Influence of Water Management on Growth and Yield of No‐till Planted Rice

To determine the feasibility of producing rice ( Oryza sativa L.) on lands equipped with sprinkler‐irrigation systems, the productivity of flood‐irrigated rice was compared with that of rice grown under two levels of sprinkler irrigation on a Mhoon silty clay (fine mont‐morillonitic, nonacid, thermic, Vertic Haplaquept) from 1984 to 1986. Two high‐quality cultivars, ‘Labelle’ and ‘Lemont,’ were compared with ‘CICA‐8,’ one of the most productive cultivars on favored upland sites in Latin America. Soil matric potential was monitored daily with tensiometers placed at 15‐ and 30‐cm soil depths. Beginning 30 d after no‐till drill planting into a killed legume cover crop, rice was flood‐ or sprinkler‐irrigated when soil matric potential at either depth fell below −20 or −50 kPa. Each cultivar matured later and yielded less, as a result of reduced numbers of seed per panicle and reduced seed weight, when grown under sprinkler irrigation compared with flooded culture. In years without severe disease, Lemont and CICA‐8 yields were reduced 10 to 25% from the 5000 to 6700 kg ha −1 produced under flood, while Labelle yield was reduced 27 to 56%. During 1986, rotten‐neck blast (caused by Pyricularia oryzae Cav.) reduced yields of all cultivars grown under sprinkler‐irrigated conditions by 60 to 80%. Biomass production and yield differed between the two levels of sprinkler irrigation only when precipitation plus irrigation was less than potential evapotranspiration (PE). It is concluded that no level of sprinkler irrigation can completely eliminate water stress in upland rice when PE is high.