**7. A new fertilization technique to promote nitrogen fixation and seed yield**

### **7.1. Nitrogen fertilization in soybean cultivation**

Nitrogen fertilization usually depresses nodulation and nitrogen fixation activity. It often re‐ sults in the same or less yield than the control treatment without fertilizer. Top dressing of N fertilizer sometimes gives positive effects on seed yield, but results are not consistent. Ta‐ kahashi et al. reported the effect of basal side-dressings of various types of controlled release nitrogen fertilizer (coated urea) on shoot growth and seed yield of soybean [73]. The yield was significantly higher in all the coated urea basal side-dressing treatments compared with control, particularly in CUS120 which releases urea from 60 to 120 days after planting with a sigmoid pattern of N release.

We have developed a new fertilization technique for soybean cultivation to supplement N during seed filling stage, without the concomitant depression of N2 fixation, by deep place‐ ment (20 cm depth from soil surface) of slow release N fertilizer coated urea and lime nitro‐ gen. We analyzed the beneficial effects from both plant nutrition and soil analysis aspects.

### **7.2. Deep placement of controlled release nitrogen fertilizer coated urea**

Takahashi et al. [4,16, 74-77] developed a new fertilization technique for soybean to supple‐ ment nitrogen during the seed filling stage without concomitant depression of symbiotic N2 fixation by deep placement of coated urea, a slow release N fertilizer. They applied 100 kg N ha-1 coated urea by deep placement (20 cm depth from soil surface) using a fertilizer injector devised by Shioya [78]. They used CU-100, a 100-day type coated urea, the commercial name "LP-100" produced by Chisso-Asahi Fertilizer Co. Ltd, Tokyo (JCAM AGRI Co. Ltd., Tokyo at present). CU-100 linearly releases urea and 80 % of which is released in 100 days in water at 25o C.

A polymer coated controlled release nitrogen fertilizer (commercial name LP in Japan or MEIS‐ TER outside Japan) has been invented by Fujita and coworkers [79]. Linear types of coated urea were first marketed in 1982. This type of fertilizer has a spherical shape of about 3mm diameter with 50-60 µm coat thickness which consists of polyolefin (polyethylene), ethylene vinyl ace‐ tate and talc mineral. Different from chemically synthesized slow release N fertilizers such as IBDU (Isobutylidene diurea) and CDU (Crotonylidene diurea), the N release rate from the coat‐ ed urea is temperature dependent and not affected by other chemical, physical and biological conditions. Therefore, the release pattern of urea can be predicted as a function of temperature and time period after application. Since the release of N from the fertilizer meets the plant N de‐ mand, and the fertilizer efficiency (recovery rate of N in plants from fertilizer N) is high, the use of coated urea can reduce environmental problems by decreasing nitrate accumulation and leaching in the soil. Also the use of coated urea saves the labor of farmers by eliminating top dressing or split dressing of fertilizer to supply N during late growth stages.

and Harper evaluated the N2 fixation potential and yield of hypernodulating soybean NOD1-3, NOD2-4 and NOD3-7 compared with the parent Williams. In the absence of N fer‐ tilizer, all hypernodulation mutants had greater N2 fixation potential than did Williams in early growth stages. However, the seed yields from the hypernodulation mutants were 10 to 30% less than that for Williams. Suganuma et al. also compared the growth and N2 fixation activity of NOD1-3 and Williams in a sandy dune field. Figure 35 shows the daily rate of N2 fixation and N absorption by Williams and the hypernodulation mutant NOD1-3. The % Ndfa was higher in NOD1-3 (65 %) than Williams (58 %), however, the rates of N2 fixation and N absorption were lower in NOD1-3 than in Williams. The hypernodulation mutant lines have not been used for cultivar improvement, but recently "Sakukei no. 4" bred from En6500 (hypernodulating line from "Enrei") and "Tamahomare" in Japan may be useful in

