Arch. Environ. Contam. Toxicol. 48, 166 –173 (2005) DOI: 10.1007/s00244-003-3125-3

A R C H I V E S O F

Environmental Contamination a n d Toxicology © 2005 Springer ScienceⴙBusiness Media, Inc.

Acute Toxicity of Six Freshwater Mussel Species (Glochidia) to Six Chemicals: Implications for Daphnids and Utterbackia imbecillis as Surrogates for Protection of Freshwater Mussels (Unionidae) C. D. Milam,1 J. L. Farris,2 F. J. Dwyer,3 D. K. Hardesty4 1 2 3 4

EA Engineering, Science, and Technology, Inc., 15 Loveton Circle, Sparks MD 21152, USA Arkansas State University, Ecotoxicology Research Facility, State University, AR 72467, USA U.S. Fish and Wildlife Service, Columbia, MO 65203, USA U.S. Geological Survey Columbia Environmental Research Center, Columbia, MO 65201, USA

Received: 10 June 2003 /Accepted: 13 October 2003

Abstract. Acute (24-h) toxicity tests were used in this study to compare lethality responses in early life stages (glochidia) of six freshwater mussel species, Leptodea fragilis, U. imbecillis, Lampsilis cardium, Lampsilis siliquoidea, Megalonaias nervosa, and Ligumia subrostrata, and two standard test organisms, Ceriodaphnia dubia and Daphnia magna. Concentrations of carbaryl, copper, 4-nonylphenol, pentachlorophenol, permethrin, and 2,4-D were used in acute exposures to represent different chemical classes and modes of action. The relative sensitivities of species were evaluated by ranking their LC50 values for each chemical. We used these ranks to determine the extent to which U. imbecillis (one of the most commonly used unionids in toxicity tests) was representative of the tolerances of other mussels. We also calculated geometric mean LC50s for the families Unionidae and Daphnidae. Rankings of these data were used to assess the extent to which Daphnidae can be used as surrogates for freshwater mussels relative to chemical sensitivity. While no single chemical elicited consistently high or low toxicity estimates, carbaryl and 2,4-D were generally the least toxic to all species tested. No species was always the most sensitive, and Daphnidae were generally protective of Unionidae. Utterbackia imbecillis, while often proposed as a standard unionid mussel test species, did not always qualify as a sufficient surrogate (i.e., a substitute organism that often elicits similar sensitivity responses to the same contaminant exposure) for other species of mussels, since it was usually one of the more tolerant species in our rankings. U. imbecillis should be used as a surrogate species only with this caution on its relative insensitivity.

The Clean Water Act, Federal Insecticide Fungicide and Rodenticide Act (FIFRA), and Toxic Substances Control Act (TSCA) have historically used aquatic toxicity testing in sup-

Correspondence to: Cristin D. Milam; email: [email protected]

port of federal protection of aquatic resources. Surrogate organisms for acute and chronic testing (Ceriodaphnia dubia, Daphnia pulex, Daphnia magna, Pimephales promelas, Selenastrum capricornutum, Americamysis bahia, and Menidia beryllina) have typically been used for monitoring whole effluents from industrial and municipal discharges through federally mandated NPDES (National Pollutant Discharge Elimination System) permits. These same organism responses have at times been used in aquatic risk assessments as surrogates for the range of sensitivities that may occur in listed (endangered or threatened) aquatic species. While test responses of surrogate organisms have been thoroughly documented, there are inherent assumptions associated with using these species to provide protection of listed species. Comparable responses between standard test organisms and suitable surrogate species can validate the applicability of toxicity assessments for the protection of species listed as threatened or endangered by the U.S. Fish and Wildlife Service. Freshwater bivalves (Unionidae) comprise the largest group of invertebrates listed under the 1973 Endangered Species Act (ESA), which is intended to provide protection for those species listed in the federal register (7 CFR, Chap. VI, Part 650.22). Determination of threshold effects of organic compounds and metals to listed species provides federal agencies critical information to properly manage species covered under the ESA (U.S. EPA 1995). Although there have been published studies to determine threshold limits using early life stages of freshwater bivalves (Milam and Farris 1998, Jacobson et al. 1997; Warren 1996; Jacobson et al. 1993; Wade et al. 1993; Huebner and Pynnonen 1992; Keller and Zam 1991), more often the assays are restricted to single metal (more specifically as total recoverable) exposures using Utterbackia imbecillis, a lentic pond mussel. Toxicity endpoints from a single species may not offer protection for all listed species exposed to a broad range of contaminants. Comparisons between freshwater mussels and surrogate test organism responses to metal complexes are also available (McKinney and Wade 1996; Masnado et al. 1995). Comparisons using these early life stages and subsequent threshold responses to pesticides is less abundant in

