Molecular Ecology Resources (2008) 8, 711?724 doi: 10.1111/j.1755-0998.2008.02108.x ? 2008 The Authors Journal compilation ? 2008 Blackwell Publishing Ltd Blackwell Publishing LtdDNA BARCODING Identification of ?extinct? freshwater mussel species using DNA barcoding DAVID C. CAMPBELL,* PAUL D. JOHNSON,? JAMES D. WILLIAMS,? ANDREW K. RINDSBERG,? JEANNE M. SERB,? KORY K. SMALL* and CHARLES LYDEARD** *Biodiversity and Systematics, Department of Biological Sciences, University of Alabama, 425 Scientific Collections Building, Box 870345, Tuscaloosa, AL 35487-0345, USA, ?Alabama Aquatic Biodiversity Center, Route 3, Box 86, Marion, AL 36756, USA, ?Florida Museum of Natural History, University of Florida, Gainesville, FL 32611, USA, ?Department of Biological & Environmental Sciences, Station 7, University of West Alabama, Livingston, AL 35470, USA, ?Department of Ecology, Evolution and Organismal Biology, Iowa State University, 253 Bessey Hall, Ames, IA 50011-1020, USA, **Smithsonian Tropical Research Institute, Smithsonian Institution, 1100 Jefferson Drive, Suite 3123, Washington, DC 20013, USA Abstract Freshwater mollusks are highly imperiled, with 70% of the North American species extinct, endangered, or at risk of extinction. Impoundments and other human impacts on the Coosa River of Alabama, Georgia and Tennessee of the southeastern USA alone are believed to have caused 50 mollusk species extinctions, but uncertainty over boundaries among several putatively closely related species makes this number preliminary. Our examination of freshwater mussels collected during an extensive survey of the upper-drainage basin, DNA barcoding and molecular phylogenetic analyses confirm the rediscovery of four morpho- species in the genus Pleurobema (Unionidae) previously thought to be extinct from the upper Coosa basin. A fifth ?extinct? form was found in an adjoining basin. Molecular data show that the Coosa morphologies represent at least three species-level taxa: Pleurobema decisum, P. hanleyianum and P. stabile. Endemism is higher than currently recognized, both at the species level and for multispecies clades. Prompt conservation efforts may preserve some of these taxa and their ecosystem. Keywords: cox1, endangered species, molecular barcode, Pleurobema, Unionidae Received 4 November 2007; revision accepted 12 December 2007 Introduction Nonmarine mollusks such as unionid mussels have dis- proportionately high rates of extinction and imperilment, but receive little conservation management compared with charismatic vertebrate species (Lydeard et al. 2004). Unionids have attracted many researchers because of their ecological significance, economic importance (chiefly in the cultured pearl industry), local abundance, complex life cycle includ- ing an obligate parasitic larva, and recent drastic decline (Strayer et al. 2004). Like other freshwater organisms such as fishes (Walsh et al. 1995), snails (Bogan et al. 1995) and crayfishes (Crandall & Templeton 1999), unionid mussels show exceptional diversity and endemism in the south- eastern USA (Williams & Neves 1995), where their varied forms have inspired colourful common names such as warrior pigtoe, painted clubshell, inflated heelsplitter, pistolgrip and spike. The upper Coosa basin extends through Tennessee, Georgia and Alabama in the USA, has an area of approxi- mately 6400 km2, and drains four physiographic provinces. The watershed has a complex and ancient geological history, dating back at least to the Cretaceous if not to the Palaeozoic, with stream capture and sea level changes producing varying connection and isolation relative to nearby drainages (Adams 1929; Conant 1964; Rindsberg 2003). Although currently part of the Mobile River system, sea level variation in the geological past has isolated the Coosa from other major rivers in the system (Fig. 1). Historically, the upper Coosa was home to over 40 species of freshwater mussels, making it one of the most biologically Correspondence: David Campbell, Fax: 205-348-6460; E-mail: amblema@bama.ua.edu 712 D N A B A R C O D I N G ? 2008 The Authors Journal compilation ? 2008 Blackwell Publishing Ltd diverse rivers in the world (Garner et al. 2004). The decline of the mussel fauna began in the early 1900s in response to the building of locks and dams and later to point-source pollution from textile and carpet-dying operations, nonpoint pollution from urban and suburban sprawl, and siltation from poor land-use practices (Mirarchi et al. 2004). The destruction and alteration of stream and floodplain habitat and reduction of water quality resulted in a catastrophic decline of freshwater mollusks. Most species suffered drastic range reductions, with about 50 mollusk species entirely eliminated. This includes all species of three snail genera (Gyrotoma, Amphigyra and Neoplanorbis) and severe reduction of two other snail genera that were thought to be extinct, but were later rediscovered (Tulotoma and Clappia) (Hershler et al. 1990; Garner et al. 2004; S. A. Clark personal commu- nication). Among unionid bivalves, the subgenus Alasmidens is presumed extinct (Clarke 1981). Pleurobema lost nine species in the Coosa. Pleurobema is one of the most speciose genera of freshwater mussel but also one of the most imperiled. The most recent comprehensive summary tentatively recognized 32 species (the majority from the Mobile River basin), of which 13 (40%) were thought to be extinct (Turgeon et al. 1998) and therefore have no official protected status. Twelve purportedly extant Pleurobema species are federally listed as Endangered in the USA (Turgeon et al. 1998). The only bivalve genus with more recently extinct species is Epioblasma, most species of which occurred in the Tennessee, Cumberland and Ohio systems. Complicating efforts to assess conservation status in freshwater mussels is the difficulty in identifying speci- mens to species for taxonomically challenging taxa (Roe 2000) including Pleurobema, whose shells often differ only by subtle characteristics. Furthermore, most species and genera are currently defined using historical morphological concepts. Although many have been included in molecular phylogenetic studies, only a few genera have been thor- oughly re-investigated phylogenetically. Recent molecular analysis of the 45 currently recognized North American genera in Ambleminae revealed that most polytypic genera are polyphyletic (Campbell et al. 2005), highlighting the problems in current classification. In particular, this means that a species-level study cannot assume that close relatives are currently assigned to the same genus. Among unionids, identifying and delimiting species within Pleurobema based on shell morphology is especially problematic (Goodrich 1913; Simpson 1914; Burch 1975; Turgeon et al. 1998). Shell shape in unionids reflects many environmental parameters, potentially over decades of growth. Pleurobema generally lacks significant shell sculpture or other distinguishing features (the exception, Pleurobema collina, usually has spines, but the present data indicate it is not a true Pleurobema), so species are currently identified by subjective assessment of shell shape. Soft-part anatomy is poorly documented, and anatomical differences between closely related species, when known, are often subtle, requiring detailed examination. Fig. 1 Biogeographical patterns in Pleuro- bema in the eastern USA. Major river systems are numbered. Mississippi, Ohio, and upper Apalachicola largely coincide with state boundaries. Tennessee (5) -Cumberland (4) -Ohio (2) -Mississippi (1) [P. clava, P. cordatum, P. gibberum (Cumberland only), P. oviforme (Tennessee and Cumberland only), P. rubrum, P. sintoxia]; Pearl (6) and Pasca- goula (7) (P. beadlianum); Tombigbee (8) - Alabama (11) (P. perovatum, P. taitianum; P. decisum); Black Warrior (9) (P. furvum, P. rubellum); Escambia (12), Yellow (13), and Choctawhatchee (14) (P. strodeanum); Coosa (10) (P. chattanoogaense, P. decisum, P. hanley- ianum, P. stabile, P. troschelianum); Apala- chicola (15), Ochlockonee (16), and Suwanee (17) (P. pyriforme); James (3) (P. collina). Dashed line indicates approximate Oligo- cene highstand shoreline, showing past isolation of major rivers (new data). D N A B A R C O D I N G 713 ? 