|Home | About | Journals | Submit | Contact Us | Français|
Freshwater bivalves in the order Unionoida are considered to be one of the most endangered groups of animals in North America. In Texas, where over 60% of unionids are rare or very rare, 15 species have been recently added to the state’s list of threatened species, and 11 are under consideration for federal listing. Due to insufficient survey efforts in the past decades, however, primary data on current distribution and habitat requirement for most of these rare species are lacking, thus challenging their protection and management. Taxonomic identification of endemic species based on shell morphology is challenging and complicates conservation efforts. In this paper we present historic and current distributional data for three rare Texas species, Fusconaia askewi, F. lananensis, and Pleurobema riddellii, collected during our 2003–2011 state-wide surveys and suggest appropriate conservation measures. In addition, we tested the genetic affinities of Fusconaia and similar species collected from eastern Texas and western Louisiana using cox1 and nad1 sequences.
We found that F. askewi still inhabits four river basins in eastern and northeastern Texas and can be locally abundant, while P. riddellii was found only in one river basin. Pleurobema riddellii was well-separated from F. askewi and grouped with the P. sintoxia clade. The sequences for F. lananensis were very similar to those for F. askewi, with a maximum difference of just over 1% for nad1 and only 0.7% for cox1, similar to the variation between F. askewi alleles. Except for one low difference (1.55%) with the partial cox1 sequence for F. burkei, all other Fusconaia populations, including those from the Calcasieu drainage, differed by over 2.3% for both genes.
Our study suggested that F. lananensis is not a valid species, and it is likely that only one Fusconaia species (F. askewi or its probable senior synonym F. chunii) is currently present in East Texas, thus simplifying conservation efforts. Distribution range of both these regional endemics (F. askewi and P. riddellii) has been reduced in the last 80years.
Molluscs are among the most threatened groups of animals on the planet , and freshwater bivalves in the order Unionoida are considered to be one of the most endangered groups of animals in North America [2-4]. Our long-term state-wide study of Texas mussels revealed that 65% of all Texas unionid species are rare, including all state and regional endemics, and most endemic species are very rare . Being one of the top states in species diversity and endemism, Texas ranks fourth in terms of the number of species extinctions . Damming, pollution, water extraction, and urban development have all negatively affected the freshwaters of Texas . Fifteen rare freshwater mussel species were recently added to the state’s list of threatened species , and 11 of those are currently under consideration for federal listing by the U. S. Fish and Wildlife Service [9,10].
Biodiversity is a fundamental component of evolutionary potential, and species are the primary targets of the U.S. Endangered Species Act. Conservation laws and methods cannot be implemented until the endangered organism is properly clarified and its geographical range is known [11,12]. In particular, some of these rare species, Fusconaia flava (Rafinesque), F. askewi (Marsh), and F. lananensis Frierson, are currently reported from several drainages west of the Mississippi [13-15], but identifying specimens using shell morphology is challenging. Morphological variation in Fusconaia in the lower Mississippi drainage is especially complex . Burdick and White  reported an unusual genetic type in Fusconaia from the northern and western Ozark region, which could represent a northern extension of F. askewi. Pleurobema riddellii (Lea) can also be very similar in shell features to F. askewi. Johnson  synonymized F. askewi with F. flava (under the name F. undata).
In light of the difficulties, we used genetic data as an additional line of evidence. We sampled Fusconaia and similar species from river systems in eastern Texas and western Louisiana to test the genetic affinities of the species, using cox1 and nad1 sequences. In this paper we describe the geographical distribution and habitat requirements of rare Fusconaia spp. and P. riddellii and results of molecular genetic analyses to define their biogeography, proper taxonomic status, and suggest appropriate conservation measures.
