Selectivity of herbicides to corn and braquiaria grass Intercropping in a crop-Livestock Integration System

The objective of this work was to evaluate the selectivity of different herbicides in the crop of corn and Brachiaria ruziziensis at crop-livestock integration. The experiment was carried out in Nova Xavantina, MT, between January and June of 2008, in soil classified like Rhodic Ferrasol. The trial was installed ar complete randomized block designing, with 11 treatments: lactofen (96 g ha(-1)); lactofen (192 g ha(-1)); carfentrazone - ethyl (8 g ha(-1)); carfentrazone - ethyl (12 g ha(-1)); flumioxazin (25 g ha(-1)); flumioxazin (50 g ha(-1)); lactofen + flumioxazin (96 + 25 g ha(-1)); lactofen + flumioxazin (192 + 50 g ha(-1)); nicosulfuron (8 g ha(-1)); nicosulfuron (40 g ha(-1)) and a check without herbicide. The treatments were applied at 30 days after the corn emergence. It were evaluated the effects of the treatments in the height, dry biomass and productivity of the corn and, also, in the dry biomass of the B. ruziziensis. The herbicides flumioxazin, lactofen and carfentrazone-ethyl didn't present potential to be used in post-emergence in this integrated system, due to low selectivity for corn. Nicosulfuron atrate of 8 a g ha(-1) provided corn yeld similar to the check with out herbicides, and a initial small suppression of B. ruziziensis; and, consequently, being the best herbicide to be used for weed management in the no-livestock system.