Long, R. A., T. M. Donovan, P. MacKay, J. S. Buzas, and W. J. Zielinski. Predicting carnivore occurrence using data collected with multiple, noninvasive methods.Landscape Ecology 26:327-340.
Abstract
Terrestrial carnivores typically require large areas of habitat and exist at low densities. As “top level” consumers, carnivores affect the biological structure and composition of ecosystems. We employed multiple, noninvasive survey methods—scat detection dogs, remote cameras, and hair snares—to collect detection-nondetection data for American black bears (Ursus americanus), fishers (Martes pennanti), and bobcats (Lynx rufus) throughout Vermont. We analyzed these data using an occupancy modeling approach that explicitly incorporated detectability as well as habitat and landscape variables. Model results were then used to predict occurrence for each species across the study area. Receiver operating characteristic (ROC) analyses of our models for species with high detectability suggest that the results of such surveys and modeling efforts may be useful for those striving to conserve or manage species at the regional or landscape scale. The noninvasive methods used for this project enabled us to collect important detection-nondetection information from low density, secretive, and wide-ranging species, without the need to capture or interact with the study subjects directly.