Population structure of Rhizobium etli-like strains nodulated with Phaseolus vulgaris in two ecoregions of China

- Haplotype/nucleotide diversities varied among rhizobial populations.
- Natural or neutral selection shaped populations depending on soil pH.
- Bean rhizobia in China derived from ancestors of South America.
- Soil factors shaped populations stronger than geographic distances.

Population genetics could reveal the evolutionary relationships of different populations in the same bacterial species or closely related species. To estimate the evolutionary history of bean-nodulating Rhizobium etli-like bacteria in China, the genetic diversity and population structures of 77 strains belonging to seven genomic species isolated from eight sites in the temperate and subtropical regions were investigated in this study. As results, seven clades were defined in the phylogenetic network based upon the sequences of eight housekeeping and adaptation genes, in which some strains of different genospecies were intermingled due to the effects of adaptation genes on the phylogeny of chromosome. The close phylogenetic relationships of the tested strains with a Columbian strain R. etli CIAT 652, and their deeply separating from the Mexican reference strains demonstrated the evolution of Chinese bean-nodulating bacteria directly from the ancestors of South America. Distinct populations with varied haplotype diversity and nucleotide diversity were observed in the alkaline, strong acid and neutral-acid soils, and they may be affected by natural selection in the alkaline/strong acid soils, or by the neutral selection in other sampling sites. The isolates in the neutral and acid soils were admixture containing more of the ancestral lineages similar to the American strains, while the isolates in strong acid (Chutan) and alkaline (Daqing) soils formed independent clusters. Soil factors shaped the rhizobial populations stronger than the geographic distance. Moreover, recombination contributed more in the evolution of R. etli-related populations than mutation.