EFFECT OF DEEP PLACEMENT OF NITROGEN FERTILIZERS ON RICE YIELD AND N USE EFFICIENCY UNDER WATER REGIMES

The experiments were conducted at the Soil Science Field Laboratory of Bangladesh Agricultural University, Mymensingh during boro season of 2013 to evaluate the effect of deep placement of nitrogen (N) fertilizers on rice yield and N use efficiency under two different water regimes [continuous flooding (CF) and alternate wetting and drying (AWD)]. After deep placement of USG and NPK briquettes and each split application of PU, the water samples were collected for five consecutive days and analyzed for ammonium-Nitrogen (NH4-N) concentration. In both water regimes the floodwater NH4-N concentration gave highest value at the second day of PU application followed by gradual decrease with time and in continuous flooding condition the release of NH 4-N was higher compared to AWD condition. Deep placement of N fertilizers also increased grain yield under AWD compared to CF condition and the maximum value was obtained in treatment T5 [USG 78 kg N ha -1 + PM 3 t ha -1 ] statistically similar to T3 [USG, 156 kg N ha -1 ]. Due to application USG and NPK briquettes, the N uptake and N use efficiency also gave higher values in AWD compared to CF condition. The results revealed that the deep placement of USG with poultry manure performed better in terms yield, N uptake and N use efficiency by BRRI dhan29 under AWD condition. So, the deep placement of USG under AWD condition could be recommended for higher production of boro rice.


INTRODUCTION
In Bangladesh, rice is grown under diverse irrigated, rainfed and deep water conditions in the three distinct seasons, namely Aus, Aman and Boro.Out of total rice production in this country about 43% comes from boro and the rest comes from Aman and Aus crops (BBS, 2011).For cultivation of rice, farmers of Bangladesh solely depend on urea fertilizer and 50% of the total demand is fulfilled by import which costs a huge amount of foreign currency (BBS, 2012).On the other hand, excessive N fertilization is one of the major concerns in sustainable agriculture for its decreased N-utilization efficiency by crops and increased N release to the environment, resulting atmosphere and water pollution (Zhu et al., 1997).The volatilization loss of prilled urea (PU) is very high and farmers lose a huge amount of money for N fertilizer.Therefore, the deep placement of urea super granule (USG) might be a good option to minimize the production cost as well as to increase crop yield.Ammonia volatilization losses in the flooded soils range from negligible to almost 60% of the applied N (Xing and Zhu, 2000).Deep placement of N fertilizers into the anaerobic soil zone is an effective method to reduce volatilization loss (Mikkelsen et al., 1978).
Again, water shortage during the boro season in Bangladesh is a growing problem due to climatic change and upstream water regulation of the major rivers in Bangladesh.One major recent advance in rice water management is termed Alternate Wetting and Drying (AWD) which combines the beneficial aspects of both aerobic and anaerobic cultivation.For AWD practices, the fields are managed as irrigated lowland rice but the top soil layer is allowed to dry out to some degree before irrigation is applied again (Belder et al., 2004).
The number of days under non-flooded soil conditions can vary depending on plant development stages and availability of water.The AWD irrigation causing alternating oxic and anoxic conditions in top soil may lead to increased N losses from coupled nitrification-denitrification (Nicolaisen et al., 2004;Liu et al., 2010).Denitrification loss could be enhanced under AWD irrigation system but ammonia (NH 3 ) volatilization is the major N loss pathway from irrigated rice field (De Datta et al., 1991).Ventura and Yoshida (1977) reported reduced NH 3 volatilization under AWD than continuous flooded condition.AWD can lower water use for irrigated rice by 35% (Zhang et al., 2009), increase rice yield by ~10% relative to permanent flooding (Yang et al., 2009;Zhang et al., 2009).Hence, the effect of deep placement of N fertilizer on yield and nitrogen use efficiency of boro rice under two different water regimes were examined in this study.

