Assessment of a Nondestructive Method for Rapid Discrimination of Moroccan Date Palm Varieties via Mid-Infrared Spectroscopy Combined with Chemometric Models

Abstract

Morocco is an important world producer and consumer of several varieties of date palm. In fact, the discrimination between varieties remains difficult and requires the use of complex and high-cost techniques. We evaluated in this work the potential of mid-IR (MIR) spectroscopy and chemometric models to discriminate eight date palm varieties. Four chemometric models were applied for the analysis of the spectral data, including principal-component analysis (PCA), support-vector machine discriminant analysis (SVM-DA), linear discriminant analysis (LDA), and partial-least-squares (PLS) analysis. MIR spectroscopic data were recorded from the wavenumber range 4000-600 cm-1, with a spectral resolution of 4 cm-1. The discriminant analysis was performed by LDA and SVM-DA with a 100% correct classification rate for the date mesocarp. PLS analysis was applied as a complementary chemometric tool aimed at quantifying moisture content; the validation of this model shows a good predictive capacity with a regression coefficient of 84% and a root-mean-square error of cross-validation of 0.50. The present study clearly demonstrates that MIR spectroscopy combined with chemometric approaches constitutes a promising analytical method to classify date palms according to their varietal origin and to establish a regression model for predicting moisture content. An alternative analytical method to discriminate date palm cultivars by FTIR-attenuated total reflection spectroscopy coupled with chemometric approaches is described.