Abatement of Agricultural Non-Carbon Dioxide Greenhouse Gas Emissions - Chapter 7 - Methane: Sources, Processes and Abatement Options

Summary
7.1 Processes
7.2 Sources of Methane
7.3 Factors Affecting Methane Emission
7.4 Mitigation
7.5 Current Research on Methane in New Zealand
References

Chapter 7 - Methane: Sources, Processes and Abatement Options

Summary

The predominant source of methane in New Zealand is the fermentation of pasture plants in the rumen of farm animals. Methane is synthesised from H₂ and CO₂ at the end of the microbial digestion chain by the methanogenic archaea, a group of microorganisms that is widely distributed in nature and is also responsible for methane synthesis in manure, effluent ponds and the soil. If H₂ is allowed to accumulate in the rumen it depresses digestion, so the archaea remove it as methane. Management of H₂ in the rumen is the key to controlling ruminant methane emissions.

There is limited data available in the world literature on methane emission from animals grazing pasture. The best set is from New Zealand, where measurements have been made over a range of pasture types and management scenarios using the SF₆ tracer technique. There is good agreement that mature dairy cows and sheep grazing high quality pastures (>75% DM digestibility) produce about 26 g methane per kg DM digested (DDMI). On poorer quality diets, dairy cows and sheep produce about 35 g methane per kg DDMI. Other sources of methane such as manure, dairy effluent ponds and the soil appear to be trivial compared to enteric digestion. The soil in fact is a major sink for methane through oxidation by methanotrophic bacteria.

Several major nutritional factors are known to have an influence on methane emission. Methane emission increases with feed intake, but the relationship is not strong because of the high degree of variation between individual animals. However, there is a stronger negative relationship between methane emitted per unit of feed intake and feed intake. So there is an advantage, in terms of methane emission, to feed animals on as high an intake as possible. It is generally accepted that digestion of cell wall carbohydrates produces more methane than the digestion of soluble carbohydrates. Protein and lipids appear to have a negative effect on methane production, but the effects are variable, and in the case of lipids toxicity to the rumen microbes can be a problem.

Many technologies have been proposed for mitigating ruminant methane emission. Livestock numbers are the major determinant of emission at the national scale. While it might be considered politically naive to advocate reducing livestock numbers, sheep farmers over the last 15 years have reduced numbers by 33% without compromising total production. This shows that farming has the inherent flexibility to respond to a meaningful economic incentive.

There are possibilities for reducing methane via improvements in animal efficiency. All animals have an obligatory maintenance requirement that results in no production, yet has an associated methane emission. The strategy must be to dilute the effects of maintenance by various measures such as increasing feed intake, manipulation of dietary composition to increase feed quality (e.g. decrease cell wall carbohydrate), increasing metabolic efficiency and genetic improvement. Dairy cows that have been selected for feed conversion efficiency produce less methane on the same diet. These efficiency improvements should form the basis for on-farm strategies to reduce methane in the short-term.

A wide range of feed additives have been proposed to reduce methane. These include alternative hydrogen acceptors (e.g. malate, fumarate), halogenated methane analogues (e.g. chloroform, bromoethanesulphonic acid), antibiotics (e.g. monensin, mevastatin), defaunating agents (e.g. manoxol, teric), and probiotics, bacteriocins and naturally occurring plant compounds (e.g. condensed tannins). There are problems with these compounds, such as toxicity to the microbes and the animal, short-lived effects due to microbial adaptation, volatility, expense and failure to meet consumer acceptance. With grazing animals, other than dairy cows, a delivery system would be required to ensure regular delivery into the rumen. Delivery by breeding into pasture plants is possible, but the time needed to get a viable plant established in the national pasture should not be underestimated.

Immunisation of animals against methanogens has been suggested by Australian scientists. This is a good concept, but we are still a long way from the delivery of an efficacious vaccine.

There are many possibilities available for manipulating the rumen microbial ecosystem to achieve methane reduction. These include targeting methanogens with microbial antibiotics, bacteriocins or phage, removing protozoa, and developing alternative sinks for H₂ such as acetogenic bacteria. Development of mitigation technologies from this type of research are well in the future because of the need to first understand the complexities of the rumen microbial ecosystem.

Investment on research into methane mitigation should cover a suite of technologies that range in their potential delivery time from short-term (on-farm systems research) to long-term (rumen microbial manipulation). A successful technology will deliver a win/win result with respect to methane reduction and increased animal production.

There are three sources of methane associated with pastoral agriculture: ruminant animals (the largest), faeces or manure (on pasture and in effluent ponds [dairying]), and the soil (which can be a source or a sink). Because enteric fermentation in animals is the predominant source of agricultural methane, this chapter will concentrate on this aspect. It will cover the processes involved in the production of methane, the sources of methane, factors affecting methane emission and possible mitigation strategies, all in the context of the grazing ecosystem.

7.1 Processes

7.1.1 Digestion in the rumen

7.1.1.1 Physiological parameters of the rumen

In ruminant animals methane is produced by microbial fermentation of the diet mainly in the reticulo-rumen (rumen) with a smaller amount in the large intestine. The rumen is essentially a fermentation vat, containing a variable amount of digesta (4-7 kg in sheep and 50-80 kg in dairy cows), determined by a balance of the input (feeding) and outflow rates. Buffering is achieved by the secretion of large volumes of saliva containing bicarbonate and phosphate, and pH is normally within the range 5.5-6.5. The temperature is closely controlled at around 39^oC. Feed particulate matter is reduced in size, predominantly by chewing during eating and rumination, and these two processes expose the plant internal structures for microbial attack. The digesta is mixed by strong regular contractions of the organ, which also aid the passage of undigested feed residues from the rumen via the reticulo-omasal orifice. Digesta particle size must be reduced to 1.0-2.0 mm in sheep and 2.0-4.0 mm in cattle before passage can occur. Mean residence time for DM in the rumen generally varies between 8 hours for a highly digestible diet to 24 hours for a hay of low digestibility. The main end products of the fermentation are the volatile fatty acids (acetic, propionic, butyric), ammonia, microbial cells and methane. The rumen can therefore be regarded as a well regulated, variable but continuous flow fermenter. Methane produced in the rumen is removed by eructation via the oesophagus and mouth.

7.1.1.2 Microbial synthesis of methane in the rumen

Methanogenesis is carried out by a specialised group of microorganisms that are not true bacteria, but are a sub-group of the archaea which are widely distributed in nature. Archaea differ from bacteria in that they lack the peptidoglycans characteristic of bacterial cell walls, their lipids are composed of glycerol joined by an ether linkage to polyisoprenoid side chains rather than glycerol esters of fatty acids, and they have distinctive ribosomal RNA sequences (Miller, 1995). Although over 66 species of archaea have been isolated from a range of anaerobic habitats, only a few have been isolated from the rumen (McAllister et al, 1996). Methanobrevibacter and Methanomicrobium have been found in large numbers in the rumen and small numbers of Methanosarcina are also present (Baker, 1998; Jarvis et al, 2000). However, non-culture methods based on 16S rRNA indicate that there is a far greater diversity of methanogenic archaea in the rumen than has previously been recognised (Whitford et al, 2001).

Details of the biochemistry of methanogenesis are given by Miller (1995) and McAllister et al (1996). This review will concentrate on the essential elements that have a bearing on possible mitigation strategies. Rumen methanogens preferably synthesise methane from H₂ and CO₂ to generate their energy requirements for growth (Miller 1995) according to the reaction:

4H₂ + CO₂ CH₄ + 2H₂O

They have also have the ability to synthesise methane from formate and, to a lesser degree, methanol, mono-, di- and tri- methylamine and acetate, but it is the reduction of CO₂ that is the preferred pathway. The anaerobic conversion of organic matter to methane in the rumen involves a consortium of rumen microorganisms, with the final step effected by the methanogens (McAllister et al, 1996). Primary digestive microorganisms such as bacteria, protozoa and fungi hydrolyse dietary starch and plant cell wall polysaccharides producing sugars, volatile fatty acids (VFA), CO₂ and H₂. The sugars and protein are then fermented by secondary microorganisms to volatile fatty acids, ammonia, hydrogen and CO₂.

Methanogens then facilitate the efficiency of processes, such as cell wall degradation, by preventing the accumulation of hydrogen (NADH) through interspecies hydrogen transfer (Wolin et al, 1997). The importance of interspecies hydrogen transfer has been demonstrated many times in in vitro co-culturing experiments (McAllister et al, 1996). For example, Williams et al (1994) showed that addition of the methanogen Methanobrevibacter smithii to cultures of Ruminococcus flavefaciens increased the rate of xylan digestion compared to R flavefaciens alone. Interspecies hydrogen transfer has also been demonstrated with rumen fungi and protozoa as well as cellulolytic bacteria (Bauchop & Mountfort, 1981; Joblin et al, 1990). One very important symbiotic association is that between methanogens and host ciliate protozoa: both as ecto- and endosymbionts (Hegarty, 1999). These microbes living on and in the protozoa may be responsible for up to 37% of rumen methane emissions (Hegarty, 1999). The process of rumen microbial digestion is a complex and finely balanced ecosystem in which the methanogenic archaea fill an important niche.

One important consequence of hydrogen utilisation by the methanogens is that they maintain a low partial pressure of hydrogen in the rumen. If hydrogen accumulates in the rumen, re-oxidation of NADH is inhibited, reduced fermentation end-products such as lactate accumulate, and forage digestion and microbial growth are reduced (Wolin et al, 1997). For digestion to proceed normally to produce acetate, propionate and butyrate as nutrients for animal production, the partial pressure of hydrogen in the rumen needs to be kept low. Consequently, reduction or elimination of methanogenesis would require other routes of electron transfer if the animal were to benefit. Otherwise hydrogen would act as an inhibitor in the fermentation process and prevent further degradation of organic matter. Management of hydrogen in the rumen is the key to controlling ruminant methane emissions (Joblin, 1999).

The type of carbohydrate in the diet has an effect on methanogenesis (van Nevel & Demeyer, 1996). High starch-containing diets lower rumen pH, and methanogens are inhibited at low pH. Few of the starch-fermenting bacteria produce hydrogen. High starch diets therefore lead to a low proportion of the gross energy diet being converted to methane. Further, the acetate/ propionate ratio in the rumen is generally decreased because high starch diets result in higher propionate production. Conversely, high fibre diets tend to produce a high acetate/propionate ratio and increased methane.

