Estimation of bioconcentration in marine species using structure‐activity models

Abstract

AbstractStructure‐activity models, which were developed to predict bioconcentration of organic chemicals in freshwater fish, were tested for use with several marine species (Cyprinodon variegatus, Lagodon rhomboides, Crassostrea virginica, Mytilus edulis). Significant linear relationships existed between the bioconcentration factor (BCF) for each marine species tested and the log of the octanol/water partition coefficient (P). The results suggest that freshwater models can be used to predict the BCF of a chemical in marine species, since the slopes of the freshwater models were within the 95% confidence intervals for marine models. Models derived from data on freshwater species were used to calculate BCF values for each of the marine species. The calculated BCF values were compared with the measured BCF values for each marine species and those measured for the fathead minnow (Pimephales promelas). The data indicate that the log BCF can be estimated for the marine species using freshwater models to within one order of magnitude for a minimum of 71% of the chemicals having a range of 3,000,000 in the partition coefficient (1.61 to 6.50 in log P). It appears that models based on data from freshwater species offer the same precision in estimation of BCF values for marine species as for freshwater fish. Good agreement existed between measured and calculated BCF values for both L. rhomboides and M. edulis, whereas the data for C. virginica were more variable and those for C. variegatus were the most variable. Tests indicated that, generally, calculated BCF values overestimate the measured BCF values. Significant linear relationships existed between measured BCF values for P. promelas and each marine species, except M. edulis.