A Selective Medium for the Isolation and Quantification of Bradyrhizobium japonicum and Bradyrhizobium elkanii Strains from Soils and Inoculants

The ecological examination of members of the family Rhizobiaceae has been hampered by the lack of a selective medium for isolation of root nodule bacteria from soil. A novel non-antibiotic-containing medium has been developed which allows selective isolation of Bradyrhizobium japonicum and B. elkanii strains from soil and inoculants. The medium, BJSM, is based on the resistance of B.japonicum and B. elkanii strains to more than 40 μg of the metals ions Zn²⁺ and Co²⁺ per ml. BJSM does not allow growth of Rhizobium sp. strains. We used BJSM to isolate bacteria from a Hubbard soil and from several commercially prepared soybean inoculants. Ninety-eight percent of the isolates obtained from Hubbard soil nodulated Glycine max cv. Kasota, and between 55 and 95% of the isolates from the commercial inoculants had the ability to nodulate soybeans. Numbers of bradyrhizobia obtained by using BJSM, strain-specific fluorescent antibodies, and the most-probable-number plant infection assay indicated that the three techniques were comparable in quantifying B. japonicum strains in soils and inoculants, although most-probable-number counts were generally 0.5 order of magnitude greater than those obtained by using BJSM. Results of our studies indicate that BJSM is useful for direct isolation and quantification of B. japonicum and B. elkanii from natural soils and inoculants. This medium may prove to be an important tool for autecological and enumeration studies of diverse populations of bradyrhizobia and as a quality control method for soybean inoculants.     

Competitive Ability and Efficiency in Nodule Formation of Strains of Bradyrhizobium japonicum

In the American Midwest, superior N₂-fixing inoculant strains of Bradyrhizobium japonicum consistently fail to produce the majority of nodules on the roots of field-grown soybean. Poor nodulation by inoculant strains is partly due to their inability to stay abreast of the expanding soybean root system in numbers sufficient for them to be competitive with indigenous bradyrhizobia. However, certain strains are noncompetitive even when numerical dominance is not a factor. In this study, we tested the hypothesis that the nodule occupancy achieved by strains is related to their nodule-forming efficiency. The nodulation characteristics and competitiveness of nine strains of B. japonicum were compared at both 20 and 30°C. The root tip marking technique was used, with the nodule-forming efficiency of each strain estimated from the average position of the uppermost nodule and the number of nodules formed above the root tip mark. The competitiveness of the nine strains relative to B. japonicum USDA 110 was determined by using immunofluorescence to identify nodule occupants. The strains differed significantly in competitiveness with USDA 110 and in nodulation characteristics, strains that were poor competitors usually proving to be inferior in both the average position of the uppermost root nodule and the number of nodules formed above the root tip mark. Thus, competitiveness was correlated with both the average position of the uppermost nodule (r = 0.5; P = 0.036) and the number of nodules formed above the root tip mark (r = 0.64; P = 0.005), while the position of the uppermost nodule was also correlated to the percentage of plants nodulated above the root tip mark (r = 0.81; P < 0.001) and the percentage of plants nodulated on the taproot (r = 0.67; P = 0.002).     

Host-Controlled Restriction of Nodulation by Bradyrhizobium japonicum Strains in Serogroup 110

