Diversity of freshwater mussel

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.


The diversity of freshwater mussels (family Unionidae) in the United States is unmatched in the rest of the world, yet this biodiversity is highly vulnerable, with more than 70% of recognized species considered to be either endangered, threatened, or of special concern (Williams et al. 1993) Most species inhabit limited ranges and have small (and/or often unknown) dispersal abilities due to reliance on specific host fish for reproduction, making them easily fragmented and particularly susceptible to the activities of humans, such as habitat modification, pollution, and over-harvesting (Neves 1993, Williams et al. 1993) Recent introductions of alien mollusks such as the zebra mussel (Dreissena polymorpha) pose an additional, large threat to this already impoverished mussel fauna.

Despite the realization that native mussel fauna are threatened, knowledge of basic biology, ecology, and taxonomy of most unionids remains lacking and limits our ability to take immediate conservation action Further complicating our knowledge of unionid species is the lack of discrete morphological characters of use in accurately diagnosing species or determining evolutionary lineages Phenotypic plasticity has often been observed in both mussel shells (conchology) and in soft-tissue anatomy (Kat 1983), and as a result, our ability to delimit species is compromised and cryptic species are often concealed At higher taxonomic levels, convergence on similar sets of shell characteristics by distantly related lineages may confound our ability to place species into higher classification schemes that accurately reflect ancestry.

Effective conservation and restoration plans require clearly definable units of management. Currently many unionid species are arbitrarily defined and managed as metapopulations, in which the principal processes controlling dynamics are extinction, migration, and colonization to establish new local populations (Hanski and Gilpin 1997) The key questions are how these processes jointly affect the dynamics and evolution of local populations and the entire metapopulation Environmental patchiness forces species to be structured into systems of local populations within which conspecifics are more likely to interact with each other than with conspecifics from other populations Isolation, however, is usually not complete and since most organisms have some power of dispersal, members of a local population have a low but positive probability of interaction with individuals from other localities (Wiens 1997) Depending on the rate of migration, demographic and genetic dynamics will be influenced by this migration as well as by local birth and death rates None of these parameters are known for most unionids.

Despite a lack of knowledge of the processes at work within and between local populations for most unionid species, restoration and supplementation efforts in unionid conservation increasingly rely on relocating mussels to other sites or into hatchery facilities for artificial propagation An example of ongoing supplementation efforts is the state of the art hatchery facility at Virginia Tech, which has been releasing juvenile mussels into the Upper Tennessee River system since 1997 (Neves et al. 2003) Through this program, the Clinch River in Southwestern Virginia has received hatchery-reared juveniles as part of a recovery attempt for 18 endangered mussel species inhabiting the area (Zimmerman and Neves, 2003) Such supplementation efforts have been described as taking advantage of, “advances in technology and proactive recovery' (McGregor and Shepard 2003), yet without including a genetic component in such work, the ‘technology' available to managers is not being fully utilized, and as a result, supplementation efforts could potentially backfire due to risks that are inherent with hatchery supplementation programs One potential risk of artificial propagation is the creation of hybrids either between populations of a species that are actually genetically distinct, or between closely related cryptic species Without knowledge of connectedness (levels of gene flow) among natural populations, the possibility exists for species that are genetically distinct yet not fully reproductively isolated to interbreed, resulting in an artificial hybrid and loss of both wild species This scenario was likely prevented through a genetic study of Potamilis inflatus, which detected large species-level genetic differences between the Black Water and Amite River populations that were about to be mixed (Roe and Lydeard 1998) The threat of hybridization between other closely related taxa that co-inhabit an area being supplemented exists as well Again, understanding the phylogenetic relationships between closely related species, and divergence between populations of one species, will be our most useful tool in assessing the risk involved in conservation efforts (Perry et al. 2002)

