Metacommunity structure of bacterioplankton communities along an urbanization gradient

Background/Question/Methods

The statement of “everything is everywhere, but the environment selects” was used for many years to explain microbial distribution, assuming the importance of local factors driving the colonization and establishment of microbial communities. Nevertheless, this idea has been challenged due to new findings of regional factors structuring the distribution of microorganisms (e.g. dispersal limitation). In this context, metacommunity theory provides a framework to study how community patterns at local scales can be related to processes at broader scales. Thus, here we studied bacterioplankton communities of 57 shallow ponds in Belgium along an urbanization gradient in a metacommunity perspective. We used a hierarchical sampling strategy with two spatial scales (200m - local and 4km - regional) and three levels of urbanization within each scale based on the percentage of built-up area present in the surrounding of each pond. This approach was used to disentangle the local and regional factors shaping the communities. We characterized the bacterial community composition (BCC) by 454 pyrosequecing of the 16s rRNA gene. We used variance partitioning procedure to estimate the importance of the regional factors (matrix based on geographical distances), local conditions (29 different environmental variables) and the urbanization categories on BCC.

Results/Conclusions

We identified 242 operational taxonomic units (OTUs) for the total dataset. Using the variation partitioning analyses, the three levels of urbanization explained a small portion of the variation on BCC for both spatial scales: regional (adj R² = 0.019, P<0.05) and local (adj R² = 0.011, P<0.05). When we included the local (environmental variables), regional factors and the urbanization categories (for both scales: local and regional), we found a total of 12% of the variance explained for all the variables included in the model. Most part of the variation explained was related to the local conditions (adj R² = 0.093, P<0.001) with four variables being selected to the model: conductivity, pH, alkalinity and total phosporus. Based on the results, BCC in shallow ponds seems to be structured by processes occuring at local scales, thus environmental conditions tend to be more important than regional processes. Regarding the metacommunity perspective, our results emphazise species sorting as the main process to shape bacterioplankton communities in small spatial scales (<4km).