Abstract
Clementine is a citrus fruit that has found a peculiar habitat in specific areas of Calabria, a region located in southern Italy. Due to its peculiar characteristics it was recently awarded with protected geographical indications (PGI) from the European Union. In this work, stepwise linear discriminant analysis (S-LDA), soft independent modeling of class analogy (SIMCA), and partial least-squares discriminant analysis (PLS-DA) were used to build chemometric models able to protect PGI Clementine from others of different origin. Accordingly, the concentration of 24-26 elements was determined in peel and juice samples, respectively, obtained from Calabrian PGI clementine and from fruits cultivated in Algeria, Tunisia, and Spain. A cross-validation procedure has shown very satisfactory values of prediction ability for both S-LDA (96.6% for juice samples and 100% for peel samples) and SIMCA (100% for both peel and juice samples). PLS-DA models also yielded satisfactory results.