Southeastern Naturalist D.F. Ford, A.D. Walters, L.R. Williams, M.G. Williams, and N.B. Ford 2016 Vol. 15, No. 1 26 2016 SOUTHEASTERN NATURALIST 15(1):26–40 Mussel Assemblages in Streams of Different Sizes in the Neches River Basin of Texas David F. Ford1,*, Ashley D. Walters2, Lance R. Williams3, Marsha G. Williams3, and Neil B. Ford3 Abstract - Freshwater mussel composition and abundance in stream ecosystems are influenced by changes in environmental and hydrologic forces related to stream size. Mussel assemblages in headwater areas tend to be depauperate but distinctive, with species richness and abundance increasing rapidly with increasing stream size. We compared the mussel assemblages in 3 subwatersheds of the Neches River Basin of east Texas: the small Attoyac Bayou (4th order), the intermediate-sized Angelina River (5th order), and the large mainstem of the upper Neches River (6th order). We conducted timed searches at 75 sites and found 12,860 mussels comprising 26 species. As in other river systems, the smallest stream segment, the Attoyac Bayou, had several species more adapted for low flow rates and stagnant conditions. The largest stream segment, the upper Neches River, had the highest diversity and abundance of unionids. We found that mussel abundances increased as the streamsegment size increased. In addition, we found a change in the species composition, with larger, more riverine species becoming more common as the size of the stream increased. A number of state-listed threatened species occur in the mainstem of the Neches River, which emphasizes the need to preserve undisturbed stretches of larger streams. Introduction Community composition in riverine ecosystems, as represented by which species are present and their individual abundances, is influenced by environmental factors that operate at different spatial and geomorphic scales (Atkinson et al. 2012). Local habitat suitability along a stream is shaped by hydrologic forces in a longitudinal pattern that is primarily related to the size of the stream at a given location (i.e., stream order or watershed size; Haag and Warren 1998, Strayer 1983). Habitat variables shift along the length of a stream as a result of the surrounding landscape and physiography; hence, changes in adjoining land usage can impact instream habitats. The influence of abiotic conditions on fish and insect communities has been extensively studied (Chick et al. 2006, Pegg and McClelland 2004, Vannote et al. 1980), but fewer studies have investigated shifts in freshwater mussel assemblages as aquatic habitats change with increasing stream size along longitudinal gradients (Atkinson et al. 2012, Haag and Warren 1998, Strayer 1983, Watters 1992). Freshwater mussels (Unionidae) are a diverse family, which often occur in dense, multispecies assemblages (Strayer 2008). The declines seen in many freshwater mollusk species have been attributed in part to their unique life-history 1Halff Associates, Inc., Richardson, TX 75081. 2Department of Biology, Miami University, Oxford, OH 45056. 3Department of Biology, University of Texas at Tyler, Tyler, TX 75799. *Corresponding author - dford@halff.com. Manuscript Editor: John Placyk Southeastern Naturalist 27 D.F. Ford, A.D. Walters, L.R. Williams, M.G. Williams, and N.B. Ford 2016 Vol. 15, No. 1 strategy (i.e., need for a larval host organism) and the intolerance of many species to pollution and habitat modifications (Brown et al. 2010, Poole and Downing 2004). Despite the paucity of information regarding habitat requirements, resource managers must often make conservation decisions for many at-risk mussel species. Although the availability of suitable habitat is widely accepted as a major factor limiting both the size and diversity of mussel populations, knowledge of how habitat variables influence unionid populations is a complex topic in need of additional research (Strayer 2008, Strayer and Ralley 1993). Microhabitat information at a specific location is recorded only when a researcher is present (i.e., during low-water events) and usually fail to reflect extreme hydrologic events. Therfore associations between the occurrence of mussel species and abiotic factors recorded in typical studies are at best only partially effective at predicting mussel species occurrences or abundances (Brim-Box et al. 2002, Cummings and Graf 2009, Gangloff and Feminella 2007). The most accurate methods for predicting unionid occurrences and abundances within a stream typically involve complex hydraulic models which compute a number of instream flow variables that potentially impact mussels during various life stages (Allen and Vaughn 2010, Gangloff and Feminella 2007). Because of the complex suite of habitat variables that influence the structuring of mussel assemblages, these models often include factors thought to be important for juvenile survival and colonization, along with other variables which influence adult survival, such as substrate stability. However, in addition to the difficulty of obtaining the measurements necessary for these types of analyses, the results often lack insight into a variable’s importance at a precise location or for a specific species during a particular life stage (Haag 2012). The importance of local factors can be explored without the use of these complex hydraulic models by addressing how longitudinal-scale influences mussel community structure in streams of different sizes within a single river basin. Streams join together as