Another potential risk involved in artificial propagation is a loss of genetic diversity relative to natural populations Genetic diversity is of fundamental concern to conservation efforts as it is required for evolutionary change to occur within populations (Frankel and Soulé 1981) In other words, genetic diversity allows populations to evolve in response to environmental change, such as new diseases, parasites, competitors, nutrient levels, pollutants, etc Genetic diversity (including polymorphism and heterozygosity) has been shown empirically to be lost at greater rates in small relative to large populations (e.g. Montgomery et al. 2000) Likewise, the importance of maintaining large effective population sizes (Ne, or numbers of effective breeding individuals) in hatcheries in order to avoid inbreeding depression (and loss of genetic variation) has been well documented (Ryman and Laikre 1991; Waples and Doh 1994; Tringali and Bert 1998) A counterintuitive yet possible consequence of supplementation is a lowering of Ne in wild populations even as numbers of animals increase This can happen when hatchery rearing artificially increases juvenile survival in selected families relative to family size in the wild and can significantly increase the variance in family size, leading to inbreeding (Ryman and Laikre 1991) For most unionid taxa, knowledge of Ne in wild populations is unknown, yet is likely low for many taxa given the rarity of animals Estimates of Ne in both wild populations and in hatchery stocks are imperative because they allow for impact predictions to be made about effects stocking and relocation programs will have on genetic variability (see Villella et al. 1998).

The clubshell (Pleurobema clava) is a freshwater bivalve that was historically widespread in the Ohio River basin and tributaries of western Lake Erie, but has experienced a 95% range reduction Despite it's listing as an federally endangered species in 1993, little is known about the ecology of P. clava, or about the connectivity between the remaining populations inhabiting widely scattered streams of declining suitability across the eastern United States P. clava is presently known to occur in rivers in six states (Table 1), but reaches maximum abundances in the Allegheny River in Pennsylvania Currently, no genetic information exists on population structure, levels of gene flow, or relatedness among geographic populations of P. clava.

Recently, several researchers have found a second morphologically distinguishable form of P. clava in the Allegheny River near Hunter's Station and in French Creek, a tributary of the Allegheny The newly identified shell type superficially resembles a non-congener that also occurs in this river, Fusconaia subrotunda (the longsolid) Interestingly, these two shell types may have been present in this region for many decades, as two shell shapes of P. clava were also noted by Ortmann (1919) Whether these distinct shell shapes represent ecophenotypic variations of P. clava or different species is unknown, yet is increasingly important as plans for artificial propagation and supplementation of P. clava in the Allegheny drainage continue to be formulated.

On a broader geographic scale, there is question about whether P. clava populations that occur in Michigan, Ohio, Indiana, and West Virginia (Table 1) are actually the same species as that found in Pennsylvania Due to the discontinuity between these geographically disjunct populations, little is known about the degree of genetic connectivity between them Given the numerous findings of significant genetic divergence among unionid populations occupying isolated drainages (discussed above), substantial sub-structuring may be present between these locations

Also of interest is the relationship between P. clava and a similar form, Pleurobema oviforme (Tennessee clubshell) that occurs in locations where P. clava has also been found in the Cumberland River in Kentucky and Tennessee Whether or not these species are distinct genetically also remains unknown Unlike P. clava, P. oviforme is not protected under the Endangered Species Act, and since P. oviforme also appears to be declining, determining the relationship between these taxa is important for conservation efforts A similar situation, with two similar forms of Pleurobema species, P. decisum (federally endangered) and P. chattanoogense (species of concern), occurs in the Mobile basin These species are currently being examined in a phylogenetic study being led by David Campbell at the University of Alabama (pers. comm.) Results from the study proposed here will be complimentary to the phylogenetic analysis of the Pleurobemini of the Mobile basin of Campbell et al., and will create a complete framework of relationships among these closely related and declining species.

