Estimating and indexing feral cat population abundances using camera traps

Context. The ability to monitor changes in population abundance is critical to the success of pest animal management and research programs. Feral cats (Felis catus) are an important pest animal, but current monitoring techniques have limited sensitivity or are limited in use to particular circumstances or habitats. Recent advances in camera-trapping methods provide the potential to identify individual feral cats, and to use this information to estimate population abundances using capture-mark-recapture (CMR) methods. Aims. Here, we use a manipulative study to test whether camera-trapping and CMR methods can be used to estimate feral cat abundances. Methods. We established a grid of infrared cameras and lure stations over three pastoral properties on Kangaroo Island, Australia, for 15 days. We then reduced the population abundance with an intensive trapping program and repeated the camera survey. We estimated population abundances using robust design CMR models, and converted abundance estimates to densities using home-range data from GPS tracking. We also calculated relative abundance indices from the same data. Key results. The CMR methods produced credible estimates of the change in population abundance, with useful confidence intervals, showing a statistically identifiable population decline from at least 0.7 cats km(-2) before trapping down to 0.4 cats km(-2) after trapping. The indexing method also showed a statistically identifiable decrease in abundance. Conclusions. Camera-trapping and CMR methods can provide a useful method for monitoring changes in the absolute abundance of feral cat populations. Camera-trap data may also be used to produce indices of relative abundance when the assumptions of CMR models cannot be met. Implications. These methods are widely applicable. The ability to reliably estimate feral cat abundances allows for more effective management than is generally available.