Southeastern Naturalist J.L. Clayton, S.A. Miller, and R. Menendez 2015 Vol. 14, Special Issue 7 261 Canaan Valley & Environs 2015 Southeastern Naturalist 14(Special Issue 7):261–275 In-Situ Bioassay Response of Freshwater Mussels to Acid Mine Drainage Pollution and its Mitigation Janet L. Clayton1,*, Shelly A. Miller1,2, and Raymond Menendez1,3 Abstract - Many West Virginia watersheds have been affected by mining activities. Runoff water, known as acid mine drainage (AMD), is acidic and tends to have a high metal content. Over the last several decades, various strategies have been employed to remediate the conditions caused by AMD and restore water quality to levels that support diverse native organisms. Located in the Canaan Valley, Tucker County, WV, the Blackwater River and its tributaries have been the focus of restoration efforts. Limestone application has proved to be among the most successful treatments to raise pH and ameliorate the effects of AMD. Our objectives were to use the introduced freshwater mussel Strophitus undulatus (Creeper) in bioassays to determine the effects of AMD and AMD neutralization on the health and survival of individuals and the potential dispersal of the species. In addition, we sought to determine the effects on mussels of limestone sediments that accumulate as a result of water treatments. Introduction Bivalve mollusks have been used effectively to assess the effects of various pollutants on aquatic systems (Gerhardt 1993, Kraak et al. 1991, Tessier et al. 1984). They are ideal organisms for biomonitoring because they are large, sessile, and filter feeding. Although several studies have been conducted on the effects of heavy metals (Hemelraad et al. 1988, Keller and Zam 1991, Winter 1996), little research has been conducted on the tolerance of bivalves to pH (Malley et al. 1988, Pynnonen 1990). It has been demonstrated that mussels are highly susceptible to sedimentation through habitat degradation. Holland-Bartels (1990) showed that mussels were generally tolerant to a variety of sediment types, but species diversity was significantly reduced at sites with finer sediments. Aldridge et al. (1987) determined that high levels of suspended solids led to mussel starvation by altering their filter-clearance rates, oxygen uptake, and nitrogenous excretion rates. Strophitus undulatus Say (Creeper) is assumed to have been introduced into the upper Blackwater River in the mid-1900s and has become well established above Beaver Creek (West Virginia Division of Natural Resources [WVDNR], unpubl. data). Due to acid mine drainage (AMD), the bivalves do not occur below this tributary (WVDNR, unpubl. data). The Blackwater River lies within Tucker and Grant counties in northeast-central West Virginia (Fig. 1). This river was once known for its valuable sport fishery; however, for over 35 years, the lower 19 km of the Blackwater River has been severely impacted by acid mine drainage 1West Virginia Division of Natural Resources, PO Box 67, Elkins, WV 26241.2Current address - Oregon Department of Fish and Wildlife, 28655 Highway 34, Corvallis, OR 97333. 3Retired. *Corresponding author - Janet.L.Clayton@wv.gov. Southeastern Naturalist J.L. Clayton, S.A. Miller, and R. Menendez 2015 Vol. 14, Special Issue 7 262 (WVDNR, unpubl. data). The first significant source of AMD to the river enters from Beaver Creek. The water quality remains about the same from Beaver Creek to the North Fork (11 km from the mouth), the influx of which further degrades the water quality of the Blackwater River and severely limits the aquatic biota (WVDNR, unpubl. data). To remediate the effects of AMD pollution, a rotary-drum limestone-treatment station was constructed on the Blackwater River approximately 1 km above the mouth of Beaver Creek (Zurbuch et al. 1997). The treatment station was designed to improve the water quality of the Blackwater River down to the North Fork, which contributes the greatest acid load. The primary purpose of this stream restoration project was to develop a trout fishery. The treatment station began operation in September 1994. Using high-grade limestone (>97% CaCO3), the treatment system was designed to add CaCO3 at 28 mg/L and add up to 8.6 metric tons of limestone slurry per day at high flows (WVDNR, unpubl. data). The system automatically adjusts slurry output to changes in stream flow, and from 14 November 1994 to 14 June 1995 supplied 1114 metric tons of limestone to the Blackwater River. A back-up, electric-powered limestonepowder doser was installed adjacent to the rotary-drum station. It was only operated periodically throughout the test period and supplied 140 metric tons of limestone (WVDNR, unpubl. data). Limestone particles accumulate on the stream bottom for a distance of approximately 2 km downstream from the doser systems, primarily within the first kilometer (WVDNR, unpubl. data). Once the low-pH, metal-contaminated water of Beaver Creek reaches the treated, buffered Blackwater River, the metals begin to precipitate and rapid dissolution of the limestone