Bradyrhizobium japonicum Mutants Allowing Improved Soybean Yield in Short Season Areas with Cool Spring Soil Temperatures Hao Zhang, Fredric D’Aoust, Trevor C. Charles, Brian T. Driscoll, B. Prithiviraj, Donald L. Smith* ABSTRACT
penetration, nodule development and function) are inhibited by suboptimal RZTs (Lindemann and Ham, 1979; Lynch and Smith, 1993). The early infection stages are the most sensitive steps (Zhang and Smith, 1994). Gibson (1971) indicated that suboptimal RZTs prevent root hair infection more than nodule initiation, nodule development, or N assimilation. The addition of genistein has proven to be an effective means of generating increases in soybean N fixation and yield (Zhang and Smith, 1996b). However, genistein is expensive. Thus, it would be beneficial to find a different approach to solve the problem of soybean nodulation at low RZTs. Our research group (1998, unpublished data) used UV mutagenesis to create 10 mutants of USDA 110 in which expression of nod genes occurred independently of genistein, or any other plant-to-bacteria signal molecule. We do not know, however, how effective these mutants will be on plant development and soybean yield under field conditions in a cool-spring, short-season area. Hence, the goal of this study was to evaluate mutants that potentially can improve soybean plant development and grain yield in short season areas with low spring soil temperatures.
In short-season production areas, cool soil temperature is a major factor potentially limiting soybean [Glycine max (L.) Merr.] plant growth and yield. Genistein (4ⴕ,5,7-trihydroxyisoflavone) is a signal compound secreted from soybean roots and is essential to establishment of the soybean–Bradyrhizobium japonicum N-fixing symbiosis. The addition of genistein to soybean inocula has proven to be an effective means of generating increases in N fixation and yield; however, genistein is costly. We used ultraviolet (UV) mutagenesis to make 10 mutants of USDA 110 that express the nod genes without exposure to genistein. A field experiment was conducted at the Lods Agronomy Research Centre in southwestern Quebec in 1998 and 1999. The treatments consisted of factorial combinations of inoculant type [no inoculant (control) and inoculants containing the mutants or the wildtypes 532 C or USDA 110] and soybean cultivar (OAC Bayfield and Maple Glen). Inoculation with mutant strains Bj 30055 and Bj 30058 resulted in greater soybean yields than inoculation with 532 C (6.2% increase, averaged across the 2 yr) or the wildtype USDA 110 (9.9% increase, averaged across the 2 yr). These increases were largely due to increases in pod and seed number. These results showed that mutants that express nod genes in the absence of plant-to-bacteria signal compounds can help to overcome the low temperature limitation of soybean nodulation leading to improved growth and yield of soybean crops grown in areas with cool spring soils.
T
he formation of effective soybean nodules is a complex and highly regulated process that requires production and exchange of specific molecular signals between the host plant and the bacterial symbiont. The most effective of the plant-to-bacteria signal molecules is the isoflavone genistein (Kosslak et al., 1987), which is secreted by soybean roots. Genistein induces B. japonicum nodulation (nod) genes, by stimulating the NodD1 or NodV proteins. The products of the nod genes are Nod factors, which have been identified as lipo-chitooligosaccharides (LCO) (Long, 1996). Nod factors are responsible for many of the early stages of nodule development (Sanjuan et al., 1992). Soybean is a subtropical legume that requires root zone temperatures (RZTs) in the 25 to 30⬚C range for optimal symbiotic activity (Jones and Tisdale, 1921; Dart and Day, 1971). At lower temperatures, the expression of the nod genes is suboptimal, resulting in a delay in the onset of nodulation (Zhang and Smith, 1996b). In regions such as the soybean production areas of Canada, low soil temperature is potentially a major factor limiting soybean growth and symbiotic N fixation (Whigham and Minor, 1978). In the establishment of the N fixing symbiosis, all stages (root hair curling, infection thread formation and