We previously reported the identification of a soybean plant introduction (PI) genotype, PI 417566, which restricts nodulation by Bradyrhizobium japonicum MN1-1c (USDA 430), strains in serogroup 129, and USDA 110 (P. B. Cregan, H. H. Keyser, and M. J. Sadowsky, Appl. Environ. Microbiol. 55:2532-2536, 1989, and Crop Sci. 29:307-312, 1989). In this study, we further characterized nodulation restriction by PI 417566. Twenty-four serogroup 110 isolates were tested for restricted nodulation on PI 417566. Of the 24 strains examined, 62.5% were restricted in nodulation by the PI genotype. The remainder of the serogroup 110 strains tested (37.5%), however, formed significant numbers of nodules on PI 417566, suggesting that host-controlled restriction of nodulation by members of serogroup 110 is strain dependent. Analysis of allelic variation at seven enzyme-encoding loci by multilocus enzyme electrophoresis indicated that the serogroup 110 isolates can be divided into two major groups. The majority of serogroup 110 isolates which nodulated PI 417566 belonged to the same multilocus enzyme electrophoresis group. B. japonicum USDA 110 and USDA 123 were used as coinoculants in competition-for-nodulation studies using PI 417566. Over 98% of the nodules formed on PI 417566 contained USDA 123, whereas less than 2% contained USDA 110. We also report the isolation of a Tn5 mutant of USDA 110 which has overcome nodulation restriction conditioned by PI 417566. This mutant, D4.2-5, contained a single Tn5 insertion and nodulated PI 417566 to an extent equal to that seen with the unrestricted strain USDA 123. The host range of D4.2-5 on soybean plants and other legumes was unchanged relative to that of USDA 110, except that the mutant nodulated Glycine max cv. Hill more efficiently. While strain USDA 110 has the ability to block nodulation by D4.2-5 on PI 417566, the nodulation-blocking phenomenon was not seen unless strain USDA 110 was inoculated at a 100-fold greater concentration than the mutant strain.     

Phenotypic Diversity among Strains of Bradyrhizobium japonicum Belonging to Serogroup 110

Thirty-four strains of Bradyrhizobium japonicum within serogroup 110 were examined for phenotypic diversity. The strains differed in their abilities to nodulate and fix dinitrogen with Glycine max (L.) Merr. cv. Williams. Thirteen strains expressed uptake hydrogenase activity when induced as free-living cultures in the presence of 2% hydrogen and oxygen. Six bacteriophage susceptibility reactions were observed. Each of the strains produced either a large, mucoid or a small, dry colony morphology, but colony type was not related to effectiveness for nitrogen fixation.     

Genotypic Diversity among Strains of Bradyrhizobium japonicum Belonging to Serogroup 110

Thirty-three strains of Bradyrhizobium japonicum within serogroup 110 were examined for genotypic diversity by using DNA-DNA hybridization analyses. The analysis of the DNA from 15 hydrogen-uptake-negative strains with the bradyrhizobial uptake hydrogenase probe pHU52 showed variation in degree of homology and restriction fragment length polymorphism of EcoRI-restricted DNA. Clustering analysis of the 33 strains on the basis of DNA-DNA hybridization analysis with four restriction enzymes and with the bradyrhizobial nodulation locus, pRJUT10, as probe indicated the existence of four groups of strains, which were less than 70% similar. Restriction digestion of genomic DNA with BamHI and DNA-DNA hybridization with pRJUT10 permitted classification of each of the strains according to a specific fingerprint pattern.     

Fractionation of the beta-Linked Glucans of Bradyrhizobium japonicum and Their Response to Osmotic Potential

Bradyrhizobium japonicum USDA 110 synthesized both extracellular and periplasmic polysaccharides when grown on mannitol minimal medium. The extracellular polysaccharides were separated into a high-molecular-weight acidic capsular extracellular polysaccharide fraction (90% of total hexose) and three lower-molecular-weight glucan fractions by liquid chromatography. Periplasmic glucans, extracted from washed cells with 1% trichloroacetic acid, gave a similar pattern on liquid chromatography. Linkage analysis of the major periplasmic glucan fractions demonstrated mainly 6-linked glucose (63 to 68%), along with some 3,6- (8 to 18%), 3- (9 to 11%), and terminal (7 to 8%) linkages. The glucose residues were beta-linked as shown by H-nuclear magnetic resonance analysis. Glucan synthesis by B. japonicum cells grown on mannitol medium with 0 to 350 mM fructose as osmolyte was measured. Fructose at 150 mM or higher inhibited synthesis of periplasmic and extracellular 3- and 6-linked glucans but had no effect on the synthesis of capsular acidic extracellular polysaccharides.     

