The central bottleneck the study addresses is well known in the dairy industry. Direct measurement of individual feed intake requires either tie-stall housing where each cow’s ration is weighed by hand, or automated electronic feeding systems that register every visit an animal makes to a feed bin. Both approaches are costly and labor-intensive, which is why they are confined to experimental facilities. On commercial farms, intake has typically been estimated with standardized prediction equations that rely on inputs such as body weight and milk composition. These equations have a well-documented weakness: they often lose accuracy at the extremes of production, particularly overestimating intake in high-producing cows, and they ignore the rich stream of behavioral data that modern wearable sensors now collect continuously.

The sensor at the heart of the new pipeline is a commercial neck collar containing a three-dimensional accelerometer that monitors head movements and activity around the clock. From these signals, the system classifies and quantifies the time each cow spends eating, ruminating, and panting. A machine learning model combines these behavioral outputs with cow-status features — parity, days in milk, and pregnancy status — and with milk production data to estimate daily intake. The prediction methodology is protected under granted patents in Europe and the United States and is already embedded in the company’s commercial software, making this one of the first instances of a sensor-based intake prediction tool validated across multiple independent farms.

To train and test the model, the researchers assembled an unusually broad dataset: 16,888 weekly observations from 320 lactating cows across five international research farms — Michigan State University, Università Cattolica del Sacro Cuore in Italy, the University of British Columbia, Penn State University, and the Institute of Animal Science in Prague. Reference intake was measured manually in tie-stalls at Michigan State or automatically with electronic feeding systems elsewhere. The raw dataset of 21,550 daily records was rigorously filtered: cows receiving more than 1.5 kilograms of supplemental concentrate beyond a total mixed ration were excluded, days with partial collar data (less than 23 hours of coverage) were discarded, cows eating less than one hour per day were removed as likely artifacts, and only records beyond nine days in milk were retained to avoid the metabolic volatility of early lactation.

The validation design was deliberately stringent. Using leave-farm-out cross-validation, the model was trained on data from four farms and tested on the fifth, held-out farm, and this was repeated so that every farm served once as the independent test set. The result was strong agreement between predicted and measured intake, with a coefficient of determination of 0.77 and a root mean square error of 1.9 kilograms per day. Crucially, the decomposition of prediction error showed that almost all of it — 99.97 percent — was random rather than systematic: mean bias was essentially zero and slope bias contributed only 0.03 percent. The calibration regression of measured on predicted intake produced a slope of 0.992, marginally but not meaningfully different from unity, indicating only a faint tendency toward regression to the mean across the observed intake range.

Performance was also remarkably uniform in relative terms across herds that differed in country, housing, diet, and management. The coefficient of variation of the error ranged from 6.9 to 8.2 percent across farms, with a pooled value of 7.3 percent. Farm-level r² values spanned a wider range, from 0.45 in Prague to 0.78 at Michigan State, but the authors show this largely reflects differences in how much cows within each herd vary from one another rather than differences in prediction error. In Prague, where the herd had the narrowest intake distribution in the study, the model achieved its lowest absolute error yet the lowest r², because there was simply less variation to explain. In the final production model trained on all five farms, the three input categories contributed comparably: collar-derived behavioral features accounted for 36.9 percent of total feature importance, milk production data 34.0 percent, and cow-status features 29.1 percent.

With the pipeline validated, the team turned to the question that gives the study its broader significance: does the collar-predicted intake capture biologically meaningful variation, or is it merely a statistical artifact? To find out, they applied the model to data from 2,096 Holstein cows on five commercial Israeli farms, entirely independent of the research farms used for validation, and examined whether predicted intake and the feed efficiency derived from it resemble each other across generations. Feed efficiency was defined as energy-corrected milk divided by predicted intake, with energy-corrected milk calculated continuously from inline milk analyzers measuring fat, protein, and lactose. The analysis exploited 665 dam–daughter pairs for predicted intake and 617 pairs for feed efficiency.

The results were consistent across multiple lines of evidence. For predicted intake, the dam–daughter association remained significant at every step of a sequential adjustment that progressively accounted for farm, temporal herd effects, and daughter parity — a critical control, since parity exerts an enormous influence on intake, with multiparous cows predicted to consume roughly four kilograms more per day than first-lactation heifers. The fully adjusted analysis yielded an indicative heritability of approximately 0.15, toward the lower end of published estimates for measured intake in dairy cattle. For the derived feed efficiency phenotype, the dam–daughter association was even more robust, surviving all adjustments with an indicative heritability of about 0.21, squarely within the published range for related efficiency traits. Complementing this, a highly significant sire effect emerged across 44 bulls with 853 daughters, and it persisted in a subanalysis restricted to sires with daughters on at least two farms, ruling out the possibility that the pattern was an artifact of farm composition.

