Mechanistic Effect Modeling Approach for the Extrapolation of Species Sensitivity

In the higher-tier environmental risk assessment of chemicals, species sensitivity distributions (SSDs) are used to statistically describe differences in sensitivity between species and derive community level endpoints. SSDs are usually based on the results from short-term laboratory experiments performed under constant environmental conditions. However, different species may be kept at different "optimal" temperatures, which influence their apparent sensitivity and thus the derivation of endpoints. Also, the extrapolation capacity of SSDs is largely limited to the tested species and conditions. Time-variable exposures and effects at higher levels of biological organization, including biological interactions, are not considered. The quantitative effect prediction at higher tiers would ultimately require the extrapolation of toxicokinetics and toxicodynamics to untested species and the involvement of population and community modeling. In this regard, we tested a toxicokinetic-toxicodynamic modeling approach to mechanistically consider and correct endpoints for ambient temperature and demonstrate the significance for SSDs. We explored correlations in toxicokinetic-toxicodynamic model parameters which would allow for the extrapolation of sensitivities to untested species. Finally, we illustrate the applicability of the approach for higher level effect predictions using an individual-based model. Our results suggest that mechanistic effect modeling approaches can reduce the uncertainties in higher tier effect assessments related to knowledge gaps.