In the aim to validate the NIRS technique for rapid characterisation of mountain cheeses made at "Stazione Sperimentale Alpina" (Sauze d'Oulx) during three trials [A=2002: compared winter feeding (Hay) vs. easter feeding (Grass); B=2003: winter feeding with hay vs. grass-silage); C=2004: as B] as compared with their fatty acid (FA) composition, thirty-nine cheeses (Toma Piemontese and Reblochon) were scanned by a PE FT-NIR IdentiCheck spectrometer. The trial C validated the previous NIRS equations with better results than cross-validation of A+B dataset: on av.ge 1-VR was 0.32 and resulted improved in FA>10% (from 0,07 to 0,32), then in FA 2?10% (from 0,18 to 0,27) and as expected in the low class FA<2% (from 0,34 to 0,36), and as a whole with R2 = 0,983, sb=±0,87%. These preliminary results while numerically limited appear to be a little satisfactory for rigorous FA analyses. The NIRS technique should be highly recommended as preliminary discriminant before starting true analyses, and for obtaining rapid technique-commercial indications about total lipids, dry residual and estimates of some FA contents.
Validazione NIRS e gascromatografica dei formaggi di montagna della Stazione Sperimentale Alpina di Sauze d'Oulx: risultati preliminari.
PEIRETTI PG;BATTELLI G;
2004
Abstract
In the aim to validate the NIRS technique for rapid characterisation of mountain cheeses made at "Stazione Sperimentale Alpina" (Sauze d'Oulx) during three trials [A=2002: compared winter feeding (Hay) vs. easter feeding (Grass); B=2003: winter feeding with hay vs. grass-silage); C=2004: as B] as compared with their fatty acid (FA) composition, thirty-nine cheeses (Toma Piemontese and Reblochon) were scanned by a PE FT-NIR IdentiCheck spectrometer. The trial C validated the previous NIRS equations with better results than cross-validation of A+B dataset: on av.ge 1-VR was 0.32 and resulted improved in FA>10% (from 0,07 to 0,32), then in FA 2?10% (from 0,18 to 0,27) and as expected in the low class FA<2% (from 0,34 to 0,36), and as a whole with R2 = 0,983, sb=±0,87%. These preliminary results while numerically limited appear to be a little satisfactory for rigorous FA analyses. The NIRS technique should be highly recommended as preliminary discriminant before starting true analyses, and for obtaining rapid technique-commercial indications about total lipids, dry residual and estimates of some FA contents.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.