Effective weed control in no-till systems is mainly dependent on herbicides. Appropriate weed management strategies, however, can reduce the amount of herbicides. In this study possible weed management was attempted in wheat ( Triticum aestivum L. emend. Fiori & Paol.) under zero-tillage (without residue) and conventional tillage between 2006-07 and 2007-08 following residual effects of 13 weed control treatments with metribuzin applied during previous rainy ( kharif) seasons, 2006 and 2007 in soybean [ Glycine max (L) Mirrill]. The experiment was laid out in a split plot design with three replications. Total density of weeds did not differ significantly between weed control measures adopted during previous rainy season. But, conventional tillage (CT) caused a significant reduction in the populations of grass, broad-leaved (e.g., Chenopodium album L. and Melilotus indica L.) and total weeds when compared with zero-tillage (ZT) without crop residues. Similarly, higher nutrient removal by weeds was under zero-tillage than under conventional tillage. In general, residual effect of all the previous season metribuzin treatments was more pronounced on the grasses ( Phalaris minor Retz. and Avena sterilis ssp. ludoviciana Dur.) when compared with Chenopodium album and Melilotus indica. Wheat yield was similar in all previous season metribuzin treatments except in weed-free check in both years. However, metribuzin at 0.5 kg/ha pre-emergence, on yield, was more effective, indicating its considerable residual effect on weeds in wheat. This treatment also gave the highest net benefit:cost ratio.
