Feed additives as a strategic approach to reduce enteric methane production in cattle: modes of action, effectiveness and safety

Keywords: greenhouse gases, methanogens, rumen function, ruminants.

The livestock sector is crucial for food and nutrition security globally, with a projected increase of 80% in consumer demand by 2050 for beef (Nadathur et al. 2017). Approximately 83% of global milk is produced from cattle (Visioli and Strata 2014) and, by the end of the decade, milk output is anticipated to have grown by 33% and 9% in developing and developed countries respectively (OECD/FAO 2018). Globally, beef is the third-most consumed meat, contributing 320 million tons of product to world food supply, representing 79% of total sourced meat (Opio et al. 2013; Ritchie and Roser 2019). Nutritional benefits from ruminants are pronounced as they have the ability to convert fibre-dense forages that are indigestible to humans into high-quality bioavailable nutrient sources. In fact, 86% of the feed consumed by livestock worldwide is not considered edible for human consumption (Mottet et al. 2017). At the same time, ruminants occupy more land than do any other livestock species and their enteric methane (CH₄) emissions contribute to total anthropogenic greenhouse gases (GHG; Knapp et al. 2014). Enteric CH₄ is under increased scrutiny due to its heightened potency compared with carbon dioxide (CO₂) in the atmosphere, and the 39% it contributes to the sector’s total emissions (Gerber et al. 2013; IPCC 2013).

Heightened attention on climate change by scientists, governments and consumers is challenging the livestock industry to reduce GHG emissions. Arguments for consumers to shift towards plant-based diets have gained traction; however, constructing diets on the basis of the level of GHG emissions will not necessarily have a positive correlation with nutritional provision (Payne et al. 2016). Dietary manipulation has been studied over the past few decades as a strategy to reduce enteric CH₄ emissions and could be assimilated into management practices, notably through feed additives (Cottle et al. 2011). Feed additives are used in livestock diets to improve feed-use efficiency, quality of animal-source foods, and animal performance and health. These additives include vitamins, amino acids, fatty acids, minerals, pharmaceutical compounds, fungal products and steroidal compounds. Recent advances in understanding methanogenesis have led to the development and discovery of feed additives that can reduce CH₄ emissions to varying degrees. The present review aims to provide a concise summary of feed additives currently available, or in development, with some potential to reduce CH₄ emissions from ruminants. The secondary objective of the review is to summarise information on mode of action, efficacy, safety and readiness for adoption of anti-methanogenic feed additives. Although the focus is on feed additives tested in vivo, some in vitro studies are also discussed if there is paucity of in vivo trials for an additive or to help explain mode of action.

Methane production can be substantial in ruminants, representing up to 12% of gross energy intake that could potentially be utilised for physiological processes, but, instead, is released into the atmosphere through eructation (Beauchemin et al. 2009a). However, CH₄ synthesis represents a significant metabolic sink for reducing equivalents (hydrogen, H₂) that would otherwise accumulate in the rumen and create an unfavourable environment for fermentative digestion processes (Morgavi et al. 2010). Hydrogen itself does not accumulate due to methanogen activity, instead, methanogens participate in interspecies H₂ transfer, and dispose of the reducing equivalents from other metabolic processes (Bergman 1990; McAllister et al. 1996). Hydrogen synthesis is a self-limiting process that relies on separate and distinct reducing equivalent consumption pathways so as to continue production. Cellulose-degrading activity in both bacteria and fungi increases in the presence of methanogens, which contributes to the principle of rumen syntrophic relationships (Bauchop and Mountfort 1981; Sasaki et al. 2012).

Rumen methanogenesis is performed strictly by archaea (Hook et al. 2010). A methanogenesis pathway is presented in a simplified diagram (Fig. 1), which includes the convergence of pathways known to occur in a Methanosarcina spp. Lambie et al. (2015) categorised methanogens on the basis of their metabolic pathways, as follows: hydrogenotrophic, acetoclastic and methylotrophic that can yield CH₄ in the rumen from Methanosarcina spp. Methanogens reduce CO₂ with H₂ (hydrogenotrophic), source a methyl group from acetate (acetoclastic), or a methyl group from compounds such as methanol, methylthiol, dimethylamine, and mono-, di-, tri- methylamine (methylotroph). Formate contributes to methanogenesis as an electron donor within the hydrogenotrophic pathway, representing ~16–18% of CH₄ in batch- and continuous-culture experiments (Seedorf et al. 2014; Ungerfeld 2015; Hungate et al. 1970). Coenzyme M requires a methyl group for the reduction to CH₄, which is provided through each of these pathways. Methane mitigation could be achieved by directly targeting methanogens or modifying the rumen environment to shift the metabolic pathways away from methanogenesis or reduce substrates for the archaea.

Fig. 1.  Simplified methanogenesis pathway from Methanosarcina barkeri CM₁, adapted from Lambie et al. (2015). The three pathways depicted include hydrogenotrophic (carbon dioxide utiliser), acetoclastic (acetate utiliser) and methylotroph (methyl-group utiliser), which all have the potential to donate a methyl group and form methane.