they flow downstream creating increasingly larger streams. This increase in size is reflected by the stream’s area at particular points along its length. By examining mussel assemblages in different size streams, the physiographic and anthropogenic factors that influence the mussel communities can be examined. Predictable habitat differences between streams of different size (watershed area) will relate to variables controlled by the volume of water, the movement of sediment, and other factors influenced by stream size (Haag 2012). Haag (2012) reviewed the available data on the impact of stream size on mussel assemblages for the Mississippi region, including the upper Mississippi River, the Mobile Basin, and the Ohio River system. In general, there was a shift from small populations with distinctive species diversity in headwater areas to a greater species richness and abundance as stream size increased. Mussel assemblages in headwater areas consisted of smaller, thinner-shelled, and shorter-lived species adapted to the variable flow rates and disturbances that typically occur in smaller streams (Atkinson et al. 2012, Haag 2012). As stream size increased, there was a rapid gain in species richness and population size due to lower extinction rates Southeastern Naturalist D.F. Ford, A.D. Walters, L.R. Williams, M.G. Williams, and N.B. Ford 2016 Vol. 15, No. 1 28 and more available and varied habitats (Haag 2012). These changes in mussel assemblages in larger streams are associated with the subsequent hydrological and habitat changes that occur as stream size increases. Haag (2012) proposed that many of the species found in intermediate-sized rivers are stream-size generalists and that the increase in the number of species found in the largest rivers is due to an increase in habitat diversity. Understanding how hydraulic variables influence the composition of mussel assemblages will provide useful information for predicting how particular mussel species are impacted by shifts in hydrology, e.g., changes that might occur in a mussel assemblage downstream of a new dam as they relate to changes in flow rates. This information is especially important for states such as Texas, where reservoir creation is a major component of the state’s water plan (Vaughan et al. 2012). Texas has a diverse unionid fauna within its numerous river basins, including several endemic and rare species (Burlakova et al. 2011, Howells et al. 1996). The eastern portion of Texas is a hotspot of mussel diversity and supports the highest mussel richness in the state (Ford and Nicholson 2006, Ford et al. 2009, Howells 1997). There are 5 drainage basins within east Texas, each consisting of complex multi-order streams, which provide opportunities to examine the relationship between stream size and mussel-assemblage structure. In this study, we compare the mussel assemblages in 3 segments of the Neches River Basin, which differ in size (reflected by watershed size): the Attoyac Bayou, the Angelina River, and the mainstem of the upper Neches River. The areas sampled occur within the same general latitude and are relatively close to each other, with similar anthropogenic effects (nearby land usage; Table 1; USGS 2015). Therefore, differences in the mussel assemblages should primarily reflect differences caused by changes in the size of the stream. This study seeks to enhance the understanding of this relationship to a province much further west than previously studied drainages, and to aid in conservation efforts throughout the state of Texas. Methods Site descriptions The Attoyac Bayou is classified as a 4th-order stream (Horizon Systems Corporation 2015) and was the smallest stream segment we surveyed. The bayou’s watershed is predominately a rural setting, situated in east Texas, with its headwaters located Table 1. Percentages of nearby land use for the Attoyac Bayou, Angelina River, and Neches River. Land-use percentages under 0.1% were considered insignificant an d removed from the table. Land-use type Attoyac Bayou Angelina River Neches River Woodland forest with limited cropland/pasture 69.0% 71.0% 59.0% Forest and woodland majority grazed 0.3% 25.0% 30.0% Sub-humid grassland and semiarid grazing land - - 0.2% Wetlands - 3.0% 3.8% Urban areas - - 0.4% Open water 0.9% 0.7% 0.2% Southeastern Naturalist 29 D.F. Ford, A.D. Walters, L.R. Williams, M.G. Williams, and N.B. Ford 2016 Vol. 15, No. 1 in the southeast corner of Rusk County. The bayou drains an area of ~1435 km2 and flows through mostly forested hardwood bottomlands (Table 1; TPWD 1974). Portions of the bayou can be very shallow, and logjams often disrupt flows throughout portions of its length. Substrate within the surveyed areas of the Attoyac Bayou consisted of sand and gravel with pockets of silt, clay, cobble, and organic detritus, along with numerous sunken logs. From its origin, the bayou flows approximately 93 km in a southeast direction, through portions of Rusk, Nacogdoches, San Augustine, and Shelby counties until it joins the Angelina River at the Sam Rayburn Reservoir (Fig. 1). The Angelina River is classified as a 5th-order stream (Horizon Systems Corporation 2015) and was the intermediate-sized stream segment we surveyed. The river begins in Rusk County, at the confluence of Barnhardt, Scober, and Shawnee Figure 1. The sites surveyed for mussels throughout the Attoyac Bayou, Angelina River, and Neches River. The inset map indicates the areas of interest in east Texas. Southeastern