Evidence of quadratic effects of traffic volume on roadkill probability
Oral Presentation | 26 Aug 13:45 | E2

Authors: Denneboom, Dror;Bar-Massada, Avi;Shwartz, Assaf;

Wildlife-vehicle collisions are considered the predominant direct negative effect exerted by roads on many species. The probability of roadkill is affected by the level of traffic volume, and theory suggests that the effect should be unimodal. However, empirical evidence of this theory is lacking. We studied the effects of traffic volume on roadkill probability for 20 species in Israel using nation-wide data, encompassing 2846 road km, while controlling for the effects of road attributes, landscape features, and species abundance (using habitat suitability as a proxy). The unimodal effect was empirically supported for the striped hyena and gray wolf. A quadratic U-shaped effect was identified for six other species. We suggest that the U-shaped effect is caused by intra-species variability in road avoidance behavior, which may be related to high tolerance of anthropogenic disturbances, causing reduced avoidance of high traffic roads. Our research shows that high mortality rates occur at all levels of traffic volume, and that low traffic roads pose a major risk of mortality for many species. These insights, together with the gleaned effects of road attributes (e.g., lighting, median separation, verge vegetation) on roadkill probability are valuable for reducing wildlife-vehicle collisions, contributing to road safety and nature conservation.