Artificial selection of stable rhizosphere microbiota leads to heritable plant phenotype changes

Abstract

Research on artificial selection of microbial community has become popular due to perspectives in improving plant and animal health1-4. However, reported results still lack consistency5-8. We hypothesized that artificial selection may provide desired outcomes provided that microbial community structure has stabilized along the selection process. In a ten-generation artificial selection experiment involving 1,800 plants, we selected rhizosphere microbiota of Brachypodium distachyon that were associated with high or low levels of leaf greenness, a proxy for plant health9. Monitoring of the rhizosphere microbiota dynamics showed strong oscillations in community structure during an initial transitory phase of five generations, with no heritability in the selected property. In the last five generations, the structure of microbial communities displayed signs of stabilization, concomitantly to the appearance of heritability in leaf greenness. Selection pressure, initially ineffective, became successful in changing the greenness index in the intended direction, especially toward high greenness values. We showed a remarkable congruence between plant traits and selected microbial community structures, highlighting two phylogenetically distinct microbial sub-communities correlating with leaf greenness, whose abundance was significantly steered by directional artificial selection. Understanding microbial community structure stabilization can thus help improve the reliability of artificial microbiota selection.

Community diversity determines the evolution of synthetic bacterial communities under artificial selection

“…2015; Jacquiod et al. 2022) to avoid a gain in the targeted function under AS through the elimination of the less performing communities (i.e., an ecological sorting between the communities) rather than through an increase in the function itself (Chang et al. 2020).…”

“…2019; Jacquiod et al. 2022) but community diversity was not intentionally manipulated. We designed an experiment that combined the approaches developed in artificial selection of microbial communities and in biodiversity‐ecosystem functioning experiments.…”

Community diversity determines the evolution of synthetic bacterial communities under artificial selection

“…64 instead of 18 species to build six replicates of the 16-species level that deeply differ in their composition). Another way to ensure sufficient compositional variability is the maintenance of multiple lineages over the experiment (Blouin et al 2015;Jacquiod et al 2021). In a recent study, (Chang et al 2020) started from a pool of 12 communities which were replicated seven or eight times each for a total of 92 communities.…”

“…It is quite common however that artificial selection produces effects on the mean of a function rather than on the slope of function change versus time (Swenson et al 2000b;Raynaud et al 2019;Chang et al 2020). The difficulty to observe a global trend in the change in a function under artificial selection is probably due to a lack of heritability, associated with the absence of stability in community structure (Chang et al 2020;Jacquiod et al 2021).…”

“…Previous artificial selection experiments were mainly conducted on complex natural microbial communities (retrieved from soil or plant leaves for example) that were then grown under laboratory conditions. Some studies assessed the changes in microbial community diversity over the course of the experiment (Raynaud et al 2019;Jacquiod et al 2021) but community diversity was not intentionally manipulated. We designed an experiment that combined the approaches developed in artificial selection of microbial communities and in biodiversity-ecosystem functioning experiments.…”