Scientific Papers

Evaluation of ruminal methane and ammonia formation and microbiota composition as affected by supplements based on mixtures of tannins and essential oils using Rusitec | Journal of Animal Science and Biotechnology

Description of Image

Currently, only a few studies focused on the use of a mixture of compounds as supplements instead of testing the single compounds, building on a potential complementary and thus associative effect. Accordingly, blends of EOC [16], a mixture of tannin and saponin extracts [43], mixtures of different types of tannin extracts [44, 45], or mixtures of EOC and tannins [14] were found to be more efficient in reducing methane or ammonia production compared to the supplementation of the individual compounds. Since the dosage and proportion between the active compounds is one of the main issues in the evaluation of the effectiveness of natural compounds as a mitigating strategy [17], the dosages of tannins and EOC blends supplied were adopted from the previous short-term in vitro experiment, expressed as gram per unit of the incubation liquid (133 mg/L of tannin extracts in addition to 100 mg/L of EOC blend; [14]). It has to be noted that the ratio of diet to incubation liquid might differ between in vitro systems, meaning that the dosage per unit of DM can be different. In the present study, the dosage per unit of DM was halved compared to the previous experiment (10 instead of 20 g tannin extracts and 7.5 instead of 15 g EOC blend per kg diet DM). The variability of the feed-to-medium ratio between the different in vitro setups has been previously considered [46], but still, this aspect has been poorly investigated in terms of determining how the supplements are dosed, and further research is needed in this sense.

To evaluate the efficacy of Q-2 and C-10, the promising mixtures selected [14], we included both a negative control and a positive control. The latter consisted of a commercial mixture of EOC that had been previously shown with a meta-analysis [10] to be able to reduce methane emissions under in vivo conditions. Therefore, and for mechanistic reasons, the effects of the newly formulated supplements are discussed separately in the following sections, in comparison to NC (basic efficiency of the mixtures) and PC (comparative effects with a known agent; here blend of EOC).

Diets Q-2 and C-10 vs. negative control diet

The usage of tannins as supplements has been repeatedly reported to adversely affect the methanogenic and fibrolytic activities in the rumen under in vivo and in vitro conditions [1, 47, 48]. In particular, tannins disrupt more Gram-positive than Gram-negative bacteria [1]. Essential oils were described as having a wide spectrum of anti-microbial activities. Due to their lipophilic nature and generally low molecular weight, they can penetrate the outer membranes of Gram-positive bacteria and protozoa and even the external membranes of Gram-negative bacteria (only low molecular weight EO) and thus cause disruption [13, 49].

To the best of our knowledge, only one other in vitro previous attempt (apart from Foggi et al. [14]) has been made to test the effect of a mixture of tannins (chestnut extract) in combination with an EO blend (bioflavonoids extracted from olives) [50]. In that study, the in vitro fermentation was conducted using the biochemical methane potential assay. The inoculum was not rumen fluid, but anaerobic mud. Moreover, the substrate fermented constituted exclusively of anhydrous glucose and no fibrous material. Therefore, the comparability with ruminal conditions is very limited. In the literature, also the availability of in vivo studies is limited. Recently, Atzori et al. [51] supplemented 1 g/d of a blend of essential oils, bioflavonoids and chestnut tannins to Sarda sheep and reported a promising mitigating effect on methane yield (−13%), even though no effect was observed on absolute methane production.

In the present study, the supplement Q-2, based on quebracho tannins and carvacrol, thymol and eugenol as EOC, was effective in substantially mitigating ammonia formation (−37%). Its reduction potential was larger than the one that had been found with Q-2 in the short-term in vitro study [14]. The effect of supplement C-10, based on chestnut tannins and EO from oregano, thyme and citrus peel, on ammonia production was also significant, but at lesser extent (−26%, similar to the 25% found previously [14]). Ammonia in the rumen can be produced by various microbial species with low specificity. In addition, a small group of microbes producing ammonia have a high specificity for substrates containing nitrogen, the so-called hyper ammonia producers [3]. In literature, there is evidence about the efficacy of EO (a blend [3] or oregano EO [52] to selectively reduce some hyper ammonia producers. Nevertheless, other EO blends seemed to be inefficient in inhibiting hyper ammonia producers (e.g., Clostridium aminophilum [5]). There might be several reasons for the greater efficiency of Q-2 compared to C-10 in mitigating ruminal ammonia production in the present study. Accordingly, a synergistic effect of specific EOC blend components of Q-2 and quebracho tannins might have a mitigating effect on specific bacteria populations. Interestingly, a lower relative abundance of Endomicrobium was found in the fluid from Q-2 treatment. Since such a bacterium was described as having an unusual nitrogenase, the decreasing of its relative abundance in the microbiota might be at least partially responsible for the lower ammonia production in Q-2, as similarly found by Mavrommatis et al. [53] supplementing a marine microalga. Another reason for lower ammonia concentration in Q-2 and C-10 incubation fluid might be the lowering of dCP compared to NC treatment. However, a different efficiency in reducing CP degradability between Q-2 and C-10 was, however, not found in the present study. Despite that, Q-2 rumen liquor had a lower BCFA concentration (mmol/L), which together with the higher depression of ammonia, suggested a reduced deamination of certain amino acids (leucine and valine) compared to C-10 [54].

