Thirty-one cultures of Rhizobium leguminosarum

Copyright C 1983, American Society for Microbiology

SHEILA N. MAYt AND B. BEN BOHLOOL*

Received 9 August 1982/Accepted 23 November 1982

The mechanisms that confer competitive ad­vantage to a strain are not fully understood. Some of the factors that have been suggested are: faster growth in the rhizosphere of the host (27); preferential selection by the host (32, 40); tolerance of climate, pH, and other soil factors (28, 31); antagonism between competing strains (37); and the relative ratio of inoculant to resi­dent strains (4).

The present study analyzes the competitive ability of effective strains of Rhizobium legumi­
† Present address: Biology Department, University of Colo­rado, Colorado Springs, Colorado 80933.

nosarum for nodulation of the roots of lentils (Lens esculenta). Experiments were designed to assess competitiveness in the absence of biologi­cal and abiological soil factors in the growth chamber and under field conditions in two soils with different properties. The fluorescent anti­body (FA) technique was used to identify the strains in the nodules.

MATERIALS AND METHODS
The strains of R. leguminosarum used in the compe­tition experiments, Hawaii 5-0, Nitragin 92A3 (previ­ously NZP5400), and Nitragin 128A12, were selected on the basis of growth chamber screening, which showed all three to be highly effective on lentils. The origin of these strains has been described previously (1). The sources of other cultures used in this study are listed in Table 1. Cultures were grown and maintained in modified yeast extract-mannitol medium (3) and routinely transferred every 6 months.

The above cultures were screened for effectiveness on lentils. Seedlings, grown in sterilized vermiculite in test tubes (25 by 200 mm) and moistened with nitro­gen-free nutrient solution (11), were inoculated in duplicate with 3-day-old cultures of each of the strains (see Table 1). Eight units were left uninoculated.

For the field experiments, the inoculum strains were grown separately in gamma-irradiated peat (31). Mix­tures of different peat cultures were made to contain equal numbers of the desired strains. The peat prepa­rations were coated onto commercial lentil seeds, using the procedure outlined by Vincent (39). Coated seeds were pelleted with calcium carbonate (8). Each seed harbored approximately 5 x 10⁴ cells of the desired rhizobial mixture, as determined by viable count on yeast extract-mannitol agar and the IF quan­titative membrane-filter technique (5, 35).

Pelleted seeds were planted at two sites: Mauka Field Station, the University of Hawaii, with an Andic Ustic Humitropept soil (Inceptisol), pH 6.1, contain­ing a low number of indigenous ineffective rhizobia (less than 100/g, as determined by the most-probable ­number method [14]); and Poamoho Experiment Sta­tion, Hawaii Institute of Tropical Agriculture and Human Resources, with a tropeptic Eustrostox soil (Oxisol), pH 5.8, characterized by low phosphorus in soil solution (0.003 ppm P) and high phosphorus-­sorption properties (16) and lack of R. leguminosarum (no nodules developed when the soil was used as an inoculum [1 g per seedling] on lentils).

In all field experiments, randomized complete-block designs with three replicates were used. Plants were harvested after 10 days, 5 weeks, and 8 weeks; there were 6 to 15 plants per replicate per treatment. Tops were removed, and the roots were tested for nitrogen­ase activity by the acetylene reduction assay (18). The roots were then washed in 0.1% Tween 80, and at least 50 nodules per replicate per treatment were picked at random for FA identification. To determine the persis­tence of the inoculum strains in the soil, uninoculated lentil seeds were planted at the same plots in the Mauka Field Station 1 year after the original competi­tion experiment. After 10 weeks, ten plants per repli­cate per treatment were harvested, and their nodules were identified as described above.

The effectiveness (acetylene reduction) pat­tern of 31 strains and isolates of R. legumino­

Preparation of FAs and immunofluorescent (IF) staining of cultures and nodules are described else­where (36). Smears from cultures and nodules were treated with gelatin-rhodamine isothiocyanate conju­gate (2) to control nonspecific staining and auto­fluorescence. Nodules from all experiments were stored dry at 55°C until use.

Stained nodule and culture smears were examined with a Zeiss universal microscope equipped for epi­fluorescence and transmitted dark field. Incident illu­mination was from an HBO-200 (OSRAM) light source with a fluoroscein isothiocyanate filterpack. Transmit­ted dark-field illumination was from a 12-V quartz­halogen lamp, using a Zeiss Ultracondenser.

For growth chamber experiments, commercial lentil seeds (Spokane Seed Co.) were surface sterilized and

(P ≤ 0.05) on the other two varieties. Table 3 also shows that incidence of double-strain occu­pancy (the same nodules containing two strains) varied from 0 to 36% of the total number of nodules, depending on the strain combinations and the host varieties used.

The inoculum strains completely dominated the indigenous rhizobia in the experiments at the Mauka Field Station (Table 4). The results show that strains Hawaii 5-0 and Nitragin 92A3 were of equal competitiveness (P = 0.01) on lentils in this field, with 38% of the nodules containing both strains. Single and mixed infections in nodules of lentils grown in the field are shown in Fig. 1. Smears from two nodules from the same plant were stained with FA against Hawaii 5-0. In a single-strain nodule, (Fig. 1A), all of the bacteroid cells stain, whereas in a double-strain nodule (Fig. 1B), only a proportion of the cells stain. The unstained cells (arrows) can be visual­ized by using a double-light system with a dark­field condenser. They were shown to be Nitragin 128A12 by staining a similar smear with anti­-Nitragin 128A12 FA. The results in Table 4 (numbers in parentheses) also show that the inoculum strains persisted well in this soil, for they still formed the majority of nodules on uninoculated plants 1 year after they were first introduced into the soil. However, Hawaii 5-0 was dominant in the single-strain nodules (P = 0.05).