MATERIALS AND METHODS Isolation of Genistein-Independent B. japonicum Mutants Strain USDA 110 was included in this work as it is widely used in inoculants in the USA. Its genetics have been well characterized, and the mutants are derived from this strain. The wildtype B. japonicum strain USDA 110 was subjected to UV mutagenesis (30 000 J) to generate mutants. A reporter plasmid (pZB32, obtained from G. Stacey, Univ. of Tennessee, Knoxville, TN), consisting of a nodY::lacZ gene fusion, was then introduced into the mutagenized cells by electroporation (Hatterman and Stacey, 1990). Tetracycline {[4S-(4alpha, 4aalpha,5aalpha,6beta,12aalpha)]-4-(dimethylamino)-1,4,4a, 5,5a,6,11,12a-octahydro-3,6,10,12,12a-pentahydroxy-6-methyl-1, 11-dioxo-2-naphthacenecarboxamide)} resistant transformants were selected on yeast-extract mannitol (YEM) plates containing X-Gal (5-Bromo-4-chloro-3-indolyl-beta-D-galactoside; 80 g mL⫺1 ), but no genistein. Blue colonies were considered to be putative mutants. To confirm the stability of the isolated putative constitutive mutants, ten candidate strains were cured of the reporter plasmid. In nod gene signal transduction, the nod D gene product, which is constitutively expressed, senses the isoflavonoids and acts as a transcription factor for Nod factor synthesis. The nod Y gene is a part of the nod YABC operon. The role of its coded protein is not clear; however, this gene is expressed only in response to isoflavonoids, and this has been proven through the use of a nod Y::lac Z construct (Stacey 1995). The plasmid was reintroduced by conjugation with Escherichia coli S 17.1 carrying pZB32, and the phenotype was verified on YEM (Vincent, 1970) and tryptone-yeast (Beringer, 1974) plates with tetracy-
H. Zhang, B. Prithiviraj, and D.L. Smith, Plant Science Dep.; F. D’Aoust and B.T. Driscoll, Dep. of Natural Resource Science, Macdonald Campus of McGill Univ., 21111 Lakeshore Road, Ste-Annede-Bellevue, QC, Canada H9X 3V9; T.C. Charles, Dep. of Biology, Univ. of Waterloo, Waterloo, ON, Canada N2L 3G1. Received 29 May 2001. *Corresponding author (
[email protected]).
Abbreviations: LCO, lipo-chito-oligosaccharides, RZT, root zone temperatures; UV, ultraviolet; YEM, yeast-extract mannitol.
Published in Crop Sci. 42:1186–1190 (2002).
1186
ZHANG ET AL.: BRADYRHIZOBIUM JAPONICUM MUTANTS AND SOYBEAN YIELD
cline and X-Gal. All putative mutants exhibited blue colonies under such conditions, confirming the mutation.
Bacterial Material The field experiment used cured forms of Bj 30050, Bj 30051, Bj 30052, Bj 30053, Bj 30054, Bj 30055, Bj 30056, Bj 30057, Bj 30058, Bj 30059, and unmutated strains 532 C and USDA 110 as inoculants. The inocula were produced by culturing the above strains and mutants in YEM broth (Vincent, 1970) in 250-mL flasks shaken at 125 rpm at 25⬚C. When the subcultures reached mid-log phase, pure medium was used to dilute the inocula to an OD620 of 0.08 (equivalent to 108 cells mL⫺1 ) (Bhuvaneswari et al., 1980).
1187
seeds along the furrow. Alcohol sterilization of the implements was used to prevent cross contamination throughout planting and all subsequent data collection procedures. Following emergence, seedlings reached a stand of 500 000 plants ha⫺1 (20 plants m⫺1 of row, with an average interplant distance of 5 cm within the row and 40 cm between rows). Where stands were excessive, they were hand thinned at the seedling stage.
Experimental Design The field experiment was a 13 by 2 factorial organized in a randomized complete block design with four blocks. Treatments were comprised of factorial combinations of 13 inocula and two soybean cultivars (Maple Glen and OAC Bayfield).
