Spatial-temporal analysis of roe deer vehicle collisions in Bavaria

In densely populated European areas, the number of wildlife-vehicle collisions is rising due to increasing traffic volumes, animal populations, and vehicle speeds. Identifying geographical and temporal accident hotspots may help to reduce wildlife-vehicle collisions (WVCs), improve human safety and animal welfare. In this study, an analysis of roe deer (Capreolus capreolus)-vehicle collisions (RDVCs) throughout Bavaria reveals ultradian, circannual, and geographical hotspots over a period of ten years (2009-2018). The highest density of RDVCs occurs in May, followed by the short but intense rutting season in August, and a third, longer lasting peak from September to November. Spring and autumn migration and the dispersal of young roe deer strongly influence the annual pattern of RDVCs. The diurnal distribution of RDVCs is interlinked with the crepuscular activity patterns of roe deer, with dusk and dawn accounting for 55.6 % of all RDVCs. The number of accidents is influenced to a small extent by the phase of the moon. Most accidents on a regional scale occur in eastern Bavaria. The fewest accidents occur in the southern Alpine region and in the north-west. Large and small cities are avoided. On a smaller landscape scale, high-risk areas for RDVCs were identified along grassland patches within densely forested areas and along forest patches within grassland areas. The majority of accidents, and the highest risk of collision, arise on tertiary roads. The results demonstrate that geographical and temporal analysis can help to identify WVC hotspots and form the basis for mitigation measures.