Testing the two-step model of plant root microbiome acquisition under multiple plant species and soil sources

Abstract

The two-step model for plant root microbiomes considers soil as the primary microbial source. Active selection of the plant’s bacterial inhabitants results in a biodiversity decrease toward roots. We collected sixteen samples of in situ ruderal plant roots and their soils and used these soils as the main microbial input for single genotype tomatoes grown in a greenhouse. Our main goal was to test the soil influence in the structuring of rhizosphere microbiomes, minimizing environmental variability, while testing multiple plant species. We massively sequenced the 16S rRNA and shotgun metagenomes of the soils, in situ plants, and tomato roots. We identified a total of 271,940 bacterial operational taxonomic units (OTUs) within the soils, rhizosphere and endospheric microbiomes. We annotated by homology a total of 411,432 (13.07%) of the metagenome predicted proteins. Tomato roots did follow the two-step model with lower α-diversity than soil, while ruderal plants did not. Surprisingly, ruderal plants are probably working as a microenvironmental oasis providing moisture and plant-derived nutrients, supporting larger α-diversity. Ruderal plants and their soils are closer according to their microbiome community composition than tomato and its soil, based on OTUs and protein comparisons. We expected that tomato β-diversity clustered together with their soil, if it is the main rhizosphere microbiome structuring factor. However, tomato microbiome β-diversity was associated with plant genotype in most samples (81.2%), also supported by a larger set of enriched proteins in tomato rhizosphere than soil or ruderals. The most abundant bacteria found in soils was the Actinobacteria Solirubrobacter soli, ruderals were dominated by the Proteobacteria Sphingomonas sp. URGHD0057, and tomato mainly by the Bacteroidetes Ohtaekwangia koreensis, Flavobacterium terrae, Niastella vici, and Chryseolinea serpens. We calculated a metagenomic tomato root core of 51 bacterial genera and 2,762 proteins, which could be the basis for microbiome-oriented plant breeding programs. We attributed a larger diversity in ruderal plants roots exudates as an effect of the moisture and nutrient acting as a microbial harbor. The tomato and ruderal metagenomic differences are probably due to plant domestication trade-offs, impacting plant-bacteria interactions.