Toxicity of Freshwater Mussels to Chemicals

the literature (Heinonen et al. 2001; Keller and Ruessler 1997; Johnson et al. 1993; Keller 1993). While reported acute responses (LC50) provide information on bivalve thresholds and measurable sensitivities to a suite of chemicals, there is a need to evaluate additional unionid species requiring different habitats. Since a majority of the federally threatened and endangered species inhabit lotic systems, inclusion of stream-dwelling unionids in developing sensitivity comparisons to contaminants should offer a wider range of protection for this group of unionids. The life history of most unionids includes critical steps between internal gametogenesis in the adult female to the release of viable glochidia (varying in number from thousands to millions) into the water column for attachment onto specific fish hosts. This post-release stage, in which the glochidia are exposed to the aquatic environment, can be the most critical (Neves 1997). High levels of mortality occur during this passage due to low incidence of fish host contact, although once attached to the gill, glochidia are relatively protected from in situ exposure to water column metal mixtures (Jacobson 1990). There is an inherent difficulty in defining mortality in any individual. In this instance, a life stage involving an obligate parasite requires that any interruption (i.e., immobilization or lethality due to exposure to a particular contaminant) of the successful attachment onto a host could be considered mortality (e.g., LC50). Exceptions to this exist with conglutinants that have evolved to persist in the water column and enhance their probability of attaching to a host. Newly released juveniles may also be particularly sensitive to water- and sedimentassociated contaminants due to their siphoning and burrowing activities once transformation occurs. Measured pesticide concentrations in receiving streams that are associated with industrial and agricultural activities have been reported as causing impairment to test organisms (Werner et al. 2000; Bailey et al. 2000; Schulz and Liess 1999) and nontarget organisms (Beauvais et al. 2000; Milam et al. 2000; Niemi et al. 1999) used in toxicity tests employing standardized methods. While pesticide exposure studies of freshwater mussels have typically been limited to adult and juvenile stages, fewer still include the range of pesticide compounds that are more often found at high concentrations in receiving streams. To understand the impact of organic contaminants in receiving streams and to provide protection for this declining group of sensitive invertebrates, it seems prudent to evaluate the effects of pesticides of highest risk to this early life stage. The objectives of this study included (1) determining relative sensitivities of six unionid species to six chemicals with differing modes of action, (2) evaluating the sensitivity of U. imbecillis as a standard test species for other mussels, (3) comparing sensitivities to surrogate test organisms (Ceriodaphnia dubia and Daphnia magna), and (4) providing additional data for determining risk of chemical exposure to listed species of freshwater mussels.

Materials and Methods Test Setup Acute toxicity tests with C. dubia and D. magna were conducted using methods from U.S. EPA’s (1993) acute guidance document. Ceri-

167

odaphnia dubia and D. magna were used as standard test organisms, and laboratory-grade synthetic water (160 –180 mg CaCO3/L) was used for dilution. Tests were incubated at 22 ⫾ 0.5°C for 24 h to replicate test conditions from previously reported toxicity tests using carbaryl (U.S. EPA 1995). Additional acute tests were conducted using glochidia from fragile papershell (Leptodea fragilis), paper pondshell (U. imbecillis), fatmucket (Lampsilis siliquoidea), plain pocketbook (L. cardium), pondmussel (Ligumia subrostrata), and washboard (Megalonaias nervosa). Water quality parameters (temperature, alkalinity, hardness, conductivity, pH, and dissolved oxygen) were measured before and after each test according to Standard Methods (APHA 1992).

Contaminants Selection Six chemicals were chosen based on their inclusion within the range of chemical classes, their exclusive modes of action, and their lack of historically published data for bivalve effects. Carbaryl (1-naphthylmethylcarbamate; Rhone-Poulenc Agricultural Co.) is generally used as an insecticide for commercial farming, home gardens, and household use and its action is cholinesterase inhibition. Its water solubility is 40 mg/L at 30°C (Kidd and James 1991). Permethrin [(3-phenoxybenzyl (1RS)-cis, trans-3-(2,2-dichlorovinyl)-2,2-dimethylcyclopropanecarboxylate); ICI Americas Inc.) is a synthetic pyrethroid used as a broad-spectrum insecticide. The water solubility of permethrin is 0.2 mg/L at 20°C (Kidd and James 1991). Pentachlorphenol (PCP; Aldrich Chemical) is in the chemical class of chlorinated hydrocarbons and is primarily used as a fungicide defoliant and general herbicide. PCP’s action includes the disruption of oxidative phosphorylation and is mainly restricted to nonagricultural uses and has a water solubility value of 80 mg/L at 20°C (Kidd and James 1991). 4-Nonylphenol (4-NP; Fluka Chemical) is a monoalkyl phenol and its mode of action includes endocrine disruption (i.e., they can block, prevent, or alter the binding onto estrogen receptors) (Nichols et al. 2001; Colburn et al. 1993). Its water solubility is 6.2 mg/L at a pH of 7 (40 CFR 796.1860). While most of the toxicity data have assessed vertebrate and invertebrate sublethal responses, 4-NP was selected because of its lack of acute toxicity data for more sensitive invertebrates. The herbicide 2,4-D (2,4-dichlorophenoxyacetic acid; T&J Farms Co.) is a phenoxy compound and is used mainly by commercial farmers as a systemic herbicide; its water solubility is 900 mg/L at 25°C (Weed Science Society of America 1994). Copper, commonly used as a molluscicide, interrupts osmoregulation and has been shown to be very toxic to larval and adult bivalves in both short- and long-term exposures (Jacobson et al. 1997; Belanger et al. 1990). Test concentrations were made from stock solutions sent by personnel from the USGS Environmental Contaminants and Research Center in Columbia, MO. Controls included synthetic water mixed with reagent-grade acetone (0.25 mL acetone/500 mL hard water) and without acetone to account for possible toxicity of carrier solvents used in stock solutions (excluding copper and 2,4-D) (U.S. EPA 1995). Concentrations for each of the six contaminants were made by pipetting known amounts of stock solutions into synthetic water for final target concentrations. All test concentrations were established using nominal values, however, stock concentrations were analyzed at Mississippi State Chemical Laboratory or ABC Laboratories using gas chromatography for chemicals (with the exception of 2,4-D) and atomic absorption spectrophotometry for copper (US EPA 1995). Overall, measured chemical stock solutions averaged 110% (n ⫽ 9); with a mean range of 63% (copper) to 160% (permethrin) of their nominal concentration (U.S. EPA 1999). As reported by Belanger et al. (1989) copper toxicity is inversely proportionate to water hardness, and our data have been recalculated for comparison to ambient WQC for copper at a hardness of 50 mg/L.

