and re-stocking.
IUCN, Gland
Jombart T (2008) Adegenet: a R package for the multivariate analysis
of genetic markers. Bioinformatics 24:1403–1405
Jombart T, Devillard S, Balloux F (2010) Discriminant analysis of
principal components: a new method for the analysis of geneti-
cally structured populations. BioMed Cent Genet 11(94):1–15
Kanno Y, Vokoun JC, Letcher BH (2011) Fine-scale population struc-
ture and riverscape genetics of brook trout (Salvelinus fontin-
alis) distributed along headwater channel networks. Mol Ecol
Kanno Y, Kulp MA, Moore SE (2016) Recovery of native brook trout
populations following the eradication of nonnative rainbow trout
in southern Appalachian mountains streams. N Am J Fish Manag
Kazyak DC, Hilderbrand RH, Keller SR, Colaw MC, Holloway AE,
Morgan IIRP, King TL (2015) Spatial structure of morphologi-
cal and neutral genetic variation in brook trout. Trans Am Fish
Soc 144:480–490
Kazyak DC, Rash J, Lubinski BA, King TL (2018) Assessing the
impact of stocking northern-origin hatchery brook trout on the
genetics of wild populations in North Carolina. Conserv Genet
Kearse M, Moir R, Wilson A, Stones-Havas S, Cheung M, Stur-
rock S, Buxton S, Cooper A, Markowitz S, Duran C, Thierer
T, Ashton B, Mentjies P, Drummond A (2012) Geneious basic:
an integrated and extendable desktop software platform for the
organization and analysis of sequence data. Bioinformatics
Keller LF, Waller DM (2002) Inbreeding effects in wild populations.
Trends Ecol Evol 17:230–241
Kelson SJ, Kapuscinski AR, Timmins D, Ardren WR (2015) Fine-scale
genetic structure of brook trout in a dendritic stream network.
Conserv Genet 16:31–42
King TL, Julian SE, Coleman RL, Burnham Curtis MK (2003) Isolation
and characterization of novel tri-and tetranucleotide microsatel-
lite DNA markers for Brook trout Salvelinus fontinalis: GenBank
submission numbers AY168187, AY168188, AY168189,
AY168191, AY168192, AY 168193, AY168194, AY168195,
AY168196, AY168197, AY168198, AY168199. https ://www. otide /. Accessed Nov 2014
King TL, Eackles MS, Letcher BH (2005) Microsatellite DNA markers
for the study of Atlantic salmon (Salmo salar) kinship, popu-
lation structure, and mixed-fishery analyses. Mol Ecol Notes
Kirchner F, Robert A, Colas B (2006) Modelling the dynamics of intro-
duced populations in the narrow-endemic Centaurea corymbosa:
a demo-genetic integration. J Appl Ecol 43:1011–1021
Kriegler FJ, McCracken GF, Habera JW, Strange RJ (1995) Genetic
characterization of Tennessee brook trout populations and associ-
ated management implications. N Am J Fish Manag 15:804–813
Krueger CC, Menzel BW (1979) Effects of stocking on genetics of
wild brook trout populations. Trans Am Fish Soc 108:277–287
Kulp MA, Moore SE (2005) A case history in fishing regulations in
Great Smoky Mountains National Park: 1934–2004. N Am J Fish
Manag 25:510–524
Laikre L, Schwartz MK, Waples RS, Ryman N, GeM Working Group
(2010) Compromising genetic diversity in the wild: unmoni-
tored large-scale release of plants and animals. Trends Ecol Evol
Lamaze FC, Sauvage C, Marie A, Garant D, Bernatchez L (2012)
Dynamics of introgressive hybridization assessed by SNP popu-
lation genomics of coding genes in stocked brook charr (Salveli-
nus fontinalis). Mol Ecol 21:2877–2895
Lande R, Barrowclough GF (1987) Effective population size, genetic
variation, and their use in population management. In: Soulé ME
(ed) Viable populations for conservation. Cambridge University
Press, Cambridge, pp87–123
Larson GL, Moore SE (1985) Encroachment of exotic rainbow trout
into stream populations of native brook trout in the southern
Appalachian mountains. Trans Am Fish Soc 114:195–203
Lennon RE (1967) Brook trout of Great Smoky Mountains National
Park. U.S. Fish and Wildlife Service, Washington, DC
Lesica P, Allendorf FW (1995) When are peripheral populations viable
for conservation? Conserv Biol 9:753–760
Letcher BH, Coombs JA, Nislow KH (2011) Maintenance of pheno-
typic variation: repeatability, heritability, and size-dependent
processes in a brook trout population. Evol Appl 4:602–615
Maillett E, Aiken R (2015) Trout fishing in 2011: a demographic
description and economic analysis. Addendum to the 2011 nation
survey of fishing, hunting and wildlife-associated recreation.
