| Sumario: | Deer-vehicle collisions (DVCs) have a negative impact on the economy, traffic safety, and the general well-being of otherwise healthy deer. To mitigate DVCs, it is imperative to gain a better understanding of factors that play a role in their spatial distribution. Much of the existing research on DVCs in the United States has been inconclusive, pointing to a variety of causal factors that seem more specific to study site and region than indicative of broad patterns. Little DVC research has been conducted in the southern United States, making the region particularly important with regard to this issue. In this study, we evaluate landscape factors that contributed to the distribution of 347 DVCs that occurred in Forrest and Lamar Counties of south Mississippi, from 2006 to 2009. Using nearest- neighbor and discriminant analysis, we demonstrate that DVCs in south Mississippi are not random spatial phenomena. We also develop a classification model that identified seven landscape metrics, explained 100 percent of the variance, and could distinguish DVCs from control sites with an accuracy of 81.3 percent.
|
|---|