Spatial patterns and persistence of woody plant species in ecological communities

The role of stochastic effects and seed dispersal limitations in maintaining the diversity of a tree community is investigated by means of a forest growth simulator called TROLL. This simulator makes it possible to investigate impacts on the spatial distribution and the extinction probability of plant species. I present the results of a two-species competition scenario. Coexistence is found to be possible for competitively dissimilar species, and the criterion for this coexistence is given analytically in the reaction-diffusion approximation. A higher frequency of tree falls implies a smaller extinction probability of least adapted species. The multispecies model is also investigated. A mathematical quantity-the persistence function-permits measurement of the ability of a species to out compete other species locally. This quantity is compared with the average time to extinction, and it is computed for the TROLL simulator. I interpret the shape of this function in light of available information on the persistence function in simple interacting particle systems. These results indicate an analogy between the dynamics of forest communities and a simple discrete and spatially explicit model, the voter model. The ecological implications of this analogy are discussed and are extended to species-rich communities.