Identification of volatiles formed in Asian pear cultivars subjected to short-term storage using multinomial logistic regression

Abstract

The differentiation of Asian pear cultivars was carried out through analysis of volatile compounds by SPME / GC / MS, with analysis of principal components as follows. Volatiles detected by logistic regression were significantly different depending on cultivar and post-harvest ripening at 20 °C. PCA components were expressed as a linear combination of selected variables to differentiate cultivars. Four compounds (-)-menthol, hexanal, 2-methoxy-4-vinylphenol and benzaldehyde were sufficient to differentiate cultivars by stepwise multinomial logistic regression. Ethyl-3E-hexenoate and 4-methyl-2-pentanol calculated using the same statistical method distinguished the fruits under post-harvest ripening period of 27 days at 20 °C.