Linking local and regional metacommunity dynamics to diversity outcomes

Background/Question/Methods

The metacommunity concept represents a major advance in ecology because it provides a framework to link ecological processes with biodiversity patterns at multiple scales. Metacommunities consist of assemblages of interacting species, typically linked over broad spatial scales. The structure of a metacommunity is an emergent property of demographic processes that are influenced by the habitat at local and regional scales, and is represented by measures of local (alpha) and regional (gamma) diversity, and species turnover (beta-diversity).  Our first objective was to simulate metacommunities under a range of dynamic constraints that represent alternative metacommunity hypotheses (species sorting, mass effects, patch dynamics, neutral models) to quantitatively assess how patch similarity, inter-patch dispersal, and species similarity relate to diversity outcomes.  Our second objective was to use simulated diversity patterns as a template to interpret the diversity observed in metacommunities from Long Term Ecological Research (LTER) sites.  These in situ metacommunities represent organisms with a large range in dispersal abilities (bacteria to fishes) from a range of ecosystem types (e.g., Florida Coastal Everglades [FCE], McMurdo Dry Valleys [MCM]).  We used an iterative, spatially explicit, lottery model to simulate metacommunity assembly under scenarios representing different metacommunity hypotheses.  We used variation partitioning to quantify the amount of beta-diversity in the metacommunity related to environmental and spatial filters for simulated and in situ (observed LTER) data sets.

Results/Conclusions

Simulated metacommunities showed that when community assembly was controlled solely by species sorting, beta-diversity was primarily correlated with environmental variation.  In a neutral scenario, dispersal limitation resulted in a relationship between spatial filters and beta-diversity independent of environmental variation.  Diversity outcomes for scenarios with interactions between local and regional dynamics and tradeoffs between dispersal and competitive traits were contingent on other factors.  Observed diversity patterns, based on fingerprinting 16S rDNA, showed cyanobacterial communities were sorted at finer spatial scales than bacteria.  We hypothesized that cyanobacteria, which form mats, may have larger effective propagule size than soil bacteria, which can become entrained in the atmosphere.  As would be predicted for passive dispersers, cyanobacteria community profiles showed finer scale spatial heterogeneity, whereas, bacteria community profiles correlated with broad spatial gradients in pH and soil moisture.  However, the relationship between propagule size and the balance of local and regional influences over diversity patterns was different for active dispersers (e.g., fishes at FCE).  The relationship between hypothesized mechanisms and different variation partitioning outcomes were illustrated using different simulation scenarios.