Division of Rhizobitoxine-Producing and Hydrogen-Uptake Positive Strains of Bradyrhizobium japonicum by nifDKE Sequence Divergence

Genomic digests from 25 strains of Bradyrhizobium japonicum,for which the phenotypes have been determined with respect toproduction of rhizobitoxine, hydrogen uptake (Hup) and compositionof extracellular polysaccharide (EPS), were hybridized withprobe DNAs of the nifDK and nifE genes of B. japonicum USDA110. The degree of the estimated base substitution in and aroundnifDKE clearly divided the strains of B. japonicum into twomarkedly divergent groups, which were designated as genotypeI and II. Moreover, a strict correlation was observed betweenthese genotypes, production of rhizobitoxine and EPS composition.The genotype I strains produced no rhizobitoxine and an EPScomposed of glucose, mannose, galactose, 4-O-methyl galactoseand galacturonic acid, whereas the genotype II strains producedrhizobitoxine and an EPS composed of rhamnose and 4-O-methylglucuronic acid. Hup⁺ strains were confined exclusively to thegenotype I. Hind III digests of genomic DNAs from the 25 strainswere hybridized with probe DNA of structural genes for the uptakehydrogenase from B. japonicum. In 23 wild-type strains, Hup⁺strains generated a 5.9-kb band that hybridized to the probeunder high-stringency conditions, while Hup^– strains didnot generate the band. These results suggest that the genotypesI and II are two highly divergent evolutionary lines that definea marked division of various phenotypes, such as productionof rhizobitoxine, EPS composition and hydrogen uptake. (Received July 10, 1989; Accepted November 4, 1989)     

Nodulation of Glycine max by Six Bradyrhizobium japonicum Strains with Different Competitive Abilities

The root nodule locations of six Bradyrhizobium japonicum strains were examined to determine if there were any differences which might explain their varying competitiveness for nodule occupancy on Glycine max. When five strains were added to soybeans in plastic growth pouches in equal proportions with a reference strain (U.S. Department of Agriculture, strain 110), North Carolina strain 1028 and strain 110 were the most competitive for nodule occupancy, followed by U.S. Department of Agriculture strains 122, 76, and 31 and Brazil strain 587. Among all strains, nodule double occupancy was 17% at a high inoculum level (10⁷ CFU pouch⁻¹) and 2% at a low inoculum level (10⁴ CFU pouch⁻¹). The less competitive strains increased their nodule representation by an increase in the doubly occupied nodules at the high inoculum level. Among all strains, the number of taproot and lateral root nodules was inversely related at both the high and low inoculum levels (r = −0.62 and −0.69, respectively; P = 0.0001). This inverse relationship appeared to be a result of the plant host control of bacterial infection. Among each of the six strains, greater than 95% of the taproot nodules formed at the high inoculum density were located on 25% of the taproot length, the nodules centering on the position of the root tip at the time of inoculation. No differences among the six strains were observed in nodule initiation rates as measured by taproot nodule position. Taproot nodules were formed in the symbiosis before lateral root nodules. One of the poorly competitive strains (strain 76) occupied three times as many taproot nodules as lateral root nodules when competing with strain 110 (nodules were harvested from 4-week-old plants). Among these six wild-type strains of B. japonicum, competitive ability evidently is not related to nodule initiation rates.     

The Soybean Rj4 Allele Restricts Nodulation by Bradyrhizobium japonicum Serogroup 123 Strains

Of nine Bradyrhizobium japonicum serogroup 123 strains examined, 44% were found to be restricted for nodulation by cultivar Hill. Nodulation studies with soybean isoline BARC-2 confirmed that the soybean Rj4 allele restricts nodulation by the same serogroup 123 isolates. Immunological analyses indicated that B. japonicum strains in serogroups 123 and 31 share at least one surface somatic antigen.     

