Differential responses of chickpea (Cicer arietinum L.) — Rhizobium combinations to saline soil conditions

Summary Chickpea cultivars (Cicer arietinum L.) and their symbiosis with specific strains of Rhizobium spp. were examined under salt stress. The growth of rhizobia declined with NaCl concentrations increasing from 0.01 to 2% (w : v). Two Rhizobium spp. strains (F-75 and KG 31) tolerated 1.5% NaCl. Of the 10 chickpea cultivars examined, only three (Pusa 312, Pusa 212, and Pusa 240) germinated at 1.5% NaCl. The chickpea — Rhizobium spp. symbiosis was examined in the field, with soil varying in salinity from electrical conductivity (EC) 4.5 to EC 5.2 dSm^-1, to identify combinations giving satisfactory yields. Significant interactions between strains and cultivars caused differential yields of nodules, dry matter, and grain. Four chickpea — Rhizobium spp. combinations, Pusa 240 and F-75 (660 kg ha^-1), Pusa 240 and IC 76 (440 kg ha^-1), Pusa 240 and KG 31 (390 kg ha^-1), and Pusa 312 and KG 31 (380 kg ha^-1), produced significantly higher grain yields in saline soil.     

Effect of high temperature on plasmid curing of Rhizobium spp. in relation to nodulation of pigeonpea [Cajanus cajan (L.) Millsp.]

Summary Fifty-six isolates of Rhizobium and Bradyrhizobium spp. (Cajanus) were studied for their plasmid profile and N₂-fixation efficacy. One to three plasmids were reproducibly detected in all the Rhizobium spp. strains but no plasmid was detected in the Bradyrhizobium spp. strains. Rhizobium sp. strain P-1 was mutagenized by Tn5 and three nod^– and six nod⁺fix^– were screened for symbiotic parameters. Neomycin-sensitive mutants were isolated by elevated temperatrue (40°C) from tranconjugants carrying Tn5 insertions. The high temperature cured these mutants from the single large plasmid present in the parent strain P-1. All these cured mutants were nod^–, indicating that the genes for nodulation were present on this plasmid, which is readily cured at a high temperature (40°C). The high temperature in the semi-arid zones of Haryana could be responsible for the low nodulation of pigeonpea because the plasmid carrying the nodulation genes is cured at 40°–45°C giving rise to non-nodulating mutants.     

Variation in nodulation and N2 fixation by the Gliricidia sepium/Rhizobium spp. symbiosis in a calcareous soil

Summary Variation in nodulation and N₂ fixation by the Gliricidia sepium/Rhizobium spp. symbiosis was studied in two greenhouse experiments. The first included 25 provenances of G. sepium inoculated with a mixture of three strains of Rhizobium spp. N₂ fixation was measured using the ¹⁵N isotope dilution method 12 weeks after planting. On average, G. sepium derived 45% of its total N from atmospheric N₂. Significant differences in fixation were observed between provenances. The percentage of N derived from atmospheric N₂ ranged from 26 to 68% (equivalent to 18–62 mg N plant^-1) and was correlated with total N in the plant (r=0.70; P=0.05). The second experiment included six strains of Rhizobium spp. and two methods of inoculation and the plants were harvested 14,35 and 53 weeks after planting. In the first harvest significant differences were found between the number of nodules and the percentage and amount of N₂ fixed. There was also a significant correlation between the number of nodules and the amount of N₂ fixed (r=0.92; P=0.05). In the final harvest no correlation was observed, although there were significant differences between the number of nodules and the percentage of N derived from the atmosphere. The amount of N₂ fixed increased with time (from an average of 27% at the first harvest to 58% at the final harvest) and was influenced by the Rhizobium spp. strain and the method of inoculation. It ranged from 36% for Rhizobium sp. strain SP 14 to 71% for Rhizobium SP 44 at the last harvest. Values for the percentage of atmosphere derived N₂ obtained by soil inoculation were slightly higher than those obtained by seed inoculation.     