Feed additives classified as CH₄ inhibitors directly act on the methanogenesis pathway (Fig. 1) in a way that can disrupt the process and reduce CH₄ production (g/day). Methanogens prevent H₂ accumulation in the rumen, which otherwise may lead to adverse effects on fibre degradability and animal performance (Ellis et al. 2008). Given the importance of efficient fibre digestion, the use of CH₄ inhibitors must balance between reducing CH₄ production and avoiding negative impacts on animal performance and welfare. Inhibition of methanogenesis requires a redirection of reducing equivalents, H₂ in this case, to alternative sinks, instead of CO₂, unless the inhibitor’s mode of action is a highly competitive electron acceptor. Malik et al. (2015) argued that H₂ clearance through pathways such as reductive acetogenesis and propionogenesis also has the advantage of energy conservation into end products such as meat and milk. Several of these alternative sinks will be reviewed herein and may also be implemented as independent feed additives or with an inhibitor. Studies have shown a decrease in CH₄ emissions paired with an increase in H₂ emissions without the addition of an alternative sink (e.g. Roque et al. 2019b), indicating that elevated H₂ concentration in the rumen may not necessarily result in decreased fermentation, and hence, productivity.

Methyl–coenzyme M reductase (MCR) is the enzyme that catalyses the final step of the methanogenesis pathway from intermediate methyl–CoM to CH₄ as illustrated in Fig. 1. As a nickel enzyme, MCR can catalyse this step only when its Ni ion is in the +1-oxidation state and can be inactivated due to the existing redox potential (Duin et al. 2016). The position of 3NOP binding to an active site of MCR places the reducing nitrate group in close proximity to Ni(I), a distance in which electrons could be transferred. Although 3NOP inhibits methanogenesis and reduces methanogen growth, it does not negatively affect other microbial groups in the rumen (Duin et al. 2016).

More than 15 studies have been conducted using 3NOP, showing a marked reduction of enteric CH₄ emissions with a range of effectiveness. 3NOP added to ruminant diets in small quantities has been shown to persistently reduce enteric CH₄ emissions by inhibiting an important step in the methanogenesis metabolic pathway, without apparent negative side effects (Hristov et al. 2015). Figure 2 shows a forest plot illustrating the effect sizes as a mean difference between the control and treatment-group mean CH₄ production. For example, Vyas et al. (2016) reported that with 0.2 g 3NOP/kg dry-matter (DM) supplementation, CH₄ production in backgrounding and finishing beef cattle reduced 37.6% and 84.3% compared with the control group, whereas Vyas et al. (2018), using the same amount of supplementation in backgrounding phase (0.2 g/kg DM) of beef cattle, found a 54.1% reduction in CH₄ production. These authors reduced the level of supplementation of 3NOP to 0.125 g/kg DM during the finishing phase and reported 53.8% reduction in CH₄ production. There was also an improvement in gain-to-feed ratio during treatment, with a 7% drop in DM intake (DMI). Similarly, Martinez-Fernandez et al. (2018) reported a decrease in CH₄ production of 38% and daily weight gain of 0.571 kg/day compared with the control in steers supplemented with 0.30 g 3NOP/kg DM. Hristov et al. (2015) demonstrated that CH₄ production in lactating cows was reduced by 30% by feeding 3NOP at 0.04–0.08 g/kg DM without affecting feed intake and milk production. Lopes et al. (2016) reported a 31% decrease in CH₄ production in lactating dairy cattle fed diets supplemented with 0.06 g/kg DM. In a meta-analysis of the anti-methanogenic effects of 3NOP, Dijkstra et al. (2018) reported that enteric CH₄ production was reduced 39% in dairy and 22% in beef cattle at a mean dose of 0.123 g/kg DM. Additive dose and the neutral detergent fibre (NDF) content of diet had a significant impact on the effectiveness of 3NOP in reducing enteric CH₄ emissions. Furthermore, an increase in 3NOP dose of 0.010 g/kg DM from the mean dose further reduced CH₄ production by 2.56 ± 0.55%. Similarly, Jayanegara et al. (2018) reported that the methanogenic archaea population was reduced through 3NOP supplementation and the magnitude of reduction was positively correlated with 3NOP dose in small and large ruminants. Addition of 3NOP is also associated with shifting H₂ production in the rumen and results in an increase in molar proportion for propionate and decreases acetate production (Haisan et al. 2014; Kim et al. 2019; Lopes et al. 2016).

Fig. 2.  Forest plot of mean difference (MD) of methane production for different feed additives, counts of studies, minimum and maximum of MD. Only studies conducted in vivo were included in the analysis.

There are no known adverse effects of supplementing 3NOP on the animal or the subsequent product. The feed additive 3NOP continues to be studied and, after approval by regulatory bodies, it is expected to be on the market in the near future.

Plant species that accumulate halogenic compounds in their tissues have been investigated for their potential to reduce enteric CH₄ emissions. Halogens are elements that hold a large, negative electron affinity and seek to combine with other compounds to reach stability through satisfaction of the valence shell in the rumen environment (Gribble 2004). Bromoform and chloroform are halogens that have been found to interfere directly with the methanogenesis pathway by serving as competitive inhibitors (or analogues) of the MCR, preventing the final catalysis step (Goel et al. 2009). The mode of action is through reacting with reduced vitamin B12 and inhibiting the cobamide-dependent methyl-transferase step of methanogenesis (Wood et al. 1968; Chalupa 1977). The B12-dependent methyl-transferases also play an important role in one carbon metabolism in acetogenic bacteria (Banerjee and Ragsdale 2003), and, therefore, halogenated compounds may have an effect on reductive acetogenesis.

