Técnicas de agricultura de precisão para avaliação da variabilidade de solo e do cafeeiro em pequenas propriedades do Sul de Minas Gerais

• Autor(a) principal: Melo, Bruno Manoel Rezende de
• Data de Publicação: 2015
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Repositório Institucional da UFLA
• Texto Completo: http://repositorio.ufla.br/jspui/handle/1/10532
• topic: Paisagem; Geoestatística; Zonas homogênea de manejo; Agricultura de precisão; Análise econômica; Landscape; Geostatistics; Homogeneous management zones; Precision agriculture; Economic analysis; Fitotecnia

Coffee growing has relevance regarding to agricultural production in Minas Gerais state. Southern of Minas Gerais produces 50% of all state coffee, being present in small and large farms, but there is less productivity by small producers when compared to corporate farmers. This low productivity is often associated to low access to technologies. In this scenario, an alternative to the reality is adequate availability of technology that meets this producers’ class. In this context the use soil sampling grids has the potential to be introduced in small farms in order to optimize the use of resources, since the conventional soil sampling process does not characterize the existing variability, so the use of inputs is done through the medium contents from the conventional sampling procedure. The objective was to assess the economic and agronomic feasibility of using precision agriculture techniques on fertilizers recommendation with different sampling grids compared to the traditional model soil sampling in coffee growing of small farms and the definition of homogeneous management zones according to chemical variables of soil and productivity. The experiment was conducted in the city of Inconfidentes - MG, on the Farm Escola property, where soil sampling grids were used with 52, 28 and 15 collection points and in the Água Limpa (small farm) where grids were tested with 49, 26 and 17 sampling points. In each property was also made conventional sampling to verify the use of inputs in the sampling process in grids and in the conventional model. Each soil sample process cost with the obtained revenue of coffee production in each field of study was related. Geostatistics was used to characterize the spatial dependence and predict attribute values in unsampled locations. It has been found that in some grids sampling was not possible to observe spatial dependence. For some soil chemical parameters, it was found that the conventional sampling model has the potential to continue to be used since the landscape is considered in determining the crops limits. In the global economic analysis study for both soil sampling procedures, it was found negative revenue. It is conclude that positive revenue was detected only for the management zone with 2.1 L/coffee/plant for the first year of production for the Farm Escola. However, total net revenue for soil sampling in grids and in traditional system were negative. For Água Limpa were not calculated production costs in view the lack of spatial dependence for the phosphorus attribute. It is observed by the productivity maps; potassium, organic matter and base saturation on the Farm Escola, these attributes enable to define two management zones in crop position function in the relief. For phosphorus and clay there was no possibility of defining management zones. In Água Limpa property there was not possible to define homogeneous management zones.