Divergence in Diversity and Composition of Root-Associated Fungi Between Greenhouse and Field Studies in a Semiarid Grassland.

Investigations of plant-soil feedbacks (PSF) and plant-microbe interactions often rely exclusively on greenhouse experiments, yet we have little understanding of how, and when, results can be extrapolated to explain phenomena in nature. A systematic comparison of microbial communities using the same host species across study environments can inform the generalizability of such experiments. We used Illumina MiSeq sequencing to characterize the root-associated fungi of two foundation grasses from a greenhouse PSF experiment, a field PSF experiment, field monoculture stands, and naturally occurring resident plants in the field. A core community consisting < 10% of total fungal OTU richness but > 50% of total sequence abundance occurred in plants from all study types, demonstrating the ability of field and greenhouse experiments to capture the dominant component of natural communities. Fungal communities were plant species-specific across the study types, with the core community showing stronger host specificity than peripheral taxa. Roots from the greenhouse and field PSF experiments had lower among sample variability in community composition and higher diversity than those from naturally occurring, or planted monoculture plants from the field. Core and total fungal composition differed substantially across study types, and dissimilarity between fungal communities did not predict plant-soil feedbacks measured in experiments. These results suggest that rhizobiome assembly mechanisms in nature differ from the dynamics of short-term, inoculation studies. Our results validate the efficacy of common PSF experiment designs to test soil inoculum effects, and highlight the challenges of scaling the underlying microbial mechanisms of plant responses from whole-community inoculation experiments to natural ecosystems.