Niche differentiation is a classical idea in community ecology, forming the basis of our understanding of species coexistence. Niche differentiation predicts that species traits should be over-dispersed: that is, species should be more dissimilar than expected by chance. However, the empirical record shows mixed trait-based evidence for niche differentiation. Here we raise the possibility that this could be due to confounding forces such as habitat filtering, immigration, and competitive asymmetries between species, and investigate whether new metrics can overcome this interference by looking for specific patterns of abundance structure along the niche axis. We perform an exploratory study using a suite of niche dynamics models where we know that niche differentiation is at play. These niche scenarios incorporate various confounding factors or biological complexities hypothesized to modulate trait pattern. We check whether four new metrics specifically designed for the purpose can distinguish species assemblages predicted by these models from neutral assemblages or assemblages where only fitness differences but no niche differentiation occur. We also examine which particular confounding forces have the strongest potential to impact detection of niche differentiation.
Results/Conclusions
Our metrics were largely capable of distinguishing between niche scenarios and neutral assemblages, but detected a wide variation in the strength of niche differentiation pattern across our niche scenarios. Pattern tends to be stronger when more niches are available for species to occupy, when there are no differences in species’ ability to grow in the absence of resource limitation (fitness differences), when resources do not go extinct, and when the competitive effect between pairs of species is symmetric. Finally, we found that our metrics can distinguish between assemblages with and without niche differentiation even when both show functional underdispersion relative to neutral assemblages. We conclude that our metrics can be further developed to be used in empirical tests of niche differentiation, but warn that the trait signal reflective of niche differentiation can be lost due to other forces of community assembly or when competitive asymmetries are strong.