Naturalist D.F. Ford, A.D. Walters, L.R. Williams, M.G. Williams, and N.B. Ford 2016 Vol. 15, No. 1 30 creeks. Two impoundments reside along the Angelina River including the Sam Rayburn Reservoir, where the Attoyac Bayou joins the Angelina River, and B.A. Steinhagen Lake (TPWD 1974). Substrate within the surveyed portions of the Angelina River consisted of mud, sand, and gravel with pockets of organic detritus and cobble. This is a larger and deeper river and lacks the shallow portions found in some areas of the bayou. Nearby land use consisted of mostly forested regions with some scattered agricultural areas (Table 1). The river flows for approximately 192 km through mostly forested bottomlands before joining the Neches River at B.A. Steinhagen Lake (Fig. 1; TPWD 1974). The Neches River was the largest of the 3 stream segments we surveyed and is classified as a 6th-order stream (Horizon Systems Corporation 2015). The river begins in Van Zandt County and flows southeast for ~670 km before emptying into the Gulf of Mexico at Sabine Lake, and flows through mostly forested bottomlands (Table 1; TPWD 1974). The river has a drainage area of ~26,000 km2, and consists of several large tributaries, including the Angelina River, which makes up approximately one-third of the Neches River Basin. There are 2 large reservoirs located on the Neches River including Lake Palestine, near the headwaters of the river and B.A. Steinhagen Lake, near the center of the river (Fig. 1). Sampling within the Neches River occurred above B.A. Steinhagen, which was at the same general latitude as the Angelina River and Attoyac Bayou. Substrate in the surveyed portions of the Neches River consisted of sand and gravel, with some clay and mud, and large amounts of organic detritus. Survey methods From 2009 to 2012, we collected mussels from various locations throughout the Attoyac Bayou, Angelina River, and upper Neches River (Fig. 1). Sites consisted of a 50-m reach containing as many different geomorphic units as possible (i.e., riffle, pool, run) so as to be representative of the habitat diversity within the reach, and we surveyed for mussels throughout the entire reach. Because we conducted surveys during an extended drought period, water levels within the Neches River Basin were significantly lower than normal (TCEQ 2015), which allowed us to survey the entire reach, from bank to bank and ~50 m downstream, without the use of SCUBA equipment. We reached each site via kayak from the nearest bridge crossing, and conducted an initial reconnaissance of the shore for mussel shells and buried individuals. We then sampled for mussels using timed tactile and visual searches throughout the entire site. This method provides the most accurate results for mussel species diversity, evenness, and richness (Hornbach and Deneka 1996, Vaughn et al. 1997). We collected, identified, enumerated, and returned to the substrate all live unionids except for voucher specimens. Vouchers were retained in the University of Texas at Tyler’s mussel collection. We also collected and counted at each site recently dead shells, indicated by the presence of tissue, shiny nacre, connected hinges and/or uneroded valves (Howells et al. 1996). However, we did not include long-dead mussels in samples because shells can be transported from Southeastern Naturalist 31 D.F. Ford, A.D. Walters, L.R. Williams, M.G. Williams, and N.B. Ford 2016 Vol. 15, No. 1 locations outside of the sampling area, falsely inflating species counts and potentially altering the species composition of a reach. We standardized surveys on a per-person-hour of searching (Strayer and Smith 2003), and conducted each for at least one person-hour depending on the prevalence of mussels at the site. After the initial person-hour, sampling continued until no mussels were located for at least 10 minutes. We calculated catch per unit effort (CPUE) by dividing the total number of live individuals of a species by the total person-hours spent surveying within each stream segment. Analysis We totaled species richness and abundance for each stream segment. Because of the unequal number of survey sites between the 3 segments, we adjusted species richness employing the rrarefy command from the vegan software package version 2.2-1 (Oksanen et al. 2015) in R version 3.1.2 (R Core Team 2014) to generate a randomly rarefied dataset for the Angelina and Neches River sites using the sample size of the Attoyac Bayou (140 individuals). This procedure made the 3 datasets comparable via the number of individuals recovered. We identified land-usage type for each site using the National Landcover Dataset (Homer et al. 2015) and obtained watershed size from 1/3-arc-second National Elevation Dataset (NED; USGS 2015) layers using ArcGIS v. 10.2.1 and the Spatial Analyst Extension Hydrology tools (ESRI 2014). We calculated stream size for each stream segment from the following locations: (1) at the mouth of the Attoyac Bayou, (2) at the mouth of the Angelina River, and (3) at the southern-most sampling point on the Neches River. Our calculated watershed areas were similar to those found by Carter et al. (1962). We tested relationships between unionid assemblages and nearby land usage (using landcover) and stream size by employing non-metric multidimensional scaling (NMDS) using the software Paleontological Statistics Software Package (PAST) v2.17 (Hammer et al. 2001) with the Bray- Curtis similarity measure. We used NMDS to spatially depict the relationships