LAUSR

Microbial communities, like their macro-organismal counterparts, assemble from multiple source populations and by processes acting at multiple spatial scales. However, the relative importance of different sources to the plant microbiome and the spatial scale at which assembly occurs remains debated. In this study, we analyzed how source contributions to the foliar fungal microbiome of a C4 grass differed between locally abundant plants and soils across an abiotic gradient at different spatial scales. Specifically, we used source-sink analysis to assess the likelihood that fungi in leaves from Panicum hallii came from three putative sources: two plant functional groups (C4 grasses and dicots) and soil. We expected that physiologically similar C4 grasses would be more important sources to P. hallii than dicots. We tested this at ten sites in central Texas spanning a steep precipitation gradient. We also examined source contributions at three spatial scales: individual sites (local), local plus adjacent sites (regional), or all sites (gradient-wide). We found that plants were substantially more important sources than soils, but contributions from the two plant functional groups were similar. Plant contributions overall declined and unexplained variation increased as mean annual precipitation increased. This source-sink analysis, combined with partitioning of beta-diversity into nestedness and turnover components, indicated high dispersal limitation and/or strong environmental filtering. Overall, our results suggest that the source-sink dynamics of foliar fungi are primarily local, that foliar fungi spread from plant-to-plant, and that the abiotic environment may affect fungal community sourcing both directly and via changes to host plant communities.