Note: We’ve included our research, calculation formulas and source links so you can explore the science in more detail. All links were current as of 19 November 2026.
Before we compare methane emissions, let’s clear up a common misunderstanding. Cattle and sheep release most of their digestive methane through their mouths when they burp. Their bums are responsible for a much smaller share.[19, 23]

Humans work differently. Methane made by microbes in our intestines can leave through our bums when we pass wind. Ordinary human burps mainly release air we have swallowed. There is one important detail, though: some methane from our intestines enters the blood, travels to the lungs and is breathed out. Saying that human methane never leaves in our breath would be incorrect. [24, 25]
When comparing livestock methane emissions with fossil fuels, we need to ask how much gas escapes and how it affects warming. We also need to recognise improvements that actually reduce emissions.
Scientists have updated some older livestock calculations to better match Australian animals and diets. Breeding research also shows that some cattle and sheep can pass lower methane traits to their offspring. These improvements matter, but methane from both farming and fossil fuels still needs careful measurement. [4, 5, 6, 10, 11, 12]
Sources of livestock and fossil fuel methane
Cattle and sheep are ruminants. They have a large stomach compartment called the rumen, where tiny living things called microbes help break down grass and other feed. Some of these microbes make methane. Scientists call this process enteric fermentation. [3, 19]
Most of that gas comes out in burps. Manure can produce more methane, depending on how it is stored and handled. [3, 23]
Fossil-fuel methane comes from coal mines and the production, processing and transport of oil and natural gas. Some gas leaks from equipment. Some is deliberately released, or vented. Methane can also escape when a flare does not burn the gas completely. [17]
Once in the air, methane traps heat whether it came from a sheep or a gas pipe. Both sources matter. The origin of the carbon and changes in emissions over time help explain their wider climate effects. [1, 13]
Recently grown grass and ancient carbon
The carbon in a grazing animal’s methane usually came from plants that recently took carbon dioxide out of the air. When methane breaks down, much of that carbon returns to carbon dioxide. This movement of carbon through plants, animals and the air is called the biogenic carbon cycle. [13, 19]
Fossil methane brings carbon from underground stores into the atmosphere. Its breakdown adds fossil carbon dioxide too. Burning coal, oil and gas releases further carbon dioxide, which also needs to be counted. [1, 17]
The natural carbon cycle does not cancel methane’s warming while the gas is in the air. Grass growing again does not automatically make a farm climate neutral.
Methane is removed gradually, with an effective atmospheric lifetime of about 12 years. This is an average, so the gas does not all disappear after exactly 12 years. Carbon dioxide behaves differently: part of a release can affect the climate for centuries or longer. These differences matter when we compare warming over time. [1, 13]
How livestock methane is calculated
There is no single calculation used equally for every cow and sheep worldwide.
The Intergovernmental Panel on Climate Change, or IPCC, provides methods with different levels of detail. Its simplest method, called Tier 1, multiplies animal numbers by an average estimate of annual methane per animal. Tier 2 uses more information about animals and their feed. Tier 3 can use more detailed models and measurements. [3]

Australia uses methods developed for local conditions. They consider animal groups, estimated feed intake, production and seasonal changes. Sheep, grazing cattle and feedlot cattle have different equations. [6]
One important term is dry-matter intake. This means the weight of feed eaten after excluding its water. Two animals eating different amounts or diets may produce different amounts of methane.
A good calculation still needs good information. If animal numbers or feed estimates are wrong, the emissions estimate can be wrong too. [3]
Older equations have already been improved
A CSIRO-led Australian study published in 2016 examined 1,034 animal records. It proposed a calculation for cattle eating mostly grass or other forage: 20.7 grams of methane for each kilogram of dry matter eaten. Using that approach would lower estimated forage-cattle emissions by 24% compared with the previous method. [4]
That was a correction to the estimate. The animals had not suddenly changed their digestion when the paper appeared.
