The mean control survival of mussels was ≥98% on test day 4 and ranged from 88 to 100% at the end of the 28-d exposures in all mussel tests. The control survival was 100% at the end of the 7-d exposures in all cladoceran tests, and the mean number of young per female in controls was ≥15 over the 7-d exposures. The control survival was 100% at the end of the Cu reference toxicant tests for both species. All treatments, therefore, met the test acceptability criteria, i.e., acute toxicity test ≥90% control survival 23, 26, and chronic toxicity test ≥80% control survival of mussels 23 and ≥90% control survival of cladocerans 27. In addition, more than 60% of the control cladocerans produced three broods over a 7-d test duration with the brood average ≥15 young per female in controls, which met the test acceptability requirements 27.
In the Cu reference toxicant tests, the 4-d Cu EC50 was 25 µg Cu/L (21–30; 95% confidence interval [CI]) for the mussel and the 2-d EC50 was 30 µg Cu/L (24–38; 95% CI) for the cladoceran in the reconstituted ASTM hard water. The mussel EC50 was within the range of the historical EC50s for the same mussel species (V. iris, 17–33 µg Cu/L 3), and the cladoceran EC50 was slightly greater than the historical EC50 for the cladoceran (C. dubia; 15 µg Cu/L 3) in ASTM hard water, indicating that the organisms tested in the present study were as sensitive to Cu as those tested in the previous studies.
Acute toxicity of Cu was evaluated based on the observations of mussel survival on test day 4 and cladoceran survival on day 2. The EC50s for the mussel ranged from 15 to 72 µg Cu/L and increased by a factor of approximately 5 across the DOC concentrations from 0.5 to 10 mg C/L, whereas the EC50s for the cladoceran ranged more widely from 25 to 267 µg Cu/L and increased by a factor of approximately 11 across the DOC concentrations from 0.5 to 10 mg C/L (Fig. 1). Acute EC50s for both species significantly increased with increasing DOC concentrations (p < 0.05; Fig. 1). These results were consistent with previous studies conducted with juvenile mussels (L. siliquoidea6) and cladocerans (C. dubia10; C. cf dubia15). Furthermore, the effect concentrations for mussels were consistently less than the effect concentrations for cladocerans across the DOC concentrations (Fig. 1), indicating that the influence of DOC on the Cu toxicity was less to the mussel than the cladoceran, and the mussel was more sensitive to acute toxicity of Cu than was the cladoceran.
Figure 1. Acute Cu EC50s and 95% CIs (error bars) for rainbow mussel (Villosa iris; 4-d exposures) and cladoceran (Ceriodaphnia dubia; 2-d exposures) at different concentrations of dissolved organic carbon (DOC). The open circle represents the EC50 for the cladoceran tested in 24-h aged exposure solutions.
Download figure to PowerPoint
The EC50s for mussel survival on test days 4, 14, 21, and 28 (Fig. 2A) or the EC50s for cladoceran survival on days 2 and 7 (Fig. 2B) were similar for each DOC-Cu treatment (with overlapping 95% CIs). Most mussels were easily identified as either alive (with foot movement) or dead (with empty shell) at the end of the 28-d exposures. However, a limited number of mussels (14%; n = 104 across all treatments) were recorded as having no foot movement and with tissue observed in shell (primarily at intermediate Cu concentrations in the 2.5 and 10 mg/L DOC treatments) at the end of the 28-d exposures. Only 25% of these mussels initially recorded as no foot movement on test day 28 were observed to exhibit foot movement after being held in control water for an additional 24 h. Therefore, the EC50s for mussel survival did not change or slightly increased (up to 29%) after the mussels were held in control water for 24 h (Fig. 2A). This was in contrast to the previous recovery study with two-month-old mussels (L. siliquoidea) after acute 4-d Cu exposures, where Wang et al. 3, 32 reported that EC50s increased by more than threefold after the mussels were held for 24 h in control water. Two-month-old mussels might be able physiologically to avoid exposure to Cu by closing their valves with limited foot movement 3, 32. It was likely that the three-month-old mussels in the present study exhibited the same avoidance behavior during the 4-d exposure. However, the similar EC50s between test days 4 and 28, and together with no change or only slight increase of the EC50s before and after the subsequent 24-h observation period in control water, indicate that the affected mussels with no foot movement in acute exposures probably died by the end of 28-d exposures. Additional studies are under way to determine how rapidly the tissue in dead mussels decomposes and results in empty shells.
