The Impact of Essential Oil Feed Supplementation on Enteric Gas Emissions and Production Parameters from Dairy Cattle

: Societal pressure to reduce enteric methane emissions from cattle continues to increase. The present study evaluated the efficacy of the commercial essential oil feed additive Agolin ® Ruminant on reducing enteric gas emissions and improving milk parameters in dairy cattle. Twenty mid-lactation Holstein cows, blocked by parity and days in milk, were randomly assigned to a top dress treatment with Agolin or an un-supplemented control for a 56-day trial. Cows were group housed and individually fed twice daily. Enteric gas emissions, including methane, carbon dioxide, ammonia, and nitrous oxide, were sampled every 14 days for a 12 h period via head chambers connected to a mobile air quality laboratory. Cows supplemented with Agolin versus the control had less methane intensity (g / period / kg energy-corrected milk (ECM); p = 0.025). Ammonia was the most affected gas, with lower ammonia production (mg / period; p = 0.028), and ammonia intensity (mg / period / kg ECM; p = 0.011) in Agolin-fed versus control-fed cows. All cow performance variables, including dry matter intake, ECM, milk fat, milk protein, or feed efficiency were similar between treatments. Further research should evaluate how Agolin impacts ruminal flora, focusing on mechanistic impacts to fermentation.


Introduction
Air pollutants have strong effects on public and environmental health. Carbon dioxide (CO 2 ), methane (CH 4 ), and nitrous oxide (N 2 O) are greenhouse gases (GHGs) that have received attention due to their contribution to climate change [1]. These GHGs, as well as ammonia (NH 3 ), can be emitted by animal agricultural processes. In an effort to regulate climate change, California passed Senate Bill 1383, mandating a 40% reduction in CH 4 emissions below 2013 levels by the year 2030 [2]. The primary focus of this bill lies on CH 4 emissions from all sources, but within the dairy sector, both CH 4 manure mitigation and enteric fermentation CH 4 can be considered. To date, the majority of research and improvements in CH 4 emissions have been in the area of manure management. With enteric fermentation accounting for 28% of total CH 4 emissions in the United States [1], reducing emissions from this source of CH 4 is important.
The diet comprised an 89-90% dry matter (DM) total mixed ration (TMR ; Table 1), with cows being fed to yield 5% refusals. Feed refusals were collected and sampled prior to each feeding at approximately 6:15 a.m. and 6:15 p.m. to determine DM content and daily DM intake (DMI). Cows were adapted to the basal diet without supplementation for 30 days prior to the start of emission sampling. Table 1. Composition (%) of the basal total mixed ration (TMR) fed to cows during the 56-day trial period, as fed (89-90% dry matter (DM)).

TMR Composition (%; As Fed)
Grain Mix 1  The AGO and control (CON) treatments were administered as a top dress. Cows fed AGO treatment received a premix composed of cornmeal + AGO, with AGO being included at a rate of 1 g/head/day (149.5 g cornmeal + 0.5 g AGO per feeding), while CON-fed cows received 150 g of cornmeal only per feeding.
Two cows were paired in each block, with each block comprising one AGO-fed and CON-fed cow. To accommodate two cows sampled for gas emission per day in the two head chambers, cow blocks were stagger-started onto their respective treatments. Blocks were randomly assigned to a respective day 0 emission sampling day and began receiving their treatments on day 1. Treatment with AGO or CON continued for 57 days.

Emission Sampling
On gas emission sampling days, the two cows were each secured in their respective head chamber (HC) using neck chains similar to a tie-stall system for a 12 h gas emissions sampling period from approximately 6:45 a.m. to 6:45 p.m. Cows were subjected to three training sessions within the HC prior to the start of the experiment, in order to become habituated to the HC. Cows were sampled every 14 days, on each of their respective days 0, 14, 28, 42, and 56 on treatment. Cows had ad libitum access to feed and water and could stand up or lie down during the HC gas emission sampling period.
Each HC consisted of a head chamber manufactured with clear polycarbonate sheeting, blowers pumping air out of the hoods, and Teflon tubing to extract emission samples. The Teflon tubing was connected from the HC to a mobile agricultural air quality laboratory, which housed all of the necessary equipment [26]. Concentrations of CH 4 , CO 2 , N 2 O, and NH 3 were analyzed using the Innova 1412 photo-acoustic multi-gas analyzer (LumaSense Technologies Inc., Ballerup, Denmark). The INNOVA 1412 analyzer had minimum detection limits of 0.4 ppm for CH 4 , 1.5 ppm for CO 2 , 0.03 ppm for N 2 O, and 1.0 ppm for NH 3 , and a maximum detection limit of 106 ppm. The continuous sampling cycle included the two HCs followed by ambient air, with each being sampled for 15 min intervals in sequence. The HCs in use were validated by Place et al. [26], and underwent both a pre-and post-trial validation in the present trial.

