Effect of glyphosate drift on marandu grass

Glyphosate drift in plants that are not resistant to the herbicide molecule can result in stimulation to certain biological features, characterizing the phenomenon of hormesis. On this basis, productive and chemical traits were evaluated in marandu grass, in a simulation of the drift effect, using sublethal doses of the herbicide glyphosate. The experiment was in laid out in a randomized-block design with split plots in time, in four replicates. The effect of sublethal doses of glyphosate acid equivalent (a.e.) (21.60, 43.20, 64.80, 86.40 and 108.00 g ha(-1)) and control was evaluated in the plots; and the effect of harvesting at 92, 113, 134 and 155 days after sowing (DAS) the grass was evaluated in the subplots. The Urochloa brizantha cv. Marandu was collected at a height of 0.20 m, at a defoliation interval of 21 days, to estimate production and chemical traits, in experimental plots with a usable area of 7.5 m(2). Leaf: stem ratio was influenced by the interaction between the evaluated factors (dose and harvest). Dose fitted a second-order polynomial model, with a hormesis effect of 21.60 to 76.50 g ha(-1) of glyphosate a.e. Harvesting at 134 DAS differed significantly from 92 DAS. The forage yield showed a linear response inversely proportional to the increasing glyphosate doses. There was a polynomial increase in leaf phosphorus content, characterizing hormesis up to the sublethal dose of 72.50 g a.e. ha(-1). Harvest influenced the neutral detergent fiber, acid detergent fiber, lignin and leaf phosphorus contents.