Serogroups of Indigenous Rhizobium japonicum for Nodulation of

Transcription

1 APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Apr. 1984, p /84/ $02.00/0 Copyright 1984, American Society for Microbiology Vol. 47, No. 4 Rhizosphere Response as a Factor in Competition Among Three Serogroups of Indigenous Rhizobium japonicum for Nodulation of Field-Grown Soybeanst H. A. MOAWAD, W. R. ELLIS, AND E. L. SCHMIDT* Department of Soil Science, University of Minnesota, St. Paul, Minnesota Received 6 October 1983/Accepted 22 December 1983 Rhizosphere response was studied as a factor in competition among indigenous Rhizobium japonicum serogroups for the nodulation of soybeans under field conditions. R. japonicum serogroups 110, 123, and 138 were found to coexist in a Waukegan field soil where they were determined to be the major nodulating rhizobia in soybean nodules. Competitive relationships among the three serogroups in that soil and in rhizospheres were examined during two growing seasons with several host cultivars with and without inoculation and with a nonlegume. Enumeration of each of the three competitors was carried out on inner rhizosphere and nonrhizosphere soil by immunofluorescence with serogroup-specific fluorescent antibodies. Rhizobia present in early- and late-season nodules were identified by fluorescent antibody analysis. Populations of each serogroup increased gradually in host rhizospheres, not exceeding 106/g of rhizosphere soil during the first few weeks after planting, whereas numbers in fallow soil remained at initial levels (104 to 105/g). The rhizosphere effects were minor in host plants during this period of nodule initiation and were about the same for all three serogroups. Although serogroup 123 gave no evidence of dominance in early host rhizospheres, it clearly dominated in nodule composition, occupying 60 to 100% of the nodules. High densities of all three serogroups were observed in host rhizospheres during flowering. Rhizosphere populations, especially of serogroup 123, were still high during pod fill and seed maturation. The rhizosphere responses of the R. japonicum serogroups were much greater with the soybean cultivars than with oats, but even in host rhizospheres the R. japonicum populations were greatly outnumbered by other bacteria. The success of serogroup 123 in achieving nodulation does not appear to be due to superior colonization of the host rhizosphere. Inoculation of soybeans with selected strains of Rhizobium japonicum has been notably unsuccessful in influencing nodulation or enhancing nitrogen fixation in regions where soybeans have been cultivated previously. The inoculant fails to nodulate the soybean to any significant extent, whereas virtually all of the nodules are occupied by strains by R. japonicum indigenous to the soil (6, 11, 20, 22). Throughout much of the major soybean production area of the United States, indigenous soil strains of R. japonicum serogroup 123 dominate nodulation of soybeans regardless ofthe inoculation practice or of the cultivar grown (5, 7, 9, 14). Other serogroups of R. japonicum also occur as indigenous soil bacteria, presumably introduced during previous soybean culture, but these indigenous strains are no more successful than the strains added as the inoculant. The success of a particular indigenous strain may be contingent on competitive interactions with other strains in the host rhizosphere before nodulation (4). The advent of immunofluorescence techniques has made it possible to examine this possibility and to study the autecology of individual serogroups of R. japonicum directly in soil and in natural rhizospheres (3, 4, 14-16). In this report we present the results of a study in which use was made of serogroupspecific fluorescent antibodies (FAs) to follow the population dynamics of R. japonicum serogroup 123 and two competing serogroups of R. japonicum within the same rhizospheres of field-grown plants during the course of two growing seasons. * Corresponding author. t Paper no in the Scientific Journal Series of the Minnesota Agricultural Experiment Station, St. Paul. 607 MATERIALS AND METHODS Plots were established in late May 1980 and 1981 in a Waukegan silt loam at the University of Minnesota, St. Paul. Soil of the 1980 plot had a ph of 5.8 and an organic matter content of 2.7% and had been planted with oats (Avena sativa) the previous year; 1981 plot soil had a ph of 6.1 and an organic matter content of 4.4% and had been planted with corn (Zea mays) the previous year. The following soybean cultivars were used: Chippewa, Hodgson 78, and plant introduction The plant introduction line was included for its reported enhanced ability to nodulate with