There is another group of rumen microorganisms, the acetogenic bacteria, which have the capacity to convert hydrogen into acetate, one of the main nutrients of the ruminant animal. Over 10 acetogenic bacteria have been isolated from the rumen (Joblin, 1999). The affinity of methanogens for hydrogen is 10 to 100 times higher than the affinity of the reductive acetogens, so the acetogens cannot compete with the methanogens in the rumen because the partial pressure of hydrogen is normally too low (Nollet et al, 1997). The nutritional requirements of the acetogens are not as limiting as those for the methanogenic archaea, so they are able to utilise numerous other substrates such as carbohydrates and aromatic compounds for their growth (Fonty & Morvan, 1996). However, when no methanogenic activity is present, the acetogens can activate their ability to utilise hydrogen and produce acetate.

7.1.2 Large intestinal methane production

Detailed reviews of large intestinal fermentation are given by Miller (1995), Immig (1996) and Moss et al (2000). The caecum and proximal large intestine can account for about 5 to 25% of digestible energy, depending on diet. The proportion digested in the large intestine generally increases when a poor quality diet is fed. Digesta flows in an intermittent pulsatile manner from the terminal ileum into the caecum, which is a blind sac, and then to the proximal and other regions of the colon. Fermentative digestion occurs mainly in the caecum and proximal colon. Residence time for digesta is much lower than for the rumen. The large intestine accounts for about 12 to 17% of VFA production and 6 to 14% of the animal's daily methane production (Immig, 1996). Murray et al (1976) showed in sheep fed 800 g lucerne chaff per day, that while 87% of methane was produced in the rumen and 13% in the lower digestive tract, >98% was excreted via the mouth and about 2% in the flatus. Of the methane produced in the lower digestive tract, 89% was absorbed and expired from the lungs. Compared to the rumen, the large intestine does not seem to produce as much methane per mole of VFA (Immig, 1996). It is likely that reductive acetogenesis occurs to some extent in the large intestine (Miller, 1995; Immig, 1996). The biochemical reactions involved in methanogenesis seem to be the same as for the rumen (Immig, 1996). Our knowledge of transactions in the large intestine is however, far from adequate.

7.1.3 Methane emission from the degradation of faeces (manure)

Conservatively 50 000 t (DM) of undigested feed residues are voided via the faeces daily in New Zealand from sheep and cattle. Except for the case of dairy shed effluent, most faeces are deposited on pasture. Methane will be emitted from these faeces in three ways: release of trapped methane originating from gastro-intestinal digestion; some residual digestion by intestinal microbes, though these are very sensitive to oxygen and reduced temperature (Williams, 1993); and, a slow subsequent release due to invasion by soil methanogenic archaea. Most of the carbon in the deposited faeces will be dispersed through oxidation to carbon dioxide by aerobic bacteria of soil origin. In the case of dairy shed effluent, there will be some anaerobic methanogenic digestion of organic matter, the amount depending on the extent to which oxidative processes are built into the effluent pond design. The biochemical reactions of anaerobic methanogenesis in deposited faeces and effluent ponds are likely to be similar to those of the rumen, although because the reaction rate will be slower there may be some switching to acetogenesis and possibly methane production from acetate.

7.1.4 Soil as a source or sink for methane

Depending mainly on its moisture status, the soil can be either a source or a sink of methane. In flooded soils fermentation of organic matter occurs, with the main products being ethanol, acetate, lactate, propionate, butyrate, H₂, N₂, CH₄ and CO₂. Methane is formed by the reduction of CO₂ with H₂, with fatty acids or alcohols as the hydrogen donor, and the transmethylation of acetic acid or methyl alcohol by the methanogenic archaea (Mosier et al, 1998). The period between flooding and the onset of methanogenesis can vary with different soils.

Microbial oxidation of methane occurs in soil and aquatic environments, where it modulates emission and is a negative feedback on the increase in atmospheric methane (Mosier et al, 1998). All methanotrophs are obligate aerobes because the enzyme responsible for the initial step in methane oxidation is a monooxygenase enzyme (MMO) that requires molecular oxygen. The affinity of a methanotroph for methane governs its ability to compete for methane at low concentrations. The MMO of methanotrophs and the ammonia monooxygenase of nitrifying bacteria have similar substrate specificities, and apparently methane and ammonia are competitive substrates for both enzymes.

7.2 Sources of Methane

7.2.1 Ruminant animals

Enteric emissions from the grazing ruminant are responsible for about 87% of New Zealand's total methane emissions (NZCCP 2001). The only other significant sources are solid waste disposal on land (7.3%) and fugitive fuel emissions (2.7%). Within agriculture, enteric emissions account for about 99% of emissions and manure management 1% (NZCCP, 2001). The inventory of ruminant methane emissions revised by Clark and Ulyatt (2002) for the 2000 year was (Gg): dairy cows 340, beef cattle 248, sheep 541, deer 40 and goats 1.8. The inventory figures are based on emission factors inferred from a limited set of data, because very few measurements of methane emission have been made from grazing ruminants. Most data available in the international literature is from the evaluation of dried feeds in experiments conducted in calorimeters. There are limitations in this data with respect to the grazing situation. The grazing animal has fresh feed, it selects its feed, it is subject to peer pressure, and its behaviour is modified by the farmer's grazing management style. Fresh pasture is metabolised with a different efficiency to the same feed that has been dried (Ekern et al, 1965). Further, animals confined indoors rarely consume as much as animals grazing the same feed outdoors (Ulyatt et al, 1988). So it is doubtful that data derived from dried feeds in a calorimeter would apply to the grazing situation. The advent of the sulphur hexafluoride (SF₆) tracer technique (Johnson et al, 1994) has meant that for the first time measurements of methane emission can be made from grazing animals.

Most of the measurements made from grazing animals to date in New Zealand using the SF₆ technique, where DM intake was also measured, are listed in Table 7.1.The ewes were mainly consistent, averaging 30.8 g/d, 19.9 g/kg DMI (DM intake) and 26.1 g/kg DDMI (digestible DM intake). The wethers grazing high (75%) digestibility pastures emitted 25.9 g/d, 15.5 g/kg DMI and 19.8 g/kg DDMI, while those grazing poor quality pastures (<61% digestible DM) had a lower total emission (18.5 g/d), a higher emission per DMI (19.3 g/kg DMI) and a substantially higher emission per DDMI (33.4 g/kg DDMI). The dairy cows' emissions ranged from 137-431 g/d, were similar at 21.0 g/kg DMI, but like the wethers emitted 25.9g/kg DDMI for high quality diets and 35.8 g/kg DDMI for the poor quality diets. Cows fed perennial ryegrass and Lotus corniculatus silages (Woodward et al, 2001) were not included in the above analysis because the diets were changed three days before the measurements and this could have led to the unusually high emissions found. Judd et al (1999) estimated 19.5 g/d for wether sheep made with a micrometeorological technique and using animal data from Ulyatt et al (unpublished; Table 7.3).

Table 7.1: Measurements of Methane Emission from Various Classes of Grazing Ruminants in New Zealand Made with the SF₆ Tracer Technique. BW = body weight, Dig = DM digestibility, CH₄ = methane, DMI = dry matter intake, DDMI = digestible dry matter intake, MY = methane yield (MJ/100 MJ gross energy intake). Dominant pasture species: PR = perennial ryegrass, WC = white clover, BT = brown top, CF = cocksfoot, Drought = completely dried off pasture, Dsa = summer grass.

	Pasture	BW	Dig	DMI	CH₄	CH₄/DMI	CH₄/DDMI	MY
		(kg)	(%)	(kg/d)	(g/d)	(g/kg)	(g/kg)
Ewes	PR/WC: Sept¹	54	82	1.51	30.6	20.3	24.7	6.1
	Nov¹	54	72	1.46	33.2	22.7	31.5	6.9
	Mar¹	62	75	1.35	27	20.1	27.0	6.1
	Jul¹	66	82	1.89	27.9	15.1	18.5	4.6
	BT/CF: Mar²	69	73	1.69	35.2	21.1	28.9	6.4
Wethers	PR/WC: Mar³	37	75	1.27	18.9	15.0	19.8	4.6
	PR/BT: Mar²	38	80	1.39	19.3	13.8	17.4	4.2
	PR/WC: Apr²	41	81	1.70	21.9	12.9	15.9	3.9
	PR/WC: Apr⁴	38	75	1.38	18.5	13.5	17.9	4.0
	PR/WC: Jun⁴	40	80	1.41	24.1	17.3	21.7	5.2
	PR/WC: Nov⁵	46	76	1.62	29.9	19.8	26.0	6.0
	PR/WC: Oct⁶	45	82	2.07	30.4	15.2	18.6	4.5
	PR/WC: Nov⁶	48	77	1.81	34.3	19.8	25.9	6.0
	PR/WC: Jan⁶	52	79	2.26	30.4	14.0	17.9	4.2
	PR/WC: Feb⁶	53	82	2.42	31.2	13.7	16.7	4.1
	Kikuyu: Feb⁷	35	61	0.76	15.6	20.7	33.8	6.3
	Drought: Feb²	47	54	1.21	21.4	17.8	32.9	5.4
Cows	PR/WC: Sept¹	475	82	19.3	431	22.4	27.3	6.8
	PR/WC/hay: Jun¹	489	63	6.8	137	20.2	25.9	6.1
	PR/WC: Mar³	483	77	12.9	263	20.4	26.4	6.2
	Kikuyu: Feb⁷	438	61	15.6	363	23.4	38.2	7.1
	Dsa/RG: Mar⁷	585	67	18.9	422	22.3	33.3	6.7
	NZ pasture/ Sep⁸	497	82	17.2	307	18.0	22.0	5.3
	Dec⁸	519	79	17.0	376	22.2	28.3	6.6
	Mar⁸	519	78	15.0	353	23.8	30.4	7.0
	OS pasture/ Sep⁸	588	82	17.7	267	15.1	18.4	4.5
	Dec⁸	601	74	17.6	345	19.9	26.8	6.0
	Mar⁸	594	78	16.3	379	23.4	30.1	6.9

¹Ulyatt et al, 2002a²Ulyatt et al, unpublished³Lassey et al, 1997⁴Pinares-Pantiño, 2000⁵Pinares-Pantiño et al, 2003a⁶Pinares-Pantiño et al, 2003b⁷Ulyatt et al, 2002b⁸Waghorn et al, 2002

For comparison, New Zealand data on methane emission from animals fed indoors is presented in Table 7.2. For the wethers, with those fed Lotus excluded, methane emission was 24.9 g/d, 18.7 g/kg DMI and 26.4 g/kg DDMI. The reason for excluding the wethers fed Lotus was that this diet appeared to suppress methane emission. It was thought that this was due to the presence of condensed tannins (CT) in the Lotus pedunculatus diet. However, dosing the sheep with polyethylene glycol, which inactivates CT, had little effect. Further, other plants with known CT or CT-like activity, such as sulla and chicory, did not produce a marked depression in methane emission. This observation suggests that there is something other than CT that reduces methane emission in animals fed Lotus. The dairy cows fed TMR in Table 7.2 had methane emissions of 437.6 g/d, 20.3 g/kg DMI and 25.9 g/kg DDMI.