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

**7. A new fertilization technique to promote nitrogen fixation and seed**

Nitrogen fertilization usually depresses nodulation and nitrogen fixation activity. It often re‐ sults in the same or less yield than the control treatment without fertilizer. Top dressing of N fertilizer sometimes gives positive effects on seed yield, but results are not consistent. Ta‐ kahashi et al. reported the effect of basal side-dressings of various types of controlled release nitrogen fertilizer (coated urea) on shoot growth and seed yield of soybean [73]. The yield was significantly higher in all the coated urea basal side-dressing treatments compared with control, particularly in CUS120 which releases urea from 60 to 120 days after planting with a

We have developed a new fertilization technique for soybean cultivation to supplement N during seed filling stage, without the concomitant depression of N2 fixation, by deep place‐ ment (20 cm depth from soil surface) of slow release N fertilizer coated urea and lime nitro‐ gen. We analyzed the beneficial effects from both plant nutrition and soil analysis aspects.

Takahashi et al. [4,16, 74-77] developed a new fertilization technique for soybean to supple‐ ment nitrogen during the seed filling stage without concomitant depression of symbiotic N2 fixation by deep placement of coated urea, a slow release N fertilizer. They applied 100 kg N ha-1 coated urea by deep placement (20 cm depth from soil surface) using a fertilizer injector devised by Shioya [78]. They used CU-100, a 100-day type coated urea, the commercial name "LP-100" produced by Chisso-Asahi Fertilizer Co. Ltd, Tokyo (JCAM AGRI Co. Ltd., Tokyo at present). CU-100 linearly releases urea and 80 % of which is released in 100 days in

**7.2. Deep placement of controlled release nitrogen fertilizer coated urea**

agricultural production by increasing the planting density.

**7.1. Nitrogen fertilization in soybean cultivation**

sigmoid pattern of N release.

water at 25o

C.

**yield**

Relationships

144

Fertilizer experiments were carried out from 1989 to 1991 in the fields, which had been convert‐ ed from a paddy rice field the previous year in Nagakura. [21]. As shown in Figure 36 the seed yield was significantly higher in the plants with the deep placement of CU-100 than the control in each year. The seed yield was from 10 (1991) to 23 % (1990) higher in deep placement than the control treatments. The promotion of leaf growth and retardation of leaf senescence were ob‐ served at the maturing stage by deep placement of CU. In 1990, the seed yield was very high, about 6 t ha-1 in deep placement, due to favorable climatic conditions. The absorption efficiency of fertilizer N determined by 15N labeled fertilizers was calculated from recovery of 15N in the shoots at the R7 stage. In 1990, the absorption efficiency at R7 from the deep placement of CU-100 was 62 %, which was much higher than the top dressing of CU-70 (33 %) and basal ap‐ plication of ammonium sulfate (9 %). It was observed that the CU-100 deep placement in‐ creased root growth and water and nutrient absorption activity revealed by the uptake of rubidium tracer in the field [75]. Owing to the promotion of subsoil root growth and N absorp‐ tion activity with supplementing N fertilizer without depression of N2 fixation, plant growth was promoted from the early vegetative stage until late maturing. Leaf area index (LAI) and chlorophyll content were always higher in CU-100 deep placement compared with the control, and the leaf senescence was retarded at R7 stage [21].

Takahashi et al. analyzed the concentration of urea, ammonium and nitrate in the upper 0-10cm and lower 15-25cm layers of the soil in control and deep placement of CU-100 treatments. In the upper layer, the concentration of urea and nitrate was very low (less than 10 mg N kg–1 soil) both in control and deep placement of CU-100 treatments. However, the accumulation of ammoni‐ um (up to 150 mg N kg-1) and nitrate (up to 50 mg N kg-1) was observed in the lower layer of deep placement of CU100 in August. Although the urea released from the coated urea was rapidly hydrolyzed to ammonia, NH4 + -N could not be easily nitrified in the deep soil layers of the con‐ verted rice field owing to the low activity of nitrification and restricted O2 supply. As a result, the nodulation and N2 fixation near the surface layer were not depressed, and instead were pro‐ moted through the improvement of plant growth and photosynthetic activity. The mechanism of promotion of deep placement of coated urea for soybean growth and seed yield is summar‐ ized in Figure 37.