168

Glochidia Acute Exposures Adult unionid mussels were collected during July to September 1997 from the Ohio River basin and shipped to the Arkansas State University Ecotoxicology Research Facility for acclimation in dechlorinated municipal water. Mussels were held for 1 week in a 60-L flow-through tank system with a water temperature ranging from 18 to 20°C (Milam and Farris 1998). They were monitored daily for indications of stress (excess mucous production, valve gaping, and aborted glochidia) and fed 1 L/day of a trialgal suspension (Chlamydomonas reinhardtii, Selenastrum capricornutum, and Ankistrodesmus sp. at a concentration of 3.6 ⫻ 106 cells/mL) (U.S. EPA 1989). Adult female unionids were checked for gravidity and returned to the flow-through holding chamber until used in toxicity tests. Viability of glochidia was determined by removing approximately 100 glochidia from the marsupium with a glass pipette (Zale and Neves 1982). The glochidia were pipetted into a glass petri dish filled with synthetic water and placed under a dissecting scope for view. Glochidia were identified as mature if they were open and free from the embryonic case or fertilization membrane. Concentrated sodium chloride (1 g NaCl/mL) was added to mature glochidia for a snap-closure response: glochidia were assumed viable if ⬎90% responded to the salt solution. Viable glochidia were excised from at least two separate females by removing the inflated marsupia and placing them into a petri dish with moderately hard synthetic water. Each marsupium was cut along the outer edge and released glochidia were held in synthetic water. Glochidia were collected from more than one female and all were pooled prior to testing to eliminate individual bias. A total of 97 tests was completed for determining the range of LC50 values for surrogate organisms (C. dubia and D. magna) and larval unionids (five genera). Test setup included the use of six-well (15-mL volume/well) polystyrene tissue plates for toxicity exposures. Although adsorption of organic compounds to various plastics has been documented (Sharom and Solomon 1980), we compared both glass and polystyrene chambers and measured no significant difference in glochidia response following 24-h exposure to these chemicals. Selection of a 24-h endpoint was used in this study due to two factors: (1) the necessity to calculate LC50’s based on aqueous concentrations while reducing the risk of pesticide volatility or degradation associated with longer test durations and (2) reduced glochidia viability (⬍90% survival) in exposures beyond 24 h. The use of a 24-h endpoint was a conservative estimate with respect to observed laboratory control mortality (72 h) in L. cardium. Mortality in this study was recognized as the coupling of two responses: (1) individuals with open values following contaminant exposure and that were nonresponsive to saline addition and (2) individuals with closed valves due to avoidance response to the contaminant and therefore lacking the ability to complete the life stage (i.e., attach to fish gill or fin). It is not known how long glochidia can survive once released from the gravid female or whether attachment (potential) onto a fish gill/fin is possible once they have been released and settle to the bottom substrates. Realistic endpoints beyond the standard 24 h may be putting a group of sensitive organisms at greater risk given the current field of knowledge. Three replicates per concentration were utilized for statistical confidence, with a series of six concentrations for each contaminant. Five milliliters of test concentration solution was placed in each of three wells and approximately 200 isolated glochidia were placed in each well with a glass pipette. All tissue plates were covered with fitted plastic lids and placed into an incubator for the 24-h test duration. A light:dark cycle of 16:8 was followed and all tests were maintained at 22°C. Initial tests were checked and noted for mortality (closed shells) at 1, 8, 12, and 24 h; however, this did not offer any definitive data since NaCl could not be added to the wells during these early observations. Mortality confirmations were conducted following 24 h by identifying 50 random glochidia/well using a dissecting microscope and reporting the number of open and closed valves. Approximately 50 ␮L of concen-

C. D. Milam et al.

trated NaCl was added to the specific well and the ratio of open/closed was counted post NaCl addition. The sum total of closed glochidia prior to salt addition and the number remaining open after salt addition were equivalent to the total number affected by the contaminant concentration (Jacobson et al. 1997). LC50 estimates were calculated from the number affected of the 50 total glochidia.

Standardized Acute Exposures Acute and chronic tests were conducted using C. dubia and D. magna for relative comparisons of organism sensitivity and use in pesticide registration assays. All tests followed U.S. EPA standardized acute method requirements (U.S. EPA 1993) but were modified for temperature (22°C) and dilution water (hard synthetic, 160 –180 mg CaCO3/ L). Daphnid tests included four replicates with two controls (acetone and hard synthetic water) and incubated for 24 h.

Statistical Analysis Results from the toxicity bioassays were statistically analyzed using Toxcalc (version 5.0.25). LC50 determinations and related confidence intervals for acute tests were determined using Probit analysis (Hamilton et al. 1977). All data were tested using ␣ ⫽ 0.05 and were tested for normality using Shapiro–Wilk’s test and Steel’s many-one rank test to determine significant differences in survival.

Results Water quality parameters measured during acute test exposures indicated that test dilutions were similar in water quality conditions: conductivity ranged from 587 to 622 (⫾9) ␮S/cm, DO ranged from 7.6 to 8.1 (⫾0.2) mg/L, pH ranged from 8.2 to 8.3 (⫾0.1) s.u., alkalinity ranged from 107 to 114 (⫾3) mg CaCO3/L, and hardness ranged from 162 to 178 (⫾10) mg CaCO3/L. No significant mortality (⬍10%) was measured in control exposures throughout the testing period. Measured acute toxic effects in unionid species and exposed Daphnia were used to generate mean LC50 values and their corresponding confidence intervals for the combined group of organisms (Table 1). Estimated LC50s were generated from a minimum of two distinct tests (using three replicates per test), with overlapping confidence intervals for each chemical. The exception to these test conditions included responses from M. nervosa, L. siliquoidea, L. cardium, and L. fragilis, where LC50s were generated from single tests. Copper and PCP exposures provided the most consistent LC50s across species responses; however, no single chemical provided consistent LC50s for all test organisms. Copper exposures resulted in the lowest overall LC50s for all test organisms and subsequently proved to be the most toxic among all contaminants tested. However, 4-NP proved most toxic of all organic compounds tested, with bivalve LC50s ranging from 0.49 mg/L for L. siliquoidea to 1.19 mg/L for L. cardium. Calculated LC50s for U. imbecillis were generally greater than estimates for more exclusive lotic species (L. cardium, L. siliquoidea, L. fragilis, and M. nervosa) throughout the suite of contaminants. For example, 2,4-D, 4-NP, and carbaryl exposures to U. imbecillis resulted in mean LC50s of

