Absorption of molecular urea by rice under flooded and non-flooded soil conditions

A pot experiment was conducted with rice to study the relative absorption of urea in molecular form compared to the other forms of N produced in soil from the applied urea. A method involving application of ¹⁴C-labelled urea and ¹⁵N-labelled urea alternately in two splits was used to quantify the absorption of molecular urea and other forms of N formed from it. Biomass production and N uptake were greater in plants grown under flooded soil conditions than in plants grown under non-flooded (upland) conditions. Absorption of N by rice increased with increasing rate of urea application up to 250 mg pot⁻¹ and declined thereafter. The absorption of urea from the flooded soil constituted 9.4% of total N uptake from applied N compared to only 0.2% from the non-flooded. Under submerged conditions, absorption of urea from topdressing was about twice that from basal application at planting. High water solubility of the fertilizer and better developed rice root system might have enhanced the absorption of molecular urea by flooded rice, especially from topdressing. Thus, in the flooded rice system, the direct absorption of molecular urea from topdressing accounted for 6.3% of the total N uptake from added urea. Under upland condition, it was 0.12%. This revised version was published online in June 2006 with corrections to the Cover Date.     

Effects of deep placement and surface application of urea on the yield of wetland rice in pot trials

Summary The effects of deep placement and surface application of urea fertilizer on the yield of rice grown in pots of alluvial clay soil covered with 5 cm water was studied under controlled conditions. Application of two levels of urea supergranules and prills (2 g and 4 g urea/0.1 m²) on the surface of submerged soil increased the vegetative growth and enhanced the grain yield as much as 85%. However, no difference in yield was found between urea prills applied in three split doses and one application of urea supergranules.Deep placement of two levels of urea supergranules in the soil at four different depths (2.5, 5.0, 10.0 and 15.0 cm) resulted in the highest yields. The fertilizer was most efficient when the highest concentration was placed in the soil at a depth of 5.0 cm. This application method increased the grain yield by 20% as compared with the soil surface application.     

Uptake and use patterns of nitrogen from urea,oxamide, and isobutylidene diurea by rice plants

Summary Two¹⁵N-labelled slow-release nitrogen (N) sources, oxamide and isobutylidene diurea (IBDU), each at two particle sizes, and¹⁵N-labelled urea were compared at two rates as sources of N for rice (Oryza sativa) under two watering regimes which simulated a transplant (continuous flood, CF) and a direct-seeded (A/F) system of paddy rice culture. Highest grain yields were obtained from −8+10-mesh oxamide particles applied at the rate of 2,000 mg of N/5 kg of soil, CF series; this yield was slightly higher than that obtained from −3+4-mesh oxamide, A/F series. Incubating the N fertilizers in moist (22% water) soil for 21 days immediately before flooding and transplanting rice greatly reduced N supply because of nitrification during the preflood period, followed by denitrification after flooding. This resulted in less plant uptake of N and less grain yield from urea, fine oxamide and IBDU, A/F series. For coarse oxamide, N release during the preflood period resulted in higher N uptake and grain yield in the A/F rather than in the corresponding CF series. The pattern of fertilizer N uptake by rice plants was affected by kind of fertilizer, particle size of oxamide and IBDU, and watering regime. Uptake of fertilizer N generally paralleled uptake of soil N throughout the growth period. Plant tops continued to accumulate some N during the period of grain filling, but much of the N in plant tops was translocated to the grain after heading. There was a large decrease in dry weight, N content, and¹⁵N content of tops after heading. Root weight and N content increased rapidly at first, and then at a diminishing rate until maturity. Unexplained N deficits occurred in the CF series (14–23% of the N applied, depending on N rate and source), and in the A/F series for IBDU (37–43% of the N applied).     

