**4. Inoculation of efficient strains of bradyrhizobia to improve soybean seed yield**

### **4.1. Inoculation of bradyrhizobia and indigenous strains in fields**

Soybean plants form root nodules by symbiosis with soil bacteria *Bradyrhizobium* (*B. japoni‐ cum, B. elkani and B. lianingense*), *Sinorhizobium* (*S. fredii, S. xinjiangense*) and Mesorhizobi‐

um (*M. tianshanense*) [49]. Only *B. japonicum, B.* elkani, and *S. fredii* are used as commercial inoculants for soybean with *B. japonicum* being the most widely employed [49]. Genus *Bra‐ dyrhizobium* belongs to α-Proteobacteria, family Bradyrhizobiaceae. *Bradyrhizobium* species are rod shaped Gram negative bacteria with a single subpolar or polar flagellum. They can be isolated from nodules and grown on a Yeast Extract Mannitol agar plate (Figure 20) [50]. Rhizobia and bradyrhizobia can exist either as a free living organism in soil or a sym‐ biotic state in the infected cells of root nodules. When bradyrhizobia inhabit a soil, they do not fix N2 and they depend on organic nitrogen. After they infect soybean roots and form root nodules, they become "bacteroids" in a symbiotic state and fix N2.

**Figure 20.** A culture of *Bradyrhizobium japonicum* USDA110 (right) and the *gus* mutant line 61A124a (left) on a Yeast Extract Mannitol agar plate which contain GUS substrate X-Gluc.

When soybean plants are cultivated in a new field where soybeans are cultivated for the first time, the inoculation of compatible strains of bradyrhizobia may significantly promote plant growth and seed yield. However, after cultivation of soybean, rhizobia will predominate throughout the soil. Therefore, soybean plants are usually nodulated with indigenous rhizo‐ bia in most fields of Japan without inoculation. Tewari et al. reported that the inoculation of the efficient strain of bradyrhizobia USDA110 in a paper pot filled with vermiculite was very effective in the first year of a soybean crop after 30 cm layer of mountain soil was dressed, where population of bradyrhizobia was very low [51]. Figure 21 shows the plant with uninoculated paper pot (left) and that with inoculated paper pot (right) cultivated in the mountain soil. The average seed weight per plant was 8.7 g in uninoculated plants and 22.3 g in inoculated plants, respectively. In a rotated paddy field in Nagakura where indige‐ nous rhizobia had been established, Tewari et al. inoculated USDA110 in a paper pot, and the plant growth and seed yield was about 10-20% higher than the uninoculated paper pot (Figure 22). Good nodulation occurred in the roots of the plant in the uninoculated paper pot, because indigenous bradyrhizobia already inhabited the field.

um (*M. tianshanense*) [49]. Only *B. japonicum, B.* elkani, and *S. fredii* are used as commercial inoculants for soybean with *B. japonicum* being the most widely employed [49]. Genus *Bra‐ dyrhizobium* belongs to α-Proteobacteria, family Bradyrhizobiaceae. *Bradyrhizobium* species are rod shaped Gram negative bacteria with a single subpolar or polar flagellum. They can be isolated from nodules and grown on a Yeast Extract Mannitol agar plate (Figure 20) [50]. Rhizobia and bradyrhizobia can exist either as a free living organism in soil or a sym‐ biotic state in the infected cells of root nodules. When bradyrhizobia inhabit a soil, they do not fix N2 and they depend on organic nitrogen. After they infect soybean roots and form

A Comprehensive Survey of International Soybean Research - Genetics, Physiology, Agronomy and Nitrogen

**Figure 20.** A culture of *Bradyrhizobium japonicum* USDA110 (right) and the *gus* mutant line 61A124a (left) on a Yeast

When soybean plants are cultivated in a new field where soybeans are cultivated for the first time, the inoculation of compatible strains of bradyrhizobia may significantly promote plant growth and seed yield. However, after cultivation of soybean, rhizobia will predominate throughout the soil. Therefore, soybean plants are usually nodulated with indigenous rhizo‐ bia in most fields of Japan without inoculation. Tewari et al. reported that the inoculation of the efficient strain of bradyrhizobia USDA110 in a paper pot filled with vermiculite was very effective in the first year of a soybean crop after 30 cm layer of mountain soil was dressed, where population of bradyrhizobia was very low [51]. Figure 21 shows the plant with uninoculated paper pot (left) and that with inoculated paper pot (right) cultivated in the mountain soil. The average seed weight per plant was 8.7 g in uninoculated plants and 22.3 g in inoculated plants, respectively. In a rotated paddy field in Nagakura where indige‐ nous rhizobia had been established, Tewari et al. inoculated USDA110 in a paper pot, and the plant growth and seed yield was about 10-20% higher than the uninoculated paper pot (Figure 22). Good nodulation occurred in the roots of the plant in the uninoculated paper

root nodules, they become "bacteroids" in a symbiotic state and fix N2.