takes place. This treatment has significantly improved the river’s water quality such that 3.9 km below the rotary drum station, the mean pH had increased from 6.2 (range = 5.3–7.3) prior to treatment to 7.2 (range = 6.2–8.3) as of December 1997 (WVDNR, unpubl. data)]. Subsequently, a marked improvement in the fishery occurred—anglers reported outstanding trout fishing in 1996 and in 1997 (WVDNR, unpubl. data). At a sampling station just above the mouth of the North Fork, fishery surveys found 11 species prior to treatment and 17 species the first year following start-up. Fish biomass increased from 3.2 to 9.1 kg/ha over the same period (WVDNR, unpubl. data). The presence of the Creeper in the Blackwater River provided a unique opportunity to determine the effects of AMD and AMD neutralization on the health and survival of individuals and the potential dispersal of the species. Another objective of this study was to determine if the limestone sediments found immediately below the doser were harmful to the bivalve population. We assessed bioavailability of metals associated with the neutralized AMD because bivalve tissues tend to accumulate them. This research will help to determine the effect of restoration strategies of degraded ecosystems on declining mussel populations. Southeastern Naturalist J.L. Clayton, S.A. Miller, and R. Menendez 2015 Vol. 14, Special Issue 7 263 Methods Study area A map of the study area is provided in Figure 1. The control site (station 1) was located in the Blackwater River above Yellow Creek and approximately 3.5 km above the treatment station. Station 2 was located within the limestone plume directly below the treatment station on the Blackwater River above the AMD input of Beaver Creek. Station 3 was located in the Blackwater River approximately 0.8 km below the mouth of Beaver Creek and 1.8 km below the treatment station. Station 4 was located within the AMD-impacted Beaver Creek, near the mouth. River water data We measured water temperature and pH in the field at the time of mussel observations using an Orion model 250A pH meter (Thermo Fisher Scientific, Waltham, MA). We measured turbidity in the laboratory immediately upon return from the field using a Hach model 2100A turbidity meter (Hach Company, Loveland, CO). Staff from West Virginia Department of Environmental Protection (WVDEP) and the WVDNR collected the water samples bimonthly as part of a larger project. The WVDEP had a water-quality station located 3.9 km downstream of the dosers, or 2.1 km downstream of our station 3. We used water-quality data from the WVDEP station to represent conditions at station 3, which was not directly sampled for extensive laboratory testing. The WVDEP samples were analyzed in the laboratory for hot acidity, alkalinity, conductivity, total and dissolved calcium, aluminum, iron, manganese, and sulfates (Clesceri et al. 1989). The results of the field data collected in this study were broken into 2 time periods because all mussels at Figure 1. Map of upper Blackwater River watershed showing freshwater mussel in-situ bioassay and chemistry stations. Southeastern Naturalist J.L. Clayton, S.A. Miller, and R. Menendez 2015 Vol. 14, Special Issue 7 264 station 4 were dead by 27 March 1995. Direct comparisons of station 4 with the other stations were only made using data through 27 March 1995. Discharge data were obtained from a US Geological Survey (USGS) gauging station located at station 3. Bioassay We collected Creeper specimens from the Blackwater River near Yellow Creek throughout the fall of 1994 for use in an in-situ bioassay. Mussels were held in cages at this location (station 1) prior to the test. We marked all mussels by attaching a numbered tag to the shell periostracum using epoxy. We randomly selected mussels and placed 10 in each of three replicate cages at every station. Shell lengths of test mussels ranged from 41.1 to 95.9 mm, and analysis between stations showed no significant difference in size (ANOVA P > 0.05). Mean shell length per station ranged from 74.5 mm at station 3 to 80.1 mm at station 1. The bioassay began on 14 November 1994, and we observed mussel survival for 8 consecutive days. Once we determined that conditions were not acutely toxic, we checked mussels for survival at least twice a week through 19 December 1994. From this time on, we checked mussels weekly for survival until the bioassay was terminated on 14 June 1995 (212 days). We classified a mussel as dead if it was gaping and unresponsive to touch. We also recorded notes on surviving mussel activity (foot movement or shell gape). Metal analyses We sacrificed 5 surviving mussels from each of stations 1, 2, and 3 for tissue analysis at the termination of the test. We also analyzed 5 mussels from station 4 (final mortality on 27 March 1995). After we removed whole-body mussel tissues from the shells, the tissues were rinsed in distilled water to remove sediments (guts were not cleared), placed in plastic bags, and frozen. We later packed samples in dry ice and shipped them overnight to the