This experiment was carried out to determine the critical periods of weed interference in beans, in response to no-tillage sowing, in the spring season. The culture sown before the bean was pearl millet, generating 2 t ha-¹ dried mass. The treatments were composed by periods of weed coexistence or weed control (0, 10, 17, 24, 31, 45, and 66 days after seedling growth). The experimental design was a complete randomized block with the treatments arranged in a 2 x 6 factorial design, with four replications and two controls (with or without weed control). The relative importance (RI) of the weed community was evaluated. The weed community consisted of sixteen species, with 10 being eudicots. Ragweed (Parthenium hysterophorus) had the highest IR. The period before weed interference was up to the phenological stage V2, and the total interference period was up to the V4 stage. Thus, the critical period of interference corresponded to the period between the V2 and the V4 bean phenological stages. Spring sowingproved to be aviable alternative. Average mass accumulation and weed density were low, causing less interference in the culture and, consequently, a short critical period of interference. The reduction of 19% in the average productivity of the beans was observed when the culture remained with the weeds throughout the cycle.
Phaseolus vulgaris; weeds; competition; Amaranthus retroflexus
