Toxicity of Aqueous Alkaline Solutions to New Zealand Mudsnails, Asian Clams, and Quagga Mussels

Infestations of aquatic invasive species often result in significant economic and environmental costs, and their presence increases the risks of further spread (Pimentel et al. 2000; Lovell et al. 2006; Connelly et al. 2007; Marbuah et al. 2014; Nakano and Strayer 2014). Nonindigenous mollusk infestations provide significant challenges to managers of water infrastructure in aquaculture facilities, power plants, water-treatment plants, irrigation canals, and industrial production systems (Rosa et al. 2011; Nakano and Strayer 2014; Sousa et al. 2014). Various biocides have been proposed and tested as tools to control or prevent the spread of mollusks (e.g., Claudi and Mackie 1994; Edwards et al. 2000; Mackie and Claudi 2009; McMillin and Trumbo 2009; Guardiola et al. 2012; Layhee et al. 2014; Oplinger and Wagner 2015). Bivalves and prosobranch snails can be especially difficult mollusks to control since they can avoid exposure to many chemicals when closed inside their shells (Checa and Jiménez-Jiménez 1998; Pereira et al. 2016; Comeau et al. 2017).

To determine appropriate control measures, managers are advised to consider the location and nature of an infestation, as a suite of compounds and approaches is possible (reviewed in GLMRIS 2012). Copper compounds can be effective (Watton and Hawkes 1984; Dwyer et al. 2003), but serious latent and long-term effects on the environment can result from their use. Elevated copper concentrations have been demonstrated to have serious physiological and biochemical effects on many freshwater organisms and their food sources (De Boeck et al. 2003; Ahlf et al. 2009; McIntyre et al. 2015; Chowdhury et al. 2016; Ghosh et al. 2016; Sharley et al. 2016). Other chemical tools include quaternary and polyquaternary compounds and microencapsulated ammonium chlorides (Britton and Dingman 2011; Calazans et al. 2013; Stout et al. 2016). Halogen-containing compounds such as sodium hypochlorite and bromine are used in disinfection of drinking water and cooling systems, but they are corrosive, and several complex derivatives such as polychlorinated dibenzodioxins, furans, and other disinfection byproducts can have harmful effects (Richardson et al. 2007; Vallejo et al. 2015). Moreover, their effectiveness is limited by organic matter.

Elevating the pH of water with NaOH in freshwater ballast systems has been reported as highly effective to kill a suite of plankton and bacterial or viral species (Moffitt et al. 2015a; Starliper et al 2015). The alkaline solutions in ballast tanks used can be neutralized with CO₂, yielding environmentally harmless compounds that have been shown to meet effluent guidelines (Elskus et al. 2015; Moffitt et al. 2015a). The pH of freshwater can also be elevated easily with addition of Ca(OH)₂ (Moffitt et al. 2015b), but with testing Ca(OH)₂ has proved difficult because of the lowered solubility of this compound. In addition, Ca(OH)₂ precipitate can clog pipes and interfere with delivery systems (Montresor et al. 2013). However, a recent study by Comeau et al. (2017) demonstrated an effective use of precipitates of Ca(OH)₂ in marine aquaculture of blue mussel Mytilus edulis where applications of Ca(OH)₂ help eliminate fouling of culture systems by the tunicate Styela clava.

Elevating freshwater pH with NaOH can be particularly appealing because the chemical is highly soluble, is available in liquid and powder forms, and is used widely in industrial and agricultural settings. Examples include the pulp and paper, food, and biodiesel industries as well as water treatments (Montresor et al. 2013; Torres-Lozada et al. 2015; Chaiyapong and Chavalparit 2016). Very little NaOH is needed to raise the pH of freshwater from neutral to a strong alkaline solution (U.S. Department of Health and Human Services 2002; ClearTech Industries Inc. 2017). To understand the effectiveness of elevated pH as a molluscicide, we conducted laboratory toxicity trials testing aqueous solutions of NaOH, pH 12, on three important aquatic nuisance mollusk species: the New Zealand mudsnail Potamopyrgus antipodarum, the Asian clam Corbicula fluminea, and the quagga mussel Dreissena rostriformis bugensis. Our objectives were to determine the time of death and model the mortality response of each species.

