Manejo de dicamba: boas práticas na descontaminação de pulverizadores e avaliação de injúria por índice de vegetação

• Ano de defesa: 2021
• Autor(a) principal: Marques, Matheus Gregorio
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Uberlândia; Brasil; Programa de Pós-graduação em Agronomia
• Palavras-chave em Português: Soja não tolerante ao dicamba; Contaminação de tanque; Deriva simulada; VANT; Agricultura digital; Non-dicamba tolerant soybean; Tank contamination; Simulated drift; UAV; Digital agriculture; CNPQ::CIENCIAS AGRARIAS::AGRONOMIA::FITOTECNIA::MANEJO E TRATOS CULTURAIS; Herbicida; Pulverizadores; Índice de vegetação
• Link de acesso: https://repositorio.ufu.br/handle/123456789/31568; http://doi.org/10.14393/ufu.di.2021.97

The use of dicamba-resistant soybean cultivars will raise concerns about the movement of this herbicide to sensitive crop areas by means of particle drift, vapor drift and the use of sprayer that contains herbicide residue. Because it is a hormonal herbicide, low concentrations are sufficient to cause injury, and this will require greater precision in identifying and predicting damage to plants. The objectives of this work were: (i) to evaluate the dicamba residue during the cleaning process of sprayers with tanks of different materials (fiberglass and polyethylene), and the effects of this residue on non-dicamba resistant soybean; and (ii) to evaluate the damage caused by simulated dicamba drift in soybean culture visually and using the Triangular Greenness Index (TGI) response in images obtained by remotely-piloted aircraft (RPA). The work was divided into two stages. The first consisted of the analysis of the dicamba residue and the spraying of the water collected in the four rinses during the decontamination of the soybean sprayers in the V3 stage. The residue analysis was performed in DIC, with three replications and a 2 x 4 factorial scheme (tank material and number of rinses), using High Efficiency Liquid Chromatography (HPLC). The assessment of the risk of injury by the residue was performed in DBC, with four repetitions and a 2 x 4 + 1 (tank material, number of rinses and a control without application) factorial scheme. The fiberglass sprayer retained more residue only on the first rinse. The dicamba was effectively removed with three rinses, regardless of the material of the tank. Injury and reduced yield were observed by applying the first and second rinses. In the second part, six doses of the herbicide (0; 0.28; 0.56; 5.6; 28 and 112 g a.e. dicamba ha-1) were applied to soybean at stage V3 in DBC with four replications. For the application of the TGI, spectral data acquired with an RGB sensor embedded in a RPA was used. The assessment of injuries caused by dicamba can be performed using the TGI vegetation index, with the ability to assess large areas in a less subjective way than visual assessment.