Content uploaded by David J. Berg
Author content
All content in this area was uploaded by David J. Berg on Jun 05, 2021
Content may be subject to copyright.
Population genetic structure of three freshwater mussel
(Unionidae) species within a small stream system:
significant variation at local spatial scales
DAVID J. BERG*, ALAN D. CHRISTIAN
†
AND SHELDON I. GUTTMAN
†
*Department of Zoology, Miami University, Hamilton, OH, U.S.A.
†
Department of Zoology, Miami University, Oxford, OH, U.S.A.
SUMMARY
1. Unionid mussels are highly threatened, but little is known about genetic structure in
populations of these organisms. We used allozyme electrophoresis to examine partitioning
of genetic variation in three locally abundant and widely distributed species of mussels
from a catchment in Ohio.
2. Within-population variation was similar to that previously reported for freshwater
mussels, but genotype frequencies exhibited heterozygote deficiencies in many instances.
All three species exhibited significant among-population variation. Evidence of isolation-
by-distance was found in Elliptio dilatata and Ptychobranchus fasciolaris, while Lampsilis
siliquoidea showed no geographical pattern of among-population variation.
3. Our results suggest that the isolating effects of genetic drift were greater in L. siliquoidea
than in the other species. Differentiation of populations occurred at a much smaller spatial
scale than has previously been found in freshwater mussels. Differences among species
may reflect differences in the dispersal abilities of fishes that serve as hosts for the
glochidia larvae of mussels.
4. Based on our results, we hypothesise that species of mussels that are common to large
rivers exhibit relatively large amounts of within-population genetic variation and little
differentiation over large geographical distances. Conversely, species typical of small
streams show lower within-population genetic variation and populations will be more
isolated. If this hypothesis can be supported, it may prove useful in the design of
conservation strategies that maintain
the genetic structure of target species.
Keywords: allozyme, dispersal, Elliptio dilatata,Lampsilis siliquoidea,Ptychobranchus fasciolaris
Introduction
Maintenance of genetic diversity has been recognised
as a key component in minimising extinction proba-
bilities of populations (Saccheri et al., 1998) and
ensuring survival of threatened species (Spielman,
Brook & Frankham, 2004). This genetic diversity is
partitioned into two components. Diversity within
populations is reflected in factors such as selection,
inbreeding probability, genetic drift and the loss or
maintenance of rare alleles. Diversity among popula-
tions reflects the geographical structuring of popula-
tions, with variation among populations due to local
selection or adaptation, and to reduced gene flow or
isolation. Both these components of total genetic
variation must be quantified in order to develop
Correspondence: David J. Berg, Department of Zoology,
Miami University, 1601 University Boulevard,
Hamilton, OH 45011, U.S.A.
E-mail: bergdj@muohio.edu
Present address: Alan D. Christian, Department of Biological
Sciences, Arkansas State University, State University,
AR 72467, U.S.A.
Freshwater Biology (2007) 52, 1427–1439 doi:10.1111/j.1365-2427.2007.01756.x
2007 The Authors, Journal compilation 2007 Blackwell Publishing Ltd 1427
strategies for adequate maintenance of genetic varia-
tion in species of conservation interest.
Freshwater mussels of the families Unionidae and
Margaritiferidae (Unionoidea) are among the largest
benthic invertebrates in freshwater ecosystems, and
they can make up a significant proportion of total
benthic biomass (Strayer et al., 2004). As a result, they
can be key elements in energy flow and nutrient
cycling within freshwater ecosystems. Large mussel
beds with densities greater than 100 individuals per
m
2
and total numbers of individuals greater than
440 000 have been recently reported in North Ameri-
can streams and rivers (Christian, 1995; Harris &
Christian, 2000). Thus, these organisms are often key
components of freshwater benthic communities, but
they are also among the most threatened faunal
elements in North America (Lydeard et al., 2004).
In fact, a large proportion of North American
freshwater mussels currently have some form of
conservation status at national or regional levels.
Declining abundance and richness of freshwater
mussels are due to combinations of habitat modifica-
tion (impoundment of rivers, alterations in land use),
pollution, commercial harvesting and the relatively
recent invasion of North America by dreissenid
mussels (Bogan, 1993). While over 70%of North
American mussel taxa are of conservation concern
(Williams et al., 1993), there is little evidence that any
mussel taxa are currently stable. Even relatively
common species have shown declines in abundance
over time (Anthony & Downing, 2001). In this sense,
all species of freshwater mussels appear to be at
elevated risk of extinction. The precipitous decline of
North America’s freshwater mussel fauna has attrac-
ted the attention of researchers and management
agencies, but effective conservation planning is still
hampered by lack of basic biological knowledge about
these organisms.
Little is known of the population genetic structure
of freshwater mussels. While a number of studies
have described genetic variation in these organisms
(Mulvey et al., 1997; Berg et al., 1998; Berg & Berg,
2000; Curole, Foltz & Brown, 2004), most have focused
on species that are already highly endangered, and in
which much of the genetic structure has presumably
been lost due to the small size of individual popula-
tions and/or the greatly reduced number of popula-
tions remaining (for examples, see Stiven &
Alderman, 1992; Machordom et al., 2003). Because
we maintain that all freshwater mussels are at
elevated risk of extinction, we believe it is important
to study species that remain relatively common across
most of their ranges and hence might still retain their
original genetic structuring.
In this paper, we use allozyme electrophoresis to
describe genetic structure of three of the most abun-
dant species of freshwater mussels in the Darby Creek
ecosystem of central Ohio. Darby Creek is a small
tributary of the Scioto River, which itself is a major
tributary of the Ohio River. All three of these species
range widely throughout the Mississippi River basin
(Parmalee & Bogan, 1998). We quantify within-
population and among-population genetic variation
of these species at multiple sites in the Darby system
and compare results among species. Based on an
earlier population genetic study of the unionid
Quadrula quadrula (Rafinesque, 1820) which found
almost no geographical pattern in allozyme variation
along approximately 1000 km of the Ohio River (Berg
et al., 1998), we predicted that species along the
<200 km of Darby Creek that we sampled would
show little or no geographical structure. By combining
our results quantifying genetic variation of small-
stream populations with the earlier results from the
large Ohio River, we develop a hypothesis to explain
genetic structure of freshwater mussel species based
on river size and the associated fish faunas.
Methods
Study areas, sample collection and electrophoresis
Freshwater mussels were collected from 15 sites on
Big Darby Creek and its largest tributary, Little Darby
Creek, in central Ohio, from 1994 to 1998. The Darby
Creek ecosystem is of considerable conservation
interest because of its diversity of freshwater mussels
(41 species reported; Watters, 1996) and fishes (>90
species; http://bigdarby.org). These numbers are
remarkable for small Midwestern streams, and have
earned the Darby Creek ecosystem a listing as one of
The Nature Conservancy’s Last Great Places. How-
ever, this outstanding aquatic resource is being
threatened by land use changes due to suburban
sprawl associated with the city of Columbus (Watters,
1996).
