Spatio-temporal hotspots of wildlife-vehicle collisions in Poland: How congruent are mammals, birds, reptiles and amphibians?

The Polish Roadkill Observation System (PROS) database, a large dataset of roadkills collected between 2000 and 2022 in Poland, was used. We calculated the total length for each road type and the main type of environment around the wildlife-vehicle collision (WVC) event, in a grid of 10 × 10 km (e.g., spatial unit). We explored the spatial congruence in WVCs among amphibians, reptiles, birds and mammals across the country, using spatially explicit correlation based on the Mantel tests. We used a) Generalized Linear Mixed Models to investigate the association between WVC and the type of dominant environment and animal group, and b) Generalized Boosted Regression Models to investigate, separately for each animal group, the association between WVC and the length and type of road in each spatial unit. A total of 19,846 roadkills were recorded in Poland, involving 28,952 individuals from different animal species: 14 amphibians, 8 reptiles, 133 birds and 52 mammals. The spatial distribution of roadkill events in the country was mainly clustered around the biggest cities. Hotspots were concentrated near cities (Warsaw, Kraków, Rrzeszów) and in areas known for high biodiversity. Coldspots - relatively smaller than hotspots - were areas characterized by a high density of housing infrastructure, with lower naturality and a predominance of single roadkill casualties. A higher spatial congruence in WVC was found between birds and mammals (71 %) than between the other animal groups. Overall, the animal group less congruent with the other groups was amphibians (13 %), while birds were most congruent with all groups. We discussed some advantages and drawbacks when working with non-systematic survey datasets of roadkills. Finally, we recommended including roadkill clusters of multiple animal groups (hotspots) in strategies for mitigating wildlife-vehicle collisions but also considering more specific strategies, which can combine the type of environment and roads, concerning each animal group.