Site Preparation and Field Layout
Harvest and Data Collection
Each plot was 4.5 ⫻ 3.2 m and consisted of eight rows of plants, with 0.4 m between rows. The space between adjacent blocks was 1 m. The space between plots was 0.8 m. The experiments were carried out at the Lods Agronomy Research Center, McGill University, Macdonald Campus in Quebec, Canada. The soil was a Chicot light sandy loam (mixed, frigid Typic Hapludalfs) in 1998 and 1999. Oat and barley were planted in 1997, and corn was planted in 1998 at the 1998 and 1999 sites, respectively. In each case, the stems and leaves were incorporated into the soil after harvest through fall plowing. In both years, K and phosphate were provided by a spring application of muriate of potash (110 kg ha⫺1 of 0-0-60 NPK) and triple superphosphate (220 kg ha⫺1 of 0-46-0 NPK) following the recommendations of a soil test.
Daily average air temperature, average soil temperatures at a depth of 5 cm, and precipitation were recorded at the Macdonald Campus weather station (Fig. 1). The classification of soybean growth stages followed those of Fehr and Caviness (1977). To collect data on the development of leaf number and leaf area, plant samples (10 plants in each) were harvested from each plot at four developmental stages: (i) V3, three nodes on the main stem with fully developed leaves beginning with the unifoliolate nodes (20 June 1998 and 24 June 1999); (ii) R1, one open flower at any node on the main stem (28 June 1998 and 2 July 1999); (iii) R4, pods 2 cm long at one of the four uppermost nodes on the main stem with a fully developed leaf (28 July 1998 and 4 Aug. 1999), (iv) R8, 95% of the pods having reached their mature pod color (14 Aug. 1998 and 18 Aug. 1999). Leaf number and leaf area per plant were determined using a Delta-T area meter (Delta-T Devices Ltd., Cambridge, UK). At harvest maturity, pod number and seed number per plant were determined from 10 plants of each plot. The remainder of the plants in the plot were harvested with a small plot combine (Wintersteiger, Salt Lake City, UT) at harvest maturity, and then oven-dried at 70⬚C for at least 48 h before being weighed for yield determination, thus the yield reported here is based on a 0% moisture basis. The N content of the seeds was determined by Kjeldahl analysis (Kjeltec system, Tecator
Planting Method Seeds of the soybean cultivars Maple Glen and OAC Bayfield were surface-sterilized in a solution containing 2% sodium hypochlorite (obtained by diluting household bleach, containing 5.25% sodium hypochlorite, with distilled water), then rinsed several times with distilled water. These seeds were planted on 15 May 1998 and 18 May 1999. Twenty milliliters of mutant or strain culture (OD620 ⫽ 0.01) or YEM (Control) per 1 m of row was applied evenly by syringe directly onto the
Fig. 1. The average daily air temperature, soil temperature (at a depth of 5 cm) and precipitation during the 1998 and 1999 soybean growing seasons at Ste-Anne-de-Bellevue, QC, Canada.
1188
CROP SCIENCE, VOL. 42, JULY–AUGUST 2002
AB, Hoganas, Sweden). Seed protein concentration was estimated by multiplying N concentration by 6.25.
Statistical Analysis Results were analyzed statistically by analysis of variance using the Statistical Analysis System computer package (SAS Institute, 1988). When analysis of variance showed significant treatment effects, the LSD test was applied to make comparisons among the means at the 0.05 level of significance (Steel and Torrie, 1980). When differences occurred at levels of significance between 0.05 and 0.1, they are noted in the text.
RESULTS Inoculation with strains Bj30050, Bj30051, Bj30052, Bj30053, Bj30054, Bj30056, Bj30057, and Bj30059 did not result in significant increases in leaf area, grain yield, grain protein, or total protein content as compared with uninoculated control treatment (data not shown). For this reason, the remainder of this section will focus on Bj 30055, Bj 30058, USDA 110, and 532 C. There were no interactions between inocula and cultivar.
