Using coverage‐based rarefaction to infer non‐random species distributions

• DOI: 10.1002/ecs2.3745
• Publication Date: 2021-09-20T06:02:35Z
• Authors: Thore Engel; Shane A. Blowes; Daniel J. McGlinn; Felix May; Nicholas J. Gotelli; Brian J. McGill; Jonathan M. Chase
• Keywords: Null model; Rarefaction (ecology); Relative abundance distribution

Abstract Understanding how species are non‐randomly distributed in space and the resulting spatial structure responds to ecological, biogeographic, anthropogenic drivers is a critical piece of biodiversity puzzle. However, most metrics that quantify diversity (i.e., community differentiation), such as Whittaker’s β‐diversity, depend on sampling effort influenced by pool size, abundance distributions, numbers individuals. Null models useful for identifying degree differentiation among communities due structuring relative expected from effects, but do not accommodate influence sample completeness proportion given sample). Here, we develop an approach makes use individual‐ coverage‐based rarefaction extrapolation, derive metric, β C , which captures changes intraspecific aggregation independently size. We illustrate metric using spatially explicit simulations two case studies: (1) re‐analysis “Gentry” plot data set consisting small forest plots spanning latitudinal gradient North South America (2) comparing large high tropical forests Barro Colorado Island, Panama, with lower temperate Harvard Forest, Massachusetts, USA. find no evidence systematic latitude these sets. As it rooted theory explicitly controls completeness, our represents important advance over existing null aggregation. Potential applications range better descriptors biogeographic patterns consolidation local regional trends current crisis.