Spatial Dimensions of the Risks of Rodenticide Use to Non-target Small Mammals and Applications in Spatially Explicit Risk Modeling

Abstract

Both target and non-target small mammals are exposed to rodenticides (AR). A better understanding of the drivers controlling this exposure is critical for the conservation of threatened small mammal species but also because they may represent important pathways of poisoning for birds of prey and carnivore mammals. Here, we consider the spatial components involved in the process of small mammal exposure to ARs with the aim to address how these can be used in spatially explicit risk assessment. We present how various drivers operate on multiple spatial scales. On continental and/or regional scales, both biogeographical distribution of small mammals and other species of conservation value and international/national regulations of AR applications (indoor vs outdoor…) could be used to identify some countries or states where exposure is more likely. For application at the local scale (i.e. few km2), we reviewed published studies that analysed the spatial pattern of small mammal exposure to ARs according to species and distance to treatments. We evidence that most of the small mammals exposed to AR are found in the immediate vicinity of treatment areas, i.e., within 100 m. Over 100 m, exposed rodents are rare but can be found until 750 m distance from treatment areas. Species traits related to spatial dimension such as habitat preferences, home range size and mobility also influence exposure. Exposure is variable, in terms of proportion of contaminated individuals and levels of residues, for species showing small home-range size and a limited spatial mobility. The level of exposure depends on whether the main habitat of the given species is similar or not to the one of the target rodent. For instance, exposure of the common vole, a grassland species, is low when ARs are used indoor while it can be highly exposed when bromadiolone is applied outdoor to control the water vole, a sympatric species. For small mammals exhibiting a relatively large home-range size and a high spatial mobility such as the wood mouse and the bank vole, the exposure is commonly reported within a lower range than target species. Although this has not been studied in details, we also address how landscape and/or habitat features may modulate exposure, suggesting that landscape management may help to mitigate the risk of ARs to small mammals. Finally, we discuss both the advantages and disadvantages of statistical, analytical or simulation models to assess potential or actual exposure of NTSM to AR in a spatially explicit way. We conclude that in order to analyse global patterns in usage and exposure risks, large scale statistical modelling should be used while for detailed site specific assessments, simulation models may be more appropriate.