Labeled nitrogen fertilizer research with urea in the semi-arid tropics

Summary Field studies with bordered microplots were conducted on an Alfisol in the semiarid tropics of India to determine (1) the fate of¹⁵N-labeled urea applied to dryland sorghum in two successive rainy seasons and (2) the effect of method of application on N fertilizer efficiency. Recoveries of¹⁵N-labeled fertilizers by above-ground plant parts ranged from 46.7% to 63.6% in 1981 when the rainfall was above the average and from 54.4% to 66.9% in 1980 when the rainfall was near the average. Small (0.014 g) pellets of urea applied twice as postemergent applications in separate 5 cm deep bands were more effective than single preemergent applications either surface applied or incorporated. Both banding and the split applications contributed to overall fertilizer efficiency. Large (1.0 g) pellets of urea (supergranules) placed at a depth of 5 cm were also superior to the incorporated, small-pellet treatment in 1981. The¹⁵N-balance data for the soil (0–90 cm in depth)-plant system in 1981 showed that the unaccounted-for fertilizer N ranged from 5.1% to 20.6%. An important finding was that high grain yields, in excess of 6,000 kg/ha, with N fertilizer losses of less than 10% could be obtained through fertilizer management during a very wet season. The data from the Alfisol experiments were compared with data from similar Vertisol experiments; N fertilizer losses resulting from incorporated and surface applications were greater for Vertisols than for Alfisols in the wetter year.     

Effects of field fertilizer practices on labeled ammonium-nitrogen transformations and its utilization by winter wheat

Summary Field experiments with winter wheat (Triticum aestivum L.) were conducted in two years at two locations using¹⁵N-enriched (NH₄)₂SO₄ on Smolan silt loam (Pachic Argiustoll) and Ost loam (Typic Arguistoll) soils. The objective was to relate differences in crop utilization of fertilizer to movement and transformations of the N in a complete factorial experiment having fall and spring applications, banded and broadcast, with and without nitrapyrin. Plant uptake of the 60 kg N/ha applied varied from 31% to 62% with greatest uptake when fertilizer was banded in the spring without nitrapyrin and least uptake from fall and spring broadcast treatments using nitrapyrin. Analysis of single factor effects showed greater crop contents of fertilizer N for spring than fall applications. That was related to immobilization of the applied N. Much more fertilizer N was in inorganic forms during the period of rapid wheat growth with spring applications than with fall. Banding the fertilizer at a depth of 0.05 m resulted in greater plant uptake than broadcasting or banding it on the soil surface. A significant portion of the applied N was immobilized near the point of application. That limited the downward movement of the N placed on the surface, making it less available to plant roots than the N placed 0.05 m deep where soil moisture was more favorable. Use of nitrapyrin resulted in lowered amounts of fertilizer N as NO₃-until mid-May for fall treatments and until harvest with spring treatments. That appeared to be the reason for lowered plant uptake when nitrapyrin was used. Published in memory of Professor R V Olson and over 40 years of contributions and service to agriculture and soil science (1919–1985).     

Fertilizer vs. organic matter contributions to nitrogen leaching in cropping systems of the Pampas: 15N application in field lysimeters

Nitrogen (N) export from soils to streams and groundwater under the intensifying cropping schemes of the Pampas is modest compared to intensively cultivated basins of Europe and North America; however, a slow N enrichment of water resources has been suggested. We (1) analyzed the fate of fertilizer N and (2) evaluated the contribution of fertilizer and soil organic matter (SOM) to N leaching under the typical cropping conditions of the Pampas. Fertilizer N was applied as ¹⁵N-labeled ammonium sulfate to corn (in a corn/soybean rotation) sown under zero tillage in filled-in lysimeters containing two soils of different texture representative of the Pampean region (52 and 78 kg N ha^-1, added to the silt loam and sandy loam soil, respectively). Total fertilizer recovery at corn harvest averaged 84 and 64% for the silt loam and sandy loam lysimeters, respectively. Most fertilizer N was removed with plant biomass (39%) or remained immobilized in the soil (29 and 15%, for the silt loam and sandy loam soil, respectively) whereas its loss through drainage was negligible (<0.01%). We presume that the unaccounted fertilizer N losses were related to volatilization and denitrification. Throughout the corn growing season, subsequent fallow and soybean crop, which took place during an exceptionally dry period, the fertilizer N immobilized in the organic pool remained stable, and N leaching was scarce (7.5 kg N ha^-1), similar at both soils, and had a low contribution of fertilizer N (0–3.5%), implying that >96% of the leached N was derived from SOM mineralization. The inherent high SOM of Pampean soils and the favorable climatic conditions are likely to propitiate year-round production of nitrate, favoring its participation in crop nutrition and leaching. The presence of ¹⁵N in drainage water, however, suggests that fertilizer N leaching could become significant in situations with higher fertilization rates or more rainy seasons.     

