Seventeen Bradyrhizobium sp. strains and one Azorhizobium strain were compared on the basis of five genetic and phenetic features: (i) partial sequence analyses of the 16S rRNA gene (rDNA), (ii) randomly amplified DNA polymorphisms (RAPD) using three oligonucleotide primers, (iii) total cellular protein profiles, (iv) utilization of 21 aliphatic and 22 aromatic substrates, and (v) intrinsic resistances to seven antibiotics. Partial 16S rDNA analysis revealed the presence of only two rDNA homology (i.e., identity) groups among the 17 Bradyrhizobium strains. The partial 16S rDNA sequences of Bradyrhizobium sp. strains form a tight similarity (>95%) cluster with Rhodopseudomonas palustris, Nitrobacter species, Afipia species, and Blastobacter denitrificans but were less similar to other members of the ␣-Proteobacteria, including other members of the Rhizobiaceae family. Clustering the Bradyrhizobium sp. strains for their RAPD profiles, protein profiles, and substrate utilization data revealed more diversity than rDNA analysis. Intrinsic antibiotic resistance yielded strainspecific patterns that could not be clustered. High rDNA similarity appeared to be a prerequisite, but it did not necessarily lead to high similarity values between RAPD profiles, protein profiles, and substrate utilization. The various relationship structures, coming forth from each of the studied features, had low compatibilities, casting doubt on the usefulness of a polyphasic approach in rhizobial taxonomy.

Effects of soil acidity on groundnut-Bradyrhizobium symbiotic performance were studied in a potted, sandy soil in a glasshouse in Zimbabwe. The soil was limed to soil-pH levels of 5.0 and 6.5. Soil acidity negatively affected plant development, measured as leaf area and plant dry weight, while nodulation was enhanced. This acidity-enhanced nodulation was most evident when nodulation was caused by the indigenous Bradyrhizobium population. Effects of soil acidity differed between groundnut cultivars and Bradyrhizobium spp. strains, the former having greater importance. Two Arachis hypogaea L. Spanish-type cultivars, Falcon and Plover, performed equally well at neutral soil pH, but Falcon was more acid tolerant. Comparison of the symbiotic performance in neutral versus acid soil of two Bradyrhizobium spp.

Eighteen Bradyrhizobium spp. strains, four Rhizobium spp. strains and one Azorhizobium caulinodans strain were grown under Fe limitation and assayed for siderophore production. It was further assessed if Fe accumulation in two groundnut cultivars was influenced by inoculant strain or nitrate fertilisation. Growth of Bradyrhizobium spp. strains nodulating groundnut was slow with mean generation times from 11-24 h. All strains, except MAR 967, showed a reduced growth rate when deprived of Fe; none of the strains showed starvation at 1/tM Fe. In the CAS (chrome azurol S)-agar assay, all strains, which formed colonies, produced siderophores as visualised by orange halos around the colonies on blue plates. Bradyrhizobium strains produced much smaller halos than the reference Rhizobium meliloti strain. In the CAS-supernatant assay, all strains, except MAR 967, gave positive responses (measured as absorbance at 630 nm) when supernatants of Fe-depleted cultures were assayed with CAS-indicator complex in comparison with Fe-supplemented cultures. Responses of all four Rhizobium spp. strains were large, while responses of all Bradyrhizobium strains, except B. japonicum MAR 1491 (USDA 110), were small and mostly insignificant. A small response, i.e. a low Fe-scavenging ability, implies either the production of small quantities of siderophores or the production of low affinity siderophores. Among the Bradyrhizobium strains, MAR 1574 and MAR 1587 gave the largest responses taken over the two assays. Fe accumulation in groundnut cultivar Falcon was seven times larger than in cultivar Natal Common. No correlation was found between the quantity of nodule tissue and Fe accumulation, making it unlikely that bacteroids are involved in Fe acquisition by groundnuts. Nitrate-fertilised plants accumulated significantly more Fe, suggesting involvement of nitrate reductase in Fe assimilation in groundnut. The two most successful Fe-scavenging Bradyrhizobium spp. strains were also the most effective in nodulating groundnut, the reverse also being true. Strain MAR 967, with the lowest Fe requirement, produced the largest nodule dry weight. These data indicate that improved Fe scavenging properties and/or reduced Fe requirement improve rhizospheric growth and with that nodulation effectiveness.

Effects of inoculating four Arachis hypogaea ssp. fastigiata cultivars with 17 Bradyrhizobium spp. strains were studied in a glasshouse experiment using a sandy soil devoid of an indigenous Bradyrhizobium population. Firstly, a wide range of parameters, indicative of symbiotic performance, were assessed for their influence on seed yield, by correlation and statistical analyses. It was found that nodule dry weight and leaf area were relevant parameters concerning seed yield. Secondly, the effects of host and strain genotype on those parameters were described.Variations in nodule dry weight did not have an effect on seed yield, except for cultivar Natal Common at lower nodule dry weight values. Therefore, it was concluded that the quantity of nitrogen fixing tissue met the demand for combined nitrogen and did not limit seed yield. This conclusion was further supported by the observation that at low nodule numbers per plant the nodule size increased to generate sufficient nitrogen fixing tissue.Leaf area, which comprises components for both photosynthetic capacity and plant development, was found to correlate well with seed yield. An increase in leaf area resulted in significant seed yield increases for all three spanish-type cultivars, but not for the valencia-type cultivar. Leaf area, thus, appeared as a factor limiting seed yield of spanish-type groundnuts.Cultivar performance concerning seed yield was significantly better for Natal Common compared to the other three cultivars, while Natal Common had a significantly lower plant (biomass excluding seed) dry weight value.Inoculation with different strains of Bradyrhizobium resulted in significantly different nodule dry weight values, but hardly led to significant differences in seed yield. This agreed with the finding that the amount of nitrogen fixing tissue appeared not to limit the availability of combined nitrogen.A large quantity of nitrogen was partitioned to the groundnut seeds: 62% to 76% of total accumulated nitrogen was located in the seeds.This study showed that testing for symbiotic effectiveness in the groundnut Bradyrhizobium symbiosis should include assessment of final (seed and biomass) yield, because parameters measured at stages prior to maturity, like nodulation parameters, may lead to flawed effectiveness ratings.280 Van Rossum et al.