Mussels were collected from a 60-km stretch of Big
Darby Creek in Union and Madison counties up-
1428 D.J. Berg et al.
2007 The Authors, Journal compilation 2007 Blackwell Publishing Ltd, Freshwater Biology,52, 1427–1439
stream of the confluence with Little Darby Creek,
from a 28-km stretch of Little Darby Creek in Madison
County, and from a single site below the confluence of
these streams in Franklin County (Fig. 1). Elliptio
dilatata (Rafinesque) were collected from five sites on
Big Darby Creek (river kilometres 92.1, 103.9, 112.9,
116.5, 123.1) and five sites on Little Darby Creek (river
kilometres 12.4, 19.4, 28.1, 31.9, 40.5). Lampsilis siliq-
uoidea (Barnes) were collected at three sites on Big
Darby Creek (river kilometres 63.5, 100.8, 106.0) and
two sites on Little Darby Creek (river kilometres 12.4,
19.4). Ptychobranchus fasciolaris (Rafinesque) were col-
lected from three sites on Big Darby Creek (river
kilometres 53.2, 100.8, 106.0) and two sites on Little
Darby Creek (river kilometres 12.4, 31.8).
At each site, 20–50 individuals per species were
collected by snorkeling in water <1.5 m deep. Most
mussels were non-destructively sampled using a
mantle biopsy (Berg et al., 1995), while a small
number of individuals (<5 per site) were destructively
sampled. Tissues were flash-frozen in liquid nitrogen
and stored at )80 C until analysed. Allozyme
electrophoresis using starch and cellulose acetate gels
was performed for a variety of enzyme systems using
standard recipes (Harris & Hopkinson, 1976; Hebert &
Beaton, 1989) and buffer systems (Selander et al., 1971;
Clayton & Tretiak, 1972; Hebert & Beaton, 1989) with
modifications. Fifteen loci were resolved for E. dilatata,
12 for L. siliquoidea and 14 for P. fasciolaris (Table 1).
Tissue samples were homogenised in 28 lLof2%
2-phenoxyethanol as a grinding buffer and centri-
fuged at 14 000 g. Supernatant liquids were then
loaded onto filter paper wicks (for starch gels) or
cellulose acetate plates. Gels were run between 40 min
and 5 h, depending on the gel and buffer systems
used, with starch gels always running much longer
than cellulose acetate gels. Alleles were identified by
assigning a 1 to the allele that migrated furthest
anodally (the fastest allele) and 2, 3, etc. to the second
fastest, third fastest, etc. When two loci were present,
Fig. 1 Map of the Darby Creek drainage system (a tributary of
the Scioto River, itself a tributary of the Ohio River) in Ohio,
showing approximate locations of sample sites as river kilo-
meters from the mouths of Big Darby (BD) and Little Darby (LD)
creeks.
Table 1 Enzyme systems and buffers used for allozyme analysis
of freshwater mussels from the Darby Creek system
System
Enzyme
number
No.
loci Gel type Buffer Species
LAP 1 Cellulose acetate CAM E, L, P
CK 2.7.3.2 1 Cellulose acetate CAM E
MDH 1.1.1.37 2 Cellulose acetate TEB E, P
2 Starch TC 6.7 L
GPI 5.3.1.9 1 Cellulose acetate TEB E
1 Cellulose acetate TG L, P
PGM 2.7.5.1 2 Cellulose acetate TEB E
2 Cellulose acetate TEM L, P
PEP* 3.4.11 1 Cellulose acetate TEB E
1 Cellulose acetate TG P
ENOL 4.2.1.11 1 Cellulose acetate TEM E
MPI 5.3.1.8 2 Cellulose acetate TEB E
1 Cellulose acetate TEB L, P
SOD 1.15.1.1 1 Cellulose acetate TG E, P
EST 3.1.1.1 2 Cellulose acetate TG E
2 Cellulose acetate TEM L, P
HEX 2.7.1.1 1 Cellulose acetate TG E
1 Cellulose acetate TEB L, P
AAT 2.6.1.1 1 Starch TC 6.7 L
1 Cellulose acetate TEB P
IDH 1.1.1.42 1 Starch TC 6.7 L, P
E¼Elliptio dilatata,L¼Lampsilis siliquoidea,P¼Ptychobranchus
fasciolaris.
*Leucine-valine as dipeptide.
Freshwater mussel population genetics 1429
2007 The Authors, Journal compilation 2007 Blackwell Publishing Ltd, Freshwater Biology,52, 1427–1439
the fastest (most anodal) was designated 1. The
number of loci analysed was limited by the small
amounts of tissue available for many individuals. At
least 20 individuals were analysed for all loci for each
population; additional individuals were analysed for
each variable locus. Loci were considered ‘polymor-
phic’ when the most common allele was present at
a frequency £0.95. Because the amount of tissue
available from each individual was often limited and
occasionally gels did not produce results, sample size
varied from 9 to 50 for each locus-by-population
combination (Tables 2–4). Sample size per locus
across all population-by-locus combinations averaged
36.4 individuals (±0.7 SE, n¼280 combinations).
