|
(21-23) Estimating acute toxicity to endangered fishes and other aquatic species. Mayer, Foster*,1, Ellersieck, Mark2, Asfaw, Amwa2, 1 U.S. Environmental Protection Agency, Gulf Breeze, FL2 University of Missouri-Columbia, Columbia, MO ABSTRACT- Predictive toxicological models, including estimates of uncertainty, are necessary to address probability-based risk assessments. A method and software were developed to aid in estimating acute toxicity of chemicals to species where data is lacking, particularly endangered species. Interspecies correlation models for acute toxicity were derived using using Model II least squares regression, where both variables are independent and subject to measurement error. Only those species with paired tests on five or more chemicals were analyzed. The resulting 618 models consisted of 53 aquatic species (1 alga, 16 invertebrates, 33 fishes, 3 amphibians), including 15 endangered ones. Acute toxicity of a chemical to one species could be predicted from toxicity to another species with known model and surrogate uncertainty. Correlations were best within a family, decreasing with increasing taxonomic distance. However, rainbow trout was the most useful of all species for acute estimations among more than one family. Key words: toxicity, acute, aquatic, correlations |