2008 The Authors Journal compilation ? 2008 Blackwell Publishing Ltd In such situations, molecular techniques such as DNA barcoding have great potential to supplement traditional taxonomic methods. Many recent studies have successfully applied these techniques to other animals (Hajibabaei et al. 2006; Kelly et al. 2007; Kerr et al. 2007). The cox1 gene has been widely used in studies on freshwater mussels over the past decade (Hoeh et al. 1997; Roe & Hoeh 2003; Araujo et al. 2005; Campbell et al. 2005; Gustafson & Iwamoto 2005; K?llersj? et al. 2005; Soroka 2005; Graf & Cummings 2006; Walker et al. 2006; Zanatta & Murphy 2006, references therein), so a good comparative data set is available. However, many species remain undocumented, limiting the potential for barcode-type approaches to identification of unknowns. Jones et al. (2006) had diffi- culty distinguishing some mussel species using other mitochondrial genes and ITS1, but microsatellites showed clear differences. Nevertheless, most studies on unionids have found mitochondrial genes to be very useful. Poten- tial pitfalls for barcoding have also been documented for other taxa and for theoretical models of speciation (Hickerson et al. 2006; Meier et al. 2006). These highlight the importance of investigating additional molecular, mor- phological and other data in addition to the barcode sequence. Although doubly uniparental inheritance of mito- chondrial DNA produces some problems for other bivalves, in unionids the male mitotype is strictly associated with the male germ line, so that sampling of somatic tissue yields only female mitotypes. Also, there is no evidence of exchange between the male and female mitotypes within Unioniformes, and the male mitotypes are so divergent from the female as to be readily recognizable (Walker et al. 2006). We sought to determine the level of molecular differ- entiation between morphological forms in Pleurobema species from the upper Coosa system. In turn, we used these data to identify molecular markers suitable for identification of problematic specimens and to place the species into a phylogenetic framework. Addition- ally, phylogenetic analyses that incorporate the actual sequence data provide a more sensitive test of patterns of molecular differentiation than simply comparing percentage differences. For molecular data, species differentiation was based on monophyly (i.e. a phylo- genetic species concept) and the per cent difference (i.e. a phenetic criterion widely used for barcoding studies). Large sample sizes are desirable to test the level of intra- population variability; however, in some cases our sample of one specimen was the entire population. Both the extreme rarity of most species and endangered species regulations limited the number of modern samples available. Within the Mobile basin, a few healthy populations are known only for P. decisum and P. perovatum, both of which are listed as Endangered. Materials and methods In 1998, an intensive programme was initiated to survey the upper Coosa River basin with an emphasis on the historically richest sub-basin, that of the Conasauga River. To date, over 700 sites in the upper Coosa River system have been surveyed for mussels and other invertebrates. Annual surveys of the Conasauga River began in 1998, but heavy rain prevented the 2002 survey. Similar surveys have examined other areas in the Mobile basin, emphasizing the few relatively undisturbed portions of larger rivers. Depen- ding on water depth, surveying required wading, snorkeling and/or scuba diving to search for mussels in the river bed. In 1998?1999, 616 mussel specimens were found in the upper Coosa basin representing 24 species, in 2002?2003, 345 mussels were found representing 18 species, and in 2005? 2006, 565 mussels of 20 species were found, for a total of 28 species, including two impoundment-tolerant species not recorded historically from the upper Coosa. Current species taxonomy is based on shell shape, colour pattern and geographical distribution. In particular, Pleu- robema species differ in degree of elongation and whether they are more oval, quadrate, or triangular. Comparison of our often eroded specimens to museum material helped verify their identity, especially when large suites, illustrating intrapopulation variation, were available. Colour pattern is somewhat variable and often obscured in older speci- mens, in addition to the influence of erosion and encrus- tation, but may be helpful if it is visible. For example, Pleurobema chattanoogaense typically has a few green spots on the early part of the shell (visible near the dorsal margin in Fig. 2) giving it the common name of painted clubshell, whereas Pleurobema stabile is all brown. To determine the taxonomic identity of unknown Pleu- robema specimens, we used molecular phylogenetic methods to construct topologies of relatedness between morpholo- gically identified species and unknown specimens. DNA was extracted from fresh, frozen, or ethanol-preserved specimens using standard cetyltrimethyl ammonium bromide (CTAB) and chloroform?isoamyl alcohol protocols (Winnepenninckx et al. 1993). Voucher specimens for all new sequences are in the University of Alabama collections except for Pleurobema pyriforme, in the North Carolina Museum of Natural History. All but two Pleurobema species with known extant populations were sequenced. We could not obtain sequences for Pleurobema plenum, an endangered species from the Tennessee and Ohio river systems closely related to Pleurobema cordatum, P. rubrum and P. sintoxia, nor specimens for P. riddelli from west Louisiana and east Texas. Taxa representing other genera of the tribe Pleurobemini, including all extant species from the Coosa, and other tribes of the subfamily Ambleminae served as outgroups. We selected two mitochondrial genes that had worked well in previous studies on unionids, cytochrome oxidase 714 D N A B A R C O D I N G ? 2008 The Authors Journal compilation ? 2008 Blackwell Publishing Ltd I (cox1) and NADH dehydrogenase subunit 1 (nadh1), and one nuclear region used in a few prior studies, the ribos- omal internal transcribed spacer I (ITS1). Primers for cox1 were 5?