Figure 2. Cu EC50s and 95% CIs (error bars) for survival of rainbow mussel (Villosa iris) over 4-, 14-, 21-, and 28-d exposures and the EC50s after the subsequent 24-h observation period in control water following the 28-d exposures in three dissolved organic carbon (DOC) treatments (A). Cu EC50s and 95% CIs (error bars) for survival of cladoceran (Ceriodaphnia dubia) over 2- and 7-d exposures in five dissolved organic carbon (DOC) treatments (B). The symbol (>) above a bar indicates that the EC50 was greater than the highest test concentration.
Download figure to PowerPoint
Chronic effect concentrations for mussels were determined based on the subsequent 24-h observation period in control water after the 28-d exposures (Table 2), and chronic effect concentrations for cladocerans were determined based on observations at the end of the 7-d exposures (Table 3). Significant reductions occurred in the survival and growth of the mussel or in the survival and reproduction of the cladoceran. Effects concentrations for mussel growth (length and weight) or biomass were generally less than the comparable effects concentrations for survival in each DOC-Cu treatment (Table 2). Similarly, effects concentrations for cladoceran reproduction also were generally less than the effect concentrations for survival (Table 3). The results indicate that the sublethal endpoints were more sensitive than the lethal endpoint. Because most EC20s were about equal to or less than the chronic value calculated as the geometric mean of the NOEC and LOEC (Tables 2, 3) and because the EC20 is not directly dependent on the dilution series of Cu solutions 19, the EC20s were used below to represent a low level of effects to evaluate the influence of DOC on chronic toxicity and to calculate acute-chronic ratios.
Table 2. Survival, shell length, dry weight, and biomass of rainbow mussel (Villosa iris) and measured concentrations of dissolved copper over a 28–d copper exposure period at different concentrations of dissolved organic carbon (DOC). Values are means with standard deviation in parentheses (n = 4 for each endpoint and 5 for copper concentration). An asterisk (*) indicates a significant reduction in an endpoint relative to control (p< 0.05). Chronic value (geometric mean [Geomean] of no-observed-effect concentration [NOEC] and the lowest-observed-effect concentration [LOEC]), 10 and 20% effect concentrations (EC10 and EC20) with 95% confidence interval (CI) are presented for each endpointa
|0.5 mg/L DOC treatment||2.5 mg/L DOC treatment||10 mg/L DOC treatment|
|Cu (µg/L)||Survival (%)||Length (mm)||Weight (mg)||Biomass (mg)||Cu (µg/L)||Survival (%)||Length (mm)||Weight (mg)||Biomass (mg)||Cu (µg/L)||Survival (%)||Length (mm)||Weight (mg)||Biomass (mg)|
|0.3||100 (0)||2.04 (0.14)||0.83 (0.18)||8.31 (1.79)||1.7||100 (0)||2.65 (0.17)||1.63 (0.28)||16.24 (2.82)||8.9||88 (19)||2.41 (0.05)||1.24 (0.12)||10.70(1.52)|
|2.6||78 (22)||2.35 (0.21)||1.23 (0.30)||9.58 (4.05)||6.2||83 (29)||2.64 (0.14)||1.61 (0.22)||12.84 (3.72)||30||88 (13)||2.32 (0.18)||1.16 (0.21)||9.94 (0.36)|
|4.9||100 (0)||2.18 (0.11)||0.87 (0.07)||8.65 (0.70)||11||95 (10)||2.36 (0.26)||1.26 (0.45)||11.98 (4.61)||51||78 (15)||2.07 (0.12)*||0.82 (0.10)*||6.40 (1.57)*|
|11||90 (8.2)||1.99 (0.21)||0.65 (0.19)||5.87 (1.77)||23||78 (9.6)*||2.01 (0.13)*||0.74 (0.12)*||5.78 (1.50)*||97||30 (14)*||1.85 (0.32)*||0.62 (0.22)*||2.06 (1.26)*|
|23||0*||–b||–||0*||46||23 (13)*||1.87 (0.32)*||0.68 (0.19)*||1.55 (0.82)*||212||0*||–||–||0*|
|NOEC||11||11||11||11|| ||11||11||11||11|| ||51||30||30||30|
|LOEC||23||>11||>11||23|| ||23||23||23||23|| ||97||51||51||51|
|Geomean||16||>11||>11||16|| ||16||16||16||16|| ||70||39||39||39|
