This study was conducted at a commercial dairy farm (Brades Farm, Lancaster, United Kingdom) with 396 dairy cows. The lactating Jersey (n = 121) and Holstein-Friesian (HF) cows (n = 275) were housed in separate barns with sandbedded freestalls. Animals were handled in accordance with the Scotland’s Rurual College’s (SRUC) ethical committee guidance on animal use for research.

A basal diet that consisted of a TMR (55% grass silage, 20% caustic wheat, 6% whey permeate; DM basis), supplemented with 19% DM of a milker’s supplement (59.5% rapeseed meal, 14% protected rape meal, 12% maize distillers, 10% soya, 4% palm kernel expeller, 0.5% magnesium oxide; DM basis) was prepared daily in a mixing wagon (Powermix Pro; Shelbourne Reynolds, Bury St Edmunds, United Kingdom) and placed in feeding troughs in both barns. Prior to mixing, feeds were stored separately in silos and both herds received the same basal diet.

Both herds received Mootral (Mootral SA, Rolle, Switzerland) incorporated in pellets, formulated by a feed mill (Dugdale Nutrition, Clitheroe, United Kingdom) containing 3% Mootral, based on previous internal research, along with other feed (58% rape meal, 14% protected rape meal, 12% maize distillers, 9% Hipro soya, 3.5% soya hulls, and 0.5% magnesium oxide; DM basis). The pellets were delivered in 25 kg bags and stored in a dry, sheltered area in the farm. The pellets were supplemented at 500 g/day/cow in the TMR during the 12 weeks of Mootral supplementation (Table 1).

All cows on the farm were milked twice a day starting at 06:15 hours and again at 18:00 hours, in a swing-over milking parlour (GEA-Westfalia, Düsseldorf, Germany). Weekly milk yield was recorded with the milk management system (GEA-Westfalia, Düsseldorf, Germany). Milk from the Jersey and Holstein-Friesian herds were stored in separate cooled bulk tanks. Milk was sampled from separate bulk tanks for each herd every week in duplicate, for milk quality analysis. Samples were sent on the same day of milking to the National Milk Laboratories (NML; Wolverhampton, United Kingdom) for analysis of milk fat, protein, lactose, urea, pH and bulk tank somatic cell counts (BTSCC), using Fourier transform infrared spectroscopy (FTIR) (Foss, Hilleroed, Denmark). Redundancy testing for the same milk attributes was made with duplicate milk samples tested at Landeskontrollverband Weser-Ems (Leer, Germany) and the results from both laboratories were evaluated. As both laboratories showed comparable results, only the results from NML have been reported here. Milk yield and results of the milk analyses during Mootral supplementation were compared with the previous farm records.

Feed samples of the total mixed ration were collected in each period and sent for nutritional analysis (Table 4) using near infrared spectroscopy (Eurofins Agro, Wageningen, Netherlands). Dry matter intake (DMI) for the TMR fed to both herds was determined weekly from feed samples using a moisture tester (Koster, Ohio, USA) and the amount of refused feed was subtracted from the total initially provided.

Enteric methane emission from a sub-set of each herd were measured using a handheld LMD (Laser Methane Mini; Tokyo Gas Engineering, Tokyo, Japan). Measurements were taken individually from 15 Jersey and 15 Holstein-Friesian cows as they returned from the milking parlour in the morning from approximately 06:15 hours. Measurements procedure was similar to the work of Sorg et al. [13] . The cows were held in either a crush or artificial insemination stall and the handheld LMD was pointed at the cow’s nostrils for 4 minutes at a distance of 1 m. LMD measurements were taken on Week 12 of the Mootral period with baseline measurements taken on week 16 (Table 1) from the same cows as identified by their unique ear tag numbers. Means of the LMD readings from each cow from these different periods were used for comparisons.

A qualified veterinarian observed the herds’ health bi-weekly, taking note of any problems related with fertility, feet problems, mastitis, and metabolic disorders.

Feed efficiency (FE; L/kg) was the ratio of milk yield (L/day) to DMI (kg/day).

Weekly means of all variables of interest were analyzed using one-way ANOVA with post-hoc test on GraphPad Prism version 6.04 (GraphPad Software, La Jolla California, USA). The fixed effect was the period, while random effects were the LMD readings, DMI, milk yield; and milk fat, protein, lactose, urea, pH and BTSCC. Analyses were considered significant at p ≤ 0.05.

Methane concentrations for both Jersey and HF cows measured with the LMD showed a significant decrease from the baseline values (p < 0.05) by 38.3% for the Jersey cows and 20.7% for the HF cows (Figure 1) with the use of the Mootral supplement. Data from 1 cow in the HF herd was omitted from analysis due to missing measurements in week 12.

