Neutral metacommunity clustering and SAR: River basin vs. 2-D landscape biodiversity patterns

Abstract

Abstract Moving from the analysis of the spatial distribution of fishes and big/small trees of the Mississippi Missouri River System, I evidenced and modeled with a neutral metacommunity model the power-law exceedence probability of the cluster-size of species and the species–area relationship (SAR). The slopes of the power-law distribution of the cluster-size P( CS ≥ c) and of the SAR vary for each taxa and life-stages underlying different spatial organization. A clear relationship exists between the slope z of the SAR and the slope ϵ of P( CS ≥ c), that is dependent on the ecosystem topology and shape, and on the dispersal kernel function. The heterogeneity of the environmental features leads to the formation of smaller clusters than in the ideal homogeneous scenario. P( CS ≥ cs) declines to an exponential distribution in dispersal-limited scenarios for which the effect of the environmental heterogeneities is stronger, the probability distribution of the local and pairwise species richness similarity is a lognormal function and the occupancy-rank is concave upward. The clustering of species has been studied on other real and artificial river networks and on 2-D non-fragmented landscapes. River networks have smaller clusters than 2-D landscapes for the same ecological dispersal scenario however the range in which P( CS ≥ c) holds is larger. The higher the elongation of the ecosystem the bigger the LSR, and the smaller the mean cluster-size. River networks due to the larger link-diameter than 2-D landscapes with the same domain are potentially more robust ecosystems, for example against invasion of invasive/exotic species and pathogens. The ecological ratio between the mean dispersal parameter and the average diameter is introduced as useful tool to compare biodiversity patterns. The influence of the dendritic structure of the river network has been reinforced. Nonetheless P( CS ≥ c), is found not invariant across different scales, and coarse-graining levels of the ecosystems. The study enhances the robustness of the stochastic birth-depth process in shaping biodiversity patterns, however I underline the strong influence of the dispersal parameter in the assemblage of species. The understanding of the relative influence of exogenous and endogenous variables is important to detect climatic and anthropic effects on the cluster-size distribution of species.