United States Fish & Wildlife Service, Washington, DC
Mantel N (1967) The detection of disease clustering and a generalized
regression approach. Can Res 27:209–220
Marie AD, Bernatchez L, Garant D (2010) Loss of genetic integrity
correlates with stocking intensity in brook charr (Salvelinus fon-
tinalis). Mol Ecol 19:2025–2037
McCracken GF, Parker CR, Guffey SZ (1993) Genetic differentiation
and hybridization between stocked hatchery and native brook
trout in Great Smoky Mountains National Park. Trans Am Fish
Soc 122:533–542
Meisner JD (1990) Effect of climatic warming on the southern margins
of the native range of brook trout, Salvelinus fontinalis. Can J
Fish Aquat Sci 47:1065–1070
Narum SR (2006) Beyond Bonferroni: less conservative analyses for
conservation genetics. Conserv Genet 7:783–787
Nathan LR, Kanno Y, Vokoun JC (2017) Population demographics
influence genetic responses to fragmentation: a demogenetic
assessment of the ‘one migrant per generation’ rule of thumb.
Biol Conserv 210:261–272
Palstra FP, Ruzzante DE (2008) Genetic estimates of contempo-
rary effective population size: what can they tell us about the
Conservation Genetics
1 3
importance of genetic stochasticity for wild population persis-
tence? Mol Ecol 17:3428–3447
Patterson N, Price AL, Reich D (2006) Population structure and eige-
nanalysis. PLoS Genet 2:2074–2093
Pavlova A, Beheregaray LB, Coleman R, Gilligan D, Harrisson KA,
Ingram BA, Kearns J, Lamb AM, Lintermans M, Lyon J, Nguyen
TT, Sasaki M, Tonkin Z, Yen JDL, Sunnucks P (2017) Severe
consequences of habitat fragmentation on genetic diversity of an
endangered Australian freshwater fish: a call for assisted gene
flow. Evol Appl 10:531–550
Peacock MM, Dochtermann NA (2012) Evolutionary potential but
not extinction risk of Lahonton cutthroat trout (Oncorhynchus
clarkia henshawi) is associated with stream characteristics. Can
J Fish Aquat Sci 69:615–626
Peakall R, Smouse PE (2005) GENALEX 6: genetic analysis in Excel.
Population genetic software for teaching and research. Mol Ecol
Pritchard JK, Stephens M, Donnelly P (2000) Inference of population
structure using multilocus genotype data. Genetics 155:945–959
R Development Core Team (2008) R: a language and environment for
statistical computing. R Foundation for Statistical Computing,
Ramasamy RK, Ramasamy S, Bindroo BB, Naik VG (2014) STRU
CTU RE PLOT: a program for drawing elegant STRU CTU RE
bar plots in user friendly interface. Springerplus. https ://doi.
Raymond M, Rousset F (1995) GENEPOP (version 1.2): population
genetics software for exact tests and ecumenicism. J Heredity
Redford KH, Amato G, Baillie J, Beldomenico P, Bennett EL, Clum
N, Cook R, Fonseca G, Hedges S, Launay F, Lieberman S, Mace
GM, Murayama A, Putnam A, Robinson JG, Rosenbaum H,
Sanderson EW, Stuart SN, Thomas P, Thorbjarnarson J (2011)
What does it mean to successfully conserve a (vertebrate) spe-
cies? BioScience 61:39–48
Rhymer JM, Simberloff D (1996) Extinction by hybridization and intro-
gression. Annu Rev Ecol Syst 27:83–109
Robinson ZL, Coombs JA, Hudy M, Nislow KH, Letcher BH, Whiteley
AR (2017) Experimental test of genetic rescue in isolated popu-
lations of brook trout. Mol Ecol 26:4418–4433
Rodriguez-Ramilo ST, Wang J (2012) The effect of close relatives
on unsupervised Bayesian clustering algorithms in population
genetic structure analysis. Mol Ecol Resour 12:873–884
Rohde FC, Arndt RG, Foltz JW, Quattro JM (2009) Freshwater fishes
of South Carolina. University of South Carolina Press, Columbia,
Ruzzante DE, McCracken GR, Parmelee S, Hill K, Corrigan A, Mac-
Millan J, Walde SJ (2016) Effective number of breeders, effec-
tive population size and their relationship with census size in
an iteroparous species, Salvelinus fontinalis. R Soc B 283:1–9
Stoneking M, Wagner DJ, Hildebrand AC (1981) Genetic evidence
suggesting subspecific differences between northern and south-
ern populations of brook trout (Salvelinus fontinalis). Copeia
Tallmon DA, Luikart G, Waples RS (2004) The alluring simplicity and
complex reality of genetic rescue. Trends Ecol Evol 19:489–496
Wang J (2004) Sibship reconstruction from genetic data with typing
errors. Genetics 166:1963–1979
Waples RS (2005) Genetic estimates of contemporary effective popula-
tion size: to what time periods do the estimates apply? Mol Ecol
Waples RS, Anderson EC (2017) Purging putative siblings from
population genetic data sets: a cautionary view. Mol Ecol
Waples RS, Do C (2010) Linkage disequilibrium estimates of con-
temporary Ne using highly variable genetic markers: a largely
untapped resource for applied conservation and evolution. Evol
Appl 3:244–262
Weeks AR, Sgro CM, Young AG, Frankham R, Mitchell NJ, Miller
KA, Byrne M, Coates DJ, Eldridge MDB, Sunnucks P, Breed
MF, James EA, Hoffmann AA (2011) Assessing the benefits and