Extract Mannitol agar plate which contain GUS substrate X-Gluc.

Relationships

132

pot, because indigenous bradyrhizobia already inhabited the field.

**Figure 21.** The effect of inoculation of *Bradyrhizobium japonicum* USDA110 on soybean growth cultivated in montain soil where the population of indigenous bradyrhizobia was very low. Left plant was cultivated with a paper pot with‐ out inoculation. Right plant was cultivated with a paper pot with inoculation of bradyrhizobia [51].

**Figure 22.** The effect of inoculation method for seed yield of soybean plants in a rotated paddy field in Nagakura. NIPP; Non-inocultated paper pot.DT; Direct inoculation of bradyrhizobia to soybean seed. IPP; Inoculated paper pot [52].

Because the soybean plant can form nodules in most of the fields including a rotated paddy rice field in Japan, farmers generally do not inoculate bradyrhizobia, except in the Tokachi

area in Hokkaido. The Tokachi Federation of Agricultural Cooperatives in Hokkaido pro‐ vides biofertilizer inoculants for soybean seeds [50]. Although many types of bradyrhizobia exist in a field [53,54], not all are efficient strains. Minamisawa et al. reported that 44 isolates of *Bradyrhizobium japonicum* from Nakazawa field in Niigata were divided into 33 genetically different groups by genetic analysis using a repeated sequence specific hybridization meth‐ od. The similar diversity of indigenous bradyrhizobia was also shown in 6 sites in Japan, in‐ cluding the Nakazawa and Nagakura fields in Niigata [54]. The Nakazawa and Nagakura fields are very near to one another, but bradyrhizobia types differ between the fields. In Na‐ gakura, *B. japonicum* hup+ (uptake hydrogenase positive) and hup- (uptake hydrogenase negative) groups made up about 80% and 20% of bradyrhizobium, respectively. On the oth‐ er hand, in the Nakazawa field, *B. japonicum* hup+ , hup and *B. elkanii* made up about 50%, 20% and 30%, respectively, of the local bradyrhizobium. Significant diversities and site-de‐ pendent variations were observed, and the fingerprints at Ishigaki island with no history of soybean cultivation were less diverse than the other sites. They suggested that soybean bra‐ dyrhizobia might be diversified in individual fields by association with host plants and local soil conditions [54].

### **4.2. Use of marker strain as an inoculant**

The ecological study of inoculated strains in soil is important to establish the efficient way of inoculation. Minagawa et al. used *gus* (β-glucuronidase gene)-marked *Bradyrhizobium* strain to estimate the number of inoculated strains in the soil. The *gus* gene from *Eschrichia coli* was introduced into *Bradyrhizobium japonicum*. This strain absorbs and hydrolyzes GUS(β-glucur‐ onidase)-substrate (X-Gluc; 5-bromo-4-chloro-3-indolyl-β-glucuronide) and precipitates an indigo pigment in the cell (Figure 20). The accumulated indigo blue metabolite in *gus*marked strain can be determined for the population of liquid cultured rhizobia using optical density at 645 nm. However, when the *gus*-marked strain was inoculated in soil, the stained bradyrhizobia were difficult to separate from the soil. Therefore, we extracted the blue pig‐ ment by phenol-water after they are incubated with X-Gluc for 4 days. The absorbance of the extracted blue pigment in the phenol-layer was measured optically at 645 nm. The initial number of bradyrhizobia in culture media or soil and the absorbance of the phenol extract of GUS metabolite was positively correlated. In addition, the occupancy by *gus*-strain in each nodule can be determined by staining the nodule slice with substrate X-Gluc (Figure 23). The nodule occupancy by *gus*-marked strain was 50% when the same population of *gus*marked strain and USDA110 were inoculated at the same time [55].

