Determinação da curva dose-resposta e atividade residual de herbicidas aplicados em pré-emergência utilizados na cultura do algodoeiro para o controle de Amaranthus e Portulaca oleracea

Abstract

The low density of the plants in the stand, the growing in wide rows and the slow growth of the cotton crop cause the delay of soil cover, creating the necessity of weed control in different moments of the crop. In the Brazilian 'cerrado', which is a great cotton producer, farmers have observed an increase of weed infestation with Amaranthus spp. and Portulaca oleracea, and found difficulties controlling such weed species. The aim of this work was to improve the recommendation of the main herbicides alternatives used in preemergence to control P. oleracea and five different species of Amaranthus in cotton crop. The work was developed in two stages, both carried out in greenhouse: the first was aimed at establishing the herbicides rates which could promote efficient control of Portulaca oleracea, Amaranthus viridis, A. hybridus, A. spinosus, A. Lividus through the rate-response curve; the second was aimed at verifying the period of residual activity provided by the herbicides rates considered efficient and selected in the first stage, and by recommended rates. In the first stage, increasing herbicides rates of alachlor, diuron, clomazone, oxyfluorfen, pendimethalin, s-metolachlor, trifluralin 600, trifluralin 450 and prometryne were used. We verified that there were differences of susceptibility among species in relation to all herbicides through the calculus of sufficient rate for 80% (C80) or 95% (C95). A. lividus was the least sensitive species to herbicides and A. Spinosus was the most sensitive. We observed that efficient rates are much inferior to those used commercially, indicating the possibility of better adequacy of recommended rates. Trifluralin 600, trifluralin 450 and clomazone did not demonstrate the same efficiency of the other herbicides in the rate level used for these herbicides. In the second stage, herbicides were applied at four different periods prior to weed sowing (30, 20, 10 and 0 days before sowing - DAS) using C95 rate selected in the first stage, and the recommended rate with the exception of clomazone, trifluralin 600 and trifluralin 450, for which only the recommended rate was used. Residual activity of alachlor, oxyfluorfen and prometryne for C95 rate, clomazone, trifluralin 600 and trifluralin 450 at the only rate applied (recommended) was reduced as the period of time between application and weed sowing was increased. The application of the com ercial rate of oxyfluorfen prolonged the residual activity of control of the species A. hybridus and A. spinosus, which had not been efficiently controlled by the C95 rate, promoting then the effective control of the species until 30 DAS. For alachlor, the increase of the C95 rate to the recommended one did not reflect a considerable increase in residual activity for all species, except to A. viridis. The recommended rate of Prometryne did not promote efficient control only in relation to A. hybridus. Diuron, pendimethalin and s-metolachlor showed effective residual activity for all species until 30 DAS at both rates used, demonstrating consistent residual activity to the soil with sandy clay loam texture (21% clay and 13,68 g dm-³ carbon).