A field experiment was carried out at the Training Farm of Ponta Grossa State University, in Ponta Grossa-PR, Brazil, during the 1996/1997 growing season to determine the critical period of weed interference in the common bean under direct seeding system, associated with the phenological stage of the common bean growth. The experimental design was a randomized complete block arranged in a 2 x 8 factorial, with four replications. The 16 treatments tested resulted from a combination of two groups of weed interference treatments : (a) relative weedy period, and (b) relative weed-free period, in seven physiological stages of bean growth: V2, V3, V4, R5, R6, R7 and R8, and a check plot with the crop in coexistence with the weeds. The experiment was carried out on an area 8 years under direct seeding system. Sowing, fertilization, and insect and disease control were performed according to the technology recommended for the crop. The weed interference critical period occurred between the V4 and R6 phenological stages of growth, and bean yield was reduced in 71% when the crop was maintained in coexistence with the weeds during all the crop cycle. Regarding weed composition, the dicotyledons class of weed represented 61.3%, being Bidens pilosa and Richardia brasiliensis the most prevalent, with 30.6% and 16.6%, respectively. Monocotyledons represented 38.7%, and Digitaria horizontalis and Brachiaria plantaginea, were the most prevalent, with 23.6% and 14.3%, respectively.
bean plant; phenology; interference; critical period; no-till