In the present study, there was only a numerical decline in methane formation by 7% with Q-2, whereas in the previous in vitro experiment the reduction in methane formation accounted for 14% [14]. On the other hand, the significant methane mitigation caused by C-10 by 12% was more similar to the one previously found (14%), even though archaeal abundance did not differ (Fig. 6), in accordance with what was reported previously when CT or HT were supplemented in vitro [55]. Overall, there were no differences in the microbial community (at least at genus level), in support of the different methane reduction extent. As such, the outcomes are not coherent with a previous study, in which a target genus (i.e., Prevotella) was found as greater abundant in rumen liquor of low buffalo emitters [56] or in rumen liquor supplemented with CT or HT tannins [55]. However, direct action against methanogens mediated either by EOC [16], by fractions obtained by HT hydrolysis [20] or by their combination cannot be fully excluded, since 16S sequencing may fail in identifying less abundant taxa but biologically meaningful [57].

Fig. 6
figure 6

Relative abundances of Euryarchaeota phylum (total of the Archaea in the present study). a,bDifferent superscript letters indicate significant differences (P < 0.05) between treatments. NC, negative control; PC, positive control

The fibre-degrading microorganisms are the main responsible for the hydrogen production which, in turn, is converted to methane by methanogenic archaea. Tannins, especially CT [58], are known to decrease fibre degradability. As such, the adverse effect of both mixtures on fibre degradability (Table 2) might have contributed to the methane mitigation in the present study. Indeed, CH4 produced per unit of disappeared NDF was higher for both mixtures in comparison to NC. However, Q-2 and C-10 did not result in a lower relative abundance of characteristic fibrolytic bacteria (Fibrobacter, Butyrivibrio, Prevotella; [49]). It can be only assumed that the reduction in the relative abundance of fungal Orpinomyces (in tendency with both Q-2 and C-10, Table S4) might indicate one reason for the reduction of fibre degradability. Indeed, a tendency to adversely affect rumen anaerobic fungi, which primarily degrade fibre components, by supplementing eucalyptus oil and mulethi root aqueous extract as an additive was previously described in rumen liquor of buffalo [59] and in vitro with the supplementation of chestnut or quebracho extracts [20]. However, in contrast to the above, the supplements (significant only for Q-2) even enhanced the abundance of Neocallimastix; thus, it is evident that several factors concurred to determine the outcome. Moreover, an effect on unclassified fibrolytic bacteria or anaerobic fungi cannot be excluded.

The more pronounced reduction of fibre degradability with the extract containing CT (Q-2) compared to HT (C-10) was associated with the lowest production of acetate (the main end-product of fibrous degradation) as previously reported [18]. Consequently, with Q-2 less total VFA were produced, partially confirming what was previously found in our in vitro screening study [14]. Similar to the present study, Buccioni et al. [60] reported that quebracho extract decreased acetate production along with the total VFA produced in the rumen of dairy sheep. The lower acetate production did not decrease the hydrogen concentration in the gas phase of Q-2 treatment sufficiently enough to clearly affect methanogenesis [61], as confirmed by the data about methane production for Q-2 treatment.

Diets Q-2 and C-10 vs. positive control diet

Outcomes from various in vitro studies may vary largely [62]. The inclusion of a common supplement as a positive control might thus help in comparing studies with different in vitro setups. Especially monensin was recommended for this purpose in studies with dairy [55], beef cattle [63], and in vitro continuous culture [64]. However, the active principle of monensin is far from that of phenolic compounds or EO. Therefore, in the present study, Agolin Ruminant, a commercial additive also based on a blend of EOC, namely coriander oil, geranyl acetate and eugenol, was employed as a positive control. In order to facilitate comparability, the dosage of the EOC from this supplement was kept the same as with Q-2 and C-10.

When compared with PC, Q-2 and C-10 had a similar effect on either ammonia (−37% by Q-2, −34% by PC) or methane emissions (−12% by C-10 and PC). This was expected being Q-2 and C-10 formulated with the same criteria (i.e., tannin extract + 3-way EOC) and having some ingredients in common (i.e., carvacrol and thymol were included in both Q-2 and C-10). Unlike PC, Q-2 and C-10 did not affect either protozoa count or hydrogen concentration in the fermentation gas. This observation suggests different mechanisms of action of Q-2 and C-10 vs. PC. In contrast to what was previously measured in rumen liquor supplemented by either CT or HT [65], it is evident that PC had a mechanism far from that suggested for tannin extracts. As such, tannins have been reported favouring Prevotella and its metabolism, which produces propionate and consumes hydrogen [65], on the contrary, propionate proportion, as well as Prevotella abundance, decreased in PC rumen liquor.

