Abstract

Although vehicle-induced mortality of wildlife is well known on roads throughout Australia, few empirical studies describe the extent of this mortality or assess the potential effects on wildlife populations. We recorded 529 roadkills of 53 vertebrate species along a 100-km circuit of three major roads during 20 weekly surveys across winter, spring and summer. This equates to 0.3 road-kills km^–1 week^–1 or one road-kill every 3.8 km week^–1. The most frequently killed native species were the northern brown bandicoot (4 per week), the mountain brushtail possum (2 per week) and the Australian magpie (2 per week). These values are underestimates because our survey technique could not detect all road-kills and ~40% of those left on the roadside disappeared within 7 days. Detailed study of the local population of the brown bandicoot is needed to determine whether such a level of road mortality is sustainable.

A logistic regression analysis was used to determine whether any of 10 road and landscape attributes were closely associated with the presence of specific groups of road-kills. Bandicoots were not associated with any measured attributes. Possums were more likely to occur along roads on mid-slopes and ridge-tops. Magpies were associated with roads on ridge-tops. Canopy-dwelling birds were more likely to be killed on 3-lane roads surrounded by dense vegetation.

Road-kill surveys such as this are needed to identify species for which road mortality is unsustainable, to determine the influence on threatened species, and to identify important crossing points where road-kills are high. The frequency of road-kill of many ground-dwelling or arboreal mammal species in this study highlights the importance of impact mitigation by road authorities.