APPLICATION OF MULTIVARIATE TECHNIQUES IN THE AGRICULTURAL LANDS APTITUDE IN CARABOBO, VENEZUELA

Abstract

Background. The selection of the best agroclimatic zones is of significant importance for the entity's agricultural, rural and forestry development, mainly because the inhabitants and local actors of the Carabobo state can already count on a study that indicates the places to produce their food. , use them to the maximum and in the best way. Objective. Analyze and interpret the suitability of agricultural land in Carabobo through multivariate techniques. Methodology. The surface data (Km2) were used for the dates of May-August of four strategic crops: corn, onion, tomato and potato. Principal Component Analysis (PCA), using Partial Least Squares Regression Discriminant Analysis (PLSDA) and Hierarchical Cluster Analysis were applied. Results. Using the PCA, the two components that accounted for 67.6% of the total variation were selected. This allowed a classification of the municipalities with the largest suitable area of crops according to the sowing date, discriminating with greater incidence the area suitable for maize in the Libertador, Carlos Arvelo, Valencia and Naguanagua municipalities, whose ideal sowing date would be established from May 11-20. Conclusion. The multivariate PCA method used represented a tool to describe and characterize the diversity of the areas studied in agriculture in Carabobo, the results of which could be the basis for agricultural development strategies.