Phylogenetically diverse Bradyrhizobium genospecies nodulate Bambara groundnut (Vigna subterranea L. Verdc) and soybean (Glycine max L. Merril) in the northern savanna zones of Ghana

Abstract

A total of 102 bacterial strains isolated from nodules of three Bambara groundnut and one soybean cultivars grown in nineteen soil samples collected from northern Ghana were characterized using multilocus gene sequence analysis. Based on a concatenated sequence analysis (glnII-rpoB-recA-gyrB-atpD-dnaK), 54 representative strains were distributed in 12 distinct lineages, many of which were placed mainly in the Bradyrhizobium japonicum and Bradyrhizobium elkanii supergroups. Twenty-four of the 54 representative strains belonged to seven putative novel species, while 30 were conspecific with four recognized Bradyrhizobium species. The nodA phylogeny placed all the representative strains in the cosmopolitan nodA clade III. The strains were further separated in seven nodA subclusters with reference strains mainly of African origin. The nifH phylogeny was somewhat congruent with the nodA phylogeny, but both symbiotic genes were mostly incongruent with the core housekeeping gene phylogeny indicating that the strains acquired their symbiotic genes horizontally from distantly related Bradyrhizobium species. Using redundancy analysis, the distribution of genospecies was found to be influenced by the edaphic factors of the respective sampling sites. In general, these results mainly underscore the high genetic diversity of Bambara groundnut-nodulating bradyrhizobia in Ghanaian soils and suggest a possible vast resource of adapted inoculant strains.

Keywords: Bradyrhizobium; Bambara groundnut; biological nitrogen fixation; housekeeping genes; multi-locus sequence analysis; soybean.

Figure 1.

Rooted maximum likelihood phylogenetic tree based on recA gene showing relationship among the new isolates and reference strains in the genus Bradyrhizobium. SH-aLRT and ultrafast bootstrap values were inferred from 1000 replicates and are indicated at the nodes when SH-aLRT value is ≥80% and ultrafast value is ≥95%. The scale bar indicates the estimated nucleotide substitution rate. R. tropici CIAT 899^T and R. etli CFN 42^T were used as outgroup.

Figure 2.

Rooted maximum likelihood phylogenetic tree based on atpD-dnaK-glnII-gyrB-recA-rpoB gene sequences showing relationship among the new isolates and reference strains in the genus Bradyrhizobium. SH-aLRT and ultrafast bootstrap values were inferred from 1000 replicates and are indicated at the nodes when SH-aLRT is ≥80% and ultrafast bootstrap is ≥95%. The scale bar indicates the estimated nucleotide substitution rate. R. tropici CIAT 899^T and R. etli CFN 42^T were used as outgroup.

Figure 3.

Maximum likelihood phylogeny based on nodA gene showing relationship among Bambara groundnut- and soybean-nodulating Bradyrhizobium test strains and references. The SH-aLRT and ultrafast bootstrap values were inferred from 1000 replicates and are indicated at the nodes when SH-aLRT is ≥80% and ultrafast bootstrap is ≥95%. The scale bar indicates the estimated nucleotide substitution rate. M. vignae BR3299^T was used as outgroup.

Figure 4.

Maximum-likelihood phylogenetic tree based on sequences of nifH gene showing the relationship among new rhizobial isolates and recognized species in the Bradyrhizobium genus. SH-aLRT and ultrafast bootstrap values ≥80% and ≥95%, respectively (based on 1000 replications), are shown at each node. The scale bar indicates the estimated nucleotide substitution rate. R. acidophilus ATCC 25092^T was included as outgroup.

Figure 5.

Biplot of the RDA on recA (rA) clusters and their soil factors from sampling sites in the northern savanna zones of Ghana. Si, silt; P, phosphorus; Sa, sand; Al, exchangeable acidity; K, potassium; Mn, manganese. Short names are the names of the sampling sites.

Figure 6.

Biplot of the RDA on genospecies (GS) and their soil factors from sampling sites in the northern savanna zones of Ghana. Al, exchangeable acidity; Si, silt; Mn, manganese; P, phosphorus; Zn, zinc; OC, organic carbon. Short names are the names of the sampling sites.