In 2025, University of New England researchers tested equations against Australian feedlot cattle eating barley- and wheat-based diets. The older Moe and Tyrrell equation came from US dairy-cattle research in 1979. In the Australian dataset, it predicted an average of 115 grams more methane per animal per day than was measured. [5]
The researchers developed improved equations. Australia’s National Inventory Report 2023, submitted in 2025, recorded adopting their equation that includes feed composition. The old method had therefore already been updated. [6]
The study covered particular feedlot cattle and diets. Its correction cannot simply be applied to every grazing animal or sheep flock.
Sheep need their own evidence
Sheep are not miniature cattle when it comes to emissions calculations. Their size, feed intake, pregnancy, milk production and management can affect methane. Australia’s documented sheep method links methane to estimated feed intake and considers different animal groups and seasons. [6]
A 2026 international analysis of 24 studies found that improving forage quality did not produce the same result in dairy cattle, other cattle and sheep. [21]
The answer also changed depending on what researchers compared. Methane per kilogram of total feed gave different results from methane per kilogram of the organic material that was actually digested. Before accepting a claim of “lower methane”, check what was measured and what it was compared with.
Better equipment still needs careful sampling
Scientists measure livestock methane in several ways. Respiration chambers are spaces where they measure gases going in and out. GreenFeed units collect repeated gas samples when animals visit a special feeding station. Portable accumulation chambers allow shorter measurements, particularly in sheep. [3, 7, 9]
A brief measurement captures only a small part of an animal’s day. Feeding times, the length of a visit and differences between instruments can affect the result.
A 2026 University of Queensland and University of New England study recorded 453 grazing cattle across four northern Australian sites. It found that visit patterns, sampling time, equipment and feed conditions affected methane estimates. [7]
Wageningen University researchers have also studied how dairy-cow methane changes through the day. Their 2025 model used chamber data from 28 cows. It worked well with those observations, but that does not prove it will work equally well in every herd. [8]
Repeated, well-planned measurements help researchers separate an animal’s usual methane output from a temporary change in what or when it ate.
Measuring methane and calculating warming are different jobs
After estimating methane in kilograms or tonnes, researchers often convert it into carbon dioxide equivalent, written CO₂-e. This is a comparison unit, not a claim that the animal released that amount of actual carbon dioxide.
The 100-year Global Warming Potential, or GWP100, compares the heat-trapping effect of a release of gas over 100 years with the same weight of carbon dioxide. Australia’s latest national inventory, submitted in April 2026, uses an IPCC methane factor of 28. One tonne of methane is therefore reported as 28 tonnes of CO₂-e. [2]
GWP*, pronounced “GWP star”, looks at ongoing methane emissions and how their rate has changed. Oxford and other researchers developed it to better estimate the effect of those changes on temperature. [13, 14]
If methane emissions stay roughly steady for a long time, new methane largely replaces methane being removed. The source maintains warming it has already caused, while adding relatively little extra warming. Growing methane emissions add more warming. Sustained cuts can reduce methane’s contribution to temperature. Continued carbon dioxide emissions keep adding to the longer-lasting problem. [13]
GWP* can also examine changes in fossil-fuel methane. Fossil carbon dioxide still needs to be counted. GWP* adds useful information, but it has not replaced GWP100 in Australia’s national inventory. [13, 14]
The IPCC warming factors were also corrected
Older explainers can miss another update.
In a correction dated 7 October 2025, the IPCC revised its Sixth Assessment Report methane table. The corrected central GWP100 estimates are 25.3 for non-fossil methane and 28.1 for fossil methane. Over 20 years, the values are 74.7 and 77.4. These replace the original table’s widely repeated 100-year values of 27.0 and 29.8. [1]
For example, a GWP100 of 25.3 compares the 100-year heat-trapping effect of one tonne of methane with 25.3 tonnes of carbon dioxide.