Survival and Competitiveness of Bradyrhizobium japonicum Strains 20 Years after Introduction into Field Locations in Poland

It was previously demonstrated that there are no indigenous strains of Bradyrhizobium japonicum forming nitrogen-fixing root nodule symbioses with soybean plants in arable field soils in Poland. However, bacteria currently classified within this species are present (together with Bradyrhizobium canariense) as indigenous populations of strains specific for nodulation of legumes in the Genisteae tribe. These rhizobia, infecting legumes such as lupins, are well established in Polish soils. The studies described here were based on soybean nodulation field experiments, established at the Poznań University of Life Sciences Experiment Station in Gorzyń, Poland, and initiated in the spring of 1994. Long-term research was then conducted in order to study the relation between B. japonicum USDA 110 and USDA 123, introduced together into the same location, where no soybean rhizobia were earlier detected, and nodulation and competitive success were followed over time. Here we report the extra-long-term saprophytic survival of B. japonicum strains nodulating soybeans that were introduced as inoculants 20 years earlier and where soybeans were not grown for the next 17 years. The strains remained viable and symbiotically competent, and molecular and immunochemical methods showed that the strains were undistinguishable from the original inoculum strains USDA 110 and USDA 123. We also show that the strains had balanced numbers and their mobility in soil was low. To our knowledge, this is the first report showing the extra-long-term persistence of soybean-nodulating strains introduced into Polish soils and the first analyzing the long-term competitive relations of USDA 110 and USDA 123 after the two strains, neither of which was native, were introduced into the environment almost 2 decades ago.     

Symbiotic Potential, Competitiveness, and Serological Properties of Bradyrhizobium japonicum Indigenous to Korean Soils

The symbiotic potential of Bradyrhizobium japonicum isolates indigenous to seven Korean soils was evaluated by inoculating soybeans with 10- and 1,000-fold-diluted soil suspensions (whole-soil inocula). At both levels, significant differences in the symbiotic potential of the indigenous B. japonicum isolates were demonstrated. The relationship between rhizobial numbers in the whole-soil inocula (x) and nitrogen fixation parameters (y) was best predicted by a straight line (y = a + bx) when the numbers in the inocula were 100 to 10,000 ml^-1, while the power curve (y = ax^b) predicted the variation when the numbers were 1 to 100 ml^-1. Thirty isolates from three soils showed wide differences in effectiveness (measured as milligrams of shoot N per plant), and several were of equal or greater effectiveness than reference strain B. japonicum USDA 110 on soybean cultivars Clark and Jangbaekkong. On both of the soybean cultivars grown in a Hawaiian mollisol, the Korean B. japonicum isolate YCK 213 and USDA 110 were of equal effectiveness; USDA 110 was the superior strain in colonization (nodule occupancy). Korean isolates YCK 117 and YCK 141 were superior colonizers compared with USDA 110. However, B. japonicum USDA 123 was the superior colonizer compared with isolates YCK 213, YCK 141, and YCK 117. In an immunoblot analysis of 97 indigenous Korean isolates of B. japonicum, 41% fell into the USDA 110 and USDA 123 serogroups. Serogroups USDA 110 and USDA 123 were represented in six of the seven soils examined. In one Korean soil, 100% of the B. japonicum isolates reacted only with antisera of YCK 117, an isolate from the same soil.     

