Mapping and predictive variations of soil bacterial richness across France

Correction: The third author’s name is spelled incorrectly. The correct name is Samuel Dequiedt. The correct citation is: Terrat S, Horrigue W, Dequiedt S, Saby NPA, Lelièvre M, Nowak V, et al. (2017) Mapping and predictive variations of soil bacterial richness across France. PLoS ONE 12(10): e0186766. https://doi.org/10.1371/journal.pone.0190128International audienceAlthough numerous studies have demonstrated the key role of bacterial diversity in soil functions and ecosystem services, little is known about the variations and determinants of such diversity on a nationwide scale. The overall objectives of this study were i) to describe the bacterial taxonomic richness variations across France, ii) to identify the ecological processes (i.e. selection by the environment and dispersal limitation) influencing this distribution, and iii) to develop a statistical predictive model of soil bacterial richness. We used the French Soil Quality Monitoring Network (RMQS), which covers all of France with 2,173 sites. The soil bacterial richness (i.e. OTU number) was determined by pyrosequencing 16S rRNA genes and related to the soil characteristics, climatic conditions, geomorphology, land use and space. Mapping of bacterial richness revealed a heterogeneous spatial distribution, structured into patches of about 111km, where the main drivers were the soil physico-chemical properties (18% of explained variance), the spatial descriptors (5.25%, 1.89% and 1.02% for the fine, medium and coarse scales, respectively), and the land use (1.4%). Based on these drivers, a predictive model was developed, which allows a good prediction of the bacterial richness (R$^2$$_{adj}$ of 0.56) and provides a reference value for a given pedoclimatic condition