Abstract
Hay milk is a traditional dairy product recently launched on the market. It is protected as "traditional specialty guaranteed" (TSG) and subjected to strict regulations. One of the most important restrictions is that the cow's feed ration must be free from silage. There is the need for analytical methods that can discriminate milk obtained from a feeding regime including silage. This study proposes two analytical approaches to assess the authenticity of hay milk. Hay milk and milk from cows fed either with maize or grass silage were analyzed by targeted GC-MS for cyclopropane fatty acid (dihydrosterculic acid, DHSA) detection, since this fatty acid is strictly related to the bacterial strains found in silage, and by HPLC-HRMS. The presence of DHSA was correlated to the presence of maize silage in the feed, whereas it was ambiguous with grass silage. HPLC-HRMS analysis resulted in the identification of 14 triacylglycerol biomarkers in milk. With the use of these biomarkers and multivariate statistical analysis, we were able to predict the use of maize and grass silage in the cow's diet with 100% recognition. Our findings suggest that the use of analytical approaches based on HRMS is a viable authentication method for hay milk.