|EC10 (CI)||12 (10–13)||11 (2.8–20)||6.2 (0.9–42)||9.2 (1.5–55)|| ||21 (18–25)||11 (4.6–25)||3.8 (0.6–25)||5.9 (2.6–13)|| ||52 (42–64)||42 (5.4–323)||21 (5.3–81)||29 (22–39)|
|EC20 (CI)||13 (11–14)||>11||8.3 (2.9–24)||10 (4.4–24)|| ||25 (22–29)||21 (10–46)||7.5 (1.9–30)||8.8 (4.7–16)|| ||61 (52–71)||83 (10–665)||35 (14–89)||38 (30–47)|
Table 3. Survival and reproduction (number of young/female) of cladoceran (Ceriodaphnia dubia) and measured dissolved copper concentration over a 7-d copper exposure period at different concentrations of dissolved organic carbon (DOC). Values are means (n = 10 for each endpoint with standard deviation in parentheses and n = 2 for copper concentration). An asterisk (*) indicates a significant reduction in an endpoint relative to control (p < 0.05). Chronic value (geometric mean [Geomean] of no-observed-effect concentration [NOEC] and the lowest-observed-effect concentration [LOEC]) and 10% and 20% effect concentrations (EC10 and EC20) with 95% confidence interval (CI) are presented for each endpoint
|0.5 mg/L DOC treatment||2.5 mg/L DOC treatment||5 mg/L DOC treatment||5 mg/L DOC (aged) treatment||10 mg/L DOC treatment|
|Cu (µg/L)||Survival (%)||No. of young||Cu (µg/L)||Survival (%)||No. of young||Cu (µg/L)||Survival (%)||No. of young||Cu (µg/L)||Survival (%)||No. of young||Cu (µg/L)||Survival (%)||No. of young|
|0.38||100||15 (3.4)||1.6||100||25 (4.0)||3.4||100||23 (3.5)||4.2||100||31 (3.0)||12||100||29 (4.1)|
|1.8||100||14 (6.2)||5.8||100||24 (6.0)||15||100||26 (3.7)||15||100||25 (6.1)*||29||100||20 (4.2)*|
|3.5||100||16 (2.1)||10||100||22 (5.9)||24||100||22 (4.4)||25||100||21 (4.8)*||54||100||20 (5.6)*|
|8.0||90||15 (2.9)||19||80||18 (8.4)*||50||90||16 (6.4)*||46||100||23 (4.3)*||98||100||16 (3.6)*|
|16||80||4.1 (5.3)*||42||80||18 (4.4)*||101||70||7.4 (5.5)*||91||90||13 (6.3)*||202||90||5.9 (3.4)*|
|NOEC||16||8.0|| ||42||10|| ||101||24|| ||91||4.2|| ||202||12|
|LOEC||37||16|| ||98||19|| ||227||50|| ||188||15|| ||412||29|
|Geomean||24||11|| ||64||14|| ||151||35|| ||131||7.9|| ||288||19|
|EC10 (CI)||10 (9.1–12)||11 (3.3–35)|| ||20 (15–26)||34 (24–48)|| ||65 (53–78)||29 (18–45)|| ||91 (84–98)||46 (30–71)|| ||200 (187–214)||25 (13–48)|
|EC20 (CI)||13 (11–14)||12 (4.9–28)|| ||34 (28–41)||42 (32–54)|| ||83 (73–93)||40 (29–56)|| ||98 (7.0–189)||59 (44–79)|| ||215 (202–230)||41 (25–68)|
Similar to the influence of DOC on acute Cu toxicity observed in the acute exposures, chronic Cu EC20s for survival of both species significantly increased with increasing concentrations of DOC (p < 0.05; Fig. 3A). The EC20s for mussel survival increased by a factor of approximately 5 and the EC20s for cladoceran survival increased by a factor of approximately 17 across the DOC concentrations from 0.5 to 10 mg C/L. However, the EC20s for cladoceran reproduction increased only by a factor of about 4 between the DOC concentrations of 0.5 and 3.0 mg C/L, and tended not to change across the higher DOC concentrations from 3.0 to 10 mg C/L (Fig. 3B). The EC20s for mussel biomass did not increase between the DOC concentrations of 0.5 to 3.0 mg C/L, but increased by a factor of approximately 4 between the DOC concentrations of 3.0 and 10 mg C/L (Fig. 3B). Overall, as observed in the acute exposures, the chronic Cu toxicity was reduced less with increasing DOC concentrations in the mussel test than in the cladoceran test (Fig. 3). Moreover, the chronic Cu effect concentrations for the mussel were equal to or less than the chronic effect concentrations for the cladoceran at the same DOC concentrations (Fig. 3), indicating the mussel was equally or more sensitive to Cu in the chronic exposure compared to the cladoceran.