Effect of Inoculation on Plant Development Inoculation with B. japonicum mutants and strains accelerated soybean leaf area development. The leaf area increase was greatest at the R4 stage (Fig. 2). The increase resulting from Bj 30055 and Bj 30058 was greater than that resulting from 532 C, and the increase resulting from 532 C was greater than that resulting from USDA 110. At R4, compared with 532 C, Bj 30055 and Bj 30058 increased leaf area by 7 to 31% (averaged across 1998 and 1999), respectively. At R4, compared
with USDA 110, Bj 30055 and Bj 30058 increased leaf area by 5 to 37% (averaged across 1998 and 1999), respectively (Fig. 2). The inoculation treatments did not affect leaf number (data not shown). The leaf area of Maple Glen was less than that of OAC Bayfield.
Effect of Inoculation Treatments on Soybean Yield and Grain Protein Content Inoculation with the mutant Bj 30055 increased grain yield by 3.2% in 1998 and 7.4% in 1999, when compared with 532 C (Tables 1 and 2), while Bj 30058 increased grain yield by 5.7% in 1998 and 8.5% in 1999. Compared with USDA 110, inoculation with mutant Bj 30055 increased grain yield by 5.4% in 1998 and 12.5% in 1999, while inoculation with mutant Bj 30058 increased grain yield by 8% in 1998 and 13.7% in 1999 (Tables 1 and 2). Inoculation with Bj 30055 and Bj 30058 increased grain protein more than inoculation with 532 C and USDA 110, by 4.3 to 13.0%, respectively (averaged across 1998 and 1999) (Tables 1 and 2). Inoculation with Bj 30055 and Bj 30058 increased total protein more than inoculation with 532 C and USDA 110, by 4.2 to 12.6%, respectively (averaged across 1998 and 1999). Analyzing the yield components, we found that pod number was increased by Bj 30055 and Bj 30058 relative to 532 C and USDA 110 in 1998 and 1999. Seed number was increased by Bj 30055 and Bj 30058 relative to 532 C and USDA 110 in 1998 (P ⫽ 0.09) and 1999. In the case of 100-seed weight, there was no difference among the mutants Bj 30055, Bj 30058, strain 532 C and USDA 110 in most cases (Tables 1 and 2). OAC Bayfield produced more grain yield than Maple
Fig. 2. The effects of inoculated mutants/strains of Bradyrhizobium japonicum on soybean (cultivars Maple Glen and OAC Bayfield) plant leaf area during the 1998 and 1999 seasons. Each value is plotted as the mean ⫹ S.E. (n ⫽ 10). The plant development stage at each harvest was: 1. V3 (35 DAP in 1998 and 36 DAP in 1999); R1 (43 DAP in 1998 and 44 DAP in 1999); R4 (73 DAP in 1998 and 77 DAP in 1999) and R8 (90 DAP in 1998 and 91 DAP in 1999).
1189
ZHANG ET AL.: BRADYRHIZOBIUM JAPONICUM MUTANTS AND SOYBEAN YIELD
Table 1. Main effects of Bradyrhizobium japonicum strains and mutants and soybean [Glycine max (L.) Merr.] cultivar grain yield, grain protein yield, total protein yield, 100-seed weight, pod number per plant, and seed number per plant at the final harvest in 1998.† Yield Treatment
Pod
Seed No. plant⫺1
532 C Bj 30055 Bj 30058 USDA 110 Uninoculated
22.5b 23.6a 24.1a 21.8b 19.6c
58.4a 59.6a 58.8a 57.2b 55.1c
Maple Glen CAC Bayfield
22.1b 23.7a
56.8b 58.1a
100-seed weight g Inoculant 20.1a 21.6a 20.8a 19.7a 17.4b Cultivar 18.9a 19.5a
Grain protein
Total protein
Grain
t ha⫺1 0.69b 0.72a 0.73a 0.68b 0.65c
1.17b 1.25a 1.23a 1.11c 1.06d
1.90b 1.96a 2.01a 1.86b 1.80c
0.67a 0.71a
1.15a 1.18a
1.87b 1.93a
† Mean yield and 100-seed weight were based on seed machine harvested from each plot at harvest maturity. Mean pod number and seed number per plant were derived from 10 plants from each plot. Means within the same column and factor followed by the same letter are not different (P ⱕ 0.05) by an ANOVA protected LSD test.
