Abstract: More than 20% of the world’s terrestrial land mass is covered in roads, fragmenting the landscape and reducing wildlife connectivity. Roads are also well-known sources of wildlife mortality, with millions of wildlife-vehicle collisions (WVC) occurring annually. Little is known, however, about how habitat fragmentation and WVC-risk intersect on a country-wide scale. Using the landscape of the United Kingdom (UK) as a case study, we investigated “roadless areas” in the context of habitat fragmentation, WVC-risk and species conservation. Using a GIS-based approach and generalized linear models, we tested for associations between roadless area distribution and size, land cover types, and WVC occurrence. We hypothesised that WVC occurrence would be highest on roads closest to smaller roadless areas, due to the greater density of roads. Preliminary results indicate that the UK is broken up into at least 6,402 roadless areas by more than 400,000km of roads, with 71% of these areas being smaller than 1km². Roadless areas differed significantly in size between England, Scotland, Wales and Northern Ireland, with, on average, the smallest roadless areas found in England (mean = 2.7km²), and the largest occurring in Scotland (mean = 19.6km²). WVC had a non-linear relationship with roadless area size, but suggested that larger areas could be acting as a source for wildlife mortality on roads. This could potentially be explained by species abundance being lower in areas of higher road density and/or species learning to avoid roads. The most dominant land cover type across all roadless areas was acid grassland (~12500km²), a relatively low biodiversity habitat type, and 60% of the extent of roadless areas fell outside of existing protected areas. In conclusion, this research highlights not only the extent to which landscapes can be fragmented by roads, but also neighbouring land cover types that pose potentially higher risks for wildlife on roads. This study also suggests a controversial issue: that large patches of land of high biodiversity value may be a source of high wildlife-vehicle mortality. Finally, we highlight the potential for utilising roadless areas to contribute towards existing protected area networks, and towards reducing habitat fragmentation.