We obtained the mudsnails used in tests from springs at Hagerman National Fish Hatchery, Hagerman, Idaho (42°45′52″N, 114°51′54″W) or from raceways at Willow Beach National Fish Hatchery, Willow Beach, Arizona (35°52′26″N, 114°39′49″W). We collected Asian clams from the Bruneau River, Idaho (42°47′37″N; 115°43′12″W). We obtained quagga mussels from hatchery raceways at Willow Beach National Fish Hatchery. We collected the mudsnails and quagga mussels from Willow Beach the day before trials, placed them into buckets, and acclimated them in filtered (65-μm mesh nylon or stainless-steel sieves) Colorado River water overnight. Field crews collected test organisms from Hagerman and the Bruneau River, packed them in wet towels inside plastic bags, and shipped them in a cooler with ice packs overnight to the University of Idaho, Moscow. We removed the mudsnails and Asian clams shipped to the University of Idaho from bags, washed them with freshwater to remove sediments, and placed them into holding and culture containers in a secured temperature-controlled isolation room. These organisms acclimated to 15°C for at least 2 d before testing.

At the University laboratory, we retained the mudsnails in 2-L containers (< 500 individuals/container) with 1 L of dechlorinated well water. Every other day we exchanged approximately half of this water. Algae and pondweed (Potamogeton spp.) shipped with the mudsnails served as food. We collected dispersal-stage mudsnails (neonates released from mature adults) from the bottom of cultures of adult mudsnails with disposable pipettes. We maintained the Asian clams used for trials in 5-L containers (< 100 per container) with supplemented air and exchanged approximately half of the water daily. We introduced pulverized brine shrimp and algae pellets daily as food source. We maintained all organisms cultured at the University lab in a natural photoperiod light cycle for the latitude of the source population (∼ 42°N), and at 2-wk intervals, we brought freshly collected organisms into the laboratory to replace cultures. At Willow Beach National Fish Hatchery, we did not culture quagga mussels and mudsnails in tests but obtained them from the hatchery water source the day before trials and held them overnight in the lab to acclimate them to test containers.

We prepared aqueous NaOH solutions for tests in 2-L increments in flasks or beakers using American Chemical Society reagent-grade NaOH (Fisher Brand lots 060432 and 994932) in dechlorinated well water (University of Idaho) or raceway water (Willow Beach) to achieve the target pH of 12.0. We transferred the test solutions into a labeled, capped, 20-L polyethylene carboy to limit exposure to atmospheric CO₂ and stored it at room temperature. We monitored air and water temperatures during each experiment with Hobo temperature loggers (Onset Computer Corporation, Bourne, MA) at hour intervals. We measured the pH, conductivity, temperature, and dissolved oxygen of test and control water before each experiment and at the time we removed test specimens (test exposure intervals) with a YSI 556 multiprobe (YSI Incorporated, Yellow Springs, OH) and a Hach HQ30D LDO probe (Hach, Loveland, CO). To further characterize the water from both laboratory locations, we quantified the total quantities of a suite of metals: barium, cadmium, calcium, chromium, cobalt, copper, iron, magnesium, manganese, molybdenum, nickel, phosphorus, potassium, sodium, vanadium, and zinc in representative water samples from both locations with inductively coupled plasma protocols at the University of Idaho Analytical Sciences Laboratory.

