Seleção de variáveis para geração de zonas de manejo e diferentes densidades de semeadura de soja

• Autor(a) principal: Camicia, Rafaela Greici da Motta
• Data de Publicação: 2018
• Tipo de documento: Tese
• Idioma: por
• Título da fonte: Biblioteca Digital de Teses e Dissertações do UNIOESTE
• Texto Completo: http://tede.unioeste.br/handle/tede/4103
• topic: Agricultura de precisão; Seleção de atributos; Unidades de manejo; Produtividade.; Precision agriculture; Selection of attributes; Management units; Productivity; CIENCIAS AGRARIAS::ENGENHARIA AGRICOLA

The selection of variables for the management zone design (MZs) generally uses productivity data from previous years; however, questions remain about the type and amount of data needed to classify these MZs, as well as whether data normalization interferes with this selection. All of these approaches merely identify the potential of the MZs; additional research is needed to test whether the identified MZs actually function as effective management zones to increase productivity, and plant density in soybean cultivation is a relevant management practice for reaching high grain yields. In this context, the present study was carried out to verify if the selection of the variables used for the design of MZs is influenced when data from one or more years of productivity are used, and if the normalization methods can influence this selection. The three main techniques of data normalization proposed in the literature with data of up to five years of productivity were evaluated. The behavior of soybean yield under different seeding densities was evaluated in two pre-established management zones. The experiments were carried out in two commercial agricultural areas, located in the state of Paraná, Brazil, where corn and soybean was grown, with data obtained between the years of 2012 and 2018. With the experiments, it was possible to conclude that the productivity did not present spatial autocorrelation in some simulations; however, this did not influence the selection of the variables. Among the studied variables, the altitude and soil mechanical resistance to penetration (RSP) correlated with soybean and corn crop productivity in both study areas; the number of harvests negatively influenced the analysis of spatial correlation between yield and soil attributes; the amplitude normalization method showed the best results of variance (VR) and ANOVA reduction and the mean one showed the greatest reduction of the coefficient of variation (CV). ZM with higher productive potential presented better but not expressive results regarding productivity, since there was no statistical difference between the means. Seed densities produced yield differences; for soybean with line spacing of 0.70 m, the density of 15 plants m-1 provided the highest yields; and, by means of economic analysis, it is confirmed that the use of this density in the whole area is the best option to maximize final yield.