Figure 3. Chronic Cu 20% effect concentration (EC20s) for survival (A) and biomass or reproduction (B) of rainbow mussel (Villosa iris, 28-d exposures) and cladoceran (Ceriodaphnia dubia, 7-d exposures) at different concentrations of DOC. The open circle represents the EC20 for the cladoceran tested in 24-h aged exposure solutions.
Download figure to PowerPoint
By definition, chronic toxicity testing takes longer than acute toxicity testing. Additionally, specific methods are not available for conducting chronic toxicity tests with many species. Therefore, acute-chronic ratios (ACRs) are useful for estimating chronic effects from acute toxicity data. In the present study an ACR was calculated by dividing an acute EC50 for survival by a chronic EC20 for mussel biomass (combined effect of survival and growth) or a chronic EC20 for cladoceran reproduction (the most sensitive endpoint) across the various DOC treatments (Table 4). Because the acute EC50 and chronic EC20 were generated from the same exposures under similar test conditions (e.g., with feeding and under flow-through condition), the acute and chronic toxicity data were comparable. The ACR for the mussel tested in various DOC waters ranged from 1.50 to 3.64 with a mean of 2.18 (Table 4), which is less than the Cu ACR of 4.70 for the same mussel species (V. iris) reported in a previous Cu toxicity study in the ASTM hard water without the addition of supplemental DOC 4. However, the ACR in the previous study was calculated based on a mean acute value of EC50s obtained under different test conditions (e.g., differences in static-renewal or flow-through testing, with or without feeding, <5-d-old or two-month-old mussels) and a chronic value (geometric mean of the NOEC and LOEC for length). The ACR for the cladoceran ranged from 2.08 to 6.51 with a mean of 3.39 (Table 4), which was slightly greater than the overall ratio of 2.85 for the same species (C. dubia) reported in the U.S. EPA AWQC for Cu 19. The mean ACRs obtained in the present study (2.18 for mussels and 3.39 for cladocerans) were close to the final ACR of 3.22 in the AWQC for Cu 19.
Table 4. Acute-chronic ratio (ACR) for copper calculated from median effect concentration (EC50) for 4-d survival and 20% effect concentration (EC20) for 28-d biomass of rainbow mussel (Villosa iris), or from the EC50 for 2-d survival and EC20 for 7-d reproduction of cladoceran (Ceriodaphnia dubia) in copper toxicity tests at different concentrations of dissolved organic carbon (DOC)
|4-d EC50||28-d EC20||ACR||2-d EC50||7-d EC20||ACR|
|5 DOC (aged)||–||–||–||147||59||2.49|
|Geometric mean|| || ||2.18|| || ||3.39|
Acute 2-d EC50s for the cladoceran exposed to the 24-h aged Cu solutions in the 5 mg/L DOC treatment were similar to the EC50s for the cladoceran exposed to the non-aged Cu solutions (Fig. 1); however, chronic 7-d EC20s for the cladoceran in the aged solution were 18% greater for survival and 48% greater for reproduction than the EC20s in the non-aged solution (Fig. 3). These results indicate that the 24-h DOC-Cu equilibration period did not influence the acute toxicity, but might influence the chronic toxicity to the cladoceran. Lack of influence of the equilibration period on the acute toxicity observed in the present study was in contrast to the findings in a previous study with the same test species (C. dubia) and the same DOM source (the Suwannee River 10) where the 24-h EC50s increased consistently with increasing Cu-DOC reaction times (1, 2, 6, and 24 h) in all four tested DOC solutions (2.5, 4.5, 7.5, and 10 mg C/L DOC 10). However, the results of the present study show that the equilibration period influenced chronic toxicity, especially reproduction of cladocerans. Because the equilibration period may influence the chronic toxicity of Cu, it might be necessary to modify proportional diluter systems to increase time for the equilibration of Cu-DOC solutions during chronic testing. For example, Besser et al. 33 equilibrated exposure water by delivering lead-spiked exposure water to intermediate equilibration chambers before water was delivered to exposure chambers in a diluter system.