Table 7.2: Methane Emission from Sheep and Dairy Cows Fed a Range of Diets Indoors, Determined with the SF₆ Tracer Technique. BW = body weight, Dig = DM digestibility, CH₄ = methane, DMI = dry matter intake, DDMI = digestible dry matter intake, MY = methane yield (MJ/100 MJ gross energy intake), PEG = polyethylene glycol, TMR = total mixed rations, NZ = New Zealand genotype, OS = overseas genotype.

	Diet	BW	Dig	DMI	CH₄	CH₄/DMI	CH₄/DDMI	MY
		(kg)	(%)	(kg/d)	(g/d)	(g/kg)	(g/kg)

Wethers	Pasture¹	57	73	1.26	25.0	19.8	27.3	6.0
	Lucerne hay¹	59	59	1.08	18.7	17.3	29.2	5.2
	Lucerne hay¹	43	64	1.18	18.8	15.9	24.8	4.8
	Pasture²	33	74	1.12	28.7	25.7	34.7	7.8
	Lucerne²	38	71	1.47	30.2	20.6	29.0	6.2
	Sulla²	38	73	1.50	26.3	17.5	24.1	5.4
	Sulla/lucerne²	38	71	1.67	31.8	19.0	26.7	5.8
	Chicory²	35	79	1.12	18.1	16.2	20.4	4.9
	Sulla²	36	63	1.17	20.5	17.5	27.7	5.3
	Chicory/sulla²	36	71	1.37	23.1	16.9	23.8	5.1
	Red clover²	44	76	1.76	31.2	17.7	23.4	5.4
	Chicory/red clover²	36	77	1.36	26.8	19.7	25.6	6.0
	Lotus²	40	70	0.94	10.8	11.5	16.4	3.5
	Lotus (+PEG)²	40	76	0.94	12.9	13.8	17.3	4.2
Ewe lambs	Lucerne hay¹	35	64	0.90	11.5	14.8	23.1	4.4
Dairy cows	NZ/TMR: Sep³	544	72	21.4	422	20.3	27.3	5.6
	Dec³	583	81	19.7	435	22.0	27.3	6.1
	Mar³	626	80	19.0	423	22.3	27.9	6.4
	OS/TMR: Sep³	610	75	27.9	446	16.0	21.5	4.4
	Dec³	646	80	23.0	448	19.5	24.3	5.4
	Mar³	693	78	21.4	452	21.4	27.1	6.2
	Ryegrass⁴		66	10.7	260	24.6	37.3	7.2
	Sulla⁴		83	13.1	253	19.5	23.5	6.1

¹Pinares-Pantiño, 2000²Waghorn et al, 2002a³Waghorn et al, 2002b⁴Woodward et al, 2002

If the emissions of the grazing and the indoor fed animals are compared there are some interesting similarities. When data from the two groups of sheep and two groups of cows grazing the high fibre kikuyu, drought and subtropical Digitaria-dominant pastures were omitted, the mean emissions of the grazing ewes and cows and indoor wethers and cows were 26.1 ± 4.94 (sd), 26.2 ± 3.83, 26.4 ± 3.65 and 25.9 ± 2.50 g/kg DDMI respectively. There was remarkable consistency, which has implications for inventory development. The grazing wethers emitted 19.8 ± 3.63 g/kg DDMI, which was lower than their indoor fed counterparts. The reason for this is unknown. When data from the two groups of sheep and two groups of cows grazing the high fibre pastures was analysed, their emissions were 33.4 and 35.8 g/kg DDMI for the wethers and cows respectively. The high fibre ryegrass in the work of Woodward et al (2002; Table 8.2) also fits this pattern. This suggests that the higher fibre, lower DM digestibility pastures resulted in more methane emitted per unit of feed digested, which is in line with the findings of Blaxter and Clapperton (1965) and Moe and Tyrell (1979).

There are a few estimates of methane emission from grazing animals published outside New Zealand. Lockyer and Jarvis (1995) and Lockyer (1997) estimated emissions of 13-14 g/d for sheep and 74.5 g/d for calves confined in a polythene tunnel on pasture in England, while Leuning et al (1999) using a mass balance and SF₆ respectively measured 11.9 g/d and 11.7 g/d in sheep in Australia. McCaughey et al (1997) used the SF₆ technique to measure the emission from steers and beef cows grazing lucerne/grass pasture mixtures in Canada (Table 7.3). While the total emissions (g/d) were within the range of similar animals measured in calorimeters (e.g. Vermorel, 1995), the values per unit of DMI, DDMI, or MY were very high for the beef cows compared to New Zealand data (Table 8.1). The reason appears to be their DMIs, which are dependent on their estimates of in vitro digestibility. The extremely low digestibilities for the beef cows would result in low DMIs and high estimates of methane emission per unit of intake. This data emphasises the crucial importance of obtaining reliable estimates of DM intake when methane measurements are made.

Table 7.3: Methane Emission Measurements Made Under Grazing Gonditions with the SF₆ Tracer Technique by ¹McCaughey et al (1997) and²McCaughey et al (1999). BW = body weight, Dig = DM digestibility, CH₄ = methane, DMI = dry matter intake, DDMI = digestible dry matter intake, MY = methane yield (MJ/100 MJ gross energy intake), HiSR = high stocking rate, LoSR = low stocking rate, of steers grazed on a lucerne (60%)/meadow bromegrass (28.6%) pasture.

	DMI	BW	Dig	CH₄	CH₄/DMI	CH₄/DDMI	MY
	(kg/d)	(kg)	(%)	(g/d)	(g/kg)	(g/kg)

Steers¹
Rotational grazing:
HiSR	14.9	392	60	188	12.6	21.0	4.1
LoSR	13.6	417	61	200	14.7	24.0	4.3
Continuous grazing:
HiSR	13.5	380	60	173	12.8	21.3	4.4
LoSR	13.2	403	58	219	16.6	28.6	5.2

Lactating beef cows²
Lucerne/grass	11.4	506	50	267	23.4	46.8	7.1
Grass	9.7	516	44	293	30.2	68.7	9.5

Source: McCaughey et al, 1997

Pavao-Zuckerman et al (1999) measured methane emission from cows and steers grazing tall fescue dominant pastures in the US (Table 7.4), with management and endophyte treatments superimposed. Unfortunately they did not measure feed intake or digestibility, so it is difficult to compare their methane estimates with New Zealand data, other than to say that most of their emissions for beef cows of around 500 kg were of the order of 140-240 g/d, compared to New Zealand dairy cows of 250-400 (Table 7.1), or the Canadian beef cows (Table 7.3) of 267-293 g/d. Steers in the work of Pavao-Zuckerman et al (1999) weighing 350-430 kg emitted 190-200 g methane per day, compared to McCaughey et al (1997; Table 7.3), where steers of around 400 kg emitted 180-220 g/d. So the Canadian and American estimates for steers are reasonably similar. The New Zealand dairy cows would be expected to have a higher methane emission than beef cows because their feed intakes would have been much higher.

Table 7.4: Measurements of Methane Emission Made With the SF₆ Tracer Technique from Cattle Grazing Tall Fescue Over Seasons at Two Sites: Blount (with (E+) and without (E-) endophyte) and Holston, where two management practices were used (unimproved pasture UP and best pasture management practices of the district BMP)

			Spring		Summer
Steers (Blount)			BW (kg)	CH₄ (g/d)	BW (kg)	CH₄ (g/d)
1997	E+		286	166	352	148
	E-		297	190	383	176
	E+/E-		289	167	356	154
	E+/clover		290	178	387	190
			Winter		Spring
1998	E+		256	110	316	147
	E-		264	120	346	153
	E+/E-		250	112	323	127
	E+/clover		268	119	337	154

*SeasonManagement (Holston)**			Spring		Summer
1997	BMP	Steer	352	201	433	190
		Cow	494	239	550	216
	UP	Steer	355	187	420	174
		Cow	495	243	546	227
			Winter		Spring
1998	BMP	Steer	279	94	322	139
		Cow			498	168
	UP	Steer	270	104	301	145
		Cow			535	147

Source: Savao-Zuckerman et al, 1999

7.2.2 Faeces

There are a few estimates available indicating the amount of methane derived from the faeces of dairy cows grazing pasture. The emission per cow was estimated by Jarvis et al (1995) in England as 0.603 g/d, Yamulki et al (1999) in England as 0.28-1.95 g/d, Flessa et al (1996) in Germany as 1.037 g/d and Tate (pers comm) in New Zealand as 1.97-3.04 g/d. While there is a five-fold range in the estimates, they do indicate that methane emission from faeces appears to be trivial compared with that emitted by cattle (c.250 g/d). The faeces of sheep are usually of higher DM content than those of dairy cows, and are often of a pelleted nature and more disposed for aerobic degradation. Thus it would be expected that methane emission per gram of faecal DM would be lower in sheep. Using an average sheep faecal DM output of 375 g/d, and the methane emission factor of 0.07 mg CH₄/g faecal DM (NIR 2002), the daily methane emission from an adult sheep in New Zealand would be about 26 mg/d for a ewe producing 30 g/d. The ratio, exhaled to manure methane, is thus 250 for cows and 1 150 for sheep. For both animals' manure would seem to be a minor source of methane and, as expected, manure would be relatively more important as a source for dairy cows than sheep.

7.2.3 Methane from effluent ponds

There are a few estimates of methane emissions from effluent ponds. The New Zealand Climate Change Programme (2001) estimated 16.93 Gg from such ponds compared to a total of 1 399 Gg from all agricultural sources.

7.2.4 Sources and sinks of methane in the soil

New Zealand's soils generally remove methane at a rate that is negatively related to both their level of disturbance (e.g. cultivation) and level of nitrogen input (Tate, pers comm). New Zealand's undisturbed natural ecosystems, which generally do not receive significant external nitrogen inputs, may exhibit very strong methane sink capacity. Price et al (2000) measured methane oxidation rates at a mountain beech forest site, which if applied to all such vegetation would indicate that 63 000 t of methane could be removed from the atmosphere per year by this type of land cover. Generally lower methane oxidation rates are measured for managed ecosystems. For example, planted pine forests and pasture soils could potentially oxidise about 11 000 and 400 t methane per year respectively (Tate, pers comm). New Zealand agricultural soils appear to have very low rates of methane emission (Judd et al, 1999) and are thus not considered a significant source.