We tested adults of the three species, dispersal-stage mudsnails (neonates), and juvenile-sized quagga mussels in a series of static bath exposures to solutions of pH 12 for a total of 12 trials. Adult mudsnails were ≥ 2 mm in length (adults) and neonates < 1 mm. Asian clam adults were ≥ 16 mm long. Quagga mussel adults were ≥ 24 mm long; juveniles averaged 11.4 mm and ranged from 6 to 15 mm. The duration of exposure of test organisms ranged from a few seconds to several days, depending on the life-history stage and species. We selected sampling intervals for removal from test solutions or controls for each experiment using log time intervals to facilitate response modeling. To prevent exposure to atmospheric CO₂ that could neutralize the NaOH and lower the pH, we filled and quickly covered all test containers to prevent exposure to the air. During test solution replacement, we exposed organisms to air for < 5 s.

We tested survival of adult mudsnails using 50-mm-diameter stainless steel tea balls (500-μm mesh) as cages for 10 adult mudsnails. We counted the mudsnails into each tea ball, and we then placed seven to eight tea balls into each of several 1.8-L containers (∼ 12 cm × 19 cm × 8 cm) with control aged water. To begin a test, we lifted each tea ball from the control water, drained each one, and placed each into one of several prefilled, 1.8-L containers with test solution or control water. We conducted tests of neonate mudsnails in 150-mL glass beakers that contained at least 10 individuals per replicate beaker that had been removed from stock containers with disposable pipettes. A neonate trial began by draining the water from replicate test beakers with neonates remaining on the bottom, and quickly introducing the test solution or control water into the beaker.

We conducted trials with Asian clams by placing 5–10 adult-sized organisms into each of several 1.8-L plastic containers (four containers for tests, one to two containers for controls for each time interval of exposure). We conducted trials with adult quagga mussels by placing four to five mussels in replicate 1.8-L containers, or placing juveniles into 150-mL beakers (two to four test containers, one to three control containers for each time interval of exposure). To begin the trial with quagga mussels or Asian clams, we drained the acclimation water from each container and replaced with the test solution or control water.

To determine the survival or mortality at preselected test intervals, two to five replicate containers or cages were removed from a test system. We then rinsed the organisms serially three times with aged water to remove any test chemical, and then placed them into containers with aged well water to recover. We drained neonate mudsnails from the test beakers through a 65-μm mesh nylon or a stainless-steel sieve, and we washed the test organisms with 120 mL of dechlorinated well water, and then back-washed them into a recovery beaker containing aged well water and retained them for a 24-h recovery. To serve as controls, at each interval we also removed at least one nonexposed container or cage from the test system and handled it similarly. We replaced the water in recovery containers of quagga mussels, Asian clams, and adult mudsnails with fresh water after 24 h. Because of the high filtration and metabolic rates of adult Asian clams, we placed supplemental aeration in recovery containers with clams.

We made a determination of live or dead individuals in each recovery container with the aid of a dissecting microscope after 24 h. We removed and recorded any obviously dead organisms from the containers. We made final assessment of survival of adult mudsnails, quagga mussels, and Asian clams after an additional 24 h, for a total of 48 h of recovery in freshwater. We considered assessment of neonates final at 24 h. To determine live status, we probed organisms not showing active movement gently with a dissecting needle to elicit a response. We scored individuals responsive to probing or observed actively crawling or moving in the recovery container as live. For quagga mussels and Asian clams, we considered organisms that were responsive to probes or remained closed and difficult to pry open live. We scored as dead organisms that were open and unresponsive, or showed signs of decomposition. We considered a trial acceptable when control mortality did not exceed 10% for adults or juvenile organisms exposed for 96 h or less, or 20% with adults in trials lasting longer than 96 h. For dispersal-stage neonates, the control mortality cutoff was 30%, following toxicity testing guidance documents of the U.S. Environmental Protection Agency (USEPA 2009).