Toxicity of Freshwater Mussels to Chemicals

169

Table 1. Mean LC50s (⫾2 SD) for freshwater mussels and standardized test organisms exposed to five chemicals and one metal LC50 value (⫾2 SD) Species Leptodea fragilis Lampsilis cardium L. siliquoidea Megalonaias nervosa Ligumia subrostrata Utterbackia imbecillis Ceriodaphnia dubia Daphnia magna

Carbaryl (mg/L)

PCP (mg/L)

4-NP (mg/L)

Permethrin (␮g/L)

2,4-D (mg/L)

Copper (mg/L)

9.1 (4.2) n⫽3 33.9 (12.6) n⫽2 31.1 (na) n⫽1 27.4 (na) n⫽1 43.1 (19.1)a n⫽2 40.2 (13.8) n⫽2 0.1 (0.0) n⫽2 1.9 (0.3) n⫽2

0.58 (0.03) n⫽3 1.33 (0.29) n⫽4 1.39 (0.05) n⫽2 0.89 (na) n⫽1 0.86 (0.19) n⫽2 0.81 (0.11) n⫽3 0.47 (0.002) n⫽2 0.68 (0.13) n⫽2

0.57 (0.22) n⫽3 1.19 (1.07) n⫽3 0.49 (0.22) n⫽2 0.56 (na) n⫽1 1.04 (0.11) n⫽2 0.77 (0.11) n⫽3 0.20 (0.01) n⫽2 0.21 (0.02) n⫽2

3515 (3169)a n⫽2 14.9 (na)a n⫽1 —

81.8 (na) n⫽1 402.3 (168.4) n⫽2 362.2 (168.0) n⫽2 247.9 (na) n⫽1 352.9 (72.5) n⫽2 436.5 (19.8) n⫽2 272.5 (64.2) n⫽3 415.7 (101.2) n⫽3

0.09 (0.02) n⫽2 0.21 (0.17) n⫽2 0.13 (na) n⫽1 0.18 (na) n⫽1 0.15 (0.03) n⫽2 0.52 (0.11) n⫽3 0.04 (0.00) n⫽2 0.12 (0.01) n⫽2

— 1740 (214)a n⫽2 1714 (202)a n⫽3 8.7 (3.5)a n⫽2 12.4 (5.8)a n⫽2

Note. na: not enough date points to fulfill the necessary requirements to calculate 95% confidence limits. (—) No tests conducted. Mean LC50 values exceed water solubility limits.

a

436.5, 0.7, and 40.2 mg/L, respectively. Exposures to L. siliquoidea resulted in lower calculated LC50s of 362.2, 0.49, and 31.1 mg/L. These trends of U. imbecillis tolerance generally apply to the range of contaminants in this study. Additionally, mean no– observed effect concentrations (NOEC) were calculated from the same organism responses used to calculate LC50s and are provided in Table 2. The only consistent ranking of sensitivity among lotic species tested was L. cardium being more tolerant than L. fragilis to the suite of six chemicals (Table 3). The average LC50 values from exposures of L. cardium to PCP and carbaryl were 1.33 and 33.9 mg/L, respectively. Responses to permethrin were relatively variable and resulted in no two tests having overlapping confidence intervals. All test exposures with copper to L. cardium resulted in an average LC50 value of 0.21 mg/L. In contrast, the most toxic pesticide exposures to L. fragilis were with PCP and 4-NP, with average LC50 values of 0.58 and 0.57 mg/L, respectively. Copper exposures resulted in the lowest thresholds among all six chemicals tested, with an average LC50 of 0.09 mg/L. When carbaryl exposures were examined solely on bivalve responses, measured L. fragilis LC50s were nearly 4.5 times lower (9.1 mg/L) than the U. imbecillis (40.2 mg/L) LC50 calculations. L. fragilis LC50 values (0.58 mg PCP/L) were the lowest among exposed bivalved for this contaminant. For the majority of the contaminants, LC50 values for C. dubia were lower than all other test organisms for five of the six chemicals evaluated (Table 1). Ceriodaphnia LC50s for carbaryl and PCP averaged 0.1 and 0.47 mg/L, respectively, which were about 150 times lower than the LC50 for carbaryl (L. fragilis ⫽ 9.1 mg/L) and only 1.2 times lower than the LC50 for PCP (L. fragilis ⫽ 0.58 mg/L). Daphnia magna comparisons to LC50s for bivalves indicated that for carbaryl, 4-NP, and permethrin, acute thresholds (1.9, 0.21, and 0.012 mg/L, respectively) were similar to the LC50s for these bivalves. Sensitivity rankings from calculated species mean acute val-

ues (SMAVs) were generated for each contaminant to provide a comparison of risk among all test organisms (U.S. EPA 2002). In general, surrogate test organisms ranked most sensitive for exposures to carbaryl, 4-NP, and permethrin. Exposures to 2,4-D, however, indicated that two unionids, L. fragilis and M. nervosa, were more sensitive to this compound than either C. dubia or D. magna. Among all exposures, U. imbecillis was considered relatively tolerant of organic compounds and copper concentrations. Although U. imbecillis was more sensitive to PCP than four other unionid species, it was not as sensitive as L. fragilis. The LC50 values generated from acute toxicity tests were used to estimate the family mean acute values (FMAVs) for Unionidae and Daphnidae (Table 4) (U.S. EPA 2002). Copper (whether or not modified for water hardness of 50 mg CaCO3/L) was the most toxic contaminant tested with Unionidae. However, among organic compounds, 4-NP (0.77 mg/L) was the most toxic and 2,4-D (313.9 mg/L) was the least toxic to bivalves. Daphnidae threshold values were generally lower than those of Unionidae for all contaminants except 2,4-D (344.1 mg/L). Utterbackia imbecillis SMAVs were subsequently compared to calculated FMAVs for Unionidae and resulted in threshold values that were generally higher than those of other bivalve species, with the exception of PCP (0.81 mg/L) and permethrin (1.7 mg/L).

