Population-level effects and recovery of aquatic invertebrates after multiple applications of an insecticide

Standard risk assessment of plant protection products (PPP) combines "worst-case" exposure scenarios with effect thresholds using assessment (safety) factors to account for uncertainties. If needed, risks can be addressed applying more realistic conditions at higher tiers, which refine exposure and/or effect assessments using additional data. However, it is not possible to investigate the wide range of potential scenarios experimentally. In contrast, ecotoxicological mechanistic effect models do allow for addressing a multitude of scenarios. Furthermore, they may aid the interpretation of experiments such as mesocosm studies, allowing extrapolation to conditions not covered in experiments. Here, we explore how to use mechanistic effect models in the aquatic risk assessment of a model insecticide (Modelmethrin), applied several times per season but rapidly dissipating between applications. The case study focuses on potential effects on aquatic arthropods, the most sensitive group for this substance. The models provide information on the impact of a number of short exposure pulses on sensitive and/or vulnerable populations and, when impacted, assess recovery. The species to model were selected based on their sensitivity in laboratory and field (mesocosm) studies. The general unified threshold model for survival (GUTS) model, which describes the toxicokinetics and toxicodynamics of chemicals in individuals, was linked to 3 individual-based models (IBM), translating individual survival of sensitive organisms into population-level effects. The impact of pulsed insecticide exposures on populations were modeled using the spatially explicit IBM metapopulation model for assessing spatial and temporal effects of pesticides (MASTEP) for Gammarus pulex, the Chaoborus IBM for populations of Chaoborus crystallinus, and the "IdamP" model for populations of Daphnia magna. The different models were able to predict the potential effects of Modelmethrin applications to key arthropod species inhabiting different aquatic ecosystems; the most sensitive species were significantly impacted unless respective mitigation measures were implemented (buffer zones resulting in reduced exposure). As expected the impact was stronger in shallow ditches as compared to deeper pond scenarios. Furthermore, as expected, recovery depended on factors such as temperature (affecting population growth rate and number of generations) and the occurence of nonimpacted aquatic ecosystems (their frequency and connectivity). These model predictions were largely in line with field observations and/or the results of a mesocosm study, providing additional evidence on the suitability and reliability of the models for risk assessment purposes. Because of their flexibility, models may predict the likelihood of unacceptable effects-based on previously defined protection goals-for a range of insecticide exposure scenarios and freshwater habitats.