The objective of this work was to evaluate the performance of residual herbicides applied pre planting in different vegetal coverings under no-till system before dry bean crop establishment in 1999 and 2000. Each herbicide was applied separately in five different coverings, totaling twenty experiments. The herbicides and respective dosages were: diclosulan (0.0125 kg ha¹), sulfentrazone (0.300 kg ha-1), dimethenamid (1.125 kg ha-1) and S-metolachlor (1.152 kg ha¹). In Santo Antônio de Goiás, Brazil, Pennisetum glaucum dry plants and dry weeds (Commelina benghalensis and Bidens pilosa) were used under the integrated control system (ICS); and in Santa Helena de Goiás sorghum dry plants were used under the (ICS) system and green Pennisetum glaucum and green weeds (Commelina benghalensis and Bidens pilosa) under the Apply and Sow system in 1999/2000. Each experiment consisted of 0, 2, 4, 6, 8, 10, 12, 14, 16, 18 e 20 t ha-1 of biomass of different vegetal coverings, each receiving the five herbicides separately in a completely randomized design, with three replicates. The amount of herbicide left in the soil was determined in a bioassay, with Sorghum bicolor being used as a control plant. The results showed that 80% of the applied doses of diclosulan and sulfentrazone were present in soil independently of the amount and type of soil covering. For dimethenamid and S-metolachlor, up to 50% of the amount applied was detected in soils with 8 t ha-1 of plant residues.
bioassay; Phaseolus vulgaris; herbicide in soil