7.2.5 Conclusions

New Zealand has the best set of measurements of methane emission from grazing animals in the world. However the data are predominantly from young wether sheep under a limited range of pastoral conditions, and a few measurements from mature ewes and dairy cows. If the inventory is to be accurate, and therefore beyond reproach for international reporting, regulatory and scientific purposes, methane emission from a wider sample of animal and land classes needs to be incorporated.

7.3 Factors Affecting Methane Emission

This subject has been reviewed recently by Johnson and Johnson (1995, 2002), Ulyatt (1996), Mathison et al (1998), Moss (2000) and Hegarty (2001).

7.3.1 Feed intake

The relationship between methane emission (g/d) and DM intake is positive, but characterised by high variability between animals (Blaxter & Clapperton, 1965; Kirchgessner et al, 1995; Lassey et al, 1997). An example of this relationship is plotted in Figure 7.1, using data from sheep grazing fresh pasture in New Zealand, showing that the absolute amount of methane emitted increases as intake increases (r=0.373; P<0.05) (Lassey et al, 1997). The notable thing about this relationship is that approximately 87% of the variation in methane emission is between animals, suggesting that differences in DM intake per se accounted for about 14% of the variation in methane emission.

Figure 7.1: Methane Emission Versus DM Intake in a Group of 50 Sheep Grazed on the Same Pasture (n=5)

Source: Lassey et al, 1997

However, when methane emission per unit of feed intake (usually expressed as MJ methane per 100 MJ gross energy intake) is plotted against DM intake for the same data (Figure 7.2), a stronger negative relationship is found (r=-0.597; P<0.01), indicating that as intake increases the percentage of dietary energy lost as methane decreases. This is a well established relationship for sets of data where animals are fed the same diet at both restricted and ad libitum intakes (Armstrong, 1964; Blaxter & Clapperton, 1965; Johnson & Johnson, 1995). This suggests that for efficient animal production and reduced methane emission it is advantageous to feed animals well above maintenance intake.

Figure 7.2: Methane Emission Per Unit of Feed Intake Plotted against DM Intake in Sheep Grazing the Same Pasture

Source: Lassey et al, 1997

7.3.2 Diet composition

The diet in ruminant agriculture in New Zealand is predominantly based on pasture. There is an increasing use of high starch supplements in the form of maize or maize silage in dairying and grain in drought conditions. Apart from these supplements, it is difficult to achieve large changes in pasture composition through plant breeding or pasture management. There are some opportunities for manipulation of plant composition through the introduction of novel molecular biological techniques (Hancock & Ulyatt, 2001).

The major constituents of the diet - sugars, starch, fibre, protein and lipid - appear to have varying impacts on methane emission. Kirchgessner et al (1995) concluded, from a regression analysis of the impact of crude nutrient fractions on methane emission from dairy cows, that on average crude fibre provides about 60%, nitrogen-free extract 30%, crude protein 10% and ether extract a minor proportion of total methane production. However, variations within and between the major classes of nutrients can cause major shifts in methane emission.

The type of carbohydrate fermented affects methane production, particularly starch and soluble sugars compared to the cell wall carbohydrates, cellulose and hemicellulose. With respect to starch, as its content in the diet increases rumen pH decreases, making the environment more hostile for methanogens to survive. Further, few of the starch fermenting bacteria produce hydrogen, so the supply of hydrogen for methanogenesis is limited. It would therefore be expected that less methane should be produced per unit of starch than per unit of digested cell wall carbohydrate. Johnson and Johnson (1995) claimed that the soluble sugars are more methanogenic than starch. The data of Blaxter and Wainman (1964), where sheep and cattle were fed variable portions of hay and maize at about maintenance and twice maintenance levels of intake, illustrates the effects of type of carbohydrate on methane emission. As the proportion of maize (and thus starch) increased in the diet from 0 to 100% there was a small reduction in methane at maintenance, and a decrease from about 7.0 to 3.5 MJ/100 MJ intake at twice maintenance.

Conversely, Blaxter and Wainman (1964) showed that as the proportion of hay increased from 0 to 100%, crude fibre in the diet increased from 2.2 to 33.8 % and methane (MJ/100MJ) showed a small decrease at maintenance, but increased from about 3.5 to 7.0 at twice maintenance. Moe and Tyrell (1979) also found little difference between carbohydrate sources at maintenance, but at higher intakes cell wall carbohydates were found to be more methanogenic than soluble carbohydrates. So any method that increases the ratio of soluble/cell wall carbohyrate should decrease methane production. It should be noted, however, that the proportion of soluble carbohydrate fed by Blaxter and Wainman (1964) and Moe and Tyrell (1979) to get depression in methane production was very high. There are also reports from experiments where large changes in the starch content of the diet had little effect on methane emission (Shiao et al, 1999; Islam et al, 2000; Cammell et al, 2002). So while the weight of evidence supports reduction of methane by increasing the ratio of cell wall to soluble carbohydrate in the diet, there is evidence that the impact of carbohydrate type on methane is complicated. It must also be emphasised that many indoor studies are conducted around maintenance intake where the effect is likely to be minimal.

The effect of protein concentration in the diet is less clear. Pelchen and Peters (1998) analysed 1 137 data sets from the literature where sheep were fed in calorimeters, and developed regression equations to predict methane emission. When crude protein was included as an independent variable it had a negative sign, indicating that increasing protein in a diet would be expected to decrease methane emission. This may indicate a direct negative effect of protein on methane, or it might reflect the replacement in the diet of methanogenic carbohydrate with protein.

Addition of lipids to the diet can reduce methane emission. Three factors - the quantity, the degree of unsaturation and the chain length of the lipid - can have an effect (Czerkawski et al, 1966a, 1966b; Johnson & Johnson, 2002). It appears that the effect of degree of unsaturation is relatively small and that the effect of lipid is mainly in depressing digestion (Johnson & Johnson, 1995; Mathison et al, 1998). Certain oils, such as coconut oil, seem to reduce methane, possibly by suppressing protozoa (Machmuller et al, 1998; Dohme et al, 1999).

7.3.3 Digestibility

Compilations of data comparing methane emission at various digestibilities exhibit a high degree of variation, e.g. the relationship of Johnson and Johnson (1995) for beef cattle (Figure 7.3).

Figure 7.3: Effect of Digestibility on Methane Emission: World Beef Cattle Data

Figure 7.3: Effect of Digestibility on Methane Emission: World Beef Cattle Data

Source: Johnson, pers comm, 2002

The main reason for this is that the results are confounded by the wide range of diets and intakes used in such comparisons. It has already been shown above that methane is dependent on both diet composition and intake level.

Blaxter and Clapperton (1965) calculated that the relationship between methane emission and digestibility is very dependent on intake level (Figure 7.4). When feed is given at low levels of intake, methane emission (MJ/100MJ) increases as digestibility increases, whereas with high intakes methane emission falls as digestibility increases.

7.3.4 Other factors

A number of other nutritional and physiological factors are known to influence methane emission, such as grinding and pelleting the diet, and frequency of feeding (Johnson & Johnson, 1995; Mathison et al, 1998), but are probably of little significance under grazing.

Figure 7.4: The Relationship between Digestibility and Methane Emission at Different Levels of Feeding

Figure 7.4: The Relationship between Digestibility and Methane Emission at Different Levels of Feeding

Source: Blaxter & Clapperton, 1965

7.4 Mitigation

There have been many recent reviews of potential mitigation methodology, for example van Nevel et al (1995), Mathison et al (1998), Hegarty (1999), Klieve and Hegarty (1999), Moss et al (2000), Ulyatt et al (2002).

A wide range of possibilities for reducing the methane emission of grazing livestock has been suggested. These include reducing livestock numbers, increasing the efficiency of animal production, genetic improvement, antimethanogenic feed additives, immunisation, manipulation of the rumen microbial ecosystem and manipulation of farm management.

7.4.1 Reduction in livestock numbers

As emissions from livestock are the predominant source of methane in countries like New Zealand and Australia, reducing livestock numbers is one option for meeting FCCC commitments. However, such countries are heavily dependent on their livestock industries for generating national income, and the imposition of regulations aimed at reducing livestock numbers would not be well received by the farming industry.

There are, however, ongoing fluctuations in livestock numbers in response to the marketplace. For example, in New Zealand sheep farming has become less profitable over the past 10 years, so farmers have reduced sheep numbers and other alternative land uses such as dairying and forestry have increased. Over the period 1990 to 2000, total sheep numbers have reduced from 57 900 000 to 45 100 000, dairy cattle have increased from 3 390 000 to 4 530 000, beef cattle have changed little (4 600 000 to 4 690 000), deer have increased from 960 000 to 1 910 000 and farmed goats have declined from 1 030 000 to 190 000 (Clark & Ulyatt, 2002). The net outcome in terms of ruminant methane emission has been a small increase over the 10 years from 1 098 to 1 171 Gg/year. Given the large changes in most categories of stock induced by economic conditions over the last 10 years, it might not be too excessive to suggest that farmers might be persuaded to change livestock numbers in response to greenhouse gases if suitable economic incentives were offered.

Livestock numbers are the major determinant of methane emission from pastoral agriculture, and it is implicit that reducing numbers is the simplest way to reduce emission.

7.4.2 Increasing the efficiency of livestock production

7.4.2.1 Introduction

The subject of efficiency of livestock production is very complicated, and efficiency related to methane emission is closely related to the maintenance requirements of the animal. The maintenance requirement of an animal is the minimum amount of energy (or feed) required to keep the animal in energy equilibrium, i.e. at constant body weight with no production. It is obligatory that the maintenance requirement must be met from the diet before any production can occur, although there are short-term exceptions such as the utilisation of tissues for milk production in early lactation in dairy cows. There is a methane emission associated with the maintenance requirement. Using the emission factor of 26 g/kg DDMI, derived from New Zealand work in Section 7.2.1 above, maintenance methane emission would be 18 g/d for a 50 kg ewe, 80 g/d for a 450 kg beef cow and 105 g/d for a 450 kg dairy cow. So the animals would emit these amounts of methane unassociated with any production. The higher the intake above maintenance, or the higher the level of production, the lower will be the methane emitted per unit of product and thus the higher the efficiency with respect to methane. Examples of this are presented in Tables 7.5 and 7.6 (below) for milk production and lamb growth. From an animal efficiency point of view, the best strategy for a farmer faced with extra feed is to increase the intake of existing animals rather than increase stocking rate. From an economic point of view, methane production and animal production per ha are the important measures of efficiency. There will be a stocking rate that maximises production per ha and minimises methane emission per ha.

