Hierarchical Reliability in Experimental Plant Assemblages

Dan F. B. Flynn,Bernhard Schmid,Jin-Sheng He,Kelly S. Wolfe-Bellin,F. A. Bazzaz
DOI: https://doi.org/10.1093/jpe/rtm004
2008-01-01
Journal of Plant Ecology
Abstract:AimsThe mechanism by which species richness affects variation in ecosystem functioning both within and among ecosystems remains a key question at the interface of community and ecosystem ecology. Statistical averaging (the smoothing of average system performance via consideration of additional components) and the insurance effect (reduced variation in system performance by inclusion of asynchronously varying components) predict that more diverse communities should vary less both between replicates and internally. We experimentally tested these theories in small plant assemblages.MethodsWe constructed plant assemblages modeled after old-field plant communities. We varied species richness, species composition and initial densities while holding functional group richness constant in replicate assemblages under glasshouse conditions.Important findingsThe inverse of the coefficient of variation of aboveground biomass production, a proxy measure of reliability, increased with higher diversity when examined at the level of the assemblage (i.e. among-replicate assemblages) but not at the levels of functional group or species. These stabilizing processes were weakest, in low-diversity, low-density assemblages. This experiment demonstrates the utility of hierarchical analysis of ecosystem reliability at the assemblage, functional group and species level.
What problem does this paper attempt to address?