risks of translocations in changing environments: a genetic per-
spective. Evol Appl 4:709–725
Wenger SJ, Isaak DJ, Luce CH, Neville HM, Fausch KD, Dunham JB,
Dauwalter DC, Young MK, Elsner MM, Rieman BE, Hamlet AF,
Williams JE (2011) Flow regime, temperature, and biotic inter-
actions drive differential declines of trout species under climate
change. Proc Natl Acad Sci 108:14175–14180
Whiteley AR, Hastings K, Wenburg JK, Frissell CA, Martin JC, Allen-
dorf FW (2010) Genetic variation and effective population size
in isolated populations of coastal cutthroat trout. Conserv Genet
Whiteley AR, Coombs JA, Hudy M, Robinson Z, Nislow KH, Letcher
BH (2012) Sampling strategies for estimating brook trout effec-
tive population size. Conserv Genet 13:577–593
Whiteley AR, Coombs JA, Hudy M, Robinson Z, Colton AR, Nislow
KH (2013) Fragmentation and patch size shape genetic structure
of brook trout populations. Can J Fish Aquat Sci 70:678–688
Whiteley AR, Coombs JA, Letcher BH, Nislow KH (2014a) Simulation
and empirical analysis of novel sibship-based genetic determina-
tion of fish passage. Can J Fish Aquat Sci 71:1667–1679
Whiteley AR, Hudy M, Robinson ZL, Coombs JA, Nislow KH (2014b)
Patch-based metrics: a cost effective method for short- and long-
term monitoring of EBTJV wild brook trout populations? In:
Carline RF, LoSapio C (eds) Wild Trout XI: looking back and
moving forward. Wild Trout Symposium, West Yellowstone
Whiteley AR, Fitzpatrick SW, Funk WC, Tallmon DA (2015) Genetic
rescue to the rescue. Trends Ecol Evol 30:42–49
Wright S (1951) The genetical structure of populations. Ann Eugen
... A subset of samples (12 collections) represents single-cohort samples that focused on age-0 (young-of-the-year) individuals. Young of the year were sampled from approximately three spatially distinct sites, each approximately 100 m in length, within contiguous stream habitat (Pregler et al. 2018). ...
... This finding highlights the importance of conserving endemic genetic diversity within the southern region, as populations are often unique and irreplaceable. Moreover, it challenges the notion that Brook Trout in the south are genetically depauperate (Pregler et al. 2018;Weathers et al. 2019). There is, in fact, high genetic diversity here, but it is spread among many populations that have had a long time to diversify and adapt to local conditions. ...
... Despite an extensive history of stocking of domesticated conspecifics, many Brook Trout populations in the southern Appalachians show little evidence of hatchery introgression (Pregler et al. 2018;Printz et al. 2018;present study). Rather, the vast majority of populations retain genetic characteristics distinct from those of hatchery strains. ...
Full-text available
Broad‐scale patterns of genetic diversity for Brook Trout remain poorly understood across their endemic range in the eastern United States. We characterized variation at 12 microsatellite loci in 22,020 Brook Trout among 836 populations from Georgia, USA, to Quebec, Canada, to the western Great Lakes region. Within‐population diversity was typically lower in the southern Appalachians relative to the mid‐Atlantic and northeastern regions. Effective population sizes in the southern Appalachians were often very small, with many estimates less than 30 individuals. The population genetics of Brook Trout in the southern Appalachians are far more complex than a conventionally held simple “northern” versus “southern” dichotomy would suggest. Contemporary population genetic variation was consistent with geographic expansion of Brook Trout from Mississippian, mid‐Atlantic, and Acadian glacial refugia, as well as differentiation among drainages within these broader clades. Genetic variation was pronounced among drainages (57.4% of overall variation occurred among Hydrologic Unit Code (HUC)10 or larger units) but was considerable even at fine spatial scales (13% of variation occurred among collections within HUC12 drainage units). Remarkably, 87.2% of individuals were correctly assigned to their collection of origin. While comparisons with fish from existing major hatcheries showed impacts of stocking in some populations, genetic introgression did not overwhelm the signal of broad‐scale patterns of population genetic structure. Although our results reveal deep genetic structure in Brook Trout over broad spatial extents, fine‐scale population structuring is prevalent across the southern Appalachians. Our findings highlight the distinctiveness and vulnerability of many Brook Trout populations in the southern Appalachian Mountains and have important implications for wild Brook Trout management. To facilitate application of our findings by conservation practitioners, we provide an interactive online visualization tool to allow our results to be explored at management‐relevant scales.
