Técnicas multivariadas na avaliação de atributos de um Latossolo vermelho submetido a diferentes manejos

The conversion of natural environments in livestock systems, especially systems of monoculture, causing changes in soil attributes. The objective of this study was to evaluate changes in chemical and physical attributes of an red Latossol under different managements (sugarcane, native forest and reforested area), with the use of multivariate statistical techniques. In each area, were collected, randomly, four samples (composed by 15 points), in the layers of 0.0-0.1, and 0.1-0.2 m. Were evaluated chemical attributes: pH, organic matter, potassium, calcium, magnesium, total acidity, aluminum, available concentrations of sulfur, boron, copper, iron, manganese and zinc, and the physical attributes: macro and microporosity, bulk density, sand, silt and clay concentrations. The data were performed to multivariate statistical analyzes. The cluster analyses and principal components have allowed to identify the formation of three groups formed by native forest, the second by area in reforestation and the third by the area cultivated with sugar-cane. The use of techniques of multivariate analysis was efficient to check the similarities and/or differences, based on chemical and physical attributes of the soil of the environments studied.