This work was carried out to evaluate the residual activity of isoxaflutole (IFT) in Brachiaria decumbens control after different simulated drought conditions following herbicide application in two soils. Six simultaneous experiments were performed, using rates of 0, 230 and 270 g ha-1 of IFT in samples of a heavy clay soil and of 0, 180 and 200 g ha-1 in samples of a sandy clay loam soil. In each experiment, rates of IFT were combined in a factorial scheme with drought periods after herbicide application. B. decumbens was used as bioindicator of the residual activity of IFT in soil. Visual control of bioindicator was evaluated at 15, 30, 45 and 60 days after sowing. IFT was very stable when applied to the heavy clay soil, even after three 20 mm simulated rain, 30 days apart from each other, and followed by 120 days of drought conditions. Stability was shorter for the sandy clay loam soil and residual activity (>80% control) ranged from 0 to 110 days for B. decumbens, depending on rate, drought periods and evaluation date. Results indicate that as period of time and number of irrigation events between herbicide application and bioindicator sowing increased, residual control of IFT decreased on the sandy clay loam soil.
bioindicator; herbicides; residues