Environmental Trace Substances Research Center, Columbia, MO, where the mussel tissue was analyzed for aluminum, arsenic, boron, barium, beryllium, cadmium, chromium, copper, iron, magnesium, manganese, molybdenum, nickel, lead, selenium, strontium, vanadium, and zinc using inductive coupled plasma (ICP) scan. Mercury was analyzed by cold vapor atomic absorption. All remaining mussels not analyzed for metals were shucked, rinsed with distilled water, and dried to determine ashfree dry weights. Statistical analyses We analyzed bioassay results with survival analysis, which uses Cox’s proportional hazard method to fit the model. This analysis modeled the probability of survival over time at each site and compared these probabilities between sites (SAS 1995). To compare metal content in tissue, the metal data were normalized (Helmelraad 1986) to the average wet weight of the test mussels by the method of least-squares. The least-squares metal content of mussel tissue from each station was adjusted to that expected for an average-sized mussel from the data set. Southeastern Naturalist J.L. Clayton, S.A. Miller, and R. Menendez 2015 Vol. 14, Special Issue 7 265 Comparisons of tissue-metal content between stations were made with ANOVA. Comparisons of water quality parameters were made with ANOVA and paired t-tests. All statistical analyses were performed with J.M.P. statistical software (SAS 1995). Results Water quality Overall water quality of the Blackwater River has greatly improved as a result of the application of CaCO3 (Zurbuch et al. 1997). The habitat data collected during the bioassay are presented in Tables 1 and 2. The pH values in the untreated section of the Blackwater River (station 1) and in the treated section immediately below Beaver Creek (station 3) were significantly higher than Beaver Creek (station 4) and significantly lower (P < 0.0001) than the treated river above Beaver Creek (station 2). Water temperatures were not significantly different between stations. The water-quality values at station 3 below Beaver Creek were similar to those of the Blackwater River above the treatment station at station 1 (Table 3). These two stations only differed significantly in conductivity (P = 0.0009) and dissolved calcium (P = 0.0109). Beaver Creek differed significantly from all other stations in hot acidity (P < 0.0001), alkalinity (P < 0.0001), conductivity (P < 0.0001), and sulfate (P = 0.0023). Station 2, immediately below the doser, differed significantly from all other stations in alkalinity (P < 0.0001) and total calcium (P = 0.0002). Levels of the metals aluminum, iron, and manganese showed no difference between stations. Table 2. Mean and range of field-chemistry values during in-situ bioassay in the Blackwater River watershed using Strophitus undulatus (Creeper). Values are for the period 14 November 1994–14 June 1995. Station 1 Station 2 Station 3 Mean Range Mean Range Mean Range Temperature (ºC) 6.7 0.7–19.9 6.6 0.8–18.5 7.1 0.7–19.5 pH 6.8 6.2–7.4 7.7 6.6–8.7 6.9 6.1–7.7 Turbidity (NTU) 4.2 2.3–11 18.2 3.1–280 5.1 3.4–13 Table 1. Mean and range of field-chemistry values during in-situ bioassay in the Blackwater River watershed using Strophitus undulatus (Creeper). Values are for the period 14 November 1994–27 March 1995, at which time station 4 was terminated. Station 1 Station 2 Station 3 Station 4 Mean Range Mean Range Mean Range Mean Range Temperature (ºC) 4.7 0.7–10.4 4.9 0.8–10.1 5.0 0.7–9.9 4.7 0.7–8.6 pH 6.7 6.2–7.4 7.4 6.6–8.6 6.8 6.1–7.7 5.0 4.6–5.7 Turbidity (NTU) 3.7 2.3–5.3 20.3 3.1–280 5.0 3.4–13 3.4 1.6–16 Southeastern Naturalist J.L. Clayton, S.A. Miller, and R. Menendez 2015 Vol. 14, Special Issue 7 266 Table 3. Results of chemical analysis (mean and range) of water samples collected within the Blackwater River drainage during an in-situ bioassay conducted during limestone treatment from 14 November 1994 to 14 June 1995. Station 3 Station 4 Station 1 Station 2 DEP park (Beaver Creek) Untreated control Below doser station Untreated AMD pH 7.24 8.39 7.38 4.86 (6.93–7.53) (8.27–8.5) (6.99–7.75) (4.61–5.36) Hot acidity (mg/L) 1.1 0.2 0.09 10.0 (less than 1–2) (0–1) (0–1) (5–21) Alkalinity (mg/L) 23.0 35.4 20.4 1.4 (15–35) (29–46) (12–29) (1–3) Total calcium (mg/L) 10.1 20.0 12.4 12.9 (6.7–16.8) (11.0–30.4) (9.1–17.6) (10.5–20.0) Dissolved calcium (mg/L) 8.9 13.2 12.4 12.7 (6.7–10.9) (10.2–16.8) (9.0–17.6) (10.3–16.8) Conductivity (μmhos) 53.7 69.15 74.46 119.92 (44.6–68.4) (60.4–95.5) (61.0–94.6) (90.6–145.0) Aluminum (mg/L) 0.60 0.50 0.97 1.50 (0.20–1.10) (0.30–0.70) (0.50–1.70) (0.90–2.50) Iron (mg/L) 0.80 0.67 0.93 1.60 (0.40–1.30) (0.30–0.90) (0.40–1.80) (0.60–3.40) Manganese (mg/L) 0.08 0.08 0.37 0.57 (0.05–0.10) (0.05–0.09) (0.14–0.81) (0.55–0.58) Sulfate (mg/L) 7.69 7.54 15.26 45.27 (7.1–8.3) (6.3–8.8) (14.2–16.3) (39.6–50.9) Bioassay We observed the first mussel mortality on 11 January 1995 (58 exposure days) at station 3 below Beaver Creek (Fig. 2). The first mortalities observed at stations 1, 2, and 4 were on days 77, 161, and 64 respectively. The