between unionid assemblages within the Attoyac Bayou, the Angelina River, and the Neches River. We performed a non-parametric multivariate analysis of variance (MANOVA) in the software PAST v2.17 to interpret whether a significant difference existed between the mussel assemblages within the 3 segments. This statistical method allowed the evaluation of the uniqueness of the mussel assemblages within each stream. Results Within the 3 stream segments, we spent a total of 185.4 person hours surveying 3630 m at 75 different sites. We found a total of 12,860 mussels representing 26 species of which 11,502 were found alive (Table 2). Six of the 26 total species found were state-listed threatened species, and include all of the state-listed threatened species known to occur within the Neches River Basin (Table 3; Howells et al. 1996). Southeastern Naturalist D.F. Ford, A.D. Walters, L.R. Williams, M.G. Williams, and N.B. Ford 2016 Vol. 15, No. 1 32 Attoyac Bayou We spent a total of 23 person-hours surveying 11 sites within the Attoyac Bayou and recorded a total of 140 mussels comprised of 14 species, of which 109 were alive (Table 2). The number of species found per site ranged from 0 to 7, and the number of mussels found per site ranged from 0 (at 1 location) to 63. Megalonaias nervosa (Rafinesque) (Washboard) was the most abundant species collected from the bayou, with a CPUE of 0.78. Amblema plicata (Say) (Threeridge) and Quadrula verrucosa (Rafinesque) (Pistolgrip) were the next most abundant species, and both species had a CPUE of 0.74. Threeridge occurred at 8 of the 11 sites, which made it the most widely distributed species. We found 2 state-listed threatened species in the bayou: Fusconaia askewi (Marsh) (Texas Pigtoe) and Fusconaia lananensis (Frierson) (Triangle Pigtoe) (Table 3). All species located in the bayou were found in low numbers, and no species had more than 18 living individuals collected. Only a single recently dead individual was collected for 2 species, Lampsilis teres (Rafinesque) (Yellow Sandshell) and Pyganodon grandis (Say) (Giant Floater). Angelina River We spent a total of 48.9 person-hours surveying 19 sites within the Angelina River and found a total of 2043 mussels comprised of 19 different species, of which 1744 were alive, and obtained an adjusted mean species richness of 16 species via rarefraction (Table 2). The number of species found per site ranged from 1 to 17, and only a single species was located at 2 sites. The number of mussels found per site ranged from 2 to 432 individuals; however, the location with the highest abundance did not have the greatest species diversity. Quadrula mortoni (Conrad) (Western Pimpleback) was the most abundant species in the Angelina River, with a CPUE of 9.76 (Table 3). Western Pimpleback, which we recovered at 15 of the 19 sites, was also the most widely dispersed species in the river. The next most abundant species was Washboard, which had a CPUE of 8.02, followed by Pistolgrip, with a CPUE of 4.85 (Table 3). These 3 species accounted for more than 60% of the total mussel abundance found in the Angelina River. We recovered 4 state-listed threatened species from the Angelina River: Texas Pigtoe, Triangle Pigtoe, Lampsilis satura (Lea) (Sandbank Pocketbook), and Pleurobema riddellii (Lea) (Louisiana Pigtoe). We also found 3 Arcidens confragosus (Say) (Rock Pocketbook); however, only a single individual occurred per site. Table 2. The number of sites surveyed, person-hours expended, meters searched, number of species, and the number of living and dead individuals found in the Attoyac Bayou, Angelina River, and Neches River. Total Total Location Total sites person-hours meters searched Total # species Alive/dead Attoyac 11 23.0 550 14 109/31 Angelina 19 48.9 950 19 1744/299 Neches 45 113.5 2130 25 9649/1028 Totals 75 185.4 3630 26 11,502/1358 Southeastern Naturalist 33 D.F. Ford, A.D. Walters, L.R. Williams, M.G. Williams, and N.B. Ford 2016 Vol. 15, No. 1 Table 3. The total number of sites where each species was located (site freq.), total number of living and dead individuals collected, and catch per unit effort (CPUE) of each species for each stream segment surveyed. The CPUE was calculated by dividing the number of living individuals of a species by the total person-hours spent surveying that segment. * signifies a Texas state-listed threatened species. Attoyac Bayou Angelina River Neches River Species Site freq. Alive/dead CPUE Site freq. Alive /dead CPUE Site freq. Alive /Dead CPUE Amblema plicata (Threeridge) 8 17/4 0.74 9 29/1 0.59 37 475/61 4.18 Arcidens confragosus (Rock Pocketbook) - - - 3 3/0 0.06 13 47/10 0.41 Fusconaia askewi (Texas Pigtoe)* 3 15/1 0.65 11 126/25 2.58 31 695/53 6.12 Fusconaia lananensis (Triangle Pigtoe)* 5 11/7 0.48 9 43/3 0.88 - - - Lampsilis hydiana (Louisiana Fatmucket) 3 2/5 0.09 3 2/1 0.04 15 42/12 0.37 Lampsilis satura (Sandbank Pocketbook)* - - - 3 9/1 0.18 30 77/22 0.68 Lampsilis teres (Yellow Sandshell) 1 0/1 - 14 18/11 0.37 29 140/66 1.23 Leptodea fragilis (Rafinesque) (Fragile Papershell) - - - 7 7/12 0.14 29 55/67 0.48 Megalonaias nervosa (Washboard) 4 18/5 0.78 13 392/98 8.02 35 878/80 7.73 No mussels 1 0/0 - - - - 2 0/0 - Obliquaria reflexa (Threehorn Wartyback) 1 4/1 0.17 12 75/13 1.53 41 1104/89 9.73 Obovaria jacksoniana (Southern Hickorynut)* - - - - - - 7 24/13 0.21 Plectomerus dombeyanus (Valenciennes) (Bankclimber) - - - 8 48/15 0.98 36 408/74 3.59 Pleurobema riddellii (Louisiana Pigtoe)* - - - 7 16/4 