Many species belonging to the tribe Pleurobemini (Bivalvia, Unionidae, Ambleminae) exhibit high amounts of variation in shell morphology, which has lead to taxonomic confusion In fact, species belonging to the genus Pleurobema have been described the following way,“the members of the genus Pleurobema are among the most difficult to identify in North America” NatureServe Explorer (ver. 1.6, 2003) Difficulty in identification of Pleurobema species is especially unfortunate in that most of the species are threatened or endangered at either the state or national level (see Table 2 for a list of conservation status of several Pleurobema species included in the proposed study)

The pink mucket, Lampsilis abrupta, is in decline, having been severely impacted by dam construction, dredging, pollution, competition with nonindigenous species (e.g., Dreissena polymorpha), and excessive harvesting L. abrupta occurs over a wide geographic area that includes the Mississippi, Tennessee, and Cumberland River systems L. abrupta has been under the protective auspices of the Endangered Species Act (ESA) since 1976 There is an approved recovery plan (1985) and USFWS has created a watershed implementation schedule for the recovery plan (1989) To plan and implement biologically sound recovery programs in the Ohio River basin, knowledge of the amount of genetic diversity present and a thorough understanding of the evolutionary relationships (e.g., levels of gene exchange) among geographic populations of L. abrupta are essential The intended use of cultured unionids as a conservation tool underscores the need to recognize the genetic composition of natural and managed populations However, no information exists on the phylogeographic relatedness among wild populations or the levels of genetic variation within the captive broodstock(s)


There are two major objectives of this comprehensive project The first objective is to utilize both mitochondrial and nuclear DNA gene regions to determine whether phylogeographic structure exists among extant populations Along with determining relatedness among populations, this work will help clarify taxonomic uncertainties involving both species

The second objective is to utilize newly developed microsatellite DNA markers to assess levels of genetic diversity and heterozygosity within selected geographic populations with the largest concentration of remaining individuals Microsatellite markers will also allow a fine-scale analysis of gene flow between main-stem and tributaries that contain populations of both species, which will allow for identification of potential sources for recruitment Furthermore, the utility of the markers will be demonstrated for use in identifying and maintaining genetic diversity during the development of captive broodstock, for choosing natural populations to supplement, and for monitoring effects on wild populations as supplementation proceeds This proposed research represents the first investigation into the population genetic structure of and will provide the baseline genetic data necessary in order to effectively manage remaining populations


Unionid Samples

Molecular methods

Fundamental to efforts to protect mussel populations is the ability to correctly delineate between species and distinct management units Molecular tools applied to systematic, phylogeographic and population genetic studies are needed to determine the scale (stream, drainage, biogeographic region) at which conservation efforts should be directed for unionid taxa of concern In fact, the use of systematic knowledge of unionids for conservation purposes was recently described this way, “Mussels may represent the best aquatic example of how an increased understanding of phylogenetic relationships will be integral in preventing the loss of biodiversity” (Perry et al. 2002) The limited number of molecular systematic studies that have been done to date on unionid taxa have shown that morphology-based identification and classification may not reflect distinct evolutionary lineages At the generic level and above, several studies have shown that species belonging to a particular genus often do not cluster together on a phylogenetic tree (Potamilus, Roe and Lydeard 1998; Lasmigona, King et al. 1999; Fusconaia, Quincuncina, Obovaria, Lydeard et al. 2000; Alasmidonta, Morrison et al. 2003) Conchological features of unionids, which are what classification schemes have been built around, may not reflect the ‘historical legacy' of species (Lydeard et al. 2000).

