A nose for cheese

NIR spectroscopy has shown itself to be a bit of a cheese connoisseur, able to to determine various attributes of cheese including geographical origin and season of production. Now, French scientists have shown that it can even determine the diet of the cows that supplied the milk.

Donato Andueza and his colleagues at the French National Institute for Agricultural Research (INRA) analysed over 300 fresh samples of three different varieties of cow's milk cheese using both visible and NIR spectroscopy. The cows that produced the milk for these cheeses had either mainly grazed on fresh grass or had been mainly been fed hay and protein-rich animal feed.

The scientists found that segments of both the visible and NIR spectra from cheeses made from milk produced by cows fed on grass were significantly different to the spectra from cheeses made from milk produced by cows fed on hay and animal feed. As they report in Food Chemistry, statistical models built from this spectral data could accurately classify the cheeses according to the cows' diets. The models built from the NIR spectra proved to be most accurate, correctly classifying over 95% of the cheeses.

According to the scientists, both the visible and NIR spectra were mainly detecting differences in the concentration of the organic pigments known as carotenoids in the cheeses. Grass contains a higher concentration of carotenoids than hay, and this is carried over into the milk and cheese. However, the NIR spectra were also detecting differences in other compounds, such as fatty acids, which explains the greater accuracy of the models built from NIR spectra.