We graphed mortality responses with scatter plots to inspect data. The goal was to model the mortality response with probit regression models to predict the lethal time to 99% mortality (LT99). We conducted analyses using the unadjusted mortality observed in test systems. We did not subtract control mortality for estimates of relative mortality. We fit models of response (fraction live or dead) to time of exposure with a normal distribution, and we developed probit models for each species to include covariates of experimental trial and test replicate. We considered trials highly accurate if the model of time of exposure was significant, the effect of test replicate was not significant, and 95% confidence intervals (CIs) could be estimated. We evaluated the responses of multiple trials with trial as a covariate, and when we detected no differences between trials we combined data and collapsed the number of covariates to determine the most parsimonious model. We analyzed all statistical models using SAS 9.3 (SAS Institute, Cary, NC).

Freshwater solutions of NaOH, pH 12, were effective in causing mortality in all three species, with responses ranging from minutes to days. The speed of the mortality response varied with mollusk species and life stage. Adult Asian clams were the most resistant to toxic exposure of NaOH. We observed some mortality of Asian clams after 1 d, but most survived for days. The probit model fit with data from both trials with Asian clams provided a significant fit (χ² = 68.0, P < 0.001), and estimated LT99 was 208.9 h; 95% CI = 190–239 h (Figure 1; Table S1).

Quagga mussels were the most sensitive of the adult species to elevated pH. The probit model fit to both trials was significant (χ² = 31.3; P < 0.001), and the estimated LT99 for the combined trials of adult quagga mussels was 30.2 h; 95% CI = 25–41 h (Figure 2; Table S2). Juvenile-sized quagga mussels were more sensitive, and the probit model of mortality over time of exposure was significant (χ² = 36.5; P < 0.001), but test replicates were a significant variable (χ² = 17.6, P < 0.001), likely due to the small number of individuals (four to five) in each test beaker. The estimated with test replicate as a covariate was 12.6 h; 95% CI = 10.4–16.5 h. When the model was collapsed without considering replicates as a covariate, a fit with time of exposure was also significant (χ² = 23.8; P < 0.001), and the estimated LT99 was 12.0 h; 95% CI = 9.6–17.8 h (Figure 2).

We conducted two trials of adult mudsnails at Willow Beach and one at the University of Idaho. We combined all trials to model time to mortality with a significant fit (χ² = 142.7; P < 0.001) and estimated LT99 of 38.6 h; 95% CI = 35.1–43.3 h (Figure 2; Table S3). We killed neonates most rapidly in solutions of pH 12. In the four trials of mudsnail neonates, we fit a significant model (χ² = 41.5; P < 0.001) to time of exposure and the estimated LT99 was 4.4 min; 95% CI of 3.6–6.0 min (Figure 2; Table S3). The mortality was rapid; the model fit (Figure 2) illustrates the uncertainty of response predicting mortality less than 50%.

For all 12 trials at two locations, the pH of test solutions averaged 12.0, with a range of 11.7 to 12.1 (Table 1). Test system room temperatures ranged from 17.9 to 23°C, and control and test containers were well oxygenated (Tables 1 and 2). All test subjects had closed opercula or shells at the time of administration of test compound, and we observed no quagga mussels or Asian clams syphoning in test vessels. We did not observe mudsnails moving during exposure to test solutions. We observed movement or siphoning in the test subjects held in controls.

The tests of adult, juvenile, and dispersal-stage mollusks presented here should be considered the high end of durations needed to achieve mortality given that test organisms had shells or opercula closed on exposure. If substances had been introduced into canals, pipes, or containers where organisms were siphoning or moving, the test animals may have ingested test solution before responding, and likely decreasing the time to death. In addition, all test systems were well oxygenated, precluding any interactions of toxicity with hypoxia. Rosa et al. (2015) found that reduced oxygen levels significantly reduced the time to mortality for Asian clams tested in solutions of niclosamide, ammonium nitrate, and dimethoate. We modeled and expressed the outcome of tests using estimates of LT99, since the desired outcome of exposure to elevated pH was complete mortality. Our estimates for LT50 of adults were 104 h, 13.4 h, and 11.4 h for adult Asian clams, mudsnails, and quagga mussels, respectively.