Viability of Bradyrhizobium japonicum bacteriods

Homogenates from soybean nodules, formed by 12 strains of Bradyrhizobium japonicum, were plated into yeast-extract mannitol agar containing 3 or 37 g mannitol 1^-1. Viable counts ranged from 8.298 to 11.265 log₁₀ cells-gram nodule^-1. When monitored over the life cycle of the symbiosis, the viability of strains USDA 110 and USDA 123 increased with days after planting (DAP), and at 70 DAP was 95% and 81%, respectively. By contrast, the viability of USDA 38 bacteroids decreased with time, and at 70 DAP was only 1.9%. At 49 DAP, nodules induced by USDA 38 had significantly fewer bacteroids per peribacteroid membrane than those formed by USDA 110 or USDA 123, and at 70 DAP, 27% of the USDA 38 bacteroids showed some degree of degeneration. Viable counts of USDA 123 and USDA 110 bacteroids, isolated from the nodules of 12 different cultivars, ranged from 10.963 to 11.463 and from 10.683 to 11.117 log₁₀ viable cells-gram nodule^-1, respectively. Varying the osmolarity of the medium had no predictable effect on bacteroid viability. When surface-sterilized nodules of IPAGO 587 (high bacteroid viability) and USDA 38 (low bacteroid viability) were inoculated into a nonsterile silt loam soil, at rates equivalent to 5.0×10⁸ and 5.0×10⁶ viable bacteroids g^-1 soil, respectively, and then incubated at 28° C for 60 days, 4.3×10⁴ and 1.5×10⁴ surviving cells g^-1 soil, respectively, were recovered. Thus, despite differences due to host and strain variation, bacteroid viability appears to be unrelated to persistence of individual strains following an annual legume crop cycle.Journal paper No. 14930, Agricultural Experiment Station University of Minnesota, St. Paul, MN 55108, USA     

Nickel accumulation and storage in Bradyrhizobium japonicum

Hydrogenase-derepressed (chemolithotrophic growth conditions) and heterotrophically grown cultures of Bradyrhizobium japonicum accumulated nickel about equally over a 3-h period. Both types of cultures accumulated nickel primarily in a form that was not exchangeable with NiCl2, and they accumulated much more Ni than would be needed for the Ni-containing hydrogenase. The nickel accumulated by heterotrophically incubated cultures could later be mobilized to allow active hydrogenase synthesis during derepression in the absence of nickel, while cells both grown and derepressed without nickel had low hydrogenase activities. The level of activity in cells grown with Ni and then derepressed without nickel was about the same as that in cultures derepressed in the presence of nickel. The Ni accumulated by heterotrophically grown cultures was associated principally with soluble proteins rather than particulate material, and this Ni was not lost upon dialyzing an extract containing the soluble proteins against either Ni-containing or EDTA-containing buffer. However, this Ni was lost upon pronase or low pH treatments. The soluble Ni-binding proteins were partially purified by gel filtration and DEAE chromatography. They were not antigenically related to hydrogenase peptides. Much of the 63Ni eluted as a single peak of 48 kilodaltons. Experiments involving immunoprecipitation of 63Ni-containing hydrogenase suggested that the stored source of Ni in heterotrophic cultures that could later be mobilized into hydrogenase resided in the nonexchangeable Ni-containing fraction rather than in loosely bound or ionic forms.     

Nickel uptake in Bradyrhizobium japonicum.