Updating scientific estimates does not automatically change official reporting rules. Australia still uses the earlier factor of 28 specified for its inventory.
A fair comparison must state which set of factors and timescale it uses. Switching factors between sectors or years can make the numbers fall without reducing the gas released.
What changing methane emissions look like
Imagine a source that released 100 tonnes of methane a year 20 years ago. The table shows three possible current levels. Both calculations use the same factor of 28, so the difference comes from the method. These are worked examples, not measured farm results. [14, 15]
| Methane now tonnes per year | GWP100 tonnes CO₂-e per year | GWP* tonnes CO₂-we per year |
| 90 | 2,520 | −484 |
| 100 | 2,800 | 784 |
| 110 | 3,080 | 2,052 |
CO₂-we means carbon dioxide warming equivalent. The minus sign shows a falling contribution to temperature compared with the source’s earlier emissions. It does not mean carbon dioxide was physically removed from the air. The two columns answer different questions and cannot be swapped as though they were the same emissions total.
“Little additional warming” also does not mean “no climate impact”. Earlier emissions caused warming that a continuing source can maintain. Scientists from institutions including the Australian National University, Duke University and Imperial College London have warned against using this distinction to justify weak methane targets. [16]
GWP* is an estimate, too. A 2025 study found limits when methane emissions fell very quickly and when the method was used far beyond the period it was designed to cover. [15]
Selective breeding can deliver lasting improvements
Selective breeding means choosing animals with useful inherited features to produce the next generation. For methane research, scientists also consider growth, fertility and health. The aim is to breed animals that release less methane while keeping other important qualities.
New Zealand’s sheep programme provides useful evidence. Its research organisation reports that high- and low-methane lines differ by close to 25% after five generations of selection. That comparison comes from specially selected research flocks. Results on commercial farms depend on which animals farmers breed and how widely they use the approach. [11]
For example, reducing methane per animal by 1% each year would add up to about 9.6% over ten years. This is an arithmetic example, not a forecast of breeding results.
Researchers can compare daily methane, methane per kilogram of feed, or methane after allowing for an animal’s size or growth. These measures can rank animals differently. A 2026 Irish study using records from 8,182 sheep showed why choosing the right measure matters. [9]
A September 2026 University of New England study also found inherited differences in cattle methane traits. It examined how these related to feed intake, growth and weight. [10]
The Zero Net Emissions from Agriculture Cooperative Research Centre supports Australian work on breeding tools using genetic information. Partners include the universities of New England, Queensland and Western Australia. Its expected future reductions are research-based projections, not cuts already achieved across all Australian herds. [12]
Lower emissions per animal and lower totals
Emissions intensity means how much gas is released for each unit of product, such as a litre of milk or kilogram of meat. Producing more with the same emissions lowers that intensity. Breeding can also reduce methane compared with the feed an animal eats.
These improvements deserve recognition, but total emissions matter too. If methane per animal falls 10% while animal numbers rise 20%, total methane increases 8%. That is an arithmetic example, not a forecast for Australia.
Calculators must also be able to recognise improvements. A model that assumes the same methane output for every kilogram of feed may miss some gains from breeding. Scientists need measurements and information about how many farmers use an approach before they can update an inventory reliably. [3, 6, 10]
Diet, animal health and management matter alongside breeding. The University of Western Australia’s 2026 report on a 17-month sheep trial described measuring animal performance and health alongside methane. An approach that works in a controlled feeding trial may perform differently in an open paddock. [20, 22]
Fossil fuel methane needs the same scrutiny
Energy companies need accurate accounting too.
