Local factors control the community composition of cyanobacteria in lakes while heterotrophic bacteria follow a neutral model

1. Neutral community models are derived from the proposition that basic probabilities of species loss (extinction, emigration) and gain (immigration, speciation) explain biological community structure, such that species with many individuals are very likely to be widespread. Niche models on the other hand assume that interactions between species and differential resource use mediate species coexistence, thus invoking environmental factors to explain community patterns. 2. In this study, we compared neutral and niche models to test how much of the spatial variability of assemblages of heterotrophic bacteria and phytoplankton in 13 lakes they could explain. Analysis of phytoplankton was restricted to cyanobacteria, so that they could be studied with the same molecular fingerprinting method, automated ribosomal intergenic spaces analysis (ARISA), as heterotrophic bacteria. We determined local biotic and abiotic lake variables as well as lake age, glacial history and distance between sites. 3. The neutral community model had a good fit to the community composition of heterotrophic bacteria (R-2 = 0.69), whereas it could not produce a significant model for the community composition of cyanobacteria. 4. The community composition of cyanobacteria was instead correlated to environmental variables. The best model, a combination of total organic carbon, biomass of eukaryotic phytoplankton, pH and conductivity, could explain 8% of the variation. In contrast, variation in the community composition of heterotrophic bacteria was not predicted by any of the environmental variables. Historical and spatial variables were not correlated to the community composition of either group. 5. The pattern found for heterotrophic bacteria suggests that stochastic processes are important. The correlation of cyanobacteria with local environmental variables alone is consistent with the niche model. We suggest that cyanobacteria, a group of organisms containing bloom-forming species, may be less likely to fit a neutral community model, since these blooms are usually triggered by a particular combination of environmental conditions.