Table 2 Allele frequencies at variable loci for Elliptio dilatata from Big and Little Darby Creeks. Population codes follow Figure 1
Locus Allele
Populations
BD 92.1 BD 103.9 BD 112.9 BD 116.5 BD 123.1 LD 12.4 LD 19.4 LD 28.1 LD 31.9 LD 40.5
LAP n ¼48* n¼49* n¼40* n¼18 n¼48* n¼50 n¼47 n¼39 n¼28* n¼50*
1 0.219 0.296 0.288 0.250 0.208 0.120 0.085 0.103 0.143 0.180
2 0.292 0.184 0.250 0.333 0.219 0.150 0.117 0.205 0.054 0.160
3 0.438 0.388 0.300 0.306 0.313 0.620 0.277 0.590 0.679 0.490
4 0.052 0.133 0.162 0.111 0.260 0.110 0.521 0.103 0.125 0.170
CK n ¼50 n¼50 n¼50 n¼47 n¼50 n¼50 n¼49 n¼15 n¼28 n¼50
1 0.033
2 1.000 1.000 1.000 1.000 1.000 1.000 1.000 0.967 1.000 1.000
MDH-1 n ¼49 n¼50 n¼50 n¼37 n¼50 n¼21 n¼49 n¼25 n¼28 n¼49
1 0.010 0.024 0.082 0.054 0.071
2 0.796 0.800 0.770 0.851 0.750 0.810 0.878 0.920 0.893 0.867
3 0.204 0.200 0.220 0.149 0.250 0.167 0.041 0.080 0.054 0.061
GPI n ¼50 n¼50 n¼50* n¼47 n¼50 n¼50 n¼49 n¼40 n¼28 n¼50
1 0.980 1.000 0.820 1.000 1.000 1.000 1.000 1.000 1.000 1.000
2 0.020 0.180
PGM-1 n ¼39 n¼39 n¼49 n¼22 n¼37 n¼23 n¼49* n¼39 n¼28 n¼40
1 0.031
2 0.923 0.808 0.898 0.909 0.878 0.696 0.837 0.962 0.982 0.963
3 0.077 0.192 0.102 0.091 0.122 0.304 0.133 0.039 0.018 0.038
PGM-2 n ¼50* n¼50* n¼49 n¼47 n¼50 n¼45 n¼49 n¼40 n¼28 n¼40
1 0.400 0.290 0.429 0.394 0.340 0.133 0.143 0.025 0.054 0.063
2 0.180 0.100 0.031 0.032 0.080 0.133 0.020 0.013 0.071 0.038
3 0.420 0.610 0.541 0.575 0.580 0.733 0.827 0.963 0.875 0.900
4 0.010
PEP n ¼34 n¼50* n¼35 n¼22 n¼40 n¼33 n¼46 n¼35 n¼28 n¼44*
1 0.941 0.790 0.957 0.886 0.950 0.909 0.663 0.829 0.839 0.568
2 0.059 0.210 0.043 0.114 0.050 0.091 0.337 0.171 0.161 0.432
SOD n ¼41 n¼44 n¼29 n¼34 n¼30 n¼50 n¼48 n¼20 n¼20 n¼50
1 1.000 1.000 1.000 1.000 0.967 1.000 1.000 1.000 1.000 1.000
2 0.033
EST-1 n ¼49 n¼50 n¼50 n¼17 n¼50 n¼50 n¼49 n¼25 n¼28 n¼50
1 1.000 1.000 1.000 1.000 1.000 1.000 1.000 0.940 1.000 1.000
2 0.060
Mean sample
size per locus
41.9 (2.8) 46.1 (1.9) 46.8 (1.7) 34.5 (3.0) 47.0 (1.6) 44.1 (2.6) 48.3 (0.2) 27.3 (2.8) 26.8 (0.8) 45.2 (2.4)
Mean number
of alleles per
locus
1.6 (0.2) 1.5 (0.2) 1.7 (0.3) 1.5 (0.2) 1.6 (0.2) 1.6 (0.3) 1.7 (0.3) 1.7 (0.2) 1.6 (0.3) 1.6 (0.3)
%P(95%
criterion)
33.3 33.3 33.3 33.3 33.3 33.3 33.3 26.7 26.7 26.7
Mean
heterozygosity
0.090 0.10 0.11 0.11 0.12 0.12 0.10 0.08 0.08 0.08
Numbers in parentheses are standard errors. n¼number of individuals.
*Locus-by-population combinations with heterozygote deficiencies.
1430 D.J. Berg et al.
2007 The Authors, Journal compilation 2007 Blackwell Publishing Ltd, Freshwater Biology,52, 1427–1439
Table 3 Allele frequencies at variable loci for Lampsilis siliquo-
idea from Big and Little Darby Creeks. Population codes follow
Fig. 1
Locus Allele
Populations
BD 63.5 BD 100.8 BD 106.0 LD 12.4 LD 19.4
LAP n ¼27* n¼37** n¼28** n¼42** n¼20
1 0.333 0.054 0.518 0.357 0.175
2 0.333 0.378 0.333 0.550
3 0.259 0.162 0.482 0.298 0.050
4 0.074 0.297 0.012 0.200
5 0.108 0.025
MDH-1 n ¼33 n¼37 n¼33 n¼45 n¼20
1 1.000 1.000 1.000 0.978 1.000
2 0.022
GPI n ¼34 n¼37 n¼33 n¼45 n¼20
1 0.118 0.041 0.152 0.133 0.325
2 0.882 0.960 0.849 0.867 0.675
PGM-1 n ¼30* n¼35* n¼31* n¼31* n¼18
1 0.283 0.471 0.484 0.613 0.417
2 0.183 0.429 0.435 0.323 0.444
3 0.433 0.100 0.081 0.065 0.111
4 0.100 0.028
PGM-2 n ¼32* n¼36 n¼34* n¼45* n¼20
1 0.078 0.153 0.118 0.056 0.050
2 0.781 0.847 0.809 0.900 0.950
3 0.141 0.074 0.044
MPI n ¼34* n¼24 n¼33 n¼45 n¼13
1 0.941 1.000 1.000 1.000 1.000
2 0.059
EST-1 n ¼28* n¼37 n¼25 n¼36* n¼20
1 0.071 0.054 0.100 0.056 0.050
2 0.500 0.608 0.740 0.458 0.600
3 0.429 0.311 0.120 0.458 0.350
4 0.027 0.040 0.028
HEX n ¼26 n¼26* n¼27 n¼45 n¼10
1 0.962 0.769 1.000 1.000 0.050
2 0.039 0.231 0.950
AAT n ¼15 n¼9n¼33 n¼45* n¼20
1 1.000 1.000 1.000 0.600 1.000
2 0.400
IDH n ¼33 n¼37 n¼33 n¼27 n¼20
1 1.000 1.000 0.985 1.000 1.000
2 0.015
Mean sample
size per locus
30.0 (1.6) 32.4 (2.5) 31.3 (0.9) 41.3 (1.8) 18.4 (1.0)
Mean number
of alleles per
locus
2.1 (0.3) 2.0 (0.4) 1.8 (0.3) 2.1 (0.3) 2.0 (0.4)
%P(95%cri-
terion)
50.0 41.7 41.7 50.0 50.0
Mean hetero-
zygosity
0.11 0.16 0.17 0.12 0.16
Numbers in parentheses are standard errors. n¼number of
individuals.
*Locus-by-population combinations with heterozygote defici-
encies.
