Selection of scenarios for landscape-level risk assessment of chemicals: case studies for mammals

Background For landscape-level risk assessments of pesticides, the choice of the scenario is a key question, since it determines the outcome of a risk assessment. Typically, the aim is to select a realistic worst-case scenario. In the present study, landscapes from an area with a high proportion of cereal fields in France were analysed and simulations with population models for wood mouse, common vole, brown hare and European rabbit were conducted to understand if the worst-case character regarding pesticide exposure and population survival can be determined based on landscape features alone. Furthermore, it was analysed which landscape features relate with population survival and the magnitude of effects due to pesticide application. Answers to these question may help to decide whether landscape scenarios can be selected based on expert decision and whether the same scenarios may be used for different species or not. Results There were species-specific landscape features relating to long-term population survival. A landscape that is worst-case for one species, was not necessarily worst-case for another. Furthermore, landscapes that were worst-case regarding population survival were often not worst-case regarding the magnitude of effects resulting from pesticide application. We also found that small landscapes were sometimes, but not always worst-case compared to larger landscapes. When small landscapes were worst-case, this was typical because of the artificial borders of the digitised landscape. Conclusions Landscape analyses can help to obtain an approximate impression of the worst-case character of a landscape scenario. However, since it was difficult to consistently and reliably do this for single landscapes, it may be advisable to use a set of different landscapes for each risk assessment, which covers the natural variability. Depending on whether population survival shall be ensured or the magnitude of effects due to pesticides, different landscape structure and composition needs to be considered to establish a worst-case landscape scenario.