The objective of this study was to evaluate the efficiency and economic viability of controlling weeds with residual herbicides, applied in pre-planting of common beans. The study was conducted in Santo Antônio de Goiás, Brazil, in 1999/2000 on three t ha-1 of dry biomass. The 60 treatments were applied on the top of the weed plants (Commelina benghalensis and Bidens pilosa) disposed as a 2x10x3 factorial, with split-plots in a randomized complete block design with three replicates. The first factor consisted of the desiccation systems in the area (Integrated Control System (ICS) and Apply and Sow). ICS consisted of sulfosate (720 g ha-1) applied 20 days before planting and paraquat 200 g ha-1 added to the residual herbicides and applied immediately after sowing. Apply and Sow consisted of (sulfosate 720 g ha-1) added to the residual herbicides, applied five days before sowing. The second factor consisted of the residual herbicides applied in g ha-1, sulfentrazone (200 and 300), dimethenamid (900 and 1.125); clomazone (360); pendimethalin (2.500), s-metolachlor (768 and 1152), diclosulan (12,45) and check; and the third factor consisted of the doses of the post-emergence herbicides: imazamox (15) + bentazon (240), imazamox (30) + bentazon (480) and the check. The results showed that the residual herbicides applied in the ICS and Apply and Plant systems reduced the initial growth of Bidens pilosa, but only the herbicides diclosulan and sulfentrazone showed the same effect on Euphorbia heterophylla. The use of residual herbicides in pre-planting was economically viable reducing the amount of herbicides applied in pre-emergence, except for dimethenamid and s-metolachlor.
Phaseolus vulgaris; herbicides; pre emergence; post-emergence
