SUPERFÍCIES GEOMÓRFICAS E ANÁLISE MULTIVARIADA DE ATRIBUTOS QUÍMICOS DO SOLO NA DEFINIÇÃO DE ZONAS DE MANEJO

Abstract

With continued population growth, food scarcity and declining quality and quantity from natural resources, current agricultural practices should be reconsidered. A better understanding of the spatial variability of soil and climatic components, as well as their numerous interactions on the yield of agricultural crops is the current approach called Precision Agriculture. Thus, identify management zones, targeting areas with relative homogeneity as a broad set of variables, can be approximate to management of spatial variability. There weregeoreferenced soil samples divided into three layers, P5 (0 - 0.05m), P10 (0.05 to 0.01 m) and P20 (0.10 - 0.20 m). The first analysis it is the landscape model, which was considered the angle of the geomorphic surface for management class delineation. The crop was segmented into four geomorphic segments, top surface, convex surface, the concave surface and plain. The second analysis was designed as mathematical technique, making use of multivariate statistical analysis, performing hierarchical cluster analysis (HCA) considering similarity between samples, and principal component analysis (PCA), the variances describing characteristics of soil attributes. For each measured layer of the soil, the samples are segmented by means of HCA into four zones management as approximate 60% similarity at Euclidean distance. To verify the significance of the segmentations,drew on comparison of average 5% by orthogonal contrasts test, considering each segment or mathematical landscape management zone as a treatment in the analysis. In the case of landscape segments considered, there were no significant differences between the means of soil attributes and yield components of soybean considered. The contrasts applied to the areas of mathematical systems differed significantly from each other for most soil properties and yield components under analysis. Thus, the use of multivariate analysis can be a tool in order to provide differentiation with regard to handling of agricultural areas.