The importance of crash reporting requirements and how they affect analyses of factors associated with wildlife-vehicle collisions

Several factors have been identified to increase wildlife-vehicle collisions (WVCs). These may be species specific, ecological, temporal, driver related or related to road characteristics. To recommend effective mitigation strategies, it is imperative to understand the underlying factors driving WVCs. Our objective was to use crash data to identify factors that may contribute to reported WVCs in North Dakota and determine if changes in crash reporting affected the results. In 2013, crash reporting requirements were changed to only include crashes with human injuries or fatalities, and in 2019 increased the property damage threshold from $1,000 to $4,000. Based on reported crashes from North Dakota Department of Transportation (NDDOT), we compared results of the full crash dataset (2005-2022), pre-2013, post-2013, and fatalities and injuries data (2005-2022). We extracted factors from crash reports and hypothesized that different reporting causes different conclusions. We analyzed species, crash time, season, driver age and sex, and road type, speed limit, and lighting conditions. Deer were involved in >95% of the 30,476 reported wildlife-vehicle accidents. Annually reported WVCs averaged 3,488 in pre-2013 data, 133 in post-2013 data, and 40 in the fatalities and injuries dataset. We found significant differences in total number of WVCs with respect to species group, season, time of day, and road features (i.e., road type, lighting conditions, speed limit). The driver's age was only significant in the fatalities and injuries dataset. Given the reduction in sample size after 2013, we did not detect statistically significant results in the post-2013 data for the effects of season. However, there were statistically significant differences in all-years, pre-2013, and fatalities and injuries datasets with respect to season. This suggests the ability to detect important factors is influenced by reduced sample size from reporting requirement changes in 2013 that limited our ability to make interferences on factors that affect WVCs in North Dakota.