-GTTCCACAAATCATAAGGATATTGG-3? and 5?-TACACCTCAGGGTGACCAAAAAACCA-3?, adapted from Folmer et al. (1994), primers for nadh1 were 5?- TGGCAGAAAAGTGCATCAGATTTAAGC-3? and 5?- GCTATTAGTAGGTCGTATCG-3? (Buhay et al. 2002; Serb & Lydeard 2003), and primers for ITS1 were 5?-AAAAA- GCTTCCGTAGGTGAACCTGCG-3? and 5?-AGCTTGCT- GCGTTCTTCATCG-3? (King et al. 1999). Polymerase chain reaction (PCR) cycles were: 92 ?C 2 min; 92 ?C 40 s 40 ?C 40 s 72 ?C 90 s 5?; 92 ?C 40 s 50 ?C 40 s 72 ?C 90 s 25?; 72 ?C 10 min; hold 4 ?C. PCR products were purified using QIAquick PCR purification kits. Cycle sequencing used ABI BigDye Terminator kits with thermal cycle parameters of 1 ?C per second ramp speed, starting with 1 min at 96 ?C followed by 26 cycles of 96 ?C for 10 s, 49 ?C for 5 s, and 60 ?C for 4 min, then 10 min at 60 ?C and hold at 4 ?C. The cycle sequencing products were purified with sephadex columns or QIAGEN DyeEx kits and then run on an auto- mated sequencer (ABI 3100). The results for each strand were compared and aligned with published sequences using bioedit (Hall 1999). No indels were found in the protein-coding genes, but ITS1 has several. New cox1 sequences have been identified as barcode data in Gen- Bank. Although ITS1 can show significant variation within individuals, all included specimens yielded sequences that were readily readable without cloning. This indicates that only one copy of the gene was amplifying, as found in some other studies on unionids (Grobler et al. 2005; Jones et al. 2006). Several other unionids have also yielded either a single sequence or else two alleles differing by a single base in a repeat region, whereas almost all gastropods we Fig. 2 Phylogram showing Bayesian analy- sis on cox1 and nadh1 sequence data. Results of the parsimony analyses were similar (Fig. 6). Burn-in was 10 000, mean ln likeli- hood was ?8252.920. Numbers are posterior probabilities. The branch uniting Pleurobema cordatum and Pleurobema rubrum and that uniting P. rubrum and P. sintoxia are too short to be visible, despite having 100% and 68% probability, respectively. Asterisks indicate figured specimens. The shells are left valves of several upper Coosa Pleurobema species from University of Alabama collec- tions. From the top: Pleurobema chattanoogaense (historical specimen), P. decisum, P. han- leyianum, P. georgianum, P. stabile (collected 1912), P. stabile (specimen collected 2001). The new P. stabile specimen is 70 mm in maximum dimension. Despite the heavy erosion in the second specimen, the posterio- ventral elongation in both specimens of P. stabile distinguishes them from the other species. D N A B A R C O D I N G 715 ? 2008 The Authors Journal compilation ? 2008 Blackwell Publishing Ltd have tried yielded multiple divergent sequences, unreadable without cloning (personal observation). Phylogenetic analyses of the sequences included heuristic parsimony searches and bootstrap analyses in paup*4.10 (Swofford 1998). Because cox1 and nadh1 yielded similar results, a partition-homogeneity test was run in paup* (PILD of Dowton & Austin 2002) with 1000 replicates of 10 random addition replicates each. The maximum number of trees per replicate was set to 10 000. This test is sensitive to other factors, such as partition size and evolutionary model, besides data compatibility (Dowton & Austin 2002), but may provide a rough idea of agreement between data sets. Despite the problems of the ILD type of tests, no better alternative has gained wide acceptance. The P value was 0.65, so the two mitochondrial genes were concatenated for further analysis. Bayesian analysis, using mrbayes 3.1.1 (Ronquist & Huelsenbeck 2003), served as an alternative phylogenetic method. Bootstrap values typically underes- timate support, whereas Bayesian probabilities tend to overestimate it (Simmons et al. 2004). Maximum parsimony analyses used 1000 random replicates, hold = 10, swap = TBR. Bootstrap analyses used 1000 replicates, each using a random parsimony search of 10 replicates. For the ITS1 data, parsimony and bootstrap analyses treated gaps as a fifth base. Bayesian analysis used 2 000 000 generations and eight chains; revmat, shape, pinvar and statefreq were unlinked. mrmodeltest 2.2 (Nylander 2004) recommended K80+g for ITS1 and GTR+I+G independently selected for both nadh1 and cox1. In the Bayesian analyses, the standard deviation of split frequencies went under 0.01 for both. All bootstrap percentages and Bayesian probabilities over 50% for branches in the maximum parsimony trees are shown. Some analyses had 55% or less bootstrap support for a clade not in the strict consensus; these are not indicated. The sequences used in this study are listed in Table 1. As we accumulated cox1 and nadh1 sequence data, all but one specimen showed close correspondence to sequences from positively identified specimens. That badly eroded specimen (Fig. 2) had been tentatively assigned to P. chat- tanoogaense, but based on the molecular data, it was highly distinct from all other sampled specimens. The anomalous molecular results prompted further morphological study of museum specimens, along with re-examination of the mystery specimen, to identify morphological characters that were not obliterated by the erosion. To determine the geological history of drainage systems, we examined stream drainage patterns, erosional features, sediment outcrop areas and other geomorphological features as well as literature data. In turn, the drainage histories were compared to the biogeographical patterns seen in the phylogenies. Results The intensive searches in the upper Coosa yielded live or freshly dead specimens still suitable for molecular genetic analysis from four supposedly extinct morphospecies of Pleurobema: painted clubshell, P. chattanoogaense (Lea 1858); Georgia pigtoe, P. hanleyianum (Lea 1852); Alabama clubshell, P. troschelianum (Lea 1852) and one badly eroded individual that, after molecular analyses and detailed analysis of museum specimens, was identified as Pleurobema stabile (Lea 1861) [often listed under the junior synonym Pleurobema murrayense (Lea 1868)] (Fig. 2). Another very eroded specimen resembles Pleurobema hartmanianum (Lea 1860), but it has not yet yielded DNA sequences. Other unusual specimens were assignable to recognized species based on DNA sequence data (Fig. 3). Molecular data also confirmed that all of the Pleurobema perovatum-like specimens found in the upper Coosa were in fact P. hanleyianum. Fig. 3 Aberrant specimens of Pleurobema decisum from the Tallapoosa system (UAUC3299, top) and from the Coosa system (UAUC471, bottom), identified based on nadh1 sequence. Contrast with the normal specimen in Fig. 2. 716 D N A B A R C O D I N G ? 2008 The Authors Journal compilation ? 