On the inference of positive and negative species associations and their relation to abundance

Abstract

The prevalence of rare species in ecosystems begs the question of how they persist. In a recent paper, Calatayuda et al. (CEA) provided a new hypothesis that rare species, in contrast to common species, share unique microhabitats and/or preferentially engage in mutualistic interactions. CEA support this hypotheses by reconstructing association networks from spatially replicated abundance data finding that rare species are over-representing in positive association networks while common species are over-representing in negative association networks. However, the use of abundance and co-occurrence data to infer true species associations is difficult and often inaccurate. Here, I show that the finding of rare species being more represented in positive association networks can be explained by statistical artifacts in the inference of species associations from abundance data. I caution against the inference of ecological association networks from abundance data alone.

Competing Interest Statement

The authors have declared no competing interest.

Footnotes

* Added an analysis of the independent swaps algorithm and improved the abstract.

* https://github.com/ajrominger/RarePlusComMinus