Variability in acute to chronic ratios (ACRs; median lethal or effect concentration divided by chronic value) has been of continuing interest in aquatic toxicology because of the reliance on ACRs to estimate chronic toxicity for chemicals and species with known acute toxicity data but with limited or no information for chronic toxicity. To investigate the variability and significant differences in ACRs, an extensive data set was compiled of 456 same-species pairs of acute and maximum acceptable toxicant concentrations for metals, narcotics, pesticides, and other organic chemicals. The overall median value for 456 aquatic invertebrate and fish ACRs analyzed in the present study was 8.3, with a 16,000-fold range in values (1.1–18,550) and a 32-fold range in 10th and 90th percentile values (2.5–79.5). Median ACRs for taxa, ambient habitat media, chronic test end point, and chemical mode of action (MOA)/class categories generally were similar but, in some cases, extremely variable (ranges of 1 to gt;10,000). No significant differences (p ⩽ 0.05) were found in median ACRs between taxa, although invertebrate ACRs generally were more variable than fish ACRs. Freshwater organisms had median ACRs significantly greater than those of saltwater species and also were more variable. No significant differences were found in median ACRs among chemical MOA/class data sets; however, ACR variance differed significantly among MOAs. Although few significant differences occurred among median ACRs for different groups, those categories that were highly variable are at an increased risk of underestimated chronic toxicity when mean or median ACRs are used.

A primary objective of threatened and endangered species conservation is to ensure that chemical contaminants and other stressors do not adversely affect listed species. Assessments of the ecological risks of chemical exposures to listed species often rely on the use of surrogate species, safety factors, and species sensitivity distributions (SSDs) of chemical toxicity; however, the protectiveness of these approaches can be uncertain. We comprehensively evaluated the protectiveness of SSD first and fifth percentile hazard concentrations (HC1, HC5) relative to the application of safety factors using 68 SSDs generated from 1,482 acute (lethal concentration of 50%, or LC50) toxicity records for 291 species, including 24 endangered species (20 fish, four mussels). The SSD HC5s and HC1s were lower than 97 and 99.5% of all endangered species mean acute LC50s, respectively. The HC5s were significantly less than the concentrations derived from applying safety factors of 5 and 10 to rainbow trout (Oncorhynchus mykiss) toxicity data, and the HC1s were generally lower than the concentrations derived from a safety factor of 100 applied to rainbow trout toxicity values. Comparison of relative sensitivity (SSD percentiles) of broad taxonomic groups showed that crustaceans were generally the most sensitive taxa and taxa sensitivity was related to chemical mechanism of action. Comparison of relative sensitivity of narrow fish taxonomic groups showed that standard test fish species were generally less sensitive than salmonids and listed fish. We recommend the use of SSDs as a distribution-based risk assessment approach that is generally protective of listed species.