Evaluation and application of ORYZA2000 for irrigation scheduling of puddled transplanted rice in north west India

Abstract

Water-saving technologies that increase water productivity of rice are urgently needed to help farmers to cope with irrigation water scarcity. This study tested the ability of the ORYZA2000 model to simulate the effects of water management on rice growth, yield, water productivity (WP), components of the water balance, and soil water dynamics in north-west India. The model performed well as indicated by good agreement between simulated and measured values of grain yield, biomass, LAI, water balance components and soil water tension, for irrigation thresholds ranging from continuous flooding (CF) to 70 kPa soil water tension. Using weather data for 40 different rice seasons (1970–2009) at Ludhiana in Punjab, India, the model predicted that there is always some yield penalty when moving from CF to alternate wetting and drying (AWD). With an irrigation threshold of 10 kPa, the average yield penalty was 0.8 t ha −1 (9%) compared with CF, with 65% irrigation water saving, which increased to 79% at 70 kPa with a yield penalty of 25%. The irrigation water saving was primarily due to less drainage beyond the root zone with AWD compared to CF, with only a small reduction in evapotranspiration (ET) (mean 60 mm). There were tradeoffs between yield, irrigation amount and various measures of WP. While yield was maximum with CF, water productivity with respect to ET (WP ET) was maximum (1.7 g kg −1) for irrigation thresholds of 0 (CF) to 20 kPa, and irrigation water productivity (WP I) increased to a maximum plateau (1.3 g kg −1) at thresholds ≥30 kPa. Because of the possibility of plant stress at critical stages known to be sensitive to water deficit (panicle initiation (PI) and flowering (FL)), treatments with additional irrigations were superimposed for 2 weeks at one or both of these stages within the 10, 20 and 30 kPa AWD treatments. Ponding for two weeks at FL was more effective in reducing the yield penalty with AWD than ponding at PI, but the biggest improvement was with ponding at both stages. This reduced the average yield loss from 9% (0.8 t ha −1) to 5% (0.5 t ha −1) for AWD with thresholds of 10 and 20 kPa. However, maximum WP I (1.1 g kg −1) was achieved with an irrigation threshold of 20 kPa combined with more frequent irrigation at FL only, but with a greater yield penalty (8%). Thus the optimum irrigation schedule depends on whether the objective is to maximise yield, WP ET or WP I, which depends on whether land or water are most limiting. Furthermore, the optimum irrigation schedule to meet the short term needs of individual farmers may differ from that needed for sustainable water resource management.