Improving management of rice in semi-arid eastern Indonesia: responses to irrigation, plant type and nitrogen

Abstract

Summary. A number of field experiments were undertaken in eastern Indonesia with the aim of improving rice production in this semi-arid region. The objectives of these studies were to examine the effects of irrigation method (raised beds under saturated soil culture v. flooded system), irrigation frequency (daily v. twice weekly) and genotype (traditional v. improved) on rice yield and components of yield, and to examine the response of rice growth on raised beds to sowing time and nitrogen fertilisation. Recent studies in northern Australia have demonstrated that rice can successfully be grown under saturated soil culture. In the Australian studies, grain yield and quality were maintained, yet saturated soil culture used 32% less water than the flooded control in both wet and dry seasons. Higher efficiencies of water use for rice production with saturated soil culture in semi-arid tropical Australia suggest that similar benefits may be realised with this method of irrigation in West Timor. The experiments in West Timor were undertaken within a low-external-input system, and all experiments were affected by drought. The central issue is one of aligning crop growth with water availability to ensure adequate quantity and quality of grain production at the end of the season. On this basis, a number of practical strategies for improving rice production under water-limited conditions in West Timor are suggested. First, time of sowing in the wet season is important, with early-sown crops escaping end-of-season drought. Significantly, the improved genotype (cv. Lemont) was only able to fill its grain adequately if sown early in the wet season, thereby avoiding drought during grain filling. Second, providing soils are sufficiently deep, rice can successfully be grown under saturated soil culture in West Timor. Importantly, preparation of raised beds before the wet season enables rice crops to be sown early, maximising the use of rainfall for crop production. Twice weekly irrigation of rice on beds was found to be more efficient than irrigating daily or flooding the bays. Third, no differences in grain yield were found between the improved short-statured genotype (cv. Lemont) and the taller traditional genotype under the low-external-input system, although differences in components of yield were observed in the wet and dry seasons. There is some evidence that the traditional genotype filled grain better when water was limiting during grain growth by restricting vegetative production and enabling the crop to finish.