Date of Award

August 2026

Document Type

Thesis

Degree Name

Master of Science (MS)

Department

Biology

First Advisor

Susan Ellis-Felege

Second Advisor

Melissa Holbrook-Schmitt

Abstract

Wildlife-vehicle collisions (WVCs) represent a significant ecological and public safety concern across the United States, yet data collection systems in many states remain insufficient to capture the true magnitude and taxonomic breadth of collision-related mortality. North Dakota, where WVCs account for approximately 18% of all reported crashes, relies almost exclusively on law enforcement collision reports governed by high property damage thresholds, a protocol that systematically excludes the majority of species and underestimates total mortality. This study addresses these data gaps through three complementary approaches conducted in North Dakota from 2021 to 2025. First, we conducted systematic weekly carcass surveys across approximately 2,136 kilometers of state-managed highway transects. Carcass detections were characterized by species, seasonality, and spatial distribution, and compared against North Dakota Department of Transportation crash report data to quantify methodological discrepancies in collision reporting. Second, the carcass data Getis-Ord Gi* hotspot analysis was conducted to identify statistically significant areas of elevated wildlife carcass density across statewide and district-level spatial extents. Finally, we evaluated wildlife use of two wildlife crossing structures and one hydrological structure on U.S. Highway 85 using camera trap imagery, quantifying species use, success rate, approaches per day, and the demographic and temporal predictors of successful movements. Within the surveyed extent and survey periods, the carcass survey recorded approximately 293 carcass detections for every reported wildlife crash, with furbearers (n = 1,123) and birds (n = 1,071) nearly absent from crash records despite comprising the majority of detections. Hotspot analysis identified priority segments for mitigation investment across multiple spatial scales. Camera trap analysis documented use of the Long X underpass by bighorn sheep, a species widely reported to avoid underpasses, across all sex and age classes, with an overall success rate of 92.2%. Together, these findings provide actionable baseline data to support evidence-based WVC mitigation and wildlife management planning by North Dakota Department of Transportation and the North Dakota Game and Fish Department.

Download

Share

COinS