Determination of competitive abilities of Bradyrhizobium japonicum strains in soils from soybean production regions in South Africa

Bradyrhizobium japonicum strain CB 1809 was recently chosen to replace strain WE 1 in commercial soybean [Glycine max (L.) Merr.] inoculants in South Africa, the selection criterion being N-2-fixing effectiveness. Nodulation competitiveness is an additional characteristic required of inoculants and was determined for CB1 809 and WB 1 as well as two other strains, USDA 110 and a Brazilian strain 965, using the gusA marker gene to identify strains. Initial experiments with plants grown in sterile sand showed that the competitive index of strain WB 1 was less than that of the other strains. Further comparisons used plants grown in five soils containing established populations of B. japonicum. When strains were applied in peat inoculum to seed at a rate of 1,000 cells per seed in a soil containing 300 rhizobia g(-1), significant differences in nodule occupancy were detected and strains ranked in the order 965>CB1809>USDA 110>WB 1. The remaining four soils each contained about 10(6) rhizobia g(-1) and 5 x 10(6) cells were applied per seed. Nodule occupancy by inoculant strains ranged from 22% to 81% between soils. In this experiment, WE 1 was consistently the poorest performer and its competitiveness was significantly less than CB 1809. The competition results supported the recent decision to replace WB 1 with CB 1809 in commercial inoculants. Although WB 1 had been used in inoculants over a period of 19 years, this strain was detected in only one soil, where it comprised 8% of isolates. In contrast, a substantial proportion (32-78%) of isolates from the soils corresponded serologically to a former inoculant strain WB 66, which had been discontinued in 1966. This illustrates the difficulty of replacing a resident population with an introduced strain. The effect of naturalized populations on the establishment of CB 1809 in South African soils will need monitoring.