A successful no-tillage system requires high mass production in the preceding to the main crop that and its crop residues remain in the soil. Pearl-millet was shown to be a good option as an autumn-winter crop before soybean, in rotation, but there are no studies of this system on soybean seed quality. The effects were investigated of three sowing times and five types of crop cut management of pearl-millet on seed quality of soybean cropped in sucession, in no-tillage system, for three crop rotation cycles (1999-2000, 2001-2002 and 2002-2003). Each pearl-millet sowing time was a trial with five treatments (M1 = cut at each flowering and no residues; M2 = cut at each flowering with residues; M3 cut at flowering and no residues; M4 = cut at flowering with residues; M5 = not cut until mature panicle harvest, then with residues) and four replications, arranged in randomized blocks. The pearl-millet trials were sown in 1999: E1 = 5 March, E2 = 25 March and E3 = 19 April; in 2001: E1 = 17 April, E2 = 7 May and E3 = 27 May; and in 2002: E1 = 25 April, E2 = 15 May and E3 = 05 June. The soybean was sown at the same time for the three trials in each crop rotation cycle, after chemical dessecation of the pearl-millet; the soybean was also harvested at the same time. The soybean seeds, cv. Embrapa-48, were evaluated by its size, weight of 100 seeds, water content, germination and vigor. Soybean seeds with better physiological quality were obtained when cropped in sucession on residues of the first pearl-millet sowing for three cycles of crops rotation, regardless of the pearl millet harvest management, whose effect was not clear.
Glycine max; no-tillage; sowing time; phytomass management