R. japonicum serogroup 110 in the presence of indigenous R. japonicum strains (11). One portion of the 1980 plot was planted with oats, A. sativa cv. Moores. Onehalf of the soybean seeds were inoculated at planting with a granular inoculant of R. japonicum serogroup 110 (Nitragin Co., Milwaukee, Wis.) at the rate of 3 x 108 cells per m of row (1980) or 5.8 x i07 cells per seed (1981) or with a suspension of R. japonicum serogroup 138 at the rate of 7.6 x 107 cells per seed (1981). Samples offallow and inner rhizosphere soils were collected during the growing seasons to determine the population densities of R. japonicum serogroups 110, 123, and 138. Plants were dug carefully, soil loosely adhering to the intact room system was removed by gentle shaking, and any large soil aggregates still adhering to the root system were removed. The remaining soil closely associated with the root surfaces was considered to be the inner rhizosphere (14). Care was taken to exclude root fragments and nodules from the rhizosphere samples. Randomly selected nodules were collected at 5 weeks and again at 12 to 15 weeks to determine

2 608 MOAWAD, ELLIS, AND SCHMIDT the strain or strains of R. japonicum within the nodule. Young female New Zealand rabbits were used to produce antibodies against somatic antigens of R. japonicum serogroups 110, 123, and 138 grown on yeast extract-mannitol medium (16). Antigen preparation, injection, and bleeding were as previously described (18), and globulin separation, conjugation with fluorescein isothiocyanate, and titration of the FA followed the procedures of Belser and Schmidt (1). Each of the FAs was serogroup specific and without heterologous cross-reactivity. The procedure of Kingsley and Bohlool (10) was used to separate R. japonicum from fallow and rhizosphere soils, starting with 10 g of fallow soil or 1 or more (up to 20 for very young plants) root systems in 95 ml of gelatin-ammonium phosphate buffer. Most inner rhizosphere samples were 0.3 to 0.5 g/100 ml of buffer. Soil-buffer mixtures were treated with 5 drops of Tween 80 (Difco Laboratories, Detroit, Mich.) and shaken for 30 min on a wrist action shaker (Burrell Corp., Pittsburgh, Pa.). After being shaken, inner rhizosphere soil suspensions were subsampled for dry weight determinations, and samples of all suspensions were sedimented by centrifugation at 200 x g for 4 min. Appropriate volumes (1 to 3 ml) of each supernatant were passed through polycarbonate membrane filters (0.4-,um pore size, 25-mm diameter; Nuclepore Corp., Pleasanton, Calif.) which had been pretreated with irgalan black (8) to reduce autofluorescence. Pretreatment (R. Smith, personal communication, Bedford Institute of Oceanography, Dartmouth, Nova Scotia, Canada) was as follows. The filters were soaked in 0.5% Wayfos detergent (Wayfos 6TD, Philip A. Hunt Chemical Corp., East Providence, R.I.) for 5 min, rinsed four times with distilled water, and soaked in 0.2% irgalan black (Irgalan Black BGL 200, CIBA-GEIGY Corp., Summit, N.J.) in 2.0% acetic acid at room temperature for 22 h and then at 90 C for 1 h. The irgalan black solution was drained off, and the filters were cooled for 5 to 10 min, transferred to a beaker, and rinsed four times with deionized water. The filters were allowed to air dry and stored flat in a dry container. Material retained on the filter was treated with gelatinrhodamine conjugate (2) before application of FAs to avoid nonspecific adsorption and to improve the background contrast of the filter. Procedures for FA staining and microscopic enumeration of fluorescing R. japonicum cells were as described previously (17). Fifty microscopic fields were counted on each of the four replicate filters per sample, and counts were converted to cells per gram of oven dried soil. The microscope was a Zeiss standard model 14 equipped with an IV FL epi-illuminator, BP exciter filter, FT 510 dichroic mirror, and LP520 barrier filter. The light source was a 12-V, 100-W halogen lamp. These procedures were found to give an average enumeration efficiency of 35%, based on techniques to measure the recovery of an R. japonicum population added to soil (16). Direct microscopic counts of total bacteria were determined for soil samples extracted as for FA examination (described above) with acridine orange staining anid fluorescence microscopy (19). Most-probable-number (MPN) estimates of total R. japonicum populations were carried out in plastic growth pouches (21) with 10-fold dilutions of soil inocula and two plants per pouch. Nodules were crushed, smeared on slides in duplicate, and analyzed serologically by the FA method to determine the serogroup(s) present, based on 50 nodules per treatment in 