The authors are careful to frame these genetic signals as preliminary. Because the analysis rests on phenotypic parent–offspring resemblance rather than pedigree- or genomic-based animal models, it does not constitute formal heritability estimation, and precise genetic parameters would require such methods applied to herds with both sensor deployment and genotyped animals. They also note an important caveat about ratio traits: feed efficiency combines collar-predicted intake with independently measured milk output, so part of its familial signal originates in the milk component rather than in the collar-predicted intake itself, and direct selection on a ratio distributes selection pressure across its component traits in ways that are difficult to predict. Predicted intake, being a linear trait, is not subject to this complication.

Even before any formal genetic evaluation, the technology has immediate practical value. Because predicted intake can be estimated at the individual level from data already collected by modern dairy management systems — collar behavior, milk records, and basic cow descriptors — it enables cow-level decisions such as identifying inefficient animals, informing culling strategy, and calculating income over feed, a direct measure of each animal’s profitability. The cohort-level mean intake needed for the model’s normalization step can be obtained from routine group feed records, so no individual intake infrastructure is required. Looking ahead, the researchers identify clear priorities: validating the approach in grazing systems, which were not represented in the training data; extending it beyond Holsteins to other breeds and crossbreds; incorporating finer-grained inputs such as milk fatty-acid profiles, body weight, and health events; and ultimately linking the predicted phenotype to enteric methane estimation, connecting scalable efficiency phenotyping to the sustainability goals that motivate much of this research. If those steps succeed, the humble collar hanging around a cow’s neck could become the gateway to breeding dairy cattle that produce more milk from less feed — a win for farmers, and for the planet.

Subject of Research: Collar-based prediction of individual dry matter intake and feed efficiency in dairy cows

Article Title: Collar-based prediction of individual dry matter intake in dairy cows: validation across international research farms and evidence of familial variation in commercial herds

Article References: Yair, R., Arazi, A., Shabtai, V., & Katz, G. (2026). Collar-based prediction of individual dry matter intake in dairy cows: validation across international research farms and evidence of familial variation in commercial herds. Frontiers in Animal Science, 7, Article 1897979. https://doi.org/10.3389/fanim.2026.1897979

Image Credits: AI Generated

DOI: 10.3389/fanim.2026.1897979

Keywords: dry matter intake, feed efficiency, dairy cattle, wearable sensors, precision livestock farming, machine learning, heritability, dam-daughter regression, energy-corrected milk, Holstein, enteric methane, breeding

The seaweed in question, Asparagopsis taxiformis and its sister species Asparagopsis armata, first stunned researchers in 2014, when scientists at James Cook University screened twenty tropical macroalgae and found that A. taxiformis suppressed methane production by 98.9 percent in laboratory fermentations. The mechanism is elegant: as a chemical defense, the alga produces halogenated compounds, dominated by bromoform, that interfere with the enzymes methanogenic archaea use to combine hydrogen and carbon in the rumen. In live cattle, subsequent studies confirmed reductions of up to 99 percent, making Asparagopsis the most potent natural antimethanogenic feed additive ever identified.

The stakes could hardly be higher. Enteric fermentation is the largest anthropogenic source of methane in the United States, accounting for 27.4 percent of the nation’s methane emissions in 2022, according to the U.S. Environmental Protection Agency, a figure that has climbed 5.2 percent since 1990 alongside growing cattle populations. Although methane survives only about a decade in the atmosphere, its warming potential is 28 times that of carbon dioxide over a century and 84 times greater over twenty years, which makes it a prime target for near-term climate action. Methane also feeds the formation of ground-level ozone, a pollutant that damages respiratory and cardiovascular health, so cutting ruminant emissions pays dividends for air quality as well as for the climate.

Yet the field has had a modeling problem. Earlier prediction models largely ignored the bromoform concentration itself, relying instead on the weight of whole seaweed fed to the animals, even though bromoform content can vary dramatically between batches. A recent meta-analysis led by Kebreab and colleagues did incorporate bromoform, but it worked from published treatment averages rather than raw animal-level data, and it extrapolated to doses as high as roughly 70 milligrams per kilogram of dry matter intake, well beyond what recent trials have actually used. Typical application rates have averaged around 37 milligrams per kilogram for beef cattle and about 16 for dairy cows, with reported maxima of 51 and 21 respectively, so models recommending far higher doses risked drifting away from both veterinary guidance and real-world practice.