Discussion Species Sensitivities There exists a relatively wide range of reported thresholds for copper to aquatic freshwater invertebrates and vertebrates. In this study, L. siliquoidea, M. nervosa, and L. subrostrata were more sensitive to copper concentrations (0.13, 0.18, and 0.15

170

C. D. Milam et al.

Table 2. Mean no– observed effect concentrations (NOECs) from acute toxicity tests conducted with unionid bivalves and standard test organisms Species Chemical

C. dubia

2,4-D (mg/L) 4-NP (mg/L) Carbaryl (mg/L) Copper PCP (mg/L) Permethrin (␮g/L) a b

a

166.2 0.1 0.05a 0.18a 0.3 8.6a

D. magna a

263.88 0.25a 2.15a 0.05a 0.5a ⬍5.4

L. cardium

L. siliquoidea

L. subrostrata

L. fragilis

M. nervosa

U. imbecillis

⬍191.7 0.2 9.3 0.085a 0.38 58.4a

a

a

⬍153.4 0.13a 3.5 0.05a 0.3a 64.7a

⬍191.7 ⬍0.18 ⬍6 0.06 0.32 —

226.45a 0.34a 3.6 0.145 0.46a 149.5a

191.7 0.24a ⬍16.7 ⬍0.06 0.66a —b

181.15 0.24 5.18a 0.06a 0.32a 471

Absolute values from calculated LC50s were used to determine mean NOECs for each test organism. (—) No acute tests conducted with these chemicals.

Table 3. Relative sensitivity of unioinids and daphnids from ranked LC50s: Species are ranked as most sensitive (1) to least sensitive (8)

1 2 3 4 5 6 7 8 a b

Carbaryl

PCP

4-Nonylphenol

Permethrin

2,4-D

Coppera

C. dubia D. magna L. fragilis M. nervosa L. siliquoidea L. cardium U. imbecillis L. subrostrata

C. dubia L. fragilis D. magna U. imbecillis L. subrostrata M. nervosa L. cardium L. siliquoidea

C. dubia D. magna L. siliquoidea M. nervosa L. fragilis U. imbecillis L. subrostrata L. cardium

C. dubia D. magna L. cardium U. imbecillis L. subrostrata L. fragilis —b —

L. fragilis M. nervosa C. dubia L. subrostrata L. siliquoidea L. cardium D. magna U. imbecillis

C. dubia L. fragilis D. magna L. siliquoidea L. subrostrata M. nervosa L. cardium U. imbecillis

Values modified for water hardness (50 mg CaCO3/L). (—) No tests conducted.

Table 4 Comparison of family mean acute values (FMAVs) and species mean acute values (SMAVs) for U. imbecillis Contaminant (mg/L)

Unionidae FMAVa

Daphnidae FMAVa

U. imbecillis SMAVb

Copper Copperc Permethrin Carbaryl PCP 4-Nonylphenol 2,4-D

0.21 0.08 1.8 30.8 0.98 0.77 313.9

0.08 0.03 0.01 1.0 0.58 0.21 344.1

0.52 0.16 1.7 40.2 0.81 0.77 436.5

a

FMAVs were calculated by averaging the LC50 values generated from all test species in their respective families. b SMAVs were calculated by averaging the LC50 values generated from single-species results. c Values modified for water hardness (50 mg/L CaCO3); CMC ⫽ e(1.1281 ln(hardness)1⫺3.828).

mg/L, respectively) than to all other contaminants. Current water quality criteria (freshwater) for the protection of aquatic life to copper (CMC, 13 ␮g Cu/L; CCC, 9 ␮g Cu/L) should provide reasonable protection for these tested species, and recent modifications (i.e., implementation of the biotic ligand model approach) to the copper criteria should account for more specific sensitivities of these aquatic organisms (U.S. EPA 1984). Effects of copper on aquatic organisms have been thoroughly examined in upper trophic levels, however, the objective of this study was to determine impacts to invertebrates and more specifically freshwater bivalves. Jacobson et al. (1997) reported the sensitivity of glochidia to copper using

four common mussel species, which generally inhabit lotic systems. The measured LC50 values in four species ranged from 26 to 347 ␮g Cu/L and associated water hardness ranged from 55 to 190 mg CaCO3/L (as hardness increased, toxicity decreased). The least sensitive species tested was Pyganodon grandis, while the most sensitive species tested was Lampsilis fasciola. The range of LC50s in the current study was similar to Jacobson’s results for copper (0.04 – 0.52 mg/L). Calculated LC50s were also above current acute and chronic WQC for copper, suggesting that the surrogate species (C. dubia) provided adequate protection. However, LC50s do not always represent a safe dose, and in certain cases, adequate protection may involve the use of NOECs and LOECs from test responses. No– observable effect concentrations were calculated in this study to provide information on the protection of early–life stage unionids afforded by current WQC. Moore et al. (1998) provides evidence that LC50s, alone, may not offer adequate protection and that the slope of an LC50 curve can greatly affect whether an organism is adequately protected using upper and lower threshold responses. The inclusion of NOEC data from this study can provide additional insight into the magnitude of toxicity that these chemicals elicit. At times a safety or application factor (i.e., a value multiplied by the criteria to provide additional protection for sensitive species) may be necessary to ensure adequate protection for unionids (AEP 1983; Milam and Farris 1996). There are relatively fewer published reports providing sensitivity responses of unionids with organic compounds. PCP and 2,4-D were the two chemicals in this study that elicited

