Null model tests of clustering of species, negative co-occurrence patterns and nestedness in meta-communities

ABSTRACT

We propose tests for patterns in meta-community structure. The tests for clustering and nestedness of the occurrences of species and negative co-occurrence patterns provide four important innovations. Firstly, they are not restricted to the analysis of communities along one-dimensional gradients or to the main axis of variation. Secondly, abundance data can also be considered in the null model whereas most previous approaches could consider only presence/absence data. And thirdly, habitat suitability and spatial autocorrelation can be incorporated in the null model so that patterns that might be caused by biotic interactions can be distinguished from patterns which are the result of differences in the suitability or accessibility of sites for the examined organisms. Finally, the test for nestedness is also appropriate if there is more than one set of nested subsets. A re-analysis of 35 data sets with these tests showed the importance of considering the autocorrelation of the occurrences of species in analyses of meta-community structure and demonstrated the advantage of abundance data for tests of clustering of species. With abundance data it could be shown that there is a significant clustering of species, i.e. there are positive associations of species in most meta-communities, even if an environmentally or spatially constrained null model is used for the test. Co-occurrence patterns that might indicate interspecific competition were found in many of the analysed presence/absence data sets. Surprisingly the analysis of abundance data sets provides less evidence for interspecific competition. A hierarchical organization of communities, i.e. nestedness, turned out to be a rare pattern, if the autocorrelation of the occurrences of species is considered.