Work

Portable NIR Spectroscopy for soil property prediction and fertility assessment in coffee and cocoa systems

Year: 2026

Type: article

Abstract: The study evaluated the potential of portable near-infrared (NIR) spectroscopy to predict soil physicochemical properties in coffee and cocoa production systems in Amazonas, Peru. A total of 126 topsoil samples were initially collected using a stratified sampling design. One coffee sample was excluded from the spectral dataset because no corresponding NIR spectral record was available, resulting in a final dataset of 125 samples (69 coffee and 56 cocoa) used for predictive modeling. Soil properties were determined using standard laboratory reference methods, and spectral data were acquired using a portable NeoSpectra spectrometer operating in the 1350–2550 nm range. Predictive models were developed using Random Forest (RF), Support Vector Machine (SVM), Neural Networks (NN), and Partial Least Squares Regression (PLSR), and their performance was evaluated using nested five-fold cross-validation. The predictive performance of portable near-infrared (NIR) spectroscopy varied according to the soil property, production system, preprocessing technique, and machine learning algorithm. Coffee soils showed higher predictive performance than cocoa soils. The best results were obtained for clay (R²CV = 0.775; RPDCV = 2.276) using Random Forest with first-derivative preprocessing, followed by phosphorus (R²CV = 0.772; RPDCV = 2.037) using Partial Least Squares Regression with band-depth preprocessing. Soil pH and silt also showed satisfactory predictive performance in coffee soils. In cocoa soils, potassium exhibited the highest predictive capability (R²CV = 0.454; RPDCV = 1.378) using Support Vector Machine with second-derivative preprocessing, although the overall model performance was lower than that observed for coffee soils. Portable NIR spectroscopy provided potentially useful estimates for selected soil properties, particularly clay, phosphorus, pH, and silt in coffee soils. However, its predictive performance was strongly dependent on the soil property, preprocessing strategy, machine learning algorithm, and production system.

Source: Soil Security

DOI: 10.1016/j.soisec.2026.100250