In this manuscript we use results of our state-wide survey of unionids in Texas, USA (latitudes 33°50′ - 26°56′, longitudes 102°08′ - 93°31′) from 2003 to 2011 [5,19]. Mussels were surveyed at 463 sub-sites that were pooled into141 major sites, distributed among 66 waterbodies belonging to 11 major drainages in Texas. The study was carried out with an appropriate Scientific Research Permit issued by the Texas Parks and Wildlife Department (TPWD), and landowner permission for wildlife research was acquired from each property owner before entering their property, if the land was privately owned. Abiotic parameters (physical and chemical) were recorded at the sites using a HACH Hydrolab Quanta, measured parameters included: temperature (°C), pH, total dissolved solids (g/L), conductivity (μS/cm), and turbidity (ed. NTU). In addition, we recorded depth and the dominant substrate type using the following classification by particle size: bedrock; large boulders (>45cm); boulders (>25 - 45cm); cobble (>6 - 25cm); gravel (>6 - 60mm); sand (0.06 - 6mm); mud/silt (<0.06mm). Substrates in sampled East Texas sites were represented by sand (32%), sand and gravel (21%), silt (15%), clay (6%), and combinations of these. Unionid sampling was conducted via hand collection of both live and dead mussels, by wading in shallow water and by snorkeling. Due to poor water visibility, tactile searches (running fingers over the sediment, usually up to 15cm deep, depending on substrate type) were used at all sites. Timed searches were used to detect the presence of mussels and species diversity [20,21] at each site, and if mussel assemblages were present, quantitative methods (from 5 to 28 randomly placed 0.25m2 quadrats at a site, in average 9 quadrats covering area of 3.75m2), or area -constrained searches (area searched were from 4 to 66m2) were used for assessments of density [22,23]. Relative species abundance was calculated as a percentage of live specimens belong to this species collected at a site from the total number of all live mussels found at the same site, and used as an indicator of the species’ dominance in mussel assemblages. Collected mussels were identified based on shell morphology, counted, measured with calipers to the nearest mm, and then carefully rebedded into the sediment from which they were taken. Ten specimens of Fusconaia sp. from the Neches drainage and 5 from the Sabine drainage were sequenced for cox1. Five Fusconaia specimens from the Neches drainage (including one not amplified for cox1) and 3 from the Sabine drainage were sequenced for nad1. Two specimens of P. riddellii from the Neches drainage were sequenced for cox1, with one of them also sequenced for nad1. Voucher specimens were deposited in the Great Lakes Center (Buffalo State College) Invertebrate Collection, in the North Carolina State Museum of Natural Sciences (Raleigh, NC), and in the Invertebrate Zoology Collection of the National Museum of Natural History (Smithsonian Institution, Washington, D.C.). All Fusconaia species identified during our study (F. askewi and F. lananensis) and historical data reported from East Texas (F. askewii[24,25], F. askewi[15,26-30], F. flava, F. lananensis[31-33], Quadrula askewi[34,35], Q. askewii, Q. chunii[25,35], Q. flava nasuta, Q. lananensis[25,34,36], Q. undata chunii, Unio askewii, U. cerinus[24,37], U. chunii[24,37,38], were considered to be F. askewi. For justification see sections “Genetic analysis” in Results and Discussion.
Specimens were preserved in ethanol in the field. DNA extraction used Qiagen DNA extraction kits. Portions of the cox1 and nad1 genes were amplified. Primers for cox1 were 5'–GTTCCACAAATCATAAGGATATTGG–3' and 5'–TACACCTCAGGGTGACCAAAAAACCA–3', adapted from Folmer et al. and primers for nadh1 were 5'–TGGCAGAAAAGTGCATCAGATTTAAGC–3' and 5'–GCTATTAGTAGGTCGTATCG–3' [40,41]. The primer LoGlyR (5’–CCTGCTTGGAAGGCAAGTGTACT–3′)  served as an alternate reverse primer for nadh1. The forward primer UNIOCOII.2 from Walker et al. and/or the reverse primer HCOout (CCAGGTAAAATTAAAATATAAACTTC ) provided good amplification for cox1 for some species. PCR cycles were: 92°C 2min; 92°C 40sec 40°C 40sec 72°C 90sec 5x; 92°C 40sec 50°C 40sec 72°C 90sec 25x; 72°C 10min; hold 4°C. PCR products were purified using Qiagen QIAquick PCR purification kits and, if necessary, Qiagen gel extraction kits. Cycle sequencing used ABI Big Dye Terminator kits with thermal cycle parameters of 1°C per second ramp speed, starting with 1min at 96°C followed by 26 cycles of 96°C for 10sec, 49°C for 5sec, and 60°C for 4min, then 10min at 60°C and hold at 4°C. The cycle sequencing products were purified with Qiagen DyeEx kits and then run on an automated sequencer.
