Null model analyses are not adequate to summarize strong associations: Rebuttal to Ulrich et al. (2022).

Mainali, Kumar P.; Slud, Eric

This article introduces a new metric called the maximum likelihood estimator (MLE) alpha-hat, which measures the degree of association between species in ecological sites. The authors argue that traditional indices of co-occurrence can be inconsistent and sensitive to fixed margins of contingency tables. They provide analyses of real and artificial datasets to support the reliability and stability of alpha-hat. The article emphasizes the importance of accurately measuring the degree of association in co-occurrence analysis and addresses criticisms of their metric. The authors also discuss the concept of affinity and its relationship to probability and the number of sites in a dataset. They compare the reliability of affinity with standardized co-occurrence count and standardized Jaccard indices, arguing that affinity is a more reliable measure of association. The authors conclude that affinity is a valuable tool for analyzing joint species occurrence.

DISTRIBUTION (Probability theory); MATHEMATICAL formulas; GENERAL semantics

Journal of Biogeography, 2024, Vol 51, Issue 2, p263

0305-0270

Academic Journal

10.1111/jbi.14750