7.4.2.2 Increasing feed intake

Increasing feed intake decreases the methane emission per unit of feed intake as shown in Figure 8.2. This can also be seen in terms of production: as milk production (Kirchgessner et al, 1995) or liveweight gain in beef cattle (McCrabb & Hunter, 1999) an asymptotic decrease in methane emission per unit of product occurs. This effect is shown in Table 8.5, where the increased intake (DDMI) of the same diet to a cow increases milk production, but decreases methane emitted per unit of milk. As intake increases, the proportion of the methane associated with maintenance declines. This is known as the dilution of maintenance as intake increases. In terms of efficiency of production it is clearly advantageous to maximise intake. A similar calculation illustrating the effect of intake on methane emission during lamb growth is shown in Table 8.6. Again, methane emission per unit of product is reduced and the proportion of methane associated with production is increased at the higher intake.

Table 7.5: A Calculation of the Proportion of the Methane Emission Attributable to Maintenance or Milk Production in 450 kg Grazing Dairy Cows

DDMI	Milk yield	CH₄	% CH₄ associated with:		CH₄/milk
(kg/d)¹	(kg/d)	(g/d)²	Maintenance	Production	(g/kg)
4.0	0	105	100	0
7.9	12	206	51	49	17.2
10.5	20	272	39	61	13.6
11.7	24	305	34	66	12.7

¹Ulyatt et al, 1976

²DDMI * 26, (Section 8.2.1)

Table 7.6: A Calculation of the Proportion of the Methane Emission Attributable to Maintenance or Growth in 30 kg Growing Lambs

DDMI (g/d)¹	Growth rate (g/d)	CH₄ (g/d)²	% methane associated with:			CH₄/growth (g/kg)
			Maintenance		Growth
517	0	10.3	100	0
832	100	16.6	62	38		166
1147	200	23.0	45	55		115
1463	300	29.3	35	65		98

¹Ulyatt et al, 1976²DDMI * 20, (Section 8.2.1)

By feeding animals ad libitum it is possible to both maximise efficiency and reduce methane emission per unit of product. This is because as intake increases the methane emission associated with the essential, but non-productive, requirements for maintenance is diluted.

7.4.2.3 Dietary manipulation

As described above (Section 7.3.2), decreasing dietary fibre and increasing starch and lipid will reduce methane emission. Generally, diets of higher digestibility have these characteristics. This effect can be seen in the calculation in Table 7.7 where dairy cows were given feeds of increasing digestibility to achieve the same level of milk production. The animals would have eaten less of the higher digestibility diets, and thus produced less total methane and reduced methane emitted per unit of milk produced. Improving the nutritive value of the feed given to grazing animals by balancing the diet with concentrates, or by breeding improved pasture plants, should result in reduced methane emission. The latter is considered to be a primary objective in the grazing ecosystem, yet the magnitude of the changes in composition needed to affect methane emission, e.g. the ratio of soluble to cell wall carbohydrate, might be outside the capability of plant breeding or management.

Table 7.7: A Calculation of the Effect of Feed Quality on Methane Emission of Cows at the Same Level of Milk Production

DM digestibility (%)	55	65	75
Milk production (kg/d)	20	20	20
Feed intake (kg DM/d)	21.6	17.5	14.6
CH₄ emission (g/d) ¹	309	296	285
g CH₄/kg milk	15.5	14.8	14.3

¹ DDMI * 26 (Section 8.2.1)

Benchar et al (2001) evaluated the effect on methane production of a range of dietary strategies using a modeling approach and predicted that a reduction of 10 to 40% can be achieved this way. Methane production (MJ/100MJ) could be reduced by increasing feed intake (-7%), increasing the concentrate proportion of the diet (-40%), replacing fibrous concentrate with starchy concentrate (-22%), with the utilisation of less ruminally degradable starch (-17%), increasing the digestibility of forage (-15%), with legume compared to grass forage (-28%) and with silage compared to hay (-20%).

7.4.2.4 Metabolic efficiency

Treatment of animals with growth promoting substances can result in increased efficiency of production. An example based on bovine somatotrophin (bST) treatment of milking cows and calculated from the data of Bauman et al (1985) is given in Table 8.8. As bST dose was increased, milk production per unit intake (efficiency) increased and methane emitted per kg milk was calculated to decrease. Growth stimulants such as steroids would be expected to have a similar effect: less feed and methane overall to achieve the same level of production. Such measures must, of course, meet the required regulatory and consumer acceptance standards.

Table 7.8: Calculated Effect of Bovine Somototrophin (bST) on Methane Emission by Lactating Cows

bST dose (mg/d)	0	13.5	27.0
Milk production (kg FCM/d)	27.9	34.4	38.0
NE intake (MJ/d)	143	154	164
kg milk/MJ NE intake	0.195	0.223	0.232
CH₄ emission (g/d)¹	365	361	351
g CH₄/kg milk	13.1	10.5	9.2

¹ Blaxter & Clapperton (1965)

All these techniques to increase efficiency use the dilution of maintenance requirements principle to achieve reduced methane emission. Their maximum effectiveness, in terms of reducing methane emission, would be in maintaining present levels of animal production with fewer animals, or in increasing animal production with the same number of animals. This would provide the farmer with options for land use that should improve profitability.

7.4.3 Genetic improvement in animals to reduce methane emission

There are two aspects of genetic improvement with respect to methane emission: genetic improvement in the efficiency of food conversion by the animals themselves; and the possibility that there are genetic differences between animals in the amount of methane they emit at the same feed intake.

In the grazing ecosystem there has been virtually no direct selection of ruminants for improved feed conversion efficiency, which is in contrast to the situation with pigs and poultry where huge gains have been made. There has, however, been considerable indirect selection for increased milk yield in dairy cows and increased liveweight gain in fattening lambs and beef cattle. Livestock statistics presented by Clark and Ulyatt (2002) show that despite large changes in livestock numbers over the 10 years 1990 to 2000, there have also been changes in production. Dairy cow weights increased from 420 to 454 kg, and milk production increased from 2 931 to 3 678 kg/cow. Beef cow weights increased from 443 to 470, heifer weights at slaughter from 397 to 420 and male weight at slaughter from 541 to 580 kg. Sheep statistics are even more interesting. Davison (2000) showed that between 1986/87 and 1999/00 breeding ewe numbers declined by 33%, but the total lamb slaughter declined at the lower rate of 19% because there was a 16.6% increase in lambing %. In 1986/87 average lamb weights were 13.20 kg compared to 16.48 kg in 1999/00, an increase of 25%. Thus the combination of increased lambing % and higher average weights resulted in a higher lamb production by 8% in 1999/00 at a time when breeding ewe numbers had declined by 33%. Davison (2000) stated that sheep are now fed at a higher level and stocking rates are 9% lower than in 1986/87. It is impossible to separate the effects of better feeding from genetic gain in this data, and the gains are undoubtedly due to a mixture of the two. Given that methane emission per unit of intake is lower at higher intakes (Tables 7.5, 7.6), the above statistics should have a positive impact on the ruminant methane inventory. They should also provide good input for farm-scale systems models (see Section 7.4.7).

Further evidence that genotype can influence methane emission comes from the genotype X diet experiment conducted at Dexcel by Dr E Kolver (Waghorn et al, 2002), summarised in Table 7.9. Dairy cows of US or New Zealand genotype were fed either a New Zealand pasture or a total mixed ration similar to those fed in the US. On both rations the US cows produced more milk and less methane. The US cows have been bred for efficiency of feed conversion and high feed intakes. They had considerably higher feed intakes in this experiment, and presumably this is why the methane emissions per unit of feed intake were lower. Similarly, Ferris et al (1999) found that methane emission (MJ/100 MJ GE intake) was higher for Holstein dairy cows of medium than high genetic merit. These experiments indicate that breeding for efficiency has the potential to lower methane emission.

Table 7.9: Methane Emission (g/kg DMI) from Dairy Cows of Two Genotypes Fed Two Different Diets. TMR = Total mixed rations.

	Days of lactation
	60	150	240
Pasture diet:
NZ genotype	18.0	22.2	23.8
US genotype	15.1	19.9	23.4
TMR diet:
NZ genotype	20.2	22.0	22.3
US genotype	16.0	19.5	21.4

Source: Waghorn et al, 2002

A notable feature of methane emission in experiments where large numbers of animals have been fed the same diet is that there are usually large differences in emission per unit of feed intake between animals. Between-animal differences account for 70-80% of the variance, with a lesser amount attributed to differences between measurement days. Such differences between animals are real (Blaxter & Clapperton, 1965; Lassey et al, 1997; Ulyatt et al, 1999) and can persist from three weeks (Lassey et al, 1997) to up to five months (Pinares-Patiño et al, 2003b). We have found that in any group of animals measured on pasture with the SF₆ tracer technique, approximately 10% are high and 10% low emitters, with the difference between these two groups approximately 40%. As methane is produced through microbial activity, the animal can only have an impact on methanogenesis by interacting with the microbes. This interaction could be via diet selection as the microbes respond to changes in the composition of the substrate (feed) presented to them. The animal could interact with the microbes through control of the fermenter (rumen) conditions via processes such as: saliva and/or salivary proteins, feed processing (e.g. increased comminution to allow faster microbial access to plant cell contents), changes in rumen volume and digesta flow rate. Pinares-Patiño et al (2003b) found that high methane emitting sheep tended to have large rumen volumes and slower digesta flow rates than low emitters. Ørskov et al (1988) identified cows with persistent differences in rumen outflow rate and concluded that they were probably genetic in origin. The evidence for a genetically determined influence of the animal on the rumen microbes is not strong. However, if it could be proven to be the case it may be possible to obtain genetic markers that could be used to select low methane emitters.

7.4.4 Feed additives

A wide range of chemicals are available that will reduce rumen methanogenesis (Chalupa, 1980; Mathison et al, 1998).

7.4.4.1 Alternative hydrogen acceptors

Sulphate/sulphite and nitrate/nitrite are potential alternative hydrogen acceptors. However Mathison et al (1998) concluded that their use is not feasible because of their toxic properties at the concentrations that would be needed to reduce methanogenesis.

Addition of unsaturated fatty acids to the rumen will decrease methane emission. Their effect is twofold: the unsaturated fatty acids are a potential alternative sink for hydrogen, and large doses are toxic to rumen microorganisms and depress digestion. Johnson and Johnson (1995) pointed out that the amount of hydrogen used in the biohydrogenation of unsaturated fatty acids is small, and concluded that the methane reduction effects of lipids are only likely to be substantial when basal digestion is inhibited. More recent work by Dong et al (1997) found that the addition of canola oil and cod liver oil to fermenters reduced methane but did not depress digestion, whereas coconut oil depressed digestion. Johnson and Johnson (2002) cite evidence to suggest that medium chain fatty acids may suppress methane more than long chain fatty acids.

