Modulation of ruminal biohydrogenation in sheep through dietary tannins or energy source

In the present thesis, four experiments were conducted to study how ruminal biohydrogenation pathways can be modulated through dietary inclusion of tannin sources and to acquire a better comprehension about the occurence of t10-shifted biohydrogenation pathways. In the first experiment, in vitro batch incubations with 100 g/kg dry matter (DM) of extracts of chestnut tannins (mostly hydrolysable tannins) and quebracho, grape seed or rockrose (Cistus ladanifer) condensed tannins, as well as a control treatment were incubated for 6 h with ruminal fluid from fistulated sheep and a dehydrated lucerne-based substrate with 60 g/kg DM of sunflower oil. Grape seed and, to a lesser extent, C. ladanifer led to a higher disappearance of 18:2n-6 with a consequent higher production of c9,t11-18:2 and t11-18:1 than chestnut, quebracho and control. There was no clear innibition of 18:0 production with any of the extracts comparing with control. In the second experiment, rumen fistulated sheep were fed tannin extracts from mimosa condensed tannins, chestnut hydrolysable tannins or their mixture (100 g/kg DM) in a complete diet with sunflower and linseed oils (40 g/kg DM), following a change-over design (3 treatments, 4 sheep and 4 periods). There was a variable inhibition of ruminal biohydrogenation and a lower "trans-/cis-18:1" ratio in bacterial fractions with mimosa than with chestnut. Mimosa led to a lower fermentative activity, as well as a lower abundance of Fibrobacter succinogenes, Ruminococcus albus, Ruminococcus flavefaciens and Butyrivibrio proteoclasticus and higher abundance of Selenomonas ruminantium with a lower bacterial biomass estimate of dimethylacetals than chestnut. In the third experiment, two rumen fistulated rams were housed in metabolic cages and adapted to a wheat-based diet with 41 g/kg DM of sunflower oil. During the first two weeks of trial, the t10-shift occurred temporarily in both animals but in different moments. These results were probably due to individual variability of rumen microbiota, since, for a selected period of the trial, a lower bacterial diversity was found for ram 1 compared to ram 2. Moreover, the t10-shift was associated with an increase of total trans-18:1 and a decrease of 18:0. There was no clear association of t10-shift with rumen pH or its expression in blood plasma. In the fourth experiment, 40 lambs were fed, for 6 weeks, with complete diets containing barley or barley completely replaced for dehydrated citrus pulp, dehydrated beet pulp or soybean hulls. All diets were supplemented with an oil blend (soybean:fish oils, 59:10 g/kg DM). Overall, the t10-/t11-18:1 ratio was above 3 in meat and subcutaneous fat, although soybean hulls increased t11-18:1 and c9,t11-18:2 comparing with the other treatments. Citrus pulp led to the lowest gene expression of fatty acid synthase, while that of stearoyl-CoA desaturase was inferior for soybean hulls and beet pulp.