Toxicity of cobalt to freshwater indicator species as a function of water hardness

Although generally rare in natural aquatic systems, cobalt concentrations can be elevated through ore and coal mining operations and discharges of certain textile dyes. The U.S. Environmental Protection Agency does not have an ambient water quality criterion for cobalt due to a lack of toxicological data. The present study determined whether freshwater acute and chronic cobalt toxicity were dependent on dilution water hardness. We specifically examined the relative toxicity of cobalt at high water hardness (≥ 200 mg/l as CaCO 3) because cobalt often occurs in carbonate ores. Acute and chronic toxicity tests using both Ceriodaphnia dubia and Pimephales promelas were performed using four different synthetic waters: 50,200. 400, and 800 mg/l hardness as CaCO 3. Diluent hardness appeared to have no effect on measured cobalt concentrations. Dissolved cobalt in water samples during testing was similar to the corresponding total recoverable cobalt in this study suggesting that most of the cobalt was in a soluble and presumably bioavailable form. LC 50 values for P. promelas could not be calculated owing to their low sensitivity to high cobalt concentrations (≥ 5 mg/l) over a 48 h period. Regression analysis of P. promelas acute NOEC values and test water hardness indicated an R 2 = 0.94 suggesting a direct relationship between water hardness and acute toxicity to cobalt for this species. C. dubia acute tests indicated water hardness effects after 48 h of exposure and non-linear dose responses in hard water (≥ 200 mg/l). Regression analysis of C. dubia 24 h LC 50 values and test water hardness indicated an R 2 = 0.91. For both cases, a simple power function gave the best regression fit. C. dubia appeared to be more sensitive than P. promelas at hardness ≥ 200 mg/l but the reverse was true in soft water ( ≤ 50 mg/l). The 7 d P. promelas chronic NOEC increased from 1232 μg/l cobalt (50 mg/l hardness) to ≥ 3833 μg/ l (800 mg/l hardness). Effects were on fish survival and not growth. Regression analysis yielded an R 2 = 0.97 based on the fish survival NOECs and the natural log of hardness. C. dubia chronic tests were inconclusive due to poor survival and reproduction in the synthetic waters. Available data suggest that within 50–200 mg/l water hardness, cobalt acute toxicily is inversely related to water hardness. Tentative cobalt acute criteria, based on power functions, are 288 μg/l and 873 μg/l for soft and hard water, respectively. Available chronic toxicity data