**Figure 36.** Comparison of the seed yield of control and deep placement of coated urea in rotated paddy fields in Na‐ gakura in 1989, 1990 and 1991. [4]

**Figure 37.** A model of the promotive effect of deep placement of coated urea on nitrogen fixation and seed yield of soybean.


**Figure 38.** Soybean plants at R7 cultivated in a newly reclaimed field with a 30 cm depth of mountain soil. Plants were planted with inoculated or non-inoculated paper pot. –N: without N deep placement, Urea: deep placement of urea at 20cm depth. Lime N: deep placement of lime nitrogen. [51]

#### **7.3. Deep placement of lime nitrogen**

**Figure 36.** Comparison of the seed yield of control and deep placement of coated urea in rotated paddy fields in Na‐

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

**Figure 37.** A model of the promotive effect of deep placement of coated urea on nitrogen fixation and seed yield of

gakura in 1989, 1990 and 1991. [4]

Relationships

146

soybean.

Recently, Tewari et al. [51,52,80,81] investigated the effects of deep placement of lime nitro‐ gen (calcium cyanamide, CaCN2) in comparison with coated urea. The fertilizer experiments were combined with a new inoculation method of bradyrhizobia using a paper pot trans‐ plantation. All the experiments were carried out in 2001 at three different sites in Niigata

Prefecture. The tests were initiated in a field reclaimed with the application of mountain soil about 30 cm depth without indigenous bradyrhizobia (Figure 38) [51], a rotated paddy field in Nagakura in Niigata Agricultural Research Institute (Figure 39) [52], and a sandy dune field of Niigata University in Ikarashi [80].

**Figure 39.** Soybean plants cultivated in a rotated paddy field in Nagakura. Plants were planted with inoculated paper pots. Control: without N deep placement, Urea: deep placement of urea at 20cm. Coated Urea: deep placement of coated urea. Lime Nitrogen: deep placement of lime nitrogen. [52]

Lime nitrogen is composed of about 60% calcium cyanamide (CaCN2) with calcium oxide and carbon, and the N content is about 20-23%. After application to the soil, the calcium cy‐ anamide is converted to urea, which is again degraded into NH3 and CO2. Dicyandiamide contained in lime nitrogen or formed during the degradation of calcium cyanamide is a po‐ tent nitrification inhibitor, which retards the oxidation of NH4 + to NO3 - . Therefore the ammo‐ nium produced by CaCN2 decomposition persists for a longer period of time and the nitrate concentration remains low in the soil. It is expected that the inhibition of nodulation and of the N2 fixation activity may be alleviated by low a level of nitrate accumulation. Also this fertilizer exerts some hormonal effects on plants and is used for controlling soil diseases caused by bacteria and fungi.

In each of these fertilizer treatments, IPP (Inoculated paper pot), DT (Direct transplantation of inoculated seedlings without paper pot) and NIPP (Non-inoculated paper pot) seedlings were transplanted in separate plots. Paper pots (height 13.5 cm, diameter 3 cm) were made of a biodegradable paper designed to break down in the field. The pots were open at the bottom to allow root expansion below the pot. A paper pot was filled with vermiculite and a seed was planted in each pot, and followed by inoculation of one ml suspension of *Bradyrhi‐ zobium japonicum* USDA110 of about 108 cells ml-1. Since the bradyrhizobium population in‐ creases about 100 times in vermiculite for a few weeks [55], efficient infection of inoculated bradyrhizobia can also be expected by paper pot inoculation with vermiculite.