1980 or 20 nodules per treatment in The presence of more than one serogroup per nodule was detected by the procedures of Lindemann et TABLE 1. Seasonal variations in the populations of total indigenous R. japonicum serogroups and indigenous R. japonicum serogroup 123 in Waukegan soila Organisms per g of soil Sampling time Total of ali R. japonicum Serogroup 123 serogroups 7/ x x 104 9/ x x / x x W 4/ x 1i4 8.2 x 103 a Total R. japonicum was determined by MPN growth pouch analysis (21), and serogroup 123 was determined by strain-specific FA analysis (17). al. (12). Nodule smears with 5% or more nonfluorescing cells in the presence of FA-positive cells was considered evidence of a mixed infection. RESULTS Samples of fallow soil taken during the year preceding the initiation of the field study were found to have a substaritial population of indigenous R. japonicum. Results of the MPN analysis for all serogroups of R. japonicum and the FA analysis for R. japonicum serogroup 123 are presented in Table 1. Populations ranged from low densities of several thousand per gram in the early spring to high densities approaching a million per gram in the early fall. Serogroup 123 was estimated to compose from 10 to 50% of the R. japonicum popualtion in the soil samples examined. The results of serotyping the nodules collected from the MPN growth pouches are presented in Table 2. These data indicate the presence of each of the three serogroups assayed for, 110, 123, and 138, as well as unidentified strains outside these serogroups. Under the conditions of the growth pouch analysis, nodules from most soil inocula were occupied predominantly by serogroup 123. Serogroup 138 was the second most commonly occurring nodule occupant, even exceeding serogroup 123 in the early-spring soil sample. Serogroup 110 occurred in the soil throughout the year but accounted for only ca. 10% of the singly infected nodules. About one-fifth of the nodules examined contained strains which did not react with the FAs used for serotyping. The population dynamics of serogroups 110, 123, and 138 were followed in the rhizospheres of field-grown plants and in fallow control soils by quantitative FA analysis over the course of two growing seasons. Results of the fallow soil counts are shown in Fig. 1, and those of the rhizosphere counts are given in Tables 3 and 4. Populations of all three serogroups remained relatively constant and basically simi- TABLE 2. APPL. ENVIRON. MICROBIOL. Identification and seasonal variations of R. japonicum serotypes in nodules of soybeans inoculated with Waukegan soil % Serogroup Sampling Others time Sa Ma S M S M S M 7/ / / / a S, Single infection; M, mixed infection.

3 VOL. 47, 1984 RHIZOSPHERE RESPONSE OF THREE R. JAPONICUM SEROGROUPS In a) - a} 5 n E c C3 0 -i -c I Weeks FIG. 1. Enumerations of three indigenous serogroups of R. japonicum in fallow Waukegan soil during the growing seasons of 1980 ( ) and 1981 (-----). Symbols: 0, serogroup 123; 0. serogroup 110; A, serogroup 138. lar. in fallow soil throughout each of the growing seasons. Counts ranged from 3 x 104 to 8 x 104/g of soil for the 1980 plots and were nearly a log higher in the 1981 location. Counts of each of the three serogroups in the rhizospheres of two soybean cultivars and oats during 1980 are reported in Table 3. Patterns of rhizosphere colonization by R. japonicum were similar for both host cultivars, were not influenced by seed inoculation with serogroup 110, and reflected no preferential growth among the three indigenous serogroups. Host rhizosphere responses were modest during the first few weeks, with a 10-fold or less increase over the fallow soil in the rhizosphere of newly emerged plants to about a 40-fold increase by week 5. At week 9, with soybeans in the late flowering stage, R. japonicum densities were surprisingly high in the rhizospheres, reaching levels of 109/g for serogroup 123 and 107 to 108/g for serogroups 110 and 138. The population of serogroup 123 remained at the level of 109/g in the soybean rhizospheres after pod fill (15 weeks), whereas serogroups 110 and 138 generally declined somewhat and were in the range of 106 to 108/g. The data for Chippewa inoculated with serogroup 110 (data not shown) were similar throughout to those reported in Table 3. The rhizosphere effect was generally slight for the R. japonicum strains in the nonhost oats at all but the 9-week sampling, when populations were comparable to those of the 5-week host rhizosphere. Data for 1981 (Table 4) are in good agreement with those reported in Table 3 with respect to the general pattern of rhizosphere response. As in 1980, the three R. japonicum serogroups responded weakly and nonselectively