Fitossociologia e manejo de plantas daninhas de difícil controle, em pré-semeadura da soja

Abstract

Soybean is one of the most important crops in the world economy and has been consolidated as the most important for Brazil, presenting several limiting factors to its potential yield, among them, weeds. The period prior to sowing and emergence of the crop is one of the most important, so pre-sowing desiccation is essential. In this sense, establishing alternatives for a greater spectrum of weed control is essential for the sustainability of the management system. In view of reactive management in crops, with the continuous use of glyphosate to control weeds, the objective was to evaluate the efficiency of herbicides in the pre-sowing phase of soybean. The experiment was carried out in a commercial crop, carried out in a randomized block design, in strip arrangements. The weed survey was carried out in pre-desiccation, through square inventory, using conventional analytical tools and precision agriculture tools (UAV). From this, phytosociological indices were estimated that encompassed the density, frequency, abundance, importance value and spatial variability of the botanical classes. The control efficiency was

estimated by conventional method and through aerial images obtained with UAV and Sentinel- 2 satellite. The use of precision agriculture tools proved to be effective for monitoring the spatial

dynamics (geostatistical analysis) and control efficiency (aerial mapping with UAV) of weeds, in soybean pre-sowing. Herbicide associations proved to be potential alternatives for rotating mechanisms of action and preventive management of weeds in soybean pre-sowing. The use of remote sensing (Sentinel-2 satellite image) in the post-desiccation phase, expressed by mean values of the NDVI index, demonstrated complementarity to the data obtained by the conventional method, in mapping the desiccation of weeds.