The new study, led by Claire Goloja of Symbrosia Inc. with collaborators including Breanna M. Roque, set out to close those gaps. Following PRISMA guidelines, the team searched PubMed, Scopus, Oxford Academic and Google Scholar for peer-reviewed, randomized, in vivo bovine trials of Asparagopsis that documented bromoform concentration, dosing, dry matter intake, feed composition and methane measurement technology. From an initial pool of 989 records, twelve studies qualified: five on beef cattle and seven on dairy cows, spanning trial durations of 14 to 200 days, wild-harvested and commercially produced seaweed, freeze-dried and oil-based formulations, and bromoform doses ranging from 5.2 to 72.6 milligrams per kilogram of dry matter intake.

A crucial methodological move was securing raw data. The authors contacted the researchers behind all twelve studies and obtained individual animal-level records from seven of them, combining those with summary statistics from the remaining five. This hybrid dataset allowed the team to capture variation between individual animals, not just between studies, a level of granularity that earlier treatment-mean analyses could not achieve. All studies reported crude protein and neutral detergent fiber values for the diets, giving the analysts dietary covariates to test alongside the bromoform dose.

The headline discovery is a threshold. Because a simple straight line could not describe the biological response across the full dose range, the team fitted a segmented regression with a breakpoint, and the data revealed a clear point of diminishing returns at 49.81 milligrams of bromoform per kilogram of dry matter intake. Below that dose, methane yield falls steeply and predictably as bromoform rises; above it, reductions plateau while the risk of overestimating benefits grows. Doses at or above the cutoff, including individual data points at 51 and above 70 milligrams per kilogram, were excluded from the final models to keep the predictions anchored in the linear, well-populated region of the dose-response curve.

Stratifying by animal type proved essential, since beef and dairy cattle responded differently, a statistically significant difference with a p-value of 0.022. The resulting eight linear mixed-effects models, which account for clustering within studies through random intercepts, deliver concrete numbers. In beef cattle, an average bromoform dose of 14.95 milligrams per kilogram of dry matter intake cuts methane yield by 34.8 percent, while the maximum modeled dose of 35.70 milligrams achieves an 83.6 percent reduction. Dairy cows are more conservative responders: an average dose of 11.25 milligrams yields a 16.9 percent reduction, and the maximum of 27.40 milligrams produces a 43.1 percent cut. Adding crude protein and neutral detergent fiber to the models marginally improved performance, especially for beef cattle, with crude protein also helping dairy predictions, meaning farmers could eventually estimate methane savings from bromoform dose and routine feed analysis alone, without running costly live trials.

The analysis also exposed an uncomfortable truth about how methane is measured. Detection technology itself significantly shaped the apparent reductions, with a p-value below 0.001. Respiration chambers, which cost over one million dollars each and house a single animal at a time, generated higher reduction estimates than the sulfur hexafluoride tracer method or the GreenFeed system, which serves roughly fifty animals per unit at a price above one hundred thousand dollars. The authors point to sensor differences as a likely culprit: older GreenFeed units use nondispersive infrared sensors with higher detection limits than the tunable diode lasers in newer models, and may simply fail to register emissions that have fallen close to baseline. When a feed additive pushes methane down by 80 or 90 percent, the precision of the instrument becomes the limiting factor in knowing what is really happening inside the cow.

The study’s authors are candid about its limits. Dairy models fit the data less tightly than beef models, raw data from five studies remained out of reach, and dietary fat, a known methane suppressant in its own right, could not be included because of sparse concentration data. They recommend that future research prioritize dairy systems, larger sample sizes and the role of fat as a predictor variable. Still, the practical upshot is transformative: for the first time, producers and regulators have a validated, dose-based framework that identifies the bromoform level where mitigation peaks, flags the point where more seaweed adds nothing, and accounts for the diet and the animal in between. As measurement technology matures toward the accuracy and affordability that large-scale verification will demand, these models offer an interim bridge, turning one of climate science’s most celebrated biological discoveries into something a feed manager can actually dose, predict and trust.