At the species level, zoogeographic (i.e. phylogeographic) patterns have often been uncovered using molecular techniques when species are examined throughout their range For example, strong geographical structuring and/or identification of cryptic species have been detected in isolated drainages (e.g. Quadrula quadrula, Berg et al. 1996; Amblema plicata/elliottii, Mulvey et al. 1997; P. inflatus, Roe and Lydeard 1998; Lasmigona subviridis, King et al. 1999; several Lampsilis species, Roe et al. 2001) The Mulvey et al. study has been criticized, however, for inadequate sample sizes, and it has been pointed out that in order to discriminate among species within a genus (especially when suggesting taxonomic changes involving common and threatened species), one must quanitify the amount of variation both among populations within a species as well as between species (Berg and Berg 2000) With an adequate sampling design along with recent advancements in statistical analyses of mitochondrial DNA sequence data such as analysis of molecular variance (AMOVA, Excoffier et al. 1992) and nested clade analysis (Templeton et al. 1995), exploration into the forces shaping intraspecific genetic diversity seen today is possible, such as historical isolation versus current ecological processes For example, allopatric fragmentation was detected using both AMOVA and nested clade analysis in Lampsilis hydiana populations from the Arkansas River drainage relative to other Eastern Highlands populations (Turner et al. 2000) The absence of phylogeographic patterning has also been found in some cases, where some species do not appear to be distinct based on analysis of DNA sequence data (e.g. Epioblasma florentina walkeri and E. capsaeformis, Buhay et al. 2002; three Megalonaias species, Mulvey et al. 1997) Again, caution must be taken in interpreting these types of studies based on small sample sizes (e.g. Berg and Berg 2000), as from a conservation standpoint, it is better to err on the side of recognizing distinct evolutionary units rather than failing to recognize them (Daughterty et al. 1990; Lydeard et al. 2000).

DNA extraction

Genomic DNA will be isolated from preserved mantle tissue using the Puregene DNA extraction kit (Gentra Systems, Inc., Minneapolis, MN), and will be resuspended in TE buffer (10 mM Tris-HCl, pH 8.0, 1 mM EDTA) DNA concentrations will be determined by fluorescence assay (Labarca and Paigen 1980) and integrity of the DNA will be visualized on 1% agarose gels (Sambrook et al. 1989).

Mitochondrial DNA COI and 16S gene regions

The majority of phylogenetic and phylogeographic studies of unionids to date have used either the mitochondrial 16S rRNA gene (Mulvey et al. 1997, plus allozymes); Turner et al. 2000; Krebs et al. 2003) or the cytochrome C oxidase I (COI) gene (Roe and Lydeard 1998; King et al. 1999, plus ITS; Graf and O'Foighil 2000; Buhay et al. 2002, plus ND1) or a combined analysis of both gene regions (Lydeard et al. 2000; Roe et al. 2001) Most importantly to this study, an ongoing phylogenetic study of the tribe Pleurobemini is well underway (D. Campbell, pers. comm.) using the 16S, COI, and ND1 gene regions The study by Campbell et al. concentrates on species found in the Mobile basin, and therefore the proposed work (concentrating on northeastern species) will compliment that study, and will soon benefit from the sequence data that will be available for many species in the Pleurobemini as that project concludes

A 710-bp fragment of the mtDNA COI gene will be amplified from genomic DNA using the polymerase chain reaction (PCR), with the COIL 1490 and COIH 2198 primers developed by Folmer et al. 1994, and PCR conditions as described in King et al. 1999 Additional internal COI primers have also been published (Mulvey et al. 1998), and may be utilized as necessary

A limited amount of COI data exists from previously published work (available in Genbank, accession numbers shown under species names, Fig. 1) for several Pleurobema (P. clava, P. decisum, and P. sintoxia), plus several Fusconaia species (F. flava, F. escambia, and F. ebena) The multiple alignment of these sequences, one P. clava sequence generated from the King lab (unpublished), was 680 bp in length, and included 203 variable and 171 phylogenetically informative sites Results of a parsimony analysis of the COI data are shown in Figure 1 This preliminary analysis suggests that P. decisum is more closely related to P. clava than is P. sintoxia We can also see that the two P. clava sequences are fairly distinct from each other by the long branches that lead to each individual relative to the branches connecting the two individuals of both P. sintoxia and F. flava In fact, the sequence divergence between the two P. clava sequences from the Allegheny (Hunter's Station) and Michigan (Hillsdale County, 1.75%) is about ten times the genetic distance between the two P. sintoxia sequences (0.15%) Divergence between species ranged from 5% between P. clava and P. decisum to 8% between P. clava and Fusconaia flava, and upwards of 20% between Pleurobemini taxa (ingroup) and Amblema or Quadrula (tribe Amblemini, data not shown) One of the three sequences for Quadrula quadrula is approximately 10% divergent from the other two sequences for this taxon, a genetic distance seen between genera in other comparisons in this data set, which could be indicative of a cryptic species It is noteworthy that when multiple individuals have been sampled for a single taxon, these individuals cluster together and receive high bootstrap support The results of this preliminary COI analysis suggest several things: first, it is likely that sub-structuring exists between geographically disjunct P. clava populations (Pennsylvania and Michigan); second, that this mitochondrial gene region is appropriate for shedding light on population sub-structuring and phylogenetic relationships among closely related taxa; and third, that sampling of individuals should include as many populations as possible in order to adequately assess the genetic variation throughout the range of a single taxon.