The *gus*-marked strains were inoculated in five different types of soil (Nagakura and Sonoki; two types of Alluvial soil from rotated paddy fields), volcanic ash soil of upland field (Na‐ kazawa), sandy dune soil (Ikarashi), and calcinic vermiculite. Soybean cultivation was fre‐ quent in Nagakura and Nakazawa fields, but soybean plants have not been grown for a long time in Sonoki and Ikarashi fields. The number of indigenous bradyrhizobia was estimated by MPN method, and the populations were as follows; Nagakura (6 x 105 cells g-1 soil), Na‐ kazawa (3 x 105 ), Sonoki (2 x 104 ), Ikarashi (8), and Vermiculite (0), respectively. The popula‐ tion of *gus*-marked strains increased over 10 times in all types of soils for 1 week after inoculation of 9 x106 cells g-1 soil (Figure 24). The population was higher in Nakazawa and Nagakura than that in Sonoki, Ikarashi, and vermiculite at 1 week after inoculation. From this result it was indicated that the population of indigenous bradyrhizobia may not restrict the growth of inoculated strains.

area in Hokkaido. The Tokachi Federation of Agricultural Cooperatives in Hokkaido pro‐ vides biofertilizer inoculants for soybean seeds [50]. Although many types of bradyrhizobia exist in a field [53,54], not all are efficient strains. Minamisawa et al. reported that 44 isolates of *Bradyrhizobium japonicum* from Nakazawa field in Niigata were divided into 33 genetically different groups by genetic analysis using a repeated sequence specific hybridization meth‐ od. The similar diversity of indigenous bradyrhizobia was also shown in 6 sites in Japan, in‐ cluding the Nakazawa and Nagakura fields in Niigata [54]. The Nakazawa and Nagakura fields are very near to one another, but bradyrhizobia types differ between the fields. In Na‐

A Comprehensive Survey of International Soybean Research - Genetics, Physiology, Agronomy and Nitrogen

(uptake hydrogenase positive) and hup-

, hup-

negative) groups made up about 80% and 20% of bradyrhizobium, respectively. On the oth‐

20% and 30%, respectively, of the local bradyrhizobium. Significant diversities and site-de‐ pendent variations were observed, and the fingerprints at Ishigaki island with no history of soybean cultivation were less diverse than the other sites. They suggested that soybean bra‐ dyrhizobia might be diversified in individual fields by association with host plants and local

The ecological study of inoculated strains in soil is important to establish the efficient way of inoculation. Minagawa et al. used *gus* (β-glucuronidase gene)-marked *Bradyrhizobium* strain to estimate the number of inoculated strains in the soil. The *gus* gene from *Eschrichia coli* was introduced into *Bradyrhizobium japonicum*. This strain absorbs and hydrolyzes GUS(β-glucur‐ onidase)-substrate (X-Gluc; 5-bromo-4-chloro-3-indolyl-β-glucuronide) and precipitates an indigo pigment in the cell (Figure 20). The accumulated indigo blue metabolite in *gus*marked strain can be determined for the population of liquid cultured rhizobia using optical density at 645 nm. However, when the *gus*-marked strain was inoculated in soil, the stained bradyrhizobia were difficult to separate from the soil. Therefore, we extracted the blue pig‐ ment by phenol-water after they are incubated with X-Gluc for 4 days. The absorbance of the extracted blue pigment in the phenol-layer was measured optically at 645 nm. The initial number of bradyrhizobia in culture media or soil and the absorbance of the phenol extract of GUS metabolite was positively correlated. In addition, the occupancy by *gus*-strain in each nodule can be determined by staining the nodule slice with substrate X-Gluc (Figure 23). The nodule occupancy by *gus*-marked strain was 50% when the same population of *gus*-

The *gus*-marked strains were inoculated in five different types of soil (Nagakura and Sonoki; two types of Alluvial soil from rotated paddy fields), volcanic ash soil of upland field (Na‐ kazawa), sandy dune soil (Ikarashi), and calcinic vermiculite. Soybean cultivation was fre‐ quent in Nagakura and Nakazawa fields, but soybean plants have not been grown for a long time in Sonoki and Ikarashi fields. The number of indigenous bradyrhizobia was estimated

tion of *gus*-marked strains increased over 10 times in all types of soils for 1 week after

), Ikarashi (8), and Vermiculite (0), respectively. The popula‐

marked strain and USDA110 were inoculated at the same time [55].

by MPN method, and the populations were as follows; Nagakura (6 x 105

), Sonoki (2 x 104

(uptake hydrogenase

cells g-1 soil), Na‐

and *B. elkanii* made up about 50%,

gakura, *B. japonicum* hup+

Relationships

134

soil conditions [54].

kazawa (3 x 105

er hand, in the Nakazawa field, *B. japonicum* hup+

**4.2. Use of marker strain as an inoculant**

**Figure 23.** Detection of *gus*-marked strain in the nodule by X-Gluc treatment.

The population of inoculated strains was determined in rhizosphere soil, non-rhizosphere soil and roots (Figure 25) [51]. The population increased about 10 times in a week both in rhizosphere and non-rhizosphere soils at one week after inoculation.