The results for each strand were compared and aligned using BioEdit . We analyzed the sequences, along with previously published sequences for other representatives of Pleurobemini with TNT . An Additional file 1 contains sequences used for genetic analysis [see Additional file 1. Maximum parsimony analyses used 500 random replicates, using all the “new technology” methods (sectorial searching, ratchet, drift, and tree fusing), which greatly speed up the process of finding optimal trees over older approaches . Jackknife analyses used 500 replicates, each using a random “new technology” parsimony search of 10 replicates.
The sequences for F. lananensis were very similar to those for F. askewi, with less than 1% difference, similar to the variation between F. askewi alleles (Tables112).2). However, the sequences for F. askewi from the Sabine and Neches drainages differed from all other Fusconaia species by over 2.3% for both genes, except for the partial cox1 sequence for F. burkei. In particular, the cox1 sequences differed by no more than 0.7% between F. askewi and F. lananensis, typical of within-species variation, but differed by a minimum of over 2.5% from all other Fusconaia sequences, except the short sequence for F. burkei, fairly normal for species-level differences. The cox1 sequences from putative F. askewi from the Calcasieu River system in Louisiana  differed from sequences for F. flava and F. cerina by less than 2% and in most cases by less than 1% (Table (Table1).1). One published sequence for F. flava (AF231733,  was identical to one of the Calcasieu sequences. Figures Figures11122 and and33 show the phylogenetic analyses. Jackknife percentages close to 100 show strong support for a particular group. As cladograms, their branching sequence provides the important information. Thus, in Figure Figure1Pleurobema1Pleurobema (Sintoxia) riddellii 186TS is modestly supported (51%) as being most closely related to the strongly supported (100%) group including P. (Sintoxia) sintoxiaP. (Sintoxia) cordatum, and P. (Sintoxia) rubrum. Those four in turn are most closely related to the group of the three Pleuronaia species. However, this association of Pleuronaia and P. (Sintoxia) received less than 50% jackknife support and was not supported by all of the analyses. The two Fusconaia lananensis have good support (84%) as being each other’s closest relative, and there is very strong support (100%) for a group including the Sabine and Neches F. askewi as well as F. lananensis. In turn, this F. askewi-lananensis group has fairly good support (78%) as being most closely related to the group including F. masoni, F. cerina, F. flava, the putative F. askewi from the Calcasieu, F. burkei, and F. escambia. The Calcasieu Fusconaia specimens are strongly supported (92%) as being most closely related to F. flava. In Figure Figure2P.2P. riddelli again appears to be most closely related to P. rubrum, P. sintoxia, and P. cordatum 2572, but yet again this result is not well-supported. Multiple branches coming from a single vertical line indicates that the relationship among those branches is unresolved. Figure Figure22 shows strong support (95%) for a group including the Sabine and Neches F. askewi and the F. lananensis specimens, but does not tell anything about relationships among those eight sequences. Relationships among the different groups within Fusconaia are not well-resolved in Figure Figure2.2. Similarly, Figure Figure33 has strong support (99%) for a group of all of the F. lananensis and Sabine and Neches F. askewi, but apart from strong support (99%) for a group of F. askewi Sab1 and Sab2, does not support any particular relationships within that group. Again, P. riddellii receives weak support as being most closely related to P. sintoxia, P. rubrum, and P. cordatum.
A total of 931 live individuals was collected during our surveys (including 774 mussels originally identified as F. askewi and 157 identified as F. lananensis) at 25 sites in 17 East Texas counties (Anderson, Angelina, Cherokee, Hardin, Harrison, Houston, Jasper, Leon, Nacogdoches, Panola, Rusk, San Augustine, Shelby, Smith, Titus, Tyler, and Upshur) (Table (Table3,3, Figure Figure4B).4B). We found F. askewi in four drainages (Neches, Trinity, Sabine, and Red river basins) in eastern and northeastern Texas. Fusconaia askewi was locally very abundant in Village Creek (Neches River basin), Neches, Sabine, Trinity and Angelina (Neches River basin) rivers, and in the Big Cypress Bayou (Red River basin). On average, F. askewi was the third most abundant species, and the number of live F. askewi collected at a particular site, on average, comprised 22% of the total number of all live mussels found at that site. Average density in mussel aggregations was 6.7m-2 (Table (Table3).3). Sites with the greatest abundance were on Village Creek and the Neches and Sabine rivers. The most typical substrate for the species was sand, then a mixture of sand and silt, and gravel with sand. Average shell length of live F. askewi was 59.2±0.6mm (mean±standard error here and elsewhere unless noted). Based on the presence of juveniles (Figure (Figure5),5), the populations in East Texas were reproducing (shell length varied from 17 to 90mm). Nevertheless we failed to find F. askewi in several waterbodies belong to the species’ former distribution range: in the San Jacinto River, its tributaries, and in Lake Houston, as well as in its historical location in Kickapoo Creek (North of Brownsboro, Henderson Co.  (Figure (Figure4).4). Likewise, we did not find the species in any of the 6 reservoirs on the Trinity River and its tributaries. Our surveys also confirmed that F. askewi has been extirpated from Lanana and Bonita creeks (type localities for F. lananensis).