PCR primers used to amplify the 16S gene region will be the 16Sar-L-myt and 16Sbr-L-myt described in Lydeard et al. 1996, which amplify approximately 500 bp at the 3' end of the mtDNA gene (see Mulvey et al. 1998 for schematic) Two internal PCR primers are also available for sequencing (16Sint2-H and 16Sint1-H; Lydeard et al. 1996) To date, there are only two 16S sequences in Genbank for Pleurobema (P. decisum and P. pyriforme) In addition, we will survey the cytochrome b and/or ND1 regions to search for polymorphisms and phylogeographic signal.

Nuclear ITS gene region

It has become standard practice to compare phylogenetic hypotheses from both nuclear and mitochondrial genes when interested in historical relationships among taxa (see Nichols 2001for a review) By examining the best hypothesis of relationships based on one gene (or several genes from the same genome, such as the mitochondrial genome that is inherited as a linked unit), the patterns observed could represent the gene's lineage, or a “gene tree”, and may or may not match the true historical relationships, or the ‘species tree” (Nichols 2001) When phylogenetic hypotheses based upon genes from different genomes agree, a more robust hypothesis results Despite acceptance of the utilizing both mitochondrial and nuclear DNA among systematists, very few systematic studies of unionids have been done One reason for this is a lack of examples of nuclear gene regions that have been amplified in unionids that demonstrate appropriate rates of change to be informative at addressing species-level phylogenetic hypotheses The only published study we are aware of deals with phylogeny of the genus Lasmigona and phylogeography of L. subviridis (King et al. 1999). This work utilizes the internal transcribed spacer (ITS) region of the tandemly repeated ribosomal RNA gene cluster that separates the 5.8S and 18S structural ribosomal genes The ITS is under less functional constraint than the structural genes neighboring it, resulting in a faster rate of evolution, making this region appropriate for species-level comparison Similar phylogenetic hypotheses for the genus Lasmigona were produced using the mitochondrial COI gene and ITS (King et al. 1999), demonstrating that this nuclear gene region will likely be useful for discerning relationships among members of the Pleurobemini as well Furthermore, ITS variation was seen between populations of L. subviridis that corresponded to differences seen in the mitochondrial COI gene, demonstrating the utility of ITS for intraspecific phylogeographic studies.

A 640-bp fragment of the nuclear ITS-1 gene will be amplified from genomic DNA using the polymerase chain reaction (PCR), with conserved primers based in the flanking 18S and 5.8S structural RNA genes (see King et al. 1999, for PCR primers and conditions) Additional ITS primers are listed in Mulvey et al. 1998 In addition to the internal transcribed spacer regions (ITS-1 and ITS-2), S7-1 and S7-2 ribosomal regions will be surveyed and tested for phylogeographic signal.

