Use of GC–MS and 1H NMR low-level data fusion as an advanced and comprehensive metabolomic approach to discriminate milk from dairy chains based on different types of forage

Abstract

As forage may affect the environmental sustainability of a given dairy chain, this study evaluated the discriminant capacity of fatty acids (FAs) and NMR metabolomic profiles of milk from three dairy chains, where forage components of cows diets were: maize silage (MS), grass-legume and maize silage (GMS), grass and lucerne hay (HAY). Canonical discriminant analysis (CDA) based on FAs and NMR metabolites highlighted a reliable discriminative performance for HAY samples that were correctly recognised, especially on the basis of C18:3n-3 and C17:0. The GMS samples were positively correlated with choline, C14:0 and C17:1 cis-9, while the MS ones were represented mainly by C16:1 cis-9. An overlap between MS and GMS samples was observed, even if a low-level fused CDA modelling improved their correct assignment. The footprint of maize silage on the milk metabolomic profile seemed not to be affected if partially replaced by a mix of legume and grass silages.