mortality observed in Beaver Creek (station 4) exceeded 10% at 64 days. At this station, mortality continued to sharply increase with no survival after 126 test-days. Mortality at the control station (station 1) exceeded 10% by test day 161. Mortality at all stations (excluding Beaver Creek, station 4) was similar until day 197 when station 2 (at the doser) exceeded 46% mortality. The proportional hazard-fit model for all stations indicated that mussels were most likely to die at station 4, Beaver Creek (χ2 = 67.81, P < 0.0001). Analysis between the remaining stations indicated no difference; however, a risk ratio of 0.769 indicated that mussels were more likely to live longer at station 1 (control) than at station 2 (below the doser). Observations of foot activity throughout the test revealed that mussels exposed to AMD remained closed tightly for the majority of observations made. Once animals became severely stressed they started to gape, and demonstrated a marked reduction in tactile response. The mussels at station 2 immediately below the doser also appeared to remain closed more than those at the control station and below Beaver Creek. Southeastern Naturalist J.L. Clayton, S.A. Miller, and R. Menendez 2015 Vol. 14, Special Issue 7 267 Cages at all stations appeared to accumulate varying amounts of silt or sediments throughout the study period. By far, the largest amounts were observed in the cages at station 2 immediately below the doser. Limestone particles continually settled within the cages though they were periodically flushed out during high-flow events and when we checked mussels for survival. Turbidity levels were also highest at this station (Tables 1 and 2). The highest tubidity values were related to periods when the backup powder-doser was in operation; pH spikes also occurred during these events. The mussels in Beaver Creek (station 4) developed an iron coating, and those at station 3 tended to become coated with a black, slimy material. The latter material never accumulated in great amounts but was generally evident. Mussel condition The physical data for the Creepers used in the in-situ bioassay are presented in Tables 4, 5, and 6. The mean shell lengths and initial wet weights were not Figure 2. Cumulative percent mortality exhibited by Strophitus undulatus during an insitu bioassay conducted 14 November 1994–14 June 1995. Table 4. Mean and range of shell length and wet weight of Strophitus undulatus (Creeper) at the onset of the in-situ bioassay in the Blackwater River watershed 14 November 1994. Station 1 Station 2 Station 3 Station 4 Shell length (mm) 80.05 74.90 74.26 74.86 (53.4–95.9) (46.9–90.4) (53.9–88.8) (41.1–93.9) Initial wet weight (g) 71.87 61.63 57.63 62.57 (24–123) (15–107) (24–93) (11–123) Southeastern Naturalist J.L. Clayton, S.A. Miller, and R. Menendez 2015 Vol. 14, Special Issue 7 268 significantly different, although those randomly selected for station 1 (control) averaged somewhat larger. Our analyses showed no difference in the dry weights of the mussels sampled at the termination of the test on 14 June 1995; however, the average weight of mussels from station 1 was higher. Although ash-free dry weights followed the same trend, the average amount of ash was greatest for samples from stations 2 and 3. In Beaver Creek (station 4), a comparison of initial wet weights to wet weights at mortality using a paired t-test indicated significant weight loss at time of death (P < 0.0001). The dry weights at death of the mussels from Beaver Creek were significantly greater (P < 0.0001) than the other 3 stations. However, this comparison is misleading because the mussels were still gravid when total mortality occurred (March), but those from the other stations collected at the end of the test in June were not gravid. We found that during this study, most mussels had discharged their glochidia by the end of April. Metal bioaccumulation Results from the analysis of whole-body tissues (mean normalized weight and measured ranges) are presented in Table 7. Mussels tested from station 4 were only exposed through 27 March 1995, by which time all mussels had succumbed to the AMD conditions of Beaver Creek. These mussels were exposed for a shorter time and were tightly closed on observation. Several metal concentrations, including barium, magnesium, manganese, mercury, and zinc, were found to be significantly lower than in mussels from the other stations (Table 7). Mussels from station 3 (below Beaver Creek) showed significantly higher levels of beryllium, nickel, and zinc than all other stations. Table 6. Physical data collected from Strophitus undulatus (Creeper) at station 4 used in the in-situ bioassay in Beaver Creek 14 November 1994–27 March 1995 (at total mortality). Variable Mean (range) Shell length (mm) at start (n = 30) 74.86 (41.1–93.9) Wet weight (g) at start (n = 30) 62.57 (11–123) Wet weight (g) at end (n = 29) 53.79 (11–101) Dry weight (g) at mortality (n = 24) 1.82 (0.29–3.01) % ash at mortality (n = 24) 28.73 (24.36–32.8) Table 5. Physical data (mean and