Quantitative metrics of overlaps in Grinnellian niches: advances and possible drawbacks

摘要: Aim  Studies of environmental niche shift/niche conservatism that are based on species distribution modelling require a quantification purity and potential overlap. Although various metrics have been proposed for this task, no comparisons their performance available yet express the linearity range shifts error-proneness. Herein, we assess six overlap using three sister pairs plethodontid salamanders as well artificial to test curves, impacts varying sizes study area sizes. Location  North America, environments. Methods  Species models were performed with Maxent, created in R environment. Potential distributions extents compared Bray–Curtis distance BC, Schoener's D, two different modifications Hellinger Imod, Icor, Pianka's O Horn's R. Niche overlaps ecological space linear discriminant analyses principal components. Results  Simulations revealed strong variations metrics. In species, BC D best, followed by O, but modified Imod showed nonlinear slope truncated range. Furthermore, simulations suggest that, proportionally small large grids, an inclusion high proportion grid cells low occurrence probabilities representing background noise may bias assessments overlaps. Main conclusions  Both salamander examples best suited compute from derived models. However, like all analysed metrics, both seriously affected numbers where probably absent, i.e. probabilities. Therefore, pre-processing eliminate grids is highly recommended.