At supplementation level of 1.50–1.59 g/kg DM (2.6 g/100 kg liveweight; mean liveweight = 288 kg) of chloroform– cyclodextrin, steers have demonstrated a 30–35% reduction in enteric CH₄ production, with no detectable differences in rumen fermentability (Martinez-Fernandez et al. 2016). Steers dosed daily with 0.267 g/kg DM of chloroform were shown to decrease 94–95% of CH₄ production within 4–5 days of treatment. However, CH₄ production has been shown to slowly recover to 62% of the pre-treatment levels by Day 42 of treatment (Knight et al. 2011). The macroalgae species Asparagopsis taxiformis and A. armata have been evaluated for their mitigation potential both in vitro and in vivo (Roque et al. 2019a, 2019b). Asparagopsis spp. contain relatively high concentrations of bromoform and other halogenated compounds such as bromochloromethane (Paul et al. 2006; Machado et al. 2016). An in vitro trial analysing effectiveness across seaweed species found A. taxiformis to be the most effective species among 20 freshwater and marine macroalgae in reducing CH₄ output (98.9%), but also reduced total gas production (62%), likely indicating inhibition of digestion (Machado et al. 2016). Increasing the dose to 5% in vitro, Roque et al. (2019a) reported a 95% reduction in the level of CH₄ production. Three papers have been published so far, reporting the effect of Asparagopsis spp. in sheep, dairy and beef cattle in vivo. Li et al. (2018) supplemented A. taxiformis at 67.5 g/kg DM (30 g/kg of organic matter, OM) in sheep diets and reported a reduction of up to 80% in enteric CH₄ production. However, rumen volatile fatty acid (VFA) concentrations in the 0%, 0.5%, 1.0%, 2.0% and 3.0% macroalgae inclusion groups declined from 92.0, to 86.5, 74.9, 69.1 and 65.4 mM respectively. Reductions in VFA concentrations are not desirable as they provide energy to the ruminant. In lactating dairy cattle, Roque et al. (2019b) observed up to 67.2% reduction in CH₄ intensity (g/kg milk produced) using A. armata at an inclusion rate of 18.3 g/kg DM (10 g/kg of OM). In Brangus beef cattle, Kinley et al. (2020) reported a reduction of enteric CH₄ production of up to 98% by supplementing a feedlot diet with A. taxiformis at 3.26 g/kg DM (2 g/kg of OM). In addition, there was an improvement of 42% in average daily gain with a supplementation level of 1.63 g/kg DM (1.0 g/kg of OM) and it went up to 53% at an inclusion rate of 3.26 g/kg DM (2.0 g/kg of OM). The study by Kinley et al. (2020) reported a greater effectiveness at a lower dose than did that of Roque et al. (2019b), which was likely due to the large differences in the bromoform concentration in A. taxiformis and A. armata, while also acknowledging the inclusion of monensin in the Kinley et al. (2020) experimental diets. The bromoform concentration in Roque et al. (2019b) study was 1.32 mg/g compared with 6.55 mg/g in the Kinley et al. (2020) study.

Sourcing naturally occurring halogens circumvents the need to use synthetic halogens. Historically, these synthetics have had detrimental effects on the environment (Gribble 2004). Kinley et al. (2020) and Roque et al. (2019b) tested for residual bromoform content in meat (or edible offal) and milk respectively. In both cases, concentrations of bromoform were either undetectable or not significantly different from the control, suggesting no safety issues arising from the active ingredient. At present, A. taxiformis is not produced commercially; so, accessibility is an issue. The use of macroalgae also needs to be approved by regulatory agencies before widespread use by producers.

Adding nitrate to ruminant diets can be an effective CH₄ mitigation strategy because nitrate competes with methanogens for H₂ in the rumen. Nitrate (NO₃^–) is reduced to nitrite (NO₂⁻; NO₃⁻ + H₂ → NO₂⁻ + H₂O) and further to ammonia (NH₄⁺; NO₂⁻ + 3H₂ + 2H⁺ → NH₄⁺ + 2H₂O) by rumen microbes. However, small quantities of nitrous oxide may also be produced (Latham et al. 2016). This pathway is highly competitive with methanogens for H₂ utilisation in the rumen due to greater changes in Gibbs energy than with methanogenesis (CO₂ + 4H₂ → CH₄ + 2H₂O) pathway (Villar et al. 2020). The result is a redirection of H⁺ flow from CO₂ to nitrate reduction, thereby reducing the generation of CH₄ (Olijhoek et al. 2016).

About 24 in vivo studies showed that the efficacy of nitrate additives varied widely, ranging from +1.25% to −29.8%, and may be affected by several factors. A meta-analysis conducted by Feng et al. (2020) investigated the potential explanatory variables for anti-methanogenic effects of in vivo nitrate supplementation in cattle. These included DMI, roughage proportion, NDF content, crude protein (CP) content, bodyweight, nitrate dose, cattle type, and CH₄ measurement methods. The authors reported that nitrate significantly reduced CH₄ emissions in a dose–response manner and the mitigating effect increased with the level of nitrate inclusion. Methane production reduced 14.6% in cattle supplemented with nitrate at 17.7 g/kg DM (Feng et al. 2020). Hulshof et al. (2012) reported that nitrate supplementation increased ammonia-nitrogen concentrations in the rumen by 34%, decreased propionate concentrations by 16%, but did not affect the total VFA concentrations. Persistency of nitrate was tested by van Zijderveld et al. (2011a), by including 21 g/kg DM during four successive 24-day periods and a consistent 16% reduction in daily CH₄ production (g/day) and yield (g CH₄/kg DMI) was demonstrated. An additive effect of nitrate and linseed oil was reported by Guyader et al. (2015a) in multiparous, non-lactating dairy cattle. These authors reported that adding 4% linseed oil to 3% calcium nitrate further reduced CH₄ production from 22.8% (nitrate only) to 33.0%.

