Plant abundance, but not plant evolutionary history, shape host specificity in foliar fungal endophytes

Austen Apigo, Ecology, Evolution, and Marine Biology, University of California, Santa Barbara, Santa Barbara, CA and Ryoko Oono, Department of Ecology, Evolution and Marine Biology, University of California, Santa Barbara, Santa Barbara, CA

Background/Question/Methods
There has been limited progress towards identifying general mechanisms that explain why horizontally-acquired microbial symbionts associate with some hosts but not others. In this study, we explored host specificity in foliar fungal endophytes that passively disperse and colonize the photosynthetic tissues of all plant lineages. Here, we tested two leading hypotheses, the Common Host and Phylosymbiosis Hypothesis, that propose endophyte host specificity is determined by plant abundance or plant evolutionary relatedness, respectively. We sampled all plant species within a single community and quantified host specificity with univariate metrics that consider host specificity in the context of plant diversity (structural specificity), plant spatial structure (beta-specificity) and plant phylogenetic relationships (phylogenetic specificity).
Results/Conclusions
In support of the Common Host Hypothesis, plant species abundance was correlated with endophytes that occupied fewer plant species (higher structural specificity) and were consistently found in the same plant species across the landscape (higher beta-specificity). There was no relationship between plant phylogenetic distance and endophyte community dissimilarity, thus we did not find support for the Phylosymbiosis Hypothesis. Instead, we found that endophyte community composition significantly varied among plant species, families and major groups supporting a plant identity effect, known as phylospecificity. In particular, endophytes in angiosperm lineages occupied narrower phylogenetic breadths of plants (higher phylogenetic specificity) relative to endophytes within conifer and fern lineages. Future studies should consider using the univariate metrics we propose in conjunction with commonly-used multivariate techniques to understand how the ubiquitous relationships between hosts and symbionts are structured.