Subject of Research: Bromoform dose-response modeling of methane mitigation from Asparagopsis seaweed feed additives in cattle

Article Title: A meta-analysis establishes bromoform dose and forage-based models to evaluate the antimethanogenic effects of Asparagopsis spp. feed additive

Article References: A meta-analysis establishes bromoform dose and forage-based models to evaluate the antimethanogenic effects of Asparagopsis spp. feed additive. (n.d.). https://doi.org/10.1186/s44399-025-00030-w

Image Credits: AI Generated

DOI: 10.1186/s44399-025-00030-w

Keywords: Asparagopsis, bromoform, enteric methane, feed additive, meta-analysis, cattle, dairy cows, beef cattle, methane mitigation, rumen fermentation, greenhouse gases, climate change

The research team, led by Setiasih Setiasih and colleagues at Indonesia’s National Research and Innovation Agency (BRIN) and partner institutions, published their findings in Results in Chemistry. Their strategy centered on a single deceptively simple question: what exactly is inside rose myrtle leaves, and how does the answer change depending on how you extract it? To find out, they harvested leaves from an experimental garden in Riau Province, dried them at 40 degrees Celsius for three days, and ground them into a flour. That flour was then subjected to maceration with three solvents spanning the polarity spectrum: 96 percent ethanol as a polar solvent, ethyl acetate as a semipolar solvent, and n-hexane as a nonpolar solvent. Each extract was subsequently analyzed by liquid chromatography coupled to high-resolution mass spectrometry, or LC-HRMS, at BRIN’s Advanced Characterization Laboratories.

The choice of solvent turned out to matter enormously. LC-HRMS profiling tentatively identified 191 distinct compounds in the ethanol extract, 226 in the ethyl acetate extract, and 206 in the n-hexane extract. Ethanol preferentially pulled out polar molecules, yielding a profile dominated by flavonoid glycosides, phenolic acids and related derivatives, including myricitrin, myricetin, quinic acid derivatives and multiple dihydroxybenzoic acids. These hydroxyl-rich compounds dissolve readily in polar solvents through hydrogen bonding and are classic free-radical scavengers, which helps explain the high antioxidant activity long associated with the plant. The ethyl acetate extract displayed the greatest metabolite diversity of all, dominated by pentacyclic triterpenoids such as ursolic acid, asiatic acid, oleanolic acid, arjungenin and glycyrrhetinic acid derivatives, alongside phenolics including ellagic acid, gallic acid, myricetin and taxifolin. The n-hexane extract, by contrast, was rich in lipophilic sesquiterpenes such as caryophyllene oxide, alpha-farnesene and curcumene, together with assorted triterpenoid derivatives.

The analytical power of the method lies in its ability to go beyond a simple mass readout. Using higher-energy collisional dissociation fragmentation, the researchers could watch molecules break apart in characteristic ways, dramatically raising confidence in their identifications. Myricitrin, for example, produced a precursor ion at mass-to-charge ratio 463.0881 that fragmented to give a product ion at 316.0224, corresponding to the loss of its rhamnose sugar and formation of the myricetin aglycone. Oleanolic acid showed a precursor at 455.3534 with a diagnostic fragment at 411.3472 reflecting loss of carbon dioxide, while caryophyllene oxide in the n-hexane extract yielded sequential ring-cleavage fragments at 161.0962 and 109.1014. The authors are careful to stress that this was qualitative screening rather than quantitative metabolomics, conducted without biological replicates, so peak areas reflect relative signal intensity within a run rather than absolute concentrations. Even so, the picture that emerges is of a plant whose chemistry shifts meaningfully with every solvent a chemist chooses.

With the chemical inventory in hand, the team turned to biological testing. Using the agar disc diffusion method, they challenged four bacterial strains with each extract at concentrations of 25, 50 and 80 percent: Escherichia coli ATCC 25922 and Salmonella typhimurium ATCC 14028 as Gram-negative representatives, and Staphylococcus aureus ATCC 25923 and Listeria monocytogenes ATCC 19115 as Gram-positive representatives. The result was a clean split. Every extract, regardless of solvent, inhibited both Gram-positive species and failed entirely to inhibit either Gram-negative one. Chloramphenicol, the reference antibiotic, produced a 14.82 millimeter zone against S. aureus, larger than the 7.11 to 7.70 millimeter zones produced by the extracts, but the n-hexane extract at 80 percent actually outperformed chloramphenicol against Listeria, generating a 7.17 millimeter zone compared with the antibiotic’s 5.23 millimeters.

The selectivity is rooted in bacterial architecture. Gram-negative bacteria wrap themselves in an outer membrane studded with lipopolysaccharides, an exceptionally effective permeability barrier that excludes hydrophobic molecules and is reinforced by selective porins and efflux pumps that expel antimicrobials before they reach lethal intracellular concentrations. Gram-positive bacteria, lacking that outer membrane, rely on a thick but porous peptidoglycan layer some 20 to 80 nanometers deep, through which small plant metabolites can diffuse relatively freely toward the cytoplasmic membrane. The researchers also observed that solvent type shaped potency: ethanol extracts, rich in phenolics and flavonoids, were most effective against S. aureus, while n-hexane extracts, loaded with membrane-disrupting terpenoids, were the champions against Listeria, a reminder that even within the Gram-positive category, membrane composition and defensive machinery differ enough to matter. The findings align with earlier reports that rose myrtle extracts and the pure compound rhodomyrtone, an antibiotic candidate with activity comparable to vancomycin, are most effective against Gram-positive organisms.

