The effect of encapsulated nitrate and monensin on ruminal fermentation using a semi-continuous culture system

Abstract

Because enteric methane (CH₄) production from ruminants represents a source of greenhouse gas emissions and an energy loss for the host animal alternatives to minimize emissions is a current research priority. Seven 37-d trials tested the effect of encapsulated nitrate (EN) and sodium monensin (MON) in diets commonly fed to dairy (DAIRY; 50:50 forage to concentrate; four trials) and beef cattle (BEEF; 15:85 forage to concentrate; three trials) on rumen fermentation and CH₄ production using a semi-continuous fermentation system. A 3 × 2 factorial arrangement was used and additives (0, 1.25, and 2.5% of EN; 0 and 4 mg/L of MON) were tested alone and combined (EN + MON) totaling six treatments. Rumen fluid was pooled from five nonadapted lactating cows fed 50:50 forage to concentrate diet 3 h after morning feeding, and 1 L of processed inoculum was transferred to 2.2-L vessels. Treatment diets were added to nylon bags which remained in the anaerobic fermentation of mixed rumen microorganisms for 48 h. Nitrate decreased CH₄ production in DAIRY (24.7 vs. 32.1 mM/d; P < 0.01) and BEEF trials (33.5 vs. 43.5 mM/d; P < 0.01). Methane production was decreased by MON in DAIRY (26.3 vs. 32.1; P < 0.01) and BEEF (26.6 vs. 43.5 mM/d; P < 0.01). The combination of EN + MON further decreased CH₄ in DAIRY (21.3 vs. 32.1 mM/d; P = 0.03) and BEEF (19.3 vs. 43.5 mM/d; P = 0.01). Nitrate did not affect major VFA production in DAIRY and BEEF trials, but significantly decreased digestion of protein (96.8 vs. 97.6%; P < 0.01) and starch (79.0 vs. 80.4%; P < 0.01) in DAIRY and NDF (29.3 vs. 32.5%; P < 0.01) and starch (88.5 vs. 90.3%; P < 0.01) in BEEF. Monensin significantly affected VFA pattern with an increase in propionate (P < 0.01) and a decrease on acetate (P < 0.01) production with consequent decrease on acetate-to-propionate ratio in DAIRY (1.6 vs. 2.0; P < 0.01) and BEEF (1.6 vs. 1.9; P < 0.01). Monensin decreased NDF digestion in BEEF only (29.3 vs. 32.5 %; P < 0.01). Significant concentrations of nitrate and nitrite were detected only for EN and EN + MON (P < 0.01). Nitrate and MON effectively decreased CH₄ production when fed separately and the combination of additives additively decreased CH₄ production.

INTRODUCTION

Enteric methane (CH₄) production represents energy loss to the animal and a source of pollution. Microbial fermentation in the rumen is an oxidative process during which reduced cofactors are reoxidized through dehydrogenation reactions that release hydrogen (H₂) in the rumen. The H₂ is used by methanogenic archaea to reduce carbon dioxide (CO₂) into CH₄ (Morgavi et al., 2010), an evolutionary essential reaction in the rumen ecosystem that prevents H₂ accumulation. However, livestock contributes with 6.3% of total anthropogenic GHG emissions, with CH₄ emissions from enteric fermentation accounting for 2.1 Gt CO₂ Eq/yr and manure management accounting for 0.99 Gt CO₂ Eq/yr (IPCC, 2014). In addition, CH₄ represents a loss of 2–12% of diet GE (Johnson and Johnson, 1995) that could be rechanneled toward usable products for the host (Grainger and Beauchemin, 2011). Therefore, assuming the efficiency of ruminal metabolism is not compromised, reducing CH₄ production by ruminants contributes to global efforts to reduce GHG emissions and can improve the efficiency of feed use.

Previous work demonstrated an additive effect of nitrate and monensin (MON) on reduction of CH₄ production, in vitro (Capelari and Powers, 2017). Microbes that reduce nitrate to ammonia (NH₃) in the rumen compete with methanogens to scavenge H₂ (Leng, 2008) and intermediates of nitrate reduction are suggested to be toxic to methanogens (Asanuma et al., 2015). Monensin reduces the number of H₂-producing bacteria (Chen and Wolin, 1979) and protozoal populations (Guan et al., 2006). Hence, nitrate and MON affect the H₂ pool in the rumen through different mechanisms. We hypothesize that encapsulated nitrate (EN) and MON additively reduce CH₄ production, and, if adaptation to MON occurs, EN would maintain CH₄ levels below that of control. The objective of this study was to determine the effect of EN and MON, alone and in combination, on ruminal fermentation parameters.

