Genetic diversity and symbiotic effectiveness of Phaseolus vulgaris-nodulating rhizobia in Kenya

Abstract

Phaseolus vulgaris (common bean) was introduced to Kenya several centuries ago but the rhizobia that nodulate it in the country remain poorly characterised. To address this gap in knowledge, 178 isolates recovered from the root nodules of P. vulgaris cultivated in Kenya were genotyped stepwise by the analysis of genomic DNA fingerprints, PCR-RFLP and 16S rRNA, atpD, recA and nodC gene sequences. Results indicated that P. vulgaris in Kenya is nodulated by at least six Rhizobium genospecies, with most of the isolates belonging to Rhizobium phaseoli and a possibly novel Rhizobium species. Infrequently, isolates belonged to Rhizobium paranaense, Rhizobium leucaenae, Rhizobium sophoriradicis and Rhizobium aegyptiacum. Despite considerable core-gene heterogeneity among the isolates, only four nodC gene alleles were observed indicating conservation within this gene. Testing of the capacity of the isolates to fix nitrogen (N₂) in symbiosis with P. vulgaris revealed wide variations in effectiveness, with ten isolates comparable to Rhizobium tropici CIAT 899, a commercial inoculant strain for P. vulgaris. In addition to unveiling effective native rhizobial strains with potential as inoculants in Kenya, this study demonstrated that Kenyan soils harbour diverse P. vulgaris-nodulating rhizobia, some of which formed phylogenetic clusters distinct from known lineages. The native rhizobia differed by site, suggesting that field inoculation of P. vulgaris may need to be locally optimised.

Keywords: MLSA; Nodulation; Phaseolus vulgaris; Phylogeny; Rhizobium.

Fig. 1

Phylogenetic tree of the 16S rRNA gene from 36 isolates (in bold) and type strains of closely related species constructed using the Maximum Likelihood method based on the Tamura 3-parameter model in MEGA6 . There was a total of 1305 positions in the final dataset, and node supports higher than 50% are labelled with a bootstrap value (1000 replicates). The sequence of Bradyrhizobium japonicum USDA 6^T was included as an outgroup. Bar indicates five nucleotide substitutions per 100 nucleotides.

Fig. 2

The phylogenetic relationship between the study isolates (in bold) and type strains of closely related species based on concatenated recA and atpD genes. The evolutionary history was inferred using the Maximum Likelihood method based on the General Time Reversible model in MEGA6 . There was a total of 736 positions in the final dataset, and node supports higher than 50% are labelled with a bootstrap value (1000 replicates). The sequence of B. japonicum USDA 6^T was included as an outgroup. Bar indicates 10 nucleotide substitutions per 100 nucleotides.

Fig. 3

Phylogenetic tree of the nodC gene from 36 isolates (in bold) and reference strains constructed using the Maximum Likelihood method based on the Tamura 3-parameter model in MEGA6 . There was a total of 504 positions in the final dataset, and node supports higher than 50% are labelled with a bootstrap value (1000 replicates). The sequence of B. japonicum USDA 6^T was included as an outgroup. Bar indicates five nucleotide substitutions per 100 nucleotides.

Fig. 4

Mean shoot dry weights of P. vulgaris cv. KK08 inoculated with 36 rhizobial isolates from Kenya expressed as a percentage of CIAT 899 treatment. N denotes the un-inoculated treatment. All plants were maintained with nitrogen-free growth media. Data are means of six plants, harvested 42 days after inoculation.