Persistence of introduced Bradyrhizobium japonicum strains in forming nodules in subsequent years after inoculation in Wisconsin soils

Nodulation of soybeans by indigenous and inoculum strains of Bradyrhizobium japonicum was studied in field experiments in Wisconsin from 1983 to 86. Aqueous suspensions of bacteria were applied to seeds at the time of planting at levels of 7 × 107–1010 bacteria per 2.5-cm row. The predominant indigenous serogroup was 123 in these soils. Six different inoculum strains were used (two from serocluster 123, two from serogroup 110, and one each from serogroups 122 and C1). Nodule occupants were identified using spontaneous antibiotic-resistant mutations in the inoculum strains, phage typing, and serotyping. In the 1983 experiment, the majority of nodules were formed by the inoculum strains in almost all cases (up to 100% in some cases), in two different soils containing 3.5 × 105 indigenous B. japonicum per gram. After 2 years without inoculation at the same two site, the inoculum strains did not form many nodules on uninoculated soybeans (less than 10% in most cases; less than 30% in all cases). In inoculation experiments carried out in 1985 and 1986, four inoculum strains were used (3 members of 123 serocluster and USDA 110str); inocula containing 108 bacteria per 2.5-cm row formed less than42%ofthe nodules in soils containing 1 × 104–4 × 104B. japonicum per gram. The major conclusions are (i) the success of inoculation in Midwestern U.S. soils is highly variable, even with members of the (highly competitive) 123 serocluster, and (ii) successful inoculation in 1 year in a Wisconsin soil does not ensure that the inoculated strain will persist in forming nodules in that field in subsequent years without further inoculation. Key words: Bradyrhizobium japonicum, strain persistence, field trials.