Résumé
Weeds cause substantive yield losses in rainfed rice, particularly direct-seeded rice (DSR). Two field experiments were conducted in 2005 in north-east Thailand to examine the effects of cutting date and timing of fertilizer application on suppressed weed, growth and yield of DSR. Treatments consisted of two fertilizer applications to the main plots (top-dressing at the date of cutting and 15 days after cutting; DAC) and four cutting dates of the sub-plots. Rice and weed cutting was performed at 30, 45, 60 and 75 days after seeding (DAS), compared with an untreated control. Cutting resulted in a higher growth rate and improved grain yield of rice compared to the untreated control. Weed cutting at 60 DAS resulted in the greatest leaf area index, which contributed to the maximum number of tiller as well as maximum plant dry mass when observed at 30 DAC. Lower densities and dry weights of weeds at 30 DAC were observed when cutting was done 60 DAS, resulting in a higher grain yield than the other cutting dates. Timing of fertilizer application had no significant effect on the density and dry weight of grasses and broad-leaved weeds, but did have a significant affected on, resulting in the lowest density of sedges at 45 DAC was observed with fertilizer applied at cutting date. Rice grain yield was higher with top-dressing of fertilizer at 15 DAC than with fertilizer applied at cutting date. The cutting treatments resulted in higher net incomes than the untreated control. These results indicate that rice and weed cutting at an appropriate growth stage is an alternative method to chemical weed control for DSR in rainfed riceland.