Mapping risk: Quantifying and predicting the risk of deer-vehicle collisions on major roads in England

Wildlife-vehicle collisions are increasing across both Europe and North America, with considerable implications for animal populations themselves, for human safety and in terms of economic cost. Deer are generally the primary species involved in large mammals-vehicle collisions. Classical mitigation measures have however proven to have a limited efficacy. The development of tools aimed at predicting the real-time risk of hitting deer on a particular stretch of road can improve both human and wildlife safety, particularly if such tools can be adopted on a large scale. We analysed data on deer-vehicle collisions (DVCs) occurring on the major roads in England between 2008-2014, collected on behalf of Highways England agency. Using zero-inflated regression models, we analysed the relationships between DVCs and data on environmental, bioclimatic and traffic-related factors, on different spatial scales and for different seasons. Traffic flow, average precipitation, and a combination of suburban areas and broadleaved forest were generally associated with increased frequency of DVCs. We used the results of these models to draw seasonal risk maps, which could potentially be used to target appropriate physical mitigation or measures aimed at increasing driver awareness.