Dairy Silage Inoculant Shows No Overall Methane Cut in 30-Cow Trial
A seven-week feeding trial in Ireland found no statistically significant overall methane reduction from a silage inoculant made with bacteriocin-producing lactic acid bacteria, despite a temporary numerical decline later in the experiment.
The peer-reviewed study, published online on October 26, 2026 in the Journal of Dairy Research, tested 30 late-lactation Holstein cows. Fifteen received inoculated silage and 15 received untreated control silage.
The full result is more cautious than the best week
Researchers measured methane, milk output and composition, body condition and feed intake. The treatment did not change dry-matter intake, concentrate intake, milk yield or body-condition score.
Methane emissions were numerically 5.8% lower in the inoculated group during weeks four through seven, with the largest weekly difference of 7.7% in week five. But the study found no overall methane reduction across the full experiment.
The peer-reviewed abstract and study record make that distinction explicit. A numerical difference can be commercially interesting, but it is not evidence that the treatment reliably caused a reduction when the statistical test for the overall effect is not significant.
The inoculant contained Lactococcus lactis subsp. lactis SL242 and Lactiplantibacillus plantarum LP58. Estimated intake was about 100 billion colony-forming units per cow each day. The cows consumed an average 12.4 kilograms of dry matter and 4.3 kilograms of concentrate per day.
Why an earlier headline could have sounded stronger
A 2025 conference abstract from the same research programme described an average reduction of roughly 5.7% later in the trial and highlighted the statistically significant 7.74% difference in week five.
The final journal paper frames the evidence more conservatively: no overall reduction, a numerical decline during the later weeks and a call for more research. That is the stronger basis for farm and policy decisions because it considers the entire experimental period rather than selecting the most favorable week.
The difference between those framings does not mean the weekly result was false. It means a transient response did not establish a stable treatment effect. Methane mitigation products must work consistently across time, animals, diets and production stages if farmers are expected to pay for them or governments to count them toward emissions targets.
The inoculant did change the rumen community
Rumen microbiome sequencing showed distinct community clustering between the treatment and control groups in weeks four and seven. Milk fat concentration and milk-solids production also rose temporarily in the treatment group during weeks four and five.
Those changes suggest that the bacteria or their bacteriocins affected fermentation. They do not prove that the altered community will deliver a persistent methane benefit. Microbial shifts can be real while the production or emissions outcome remains small, temporary or variable.
The absence of lower feed intake is useful because a methane decline caused simply by cows eating less would not be an attractive mitigation strategy. Milk yield was also maintained. The unresolved question is whether the biological effect can be made durable enough to matter.
A 30-cow, late-lactation result has limits
The study involved one production stage, one seven-week period and 15 cows per treatment. Late-lactation animals may respond differently from fresh or peak-lactation cows, and a housed feeding system does not represent every pasture-based or mixed ration farm.
The experiment also cannot establish commercial economics. The published abstract does not state the inoculant cost per tonne of silage, storage stability, application losses or the price a farmer would pay per tonne of methane avoided.
A larger trial could detect a modest effect that this experiment lacked the power to confirm. It could also show that the late numerical decline was a chance or time-specific result. Both possibilities remain open.
What needs to happen before adoption
The next test should be preregistered around a full-period methane endpoint, include more animals and repeat the treatment across lactation stages, diets and silage batches. Researchers also need to determine whether changing the dose, strain combination or delivery method makes the response more consistent.
Teagasc's October methane research overview emphasizes measurement and validation across multiple mitigation approaches. That standard is especially important for feed interventions whose effects may shift week by week.
For now, the inoculant is a credible research lead rather than a proven farm methane solution. It altered the rumen community and produced a potentially interesting temporary signal without reducing feed intake or milk yield. The full trial, however, did not demonstrate an overall methane cut.
Reporting method: Agrilinkage reviewed the October 26 peer-reviewed study record, the journal listing, the earlier British Society of Animal Science proceedings and Teagasc's methane research context. No interviews were claimed or conducted. Percentage changes are those reported by the researchers.
Comments