Multivariate analysis and geostatistics for hydro-physical variables of different soils

Abstract

The research aimed the use of multivariate analysis, in order to reduce the dimensionality of the hydro-physical variables set of different soils in the Northwestern Paraná, to facilitate the interpretation of the interdependence between them and also to identify the similarity between the soils. The experiment was conducted in a regular grid of 36 sampling points, in three soils: an Alfisol and two distrophic Oxisols; under different managements. The data were subjected to the principal component analysis, to transformation in new variables, and after to the geostatistics, to quantify the spatial dependence degree of the major components and of the soil water infiltration rate. The principal component analysis supported in understanding of the differences and similarities among the hydro-physical variables of the soil. With the principal components analysis, there was a reduction from thirteen soil hydro-physical variables to two new variables, that were soil porosity and infiltration rate, which explained 78.4% of the variation in data.