Animal-vehicle collisions (AVCs) substantially contribute to wildlife population decline and adversely impact threatened species with already limited numbers. In this study, we aimed to investigate the factors contributing to AVCs involving two threatened species in the Republic of Korea: the leopard cat (Prionailurus bengalensis) and Eurasian otter (Lutra lutra). These two species are the most frequently affected by AVCs among threatened mammals in Korea. To achieve this, we used data from the Korea Roadkill Observation System (KROS), a government-sponsored web-based AVC monitoring system. Our analysis focused on 17 variables, categorized into bioclimatic, landscape, and traffic factors, using machine-learning-based predictive modeling. From 2019 to 2021, 589 AVC incidents of P. bengalensis and 228 AVC incidents of L. lutra were recorded in KROS. Our findings indicate that AVC frequencies peaked during the most active seasons of the year, corresponding to the mating or dispersal periods, with P. bengalensis showing the highest AVC frequency in fall and L. lutra peaking in summer. For P. bengalensis, habitat suitability consistently exerted a strong influence on AVC risk across all seasons, whereas for L. lutra, the key factors affecting AVC risk highly varied seasonally. These results underscore the importance of seasonal and species-specific approaches for AVC mitigation. Targeted strategies, including wildlife corridors, underpasses, and speed reduction zones in high-risk areas, are recommended to mitigate AVC risks. By identifying the key factors and their seasonal dynamics, this study provides critical insights for conserving threatened wildlife and effectively reducing AVC incidents.
Kim et al. (Sat,) studied this question.