0.33 34 421/32 3.71 Potamilus amphichaenus (Texas Heelsplitter)* - - - - - - 7 6/5 0.05 Potamilus purpuratus (Lamarck) (Bleufer) 2 1/1 0.04 14 31/14 0.63 36 153/80 1.35 Pyganodon grandis (Giant Floater) 1 0/1 - 3 5/0 0.10 8 11/2 0.10 Quadrula apiculata (Say) (Southern Mapleleaf) - - - 12 56/4 1.15 30 144/23 1.27 Quadrula mortoni (Western Pimpleback) 4 17/2 0.74 16 477/31 9.76 42 3202/141 28.21 Quadrula nobilis (Gulf Mapleleaf) - - - - - - 13 54/4 0.48 Quadrula verrucosa (Pistolgrip) 1 6/0 0.26 14 237/33 4.85 40 1175/116 10.35 Strophitus undulatus (Creeper) 2 3/0 0.13 - - - 6 7/1 0.06 Toxolasma texasensis (Texas Lilliput) 1 1/0 0.04 - - - 2 4/0 0.04 Truncilla donaciformis (Fawnsfoot) - - - - - - 11 35/14 0.31 Truncilla truncata Rafinesque (Deertoe) - - - 10 170/30 3.48 34 481/61 4.24 Uniomerus tetralasmus (Pondhorn) 1 14/3 0.61 - - - 1 1/0 0.01 Villosa lienosa (Little Spectaclecase) - - - 1 0/3 - 7 10/2 0.09 Totals 11 109/31 4.73 19 1744/299 35.68 45 9649/1028 85.00 Southeastern Naturalist D.F. Ford, A.D. Walters, L.R. Williams, M.G. Williams, and N.B. Ford 2016 Vol. 15, No. 1 34 Three recently dead Villosa lienosa (Conrad) (Little Spectaclecase) were located at a single site (Table 3). Neches River We spent a total of 113.5 person-hours surveying 45 sites within the mainstem of the Neches River and recovered a total of 10,677 mussels comprised of 25 different species, of which 9649 were alive, and obtained an adjusted mean species richness of 16 species via rarefraction (Table 2). The number of species found per site ranged from 0 to 21, with no mussels recovered from 2 locations. The number of individuals found per site ranged from 0 to 835, though the location with the highest abundance did not have the greatest species diversity. Western Pimpleback was the most abundant species within the Neches River, with a CPUE of 28.21 (Table 3). It was located at 42 of the 45 sites, which made it the most widely distributed species as well. The next most abundant species was Pistolgrip with a CPUE of 10.35, followed by Obliquaria reflexa Rafinesque (Threehorn Wartyback) with a CPUE of 9.73 (Table 3). These 3 species accounted for more than 50% of all the mussels found in the Neches River. We located 5 state-listed threatened species in the Neches River: Texas Pigtoe, Sandbank Pocketbook, Obovaria jacksoniana (Frierson) (Southern Hickorynut), Louisiana Pigtoe, and Potamilus amphichaenus (Frierson) (Texas Heelsplitter) (Table 3). Four species found at this segment were not collected at either of the other 2 segments: Southern Hickorynut, Texas Heelsplitter, Quadrula nobilis (Conrad) (Gulf Mapleleaf), and Truncilla donaciformis (Lea) (Fawnsfoot). Texas Heelsplitter was one of the rarest species in our collections, with just 6 living and 5 dead individuals recovered, and was only found in the Neches River (Table 3). We recovered just 7 live individuals of Strophitus undulatus (Say) (Creeper). In addition, we found 10 living specimens of the poorly studied Little Spectaclecase and collected 2 other species in low densities: Toxolasma texasense (Lea) (Texas Lilliput) and Uniomerus tetralasmus (Say) (Pondhorn). NDMS Results The NMDS analysis resulted in 2 dimensions with a stress of 0.16, and the proportion of variance represented by each axis (R2) was 0.22 and 0.29, respectively. A strong pattern was detected for stream size, suggesting that size is an important predictor of unionid assemblages in all 3 stream segments surveyed (Fig. 2). However, a strong pattern was not detected for landcover type. The NMDS ordination suggests that the mussel assemblages in the 3 systems are similar; however, the Neches River is more similar to the Angelina River than the Attoyac Bayou. Non-parametric MANOVA (F = 7.90, P < 0.001) indicated a significant difference between the mussel assemblages within the 3 streams. Discussion Our findings were consistent with studies of smaller-sized streams in more eastern regions of the US. The smallest stream segment of the Neches River Basin surveyed, the Attoyac Bayou, had a depauperate mussel assemblage with low Southeastern Naturalist 35 D.F. Ford, A.D. Walters, L.R. Williams, M.G. Williams, and N.B. Ford 2016 Vol. 15, No. 1 species abundances. Mussel assemblages within the bayou were typically composed of short-lived, fast-reproducing species capable of surviving stagnant and low-flow conditions, such as Lampsilis hydiana (Lea) (Louisiana Fatmucket), Yellow Sandshell, Giant Floater, Creeper, and Pondhorn (Table 3; Howells et al. 1996). All species were found in low numbers, and even the most abundant species in the bayou, the Washboard, had only 18 individuals collected. Washboard, typical of its nature as a deeper-water species (Howells et al. 1996), was found in some of the deeper sections of the bayou. Likely, these individuals are survivors from floods and periods when the Attoyac Bayou held more water. In addition, we found distinctive mussel assemblages in the northern and southern sites of the bayou. The bayou gains in size and depth as it travels downstream, and these distinctive mussel communities likely reflect this. Our findings from the Attoyac Bayou corroborate those summarized by Haag (2012), which indicated that mussel assemblages in smaller streams and headwater areas consist of low densities of Figure 2. Non-metric multidimensional scaling plot for the Attoyac Bayou, Angelina River, and Neches River. The Attoyac Bayou sites are represented with circles, the Angelina River sites are represented