Analytical protocols for precision agriculture field trials

CÓRDOBA, M.; BRUNO, C.; BALZARINI, M.

Valle Hermoso, Córdoba

Workshop; Taller Megadatos en Agricultura de Precisión / Workshop Big Data in Precision Agriculture; 2014

UNC-UNdeC-UNRC-INTA

Uniform management of agricultural fields has been increasingly replaced by environmentally based management, which is benefited by the identification of homogeneous zones within crop fields. Several variables are usually measured and georeferenced for this purpose at multiple sites within the field. However, methods for multivariate analysis of those spatial data are not common and a by-variable analysis is generally used. Site classification is only one step in the zone delineation process, and it needs to be complemented with other methods for treating spatial data. Here we present a protocol for the multivariate analysis of a set of site covariates with the aim of delineating homogeneous zones. The analysis sequence includes methods for data preprocessing, multivariate site classification and use of filters to smooth and promote zoning. The protocol has been illustrated using a rainfed wheat crop from the Argentine pampas, with apparent electrical conductivity, elevation and soil depth as master variables for zoning and yield as check trait, but could be applied to other production systems under precision agriculture.