The effect of deep placement and surface application of nitrogen fertilizers at different light intensities on growth and yield of wetland rice

Summary Lowland rice (RD 3) was cultivated in containers of clay soil submerged with 5 cm water under controlled conditions in the phytotron. Deep placement of urea supergranules 5 cm in the soil significantly enhanced both plant growth and fertilizer efficiency when the plants were cultivated under high light intensity (70 Wm^–2). At the highest urea level grain yield increased 119% above the control level, while growth and fertilizer efficiency was not as high when deep placement of calcium nitrate was used.The application of urea prills and calcium nitrate (18.4g Nm^–2) in two split doses on the soil surface increased grain yield as much as 91% above the control level. At the lower nitrogen concentration (9.2 g N m^–2), the urea prills were more efficient than calcium nitrate as indicated by the grain yield. The height of those plants fertilized by surface application was affected by the concentration and not the type of fertilizer. The number of tillers, however, was significantly higher on urea fertilized plants.When the rice plants were cultivated under low light intensity 930 Wm^–2), neither the nitrogen fertilizers nor the method of application had a significant effect on growth and yield.     

Labeled nitrogen fertilizer research with urea in the semi-arid tropics

Summary As part of a research program to determine the fate of N fertilizers applied to dryland sorghum in the semi-arid tropics,¹⁵N balance studies were conducted with various N sources in the greenhouse. Two American soils, Houston Black clay (Udic Pellustert) and Windthorst sandy loam (Udic Paleustalf), similar in properties to the Vertisol and Alfisol in the semi-arid tropics of India, were employed. Experiments were conducted with large pots containing 20 or 60 kg of soil which was subjected to several watering regimes. The¹⁵N not accounted for in the plant and soil was presumably lost. Losses of N on calcareous Houston Black clay were greatest for broadcast urea, 16%–28%. Amendment of broadcast urea with 2% phenyl phosphorodiamidate, a urease inhibitor, reduced N losses only slightly to 15%–20%. Point placement of urea at a 6 cm soil depth reduced losses to 1%–11%. Granule size had no effect on N loss from point-placed urea. Ammonia volatilization was apparently the main N loss mechanism, since N losses from sodium nitrate were less than 7%, except when the soil surface was waterlogged. N losses on the Windthorst soil averaged 30% from urea and 11% from ammonium nitrate. Amendment of urea with urea phosphate to form a 27% N and 13% P product reduced fertilizer N losses but did not increase grain yield on Windthorst soil. N losses were also less from ammonium nitrophosphate than from urea. Band and point placement of urea 6 cm below the soil surface were equally effective in reducing N loss on Houston Black clay. The findings give credence to the recommendation of deep band placement for urea in the semi-arid tropics.     

The partitioning of fertilizer-N between soil and crop: Comparison of ammonium and nitrate applications

Field experiments were carried out in 1987 on winter wheat crops grown on three types of soil. ¹⁵N-labelled urea, ¹⁵NH₄NO₃ or NH₄ ¹⁵NO₃ (80 kg N ha^-1) was applied at tillering. The soils (chalky soil, hydromorphic loamy soil, sandy clay soil) were chosen to obtain a range of nitrogen dynamics, particularly nitrification. Soil microbial N immobilization and crop N uptake were measured at five dates. Shortly after fertilizer application (0–26 days), the amount of N immobilized in soil were markedly higher with labelled urea or ammonium than that with nitrate in all soils. During the same period, crop ¹⁵N uptake occurred preferentially at the expense of nitrate. Nitrification differed little between soils, the rates were 2.0 to 4.7 kg N ha^-1 day^-1 at 9°C daily mean temperature. The differences in immobilization and uptake had almost disappeared at flowering and harvest. ¹⁵N recovery in soil and crop varied between 50 and 100%. Gaseous losses probably occurred by volatilization in the chalky soil and denitrification in the hydromorphic loamy soil. These losses affected the NH₄ ⁺ and NO₃ ^- pools differently and determined the partitioning of fertilizer-N between immobilization and absorption.     