MATERIALS AND METHODS
Two experiments were conducted under continuous flooded (CF) and AWD conditions with similar N fertilizer treatment at the Soil Science Field Laboratory of Bangladesh Agricultural University, Mymensingh to evaluate the effect of water regimes on the yield and N use efficiency in boro rice.The soil belongs to Sonatala series under the AEZ-9 (Old Brahmaputra Floodplain).The soils were silt loam in texture.In AWD condition, the soil had pH 6.07, organic matter content 1.10%, total N 0.055%, available P 3.8 ppm, exchangeable K 0.24 meq% and available S 12.56 ppm.In continuous flooding condition the soil had pH 6.27, organic matter content 1.95%, total N 0.136%, available P 3.16 ppm, exchangeable K 0.095 meq% and available S 10.5 ppm.There were 7 treatments consisting of different sources of N and these include-T 1 = Control, T 2 = PU 156 kg N ha -1 , T 3 = USG 156 kg N ha -1 , T 4 = USG 104 kg N ha -1 , T 5 = USG 78 kg N ha -1 + PM 3 t ha -1 , T 6 = USG 78 kg N ha -1 + CD 3 t ha -1 and T 7 = NPK briqutte102 kg ha -1 .The experiment was laid out in a Randomized Complete Block Design (RCBD) with three replications.Forty day old rice seedlings were transplanted to the plots maintaining a spacing of 20 cm × 20 cm.Fertilizers were applied as per treatment.All the treatments except T 7 received 20 kg P and 50 kg K ha -1 from TSP and MoP, respectively.T 7 received NPK briquette fertilizer.Gypsum was applied to all the plots as basal dose as sulphur.PU was applied in three equal splits as top dress at 10 days after transplanting (DAT); 35 DAT (active tillering stage) and at 55 DAT (panicle initiation stage).USG and NPK briquettes were applied at 10 DAT and were placed at 8-10 cm depth between four hills at alternate rows.
Different intercultural operations such as weeding, pest control etc was done when required.In continuous flooded plots water was always maintained at 6 cm depth and thus irrigations were applied as and when necessary.Alternate wetting and drying (AWD) condition was maintained through limited water supply.For proper monitoring a PVC pipe was inserted in control plots and irrigation was provided when the water level goes below 15 cm down from the surface.The first alternating wetting and drying cycle is deployed 10-15 days after transplanting and continued until the commencement of flowering.The wetting/drying cycle consists of flooding the field followed by dry out 15 cm below the soil surface (as observed in the tubes); the field is then re-flooded to 2 cm above the soil surface before next drying cycle begins.
Before collecting water sample, irrigation was given to maintain a water depth of about 6 cm and for this reason time to time irrigation was done.T he field was supposed to irrigate after 3-5 days of depletion of water from the field in order to maintain reduced water condition (alternate wetting and drying condition), although, it was not always feasible practically due to frequent rains.Water samples were collected for 5 consecutive days after deep placement of USG and NPK briquettes and each split application of PU.The concentration of NH 4 -N was determined by Phenol-hypochlorite method (Solorzano, 1969).
The crop was harvested at full maturity and the data on grain and straw yields were recorded.The grain yield was expressed at 14% moisture basis and straw yield was recorded on sundry basis.The grain and straw samples were analyzed for N content following semi-micro Kjeldahl method (Bremner and Mulvaney, 1982).The N uptake by grain and straw was determined from N content and yield data.The N use efficiency (kg grain yield increase kg -1 N applied) was determined by the following formula: NUE = (Gy + N -Gy ON ) / FN, Where Gy + N = grain yield in treatment with N application; Gy ON = grain yield in treatment without N application and FN = amount of fertilizer N applied (kg ha -1 ).The apparent N recovery was calculated by the following formula: ANR (kg ha -1 ) = (UN + N -UN 0N ) /FN; where, UN +N is total N uptake (kg ha -1 ) with grain and straw; UN 0N is the N uptake (kg ha -1 ) in control; FN is amount of fertilizer N applied (kg ha -1 ).All the data were statistically analyzed by F-test and the mean differences were ranked by DMRT at 5% level (Gomez and Gomez, 1984).

