This study aims to verify the effect of crop varietal architecture and straw removal from planting rows in the occurrence of weeds in sugarcane and infer about the sustainability of the production system with no herbicide application. The experiment was established in 2011 in a randomized block design with split plots and four replications. Main plots were varieties IACSP95 5000 and SP91 1049. In the sub-plots there the straw removal was allocated (evenly scattered in the area, or concentrated in inter-rows). Assessments were conducted in 2012 and 2013 and the absolute levels of infestation, density, frequency and dominance of weed species were obtained. Areas were intra-characterized by the coefficients of Simpson and Shannon-Weiner and sustainability inferred by the SEP coefficient. Areas were grouped by the similarity coefficient of Jaccard. Other factors besides leaf architecture were more significant for level of infestation. Treatments with straw removal from planting rows were more infested than those with evenly scattered straw. In the second year of cultivation, those species most adapted to the system increased their importance value. Wild poinsettia was the dominant weed in all treatments, deserving attention from pre-planting on, to reducing its occurrence in the soil seed bank. Species diversity was higher where straw was evenly scattered due to the occurrence of species other than wild poinsettia. Sustainability was reduced from the first to the second year, indicating that only cultural practices are not enough, even with high shading provided by crop and straw production, demanding herbicides.
Saccharum spp.; phytosociology; sustainability