A metacommunity approach to urban plant community assembly

Background/Question/Methods

There is great value to conducting basic ecological research in urban environments. Classical ecology, however, has avoided human-dominated landscapes in favor of “pristine” or “more natural” landscapes, and traditionally characterized human inputs to natural systems as a “disturbance.” It is becoming clear, however, that characterizing all urban habitats as ‘disturbed’ does not accurately describe the complex patterns of species turnover that result from how humans influence the interacting processes that shape community assembly. Herbaceous plant surveys were conducted in Baltimore, MD, during the summer of 2009 in habitat patches that fall along a gradient of human management from “low” (the building footprint section of vacant lots) to “medium” (remnant back gardens in vacant lots).  These sites were resurveyed during the summer of 2012, and the management gradient was extended to include a “high” category of human management (active community gardens).  Using a metacommunity framework, we hypothesize that humans influence both local and regional processes, and that this results in patterns of species diversity and composition that are not well explained by models that do not take human impacts into account.  We ask 1) does the spatial partitioning of species diversity vary between management groups? and 2) How does species diversity vary with environmental distance and geographic distance between sites?

Results/Conclusions

We show significant differences in the spatial partitioning of diversity between “low” and “medium” management categories.  “Low” management patches showed high local diversity but low turnover between local communities.  In contrast, “medium” management patches showed significantly lower local diversity than the “low” management patches, but higher turnover between patches, despite the fact that both patch-types shared a common regional species pool. There was no relationship found between geographic distance between sites and site compositional similarity, which we suggest may result from mass effects caused by high levels of dispersal between sites.  We suggest alternative hypotheses to explain these patterns of diversity, and present the results of preliminary experimental work to test the potential processes that shape community assembly in this system. We contend that urban ecosystems are an excellent study system for applying and extending basic ecological theory, and that a process-based and spatially-explicit approach to explaining patterns of diversity is essential to better understand the role of humans in shaping urban ecosystems.