Above and Belowground Microbes Are Associated With Variation of the Leaf Metabolome; But Herbivory Has Nominal Effects

Abstract Background Microbes are associated with nearly every plant surface and affect the host’s functioning. While many studies to date have investigated the composition of the plant microbiome and compared it with the plant’s metabolome (a measure of the host functioning), a holistic understanding which considers interactions both above and belowground is lacking. In this work, we hypothesise that soil is a reservoir for both root and leaf microbes, and that herbivory disrupts the regulation of the leaf microbial community. We further hypothesise that plant-associated microbial communities will correlate with the leaf metabolome, but the root microbiota will better correlate with the leaf metabolome than the leaf microbiota. To test these hypotheses, fungal and bacterial communities of herbivore-damaged and undamaged leaves, roots and the surrounding soil was characterised across 15 Plantago major populations sampled from across geographical and environmental gradients from Denmark. Microbial communities were then compared against untargeted and targeted (anti-herbivory/microbials) leaf metabolomic data. Results We found that the fungal communities showed significant inter-connectivity above and belowground, while bacteria were mainly specialised to each sample type. Herbivory had no effect on leaf microbiome or the metabolome. Fungal root endophytes (Glomeromycotina arbuscular mycorrhizal fungi) and leaf bacteria correlated most strongly with the overall leaf metabolome. However, there were specific pathogenic fungi found in the leaf associated with the abundance of specific metabolites. Conclusions In this study, demonstrate the complexity of plant metabolome-microbiome interactions, and the need to fully integrate the microbiome of different tissue types to fully understand plant’s secondary metabolome. The decreasing costs of -omics methods and high-performance computing costs has made it possible to perform multi-omic, multi-tissue approaches at large-scale, as in this work. It is therefore possible to identify novel plant-microbe interactions, even from plants sampled outside of controlled environments.