As expected, the juvenile and neonate life stages tested were less tolerant of elevated pH exposure over larger adults. In smaller organisms, thin shells and larger surface area-to-body-size ratios reduce the protection from biocides. Harrison et al. (1984) found that sensitivity to copper exposure decreased with successional life stages of Asian clams, with juveniles being less sensitive to copper than veligers. Edwards et al. (2000) found that post-D veligers (veligers that have formed shells) of zebra mussels were less sensitive then preshell veligers. Nielson et al. (2012) reported significantly faster mortality of neonate mudsnails over adults when exposed to elevated partial pressures of CO₂.

Temperature as well as organismal condition will affect the mortality response times. We selected test organisms that were apparently healthy, but did not examine the effect of temperature on the toxicity response. It is well established that metabolic rates of mollusks are lowered by cooler temperatures (Polhill and Dimock 1996; Johnson and McMahon 1998; Braby and Somero 2006). Moffitt et al. (2015b) observed significant effects of temperature on the mortality of quagga veligers in solutions of pH 12, finding that veligers held at 16°C took longer to die than those tested at 20°C.

The studies we present here confirm the exceptional resilience of Asian clams to NaOH as a biocide, as our estimates of LT99 were five to six times longer than estimates of time to mortality for mudsnails and quagga mussels. Other studies support this difference in scope of the lethal response in Asian clams to other reagents. Bidwell et al. (1995) found that Asian clams exposed to dodecylguanidine hydrochloride and n-alkyl dimethylbenzyl ammonium chloride at 20 to 25°C were significantly more resistant to mortality than were zebra mussels Dreissena polymorpha. They found little change in mortality from dodecylguanidine hydrochloride and n-alkyl dimethylbenzyl ammonium chloride, even when clam valves were forced to remain open with pieces of toothpick. Their results suggested that Asian clams have an advanced ability to tolerate stress associated with contaminant exposure. The survival of Asian clams in toxic environments may be related to their ability to reduce their metabolic rate by 90% during valve closure and stress, and remain aerobic for up to 9 h (Ortmann and Grieshaber 2003). Matthews and McMahon (1995, 1999) evaluated the effects of temperature on extreme hypoxia (O₂ < 3% of saturation) in both zebra mussels and Asian clams, and found that the Asian clams were two to seven times more tolerant of hypoxia than zebra mussels exposed to same conditions at the same temperatures. Asian clams held in extreme hypoxia at 25°C survived a mean of 11.8 d, 35.1 d at 15°C, and without mortality for 84 d at 5°C (Matthews and McMahon 1999). Valve closure and an ability to reduce metabolism may explain why Asian clams appear to remain unaffected by elevated pH for such a long period of time.

Our experimental approach to testing toxicity of elevated pH was not typical, as we examined time to mortality in a specific pH, rather than testing an exact quantity of the test reagent. We conducted our experiments with mudsnails in two water sources with different specific conductivities (0.30–0.35 mS/cm at the University of Idaho vs. 0.88–0.98 mS/cm at Willow Beach; Table 2). These differences in conductivity were largely driven by differences in the calcium, sodium, and zinc concentrations (Table 3). Test waters from Willow Beach contained 76, 98, and 0.72 mg/L for calcium, sodium, and zinc, respectively, compared with 23, 24, and 0.03 mg/L, respectively, for the University of Idaho water. Even though the baseline conductivity and metals profile for the two water sources differed, we found similar outcomes in tests of mudsnails in pH-adjusted test solutions. However, the quantity of NaOH we added to the source water to achieve pH 12 target levels averaged 676 mg/L and 592 mg/L for Willow Beach and the University of Idaho, respectively.