Free-living Bradyrhizobium japonicum grown heterotrophically with 1 microM 63Ni2+ accumulated label. Strain SR470, a Hupc mutant, accumulated almost 10-fold more 63Ni2+ on a per-cell basis than did strain SR, the wild type. Nongrowing cells were also able to accumulate nickel over a 2-h period, with the Hupc mutant strain SR470 again accumulating significantly more 63Ni2+ than strain SR. These results suggest that this mutant is constitutive for nickel uptake as well as for hydrogenase expression. The apparent Kms for nickel uptake in strain SR and strain SR470 were found to be similar, approximately 26 and 50 microM, respectively. The Vmax values, however, were significantly different, 0.29 nmol of Ni/min per 10(8) cells for SR and 1.40 nmol of Ni/min per 10(8) cells for SR470. The uptake process was relatively specific for nickel; only Cu2+ and Zn2+ (10 microM) were found to appreciably inhibit the uptake of 1 microM Ni, while a 10-fold excess of Mg2+, Co2+, Fe3+, or Mn2+ did not affect Ni2+ uptake. The lack of inhibition by Mg2+ indicates that nickel is not transported by a magnesium uptake system. Nickel uptake was also inhibited by cold (53% inhibition at 4 degrees C) and slightly by the ionophores nigericin and carbonyl cyanide m-chlorophenylhydrazone. Other ionophores did not appreciably affect nickel uptake, even though they significantly stimulated O2 uptake. The cytochrome c oxidase inhibitors azide, cyanide, and hydroxylamine did not inhibit Ni2+ uptake, even at concentrations (of cyanide and hydroxylamine) that inhibited O2 uptake. The addition of oxidizable substrates such as succinate or gluconate did not increase nickel uptake, even though they increased respiratory activity. Nickel update showed a pH dependence with an optimum at 6.0. Most (approximately 85%) of the 63Ni2+ taken up in 1 min by strain SR470 was not exchangeable with cold nickel.     

Host Plant Effects on Nodulation and Competitiveness of the Bradyrhizobium japonicum Serotype Strains Constituting Serocluster 123

Strains in Bradyrhizobium japonicum serocluster 123 are the major indigenous competitors for nodulation in a large portion of the soybean production area of the United States. Serocluster 123 is defined by the serotype strains USDA 123, USDA 127, and USDA 129. The objective of the work reported here was to evaluate the ability of two soybean genotypes, PI 377578 and PI 417566, to restrict the nodulation and reduce the competitiveness of serotype strains USDA 123, USDA 127, and USDA 129 in favor of the highly effective strain CB1809 and to determine how these soybean genotypes alter the competitive relationships among the three serotype strains in the serocluster. The soybean genotypes PI 377578 and PI 417566 along with the commonly grown cultivar Williams were planted in soil essentially free of soybean rhizobia and inoculated with single-strain treatments of USDA 123, USDA 127, USDA 129, or CB1809 and six dual-strain competition treatments of USDA 123, USDA 127, or USDA 129 versus CB1809, USDA 123 versus USDA 127, USDA 123 versus USDA 129, and USDA 127 versus USDA 129. PI 377578 severely reduced the nodulation and competitiveness of USDA 123 and USDA 127, while PI 417566 similarly affected the nodulation and competitiveness of USDA 129. Thus, the two soybean genotypes can reduce the nodulation and competitiveness of each of the three serocluster 123 serotype strains. Our results indicate that host control of restricted nodulation and reduced competitiveness is quite specific and effectively discriminates between B. japonicum strains which are serologically related.     

Methods To Alter the Recovery and Nodule Location of Bradyrhizobium japonicum Inoculant Strains on Field-Grown Soybeans

Three strains of Bradyrhizobium japonicum, I17, 110, and 61A76, were evaluated for their ability to form nodules on field-grown soybeans in soil with a highly competitive indigenous B. japonicum population. The predominant indigenous strain, 0336, in the field site used was unlike the more common isolates from Midwestern soils which belong to the 123 or 138 serogroups. This strain persisted in the soil for at least 30 years without any soybean crops. The three inoculant strains differed in their ability to compete with indigenous strains for nodule formation. Four different inoculation treatments were tested in three adjacent fields. When the amount of inoculum was increased, a higher proportion of nodules contained the inoculant strain. The most competitive inoculant strain was I17, a recent field isolate. Strain 61A76 was better than 110. There was no difference in recovery of the inoculant strains on the Hodgson or Corsoy soybean cultivars, nor was there a difference in recovery of the inoculant strains during the growing season. The vertical distribution of nodules containing the inoculant strains was affected by the method of adding the inoculant to the soil. Inoculant added to the seed furrow produced nodules mainly in the top region of the soybean root. Inoculant tilled into the soil produced nodules primarily in the bottom part of the root. The nodules that were produced in the bottom part of the root are younger and may contribute significant amounts of fixed nitrogen to the soybean during seed formation.     

