Multiple individual southern brown bandicoots (Isoodon obesulus fusciventer) and foxes (Vulpes vulpes) use underpasses installed at a new highway in Perth, Western Australia

Context. Although wildlife crossing structures are often included when new roads are built, their effectiveness at reconnecting wildlife populations is still largely unknown. A new highway was built in 2005 through an area of remnant vegetation in Perth, Western Australia. Assessment of the area before construction identified potential impacts on a population of southern brown bandicoots (Isoodon obesulus fusciventer). Aims. We aimed to determine the use by bandicoots of three underpasses constructed to provide a linkage between habitats that were fragmented by the highway, focussing on how many different individuals used them, which is an essential step to demonstrate their effectiveness at reconnecting fragmented populations. Methods. We used detection of tracks in sand pads for 1 year to establish the use of the underpasses by bandicoots. We then captured 56 bandicoots and fitted them with passive integrated transponders (PIT), and installed a Trovan 650 scanner/decoder within the most frequently used underpass to establish whether multiple individuals used it. Key results. By using sand pads, we demonstrated that bandicoots used the underpasses, with a total of 278 passes between August 2005 and August 2006. One underpass accounted for 71% of these passes and was used already during construction. Eight different bandicoots were recorded using this underpass between August 2006 and August 2007, demonstrating use by multiple individuals. A dramatic decline in the use of this underpass was observed after foxes (Vulpes vulpes) also started using it in August 2006, and a fox built a den near the entrance of this structure. Because we also failed to recapture any of the bandicoots implanted with PITs we suspect that they had been killed by foxes. Conclusions. A severe decline in bandicoots coinciding with underpass use by foxes raises questions as to the long-term success of fauna crossings. Clearly, the relationship between underpass use by predators and the target species, in this case bandicoots, needs to be examined further. Implications. Our work demonstrated that although underpasses have the potential to reconnect populations because multiple individuals used them, their installation may be detrimental to wildlife populations if predators are not controlled.