average) collected from Strophitus undulatus (Creeper) that survived the in-situ bioassay in the Blackwater River watershed 14 November 1994–14 June 1995. Length given in mm and weight in g. Station 1 (n = 15) Station 2 (n = 9) Station 3 (n = 10) Shell length at start 76.23 (53.4–93.6) 68.31 (50.8–78.7) 74.69 (61.9–88.8) Wet weight at start 63.33 (24–108) 46.00 (10–64) 56.50 (33–93) Dry weight at end 0.96 (0.39–1.58) 0.64 (0.26–0.92) 0.80 (0.43–1.44) % ash at end 30.12 (22.96–37.3) 32.22 (28.19–35.45) 30.76 (24.88–34.79) Southeastern Naturalist J.L. Clayton, S.A. Miller, and R. Menendez 2015 Vol. 14, Special Issue 7 269 Table 7. Weight-adjusted mean metal concentrations of mussels and results of significance testing for treatment effects. Weight is least square means (LSM) weight of mussels; metal concentration is the LSM metal concentration in a mussel of the mean weight for that location. P is the probability of significant difference between LSM metal values across stations (n = 5 per station). Measured range of metal concentrations (μg/g dry weight) is presented in parentheses. Selenium was below detection limits at all stations. Metals with the same letter were not significantly different (P less than 0.05) from each other. Station 1 Station 2 Station 3 Station 4 P-value Weight 22.01 13.91 16.65 25.48 Aluminum 2631.69 AB 588.31A 3568.17B 728.30AB 0.0008 (593–4400) (290–1020) (1930–5590) (537–1240) Arsenic 5.38 7.86 3.87 5.85 0.1355 (0–8.4) (7.2–9.2) (0–8) (4–7) Barium 1380.0A 1365.0A 1396.0A 763.0B 0.0002 (1180–1570) (1160–1570) (1200–1860) (518–902) Beryllium 0.216A 0.146A 0.566B 0.186A less than 0.0001 (0.12–0.29) (0.10–0.23) (0.35–0.79) (0.16–0.25) Boron 0.20 -0.01 0.79 0.00 0.0524 (0–1) (0–0) (0–2) (0–0) Cadmium 2.881 2.770 2.681 1.867 0.3894 (2.5–3.1) (2.0–6.1) (2.4–3.5) (1.3–2.2) Chromium 2.94 86.26 14.24 15.11 0.2012 (3.0–6.1) (3.0–264.0) (3.5–9.2) (1.0–3.0) Copper 8.23 133.82 38.83 21.00 0.4072 (3.3–4.4) (4.9–481.0) (4.9–95.4) (6.9–8.6) Iron 4680.88AB 2980.91AB 6236.99A 1957.21B 0.0001 (2700–5990) (1820–4370) (4400–8010) (1130–2740) Magnesium 1204.96A 1015.74A 1213.47A 578.82B less than 0.0001 (1020–1420) (854–1180) (1010–1420) (514–614) Manganese 14454.42A 13878.97A 15132.23A 7796.38B 0.0001 (12,900–16,600) (12,400–16,600) (13,800–19,100) (5360–9680) Mercury 0.508A 0.493A 0.477A 0.242B 0.0001 (0.464–0.539) (0.451–0.604) (0.423–0.534) (0.17–0.29) Molybdenum 1.701 3.279 2.153 0.904 0.0760 (1.7–1.9) (1.6–6.6) (1.9–2.3) (0.99–1.3) Nickel 3.12A 4.79A 11.77B 2.51A 0.0012 (2.1–4.1) (1.7–8.8) (7.3–19) (2.6–3.9) Lead 2.051 0.685 2.541 0.122 0.1531 (0–3) (0–4) (0–6) (0–0) Strontium 165.87A 174.73A 165.66A 131.94B 0.0006 (154–176) (158–212) (154–207) (104–147) Vanadium 1.15 0.07 1.38 0.07 0.2130 (0–3.6) (0–0) (0–4.1) (0–0) Zinc 237.237A 219.492A 299.103B 144.567C less than 0.0001 (212–258) (189–313) (263–343) (114–155) Southeastern Naturalist J.L. Clayton, S.A. Miller, and R. Menendez 2015 Vol. 14, Special Issue 7 270 Discussion The Blackwater River has shown significant improvements in water quality and biological diversity as a result of limestone treatment (Zurbuch et al. 1997). The acid-polluted waters of Beaver Creek were successfully neutralized upon entering the treated section of Blackwater River such that pH levels below Beaver Creek equaled those of station 1. The minimum pH recorded at the WVDEP station downstream (station 3) was 5.36 prior to treatment and 6.99 during treatment. Likewise, alkalinity significantly increased, and hot acidity significantly decreased. Prior to treatment, a fish kill occurred in the Black Fork and Cheat Rivers below the mouth of the Blackwater River, resulting in significantly reduced fish populations (WVDNR, unpubl. data). Prior to treatment, the Blackwater River consumed nearly all excess alkalinity of the Dry Fork at their confluence. Following treatment, alkalinity increased significantly (Zurbuch et al. 1997). We found that Creepers were able to survive in water with pH values from 4.6 to 5.4 for more than 9 weeks at temperatures ranging from 0.8 to 8.6 °C. Previously, Pynnonen (1990) found no toxicity demonstrated in unionids exposed to pH levels between 4 and 4.5 for up to four weeks at temperatures of 10–11 °C. The only way for mussels to avoid acidic conditions or other contaminants is by valve closure. Several authors have noted that bivalves can withstand several days of anoxic conditions and subsequently diminish the influx of hydrogen ions and the efflux of electrolytes (Holwerda and Vennhof 1984, Pynnonen 1995). Valve activity has been shown to decrease with decreased pH values, with extended closures occurring around pH 5.0 (Pynnonen and Huebner 1995). Another threat to mussels in acid-polluted streams is the reduction of calcium reserves to counteract acid effects. Pynnonen (1995) noted that bivalves can utilize large CaCO3 reserves, yet even a minor decrease in the CaCO3 reserves of individuals living in waters low in calcium could severely affect both reproduction and shell growth. Beaver Creek, however, averaged 12.7 mg/L of dissolved