Improved β‐diversity estimators based on multiple‐site dissimilarity: Distinguishing the sample from the population. Issue 6 (13th April 2020)

• Record Type: Journal Article
• Authors: Fortin, Mathieu; Kondratyeva, Anna; Van Couwenberghe, Rosalinde
• Editors: Leprieur, Fabien
• Is Part Of: Global ecology & biogeography. Volume 29:Issue 6(2020)
• Page Start: 1073
• Page End: 1084
• Publication Date: 2020-04-13
• Subjects: beta diversity (β) -- jackknife variance -- multiple‐site dissimilarity -- population size -- sampling -- statistical estimators; Ecology -- Periodicals; Biogeography -- Periodicals; Biodiversity -- Periodicals; Macroevolution -- Periodicals; 577
• DOI: 10.1111/geb.13080
• Languages: English
• ISSNs: 1466-822X

Abstract

Abstract: Aim: β‐diversity is often measured through incidence‐based dissimilarity indices, such as the Simpson and Sørensen indices. Multiple‐site versions of these indices, which have been recently developed, enhance the accountability of global heterogeneity relative to the original formulations. However, they are known to be sensitive to the number of sites, which hinders the comparison of β‐diversity between two populations of unequal sizes. Moreover, the populations are never completely measured and the dissimilarity must be estimated from a sample of sites. Innovation: In this study, we propose adapted multiple‐site versions of the Simpson, Sørensen and nestedness indices that are population size‐independent as well as estimators of these indices. The properties of the indices and their estimators were tested through simulation studies. These simulation studies show that the adapted indices were population size‐independent. The estimators and their estimated standard errors were nearly unbiased for moderate sample sizes ( n ≥ 25). Main conclusions: These estimators of these adapted multiple‐site dissimilarity indices now make it possible to compare the β‐diversity of two populations of unequal sizes.