Nodule-specific expression of Rhizobium meliloti symbiotic promoters P1 and P2 in chickpea-Rhizobium sp. symbiosis

Developmentally specific expression of Rhizobium spp. genes involved in symbiotic N₂ fixation is known to operate through cascade regulation of various nif and fix operons. Fusion constructs of lacZ under symbiotic promoters P1 (for nifHDK operon) and P2 (for fixABCX operon) of Rhizobium meliloti were mobilized into Rhizobium spp. (Cicer) strains Rcd301 and RCR13. The assays for -galactosidase activity to monitor the expression of lacZ under these promoters was performed in host backgrounds of Escherichia coli, R. meliloti, and Rhizobium spp. (Cicer). The enzyme assays indicated significant levels of expression from P1 and P2 promoters in chickpea rhizobia, specifically in symbiotic cells from nodules. However, as in R. meliloti, these promoters did not induce strong expression in free-living cells of Rhizobium spp. (Cicer). This indicates functional homology of R. meliloti promoters in rhizobium spp. (Cicer). Functional cross-reactivity of trans regulatory factors like NtrA, NtrC, and NifA between these rhizobia seems evident from the nodule-specific expression of P1 and P2 cis elements.     

Nitrogen-fixing potential of Acacia mangium and Acacia auriculiformis seedlings inoculated with Bradyrhizobium and Rhizobium spp.

Summary Two Australian Acacia species, A. mangium and A. auriculiformis were inoculated in vitro with eight strains of Bradyrhizobium spp. and two strains of Rhizobium spp. On the two plant species, only Bradyrhizobium spp. strains formed effective N₂-fixing nodules. A. mangium, which nodulates effectively with a restricted range of Bradyrhizobium spp. strains, is a specific host compared to A. auriculiformis. A. auriculiformis is assumed to be a promiscuous host because it nodulates effectively with a wide range of Bradyrhizobium spp. strains. Nodule efficiency as expressed by the ratio of N₂ fixed to nodule dry weight appeared to be higher in A. auriculiformis (0.44–0.81) than in A. mangium (0.23–0.55).     

Response of the pigeonpea-Rhizobium symbiosis to salinity stress: variation among Rhizobium strains in symbiotic ability

Summary There were significant differences among pigeonpea [Cajanus cajan (L.) Millsp] Rhizobium sp. strains (IC 3506, IC 3484, IC 3195, and IC 3087) in their ability to nodulate and fix N₂ under saline conditions. Pigeonpea plants inoculated with IC 3087 and IC 3506 were less affected in growth by salinity levels of 6 and 8 dS m^-1 than plants inoculated with the other strains. For IC 3506, IC 3484, and IC 3195, there was a decrease in the number of nodules with increasing salinity, while the average nodule dry weight and the specific nitrogenase activity remained unaffected. However, in IC 3087, the number of nodules increased slightly with increasing salinity. Leaf-P concentrations increased with salinity in the inoculated plants irrespective of the Rhizobium sp. strain, and leaf-N concentrations decreased with increasing salinity in IC 3484 and IC 3195 only. Shoot-Na and-Cl levels were further increased in these salt-sensitive strains only at 8 dS m^-1. Therefore there may be scope for selecting pigeonpea Rhizobium sp. symbioses better adapted to saline conditions. The Rhizobium sp. strains best able to form effective symbioses at high salinity levels are not necessarily derived from saline soils.Submitted as JA No. 919 by the International Crops Research Institute for the Semi-Arid Tropics (ICRISAT)     

Direct monitoring of Rhizobium sp. (Cicer) in nodules and soil using lac Z fusions

We constructed lacZ fusions in Rhizobium sp. (Cicer) by random Tn5-lacZ mutagenesis. The lacZ⁺ fusants formed blue colonies on a Rhizobial minimal medium containing 5-bromo-4-chloro-3-indolyl--D-galactopyranoside (X-gal). Rhizobium sp. (Cicer) fusant HSL-2 was identified in nodules and soil in a mixed population on the basis of the lacZ⁺ phenotype. Nodule occupancy of inoculated Rhizobium sp. (Cicer) HSL-2 (lacZ⁺) was assessed by directly streaking the nodule sap on X-gal plates. This method revealed differences between rhizobia carrying identical antibiotic markers. The rhizobial population in soil was estimated by direct plate counts using a medium containing X-gal. Introduction of lacZ into the Rhizobium sp. thus provided a simple and direct method for identifying strains from nodules and soil.     

