Plants expanding into currently-occupied habitats face challenges from the current occupants. Among these challenges are the soil microbial communities that have been cultivated by the current occupants, and these "plant-soil feedbacks" from the current occupants' microbes can affect the performance of the newly-arriving plants. Here we explore how this kind of cross-species plant-soil feedback may affect the performance of non-grass plants in prairie restorations, which sometimes get stuck in an undesirable grass-dominated state. Starting from an inoculum of tall-grass prairie soil, we allowed four native grass species to cultivate the microbial communities in a pot-based greenhouse study. Into these grass-cultivated soils we planted individuals from nine different prairie forb and legume species representing low-, mid-, or high-conservation groups, and we monitored the performance of these response plants. We also conducted high-throughput DNA sequencing of bacteria and fungi from the bulk soil and rhizospheres of all plants in order to determine whether cultivation by different grasses affected the composition of microbes that colonized the rhizospheres of the target plants.
Results/Conclusions
The different grass species used in our study cultivated distinct soil microbial communities (R-square = 0.11, p = 0.002), and this led to variation in the rhizosphere communities of the response plants. For any given response plant, roughly 20-25% of the variance in its rhizosphere community was linked to the identity of the grass species that cultivated its soil. In general, a majority of the response plants grew better in grass-cultivated soil when compared to sterilized soil, but the effect of microbes on plant growth varied considerably across different pairs of cultivator and response species. For example, Rudbeckia hirta, a low conservation group species, performed more poorly in grass-conditioned soil than in sterile soil, while Symphyotrichum laeve, a high conservation group species, benefitted from microbes cultivated by Bouteloua gracilis and Elymus virgincus but was harmed by microbes from the other two grasses. There was no clear pattern in plant response to grass cultivation across the different conservation groups of response plant species. Our results indicate that soil cultivation by one plant species can persist to affect the performance of subsequent plants.