The selectivity and agronomic efficiency of clethodim and sethoxydim herbicides on D. horizontalis control in watermelon crop was evaluated. The experiment was carried out at Floresta, Paraná State, Brazil, in 1995/96. The watermelon cultivar used was Crimson Sweet and the experimental design was a randomized block with four replicates. The treatments evaluated were clethodim (240 g/L) in the dosage of 84; 96, 108 g i.a./ha with 1.0 L/ha of mineral oil (Dytrol); sethoxydim (184 g/L) in the dosage of 230 g i.a./ha with 1.0 L/ha of mineral oil (Assist); free of weeds and weedy. The herbicides were applied in post-emergency, when the watermelon had branches up to 1.4 m long and the D. horizontalis presented three leaves and three tillers, and density of 240 plants/m². The injury evaluations (E.W.R.C.), control (0-100%) and weed control. were performed at 7. 15 and 30 days after application of treatments. The clethodim herbicide, in the dosage of 96 g i.a./ha or higher, controlled efficiently the D. horizontalis (100), being more efficient than sethoxydim in controlling weeds. None of chemical treatments affected the watermelon development and production. Weedy plots with D. horizontalis interfered in the culture development, resulting in a 36% reduction on watermelon production.
Digitaria horizontalis; watermelon; clethodim; sethoxydim
