Ability of 3 tanniferous forage legumes to modify quality of milk and Gruyère-type cheese

Condensed tannins (CT) may affect ruminal biohydrogenation of dietary polyunsaturated fatty acids. A feeding experiment was conducted with 24 Holstein cows to evaluate whether diets containing CT from different forage legumes can increase polyunsaturated fatty acids, especially n-3 fatty acid content in milk and cheese, without affecting negatively their physicochemical and sensorial properties. Cows were assigned to 4 treatment groups (n = 6) for 52 d, divided into 2 periods: a control period (CoP) and an experimental period (ExP). During the CoP, cows received a basal diet composed of hay, corn silage, ExtruLin (Trinova Handel & Marketing AG, Wangen, Switzerland), concentrate, and alfalfa (AF) in a ratio of 45:25:5:7:18. In the ExP, in 3 of the 4 groups AF was replaced by either sainfoin (SF; 19% CT in dry matter) or 1 of 2 cultivars of birdsfoot trefoil [Polom (BP), 3% CT; Bull (BB), 5% CT]. At the end of each period, milk was collected on 3 consecutive days and analyzed for milk gross composition and fatty acid profile and was processed to Gruyère-type cheese. A trained panel assessed the sensory quality of raw milk and cheese using discriminative and descriptive tests. This experimental design consisting of AF in both the CoP and ExP allowed us to quantify effects due to lactation stage and experimental diets. In both the CoP and ExP, dry matter intake and milk yield did not differ among treatment groups. From the CoP to the ExP, milk urea content was reduced by 23% with SF, remained unchanged with BP, and tended to increase with AF and BB. The odor of the raw BB milk was judged to be different from AF milk. With SF, switching from the CoP to the ExP resulted in a 17% increase of the 18:3n-3 proportion in milk and cheese lipids. In BP cheese, the increase was 3%, whereas it tended to decrease in BB cheese. Additionally, the 20:5n-3 and 22:5n-3 proportions tended to increase in SF cheese from the CoP to the ExP. Compared with the AF cheeses, cheeses from cows fed CT-containing legumes were judged harder and tended to be less adhesive to the palate. In addition, SF and BP cheeses had less rind. In conclusion, feeding SF compared with BB and BP increased the content of 18:3n-3 in the milk and the cheese without a negative effect on flavor of the cheese. Despite a similar CT content, the 2 birdsfoot trefoil cultivars had opposite effects on milk urea and 18:3n-3 deposition, suggesting that, besides the content, the chemical structure may have had an important effect on the CT efficacy.