Only one dead shell and one valve of mussels identified as F. flava were found during our surveys, at two sites in the Sulphur River (Red River drainage), in Red River County and in Delta/Hopkins counties. Live individuals resembling F. flava have recently been collected in the East Fork of the Trinity River approximately 70km from Dallas . Mussels from the Sulphur River and the Trinity River have not been genetically tested yet.
During our surveys, we found 132 live P. riddellii at 10 sites in 5 Texas counties (Anderson, Angelina, Cherokee, Hardin, and Nacogdoches), in the Neches, and Angelina rivers, and in Village Creek (Figure (Figure6B,6B, Table Table3).3). Average density of P. riddellii was 1.9m-2, and the species was not dominant in local unionid assemblages (the average relative abundance of P. riddellii was 5%, Table Table3).3). Most often P. riddellii was found in sand, silty sand, and sometimes in a mixture of sand and clay. Mean and median P. riddellii length were 52.4 ±1.1mm, range - 39–82mm (Figure (Figure5).5). The largest density was found in the Neches River south of Neches (Anderson Co.) in sand and gravel; this population had many juveniles (< 25mm long) in 2009 (Barclay unpublished data).
We found that F. askewi and P. riddellii have similar distribution (Table (Table3)3) and very similar habitat requirements. All these species were found exclusively in lotic waters, in relatively shallow areas (at 0.2 - 1.5m depth), and the most preferable substrates for both F. askewi and P. riddellii were sand, and combinations of sand with gravel and silt. Total dissolved solids among waterbodies studied varied from 0.10 to 0.15g/L, turbidity – from 18.9 to 66.9 ed. NTU, pH – from 6.38 to 8.21. The lowest pH was recorded in Village Creek (average of 4 measurements in 2005 and 2007: 6.64±0.24 (standard deviation), minimal 6.38±0.12) and in Sandy Creek (6.69±0.006). Minimal pH value for the studied rivers and creeks recorded from 1973 to 2009 was 4.8 (4.8 for Village Creek, 5.4 for the Angelina River, 5.6 for the Neches River, and 5.7 for Attoyac Bayou; data from the Texas Commission on Environmental Quality database (TCEQ Data Management and Analysis, Water Quality Planning Division), measured 4–12 times a year). This low pH caused heavy erosion of F. askewi shells, as it was previously recorded for Corbicula fluminea inhabiting acidic waters (streams with pH 5.6) . In a few extreme cases, shells were eroded to the extent that the mussels’ soft tissues were visible.
Our surveys documented the current distribution and change in historical range, densities, and preferred habitat of rare Texas species. Genetic analysis revealed that: (1) F. lananensis is not a valid species; (2) it is likely that only one Fusconaia species (F. askewi) is currently found in East Texas; (3) the presence of F. flava in East Texas is unlikely, however the species may still persist in the Red River basin and upper Trinity River; (4) P. riddellii was well-separated from F. askewi and instead grouped with the P. sintoxia clade.
We found that the specimens from the Sabine and Neches drainages were genetically distinct from all other currently recognized Fusconaia species, as well as from the unusual sequences obtained by Burdick and White , and represented a distinct species. The relatively low percent difference from F. burkei reflects the shorter sequence for F. burkei, which consistently has a low difference from other sequences. Apart from it, all other Fusconaia cox1 sequences differed from F. askewi and F. lananensis by more than 3.5 times as much as the largest difference within the F. askewi-F. lananensis group. In contrast, putative F. askewi sequences from the Calcasieu River in Louisiana matched closely sequences for F. flava, strongly suggesting that this population belongs in F. flava rather than F. askewi. The Calcasieu River runs between the Mississippi (specifically, the Red River) and the Sabine drainages, so faunal exchange could occur in either direction. Study of additional populations would be necessary to determine whether F. askewi is also present in the Calcasieu system or anywhere else east of the Sabine drainage.
