**4. Discussion**

Multidisciplinary studies (agronomic, nutritional, parasitological, and chemical) have identified many benefits to animal health of tanniferous forages and legumes used as feed for ruminants [40,41]. Sainfoin generally contains mainly CTs (approximately 42–50 g CTs/kg DM), which have been well studied [42–44]. The consumption of sainfoin disturbs various stages of parasitic life cycles, mainly due to its high tannin content [45]. The anthelmintic activity of CTs, however, can be increased by the addition of flavonoids, which also interfere with the biology of GINs [15,16]. We therefore focused on the analysis of flavonoids and phenolic acids.

Quantitative UHRMS analyses of the bioactive components in the SFPs identified more than 32.0 g/kg DM flavonoids and 4.5 g/kg DM phenolic acids. The main flavonoid was rutin (18.92 mg/g DM), which has multiple pharmacological activities with metabolic

health benefits [46]. Rutin can alter the ruminal microbiome and reduce the population of methanogenic bacteria [47]; and adding rutin (3.0 mg/kg) to feed dairy cows for 11 weeks improved the efficiency of carbohydrate fermentation in the rumen and the ability to synthesize protein [48]. Our results with SFPs indicated reductions in methane concentrations and *Archaea* population sizes of 11 and 33% and 27 and 38% in in vitro and in vivo treatments, respectively. Ruminal contents in the SFP group indicated also smaller populations of *Methanomicrobiales* (in vitro and in vivo) and *Methanobacteriales* (in vivo) than the control group. In another study, dietary supplementation with dry fumitory, mallow, wormwood, and chamomile with flavonoids (0.4–12.2 g/kg DM) for 70 d had no antimethanogenic effect in lambs [19]. The anti-methanogenic effect of SFPs observed in our study may have been due to the direct effect of either the CTs [6,49] or rutin [50] or both on methanogenesis in the SFP group. Data on reducing methane concentration by rutin supplementation in vitro and in vivo, however, have been inconsistent [51–53]. Acceptably low methane emissions can therefore be achieved by a suitable choice of the vegetative stage of sainfoin [54]. The stimulation of gas concentration by rutin (50 mg/g DM) in ruminal fermentation in vitro led to an increased CO2 concentration and a decreased methane concentration, probably because rutin is a substrate for nonmethanogenic microbiota [50]. However, the consumption of SFPs by infected lambs for 14 d in our experiment affected ruminal methanogens and consequently reduced methane emission without adverse changes in the ruminal microbiome.

The relative abundance of *B. fibrisolvens* in our study was higher in the SFP than the control group, because the replacement of MH by SFPs was probably associated with the increased demand for microbial degradation of fiber in the SFP group. Other bacterial species were not significantly affected, probably due to the relatively short SFP treatment. *B. fibrisolvens* plays an important role in the ruminal fermentation of polysaccharides that participate in cellulolytic processes in the rumen but do not have an autonomous cellulolytic capacity [55]. Changes in the relative abundance of *B. fibrisolvens* sugges<sup>t</sup> that the SFPs did not negatively affect ruminal fermentation (e.g., total and individual VFAs). A diet with plant bioactive compounds can affect the ruminal microbiome, the kinetics of fermentation, and the response and adaptation to anti-methanogenic compounds, sometimes leading to inconsistent efficacy of phytochemicals [56–58]. We cannot sufficiently confirm the effect of SFPs on ruminal fermentation in lambs, because the SFPs were consumed for only 14 d. We also observed no effects of short-term SFP feeding on total protozoal counts in vitro and in vivo. The in vitro experiments on the effects of CT fractions of differing molecular weights on *Leucaena leucocephala* identified a lowering of the total number of ciliate protozoa with changes in counts of community members [59]. Similar effects were observed in vivo after long-term feeding of lambs with extracts of *Acacia negra* and *Uncaria gambir*, sources rich in CTs [60]. The feeding of dry leaves of *L. leucocephala* (12–36% of DM intake), however, did not affect ruminal protozoan, bacterial, or archaeal populations in crossbred heifers [61]. Supplementation of rutin for three weeks in dairy cows (3.0 mg/kg diet) did not significantly decrease counts of ruminal protozoa [49]. The amount and composition of the CTs and the length of treatment are likely the main factors influencing the effects of CT on the ruminal microbiome. In our experiment, even feeding only SFPs did not affect the protozoan counts. We can speculate about the amount of intake of CTs in the diet of the control group. Several studies have reported the CT compositions of various kinds of forages and MH from permanent pastures [6,62]. Unfortunately, we did not measure the CT content of the control MH diet.

Our study confirmed a significant reduction in RBC count, HGB level, and HCT from D23 in both groups of lambs infected with *H. contortus*, consistent with our previous results [20,28,63]. The reductions were likely due to damage caused by the GIN parasites, but SFPs as a replacement for hay did not affect the RBC parameters. The intensity and duration of hematological disorders depend on the nutritional status of infected sheep because protein-enriched diets induce resistance to infection associated with the improved regenerative capacity of bone marrow [64]. In addition to the nutritional status of the

host, differences in RBC disorders during GIN infection may be affected by the species of nematode, the severity of the infection, the iron stores and bodily reserves of the host, and the susceptibility of the host breed [65]. The basophil level in our study was higher in the infected lambs of the SFP group. Basophils are generally relatively rare and short-lived cells and probably played an important role in the immune response in the SFP group to GIN infection in our experiment [66]. The number of monocytes, eosinophils, and basophils was differentially affected by the time after infection, consistent with the blood variables during a subclinical *H. contortus* infection [67].

The development of GINs in host abomasa causes pathology, with mucosal damage and gastropathy with protein loss, followed by inflammatory immune responses of the host [68]. Our observations included microscopic changes in the abomasa, such as mucosal hypertrophy, damage to epithelial cells, mucus-producing cell hyperplasia, glandular dilatation, glandular damage, inflammatory cell infiltration, and submucosal edema. A roughened and hyperemic abomasal mucosa with enlarged glands and globular leukocytes have been described in lambs infected with *H. contortus* [69]. In our experiment, glandular damage in the SFP group differed significantly between the experimental groups, but the other changes in the abomasal mucosa were essentially the same for both infected groups. Mucosal hypertrophy was also more pronounced but not significant in the lambs fed with SFPs. An increased percentage of abomasal injuries and mucosal hypertrophy in the SFP group was attributed to regeneration, which is more common in abomasal tissue due to herbal treatment [70]. The histopathological changes in lambs in our previous study infected with *H. contortus* and supplemented with dry *Artemisia absinthium* and *Malva sylvestris* were predominantly on the mucosal membrane, with inflammatory cell infiltrates (mainly lymphocytes and macrophages with a mixture of eosinophils, plasma cells, and mast cells) [70]. Subclinical *H. contortus* infections generally damage the abomasal mucosa, which was very similar in the two infected groups, i.e., with and without SFPs.
