FITOSSOCIOLOGIA DE PLANTAS DANINHAS E ATRIBUTOS PRODUTIVOS DE SOJA EM DIFERENTES SISTEMAS DE CULTIVO EM LATOSSOLO

Abstract

Understanding the effects of management applied in different agricultural production systems on the ecological interactions between crops and weeds is fundamental for sustainable management. The identification of the flora present in production areas is the basis for the development of effective weed control strategies and is obtained through phytosociological surveys. In view of the above, the aim of this study was to identify and quantify the predominant weed species in soybean cultivation areas under three different forms of management: conventional system (SC), direct seeding system (SSD) and crop integration livestock-forest (ILPF), in Oxisol. Furthermore, there were evaluated the productive aspects of soybean in each of these systems, aiming to contribute to a more comprehensive and improved understanding of the interactions between weeds and the main crop. The soybean cultivar used in the two agricultural years was TMG 7062 IPRO. A randomized block experimental design was used with three treatments and 8 replications. The treatments were: conventional system (SC), direct seeding system (SSD) and crop, livestock, and forest integration system (ILPF). The total area of each experimental plot was 0.06 ha in SC and SSD and 0.15 ha in ILPF. The total area of the experiment was 2.5 ha, with 0.50 ha in the SSD, 0.5 ha in the SC and 1.5 ha in the ILPF. The phytosociological survey was carried out in the second agricultural year. Phytosociology was carried out at three different times: at 5 and 30 days after emergence (DAE) and pre-harvest. Weed plants were identified and quantified using the inventory square method (1.0 x 1.0 m). Five sampling points were carried out in each experimental plot, for which the following phytosociological parameters were determined: density (D), relative density (Dr), frequency (F), relative frequency (Fr), abundance (A), relative abundance (Ar), relative importance index (Ir) and the Similarity Index (IS). The following biometric variables were also evaluated in soybeans: plant height (AP), number of pods per plant (NVPP), number of grains per plant (NGPP) and productivity in kilograms per hectare (P kg.ha-1). Throughout the different evaluation periods in the three cultivation systems analyzed, 13 species of weeds were identified, belonging to 7 different botanical families. Among the families identified, Poaceae stood out, presenting the highest number of individuals (NI) in the three cultivation systems, regardless of the evaluation period. Of the weeds identified, the most abundant in the three cultivation systems were Portulaca oleracea, Cenchrus echinatus, Amaranthus retroflexus and Digitaria horizontalis, at the evaluation times. Among the management systems, the SC had a greater number of individuals when compared to the SSD and ILPF. The ILPF presented weeds with lower importance index values. The species found in the survey revealed high variability in phytosociological parameters, however, there was similarity between these species and the cultivation systems, as well as the evaluation period. Higher soybean productivity was obtained in SSD (5306.47 kg.ha-1) and in ILPF (4603.69 kg.ha-1).