All analyses strongly supported a group of Fusconaia lananensis and F. askewi (excluding the Calcasieu specimens). None of the analyses separated F. askewi from F. lananensis. Along with the low percentage difference (especially within the Neches drainage) and presence of morphologically intermediate specimens, this suggests that the F. lananensis is a subjective junior synonym of F. askewi. The distinguishing features noted by Frierson  would represent individual variation. Conversely, the specimens from the Calcasieu drainage are consistently strongly supported as closely related to F. flava and F. cerina. Current molecular data do not clearly distinguish between F. cerina and F. flava[17,47], so the Calcasieu population should probably be regarded as representing F. flava. The variations between Figures Figures11122233 show that relationships within Fusconaia are not well-resolved. Although the support is not strong, all analyses agree that F. subrotunda is basal, followed by a clade of F. cor and F. cuneolus. The remaining Fusconaia species, including F. askewi and F. lananensis, form a group with generally poorly resolved internal relationships. Thus, F. askewi and F. lananensis clearly belong in Fusconaia, are distinct from other currently recognized species (except each other), and are most closely related to the F. cerina-F. flava group, the F. escambia-F. burkei group, F. masoni, F. ozarkensis, and the unidentified flava-like Fusconaia from the Ozark region (hebetata?). Support for the genus Fusconaia is modest in the cox1 only analysis (perhaps due to the partial sequences) but very high in the others. However, relationships of Fusconaia to other genera of Pleurobemini are poorly resolved, and the weakly supported relationships between genera are not consistent between analyses.
Pleurobema riddellii shows consistent but weakly supported affinity for members of the subgenus Sintoxia-P. sintoxia, P. rubrum, and P. cordatum. However, the cox1 analysis shows that other specimens identified as P. cordatum are more distantly related to this group. This may reflect the difficulties of identifying species in the P. cordatum group. Ongoing genetic work on this group  shows further complications, but the morphological similarities of P. riddellii to the P. cordatum group  supports a relationship. Additionally, the only species of Pleurobema that occur in the lower Mississippi drainage are from the P. cordatum group , so the relationship also makes biogeographic sense.
At least four names older than F. askewi are available for Fusconaia species west of the Mississippi, besides F. flava, which was described from the Ohio drainage but occurs also in the upper Mississippi and west of it. Fusconaia ozarkensis (Call) is genetically and morphologically distinctive, but the remaining species have all been synonymized with or confused with F. flava: Fusconaia fulgidus (Lea), from the Red River at Alexandria, Louisiana; F. hebetata (Conrad), from Missouri (unfortunately, no information on which drainage); F. chunii (Lea), from the Trinity River at Dallas, Texas; and F. friersoni (Wright), from Bayou Pierre in the Red River system, De Soto Parish, Louisiana. Although the first three are generally regarded as synonyms of F. flava, as older names they would have priority over F. askewi; F. friersoni was published just before F. askewi, but appears to be a synonym of P. riddellii instead . Burdick and White  sampled one population from the lower Red River drainage near Alexandria and found it genetically similar to F. flava. The present results for the Calcasieu system also suggest that F. flava occurs in the lower Red River system. Graf and Cummings  suggested that F. hebetata might be a valid species. Study of the populations in the Ozark region, building on the work of Utterback  and Graf , should determine whether the conchological variation in populations in this region can be correlated with the genetic divergence found by Burdick and White . If so, F. hebetata and other names based on material from the Ozark region can be assigned to the appropriate population. However, as Burdick and White’s  sequences are quite distinct from those obtained in the present study for F. askewi, it seems safe to assume that F. hebetata is not applicable to the present material from Texas and Louisiana.