The International Energy Agency’s Global Methane Tracker 2026 estimated that fossil-fuel operations released 124 million tonnes of methane in 2025: 45 million from oil, 43 million from coal and 36 million from natural gas. Its estimates draw on industry information, measurements, scientific studies and satellites. [17]

The Global Methane Budget published in 2025 reported inventory estimates of 114–124 million tonnes from livestock digestion and manure in 2020. That includes several livestock species, not only cattle and sheep. The different years and groups mean these figures cannot provide an exact same-year comparison of cattle and sheep with fossil fuels. [19]
A July 2026 Harvard-led satellite study estimated US oil-and-gas methane emissions at 64% above the government inventory used for comparison. It also found a modest fall in that sector’s total emissions during 2019–2024. Measurements can reveal both undercounting and genuine progress. These US findings cannot simply be applied to Australia. [18]
For every sector, estimates should be checked against evidence. Reports should explain what they include and how certain the numbers are.
Fair accounting must lead to real reductions
Farmers who make measurable improvements through breeding, management and animal health deserve recognition. Fossil-fuel operators should also find and reduce methane releases, with independent checks.
A full climate comparison must look beyond methane. Livestock systems can also produce nitrous oxide, use fossil energy and involve land clearing. Fossil-fuel systems release large amounts of carbon dioxide as well as methane. One gas cannot tell us the whole environmental story. [2, 3, 17]
At Habitat for the Future, we support accurate science alongside habitat protection. Native trees and woodland matter for wildlife and farming landscapes. Carbon storage should be measured separately, rather than assumed from a few trees in a paddock.
How much gas reached the atmosphere? Has that amount fallen? What does the emissions history mean for warming? Are we protecting the habitats around us while making improvements?
Better calculations should help us answer those questions honestly and act on the answers.
One Planet. One Home. One Chance.
Appendix A Calculation methods
The equations below show how scientists estimate methane and compare warming. Each letter stands for a quantity explained beside it. The results apply to the types of animals and diets for which each method was developed.
Methane released by livestock
M is daily methane in grams per animal. I is dry-matter intake in kilograms per animal per day. In the feedlot equations, EE is ether extract, a measure of fat, and NDF is neutral detergent fibre. Both are percentages of feed dry matter, entered as percentage points, for example 4 and 20.
Australian forage cattle relationship published in 2016

This equation was developed for cattle eating mostly grass or other forage. It estimates an average relationship and cannot predict every animal exactly. [4]
Older feedlot equation evaluated in 2025

SR, HC and CEL are the daily amounts of different parts of the feed, measured in kilograms: soluble residue, hemicellulose and cellulose. The last two are types of plant fibre. This is the older equation as printed in Table 2 of the Australian study. [5]
Diet sensitive Australian feedlot equation adopted in 2025

The national inventory documents adoption of this equation, rather than the paper’s simpler intake-only alternative. For an illustrative intake of 10 kg/day, EE of 4% and NDF of 20%, it predicts 80.3 g methane/day. This is a calculation, not a measured result. [5, 6]
Sheep relationship documented in the 2023 inventory

This expresses the documented kilogram-based equation in grams. Sheep intake and annual totals are estimated separately by animal class and conditions. [6]
Scaling daily estimates to a herd or flock

For each class, N is the average number of animals, D is days in that class and T is methane in tonnes over that period. Add the classes using consistent boundaries. Physical reductions and revised estimates are different. [3]
Comparing the warming algorithms
The article keeps the reporting factor at 28 in both calculations so that only the metric changes. M(t) is methane emitted in year t, in tonnes per year, and M(t − 20) is the rate 20 years earlier. E100 is the GWP100 result in tonnes CO₂-e per year.

The earlier equation developed by Cain and colleagues gives more weight to changes in emissions, with a smaller term for ongoing emissions. [15]

The improved equation below, with rounded numbers, was used for the article’s table. [14, 15]

For a stable 100 tonnes of methane per year, the refined formula gives 784 tonnes CO₂-we per year. At 90 tonnes now versus 100 tonnes 20 years earlier, it gives −484.4 tonnes CO₂-we per year. These are approximate warming equivalents relative to history, not emissions removals. The formulation and factor must be stated when comparing results.