MATERIALS AND METHODS

Experiments were conducted at Michigan State University and consisted of two studies one with four replications (DAIRY) and the other with three replications (BEEF), where a semi-continuous culture of mixed rumen microorganisms was used to assess the effect of EN, MON, and EN + MON on rumen fermentation parameters (total gas production, CH₄ production, nitrous oxide (N₂O) production, VFA production, pH, and nutrient disappearance) in two diets (DAIRY, 50:50 forage to concentrate; and BEEF, 15:85 forage:concentrate). Because MON resistance of microbial species have been reported to occur during the initial 4 wk following supplementation (Carmean and Johnson, 1990; Johnson et al., 1997; Guan et al., 2006), each trial consisted of 37 d total, with the initial 7 d for stabilization of the system (steady-state) followed by 30 d of data collection. All procedures were consistent for each trial, unless otherwise stated. Animal procedures were approved and followed internal guidelines recommended by the Animal Care and Use Committee of Michigan State University (IACUC study number 11/15-175-00).

Experimental Design and Treatments

Trials were arranged as a completely randomized design in a 2 × 3 factorial arrangement of treatments. Nitrate (0, 1.25, and 2.5% of EN; GRASP Ind. & Com. LTDA EW|Nutrition GmbH, Curitiba, Brazil) and sodium MON (0 and 4 mg/L; Elanco Animal Health, Greenfield, IN) were tested alone and in combination, totaling six treatments. The EN source was calcium ammonium nitrate decahydrate (5Ca(NO₃)₂ ∙ NH₄NO₃ ∙ 10H₂O; 85.6% DM, 17.6% N, 19.6% Ca, and 71.4% nitrate on a DM basis) designed to release 50, 80, and 100% of nitrate within 4, 12, and 30 h, respectively (El-Zaiat et al., 2014; Lee et al., 2015a, 2015b, 2017). Urea and calcium carbonate were added when necessary to compensate for the additional nonprotein nitrogen and Ca provided with EN, therefore maintaining N and Ca levels similar across treatments. The fermentation system consisted of 18 vessels (see detailed description of the system below); each treatment was replicated three times per trial. Vessel within trial was considered the experimental unit. The MON level tested (4 mg/L of culture media) was chosen because it represents an intermediary level of that tested previously, where the higher MON level caused negative responses to IVDMD when combined with the higher dose of nitrate (Capelari and Powers, 2017). The nitrate levels tested were chosen because 2.5% of DM was considered as a threshold for nitrate supplementation due to inhibition of feed intake (Lee et al., 2017) while the same level has shown to consistently reduce CH₄ levels, in vivo (Lee and Beauchemin, 2014).

Two contrasting basal diets were used for the seven trials. DAIRY trials (four total) were conducted with a basal diet consisting of a 50:50 forage to concentrate ratio commonly fed to high producing dairy cows in the upper Midwest, and BEEF trials (three total) used a high concentrate diet (15:85 forage to concentrate ratio; Table 1). Sufficient diet substrate for both DAIRY and BEEF trials were prepared in a single day in order to avoid changes in ingredients chemical composition. Following preparation as a total mixed ration (TMR), basal diets were dried for 48 h at 55 °C, ground through a 5-mm screen (Wiley mill model 4, Thomas Scientific, Swedesboro, NJ) and frozen at −20 °C until use. Prior to initiation of each trial basal diets were thawed, basal diet aliquots (20 g) were weighed and additives (EN and MON), urea and calcium carbonate were added to basal diet as needed to obtain final treatment diets (Table 2). Treatment diets were stored in prelabeled and preweighed nylon bags (10 × 20 cm; 50 µm mesh size; ANKOM Technology, Macedon, NY) until the day of use.

Table 1.

Ingredients and chemical composition of basal diets for DAIRY (50:50 roughage: concentrate) and BEEF (15:85 roughage: concentrate) trials

Item (DM, %)	DAIRY	BEEF
Ground corn	18.1
Alfalfa haylage	17.9
Soybean meal	15.8
High moisture corn	15.7	73.0
Corn silage	17.9	15.0
Dry distillers		8.0
Cottonseed	6.5
Wheat straw	2.7
By pass fat1	2.0
Premix2	2.0	4.0
Sodium bicarbonate	0.75
Calcium carbonate3	0.66
Analyzed chemical composition (% of DM)
DM	56.4	64.0
CP	17.5	12.1
NDF	25.0	15.4
NFC	46.6	64.4
Starch	29.2	57.5
Ash	5.5	3.9

¹Megalac (97% DM; 6.83 Mcal/kg DM; 84.5% fat; 15.5% ash).