to the rhizosphere conditions of early plant growth. Also as in 1980, all three serogroups colonized the rhizosphere vigorously during host flowering, with serogroup 123 stimulated the most, whereas during pod fill and at plant maturity, only serogroup 123 remained at consistently high rhizosphere/soil (R/S) ratios. Seed inoculation in 1981 again had no effect on rhizosphere colonization. Additional data (not shown) were obtained for the cultivar Hodgson and for the three cultivars (Chippewa, Hodgson, and ) after inoculation with serogroup 138; all of these data were similar to those reported in Table 4. The 1981 data varied quantitatively from those of the previous year in that rhizosphere responses were lower throughout the growing season. The combined populations of the three R. japonicum serogroups made up only a small percentage of the total bacteria present in the host rhizospheres (Table 5). R. japonicum generally increased to some extent in the early rhizosphere but, nonetheless, made up less than 1% of all of the rhizosphere bacteria during the period of active nodulation. The very high values of 5 to 20% observed at plant maturity (15 weeks) probably reflect the release of bacteroids into the rhizosphere from mature root systems. Results of serotyping of nodules designated as early (5 weeks after planting) and late (12 or 15 weeks after planting) are summarized in Table 6. The dominance of serogroup 123 in nodule occupancy is striking. This dominance was expressed both years, in both early and late nodules, and in all soybean cultivars and was not influenced materially by TABLE 3. Wk after Population densities of three indigenous serogroups of R. japonicun in host and nonhost rhizospheres of field-grown plants during the 1980 season Density (no. x 106/g) Soybean cultivar Nonhost Oats seeding Chippewa inoc' NDh ND ND ND ND ND ND ND ( , " Seed inoculated with R. japonicum serotype 110 at planting. b ND, Not determined.

4 610 MOAWAD, ELLIS, AND SCHMIDT APPL. ENVIRON. MICROBIOL. TABLE 4. Population densities of three indigenous serogroups of R. japonicum in the rhizospheres of field-grown soybean cultivars during the 1981 season Density (no. x 106/g) of cultivar and serogroup Days after Chippewa Chippewa inoc' inoca seeding a Seed inoculated with R. japonicum serotype 110 at planting. inoculation. The indigenous R. japonicum soil population obviously included serogroups other than the three examined, since more than a third of the nodules in some instances contained rhizobia that did not cross-react with the FAs used. DISCUSSION R. japonicum serogroup 123 clearly dominated the nodulation of field-grown soybeans (Table 6), as was anticipated from several earlier investigations (6, 14, 15). This dominance was consistent for each year of the study and for all cultivars examined, including the experimental cultivar chosen for its possible inoculant-selective properties (11). This dominance was not changed materially either by inoculation with R. japonicum of another serogroup or by plant maturity. Nodule occupancy by serogroup 123 was expressed in the presence of an indigenous population of R. japonicum within the soil that included strains of serogroups 110 and 138 and at least one other (probably several) unidentified serogroup(s) in addition to 123. All three components of the indigenous population were present in the soil at about the same cell density (Fig. 1), so that the ultimate success of serogroup 123 could not be attributed to a numerical predominance at the time of planting. The hypothesis tested in this study, that the success of serogroup 123 is due to an ability to outgrow other indigenous R. japonicum serogroups in the host rhizosphere, was not borne out. Early development of the seedlings was usually accompanied by enhanced growth of R. japonicum in the rhizospheres as compared with that in the fallow soil. The rhizosphere effects were slight, however, and no greater for serogroup 123 than for its competitors (Tables 3 and 4). TABLE 5. Combined populations of indigenous R. japonicum serogroups 110, 123, and 138a as a percentage of total bacteriab in 1980 and 1981 field plots % of total bacteria at indicated wk after planting Cultivar.. rhizosphere or fallow inocc NDd Chippewa inocc ND Chippewa Fallow a Total of specific FA counts. b Total bacteria as determined by acridine orange direct microscope count on membrane filters (19). c Seed inoculated with R. japonicum serogroup 110 at d planting. ND, Not determined. Since soybean nodules are already evident macroscopically within 2 to 3 weeks of seeding, the early rhizosphere samples should reflect the circumstances under which the nodulation was initiated, given the slow