利用微生物防除根寄生杂草列当

根寄生杂草列当(Orobanche spp.)已经严重制约全球许多地区的农业发展,寻找有效防除措施迫在眉睫。由于列当具有特殊生活史且与寄主关系密切,常规防除杂草措施难以达到理想防效。目前,尚无既能有效防除列当又不对寄主造成危害且便于大规模推广应用的列当防除措施。在众多列当防除措施中,微生物防除越来越引起关注和重视。本文对微生物防除列当的国内外研究进展及防除机理进行了综述。目前,列当生防微生物的研究主要集中在镰刀菌(Fusarium spp.)等列当病原菌和根瘤菌(Rhizobium spp.)等列当寄主植物共生菌上。微生物防除列当的机制主要包括两方面:一是通过产生代谢产物直接影响列当的萌发和生长,或通过降解列当种子萌发诱导物质间接影响列当的萌发;二是通过提高寄主植物自身对列当的抗性间接影响列当的寄生和生长。此外,本文还重点介绍了植物土传病害的土壤拮抗微生物防除列当杂草的可行性及研究进展。植物土传病害病原菌和列当均首先通过在地下侵染作物的根系进而危害作物正常生长,而作物抗土传病害的机理也与抗列当的机理类似。因此,存在于土壤中具有防治植物土传病害能力的微生物可能也具有防除根寄生杂草列当的功能。本团队前期试验从植物土传病害的土壤拮抗微生物中筛选到在盆栽试验中能够有效防除向日葵列当(O.cumana Wallr.)和瓜列当(O.aegyptiaca Pers.)的放线菌各1株,分别为淡紫褐链霉菌(Streptomyces enissocaesilis Sveshnikova)和密旋链霉菌(Streptomyces pactum Bhuyan B.K)。其中,密旋链霉菌的菌剂在田间试验中既降低了瓜列当的出土数量又增加了番茄的产量。总之,微生物是防除根寄生杂草列当的一条有效途径。     

Rhizobium sp. exopolysaccharide production,nodulation, and acetylene reduction activity (N2-fixation) in pigeon pea (Cajanus cajan L.)

Summary There was no correlation between the quantity of exopolysaccharide produced and acetylene reduction activity by Rhizobium spp. or by Bradyrhizobium spp. (Cajanus). The exopolysaccharide-defective mutants of Rhizobium sp. strain P 116 either failed to nodulate or showed a decrease in effectiveness. The deficiency in exopolysaccharide production was corrected by the addition of purified exopolysaccharide from the parent strain, or from Bradyrhizobium sp. strain P 149 or S24 isolated from pigeonpea (Cajanus cajan) and mungbean (Vigna radiata), respectively. However, the nodules so formed were not fully effective compared to those formed by the parent strain.     

Effect of nitrogen levels and Rhizobium strains on symbiotic N2 fixation and grain yield of Phaseolus vulgaris L. genotypes in normal and saline-sodic soils

Summary Following screening, selection, characterization, and symbiotic N₂ fixation with 12,5, 25.0, and 40.0 mg N kg^–1 in normal and saline-sodic soils, only two Phaseolus vulgaris genotypes (HUR 137 and VL 63) and two Rhizobium spp. strains (ND 1 and ND 2) produced maximum nodulation, nitrogenase activity, plant N contents, and grain yields in saline-sodic soil, with 12.5 mg N kg^–1, compared with the other strains. However, interactions between strains (USDA 2689, USDA 2674, and ND 5) and genotypes (PDR 14, HUR 15, and HUR 138) were significant and resulted in more nodulation, and greater plant N contents, nitrogenase activity, and grain yields in normal soils with 12.5 mg N kg^–1 compared with salt-tolerant strains. Higher levels of N inhibited nodulation and nitrogenase activity without affecting grain yields. To achieve high crop yields from saline-sodic and normal soils in the plains area, simultaneous selection of favourably interacting symbionts is necessary for N economy, so that bean yields can be increased by the application of an active symbiotic system.     

Competitive interaction among strains of Rhizobium leguminosarum bv. phaseoli in the nodulation of kidney beans (Phaseolus vulgaris L.)

We examined the competitiveness of five effective Rhizobium leguminosarum biovar phaseoli strains in the nodulation of kidney beans (Phaseolus vulgaris L.), either alone or in pairwise combination, against the indigenous strains. The results showed that the introduced Rhizobium sp. strains (B2, B17, B36, T2, or CIAT 652) occupying 64–79% of the total nodules (as single inocula) were more competitive in nodulation than the native rhizobia. However, the competitiveness of the individual Rhizobium sp. strain either increased or decreased when used in a pairwise combination of double-strain inocula. For example, strain B17, although quite competitive against the indigenous population (68% nodule occupancy), became poorly competitive in the presence of strain B2 (reduced from 68 to 2.5%). A similar reduction in nodule occupancy by strain B17 was observed in the presence of B36 or CIAT 652, indicating that two competitive strains may not always be compatible. These results suggest that it is important to co-select competitive as well as compatible rhizobia for multistrain inoculant formulation.     