The second half of the study addresses climate. Ruminant livestock account for roughly a quarter of anthropogenic greenhouse gas emissions measured by methane’s contribution, with enteric fermentation in the digestive tract responsible for 80 to 89.5 percent of that methane and manure contributing the rest. Methanogenic archaea in the rumen generate the gas, and feed additives that suppress them are among the most actively pursued mitigation strategies. Because flavonoids, tannins and saponins have documented anti-methanogenic effects, the team tested the ethyl acetate extract, the richest in bioactive triterpenoids and phenolics, in an in vitro rumen fermentation system using Brachiaria ruziziensis grass as the basal feed, at inclusion levels of 1 to 4 percent, with methane measured by gas chromatography over 48 hours.

The outcome was striking in one particular way. Total gas production was unaffected across all treatment levels, meaning fermentation efficiency was preserved, but methane output fell steadily as the extract dose increased. The methane percentage of total gas dropped from 7.72 percent in the untreated control to 3.77 percent at the 4 percent inclusion level, a statistically significant reduction, and methane volume declined as well with a P value of 0.056, right at the threshold of significance. The chemical fingerprint offers a plausible mechanism in several complementary parts. Triterpenoids such as ursolic acid are membrane-active against rumen protozoa, which maintain physical symbioses with methanogenic archaea and supply them with hydrogen; suppressing protozoa often drags methane down with it. Phenolic compounds can selectively inhibit methanogens while redirecting metabolic hydrogen toward propionate synthesis, an energetically favorable route for the animal. Terpenoids such as caryophyllene oxide may interfere with membrane integrity and ion gradients in anaerobic microbes. The authors also cite prior molecular docking work showing that rose myrtle compounds bind strongly to methyl coenzyme M reductase, the enzyme at the heart of methanogenesis, with rhodomyrtial A achieving a predicted binding affinity of minus 10.4 kilocalories per mole.

The researchers are appropriately measured about what these results do and do not prove. The antibacterial assay was a preliminary screen, and they recommend follow-up with minimum inhibitory and bactericidal concentration determinations, plus in vivo testing, to establish therapeutic potential; earlier work suggests possible veterinary applications including mastitis treatment in dairy cows. Similarly, they emphasize that the methane findings provide a chemical rationale rather than a demonstrated field benefit, and that in vivo feeding trials will be required to confirm efficacy and safety. Nevertheless, the study is the first LC-HRMS-based phytochemical characterization of rose myrtle in the context of ruminant nutrition, and it reveals an intertwined network of triterpenoids, phenolics, flavonoids and lipophilic terpenoids whose combined effects may prove more stable than any single isolated compound. In an era when agriculture is under intense pressure to cut emissions without compromising animal productivity, a shrub that grows wild across tropical Asia, fights Gram-positive pathogens in the laboratory, and measurably dampens methane in simulated rumen fluid is exactly the kind of natural lead worth chasing.

Subject of Research: Phytochemical characterization of Rhodomyrtus tomentosa leaves and their antibacterial and methane-mitigating bioactivities

Article Title: LC–HRMS-based phytochemical characterization and biological activities of Rhodomyrtus tomentosa leaves extracts: Antibacterial and methane mitigation potentials

Article References: Setiasih, S., Anggraeny, Y. N., Puspito, S., Widodo, S., Wardi, W., Prihandini, P. W., Antonius, A., Shiddieqy, M. I., Sasongko, W. T., Istiqomah, N., Haryanto, B., Kurniawati, A., Prihartini, I., Wulansari, W. I., & Ernawati, A. (2026). LC–HRMS-based phytochemical characterization and biological activities of Rhodomyrtus tomentosa leaves extracts: Antibacterial and methane mitigation potentials. Results in Chemistry, 30, Article 103812. https://doi.org/10.1016/j.rechem.2026.103812

Image Credits: AI Generated

DOI: 10.1016/j.rechem.2026.103812

Keywords: Rhodomyrtus tomentosa, rose myrtle, LC-HRMS, phytochemical profiling, antibacterial activity, triterpenoids, flavonoids, enteric methane, ruminant nutrition, feed additives, methanogens, Results in Chemistry