with crosses, and the Neches River sites are represented with squares. There was a strong pattern for watershed size, indicating that stream size is an important predictor of unionid assemblages. However, a strong pattern was not detected for landcover type, which suggests that this variable does not influence union id assemblages. Southeastern Naturalist D.F. Ford, A.D. Walters, L.R. Williams, M.G. Williams, and N.B. Ford 2016 Vol. 15, No. 1 36 species that are capable of surviving in areas which experience greater environmental variability. The mussel assemblages of the intermediate-sized Angelina River also followed the findings summarized by Haag (2012), and both species richness and abundance increased with increasing stream size. We recovered 8 additional species from the Angelina River that were not located within the bayou (Table 3). We found only 3 species in the Attoyac Bayou that were not located in the Angelina River: Creeper, Texas Lilliput, and Pondhorn (Table 3). These species tend to be tolerant of lentic habitats that can become stagnant and fragmented during warmer months (i.e., streams with low flow rates and low oxygen content; Howells et al. 1996). It has been suggested that these species are outcompeted in larger streams by more riverine mussel species, which can withstand higher flow rates and different environmental conditions (Haag 2012), and may explain why these species were recovered in the larger Angelina River. Like the Attoyac Bayou, the mussel communities of the Angelina River separated into 2 unique groups, with a distinct mussel assemblage in the smaller northern section, and another distinct assemblage in the larger southern section of the river. These changes in the mussel assemblages of the Angelina River likely relate to longitudinal variations in hydrology with increasing stream size (Haag 2012), since the Angelina River greatly increases in width and depth as it travels downstream. The largest stream segment surveyed, the upper Neches River, had the largest watershed and the highest diversity and abundance of mussels, and followed the longitudinal trend noted by Haag (2012) that the largest streams have the highest mussel abundances and species diversity. We found 4 species only in the Neches River: the state-listed Threatened Southern Hickorynut and Texas Heelsplitter, and the non-listed Gulf Mapleleaf and Fawnsfoot (Table 3). Triangle Pigtoe was the only species found in the smaller Attoyac Bayou and Angelina River but not within the Neches River (Table 3). Three lentic mussel species, Creeper, Texas Lilliput, and Pondhorn (Howells et al. 1996) were located in both the Attoyac Bayou and the Neches River but were not found in the Angelina River (Table 3). These species tend to be more commonly found in headwaters, and were recovered in the Neches River only in small numbers and with CPUEs lower than those in the Attoyac Bayou. Within the Neches River mainstem, these species were only found in small backwater reaches containing lentic microhabitats where riverine species could not survive (Haag 2012) and which mirrored the habitat typically seen throughout the Attoyac Bayou. Unlike the Attoyac Bayou and Angelina River, the mainstem Neches River exhibited no differences in the mussel assemblages between the sites surveyed within it. Both the Attoyac Bayou and Angelina River gain in size, width, and depth as they travel downstream towards the Neches River, while the mainstem of the Neches River remains fairly uniform throughout (TPDW 1974). Because land usage was fairly uniform throughout the 3 streams (Table 1), the differences in the mussel assemblages between the northern and southern portions of the Attoyac Bayou and Angelina River were likely caused by successional changes down the longitudinal Southeastern Naturalist 37 D.F. Ford, A.D. Walters, L.R. Williams, M.G. Williams, and N.B. Ford 2016 Vol. 15, No. 1 gradient as has been described in streams of the Mississippian Region (Haag 2012). Rivers with fairly constant dimensions, such as the mainstem of the Neches River, are predicted to have less-noticeable changes in mussel assemblage structure (Haag 2012), and our findings within the Neches River corroborate this . After adjustments for sample size were made, our analysis indicated the mussel assemblage composition in the Neches River Basin was influenced by stream size (watershed area), as seen in other studies (Atkinson et al. 2012, Haag 2012, Strayer 2008). Mussel diversity and abundance were highest within the largest stream surveyed (mainstem Neches River) followed by the intermediatesized stream (Angelina River), and much lower in the smallest stream (Attoyac Bayou). Typically smaller streams have more unstable or variable habitats than larger streams, and as expected, lower mussel abundance and species diversity were found in these environments, though the species present tended to be unique to these locations (Brown 1984, Haag 2012). Flow rates in small streams increase rapidly after rainfall, resulting in brief but turbulent flood events. These more extreme flows can radically reconfigure a small stream and cause mortality in species not adapted to these variable conditions (Haag 2012). Small streams also tend to be more vulnerable to fragmentation during droughts, and some mussel species are more sensitive to desiccation, low oxygen content, and high temperatures (Spooner and Vaughn 2008). Thus, the mussel assemblages in smaller streams should consist of those species