A review of null models in community ecology: a different robust viewpoint for understanding statistical community ecology

One classic definition of ecology is that it is the science that studies the distribution and abundance of biotic components and their relationship with abiotic components. The use of statistical tools is very important for understanding ecology, especially the distribution and abundance of biotic components. The classic statistical viewpoint was that an ecological community (an interaction of different species in defined time and space) has a determined structure due to biotic and abiotic interactions. Nevertheless, this classic viewpoint has the risk of proneness to type I errors or “false positives”. In this situation, null models were proposed that have the premise that community ecology is random, meaning the absence of structure, and the null hypothesis for these models is the absence of regular structured patterns. The present study is a review of null models and their application to aquatic environments. These null models include three main models: for species co-occurrence, asserting that species associations are random; for size overlap, asserting that the size structure of species in the community is random, as a strategy for use of ecological niche; and for niche overlap, asserting that species in a community can share a defined ecological niche with consequent interspecific competition.