2008 Blackwell Publishing Ltd Some of these species have not been reported as alive for decades (Evans 2001). Pleurobema stabile was last collected and reliably identified in 1958 by H. Athearn (Museum of Fluviatile Mollusks collections, P.D.J., personal obser- vation). Additional sampling in other parts of the Mobile basin yielded the warrior pigtoe, Pleurobema rubellum, in the upper Black Warrior River system. This species has historically been reported from the upper Coosa and cur- rently is listed as Extinct (Turgeon et al. 1998). However, no specimens of the western Mobile basin species Pleurobema curtum or Pleurobema marshalli were found, despite their current listing as Endangered rather than Extinct. Live specimens of these have not been found since the 1980s. Both mitochondrial genes yielded similar results, showing several well-supported clades within Pleurobema (Figs 2 and 6). Intraspecies variation is low. The results from ITS1 are generally less well-resolved and less well-supported than from the other two genes (Figs 4 and 5), with more intraspecies variation, but it provides evidence for the dis- tinctiveness of some species. Pleurobema does not appear to be monophyletic. When cox1 or nadh1 were analysed sepa- rately, including sequences from specimens that amplified for only one gene, all sequences for a species placed in the same clade, and those clades had at least 89% bootstrap support (not shown). Table 2 shows the per cent difference between various taxa for each gene region. Discussion The tribe Pleurobemini includes approximately 90 species in the genera Elliptio, Fusconaia, Hemistena, Lexingtonia, Plethobasus, Pleurobema and Quincuncina (Turgeon et al. 1998; Campbell et al. 2005). Molecular analyses indicate that the current generic classification of freshwater mussels Fig. 4 Strict consensus of 861 maximum parsimony trees, length 492, ITS1 data. Numbers indicate bootstrap percentage if over 50%. D N A B A R C O D I N G 717 ? 2008 The Authors Journal compilation ? 2008 Blackwell Publishing Ltd requires extensive revision, and some species are even assigned to the wrong tribe (Roe & Lydeard 1998; King et al. 1999; Buhay et al. 2002; Serb et al. 2003; Campbell et al. 2005). Most extant species currently assigned to Pleurobema, including its type species, Pleurobema clava (Lamarck 1819), comprise a clade, thus largely but not entirely supporting current taxonomy. For the mitochondrial genes, both phenetic distance and phylogenetic placement generally did well at sorting out morphologically distinct species. Kandl et al. (2001) like- wise was able to separate problematic Pleurobema species in the Gulf Coast drainages east and south of the Mobile basin, using a short segment of the cox1 gene. Their sequences place in the same clades as our sequences for the same species (personal observation), but because they are much shorter they were not included in the present analyses. Within Pleurobema, several smaller clades are largely congruent with major rivers (Figs 1 and 2), reflecting their long independent histories over geological time. Except for Pleurobema decisum, most species are confined to a single basin or group of associated river systems. This contrasts with existing groupings based on shell morphology, which range across drainages. For example, Pleurobema taitianum from the Tombigbee and Alabama systems resembles the Pleurobema sintoxia group from the Tennessee, Ohio, and Mississippi systems in its relatively triangular, heavy shell, and P. decisum and Pleurobema chattanoogaense from the Coosa resemble the Pleurobema clava-P. oviforme group from the Tennessee and Ohio systems in their elongate shape. In turn, the species within a river are typically more closely related to each other than to species from other river systems. Although most rivers have a single clade, and the large Tennessee?Ohio?Mississippi system has two (three if one counts the Cumberland species Pleurobema gibberum that belongs in a different genus based on present results), three clades of species occur in the small but ecologically rich Fig. 5 Phylogram showing Bayesian analysis on ITS1 sequence data. Burn-in was 30800, mean ln likelihood was ?1978.885. Numbers are posterior probabilities. The branch uniting Pleurobema athearni and P. georgianum is too short to be visible, despite having 68% probability. Interiors of left valves of specimens from Fig. 2 shown. From the top: Pleurobema chattanoogaense, P. decisum, P. hanleyianum, P. georgianum, P. stabile (museum specimen, collected 1912), P. stabile (specimen collected 2001). The new Pleurobema stabile specimen is 70 mm in maximum dimension. The different position of the posterior (left) adductor muscle scar in P. stabile vs. P. chattanoogaense shows a different body configuration in the shell despite having similarly ovate outlines. 718 D N A B A R C O D I N G ? 2008 The Authors Journal compilation ? 2008 Blackwell Publishing Ltd Table 1 Taxa and GenBank accession numbers. Type species of genera are indicated by T. New sequences generated in the present study are indicated by *. Most mitochondrial sequences (the unstarred ones starting with AY) were also generated in this study but previously published in Campbell et al. (2005) Species nadh1 cox1 ITS1 Cyrtonaias tampicoensis (Lea 1838) T AY655090 AF231749 DQ383436* Elliptio arca (Conrad 1834) AY655093 AY654995 DQ383437* Elliptio arctata (Conrad 1834) no data DQ383427* DQ383438* Elliptio crassidens (Lamarck 1819) T AY613788 DQ383428* DQ383439* Elliptio dilatata (Rafinesque 1820) DQ385872* AF231751 DQ383440* Fusconaia cerina (Conrad 1838) AY613792 AY613823 DQ383441* Fusconaia flava (Rafinesque 1820) T AY613793 AF156510 DQ383442* Hemistena lata (Rafinesque 1820) T AY613796 AY613825 DQ383443* Plectomerus dombeyanus (Valenciennes 1827) T AY655110 AY655011 DQ383444* Plethobasus cyphyus (Rafinesque 1820) T AY613799 AY613828 DQ383445* Pleurobema athearni (Gangloff et al. 2006) AY655114 AY655015 DQ383446* Pleurobema beadlianum (Lea 1861) DQ385873* DQ383429* DQ383447* Pleurobema beadlianum AY613800* no data no data Pleurobema chattanoogaense (Lea 1858) AY613801 AY613829 no data Pleurobema chattanoogaense AY655111 AY655012 DQ383448* Pleurobema chattanoogaense no data DQ383430* no data Pleurobema clava (Lamarck 1819) T AY613802 AY655013 DQ383449* Pleurobema clava T no data AF231754 no data Pleurobema collina (Conrad 1837) AY613803 AY613830 DQ383450* Pleurobema cordatum (Rafinesque 1820) AY613804 AY613831 DQ383451* Pleurobema decisum (Lea 1831) Coosa1 AY613805 AY613832 DQ383452* Pleurobema decisum Coosa2 DQ383467* no data no data Pleurobema decisum Sipsey1 no data AF232801 no data Pleurobema decisum Sipsey2 no data DQ383431* DQ383453* Pleurobema decisum Tallapoosa1 AY655112 AY655014 no data Pleurobema decisum Tallapoosa2 DQ383466* no data DQ383454* Pleurobema furvum (Conrad 1834) AY613806 AY613833 DQ383455* Pleurobema georgianum (Lea 1841) AY613807 AY613834 DQ383456* Pleurobema georgianum AY655113 no data DQ383457* Pleurobema gibberum (Lea 1838) DQ385874* AY613835 DQ383458* Pleurobema gibberum AY613808 no data no data Pleurobema hanleyianum (Lea 1852) AY655115 AY655016 DQ470003* Pleurobema hanleyianum AY613809 AY613836 DQ383459* Pleurobema oviforme (Conrad 1834) AY613810 AY655017 DQ470004* Pleurobema oviforme AY655116 AY613837 DQ383460* Pleurobema perovatum (Conrad 1834) AY613811 AY613838 no data Pleurobema perovatum no data DQ383433* no data Pleurobema pyriforme (Lea 1857) AY613812 AY613839 no data Pleurobema pyriforme DQ383468* no data DQ383461* Pleurobema rubellum (Conrad 1834) AY613813 AY613840 DQ383462* Pleurobema rubrum (Rafinesque 1820) AY655117 AY655018 no data Pleurobema rubrum AY613814 AY613841 DQ470005* Pleurobema sintoxia (Rafinesque 1820) AY613815 AY655019 DQ470006* Pleurobema sintoxia no data AF156508 no data Pleurobema stabile (Lea 1861) AY613816* AY613842* DQ383463* Pleurobema strodeanum (Wright 1898) AY613817 AY613843 no data Pleurobema strodeanum no data DQ383434* no data Pleurobema taitianum (Lea 1834) AY613818 AY613844 no data Pleurobema troschelianum (Lea 1852) AY613819 AY613845 DQ383464* Uniomerus declivus (Say 1831) no data AY613846 DQ383435* D N A B A R C O D I N G 719 ? 