Prefecture. The tests were initiated in a field reclaimed with the application of mountain soil about 30 cm depth without indigenous bradyrhizobia (Figure 38) [51], a rotated paddy field in Nagakura in Niigata Agricultural Research Institute (Figure 39) [52], and a sandy dune

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

**Figure 39.** Soybean plants cultivated in a rotated paddy field in Nagakura. Plants were planted with inoculated paper pots. Control: without N deep placement, Urea: deep placement of urea at 20cm. Coated Urea: deep placement of

Lime nitrogen is composed of about 60% calcium cyanamide (CaCN2) with calcium oxide and carbon, and the N content is about 20-23%. After application to the soil, the calcium cy‐ anamide is converted to urea, which is again degraded into NH3 and CO2. Dicyandiamide contained in lime nitrogen or formed during the degradation of calcium cyanamide is a po‐

nium produced by CaCN2 decomposition persists for a longer period of time and the nitrate concentration remains low in the soil. It is expected that the inhibition of nodulation and of the N2 fixation activity may be alleviated by low a level of nitrate accumulation. Also this fertilizer exerts some hormonal effects on plants and is used for controlling soil diseases

In each of these fertilizer treatments, IPP (Inoculated paper pot), DT (Direct transplantation of inoculated seedlings without paper pot) and NIPP (Non-inoculated paper pot) seedlings were transplanted in separate plots. Paper pots (height 13.5 cm, diameter 3 cm) were made of a biodegradable paper designed to break down in the field. The pots were open at the bottom to allow root expansion below the pot. A paper pot was filled with vermiculite and a seed was planted in each pot, and followed by inoculation of one ml suspension of *Bradyrhi‐*

+ to NO3 -

cells ml-1. Since the bradyrhizobium population in‐

. Therefore the ammo‐

field of Niigata University in Ikarashi [80].

Relationships

148

coated urea. Lime Nitrogen: deep placement of lime nitrogen. [52]

caused by bacteria and fungi.

*zobium japonicum* USDA110 of about 108

tent nitrification inhibitor, which retards the oxidation of NH4

In regard to the inoculation method, IPP (inoculated paper pot) tended to show the highest seed yield than DT and NIPP treatments. Especially in the newly reclaimed field having mountain soil with no indigenous rhizobia at the 30 cm depth, the inoculation by IPP or DT promoted seed yield to a level more than twice as much as the control treatment. Among the inoculation methods, the IPP and DT seedlings showed a higher seed yield than the NIPP seedlings.

Significantly higher seed yields in the rotated paddy field were obtained with the deep placement of CaCN2 IPP (6.12 t ha-2) and CU-100 IPP (6.04 t ha-2), compared with the Urea IPP (4.67 t ha-2) and Control IPP (3.31 t ha-2) treatments (Figure 39) [51]. A similar effect was observed in a reclaimed field and sandy dune field, where deep placement of lime nitrogen gave the same or better seed yields compared with coated urea. Recently, Sakashita et al. re‐ ported the promotive effect of deep placement of lime nitrogen in 8 sites of farmers' field in 2008, 2009 and 2010. Seed yields increased about 30% on average by deep placement of lime nitrogen.

The mechanism of yield promotion by deep placement of lime nitrogen for soybean growth and seed yield is summarized in Figure 40.

**Figure 40.** A model for the promotive yield effect of deep placement of lime nitrogen on nitrogen fixation and seed yield of soybean.

**a.** Deep placement of lime nitrogen is hydrolyzed to urea, then to ammonium and carbon dioxide. The ammonium does not easily leach out from the fertilization sites at 20 cm depth. Dicyandiamide contained in lime nitrogen or formed in the soil from cyanamide

> depresses nitrification to prevent ammonium oxidation to nitrate. As a result, nitrate leaching and denitrification is reduced and the ammonium can be sustained in the sub‐ soil for a long time.

