Neutral and non-neutral macroecology

Because of the multiscalar nature of processes underlying biodiversity dynamics, macroecology has emerged as a discipline that seeks to build an understanding of this complexity by examining statistical patterns in large assemblages of species in geographic space and ecological time. Models that assume individual organisms within trophically defined assemblages are ecologically equivalent can produce many patterns identified by macroecology. Neutral models predict two important dynamical patterns that can be tested in real assemblages. First, they predict that species diversity will decline within an assemblage over time. The rate of this decay in species diversity can be predicted from estimates of migration rates from a "metacommunity" or species pool. Second, neutral models predict a divergence of species composition among local communities over time. The rate and degree of divergence among communities also depend on the migration rate. The few studies that have been done to date imply that the rate of migration in real species assemblages is much lower than that required to explain the degree of community similarity maintained in space and time. There are at least two alternative ways to extend neutral models to incorporate more biological realism. First, competitive asymmetries among species may be introduced to allow for the possibility that individuals of some species may have an advantage in replacing individuals that die. Second, environmental heterogeneity can be introduced by assuming sites available to individuals differ in quality to individuals of different species. The neutral model, because of its conceptual simplicity and rigor, should be considered as a null model for baseline comparison to actual patterns of distribution, abundance, species composition, and beta diversity. (C) 2004 Elsevier GmbH. All rights reserved.