**Figure 24.** Changes in population of inoculated *gus*-marked strain in various types of soils in Niigata.

The proliferation and mobility of inoculated *gus-*marked strains were examined in a rhizo‐ box containing various soil types (Figure 26), where a soybean plant was cultivated in the center of the box. Inoculation and watering were supplied from one side of the box. The pro‐ liferation rates of the *gus*-marked strain in a whole box at 25 days after planting were differ‐ ent in various types of soils; Nagakura (x 1,218), Nakazawa (x 538), Sonoki (x 513), Ikarashi, (x173) and vermiculite (x 98). In all the soil types, bradyrhizobia were distributed in many compartments of the rhizobox (Figure 27). It was observed that most of bradyrhizobia at‐ tached to the soil particle, however, some of them moved by water flow through soil aper‐ tures or along with the root elongation.

**Figure 25.** Changes in population of inoculated *gus*-marked strains in non-rhizosphere soil, rhizosphere soil and roots.

**Figure 26.** Rhizobox experiment system.

A soybean plant was grown in the center of the box, and inoculation of the *gus*-marked strain and the watering was carried out from one side of the box (arrow).

For soils lacking indigenous bradyrhizobia (vermiculite) or having a very low density of in‐ digenous *Bradyrhizobia* (Sonoki and Ikarashi), the nodules were formed almost exclusively by the inoculated *gus*-marked strain Sonoki (100%; occupancy rate by *gus*-strain), Ikarashi (98%) and vermiculite (100%). However, the major nodules were formed by indigenous strains in the soil types, which contain a high population of indigenous *Bradhrhizobia*, in Na‐ gakura (25% occupancy rate by *gus*-marked strain), Nakazawa (35%*gus*-marked strain), al‐ though the inoculated strain proliferated very well. The results suggested that nodule occupancy may be simply related to the population rate of inoculated strains vs. indigenous strains rather than the competition between inoculated and indigenous strains.

**Figure 27.** Population of *gus*-marked strains in 9 compartments of each rhizobox (color of each compartment), and the percentage of nodule occupancy by *gus*-marked strains (number in the center of each compartment) [55]. Arrow indicates the site of inoculation and watering.

#### **4.3. Survival of inoculated** *gus***-marked strain**

The proliferation and mobility of inoculated *gus-*marked strains were examined in a rhizo‐ box containing various soil types (Figure 26), where a soybean plant was cultivated in the center of the box. Inoculation and watering were supplied from one side of the box. The pro‐ liferation rates of the *gus*-marked strain in a whole box at 25 days after planting were differ‐ ent in various types of soils; Nagakura (x 1,218), Nakazawa (x 538), Sonoki (x 513), Ikarashi, (x173) and vermiculite (x 98). In all the soil types, bradyrhizobia were distributed in many compartments of the rhizobox (Figure 27). It was observed that most of bradyrhizobia at‐ tached to the soil particle, however, some of them moved by water flow through soil aper‐

A Comprehensive Survey of International Soybean Research - Genetics, Physiology, Agronomy and Nitrogen

**Figure 25.** Changes in population of inoculated *gus*-marked strains in non-rhizosphere soil, rhizosphere soil and roots.

tures or along with the root elongation.

Relationships

136

**Figure 26.** Rhizobox experiment system.

The survival of the inoculated strain is very important to improve indigenous strains in a field. We investigated the survival of *gus*-strains in the first soybean cultivation just after in‐ oculation of the *gus-*strain, and in the second year in the pot filled with five soil types. The nodule occupancy rate (%) by the *gus-*marked strain in the first and second year are shown in Figure 28. In the Ikarashi and vermiculite soils with low densities of indigenous bradyrhi‐ zobia, the nodule occupancy rates by the *gus*-marked strains were high (about 80-100%) both in the first and second year. However, in Nagakura, Nakazawa and Sonoki soils, the nodule occupancy rates were lower in the second year than the first year. This result sug‐

gests that the survival ability of the inoculated single strain may be inferior to the indige‐ nous strains. The genetic diversity of the indigenous strain may be related to the competition between inoculated and indigenous strains.

**Figure 28.** Percentage of nodules occupied by *gus*-marked strain (GUS+) and indigenous strain (GUS-) in various soils in Niigata in the first and second cropping years after inoculation.
