OOS 35-2 - Using plant-soil feedback theory to predict changes in plant community composition in a changing world

Background/Question/Methods
Climate change can alter ecological communities through changes in biotic interactions. By describing reciprocal interactions between plants and soil microbes, plant-soil feedback (PSF) has emerged as a powerful framework for predicting plant species coexistence and community dynamics. However, the effects of climate change on PSF are relatively unknown. Importantly, climate-mediated changes in PSF could lead to changes in plant community composition or the erosion of plant diversity. First, we used a greenhouse experiment with eight grassland plant species to test how PSF changes in response to water availability and parameterized a simulation model to predict changes in plant community composition. Next, we sequenced fungal communities from the greenhouse experiment and conducted follow-up experiments to help determine what changes in microbial communities might be influencing responses. Then, we conducted a reciprocal soil transplant experiment in the field to determine whether PSF responses are similar across the landscape.

Results/Conclusions
In the greenhouse experiment, we found that coexistence-stabilizing negative PSF at drier than average conditions shifted to coexistence-destabilizing positive PSF under wetter than average conditions (P < 0.05). A simulation model parameterized with the experimental results supports the prediction that more positive PSF accelerates the erosion of diversity within communities while decreasing the predictability in plant community composition. Fungal community composition was more dissimilar among plant species in the high watering treatment relaitve to the low watering treatment (P < 0.05), and an additional experiment suggested that the specificity of microbe effects may be greater in wetter conditions (P < 0.05), both of which can lead to stronger PSFs. In the field, PSF strongly responded to the location of the soil common garden (P < 0.05) but not the original soil origin (P > 0.01). This suggests that PSFs across a landscape may respond similarly to changes in climate, aiding our ability to predict changes in plant community composition driven by PSFs.