Glen in 1998 and 1999. Analysis of the grain yield components indicated that the grain increases came from the combination of pod number, seed number and, in one case, 100-seed weight increases (Tables 1 and 2), showing that the benefits of the mutants extended from at least the beginning of pod set to late grain filling. In 1998 and 1999, grain yield, pod, and seed numbers were greater for OAC Bayfield than Maple Glen. The 100-seed weight was greater for OAC Bayfield than Maple Glen in 1999 but not in 1998 (Tables 1 and 2). There were no differences between OAC Bayfield and Maple Glen for grain protein, while OAC Bayfield produced more total protein than Maple Glen.
DISCUSSION In North American soybean production areas, such as those in Canada, low soil temperature is one of the main potential limitations to soybean N fixation (Zhang and Smith, 1994, 1996a). Further, low soil temperature is a problem in areas where early production practice is implemented, for example, early planting in the lower Mississippi delta, to avoid drought (Heatherly and Hodges, 1999). In the present experiment, the seasonal soil temperature data from 1998 and 1999 showed that the average daily RZTs at a depth of 5 cm were below 25⬚C, especially during early stages of growth (Fig. 1). Genistein is an important plant-to-bacteria signal se-
creted from soybean roots. It can induce nod gene expression, but at low temperatures, higher genistein concentrations were required for maximum nod gene expression than at an optimal temperature (25⬚C) (Zhang et al., 1996). This means that the efficiency of genistein, in terms of inducing nod gene expression, is decreased under cool temperature conditions. At the same time, genistein concentration in soybean roots is low (Zhang and Smith, 1996b) at cool temperatures. Thus, lack of genistein or related plant-to-bacteria signal compounds can be a limiting factor during the establishment of the soybean N fixing symbiosis under low temperature conditions. For these reasons, we developed mutants that can express nod genes in the absence of genistein or similar plant-to-bacteria signals. Our laboratory results (1999, unpublished data) showed that all of the mutants had the capacity to form nodules on soybean roots at room temperature. However, while they were shown not to produce Nod factor independently of genistein addition, they produced more Nod factor than 532 C and USDA 110 at low temperatures (15 and 17⬚C); the increase in Nod factor production ranged from one to three fold (1999, unpublished data). The role of Nod factors in nodulation has been extensively studied (Denarie et al. 1996; Long 1996). Skorupska and Krol (1995) found that mutants of Rhizobium leguminosarum bv. Trifolii 24.1 synthesized no detectable acidic exopolysaccharide and were unable to induce
Table 2. Main effects of Bradyrhizobium japonicum strains and mutants and soybean [Glycine max (L.) Merr.] cultivar on soybean grain yield, grain protein yield, total protein yield, 100-seed weight, pod number per plant, and seed number per plant at the final harvest in 1999.† Yield Treatment
Pod
Seed No. plant⫺1
532 C Bj 30055 Bj 30058 USDA 110 Uninoculated
24.2b 25.4a 26.1a 23.2b 21.3c
57.1b 58.3a 58.7a 56.2b 53.1c
Maple Glen CAC Bayfield
24.4b 25.7a
55.7b 57.4a
100 seed weight g Inoculant 19.1b 19.3b 21.6a 18.2b 16.7c Cultivar 18.7b 20.4a
Grain protein
Total protein
Grain
t ha⫺1 0.73b 0.76a 0.78a 0.69c 0.67c
1.18b 1.23a 1.26a 1.16b 1.08c
1.76b 1.89a 1.91a 1.68c 1.57d
0.70a 0.72a
1.07a 1.10a
1.76b 1.85a
† Mean yield and one hundred seed weight were based on seed machine harvested from each plot at harvest maturity. Mean pod number and seed number per plant were derived from ten plants from each plot. Means within the same column and factor followed by the same letter are not different (P ⱕ 0.05) by an ANOVA protected LSD test.
