Experimenting with artificial shelters to assist habitat restoration for medium‐sized ground‐dwelling mammals

Many terrestrial vertebrates require microhabitat shelter structures for survival. Where anthropogenic or environmental disturbances have degraded or depleted shelter, artificial shelters are increasingly used to provide supplementary habitat for various taxa. However, their application to medium‐sized ground‐dwelling mammals (MGMs) remains largely unexplored. We installed rudimentary artificial shelters in a conservation reserve to emulate the vegetative cover used as refuge by three Australian MGMs: the long‐nosed ( Perameles nasuta ) and northern brown bandicoots ( Isoodon macrourus ), and the vulnerable long‐nosed potoroo ( Potorous tridactylus ). We used multi‐method occupancy modeling and behavioral analysis to compare the detections and behaviors of the target species with those of four non‐target species. Our study design included three plot types (treatments)—artificial shelter, baited lure, and disturbance control—enabling unambiguous evaluation of responses to the shelters. The bandicoots showed no difference in detection among treatments, whereas detection of the potoroo was highest at control plots. Detection of non‐target species was generally highest at baited plots. The target species demonstrated a much higher willingness to enter the shelters compared to the non‐target species. The use of the structures appeared to be exploratory, with no evidence that individuals remained in shelters during the day. There was limited evidence that the shelters reduced the target species' perceptions of predation risk and no evidence that predators were attracted to the shelters. The high availability of natural shelter and our small sample size likely influenced these findings. Nevertheless, they lay the foundation for research and refinement into more effective shelter designs.