Scientific factor updates and reporting conventions
| Basis | Non fossil methane GWP100 | Fossil methane GWP100 |
| AR5 factor used in the Australian 2026 inventory | 28 | 28 |
| AR6 original Table 7.15 published in 2021 | 27.0 | 29.8 |
| AR6 corrected Table 7.15 in the October 2025 errata | 25.3 | 28.1 |
AR5 means the IPCC Fifth Assessment Report; AR6 means its Sixth Assessment Report. The numbers compare one tonne of methane with tonnes of carbon dioxide over 100 years. A corrected scientific estimate does not automatically change the factor required by official reporting rules. [1, 2]
Appendix B Sources of information and research notes
Climate research was checked on 18 November 2026. The added explanations of livestock burps and human methane were checked on 19 November 2026. Sources include recent peer reviewed research, Australian emissions reports, official scientific corrections and research programme updates. Earlier studies explain where the methods came from. Research findings, future projections and policy statements are identified separately.
The publication year is not always the emissions year. In particular, the Australian 2026 annual submission reports through 2023–24, the IEA 2026 fossil-fuel estimate describes 2025, and the 2025 Global Methane Budget describes data through 2020. Livestock studies also include both government and industry-funded research; conclusions here are tied to their stated populations and methods.
[1] IPCC (2025). Errata to Climate Change 2021 The Physical Science Basis. Version 7 October 2025, corrected Table 7.15.
Official scientific correction | Primary source
Corrects methane GWP100 central values to 25.3 for non-fossil and 28.1 for fossil methane, and GWP20 to 74.7 and 77.4. The errata state that the listed changes had not been implemented in the published documents, so the original chapter table can still show superseded values.
[2] DCCEEW (2026). National Inventory Report 2024 Volume 1. Submitted April 2026.
Current Australian national inventory | Primary source | Supporting record or full text
Latest annual inventory identified at the research date; covers financial years 1989–90 to 2023–24. Table ES.01 specifies the AR5 methane GWP100 factor of 28. This is the reporting convention used in the article’s worked warming example.
[3] IPCC (2019). 2019 Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories. Volume 4 Chapter 10 Emissions from Livestock and Manure Management.
International inventory methodology | Primary source
Explains livestock classification, feed and energy inputs, methodological tiers, enteric methane and manure accounting. Default factors and detailed methods serve different data conditions; a new research model is not automatically a universal reporting replacement.
[4] Charmley E and colleagues (2016). A universal equation to predict methane production of forage-fed cattle in Australia. Animal Production Science 56, 169–180. DOI 10.1071/AN15365.
Peer reviewed Australian model development | Primary source | Supporting record or full text
CSIRO-led analysis including UNE and government researchers. The 1,034-record analysis supported a forage-cattle relationship of 20.7 g methane per kg dry-matter intake. Its estimated 24% inventory reduction was a recalculation against the preceding method, not a measured reduction caused by intervention.
[5] Almeida AK, McMeniman JP, Van der Saag MR and Cowley FC (2025). Evaluation of methane prediction equations for Australian feedlot cattle fed barley and wheat-based diets. Animal Production Science 65, AN24212. DOI 10.1071/AN24212.
Peer reviewed Australian feedlot model evaluation | Primary source | Supporting record or full text
Database of 713 measurements from 175 cattle in 12 studies. Evaluated the older equation and developed local equations. Table 2 supplies the older formula; the paper gives the updated diet-sensitive model. Diet and validation boundaries matter.
[6] DCCEEW (2025). National Inventory Report 2023 Volume 1. Executive summary and Chapter 5 Agriculture.
Documented Australian method adoption | Primary source
Explicitly records implementing Almeida et al. (2025). Equations 3.A.1c_1 and 3.A.2_5 document the feedlot and sheep calculations; the grazing-cattle method uses 20.7 g methane per kg intake. This establishes adoption by the 2025 submission, rather than treating the research as still awaiting adoption.