1190
CROP SCIENCE, VOL. 42, JULY–AUGUST 2002
N fixing nodules on clover plants. As to the 8 mutants having no effect in the field, we supposed that the change(s) in gene structure that allowed them to produce more Nod factor under low temperature, and perhaps, other unknown changes in gene structure, made them ineffective under field conditions. Perhaps they were unable to survive the much more demanding conditions of the field environment, or to compete effectively with indigenous B. japonicum strains. Zhang and Smith (1994) demonstrated that low RZTs delay all of the steps in the infection of soybean roots by B. japonicum. Added genistein activated B. japonicum nod genes and soybean inoculation, nodulation events, and N fixation began 2 to 5 d earlier at the suboptimal RZTs (Zhang and Smith, 1995). Sprent (1979) speculated that, because of the initial exponential rise, an increase of 10% in the period of nodule activity of a grain legume, particularly between the onset of N fixation and the attainment of maximum fixation, could double the season level of N fixed. We suppose inoculation with Bj 30055 and Bj 30058 leads to better plant growth at low temperatures than inoculation with 532 C and USDA 110 because the mutants can express nod genes without the addition of genistein, and because they are able to produce higher levels of LCOs in the presence of moderate concentrations of genistein at low temperatures. In other words, inoculation with Bj 30055 and Bj 30058 should result in more N fixation than with 532 C and USDA 110 because more LCO is produced under low temperature conditions. This possibility was supported by greater leaf areas for plants inoculated with Bj 30055 and Bj 30058 than 532 C and USDA 110 (Fig. 2) and greater seed protein and total protein levels (Tables 1 and 2). In comparison with USDA 110, mutants Bj 30055 and Bj 30058 increased nodule number by 37 and 29% in 1998, respectively, and 32 and 20% in 1999, respectively, resulting in higher amounts of N fixed (1998–1999, unpublished data). Addition of genistein preincubated B. japonicum or direct application of genistein to soil also resulted in significant increases in the yield in both sterilized and unsterilized soils under cool season conditions. At the unsterilized site, the increase in yield due to preincubated B. japonicum strain USDA 110 with genistein was 25.5%, while at the sterile site the increase was 15.7%, suggesting that the interaction is more complex (Zhang and Smith, 1996a). Thus, in the present study factors other than the direct effect of inoculation of genistein independent mutants might have contributed to the observed increase in soybean yield. Further, it would be interesting to evaluate the effects of inoculation of these mutants under warm season conditions. In summary, this is the first report indicating that when soybean is planted in an area with cool spring soil conditions, mutants able to express nod genes without plant-to-bacteria signal molecules increased leaf area, total protein, grain protein, and grain yield compared with currently used strains. REFERENCES Beringer, J.E. 1974. R-factor transfer in Rhizobium leguminosarum. J. Gen. Microbiol. 84:188–198.