²Premix contained 17.0% Ca, 8.0% P, 10.0% Mg, 3.000 mg/Kg Fe, 75 mg/Kg I, 8.000 mg/Kg Zn, 2.500 mg/Kg Cu, 2.000 mg/Kg Mn, 45 mg/Kg Co, 2.000 mg/Kg F, 1.666.000 IU/Kg Vit. A, 366.666 IU/Kg Vit. D, and 20.000 IU/Kg Vit. E.

³Calcium carbonate contained 38% Ca.

Table 2.

Composition of experimental treatments of the Rusitec DAIRY (50:50 roughage: concentrate) and BEEF (15:85 roughage: concentrate) trials¹

Item, DM, % (unless otherwise stated)	0 EN		1.25 EN		2.5 EN
	0 MON	4 MON	0 MON	4 MON	0 MON	4 MON
Basal diet	97.5	97.5	97.5	97.5	97.5	97.5
Nitrate source2	0	0	1.25	1.25	2.5	2.5
Monensin,3 mg/L	0	4	0	4	0	4
Urea	1.2	1.2	0.65	0.65	0	0
Calcium carbonate4	1.3	1.3	0.6	0.6	0	0

¹EN = encapsulated calcium nitrate (0, 1.25 and 2.5% of diet DM); MON = sodium monensin (0 and 4 mg/L).

²Nitrate source was manufactured by GRASP Ind. & Com. LTDA and EW|Nutrition GmbH. The source of nitrate was the double salt of calcium ammonium nitrate decahydrate [5Ca(NO₃)₂ ∙ NH₄NO₃ ∙ 10H₂O]; 85.6% DM; 17.6% N, 19.6% Ca, and 71.4% nitrate on a DM basis.

³Monensin sodium (C₃₆H₆₁NaO₁) was manufactured by Elanco Animal Health as Rumensin 90; 198 g of active ingredient per kg.

⁴Calcium carbonate contained 38% Ca.

Rumen Simulation System

The rumen simulation system was designed and constructed at Washington State University Department of Animal Science Farm Shop and represented an adaptation of the original Rusitec system (Czerkawski and Breckenridge, 1977). The system had the capacity to administer artificial saliva, collect overflow and gas collection, and accommodate feeding of different diets. The equipment was similar to that described by Czerkawski and Breckenridge (1977), but it contained 18, 2.2 L vessels and located the artificial saliva inflow on top of vessels. Vessels were collectively located in a large water bath (approximately 400 L volume) with two immersion-circulating heaters that maintained the temperature at 39 ± 0.5 °C. Vessels were gas-tight with a cap that, when fixed, maintained the fermentation vessels thermodynamically as a closed system. Located on the top of caps was an inlet for artificial saliva, an outlet for the gas that accumulated in the headspace, and an outlet for effluent. The artificial saliva reservoir was connected to vessels through Tygon tubing (3.1 mm ID × 6.3 mm ED; Cole-Parmer Lab Supplies, Vernon Hills, IL) connected to a peristaltic pump that was installed at the top of the apparatus. In order to store the gas for a 24-h period, gas outlets were connected to 10-L Tedlar sampling bags (Supelco Inc, Bellefonte, PA; Rahman et al., 2013). Liquid effluent was stored in 2 L plastic containers connected to each vessel by Tygon tubing (9.5 mm ID × 12.7 mm ED; Cole-Parmer Lab Supplies).

A hydraulic pump was used to simulate rumen movements. A perforated feed container containing diet bags with treatment was located in each vessel and connected to the hydraulic pump in order to mix vessel contents, avoiding the formation of dead spaces, and allowing contact between liquid rumen fluid and feed bags.