**Locus-by-population combinations with heterozygote excesses.
Table 4 Allele frequencies at variable loci for Ptychobranchus
fasciolaris from Big and Little Darby Creeks. Population codes
follow Fig. 1
Locus Allele
Populations
BD 53.2 BD 100.8 BD 106.0 LD 12.4 LD 31.8
LAP n ¼28* n¼46* n¼29* n¼40 n¼25
1 0.089 0.120 0.069 0.100 0.360
2 0.768 0.533 0.379 0.900 0.600
3 0.143 0.348 0.552 0.040
MDH-1 n ¼39 n¼49 n¼31 n¼36 n¼45
1 1.000 1.000 0.936 1.000 1.000
2 0.065
MDH-2 n ¼39 n¼39 n¼31 n¼31 n¼33
1 1.000 1.000 0.968 1.000 1.000
2 0.032
GPI n ¼37 n¼49 n¼30 n¼40 n¼45
1 0.068 0.112 0.050 0.100 0.133
2 0.162 0.163 0.150 0.113 0.133
3 0.419 0.388 0.400 0.488 0.578
4 0.324 0.265 0.317 0.300 0.156
5 0.027 0.071 0.067
6 0.017
PGM-1 n ¼38* n¼46* n¼29* n¼39 n¼32
1 0.342 0.261 0.431 0.449 0.563
2 0.461 0.609 0.431 0.551 0.422
3 0.197 0.120 0.138 0.016
4 0.011
PGM-2 n ¼38* n¼48* n¼30* n¼41 n¼38*
1 0.487 0.458 0.533 0.500 0.500
2 0.382 0.427 0.350 0.500 0.487
3 0.118 0.115 0.067 0.013
4 0.013 0.050
PEP n ¼37* n¼49 n¼31* n¼40 n¼26
1 0.865 1.000 0.903 1.000 0.923
2 0.135 0.065 0.077
3 0.016
4 0.016
MPI n ¼35 n¼49 n¼31 n¼40 n¼31
1 0.914 0.888 0.952 1.000 1.000
2 0.086 0.112 0.048
EST-1 n ¼39 n¼49* n¼31 n¼41 n¼45
1 1.000 0.949 0.968 0.951 1.000
2 0.051 0.032 0.049
EST-2 n ¼31* n¼29 n¼27* n¼33 n¼36
1 0.855 1.000 0.926 1.000 1.000
2 0.145 0.074
HEX n ¼35 n¼47 n¼31* n¼40 n¼32
1 1.000 0.957 0.742 1.000 1.000
2 0.043 0.258
IDH n ¼38 n¼39 n¼31 n¼32 n¼36
1 0.013
2 0.987 1.000 1.000 1.000 1.000
Mean sample
size per locus
36.0 (1.0) 44.7 (1.6) 30.3 (0.3) 36.9 (1.1) 35.4 (1.7)
Mean number
of alleles per
locus
2.1 (0.3) 2.0 (0.3) 2.5 (0.4) 1.5 (0.2) 1.7 (0.3)
Freshwater mussel population genetics 1431
2007 The Authors, Journal compilation 2007 Blackwell Publishing Ltd, Freshwater Biology,52, 1427–1439
Statistical analyses
Electrophoretic data were analysed using standard
population genetic techniques contained in Tools for
Population Genetic Analysis (TFPGA; Miller, 1997).
Descriptive statistics calculated for each population
included mean number of alleles per locus, percent-
age polymorphic loci (95%level), and average direct-
count heterozygosity. Comparisons of measured
genotype frequencies with Hardy–Weinberg (H-W)
expectations were evaluated using the Monte Carlo
simulation method of TFPGA. To minimise the
probability of a type 1 error, we adjusted significance
levels to an experimentwise error rate of a¼0.05
using the ‘sharper’ sequential-comparison Bonferroni
(s-cB) technique described by Lessios (1992).
For polymorphic loci, allele frequencies among
populations were compared using a Monte Carlo
simulation of Fisher’s Exact Test for RxC contingency
tables in TFPGA, followed by s-cB to hold the exper-
imentwise error rate at a¼0.05. Among-population
genetic variation was further analysed by performing
an analysis of molecular variance (
AMOVAAMOVA
) using
the standard F
ST
values estimation procedure in
GENALEX6GENALEX6
(Peakall & Smouse, 2005) with interpolation
of missing data and by calculating hierarchical values
of h, the F
ST
analogue of Weir & Cockerham (1984), for
polymorphic loci. These values of hwere bootstrapped
(1000 replicates) to obtain 95%confidence intervals.
Allele frequency differences were integrated across loci
by calculating unbiased genetic distances for all pairs
of populations within a species (Nei, 1978). Genetic
similarity of populations was determined by construc-
tion of dendrograms using genetic distance and the
unweighted-pair-group method of analysis (UPGMA)
to cluster populations (Sokal & Sneath, 1963). Bootstrap
values for the nodes of the dendrogram were based on
1000 replicates. Mantel tests were performed between
matrices of hand geographical distance (calculated as
river kilometres from the mouth of Big Darby or from
the mouth of Little Darby) among pairs of populations
to evaluate the likelihood of isolation-by-distance and
between matrices of residuals of hand geographical
distance to estimate the relative contributions of gene
flow and genetic drift in determining population
genetic structure (Hutchison & Templeton, 1999).
Estimates of hfor this analysis were obtained using
the coancestry distances available from TFPGA (Miller,
Blinn & Keim, 2002).
Results
Within-population variation
Five of the 15 loci sampled were variable in all
populations of E. dilatata (Table 2). Four loci had one
or two variable populations, and six loci were fixed
for a single allele each. Twelve of 47 locus-by-
population combinations, representing five different
loci, showed genotype frequencies that did not match
H-W expectation because of heterozygote deficits.
Of the 12 loci sampled for L. siliquoidea, five were
polymorphic in all populations (Table 3), five were
variable in 1–3 populations, and two were fixed.
Sixteen of 29 locus-by-population combinations
showed genotype frequencies that were different
from H-W expectation. Thirteen of these were hetero-
zygote deficits. Seven different loci exhibited a hetero-
zygote deficit in at least one population, while all
heterozygote excesses were at the LAP locus.
Variation was detected at 12 of 14 loci sampled for
P. fasciolaris (Table 4). Four loci were polymorphic in
all five populations, three loci had three variable
populations, and five loci had 1–2 variable popula-
tions. Two loci were fixed. Sixteen of 30 locus-by-
population combinations had genotype frequencies
that were different from H-W expectation. These
combinations involved six different loci and all differ-
ences were due to heterozygote deficits. Fifteen of these
differences were found in Big Darby populations.
Among-population variation
For E. dilatata populations, pairwise differences in
allele frequencies at variable loci were low between
Table 4 (Continued)
Locus
Populations
BD 53.2 BD 100.8 BD 106.0 LD 12.4 LD 31.8
%P(95%
criterion)
50.0 42.9 57.1 28.6 35.7
Mean
hetero-
zygosity
0.12 0.13 0.11 0.11 0.12
Numbers in parentheses are standard errors. n¼number of
individuals.
*Locus-by-population combinations with heterozygote defici-
encies.
1432 D.J. Berg et al.
2007 The Authors, Journal compilation 2007 Blackwell Publishing Ltd, Freshwater Biology,52, 1427–1439
populations from the same creek, but high when Big
Darby populations were compared with those from
Little Darby. For Big Darby Creek, nine of 90 (10%)
locus-by-population-pair combinations showed signi-
ficant differences in allele frequencies. Among Little
Darby Creek comparisons, 13 of 90 (14%) were
significant. Seventy-two of 225 (32%) comparisons of
Big Darby populations with Little Darby populations
showed significant shifts in allele frequencies. Indi-
vidual locus values (not shown) for hwere low but
variable (all £0.181), while overall values were signi-
ficant (Table 5). The
AMOVAAMOVA
revealed that 90%of
variation was found within populations, while 4%
was among populations within streams, and 6%was
among streams. All these estimates were significant
(P¼0.001, 999 permutations). Unbiased genetic dis-
tances averaged 0.0118 (n¼45 pairs; range 0.0005–
0.0278) among all pairs of populations. Average
distances among populations within streams were
0.0024 (n¼10) for Big Darby and 0.0095 (n¼10) for
Little Darby, while average distance for pairs of
populations consisting of one population from each
stream was 0.0165 (n¼25). A dendrogram (Fig. 2)
based on genetic distance showed two major bran-
ches, one containing all Big Darby populations and
the other containing the Little Darby populations.
