Creating a model for the prediction of roadkill Kruger National Park, South Africa

Brilliant Mashao, University of Venda & Endangered Wildlife Trust
Eduard M. Stam, Matjutla J.Mokgoebo (Univen), Daniel M.Parker(UMP), Wendy J. Collinson (EWT)
Topic Area
Sustainability and Resilience in Transportation Systems

Roads not only create barriers for animals by preventing the free movement of individuals between populations, but also cause mortalities through collisions with vehicles (i.e. roadkill). In South Africa, most of our understanding of roadkill events stems mainly from research on national and regional roads, with very little known about the impacts of roads in protected areas. Despite ad hoc roadkill reports in protected areas by members of the public on social media, there have been little systematic roadkill surveys undertaken in South African protected areas and, road signage aside, no effective mitigation measures have been applied except the traffic officials of whom they are unable to monitor all roads simultaneously. Our study will form part of a five-year project to undertake an assessment of roadkill in protected areas and assess mitigation measures to reduce it. The present two-year project will record temporal and spatial roadkill in one section of the Kruger National Park, on both paved and unpaved roads. Factors, such as surrounding habitat, proximity of a water source and traffic volumes and speeds will be recorded. Using these data we will identify the predictors of where roadkill is most likely to occur, which will then be used to develop a 'Roadkill Risk Map'. The procedure can then be tested in other protected areas, and used as a future, cost-and time effect model for roadkill predictions.

Abstract Keywords
Roadkill
Modelling & Roadkill risk map