Prioritizing road-kill mitigation areas: a spatially explicit national-scale model for an elusive carnivore

Aim: Roads impact wildlife in different ways, among which road mortality
has been the most studied. Budgets in conservation biology are usually
small, and macroecological approaches have been employed in recent years
as the first steps towards guiding management. Carnivores are particularly
vulnerable to mortality on roads due to their elevated ecological needs
(low population density, often low fecundity, and relatively large home
ranges). Our aim was to develop a ranking methodology to prioritize
specific areas for road-kill mitigation. Location: Continental Italy.
Methods: We studied 271 occurrences of live polecats (Mustela
putorius) and 212 polecat road-kill sites. We used the former to generate
a species distribution model (SDM), and the latter to identify the
variables that determined the road-kill risk. Habitat suitability was
derived from a spatial distribution model that combined the polecat
occurrence data with a set of environmental variables. Prey availability
was derived from the combination of suitability maps of 26 prey species.
We used generalized linear modelling to identify the set of variables that
best explained the occurrence of road-kills. The variables included in the
best-performing model were combined to produce the road risk map and to
identify the areas with the highest densities of road sections with
highest risk. Results: Road-kills were positively associated with the road
sections with higher broad-leaved forest coverage. The number of
casualties was found to be higher than expected on the national and
provincial roads and lower than expected on the local roads. Main
conclusions: This approach allowed us to identify the 10x10 km cells where
mitigation actions to prevent road-kills should be prioritized. As
mitigation actions (wildlife passage construction, fencing) are expensive,
measures should be prioritized on the specific high-risk road sections
inside these selected cells, avoiding generalized mitigation plans.