RESULTS AND DISCUSSION NH 4 -N in flood water
Ammonium concentration in the water sample was significantly influenced by the application of different treatments have been depicted in figure A. Due to application of PU, USG and NPK briquettes, the concentration of NH 4 -N in flood water varied widely.According to figure 1, NH 4 + concentration rocketed just after the application of first split of PU and in the second day, NH 4 + concentration in both CF and AWD conditions reached at its peak.On the other hand, NH 4 + concentration remained stable in all USG and NPK briquettes treated plots (T 3 , T 5 , T 6 , and T 7 ) except T 5 in CF condition and T 7 in AWD condition.After fourth day the concentration of ammonium in the all treatments became almost similar.This reveals that the USG and NPK briquettes treated plots (T 3 , T 4 , T 5 , T 6 , 7 ) had much lower concentration of ammonium in flood water.+ concentration (6 ppm in CF and 8 ppm in AWD) was recorded for the treatment T 2 [PU, 156 kg N ha -1 ].After second day, the NH 4 + concentration started to decrease slowly in PU treated plot and at fifth day it was around 1-2 ppm in both CF and AWD condition.The NH 4 + concentration in T 3 , T 4 , T 5 , T 6 , and T 7 remain around 1ppm for the whole time.+ concentrations was produced for the treatment T 2 (PU, 156 kg N ha -1 ).After that, NH 4 + concentration decreased slowly and at fifth day, the concentration in all treatments became almost similar.The treatment T 2 gave higher NH 4 + concentrations compared to the plots treated with USG (T 3 , T 4 , T 5 , and T 6 ) and NPK briquettes (T 7 ).
The results on ammonium concentration of boro rice field water demonstrate that the PU treated plots (T 2 ) had much higher concentration of ammonium in water compared to the plots treated with USG (T 3, T 4 , T 5 , and T 6 ) and NPK briquette (T 7 ) as shown in figure 1.The treatment T 2 (PU, 156 kg N ha -1 ) showed the maximum NH 4 -N concentration followed by treatment T 4 (USG, 104 kg N ha -1 ).Within five days, the NH 4 -N concentration in all the treatments became almost similar.The USG treated plots (T 3 , T 4, T 5 and T 6 ) and NPK briquette treated plots (T 7 ) had much lower concentration of ammonium in flood water indicating that deep placement of urea fertilizer can reduce N loss by ammonia volatilization.Deep placement of N reduces ammonium N in floodwater.This not only improves N use efficiency in rice but also minimizes N loss resulting from ammonia volatilization and denitrification (Savent and Stangel, 1990;Mohanty et al., 1999).This finding is also accorded with Xiang et al. (2013).There were sharp decline in the amount of NH 4 -N; 2 to 3 days after broadcasting of PU may be attributed to volatilization loss, diffusion of NH 4 -N into soil, and / or nitrification losses.

Grain yield
The grain yield boro rice was responded significantly due to water management and the application of PU, USG and NPK briquette (Figure 2).In AWD condition the grain yield ranged from 3033 to 6714 kg ha -1 and in continuous flooded condition it was varied from 3261 to 6425 kg ha -1 .In both condition, the highest grain yield was recorded in T 5 (USG 78 kg N ha -1 + PM 3 t ha -1 ) and the lowest in T 1 (control).USG with poultry manure followed by USG performed better in increasing grain yield of rice compared to PU and NPK briquette alone.On the other hand, higher grain yield is found with AWD condition compared to CF condition in every treatment except T 1 and T 7.

Straw yield
Straw yield of boro rice also responded significantly to different treatments (Figure 3).In AWD condition, the highest straw yield (6516 kg ha -1 ) was found in T 5 (USG, 78 kg N ha -1 + PM 3 t ha -1 ) and the lowest value (3214 kg ha -1 ) in T 1 (control).The straw yield produced by the treatment T 3 was identical to T 5 .In case of continuous flooded condition, the maximum straw yield (7428 kg ha -1 ) was obtained from the treatment T 5 which was identical with that of treatment T 2 (PU, 156 kg N ha - 1 ), T 3 and T 4 (USG, 104 kg N ha -1 ).Every treatment except T 3 under CF condition gave better straw yield compared to AWD condition.

Nitrogen uptake
It was revealed from the figure 4 that the total N uptake by boro rice was influenced significantly by the application of PU, USG and NPK briquettes under CF and AWD condition.In CF condition, the highest total N uptake was observed in treatment T 6 (78 k g N ha -1 from USG + CD 3 t ha -1 ) was statistically identical to T 7 (102 kg N ha -1 as NPK briquette), T 5 (78 kg N ha -1 from USG + PM 3 t ha -1 ) and T 3 (156 kg N as USG).In AWD condition, the maximum total N uptake was recorded in T 5 [USG, 78 kg N ha -1 + PM 3 t ha -1 ] followed by T 3 [USG, 156 kg N ha -1 ].On the other hand, treatments T 2 (PU, 156 kg N ha -1 ), T 4 (USG, 104 kg N ha -1 ) and T 6 (USG, 78 kg N ha -1 + CD 3 t ha -1 ) gave statistically similar total N uptake.The results revealed that that the total N uptake by rice grain and straw was increased significantly with the deep placement of USG combined with organic manure in all the treatments but T 7 and T 1 gave significantly higher N uptake in AWD condition compared to CF.