calcium throughout the test period (a level not considered low in this system). Many of the shells collected for the study had damaged periostraca. Although some mussels in all stations developed holes completely through the shell, those in Beaver Creek more readily developed holes during the test. In our study, some mortality may have been a result of secondary factors such as microbial contamination due to these holes. Kat (1982) found prolonged exposure to acid conditions damaged the periostracum, which led to microbial contamination. We believe the bivalves in Beaver Creek (station 4) most likely succumbed to the effects of acidic conditions rather than metal contamination. In most cases, the tissue-metal concentrations of these mussels were less than at the other stations even though metal concentrations (aluminum, iron, and manganese) in the water column tended to be higher at this station. There are several possible explanations for this finding. Simple relationships between trace-metal concentrations in the environment and organisms are seldom found, although direct Southeastern Naturalist J.L. Clayton, S.A. Miller, and R. Menendez 2015 Vol. 14, Special Issue 7 271 proportions between these concentrations are frequently assumed (Tessier et al. 1984). In addition, there are interactive effects between trace elements that affect accumulation. Sediment constituents expected from AMD, such as amorphous iron oxyhydroxides, have been shown to have a protective or competitive effect against other metal contaminants (Tessier et al. 1984). Hemlraad et al. (1987) found that zinc reduced the accumulation of cadmium in whole animals to about one half the value for cadmium exposure alone. Increased water hardness may also have an ameliorative affect on metal toxicity (Keller and Zam 1991, Markich and Jeffree 1994, Winter 1996). Markich and Jeffree (1994) speculated that this moderating effect most likely results from the competitive binding of calcium. Increased concentration of calcium reduces the rate of trace-metal absorption, and hence, the resultant toxicity of trace metals. While most lab studies have dealt with toxicity of dissolved metal concentrations, fewer studies have investigated the effects of the particulate fraction of the total metal concentration. At station 3, the mussels were exposed to higher concentrations of particulate metals that precipitated following the rise in pH compared to station 4, which was exposed to the metals primarily in the dissolved form. Though many studies have determined that the most likely route of tracemetal uptake is through the gills and mantle (Tessier et al. 1984), we suspect that a large portion of the metals associated with stations 1, 2, and 3 were actually particulate fractions collected in the gut and not incorporated in the tissues. Anderson (1977) listed viscera secondarily to gills in the ability to accumulate metals, and Adams et al. (1981) described the digestive gland as containing the highest concentration of cadmium. There are a few other possibilities for the non-significant differences in tissuemetal concentrations between stations. Although allowing mussels to purge their guts is an accepted practice, it may also allow partial elimination of some biologically incorporated metals. Metals have been demonstrated to decrease after only two days of gut clearing (Latouche and Mix 1982). Renzoni and Bacci (1976) demonstrated the loss of mercury over several weeks after removing mussels from contaminated water. This type of metal loss may also be variable depending upon element type. The mussels in Beaver Creek (station 4) were mostly inactive during the test period and had significant weight loss; therefore, they may also have exhibited metal loss. Phillips (1977) reported that the incidence of significant relationships between tissue weight and metal concentration varied with season and between metals. Because mussels are cold-blooded animals, their metabolisms are greatly reduced during the winter months, and metabolism and gut contents would increase once water temperatures begin to rise. This phenomenon may have resulted in the lower metal concentrations we observed in Beaver Creek mussels, which were exposed from November to March while all other stations were exposed through June. This lack of feeding may have also resulted in the lower percent ash than we found in mussels from the other stations. Tessier et al. (1994) did not find that temperature affected the cadmium and mercury concentrations Southeastern Naturalist J.L. Clayton, S.A. Miller, and R. Menendez 2015 Vol. 14, Special Issue 7 272 of Elliptio complanata Lightfoot (Eastern Elliptio); however, individuals were allowed to clear their guts prior to analysis. Jones and Walker (1979) found little variation in weight-normalized metal concentrations due to season. Our study mussels may have experienced a threshold of exposure after which mortality increased substantially. Hemelraad et al. (1986) found that mortality was negligible for 2 Anodonta species exposed to 25 μg/L cadmium for 16 weeks (bioaccumulation was less than 30 μg/g) and then drastically increased between 16 and 22 weeks. The slightly lower metal concentrations found in mussels from station 4 could be due to the gravid condition of these mussels. Mussels at other stations were no longer gravid. Although gills tend to show the highest metal accumulation, glochidia-bearing has been shown to have a dilution effect on the concentrations of some metals (V.