Calibration and validation of Veris MSP3 on two soils in Guanacaste, Costa Rica

Abstract

Introduction. It is necessary to implement and validate precision agriculture (PA) technologies in tropical regions. Objective. The objective of this study was to calibrate and validate Veris MSP3 equipment in two different fields located in Guanacaste, Costa Rica. Materials and methods. The data obtained by the sensors integrated into the Veris MSP3 equipment was correlated with laboratory analysis of soil samples with the intent to develop simple or multiple linear regressions to predict soil texture, organic matter, nitrogen in the soil, total cation exchange capacity, and pH, in order to obtain a model that best suits each of the parameters. Results. The regressions that resulted with the best models were: i) apparent electrical conductivity (CEa) with soil texture, ii) optic ratio and slope with organic matter, iii) optic ratio with N, iv) CEa with total cation exchange capacity (total CIC), and v) Veris pH with pH in water and KCl. The higher determination coefficient was obtained between CEa and sand percentage with r2 of 0,82. In addition, the r2 for the rest of parameters ranged from 0,28 to 0,82. Conclusion. The calibration method used gave reasonably precise correlations (r2≥0,55) for soil texture at depth from 0 to 30 cm and organic matter variables. However, from 30 to 90 cm soil texture, N, total CIC, and pH anoter calibration methodology should be considered because of imprecise correlations (r2≤0,55).