Road networks continue to expand globally with predictable effects on ecological systems. Research into the effectiveness of road underpasses and overpasses for wildlife has been concentrated in North America and Europe. In Australia, most studies of underpasses have been of relatively short duration and without reference sites to give context to the measured rates of use. We studied 5–7 road underpasses at two locations in eastern Australia over 2–3 years, comparing camera trap detections of animals in underpasses with those at nearby forest sites. Three species of large macropod (wallabies and kangaroos) were frequently detected in the underpasses, with some underpasses traversed 1–4 times per week, and in many cases exceeded detections in the forest. The lace monitor (Varanus various) was detected in all underpasses, often once per week during spring and summer, and infrequently in the forest. At each location a different small macropod species, including one regionally threatened, showed a higher probability of detection in one underpass compared to several of the forest sites. The vulnerable koala (Phascolartos cinereus) was detected infrequently in underpasses and in the adjoining forest. The short-beaked echidna (Tachyglossus aculeatus) had a high probability of detection in a single underpass. The ‘prey-trap hypothesis’ postulates that predators will exhibit increased activity at underpasses as a consequence of prey being funnelled. We found the red fox (Vulpes vulpes) had high activity in some underpasses. However, its activity coincided less than expected with the activity of the mammals most at risk to it. Our results provide no consistent support for the ‘prey-trap hypothesis’. Instead, our study confirms the generic value of underpasses for a range of medium-large mammals as well as one large reptile. Habitat adjoining underpasses exerts a strong influence on their use and requires greater consideration to maximise underpass use.  

Camera traps operated in the underpasses at Grafton from February 2013 until January 2016 (i.e. 135 weeks). Camera traps operated in the forest at Grafton from December 2015 until February 2017 (i.e. 76 weeks).

Camera traps operated in the underpasses at Port Macquarie from June 2013 until August 2016 (i.e. 165 weeks). Camera traps operated in the forest at Port Macquarie from October 2014 until August 2016 (i.e. 98 weeks).

Weekly camera detections were collated for each species at the two locations.