However, bootstrap support for the best tree was
minimal, with only two of eight nodes supported in
more than half of all permutations. For pairs of
populations, hand geographical distance were signi-
ficantly correlated (Fig. 3; r¼0.672, P¼0.0040, 999
permutations), but residuals of hwere not correlated
with geographical distance (r¼)0.233, P¼0.9320,
999 permutations).
For L. siliquoidea, an average of 27%(±2.1%SE,
range ¼20–40%,n¼100 comparisons) of variable
loci had significant differences in allele frequencies for
pairs of populations. Individual locus values (not
shown) for hwere highly variable, ranging from 0.000
to 0.558. While h
PS
was significantly different
from zero, h
S
was not (Table 5). Within-population
variation accounted for 83%of total variation, while
variation among populations within streams com-
prised the remaining 17%. Both of these values were
significant (
AMOVAAMOVA
,P¼0.001, 999 permutations),
while variation among streams was not significantly
different from zero (P¼1.000). Over all pairs of
populations, average unbiased genetic distance was
0.0642 (n¼10 pairs; range ¼0.0234–0.1392). Genetic
distance within stream was greater for Little Darby
(0.1345, n¼1) than Big Darby (average ¼0.0289, n¼
3), while comparisons between Big and Little Darby
populations were intermediate (average ¼0.0701,
n¼6). A dendrogram based on these distances
showed no pattern and had little bootstrap support
(tree not shown; 1 node 62%support, all others <27%
support). Geographical distance was not correlated
with hamong pairs of populations (Fig. 3; r¼0.1146,
P¼0.2980, 999 permutations), nor with residuals of h
(r¼0.1304, P¼0.4220, 999 permutations).
Patterns of allele frequencies among populations of
P. fasciolaris were similar to that of E. dilatata.
Comparisons between Big Darby populations and
between Little Darby populations rarely showed
Table 5 Hierarchical F-statistics for freshwater mussels from Darby Creek. Values of Fcorrespond to Wright’s F
IT
;q
PS
is differenti-
ation among populations within a stream, q
S
is differentiation among streams (both are components of Wright’s F
ST
); values of f are
inbreeding coefficients (Wright’s F
IS
). Table entries are mean values over all loci (lower, upper bounds of 95%confidence intervals)
Species Fh
PS
h
S
f
Elliptio dilatata 0.2971 (0.1762, 0.3917) 0.1005 (0.0507, 0.1571) 0.0644 (0.0161, 0.1332) 0.2186 (0.1000, 0.3253)
Lampsilis siliquoidea 0.4214 (0.1991, 0.7245) 0.1244 (0.0541, 0.2587) )0.1004 ()0.4003, 0.0411) 0.3392 (0.1021, 0.6559)
Ptychobranchus fasciolaris 0.4249 (0.2219, 0.7053) 0.0592 (0.0141, 0.1404) 0.0321 (0.0073, 0.0785) 0.3887 (0.2014, 0.6542)
Fig. 2 Nei unbiased genetic distances (Nei, 1978) for popula-
tions of Elliptio dilatata from the Darby Creek catchment. Popu-
lation codes follow Fig. 1. Distances were calculated using 15
allozyme loci and a tree was constructed using UPGMA. Values
at the nodes represent the percentages of 1000 bootstrapped
trees that produced these nodes. Unlabelled nodes had values
<50%.
Freshwater mussel population genetics 1433
2007 The Authors, Journal compilation 2007 Blackwell Publishing Ltd, Freshwater Biology,52, 1427–1439
significant differences in allele frequencies (2 of 12
variable loci for BD, 1 of 12 for LD). Population BD
53.2 (the site below the confluence of the two streams)
was similar to the other Big Darby populations
(differences at 2 of 12 loci for each comparison).
Allele frequencies were much more variable when
comparing Little Darby populations with Big Darby
populations; of 72 variable-locus-by-population pairs,
22 (31%) had significant differences in allele frequen-
cies. Individual locus values for hwere all <0.196; over
all loci, both components of hwere significantly
different than zero (Table 5). Analysis of molecular
variation revealed that 94%of total variation was
found within populations, 3%was among popula-
tions within streams, and 3%was among streams. All
these values were significant (
AMOVAAMOVA
,P¼0.001, 999
permutations). Average genetic distance was 0.0144
(n¼10 pairs; range ¼0.0061–0.0312) when compar-
ing all pairs of populations. Within stream distances
were similar (0.0088 for Big Darby and 0.0074 for Little
Darby), as were values for the lower Big Darby
compared with each of these (average ¼0.0144 versus
Big Darby and 0.0087 versus Little Darby; n¼2
comparisons each). Comparison of Big Darby versus
Little Darby showed the greatest distances (aver-
age ¼0.0212, n¼4). No pattern was seen in the
dendrogram based on genetic distances and bootstrap
support was very low (tree not shown; 1 node 50%
support, all others <47%support). The correlation of
geographical distance with hfor pairs of populations
was marginally insignificant with P-values varying
between 0.0477 and 0.0535 (Fig. 3; r¼0.675; we ran
five sets of 10 000 permutations each because of the
‘borderline’ values). Geographical distance was not
significantly correlated with residuals of h(r¼0.4369,
P¼0.0910, 999 permutations).
Discussion
Our measures of within-population genetic variation
are similar to those previously reported for freshwa-
ter mussels. For E. dilatata, our measure of average
heterozygosity (H)¼0.10 is similar to a value of 0.11
for a single population (Davis, 1984). Lampsilis
siliquoidea average heterozygosities of 0.14 for our
study were similar to those of 0.11 previously
reported by Davis (1984). Our average values for
both these species and for P. fasciolaris (H¼0.12) are
in the middle of the ranges reported from several
surveys of heterozygosity in unionoid bivalves (Berg
et al., 1998; Curole et al., 2004). Deviations of geno-
type frequencies from H-W expectations occurred in
a significant proportion (25–55%) of polymorphic
(a)
(b)
(c)
Fig. 3 Population differentiation (h) as a function of geograph-
ical distance for three species of freshwater mussels from the
Darby Creek catchment (a) Elliptio dilatata; (b) Lampsilis siliquo-
idea; (c) Ptychobranchus fasciolaris. Correlations (r¼0.67, 0.11 and
0.68, respectively) were significant for E. dilatata and P. fasciolaris
(Mantel test, P£0.05).
1434 D.J. Berg et al.
2007 The Authors, Journal compilation 2007 Blackwell Publishing Ltd, Freshwater Biology,52, 1427–1439
locus-by-population combinations. Reported values
for freshwater mussels have been quite variable, with
anywhere from 0%to 75%of locus-by-population
combinations exhibiting deviation from expected
heterozygote frequencies (summarised in Berg et al.,
1998 and Johnson, Liang & Farris, 1998). Like our
results, most of these deviations are caused by
heterozygote deficiencies.
Such heterozygote deficiencies have been noted in
many studies of marine bivalves utilising both
allozyme (reviewed by Berger, 1983) and microsatel-
lite markers (Launey & Hedgecock, 2001). In our
study, mutation is an unlikely factor because allo-
zymes are relatively conserved genetic markers.