Increased growth or reproduction of both species in controls occurred in waters with addition of the DOM concentrate compared to the growth or reproduction in water without addition of the DOM concentrate (Tables 2, 3); mean growth of the mussel (length and weight) and mean reproduction of the cladoceran in controls in the 2.5 to 10 mg/L DOC treatments were significantly greater than the growth or reproduction in controls in the 0.5 mg/L DOC treatment (Tukey's test, p < 0.05). However, growth and reproduction among the 2.5 to 10 mg/L DOC treatments were not significantly different (Tukey's test, p > 0.05). It is not clear whether the increased growth or reproduction in waters with addition of the DOM concentrate was attributed to elevated concentrations of DOC or attributed to other additional materials from the DOM concentrate, such as essential trace metals or some nutritional benefit (perhaps bacteria). For example, the measured concentrations of Cu in controls of all treatments increased with additional amount of DOM concentrate into the test waters (Tables 2, 3). The concentrations of other chemicals including Al, B, Fe, K, and Na were also slightly elevated in waters with addition of the DOM concentrate (Supplemental Data, Table S4). In addition, DOM supports aquatic food webs, including bacteria that in turn support filter-feeding invertebrates such as mussels 34, 35. Therefore, it was likely that more bacteria (additional nutrition to the test organisms) were available in test waters for the 2.5 to 10 mg/L DOC treatments than for the 0.5 mg/L DOC treatment.
The use of length as a measurement endpoint for growth may be challenging to interpret in toxicity tests such as our mussel test because of the limited range of response. However, if the mussels reach maturity, relative differences in length may predict fecundity (Supplemental Data, Table S5). In that case, length is a sublethal endpoint that may be relevant to reproductive effects in mussels much in the way that growth predicts reproductive success in other animals (e.g., 36). For instance, a 10% reduction in length of mussels would predict an approximate 19 to 44% reduction in fecundity, based on length-fecundity regressions from field studies with different freshwater mussel species (37, 38; Supplemental Data, Table S5); that is, an EC10 for length, which can be measured in a 28-d test, could be thought of as being comparable to an EC21 to EC44 level of effect for fecundity. Fecundity would be a difficult endpoint to measure in laboratory toxicity tests with long-lived mussels. However, the estimate of effect concentration for length may be uncertain because nonlinear curve fitting techniques (e.g., logistic regression) function best for estimating EC values when the biological variable of interest is the highest under control conditions, and declines to zero with increasing exposure concentrations. Survival, biomass, and fecundity are examples of data types that tend to behave well for this type of ECp estimation. In contrast, for the length measurements presented in Table 2 the percent declines in length in the highest treatments, in which some mussels survived, was only 13 to 29% relative to controls or low treatments with the highest growths. The data could have been transformed and standardized to better fit the regression assumptions. However, data transformations affect the analytical outcomes, and when the raw data are already expressed on a scale that is of biological interest, data transformations introduce additional complications in interpretation. Thus, because the mussel length data are related to fecundity, transforming the length data to obtain better ECp estimates in terms of smaller confidence intervals could actually produce worse estimates in terms of biological relevance.
Predictions from the BLM, which were used to derive the U.S. EPA AWQC for Cu 19, explained over 90% of the variation in acute and chronic endpoints for mussels and in acute endpoints for cladocerans (i.e., had r2 coefficients of determination >0.9); however, only 46% of the variation in the chronic EC20s for cladoceran reproduction was explained (Fig. 4). With mussels, the BLM tended to underestimate the Cu toxicity at elevated concentrations of DOC and to overestimate the toxicity at low concentrations of DOC. However, with cladocerans little bias occurred in the predictions associated with DOC, except perhaps at lower DOC treatments (more details are available in Supplemental Data, Fig. S1). When all comparisons of empirical versus BLM-predicted effects were pooled, regardless of species or endpoint, the BLM predictions accounted for about 80% of the total variation in the empirical results (Fig. 4).