Apparent N recovery
The apparent N recovery (ANR) indicates the absorption efficiency of applied N. According to figure 5 mean apparent recovery of N by rice ranged from 36.26 to 55.79% in different treatments under AWD condition and 26.97 to 52.77% in CF.The maximum value of ANR was obtained with the application of USG in treatment T 5 (USG, 78 kg N ha -1 + PM 3 t ha -1 ) followed by T 3 (USG, 156 kg N ha -1 ) under AWD condition whereas (78 kg N as USG + CD 3 t) followed by T 7 (102 kg N as NPK briquette) gave the maximum ANR value under CF condition.The data clearly indicate that the deep placement of USG enhanced the recovery of applied N compared to broadcast application of NPK fertilizers.

Nitrogen use efficiency
Nitrogen use efficiency (NUE) represents the response of rice plant in terms of grain yield to N fertilizer.NUE responded significantly due to different treatment combinations in CF and AWD condition.The range of NUE varied from 12.86 to 28 kg grain increase per kg N applied (Figure 6) in CF condition.The maximum value (28 kg grain increase per kg N applied) was obtained in T 7 (102 kg N ha -1 from NPK briquette) followed by T 5 (25.11kg grain increase per kg N applied), T 6 (24.06 kg grain increase per kg N applied), T 4 (18.29 kg grain increase per kg N applied) and the lowest value in T 3 (156 kg N ha -1 from USG).In AWD condition, the maximum value of NUE (29.21 kg grain increase per kg N applied) was obtained in T 5 (USG, 78 kg N ha -1 + PM 3 t ha -1 ) followed by T 4 (25.57kg grain increase per kg N applied), T 6 (25.57kg grain increase per kg N applied), T 3 (23.54kg grain increase per kg N applied), T 7 (20.82kg grain increase per kg N applied) and the minimum value was found in T 2 (18.62 kg grain increase per kg N applied).These results indicates that application of NPK briquettes and USG in combination with organic manure in rice field decreases the losses of N or the rate of N, leading to efficient uptake and utilization of applied N under AWD condition compared to CF condition.

CONCLUSION
The overall results indicate that the treatment comprising T 2 , T 3 , T 4 , T 5 produced the higher grain yield under AWD condition.Similarly, the highest N recovery and N use efficiency were also found in the same treatments as compared to CF condition.This reveals that the deep placement of USG fertilizer under AWD condition enhanced the N use efficiency.On the other hand, T 6 , T 7 gave the maximum grain yield, N recovery and N use efficiency under CF condition compared to AWD condition.

Figure
Figure B shows the effect of USG, PU and NPK briquette on NH 4 + concentration of water sample collected during 10-14 March, 2013.After two days of N fertilizer application the highest amount of NH 4+ concentration (6 ppm in CF and 8 ppm in AWD) was recorded for the treatment T 2 [PU, 156 kg N ha -1 ].After second day, the NH 4 + concentration started to decrease slowly in PU treated plot and at fifth day it was around 1-2 ppm in both CF and AWD condition.The NH 4 + concentration in T 3 , T 4 , T 5 , T 6 , and T 7 remain around 1ppm for the whole time.

Figure
Figure C shows the effect of PU, USG and NPK briquettes on NH 4 + concentrations in water samples collected during 1-5 April, 2013.At the second day of N fertilizer application, the highest amount of NH 4+ concentrations was produced for the treatment T 2 (PU, 156 kg N ha -1 ).After that, NH 4 + concentration decreased slowly and at fifth day, the concentration in all treatments became almost similar.The treatment T 2 gave higher NH 4 + concentrations compared to the plots treated with USG (T 3 , T 4 , T 5 , and T 6 ) and NPK briquettes (T 7 ).

Figure 1 .
Figure 1.Ammonium concentration in floodwater after deep placement of N fertilizers under Continuous Flooding (CF) and Alternate Wetting and Drying (AWD) Condition.[A: First Sampling; B: Second Sampling; C: Third Sampling]