The accelerated bioconcentration test is commonly used in aquatic toxicology to calculate the bioconcentration factor (BCF) for waterbprne xenobiotics. This method assumes that when aquatic organisms accumulate chemicals, they behave as a single compartment with the chemical coming to an equilibrium between the organism compartment and the water. Experiments have shown, however, that for many lipophilic substances, the aquatic organism must be modeled as two or more compartments to characterize the accumulation process adequately. We studied the errors associated with the estimation of the BCF when the accelerated test is applied to substances that require two compartments to represent the fish.

Body level/time data for xenobiotic uptake and depuration were simulated using the computer simulation program Minidynamo. The aquatic organism was modeled as two compartments; the amount of xenobiotic in the organism was a biexponential function of time; that is, for the uptake phase, X = Ae-αt + Be-βt where X is the amount in the organism, t is time, and A, B, α, and β are constants. These data sets were then fitted using NONLIN, an iterative nonlinear least-squares computer program, to the single-compartment model of the accelerated bioconcentration test to determine the apparent uptake and depuration rate constants, k₁ and k₂, and to estimate the apparent BCF from their ratio. The larger the ratio of the 0 time intercepts of the two components of the biexponential curve, A/B, the larger was the error in estimating BCF. Estimates for the uptake rate constant, k₁, were generally good but 10 to 30% smaller than the actual values, while the depuration rate constant, k₂, was poorly estimated, even for cases where BCF was accurately estimated. In extreme cases, the calculated BCF was less than 5% of the actual BCF. Testing for fit to multicompartmental models and knowledge of multicompartmental characteristics and the infusion input can aid in reducing errors involved in estimates of BCF, k₂, and k₂ by the accelerated bioconcentration test.