A 1000-cow study across four European countries was undertaken to understand to what extent ruminant microbiomes can be controlled by the host animal and to identify characteristics of the host rumen microbiome axis that determine productivity and methane emissions. A core rumen microbiome, phylogenetically linked and with a preserved hierarchical structure, was identified. A 39-member subset of the core formed hubs in co-occurrence networks linking microbiome structure to host genetics and phenotype (methane emissions, rumen and blood metabolites, and milk production efficiency). These phenotypes can be predicted from the core microbiome using machine learning algorithms. The heritable core microbes, therefore, present primary targets for rumen manipulation toward sustainable and environmentally friendly agriculture.

Nutritional and animal-selection strategies to mitigate enteric methane (CH4) depend on accurate, cost-effective methods to determine emissions from a large number of animals. The objective of the present study was to compare 2 spot-sampling methods to determine CH4 emissions from dairy cows, using gas quantification equipment installed in concentrate feeders or automatic milking stalls. In the first method (sniffer method), CH4 and carbon dioxide (CO2) concentrations were measured in close proximity to the muzzle of the animal, and average CH4 concentrations or CH4/CO2 ratio was calculated. In the second method (flux method), measurement of CH4 and CO2 concentration was combined with an active airflow inside the feed troughs for capture of emitted gas and measurements of CH4 and CO2 fluxes. A muzzle sensor was used allowing data to be filtered when the muzzle was not near the sampling inlet. In a laboratory study, a model cow head was built that emitted CO2 at a constant rate. It was found that CO2 concentrations using the sniffer method decreased up to 39% when the distance of the muzzle from the sampling inlet increased to 30cm, but no muzzle-position effects were observed for the flux method. The methods were compared in 2 on-farm studies conducted using 32 (experiment 1) or 59 (experiment 2) cows in a switch-back design of 5 (experiment 1) or 4 (experiment 2) periods for replicated comparisons between methods. Between-cow coefficient of variation (CV) in CH4 was smaller for the flux than the sniffer method (experiment 1, CV=11.0 vs. 17.5%, and experiment 2, 17.6 vs. 28.0%). Repeatability of the measurements from both methods were high (0.72-0.88), but the relationship between the sniffer and flux methods was weak (R(2)=0.09 in both experiments). With the flux method CH4 was found to be correlated to dry matter intake or body weight, but this was not the case with the sniffer method. The CH4/CO2 ratio was more highly correlated between the flux and sniffer methods (R(2)=0.30), and CV was similar (6.4-8.8%). In experiment 2, cow muzzle position was highly repeatable (0.82) and influenced sniffer and flux method results when not filtered for muzzle position. It was concluded that the flux method provides more reliable estimates of CH4 emissions than the sniffer method. The sniffer method appears to be affected by variable air-mixing conditions created by geometry of feed trough, muzzle movement, and muzzle position.

Concentrations of milk urea N (MUN) are influenced by dietary crude protein concentration and intake and could therefore be used as a biomarker of the efficiency of N utilization for milk production (milk N/N intake; MNE) in lactating cows. In the present investigation, data from milk-production trials (production data set; n=1,804 cow/period observations from 21 change-over studies) and metabolic studies involving measurements of nutrient flow at the omasum in lactating cows (flow data set; n=450 cow/period observations from 29 studies) were used to evaluate the influence of between-cow variation on the relationship of MUN with MNE, urinary N (UN) output, and diet digestibility. All measurements were made on cows fed diets based on grass silage supplemented with a range of protein supplements. Data were analyzed by mixed-model regression analysis with diet within experiment and period within experiment as random effects, allowing the effect of diet and period to be excluded. Between-cow coefficient of variation in MUN concentration and MNE was 0.13 and 0.07 in the production data set and 0.11 and 0.08 in the flow data set, respectively. Based on residual variance, the best model for predicting MNE developed from the production data set was MNE (g/kg)=238 + 7.0 × milk yield (MY; kg/d) - 0.064 × MY(2) - 2.7 × MUN (mg/dL) - 0.10 body weight (kg). For the flow data set, including both MUN and rumen ammonia N concentration with MY in the model accounted for more variation in MNE than when either term was used with MY alone. The best model for predicting UN excretion developed from the production data set (n=443) was UN (g/d)=-29 + 4.3 × dry matter intake (kg/d) + 4.3 × MUN + 0.14 × body weight. Between-cow variation had a smaller influence on the association of MUN with MNE and UN output than published estimates of these relationships based on treatment means, in which differences in MUN generally arise from variation in dietary crude protein concentration. For the flow data set, between-cow variation in MUN and rumen ammonia N concentrations was positively associated with total-tract organic matter digestibility. In conclusion, evaluation of phenotypic variation in MUN indicated that between-cow variation in MUN had a smaller effect on MNE compared with published responses of MUN to dietary crude protein concentration, suggesting that a closer control over diet composition relative to requirements has greater potential to improve MNE and lower UN on farm than genetic selection.