Emission Calculations
The measured gas concentrations of the outgoing air samples from the HC for each 15 min period were truncated to remove the first five minutes and last two minutes of each sample period for the prevention of carry-over effects. The total flux of gases in mg/h were calculated according to the equation outlined in Peterson et al. with an ambient air flow rate (FL) of 2300-2500 L min −1 [27]. The emission rate by cow for the HC (mg/h/head) was the same as the total flux of gases, as there was only one cow housed in each HC.

Milk Yield and Analysis
Milk yield for each cow was recorded daily at both the a.m. and p.m. milking sessions. Samples of milk were collected at consecutive a.m. and p.m. milking sessions on a weekly basis, and were sent for component analysis of milk fat, protein, and milk urea nitrogen (MUN) (Central Counties Dairy Herd Improvement Association, Atwater, CA, USA). Energy-corrected milk (ECM) was calculated according to the following formula [28]: ECM = (0.327 × milk kg) + (12.95 × milk fat kg) + (7.65 × milk protein kg). (1)

Blood Sampling
Blood samples were collected from each cow following the morning milking (hour 0) and before feeding on their respective days 1, 15, 29, 43, and 57 on treatment. Animals were secured in a chute and blood samples collected into 9 mL serum separator tubes (Corvac™ Serum Separator Tubes, Covidien, Mansfield, MA, USA) from the coccygeal vein prior to returning them to the free-stall pen. Cow blood samples were immediately centrifuged, and serum was stored at −18 • C. Frozen samples were transported to the UC Davis Veterinary Medical Teaching Hospital Clinical Diagnostic Laboratory Services (Davis, CA, USA) for analysis of SUN concentration.

Data Analysis
Gas emissions over the 12 h gas emission sampling period were summed to determine total gas production for the sampling period. Gas emissions were analyzed for intensity by calculating gas production per unit of energy-corrected milk (gas production/ECM), using ECM from the afternoon milking session which immediately followed the gas emission measurements. Gas yield was defined as the total production of the gas (i.e., summative emission measurement) per unit of DMI. Gas yield was calculated using DMI only while the cow was undergoing gas emission sampling in the HC (HC yield). Feed efficiency was calculated as kg ECM/kg daily DMI. Data were analyzed as a randomized complete block design with repeated measures, using the "nlme" and "emmeans" packages in R [29][30][31]. Blocks refer to each pair of parity and days in milk-matched cows. Gas emissions and production parameters were analyzed according to the following base model: where µ = the overall mean of the response variable in question; β i = overall mean of day 0 for the response variable in question; β k = cow (random) which was nested within β j = block (random); β l = treatment; β m = day; ε ijklm = the error term. Serial correlation structures and model selection were determined based on the Akaike information criterion, Bayesian information criterion, and log-likelihood [32]. Day 0 afternoon ECM, which was used to calculate gas intensity, was unavailable for one cow; following the model selection criteria and model fit, day 0 was excluded from the gas intensity and the head chamber ECM models. The data for each of the response variables were further verified for assumptions of normality by the Shapiro-Wilk method, with outliers removed accordingly where normality was not met. All means are presented as least squares means (LSMs) based on "emmeans" and comparisons between treatment LSMs were completed using the "anova" function. Test day means were compared using Tukey's test pairwise comparisons using "glht" and "cld" in "multcomp" [33]. Differences were declared significant at p < 0.05 and a trend toward significance at 0.05 ≤ p < 0.10.

