Abstract
Conducting population studies on elusive species can be challenging. Despite intensive sampling effort, individuals can vary markedly in their trappability. These differences in detection probabilities are problematic for population estimates, where models incorporate assumptions that, if they are present within an area, all animals have an equal probability of being trapped. This study investigates the use of remote cameras to estimate detectability of a medium-sized macropod. We conducted intensive monitoring of quokkas (Setonix brachyurus) at four riparian sites in southwest Western Australia. Quokkas were trapped (seasonally over 18 months) and marked. Camera traps were placed out for 12 months in the vicinity of the trapping area. Camera trap events were reviewed and animals were classified as marked or not-marked. Animals were identified from photos as adult or juvenile and sexed (presence of pouch young, head size) where possible. Cameras indicated the presence of trap-shy animals which had not been marked during the course of the 18 months trapping, and allowed the estimation of detection probabilities for different cohorts. We conclude that camera trapping can enhance the robustness of population estimates of these elusive animals.