Natural selection in urban environments: The role of plant-animal interactions

Background/Question/Methods

Some of the most rapidly increasing habitats are those dominated by humans. The ecological consequences of land-use change associated with housing development (hereafter suburbanization) include lower native species diversity, increased density of the remaining species, shifts in community composition, and changes in hydrology and nutrient cycling. These ecological changes likely have implications for contemporary selection pressures experienced by native species. Yet, little work to date addresses the combined ecological and evolutionary consequences of suburbanization. Here we develop a conceptual model of how suburbanization is likely to affect natural selection on native species. We focus on how suburbanization alters plant-pollinator and plant-herbivore interactions, with plants as the target of selection, because these interactions have played a central role in natural selection and evolution of plant and floral traits. Moreover, recent studies in urban ecology document impacts on the richness and abundance of plants and the animals that interact with them. We apply components of the conceptual model to a case study in the southeastern United States. Our work there addresses how suburbanization affects plant-animal interactions and patterns of natural selection on floral attractive and defensive traits in the native vine Gelsemium sempervirens. 

Results/Conclusions

We found that suburban forests harbored significantly more bee pollinators than natural forests. However, plants growing in suburban forests also received significantly more nectar robbing and floral herbivore damage than plants in natural forests. We found no evidence that suburbanization altered how mutualists and antagonists responded to Gelsemium floral or defensive traits. However, some floral traits affected the intensity of species interactions across suburban and natural forests, suggesting that species interactions may be important in driving differential trait selection as a function of suburbanization. Moreover, we found significant differences in trait expression between suburban and natural forests that partially matched predicted evolutionary responses to selection driven by species interactions. However, genetic correlations among traits may constrain the degree to which Gelsemium in suburban vs. natural forests can respond to any one interactor and may prevent the evolution of optimal trait expression. The conceptual model and results from this case study suggest that consideration of the broader context in which organisms interact, in both suburban and natural areas, is fundamental to our understanding the forces that shape contemporary selection pressures in native species.