Rapid Quantification of Phenolic Compounds in Virgin Olive Oil Using Near-Infrared Spectroscopy: A Tool for Breeding and Quality Assessment

Abstract

Phenolic compounds play a key role in virgin olive oil (VOO), affecting both sensory quality and health-promoting properties. Their content and composition are crucial for nutritional evaluation and for characterizing new cultivars in breeding programs. Although high-performance liquid chromatography (HPLC) is the standard method for phenolic analysis, it is laborious, time-consuming, and impractical for the large sample numbers required in genetic studies. Near-infrared spectroscopy (NIRS) offers a fast, non-destructive alternative, yet its use for VOO phenolic profiling remains limited. A comprehensive three-year dataset with broad genetic and environmental variability was assembled to calibrate NIRS models. A subdataset, collected during the same period from four cultivars grown in high-density hedgerow systems across four environments, was reserved for external validation and for estimating broad-sense heritability (H²). To enhance the variability of the calibration set, a core subset from the world germplasm collection was also included. NIRS spectra were acquired using an FT-NIR MPA instrument (1100–2200 nm). Partial Least Squares (PLS) models demonstrated the best predictive performance for oleocanthal, oleacein, and oleuropein aglycone, with high coefficients of determination (R²). Oleocanthal and oleuropein aglycone achieved residual predictive deviation (RPD) ≥2.5 and range error ratio (RER) >10, while oleacein and ligstroside aglycone also performed well (RPD = 2.49 and 1.71; RER = 9.03 and 7.75, respectively). Variance component analysis revealed that genotype was the main contributor to ligstroside aglycone variance, whereas environment dominated for oleuropein aglycone, oleacein, and oleocanthal. The highest broad-sense heritability (H²) values were found for ligstroside and oleuropein aglycones, and the lowest for oleocanthal. These findings demonstrate that FT-NIR MPA spectroscopy is a powerful, high-throughput tool for predicting VOO phenolic composition in breeding programs.