Ethanol and acetic acid are common end products from silages. The main objective of this study was to determine whether high concentrations of ethanol or acetic acid in total mixed ration would affect performance in dairy cows. Thirty mid-lactation Holstein cows were grouped in 10 blocks and fed one of the following diets for 7 wk: (1) control (33% Bermuda hay + 67% concentrates), (2) ethanol [control diet + 5% ethanol, dry matter (DM) basis], or (3) acetic acid (control diet + 5% acetic acid, DM basis). Ethanol and acetic acid were diluted in water (1:2) and sprayed onto total mixed rations twice daily before feeding. An equal amount of water was mixed with the control ration. To adapt animals to these treatments, cows were fed only half of the treatment dose during the first week of study. Cows fed ethanol yielded more milk (37.9 kg/d) than those fed the control (35.8 kg/d) or acetic acid (35.3 kg/d) diets, mainly due to the higher DM intake (DMI; 23.7, 22.2, and 21.6 kg/d, respectively). The significant diet × week interaction for DMI, mainly during wk 2 and 3 (when acetic acid reached the full dose), was related to the decrease in DMI observed for the acetic acid treatment. There was a diet × week interaction in excretion of milk energy per DMI during wk 2 and 3, due to cows fed acetic acid sustained milk yield despite lower DMI. Energy efficiency was similar across diets. Blood metabolites (glucose, insulin, nonesterified fatty acids, ethanol, and γ-glutamyl transferase activity) and sensory characteristics of milk were not affected by these treatments. Animal performance suggested similar energy value for the diet containing ethanol compared with other diets. Rumen conversion of ethanol to acetate and a concomitant increase in methane production might be a plausible explanation for the deviation of the predicted energy value based on the heat of combustion. Therefore, the loss of volatile compounds during the drying process in the laboratory should be considered when calculating energy content of fermented feedstuffs.

A meta-analysis based on an individual-cow data set was conducted to investigate the effects of between-cow variation and related animal variables on predicted CH emissions from dairy cows. Data were taken from 40 change-over studies consisting of a total of 637 cow/period observations. Animal production and rumen fermentation characteristics were measured for 154 diets in 40 studies; diet digestibility was measured for 135 diets in 34 studies, and ruminal digestion kinetics was measured for 56 diets in 15 studies. The experimental diets were based on grass silage, with cereal grains or by-products as energy supplements, and soybean or canola meal as protein supplements. Average forage:concentrate ratio across all diets on a dry matter basis was 59:41. Methane production was predicted from apparently fermented substrate using stoichiometric principles. Data were analyzed by mixed-model regression using diet and period within experiment as random effects, thereby allowing the effect of experiment, diet, and period to be excluded. Dry matter intake and milk yield were more repeatable experimental measures than rumen fermentation, nutrient outflow, diet digestibility, or estimated CH yield. Between-cow coefficient of variation (CV) was 0.010 for stoichiometric CH per mol of volatile fatty acids and 0.067 for predicted CH yield (CH/dry matter intake). Organic matter digestibility (OMD) also displayed little between-cow variation (CV = 0.013), indicating that between-cow variation in diet digestibility and rumen fermentation pattern do not markedly contribute to between cow-variation in CH yield. Digesta passage rate was much more variable (CV = 0.08) between cows than OMD or rumen fermentation pattern. Increased digesta passage rate is associated with improved energetic efficiency of microbial N synthesis, which partitions fermented substrate from volatile fatty acids and gases to microbial cells that are more reduced than fermented carbohydrates. Positive relationships were observed between CH per mol of volatile fatty acids versus OMD and rumen ammonia N concentration versus OMD; and negative relationships between the efficiency of microbial N synthesis versus OMD and digesta passage rate versus OMD, suggesting that the effects of these variables on CH yield were additive. It can be concluded that variations in OMD and efficiency in microbial N synthesis resulting from variations in digesta passage contribute more to between-animal variation in CH emissions than rumen fermentation pattern.