Abstract
Direct wet seeding of rice (DWSR) and alternate wetting and drying (AWD) system of water management became a proven resource-conservation (RC) technology for rice production. Optimum doses of nitrogen (N) and weed management (WM) practice in the RC technology have not been adequately addressed. The present investigation aimed to determine optimum doses of N and weed management strategies in DWSR under AWD irrigation system. The study involved six field experiments at the Bangladesh Rice Research Institute, Gazipur during dry (Boro) seasons of three consecutive years of 2009-10, 2010-11 and 2011-12. The experiments separately compared six rates of nitrogen 0 - 200 kg ha-1 with a 40 kg ha-1 interval, seven weed management options and four nitrogen management practices. All the experiments used BRRI dhan29 as a test crop and grown as direct wet seeded crop followed by AWD water management. Experiment with different rates of N revealed the economic optimum rate of N as 166 and 155 kg ha -1 in first and second year, respectively, with the corresponding yield of 7.1 and 6.5 t ha1. At the N rate of 160 kg ha-1, average agronomic N use efficiency, physiological N use efficiency and i ecovery of N appeared as 25 kg kg-1, 97 kg kg-1 and 43% respectively. Gross return over fertilizer (GRF) reached the highest with 160 kg N ha-1 (Tk. 692 in 2009-10 and Tk.489 in 2010-11). Another experiment compared different methods of N management including leaf-color chart based real time N management (RTNM) and fixed time N management (FTNM), prilled urea (PU) and urea super granule (USG) with an N-control treatment. The application of N according to different methods gave similar rice yield in both the years, however, leaf color chart (LCC) based N application involved lower N doses than PU and USG. The FTNM and RTNM gave higher N use efficiencies than PU and USG. Nitrogen management following FTNM (142.5 kg N ha-1) was more profitable which gave the highest GRF (Tk. 556 in 2010-11 and Tk. 609 in 2011-12). In weed management experiment, Scirpus juncoides, Echinochloa crus-galli and Cynodon dactylon appeared as the most important weed species, whereas Fimbristylis miliaceae, Cyperus iria and Lindernia floribunda appeared in the least important group. Weeds interfere up to 55 days after seeding (DAS) produced significant impact on rice growth and yield. With the advancement of time weed dominance ranking had changed.The application of mefenacet+bensulfuron methyl (M+BM) along with 1HW and pyrazosulfuron ethyl (PE) with 1HW effectively reduced weed biomass and number which lead to achieved higher weed control efficiency. These two weed management treatments reduced weed persistence index (WPI), weed management index (WMI) and agronomic management index (AMI) and increased crop resistance index (CRI) at 30, 55 and 80 DAS. Weed index showed about 100% variation in grain yield in WSR grown under AWD conditions. The M+ BM and PE with 1 HW gave the highest LAI, CGR, LAD and dry matter. M+BM with 1HW and PE with 1HW eventually produced significantly higher number of panicles m2, grains panicle-1, panicle length and higher grain and straw yields than these other weed management treatments. The treatment PE with HW received the highest MRR (Tk. 4663 in 2009-10 and 5426 in 2010-11) which was more profitable than that of other treatments. The results suggest that BRRI dhan29 can be successfully established in direct wet seeded condition under AWD irrigation system. The crop should receive a 155-166 kg N ha-1, but the N rate could be reduced by about 10% with the use of LCC for FTNM approach. The great challenge of weed management for DWSR could be overcome by applying either M+MB or PE followed by only one hand weeding.