The effect of intermittent flooding on the growth and yield of wetland rice and nitrogen-loss mechanism with surface applied and deep placed urea

Summary Wetland rice was cultivated in pots of puddled soil under continuous and intermittent flooding conditions. The soil was either fertilized with the surface application of prilled urea in three split doses or once with urea supergranules applied at different soil levels.The grain yield, fertilizer efficiency and percent nitrogen recovery by the grains were increased by deep placement of urea supergranules independent of the water regime. Grain yield was always lower with intermittent flooding, particularly when the plants were fertilized with the surface application of prilled urea.Nitrogen loss by ammonia volatilization, measured in a closed cuvette system, was reduced from 24% with the surface application of urea prills and 20% with surface application of urea supergranules to approximately 2% with deep placement of urea supergranules. Intermittent flooding created conditions which promoted additional nitrogen loss by nitrification and denitrification processes. The total nitrogen loss, measured in an open cuvette system, was about 38% with the surface application of urea supergranules, whereas this loss was reduced to 10% with deep placement of urea supergranules. Furthermore, deep placement of urea fertilizer reduced the nitrogen loss irrespective of water regime.     

Ammonia volatilization losses from prilled urea,urea supergranules (USG) and coated USG in rice fields

Summary About 8.4 per cent of applied nitrogen was lost as ammonia during a week after application when prilled urea was broadcast or banded and incorporated in soil 20 days after sowing of rice. Ammonia volatilization was reduced to 3.3 per cent when urea supergranules (USG) were used. Coating of USG with DCD or neem cake showed no advantage. Ammonia volatilization was only 0.7 to 1.6 per cent when fertilizer was applied at panicle initiation stage of rice; highest values were again obtained with prilled urea. The experiments were carried out in closed cages.     

Nitrogen uptake by crops,soil distribution and recovery of urea-N in a sorghum—wheat rotation in different soils under Mediterranean conditions

The N uptake by crops, soil distribution and recovery of 15N labelled urea-N (100 kg N ha-1) were investigated in a sorghum-wheat rotation in two silty clay soils (Foggia and Rieti Casabianca) and one silt loam soil (Rieti Piedifiume) under different mediterranean conditions. Non-exchangeable labelled NH4-N represented an important pool at both Rieti sites with higher values (p<0.05) under sorghum (14.0 and 24.6% of the urea N in the 0-20 cm layer at the end of the cropping season) than wheat whereas it was much less important in the Foggia soil (10.0% in the surface soil under sorghum). This is probably related to the clay minerals composition of the three soils; because vermiculite was present in both Rieti sites but not in the Foggia soil. At harvest from 4.4 to 5.3% of the urea N initially applied was present as microbial biomass N in the surface soil layer with no generally significant differences due to location and type of crops. Both sorghum and wheat N yields were higher in the driest site (Foggia) probably due to better light conditions, higher temperatures and irrigation during summer of the sorghum cropping period. The recovery of plant fertilizer N (about 21% for sorghum and 27% for wheat) and the percentage of N in the plant derived from the fertilizer (NDFF) were the lowest at Rieti-Casabianca probably as the result of the protection of immobilized fertilizer N against microbial mineralization by the swelling clays. The fertilizer N unaccounted for was nil or very low (10.8% at Rieti-Casabianca under wheat and 11.8 and 4.9% at Rieti-Piedifiume under sorghum and wheat, respectively). Urea-N losses occurred when Rieti Piedifiume and Rieti Casabianca soils were kept bare. In this case the urea N unaccounted for ranged from 12 to 56% of the urea N with higher losses in Rieti-Piedifiume than in Rieti-Casabianca. The higher recoveries in the latter soil were probably confirmed by the stabilizing effect of clays on the immobilized urea N. This revised version was published online in July 2006 with corrections to the Cover Date.     

