Derivation of a toxicity-based model to predict how water chemistry influences silver toxicity to invertebrates

The effect of altering water chemistry on acute silver toxicity to three invertebrate species, two Daphnids, Daphnia magna and Daphnia pulex, as well as an amphipod Gammarus pulex was assessed. In addition, the physiological basis of Ag(I) toxicity to G. pulex was examined. Daphnia magna and D. pulex were more sensitive than G. pulex and 48 h LC(50) values in synthetic ion poor water were 0.47, 0.65 and 2.1 microg Ag(I) l(-1), respectively. Increasing water [Cl(-)] reduced Ag(I) toxicity in all species, and increasing water [Ca(2+)] from 50 to 1,500 microM reduced Ag(I) toxicity in G. pulex. Whole body Na(+) content, but not K(+) or Ca(2+) was significantly reduced in G. pulex exposed to 6 microg Ag(I) l(-1) for 24 h, but there was no inhibition of whole body Na(+)/K(+)-ATPase activity. Both increasing water [Cl(-)] and [Ca(2+)] reduced this Ag(I)-induced Na(+) loss. For D. magna, the presence of 10 mg l(-1) humic acid or 0.5 microM 3-mercaptoproprionic acid (3-MPA) increased the 48 h LC(50) values by 5.9 and 58.5-fold, respectively, and for D. pulex the presence of 1 microM thiosulfate increased the 48 h LC(50) value by four-fold. The D. magna toxicity data generated from this study were used to derive a Daphnia biotic ligand model (BLM). Analysis of the measured LC(50) values vs. the predicted LC(50) values for toxicity data from the present and published results where water Cl(-), Ca(2+), Na(+) or humic acid were varied showed that 91% of the measured toxicity data fell within a factor of two of the predicted LC(50) values. However, the daphnid BLM could not accurately predict G. pulex toxicity. Additionally, the Daphnia BLM was under-protective in the presence of the organic thiols 3-MPA or thiosulphate and predicted an increase in the LC(50) value of 114- and 74-fold, respectively. The Daphnia toxicity based BLM derived from the present data set is successful in predicting Daphnia toxicity in laboratory data sets in the absence of sulfur containing compounds, but shows its limitations when applied to waters containing organic thiols or thiosulphate.