Estimating cougar predation rates from GPS location clusters

Anderson and Lindzey pioneered the use of GPS location clusters to detect cougar kills, exploiting the fact that a lion lingers at a carcass to feed over one or more days, producing a tight cluster of fixes. By visiting clusters in the field, they verified kills, identified prey, and developed predictive models relating cluster characteristics, such as number of nights and fix density, to the probability that a cluster represented a kill and to prey size. The approach made it feasible to estimate predation rates and prey composition across many individuals without continuous snow-tracking. It became a standard method in carnivore ecology, automating the kill-site detection that earlier studies achieved laboriously by hand and scaling predation studies to the GPS era.