Selection cannot be ruled out; in particular, the LAP
locus seems to show consistent deviation from H-W
expectation in all three species, although these devi-
ations include heterozygous excesses in three popu-
lations of L. siliquoidea, but a deficit in a fourth.
Selection against deleterious recessive alleles has been
shown to cause heterozygote deficits in Pacific oysters
(Launey & Hedgecock, 2001) and blue mussels
(Myrand, Tremblay & Se
´vigny, 2002). Genetic drift
coupled with restricted gene flow and inbreeding is
another potential cause of heterozygote deficiencies.
Studies of genetic variation within multiple popula-
tions of Q. quadrula and Amblema plicata (Berg et al.,
1998; Elderkin et al., 2007) showed few or no devia-
tions from H-W expectation. Given that genetically
effective population size is likely to be larger in large
river systems, mussel populations in these large rivers
should be less susceptible to the effects of genetic drift
and inbreeding. This is consistent with our observa-
tion of much greater occurrence of deviation from
H-W expectation in Darby Creek mussels than in
mussels that are typical of larger rivers. Of the three
species we sampled, Darby Creek populations of L.
siliquoidea (4–65 individuals representing 1–27%of all
mussels collected at a site) were generally smaller
than those of E. dilatata (45–400 individuals or 18–73%
of all mussels) or P. fasciolaris (32–70 individuals or 6–
27%of all mussels) (A.D. Christian, unpublished),
and our analysis indicated that the former was
furthest from gene flow–genetic drift equilibrium
(see discussion below). Combinations of restricted
gene flow, inbreeding, and other factors have been
implicated as the cause for heterozygote deficits in
coot clams (Gaffney et al., 1990). Overall, Darby Creek
populations of all three species of freshwater mussels
contain moderate levels of within-population genetic
variation, with many occurrences of heterozygote
deficiencies.
The pattern of among-population genetic variation
was different for L. siliquoidea than for the other two
species. While
AMOVAAMOVA
revealed significant variation
among populations for all three species, both E. dilatata
and P. fasciolaris also showed significant variation
among streams. This pattern was more pronounced in
E. dilatata, but we found some evidence of isolation-
by-distance (indicated by the correlations of geo-
graphical distance and h) in both species. Conversely,
we found no evidence for any sort of geographical
pattern in among-population variation for L. siliquo-
idea. Residuals of hwere larger for L. siliquoidea than
for E. dilatata and P. fasciolaris, and they were not
correlated with geographical distance, indicating that
random effects of genetic drift were greater than the
homogenising effects of gene flow (Hutchison &
Templeton, 1999). For P. fasciolaris, residuals tend to
increase with geographical distance so that the lack of
a significant correlation may be due to the small
sample size (10 comparisons). In this case, isolation-
by-distance and gene flow–genetic drift equilibrium
may be established at the within-basin spatial scale we
examined (Hutchison & Templeton, 1999). Isolation-
by-distance has been reported for other freshwater
mussel species at large geographical scales (>1000 km;
Berg et al., 1998; Elderkin et al., 2007). Geographical
distance among populations was a significant factor in
determining among-population genetic variation
within drainage systems at allozyme loci in a
stream-dwelling fish from Australia (McGlashan,
Hughes & Bunn, 2001). A study using amplified
fragment length polymorphisms to estimate isolation-
by-distance in four species of aquatic insects found
that dispersal ability was negatively correlated with
degree of isolation of populations (Miller et al., 2002).
Given that mussel distributions are not continuous
(they live in ‘beds’), an isolation-by-distance pattern is
best explained by a stepping-stone model of dispersal.
While isolation-by-distance and presence along a
linear stretch of stream appear to be important factors
for E. dilatata and P. fasciolaris, differences among
populations of L. siliquoidea appear to occur more
randomly.
Few other studies have examined among-popula-
tion variation of freshwater mussels within individual
stream systems; most studies have examined such
Freshwater mussel population genetics 1435
2007 The Authors, Journal compilation 2007 Blackwell Publishing Ltd, Freshwater Biology,52, 1427–1439
variation across wider geographical ranges and mul-
tiple drainage basins. While all three species in the
Darby Creek system exhibited significant among-
population variation at distances no greater than 200
river km, populations of the freshwater mussel
Q. quadrula showed no such differentiation along a
stretch of the Ohio and Tennessee rivers greater than
1000 river km (Berg et al., 1998). In fact, these latter
populations showed no significant differences in allele
frequencies among populations along this length of
river, and values of hwere very low, even among
populations separated by >2500 river km (overall h¼
0.031). Amblema plicata, a mussel species common to
both large and small rivers, exhibits even lower
variation among populations within large rivers
(h¼0.017; Elderkin et al., 2007). However, for both
of these species, geographical distances were signifi-
cantly correlated with genetic distances. Based on our
study and these others, it appears that the spatial scale
at which significant allozyme variation occurs among
freshwater mussel populations is quite variable.
While Darby Creek populations are differentiated
within short distances in a small stream system and
this differentiation may reflect effects of genetic drift
and inbreeding, Q. quadrula and A. plicata populations
are similar over much greater distances in larger river
systems.
The complex life cycle of unionoid mussels includes
a stage as glochidia larvae that are most often parasitic
on freshwater fishes. As adult mussels are essentially
sessile, dispersal among populations of freshwater
mussels is a function of host movement and these
movements can determine mussel distributions (Lee,
DeAngelis & Koh, 1998). Because dispersal promotes
gene flow among populations, host fish vagility
should be negatively correlated with measures of
differentiation (h, genetic distance) among mussel
populations. We propose that the lack of genetic
differentiation seen in a large-river mussel species
such as Q. quadrula is probably because of high
vagility of hosts; mussels from large rivers are likely
to use large, highly mobile fishes as hosts – flathead
catfish, Pylodictus olivaris (Rafinesque), in the case
of Q. quadrula [Ohio State University Museum of
Biological Diversity (MBD) Mussel/Host Database;
http://www.biosci.ohio-state.edu/molluscs/OSUM2/].
Conversely, populations of mussels from small
streams, such as those from Darby Creek, would tend
to rely on host fishes that generally move shorter
distances or are unlikely to survive passage in a
mainstem river when moving from one tributary to
another. Lampsilis siliquoidea utilises centrarchids and
minnows (MBD database), while mussels of the genus
Ptychobranchus (the database has no entries for P.
fasciolaris) parasitise darters (MBD database). Purpor-
ted hosts for E. dilatata include both centrarchids and
some fishes that are typical of larger rivers, but the
nature of the evidence in the database casts some
doubt as to which fishes are the relevant hosts for this
species. Other species of the genus Elliptio have been
reported to use darters and other percids (MBD
database), which are more likely to be found in our
study area than are the centrarchids reported in the
MBD database. Individual flathead catfish show
movements of tens or hundreds of kilometers (Vok-
oun & Rabeni, 2005), while both darters and sunfishes
move <1 km (Freeman, 1995). Smallmouth bass,
Micropterus dolomieui (Lacepe
´de), another centrarchid
host for L. siliquoidea, disperse an average of 6.5 km in
streams and individuals occasionally migrate up to
75 km during severe winters (Lyons & Kanehl, 2002).
