Vehicular collisions with wildlife are one of the most widespread and persistent human-wildlife conflicts that exist throughout the United States. An estimated 1-2 million wildlife-vehicle collisions (WVCs) occur each year, and that number is increasing annually. The total annual cost associated with WVCs is estimated to be >8.3 billion dollars, as well as the loss of millions of animals. Despite the magnitude of consequences from WVCs, relatively few options exist for reducing the frequency of collisions. A number of reasons can explain the shortage of options. First, identifying the most critical locations for mitigation is not straightforward. Hotspots of WVC locations are loosely defined, even though hotspots provide the best opportunity for cost-effective mitigation. Second, reporting of WVCs is inconsistent, resulting in incomplete information for studies that analyze where collisions occur. Inferences from these studies could be unreliable because of incomplete data. Finally, there is a lack of knowledge regarding large-scale and long-term trends in the frequencies of WVCs. Larger geographic and temporal studies are needed to understand the environmental influences for those trends. My overall objective was to enhance the current approaches for examining WVCs and provide more reliable inferences for reducing collisions. In Chapter 1, I address the issue of inconsistency and subjectivity in delineating hotspots of WVCs. In Chapter 2, I address the issue of sensitivity in statistical inferences from underreporting or studies of WVCs. In Chapter 3, I address the issue of understanding large-scale, dynamic processes that influence white-tailed deer (Odocoileus virginianus)-vehicle collisions (DVCs) through space and time. The collective works in these chapters contribute 3 primary conclusions for better understanding the influences of WVCs. First, the landscape can be used to objectively delineate hotspots. This new approach indicates that hotspots are larger than previously reported. Second, analyses of the influences of WVCs are highly robust to underreporting likely because WVCs occur in highly predictable patterns (i.e., hotspots). Therefore, relatively few reports are required for reliably understanding the environmental influences on where hotspots occur. Third, the large-scale, ecological drivers of DVCs through are related to suburbanization. The suburb effect consists of a unique combination intermediate to high traffic volume, high abundances of deer, and a highly fragmented landscape with high proportions of croplands. The suburb effect did not change through time, indicating high spatiotemporal predictability for DVCs. These collective works suggest large hotspots associated with suburban landscapes account for the highest frequencies of collisions, therefore these locations should be targeted for mitigation. Identifying the most critical locations to mitigate can be accomplished with relatively few reports of collisions if collected in a consistent manner. Managers should consider investing in long-term mitigation strategies (i.e., underpasses) to reduce WVCs for many years, because the ecological drivers of hotspots do not appear to change.
Read