[7] Factors affecting recording methane emission phenotypes of composite and crossbreed beef cattle grazing tropical and subtropical rangelands of Northern Australia (2026). Journal of Animal Science 104, skag050. DOI 10.1093/jas/skag050.
Peer reviewed grazing measurement study | Primary source
UQ and UNE collaboration, published 19 February 2026. Analyses 453 cattle across four trial sites and examines GreenFeed visitation, sample duration, equipment and environmental influences. Supports careful sampling, rather than a new single national coefficient.
[8] Modelling methane production of dairy cows A hierarchical Bayesian stochastic approach (2025). Computers and Electronics in Agriculture 228, 109683. DOI 10.1016/j.compag.2024.109683.
Peer reviewed measurement modelling | Primary source
Wageningen University and Research record. Models changing methane production through the day using respiration-chamber observations from 28 cows. Strong fit to this dataset is distinct from external validation across grazing systems or countries.
[9] Assessing alternative metrics of methane output measured in a multi-breed, pasture-based sheep population (2026). Journal of Animal Science 104, skag145. DOI 10.1093/jas/skag145.
Peer reviewed sheep trait comparison | Primary source
Irish Teagasc-led research, published 2 May 2026. Uses 15,385 records from 8,182 sheep to compare absolute, ratio and residual methane traits. These are animal trait definitions, rather than alternatives to GWP100 or GWP*.
[10] The genetic relationships between different definitions of emissions outputs and feedlot production traits (2026). Genetics Selection Evolution. DOI 10.1186/s12711-026-01086-1.
Peer reviewed cattle genetics | Primary source | Supporting record or full text
UNE research published September 2026. Reports genetic variation and relationships between methane traits and production. Its minimum-sample findings relate to the study’s genetic evaluation dataset; they should not be turned into universal monitoring requirements.
[11] Ag Emissions Centre. Breeding low-emitting sheep. Research programme update accessed 18 November 2026.
New Zealand research programme evidence | Primary source | Supporting record or full text
Describes low- and high-methane selection lines, inheritance and the Cool-Sheep programme. Reports a close-to-25% difference between selection lines after five generations. This is not a 25% reduction already achieved across the national flock.
[12] Zero Net Emissions Agriculture CRC. Selecting for more methane efficient cattle and sheep. Project information accessed 18 November 2026.
Active Australian research programme | Primary source | Supporting record or full text
UNE-led work with UQ, UWA, government and industry partners to improve genomic breeding values. Programme estimates of 7–12% cattle and up to 15% sheep methane reductions by 2040 are projections, not demonstrated industry-wide outcomes.
[13] Allen M, Lynch J, Cain M and Frame D (2022 update). Climate metrics for ruminant livestock. Oxford Martin School.
Researcher scientific briefing | Primary source
Explains short-lived methane, long-lived carbon dioxide and changing emissions rates. Its temperature-pathway discussion is useful alongside conventional inventories; maintaining existing warming is distinct from having caused no warming.
[14] Smith MA, Cain M and Allen MR (2021). Further improvement of warming-equivalent emissions calculation. npj Climate and Atmospheric Science 4, 19. DOI 10.1038/s41612-021-00169-8.
Peer reviewed GWP star refinement | Primary source | Supporting record or full text
Refines the warming-equivalent formulation. The worked example uses the rounded coefficients 0.28 and 4.25 also stated in the 2025 testing paper [15], with the same AR5 factor in both metric calculations.
[15] Testing GWP* to quantify non-CO2 contributions in the carbon budget framework in overshoot scenarios (2025). npj Climate and Atmospheric Science 8, 101. DOI 10.1038/s41612-025-00980-7.
Peer reviewed metric evaluation | Primary source | Supporting record or full text
Tests GWP* against climate-model temperature outcomes. Identifies limitations for very rapid methane reductions and beyond 2100. The accepted manuscript gives both the earlier 0.75/0.25 weighting and the refined formulation used here.
