| Abstrakt: |
Thirty-seven olive oil samples belonging to cultivars: Ntopia, Koroneiki, Thiaki, Asprolia and Lianolia were collected from four Western Greek islands. Samples were subjected to the following physicochemical analyses: acidity, peroxide value, K232, K270, ΔK indices, chlorophyll and carotenoid contents, along with fatty acids in an effort to characterize and mainly classify olive oil samples according to cultivar. Multivariate analysis of variance (MANOVA) showed that the following parameters: acidity, chlorophyll, carotenoid, myristic acid, margaric acid, stearic acid, arachidic acid, and eicosenoic acid were significant (p < 0.05) for the classification of olive oil cultivars. The aforementioned parameters were subjected to linear discriminant analysis (LDA) providing a correct classification rate of 91.9% and 81.1% using the original and cross-validation methods, respectively. Finally, the application of quality control analysis such as Pareto chart showed that with only two variables namely acidity and chlorophyll content, the investigated cultivars could be differentiated, providing thus, a rapid and costless methodology for olive oil cultivar differentiation. [ABSTRACT FROM AUTHOR] |
Full text from SpringerLink