Toxicity of Freshwater Mussels to Chemicals

bivalve responses that were generally more sensitive than surrogate test organisms (D. magna, C. dubia, and U. imbecillis). Ligumia subrostrata [(more typically a quiet-water species (Parmalee and Bogan 1998)] was more sensitive to PCP (0.86 mg/L) than to all other chemical exposures. However, calculated LC50s for all unionids were well above (e.g., approximately 50 times) the PCP acute criterion (CMC). Heinonen et al. (2001) reported that 28-day exposures of juvenile Pisidium amnicum (freshwater clam) to 100 and 300 ␮g PCP/L resulted in reductions in valve movement and siphoning activity. Keller and Ruessler (1997) reported a mean guidance value of 3.3 mg PCP/L from laboratory exposures to V. nebulosa, L. fasciola, and V. iris and suggested their use as suitable surrogate species. Earlier acute tests conducted by Keller (1993) indicated that juvenile U. imbecillis were more sensitive (LC50 value, 1.11 mg PCP/L) than adult unionid tests conducted in the 1997 studies, which may indicate the appropriateness of U. imbecillis as a surrogate organism. Although the calculated SMAV using U. imbecillis from this study was protective of other unionids for exposures to PCP, it would not be a suitable surrogate for the other unionid species used in this study because of its relatively high tolerance to most other organic compounds. Unionid exposures to 2,4-D have not been reported in the literature, perhaps due to more prevalent organic compounds that are typically associated with risk in aquatic ecosystems (e.g., malathion and atrazine). The data in this study, however, indicate that bivalves are acutely affected by the herbicide at concentrations ranging from 82 to 402 mg 2,4-D/L. Unionids inhabiting lotic systems (e.g., L. fragilis, M. nervosa, L. siliquoidea, and L. cardium) are generally more sensitive than surrogate organisms used in the registration of pesticides.

Surrogate Use Our study showed that the use of C. dubia as a surrogate test organism was generally protective in short-term exposures for the majority of these organic and inorganic contaminants (with the exception of exposures to 2,4-D). The conventional use of D. magna in the pesticide registration process, however, would not seem to offer adequate protection of unionid species for exposures to PCP, 2,4-D, and perhaps others not tested in this study. Daphnia magna has traditionally been utilized as a standard test organism for the registration of pesticides in the United States; however, toxicity data from this study indicate that this species may not be suitable for the protection of early life stages of bivalve species. Calculated SMAVs for D. magna were not protective for unionid exposures to 2,4-D, PCP, and copper, but calculated FAVs for Cladocera (Daphnidae) were protective of all unionids with the exception of 2,4-D. USFWS (1980) reported LC50s of vertebrates using pesticide mixtures of 2,4-D and 2,4,5-T and determined that they were toxic to bluegill, with calculated LC50s of 23 ␮g/L. While our study did not address pesticide mixtures or the use of vertebrates as test organisms, these reported effects on bluegill as low as 23 ␮g/L would indicate that 2,4-D is somewhat toxic to a relatively tolerant vertebrate. Since centrarchids are often represented as generalized hosts for many of the lampsiline species, the pro-

171

tection of fish hosts is critically linked to the adequate protection of unionids. Measured LC50s reported in the EPA/ORD pesticide registration database for 24-h laboratory exposures to D. magna include 0.56 ␮g permethrin/L, 438.75 ␮g PCP/L, 7.41 ␮g carbaryl/L, and 25 mg 2,4-D/L. While these data are within the range of the LC50 values from this study, consideration should be given to the level of protection afforded by daphnid responses for the relatively complex unionid life history. For instance, acute exposures to contaminants could inhibit glochidia attachment onto the fish gill (a requirement in the transformation and life history of most freshwater mussels) and subsequent transformation of the juvenile stage. A surrogate species with similar life history patterns would perhaps provide a better assessment of exposure and protection of organic compounds to the varying unionid lifestages (e.g., glochidium, juvenile, adult). Jacobson (1990) reported that host fish exposures to copper concentrations did not inhibit the attachment and transformation of juvenile bivalves. There are currently no reports for similar exposures to organic contaminants. Warren (1996) suggested the use of U. imbecillis as a surrogate bivalve with copper and cadmium. While our data support the current acute water quality criteria for copper as adequately protective of bivalves, the data do not support the recommendation of U. imbecillis as a standard test organism for bivalves, since it was the least sensitive of all tested species to copper and organics. Reported thresholds for early life stages of freshwater mussels exposed to various pesticides are generally lacking in the literature. However, Keller and Ruessler (1997) related impacts of malathion to glochidia, juvenile, and adult unionids. They reported mean LC50 values of 7 mg malathion/L for L. siliquoidea glochidia and 324 mg malathion/L for U. imbecillis glochidia. Their data suggest that the use of U. imbecillis glochidia as a standard test organism would not protect other, more sensitive species. In our study, U. imbecillis was less sensitive than the majority of other tested bivalves when exposed to carbaryl, 4-NP, 2,4-D, and copper. However, acute responses to PCP and permethrin by U. imbecillis were higher than those of L. fragilis and L. cardium. 4-Nonylphenol exposures were more toxic to L. siliquoidea (0.49 mg/L) and M. nervosa (0.56 mg/L) than to all other tested bivalve species.

Environmental Realism The relative comparisons of these laboratory pesticide/metal exposures to published expected environmental concentrations (EECs) indicate that there is low risk of acute mortality to glochidia of the tested unionid species. However, determination of hazard levels in this study indicate that concentrations of PCP and 2,4-D may have some impact upon bivalves evaluated in this study, and these data may support the use of a more sensitive indicator (C. dubia or unionid) of effect for some organic compounds. Comparisons of EECs to calculated LC50s in this study indicate that carbaryl [5.2 ␮g/L (Larson et al. 1999)], 4-NP [⬍5 ␮g/L (Environmental Canada 2000)], permethrin [0.02 ␮g/L (Larson et al. 1999)], and copper [13 ␮g/L CMC (U.S. EPA 1984)] have little impact on this particular bivalve life stage. Established water quality criteria or

