Visualizing ecotoxicology’s extrapolation dilemma: A discussion of case study data for 17α-ethynylestradiol (EE2) and implications for Ecological Risk Assessment

Abstract

Predicting and preventing negative environmental impacts from exposure to human-related chemicals stressors is necessary for long-term preservation of global biodiversity and ecosystem function. Anticipating the effects of chemical pollutants, however, remains a pervasive challenge for ecotoxicologist and environmental risk assessors who must rely on controversial extrapolation methods to extend toxicity predictions from a handful of short-term lab studies for standard test organisms to long-term chronic effects for entire ecosystems. Using data from US EPAs ECOTOX database for a synthetic estrogen, 17α-ethynylestradiol (EE2), I illustrate these extrapolation challenges and discuss potential implications for ecological risk assessment and environmental management. Analyses with EE2 data show there is limited opportunity to compare or validate effects across lab and field toxicology studies. Endpoints and effect concentrations for EE2 do not vary predictably within species or among related species, and well-studied test organisms tend to have a lower minimum and overall wider range of effect concentrations than less-studied organisms. Additionally, results from a multi-year whole-ecosystem field experiment demonstrates how species sensitivity distribution curves do not account for interspecific interactions and may therefore result in an underestimation of ecosystem level effects.