Desenvolvimento de um sistema de apoio à decisão para definir zonas de manejo em cafeicultura de precisão

• Ano de defesa: 2010
• Autor(a) principal: Valente, Domingos Sárvio Magalhães
• Orientador(a): Queiroz, Daniel Marçal de
• Banca de defesa: Pereira, Antonio Alves , Vieira, Carlos Antonio Oliveira , Teixeira, Mauri Martins
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Viçosa
• Programa de Pós-Graduação: Doutorado em Engenharia Agrícola
• Departamento: Construções rurais e ambiência; Energia na agricultura; Mecanização agrícola; Processamento de produ
• País: BR
• Palavras-chave em Português: Agricultura de precisão; Café; Zonas de manejo
• Palavras-chave em Inglês: Precision agriculture; Coffee; Management zones
• Área do conhecimento CNPq: CNPQ::CIENCIAS AGRARIAS::ENGENHARIA AGRICOLA::MAQUINAS E IMPLEMENTOS AGRICOLAS
• Link de acesso: http://locus.ufv.br/handle/123456789/674

The variable rate application of fertilizers and lime based on chemical and physical properties of soil requires a dense sampling for determining the spatial variability in the field. One technique to reduce the number of samples is defining the management zones. Some researchers have demonstrated the importance played by the soil electrical variables to explain the variability of soil properties. Thus, the objective of this study is to develop and to evaluate a decision support system for defining management zones based on soil apparent electrical conductivity in the mountain coffee production fields. The decision support system was structured to perform semivariograms, to generate maps by kriging and to define management zones using fuzzy k-means algorithm. To find the optimal number of classes, two indices were calculated: the Fuzzy Performance Index (FPI) and Modified Partition Entropy (MPE). The electrical conductivity of soil was determined at 0.20 m (CE20) and 0.40 m (CE40)using a portable meter ERM-02 made by Landviser. The data were grouped into six management zones defined from the map of elevation (ZMA), the map of CE20 (ZM20), the map of CE40 (ZM40), the maps of CE20 and elevation (ZM20A), the maps of CE40 and elevation (ZM40A) and the maps of altitude, CE20 and CE40 (ZM2040A). For each case, the field was classified in two, three, four and five classes. To analyze the correlation between the management zones and soil properties, the Kappa coefficient was calculated. The mean values of CE20 and CE40 were 1.80 mS m-1 and 1.22 mS m-1, respectively. The CE20 and CE40 showed low correlation with soil properties. The highest correlation was obtained for the remaining Phosphorus with values of 0.427 and 0.465 for CE20 e CE40, respectively. Maps of CE20 and CE40 presented high similarity to each other, with a correlation coefficient of 0.969. The map of Potassium content was more closely correlated with the map of CE20. For values obtained of FPI and MPE, it was observed that the ZMA requires additional analysis to determine the optimal number of classes. For ZM20 and ZM40, the optimal number of classes was two. For ZM20A and ZM40A, the optimal number of classes was three. The Kappa coefficient for ZMA was better when classifying zinc, base saturation, pH, organic matter, potential acidity and fine sand, in relation to other management zones. For the other soil properties, the highest values of Kappa coefficient were obtained using the ZM20A and ZM40A, and they were statistically equal. By analyzing the Kappa coefficient, it was concluded that ZM20A, ZM40A and ZM2040A were the best way to classify soil properties. However, ZM2040A was not better than ZM20A and ZM40A.