Concerns about the toxicity of the intermediate product of nitrate, namely nitrite, to ruminants necessitate management, as animal poisoning may occur via methaemoglobinemia (Latham et al. 2016). Nitrite is toxic in blood because it converts haemoglobin to methaemoglobin, which is incapable of carrying oxygen. Blood methaemoglobin concentrations in ruminants increase with a greater nitrate consumption and could cause nitrate poisoning (Lee and Beauchemin 2014). Apparent nitrate-poisoning symptoms such as depressed feed intake, slow or no weight gain, reproduction failure, respiratory distress, coma and death have been reported in previous studies with methaemoglobin concentrations of 30–40% of total haemoglobin (Bruning-Fann and Kaneene 1993). Lee and Beauchemin (2014) discussed several critical factors related to nitrate toxicity, including the dietary nitrate concentrations, nitrate consumption rate, incomplete reduction of nitrate and nitrite to ammonia, and rumen outflow rates. Toxic effects of nitrite on the populations of main cellulolytic bacteria, which may be caused by the negative effects of nitrate/nitrite on cellulolytic and xylanolytic activity, have also been observed (Iwamoto et al. 2002; Asanuma et al. 2015; Granja-Salcedo et al. 2019). However, the risk of nitrate toxicity can be reduced by gradual acclimation of ruminants to dietary nitrate or utilisation of encapsulated nitrate (Lee and Beauchemin 2014). Currently, nitrate inclusion may not be advisable in commercial operations due to its potential toxicity. However, a denitrifying probiotic, Paenibacillus fortis, that can enhance nitrite detoxification in nitrate treated ruminants, has been identified (Latham et al. 2019). If successful, nitrate and the probiotic might be a practical mitigation strategy to reduce CH₄ production from ruminants.

The rumen environment can be modified with feed additives to limit the growth of methanogens and to suppress CH₄ production, without targeting the specific methanogenesis pathway. The factors influencing CH₄ production include those involved in H₂ and carbohydrate metabolism (Morgavi et al. 2010). Understanding rumen metabolic processes that affect CH₄ formation is still advancing; however, feed additives were used to modify the rumen environment to reduce CH₄ production without compromising animal health or productivity. This section discusses feed additives that can potentially reduce CH₄ production by modifying the rumen environment.

Dietary lipids modify the rumen environment in several ways, including (1) toxic characteristics on methanogens and protozoa, (2) hydrogenation of unsaturated fatty acids (alternative H₂ sink) and (3) shifts to propionic production, leading to reduction of enteric CH₄ production (Johnson and Johnson 1995; Beauchemin et al. 2008, 2009b). Efficacy of lipids to reduce CH₄ emissions are dependent on the form and level of supplementation, as well as the source and fatty acid profile (Beauchemin et al. 2008; Eugène et al. 2008). Several meta-analyses were conducted to estimate the impact of dietary lipids on CH₄ production (e.g. Beauchemin et al. 2008; Eugène et al. 2008; Martin et al. 2010). For example, Beauchemin et al. (2008) evaluated 17 studies in sheep, beef and dairy cattle and reported a 5.6% reduction in CH₄ production for every 1% additional inclusion of supplemental fat. In dairy cattle, Eugène et al. (2008) reported a decrease of 9% through lipid-supplementation (average 6.4%) compared with control diets (average 2.5%), mostly as a consequence of reduced DMI. Similarly, Patra (2013) reported 3.77% decline in CH₄ emissions for each percentage inclusion of lipid in dairy cattle diets. Prediction inconsistencies by the inclusion of supplemental lipid are likely to be due to differences in lipid source and diet composition. In a review, Rasmussen and Harrison (2011) reported that the most effective fatty acid profiles that reduce CH₄ production were medium-chain (8–16 carbon chains; MCFA) and polyunsaturated (PUFA) fatty acids. However, reductions in DMI due to high levels of dietary lipids are well characterised and ration formulation programs often are set not to exceed 6–7% of total DMI (NRC 2001).

These include lauric, myristic, capric and caprylic acids (Hollmann et al. 2012). In vitro studies have reported coconut oil, which contains 75% of MCFA, to reduce CH₄ production by 43–85% (Dong et al. 1997; Machmüller et al. 1998). Application of coconut oil in in vivo trials also showed similar patterns in CH₄ reduction (Hollmann et al. 2012). Ruminants fed diets containing 13, 27 and 33 g coconut oil/kg DM had 3%, 37% and 45% reduction in CH₄ output compared with the control respectively. DMI, solids-corrected milk yield, and milk fat yield (no difference between the two greatest levels of inclusion on milk fat yield) decreased linearly with an increase in coconut oil application. Inclusion of myristic acid at a rate of 50.0 g/kg DM in dairy cattle diets reduced CH₄ production by 36%, but also reduced milk fat by 2.4%, with a tendency to reduce DMI (Odongo et al. 2007). Lauric acid had no negating effects on methanogenesis in dairy cattle when they received it at 10.0 g/kg DM (Hristov et al. 2009). Within the same trial, the treatment group receiving 21.6 g/kg DM of coconut oil reduced their CH₄ production by 61% compared with the control.