... There are numerous and substantial pressures that make it critical to document relationships among salmonid populations and to characterize potential threats to the genetic integrity of extant populations, especially for Brook Trout. With increasing interest in Brook Trout, numerous studies have addressed eastern Brook Trout genetics (e.g., Hayes et al. 1996;Danzmann 1997, 1998;Danzmann et al. 1998;Hall et al. 2002;Stauffer and King 2014;Aunins et al. 2015;Kazyak et al. 2015Kazyak et al. , 2016Buonaccorsi et al. 2017;Bruce et al. 2018;Nathan et al. 2018;Pregler et al. 2018;Weathers et al. 2018). However, key spatial gaps in genetic structure remain unaddressed throughout the native range of eastern Brook Trout, particularly in the mid-Atlantic region, consisting of New York; New Jersey; Pennsylvania; Delaware; Maryland; Washington, D.C.; Virginia; and West Virginia. ...
... Based on the microsatellite comparison of Maryland and eight hatchery collections in both STRUCTURE analysis and PCA, our study suggests that stocking has not had any widespread homogenizing influence on native Brook Trout, as also noted by Hall et al. (2002). In general, other studies on Brook Trout stocking have also noted that hatchery introgression in the eastern United States is not nearly as pervasive as previously thought (Annett et al. 2012;Kazyak et al. 2018;Pregler et al. 2018;White et al. 2018;Beer et al. 2019). This finding is in contrast to the observation that other Brook Trout populations along the Appalachians and Midwest have been influenced by hatchery stocks (e.g., Wisconsin, Krueger and Menzel 1979;southern Appalachians, Hayes et al. 1996;Pennsylvania, Buonaccorsi et al. 2017;New York, Beer et al. 2019). ...
Full-text available
Brook Trout Salvelinus fontinalis have declined across their native range due to multiple anthropogenic factors, including landscape alteration and climate change. Although coldwater streams in the State of Maryland (eastern United States) historically supported significant Brook Trout populations, only fragmented remnant populations remain with the exception of the upper Savage River watershed in western Maryland. Using microsatellite data from 38 collections, we defined genetic relationships of Brook Trout populations in Maryland drainages. Microsatellite analyses of Brook Trout indicated the presence of five major discrete units, defined as the Youghiogheny (Ohio), Susquehanna, Patapsco/Gunpowder, Catoctin, and the Upper Potomac, with a distinct genetic subunit present in the Savage River (Upper Potomac). We did not observe evidence for widespread hatchery introgression with native Brook Trout. However, genetic effects due to fragmentation were evident in several Maryland Brook Trout populations, resulting in erosion of diversity that may have negative implications for their future persistence. Our current study supplements an increasing body of evidence that Brook Trout populations in Maryland are highly susceptible to multiple anthropogenic stresses, and many populations may be extirpated in the near future. Future management efforts focused on habitat protection and potential stream restoration, coupled with a comprehensive assessment framework that includes genetic considerations may provide the best outlook for Brook Trout populations in Maryland.
... Kazyak et al. [68] observed limited effects of hatchery stocking, but introgression did not affect the overall broad-scale signal pattern of genetic structure. In a study of rear edge populations of Brook Trout in South Carolina, Pregler et al. [71] found effects of hatchery introductions in 6 of 18 streams sampled, and postulated management options for enhancing or restoring these rear edge populations. ...
Full-text available
The determination of endangered species is problematic. If one considers a species to be ontological individuals, then if a species goes extinct, it is gone forever. The Brook Trout is used as an example of a “species” which may be comprised of several unique entities that warrant a specific status. In addition to determining the specific status, it is difficult to determine how to place a monetary value on endangered species that do not have a general appeal to the public (e.g., many bird species), a commercial value, no known medical properties (e.g., deep water sponges vs. cancer), or generate monies for recreation. Perhaps if we could identify the unique information carried by a particular species, we could place a value on that information and assess the monetary value of the information lost.
... Lastly, hatchery-reared Brook Trout of non-native ancestry (primarily northeastern US stocks) were introduced throughout the southern Appalachian Mountains during the last century (Lennon 1967;Habera and Moore 2005;Kazyak et al. 2018). Various facets of Brook Trout life history and ecology have been studied, including the distribution of populations across the landscape (i.e., habitat requirements and land use impacts; Hudy et al. 2008) and the extent of hatchery introgression (Guffey 1998;McCracken et al. 1993;Kriegler et al. 1995;Hayes et al. 1996;Galbreath et al. 2001;Dunham et al. 2002;Seehorn 2004;Kazyak et al. 2018;Pregler et al. 2018;Weathers et al. 2019). However, no formal work has attempted to relate the relative influence of habitat characteristics, land use practices, and hatchery introgression on levels of genetic diversity in Brook Trout populations. ...
