Home range size (HRS) is the fundamental measure of space use by animals. Despite the importance of the home range concept, there is no consensus on how to estimate the HRS of animals. Assessments of the performance of commonly applied HRS estimators have largely been based on simulated data or on location data of few sample individuals occupying one study area. To empirically evaluate the impact of supplementary feeding, habitat composition, red deer sex, and estimation method (minimum convex polygon (MCP), kernel density estimator (KDE) and α-local convex hull (α-LoCoH)) on HRS, we analysed the data of 183 annual red deer home ranges using a mixed modelling approach. Red deer HRSs were smallest in areas with substantial supplementary feeding, intermediate in areas with closed forest cover but no supplementary feeding, and largest in fragmented landscapes where supplementary feeding rarely occurs. Consistently, male HRSs were larger than female HRSs. While MCP- and KDE-HRS estimates were roughly similar, estimates from the α-LoCoH method were substantially smaller than those of MCP and KDE. Analyses of 342 seasonal HRS largely reflected patterns of annual HRS. However, seasonal HRS differed between seasons and red deer sex. In areas with no or little feeding, red deer adjusted HRS seasonally, whereas red deer supplied with supplementary food during winter did not alter their HRS seasonally. Our study suggests that supplementary feeding and habitat configuration strongly affect the spatial ecology of red deer; this might have considerable sanitary and ecological implications. We suggest that sex differences in annual space use extent are proportional along a resource gradient but are mediated by seasons. Finally, method-related variation in space use studies of animals needs to be considered more cautiously.