better adapted to deal with the stresses associated with dewatering and variable flow events (Atkinson et al. 2012, Galbraith et al. 2010, Haag 2012) as opposed to those found in larger streams which are adapted for continuous flows and higher rates of competition (Haag 2012). The more unique mussel assemblage inhabiting the Attoyac Bayou and the small backwater areas of the Neches River support this hypothesis, consisting of species likely to survive extreme changes in water depth, flow rates, and temperature when compared to the mussel assemblages of the Angelina River and the larger portions of the Neches River (Howells et al. 1996). Conversely, larger streams tend to have larger mussel populations with greater species diversity (Strayer 2008), and our findings within the Neches River Basin corroborate this research. The mainstem of the Neches River, the largest of the 3 streams surveyed, had the highest diversity and abundance of mussels, with nearly double the number of mussel species as the smaller Attoyac Bayou (Table 3). In large streams, flood events are attenuated and have a longer duration, and typically result in more uniform streambed profiles (Gordon et al. 1992). In addition, large streams provide additional habitats, and the greater volume of water acts as a buffer against thermal extremes (Atkinson et al. 2012, Haag 2012). These characteristics allow a greater number of species to survive and in higher numbers within larger streams, including those species not capable of adapting to extreme flow events or temperature fluctuations. Our findings within the Neches River Basin support these presumptions. Our study is the first in east Texas to examine the influence of increasing stream size on mussel assemblages and provides supportive evidence for the hypothesis Southeastern Naturalist D.F. Ford, A.D. Walters, L.R. Williams, M.G. Williams, and N.B. Ford 2016 Vol. 15, No. 1 38 that stream size (watershed size) influences mussel assemblages within a stream in predictable ways. However, we did not investigate other variables that potentially vary as stream size increases down the longitudinal extent of a river. For example, although the geomorphic and environmental variables that change with increasing stream size have been examined in other studies, more complex issues, such as the distribution of host organisms, likely play a role in the structuring of mussel assemblages in a stream (Schwalb et al. 2012). Informed species management requires the identification of environmental variables that potentially limit mussel populations, and stream size is an important factor which needs to be considered in conservation and management efforts. Here, we have shown that in this east Texas river basin the most abundant and specious mussel populations occur within the largest stream segment of the basin. This information will aid in future conservation projects throughout Texas, as it suggests that undammed stretches of large streams need to be preserved for mussel conservation efforts. Acknowledgments We thank Judith Bilyea, Joel Hunt, and Kaitlyn Holden for their help in the field. We are also grateful to Dr. Kate Hertweck for assistance running the rarefaction analysis. This research was supported by grants to NBF and LRW from Texas Parks and Wildlife Department, the US Fish and Wildlife Service, and North American Coal (Sabine mine). Literature Cited Allen, D.C., and C.C. Vaugh. 2010. Complex hydraulic and substrate variables limit freshwater mussel species richness. Journal of the North American Benthological Society 29:383–394. Atkinson, C. L., J. P. Julian, and C.C. Vaughn. 2012. Scale-dependent longitudinal patterns in mussel communities. Freshwater Biology 57:2272–2284. Brim-Box, J.B., R.M. Dorazio, and W.D. Liddell. 2002. Relationships between streambed substrate characteristics and freshwater mussels (Bivalvia: Unionidae) in Coastal Plain streams. Journal of the North American Benthological Society 21:253–260. Brown, J.H. 1984. On the relationship between abundance and distribution of species. American Naturalist 124:255–279. Brown, K.M., G. George, and W. Daniel. 2010. Urbanization and a threatened freshwater mussel: Evidence from landscape-scale studies. Hydrobiologia 655:189–196. Burlakova, L.E., A.Y. Karatayev, V.A. Karatayev, M.E. May, D.L. Bennett, and M.J. Cook. 2011. Biogeography and conservation of freshwater mussels (Bivalvia: Unionidae) in Texas: Patterns of diversity and threats. Diversity and Distributions 17:393–407. Carter, J.D., O.F. Dent, and H.A. Beckwith. 1962. Drainage areas of Texas streams: Neches river basin and Neches-Trinity coastal area. Texas Water Commission Circular No. 62-03. Chick, J.H., M.A. Pegg, and T.M. Koel. 2006. Spatial patterns of fish communities in the Upper Mississippi River System: Assessing fragmentation by low-head dams. River Research and Applications 22:413–427. Cummings, K.S., and D.L. Graf. 2009. Mollusca: Bivalvia. Pp. 308–384, In J.H. Thorp and A.P. Covish (Eds.). Ecology and Classification of North American Freshwater Invertebrates. 3rd Edition. Academic Press, San Diego, CA. 1021 pp. Southeastern Naturalist 39 D.F. Ford, A.D. Walters, L.R. Williams, M.G. Williams, and N.B. Ford 2016 Vol. 15, No. 1 Environmental Systems Research Institute (ESRI). 2014. ArcGIS software package version 10.2.1. Redlands, CA. Ford, N.B., and M. Nicholson. 