How do plants select rhizosphere communities. Community succession

A multi-generation experimental selection was developed to separate the roles of plant type and soil in shaping rhizosphere microbiomes of six plants from three families (models Arabidopsis thaliana, Medicago truncatula and Brachypodium distachyon and crops Brassica rapa, Pisum sativum and Triticum aestivum). The microbiomes, selected over three generations from a common microbial inoculum, were distinct for each plant but also depended on soil type, with around 92% of the rhizosphere selected OTUs being different between soil and compost grown plants. Microbiome diversity collapsed in the third generation of plants grown in sand but not those in compost, with cross-generational sweeps of bacterial opportunists and autotrophs. Plants bias the rhizosphere microbiome but its stability and diversity depend on soil structure and organic matter. Our multi-generation approach mimicked crop monoculture with the fungal pathogen Olpidium brassicae specifically sweeping through its susceptible host B. rapa. This was accompanied by an increase in bacteria from the genus Massilia, which are known to degrade fungal chitin. Compared with wild type A. thaliana, the rhizosphere microbial community composition was significantly altered by mutants defective in innate immune recognition, despite this species having the weakest effect on the rhizosphere microbiome. Actinobacteria were less abundant in rhizosphere microbiomes of mutant plants than those of wild type, suggesting their recruitment to outcompete less desirable microbes. Thus the microbiome is shaped by plant species and immunity but its composition and stability also depends on the soil.