Carbon metabolism in Bradyrhizobium japonicum bacteroids

Abstract Carbon metabolism in Bradyrhizobium japonicum bacteroids is reviewed. Additionally, the bacteroid tricarboxylic acid (TCA) cycle and its regulation under oxygen-limited conditions is considered, with emphasis on possible sites of TCA cycle rate-limiting reactions. Furthermore, we consider other adaptive pathways that may be employed by these organisms while in symbiosis. These pathways include: (1) a poly-β-hydroxy-butyrate shunt, (2) a malate-aspartate shuttle, (3) an α-ketoglutarate-glutamate shunt, (4) a modified dicarboxylic acid cycle, and (5) fermentation pathways leading to lactate, acetaldehyde and ethanol. The effects of oxygen limitation on host carbon metabolism are also considered briefly.     

Colony Dimorphism in Bradyrhizobium Strains

Ten isolates of Bradyrhizobium spp. which form two colony types were studied; the isolates originated from a range of legume species. The two colony types differed in the amount of gum formed or size or both, depending on the strain. Whole 7-day-old colonies of each type were subcultured to determine the proportion of cells which had changed to the other type. An iterative computerized procedure was used to determine the rate of switching per generation between the two types and to predict proportions reached at equilibrium for each strain. The predicted proportions of the wetter (more gummy) or larger colony type at equilibrium differed significantly between strains, ranging from 0.9999 (strain CIAT 2383) to 0.0216 (strain CIAT 2469), because some strains switched faster from dry to wet (or small to large) and others switched faster from wet to dry (or large to small). Predicted equilibrium was reached after about 140 generations in strain USDA 76. In all but one strain (CIAT 3030) the growth rate of the wetter colony type was greater than or similar to that of the drier type. The mean difference in generation time between the two colony types was 0.37 h. Doubling times calculated for either colony type after 7 days of growth on the agar surface ranged from 6.0 to 7.3 h. The formation of two persistent colony types by one strain (clonal or colony dimorphism) may be a common phenomenon among Bradyrhizobium strains.     

Autecology of Bradyrhizobium japonicum in soybean-rice rotations

The effect of rice culture on changes in the number of a strain of soybean root-nodule bacteria, (Bradyrhizobium japonicum CB1809), already established in the soil by growing inoculated soybean crops, was investigated in transitional red-brown earth soils at two sites in south-western New South Wales. At the first site, 5.5 years elapsed between the harvest of the last of four successive crops of soybean and the sowing of the next. In this period three crops of rice and one crop of triticale were sown and in the intervals between these crops, and after the crop of triticale, the land was fallowed. Before sowing the first rice crop, the number of Bradyrhizobium japonicum was 1.32×10⁵ g^–1 soil. The respective numbers of bradyrhizobia after the first, second and third rice crops were 4.52 ×10⁴, 1.26×10⁴ and 6.40×10² g^–1 soil. In the following two years the population remained constant. Thus sufficient bradyrhizobia survived in soil to nodulate and allow N₂-fixation by the succeeding soybean crop. At the second site, numbers of bradyrhizobia declined during a rice crop, but the decline was less than when the soil was fallowed (400-fold cf. 2200-fold). Multiplication of bradyrhizobia was rapid in the rhizosphere of soybean seedlings sown without inoculation in the rice bays. At 16 days after sowing, their numbers were not significantly different (p<0.05) from those in plots where rice had not been sown. Nodulation of soybeans was greatest in plots where rice had not been grown, but yield and grain nitrogen were not significantly different (p<0.05). Our results indicate that flooding soil has a deleterious effect on the survival of bradyrhizobia but, under the conditions of the experiments, sufficient B. japonicum strain CB 1809 survived to provide good nodulation after three crops of rice covering a total period of 5.5 years between crops of soybean.