Full-text available
Throughout their range, Brook Trout (Salvelinus fontinalis) occupy thousands of disjunct drainages with varying levels of disturbance, which presents substantial challenges for conservation. Within the southern Appalachian Mountains, fragmentation and genetic drift have been identified as key threats to the genetic diversity of the Brook Trout populations. In addition, extensive historic stocking of domestic lineages of Brook Trout to augment fisheries may have eroded endemic diversity and impacted locally adapted populations. We used 12 microsatellite loci to describe patterns of genetic diversity within 108 populations of wild Brook Trout from Tennessee and used linear models to explore the impacts of land use, drainage area, and hatchery stockings on metrics of genetic diversity, effective population size, and hatchery introgression. We found levels of within-population diversity varied widely, although many populations showed very limited diversity. The extent of hatchery introgression also varied across the landscape, with some populations showing high affinity to hatchery lineages and others appearing to retain their endemic character. However, we found relatively weak relationships between genetic metrics and landscape characteristics, suggesting that contemporary landscape variables are not strongly related to observed patterns of genetic diversity. We consider this result to reflect both the complex history of these populations and the challenges associated with accurately defining drainages for each population. Our study highlights the importance of genetic data to guide management decisions, as complex processes interact to shape the genetic structure of populations and make it difficult to infer the status of unsampled populations.
... Brook Trout Salvelinus fontinalis in southern Appalachian Mountains streams occur at the southernmost limit of their native range in eastern North America ( Figure 1) (Moore et al. 1986;Hudy et al. 2008;Kanno et al. 2016a). Small, isolated Brook Trout populations in southern Appalachian Mountains streams show signs of genetic drift including decreased genetic diversity, low effective population size, and increased extinction risk (Pregler et al. 2018;Weathers et al. 2019). Remaining populations occur in low-productivity, high-elevation headwater streams (i.e., suboptimal habitat), where limited food availability and habitat space lead to small adult size (Konopacky and Estes 1986;Ensign et al. 1990;Kulp and Moore 2005;Knoepp et al. 2016). ...
Full-text available
Brook Trout Salvelinus fontinalis in southern Appalachian Mountains streams of the United States occur at the southernmost portion of their native range, and occupy small, isolated, and low-productivity headwater streams. The existing standard weight (Ws) equation is applicable only to Brook Trout > 120 mm total length (TL), but many individuals in the region are smaller than this minimum size threshold due to their habitat characteristics. Here, we developed a new Ws equation for Brook Trout in southern Appalachian Mountains streams using length–weight data on 72,502 individuals. The weighted quadratic empirical-percentile method minimized length-related bias in relative weight compared to the regression-line-percentile and weighted linear empirical-percentile methods. The proposed Ws equation was: log10W = −3.364 + 1.378 × log10L + 0.397 × (log10L)2, where W was weight (g) and L was TL (mm). The new equation characterized body condition of Brook Trout in southern Appalachian Mountains streams more accurately than the existing equation.
... Thus, there is some risk that stocking hatchery-strain Brook Trout where native conspecifics occur will result in the erosion of the native genetic resources. Overall, our findings are generally consistent with the results of recent research in other places, which suggests that hatchery introgression is not nearly as prevalent as previously thought but also can be substantial in some populations (Kazak et al. 2017;Pregler et al. 2018;White et al. 2018;Beer et al. 2019). Future research should consider why hatchery introgression occurs in some populations but not in others. ...
Full-text available
Following centuries of declines, there is growing interest in conserving extant wild populations and reintroducing Brook Trout (Salvelinus fontinalis) populations of native ancestry. A population genetic baseline can enhance conservation outcomes and promote restoration success. Consequently, it is important to document existing patterns of genetic variation across the landscape and translate these data into an approachable format for fisheries managers. We genotyped 9,507 Brook Trout representing 467 wild collections at 12 microsatellite loci to establish a genetic baseline for North Carolina, USA. Rarefied allelic richness and observed heterozygosity, which reflect within‐population diversity, were low to moderate relative to levels typically observed at higher latitudes (means = 3.12 and 0.42, respectively). Effective population sizes varied widely, but were often very low (151 collections with an estimated Ne < 10). Despite decades of intensive stocking across the state, we found little to no evidence of hatchery introgression in most populations. Although genetic variation was significant at a variety of spatial scales (mean pairwise F’ST = 0.73), substantial genetic variation occurred between patches within individual watersheds. Analysis of molecular variance (AMOVA) found that a substantial portion (28.5%) of the observed genetic variation was attributed to differences among populations, with additional genetic variation among hydrological units (HUCs; 16.0%, 16.6%, 12.1%, and 9.4% of the overall variation among twelve‐, ten‐, eight‐, and six‐digit HUCs, respectively). We discuss a suite of potential applications for this type of genetic data to enhance management outcomes, such as conservation prioritization and selection of source stocks for reintroductions or genetic rescue.