This leaves F. chunii as a possible senior synonym of Fusconaia askewi and F. lananensis. Howells et al. synonymized F. chunii with F. flava, but Graf  identified their illustrated F. “flava” from Texas as different from true F. flava. We were unable to obtain live specimens from the Red River systems in Texas for genetic analyses. Specimens suggestive of F. flava from the Neches drainage, sampled in the present study, placed genetically with F. askewi. The Trinity system is immediately west of the Neches and the headwaters of the Sabine, and could easily have exchanged species through stream capture or other interaction. Stream capture occurs when a stream previously connected to one drainage system becomes connected to another, eventually becoming a part of the second drainage system . However, the Trinity River headwaters also adjoin the Red River system in northern Texas. The lower Red River system in Louisiana has F. flava. To the north of the Red River system is the Arkansas system, and the possible F. hebetata haplotype occurs in an Arkansas tributary. The picture is thus very complex, but it seems most likely that F. chunii is a senior synonym of F. askewi.
In contrast to the varying opinions on Fusconaia species, authors have generally agreed on recognizing Pleurobema riddellii. However, there has been some uncertainty about its affinities . The present results provided moderate support for Frierson’s  suggestion that it is relatively closely related to the Pleurobema cordatum group. Most other work on this group has focused exclusively on the Mississippi drainage species and does not mention P. riddellii.
F. askewi is a regional endemic, historically known from the Sabine, Neches, Trinity and San Jacinto rivers in Texas  (Table (Table3,3, Figure Figure4A),4A), and from Louisiana . Simpson  lists F. askewi range from western Louisiana to eastern Texas with type locality as Village Creek, Hardin Co., and the Sabine River, Texas. Strecker  recorded this species in the Angelina, Sabine and Navasota rivers, and from Kickapoo Creek. Neck  reported F. askewi as locally common, but noted that the status over its entire range was unclear. During our surveys we found live F. askewi in four drainages in eastern and northeastern Texas (Table (Table3,3, Figure Figure4B).4B). This species was locally abundant, often dominated mussel assemblages, and several populations were reproducing. The most typical substrate for the species was sand, sand and silt, and gravel with sand.
Fusconaia lananensis was described by Frierson in 1901 , after the first account of Texas unionids was published . Frierson collected 200 specimens of F. lananensis from Lanana and Bonita creeks near Nacogdoches, Texas . Strecker  found live F. lananensis in Lanana Creek, and in the San Jacinto River. In 1990s, few live mussels were found in Attoyac Bayou and Sandy Creek (Angelina River drainage) , and 36 live mussels were found in Village Creek . We found live mussels that fit the description of “F. lananensis” in several waterbodies in East Texas. Due to the similar shell morphologies of F. askewi and F. lananensis, field identification between the two nominal species was very challenging, which is not surprising considering their genetic similarity. Frierson  reports that “Q[uadrula] lananensis is closely allied to Q. askewi Marsh, both by its conchological and anatomical characteristics. It may be differentiated from that shell by being longer, more compressed, more oblique, and its shell is never so inflated and thickened in front as askewi and not so acutely angled on the posterior ridge. Internally, lananensis is rose-colored nearly invariably and the color is uniformly spread over its surface. Askewi is mostly white, and, when colored (pink) the color is almost always confined exterior to the pallial line. Finally, Q. askewi never possess those peculiar pearly excrescences, which seem to belong to lananensis”. We observed several patterns in nacre coloration of Fusconaia from East Texas drainages. There were three forms recorded in the Neches drainage: with entirely white nacre, solid rose/pink, and the form with the pink extrapallial ring described by Frierson . Practically the entire Fusconaia population in the Sabine River had white nacre, while almost none of the Trinity Fusconaia showed the pink extrapallial ring (most of them were white, and a few - solid pink). Therefore, we saw the same features (e.g., pearly excrescences and rose-colored nacre) in both species, with many intermediate forms that were impossible to separate, suggesting that F. lananensis may not be a valid species. This suggestion was supported by our genetic analysis. Habitat and substrate preferences of both Fusconaia spp. were found to be similar as well.
This species is a regional endemic, found in Texas and Louisiana [14,51]. Singley  recorded P. riddellii in Village Creek only; Strecker recorded the species from the Angelina, Sabine, San Jacinto and Trinity rivers in East Texas  (Figure (Figure6A).6A). NatureServe reports a substantial recent decline in this species . During our surveys, we found a total of 132 live P. riddellii in one East Texas river basin (the Neches River), but not at the sites we surveyed on the Trinity River (Figure (Figure6B).6B). Pleurobema riddellii has probably been extirpated from the San Jacinto River. This species was not locally abundant, and not dominant in mussel assemblages. Although most populations were comprised of older animals, several populations were reproducing. Pleurobema riddellii was found exclusively in lotic waters, in relatively shallow areas, most often in sand, or in a mixture of sand, gravel and silt.
