Time and mode of nitrogen fertilizer application to tropical wetland rice

Abstract

In experiments with transplanted rice (Oryza sativa L.) at the International Rice Research Institute, Philippines, two methods of split application of urea and ammonium sulfate were compared with deep, point placement (10 cm) of urea supergranules and broadcast application of a slow-release fertilizer sulfur-coated urea (SCU). Comparisons were made in the wet and dry seasons and were based on rice yield and N uptake. Urea- and ammonium-N concentrations and pH of the floodwater were measured to aid interpretation of the results. Split applications of urea were generally less efficient than ammonium sulfate. The split in which the initial fertilizer dose was broadcast and incorporated into the soil before transplanting was more effective than the split in which the fertilizer was broadcast directly into the floodwater 21 days after transplanting. Both split applications were inferior to the urea supergranules and SCU, in terms of both yield and N uptake efficiency; average apparent N recoveries ranged from 30% for the delayed split urea to 80% for the urea supergranule. Broadcast applications of urea and ammonium sulfate produced high floodwater concentrations of urea- and ammonium-N, which fell to zero within 4–5 days. Floodwater pH was as high as 9.3 and fluctuated diurnally due to heavy algal growth. Ammonia volatilization and algal immobilization of N in the floodwater were probably responsible for the poor efficiency of the split applications; the supergranules and SCU on the other hand produced low floodwater N concentrations and were efficiently used by the rice crop.