Epsilon Archive for Student Projects

Abstract

This study evaluated whether the density and design of conventional bridges affected the frequency of reported deer-vehicle collisions along a given road or railroad section. I used official accident data from roads for 2008-2010 and from railways for 2001-2010, data on bridges and infrastructure together with digital topographic information, and estimates of regional abundances of moose (Alces alces) and roe deer (Capreolus capreolus) to develop generalized linear models of the parameters influencing the occurrence of deer-vehicle collisions. Among the obtained models, the most parsimonious were distinguished using the Akaike’s Information Criterion. Single regressions revealed that the density of bridges was negatively related to the occurrence of accidents along railways but not on roads. Traffic volume on both the barrier infrastructure and inside the passage, as well as other bridge characteristics such as use, width, integrity type and shape had some but not consistent effect on the occurrence of wildlife-vehicle collisions. However, as expected, multiple regression analyses revealed that environmental variables were the main factors influencing the occurrence of accidents on railways and roads, with forest cover, the density of buildings, of infrastructure and other linear features such as watercourses as the most relevant parameters. I conclude that even conventional bridges, especially if widened and placed appropriately, can contribute to the safe crossing of barrier infrastructure by animals, reducing the occurrence of deer-vehicle collisions.