Darters are typically considered a minimally disper-
sing group with movements of metres (Scalet, 1973),
and limited dispersal by darter hosts has been
implicated as a factor leading to patchy distribution
and poor colonisation ability of the endangered
mussel Alasmidonta heterodon (I. Lea) (McLain & Ross,
2005). Moreover, variation in species richness and
composition of freshwater mussel communities is
much greater in small streams than in larger streams
or across catchments (Watters, 1992; Haag & Warren,
1998; Vaughn & Taylor, 2000). These patterns in which
dispersal characteristics appear to be a major deter-
minant of among-population genetic structure are
consistent with those found for a variety of aquatic
invertebrates (Bilton, Freeland & Okamura, 2001).
We hypothesise that mussels from large rivers
should show relatively little among-population gen-
etic variation, while populations from headwaters and
other small streams should show genetic differenti-
ation that is randomly distributed across small spatial
scales unless gene flow (which is likely a function of
host fish vagility) is very high. At the same time, we
predict that the combination of gene flow among
large-river populations, larger genetically effective
population sizes, and more stable flow regimes
should maintain higher levels of within-population
genetic variation in large rivers than in small streams.
1436 D.J. Berg et al.
2007 The Authors, Journal compilation 2007 Blackwell Publishing Ltd, Freshwater Biology,52, 1427–1439
Conversely, genetic drift and inbreeding, resulting
from factors such as smaller genetically effective
population sizes and/or less stable environmental
conditions, should lead to lower within-population
genetic variation in small streams and more frequent
deviation of genotype frequencies from H-W expec-
tation. Given that dispersal ability is a key factor
structuring freshwater invertebrate communities (Bil-
ton et al., 2001), these same forces are likely to be at
play in shaping patterns of genetic diversity in other
aquatic macroinvertebrates. If so, our hypothesis may
explain patterns of genetic variation for many differ-
ent taxa within stream systems of different sizes.
If our hypothesis does accurately explain patterns
of genetic variation in freshwater mussels, it can
inform the development of conservation strategies for
these organisms. In general, among-population gen-
etic variation will be high in species typical of small
streams. Therefore, any given population is likely to
contain a relatively small proportion of total genetic
variation. In order to preserve much of the total
variation, a large number of populations that are
relatively near to each other will need to be protected
across the entire range of the species. Conversely,
single populations of large-river species should con-
tain much of the total genetic variation present in the
species and protection of a comparatively small
number of populations should occur over large
geographical ranges. Because these differences are
likely to result from differences in host-fish charac-
teristics, conservation of unionoid mussels is inevit-
ably tied to the presence of such fishes. Further testing
of this hypothesis should include use of additional
genetic markers, more species of mussels including
those that utilise a variety of strategies for host
selection (Haag & Warren, 2003), and a variety of
geographical areas.
Acknowledgments
We thank Jan Trybula, Vivianaluxa Gervasio, Chad
Garland, and a number of Miami University under-
graduate students for assisting with field work and
performing electrophoretic analyses. Curt Elderkin
has engaged in discussions with us throughout this
project. Todd Levine, Emy Monroe, Makiri Sei, Rich-
ard Seidel, and several anonymous reviewers greatly
improved the quality of this manuscript. This research
was funded by grants from the Darby Creek program
of the Ohio Chapter of The Nature Conservancy and
the Ohio Sea Grant Development Fund.
References
Anthony J.L. & Downing J.A. (2001) Exploitation
trajectory of a declining fauna: a century of freshwater
mussel fisheries in North America. Canadian Journal of
Fisheries and Aquatic Sciences,58, 2071–2090.
Berg D.J. & Berg P.H. (2000) Conservation genetics of
freshwater mussels: comments on Mulvey et al. Con-
servation Biology,14, 1920–1923.
Berg D.J., Cantonwine E.G., Hoeh W.R. & Guttman S.I.
(1998) Genetic structure of Quadrula quadrula (Bivalvia:
Unionidae): little variation across large distances.
Journal of Shellfish Research,17, 1365–1373.
Berg D.J., Haag W.R., Guttman S.I. & Sickel J.B. (1995)
Mantle biopsy: a technique for nondestructive tissue-
sampling of freshwater mussels. Journal of the North
American Benthological Society,14, 577–581.
Berger E.M. (1983) Population genetics of marine gastro-
pods and bivalves. In The Mollusca, Volume 6: Ecology
(Ed. Russell-Hunter W.D.), pp. 563–596. Academic
Press: Orlando, FL.
Bilton D.T., Freeland J.R. & Okamura B. (2001) Dispersal
in freshwater invertebrates. Annual Review of Ecology
and Systematics,32, 159–181.
Bogan A.E. (1993) Freshwater bivalve extinctions (Mol-
lusca: Unionoida): a search for causes. American Zoolo-
gist,33, 599–609.
Christian, A.D. (1995). Analysis of the commercial mus-
sel beds in the Cache and White rivers in Arkansas. MS
thesis, Arkansas State University, State University, AR,
197pp.
Clayton J.W. & Tretiak D.N. (1972) Amine citrate buffers
in starch gel electrophoresis. Journal of the Fisheries
Research Board of Canada,29, 1169–1172.
Curole J.P., Foltz D.W. & Brown K.A. (2004) Extensive
allozyme monomorphism in a threatened species of
freshwater mussel, Margaritifera hembeli Conrad (Bi-
valvia: Margaritiferidae). Conservation Genetics,5, 271–
278.
Davis G.M. (1984) Genetic relationships among some
North American Unionidae (Bivalvia): sibling species,
convergence, and cladistic relationships. Malacologia,
25, 629–648.
Elderkin C.L., Christian A.D., Vaughn C.C., Metcalfe-
Smith J.L. & Berg D.J. (2006) Population genetics of
the freshwater mussel, Amblema plicata (Say 1817)
(Bivalvia: Unionidae): evidence of high dispersal
and post-glacial colonization. Conservation Genetics,8,
355–372.
Freshwater mussel population genetics 1437
2007 The Authors, Journal compilation 2007 Blackwell Publishing Ltd, Freshwater Biology,52, 1427–1439
Freeman M.C. (1995) Movements by two small fishes in a
large stream. Copeia,1995, 361–367.
Gaffney P.M., Scott T.M., Koehn R.K. & Diehl W.J. (1990)
Interrelationships of heterozygosity, growth rate and
heterozygote deficiencies in the coot clam, Mulinia
lateralis.Genetics,124, 687–699.
Haag W.R. & Warren M.L., Jr (1998) Role of ecological
factors and reproductive strategies in structuring
freshwater mussel communities. Canadian Journal of
Fisheries and Aquatic Sciences,55, 297–306.
Haag W.R. & Warren, M.L., Jr (2003) Host fishes and
infection strategies of freshwater mussels in large
Mobile Basin streams, USA. Journal of the North
American Benthological Society,22, 78–91.
