Freshwater to saltwater toxicity extrapolation using species sensitivity distributions

Abstract There is generally a lack of saltwater ecotoxicity data for risk assessment purposes, leaving an unknown margin of uncertainty in saltwater assessments that utilize surrogate freshwater data. Consequently, a need for sound scientific advice on the suitability of using freshwater data to extrapolate to saltwater effects exists. Here we use species sensitivity distributions to determine if freshwater datasets are adequately protective of saltwater species assemblages for 21 chemical substances. For ammonia and the metal compounds among these data, freshwater data were generally protective because freshwater organisms tended to be more sensitive. In contrast, for pesticide and narcotic compounds, saltwater species tended to be more sensitive and a suitable uncertainty factor would need to be applied to surrogate freshwater data. Biological and physicochemical factors contribute to such differences in freshwater and saltwater species sensitivities, but the species compositions of datasets used are also important.