Pattern recognition of peach cultivars (Prunus persica L.) from their volatile components

The volatile compounds of four peach cultivars (Prunus persica L) were studied: Sudanell, San Lorenzo, Miraflores and Calanda (two clones, Calante and Jesca). 17-23 Samples of each cultivar with the same maturity level were analyzed, measuring color, firmness, and soluble solids content. The pulp was crushed and mixed with water prior to HS-SPME analysis, and GC-MS was used to determine the volatile compounds. Sixty-five compounds were identified using spectral library matching, Kovat's indices and, when available, pure standards. The main components were lactones and C6 compounds. From the distribution of these compounds, Principal Component Analysis led to the clustering of the samples according to their different cultivars. Finally, Canonical Component Analysis was used to create a classification function that identifies the origin of an unknown sample from its volatile composition. The results obtained will help to avoid fraud and protect the European Designation of Origin 'Melocoton de Calanda'. (C) 2012 Elsevier Ltd. All rights reserved.