The dicarboxylic organic acids, malate and fumarate, have been suggested as potential alternative hydrogen acceptors (Martin, 1998; Lopez et al, 1999; Carro et al, 1999; Asanuma et al, 1999; Krause 2002). Some anaerobic bacteria synthesise propionate from fumarate or malate using a reverse citric acid cycle. Malate must be converted to fumarate, which is in turn reduced to succinate, a process that requires hydrogen, and the succinate is then decarboxylated to form propionate. The tactic would be to increase the utilisation of hydrogen by rumen bacteria that can use hydrogen in the above reaction. There would be a decline in the availability of hydrogen for methanogenesis in the rumen. Most of the experiments to date have been conducted in vitro. Addition of fumarate to fermenters reduced methane, enhanced propionate production (Asanuma et al, 1999; Lopez et al, 1999), and increased the numbers of cellulolytic bacteria and increased DM digestibility (Lopez et al, 1999). Similarly, addition of malate reduced methane, increased propionate, and increased the digestibility of hemicellulose (Carro et al, 1999). Malate and fumarate are expensive chemicals, so it is doubtful that the amount required could be used as a feed additive. The concentrations of these organic acids vary naturally in plants. Whether their concentrations in pasture plants can be increased by selection or molecular genetics to the level that methanogenesis is reduced has yet to be determined.

7.4.4.2 Halogenated methane analogues

Many halogenated methane analogues such as chloroform, carbon tetrachloride, chloral hydrate, bromochloromethane and bromoethanesulphonic acid can be very potent methane inhibitors (van Nevel & Demeyer, 1996; Mathison et al, 1998). While some of these compounds are volatile and difficult to administer, McCrabb et al (1997) claimed success in inhibiting methane in cattle with bromochloromethane complexed with _-cyclodextrin, which reduced volatility. Mathison et al (1998) concluded that halogenated methane analogues have potential as methane inhibitors, provided that problems such as adaptation by rumen microbes, host toxicity and suppression of digestion can be overcome. In the pastoral environment, a cost-effective delivery system would also be needed.

7.4.4.3 Antibiotics

The effects of a wide range of antibiotics, including the ionophores, on methane production have been reviewed by van Nevel and Demeyer (1995, 1996) and Johnson and Johnson (2002). For many years antibiotics have been used routinely overseas as growth promotants at very low dose levels. Their effect on methane production has been inconsistent (van Nevel & Demeyer, 1995). However, avoparcin and the ionophores are known to inhibit methanogenesis and shift VFA patterns towards higher propionate. Monensin, an ionophore, inhibits methane in vivo by about 25% (van Nevel & Demeyer, 1995). There are, however, reports that the effect appears to be short-lived as the rumen microbes adapt to the additive within two weeks (Johnson & Johnson, 1995, 2002). Monensin has been used routinely as a bloat preventative in New Zealand for many years, but its effect on methane emission under our grazing conditions is unknown.

Miller and Wolin (2001) have recently described the in vitro inhibition of methanogen growth and methane by the two hydroxymethylglutaryl~SCoA reductase inhibitors, mevastatin and lovastatin. Archaea are unique among rumen microbes in having membrane lipids that contain glycerol joined by ether linkages to long chain isoprenoid alcohols. A key precursor in the synthesis of isoprenoid units is mevalonate, which is formed by the reduction of HMG-CoA. The enzyme HMG-CoA reductase catalyses the formation of mevalonate and is the specific target of the antibiotics mevastatin and lovastatin. These antibiotics inhibit the growth of rumen methanogens, but do not inhibit the growth of the rumen bacteria responsible for fermenting cellulose, starch and other plant polysaccharides. These drugs are also used to lower cholesterol in humans, so it is argued that they would be a harmless addition to the animal's diet.

7.4.4.4 Defaunating agents

Methanogens living in a symbiotic relationship with protozoa can account for about 40% of rumen methane emissions (Hegarty, 1999), and defaunation results in reductions in emission of about 20-50% (Kreuzer et al, 1986).

Defaunating agents such as manoxol, teric, alkanate 3SL3 and sulphosuccinate can reduce methane emission (Mathison et al, 1998). They appear to act by disrupting the close symbiotic relationship between methanogenic bacteria and protozoa. Complete defaunation is difficult to achieve on a large scale, and there is a fine line between killing the protozoa and killing the animal. The toxicity of many of these defaunating agents restricts their routine use.

7.4.4.5 Probiotics

Microbial feed additives, especially those based on Saccheromyces cerevisiae and Aspergillus oryzae, are widely used in animal feeding in the northern hemisphere, particularly with high grain diets. There are mixed reports as to whether these probiotic additives can reduce methane emission (van Nevel & Demeyer, 1995; Moss et al, 2000). It would appear that more research is needed to evaluate whether probiotics have any role in methane mitigation strategy, although it seems unlikely that they would be effective with animals grazing pasture.

7.4.4.6 Bacteriocins

Bacteriocins are antibiotics, generally protein or peptide in nature, produced by bacteria. Research is ongoing to see if these compounds can be used to manipulate the rumen ecosystem (Klieve & Hegarty, 1999; Attwood, pers comm). Callaway et al (1997) used the bacteriocin nisin, which is produced by Lactococcus lactis, to produce a 36% reduction of methane production in vitro. Further research is required to evaluate the efficacy of bacteriocins.

7.4.4.7 Naturally ocurring plant compounds

There are naturally occurring compounds in some forages that appear to have antimethanogenic properties. Johnson and Johnson (2002) cite a number of plant compounds that appeared to have this effect. Gupta et al (1993) claimed that the leaves of the tropical plant Enterolobium timbouva defaunated the rumen of buffalo. Woodward et al (2001) found a depression of methane emission by feeding sheep and dairy cows the condensed tannin-containing legume Lotus corniculatus, as did Waghorn et al (2002) when L pedunculatus was fed to sheep and Woodward et al (2002) when sulla (Hedysarum coronarium) was fed to dairy cows. Ulyatt et al (2002) found that under some conditions methane emission was severely reduced in sheep and dairy cows grazing kikuyu grass (Pennisetum clandestinum), suggesting the presence of yet unidentified suppressing compounds. Such observations suggest that there are compounds to be found in pasture plants that offer the prospect of methane reduction in the grazing environment if they can be bred into competitive pasture plants.

It is important to note that despite a huge amount of research there is not one feed additive that is currently used in commercial agriculture for the sole purpose of reducing methane.The main problems with chemical additives are that many are toxic to the animal, toxic to rumen microflora and therefore reduce digestion and food intake, have short-lived effects because the rumen microbes adapt, are volatile and thus difficult to administer, are expensive, or would fail to meet consumer product acceptance.

With grazing animals, especially under extensive conditions, slow release devices would be required to ensure regular delivery into the rumen. If the antimethanogenic agent were to be built into a pasture plant through selection or molecular genetics, the time required to achieve this and get the plant accepted and distributed through the national pasture should not be underestimated. Condensed tannins are an example. It is thirty years since CT was shown to prevent bloat in cattle (Jones et al, 1973) and have the potential to improve the efficiency of protein utilisation by the animal (Reid et al, 1974). Existing plants containing CT, such as Lotus species, do not compete well with temperate pasture species under high soil fertility conditions and they have been slow to improve through selection. Molecular genetic techniques have had no success in introducing genes that produce CT into more agronomically competitive plants. Thirty years on we still do not have a competitive CT-containing pasture plant.

7.4.5 Immunisation

Scientists in Australia have registered patents for immunisation procedures that are claimed to reduce methane emission. They have developed a vaccine containing an antigen derived from rumen methanogenic microorganisms (Baker, 1998), and an immunogenic preparation that reduces the activity of rumen protozoa (Baker et al, 1997). The antimethanogenic vaccine is claimed to reduce methane in in vitro incubations, and significantly increase DM intake and wool growth. CSIRO plan to release a suitable vaccine on the market in 2007. A vaccine would be a valuable tool in providing a cost-effective and long-acting treatment to reduce methane emission and enhance animal production under grazing.

7.4.6 Manipulation of the rumen microbial ecosystem

The methanogenic archaea, which are highly efficient scavengers of hydrogen, are the main, but not the only, agents for converting hydrogen to methane in the rumen (Joblin, 1999). There is also evidence that the rumen can function satisfactorily in the absence of methanogens (Joblin 1999). There are many potential opportunities for mitigating methane through microbial intervention in the rumen such as: targeting methanogens with antibiotics, bacteriocins, or phage; removing protozoa from the rumen; and the development of alternative sinks for hydrogen such as reductive acetogenesis. All these opportunities are possible through microbial intervention. However, it is very early days in the realisation of these possibilities.

While production of methane in the rumen from carbon dioxide and hydrogen is carried out by the methanogenic archaea, there is a class of bacteria present in the rumen, the acetogens, whiich utilise carbon dioxide and hydrogen to produce acetic acid, a major nutrient of the ruminant. Acetogens do not compete well in the rumen compared to methanogens, so experiments are in progress to see if the microbial ecosystem can be manipulated to enhance acetogen activity (Joblin, 1999). One strategy is to genetically modify acetogens so that they can compete more effectively in the rumen.

7.4.7 Management of methane emissions at the farm-scale

There are many ways in which existing and new technologies could be applied through a farm systems approach to reduce methane emission. These very important mitigation strategies are dealt with in detail in Chapter 8.

7.4.8 Conclusions

The best short-term possibility for mitigating methane emissions would seem to be via improvements in animal efficiency. Various measures can be taken to dilute the amount of methane associated with the animal's maintenance requirement, such as increasing feed intake or decreasing the proportion of cell wall carbohydrate in the diet. These factors should be evaluated with systems and farm-scale research. There are possibilities for mitigation by exploiting variation between animals via genetic selection, but these await verification of the effect and selection of markers.

A wide range of feed additives for reducing methane has been tested over the last 40 years. Not one has been incorporated into farm practice for many reasons: toxicity, adaptation by the rumen microbes, expense, and lack of a suitable delivery system. It is pointless pursuing this type of research, unless at the outset of the research a delivery system is proposed and evaluated in a cost-effective way. In the grazing situation, delivery of compounds with antimethanogenic activity via pasture plant breeding is a possibility, but the estimation of the time required must be realistic.