Cyanobacterial inoculation and nitrogen fertilization in rice

During three rice-growing seasons in Uruguay, field experiments were conducted to study the contribution of cyanobacterial inoculation and chemical N fertilization to rice production. Neither grain yield nor fertilizer recovery by the plant were affected by inoculation with native cyanobacterial isolates. A low fertilizer use efficiency (around 20%) was observed when labelled (NH₄)₂SO₄ was applied at sowing. Recovery of applied ¹⁵N by the soil–plant system was 50%. Inoculation did not modify ¹⁵N uptake by the plant when the fertilizer was three-split applied either. The total N-fertilizer recovery was higher when the fertilizer was split than when applied in a single dose. Plant N-fertilizer uptake was higher when the fertilizer was applied at tillering. Uptake of ¹⁵N from cyanobacteria by rice was studied in a greenhouse pots experiment without chemical nitrogen addition. Recovery of ¹⁵N from labelled cyanobacteria by rice in greenhouse growth conditions was similar to that of partial recovery of (NH₄)₂SO₄ applied at sowing in the field. Cyanobacterial N mineralization under controlled conditions was fast as cyanobacterial N was detected in plants after 25 days. Moreover 40 days after inoculation non-planted and inoculated soil had more inorganic N than the non-inoculated one.     

The significance of denitrification of applied nitrogen in fallow and cropped rice soils under different flooding regimes I. Greenhouse experimènts

The role of nitrification-denitrification in the loss of nitrogen from urea applied to puddled soils planted to rice and subjected to continuous and intermittent flooding was evaluated in three greenhouse pot studies. The loss of N via denitrification was estimated indirectly using the¹⁵N balance, after either first accounting for NH₃ volatilization or by analyzing the¹⁵N balance immediately before and after the soil was dried and reflooded. When urea was broadcast and incorporated the loss of¹⁵N from the soil-plant systems depended on the soil, being about 20%–25% for the silt loams and only 10%–12% for the clay. Ammonia volatilization accounted for an average 20% of the N applied in the silt loam. Denitrification losses could not account for more than 10% of the applied N in any of the continuously flooded soil-plant systems under study and were most likely less than 5%. Intermittent flooding of soil planted to rice did not increase the loss of N. Denitrification appeared to be an important loss mechanism in continuously flooded fallow soils, accounting for the loss of approximately 40% of the applied¹⁵N. Loss of¹⁵N was not appreciably enhanced in fallow soils undergoing intermittent flooding. Apparently, nitrate formed in oxidized zones in the soil was readily denitrified in the absence of plant roots. Extensive loss (66%) of¹⁵N-labeled nitrate was obtained when 100 mg/pot of nitrate-N was applied to the surface of nonflooded soil prior to reflooding. This result suggests that rice plants may not compete effectively with denitrifiers if large quantities of nitrate were to accumulate during intermittent dry periods.     

Nitrogen uptake and recovery from urea and green manure in lowland rice measured by15N and non-isotope techniques

In the recent past considerable attention is paid to minimize dependence on purchased inputs such as inorganic nitrogen fertilizer. Green manure in the form of flood-tolerant, stem-nodulatingSesbania rostrata andAeschynomene afraspera is an alternative N source for rice, which may also increase N use efficiency. Therefore research was conducted to determine the fate of N applied to lowland rice (Oryza sativa L.) in the form ofSesbania rostrata andAeschynomene afraspera green manure and urea in two field experiments using¹⁵N labeled materials.¹⁵N in the soil and rice plant was determined, and¹⁵N balances established. Apparent N recoveries were determined by non-tracer method. ¹⁵N recoveries averaged 90 and 65% of N applied for green manure and urea treatments, respectively. High partial pressures of NH₃ in the floodwater, and high pH probably resulted from urea application and favoured losses of N from the urea treatment. Results show that green manure N can supply a substantial proportion of the N requirements of lowland rice. Nitrogen released fromSesbania rostrata andAeschynomene afraspera green manure was in synchrony with the demand of the rice plant. The effect of combined application of green manure and urea on N losses from urea fertilizer were also investigated. Green manure reduced the N losses from¹⁵N labeled urea possibly due to a reduction in pH of the floodwater. Positive added N interactions (ANIs) were observed. At harvest, an average of 45 and 25% of N applied remained in the soil for green manure and urea, respectively.Contribution from IRRI, Los Baños, Philippines and Justus-Liebig-University, Giessen, GermanyContribution from IRRI, Los Baños, Philippines and Justus-Liebig-University, Giessen, Germany     