2008 The Authors Journal compilation ? 2008 Blackwell Publishing Ltd upper Coosa system. Pleurobema decisum, P. chattanoogaense, P. hanleyianum and P. troschelianum form one clade of species, all confined to the upper Coosa system except for P. decisum, which ranges throughout the Mobile basin. The molecular evidence thus strongly suggests that P. decisum originated in the upper Coosa basin and recently spread to other rivers. This idea is also supported by the genetic diversity found in P. decisum samples from the Coosa, vs. minimal variation within the populations from other parts of the Mobile basin. A second clade includes only Pleurobema stabile, although extinct species such as P. fibuloides (Lea 1859) might belong here based on shell features. The third clade present in the upper Coosa includes Pleurobema georgianum and the newly described Pleurobema athearni (Gangloff et al. 2006) from the middle Coosa. Both the phenetic distances and the phylogenetic results indicate that the upper Coosa forms are distinct from the species endemic to the western Mobile basin. These results contradict earlier morphological studies that suggested that Coosa species might be synonyms of taxa described from other river basins. In particular, P. hanleyianum has previously been confused with Pleurobema perovatum (Par- malee & Bogan 1998), and P. stabile has been synonymized with Pleurobema rubellum (Frierson 1927). In contrast, our molecular results indicate that P. perovatum and P. rubellum are part of a large clade centred on the western Mobile basin, not closely related to Coosa natives. Some species pairs show minimal molecular divergence (Table 2) and may represent ecophenotypes of a single species or extremely close relatives, including P. chattanoog- aense and P. decisum, P. hanleyianum and P. troschelianum, P. furvum and P. rubellum, P. georgianum and P. athearni, P. perovatum and P. taitianum, and P. clava and P. oviforme. However, ITS1 showed greater differences between P. hanleyianum and P. troschelianum and between P. clava and P. oviforme than did the mitochondrial genes. ITS1 also showed high intraspecific variation within P. hanleyianum and P. oviforme, so the significance of the differences for these species is unclear. The species pair P. georgianum and P. athearni and the pair P. perovatum and P. taitianum are morphologically quite distinct with mitochondrial per cent differences that are lower than average for interspecies comparisons and higher than average for intraspecies com- parisons. They thus seem to represent recently diverged but separate taxa. The remaining species pairs are morpho- logically more similar and have individuals of each type with nearly identical genotypes, suggesting that they may be synonyms. A complete nomenclatural revision is in preparation (J.D. Williams, personal communication). Many of these pairs also failed to resolve as reciprocally mono- phyletic. The Pleurobema cordatum group (P. cordatum, P. plenum, P. rubrum, P. sintoxia) also is poorly resolved, but more intensive sampling throughout the Mississippi- Ohio-Tennessee river system is needed to understand this clade. Because it is both geographically and phylogenetically separate from the upper Coosa forms, we did not pursue them in detail. The molecular data and phylogenetic anal- yses thus suggest that current, morphological classification has slightly oversplit Pleurobema. But even if each of these species pairs were combined for conservation purposes, all the species would remain highly imperiled. Moreover, current nomenclature and literature fail to capture the diversity of supraspecific clades. Present data indicate higher levels of endemism than previously recognized, with species and higher clades generally each confined to a single river system. High endemism within drainages is also Table 2 Percentage differences between taxa. Pleurobema s.s. excludes P. collina, P. cordatum group (P. cordatum, P. rubrum, P. sintoxia), P. gibberum and P. stabile. Between species comparison excludes the possibly conspecific close pairs, separately enumerated (P. chattanoogaense? P. decisum, P. clava-P. oviforme, P. furvum?P. rubellum, P. georgianum?P. athearni, P. hanleyianum?P. troschelianum). Numbers given are mean and range of raw percentages (gaps treated as missing data for ITS1). If only a single sequence was available for each species in a comparison, only a single value is given Comparison cox1 nadh1 ITS1 Between genera of Pleurobemini 8.97 (5.39?12.28) 10.18 (7.38?13.85) 1.86 (0.20?4.36) Between species of Pleurobema s.s. 5.63 (1.16?9.08) 5.98 (2.65?9.03) 0.90 (0.19?1.96) Within species of Pleurobema s.s. 1.18 (0.00?2.74) 0.82 (0.13?2.07) 0.42 (0.00?1.17) Elliptio dilatata?other Elliptio 8.12 (7.37?8.51) 11.89 (11.50?12.27) 1.13 (1.01?1.20) ?Pleurobema? collina?other Pleurobema 9.27 (8.08?11.27) 11.37 (9.65?13.34) 2.90 (2.39?3.63) ?Pleurobema? cordatum group?other Pleurobema 6.90 (3.83?10.04) 8.29 (6.85?11.50) 0.96 (0.39?1.40) ?Pleurobema? gibberum?other Pleurobema 9.30 (7.44?11.30) 9.70 (8.70?11.11) 1.40 (0.80?1.97) ?Pleurobema? stabile?other Pleurobema 7.07 (5.52?9.30) 10.27 (9.04?12.15) 1.14 (0.79?1.58) Pleurobema chattanoogaense?P. decisum 1.57 (0.49?2.38) 1.07 (0.26?1.81) 0.00 (0.00?0.00) Pleurobema clava?P. oviforme 1.48 (0.67?2.22) 1.00 (0.84?1.16) 0.59 (0.20?0.98) Pleurobema furvum?P. rubellum 0.50 0.39 0.19 Pleurobema georgianum?P. athearni 1.29 1.04 (0.90?1.18) 0.48 (0.19?0.77) Pleurobema hanleyianum?P. troschelianum 0.64 (0.53?0.75) 0.32 (0.26?0.39) 0.58 (0.20?0.97) 720 D N A B A R C O D I N G ? 2008 The Authors Journal compilation ? 2008 Blackwell Publishing Ltd increasingly recognized in the fish of the region (Boschung & Mayden 2004). The present results reflect both the potential and the pitfalls of molecular barcoding approaches (Smith 2005). Molecular data provided key evidence that one unknown specimen was a different species from all other sampled species, leading to its recognition as P. stabile, and other problematic specimens were readily assigned to species based on DNA sequence. However, identification of the species required careful morphological studies to supply reliable reference DNA sequences. The lack of molecular data for many of the rarest species (especially other, probably extinct forms) make morphological examination of voucher material essential to identification of unusual specimens (De Ley et al. 2005). Nadh1 yielded results almost identical to those for cox1 for barcoding purposes and only a few differences in the phylogeny when analysed separately. Given the apparent fixed differences between male and female mitotypes in unionids (Hoeh et al. 2002; Curole & Kocher 2005), all mito- chondrial genes are expected to show similar evolutionary patterns. As nadh1 has a slightly higher level of interspecies variability and slightly lower intraspecies variability than cox1, it might be a better barcode choice than cox1 in unionids. However, Jones et al. (2006) showed that mitochondrial sequence data alone (as in Buhay et al. 2002, a previous study on the same taxa) does not capture some of the species diversity in unionids. Grobler et al. (2007) found apparent mitochondrial introgression or ancestral poly- morphism, indicating further risks for reliance solely on a Fig. 6 Strict consensus of four trees, length 1473, nadh1 and cox1 data. Numbers indicate bootstrap percentage if over 50%. D N A B A R C O D I N G 721 ? 