... Brook Trout Salvelinus fontinalis in southern Appalachian Mountains streams occur at the southernmost limit of their native range in eastern North America ( Figure 1) (Moore et al. 1986;Hudy et al. 2008;Kanno et al. 2016a). Small, isolated Brook Trout populations in southern Appalachian Mountains streams show signs of genetic drift including decreased genetic diversity, low effective population size, and increased extinction risk (Pregler et al. 2018;Weathers et al. 2019). Remaining populations occur in low-productivity, high-elevation headwater streams (i.e., suboptimal habitat), where limited food availability and habitat space lead to small adult size (Konopacky and Estes 1986;Ensign et al. 1990;Kulp and Moore 2005;Knoepp et al. 2016). ...
Full-text available
Brook Trout Salvelinus fontinalis in southern Appalachian Mountains streams of the USA occur at the southernmost portion of their native range, and occupy small, isolated, and low-productivity headwater streams. The existing standard weight (Ws) equation is applicable only to Brook Trout > 120 mm total length (TL), but many individuals in the region are smaller than this minimum size threshold due to their habitat characteristics. Here, we developed a new Ws equation for Brook Trout in southern Appalachian Mountains streams using length-weight data on 72,502 individuals. The weighted quadratic empirical-percentile method minimized length-related bias in relative weight compared to the regression-line-percentile and weighted linear empirical-percentile methods. The proposed Ws equation was: log10W = -3.364 + 1.378 x log10L + 0.397 x (log10L)2, where W was weight (g) and L was TL (mm). The new equation characterized body condition of Brook Trout in southern Appalachian Mountains streams more accurately than the existing equation.
... In continental Europe, the most genetically impoverished and distinct populations were found along the southern (Greece) and western (Switzerland) peripheries of the natural distribution (Triantafyllidis et al., 2002). Although not always observed empirically, this is a pattern frequently expected for peripheral populations due to their geographical isolation combined with lower N e than in more benign environments, possibly reinforced by past founder events, population bottlenecks in combination with selection on locally adapted traits (Eckert et al., 2008;Johannesson & Andre, 2006;Pregler et al., 2018). ...
Using ten polymorphic microsatellites and 1251 individual samples (some dating back to the early 1980s), genetic structure and effective population size in all native and introduced Swedish populations of the European wels catfish or Silurus glanis were studied. Levels of genetic variability and phylogeographic relationships were compared with data from a previous study of populations in other parts of Europe. The genetically distinct Swedish populations displayed comparably low levels of genetic variability and according to one‐sample estimates based on linkage disequilibrium and sib ship‐reconstruction, current local effective population sizes were lower than minimum levels recommended for short‐term genetic conservation. In line with a previous suggestion of postglacial colonisation from a single refugium, all Swedish populations were assembled on a common branch in a star‐shaped dendrogram together with other European populations. Two distinct subpopulations were detected in upper and lower habitats of River Emån, indicating that even minor dispersal barriers may restrict gene flow for wels in running waters. Genetic assignment of specimens encountered in the brackish Baltic Sea and in lakes where the species does not occur naturally indicated presence of long‐distance sea dispersal and confirmed illegal translocations, respectively.
Full-text available
Due to geologic processes and recent anthropogenic introductions, patterns of genetic and morphological diversity within the Smallmouth Bass (Micropterus dolomieu), which are endemic to the central and eastern United States (USA), are poorly understood. We assessed genetic and morphological differentiation between the widespread Northern Smallmouth Bass (M. d. dolomieu) and the more restricted Neosho Smallmouth Bass (M. d. velox) where their ranges meet in the Central Interior Highlands ecoregion (CIH). Data from 14 microsatellite loci were used to conduct Structure and principal components analyses to evaluate diversity across populations and screen for hybridization with sympatric Spotted Bass (M. punctulatus). We also tested for morphological differences using five morphometric traits and one meristic trait. We found support for three genetic clusters corresponding to previously described taxonomic variation; five clusters largely corresponding to river systems; and nine clusters representing hierarchical population structure within both ranges. We found evidence of a unique genetic cluster in tributaries of the White River within the Northern Smallmouth Bass range and admixture between the subspecies throughout the Neosho range. We also found evidence of morphological differentiation between subspecies; Neosho Smallmouth Bass exhibited larger head length than Northern Smallmouth Bass relative to total length, and there was a significant interaction of subspecies and orbital length, possibly indicating differential growth patterns between subspecies. Our results reveal multiple levels of divergence, suggesting the CIH harbors greater and more complex Smallmouth Bass diversity than previously thought.
