Monday, August 3, 2009

PS 2-16: Dynamic neutral model, null hypotheses for large-scale patterns of diversity and distribution

Paulina Trejo1, Hector Arita2, and Fabricio Villalobos Camacho1. (1) Instituto de Ecología, UNAM, (2) Instituto de Ecologia, UNAM

Background/Question/Methods

In macroecology there has been a long tradition that works with null models for the study of biodiversity and geographic distribution, but this kind of work does not imply a mechanism of community assembly. In this study we use a neutral model to incorporate a dynamic mechanism that generates quantitative predictions to be compared with patterns of diversity and geographic distribution of species in a real data set as a null model. The model incorporates a demographic process with speciation and dispersal parameters where all the individuals have identical demographic probabilities, in a spatial context. We compare the results of species distributions generated by simulations with the geographic distribution of the 143 species of the family Phyllostomidae, a neo-tropical bat family limited to the American Continent. The geographic analysis was made for the entire family at a continental scale considering two recently described concepts, diversity field, species diversity of all sites in which a particular species occurs and dispersion field, set of geographic ranges of species occurring in a given site.

Results/Conclusions

Neutral dynamics produces spatial patterns in species richness and occupancy distribution showing a particular structure, right skewed frequency distribution of range size and heterogeneous distribution of richness among sites. Speciation process is essential for the maintenance of diversity in the model; the distance of dispersion determines the distribution of diversity, the range size and the overlapping of species. The empirical results present the same mean values that the simulations and variation in the extreme values of the distribution of the dispersion and diversity fields, that as effect of the specific composition of each site. These results are evidence that neutral models are capable of providing appropriate null hypotheses for the evaluations of macroecological questions.