172

reported EECs for PCP [19 ␮g/L CMC (U.S. EPA 1984)] and 2,4-D [15 ␮g/L (Larson et al. 1999)], however, may not provide adequate protection for all unionid species as noted by the sensitivities calculated in this study. Water solubility of permethrin (0.2 mg/L) suggests that its partition coefficient is relatively high. Calculated LC50 values for U. imbecillis, L. subrostrata, and L. fragilis reflected this relationship with water solubility limits and suggest that permethrin is most often adsorbed and unavailable to bivalves. While there are no current water quality criteria established for permethrin, environmental concentrations as high as 0.03 ␮g/L have been reported in Mississippi surface waters (http://www.usgs.gov). Based on these limited data, comparisons to tested organisms in this study suggest that the Unionidae FMAV for permethrin (1.7 mg/L) is well above reported concentrations of permethrin in the environment. Peterson et al. (2001) reported measured carbaryl concentrations in Oregon streams that ranged from 0.11 to 2.0 ␮g/L, with increased concentrations in brooks as high as 7.8 ␮g carbaryl/L following aerial application. In this study, unionid LC50s ranged from 9 to 43 mg carbaryl/L, suggesting that the risk of carbaryl exposures to freshwater mussels is relatively low. Johnson et al. (1993) conducted acute toxicity tests with juvenile U. imbecillis with atrazine, carbaryl, and cyhalothrin, with reported mean LC50s of 60, 24, and 1 mg/L, respectively. Glochidia comparisons of the same species in this study (mean LC50, 40.2 mg carbaryl/L) provide further evidence of the sensitivities for the earlier lifestage. In general, EECs are lower than measured LC50s and NOECs calculated in this study. The importance of developing sensitivity rankings for unionids, however, is not only to provide reference data that are currently unavailable, but also to raise awareness of the risk associated with possible short-term elevated concentrations from spills or unregulated discharges. Reliance on controlled conditions in standardized laboratory bioassays to represent the range of dynamic environmental conditions faced by freshwater mussels may especially limit the evaluation of pesticide risk. The use in this case of an unconventional test species (e.g., unionid) may provide a better understanding of the risk-based approach for aquatic populations other than commonly tested surrogate organisms and may help define ecological consequences related to pesticide application, which could otherwise be masked by traditional toxicity tests (Kapustka et al. 1996). As a result, the development of standardized freshwater bivalve toxicity tests is a first step in facilitating the evaluation of contaminant impacts to unionids. Test protocols for glochidia should consider the relatively short time span (72–96 h) of this particular lifestage when interpreting effective endpoints.

Acknowledgment. Support for this project was provided by Arkansas State University’s Ecotoxicology Research Facility in cooperation with USGS Environmental Contaminants and Research Center. The authors thank C. Ingersoll for his guidance throughout the project and critical review of the manuscript. J. Maul and two anonymous reviewers provided valuable comments for the completion of the manuscript. Additional thanks to H. Dunn, B. Sietman, D. Kelner, A. Christian, and D. Hubbs for their collection of unionids used for testing.

C. D. Milam et al.

References American Electric Power Service Corporation (1983) Reappraisal of the appropriateness of the iron water quality criterion in West Virginia: Technical Report. Appalachian Power Company, Roanoke, VA American Public Health Association (1992) Standard methods for the examination of water and wastewater, 18th ed. American Public Health Association, American Water Works Association, and Water Pollution Control Federation, Washington, DC. Bailey HC, James RK, Elphick R, Mulhall AM, Hunt P, Tedmanson L, Lovell A (2000) Whole effluent toxicity of sewage treatment plants in the Hawkesbury-Nepean watershed, New South Wales, Australia, to Ceriodaphnia dubia and Selenastrum capricornutum. Environ Toxicol Chem 19:72– 81. Beauvais SL, Jones SB, Brewer SK, Little EE (2000) Physiological measures of neurotoxicity of diazinon and malathion to larval rainbow trout (Oncorhynchus mykiss) and their correlation with behavioral measures. Environ Toxicol Chem 19:1875–1880 Belanger SE, Farris JL, Cherry DS (1989) Effects of diet, water hardness, and population source on acute and chronic copper toxicity to Ceriodaphnia dubia. Arch Environ Contam Toxicol 18:601– 611 Belanger SE, Farris JL, Cherry DS, Cairns J Jr (1990) Validation of Corbicula fluminea growth reductions induced by copper in artificial streams and river systems. Can J Fish Aquat Sci 47:904 –914 Colborn T, vom Saal FS, Soto AM (1993) Developmental effects of endocrine-disrupting chemicals in wildlife and human. Environ Health Perspect 101:378 –384 Environment Canada (2000) Priority substances list assessment report: Nonphynol and its ethoxylates. Draft Report. Health Canada, Montreal Hamilton MA, Russo RC, Thurston RV (1977) Trimmed SpearmanKarber method for estimating median lethal concentrations. Environ Sci Tech 11:714 –719 Heinonen J, Kukkonen VK, Holopainen IJ (2001) Temperature- and parasite-induced changes in toxicity and lethal body burdens of PCP in the freshwater clam Pisidium amnicum. Environ Toxicol Chem 20:2778 –2784 Huebner JD, Pynnonen KS (1992) Viability of glochidia of two species of Anodonta exposed to low pH and selected metals. Can J Zool 70:2348 –2355 Jacobson PJ (1990) Sensitivity of early lifestages of freshwater mussels (Bivalvia: Unionidae) to copper. MS thesis. Virginia Polytechnic Institute and State University, Blacksburg Jacobson PJ, Farris JL, Cherry DS, Neves RJ (1993) Juvenile freshwater mussel (Bivalvia: Unionidae) responses to acute toxicity testing with copper. Environ Toxicol Chem 12:879 – 883 Jacobson PJ, Neves RJ, Cherry DS, Farris JL (1997) Sensitivity of glochidial stages of freshwater mussels (Bivalvia: Unionidae) to copper. Environ Toxicol Chem 16:2384 –2392 Johnson IS, Keller AE, Zam SG (1993) A method for conducting acute toxicity tests with the early lifestages of freshwater mussels. Standard Technical Publication 1179. ASTM, Philadelphia, PA Kapustka LA, Williams BA, Fairbrother A (1996) Evaluating risk predictions at population and community levels in pesticide registration—Hypotheses to be tested. Environ Toxicol Chem 15: 427– 431 Keller AE (1993) Acute toxicity of several pesticides, organic compounds, and wastewater effluent to the freshwater mussel, Anodonta imbecillis, Ceriodaphnia dubia, and Pimephales promelas. Bull Environ Contam Toxicol 51:696 –702 Keller AE, Ruessler SD (1997) The toxicity of malathion to unionid mussels: Relationship to expected environmental concentrations. Environ Toxicol Chem 16:1028 –1033 Keller AE, Zam SG (1991) The acute toxicity of selected metals to the