generation times of R. japonicum. The earliest rhizosphere samplings were at 1 and 2 weeks after planting, with seedlings only newly emerged and with the first trifoliate leaves still forming. Although unlikely, the possibilities cannot be excluded that serogroup 123 populations either were higher than those of competitors at a very early stage or had responded to the rhizosphere more rapidly than had the competitors and that these effects were no longer evident in the 1-week rhizosphere. These possibilities are now under investigation. The pattern of host rhizosphere colonization by the three populations of competing R. japonicum was similar for each of the two growing seasons. Rhizosphere effects were slight for newly emerged soybeans, increased gradually with increasing root development before flowering, increased markedly during flowering, and then declined at plant maturity. The pattern before flowering was similar for all three populations, but R/S ratios were much higher for serogroup 123 than for competitors during and after flowering. Quantitative differences seen in the rhizosphere enumerations from year to year can be accounted for in terms of differences in plot location, growing season, sampling times, and soil moisture content as affecting the adherence of rhizosphere soil to the roots. Early RIS ratios were very low both years, but they were higher in 1980 than in At 1 week after planting, the 1980 R/S values varied from 4 to 32 for the individual serogroups, and they varied from 14 to 50 after 20 weeks; R/S values for 1981 were substantially lower. It is noteworthy that during this period of nodule initiation, the dominant nodule occupant, serogroup 123, was less abundant in the rhizosphere than were serogroups 110 and 138. Moreover, the three groups combined made up only a small minority of the rhizosphere bacteria (Table 5). These data are consistent with the modest R/S ratios of rhizobia in early rhizospheres of host plants observed previously for soybeans (14) and for Phaseolus vulgaris (16), and they further indicate that extensive multiplication of the microsymbiont in the host rhizosphere is not a prelude to the initiation of nodulation as has been postulated previously (13). The factors responsible for the high populations of indigenous R. japonicum in the rhizospheres of host plants during flowering and pod fill are not clear. It is possible that all three serogroups, and especially 123, were responsive to the rhizosphere conditions imposed at this highly active stage of plant development. Such pronounced host rhizosphere effects may be a mechanism whereby diverse R. japonicum

5 VOL. 47, 1984 RHIZOSPHERE RESPONSE OF THREE R. JAPONICUM SEROGROUPS 611 TABLE 6. R. japonicum serogroup occupancy of early (5-week) and late (12- or 15- week) nodules on several soybean cultivars studied in the 1980 and 1981 field plots % of serogroup present in nodulesa Cultivar Inoculantb Others Ec Lc E L E L E L Chippewa None 84 (95)d 72 (93) 12 (0) 12 (18) 4 (3) 10 (10) 12 (5) 14 (5) (90) 78 (100) 14 (5) 4 (0) 14 (10) 24 (0) 18 (10) 26 (0) 138 (75) (90) (0) (10) (45) (30) (15) (0) Hodgson None 70 (93) 70 (88) 8 (10) 8 (3) 12 (3) 22 (10) 26 (8) 16 (13) (90) 62 (95) 4 (0) 14 (20) 20 (25) 6 (5) 24 (5) 38 (0) 138 (95) (90) (20) (0) (5) (25) (0) (5) None 84 (98) 78 (100) 6 (15) 6 (5) 6 (5) 10 (13) 18 (3) 38 (0) (90) 88 (95) 6 (0) 6 (20) 14 (25) 12 (5) 18 (5) 16 (0) 138 (95) (90) (20) (0) (5) (25) (0) (5) a Values listed report the serogroup present in both single- and mixed-infection nodules as determined by immunofluorescence. Mixedinfection frequency in 1980: serogroup 123, 14.5%; serogroup 110, 6.3%; serogroup 138, 2.0%. b Seeds inoculated with designated strain of R. japonicum at time of planting. c E, Early, 5-week nodules; L, late, 12- or 15-week nodules. d Numbers in parentheses report percentages of nodules occupied by a given serotype in 1981; numbers not in parentheses are for strains maintain a population base in the soil despite their inability to compete successfully in nodulation. Similarly, the ability to colonize nonhost rhizospheres, as in oats (Table 3), may also serve to maintain indigenous R. japonicum (though possibly at lower population densities) in the absence of host crops. The concept of a host rhizosphere intensively colonized during the flowering stage by diverse strains of microsymbionts along with other rhizosphere bacteria is supported by several aspects of this study. Rhizosphere responses were qualitatively consistent for the 2 years and were not restricted to the principal nodule occupant, all rhizosphere samples were carefully prepared to avoid nodules and root fragments, and all nodules appeared to be healthy