Harris J.L. & Christian A.D. (2000) Current Status of the
Freshwater Mussel Fauna of the White River, Arkansas,
River Miles 10-255. Final Report, Memphis District, U.S.
Army Corps of Engineers, Memphis, TN.
Harris H. & Hopkinson D.A. (1976) Handbook of Enzyme
Electrophoresis in Human Genetics. American Elsevier
Publishing Company, New York.
Hebert P.D.N. & Beaton M.J. (1989) Methodologies for
Allozyme Analysis Using Cellulose Acetate Electrophoresis.
Helena Laboratories, Beaumont, TX.
Hutchison D.W. & Templeton A.R. (1999) Correlation of
pairwise genetic and geographic distance measures:
inferring the relative influences of gene flow and drift
on the distribution of genetic variability. Evolution,53,
1898–1914.
Johnson R.L., Liang F.Q. & Farris J.L. (1998) Genetic
diversity among four Amblemini species (Bivalvia:
Unionidae) in the Cache and White Rivers, Arkansas.
The Southwestern Naturalist,43, 321–332.
Launey S. & Hedgecock D. (2001) High genetic load in the
Pacific oyster Crassostrea gigas.Genetics,159, 255–265.
Lee H.-L., DeAngelis D. & Koh H.-L. (1998) Modeling
spatial distribution of the unionid mussels and the
core-satellite hypothesis. Water Science and Technology,
38, 73–79.
Lessios H.A. (1992) Testing electrophoretic data for
agreement with Hardy-Weinberg expectations. Marine
Biology,112, 517–523.
Lydeard, C., Cowie R.H., Ponder W.F. et al. (2004) The
global decline of nonmarine mollusks. Bioscience,54,
321–330.
Lyons J. & Kanehl P. (2002) Seasonal movements of
smallmouth bass in streams. In: Black Bass: Ecology,
Conservation, and Management (Ed. Philipp D.P. &
Ridgeway M.S.), pp. 149–160. Symposium 31, Ameri-
can Fisheries Society, Bethesda, MD.
Machordom A., Araujo R., Erpenbeck D. & Ramos M.A.
(2003) Phylogeography and conservation genetics of
endangered European Margaritiferidae (Bivalvia:
Unionoidea). Biological Journal of the Linnean Society,
78, 235–252.
McGlashan D.J., Hughes J.M. & Bunn S.E. (2001) Within-
drainage population genetic structure of the fresh-
water fish Pseudomugil signifer (Pseudomugilidae) in
northern Australia. Canadian Journal of Fisheries and
Aquatic Sciences,58, 1842–1852.
McLain D.C. & Ross M.R. (2005) Reproduction based on
local patch size of Alasmidonta heterodon and dispersal
by its darter host in the Mill River, Massachusetts,
USA. Journal of the North American Benthological Society,
24, 139–147.
Miller M.P. (1997) Tools for Population Genetic Analysis
(TFPGA), Ver. 1.31. http://www.marksgeneticsoftware.
net.
Miller M.P., Blinn D.W. & Keim P. (2002) Correlations
between observed dispersal capabilities and patterns
of genetic differentiation in populations of four aquatic
insect species from the Arizona White Mountains,
U.S.A. Freshwater Biology,47, 1660–1673.
Mulvey M.C., Lydeard C., Pyer D.L., Hicks K.M., Brim-
Box J., Williams J.D. & Butler R.S. (1997) Conservation
genetics of North American freshwater mussels
Amblema and Megalonaias.Conservation Biology,11,
868–878.
Myrand B., Tremblay R. & Se
´vigny J.-M. (2002) Selection
against blue mussel (Mytilus edulis L.) homozygotes
under stressful conditions. The Journal of Heredity,93,
238–248.
Nei M. (1978) Estimation of average heterozygosity and
genetic distance from a small number of individuals.
Genetics,89, 583–590.
Parmalee P.W. & Bogan A.E. (1998) The Freshwater
Mussels of Tennessee. The University of Tennessee
Press, TN, Knoxville.
Peakall R. & Smouse P.E. (2005) GenAlEx 6: Genetic
Analysis in Excel. Population Genetic Software for Teaching
and Research. Australian National University, Canberra,
Australia. http://www.anu.edu.au/BoZo/GenAlEx/.
Saccheri I., Kuussaari M., Kankare M., Vikman P.,
Fortelius W. & Hanski I. (1998) Inbreeding and
extinction in a butterfly metapopulation. Nature,392,
491–494.
Scalet, C.G. (1973) Stream movements and population
density of the orangebelly darter, Etheostoma radiosum
cyanorum (Osteichthyes: Percidae). Southwestern Natur-
alist,17, 381–387.
Selander R.K., Smith M.H., Yang S.Y., Johnson W.E. &
Gentry J.B. (1971) Biochemical polymorphism and
systematics in the genus Peromyscus. 1. Variation in
the old field mouse (Peromyscus polionotus). Studies
in Genetics VI. University of Texas Publications,7103,
49–90.
1438 D.J. Berg et al.
2007 The Authors, Journal compilation 2007 Blackwell Publishing Ltd, Freshwater Biology,52, 1427–1439
Sokal R.R. & Sneath P.H.A. (1963) Principles of Numerical
Taxonomy. W.H. Freeman, San Francisco, CA.
Spielman D., Brook B.W. & Frankham R. (2004) Most
species are not driven to extinction before genetic
factors impact them. Proceedings of the National Academy
of Sciences of the United States of America,101, 15261–
15264.
Stiven A.E. & Alderman J. (1992) Genetic similarities
among certain freshwater mussel populations of the
Lampsilis genus in North Carolina. Malacologia,34, 355–
369.
Strayer D.L., Downing J.A., Haag W.R., King T.L., Layzer
J.B., Newton T.J. & Nichols S.J. (2004) Changing
perspectives on pearly mussels, North America’s most
imperiled animals. Bioscience,54, 429–439.
Vaughn C.C. & Taylor C.M. (2000) Macroecology of a
host-parasite relationship. Ecography,23, 11–20.
Vokoun J.C. & Rabeni C.E. (2005) Variation in annual
movement cycle of flathead catfish within and between
two Missouri watersheds. North American Journal of
Fisheries Management,25, 563–572.
Watters G.T. (1992) Unionids, fishes, and the species-area
curve. Journal of Biogeography,19, 481–490.
Watters G.T. (1996) 1996 Freshwater Mussel Survey of Big
Darby Creek. Report to The Ohio Chapter of The Nature
Conservancy. Columbus, OH.
Weir B.S. & Cockerham C.C. (1984) Estimating F-
statistics for the analysis of population structure.
Evolution,38, 1358–1370.
Williams J.D., Warren, M.L., Jr, Cummings K.S., Harris
J.L. & Neves R.J. (1993) Conservation status of fresh-
water mussels of the United States and Canada.
Fisheries,18, 6–22.
(Manuscript accepted 13 February 2007)
Freshwater mussel population genetics 1439
2007 The Authors, Journal compilation 2007 Blackwell Publishing Ltd, Freshwater Biology,52, 1427–1439