Polyunsaturated fatty acids have also been shown to reduce CH₄ production. For example, Bayat et al. (2015) found that enteric CH₄ production reduced by 29.5% with supplementation of 60 g/kg DM of camelina oil, but other parameters such as milk yield and milk components were compromised. In contrast, Duthie et al. (2018) did not find significant differences in enteric CH₄ production in steers fed increasing amounts of dietary lipid sourced from maize distillers dark grains, which increased diet ether extract from 24 to 37 g/kg DM for 17 weeks. Supplementation of diets with cottonseed oil has been shown to decrease enteric CH₄ production by ~42% (Nogueira et al. 2020). These authors suggested that bio-hydrogenation of lipids served as an alternative H₂ sink, and with each percentage point of lipid added to the diet, CH₄ production was reduced by 8%. Further characterisation and understanding of the impact and longevity of dietary lipid inclusion on methanogenesis would be valuable in selecting plant sources and estimating their impact. Dietary lipid additives (both MCFA and PUFA) show substantial decreases in CH₄ production with a wider range of effectiveness compared with other feed additives (Fig. 2).

Microorganisms included in diets are often referred to as probiotics, cultures, or direct-fed microbials. Introducing microorganisms to a digestive microbiome is practiced on farms to influence the rumen flora for improved digestion. Results of feeding fungi, yeast or bacteria to reduce CH₄ production have not been consistent in studies conducted in vitro or in vivo. Application of live yeast cultures (various strains of Saccharomyces cerevisiae) have not been shown to significantly change CH₄ production, rumen fermentation or apparent total tract nutrient digestibility in dairy cattle (Bayat et al. 2015). Additionally, inclusion of either a dead or live form of S. cerevisiae has little to no impact on nutrient digestibility or rumen fermentation patterns in beef heifers (Vyas et al. 2014). A meta-analysis by Darabighane et al. (2019) using data from 1990–2016 observed no significant reduction in CH₄ production through the use of probiotics.

Introducing propionate-producing bacteria has been evaluated as a possible solution because propionate production consumes H₂ as a reducing equivalent and, thereby, competes with methanogenesis (Ungerfeld 2013). This has not been effective with all strains of bacteria but Propionibacterium thoenii T159 reduced CH₄ production by 20% and increased VFA production by 21% in a study that screened 31 different strains within in vitro models (Chen et al. 2020). However, in lactating primiparous cows, P. freudenreichii 53-W was shown to increase CH₄ production by 27% (Jeyanathan et al. 2019). The mechanisms of reduction in CH₄ production (if any) are still unknown and could be either directly by microbes or indirectly through metabolites that affect the rumen microbiome (Doyle et al. 2019). Jeyanathan et al. (2019) found no effect on CH₄ output when feeding Lactobacillus pentosus D31, and L. bulgaricus D1 in vivo. Currently, there is no concrete evidence that probiotics are an effective method of CH₄ mitigation.

Acetogenesis, or reductive acetogenesis, is another H₂-utilising metabolic pathway in which acetogens utilise CO₂ and H₂ as substrates to produce acetate. While more prevalent in other mammalian guts, acetogens cohabit with methanogens in the rumen, but are either lacking a substantial population density, preferred environment conditions, or the competitiveness to be the favourable pathway of H₂ ‘disposal’ (Joblin 1999). Redirection of H₂ into the acetogenesis pathway to yield acetate would allow the recapture of energy compared with the loss due to methanogenesis. Enhancing this pathway in the rumen may be approached by sourcing acetogens from other ecosystems and transplanting them into the rumen (Gagen et al. 2014) or uncovering a method to enhance the existing rumen acetogen population if they can outcompete native methanogens.

Organic matter that has undergone pyrolysis, commonly known as biochar, has a wide range of impacts on livestock systems due to its unique characteristics. Biochar has been utilised for generations as a remedy for digestive disorders and is sourced by the livestock industry to address issues surrounding animal husbandry, metabolism and waste management (Kalus et al. 2019; Schmidt et al. 2019). Abatement of CH₄ production through the application of biochar has been shown in soil (Yu et al. 2013) and compost (Sonoki et al. 2013). Considering that there is already an existing market for biochar as a beneficial feed additive, in vivo evidence for GHG mitigation will be significant (Schmidt et al. 2019). Possible mechanisms have been elucidated through a study that observed that application of biochar to paddy soils stimulated methanotrophic proteobacteria and reduced CH₄, despite methanogens also being stimulated (Feng et al. 2012). Additionally, biochar may provide a habitat for methanogens or possibly absorb gases when consumed due to its porous nature, but the mechanisms of action for CH₄ mitigation in cattle are not well understood (Terry et al. 2019; Man et al. 2020).

Rice husks sourced for biochar and fed at an inclusion rate of 6 g/kg DM reduced CH₄ production by 22%, increased liveweight gain by 25%, and had no impact on DMI over a 98-day period (Leng et al. 2012). Biochar supplemented at 8 g/kg DM reduced CH₄ production by 9.5% in growing steers and 18.4% in finishing steers (Winders et al. 2019). Contrary to these findings, inclusion levels of ‘pine-enhanced biochar’ at 5, 10 and 20 g/kg DM in the diets of Angus × Hereford heifers did not reduce CH₄ emissions (Terry et al. 2019). However, it altered the microbiota, notably selecting against Fibrobacter species, which is one of the dominating phyla of the rumen responsible for cellulose degradation (Béra-Maillet et al. 2004). The wide variation in effectiveness precludes biochar as proven feed additive to reduce CH₄ production at present. More research, particularly in vivo, is required to understand the conditions under which biochar can mitigate CH₄ production.

Ionophores, such as monensin, alter rumen microbial populations to improve digestive efficiency by depriving methanogens of substrates that are typically provided by Gram-positive bacterial and ciliate protozoal populations (Russell and Strobel 1989; Hook et al. 2010). This fermentation shift favours the production of propionate over acetate, which reduces the amount of H₂ available for methanogens.