Effect of AGO on GHG Emissions
Methane production was found to be similar between AGO-treated versus CON-treated cows (p > 0.05; Table 2). Methane production differed by day (p < 0.001; Table S1), whereas the interaction of treatment by day was not significant. Our findings are dissimilar to those of Hart et al., who found a significant decrease in enteric CH 4 production when cows were supplemented with AGO [22]. Hart et al. separated the AGO-treated from the CON-treated cows in group-fed and group-treated pens, rather than individually feeding and applying the treatment to the cow. The researchers could therefore not ensure that each cow consumed the allocated 1 g/head/day of AGO in this model [22], which makes extrapolation of their findings difficult. Similarly, Castro-Montoya found that enteric CH 4 production tended to decrease when cows were supplemented with AGO [21]. Castro-Montoya et al. used each cow's respective day 0 as a control in their experiment; it is therefore possible that temporal changes could have affected CH 4 production in each cow. Klop et al. noted a brief reduction in CH 4 production (p < 0.05) in the first 14-day period after AGO supplementation began, compared with pre-treatment CH 4 production. Although DMI was unaffected by AGO supplementation, CH 4 production increased and was no longer different from pre-treatment CH 4 production by the third 14-day period in Klop et al. [23].
In general, a strong positive correlation has been found between CH 4 production and individual animal DMI [3,34]. Cows that consume higher levels of DM have more substrate available for fermentation and more hydrogen available for methanogenesis, and are therefore generally associated with higher daily CH 4 emissions [35,36]. Gas yield (gas emissions/DMI) is therefore an important outcome to measure [36]. In the present trial, AGO-versus CON-fed cows showed similar HC yields for CH 4 , CO 2 , N 2 O, and NH 3 (p > 0.05; Table 2). Klop et al. found a reduction in CH 4 yield in AGO-supplemented cows when comparing the pre-treatment period to the first period (periods were 14 days in length); however, the difference was no longer present when comparing the pre-treatment period to the third or the fifth period [23]. Klop  In the present study, cows supplemented with AGO versus CON showed lower CH 4 intensity (p = 0.025; Table 2). The effect of day was found to be significant for CH 4 intensity (p < 0.001; Table S1), while the interaction of treatment by day was not significant (p > 0.05). Our findings are consistent with those of Hart et al., who found a reduction in CH 4 intensity in AGO-versus CON-treated cows [22]. Klop et al. similarly noted a decrease in CH 4 intensity when comparing the period in which cows were on AGO treatment to the cow's respective pre-treatment period [23]. Our findings are contrary to those by Castro-Montoya et al., who found no differences in CH 4 intensity when cows were supplemented with AGO [21], although they used actual kg milk instead of ECM. A cow could be more productive with respect to CH 4 intensity; however, this is diminished as herd size increases [37].
In the present trial, no differences between AGO-versus CON-treated cows were detected for CO 2 production, CO 2 HC yield, or CO 2 intensity (p > 0.05; Table 2). The effect of day was significant for CO 2 production (p < 0.001) and intensity (p < 0.001; Table S1), whereas the interaction of treatment by day was not significant for any of the CO 2 emission measurements (p > 0.05). Rumen methanogens have long been regarded nutritionally discriminatory, consuming select substrates such as CO 2 as a source of carbon, and H 2, formate, and acetate as sources of hydrogen [38]. Based on this, we would expect CO 2 emissions to either increase or remain unchanged by AGO supplementation. Although this was not the case, our findings were consistent with those of Melgar et al., who found decreased CO 2 production and no changes in CO 2 yield when dairy cows were supplemented with 3-nitrooxypropanol (3NOP) to reduce CH 4 emissions [39]. Hristov et al. saw no changes in CO 2 production in instances where CH 4 production was reduced in cows supplemented with 3NOP [39,40]. The dosing level of 3NOP was found to affect the CO 2 emission response, with CO 2 increasing as dosing levels increased [41]. Further research is therefore needed to determine if the dosage level of AGO could similarly affect CO 2 emissions in dairy cows. Table 2. Treatment least squares means (LSMs) for gas production, gas head chamber (HC) yield, and gas intensity of methane (CH 4 ), carbon dioxide (CO 2 ), nitrous oxide (N 2 O), and ammonia (NH 3 ) from Holstein dairy cattle to which Agolin (AGO) vs. untreated control (CON) diets were supplemented (n = 10 per treatment). Period = 12 h gas emission sampling period; 1 gas production per period × (1/kg dry matter intake (DMI) from the sampling period while in the HC); 2 gas production per period × (1/kg energy-corrected milk from the afternoon milking session).

