Comparison of camera trapping and live trapping of mammals in Tasmanian coastal woodland and heathland

Camera traps have potential as a cost-and time-efficient technique to survey mammal species diversity and to monitor changes in abundance. During a survey of mammals in Tasmanian coastal woodland and heathland, we compared species richness and indices of abundance obtained from, and survey effort for, baited camera and live traps at the same trap stations on four trapping blocks over 3 nights during two survey periods (March and August-September). More mammal species were detected by camera traps than live traps, partly because of the greater size range of species susceptible to camera traps; yet other rarely live-trapped and some commonly live-trapped species were recorded only on cameras. All but one species, the swamp rat (Rattus lutreolus), were detected at more trap stations by camera traps than by live traps. For those species detected by both techniques, on all four blocks combined, indices of abundance were positively correlated. For the most commonly caught species, the long-nosed potoroo (Potorous tridactylus), there were positive correlations between numbers live trapped and numbers of visits caught on camera at both the spatial scale of the block and of the trap station; however, this relationship was not consistent between survey periods. Higher numbers of detections of P. tridactylus on camera traps allowed finer grained interpretation of results, implying that live trapping reflects the mammal community less efficiently than camera trapping. Camera trapping required similar to 50% of the person-hours needed for live trapping. Compared with live trapping, camera traps are a more time-efficient method of detecting differences and monitoring changes in abundance of Australian mammals.