A meta-analysis by Appuhamy et al. (2013) quantitatively determined the impact of monensin in cattle. In beef cattle supplemented with monensin at an average monensin dose of 0.032 g/kg DM, CH₄ production was reduced by 19 g/day, which was further reduced as the NDF content of the diet increased. In dairy cattle, CH₄ production was reduced by 6 g/day at the same average dose and was further reduced as the dietary lipid content increased. Appuhamy et al. (2013) concluded that although there were reductions in CH₄ production through supplementation with monensin, the effect was transient, lasting ~6 weeks. In contrast, Benchaar (2020) reported no suppression effect of monensin on CH₄ output when it was administered to dairy cattle (0.024 g/kg DM), but there was an increase in the proportion of a biohydrogenation intermediate, thus altering rumen metabolism patterns.

The antimicrobial nature of ionophores has caused a concern to human health (Guan et al. 2006; Hook et al. 2010). Long-term use of ionophores is limited due to a low efficacy, transient nature and safety concerns.

Plant secondary compounds are primarily synthesised in response to their environmental conditions and not for specific physiological function (Morrissey 2009). Some plant secondary compounds that may possess antimethanogenic properties are variable in composition due to environmental condition in which they are grown. Seasonal variation, pollution, diseases, pests, storage, injuries and pollination activity influence secondary-compound production and composition (Figueiredo et al. 2008). These compounds are not commonly extracted or isolated before feeding to ruminants because of time and cost considerations, which may contribute to their concentrations being inconsistent. These obstacles present a challenge in determining or predicting efficacy.

Tannins are soluble, phenolic compounds that accumulate within plant tissues likely due to ongoing metabolic processes and contribute to the plant defence system (Swanson 2003). The CH₄ mitigation mechanisms of tannins are not well understood but may be due to a combination of factors, including a reduction in fibre digestibility (decrease in H₂ production) or a direct inhibition of methanogens (Tavendale et al. 2005).

Jayanegara et al. (2012) conducted a meta-analysis describing the relationship between rumen CH₄ formation and the level of dietary tannin (hydrolysed or condensed) inclusion between in vivo and in vitro models. These authors reported that low levels of inclusions of tannins in animal experiments often yielded inconsistent results on CH₄ production, but that variability seemed to diminish at higher doses, leading to setting the threshold for detecting treatment differences in animals to be >20 g/kg DM of tanniferous inhibitors. Furthermore, reduction in CH₄ production was often followed by a suppression in OM and fibre digestibility. Methane measurements from goats fed Kobe lespedeza, a forage containing condensed tannins at 151, 101 and 49.9 g/kg of DM led to a 54%, 52% and 32% reduction compared with the control group respectively (Animut et al. 2008). Supplementing beef cattle diets with tannic acid at a 26 g/kg DM inclusion rate, CH₄ production decreased 33.6%, but the digestibility of DM and CP, and the concentration of VFA were negatively affected (Yang et al. 2017). Investigating different tannin-containing hays, Stewart et al. (2019) found small burnet (Sanguisorba minor) fed to Angus cows and heifers to reduce CH₄ production in comparison to a diet containing alfalfa hay (209 vs 289 g CH₄/day respectively). However, CP and DM digestibility was affected negatively.

Grape marc or pomace contains high concentrations of condensed tannins and it is a readily available biowaste from the viticulture industry. Moate et al. (2014) fed dried pelleted (274 g/kg DM) or ensiled grape marc (269 g/kg DM) to dairy cattle and found that the dried form was the most effective in reducing CH₄. The authors reported that the CH₄ production in dairy cattle fed the control, dried and ensiled grape marc was 470, 375 and 389 g CH₄/day. More recently, Caetano et al. (2019) fed ensiled grape marc at a rate of 31.2 g/kg DM, which equates to ~3–4 kg/day of ensiled grape marc (estimated on the basis of reported DMI). Treatment inclusion in the study of Caetano et al. (2019) study led to a 14% reduction in CH₄ production; however, it ultimately decreased the energy availability of the diet due to the greater contents of lignin and acid detergent fibre in the treatment diet. Cattle have exhibited intoxication sensitivities to tannins, particularly if diets do not meet nutrient requirements for growth or milk production. However, such issues can be avoided through appropriate dosages and adaptation periods paired with properly formulated diets (Doce et al. 2013).

Flavonoids are not known to have extensive CH₄ reduction potential, but anti-microbial properties of the compounds have been reviewed (Patra and Saxena 2010). Several in vitro trials (Oskoueian et al. 2013; Kim et al. 2015) have been conducted to gain a better understanding of antimicrobial characteristics and its relation to methanogenesis, but studies utilising in vivo models are scarce. Kim et al. (2015) studied the mitigation potential of four plants containing flavonoids in vitro, by using rumen fluid sourced from a single cow. In all treatments, CH₄ production was reduced by 39–48%; however, results such as this have not yet been translated into animal models. Flavonoids derived from mulberry leaves (~1.3 g/kg DM) did not influence methanogenesis to a detectable level in sheep, but they increased digestibility (Chen et al. 2015). Rutin trihydrate, a flavonoid, was given to dairy cattle at a dose of 100 mg/kg bodyweight, which led to an elevated plasma glucose, β-hydroxybutyrate and albumin, but did not suppress CH₄ production (Stoldt et al. 2016).