Hawaii 5-0 occupied the majority of nodules at all treatments at the Poamoho field sites, except at high P levels after 8 weeks, when its nodule occupancy was not significantly different from that of the other strains (Table 5). The results in Table 5 also show that the incidence of double­strain occupancy varied depending on the strains used. In the majority of the cases, there were higher proportions of Hawaii 5-0-Nitragin 92A3 and Hawaii 5-0-Nitragin 128A12 in dou­ble-strain nodules than other combinations.

The uninoculated plants remained free of nod­ules in samplings 1 and 2 at Poamoho. However, by the harvest at 8 weeks, approximately 50% of the uninoculated control plants harbored nod­ules. Immunofluorescence typing of these nod­ules revealed that all of these nodules were formed by one or more of the three inoculum strains used in this experiment.
DISCUSSION
The three strains of R. leguminosarum used in these studies were among the most efficient (C₂H₂ reduction) of the 31 cultures screened (Table 1). They were also shown to be distinct by immunofluorescence staining (Table 2). This distinction was previously demonstrated by the enzyme-linked immunosorbent assay for the same organisms (1). It was thus possible to use



962 MAY AND BOHLOOL
sarum on lentils (commercial variety) is shown in Table 1. Eleven of the cultures proved ineffec­tive, five highly effective, and the rest intermedi­ate. Three of the cultures from the highly effec­tive group, Hawaii 5-0, Nitragin 92A3, and Nitragin 128A12, were chosen for further field testing and competition studies.

The FAs prepared against the somatic anti­gens of Hawaii 5-0, Nitragin 92A3, and Nitragin 128A12 exhibited a high degree of strain speci­ficity, as shown in Table 2. The FAs reacted strongly with the homologous bacteria in culture and with the bacteroids in nodules. They did not cross-react with one another and cross-reacted only weakly with 5 of the 41 other R. legumino­sarum cultures tested.

The competition pattern of strains Hawaii 5-0, Nitragin 92A3, and Nitragin 128A12 on four varieties of lentils grown in sterilized vermiculite is given in Table 3. Nitragin 92A3, whenever included in the inoculum, occupied the majority of nodules. This was significant (P < 0.05) in two-strain competitions against Nitragin 128A12 on all varieties and against Hawaii 5-0 on the Chilean variety. Hawaii 5-0 was of equal com­petitiveness (P = 0.01) with Nitragin 128A12 on the Chilean variety, more competitive (P = 0.01) on the commercial variety, and less competitive




these strains in competition against one another and easily identify the one(s) present in the nodules.

Under bacteriologically defined conditions in the growth chamber, one of the strains (Nitragin 92A3) dominated the nodules in two-strain com­petition trials (Table 3). The competitive advan­tage of Nitragin 92A3 was consistent, but signifi­cant (P = 0.01) only against Nitragin 128A12 on all four varieties and against Hawaii 5-0 on the Chilean variety. Vincent and Waters (40) and Russell and Jones (32) have reported that some clover species preferentially select one strain over another. Our data indicate that the overall pattern of competition of the strains used in this study varies depending on the host variety.

In the field that contained low numbers of resident R. leguminosarum (Mauka Station), the inoculum strains outcompeted the resident rhi­zobia completely. This might have been due, in the first year, to the high numbers of the inocu­lum strains applied to the seeds. However, even after 1 year in the soil, when these strains no longer had the advantage of direct seed place­ments, they still formed the majority of nodules on uninoculated plants (Table 4).

The local isolate, Hawaii 5-0, performed bet­ter than the other inoculum strains in an Oxisol field (Table 5). Thornton (38) has also reported that local strains survive in the same soil better than commercial strains in two of three in­stances.

Many investigators (12, 21, 26, 28, 30, 40) have suggested the use of inocula containing several strains adapted to a wider range of soil and plant differences. Some workers (7, 12, 15) have even shown that plants inoculated with mixtures of effective strains give better yield and higher nitrogen content than those with a single­strain inoculum. Considering that our data (Ta­

Lindemann et al. (24) used immunofluores­cence to provide evidence for double infection in nodules of soybeans in sterilized sand and re­ported a frequency of such occurrence as high as 32%. Other investigators have reported inci­dences of double-strain occupancy exceeding 20% in nodules on plants grown under bacterio­logically controlled conditions (10, 17, 20, 22, 25), but only <1.0 to 3.5% in the field (9, 10).

In our studies, the incidence of double-strain occupancy in nodules varied from 0 to 36% in vermiculite (Table 3), depending on the strains in the mixture and the host variety, and from 0 to 38% in the field (Tables 4 and 5), depending on the strain mixture and the soil type. The data also suggest a close relationship between the competitiveness of a strain and its involvement in the formation of doubly infected nodules.

VOL. 45, 1983