Vehicle tracks are predator highways in intact landscapes

Roads and other forms of linear infrastructure are rapidly proliferating worldwide, yet little is known about how roads affect the distribution and abundance of predators, particularly in relatively intact landscapes. We used a combination of motion-sensor cameras and spoor surveys to compare dingo, fox and feral cat activity on unsealed vehicle tracks (hereafter: roads) and up to 3 km away, in relatively intact landscapes of the Great Western Woodlands in south-western Australia. We compared predator activity as indicated by independent sightings and spoor observations, in woodlands and shrublands: vegetation types with contrasting permeabilities. Predator activity was observed between 12 and 261 times more frequently on roads compared with off-road for all species studied. Roads also appeared to affect predator activity up to 2.5 km away. Even poorly formed and abandoned roads concentrated predator activity and affected landscape-scale rates of predator observations. The effect of road proximity on predator activity was non-linear and different between vegetation types for dingoes and cats but not foxes. Our results provide new evidence of the effects of roads on predator activity in surrounding landscapes, with interacting effects of vegetation. They also reinforce previous findings e.g. stronger roads preference displayed by dingoes and foxes, than by cats. Roads and other linear infrastructure have strong effects on predator activity within intact landscapes, although further research is needed to characterise the implications for prey species. Road planning or approvals, as well as habitat restoration programs for threatened species, should account for the effects of roads on predator activity.