Bhuvaneswari, T.V., R.N. Goodman, and W.D. Bauer. 1980. Early events in the infection of soybean (Glycine max L. Merr.) by Rhizobium japonicum. I. Location of infectible root cells. Plant Physiol. 66:1027–1031. Dart, P., and J. Day. 1971. Effects of incubation temperature and oxygen tension on nitrogenase activity of legume root nodule. Plant Soil (special volume):167–184. Denarie, J., F. Debelle, and J.C. Prome. 1996. Rhizobium lipo-chitooligosaccharide nodulation factors: signaling molecules mediating recognition and marphogenesis. Ann. Rev. Biochem. 665:503–535. Fehr, W.R., and C.E. Caviness. 1977. Stages of soybean development. Special Report, Agric. and Home Economics Exp. Stn., Iowa State Univ., Iowa City, IA. Gibson, A.H. 1971. Factors in the physical and biological environment affecting nodulation and nitrogen fixation by legumes. Plant Soil 35:139–152. Hatterman, D.R., and G. Stacey. 1990. Efficient DNA transformation of Bradyrhizobium japonicum by electroporation. Appl. Environ. Microbiol. 56:833–836. Heatherly, L.G., and H.F. Hodges. 1999. Soybean production in the midwest. CRC Press, Boca Raton, FL. Jones, F.R., and W.B. Tisdale. 1921. Effect of soil temperature upon the development of nodules on the roots of certain legumes. J. Agric. Res. 22:17–37. Kosslak, R.M., R. Bookland, J. Barkei, H. Paaren, and E.R. Appelbaum. 1987. Induction of B. japonicum common nod genes by isoflavones isolated from Glycine max. Proc. Natl. Acad. Sci. USA 84:7428–7432. Lindemann, W.C., and G.E. Ham. 1979. Soybean plant growth, nodulation, and nitrogen fixation as affected by root temperature. Soil Sci. Soc. Am. J. 43:1134–1137. Long, S.H. 1996. Rhizobium symbiosis: Nod factors in perspective. Plant Cell 8:1885–1898. Lynch, D.H., and D.L. Smith. 1993. Soybean (Glycine max ) nodulation and N2-fixation as affected by exposure to low root-zone temperature. Physiol. Plant. 88:212–220. Sanjuan, J., R.W. Carlson, H.P. Spaink, U.R. Bhat, W.M. Barbour, J. Glushka, and G. Stacey. 1992. A 2-O-methylfucose moiety is present in the lipo-oligosaccharide nodulation signal of Bradyrhizobium japonicum. Proc. Nat. Acad. Sci. USA 89:8789–8793. SAS Institute. 1988. SAS user’s guide. Version 6, 4th ed. SAS Inst., Cary, NC. Skorupska, A., and J. Krol. 1995. Nodulation response of exopolysaccharide deficient mutants of Rhizobium leguminosarum bv. trifolii to addition of acidic exopolysaccharide. Microbiol. Res. 150:297– 303. Sprent, J.I. 1979. The biology of nitrogen fixing organisms. McGrawHill, NY. Stacey, G. 1995. Bradyrhizobium japonicum nodulation genetics. FEMS Microbiol. Lett. 127:1–9. Steel, R.G.D., and J.H. Torrie. 1980. Principles and procedures of statistics: A biometric approach. McGraw-Hill, New York. Vincent, J.M. 1970. A manual for the practical study of root nodule bacteria. Blackwell Scientific Publ., Oxford, UK. Whigham, D.K., and H.C. Minor. 1978. Agronomic characteristics and environmental stress. p. 77–118. In G.A. Northman (ed.) Soybean physiology, agronomy, and utilization. Academic Press, New York. Zhang, F., T.V. Charles, B. Pan, and D.L. Smith. 1996. Inhibition of the expression of Bradyrhizobium japonicum nod genes at low temperatures. Soil Biol. Biochem. 28:1579–1583. Zhang, F., and D.L. Smith. 1994. Effects of low root zone temperature on the early stages of symbiosis establishment between soybean [Glycine max. (L.) Merr] and Bradyrhizobium japonicum. J. Exp. Bot. 279:1467–1473. Zhang, F., and D.L. Smith. 1995. Preincubation of Bradyrhizobium japonicum with genistein accelerates nodule development of soybean [Glycine max (L.) Merr.] at suboptimal root zone temperatures. Plant Physiol. 180:961–968. Zhang, F., and D.L. Smith. 1996a. Inoculation of soybean [Glycine max (L.) Merr.] with genistein-preincubated Bradyrhizobium japonicum or genistein directly into soil increases soybean protein and dry matter yield under short season conditions. Plant Soil 179: 233–241. Zhang, F., and D.L. Smith. 1996b. Genistein accumulation in soybean (Glycine max [L.] Merr.) root systems under suboptimal root zone temperatures. J. Exp. Bot. 47:785–792.