Toxicity of Freshwater Mussels to Chemicals

freshwater mussel, Utterbackia (formally Anodonta) imbecillis. Environ Toxicol Chem 10:539 –546 Kidd H, James DR (eds) (1991) The agrochemicals handbook, 3rd ed. Royal Society of Chemistry Information Services, Cambridge, UK, pp 3–11 Larson SJ, Gilliom RJ, Capel PD (1999) Pesticides in streams of the United States: Initial results from the National Water-Quality Assessment Program U.S. Geological Survey Water-Resources Investigations Report 98 – 4222, Sacramento, CA Masnado RG, Geis SW, Sonzogni WC (1995) Comparative acute toxicity of a synthetic mine effluent to Ceriodaphnia dubia, larval fathead minnow, and the freshwater mussel Anodonta imbecillis. Environ Toxicol Chem 14:1913–1920 McKinney AD, Wade DC (1996) Comparative response of Ceriodaphnia dubia and juvenile Anodonta imbecillis to pulp and paper mill effluents discharged to the Tennessee River and its tributaries. Environ Toxicol Chem 15:514 –517 Milam CD, Farris JL (1998) Risk identification associated with irondominated mine discharges and their effect upon freshwater bivalves. Environ Toxicol Chem 17:1611–1619 Milam CD, Farris JL, Wilhide JD (2000) Evaluating mosquito control pesticides for effect on target and non-target organisms. Arch Environ Contam Toxicol 39(3):324 –328 Moore MT, Huggett DB, Gillespie WB Jr, Rodgers JH, Cooper CM (1998) Comparative toxicity of chlordane, chlorpyrifos, and aldicarb to four aquatic testing organisms. Arch Environ Contam Toxicol 34:152–157 Neves RJ (1997) A state-of-the-unionids address. In Cummings KS, Buchanan AC, Mayer CA, Naimo TJ (eds) Conservation and management of freshwater mussels II: Initiatives for the future. Proceedings of a UMRCC symposium, October 1995, St. Louis, MO. Upper Mississippi River Conservation Committee, Rock Island, IL Nichols KM, Snyder EM, Snyder SA, Pierens SL, Miles-Richardson SR, Giesy JP (2001) Effects of nonylphenol ethoxylate exposure on reproductive output and bioindicators of environmental estrogen exposure in fathead minnows, Pimephales promelas. Environ Toxicol Chem 20:510 –522 Niemi GJ, Hershey AE, Shannon L, Hanowski JM, Lima A, Axler RP, Regal RR (1999) Ecological effects of mosquito control on zooplankton, insects, and birds. Environ Toxicol Chem 18:549 –559 Parmalee PW, Bogan AE (1998) The freshwater mussels of tennessee. University of Tennessee Press, Knoxville Peterson JL, Jepson PC, Jenkins JJ (2001) A test system to evaluate the

173

susceptibility of Oregon, USA, native stream invertebrates to triclopyr and carbaryl. Environ Toxicol Chem 20:2205–2214 Schulz R, Liess M (1999) Validity and ecological relevance of an active in situ bioassay using Gammarus pulex and Limnephilus lunatus. Environ Toxicol Chem 18:2243–2250 Sharom MS, Solomon KR (1980) Adsorption and desorption of permethrin and other pesticides on glass and plastic materials used in bioassay procedures. Can J Fish Aquat Sci 38:199 –204 U.S. Environmental Protection Agency (1984) Ambient water quality criteria for copper. Office of Water EPA 440/5-84-031 U.S. Environmental Protection Agency (1989) Short-term methods for estimating the chronic toxicity of effluents and receiving waters to freshwater organisms, 2nd ed. EPA/600/4-89/001 U.S. Environmental Protection Agency (1993) Methods for measuring the acute toxicity of effluents and receiving waters to freshwater and marine organisms, 4th ed. EPA/600/4-90/027F U.S. Environmental Protection Agency (1995) Use of surrogate species in assessing contaminant risk to endangered and threatened fishes. EPA/600/R-96/029 U.S. Environmental Protection Agency (1999) Assessing contaminant sensitivity of endangered and threatened species: Toxicant classes. EPA/600/R-99/098 U.S. Environmental Protection Agency (2002) National recommended water quality criteria: 2002. EPA-822-R-02-047 U.S. Fish and Wildlife Service (1980) Handbook of acute toxicity of chemicals to fish and aquatic invertebrates. Resource publication 137. USFWS, Washington, DC Wade DC, Hudson RG, McKinney AD (1993) Comparative response of Ceriodaphnia dubia and juvenile Anodonta imbecillis to selected complex industrial whole effluents. Conservation and Management of Freshwater Mussels. Proceedings of a UMRCC Symposium. KS Cummings, AC Buchanan, and LM Koch, 109-12. Upper Mississippi River Conservation Committee, Rock Island, IL Warren LW (1996) The use of juvenile mussels, Utterbackia imbecillis Say (Bivalvia: Unionidae), as a standardized toxicity testing organism. PhD dissertation, Clemson University, Pendleton, SC Werner I, Deanovic LA, Connor V, Vlaming V, Bailey HC, Hinton DE (2000) Insecticide-caused toxicity to Ceriodaphnia dubia (Cladocera) in the Sacramento-San Joaquin River Delta, California, USA. Environ Toxicol Chem 19:215–227 Zale AV, Neves RJ (1982) Fish hosts of four species of lampsiline mussel (Mollusca: Unionidae) in Big Moccasin Creek, Virginia. Can J Zool 60:2535–2542