Figure 4. Relationship between measured (empirical) and biotic ligand model (BLM) predicted dissolved Cu 50% effect concentrations (EC50s) in acute exposures and 20% effect concentration (EC20) in chronic exposures with rainbow mussel (Villosa iris) and cladoceran (Ceriodaphnia dubia). The solid line indicates the 1:1 line of perfect agreement, and dashed lines indicate 1:2 and 2:1 lines, i.e., bounds for predicted values being within twofold of observed values. Coefficient of determination (r2) values from linear regression predictions of empirical effect values are also shown.
Download figure to PowerPoint
To compare the generality of the BLM predictions to chronic cladoceran reproductive effects and to compare the sensitivity of the cladoceran in the present study, the toxicity for an independent dataset was evaluated. Using the CA20 previously derived for the 7-d reproduction effects on the cladoceran (C. dubia), the toxicity of Cu to the cladoceran was predicted in six water samples collected from diverse surface waters across Canada and a dechlorinated Ottawa city tap water 22. For the six natural water samples, the BLM predictions derived from our data were in good agreement with the empirical EC25 values (more details are available in Supplemental Data, Fig. S2). For the tap water sample, the BLM overestimated reproductive impairment (Supplemental Data, Fig. S2). Thus, at least, for the additional studies summarized in Figure S2, the BLM predicted chronic toxicity of Cu to the cladoceran reasonably well. This is an important observation because the BLM used in the 2007 AWQC for Cu was developed as an acute BLM 18 and later extrapolated to establish chronic criterion 19 without any published validation based on chronic toxicity data. Because there has been debate whether BLM-based Cu criteria require separate acute and chronic BLMs 20, 22, 39, more comparisons, similar to those in the Supplemental Data, Figure S2, would be useful.
Sensitivity to Cu
The BLM-normalized EC50s for Cu in acute 4-d mussel tests ranged widely, from 2.1 to 12 µg Cu/L, among the four DOC treatments (0.5–10 mg C/L DOC) and the reference toxicant test (0.3 mg C/L DOC; Table 5). These BLM-normalized values were within the range of BLM-normalized acute 4-d EC50s (1.7–26 µg Cu/L) reported for another mussel species (L. siliquoeda) tested in various natural and reconstituted waters with DOC concentrations of 0.3 to 11 mg C/L 6. The geometric mean of the BLM-normalized acute 2-d Cu EC50s for the cladoceran in the four DOC treatments and the reference toxicant test (8.8 µg Cu/L, ranged from 6.7–13 µg Cu/L, n = 5; Table 5) was slightly higher than the geometric mean derived for this species for use in the U.S. EPA 2007 Cu criteria (5.9 µg Cu/L, range from 2.5–11 µg Cu/L, n = 24; 19).
Table 5. Biotic ligand model normalized acute median effect concentration (EC50) or chronic 20% effect concentration (EC20) for copper in reference toxicant test (Reference test), in 28-d toxicity test with rainbow mussel (Villosa iris), and in 7-d toxicity test with cladoceran (Ceriodaphnia dubia) at different concentrations of dissolved organic carbon (DOC). An asterisk (*) indicates the EC50 below the final acute value 4.7 µg Cu/L or the EC20 equal to or below the final chronic value 1.5 µg Cu/L in the U.S. Environmental Protection Agency ambient water quality criteria for copper 19
|4-d survival||28-d survival||28-d biomass|
| ||0.5 DOC||8.4||4.5||3.5|
| ||2.5 DOC||3.0*||2.3||0.8*|
| ||5 DOC||3.0*||–||–|
| ||10 DOC||2.1*||1.7||1.0*|
| || ||2-d survival||7-d survival||7-d reproduction|
| ||0.5 DOC||13||4.8||4.5|
| ||2.5 DOC||NDb||2.5||1.1*|
| ||5 DOC||7.3||3.3||1.5*|
| ||5 DOC (aged)||6.7||4.0||2.3|
| ||10 DOC||7.0||5.3||0.8*|
The BLM-normalized acute EC50 for both species in the reference toxicant tests was 12 µg Cu/L in the ASTM reconstituted hard water and was greater than the EC50s from all DOC treatments except the 0.5 mg/L DOC treatment with the cladoceran (Table 5). This was consistent with the findings in a previous study with another mussel species (L. siliquoidea6), where the BLM-normalized EC50s (5.0–26 µg Cu/L) in ASTM reconstituted waters with a very low DOC concentration (≈0.3 mg C/L) were generally greater than the normalized EC50s (1.7–7.7 µg Cu/L) in natural waters with DOC concentrations ranging from 0.8 to 11 mg C/L. In addition, the normalized acute EC50s and chronic EC20s for both species in the present study were substantially greater (up to 5.6-fold) at low DOC concentration in the 0.5 mg/L DOC treatment than at higher DOC concentrations in the 2.5, 5, and 10 mg/L DOC treatments (Table 5). These results provided additional evidence that the toxicity of Cu to mussels and presumably other organisms may be underestimated when using BLM-normalized EC50s from tests conducted with waters containing very low concentrations of DOC 6. Wang et al. 6 recommended that a small amount of a standardized source of natural organic matter might be added to the reconstituted water recipes in order to provide about 1 mg/L DOC in the final test water to increase comparability of data across studies.
