Spectroscopic Analysis of Proximal Leaves as a Method for Studying Nectarine Ripening

Traditional methods for fruit quality assessment are labor-intensive, destructive, and result in the loss of marketable produce. Spectroscopy, especially near-infrared (NIR) and mid-infrared (MIR), has helped in the analysis of fruit quality, despite being nondestructive, as it can leave some marks on the fruit. This study investigates the potential of NIR and MIR spectroscopy for monitoring nectarine ripening through the analysis of proximal leaves, leveraging their biochemical and physiological changes during ripening as a practical and truly noninvasive alternative to predict key fruit attributes. Spectral data were analyzed using ANOVA-Simultaneous Component Analysis (ASCA) to determine the key factors influencing spectral variability. The results indicated that the evolution of the spectra was the primary contributor to spectral changes, reflecting physiological dynamics during fruit ripening. Partial Least Squares (PLS) regression models were employed to predict key fruit properties (weight, firmness, sugar content, pH and acidity). The models showed acceptable performance for indirect prediction with R2CV values ranging from 0.4 to 0.7, RPD values from 1.41 to 1.88, and RER values from 5.56 to 10.21. Predictions were good for nectarine properties like weight and firmness, with leaf spectra effectively predicting these fruit characteristics, though predictions for acidity and pH were less robust. Key findings suggest that combining spectral data from both sides of the leaf provides models with good performance, offering a practical noninvasive alternative to destructive fruit quality analysis methods and providing valuable insights for precision agriculture. This approach has great potential to redefine ripening assessments in fruit production and monitoring practices.