Plant traits play an active role in shaping the composition of their microbiomes, the diverse communities of bacteria and fungi that live on and in plant tissues and influence plant productivity and health. For example, traits related to the production of secondary compounds or the presence of physical barriers can limit colonization by some microbes but not others. As a result, plant microbiome composition varies both among and within plant species and can be viewed as a complex heritable trait. Environmental context is also a strong driver of microbiome composition and, as a result, complex genotype-by-environment (GxE) interactions are likely to underlie plant microbiome community assembly. However, direct evidence for GxE interactions governing plant microbiome assembly are lacking, potentially limiting the ability to predict how environmental change will affect plant microbiome composition. To explicitly tease apart the relative impact of plant genes, environmental conditions, and their interactions we used a reciprocal common garden approach, focusing on the foliar fungal microbiome of the model woody tree species, Populus trichocarpa. We used Illumina metabarcoding to characterize fungal community composition in the leaves of 1000 previously sequenced P. trichocarpa genotypes, growing in two locations separated by ca. 200 km in eastern Oregon.
Results/Conclusions
We found that more similar P. trichocarpa genotypes (e.g., originating from the same river basin) tended to have more similar foliar microbiome composition in both common gardens. However, fungal community composition varied substantially between location. In addition, the specific pattern of fungal community similarity among plant genotypes was not consistent between the two gardens, suggesting that GxE effects do influence P. trichocarpa leaf microbiome assembly. Specifically, fungal community composition was correlated with the mean annual temperature of the genotype origin in one location, but not the other. In order to provide a mechanistic explanation for these results we are currently developing novel multivariate GWAS methods to identify plant genes that dictate microbiome composition in different environments. By linking the fields of plant genomics and microbial community ecology, we expect this work to facilitate progress towards more generalizable predictions about the links between plant genes, environmental conditions, and microbiome composition.