Gas Production
No differences were found between AGO-versus CON-treated cows for N 2 O production, N 2 O HC yield, or N 2 O intensity in the present study (p > 0.05; Table 2). The effect of day was significant for N 2 O HC yield (p = 0.012), but not for N 2 O production or intensity (p > 0.05; Table S1). Similar to the other parameters, the interaction of treatment by day was not significant (p > 0.05). Despite making a smaller contribution to overall emissions from enteric fermentation, enteric N 2 O production has been quantified in the literature [42,43]. However, previous research with EO supplementation in dairy cows has not quantified enteric emissions of N 2 O.
Enteric NH 3 was the most impacted gaseous emission in the present trial, with NH 3 production (p = 0.028), and NH 3 intensity (p = 0.011) being lower among AGO-versus CON-treated cows. No difference was found for NH 3 HC yield ( Table 2). The effect of day was highly significant for NH 3 production, NH 3 HC yield, and NH 3 intensity (p < 0.001; Table S1), and the interaction of treatment by day was not significant for any of the parameters (p > 0.05). These findings are consistent with those of Castillejos et al., who found that the inclusion of eugenol led to a decrease in ruminal ammonia-N concentration when investigated in a batch fermentation system [44]. Coriander seed oil was also found to reduce ruminal ammonia-N concentration when compared to control-and salinomycin-treated cows [45]. The decrease in ammonia could be the result of the sensitivity of hyper-NH 3 -producing bacteria to EOs [11]. Working with another commercial EO, McIntosh et al. found that EOs may specifically affect the deamination of amino acids, which is the final step in protein catabolism [46]. The deamination of amino acids lead to more NH 3 being produced than can generally be consumed by ruminal microorganisms, resulting in nutritional losses [47]. Although NH 3 was measured within the ruminal fluid content for many of these studies, previous literature noted that NH 3 gas can form and be eructated from the rumen [48]. The reduction of NH 3 gas in the present trial may therefore be due to more nitrogen being retained by the animals, resulting in less nutritional loss.
Essential oils have demonstrated diverse mechanisms of action, which are used to interact with ruminal microorganisms. For example, some EOs interact with the external membranes of bacterial cells, which leads to conformational changes and the loss of stability of the cell membrane [14]. Other EOs act on microorganisms by coagulating the material within the cytoplasm of the cell [49]. The specific mechanism of action of AGO remains unclear. Further research should assess how the blend of EO within AGO individually and collectively interacts with and affects ruminal microorganisms.

Effect of AGO on Production Parameters
In the present study, daily DMI and head chamber DMI were similar between AGO-versus CON-treated cows (p > 0.05; Table 3). The effect of day on DMI was significant (p = 0.003; Table S2), whereas the interaction of treatment by day was not significant. Although Hart et al. found that AGO increased DMI [22], our present findings are consistent with those of both Elcoso et al. and Guasch et al., who saw no differences in DMI between treatment groups [18,25]. Table 3. Treatment least squares means (LSMs) for feed efficiency, daily dry matter intake (DMI), head chamber DMI, head chamber energy-corrected milk (ECM), ECM, milk fat, milk protein, milk urea nitrogen (MUN), and serum urea nitrogen (SUN) from Holstein dairy cattle fed Agolin (AGO) vs. untreated control (CON) (n = 10 per treatment). In the present trial, all production parameters, such as ECM, head chamber ECM, milk fat, and milk protein, were similar between treatments (p > 0.05; Table 3). For each of these production parameters, the effect of day was significant (p < 0.05), while the interaction of treatment by day was not (p > 0.05). Although they both focused on actual milk yield instead of ECM, our present findings are consistent with those of Castro-Montoya et al. and Santos et al., who found no differences in milk yield with AGO supplementation [21,50]. Effects of increased milk fat (kg/d) [22,50], and protein yield (kg/d) [22] have been found in previous AGO supplementation experiments; however, Castro-Montoya et al. and Elcoso et al. found no differences with respect to milk fat or protein, which is in agreement with our present findings [18,21]. In the case of Santos et al., it should be noted that the AGO treatment was applied to the pen and not the cow and the increase in milk fat yield with EO was just 0.03 kg/cow [50]. A meta-analysis conducted by Belanche et al. showed that supplementation with 1 g/head/day of AGO to dairy cattle improved ECM (referred to as FPCM) (response ratio = 1.031; p < 0.001) across the 20 studies that had addressed this parameter [24]. However, it is important to note that in addition to the published literature, the meta-analysis also incorporated unpublished experiments and information from on-farm trials.