Macrobiofouling of infrastructures by invasive mollusks often results in extensive economic costs (Aldridge et al. 2006; Mackie and Claudi 2009; Rosa et al. 2011, 2015; Pereira et al. 2016) and proposed mitigative treatments include consideration of costs and environmental feasibility. Because of the simplicity of elevating water-system pH with NaOH, its potential use in confined pipes and in channels and tanks infested with nuisance mollusks may hold promise. Tests of ballast disinfection using additions of NaOH (Elskus et al. 2015; Moffitt 2015a; Starliper et al. 2015) suggested that not only was this chemical effective, but that after neutralization with CO₂, the effluents would meet regulatory guidelines for safe discharge. Sodium hydroxide was registered as a herbicide many decades ago for control of tree roots in sewer systems, as a fungicide and algicide for use on water-well casings, and as a disinfectant in various indoor settings (USEPA 1992). The uses were re-examined in the reregistration process and the findings of a partial review were documented in the Federal Register (2009). Moreover, alkaline treatments with NaOH are widespread in many industrial and agricultural applications (e.g., Occidental Chemical Corporation 2013; Torres-Lozada et al. 2015; Chaiyapong and Chavalparit 2016). However, any proposed registered or approved use of NaOH will need to include details for containment and adequate neutralization steps. An experimental treatment with NaOH was shown effective as a management tool for control of Asian clams in combination with impermeable benthic barriers such as those applied in Lake Tahoe to control Asian clams (Wittmann et al. 2012). In Lake Pend Oreille, Idaho, applications of NaOH placed beneath impermeable barriers retained in place for 7 mo showed efficacy in controlling a localized infestation of Asian clams (Moffitt and Wilhelm, 2017; See Reference S4). Additional studies of elevated pH will help provide more toxicity data on other target and nontarget species, and studies should include field trials to characterize the effectiveness and safety to infrastructure and applicators.

Please note: The Journal of Fish and Wildlife Management is not responsible for the content or functionality of any supplemental material. Queries should be directed to the corresponding author for the article.

Table S1. A spreadsheet of data in probit model of survival of Asian clams Corbicula fluminea in solutions of pH 12. Models and predicted 95% confidence intervals are plotted in Figure 1. We conducted trials at the University of Idaho in 2013.

Found at DOI: http://dx.doi.org/10.3996/022017-JFWM-013.S1 (13 KB XLSX).

Table S2. A spreadsheet of data used in probit models of survival in solutions of pH 12 for two life stages of quagga mussels Dreissena rostriformis bugensis. Models and predicted 95% confidence intervals are plotted in Figure 2. We conducted trials at Willow Beach National Fish Hatchery, Arizona in 2012.

Found at DOI: http://dx.doi.org/10.3996/022017-JFWM-013.S2 (10 KB XLSX).

Table S3. A spreadsheet of data used in probit models of survival in solutions of pH 12 for two life stages of New Zealand mudsnails Potamopyrgus antipodium. Models and predicted 95% confidence intervals are plotted in Figure 2. We conducted trials at the University of Idaho, and also at Willow Beach National Fish Hatchery, Arizona in 2012 and 2013.

Found at DOI: http://dx.doi.org/10.3996/022017-JFWM-013.S3 (13 KB XLSX).

Reference S1. ClearTech Industries Inc. 2017. Safety Data Sheet sodium hydroxide. ClearTech Industries. Saskatoon, Saskatchewan, Canada.

Found at http://dx.doi.org/10.3996/022017-JFWM-013.S4; also available at http://www.cleartech.ca/ckfinder/userfiles/files/Sodium%20Hydroxide%20Solution(11).pdf (274 KB PDF).

Reference S2. [GLMRIS] Great Lake and Mississippi River Interbasin Study. 2012. Inventory of available controls for aquatic nuisance species of concern. GLMRIS Team, Chicago, IL.

Found at http://dx.doi.org/10.3996/022017-JFWM-013.S5; also available at http://glmris.anl.gov/glmris-report/ (16155 KB PDF).