Incubation Conditions and Initial Experimental Procedures

A mixture (1:1; 1.2 L) of artificial saliva (McDougall, 1948; pH = 8.3; 9.8 g/L of NaHCO₃, 3.72 g/L of Na₂HPO₄, 0.47 g/L of NaCl, 0.57 g/L of KCl, 0.053 g/L of CaCl₂·2H₂O, 0.128 g/L of MgCl₂·6H₂O, and 0.3 g/L of (NH₄)₂SO₄) and water was added to all vessels on day 8. Ruminal inoculum donors were housed at the Michigan State University Dairy Cattle Teaching and Research Facility. On day 7, ruminal inoculum was collected from five ruminally fistulated lactating cows (different cows in each trial; there were no available animals that could be fed the BEEF diet and serve as inoculum donors) consuming a 50:50 roughage: concentrate diet similar to that used for DAIRY trials (DM basis—approximately 18% ground corn, 18% alfalfa haylage, 15% soybean meal, 15% high moisture corn, and 18% corn silage) in the form of TMR. Diets of rumen fluid donors did not contain any ionophore or supplementary nitrate. Rumen fluid was manually collected 3 h following morning feeding from four different rumen locations of each donor and transferred to three preheated 30 L insulated containers. The ruminal inoculum was immediately transported to Michigan State University Animal Air Quality Research Facility where it was blended and filtered through two layers of cheesecloth. The processed fluid (1 L) was transferred under continuous CO₂ flow to each vessel. Approximately 40 g of wet solid rumen contents from inoculum donors were weighed into a prelabeled bag and added to feed containers on day 7 along with 20 g of basal diet. Following incubation with rumen inoculum, vessels were closed and the peristaltic pump and hydraulic pump were started. Artificial saliva was prepared daily with a dilution rate set to 0.7 d⁻¹ (1,600 mL/d) and hydraulic pump set for three complete rotational movements per minute. Bags containing solid rumen contents were removed from feed containers on day 6 and replaced with a second bag of basal diet. From day 6 through day 0, two bags with 20 g of basal diet was present in each vessel for 48 h. Each morning at 0600 h, a new feed bag was added to all vessels and the bag that had been in the vessel 48 h was removed. On day 0, the feed bag added contained the assigned experimental treatment.

Sampling and Analytical Procedures

Daily, before feed bag exchange, effluent volume was measured and bags containing gas produced over 24 h were closed and replaced by a second set of bags. Following, all vessels were opened (three at a time), flushed with pure CO₂ to maintain anaerobic conditions, and a 5 mL sample was taken from liquid phase for VFA analysis. After a new feed bag was added to feed containers, vessels were immediately closed and CO₂ was flushed in excess for 5 s through the gas outlet port. Gas samples were collected from gas bags with a 10 mL syringe (Model 1010 C, Hamilton Company, Reno, NV) and transferred to 20 mL gas chromatography (GC) vials containing 15 mL of ultra-high purity N₂ gas for off-site analysis for CH₄, N₂O, and CO₂ concentration. The analytical procedures used for gas and VFA measurements considered total gas and effluent volume, respectively, and concentration of analytes obtained by GC, so total daily production of fermentation products could be measured. A detailed description of the GC method used was described in detail previously (Capelari and Powers, 2017). Total gas production was calculated by difference between relative pressure, corrected for temperature and humidity, and measured using a relative pressure gauge (Model Media Gauge, SSI Technologies, Jonesville, WI). A homemade container system (5 L) enabled the transfer and measurement of gas samples accumulated in the bags. To determine gas volume by pressure difference, the container was connected to a vacuum pump and the pressure gauge. The vacuum pump was started and when the pressure reading stabilized (approximately 3.44 kPa) a gas valve was closed and the value noted. Each bag was then connected and the container was opened with the gas content of bags transferred via negative pressure to the container. The relative pressure of container was measured and recorded for a second time and the value used in the equation as follows:

$$V_{\text{gas}}=(V_{\text{c}}/P_{\text{EL}})\times P_{\text{BT}}$$ (1)

where V_gas is the volume of gas production, ml; V_c is the volume of solid container, 5,000 mL; P_EL is the atmospheric pressure in East Lansing, MI, 101.3 kPa; and P_BT are the pressure measurements from the gauge, psi (final read − initial read).

