Analysis of toxicity tests on fish growth

We present a statistical analysis of bioassays for fish growth, such as the routine toxicity test that is described in the OECD guideline 210. The analysis is based on the Dynamic Energy Budget theory and a one-compartment kinetics for the toxic compound. It is fully process oriented. We compare a formulation in terms of direct effects on growth with indirect effects via assimilation and maintenance. All formulations characterize the effects by a no-effect concentration, a tolerance concentration and the elimination rate. Simplified formulations are obtained for very small and very large elimination rates. The accuracy of estimates for the no-effect concentration is judged with the profile likelihood function. The method is illustrated with applications to several data-sets for body size versus concentration of toxicant.