Immunisation against methanogens is an interesting concept, but proof of efficacy is not available. A 20% reduction in methane would be helpful, but far from optimum.

Increased understanding of rumen microbiology is the most likely pathway to control of methanogenesis, because this work will address the biosynthesis of methane by the archaea. For the reason that this remains the major unknown in the area of mitigation research, it is by default the area of research with the best chance of finding a solution. However, the rumen ecosystem is extremely complex and such research will be long-term, expensive and high risk.

7.5 Current Research on Methane in New Zealand

Research into methane emissions is carried out predominantly at AgResearch, NIWA and Landcare. There is an informal group, MetNet, comprising the three major players and several other interested groups that meet occasionally to discuss common interests. Current research trends are summarised below and details are given in Appendix 2.

7.5.1 Inventory

The primary driver of inventory development has been the need of MfE and MAF to report to the UNFCCC on a regular basis. Until recently, reports have been based on a model developed by Ulyatt et al (1991). In 1995 NIWA invited a team from Washington State University, led by Dr Kristen Johnson, to New Zealand to demonstrate their SF₆ tracer technique for measuring methane emission from grazing animals. Since that time SF₆ has been used to measuremethane emission from cattle and sheep over a wide range of pastoral conditions in a series of collaborative experiments between NIWA and AgResearch that have also involved Landcare and Dexcel. On several occasions the opportunity has been taken to measure concurrently methane emission from soil, grazing animals (SF₆) and with a range of meteorological techniques. Recently, Clark and Ulyatt (2002) revised the official inventory and incorporated methane emission factors derived from the measurement programme.

Programmes to refine the inventory model by measuring emission from various types and classes of livestock and pasture types continue, led by AgReseach.

7.5.2 Mitigation

Most of the current work on methane mitigation is carried out at AgResearch, Palmerston North. The ongoing research programme will form the cornerstone of the research portfolio of the PGGRC (see Chapter 3). The AgResearch group will also be conducting collaborative work with NIWA and Dexcel in the area of forages and diet supplements, the effect of animal maturity on its methane emission, and will possibly look for genes associated with low methane emission in dairy cows with Via Lactia.

7.5.3 Methodology

There is considerable expertise in New Zealand in meteorological methods for assessing greenhouse gas emissions. This work continues to be developed at both NIWA and Landcare. These technologies are aimed at estimating large scale or regional emissions and can be used to verify inventory calculations. Satellite approaches to estimate emission that correlate with pasture quality via NIR measurements are also being investigated.

Dexcel and NIWA are looking for new methods for determining the emissions of individual animals to replace the costly SF₆ methodology.

7.5.4 Models and systems approaches

There is an urgent need for a decision support model that is capable of evaluating the effectiveness of both existing and new technologies for reducing methane emission. This model needs to be used in conjunction with on-farm systems experiments to test alternative strategies. Several groups within New Zealand are either working on such models or contemplating such work. Thus AgResearch have developed the methane inventory model, are evaluating an Australian dairy model and looking to adapt the existing Stockpol and OVERSEER models, while Dexcel are developing a decision support model for dairying and MOTU are developing a land use model for New Zealand. There is an urgent need for all the modeling people to collaborate to produce the minimum number of models needed to do the job effectively. The resources in a small country are not large enough for the present apparent ad hoc approach.

References

Armstrong DG (1964): Evaluation of artificially dried grass as a source of energy for sheep. II - The energy value of cocksfoot, timothy and two strains of rye-grass at varying stages of maturity. Journal of Agricultural Science, Cambridge 62: 399-416.

Asanuma A, Iwamoto M & Hino T (1999): Effect of the addition of fumarate on methane production by ruminal microorganisms in vitro. Journal of Dairy Science, 82: 780-787.

Baker SK (1998): Method for Improving Utilisation Nutrients by Ruminants or Ruminant-like Animals. Patent application: international publication number WO 95/11041.

Baker SK, Gnanasampanthan G, Purser DB & Hoskinson RM (1997): Immunogenic preparation and method for improving the productivity of ruminant animals. Patent application: international publication number WO 97/00086.

Bauchop T & Mountfort DO (1981): Cellulose fermentation by a rumen anaerobic fungus in both the absence and presence of rumen methanogens. Applied and Environmental Microbiology, 42: 1103-1110.

Bauman DE, Eppard PJ, de Geeter MJ & Lanza GM (1985): Responses of high-producing dairy cows to long-term treatment with pituitary somatotrophin and recombinant somatotrophin. Journal of Dairy Science, 68: 1352-1362.

Benchaar C, Rivest J, Pomar C & Chiquette J (1998): Prediction of methane production from dairy cows using existing mechanistic models and regression equations. Journal of Animal Science, 76: 617-627.

Benchar C, Pomar C & Chiquette J (2001): Evaluation of diet strategies to reduce methane production in ruminants: a modeling approach. Canadian Journal of Animal Science, 81: 563-574.

Blaxter KL & Clapperton JL (1965): Prediction of the amount of methane produced by ruminants. British Journal of Nutrition, 19: 511-522.

Blaxter KL, Wainman FW (1964): The utilisation of the energy of different rations by sheep and cattle for maintenance and for fattening. Jounal of Agricultural Science, Cambridge 63: 113-128.

Callaway TR, Martin SA, Wampler JL, Hill NS & Hill GM (1997): Malate content of forage varieties commonly fed to cattle. Journal of Dairy Science, 80: 1651-1665.

Cammell SB, Beever DE, Sutton JD, France J, Alderman G & Humphries DJ (2002): An examination of energy utilisation in lactating dairy cows receiving a total mixed ration based on maize silage. Animal Science, 585-596.

Carro MD, Lpez S, Valdes C & Ovejero FJ (1999): Effect of dl-malate on mixed ruminal microorganism fermentation using the rumen simulation technique (RUSITEC). Animal Feed Science and Technology, 79: 279-288.

Chalupa W (1980): Chemical control of rumen microbial activity. In Digestive Physiology and Metabolism in Ruminants: 325-347. (Eds) Ruckebusch Y & Thivend P. MTP Press, Lancaster, England.

Clark H & Ulyatt MJ (2002): A Recalculation of Enteric Methane Emissions from New Zealand Ruminants 1990-2000 with Updated Emission Predictions for 2010. A report prepared for MAF by AgResearch, Palmerston North. 38p.

Czerkawski JW, Blaxter KL & Wainman FW (1966a): The metabolism of oleic, linoleic and linolenic acids by sheep with reference to their effects on methane production. British Journal of Nutrition, 20: 349-362.

Czerkawski JW, Blaxter KL & Wainman, FW (1966b): The effect of linseed oil and of linseed oil fatty acids incorporated in the diet on the metabolism of sheep. British Journal of Nutrition, 20: 485-494.

Davison, RM (2000): Genetic Improvement: Importance to Sheep and Beef Farming. Paper No. G2159, Meat and Wool Economic Service of New Zealand. Wellington. 6p.

Dohme F, Machmuller A, Estermann BL, Pfister P, Wasserfallen A & Kreuzer M (1999): The role of the rumen ciliate protozoa for methane suppression caused by coconut oil. Letters in Applied Microbiology, 29: 187-192.

Dong Y, Bae HD, McAllister TA, Mathison GW & Cheng KJ (1997): Lipid-induced depression of methane production and digestibility in the artificial rumen system (RUSITEC). Canadian Journal of Animal Science, 77: 269-278.

Ekern A, Blaxter KL & Sawers D (1965): The effect of artificial drying on the energy value of grass. British Journal of Nutrition, 19: 417-434.

Ferris CP, Gordon FJ, Patterson DC, Porter MG & Yan T (1999): The effect of genetic merit and concentrate proportion in the diet on nutrient utilisation by lactating dairy cows. Journal of Agricultural Science, 132: 483-490.

Flessa H, Dorsch P, Beese F, Konig H & Bouwman AF (1996): Influence of cattle wastes on nitrous oxide and methane fluxes in pasture land. Journal of Environmental Quality, 25: 1366-1370.

Fonty G & Morvan, B (1996): Ruminal methanogenesis and its alternatives. Annales de Zootechnie, 45: 313-318.

Gupta R, Chaudhary UB & Chauhan TR (1993): Effect of protein supplementation and defaunation on the growth rate of young buffalo heifers. Indian Journal of Animal Sciences, 63:179-183.

Hancock KR & Ulyatt MJ (2001): Opportunities in molecular biology: the enhancement of the nutritive value of forages. In Proceedings of the XIX International Grassland Congress: pp627-632.

Hegarty RS (1999): Reducing rumen methane emissions through elimination of rumen protozoa. Australian Journal of Agricultural Research, 50: 1321-1327.

Hegarty RS (2001): Greenhouse Gases from the Australian Livestock Sector: What do we Know, What can we Do? The Australian Greenhouse Office, Canberra. 32p

Immig I (1996): The rumen and hindgut as a source of ruminant methanogenesis. Environmental Monitoring and Assessment, 42: 57-72.

Islam M, Abe H, Terada F, Iwasaki K & Tano R (2000): Effects of levels of feed intake and inclusion of corn on rumen environment, nutrient digestibility, methane emission and energy and protein utilisation by goats fed alfalfa pellets. Asian Australian Journal of Animal Sciences, 13: 948-956.

Jarvis SC, Lovell RD & Panayides R (1995): Patterns of methane emission from excreta of grazing animals. Soil Biology and Biochemistry, 27: 1581-1588.

Jarvis GN, Strompl C, Burgess DM, Skillman LC, Moore ERB & Joblin KN (2000): Isolation and identification of ruminal methanogens from grazing cattle. Current Microbiology, 40: 327-332.

Joblin KN (1999): Rumen acetogens and their potential to lower ruminant methane emissions. Australian Journal of Agricultural Research, 50: 1307-1313.

Joblin KN, Naylor GE & Williams AG (1990): The effect of Methanobrevibacter smithii on the xylanolytic activity of anaerobic rumen fungi. Applied and Environmental Microbiology, 56: 2287-2295.

Johnson KA & Johnson DE (1995): Methane emissions from cattle. Journal of Animal Science, 73: 2483-2492.

Johnson DE & Johnson KA (2002): Recent Developments in Understanding Enteric Methane Production by Ruminants; Implications for Mitigation. International Non-CO₂ GHG Mitigation Workshop, Washington DC, 2-3 December 2002.

Johnson K, Huyler M, Westberg H, Lamb B & Zimmerman P (1994): Measurement of methane emissions from ruminant livestock using an SF₆ tracer technique. Environmental Science and Technology, 28: 359-362.

