Quantifying Home-Range Overlap: The Importance of the Utilization Distribution

Abstract

The concept of an animal's home range has evolved over time, as have methods for estimating home-range size and shape. Recently, home-range estimation methods have focused on estimating an animal's utilization distribution (UD; i.e., the probability distribution defining the animal's use of space). We illustrate the importance of the utilization distribution in characterizing the degree of overlap between home ranges (e.g., when assessing site fidelity or space-use sharing among individuals). We compare several different statistics for their ability to accurately rank paired examples in terms of their degree of overlap. These examples illustrate limitations of indices commonly used to quantify home-range overlap and suggest that new overlap indices that are a function of the UD are likely to be more informative. We suggest 2 new statistics for measuring home-range overlap: (1) for a measure of spaceuse sharing, we suggest a generalization of Hurlbert's (1978) E/E uniform statistic, which we term the utilization distribution overlap index (UDOI), and (2) for a general measure of similarity between UD estimates, we suggest Bhattacharyya's affinity (BA;Bhattacharyya 1943). Using a short simulation study, we found that overlap indices can accurately rank pairs of UDs in terms of the extent of overlap, but estimates of overlap indices are likely to be biased. The extent of the bias depended on sample size and the degree of overlap (UDs with a high degree of overlap resulted in statistics that were more biased [low]), suggesting that comparisons across studies may be problematic. We illustrate the use of overlap indices to quantify the degree of similarity among UD estimates obtained using 2 different data collection methods (Global Positioning Systems [GPS] and very high frequency [VHF] radiotelemetry) for an adult female northern white-tailed deer (Odocoileus virginianus) in north-central Minnesota. JOURNAL OF WILDLIFE MANAGEMENT 69(4):1346-1359; 2005

Helpers influence on territory use and maintenance in Alpine marmot groups

“…We used the Kernel Density Estimator (KDE), representing the intensity with which a given area is used (Fieberg & Kochanny 2005), because it allows for the estimation of multiple centres of activities (Worton 1989;Kernohan et al 2001). However, KDE is sensitive to the choice of the smoothing parameter h (Fieberg 2007).…”

“…To evaluate the effect of helpers on the overlap areas, we estimated the proportion of area shared by each group calculating the ratio between their overlap area and territory size (Fieberg & Kochanny 2005) using the kerneloverlap function in the "adehabitat" R package. Territorial aggressive interactions usually occurred between animals of two different social groups (Pasquaretta pers.…”

Escaping the oligotrophic gyre? The year-round movements, foraging behaviour and habitat preferences of Murphy’s petrels

“…We investigated whether space use differed between sexes and years for each breeding stage. We calculated observed overlap in core and general use areas using Bhattacharyya's affinity (BA), which is the most appropriate measure of quantifying similarity among UD estimates (Fieberg & Kochanny 2005). BA estimates range between 0 and 1, representing no similarity between UDs (or overlap) and identical UDs (or complete overlap), respectively.…”

Disentangling social networks from spatiotemporal dynamics: the temporal structure of a dolphin society

“…Although we did not obtain sufficient samples to estimate the home range for some individuals, we used the home range framework to estimate individuals' ranges and then used these polygons to generate a proxy of spatial overlap between pairs of individuals. We estimated spatial overlap using the utilization distribution overlap index (UDOI), a generalization of Hulbert's niche overlap concept (Fieberg & Kochanny 2005). A correlation analysis was performed between the association matrix (HWI) and the matrices of individual ranges and core areas of overlap (Mantel correlation, 1000 permutations).…”

“…A correlation analysis was performed between the association matrix (HWI) and the matrices of individual ranges and core areas of overlap (Mantel correlation, 1000 permutations). In addition, we performed the same correlation analysis between HWI and a probabilistic measure of space sharing (PHR ij ), the probability of individual j occurring within the range of individual i (Fieberg & Kochanny 2005). Analyses were performed in the R environment (R Development Core Team 2011) using the adehabitatHR package (Calenge 2006).…”