Yield constraint analysis of rainfed lowland rice in Souteast Asia

<b>Keywords:</b> Fertilizer application, field hydrology, plant nutrient uptake, toposequence, weed control, yield loss. Rainfed lowland rice yields are low and unstable due to uncertain water supply, low soil fertility, and pest infestation. To design management interventions aimed at increasing rainfed rice production, the magnitude of and variation in yield gaps associated with various constraining factors need to be assessed. This study aimed at improved understanding of spatial and temporal variations in water status and nutrient availabilities, and their effects on crop growth and yield for rainfed rice along toposequences in SE Asia, as a basis for the design of improved crop management strategies. To realize these objectives, statistical and systems analyses were used to quantify the effects of yield-determining, yield-limiting, and yield-reducing factors and to identify management options along toposequences in sloping rainfed rice areas in Indonesia and Thailand. Field experiments were conducted from 1995–2000 at Jakenan Experiment Station, Indonesia and from 2000–2002 in farmers’ fields in rainfed rice areas in Indonesia and Thailand. These areas have similar rainfall levels and represent the common single rainfed rice system (in Thailand) and the (one) more intensive double rainfed rice system (in Indonesia) The ORYZA2000 crop growth simulation model adequately simulated rice yields at various levels of fertilizer N under irrigated and rainfed conditions in the study area in Indonesia. Simulated potential yields ranged from 3.74 (for crops sown in October) to 5.53 Mg ha⁻¹ (for crops sown in July). Groundwater depth had a significant impact on the effects of establishment date, deep tillage, fertilizer application, and supplementary irrigation performance of on rainfed rice sown towards the end of the wet season. In farmers’ fields in Indonesia and Thailand, substantial differences were recorded in field hydrology, exchangeable K, organic C and clay content among different toposequence positions. Differences were also observed in yield and in the magnitude of yield increase due to intensive weed control and/or application of recommended fertilizer doses, but these differences were not consistent across countries, seasons, and years. In Indonesia, the yield constraints of rainfed lowland rice in the research station and in farmers’ fields were identified based on analyses of yield gaps associated with the interactive effects of water and nutrient stress, taking into account the incidence of rice pests. There were consistently wide yield gaps due to water, N and K limitations, and pest infestations. These results indicate substantial scope for increasing rainfed rice yields through timely establishment, deep tillage, supplementary irrigation, application of N and K fertilizers and pest control. Combining various methodologies in estimating reference yields provided increased understanding of the constraints to rainfed lowland rice production and this set of tools could be applied to other locations to identify yield constraints and management options to increase production, provided that the necessary experimental data required for such a comprehensive analysis are available