Towards the comparison of home range estimators obtained from contrasting tracking regimes: the wild boar as a case study

Abstract

Traditional VHF radio-tracking is gradually being replaced by GPS tracking devices in spatial ecology studies, although both technologies continue to be used. Differences between tracking regimes (time and fix frequency) may lead to home range estimates that are not directly comparable. Our primary aim was to test the reliability of comparisons in home range estimates using different estimators (minimum convex polygon [MCP], fixed kernels [KDE] and dynamic Brownian Bridge Models [dBBMM]) and tracking regimes (see above) and to provide an empirical basis for linking VHS and GPS tracking data analysis, using the wild boar (Sus scrofa) as a study model. Data were obtained from 15 GPS-collared individuals from three areas in Mediterranean Spain. Using tracking data, we simulated different tracking regimes (approaches), from three relocations per week during the daytime to one location per hour in the case of more intensive continuous monitoring. Results suggested that MCP produces the greatest differences between approaches, while 95% kernel home range (Khr95) the lowest, both in size and utilization distribution overlap. From the 50% kernel core range (Kcr50), similar estimates were also obtained for the less intensive approaches (typical of VHF tracking regimes). Using the most intensive sampling scheme, dBBMM estimates did not differ significantly from those obtained with Khr95 for all approaches. The standardization of home range estimates is very important for wildlife management and conservation. The proposed methodology allows direct comparison among studies using different approaches for estimating home ranges.