Competition between inoculated and indigenous Rhizobium/Bradyrhizobium spp. strains for nodulation of grain and fodder legumes in Pakistan

Summary The competitive ability of inoculated and indigenous Rhizobium/Bradyrhizobium spp. to nodulate and fix N₂ in grain legumes (Glycine max, Vigna unguiculata, Phaseolus vulgaris) and fodder legumes (Vicia sativa, Medicago sativa, and Trifolium subterraneum) was studied in pots with two local soils collected from two different fields on the basis of cropping history. The native population was estimated by a most-probable-number plant infectivity test in growth pouches and culture tubes. The indigenous rhizobial/bradyrhizobial population ranged from 3 to 2×10⁴ and 0 to 4.4×10³ cells g^-1 in the two soils (the first with, the second without a history of legume cropping). Inoculated G. max, P. vulgaris, and T. subterraneum plants had significantly more nodules with a greater nodule mass than uninoculated plants, but N₂ fixation was increased only in G. max and P. vulgaris. A significant response to inoculation was observed in the grain legume P. vulgaris in the soil not previously used to grow legumes, even in the presence of higher indigenous population (>10³ cells g^-1 soil of Rhizobium leguminosarum bv phaseoli). No difference in yield was observed with the fodder legumes in response to inoculation, even with the indigenous Rhizobium sp. as low as <14 cells g^-1 soil and although the number and weight of nodules were significantly increased by the inoculation in T. subterraneum. Overall recovery of the inoculated strains was 38–100%, as determined by a fluorescent antibody technique. In general, the inoculation increased N₂ fixation only in 3 out of 12 legume species-soil combinations in the presence of an indigenous population of rhizobial/bradyrhizobial strains.     

Displacement of native rhizobia nodulating chickpea (Cicer arietinum L.) by an inoculant strain through soil solarization

Summary Soil solarization greatly reduced the native chickpea Rhizobium population. With inoculation, it was possible to increase the population of the Rhizobium in solarized plots. In the 1st year, 47% nodulation was obtained with chickpea inoculant strain IC 59 when introduced with a cereal crop 2 weeks after the soil solarization and having a native Rhizobium count of <10 g^-1 soil, and only 13% when introduced 16 weeks after solarization at the time the chickpeas were sown, with 2.0×10² native rhizobia g^-1 soil. In the non-solarized plots inoculated with 5.6×10³ native rhizobia g^-1 soil, only 6% nodulation was obtained with the inoculant. In the succeeding year, non-inoculated chickpea was grown on the same plots without any solarization or Rhizobium inoculation. The treatment that showed good establishment of the inoculant strain in year 1 formed 68% inoculant nodules. Other treatments indicated a further reduction in inoculant success, from 1%–13% to 1%–9%. Soil solarization thus allowed an inoculant strain to successfully displace the high native population in the field and can serve as a research tool to compare strains in the field, irrespective of competitive ability. In year 1, Rhizobium inoculation of chickpea gave increased nodulation and increased plant growth 20 and 51 days after sowing, and increased dry matter, grain yield, and grain protein yield at maturity. These beneficial effects of inoculation on plant growth and yield were not measured in the 2nd year.Submitted as Journal Article No. JA 945 by the International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), Patancheru, Andhra Pradesh 502 324, India     

<Emphasis Type="Italic">Rhizobium</Emphasis>,<Emphasis Type="Italic"> Bradyrhizobium</Emphasis> and<Emphasis Type="Italic"> Agrobacterium</Emphasis> strains isolated from cultivated legumes

