Examining the link between competition and negative co-occurrence patterns
1 May 2019Brazeau, Hannah; Schamp, Brandon
Negative species co-occurrence patterns have long intrigued ecologists because of their potential link to competition. Although manipulative field experiments have consistently revealed evidence of competition in natural communities, there is little evidence that this competition produces negative co-occurrence patterns. Evidence does suggest that abiotic variation, dispersal limitation, and herbivory can contribute to patterns of negative co-occurrence among species; it is possible these influences have obscured a link with competition. Here, we test for a connection between negative co-occurrence and competition by examining a small-scale, relatively homogeneous old-field plant community where the influence of abiotic variation was likely to be minimal and we accounted for the impact of herbivory with an herbivore exclosure treatment. Using three years of data (two biennial periods), we tested whether negatively co-occurring pairs of species, when occasionally found together, experienced asymmetric abundance decline more frequently than positively co-occurring pairs, for which there is no such expectation. We found no evidence that negatively co-occurring pairs consistently suffered asymmetric abundance decline more frequently than positively co-occurring pairs, providing no evidence that competition is a primary driver of negative co-occurrence patterns in this community. Our results were consistent across control and herbivore exclosure treatments, suggesting that herbivores are not driving patterns of negative species co-occurrence in this community. Any influence of competition or herbivory on co-occurrence patterns is small enough that it is obscured by other factors such as substrate heterogeneity, dispersal, and differential species responses to climatic variation through time. We interpret our results as providing evidence that competition is not responsible for producing negative co-occurrence patterns in our study community and suggest that this may be the case more broadly.