Improving water productivity in moisture-limited rice-based cropping systems through incorporation of maize and mungbean: A modelling approach.

Crop and water productivities of rice-based cropping systems and cropping patterns in the irrigated lowlands of Sri Lanka have not been researched to the degree warranted given their significance as critical food sources. In order to reduce this knowledge gap, we simulated the water requirement for rice, maize, and mungbean under rice-based cropping systems in the Dry Zone of Sri Lanka. We evaluated the best combinations of crops for minimum water usage while reaching higher crop and water productivities. We also assessed the risk of cultivating mungbean as the third season/sandwich crop (i.e. rice-mungbean-rice) in different regions in Sri Lanka. In the simulation modelling exercise, APSIM-Oryza (rice), APSIM-maize and APSIM-mungbean modules were parameterised and validated for varieties grown widely in Sri Lanka. Moreover, crop productivities and supplementary irrigation requirement were tested under two management scenarios i.e. Scenario 1 : irrigate when plant available water content in soil fell below 25% of maximum, and Scenario 2: irrigate at 7-day intervals (current farmer practice). The parameterised, calibrated and validated model estimated the irrigation water requirement (number of pairs of observations (n) = 14, R 2 > 0.9, RMSE = 66 mm season −1 ha −1 ), and grain yield of maize (n = 37, R 2 > 0.95, RMSE = 353 kg ha −1 ) and mungbean (n = 26, R 2 > 0.98, RMSE = 75 kg ha −1 ) with a strong fit in comparison with observed data, across years, cultivating seasons, regions, management conditions and varieties. Simulated water requirement during the cropping season reduced in the order of rice (1180–1520 mm) > maize and mungbean intercrop = maize sole crop (637–672 mm) > mungbean sole crop (345 mm). The water productivity of the system (crop yield per unit water) could be increased by over 65% when maize or mungbean extent was increased. The most efficient crop combinations to maximise net return were diversification of the land extent as (i) 50% to rice and 50% to mungbean sole crops, or (ii) 25%, 25% and 50% to rice, maize and mungbean sole crops, respectively. Under situations where water availability is inadequate for rice, land extent could be cultivated to 50% maize and 50% mungbean as sole crops to ensure the maximum net return per unit irrigation water (115 Sri Lankan Rupees ha −1 mm −1 ). Regions with high rainfall during the preceding rice cultivating season are expected to have minimum risk when incorporating a third season mungbean crop. Moisture loss through evapotranspiration from the third season mungbean crop was similar to that of a fallowed site with weeds. [ABSTRACT FROM AUTHOR]