QUANTIFYING SPATIAL PATTERN WITH EVENNESS INDICES

Quantifying the spatial distributions of organisms in simple and meaningful ways is important for understanding the ecology and habitat needs of species subject to anthropogenic disturbances. Most multi‐species conservation and management programs do not yet account for the variation of space‐use patterns exhibited by species preferring the same habitat type. To measure species' space‐use patterns as a step toward determining habitat needs, we suggest using evenness indices. Although commonly used in characterizing communities (i.e., as a measure of species diversity), these indices are suitable for characterizing many other ecological patterns, including patterns of site use by individuals. We investigate the statistical properties of five indices (Camargo's index of evenness, E′; Simpson's index of evenness, E1/D^; Lloyd's index of mean crowding, J; Smith‐Wilson index Evar; and Dispersion index, DL, a variant of the Shannon diversity index) to evaluate their utility for quantifying broad‐scale spatial patterns of migratory shorebirds. We use a Monte Carlo simulation approach to compare these indices for their (1) ability to characterize a wide range of spatial patterns (from even to patchy); (2) ability to discriminate among distributions; and (3) robustness to incomplete sampling. In addition, we compare the ability of these indices to characterize spatial dispersion for four species of shorebirds migrating through the United States: Killdeer ( Charadrius vociferus ), Semipalmated Plover ( C. semipalmatus ), Sanderling ( Calidris alba ), and Red Knot ( C. canutus ). Overall, we recommend Camargo, Simpson, and Lloyd indices for quantifying spatial dispersion. All three indices gave precise and unbiased estimates once 35 or more sites (3.5% of 1000 simulated sites) were sampled randomly, regardless of the degree of patchiness. Nonrandom sampling resulted in biased estimates until a much greater proportion of sites (at least 30–50%) were sampled, highlighting the importance of site selection in sampling programs. Our analysis of shorebird spatial patterns revealed that evenness indices discriminate well among the species we considered, ranking Killdeer as most dispersed and Red Knot as most aggregated. Our results agree with earlier (independent) assessments of these species' migration strategies and illustrate how simple univariate metrics may be useful tools for characterizing complex spatial patterns.