The weed supression capacity of cover crops is well known and explored, although only a few works have been conducted on the relative importance of the physical and allelopathic effects on this phenomenon. Two trials were carried out in the field, in 1999/2000 and 2000/2001, at the experimental area of the University of Rio Grande do Sul, Brazil, arranged in a randomized experimental block design with four replications to evaluate the effects of sorghum and pearl millet mulch on weedsupression. In both experimental years, the treatments resulted in a factorial, where factor A was composed by sorghum genotypes RS 11, BR 601, and BR 304, representatives of three classes of hydrophobic root extract production in the laboratory; the pearl millet genotype Comum RS; and by a check without cultures. Factor B was formed by different levels of straw of each genotype over the soil. In 1999/2000, sorghum straw levels of 1.3 t ha-1 were sufficient to decrease Brachiaria plantaginea (BRAPL) and Sida rhombifolia (SIDRH) infestations by 50%. In 2000/2001, 4 t ha-1 of sorghum or pearl millet straw were sufficient to decrease 91, 96, and 59% of the total SIDRH, BRAPL, and Bidens pilosa population, respectively. The presence of sorghum shoot residues is more important in weed supression than the presence of root residues of this crop.
Sorghum spp.; Pennisetum americanum; straw levels; allelopathy; physical factors
