LAUSR

Abstract Authentication of dairy systems is of growing interest for the dairy industry and we investigated the potentiality of using volatile fingerprinting of ripened cheeses by proton transfer reaction time-of-flight mass spectrometry. A total of 1,075 individual model cheeses made from milk of individual Brown Swiss cows of 72 farms were analysed. Using a linear discriminant analysis, cows and herds were assigned to 3 or 5 dairy systems differing in management, available facilities, and diets. We obtained variable discrimination abilities (up to 77% of correct classification of cheeses and 70% of farms with cross-validation). We found m/z 61,028 (acetic acid), 109,070 (pyrazine), and m/z 137,132 (terpene) characterising model cheeses from traditional dairy systems and m/z 71,086 (3-methyl-butan-1-ol, 3-methyl-3-buten-1-ol, pentan-1-ol), m/z 101,097 (hexan-2-one, hexanal), m/z 123,117 (nonenal), m/z 129,127 (octan-1-one, octanal), and two unidentified peaks m/z 83,071 and m/z 93,090 characterising model cheeses from the modern farms. In conclusion, it seems possible to discriminate between a range of dairy systems using fast volatile fingerprinting of ripened cheeses but a proper validation of results obtained is needed. Highlights Mass spectrometry technique (PTR-ToF-MS) was able to discriminate between dairy systems. We found m/z 61,028 (acetic acid), 109,070 (pyrazine), and m/z 137,132 (terpene) characterising model cheeses from traditional dairy systems. We found m/z 71,086 (3-methyl-butan-1-ol, 3-methyl-3-buten-1-ol, pentan-1-ol), m/z 101,097 (hexan-2-one, hexanal), m/z 123,117 (nonenal), m/z 129,127 (octan-1-one, octanal), and two unidentified peaks m/z 83,071 and m/z 93,090 characterising model cheeses from the modern farms.