WEED SUPPRESSION IN SUSTAINABLE INTEGRATED AGRICULTURAL SYSTEMS

Development of integrated agricultural systems emerged in order to increase production and promote several long-term benefits. We aimed to evaluate weed dynamics in agricultural integrated production systems in Brazil, through an experiment implemented in 2009 at Embrapa Western Agriculture, in the city of Ponta Porã, Brazil, with 10 treatments, being two areas under no-till planting system; one conventional tillage area; one forest production area; two crop-livestock integration areas, and four crop-livestock-forest integration areas. Evaluations were performed in the early post-emergence of the 2011/12 cropping season, in completely randomized design. A quadrat with 0.5 m side was released randomly 10 times in each area, and weeds inside each quadrat were counted and collected for determination of dry mass, being estimated the density, frequency, dominance and importance value of each species in each cropping system. Areas were intra-analyzed for diversity by the coefficients of Simpson and Shannon-Weiner, and compared among them for similarity of species by Jaccard s coefficient, being areas grouped by multivariate cluster analysis. The overall most important weed species in the area were Bidens pilosa, Commelina benghalensis, Digitaria horizontalis and Raphanus sativus; there was change in the number and species of individuals according to management type, being broadleaved species predominant. Areas planted with Brachiaria were less infested by weeds. The cultivation of corn intercropped with B. brizantha cv. Xaraés in the second cropping season, post-soybean, between rows of Eucalyptus (Integration Crop-Livestock-Forest - iCLF) proved as the best option to reduce the incidence of weeds in long-term sustainable cropping systems.