Abstract
Raman, near-infrared, and mid-infrared (MIR) spectroscopic techniques were evaluated for the potential to non-destructively quantify bioactive compounds in commercially grown New Zealand plums. These spectroscopic data sets were qualitatively assessed using principal component analysis where groupings based on the cultivar and growing location were observed. The different techniques separated based on different compositional differences. Four cultivars grown in two different growing locations were assessed for total anthocyanin content, vitamin C content, phenolic content, and total antioxidant capacity. These parameters (measured using traditional methods) were correlated to the spectroscopic data using partial least squares regression and support vector machine (SVM) classification where appropriate. Raman and MIR showed the most promise for quantifying vitamin C content, and Raman showed the most promise for quantifying anthocyanin content. Some parameters were better suited to classification, and in this regard, Raman spectroscopy provided a robust method for classifying growing location and MIR for cultivar classification.