Saponins have been studied for their capacity to alter rumen fermentation by reducing protozoal communities, thus lowering H₂ availability and the production of CH₄ (Hess et al. 2003). Saponins are commonly found in low quantities in legume plants such as kidney beans, soya beans, chickpeas and green peas (Shi et al. 2004). Holtshausen et al. (2009) conducted a two-part study on saponins derived from Yucca schidigera and Quillaja saponaria and their effect on CH₄ production in vitro and in vivo. Inclusions of 15, 30, or 45 g/kg DM of Y. schidigera and Q. saponaria decreased CH₄ production ranging from 6 to 26% in vitro. However, in vivo study in dairy cattle using whole-plant Y. schidigera and Q. saponaria powders at 10 g/kg of DM did not show an impact on rumen fermentation. Cross-bred cattle supplemented with soapnut, a saponin-containing plant, did not have significant reductions in CH₄ production (Poornachandra et al. 2019). Tea saponins offered to ewes led to a decrease in CH₄ production if scaled to metabolic weight; otherwise, no differences were observed in absolute values (Liu et al. 2019). The same supplement was offered to steers at 2.44 and 3.85 g/kg DM, but no impact on gas output was observed (Ramírez-Restrepo et al. 2016). Lack of results in reducing CH₄ production may be linked to low concentrations of saponins within additives. However, in some circumstances, high concentrations of saponins have been linked to bloat through foaming properties, but no strong conclusions have been drawn (Lindahl et al. 1954; Sen et al. 1998). Low-level inclusions may have antiprotozoal and mild antibacterial characteristics and can be incorporated into livestock diets through a variety of plant options.

Essential oils (EO) are naturally occurring chemical compounds extracted from plants and used in fragrances and cosmetics and, to a lesser extent, pharmaceutical products for humans and animals. Volatile in nature, the EO contribute to the phenotypic expression of the plant including colour and scent (Edris 2007; Benchaar et al. 2008). Consumption of EO has been observed to affect rumen microbial communities and fermentation patterns in a varying manner, depending on the EO source (Benchaar and Greathead 2011). Many EO hold a high affinity for lipid and bacterial membranes, leading to disruption, but the broad antimicrobial effect is likely to be due to a combination of mechanisms (Helander et al. 1998). EO are non-specific in nature; therefore, there is a concern for their inclusion in diets because they may affect favourable microbe populations, leading to a decrease in feed efficiency. Numerous plants such as cinnamon, lemongrass, ginger, garlic, juniper berries, eucalyptus, thyme, citrus, oregano, mint, rosemary and coriander have been screened in vitro (Becnhaar et al. 2008; Nanon et al. 2015). However, only few have been studied in vivo. Some studies include the whole plant (Olijhoek et al. 2019) into a diet, while other extract the EO before inclusion in a more concentrated treatment (Lejonklev et al. 2016), which introduces another level of variability.

Oregano contains EO compounds carvacrol and thymol that may stimulate general antimicrobial properties in the rumen (Kolling et al. 2018). Only two in vivo studies (Tekippe et al. 2011; Hristov et al. 2013) have shown reduction of CH₄ production of up to 40% in dairy cattle. Hristov et al. (2013) did not observe any adverse effects of supplementation (8.7, 18.9 and 28.2 g Origanum vulgare leaves/kg DM) on feed efficiency, rumen pH or VFA concentrations. In contrast, several other studies have shown no significant impact of supplementing oregano on CH₄ production. For example, lactating dairy cattle supplemented with oregano oil and carvacrol at 0.05 g/kg DM did not express any anti-methanogenic properties (Benchaar 2020). Kolling et al. (2018) reported a reduction in CH₄ yield (in g/kg digestible DMI), but no reductions surrounding other CH₄ emission parameters such as protozoal count, by using 0.56 g oregano extract/kg DM in lactating dairy cattle. Olijhoek et al. (2019) reported no significant reduction in dairy cows supplemented with either 18–53 g oregano plant/kg DM from Origanum vulgare ssp. vulgare containing 0.12% EO of oregano DM, or 7–21 g oregano DM/kg of DM from Origanum vulgare ssp. hirtum containing 4.21% EO of oregano DM. The authors speculated that the differences in reported effectiveness could be related to the duration of measurement (1–8 h post intake in those that reported reductions vs >24 h in studies with no effect). The observation by Hristov et al. (2013) who reported a linear decline in effectiveness after feed intake lends support to measurement duration contributing to differences in reported effectiveness.

Garlic (Allium sativum) contains organosulfur compounds, specifically diallyl disulfide, as its main EO component. Organosulfur compounds are suspected of having a toxic effect on the enzyme system of the methanogenic archaea, inhibiting their activity, while also suppressing protozoal populations (Busquet et al. 2005a; Soliva et al. 2011). Soliva et al. (2011) reported a 91% reduction in CH₄ with 300 mg/L garlic oil in vitro, associated with an increase in bacterial counts and reduction in protozoa. Similarly, Busquet et al. (2005a) observed a 73.6% reduction in CH₄ production in vitro by using similar concentrations of garlic oil. However, most in vivo cattle studies have not found an impact of garlic oil on CH₄ production. For example, van Zijderveld et al. (2011b) used diallyl disulfide at 0.056 g/kg DM in dairy cattle and observed no reduction in CH₄ production. Staerfl et al. (2012) using dried garlic bulbs (treatment standardised for 15 g allicin/kg DM) in feedlot cattle reported no significant effect on CH₄ production measured at 5, 9 and 11 months of age. Similarly, Meale et al. (2014) reported no detectable differences in enteric CH₄ or CO₂ production in animals supplemented with garlic oil (15 g allicin/kg DM). Sheep models have reported similar results of no detectable difference in enteric CH₄ production (Patra et al. 2011; Klevenhusen et al. 2011); however, goat models supplemented with L propyl–propane–thiosulfinate, another organosulfur compound found in garlic, suppressed CH₄ production by roughly 33% (Martinez-Fernandez et al. 2013). Nevertheless, in their subsequent experiment, Martinez-Fernandez et al. (2014), using the same compound in goats in vivo, did not find a significant reduction in enteric CH₄ production.