Jones WT, Anderson LB & Ross MD (1973): Bloat in cattle XXXIX. Detection of protein precipitants (flavolans) in legumes. New Zealand Journal of Agricultural Research, 16: 441-446.

Judd MJ, Kelliher FM, Ulyatt MJ, Lassey KR, Tate KR, Shelton ID, Harvey MJ & Walker CF (1999): Net methane emissions from grazing sheep. Global Change Biology, 5: 647-657.

Kirchgessner M, Windisch W, Muller HL (1995): Nutritional factors for the quantification of methane production. In Ruminant Physiology: Digestion, Metabolism, Growth and Reproduction: 333-348. (Eds) Engelhardt W, Leonhard-Marek S, Breves G & Gieseke D. Ferdinand Enke Verlag, Stuggart.

Klieve AV & Hegarty RS (1999): Opportunities for biological control of ruminal methanogenesis. Australian Journal of Agricultural Research, 50: 1315-1319.

Krause M (2002): Effect of Malate and Fumarate Supplementation on Methane Production by Ruminal Microorganisms. Report to AgResearch. 14p.

Lassey KR, Ulyatt MJ, Martin RJ, Walker CF & Shelton ID (1997): Methane emissions measured directly from grazing livestock in New Zealand. Atmospheric Environment, 31: 2905-2914.

Lockyer DR (1997): Methane emissions from grazing sheep and calves. Agriculture, Ecosystems and Environment, 66: 11-18.

Lockyer DR & Jarvis (1995): The measurement of methane losses from grazing animals. Environmental Pollution, 90: 383-390.

Lopez S, Valdes C, Newbold CJ & Wallace RJ (1999): Influence of sodium fumarate addition on rumen fermentation in vitro. British Journal of Nutrition, 81: 59-64.

Machmuller A, Ossowski DA, Wanner M & Kruezer M (1998): Potential of various fatty feeds to reduce methane release from rumen fermentation in vitro (RUSITEC). Animal Feed Science and Technology, 71: 117-130.

Martin SA (1998): Manipulation of ruminal fermentation with organic acids: a review. Journal of Animal Science, 76: 3123-3132.

Mathison GW, Okine EK, McAllister TA, Dong Y, Galbraith J, & Dmytruk OIN (1998): Reducing methane emission from ruminant animals. Journal of Applied Animal Research, 14: 1-28.

McAllister TA, Okine EK, Mathison GW & Cheng KJ (1996): Dietary, environmental and microbiological aspects of methane production in ruminants. Canadian Journal of Animal Science, 76: 241-243.

McCaughey WP, Wittenberg K & Corrigan D (1997): Methane production by steers on pasture. Canadian Journal of Animal Science, 77: 519-524.

McCaughey WP, Wittenberg K & Corrigan (1999): Impact of pasture type on methane production by lactating beef cows. Canadian Journal of Animal Science, 79: 221-226.

McCrabb GJ & Hunter RA (1999): Prediction of methane emissions from beef cattle in tropical production systems. Australian Journal of Agricultural Research, 50: 1135-1139.

McCrabb GJ, Berger KT, Magner T, May C & Hunter RA (1997): Inhibiting methane production in Brahman cattle by dietary supplementation with a novel compound and effects on growth. Australian Journal of Agricultural Research, 48: 323-329.

Miller TL (1995): Ecology of methane production and hydrogen sinks in the rumen. In Ruminant Physiology: Digestion, Metabolism, Growth and Reproduction: 317-331. (Eds) von Engelhardt W, Leonhard-Marek S, Breves G & Gieseke D. Ferdinand Enke Verlag, Stuttgart.

Miller TL & Wolin MJ (2001): Inhibition of growth of methane-producing bacteria of the ruminant forestomach by hydroxymethylglutaryl~SCoA reductase inhibitors. Journal of Dairy Science, 84: 1445-1448.

Moe PW & Tyrell HF (1979): Methane production in dairy cows. Journal of Dairy Science, 62: 1583-1586.

Mosier AR, Duxbury JM, Freney JR, Heinemeyer O, Minami K & Johnson DE (1998): Mitigating agricultural emissions of methane. Climatic Change, 40: 39-80.

Moss AR, Jouany JP & Newbold J (2000): Methane production by ruminants: its contribution to global warming. Annales de Zootechnie, 49: 231-253.

Murray RM, Bryant AM & Leng RA (1976): Rates of production of methane in the rumen and large intestine of sheep. British Journal of Nutrition, 36: 1-14.

Nollet L, Demeyer DI & Verstraete W (1997): Effect of bromoethanesulfonic acid and Peptostreptococcus productus ATCC 35244 addition on stimulation of reductive acetogenesis in the ruminal ecosystem by selective inhibition of methanogenesis. Applied and Environmental Microbiology, 63: 194-200.

NZCCP (2001): National Communication 2001: New Zealand's Third National Communication under the Framework Convention on Climate Change. Ministry for the Environment, Wellington. 105p.

Ørskov ER, Ojwang I & Reid GW (1988). A study on consistency of differences between cows in rumen outflow rate of fibrous particles and other substrates and consequences for digestibility and intake of roughages. Animal Production, 47: 45-51.

Pavao-Zuckerman MA, Waller JC, Ingle T & Fribourg HA (1999): Methane emissions of beef cattle grazing tall fescue pastures at three levels of endophyte infection. Journal of Environmental Quality, 28: 1963-1969.

Pelchen A & Peters KJ (1998): Methane emissions from sheep. Small Ruminant Research, 27: 137-150.

Pinares-Patiño C (2000): Methane Emission from Forage-fed Sheep: A Study of Variation Between Animals. PhD thesis, Massey University. 226p.

Pinares-Pantiño C, Ulyatt MJ, Lassey KR, Barry TN, Holmes CW & Johnson DE (2003a): Methane emission by alpaca and sheep fed on lucerne hay or grazed on pastures of perennial ryegrass/white clover or birdsfoot trefoil. Journal of Agricultural Science, Cambridge, (in press).

Pinares-Patiño CS, Ulyatt MJ, Lassey KR, Barry TN & Holmes CW (2003): Persistence of the between-sheep variation in methane emission. Journal of Agricultural Science, Cambridge, (in press).

Reid CSW, Ulyatt MJ & Wilson JM (1974): Plant tannins and nutritive value. Proceedings of the New Zealand Society of Animal Production, 34: 82-92.

Shiao TF, Liu SC, Chen CP & Lee CF (1999): Effects of dietary grain to forage ratio on methane production in the digestive tract of Holstein dairy cows. Journal of the Chinese Society of Animal Science, 28: 437-449.

Taylor R & Smith I (1997): The State of New Zealand's Environment. The Ministry for the Environment, GP Publications, Wellington. pp5.30-5.32.

Ulyatt MJ (1996): Current Status of Non-CO₂ Greenhouse Gas Emissions from Agriculture in New Zealand: Potential for Mitigation. National Science Strategy Committee for Climate Change. Information Series 8. Wellington, SIR Publishing. 21p.

Ulyatt MJ, Fennessy PF, Rattray PV, Jagusch KT (1976): The nutritive value of supplements. In Supplementary Feeding: 157-184. (Eds) Drew KR, Fennessy PF. Occasional Paper No. 7. New Zealand Society of Animal Production.

Ulyatt MJ, Thomson DJ, Beever DE, Evans RT & Haines MJ (1988): The digestion of perennial ryegrass (Lolium perenne cv. Melle) and white clover (Trifolium repens cv. Blanca) by grazing cattle. British Journal of Nutrition, 60: 137-149.

Ulyatt MJ, Lassey KR, Shelton ID & Walker CF (2002a): Methane emission from dairy cows and wether sheep fed sub-tropical grass-dominant pastures in mid summer in New Zealand. New Zealand Journal of Agricultural Research, (in press).

Ulyatt MJ, Lassey KR, Shelton ID & Walker CF (2002b): Seasonal variation in methane emission from dairy cows and breeding ewes grazing ryegrass/white clover pasture in New Zealand. New Zealand Journal of Agricultural Research, (in press).

Ulyatt MJ, Baker SK, McCrabb GJ & Lassey KR (1999): Accuracy of the SF₆ tracer technology and alternatives for field measurements. Australian Journal of Agricultural Research, 50: 1329-1334.

van Nevel CJ & Demeyer DI (1995): Feed additives and other interventions for decreasing methane emissions. Biotechnology in Animal Feeds and Feeding, 17: 329-349.

van Nevel CJ & Demeyer DI (1996): Control of rumen methanogenesis. Environmental Monitoring and Assessment, 42: 73-97.

Vermorel M (1995): Emissions annuelles de methane d'origine digestive par les bovins en France. INRA Productions Animales, 8: 2654-272.

Waghorn GC, Tavendale MH & Woodfield DR (2002a): Methanogenesis from forages fed to sheep. Proceedings of the New Zealand Grassland Association, 64: 167-171.

Waghorn GC, Lassey KR, Kolver E & Molano G (2002b): Methane Emission from Dairy Cows Fed either Pasture or Balanced Rations over one Lactation. A report prepared for the New Zealand Dairy Board Global Programme, Palmerston North. 35p.

Whitford MF, Teather RM & Forster RJ (2001): Phylogenetic analysis of methanogens from the bovine rumen. BMC Microbiology, www.biomedcentral.com/1471-2180/1/5.

Williams DJ (1993): Methane emissions from manure of free-range dairy cows. Chemosphere, 26: 179-187.

Williams AG, Withers SE & Joblin KN (1994): The effect of cultivation with hydrogen-consuming bacteria on xylanolysis by Ruminococcus flavifaciens. Current Microbiology, 29: 133-138.

Wolin MJ, Miller TL & Stewart CS (1997): Microbe-microbe interactions. In The Rumen Microbial Ecosystem: 467-491. (Eds) Hobson PN & Stewart CS. Blackie Academic and Professional, London.

Woodward SL, Waghorn GC, Ulyatt MJ & Lassey KR (2001): Early indications that feeding Lotus will reduce methane emissions from ruminants. In Proceedings of the New Zealand Society of Animal Production, 61: 23-26.

Woodward SL, Waghorn GC, Lassey KR & Laboyrie PG (2002): Does feeding sulla (Hedysarum coronarium) reduce methane emissions from dairy cows? In Proceedings of the New Zealand Society of Animal Production, 62: 227-230.

Yamulki S, Jarvis SC & Owen P (1999): Methane emission and uptake from soils as influenced by excreta deposition from grazing animals. Journal of Environmental Quality, 28: 676-682.

Contact for Enquiries

MAF Information Services
Pastoral House
25 The Terrace
PO Box 2526
Wellington, NEW ZEALAND

Fax: +64 4 894 0721
Contact this person