DNA sequencing

PCR products from both nuclear and mtDNA genes will be purified with Exonuclease I and Shrimp Alkaline Phosphatase (Promega Corp.) and then used as templates in sequencing reactions with the ABI Prism BigDye Terminator Cycle Sequencing Ready Reaction Kit (Applied Biosystems) Sequencing reactions will be run using a PTC-200 Thermal Cycler (MJ Research, Watertown, MA, USA) and electrophoresed on an ABI Prism 3100ä Genetic Analyzer in Tim King's lab at the Leetown Science Center Forward and reverse sequencing reactions will be assembled forming a consensus sequence for each individual and gene using Sequencher 4.0 (GeneCodes Corporation, Ann Arbor, MI) Multiple alignments will be performed for each gene region using Clustal X ver. 1.4b (Thompson et al. 1994).

Data analysis

Phylogeographic analyses

The extent of DNA polymorphism within and divergence between populations of P. clava and L. abrupta for each gene region sequenced will be estimated using DnaSP version 3 (Rozas and Rozas 1999) The average number of nucleotide differences per site (p, Nei 1987, Eq. 10.5 or 10.6) will be calculated for individuals within populations for each gene region separately The number of net nucleotide substitutions per site with sampling variances (Da, Nei 1987, Eq. 10.21-10.24) plus Jukes and Cantor correction (1969) between populations of P. clava and L. abrupta will be estimated for each gene region using DnaSP.

Gene geneologies will be estimated for populations of P. clava and L. abrupta based upon mitochondrial COI and 16Sdata separately using the TCS computer program (Clement et al. 2000), which follows the parsimony-based approach of Templeton, Crandall and Sing (1992, ‘TCS' method, after authors) Nesting of haplotypes within subnetworks will be carried out following the nesting rules given in Templeton et al., (1987) and Templeton and Sing (1993).

The genetic structure among populations was examined by pairwise FST statistics (Excoffier et al. 1992) using Arlequin ver. 2.000 (Schneider et al. 2000) The significance of pairwise FST values was calculated by permuting haplotypes between populations 1,000 times.

Phylogenetic analyses

For each molecular data partition (species and gene), maximum parsimony (MP), distance-based neighbor-joining, and maximum likelihood (ML) analyses will be run using PAUP* 4.0b10 (Swofford 2002) For parsimony analyses, heuristic searches will be run using unweighted, parsimony-informative (PI) characters with the following settings: starting trees for branch swapping obtained via stepwise addition, 100 random additions of sequences per run, and tree bisection-reconnection (TBR) branch swapping on best trees For 16S and ITS data partitions, gaps will either be coded as missing data or as a 5th base in parsimony-based analyses, and results from these settings will be compared An estimate of support for each node on parsimony trees will be assessed using bootstrap resampling (Felsenstein 1985) with 1,000 replicates and the full heuristic search algorithm Additionally, the Bremer support, or decay index (Bremer 1988, 1994) will be calculated for nodes occurring in strict consensus trees using TreeRot (Sorenson 1999).

We will determine the most appropriate model of DNA substitution for each species and each DNA region independently using hierarchical likelihood ratio tests (Posada and Crandall 2001) in Modeltest 3.06 (Posada and Crandall 1998) ML analyses were run in PAUP* with the best-fit model from Modeltest and heuristic searches with settings as in parsimony except the ‘as-is' option for the addition of sequences. ML bootstrap analyses will be run with the “fast” stepwise addition algorithm and 500 replicates.

Population structure and kinship

Observed genotype frequencies will be tested for consistency with Hardy-Weinberg and linkage equilibrium expectations using randomization tests available in genepop 3.1 (Raymond and Roussett 1995) Pairwise genetic distances between sampled populations will be described by Da of Nei et al. 1983 Statistical significance of genetic differences will be tested using the genetic differentiation randomization test available in genepop Da distances will be visualized in a neighbor-joining tree using TREEVIEW (Page 1996) Isolation by distance will be examined by comparing values between samples and their geographic river distance using a Mantel randomization test available in Arlequin Assignment tests will be used to determine the likelihood of an individual's genotype being found at the sample site from which it was collected Paetkau 1995; Cornuet et al. 1999) The computer program GENECLASS (ver. 1.0.02) will be used for assignment tests (http://www.montpellier.inra.fr/URLB/geneclass/geneclass.html).