The American Fisheries Society considers F. askewi and F. lananensis to be of special concern , and both species are currently listed as state threatened  and as near-threatened by the IUCN . Our recent surveys classified these species as rare (species that were found at low densities in 1 to 9 Texas waterbodies) based on their occurrence and density . The U.S. Fish and Wildlife Service found that substantial scientific information was presented indicating that listing of F. lananensis may be warranted due to the present or threatened destruction, modification, or curtailment of its habitat or range , and a status review for the species was initiated in 2009. However, our study suggested that F. lananensis is not a valid species and it is likely that only one Fusconaia species (F. askewi, senior synonym F. chunii) is currently present in East Texas, thus simplifying conservation efforts. Although we found that F. askewi still inhabits four river basins in eastern and northeastern Texas and can be locally abundant, its distribution range has been reduced in the last 80years: the species have been extirpated from a number of waterbodies in Texas, including Lanana and Bonita creeks, the San Jacinto and Navasota rivers, and Kickapoo Creek (Figure (Figure4).4). The distribution of F. askewi in the Trinity River has been also reduced in the last 40years (Figure (Figure4).4). The species has been extirpated from much of its former range in the upper Trinity River north of SR-7 (Leon/Houston Counties), and appears to be completely absent from the river south of Lake Livingston (D. Barclay, personal observations).
This species was found in only one East Texas drainage (the Neches River), and at very low densities. During the last 80years the distribution range of P. riddellii has been dramatically reduced, and this species has been extirpated from several East Texas waterbodies where it occurred historically (Figure (Figure6).6). Notably, some of these waterbodies (e.g., San Jacinto River) that lost both F. askewi and P. riddellii, are the most highly populated in Texas . At the beginning of 20th century, the San Jacinto River was a home for 29 unionid species, but due to extensive mining, deforestation, damming and urbanization, it lost almost 70% of its former unionid diversity . The U.S. Fish and Wildlife Service found that listing of P. riddellii as threatened or endangered may be warranted due to the present or threatened destruction, modification, or curtailment of its habitat or range resulting from general human modification of the water and adjacent land, siltation, impoundments, and water pollution [9,10], however it is currently listed as threatened only at the state level .
Currently East Texas has predominantly forested watersheds with little urbanization, both factors being important for maintaining the health of aquatic environments . Not surprisingly, this part of Texas is the hotspot for the state’s unionid diversity where almost every river supports from 17 to 28 species . However, Texas is one of the fastest growing states in the nation. The urban population in Texas nearly doubled in the last 30years , with a 21% increase in urbanization since 1990 . Along with growing urbanization, it is predicted that>20 million ha of U.S. forest will be developed over the next 50years [66,67], and>11% of private forests, mostly in the South, could experience substantial increases in housing density by 2030 [68,69]. Considering growing development and water demand, the best measure for conservation of both F. askewi and P. riddellii would be by controlling deforestation, urbanization and water diversion in East Texas watersheds, and particularly the Neches River.
The authors declare that they have no competing interests.
LEB and AYK designed the study and surveyed sites state-wide. DB surveyed additional sites in East Texas. DC carried out the molecular genetic studies and their interpretation. LEB, AYK and DC led, and DB edited the writing. All authors read and approved the final manuscript.
Funding for this study was provided by the U.S. Fish and Wildlife Service’ State Wildlife Grant Program through the Texas Parks and Wildlife Department (2004–2010: PIs LEB, AYK; 2011–2012: PIs LEB, AYK, and M. E. May, M. D. Warriner, and B. Gottfried, TPWD). The ABI 3100 automated sequencer was funded by a NSF equipment grant to C. Lydeard, R. Mayden, M. Powell, and P. Harris (DBI-0070351). We would like to acknowledge the help of Vadim and Dimitry Karatayev, Jesse, David, and Regan Barclays, and Daniel Bennett (TPWD) in data collection, and Sandy Birnbaum (Texas Natural Diversity Database manager) for help with GIS maps. The TNT program was available with the sponsorship of the Willi Hennig Society.