Methane gas readings from the LMD in this study was low compared to other reported LMD measurements which ranged from 100 to 400 ppm-m [8] [13] [14] . Variations of methane emissions were influenced by the cow’s activity, as demonstrated in an earlier work [14] , and would explain the low methane

measurements. As the LMD measurements were taken in the early morning before the cows had been fed, and with low rumen fill, fermentation would not have been at its peak. Measurements made in enclosed areas later in the day after rumen fill, as with respiration chambers, tended to record higher methane concentrations, as seen in LMD readings reported in an earlier work [8] . However, the LMD measurements were taken in a sheltered area with no noticeable effect of wind movements. When compared with other methods of methane measurements, the LMD still ranked intensity of methane similarly [13] .

The results of enteric methane reduction previously reported for garlic additions in diets for ruminants have ranged from 6% to 18% [15] [16] [17] , while that for citrus pulp had no effect on methane mitigation [18] . The bitter orange extract in Mootral would naturally contain flavonoids and as Graz and Miller [11] had shown, flavonoids alone did very little in reducing methane gas during rumen fermentation. However, when allicin from garlic was present with the bitter orange extract, there was notable methane reduction. This indicates that the Mootral composition could have synergistic effects in reducing enteric methane production.

The milk yield for the Jersey herd increased significantly (p < 0.05) by 5% during the Mootral period compared to before Mootral supplementation. The increased milk yield during Mootral supplementation was also recorded for the HF herd, which had a significant (p < 0.05) increase of 7.8% compared to before Mootral supplementation. Both herds’ milk yield remained at the heightened level for 4 weeks after Mootral supplementation stopped (Table 2).

Milk constituents were monitored at pre-Mootral and during Mootral supplementation (Table 3) with most of the milk constituents (milk fat, protein,

Superscripts within the same row which differ denote significant differences (p < 0.05).

¹BTSCC = bulk tank somatic cell counts in milk, ×1000 cells/ml. Asterisks (*) denote statistically significant differences, p < 0.01.

and pH) analyzed in both herds showed no significant differences between periods, except for urea and BTSCC. This was unsurprising, as previous studies had not seen any negative effects of garlic on rumen fermentation [19] [20] . Although the urea content of the milk in both herds increased significantly during Mootral supplementation by 8.5 to 10.5 mg/dL (p < 0.01), the values were within the expected range (15 to 30 mg/dL) for cows in mid-lactation [21] .

Only the BTSCC for the HF herd decreased significantly by 32.6% (p < 0.01) during the Mootral period. SCC reductions have been previously reported for a similar garlic-based supplement [22] [23] . In addition, numerous previous studies have suggested garlic and its sulphur constituents had an immune-modulating effect [24] [25] [26] , which could have attributed to the lowering of SCC seen in this study. BTSCC is a quick gauge of the herd’s health, whereby high numbers indicate an immune response to some likelihood of infection in the herd [27] [28] . The BTSCC for both herds are much lower than the European Commission’s limit of ≤400,000 cells/ml [29] and milk processors often provide incentives to keep the BTSCC lower (≤250,000 cells/ml). If the effects of lower BTSCC from the Mootral supplement could be repeated in other dairy herds, including those with high SCC, it may prove economically attractive for farmers.

This simple metric was used to gauge the herd’s efficiency in converting feed into milk, which would eventually impact the farmer’s net financial margin. The Jersey cows in this study showed a significant increase in FE during Mootral supplementation by 13% and increase of 24% after Mootral was withdrawn (Figure 2(a)), compared to the pre-Mootral period (p < 0.05). FE for the HF herd remained the same during Mootral supplementation compared to the pre-Mootral period, however increased by 8% (p < 0.05) during the post-Mootral period (Figure 2(b)). On average, the Jersey cows had a higher FE than the HF cows, which had been reported in a previous study [30] . Although the ruminal microbiome is responsible for the degradation of feed matter to provide energy for the cows, their population are not thought to be affected by the animals’ genetics [31] . Though there was a difference in the FE trends observed during this study, it is not immediately understood and would warrant future investigation.

The significant reduction of enteric methane emission in the Jersey herd with Mootral supplementation might have contributed to the increase in their FE by 10% during the post-Mootral period. In the HF herd, the significant decrease of enteric methane emission did not have such a pronounced effect on their FE during Mootral supplementation, but the increase of their FE by 6% in the following 4 weeks after withdrawal of supplementation could be attributed to a persistent effect from the supplementation. Eger et al. [12] had shown that in the

¹ME = Metabolized energy. ²NEl = Net energy of lactation.

presence of Mootral, the percentage of methanogens reduced while bearing no negative effects on rumen fermentation. Our study here confirms these earlier findings especially with the lactating Jersey cows and to some extent in lactating HF cows with an improved FE in the longer term.

Results of feed analysis showed some changes in DM, crude ash, crude protein, NDF and ADF in the feed analysis but the changes were not statistically significant. The ME and NEL content were more consistent throughout the periods (Table 4).

There were no reports of adverse health events in both the Jersey and HF herds, and none of the cows had fertility, feet, mastitis, or metabolic issues during the study.