Management zones delineation using multivariate geostatistics

Optimum benefits on profitability and environment protection depend on how well land use and agricultural practices are fitted to variable soil conditions. Soil variability is the result of both natural processes and management practices, acting at different spatial and temporal scales, so it is critical to characterize soil with precision, both quantitatively and spatially (Castrignanò et al., 2000). Adequate techniques of data analysis are then necessary to put in evidence important spatial relationships and to identify those factors that control the variability of soil properties. Multivariate geostatistics uses the information coming from relationships among variables in order to subdivide an agricultural field into smaller, more homogeneous units, with respect to soil physical and chemical properties, as required by the application of site-specific techniques and more modern technologies. Some of the several possible factors that govern soil variations are likely to have a short-range action, w