Full-text available
In 2006, Narum published a paper in Conservation Genetics emphasizing that Bonferroni correction for multiple testing can be highly conservative with poor statistical power (high Type II error). He pointed out that other approaches for multiple testing correction can control the false discovery rate (FDR) with a better balance of Type I and Type II errors and suggested that the approach of Benjamini and Yekutieli (BY) 2001 provides the most biologically relevant correction for evaluating the significance of population differentiation in conservation genetics. However, there are crucial differences between the original Benjamini and Yekutieli procedure and that described by Narum. After carefully reviewing both papers, we found an error due to the incorrect implementation of the BY procedure in Narum (Conserv Genet 7:783–787, 2006) such that the approach does not adequately control FDR. Since the incorrect BY approach has been increasingly used, not only in conservation genetics, but also in medicine and biology, it is important that the error is made known to the scientific community. In addition, we provide an overview of FDR approaches for multiple testing correction and encourage authors first and foremost to provide effect sizes for their results; and second, to be transparent in their descriptions of multiple testing correction. Finally, the impact of this error on conservation genetics and other fields will be study-dependent, as it is related to the number of true to false positives for each study.
Full-text available
The release of hatchery-origin fish into streams with endemics can degrade the genetics of wild populations if interbreeding occurs. Starting in the 1800s, brook trout descendent from wild populations in the northeastern United States were stocked from hatcheries into streams across broad areas of North America to create and enhance fishery resources. Across the southeastern United States, many millions of hatchery-origin brook trout have been released into hundreds of streams, but the extent of introgression with native populations is not well resolved despite large phylogeographic distances between these groups. We used three assessment approaches based on 12 microsatellite loci to examine the extent of hatchery introgression in 406 wild brook trout populations in North Carolina. We found high levels of differentiation among most collections (mean F′ST = 0.718), and among most wild collections and hatchery strains (mean F′ST = 0.732). Our assessment of hatchery introgression was consistent across the three metrics, and indicated that most wild populations have not been strongly influenced by supplemental stocking. However, a small proportion of wild populations in North Carolina appear to have been strongly influenced by stocked conspecifics, or in some cases, may have been founded entirely by hatchery lineages. In addition, we found significant differences in the apparent extent of hatchery introgression among major watersheds, with the Savannah River being the most strongly impacted. Conversely, populations in the Pee Dee River watershed showed little to no evidence of hatchery introgression. Our study represents the first large-scale effort to quantify the extent of hatchery introgression across brook trout populations in the southern Appalachians using highly polymorphic microsatellite markers.
Full-text available
Genetic diversity underpins the ability of populations to persist and adapt to environmental changes. Substantial empirical data show that genetic diversity rapidly deteriorates in small and isolated populations due to genetic drift, leading to reduction in adaptive potential and fitness and increase in inbreeding. Assisted gene flow (i.e. via translocations) can reverse these trends, but lack of data on fitness loss and fear of impairing population ‘uniqueness’ often prevents managers from acting. Here we use population genetic and riverscape genetic analyses and simulations to explore the consequences of extensive habitat loss and fragmentation on population genetic diversity and future population trajectories of an endangered Australian freshwater fish, Macquarie perch Macquaria australasica. By using guidelines to assess the risk of outbreeding depression under admixture, we develop recommendations for population management, identify populations requiring genetic rescue and/or genetic restoration and potential donor sources. We found that most remaining populations of Macquarie perch have low genetic diversity, and effective population sizes below the threshold required to retain adaptive potential. Our simulations showed that under management inaction, smaller populations of Macquarie perch will face inbreeding depression within a few decades, but regular small-scale translocations will rapidly rescue populations from inbreeding depression and increase adaptive potential through genetic restoration. Despite the lack of data on fitness loss, based on our genetic data for Macquarie perch populations, simulations and empirical results from other systems, we recommend regular and frequent translocations among remnant populations within catchments, in order to emulate the effect of historical gene flow and improve population persistence through decrease in demographic and genetic stochasticity. Increasing population genetic connectivity within each catchment will help to maintain large effective population sizes and maximize species adaptive potential. The approach proposed here could be readily applicable to genetic management of other threatened species to improve their adaptive potential. This article is protected by copyright. All rights reserved.
1. In the context of the restoration of an endangered species, population viability analysis represents a useful tool for assessing the effectiveness of different possible management strategies before implementation. However, despite the consensus that demographic and genetic mechanisms are both involved and interact in the process of extinction, few attempts have been made to examine their combined impacts on population viability in a particular species. 2. We integrated specific data resulting from 10-year multidisciplinary investigations into a descriptive model to simulate the dynamics of an introduced population of the rare self-incompatible plant species Centaurea corymbosa. The model allowed us to examine the interplay between demographic processes and genetic self-incompatibility in the particular habitat conformation of the species, alternating suitable and unsuitable sites within a population along cliffs. Population growth and extinction risk were compared for different introduction strategies. 3. Population persistence mainly depended on the number of introduced seeds and on their initial spatial distribution within the population (single vs. multisite introduction). In most cases, a multisite introduction resulted in faster population growth and higher viability than a single-site introduction. 4. As expected, a strong negative impact of the self-incompatibility system was observed on population dynamics and viability. However, because of positive feedback between demographic and genetic processes, this impact differed among introduction strategies: it was less severe when seeds were distributed among suitable sites, which also limited the loss of self-incompatibility alleles. Moreover, self-incompatibility contributed to the positive relationship between flowering plant density and fertilization rate. 5. Synthesis and applications. Our results provide strong management guidelines for future introductions of C. corymbosa regarding the number of seeds required (> 800) and the benefits of introducing them into several sites to achieve population persistence. Further, the study highlights the general importance of integrating demography and genetics to compare the effectiveness of different management strategies.