-Balogh and Mastala 1994). The sedimentation of limestone below the doser appeared to affect the bivalves. Mussels at this station displayed greater weight loss and higher percent ash and mortality. This finding is similar to results of Aldridge et al. (1987) who reported that mussels exposed intermittently to relatively high concentrations of suspended solids had reduced food-clearance rates. In contrast, Roper and Hickey (1995) found that after bivalves were subjected to silt for 3 weeks, body condition was not affected, though they speculated that when food levels are low, increases in silt loadings stimulate pumping and may increase the potential for contaminant uptake. However, although impacts were noted, the benefits of the neutralization process outweigh the negative impacts of sedimentation because the area rehabilitated is much greater than that impacted. In this case, only 1 km of mussel habitat is being impacted by sediments, and over 18 kilometers have been restored or improved. These limestone sediments are beneficial in sustaining water quality, and increased dissolution is evident during high flows. Conclusions Our in-situ mussel bioassay appears to be an effective tool in determining the success of limestone treatment of acid mine drainage. We observed no significant mortality in our control mussels for the first 7 months of the study, at which time mortality exceeded 10%, likely due to caging that prevented the mussels from burying into the substrate. We found that AMD was toxic to the Creeper, though this species of freshwater mussel was able to survive in Beaver Creek with no significant mortality for nearly two months in conditions with pH values ranging from 4.6 to 5.7 at temperatures from 0.8 to 8.6 ºC. The addition of CaCO3 via a limestone doser successfully neutralized the toxicity from this tributary in the mainstem of Blackwater River. Though it appears that the limestone sediments that accumulate within the short segment of stream may negatively impact mussels, the benefits outweigh the harm. Many kilometers of stream that were severely impacted by AMD for over 35 years now have the potential for restoration of biological diversity. There has been concern over Southeastern Naturalist J.L. Clayton, S.A. Miller, and R. Menendez 2015 Vol. 14, Special Issue 7 273 the possibility of metal contamination and metal precipitation within the area of AMD neutralization. The mussels at a station within this area did not exhibit any significantly higher mortality than mussels at the control station, nor did they exhibit any increase in metal bioaccumulation. Sediment toxicity may however, affect juvenile mussels which can feed from overlying and interstitial water as well as using a ciliated foot to sweep in collected particles (Yeager et al. 1994). Mussels are only one component of the biological diversity existing in the Blackwater River. The completion of ongoing assessments of fish and aquatic insects will provide a more thorough analysis of the effects of AMD and limestone treatments on the biodiversity of Blackwater River. Acknowledgments This project was conducted through funding provided by the West Virginia Department of Environmental Protection, Charleston, WV, and the US Geological Survey, Biological Resources Division, Leetown, WV. Literature Cited Adams, T.G., G.J. Atchison, and R.J. Vetter. 1981. The use of the Three-ridge Clam (Amblema perplicata) to monitor trace-metal contamination. Hydrobiologia 83:67–72. Aldridge, D.W., B.S. Payne, and A.C. Miller. 1987. The effects of intermittent exposure to suspended solids and turbulence on three species of freshwater mussels. Environmental Pollution 45:17–28. Anderson, R.V. 1977. Concentration of cadmium, copper, lead, and zinc in six species of freshwater clams. Bulletin of Environmental Contamination and Toxicology 18:492–496. Clesceri, L.S. (Chairman), A.E. Greenberg, R.R. Trussel, and M.A.H. Franson. (Eds.). 1989. Standard Methods for the Examination of Waste Water, 17th Edition. American Public Health Association, Washington, DC. Gerhardt, A. 1993. Review of impact of heavy metals on stream invertebrates, with special emphasis on acid conditions. Water, Air, and Soil Pollution 66:289–314. Hemelraad, J., D.A. Holwerda, K.J. Teerds, H.J. Herwig, and D.I. Zandee. 1986. Cadmium kinetics in freshwater clams. II. A comparative study of cadmium uptake and cellular distribution in the Unionidae Anodonta cygnea, Anodonta anatina, and Unio pictorum. Archives of Environmental Contamination and Toxicology 15:9–21. Hemelraad, J., H.A. Kleinveld, A.M. de Roos, D.A. Holwerda, and D.I. Zandee. 