Comparisons of acute toxicity of selected chemicals to rainbow trout and rats

Median lethal concentration (LC(50)) and dose (LD(50)) for 217 chemicals were compared over several exposure routes in rainbow trout and rats. Data were compiled primarily from on-line databases and included octanol-water partition coefficient (K(ow)), trout LC(50) and trout and rat LD(50) for oral, dermal, and intraperitoneal (IP) routes. K(ow) and molar lethality data were log-transformed for correlation and regression analyses. Median ratios (rat oral LD(50)/trout LC(50) = 156, rat/trout oral LD(50) = 0.80, rat/trout dermal LD(50) = 0.43, rat/trout IP LD(50) = 0.55) demonstrated the effect of similar versus dissimilar exposure routes on lethality, reflecting toxicokinetic differences. Simple linear regressions revealed significant correlations (Bonferroni P<0.05) for trout LC(50) versus rat oral LD(50) and for trout LD(50) versus rat LD(50) when matched on exposure route. When K(ow) was included with multiple regression to consider the effect of hydrophilic/lipophilic partitioning on uptake and toxicity, correlations improved only for trout LC(50) versus rat oral LD(50) and for trout IP LD(50) versus rat IP LD(50). Stratification of data into pesticides and nonpesticides failed to refine prediction of trout LC(50) from rat oral LD(50). Simple regressions reported here and in the literature relating fish LC(50) and rat oral LD(50) exhibited varying degrees of correlation (r = 0.26-0.99), reflecting differences in interspecies sensitivity, chemical group evaluated, and experimental methodologies. (C) 1998 Academic Press.