Abstract
During two successive rabi seasons 2019-20 and 2020-21, a field experiment was undertaken at the Regional Station of Bangladesh Wheat and Maize Research Institute (BWMRI), Gazipur to establish the optimum amount of herbicide application (Gramoxone) before sowing wheat under late-planted conditions. Four optimum doses of application were applied as pre-emergence. Herbicide (Gramoxone) spraying at 6 liter/ha for 5 days before sowing provided the best weed control. The experiment was laid out in a randomized complete block design (RCBD) with 3 replications. The major weeds in the experimental plot such as Biskatali (Polygonium hydropiper), Chapra (Elusine indica), Bathua (Chenopodium album), Banpalong (Sanchus arvensis) and Banmasure (Vicia sativa) were found. Weed count was 92 m2 and 89 m2 in 2019-20 and 2020-20 respectively, when herbicide (Gramoxone) was sprayed @ 12 liter/ha at 5 days before sowing (DBS). Highest weed monitoring efficiency (78%) in T2=herbicide (Gramoxone) treatment @ 6 liter/ha at 5 DBS. Grain yield and wheat biological yield were statistically diverse, while the other metrics were insignificant. When herbicide (Gramoxone) was treated at 6 liter/ha at 5 DBS, a numerically greater average grain yield (2.57 t/ha) was obtained. The control plot produced a numerically lower average grain yield (1.56 t/ha). When herbicide (Gramoxone) was administered at 6 liter/ha at 5 DBS, the maximum benefit-cost ratio (BCR) was 1.30 which is the best method for weed management prior to wheat application.