OOS 18-1
Plant species coexistence driven by negative plant-soil feedback: Theory and tests
There is growing evidence that pathogens play an important role in mediating plant species coexistence and in maintaining plant diversity. Soil pathogens have been found to be particularly important and much of this evidence has emerged through the now common practice of evaluation of plant-soil feedbacks. Basic theory of plant-soil feedbacks identify that negative feedbacks can contribute to local scale plant species coexistence, even between strong competitors. Plant-soil feedbacks have been found to be commonly negative, with accumulation of host-specific soil pathogens being likely causes. But are these negative feedbacks sufficient to overcome competitive differences between plant species?
I present a test of the relative importance of plant-soil feedbacks and interspecific competition using a pair of co-occurring grass species, Anthoxanthum odoratum (non-native) and Panicum sphaerocarpon (native). We had previously demonstrated negative feedback between these species and this negative feedback is due to both host-specific pathogens and host-specific changes in rhizosphere composition. A series of greenhouse experiments evaluated the strength, consistency, and robustness of negative feedback. Competitive outcomes were evaluated by growing these plant species in mixtures at a range of densities and frequencies, while manipulating the soil community. The frequency and density dependence of plant relative growth rates was evaluated using general linear models and compared to model predictions.
Results/Conclusions
We consistently found strong negative feedback between Anthoxanthum and Panicum, and that this negative feedback was robust to cross training and was consistent whether the plants were grown singly or in mixture. We found Anthoxanthum was a better competitor than Panicum overall. There was significant negative frequency dependence in plant performance at intermediate plant densities in live soils but not in sterile soils. Negative frequency dependence was enhanced in trained soil communities. In addition, we found significant overyielding in live soil, but not in sterile soil, and overyielding was enhanced in trained soils.
These results are consistent with soil microbial dynamics, and the dynamics of soil pathogens in particular, being more important than resource competition for plant species coexistence. Simulations predict that Anthoxanthum would exclude Panicum in the absence of soil microbial dynamics, while the two species would coexist with soil microbial dynamics. This intensively studied system provides evidence that dynamics occurring within the soil community can play a critical, though easily overlooked, role in plant community processes, including maintaining plant diversity.