The present study was conducted to isolate and characterize rhizobial strains from root nodules of cultivated legumes, i.e. chickpea, mungbean, pea and siratro. Preliminary characterization of these isolates was done on the basis of plant infectivity test, acetylene reduction assay, C-source utilization, phosphate solubilization, phytohormones and polysaccharide production. The plant infectivity test and acetylene reduction assay showed effective root nodule formation by all the isolates on their respective hosts, except for chickpea isolate Ca-18 that failed to infect its original host. All strains showed homology to a typical Rhizobium strain on the basis of growth pattern, C-source utilization and polysaccharide production. The strain Ca-18 was characterized by its phosphate solubilization and indole acetic acid (IAA) production. The genetic relationship of the six rhizobial strains was carried out by random amplified polymorphic DNA (RAPD) including a reference strain of Bradyrhizobium japonicum TAL-102. Analysis conducted with 60 primers discriminated between the strains of Rhizobium and Bradyrhizobium in two different clusters. One of the primers, OPB-5, yielded a unique RAPD pattern for the six strains and well discriminated the non-nodulating chickpea isolate Ca-18 from all the other nodulating rhizobial strains. Isolate Ca-18 showed the least homology of 15% and 18% with Rhizobium and Bradyrhizobium, respectively, and was probably not a (Brady)rhizobium strain. Partial 16S rRNA gene sequence analysis for MN-S, TAL-102 and Ca-18 strains showed 97% homology between MN-S and TAL-102 strains, supporting the view that they were strains of B. japonicum species. The non-infective isolate Ca-18 was 67% different from the other two strains and probably was an Agrobacterium strain.     

Genetic diversity of rhizobia associated with indigenous legumes in different regions of Flanders (Belgium)

We investigated the diversity of rhizobia isolated from different indigenous legumes in Flanders (Belgium). A total of 3810 bacterial strains were analysed originating from 43 plant species. Based on rep-PCR clustering, 16S rRNA gene and recA gene sequence analysis, these isolates belonged to Bradyrhizobium, Ensifer (Sinorhizobium), Mesorhizobium and Rhizobium. Of the genera encountered, Rhizobium was the most abundant (62%) and especially the species Rhizobiumleguminosarum, followed by Ensifer (19%), Bradyrhizobium (14%) and finally Mesorhizobium (5%). For two rep-clusters only low similarity values with other genera were found for both the 16S rRNA and recA genes, suggesting that these may represent a new genus with close relationship to Rhodopseudomonas and Bradyrhizobium. Primers for the symbiotic genes nodC and nifH were optimized and a phylogenetic sequence analysis revealed the presence of different symbiovars including genistearum, glycinearum, loti, meliloti, officinalis, trifolii and viciae. Moreover, three new nodC types were assigned to strains originating from Ononis, Robinia and Wisteria, respectively. Discriminant and MANOVA analysis confirmed the correlation of symbiosis genes with certain bacterial genera and less with the host plant. Multiple symbiovars can be present within the same host plant, suggesting the promiscuity of these plants. Moreover, the ecoregion did not contribute to the separation of the bacterial endosymbionts. Our results reveal a large diversity of rhizobia associated with indigenous legumes in Flanders. Most of the legumes harboured more than one rhizobial endosymbiont in their root nodules indicating the importance of including sufficient isolates per plant in diversity studies.     

Characterization of rhizobia from <Emphasis Type="Italic">Sesbania</Emphasis> species native to seasonally wetland areas in Uruguay

Naturally growing Sesbania species with tolerance to unfavourable habitats are widely distributed in non-cultivated seasonally wetland areas in Uruguay. We investigated the relative abundance, diversity and symbiotic efficiency of Sesbania punicea and S. virgata rhizobia in three ecologically different undisturbed and water-logged sites in Uruguay. Numbers of native-soil rhizobia infective on S. punicea or S. virgata were low, with higher numbers associated with the presence of S. virgata. Plants of S. virgata inoculated with soil suspension showed aerial and nodule biomass greater than that obtained with S. punicea. The rhizobia nodulating Sesbania species in water-logged lands in different regions of Uruguay were diverse differing in growth rates, acid production, growth at 39°C and in LB medium, host range and symbiotic efficiency. Seventeen representative strains clustered into four groups on the basis of phenotypic characteristics, ARDRA and DNA fingerprinting (GTG₅-PCR). Partial sequence of 16S rRNA from eight of these strains classified them into at least two genera with four species: Azorhizobium doebereinerae, Rhizobium sp. related to R. etli and two different Rhizobium sp.-Agrobacterium. Our results confirm the presence of the specie Azorhizobium doebereinerae as microsymbionts of S. virgata in South America. No strain of Rhizobium etli has previously been reported as a microsymbiont of Sesbania, though R. etli like organisms have also been recovered from Dalea purpurea and Desmanthus illinoensis. Significant increases in dry matter production were obtained with S. virgata plants inoculated with selected rhizobial strains under growth chamber conditions.     