2008 The Authors Journal compilation ? 2008 Blackwell Publishing Ltd mitochondrial barcode. ITS1 showed a lower percentage variation and poor resolution, with some discrepancies with the mitochondrial data. ITS1 often has multiple alleles within a single individual (Campbell et al. 2004), leading to the potential of reticulate evolution through recombination between multiple ancestral alleles, lineage sorting, or other confounding effects. The lack of variation in ITS1 sequence within the P. decisum-P. chattanoogaense clade suggests that the region has undergone concerted evolution occasionally within the Unionidae, but the lack of clear pattern in most other sampled species suggests that such events have been infrequent within Pleurobemini. Unfortunately, better nuclear genetic markers have not yet been identified for most mollusks. The degree of variation shown in each sequence region varied from taxon to taxon. This supports cautions about the reliability of genetic barcoding that uses a single DNA region and a universal percentage cut-off for species recognition (Hickerson et al. 2006; Meier et al. 2006; Nielsen & Matz 2006). Nevertheless, the present barcode data set provides a powerful identification tool for the often problematic species in Pleurobema. The high level of endemism has important implications for conservation. Although preservation of the few remain- ing localities with good faunas is crucial, focus on only the most diverse faunas will fail to protect many species. Restoration and protection of habitat in each of the river systems is necessary to protect all the taxa. In particular, incorrectly treating P. stabile and P. hanleyianum as synonyms of species found in other drainages could allow the Coosa species to go extinct under the mistaken belief that they were protected elsewhere. Also, the evolutionary lineages identified in this study may influence conservation deci- sions. Given the unfortunate reality of limited conservation funding, some authors propose prioritizing phylogeneti- cally distinctive taxa (Vane-Wright et al. 1991; Faith 1992). The genetic distinctiveness of P. stabile would make its conservation a high priority. The rediscovery of these species provides a new oppor- tunity for their conservation. Because they were presumed to be extinct, none of these species currently have any legal protection. Reviews of their status and proposals to add them to the Endangered species list are in preparation. The taxonomic confusion that existed before our analyses hindered assessment of conservation needs. The present data provide better justification of alpha taxonomy and molecular tools to help in identification. Also, preservation of genetic diversity within a species provides greater evo- lutionary resiliency and avoids inbreeding problems. The concentration of genetic diversity in P. decisum (including P. chattanoogaense) in the remnant upper Coosa population suggests that this region is exceptionally important to the total diversity of the species. The discovery of living individuals of several Pleurobema species raises some hope of preserving them from extinc- tion if prompt efforts are made to protect their environ- ment. Habitat restoration in the upper Coosa system, such as establishment of riparian buffer zones or restoration of a more natural flow regime below dams, would provide natural or restocked populations with better opportunities to survive and recover. Without such changes, the future of these species will be tentative at best. Regulation of point- source pollution has already ameliorated water quality. The Conasauga River, a Coosa tributary, was known a few decades ago as the ?Rainbow River? because its colour constantly varied because of factories discharging waste dyes. Such dramatic insults are gone, but the subtler effects of nonpoint pollution, excessive siltation, and unchecked suburban sprawl could easily eliminate the few survivors. For P. stabile, searches of recent collections yielded only one other specimen, also badly eroded externally, collected as a freshly dead shell. No more specimens have been found since 2001, live or dead. The situation for P. hanleyianum (whether or not Pleurobema troschelianum is treated as a synonym) is not much better, last found freshly dead in 2003. Mussels may live for decades, so a slowly dwindling, nonreproducing population may exist long after it is no longer self-perpetuating (Strayer et al. 2004). Also, the limited legal protection of the river systems and high (and increasing) anthropogenic impact lead to continu- ing habitat degradation. Although the Conasauga River has the highest remaining concentration of severely im- periled species in the upper Coosa system (eight Federally listed species, one candidate for Federal listing, and several species either endemic or extirpated from all other localities), it has received little conservation attention. Much of the historic range in the Coosa River system is now unsuitable habitat due to impoundment, unnatural flow regimes caused by inadequately regulated hydroelec- tric dam releases, siltation from poor land use and other detrimental modification (Burkhead et al. 1997; Mirarchi et al. 2004; Gangloff & Feminella 2007; Poff et al. 2007). In drought years, water demand from growing urban centres, especially Atlanta, poses a new threat. Almost the entire upper Coosa system lies within 150 km of Atlanta, putting many species at high risk of disturbance through- out their range. Competition by the introduced Asian clam, Corbicula leana (Prime 1864), may also affect the Coosa bivalve fauna. Most current threats could be reduced by proactive planning and better watershed practices. However, freshwater mollusks seem highly vulnerable to the effects of global warming (Mouthon & Daufresne 2006), making international as well as regional action important. Similar threats face freshwater systems worldwide. As a result, nonmarine mollusks rank globally among the most imperiled organisms (Lydeard et al. 2004). The rediscovery of multiple species on the brink of extinction highlights the urgent need for protection and study of freshwater faunas, 722 D N A B A R C O D I N G ? 2008 The Authors Journal compilation ? 