Peter H. Graham

Active N₂ fixation in soybean [Glycine max (L.) Merr.] in Minnesota is limited by cool early‐season soil temperatures and by postflowering nodule senescence. This study examined variation in onset of nodule senescence among Maturity Group I soybean cultivars and sought traits associated with this variation. Host genotype markedly affected onset of crown nodule senescence. For most cultivars, crown nodule fresh weight and specific nodule activity (SNA) peaked 31 to 38 d after emergence (DAE) and declined rapidly thereafter. In contrast, maximum crown nodule fresh weight in ‘Hardin’ and ‘Hodgson 78’ did not occur until 52 DAE, and SNA was still high 45 to 52 DAE. Two cultivars, Chippewa and Alpha, that exhibited early change in crown nodule mass and SNA, accumulated glyceollin I 10 to 45 DAE at rates significantly greater than for Hardin and Hodgson 78. The four cultivars also differed in phenylalanine ammonia lyase (PAL) and chalcone synthase (CHS) gene expression, nodule protease activity, and polyamine accumulation. Morphological changes within the nodule paralleled the biochemical differences, with Chippewa nodules 45 DAE showing more conspicuous deterioration than was evident in Hardin. Because crown‐nodule mass and nitrogenase activity in Hardin and Hodgson 78 declined later than in other Maturity Group I cultivars, with less evidence of host–strain incompatibility, these two lines may have value in breeding programs to extend the period of active nodulation and N₂ fixation in soybean.

We have previously reported a soybean plant introduction, PI 437153A, which was ineffectively nodulated (definition of G. Vest, D. F. Weber, and C. Sloger, Agronomy 16:353–390, 1973) by Bradyrhizobium japonicum strain UMR (University of Minnesota Rhizobium) 161. In this study, we further characterize this nodulation system by (i) examining the nodulation of 21 B. japonicum strains, representing 15 serogroups, with PI 437153A, (ii) ascertaining the inheritance of the ineffective nodulation, and (iii) characterizing the symbiosis both morphologically and physiologically. Only USDA 126 (serogroup 125) and UMR 161 failed to produce effective nodules with PI 437153A. Segregation among F₂ and F₃ plants from a cross between cv. Lambert and PI 437153A was consistent with control of this trait by a single, dominant gene. In reciprocal grafts between cv. Lambert and PI 437153A, ineffective nodulation was determined by the root genotype, and was independent of plant growth temperature and inoculant concentration. The nodule dry weight of UMR 161-inoculated plants of PI 437153A plants was approximately 10-fold less than that of comparable plants inoculated with USDA 110; nodule soluble protein levels were also significantly reduced. Nitrogenase activity and leghemoglobin could not be detected in the ineffectively nodulated phenotype, but 1 to 6 days after inoculation (DAI) phenylalanine ammonialyase RNA level was higher in the ineffectively nodulated phenotype than in the effectively nodulated one. Microscopic examination of effective and ineffective nodules 20 and 35 DAI revealed major morphological differences. Ineffective nodules contained large, deteriorating regions near apparently infected plant cells at 35 DAI, and schlerenchyma cells that were stained red, suggestive of lignin deposition. These changes could be the result of incompatibility between host and Bradyrhizobium, though the lack of chalcone synthase expression and the appearance of infected cells within the ineffective nodules suggest that some later steps in the infection process do occur.