With the aim to test the effects of roughage source and species on fatty acid (FA) composition of milk, an experiment was initiated on 10 buffalo and 12 bovine cows, allotted by the analog method into control (fed silage) and experimental (fed hay) groups. Milk samples were taken individually at three test days prior to experiment (TD0), after one (TD1) and two (TD2) months from the start and subjected to fatty acid analysis. The effects and co-effects were tested by two-way ANOVAs for TD1 and TD2 data, while for the assessment of the differences between species at TD0 by one-way ANOVA. Compared to the bovine, at TD0 the buffalo milk displayed significantly lower contents of C8:0-C14:0 and higher of C18:1, total SFA being lower; PUFA were significantly lower, concerning also C18:2 (P≤0.01), while MUFA (especially C18:1) were higher (P≤0.05). In the two experimental groups at TD2 PUFA and particularly C18:2 were higher (P≤0.001). In the buffaloes fed hay C18:1 and C18:3 were also established higher (P≤0.001) with noticeably lower SFA, as compared to silage diet. It was contrary in the bovine cows, the experimental group having higher C8:0C15:0 (P≤0.001) and lower UFA portion. Despite having similar FA profile to the bovine milk, the bubaline milk displayed better n6/n3 ratio; but lower PUFA fraction with tendency to improve, together with MUFA, when roughage source is shifted from silage to hay.
Key words: Bovine, Buffalo, Fatty acids, Hay, Silage