Reference S3. Matthews MA, McMahon RF. 1995. Survival of zebra mussels (Dreissena polymorpha) and Asian clams (Corbicula fluminea) under extreme hypoxia. U.S. Army Corps of Engineers, Waterways Experiment Station, Vicksburg, MS. Technical Report EL-95-3.

Found at http://dx.doi.org/10.3996/022017-JFWM-013.S6; also available at http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.562.2627&rep=rep1&type=pdf (1156 KB PDF).

Reference S4. Moffitt CM, Wilhelm FM. 2017. Risks and effectiveness of benthic barriers as tools to eradicate infestations of Asian clams (Corbicula fluminea) in the Ellisport Bay Treatment Areas of lake Pend Oreille, Idaho. Final Report SSP Project, WRO #162, Idaho FWIN #2722. University of Idaho, Boise, ID. Found at http://dx.doi.org/10.3996/022017-JFWM-013.S7 (1577 KB PDF).

Reference S5. Occidental Chemical Corporation. 2013. Product information sodium hydroxide. Occidental Chemical Corporation, Dallas, TX. Found at http://dx.doi.org/10.3996/022017-JFWM-013.S8; also available at http://www.oxy.com/OurBusinesses/Chemicals/ResponsibleCare/Documents/Sodium%20Hydroxide.pdf (125 KB PDF).

Reference S6. U.S. Department of Health and Human Services. 2002. Sodium hydroxide. CAS #1310-73-2. Atlanta, Georgia: Division of Toxicology ToxFAQs. Agency for Toxic Substances and Disease Registry.

Found at http://dx.doi.org/10.3996/022017-JFWM-013.S9; also available at https://www.atsdr.cdc.gov/toxfaqs/tfacts178.pdf (140 KB PDF).

Reference S7. U.S. Environmental Protection Agency. 1992. Reregistration eligibility document (RED). sodium hydroxide. Washington, D.C.: Office of Prevention, Pesticides and Toxic Substances. EPA-738-F-92-008.

Found at http://dx.doi.org/10.3996/022017-JFWM-013.S10; also available at https://archive.epa.gov/pesticides/reregistration/web/pdf/4065fact.pdf (3468 KB PDF).

Reference S8. U.S. Environmental Protection Agency. 2009. Ecological effects test guidelines. Bivalve acute toxicity test (embryo-larval). Washington, D.C.: OSPP 850.1055. enfiro712-C-16006. Posted October 2016.

Found at http://dx.doi.org/10.3996/022017-JFWM-013.S11; also available at https://www.regulations.gov/document?D=EPA-HQ-OPPT-2009-0154-0036 (380 KB PDF).

M. Olson, T. Frew, S. Karpowicz, G. Cappellii, A. Flaten, J. Saccomanno, and S. Peterson at Willow Beach National Fish Hatchery provided essential facilities, support, and assistance. T. Woolf, N. Zurfluh, and B. Muffled, Idaho Department of Agriculture, collected and provided Asian clams; J. Trimpey and staff at Hagerman National Fish Hatchery collected and shipped mudsnails. University of Idaho students and interns J. Shearer, M. Torres, L. Hughes, C. Withers-Haley, T. Venable, J. Davis, J. C. Ortiz-Perez, K. Wilcox, M. F. Babrowicz, B. Snow, T. McClure, M. Gartiez, L. C. de Gliniewicz, and Z. Penny provided laboratory and field assistance. Support for interns was provided by the National Science Foundation-funded Center for Invasive Species and Small Populations Research Experiences for Undergraduates, and the Helping Orient Indian Students and Teachers program. B. J. Watten, Leetown Science Center, U.S. Geological Survey, provided expertise and advice for trials. We are grateful for the insights of anonymous reviewers of earlier drafts of this manuscript. Funding was provided by the U.S. Geological Survey and the U.S. Fish and Wildlife Service.

Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the U.S. Government.