Characterization and recognition of citrus fruit spoilage fungi using Raman scattering spectroscopic imaging

Abstract

Citrus fruit is cultivated globally with a high production every year. However, it is easily infected by spoilage fungi with toxic metabolites. The mechanism of changes in peel tissues of mandarins was revealed by Raman microscopic imaging. Single Raman spectra, multiple Raman spectra and microscopic images of fresh fruits and infected fruits by spoilage fungi were acquired by confocal Raman microscopy. Intensity reduction of Raman characteristic peaks at 1008 cm−1,1154 cm−1 and 1525 cm−1 indicated the content of carotenoids in citrus fruit peels decreased significantly during the periods of fruit spoilage. The distribution and composition of carotenoids in the mandarins were directly illustrated on the visualized microscopic images. Classification algorithms were operated in discrimination on fresh and spoilage samples based on Raman spectra. Classification on fresh fruits and all spoiled fruits were better than that of identification on four levels (fresh samples, slight spoilage samples, medium spoilage samples and serious spoilage samples). Liner discriminant analysis (LDA) of mandarins infected by A. alternate achieved the optimal result, with correlation coefficients of calibration set and prediction set both reaching 100%. The correlation coefficients of samples infected by A. niger achieved 0.925 for calibration set and 0.900 for prediction set. Meanwhile, infected samples by P. italicum reached 0.842 and 0.800. Raman microscopic imaging was confirmed to be a powerful tool to identify compositional changes in fruits caused by spoilage fungi without chemical treatments.