2008 Blackwell Publishing Ltd especially in areas of high endemism such as southeastern North America. Acknowledgements A grant from the US Fish and Wildlife Service to C. Lydeard sup- ported this work. The ABI 3100 automated sequencer was funded by an NSF equipment grant to C. Lydeard, R. Mayden, M. Powell, and P. Harris (DBI-0070351). A Howard Hughes Medical Institute Undergraduate Biological Sciences Education Program grant to the University of Alabama supported K. K. Small as a Hughes Undergraduate Research Intern, as well as providing some funding for supplies. In addition to material collected by the authors, S. Ahlstedt, S. Bakalety, J. E. Buhay, P. Burgess, R. Butler, S. A. Clark, A. M. Commens, R. R. Evans, S. Fraley, M. Gangloff, J. T. Garner, W. R. Haag, P. Hartfield, M. Hughes, H. McCullagh, M. A. McGregor, J. G. McWhirter, C. R. Merrill, K. J. Roe, S. Shively, D. Thurmond, R. Towes, A. Wethington, and A. Wyss collected specimens used in this study. This manuscript was developed to some extent while C. Lydeard served as a Program Officer at the National Science Foundation under the Intergovernmental Personnel Agreement Act and was supported in part by the IR/D program. References Adams GI (1929) The streams of the Coastal Plain of Alabama and the Lafayette problem. Journal of Geology, 37, 193?203. Araujo R, Gomez I, Machordom A (2005) The identity and biology of Unio mancus Lamarck, 1819 (= U. elongatulus) (Bivalvia: Unionidae) in the Iberian Peninsula. Journal of Molluscan Studies, 71, 25?31. Bogan AE, Pierson JM, Hartfield P (1995) Decline in the freshwater gastropod fauna in the Mobile Bay Basin. In: Our Living Resources: A Report to the Nation on the Distribution, Abundance, and Health of U.S. Plants, Animals, and Ecosystems (eds LaRoe ET, Farris GS, Puckett CE, Doran PD, Mac MJ), pp. 249?252. US Department of the Interior, National Biological Service, Washington, DC. Boschung HT, Mayden RL (2004) Fishes of Alabama. Smithsonian Books, Washington, DC. Buhay JE, Serb JM, Dean CR, Parham Q, Lydeard C (2002) Con- servation genetics of two endangered unionid bivalve species, Epioblasma florentina walkeri and E. capsaeformis (Unionidae: Lampsilini). Journal of Molluscan Studies, 68, 385?391. Burch JB (1975) Freshwater Unionacean Clams (Mollusca: Pelecypoda) of North America, revised edn. Malacological Publications, Hamburg, Michigan. Burkhead NM, Walsh SJ, Freeman BJ, Williams JD (1997) Status and restoration of the Etowah River, an imperiled southern Appalachian ecosystem. In: Aquatic Fauna in Peril: the Southeastern Perspective. Special Publication no. 1, Southeast Aquatic Research Institute (eds Benz GW, Collins DE), pp. 375?444. Lenz Design & Communications, Decatur, Georgia. Campbell DC, Serb JM, Buhay JE, Roe KJ, Minton RL, Lydeard C (2005) Phylogeny of North American amblemines (Bivalvia, Unionoida): prodigious polyphyly proves pervasive across genera. Invertebrate Biology, 125, 131?164. Campbell MR, Steiner G, Campbell LD, Dreyer H (2004) Recent Chamidae (Bivalvia) from the Western Atlantic Ocean. Malacologia, 46, 381?416. Clarke AH (1981) The tribe Alasmidontini (Unionida: Anodontinae), part I: Pegias, Alasmidonta, and Arcidens. Smithsonian Contributions to Zoology, 326, I?iii+1?101. Conant LC (1964) General Remarks on the Pre-Selma Cretaceous Strata of Western Alabama. US Geological Survey, Bulletin 1160-F, Government Printing Office, Washington, DC. Crandall KA, Templeton AR (1999) The zoogeography and centers of origin of the crayfish subgenus Procericambarus (Decapoda: Cambaridae). Evolution, 53, 123?134. Curole JP, Kocher TD (2005) Evolution of a unique mitotype-specific protein-coding extension of the cytochrome c oxidase ii gene in freshwater mussels (Bivalvia: Unionoida). Journal of Molecular Evolution, 61, 381?389. De Ley P, Tandingan-De Ley I, Morris K et al. (2005) An integrated approach to fast and informative morphological vouchering of nematodes for applications in molecular barcoding. Philosophical Transactions of the Royal Society of London. Series B: Biological Sciences, 360, 1945?1958. Dowton M, Austin AD (2002) Increased congruence does not necessarily indicate increased phylogenetic accuracy ? the behavior of the incongruence length difference test in mixed- model analyses. Systematic Biology, 51, 19?31. Evans RR (2001) Historical and contemporary distributions of aquatic mollusks in the Upper Conasauga River System of Georgia and Tennessee. MS Thesis, University of Tennessee at Chat- tanooga, Chattanooga, Tennessee. Faith DP (1992) Conservation evaluation and phylogenetic diversity. Biological Conservation, 61, 1?10. Folmer O, Hoeh WR, Black MB, Vrijenhoek RL (1994) DNA primers for amplification of mitochondrial cytochrome c oxidase subunit I from diverse metazoan invertebrates. Molecular Marine Biology and Biotechnology, 3, 294?299. Frierson LS (1927) A Classification and Annotated Check List of the North American Naiades. Baylor University Press, Waco, Texas. Gangloff MM, Feminella JW (2007) Stream channel geomorphology influences mussel abundance in southern Appalachian streams, USA. Freshwater Biology, 52, 64?74. Gangloff MM, Williams JD, Feminella JW (2006) A new species of freshwater mussel (Bivalvia: Unionidae), Pleurobema athearni, from the Coosa River drainage of Alabama, USA. Zootaxa, 1118, 43?56. Garner JT, Blalock-Herod HN, Bogan AE et al. (2004) Freshwater mussels and snails. In: Alabama Wildlife. Volume 1. A Checklist of Vertebrates and Selected Invertebrates: Aquatic Mollusks, Fishes, Amphibians, Reptiles, Birds, and Mammals (ed. Mirarchi RE), pp. 13?58. University of Alabama Press, Tuscaloosa, Alabama. Goodrich C (1913) Spring collecting in southwest Virginia. The Nautilus, 27, 81?82. Graf DL, Cummings KS (2006) Palaeoheterodont diversity (Mollusca: Trigonioida + Unionoida): what we know and what we wish we knew about freshwater mussel evolution. Zoological Journal of the Linnaean Society, 148, 343?394. Grobler PJ, Jones JW, Johnson NA et al. (2005) Patterns of genetic differentiation and conservation of the slabside pearlymussel, Lexingtonia dolabelloides (Lea, 1840) in the Tennessee River drainage. Journal of Molluscan Studies, 72, 65?75. Grobler PJ, Jones JW, Johnson NA, Neves RJ, Hallerman EM (2007) Population genetic structure of the endangered fanshell pearly- mussel (Cyprogenia stegaria) in the Ohio and Tennessee River drainages. Freshwater Mollusk Conservation Society 2007 Program and Abstracts, 97. D N A B A R C O D I N G 723 ? 2008 The Authors Journal compilation ? 2008 Blackwell Publishing Ltd Gustafson RG, Iwamoto EM (2005) A DNA-based identification key to Pacific Northwest freshwater mussel glochidia: importance to salmonid and mussel conservation. Northwest Science, 79, 233?245. Hajibabaei M, Janzen DH, Burns JM, Hallwachs W, Hebert PDN (2006) DNA barcodes distinguish species of tropical Lepidoptera. Proceedings of the National Academy of Sciences, USA, 103, 968? 