Comparative efficieny of Rhizobium/Bradyrhizobium spp. strains in nodulating Cajanus cajan in relation to characteristic metabolic enzyme activities

The comparative symbiotic properties of Rhizobium spp. and Bradyrhizobium spp. strains infecting pigeon pea were evaluated. Bradyrhizobium strains (Cajanus) were found to be superior to Rhizobium strains (Cajanus) and the superiority was ascertained to be due to the higher enzyme activity of the tricarboxylic acid (TCA) cycle in comparison to Rhizobium spp. strains. Moreover, metabolic superiority or rapid growth rate does not necessarily correlate with symbiotic effectiveness. The symbiotic performance of isolates varied with the host cultivar. The dry matter accumulation could be correlated with the total acetylene reduction activities rather than nodule number or nodule fresh weight per plant. Received: 3 March 1993     

Effects of Rhizobium and vesicular-arbuscular mycorrhiza inoculations on nodulation,symbiotic nitrogen fixation and soybean yield in subtropical-tropical fields

Summary Field experiments were carried out to determine the effects of single and mixed inoculations with Rhizobium and vesicular-arbuscular mycorrhiza (VAM) on nodulation, symbiotic N₂ fixation and yield of soybeans in six Taiwan subtropical-tropical sites. Inoculation with Rhizobium alone significantly increased nodulation, nodule weight and nitrogenase activity of nodules in three out of six experimental fields, and affected soybean yields in the range –13% to + 134%. Inoculation with VAM fungi alone did not have a significant effect on nodulation and nitrogenase activity. Mycorrhiza inoculation affected soybean yields in the range –13% to + 65%, but only the yield increases at one out of six sites with N application were statistically significant. Mixed inoculation with Rhizobium and mycorrhiza affected yields in the range –8% to + 145% A synergistic effect from mixed inoculation of Rhizobium-mycorrhiza on soybean yields was found in one out of six experimental fields. The yield response to N application (40 kg N ha^–1) in these six paddy-field trials was not significant. These results suggest that single or mixed inoculation of rhizobia can greatly assist soybean grain yields and can replace N fertilizers.     

Evaluation of nitrogen fixation by different strains of theAzolla-Anabaena symbiosis in the presence of a high level of ammonium

Summary The proportion of N derived from N₂ fixation for 99 strains ofAzolla spp. (comprising all known species) in the presence of ammonium (40 mg/1) was assessed using a¹⁵N-dilution technique. The percentage of N derived from air varied from 29.5% to 79.9%. Although the N concentration ofAzolla spp. was not correlated with fertilizer N, it correlated fairly well with N₂ fixation. Regression analysis suggests that the N yield ofAzolla spp. is more dependent on N₂ fixation than on ammonium assimilation. The high correlation between N yield and isotopically determined, fixed N₂ indicates that the N yield could be used as a parameter in the selection ofAzolla spp. strains that are capable of maintaining high N₂ fixation in the presence of a high level of ammonium.     

Effect of successive cuttings on uptake and partitioning of 15N among plant parts of Leucaena leucocephala

Summary We studied the effect of three successive cuttings on N uptake and fixation and N distribution in Leucaena leucocephala. Two isolines, uninoculated or inoculated with three different Rhizobium strains, were grown for 36 weeks and cut every 12 weeks. The soil was labelled with 50 ppm KNO₃ enriched with 10 atom % ¹⁵N excess soon after the first cutting. Except for the atom % ¹⁵N excess in branches of K28 at the second cutting, both the L. leucocephala isolines showed similar patterns of total N, fixed N₂, and N from fertilizer distribution in different parts of the plant at each cutting. The Rhizobium strain did not influence the partitioning of ¹⁵N among the different plant parts. Significant differences in ¹⁵N enrichment occurred in different parts. Live nodules of both isolines showed the lowest atom % ¹⁵N excess values (0.087), followed by leaves (0.492), branches (0.552), stems (0.591), and roots (0.857). The roots contained about 60% of the total plant N and about 70% of the total N derived from fertilizer over the successive cuttings. The total N₂ fixed in the roots was about 60% of that fixed in the whole plant, while the shoots contained only 20% of the fixed N₂. We conclude that N reserves in roots and nodules constitute another N source that must be taken into account when estimating fixed N₂ or the N balance after pruning or cutting plants. ¹⁵N enrichment declined up to about fivefold in the reference and the N₂-fixing plants over 24 weeks following the ¹⁵N application. The proportion and the amounts of N derived from fertilizer decreased, while the amount derived from N₂ fixation increased with time although its proportion remained constant.