The package adegenet for the R software is dedicated to the multivariate analysis of genetic markers. It extends the ade4 package of multivariate methods by implementing formal classes and functions to manipulate and analyse genetic markers. Data can be imported from common population genetics software and exported to other software and R packages. adegenet also implements standard population genetics tools along with more original approaches for spatial genetics and hybridization. Availability: Stable version is available from CRAN: Development version is available from adegenet website: Both versions can be installed directly from R. adegenet is distributed under the GNU General Public Licence (v.2). Supplementary information:Supplementary data are available at Bioinformatics online.
Genetic rescue is an increasingly considered conservation measure to address genetic erosion associated with habitat loss and fragmentation. The resulting gene flow from facilitating migration may improve fitness and adaptive potential, but is not without risks (e.g. outbreeding depression). Here, we conducted a test of genetic rescue by translocating ten (five of each sex) brook trout (Salvelinus fontinalis) from a single source to four nearby and isolated stream populations. To control for the demographic contribution of translocated individuals, ten resident individuals (five of each sex) were removed from each recipient population. Prior to the introduction of translocated individuals, the two smallest above-barrier populations had substantially lower genetic diversity, and all populations had reduced effective number of breeders relative to adjacent below-barrier populations. In the first reproductive bout following translocation, 31 out of 40 (78%) of translocated individuals reproduced successfully. Translocated individuals contributed to more families than expected under random mating, and generally produced larger full-sibling families. We observed relatively high (>20%) introgression in three of the four recipient populations. The translocations increased genetic diversity of recipient populations by 45% in allelic richness, and 25% in expected heterozygosity. Additionally, strong evidence of hybrid vigor was observed through significantly larger body sizes of hybrid offspring relative to resident offspring in all recipient populations. Continued monitoring of these populations will test for negative fitness effects beyond the first generation. However, these results provide much-needed experimental data to inform the potential effectiveness of genetic rescue motivated translocations. This article is protected by copyright. All rights reserved.
Fragmented landscapes reduce gene flow and impair long term population viability. Stream networks are particularly susceptible to fragmentation because dispersal is constrained to linear upstream and downstream movements. Despite these potential effects, infrequent migrations can maintain genetic diversity and as few as one migrant per generation (OMPG) is commonly suggested as sufficient gene flow to minimize losses in genetic diversity. However, demography varies by taxa, space and time, making such a generalized rule of thumb unlikely to be applicable across a diverse array of fragmentation scenarios and species. We utilized a demogenetic model to evaluate the OMPG rule and simulate the influence of population demographics on the rate of genetic changes following fragmentation in a headwater meta-population of brook trout (Salvelinus fontinalis). A single migrant per generation increased allelic diversity by an average of 15% and decreased genetic differentiation by 31% following 40 years of simulations compared to complete isolation, however OMPG was not sufficient to prevent significant changes in within- or between-population genetic metrics in all but the largest population scenario (N = 500). Less than 10 individuals were typically required to achieve no changes in both genetic metrics, yet this pattern was dependent on the source populations and will be context-specific given the population sub-structuring in a given stream network. Sensitivity analyses indicated the parameter controlling the proportion of mature females spawning annually was the most influential on population genetic responses in isolated populations, suggesting that when fewer females contribute to each generation the population is more likely to experience rapid changes in allelic frequency through genetic drift. This finding supports the use of metrics such as effective population size and the number of effective breeders in predicting population stability and viability following fragmentation. Variability in population dynamic processes and associated responses to fragmentation suggest that generalized rule of thumbs for management should be used with caution. Particularly when violations of the underlying theoretical assumptions exist, consideration of demographic processes (i.e. vital rates, species specific life history strategies and dispersal) and genetic structuring will allow for more appropriate conservation recommendations.
We describe a model-based clustering method for using multilocus genotype data to infer population structure and assign individuals to populations. We assume a model in which there are K populations (where K may be unknown), each of which is characterized by a set of allele frequencies at each locus. Individuals in the sample are assigned (probabilistically) to populations, or jointly to two or more populations if their genotypes indicate that they are admixed. Our model does not assume a particular mutation process, and it can be applied to most of the commonly used genetic markers, provided that they are not closely linked. Applications of our method include demonstrating the presence of population structure, assigning individuals to populations, studying hybrid zones, and identifying migrants and admixed individuals. We show that the method can produce highly accurate assignments using modest numbers of loci—e.g., seven microsatellite loci in an example using genotype data from an endangered bird species. The software used for this article is available from