Wildlife road crossings: a mechanistic trait- and landscape-based model

The expanding global road network creates growing challenges for wildlife due to habitat fragmentation, forcing frequent road crossings during their movements. The interactions between animals and roads frequently lead to altered wildlife movement patterns and increased wildlife-vehicle collisions. Mitigating these impacts is crucial, but since financial resources are limited, road mitigation must be spatially prioritised in places where it will be most effective. Our objective was to provide a mechanistic understanding of how landscape structure (habitat amount and configuration) affects wildlife road crossings, while considering matrix permeability and two animal navigation-related traits: perceptual range and vision angle. To accomplish this, we devised individual-based models capable of representing movement dynamics for a broad range of taxa and landscape configurations, allowing us to manipulate and analyse these parameters. Our model enabled the identification of variations in wildlife road crossing concentration and total number of crossings among different habitat amounts and configurations, as well as matrix permeability, perceptual range, and vision angle. A higher concentration of road crossings was observed under two conditions: first, in landscapes characterised by intermediate amounts of habitat or with fewer patches; and second, in simulations with lower matrix permeability or with higher perceptual range or vision angle. These sets of landscape features and animal navigation-related traits that resulted in higher concentration of crossings should be prioritised in efforts to mitigate road impacts such as connectivity and wildlife-vehicle collisions.