AGO CON Treatment
In our trial, feed efficiency was similar between AGO-and CON-treated cows (Table 3). For feed efficiency, the effect of day was found to be highly significant (p < 0.001; Table S2), while the interaction of treatment by day was not significant (p > 0.05). Elcoso et al. and Guasch et al. saw increased feed efficiency in AGO-versus CON-treated cows, which is not consistent with our present findings [18,25]. The meta-analysis conducted by Belanche et al. showed an overall improvement in feed efficiency (response ratio = 1.030; p = 0.002) across 16 trials, when dairy cows were supplemented with 1 g/head/day of AGO [24]. This improvement in feed efficiency appears fairly common across various EO supplements. Another commercially available blend of EO, containing eugenol, cinnamaldehyde, and capsicum, was also found to improve feed efficiency in lactating Holstein cows [51]. Supplementing cows with an EO blend containing eugenol, thymol, and m-cresol and 10 other volatile compounds, Joch et al. noted a trend towards improved feed efficiency [52]. Braun et al. also found an increase in feed efficiency when supplementing Holstein dairy cows with a commercial blend of menthol, eugenol, and anethol [53].
In the present trial, MUN was similar between dietary treatments (p > 0.05; Table 3). The effect of day was highly significant for MUN (p < 0.001; Table S2), whereas the interaction of treatment by day was not significant (p > 0.05). Previous studies reported varying results when supplementing cows with EOs. Benchaar et al. similarly found no differences with respect to MUN when cows were supplemented with eugenol [54], which was also confirmed by Joch et al. [52]. However, in a series of experiments where cows were supplemented with Xtract 6965 (consisting of eugenol and cinnamaldehyde) at the same dosage levels, Tekippe et al. showed increased MUN concentrations when the supplement was mixed into the mineral premix (p < 0.001), but not when the supplement was administered as a top dress (p = 0.50) [55]. Dairy cow MUN may therefore differ based on the method in which the EO is supplemented.
In the present experiment, SUN concentrations were also found to be unaffected in AGO-versus CON-treated cows (Table 3). Test day was found to be highly significant for SUN (p < 0.001), and the interaction of treatment by day was not significant (p > 0.05). Experiments conducted on Holstein dairy heifers supplemented with cinnamaldehyde demonstrated no difference in plasma urea nitrogen (PUN) between EO-versus control-supplemented cows [56]. Supplementation of eugenol and cinnamaldehyde to multiparous Holstein dairy cows resulted in inconclusive results, with significantly higher PUN concentrations when the EO was mixed into the premix (p < 0.001) and significantly lower PUN when the EO was applied as a top dress (p = 0.03) [55].

Conclusions
Societal pressure and legislation have resulted in a need for California's dairy industry to reduce GHG emissions. Our present findings suggest that supplementing lactating dairy cow rations with 1 g/head/day of AGO may be part of an effective enteric CH 4 intensity mitigation strategy. Agolin also demonstrated a potential for reducing nitrogen-based gas emissions in mid-lactation dairy cattle, although additional research is needed to elucidate AGO's impact on nitrogen utilization. In order to form a more comprehensive understanding of the benefits of supplementation, future research should assess AGO's impact on ruminal microorganisms, and determine the EO blend's specific mode of action.