Lemongrass (Cymbopogon spp.) has been assessed in vitro for potential antimicrobial effects due to citral, an aldehyde sourced from the EO fraction contributing to aromatic characteristic of the plant (Pawar et al. 2014; Joch et al. 2016; Singh et al. 2018). While CH₄ was not measured, Wanapat et al. (2008) detected an improvement in microbial protein supply, DM digestibility and microbial populations when Brahman-native beef cattle consumed 18.5 g lemongrass powder/kg DM. In lactating Barki goats, 4 g/kg DM elicited a slight increase in protozoal counts and CH₄ production (Khattab et al. 2017).

Supplementing cinnamaldehyde and cinnamon oil (containing 78% cinnamaldehyde) to dairy cattle diet at inclusion rates ranging from 0.003 to 0.16 g/kg did not reduce CH₄ production (Benchaar 2016). Methanogen numbers decreased in a study adding 0.5 g/kg DM of cinnamon oil, but the study did not measure gases directly, so any CH₄ reduction was speculative (Khorrami et al. 2015). Eugenol, an active EO component of cinnamon, was added to diets at 0.025, 0.050 or 0.075 g/kg DM, but no treatment group demonstrated a difference in enteric CH₄ compared with the control (Benchaar et al. 2015). Shifts away from acetate production and towards propionogenesis have been observed in artificial conditions when cinnamon-sourced additives were introduced (Busquet et al. 2005b). Inclusion of EO in livestock diets has not rendered any safety concerns for animal husbandry or consumption of subsequent products.

Taking advantage of the unique composition among plants, some studies have used an EO ‘blend’ or ‘complex’ containing extracts from multiple plants. The antimicrobial nature of a variety of the EO may imply a capacity to modify rumen fermentation. EO blends have demonstrated a greater feed efficiency and a higher production of energy-corrected milk in dairy cattle through modification of rumen fermentation (Elcoso et al. 2019; Silva et al. 2020). Blends have become commercially available, typically containing at least two different EO. For example, Agolin Ruminant (Agolin, Bière, Switzerland; AR) contains a blend of eugenol, geranyl acetate and coriander EO. Agolin Ruminant is an antimicrobial EO product and has shown 20% reduction in CH₄ intensity in dairy cattle (Hart et al. 2019). Klop et al. (2017) alternated AR (0.17 g/kg DM) with lauric acid (0.65 g/kg DM) for 2-week periods over 10 weeks, but CH₄ production was not altered. Elcoso et al. (2019) estimated 15% lower CH₄ production in lactating dairy cattle consuming AR. Castro-Montoya et al. (2015) fed 0.0128 and 0.0240 g AR/kg DM to dairy and beef cattle respectively, but detected only tendencies towards CH₄ reduction in both groups, with no significant differences occurring.

Mootral^© is synthesised from natural products including garlic- and flavonoid-containing citrus extract and has demonstrated anti-methanogenic properties (Eger et al. 2018; Roque et al. 2019c; Vrancken et al. 2019). The garlic component in Mootral^© targets methanogenic archaea populations and protozoal communities in the rumen and has led to nearly complete inhibition of CH₄ production in vitro at a dosage of 2 g experimental mixture/day, without compromising bacterial population (Eger et al. 2018). The experimental mix contained 1.5% (w/w) allicin and 45% (w/w) polyphenolics (Eger et al. 2018). A 23.2% decrease in CH₄ yield (26.8% expressed in CH₄ production) was observed in Angus × Hereford crosses after 12 weeks of treatment by supplementing Mootral^© at 1.58 g/kg DM (Roque et al. 2019c). Adverse effects on DMI, ADG and feed efficiency were not detected over the 12-week trial. Lactating cattle offered Mootral incorporated in pellets at a rate of 0.640 g/kg DM for Holstein-Friesian and 1.21 g/kg DM for Jersey herd experienced suppression of CH₄ of 20.7% and 38.3% respectively (Vrancken et al. 2019). Additionally, 3–5% increase in milk yield across breeds was observed with increased feed efficiency in the Jersey cattle. Further research is required to determine the effective dose and magnitude of reduction from ruminants supplemented with Mootral^©.

Several feed additives provide a promising option that could increase the sustainability of animal-sourced foods by substantially reducing enteric CH₄ emissions. Rumen inhibitors have shown potential of up to 98% reduction in enteric CH₄ production, although they differ in accessibility and risk to animal welfare. Although none of the inhibitors are currently on the market, on the basis of the volume of available literature, 3NOP may be offered to producers in the near future, with nitrate and microalgae to follow after further research. Rumen modifiers including EO, tannins, saponins, biochar and lipids can be sourced globally but vary in composition and are not always effective. Consistency is a factor to consider with plant-based feed additives, but it can be addressed, as demonstrated, in commercial applications such as Mootral^© and Agolin Ruminant. Direct-fed microbials or probiotics have not demonstrated strong evidence to be considered a rumen modifier to suppress CH₄ production. Due to increased interest in this area, research is expected to accelerate in production of feed additives that reduce enteric CH₄ production.

The authors declare no conflicts of interest.