An experiment was carried out to evaluate the selectivity of oxyfluorfen for physic nut culture. The experiment was arranged in a completely randomized design with five replications in a factorial 2 x 2 + 1, two herbicide doses (600 and 1,200 g ha-1 of active ingredient), two modes of application (on the plant and soil) and a control without herbicide application.. After transplanting the seedlings into plastic pots with 10 liters of soil, the treatments were applied using a bucket. When applied on the plants, oxyfluorfen was shown to be toxic in the doses tested, with more intense symptoms in the first days after treatment. After this period, the seedlings recovered strength and new shoots came out showing no symptoms of toxicity. These symptoms consisted of whitish spots on the leaves progressing to necrosis. In plants where the herbicide was applied in the soil, no visual symptoms of toxicity were observed. There was a reduction in fresh and dry weight of the leaves when 600 g ha-1 of the herbicide was applied on the plant or 1,200 g ha-1 was applied on the soil. However, leaf area and number of leaves did not differ significantly between the doses used and modes of application. Of the four weeds sown (Brachiaria decumbens, Brachiaria plantaginea, Sida rhombifolia and Bidens pilosa) only Bidens pilosa was not controlled by oxyfluorfen applied in pre-emergence.
chemical control; herbicide; Jatropha curcas; toxicity