and intact during flowering and pod fill. Despite this, however, the elevated host rhizosphere response might still be an artifact, since a single undetected or disrupted nodule in the rhizosphere sample could account for the high populations observed. The impact of nodule degradation is illustrated in the 15-week samples (Table 3) taken from mature and drying plants. Here the less successful competitors were strikingly less abundant than serogroup 123, which occupied nearly all of the nodules subject to disruption and bacteroid release. Studies of competition based on growth pouch data must be interpreted with caution. The dominance of serogroup 123 with respect to nodule occupancy observed so clearly in the analyses of field-grown nodules (Table 6) was much less clear-cut in pouch-grown nodules (Table 2). Serogroup 123 was less competitive by pouch analyses, whereas others, serogroup 138 for example, appeared much more competitive in pouches than in soil. Reasons for the discrepancies are not known, but clearly competition in the growth pouch under our conditions did not mimic the natural environment. ACKNOWLEDGMENTS This work was supported by grant for the U.S. Department of Agriculture Competitive Research Grants Office and by a grant to one of us (H.A.M.) from the International Development Research Centre, Canada. LITERATURE CITED 1. Belser, L. W., and E. L. Schmidt Serological diversity within a terrestrial ammonia-oxidizing population. Appl. Environ. Microbiol. 36: Bohlool, B. B., and E. L. Schmidt Nonspecific staining: its control in immunofluorescence examination of soil. Science 162: Bohlool, B. B., and E. L. Schmidt Persistence and competition aspects of Rhizobium japonicum observed in soil by immunofluorescence microscopy. Soil Sci. Soc. Am. Proc. 37: Clayet-Marel, J. C., and Y. Crozat Etude ecologique en immunofluorescence de Rhizobium japonicum dans le sol et al rhizosphere. Agronomie 2: Damirgi, S. M., L. R. Frederick, and I. C. Anderson Serogroups of Rhizobium japonicum in soybean nodules as affected by soil types. Agron. J. 59: Ham, G. E., V. B. Cardwell, and H. W. Johnson Evaluation of Rhizobium japonicum inoculants in soils containing naturalized populations of rhizobia. Agron. J. 63: Ham, G. E., L. R. Frederick, and I. C. Anderson Serogroups of Rhizobium japonicum in soybean nodules sampled in Iowa. Agron. J. 63: Hobbie, J. E., R. J. Daley, and S. Jasper Use of Nuclepore filters for counting bacteria by fluorescence microscopy. Appl. Environ. Microbiol. 33: Kapusta, G., and D. L. Rouwenhorst Interaction of selected pesticides on Rhizobium japonicum in pure culture under field conditions. Agron. J. 65: Kingsley, M. T., and B. B. Bohlool Release of Rhizobium spp. from tropical soils and recovery for immunofluorescence enumeration. Appl. Environ. Microbiol. 42: Kvien, C. S., G. E. Ham, and J. W. Lambert Recovery of introduced Rhizobiumjaponicum strains by soybean genotypes. Agron. J. 73: Lindemann, W. C., E. L. Schmidt, and G. E. Ham Evidence for double infection within soybean nodules. Soil. Sci. 118: Nutman, P. S The relation between nodule bacteria and the legume host in the rhizosphere and in the process of infection, p In. K. F. Baker and W. C. Snyder (ed.), Ecology of soilborne plant pathogens. University of California Press, Berkeley. 14. Reyes, V. G., and E. L. Schmidt Population densities of Rhizobium japonicum strain 123 estimated directly in soil and rhizospheres. Appl. Environ. Microbiol. 37: Reyes, V. G., and E. L. Schmidt Populations of Rhizobium japonicum associated with the surfaces of soil-grown roots. Plant Soil 61: Robert, F. M., and E. L. Schmidt Population changes and

6 612 MOAWAD, ELLIS, AND SCHMIDT persistence of Rhizobium phaseoli in soil and rhizospheres. Appl. Environ. Microbiol. 45: Schmidt, E. L Quantitative autecological study of microorganisms in soil by immunofluorescence. Soil Sci. 118: Schmidt, E. L., R. 0. Bankole, and B. B. Bohlool Fluorescent-antibody approach to study of rhizobia in soil. J. Bacteriol. 95: Schmidt, E. L., and E. A. Paul Microscopic methods for soil microorganisms, p In A. L. Page (ed.), Methods of soil analysis, part 2. Monograph 9, 2nd ed. American Society APPL. ENVIRON. MICROBIOL. for Agronomy, Madison, Wis. 20. Vest, G., D. F. Weber, and C. Sloger Nodulation and nitrogen fixation, p In B. E. Caldwell (ed.), Soybeans: improvement, production, and uses. Monograph 16. American Society for Agronomy, Madison, Wis. 21. Weaver, R. W., and L. R. Frederick A new technique for most-probable-number counts of rhizobia. Plant Soil 36: Weaver, R. W., and L. R. Frederick Effect of inoculum rate of competitive nodulation of Glycine max L. Merrill. II. Field studies. Agron. J. 66: