One of the technology options that can help farmers cope with water scarcity at the field level is alternate wetting and drying (AWD). Limited information is available on the varietal responses to nitrogen, AWD, and their interactions. Field experiments were conducted at the International Rice Research Institute (IRRI) farm in 2009 dry season (DS), 2009 wet season (WS), and 2010 DS to determine genotypic responses and water use efficiency of rice under two N rates and two water management treatments. Grain yield was not significantly different between AWD and continuous flooding (CF) across the three seasons. Interactive effects among variety, water management, and N rate were not significant. The high yield was attributed to the significantly higher grain weight, which in turn was due to slower grain filling and high leaf N at the later stage of grain filling of CF. AWD treatments accelerated the grain filling rate, shortened grain filling period, and enhanced whole plant senescence. Under normal dry-season conditions, such as 2010 DS, AWD reduced water input by 24.5% than CF; however, it decreased grain yield by 6.9% due to accelerated leaf senescence. The study indicates that proper water management greatly contributes to grain yield in the late stage of grain filling, and it is critical for safe AWD technology.
Rice (
Continuous flooding (CF) provides a favorable water and nutrient supply under anaerobic conditions. However, the conventional system consumes a large amount of water [
Water use efficiency (WUE) is defined as the units of yield produced per unit of available water [
Nitrogen is one of the most important agricultural inputs to increase yield, and its use and uptake are affected by availability of water. Fertilizer application can improve both the crop yield and WUE. Hatfield [
Water deficiency can accelerate plant senescence and lead to a faster and better remobilization of carbon from vegetative tissues to the grain [
In our current study, we compared hybrid rice varieties and inbred varieties under two N rates (low N, high N) and two water management treatments (AWD, CF). The objectives of this study were (1) to determine rice yield potential and water use efficiency under two N rates and two water management methods, (2) to identify the factors that contribute to increased yield and water productivity under these conditions, and (3) to determine if there exist interactions among N, water management, and varieties.
The field experiments were conducted for three consecutive seasons (2009 dry season (DS), 2009 wet season (WS), and 2010 DS) in the same field at the International Rice Research Institute (IRRI) farm, Los Baños (14°11′N, 121°15′E, and 21 m als), Philippines. The soil was an Aquandic Epiaquoll with pH 6.2; 20.0 g kg−1 organic C; 2.0 g kg−1 total N; 11.4 mg kg−1 Olsen P; 0.43 cmol kg−1 exchangeable K and 34.5 cmol kg−1 cation exchange capacity; and 58.3% clay, 34.0% silt, and 8.0% sand. The soil test was based on samples taken from the upper 20 cm of the soil before transplanting in 2010 DS.
The experimental design was split-split plot with four replications in the three seasons. The main plots were two water management treatments (AWD and CF). The subplots were two N treatments: low N rate (60 kg ha−1 in WS, 100 kg ha−1 in DS) and high N (120 kg ha−1 in WS, 200 kg ha−1 in DS). The sub-subplots were four rice varieties; they belong to two groups: hybrid rice (IR72 and PSBRc80) and inbred varieties (IR82372H and Mestizo7). SL8-H was replaced with Mestizo7 because of its disease susceptibility in the WS.
In the CF plots, ponded water was kept with a depth of 3–5 cm during the 7 days after transplanting until the 7 days before maturity. In the AWD plots, soil water potential was measured with two porous-cup tensiometers installed at 20 cm and 40 cm depth. The depth of groundwater table was monitored using piezometers in open-bottom PVC tubes installed at a depth of 100 cm. Holes were perforated on all sides of the tube. When the ponded water dropped to 15 cm below the soil surface, then irrigation was applied to reflood the field up to 5 cm in AWD treatment. This cycle was repeated throughout the season. The first AWD treatment was initiated in the 3 weeks after transplanting. The irrigated water of each plot was measured using a 90° boxed Weir connected to an irrigation outlet. Daily mean temperature and rainfall were recorded from the weather station adjacent to the experimental site. Total water input = the amount of irrigated water applied + rainfall. Water productivity = grain yield/total amount of water supplied.
Pregerminated seeds were sown in seedling trays to produce uniform seedings. Fourteen-day-old seedlings were manually transplanted on January 6, June 10, and January 14 for 2009 DS, 2009 WS, and 2010 DS, respectively. Four seedlings per hill were transplanted at a hill spacing of 20 cm × 20 cm. Insects, diseases, and weeds were intensively controlled by using approved pesticides to avoid biomass and yield loss. Fertilizers were manually broadcasted and incorporated during basal application: 30 kg P ha−1, 40 kg K ha−1, and 5 kg Zn ha−1 in the DS and 15 kg P ha−1, 20 kg K ha−1, and 2.5 kg Zn ha−1 in the WS. Nitrogen in the form of urea was applied. During DS, low N rate was supplied with 40, 20, 40, and 20 kg N ha−1 at basal, midtillering, panicle initiation, and booting, respectively. High N rate corresponded to 60-40-60-40 kg N ha−1. During WS, low N rate was lowered to 20-10-20-10 kg N ha−1 while the rate for high N was reduced to 30-20-30-20 kg N ha−1. In the dry season, total N rate was 120 and 200 kg ha−1 for the low and high N rates, respectively. In the wet season experiments, total N rate was 60 and 100 kg ha−1 for the low and high N rates, respectively.
The soil water content (SWC) of the soil was monitored when water was deficient in the AWD treatment in 2010 DS. In each plot, soil samples were taken every 2 days using a core sampler. Fresh weight of the soil samples was measured immediately. Dry weight was obtained after oven drying at 105°C for 24 h. The soil water content was calculated following the equation: SWC = 100 × (fresh weight − dry weight)/fresh weight. Three varieties (IR72, IR82372H, and SL-8H) were used to measure grain filling and SPAD value. At the onset of flowering, 150 panicles headed on the same day were initially tagged from the high N plots. Among these panicles, ten were taken every two days from heading until maturity. The SPAD value of its flag leaf was also measured before sampling. Dry weights of the spikelets were determined after oven drying at 70°C to constant weight.
For growth analysis, 12 hills were sampled from each plot at flowering to measure plant height, stem number, leaf area index, and aboveground total dry weight. Plant height was measured from the plant base to the tip of the highest leaf. Plants were separated into green leaves and stems. Green leaf area was measured with a leaf area meter (LI-3000, LI-COR, Lincoln, NE, USA) and expressed as leaf area index. The dry weight of each component was determined after oven drying at 70°C to constant weight. Total dry weight was the sum of the weights of green leaves and stems. At maturity, 12 hills were taken diagonally from a 5 m2 area in each plot where grain yield was determined to measure the above ground total dry weight, harvest index, and yield components. Panicles of each hill were counted to determine the panicle number per m2. Plants were separated into straw and panicles. Straw dry weight was determined after oven drying at 70°C to constant weight. Panicles of all 12 hills were hand threshed and filled spikelets were separated from unfilled spikelets by submerging them in tap water. Three subsamples each of 30 g filled spikelets and 2 g unfilled spikelets were taken to determine the number of spikelets. Dry weights of rachis and filled and unfilled spikelets were measured after oven drying at 70°C to constant weight. Aboveground total dry weight was the total dry matter of straw, rachis, and filled and unfilled spikelets. Spikelets per panicle, grain filling percentage (100 × filled spikelet number/total spikelet number), and harvest index (100 × filled spikelet weight/aboveground total dry weight) were calculated. Grain yield was determined from a 5 m2 area in each plot and adjusted to the standard moisture content of 0.14 g H2O g−1 fresh weight. Grain moisture content was measured with a digital moisture tester (DMC-700, Seedburo, Chicago, IL, USA).
Data were analyzed following the analysis of variance (SAS Institute) and means were compared based on the least significant difference test (LSD) at the 0.05 probability level [
Average temperatures during the growing season in 2009 DS were 1.1–1.3°C higher than that in the 2009 WS (Figure
Daily maximum and minimum temperatures and solar radiation during rice-growing seasons at the IRRI farm in 2009 DS (a), 2009 WS (b), and 2010 DS (c).
Total rainfall of each season was 349, 1079, and 92 mm in 2009 DS, 2009 WS, and 2010 DS, respectively (Table
Rainfall and total water supply (irrigation plus rainfall) of the four rice varieties grown under two water management treatments and two N rates at IRRI farm in the three consecutive seasons.
N | Irrigation (mm) | Rainfall | Total water input (mm) | Reduction | ||
---|---|---|---|---|---|---|
AWD | CF | AWD | CF | |||
2009 DS | ||||||
LN | 517 | 606 | 349 | 866 | 955 | 9.3 |
HN | 537 | 584 | 349 | 886 | 933 | 5.0 |
2009 WS | ||||||
LN | 107 | 168 | 1079 | 1186 | 1247 | 4.9 |
HN | 102 | 172 | 1079 | 1181 | 1251 | 5.6 |
2010 DS | ||||||
LN | 753 | 1024 | 92 | 845 | 1116 | 24.3 |
HN | 729 | 998 | 92 | 821 | 1090 | 24.7 |
Data are the means across four varieties. Variety had insignificant effect on the amount of water supply.
Soil water content during the growing period under AWD in 2010 DS was shown in Figure
Change in soil water content during the growing season under AWD in 2010 DS at the IRRI farm. Data were the means across two rates and four varieties; N rate and variety had no significant effect on soil water content.
Interaction effects of variety, water management, and N rate in all the three experiments were not significant. Grain yield was not significantly different between AWD and CF across the three seasons (Table
Analysis of variance for grain yield and water use efficiency (WUE) in the three consecutive seasons at IRRI farm, Philippines.
Year | 2009 DS | 2009 WS | 2010 DS | |||
---|---|---|---|---|---|---|
Source of variation | Yield | WUE | Yield | WUE | Yield | WUE |
Water regime (W) | ns | ns | ns | ns | ns | * |
Nitrogen (N) | * | ns | ** | * | ** | * |
Variety (V) | * | * | * | ns | * | * |
W × N | ns | ns | ns | ns | ns | ns |
W × V | ns | ns | ns | ns | ns | ns |
N × V | ns | ns | ns | ns | ns | ns |
W × N × V | ns | ns | ns | ns | ns | ns |
*Significance at the 0.05 level based on analysis of variance.
**Significance at the 0.01 level based on analysis of variance.
ns: denotes nonsignificance based on analysis of variance.
Grain yield and water productivity of the four rice varieties grown under two water management treatments and two N rates at the IRRI farm for the three consecutive seasons.
Variety | Grain yield (t ha−1) | Water productivity (kg m−3) | ||||||
---|---|---|---|---|---|---|---|---|
LN | HN | LN | HN | |||||
AWD | CF | AWD | CF | AWD | CF | AWD | CF | |
2009 DS | ||||||||
IR72 | 7.45a | 7.80b | 7.74ab | 7.75b | 0.86a | 0.82ab | 0.87ab | 0.83b |
PSBRc80 | 7.52a | 7.45b | 7.67ab | 8.29ab | 0.87a | 0.78b | 0.87ab | 0.89a |
IR82372H | 7.66a | 8.16ab | 8.55a | 9.09a | 0.88a | 0.85a | 0.97a | 0.97a |
SL-8H | 7.90a | 8.23a | 7.30b | 8.48ab | 0.91a | 0.86a | 0.82b | 0.91a |
| ||||||||
Mean | 7.63 | 7.91 | 7.82 | 8.40 | 0.88 | 0.83 | 0.88 | 0.90 |
| ||||||||
2009 WS | ||||||||
IR72 | 4.96a | 5.05a | 5.6a | 5.63a | 0.42a | 0.40a | 0.47a | 0.45a |
PSBRc80 | 5.11a | 4.92a | 5.03a | 5.29a | 0.43a | 0.39a | 0.43a | 0.42a |
IR82372H | 4.73a | 4.97a | 5.30a | 5.47a | 0.40a | 0.40a | 0.45a | 0.44a |
Mestizo7 | 4.64a | 5.12a | 5.22a | 5.34a | 0.39a | 0.41a | 0.44a | 0.43a |
| ||||||||
Mean | 4.86 | 5.02 | 5.29 | 5.43 | 0.41 | 0.4 | 0.45 | 0.44 |
| ||||||||
2010 DS | ||||||||
IR72 | 7.51a | 7.69b | 8.93a | 8.74b | 0.89a | 0.69b | 1.09a | 0.80b |
PSBRc80 | 7.57a | 7.94b | 8.86a | 9.50a | 0.90a | 0.71b | 1.08a | 0.87a |
IR82372H | 7.21a | 7.97b | 8.35b | 8.74b | 0.85a | 0.71b | 1.02a | 0.80b |
SL-8H | 7.53a | 8.92a | 8.14b | 9.37a | 0.89a | 0.80a | 0.99a | 0.86a |
| ||||||||
Mean | 7.46 | 8.13 | 8.57 | 9.09 | 0.88 | 0.73 | 1.05 | 0.86 |
Data are the means across two N rates. Within a column for each season, means followed by the same letters are not significantly different according to LSD (0.05).
Nitrogen rate had a significant effect on grain yield in all the three experiments. In this study a significantdifference in water productivity between N treatments onlyin the normal dry season such as 2010 DSwas observed. In the 2010 DS, CF received 16 irrigations from transplanting to maturity, while 10 irrigations were applied to AWD. The number of irrigation was reduced in 2009 WS, when 2 and 1 irrigations were applied to CF and AWD, respectively. The differences in water productivity between AWD and CF treatments were insignificant. Water productivity in the two DS ranged from 0.78 to 1.09, which was 2.0–2.4 times higher than in 2009 WS. No significant interactions were observed in terms of variety, water management and, N rate. Varieties with higher yield had greater WUE. AWD received higher WUE than CF due to the decrease in water input. Using high nitrogen fertilization and high yield varieties were the two ways to improved water productivity in this study, as discussed by Hatfield [
The difference in grain yield between the hybrid and inbred varieties was relatively slight, except in 2010 DS. Nitrogen rate had a significant effect on grain yield in all the three experiments. Significant differences in grain weight between the AWD and CF treatments were observed in 2010 DS. Panicles per m2 and spikelets per m2 were significantly higher in high N than low N. Among the four varieties, the hybrid ones had more spikelets number per m2 compared with inbreds ones (Table
Yield components of the four rice varieties grown under two water management treatments and two N rates at the IRRI farm for the three consecutive seasons.
Spikelets panicle−1 | Panicles m2 | Grain filling (%) | Grain weight (mg) | |||||
---|---|---|---|---|---|---|---|---|
AWD | CF | AWD | CF | AWD | CF | AWD | CF | |
2009 DS | ||||||||
IR72 | 80.2c | 83.4c | 412.3a | 441.2a | 87.9a | 85.5a | 23.2d | 23.0d |
PSBRc80 | 101.7b | 104.6b | 370.3b | 363.8b | 82.8ab | 81.6b | 23.9c | 23.6c |
IR82372H | 120.0a | 117.4a | 330.2c | 344.8b | 78.3b | 77.1c | 24.4b | 24.5b |
SL-8H | 119.8a | 124.8a | 228.3d | 301.1c | 81.3b | 83.0ab | 26.8a | 27.1a |
| ||||||||
Mean | 105.4 | 107.6 | 335.3 | 362.7 | 82.6 | 81.8 | 24.6 | 24.6 |
| ||||||||
2009 WS | ||||||||
IR72 | 79.6b | 81.0b | 351.1a | 351.6a | 74.0a | 75.9ab | 21.9c | 22.1c |
PSBRc80 | 94.3a | 96.9a | 295.6b | 296.1b | 74.9a | 76.7a | 22.7b | 22.5b |
IR82372H | 98.5a | 99.6a | 307.8b | 298.7b | 66.6b | 71.5b | 23.2a | 23.4a |
Mestizo7 | 100.4a | 95.2a | 291.2b | 298.2b | 72.1a | 74.7ab | 23.4a | 23.5a |
| ||||||||
Mean | 93.2 | 93.2 | 311.4 | 311.2 | 71.9 | 74.7 | 22.8 | 22.9 |
| ||||||||
2010 DS | ||||||||
IR72 | 70.5b | 72.9c | 517.5a | 504.5a | 90.0a | 89.8a | 22.4d | 22.7d |
PSBRc80 | 97.4a | 98.8b | 424.0b | 420.1b | 81.9bc | 83.6ab | 23.0c | 23.1c |
IR82372H | 108.9a | 106.6ab | 390.3c | 407.0b | 79.0c | 82.4b | 23.5b | 23.8b |
SL-8H | 105.9a | 110.1a | 346.9d | 344.0c | 84.8b | 85.9ab | 26.2a | 26.5a |
| ||||||||
Mean | 95.7 | 97.1 | 419.7 | 418.9 | 83.9 | 85.4 | 23.8 | 24.0 |
Data are the means across two N rates. Within a column for each site, means followed by the same letters are not significantly different according to LSD (0.05).