Fate and efficiency of N fertilizers applied to pearl millet in Niger

Field studies were conducted in Niger using ¹⁵N-labeled fertilizers to assess the fate and efficiency of fertilizer N in pearl millet (Pennisetum glaucum [L.] R.Br.) production. Total plant uptake of fertilizer N was low in all cases (20%–37%), and losses were severe (25%–53%). The majority of N remaining in the soil was found in the 0- to 15-cm layer though some enrichment at lower depths was found when the N fertilizer was calcium ammonium nitrate (CAN). In a comparison of urea placement methods (band, broadcast, or point placement), no significant differences in ¹⁵N uptake or yield were noted though point placement did exacerbate ¹⁵N loss. The mechanism of N loss is believed to have been ammonia volatilization. Yields were similar whether urea or CAN was used, but ¹⁵N uptake from CAN was higher. A statistical model was developed relating millet yield and N response to midseason rainfall. In drought years, no N response was found, whereas in years of good rainfall a response was found of 15 kg grain for each kilogram of N applied (at 30 kg N ha^-1 rate).     

Urea fertilization and the N-cycle of rice-fields in the Camargue (S. France)

The use of urea as N-fertilizer for rice production has been studied for European conditions. In field and pot experiments, urea appeared to be more efficient than ammonia or nitrate. The rice yield could be maintained at maximal local production values by broadcasting 100 kg ha^-1 of urea-N instead of using 200–200 kg ha^-1 of ammonia or nitrate nitrogen. The best results were obtained by adding the urea about 5–8 weeks after sowing. When this was done, splitting the dose did not increase the efficiency. In laboratory experiments it was shown that urea diffuses and hydrolyses rapidly to ammonia in the anoxic layers of the flooded soils. Because this happens in the anoxic zones, the ammonia cannot be oxidised to nitrate, and is therefore less prone to denitrification, but it is adsorbed onto the soil particles. The upward diffusion of ammonia is slow, rendering it available to the rice roots for a relatively long time. The diminished use of N fertilizer is profitable for the rice farmer and helps to protect the surrounding environment.     

Fate of nitrogen applied as Azolla and blue-green algae (Cyanobacteria) in waterlogged rice soils—a15N tracer study

Summary ¹⁵N tracer was used to detect the extent to which nitrogen of appliedAzolla caroliniana, Anabaena variabilis andNostoc muscorum was available for assimilation by the growing rice plants in pots under 4 cm flood water for 60 days. The rate of release of nitrogen from the above biofertilizers, the amount of nitrogen remaining in the soils and the amount that was lost from the soils during this period were also examined. Previously¹⁵N-labelled biomass of Azolla, Anabaena and Nostoc to provide 40 mg N was mixed thoroughly with 0.5 kg silt loam Bangladesh soil (Sonatola series) in each of three pots used for a single treatment. Each pot received four 16 days old IR8 rice seedlings. A parallet set of experiments was conducted without rice plants.It was found that nitrogen uptake in the rice plants was increased by 91, 176 and 215% on using Azolla, Anabaena and Nostoc which resulted in increased total dry matter yields (shoot plus root) of 74, 105 and 125%, respectively. Of the total¹⁵N applied at the start, 26, 49 and 53% was released from Azolla, Anabaena and Nostoc; about 7, 14 and 13% was lost by denitrification and 74, 51 and 47% remained in the soils as the undecomposed part of the biofertilizers, respeciively, after 60 days. Of 15.76, 22.72 and 25.92 mg N assimilated by the rice plants, 48, 61 and 62% was supplied by Azolla, Anabaena and Nostoc, respectively. The rest was obtained from the soil used.In the absence of the rice plants 30, 43 and 45% of applied¹⁵N of Azolla, Anabaena and Nostoc was released, respectively, in 60 days of which 93–96% was lost as N₂ through denitrification.     

