Yield loss prediction for integrated weed management in direct-seeded rice.
In field trials conducted at CIAT during 1989-1990, the critical period of crop-weed competition was evaluated for rice cv. Oryzica 1 and CICA 8. Weed species present were Eleusine indica, Echinochloa crus-galli, Echinochloa colonum, Cyperus difformis, Cyperus esculentus, Cyperus iria, Leptochloa filiformis and Eclipta alba. Competition effects were similar for both varieties. Weeds emerging with the crop were the most damaging ones. Rice yields increased with longer weed-free periods up to 70 d after emergence (DAE). Weed emergence and growth after 70 DAE was suppressed by the crop. Both rice cultivars required a critical height of 44-50 cm to suppress weed growth at 70 DAE. Both varieties were equally weed suppressive by 70 DAE, but the leaf area index of CICA 8 was much greater than that of Oryzica 1 when rice grew weed-free after emergence. Economic analysis showed that under heavy weed infestations, in a system where rice cannot be early and continuously flooded, 3 post-em. herbicide applications (propanil + butachlor + pendimethalin at 1 + 1.2 + 1.3 kg/ha applied at 9 DAE, thiobencarb + propanil + picloram + 2,4-D at 2.2 + 2 + 3.3 + 0.09 kg at 18 DAE, and fenoxaprop-ethyl + bentazone at 0.1 + 0.5 kg at 44 DAE followed by spot applications of paraquat at 0.2 kg) were justified.