How TK-TD and population models for aquatic macrophytes could support the risk assessment for plant protection products.

This case study of the SETAC workshop MODELINK demonstrates the potential use of mechanistic effects models for macrophytes to extrapolate from effects of a plant protection product observed in laboratory tests to effects resulting from dynamic exposure on macrophyte populations in edge-of-field water bodies. A standard EU risk assessment for an example herbicide based on macrophyte laboratory tests indicated risks for several exposure scenarios. Three of these scenarios are further analysed using effect models for two aquatic macrophytes, the free-floating standard test species Lemna sp. and the sediment-rooted submerged additional standard test species Myriophyllum spicatum. Both models include a toxicokinetic (TK) part, describing uptake and elimination of the toxicant, a toxicodynamic (TD) part, describing the internal concentration-response function for growth inhibition, and a description of biomass growth as a function of environmental factors to allow simulating seasonal dynamics. The TK-TD models are calibrated and tested using laboratory tests whereas the growth models were assumed to be fit for purpose based on comparisons of predictions with typical growth patterns observed in the field. For the risk assessment, biomass dynamics are predicted for the control situation and for several exposure levels. Based on specific protection goals for macrophytes, preliminary example decision criteria are suggested for evaluating the model outputs. The models refined the risk indicated by lower tier testing for two exposure scenarios, whilst confirmed the risk associated for the third. Uncertainties related to the experimental and the modelling approaches and their application in the risk assessment are discussed. Based on this case study and the assumption that the models prove suitable for risk assessment once fully evaluated, we recommend that i) ecological scenarios be developed that are also linked to the exposure scenarios and ii) quantitative protection goals be set to facilitate the interpretation of model results for risk assessment.