The LAI at flowering was significantly higher in high N than low N in the two DS. LAI in hybrids was higher than in inbreds in 2009 WS and 2009 DS (Table
Growth duration, leaf area index (LAI) at flowering, harvest index, and total dry weight of the four rice varieties grown under two water management treatments and two N rates at the IRRI farm for the three consecutive seasons.
Growth duration (days) | LAI at flowering |
Total dry weight |
Harvest index | |||||
---|---|---|---|---|---|---|---|---|
AWD | CF | AWD | CF | AWD | CF | AWD | CF | |
2009 DS | ||||||||
IR72 | 104 | 104 | 5.49b | 5.35c | 1481a | 1557a | 45.5b | 46.4c |
PSBRc80 | 106 | 104 | 5.56b | 5.58bc | 1481a | 1459b | 50.1a | 50.1b |
IR82372H | 100 | 100 | 6.63a | 6.26ab | 1441b | 1464b | 50.6a | 52.2a |
SL-8H | 106 | 106 | 6.82a | 6.71a | 1478a | 1609a | 52.1a | 52.4a |
| ||||||||
Mean | 104 | 104 | 6.13 | 5.98 | 1470 | 1522 | 49.6 | 50.3 |
| ||||||||
2009 WS | ||||||||
IR72 | 101 | 102 | 3.16b | 3.31a | 1089a | 1117a | 41.7b | 42.7c |
PSBRc80 | 103 | 103 | 3.30ab | 3.53a | 1078a | 1119a | 43.8b | 44.3bc |
IR82372H | 102 | 102 | 3.56a | 3.50a | 1078a | 1105a | 43.5b | 44.9b |
Mestizo7 | 100 | 100 | 3.55a | 3.71a | 1045b | 1055a | 47.2a | 47.1a |
| ||||||||
Mean | 102 | 102 | 3.39 | 3.51 | 1073 | 1099 | 44.1 | 44.8 |
| ||||||||
2010 DS | ||||||||
IR72 | 105 | 106 | 4.70a | 5.15a | 1545b | 1560a | 47.7b | 48.1b |
PSBRc80 | 107 | 107 | 4.63a | 5.40a | 1594b | 1578a | 48.6ab | 50.7ab |
IR82372H | 100 | 100 | 4.72a | 5.05a | 1554b | 1631a | 50.4a | 50.0a |
SL-8H | 111 | 111 | 5.04a | 5.01a | 1658a | 1667a | 49.1ab | 51.7ab |
| ||||||||
Mean | 106 | 106 | 4.77 | 5.15 | 1588 | 1609 | 49.0 | 50.1 |
Data are the means across two N rates. Within a column for each season, means followed by the same letters are not significantly different according to LSD (0.05).
SPAD values were significantly different between AWD and CF in three varieties at the grain filling stage in 2010 DS (Figure
SPAD values after flowering under AWD and CF in 2010 DS at the IRRI farm. Three varieties IR72 (a), IR82372H (b), and SL-8H (c) were used in the experiment at high N level.
Grain filling rates (mg · grain−1 · day−1) were significantly different between AWD and CF in all the three varieties in grain filling stage (Figures
Grain weight after flowering under AWD and CF in 2010 DS. Three varieties IR72 (a), IR82372H (b), and SL-8H (c) were used in the experiment at high N level.
Grain filling rate after flowering under AWD and CF in 2010 DS. Three varieties IR72 (a), IR82372H (b), and SL-8H (c) were used in the experiment at high N level.
In conclusion, interaction effects among variety, water management, and N rate were not significant under tropical condition. Grain yield was not significantly different between AWD and CF in all the seasons through saving water input. Using high nitrogen fertilization and high yield varieties were the two ways to improve water productivity in this study; severe water stress during late grain filling stage accelerated grain filling rate, shortened the grain filling period, and enhanced whole plant senescence, thus reducing grain weight. The study indicated that proper water management greatly contributed to grain yield in the late grain filling stage, and it was critical for safe AWD technology.
National Basic Research Program of China (Project no. 2009CB118603) is acknowledged for funding the doctoral studies of the first author, and the Ministry of Science and Technology in China (Project nos. 2011BAD16B14 and 2012BAD04B00) is acknowledged for their financial support. Alex and Eddie are acknowledged for the excellent crop management.