Effects of sampling effort based on GPS telemetry on home-range size estimations

Home-range size is widely used in wildlife biology to assess animal-habitat relationships. But home-range size estimation largely depends on the estimator and sample size used. Using 3 different estimators (minimum convex polygons, fixed-kernels, clusters), we used data collected on moose (Alces alces), fitted with Global Positioning System (GPS) collars (which allow automatic recording of large data sets with fewer field constraints than VHF telemetry) to investigate to what extent increasing the number of locations affects home-range size estimations. Our results indicated that 100 to 300 locations per animal annually and 30 to 100 locations seasonally were needed to reach an asymptote. High biases occurred below this asymptotic value. Minimum convex polygons consistently underestimated home-range sizes, but fixed-kernel and cluster estimators followed variable trends that often overestimated home-range sizes. Low sampling efforts also affected the number of animals needed to ensure an adequate statistical power of analysis to compare space rise between 2 groups. Despite the higher cost, our findings indicate that in most instances, GPS telemetry is better suited than conventional radiotelemetry to estimate home-range sizes precisely and accurately. Moreover, this tracking technique is not limited by meteorological constraints and allows for the collection of similar sample sizes for all tracked animals, which is of major importance for further comparisons of space rise among individuals.