Relationship between Sensory Scores and near Infrared Absorptions in Characterising Bitto, an Italian Protected Denomination of Origin Cheese

Abstract

Historically, specific types of cheese are made in certain geographic areas. Often they have unique flavour characteristics. Studies have suggested the role of local pastures in determining cheese aroma. Bitto is a protected denomination of origin cheese produced in summer in Valtellina (Lombardy, Italy). The aim of this paper was to study the relationship between sensory scores assigned to Bitto cheese by a highly trained panel of experts and near infrared (NIR) data in verifying the NIR ability in predicting sensory characteristics. Bitto moulds (39), with assigned sensory scores, were analysed in the whole NIR range by an Fourier transform-NIR spectrometer. Grated cheese spectra (156) were recorded in reflectance mode. Spectra were grouped in two independent sets (calibration/prediction set = 30 samples; test set = nine samples). PLS1 and PLSD were applied. PLS1 results allowed a satisfactory prediction of total sensory score ( RMSEP 2.53; bias 1.2; slope 0.814) and an acceptable prediction of “Taste and Flavour” score ( RMSEP 0.60; bias 0.51; slope 0.565) for the nine samples in the test set. An acceptable preliminary classification of samples into three “quality classes” was also obtained applying PLSD. Factor loadings plots allowed the identification of some NIR absorption bands related to the development of cheese taste and flavour.