2006. Surveys of the freshwater mussels of the Old Sabine Wildlife Management Area, Smith County, Texas. Texas Journal of Science 58:243–254. Ford, N.B., J. Gullett, and M.E. May. 2009. Diversity and abundance of unionid mussels in three sanctuaries on the Sabine River in northeast Texas. Texas Journal of Science 61:279–294. Galbraith, H.S., D.E. Spooner, and C.C. Vaughn. 2010. Synergistic effects of regional climate patterns and local water management on freshwater mussel communities. Biological Conservation 143:1175–1183. Gangloff, M.M., and J.W. Feminella. 2007. Stream-channel geomorphology influences mussel abundance in southern Appalachian streams, USA. Freshwater Biology 52:64–74. Gordon, N.D., T.A. McMahon, B.L. Finlayson, C.J. Gippel, and R.J. Nathan. 1992. Stream Hydrology: An Introduction for Ecologists. John Wiley Publishing, Chichester, UK. 448 pp. Haag, W.R. 2012. North American Freshwater Mussels: Ecology, Natural History, and Conservation. Cambridge University Press, New York, NY. 538 pp. Haag, W.R., and M.L. Warren. 1998. Role of ecological factors and reproductive strategies in structuring freshwater mussel communities. Canadian Journal of Fisheries and Aquatic Sciences 55:297–306. Hammer, ø., D.A.T. Harper, and P.D. Ryan. 2001. PAST: Paleontological statistics software package for education and data analysis. Palaeontologia Electronica 4(1):1–9. Horizon Systems Corporation. 2015. NHDPlus Version 2. Available online at http://www. horizon-systems.com/nhdplus/NHDPlusV2_home.php. Accessed 15 January 2015. Hornbach, D.J., and T. Deneka. 1996. A comparison of qualitative and quantitative collection method for examining freshwater mussel assemblages. Journal of the North American Benthological Society 15:587–596. Howells, R.G. 1997. Status of freshwater mussels (Bivalvia: Unionidae) of the Big Thicket region of eastern Texas. Texas Journal of Science Supplement 49:21–24. Howells, R.G., R.W. Neck, and H.D. Murray. 1996. Freshwater Mussels of Texas. Texas Parks and Wildlife Press, Austin, TX. 224 pp. Oksanen, J., F.G. Blanchet, R. Kindt, P. Legendre, P.R. Minchin, R.B. O’Hara, G.L. Simpson, P. Solymos, M.H.H. Stevens, and H. Wagner. 2015. Vegan: Community Ecology Package. R package version 2.2-1. Available online at http://CRAN.R-project.org/ package=vegan. Accessed 27 February 2015. Pegg, M.A., and M.A. McClelland. 2004. Spatial and temporal patterns in fish communities along the Illinois River. Ecology of Freshwater Fish 13:125–135. Poole, K.E., and J.A. Downing. 2004. Relationship of declining mussel biodiversity to stream-reach and watershed characteristics in an agricultural landscape. Journal of the North American Benthological Society 23:114–125. R Core Team. 2014. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. Available online at: http://www.R-project. org/. Accessed 27 February 2015. Schwalb, A.N., T.J. Morris, N.E. Mandrak, and K. Cottenie. 2012. Distribution of unionid freshwater mussels depends on the distribution of host fishes on a regional scale. Biodiversity Research 19(4):446–454. Spooner, D.E., and C.C. Vaughn. 2008. A trait-based approach to species’ role in stream ecosystems: Climate change, community structure, and material cycling. Oecologia 158:307–317. Southeastern Naturalist D.F. Ford, A.D. Walters, L.R. Williams, M.G. Williams, and N.B. Ford 2016 Vol. 15, No. 1 40 Strayer, D.L. 1983. The effects of surface geology and stream size on fresh-water mussel (Bivalvia, Unionidae) distribution in southeastern Michigan, USA. Freshwater Biology 13:253–264. Strayer, D.L. 2008. Freshwater Mussel Ecology: A Multifactor Approach to Distribution and Abundance. University of California Press, Berkeley, CA. 216 pp. Strayer, D.L., and J. Ralley. 1993. Microhabitat use by an assemblage of stream-dwelling unionaceans (Bivalvia), including two rare species of Alasmidonta. Journal of the North American Benthological Society 12:247–258. Strayer, D.L., and D.R. Smith. 2003. A guide to sampling freshwater mussel populations. American Fisheries Society Monograph 8, Bethesda, MD. 97 pp. Texas Commission on Environmental Quality (TCEQ). 2015. Texas drought information. Available online at http://www.tceq.state.tx.us/response/drought. Accessed 10 June 2015. Texas Parks and Wildlife Department (TPWD). 1974. An analysis of Texas waterways: A report on the physical characteristics of rivers, streams, and bayous in Texas. Texas Parks and Wildlife Press, Austin, TX. Available online at http://tpwd.texas.gov/publications/ pwdpubs/pwd_rp_t3200_1047/index.phtml/#F. Accessed 25 May 2015. US Geological Survey (USGS). 2015. The national map. Available online at http://nationalmap. gov/index.html. Accessed 2 August 2015. Vannote, R.L., G.W. Minshall, K.W. Cummins, J.R. Sedell, and C.E. Cushing. 1980. The river continuum concept. Canadian Journal of Fisheries and Aquatic Sciences 37:130–137. Vaughan, E.G., J.M. Crutcher, T.W. Labatt III, L.H. McMahan, B.R. Bradford Jr., and M. Cluck. 2012. Water for Texas: 2012 state water plan. Texas Water Development Board. Available online at http://www.twdb.texas.gov/publications/state_water_ plan/2012/2012_SWP.pdf. Accessed 25 May 2015. Vaughn, C.C., C.M. Taylor, and K.J. Eberhard. 1997. A comparison of the effectiveness of timed searches vs quadrat sampling in mussel surveys. Pp. 157–162, In K.S. Cummins, A.C. Buchanan, C.A. Mayer, and T.J. Naimo (Eds.). Conservation and Management of Freshwater Mussels II: Initiatives for the Future. Proceedings of a UMRCC symposium, 16–18 October 1995, St. Louis, MO. Watters, G.T. 1992. Unionids, fishes, and the species-area curve. Journal of Biogeography 19:481–490.