Understanding the underlying microbial causes of plant-soil feedbacks is necessary for the development of techniques to control invasive species in natural habitats. Lespedeza cuneata is a leguminous grassland invader that alters both plant and microbial community composition at invasion sites. The extent to which plant-soil feedbacks influence L. cuneata growth and competition with native plants remains unclear. We used a greenhouse experiment to examine the feedbacks that exist between L. cuneata and two native plants, L. virginica and Panicum virgatum. First, we conditioned sterile soil with a prairie soil inoculum by growing each plant species for two months. Second, we calculated feedbacks by comparing the biomass of plants grown in soil conditioned by each plant species. We sequenced bacterial and fungal communities from subsamples of soil taken throughout the experiment. We identified microbes that were specifically selected by each plant species, and we characterized microbes as beneficial or deleterious to each plant species using partial least squares regression of microbial community composition against plant biomass.
Results/Conclusions
Conditioning with different plant species had significant effects on plant biomass, and we could explain these effects based on the build up of particular beneficial or deleterious microbes. We found evidence of pathogen release for both L. cuneata and L. virginica when grown in P. virgatum-conditioned soil. However, L. cuneata-conditioned soil contained a high abundance of a Burkholderia bacterium that was beneficial for L. cuneata but deleterious for L. virginica, causing L. virginica to perform poorly in soil conditioned with L. cuneata. Conversely, P. virgatum benefits from the attraction of beneficial microbes that accumulate in P. virgatum-conditioned soil. Overall, the results from this study suggest that plant-soil feedbacks may be caused by shifts in abundance of a few influential microbes.