Food partitioning and the evolution of non-randomly structured communities in tailed amphibians: a worldwide systematic review

The evolutionary response to interspecific competition is the partitioning of the resources used by the competing species. At the community level, the ultimate outcome of resource partitioning determines how communities evolve in terms of their species composition, explaining the distribution and abundance of organisms within natural assemblages. Despite amphibians having attracted considerable attention from population biologists, niche partitioning patterns have never been explored with a systematic approach in tailed amphibians (order Urodela). To fill this gap we tested whether Urodela communities are assembled in a non-random fashion along the trophic niche and which determinants (environmental factors and interspecific interactions) explained the observed patterns. We used a suite of statistical tools, including Generalized Linear Models and Null Models with Monte Carlo simulations, in order: (1) to explore the occurrence of non-random assemblage structure among the Urodela communities, and (2) to test environmental features (habitat) and interspecific interactions (exemplified by the overlap in species’ body size) as predictors of non-random assemblage composition. Our systematic review pointed out that: (1) the evolution of non-random structure in Urodela assemblages was determined by both biotic and abiotic factors, and (2) the food niche plays a key role in the coexistence of Urodela syntopic species. The considered assemblages showed a non-random structure influenced by diet, habitat and morphological features acting in an interactive fashion. Lawlor's RA2 algorithm was more effective than any other randomization algorithms to reveal community structure based on interspecific competition. The assemblage structure was influenced by the niche breadth of the species in the generalist–specialist nature of the resource utilization: the much higher the niche breadth similarity among species, the most likely was a non-random structure in the assemblage. In an aquatic habitat, where species can partition the space in three dimensions, assemblages composed by species overlapping in body size and generalist in feeding behaviour show non-random structure. Conversely, in terrestrial habitats, probably due to the lack of vertical stratification, differently sized species composed assemblages in a non-random fashion