Influence of urea on biological N2 fixation and N transfer from Azolla intercropped with rice

The N₂ fixed by Azolla before and after urea application during the rice cycle, the mineralisation of Azolla-N as well as its availability to rice was studied in two greenhouse experiments conducted in 1996 and 1997 and in June 1998 in Goettingen (Germany). Dry matter production of the various rice parts of experiment 1 showed a clear positive synergism between treatment with Azolla and urea with a resulting apparent N recovery by rice increasing from 40% (without Azolla) to 57% in the presence of Azolla. Part of this increase may be due to N fixed biologically by Azolla and transferred to the rice. The second experiment shed some light on the role of BNF. Using an iterative method of estimation, the daily rate of N fixation was estimated at 0.6 – 0.7 kg N ha^–1. The rate was not so much affected by the age of the Azolla crop. At this rate, the BNF would amount to up to 100 kg N ha^–1 over a 130-day season. Assuming that BNF may be inhibited for a period of 5 – 10 days following urea application due to high levels of N in the floodwater, this might reduce the BNF by between 6 and 14 kg N ha over the season. Using the mean-pool-abundance concept, it was estimated that around 75 – 80% of the Azolla-N mineralized during the growth period was actually absorbed by the rice plants. Of the N taken up by rice around 28% was derived from the biologically fixed Azolla N, the remainder was urea N cycled through the Azolla. Azolla also seems to help sustain the soil N supply by returning N to the soil in quantities roughly equal to those extracted from the soil by the rice plant.     

Effectiveness of Sesbania rostrata and Phaseolus calcaratus as green manure for upland rice grown in acidic soil

Two field experiments on green manuring were conducted under upland acidic soil (pH = 4.35) conditions with the following objectives: (1) to determine the influence of inoculation site, P fertilization, and liming on the biomass production, N content, N accumulation, and N availability of S. rostrata grown in an acidic soil, (2) to compare the effectiveness of S. rostrata, P. calcaratus and urea as N sources for upland rice as affected by liming and N source-sowing time combination, and (3) to assess the effect of liming and N source-sowing time combination on % Ndff (N derived from the fertilizer), % Ndfs (N derived from soil), % FNU (fertilizer N utilization), and FNY or fertilizer N yield (kg N ha^–1) of upland rice grown in acidic soil. At 2 weeks after incorporating S. rostrata (95 days after lime application), liming significantly increased N availability by more than 2-fold suggesting that the decomposition of S. rostrata by soil microflora was stimulated by lime. Liming, phosphorus application, and inoculation site improved significantly the dry biomass production, N content and N accumulation of S. rostrata; thus, enhancing its green manuring potential. Regardless of liming, S. rostrata whether applied at 0 week or 2 weeks before sowing was superior to urea in improving grain and straw yields. P. calcaratus when applied at 2 weeks before sowing also produced higher grain yield than urea. Immediate sowing of upland rice after green manure incorporation did not affect negatively the growth and development of upland rice; hence, farmers could save at least 2 weeks in their cropping calendar. N source-sowing time combination had a highly significant influence on % Ndff, % Ndfs, % FNU, N uptake, and fertilizer N yield of upland rice. However, only N uptake was influenced significantly by liming. The rice plant obtained significantly higher % Ndfs from the soils treated with green manure than those treated with urea regardless of liming. The % FNU and % Ndff from the green manures were 11-37% and 9-25%, respectively. These values are much lower than those obtained under continuously flooded soil conditions possibly because of the differences in the organic matter decomposer populations and N loss mechanisms between sloping upland conditions and continuously flooded conditions.