[16] Scientific Statement on Methane Temperature Neutrality and GWP*. Accessed 18 November 2026.
Primary scientific policy statement | Primary source
Signed by researchers including representatives of ANU, Duke, Imperial College London and other institutions. Recognises legitimate science behind GWP* while warning about its misuse to justify weak targets. This is a policy statement, not a new measurement experiment.
[17] International Energy Agency (2026). Global Methane Tracker 2026 Key findings and methodology.
Current global fossil fuel methane assessment | Primary source | Supporting record or full text
Reports 124 Mt fossil-fuel methane in 2025, separately identifying oil, coal and natural gas. Bioenergy is additional and is not included in that fossil-fuel total. Explains the use of operational data, measurements and satellites.
[18] Estrada LA and colleagues (2026). Quantifying national state and oil gas field methane emissions and trends in the US 2019–2024 through high resolution inversion of satellite observations. Atmospheric Chemistry and Physics 26, 10629–10646. DOI 10.5194/acp-26-10629-2026.
Peer reviewed satellite inversion study | Primary source
Harvard-led study published 30 July 2026. Estimates oil/gas methane 64% above the EPA inventory used as its prior, while finding declining oil/gas emissions over 2019–2024. Geographic and inventory-version boundaries prevent extrapolating that percentage to Australia.
[19] Saunois M and colleagues (2025). Global Methane Budget 2000–2020. Earth System Science Data 17, 1873–1958. DOI 10.5194/essd-17-1873-2025.
Peer reviewed global methane synthesis | Primary source
International Global Carbon Project assessment. Reports 2020 inventory estimates of 114–124 Mt methane from enteric fermentation and manure across livestock species. Publication year and emissions year differ; the article keeps those dates explicit.
[20] University of Western Australia (2026). UWA Farm Ridgefield trial delivers powerful insights on methane reduction. 13 April 2026.
University report on completed field trial | Primary source
Describes a 17-month grazing-sheep trial with methane, growth, reproduction and health monitoring. The news item does not provide a universal percentage reduction; the article uses it to explain why field validation and animal performance matter.
[21] van Gastelen S and colleagues (2026). Forage characteristics as drivers of variation in enteric methane emission in cattle and sheep A meta-analysis. Animal, 101910. DOI 10.1016/j.animal.2026.101910.
Peer reviewed international forage analysis | Primary source | Supporting record or full text
International collaboration represented in Aarhus University’s research record. Analyses 24 studies and 109 treatment means. Responses differ among dairy cattle, other cattle and sheep, and depend on whether feed intake or digestible organic matter is used as the denominator.
[22] DCCEEW (2026). Reducing methane from livestock. Updated 27 August 2026.
Australian research programme overview | Primary source
Summarises MERiL feed-additive and forage research and distinguishes controlled feeding from grazing outcomes. Useful context for delivery and field-performance limitations; this article does not assess the safety or suitability of individual products.
[23] CSIRO. FutureFeed. Accessed 19 November 2026.
Australian scientific organisation | Primary source
Explains enteric fermentation in cattle and sheep and confirms that digestive methane is mostly released in burps. Used here for the route of gas release, alongside the detailed livestock sources already cited.
[24] US National Institute of Diabetes and Digestive and Kidney Diseases. Symptoms and Causes of Gas in the Digestive Tract. Accessed 19 November 2026.
Official human digestion information | Primary source
Explains burping, swallowed air and passing gas from the intestines. Supports the distinction between ordinary human burps and gas produced lower in the digestive system.
[25] University Hospitals Plymouth NHS Trust (2025). Patient information sheet for the hydrogen methane breath test. Posted 5 November 2025, accessed 19 November 2026.
Official explanation of breath testing | Primary source
Explains that some hydrogen and methane produced in the bowel passes into the blood, travels to the lungs and is breathed out. Human methane therefore does not leave only through flatulence.