In test waters with elevated concentrations of DOC (2.5–10 mg/L DOC treatments), the BLM-normalized acute EC50s for mussel survival were less than the final acute value for Cu and the chronic EC20s for mussel biomass were less than the final chronic value in the U.S. EPA AWQC for Cu (Table 5). In contrast, all acute EC50s for cladoceran survival were above the final acute value and the chronic EC20s for the cladoceran reproduction were equal to or slightly less than the final chronic value in waters with elevated concentrations of DOC (2.5–10 mg/L DOC treatments; Table 5). These results indicate that the tested mussel species had the same or greater sensitivity to Cu compared to the cladoceran and most other organisms used to derive the AWQC for Cu; that the acute AWQC for Cu might not be protective of the mussel tested; and that the chronic AWQC might not be protective of the mussel and the cladoceran tested.
The results of the present study, which was conducted over a wide range of DOC concentrations, were consistent with previous findings that mussels were generally more sensitive to Cu than other tested aquatic organisms 1–4, 6 and that Cu effect concentrations for mussels were often at or below the final acute or chronic values used to derive previous hardness-dependent AWQC for Cu 2–4 or the current BLM-based AWQC for Cu 6. Mussel toxicity data generated from these studies have not been included in the current AWQC for Cu 19. Wang et al. 6 calculated the genus mean acute values (GMAVs) on the basis of BLM-normalized EC50s for several mussel species from acute Cu toxicity tests which met the ASTM test acceptability requirements 23; five of six GMAVs for mussels were within the range of GMAVs for the 10 most sensitive genera used to derive the acute AWQC for Cu. This conclusion would not be changed by adding the additional acute mussel toxicity data from the present study. Incorporating mussel toxicity data from the previous studies and the present study into a revision of the AWQC would decrease the final acute value from 4.7 (n = 27) to 4.5 (n = 32) µg/L.
The cladoceran tested in the present study is a commonly tested species that is highly sensitive to Cu (the second most sensitive species to Cu in the acute AWQC for Cu 19). More than 20 acute values for this species have been included to derive the acute AWQC for Cu 19. Thus, it is not surprising that the final acute value in the AWQC for Cu was less than the acute values for the cladoceran observed in the present study. However, most chronic effect concentrations for cladoceran reproduction were equal to or less than the final chronic value (Table 5). The final chronic value for Cu is calculated based on the final acute value estimated from the BLM and a fixed overall ACR 19. The results of the cladoceran toxicity tests in the present study indicate that the chronic Cu AWQC based on the ACR may not adequately protect the cladoceran.
The cladoceran (C. dubia) is one commonly tested species to estimate the acute and chronic toxicity of effluents and receiving waters 24, 25. The results of the present study indicate that the mussel was more sensitive to acute toxicity of Cu and was equal to or more sensitive to chronic Cu toxicity compared to the cladoceran. Thus, although the cladoceran is among the most metal-sensitive test species, an effluent or receiving water test with cladocerans may not provide specific information regarding protection of mussels, especially for metals. An ongoing study with 7-d effluent toxicity testing is designed to evaluate the relative sensitivity of newly transformed mussels compared to commonly tested C. dubia or fathead minnow (Pimephales promelas; Ning Wang, unpublished data).