Abstract
The crop management in direct-seeded rice to promote growth during the reproductive stage was evaluated in weed-controlled toposequentially intermediate fields in a small watershed in Northeast Thailand. In 2004, the effectiveness of topdressing with 21 kg ha-1 of nitrogen at the panicle initiation stage were examined using two genotypes, KDML105 and IR57514-PMI-5-B-1-2 (IR57514) seeded at the rates of 500, 250 and 125 seeds m-2. In 2005 and 2006, the effectiveness of a new management (seeding rate of 125 seeds m-2 and nitrogen application of 90-101 kg ha-1; CM2) was compared with that of conventional management (seeding rate of 500 seeds m-2 and nitrogen application of 50 kg ha-1; CM1) using 3 genotypes (KDML105, IR57514 and HY71) seeded in May and June. In 2004, the number of spikelets on the tertiary pedicel at a low seeding rate in KDML105 was greatly increased by topdressing. In 2005 and 2006, CM2 had higher grain yield than CM1 (346 vs. 235 g m-2), owing to its larger spikelet number per panicle, heavier shoot dry weight and greater nitrogen uptake. May-seeding resulted in longer non-flooded period in the seedling to tillering stage, lower SPAD reading value around heading and less shoot dry weight increase from heading to maturity, and had lower grain yield than June-seeding (253 vs. 328 g m-2). This reduction in grain yield was larger for late-heading KDML105 than in early heading IR57514. These results indicated the effectiveness of the new crop management (CM2) for direct-seeded rice in toposequentially intermediate fields with less weed infestation or weed-controlled conditions.
