Effects of predation and variation in species relative abundance on the parameters of neutral models

Hubbell (2001) proposes that random demographic processes (i.e., neutral dynamics) can explain observed levels of variation in the richness and abundance of species within and among communities. Hubbell’s neutral models have drawn attention because they reproduce several characteristic features of natural communities. But neutral models are criticized for ignoring nonrandom processes known to cause species densities to fluctuate. We parameterized neutral models using the population counts of 64 species of aquatic invertebrates collected from 49 discrete rock pools over a 13 year period. We used Hubbell ‘s numerical modeling approach to evaluate the effect of natural population fluctuations on the parameter settings. We also analyzed the effect of observed variation on the species proportional abundance predicted by neutral models. We find that observed levels of variation in abundance are much higher than predicted by neutral models, forcing estimates of the migration probability, m, and fundamental biodiversity parameter, θ, to fluctuate over time. Much of the observed variation is mediated by predator-prey interactions. Low predator densities are associated with fewer species and less even relative abundances of species, resulting in lower estimates of m and θ compared to periods of high predator densities. Our results show that by assuming an identical survival probability for all species, neutral models misrepresent substantial aspects of community dynamics.