Unidades de Gestão Diferenciada por meio de índices de vegetação e mapas de produtividade

Abstract

Many studies indicate that the population in the world may have a large increase in a few years and, in addition, the demand for food also tends to increase. Moreover, the space for food production is close to a limit, with many areas already in degradation and no longer being able to supply this need. Therefore, an alternative to traditional production methods is necessary, such as precision agriculture, which aims to increase productivity by optimizing the use of resources and available space. However, it is necessary to collect data on all parameters that may interfere with productivity and, even more importantly, the best way to interpret this data to make decisions. The present work aims to evaluate several vegetation indexes, through harvest maps, for the delineation of management zones (ZMs), since productivity data is generally difficult to access among most small and medium producers. For this, ten vegetation indexes were selected for four different dates, referring to two soybean crops and two corn crops. Such crops have harvest records, which were converted into maps and used for comparison and validation of maps generated by vegetation indexes. The management zones were generated using the Fuzzy C-means algorithm, which performs similar data grouping according to a predetermined number of classes. For determination of the best procedure, the data were submitted to analysis of variance and fuzzy validation indexes. To determine the best delimitation between the ZMs of the vegetation indexes, the Kappa index was used. The results indicated a better grouping of the data in four classes, generating four ZMs. The comparison between the ZMs generated by productivity and the ZMs generated by vegetation index indicated good agreement for four classes (K = 0.64 to 0.70). It is expected, with this work, a contribution in the transformation process of agriculture in regards to making information easier to access and increasing anticipation in decision making.