971. Hall TA (1999) bioedit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/ NT. Nucleic Acids Symposium Series, 41, 95?98. Hershler R, Pierson JM, Krotzer RS (1990) Rediscovery of Tulotoma magnifica (Conrad) (Gastropoda: Viviparidae). Proceedings of the Biological Society of Washington, 103, 815?824. Hickerson MJ, Meyer CP, Moritz C (2006) DNA barcoding will often fail to discover new animal species over broad parameter space. Systematic Biology, 55, 729?739. Hoeh WR, Stewart DT, Guttman SI (2002) High fidelity of mito- chondrial genome transmission under the doubly uniparental mode of inheritance in freshwater mussels (Bivalvia: Unionoi- dea). Evolution, 56, 2252?2261. Hoeh WR, Stewart DT, Saavedra C, Sutherland BW, Zouros E (1997) Phylogenetic evidence for role-reversals of gender-associated mitochondrial DNA in Mytilus (Bivalvia: Mytilidae). Molecular Biology and Evolution, 14, 959?967. Jones JW, Neves RJ, Ahlstedt SA, Hallerman EM (2006) A holistic approach to taxonomic evaluation of two closely related endan- gered freshwater mussel species, the oyster mussel Epioblasma capsaeformis and tan riffleshell Epioblasma florentina walkeri (Bivalvia: Unionidae). Journal of Molluscan Studies, 72, 267?283. K?llersj? M, von Proschwitz T, Lundberg S, Elden?s P, Ers?us C (2005) Evaluation of ITS rDNA as a complement to mitochon- drial gene sequences for phylogenetic studies in freshwater mussels: an example using Unionidae from north-western Europe. Zoologica Scripta, 34, 415?424. Kandl KL, Liu HP, Butler RS, Hoeh WR, Mulvey M (2001) A genetic approach to resolving taxonomic ambiguity among Pleurobema (Bivalvia: Unionidae) of the eastern Gulf Coast. Malacologia, 43, 87?101. Kelly RP, Sarkar IN, Eernisse DJ, Desalle R (2007) DNA barcoding using chitons (genus Mopalia). Molecular Ecology Notes, 7, 177? 183. Kerr KCR, Stoeckle MY, Dove CJ, Weigt LA, Francis CM, Hebert PDN (2007) Comprehensive DNA barcode coverage of North American birds. Molecular Ecology Notes, 7, 535?543. King TL, Eackles MS, Gjetvaj B, Hoeh WR (1999) Intraspecific phylogeography of Lasmigona subviridis (Bivalvia: Unionidae): conservation implications of range discontinuity. Molecular Ecology, 8 Supl., 1, S65?S78. Lydeard C, Cowie RH, Ponder WF et al. (2004) The global decline of nonmarine mollusks. Bioscience, 54, 321?330. Meier R, Shiyang K, Vaidya G, Ng PKL (2006) DNA barcoding and taxonomy in Diptera: a tale of high intraspecific variability and low identification success. Systematic Biology, 55, 715?728. Mirarchi RE, Bailey MA, Garner JT et al. (ed.) (2004) Alabama Wildlife. Volume 4. Conservation and Management Recommendations for Imperiled Wildlife. University of Alabama Press, Tuscaloosa, Alabama. Mouthon J, Daufresne M (2006) Effects of the 2003 heatwave and climatic warming on mollusc communities of the S?one: a large lowland river and of its two main tributaries (France). Global Change Biology, 12, 441?449. Nielsen R, Matz M (2006) Statistical approaches for DNA barcoding. Systematic Biology, 55, 162?169. Nylander JAA (2004) MRMODELTEST v2. Program distributed by the author. Evolutionary Biology Centre, Uppsala University, Uppsala, Sweden. Parmalee PW, Bogan AE (1998) The Freshwater Mussels of Tennessee. University of Tennessee Press, Knoxville, Tennessee. Poff NL, Olden JD, Merritt DM, Pepin DM (2007) Homogenization of regional river dynamics by dams and global biodiversity implications. Proceedings of the National Academy of Sciences, USA, 104, 5732?5737. Rindsberg AK (2003) Shoreline and rivers in Alabama during the Oligocene highstand. Geological Society of America Abstracts with Programs, 35 (1), 70?71. Roe KJ (2000) The utility of DNA sequences to aid in the identifi- cation of rare or problematic species of freshwater mussels. In: Freshwater Mollusk Symposia Proceedings. Part II. Proceedings of the First Freshwater Mollusk Conservation Society Symposium (ed. Tankersley RA, Warmolts DI, Watters GT, Armitage BJ, Johnson PD, Butler RS), pp. 197?202. Ohio Biological Survey Special Publication, Columbus, Ohio. Roe KJ, Hoeh WR (2003) Systematics of freshwater mussels (Bivalvia: Unionoida). In: Molecular Systematics and Phylogeography of Mollusks (eds Lydeard C, Lindberg DR), pp. 91?122. Smithsonian Books, Washington, DC. Roe KJ, Lydeard C (1998) Molecular systematics of the freshwater mussel genus Potamilus (Bivalvia: Unionidae). Malacologia, 39, 195?205. Ronquist F, Huelsenbeck JP (2003) mrbayes 3: Bayesian phylo- genetic inference under mixed models. Bioinformatics, 19, 1572? 1574. Serb JM, Lydeard C (2003) Complete mtDNA Sequence of the North American freshwater mussel, Lampsilis ornata (Unionidae): An examination of the evolution and phylogenetic utility of mitochondrial genome organization in Bivalvia (Mollusca). Molecular Biology and Evolution, 20, 1854?1866. Serb JM, Buhay JE, Lydeard C (2003) Molecular systematics of the North American freshwater bivalve genus Quadrula (Unionidae: Ambleminae) based on mitochondrial ND1 sequences. Molecular Phylogenetics and Evolution, 28, 1?11. Simmons MP, Pickett KM, Miya M (2004) How meaningful are Bayesian support values? Molecular Biology and Evolution, 21, 188?199. Simpson CT (1914) A Descriptive Catalogue of the Naiades, or Pearly Fresh-water Mussels. Parts I?III. Bryant Walker, Detroit, Michigan. Smith VS (2005) DNA Barcoding: perspectives from a ?Partner- ships for Enhancing Expertise in Taxonomy? (PEET) Debate. Systematic Biology, 54, 841?844. Soroka M (2005) Genetic variability among freshwater mussel Anodonta woodiana (Lea, 1834) (Bivalvia: Unionidae) populations recently introduced in Poland. Zoological Science (Tokyo), 22, 1137?1144. Strayer DL, Downing JA, Haag WR et al. (2004) Changing perspec- tives on pearly mussels, North America?s most imperiled animals. Bioscience, 54, 429?439. Swofford DL (1998) PAUP*. Phylogenetic Analysis Using Parsi- mony (*and Other Methods). Sinauer Associates, Sunderland, Massachusetts. Turgeon DD, Quinn JF, Bogan AE et al. (1998) Common and Scientific Names of Aquatic Invertebrates from the United States and Canada: Mollusks. Special Publication 26. American Fisheries Society, Bethesda, Maryland. 724 D N A B A R C O D I N G ? 2008 The Authors Journal compilation ? 2008 Blackwell Publishing Ltd Vane-Wright RI, Humphries CJ, Williams PH (1991) What to protect? ? systematics and the agony of choice. Biological Conservation, 55, 235?254. Walker JM, Curole JP, Wade DE, Chapman EG, Bogan AE, Watters GT, Hoeh WR (2006) Taxonomic distribution and phylogenetic utility of gender-associated mitochondrial genomes in the Unionoida (Bivalvia). Malacologia, 48, 265?282. Walsh SJ, Burkhead NM, Williams JD (1995) Southeastern fresh- water fishes. In: Our Living Resources: A Report to the Nation on the Distribution, Abundance, and Health of U.S. Plants, Animals, and Ecosystems (eds LaRoe ET, Farris GS, Puckett CE, Doran PD, Mac MJ), pp. 144?147. US Department of the Interior, National Biological Service, Washington, DC. Williams JD, Neves RJ (1995) Freshwater mussels: a neglected and declining resource. In: Our Living Resources: A Report to the Nation on the Distribution, Abundance, and Health of U.S. Plants, Animals, and Ecosystems (eds LaRoe ET, Farris GS, Puckett CE, Doran PD, Mac MJ), pp. 177?179. US Department of the Interior, National Biological Service, Washington, DC. Winnepenninckx B, Backeljau T, De Wachter R (1993) Extraction of high molecular weight DNA from molluscs. Trends in Genetics, 9, 407. Zanatta DT, Murphy RW (2006) Evolution of active host- attraction strategies in the freshwater mussel tribe Lampsilini (Bivalvia: Unionidae). Molecular Phylogenetics and Evolution, 41, 195?208.