Analysis of spatial relationships between soil and crop variables in a durum wheat field using a multivariate geostatistical approach

For important crop and soil properties, temporal variability is generally higher than spatial variability and the definition of stable low- and high-yield potential zones, for site-specific management, is very difficult. In this study the application of a multivariate geostatistical methodology, factorial kriging analysis (FKA), is proposed for this purpose, allowing simultaneous processing of several layers of information on spatially and temporally variable crop and soil properties. The methodology was applied to measurements carried out in a durum wheat field in Viterbo (Central Italy). Soil properties, plant development and biomass, LAI and normalized difference vegetation index (NDVI) were measured following a grid sampling scheme. Yield components were assessed at the same points at harvest. Coregionalization analysis was carried out and FKA was applied in order to clarify the spatial relationships between the different variables acting at the different scales. The application of FKA to soil, plant and yield properties allowed to discriminate between variables with a different rate of variation, pointing out at those more stable which could be used as a basis to site-specific management. (c) 2007 Elsevier B.V. All rights reserved.