1987. Cadmium kinetics in freshwater clams. III. Effects on zinc on uptake and distribution of cadmium in Anodonta cygnea. Archives of Environmental Contamination and Toxicology 16:95–101. Hemelraad, J., H.J. Herwig, E.G. van Donselaar, D.A. Holwerda, and D.A. Zandee. 1988. Cadmium kinetics in freshwater clams. IV. Histochemical localization of cadmium in Anodonta cygnea and Anodonta anatina exposed to cadmium chloride. Archives of Environmental Contamination and Toxicology 17:333–343. Holland-Bartels, L.E. 1990. Physical factors and their influence on the mussel fauna of a main-channel border habitat of the upper Mississippi River. Journal of the North American Benthological Society 9:327–335. Southeastern Naturalist J.L. Clayton, S.A. Miller, and R. Menendez 2015 Vol. 14, Special Issue 7 274 Holwerda, D.A., and P.R. Vennhof. 1984. Aspects of anaerobic metabolism in Anodonta cygnea L. Comparative Biochemistry and Physiology 78B:707–711. Jones, W.G., and K.F. Walker. 1979. Accumulation of iron, manganese, zinc, and cadmium by the Australian freshwater mussel Velesunio ambiguus (Phillipi) and its potential as a biological monitor. Australian Journal Marine and Freshwater Research 30:741–51. Kat, P.W. 1982. Shell dissolution as a significant cause of mortality for Corbicula fluminea (Bivalvia: Corbiculidae) inhabiting acidic waters. Malacological Review 15:129–134. Keller, A.E., and S.H. Zam. 1991. The acute toxicity of selected metals to the freshwater mussel, Anodonta imbecilis. Environmental Toxicology and Chemistry 10:539–546. Kraak, M.H.S., M.C.T. Scholten, W.H.M. Peeters, and W.C. de Kock. 1991. Bomonitoring of heavy metals in the western European Rivers Rhine and Meuse using the freshwater mussel Dreissena polymorpha. Environmental Pollution 74:101–114. Latouche,Y.D., and M.C. Mix. 1982. The effects of depuration, size, and sex on tracemetal levels in bay mussels. Marine Pollution Bulletin 13:27–29. Malley, D.F., J.D. Huebner, and K. Donkersloot. 1988. Effects of ionic composition of blood and tissues of Anodonta grandis grandis (Bivalvia) of an addition of aluminum and acid to a lake. Archives of Environmental Contamination and Toxicology 17:479–491. Markich, S.J., and R.A. Jeffree. 1994. Absorption of divalent trace metals as analogues of calcium by Australian freshwater bivalves: An explanation of how water hardness reduces metal toxicity. Aquatic Toxicology 29:257–290. Phillips, D.J. 1977. The use of biological indicator organisms to monitor trace-metal pollution in marine and estuarine environments: A review. Environmental Pollution 13:281–317. Pynnonen, K. 1990. Physiological responses to severe acid stress in four species of freshwater clams (Unionidae). Archives of Environmental Contamination and Toxicology 19:471–478. Pynnonen, K. 1995. Changes in acid-base status, gases, and electrolytes in the hemolymph of freshwater unionids during continuous and intermittent exposure to acid water. Annuales Zoologici Fennici 32:355–363. Pynnonen, K., and J. Huebner, 1995. Effects of episodic low pH exposure on the valve movements of the freshwater bivalve Anodonta cygnea L. Water Research 29:2579–2582. Renzoni, A., and E. Bacci. 1976. Bodily distribution, accumulation, and excretion of mercury in a fresh-water mussel. Bulletin of Environmental Contamination and Toxicology 15:366–73. Roper, D., and C.W. Hickey. 1995. Effects of food and silt on filtration, respiration, and condition of freshwater mussel Hyidella menziesi (Unionacea: Hyriidae): Implications for bioaccumulation. Hydrobiologia 312:17–25. SAS Institute, Inc. 1995. JMP User’s Guide, version 3.1. SAS Institute, Cary, NC. Tessier, A., P.G.C. Campbell, J.C. Auclair, and M. Bisson. 1984. Relationships between the partitioning of trace metals in sediments and their accumulation in the tissues of the freshwater mollusc Elliptio complanata in a mining area. Canadian Journal of Fish and Aquatic Sciences 41:1463–1472. Tessier, L., G. Vaillancourt, and L. Pazdernik. 1994. Temperature effects on cadmium and mercury kinetics in freshwater molluscs under laboratory conditions. Archives of Environmental Contamination and Toxicolgy 26:179–184. Southeastern Naturalist J.L. Clayton, S.A. Miller, and R. Menendez 2015 Vol. 14, Special Issue 7 275 V-Balogh, K., and Z. Mastala. 1994. The influence of size and glochidiabearing upon the heavy metal accumulation in gills of Anodonta piscinalis (Nilss). Chemosphere 28:1539–1550. Winter, S. 1996. Cadmium-uptake kinetics by freshwater mollusc soft body under hard and soft water conditions. Chemosphere 32:1937–1948. Yeager, M.M., D.S. Cherry, and R.J. Neves. 1994. Feeding and burrowing behaviors of juvenile Rainbow Mussels, Villosa iris (Bivalvia:Unionidae). Journal of the North American Benthological Society 13:217–222. Zurbuch, P.E., R. Menendez, J.L Clayton, and S.A. Miller. 1997. Restoration of two West Virginia rivers from effects of acid mine drainage. Pp. 19–28, In Proceedings, 19th Annual Conference, National Association of Abandoned Mine Lands Programs, Davis, West Virginia, 17–20 August 1997.