Soil properties prediction of western Mediterranean islands with similar climatic environments by means of mid‐infrared diffuse reflectance spectroscopy

This study investigated the suitability of mid‐infrared diffuse reflectance Fourier transform (MIR‐DRIFT) spectroscopy, with partial least squares (PLS) regression, for the determination of variations in soil properties typical of Italian Mediterranean off‐shore environments. Pianosa, Elba and Sardinia are typical of islands from this environment, but developed on different geological substrates. Principal components analysis (PCA) showed that spectra could be grouped according to the soil composition of the islands. PLS full cross‐validation of soil property predictions was assessed by the coefficient of determination ( R 2 ), the root mean square error of cross‐validation and prediction (RMSECV and RMSEP), the standard error (SECV for cross‐validation and SEP for prediction), and the residual predictive deviation (RPD). Although full cross‐validation appeared to be the most accurate ( R 2 = 0.95 for organic carbon (OC), 0.96 for inorganic carbon (IC), 0.87 for CEC, 0.72 for pH and 0.74 for clay; RPD = 4.4, 6.0, 2.7, 1.9 and 2.0, respectively), the prediction errors were considered to be optimistic and so alternative calibrations considered to be more similar to ‘true’ predictions were tested. Predictions using individual calibrations from each island were the least efficient, while predictions using calibration selection based on a Euclidian distance ranking method, using as few as 10 samples selected from each island, were almost as accurate as full cross‐validation for OC and IC ( R 2 = 0.93 for OC and 0.96 for IC; RPD = 3.9 and 4.7, respectively). Prediction accuracy for CEC, pH and clay was less accurate than expected, especially for clay ( R 2 = 0.73 for CEC, 0.50 for pH and 0.41 for clay; RPD = 1.8, 1.5 and 1.4, respectively). This study confirmed that the DRIFT PLS method was suitable for characterizing important properties for soils typical of islands in a Mediterranean environment and capable of discriminating between the variations in soil properties from different parent materials.