Chemical structure is decisive for activity in ruminal fermentation, and the antimicrobial activities of several EOCs with a wide chemical spectrum hence makes predicting the regulatory effect on rumen fermentation often difficult [13, 49]. At present, specific mechanisms of action have been proposed only for a few individual EOC compounds [3] and commercial products, which have already been validated for their mitigating effect in vivo (such as the product used for PC treatment) but were limitedly characterised for the mechanisms of action [10, 11]. In the present study, the severe defaunation caused by PC might partly explain the significant reduction in the production of ammonia, methane and total VFA (−17%) in comparison to NC. Moreover, the decline in methane formation with PC was associated with the lowest hydrogen concentration in the gas phase of all groups corroborating the explanation for the methane mitigation achieved with this product [61]. The ciliate protozoa may represent up to 50% of the microbial biomass in the rumen [66], in which they engulf organic matter particles and bacteria, and actively degrade fibre and other nutrients to produce VFA (particularly acetate and butyrate) and large amounts of hydrogen [1, 66, 67]. The latter is the reason for intensive associations of methanogens with the protozoa. Indeed, in the present study, archaeal abundance was lower in PC in comparison to Q-2 and C-10. The highly disruptive effect on the protozoa population as found with PC was not previously reported with Agolin Ruminant in vitro, at least not at this severity (only −15% according to a meta-analysis [10]). Still, a recent study proposed the change of relative abundance of protozoa as the principal mechanism of Agolin Ruminant in reducing in vivo methane [11]; besides, they did not report a significant effect on Archaea, which might have furtherly supported their methane reduction (−8.8% CH4/kg DM intake, [11]) and despite the present findings.

Compared to PC, dDM and dOM (expressed as % of OM supplied) were greater for Q-2 and C-10, whereas the degradability of NDF was smaller (Q-2) or tended to be smaller (C-10). Overall, this pattern of degradability was probably due to the inhibitory effect of PC on microorganisms degrading non-structural carbohydrates, including protozoa. As a matter of fact, Butyrivibrio, Lachnospira and Treponema (from 2 to 5-fold higher in PC than in Q-2 and C-10), which are known as fibre degrading bacteria [1, 49], were reported herein as having a higher relative abundance. However, it must be noted that NDF degradability was reduced by all treatments compared to NC, and, in the case of PC, this might have been partially due to the lower abundances of the Fibrobacter genus (−95%) and the Prevotellaceae family (−28%) in comparison to Q-2 and C-10. On the other hand, fungi relative abundance seemed not directly involved. Overall, the depression of feed degradation caused by PC and Q-2 compared to C-10 and NC was supported by the concomitant depression of total VFA produced. Unlike the present study, previous in vivo and in vitro studies [9, 10, 68] did not find a significant effect of Agolin on VFA production, maybe due to the lower concentration tested. The lowest molar proportion of propionate was found for PC, confirming what was previously reported by Pirondini et al. [68], using Agolin Ruminant. The lowest proportion of acetate was associated with the Q-2 treatment. Consequently, the acetate-to-propionate ratio was highest with PC, pointing towards a lower energetic value of the substrates obtained from fermentation in the animal. This modification of the VFA profile by PC was associated with a decrease in the succinate producers (i.e., Fibrobacter), usually considered to be involved in the fermentative pathway of propionate [69]. It is known that the effects on molar proportions of VFA vary between types of EO supplemented [15, 16], but the present study indicates that duration of the fermentation, diet composition and especially EOC dosage affects the exhibition and extent of such effects in vitro.

Some of the effects on nutrient degradability were not supported by the expected changes in abundance of the respective families or genera of bacteria or fungi. However, it has to be noted that the applied sequencing cannot define enough accurately the abundance of targeted species of individual microbe species. Therefore, a decrease in key fibrolytic bacteria species reported previously [60, 70] could not sufficiently be distinguished. In addition, the complex interaction of the EOC with the tannins, and of CT and HT with the feed particles, as they could have been complexed nutrients differently, might have caused differences in fibre degradability, methane and ammonia formation, which are not clearly reflected in the composition of the microbial communities [48].

Finally, the biodiversity (both richness and evenness) of the prokaryotes was highly reduced by PC (Fig. 1), if compared to both Q-2 and C-10 treatments, probably due to the far-going defaunation [71]. A similar, however less severe, trend was also reported for fungal richness (Fig. 2). Similarly, a recent study reported a lower α-diversity of rumen microbiome of dairy cows supplemented with 1 g/d of Agolin Ruminant [11]. The robustness and resilience of microbiota is a biomarker for animal health, and it increases with higher α-diversity and network complexity [72]. Thus, if the present in vitro data will be confirmed by in vivo experiments, the severe defaunation observed in the present study at the dosage of Agolin Ruminant used could cause substantial changes in the microbiota community and an undesired decline of the bacterial and fungal richness.

Description of Image

Source link