Samples for analysis of NH₃-N, nitrate, and nitrite (NO₂) in rumen fluid were collected on days 0, 2, 9, 16, 23, and 30. For NH₃-N, an aliquot of 30 mL was sampled directly from the vessels before feed bag exchange and transferred to 50 mL tubes containing 1 mL of concentrated sulfuric acid. The samples were frozen at −20 °C until analysis of NH₃-N using a Kjeldahl distillation apparatus (KJELTEC system 1002 distilling unit, FOSS, Eden Prairie, MN). Samples for nitrate and NO₂ analysis (5 mL) were centrifuged (15,000 × g for 15 min at 4 °C) and concentration was measured with a colorimetric assay kit (Cayman Chemical Company, Ann Arbor, MI). On the same days, pH of the fermentation vessels was measured using a pH meter (Model HQ40d Portable pH meter, HACH Co, Reno, NV).

Nutrient disappearance over 48 h was measured twice weekly on days −1, 1, 2, 8, 10, 15, 17, 22, 24, 28, and 29. Feed bags withdrawn from the vessels were gently washed with water until the effluent ran clear. The bags were dried for 48 h at 55 °C for DMD determination. The residues from the 11 sampling days were then pooled, equally, and ground through a 1 mm screen (A.H. Thomas, Philadelphia, PA, USA). Samples were further dried at 105 °C for 2 h. Composite samples from residues were then analyzed for DM, NDF, starch, and CP. Chemical analyses were performed on each sample, in triplicate, and when the coefficient of variation was greater than 5% the analysis was repeated. The NDF was determined according to Van Soest et al. (1991) with heat-stable amylase and sodium sulfite used in the procedure. Starch content was determined by enzymatic hydrolysis of α-linked glucose polymers as described by Karkalas (1985). Total Kjeldahl N content was analyzed according to standard method (AOAC, 1990), and CP was calculated as N × 6.25. Microbial protein production was not measured because of internal limitations in the materials available at the ruminant nutrition laboratory at Michigan State University. Samples of ingredients used in the TMR in DAIRY and BEEF trials were analyzed for DM, CP, NDF, nonfiber carbohydrates, starch, and ash by wet chemistry (Dairy One Forage Analysis Laboratory, Ithaca, NY).

Fermentation Balance Calculations

All equations used here are described in detail by Wolin (1960), Demeyer and Tamminga (1987), and Demeyer (1991). In summary, the ratio of H₂ utilized and H₂ produced (µ mol/mL; Eq. 2 and 3, respectively) is used to calculate percent H₂ recovery (%; Eq. 4). The H₂ produced as fermentation end products and H₂ consumed to form CH₄ and VFA were determined from molar concentration of acetate (C₂), propionate (C₃), butyrate (C₄), isovalerate (Ci₅), valerate (C₅), and CH₄. The equations do not account for H₂ released in the gaseous form, lactate, microbial mass, and potential acetate produced via reductive acetogenesis.

$${\text{H}}_2\text{ utilized}=\left(\text{2}\times {\text{C}}_3\right)+\left(\text{2}\times {\text{C}}_4\right)+\left(\text{4}\times {\text{CH}}_{\text{4}}\right)+{\text{Ci}}_{\text{5}}\text{,}$$ (2)

in which H₂ utilized is expressed as micromoles per milliliters,

$${\text{H}}_2\text{ produced}=\left(\text{2}\times {\text{C}}_2\right)+{\text{C}}_3+\left(\text{4}\times {\text{C}}_4\right)+\left(\text{2}\times {\text{C}}_5\right)+\left(\text{2}\times {\text{Ci}}_{\text{5}}\right)\text{,}$$ (3)

in which H₂ produced is expressed as micromoles per milliliters,

$${\text{H}}_2\text{ recovery}=\left({\text{H}}_2\mathrm{utilized}/{\text{H}}_2\mathrm{produced}\right)\times \text{1}00\text{,}$$ (4)

in which H₂ recovery is expressed as a percent,

The amount of hexose (C₆) fermented was calculated as follows:

$$\text{Hexose fermented}=\left(0.\text{5}\times {\text{C}}_2\right)+\left(0.\text{5}\times {\text{C}}_3\right)+{\text{C}}_4+{\text{C}}_5\text{,}$$ (5)

in which hexose fermented is expressed in micromoles per milliliter.

Fermentation efficiency was calculated by considering the heat of combustion of glucose, acetate, propionate, and butyrate and their molar concentration (mM), as follows:

$$\text{Fermentation efficiency}=\left\{\left[\left(0.\text{62}\times {\text{C}}_2\right)+\left(\text{1}.0\text{9}\times {\text{C}}_3\right)+\left(0.\text{78}\times {\text{C}}_4\right)\right]/\left({\text{C}}_2+{\text{C}}_3+{\text{C}}_4\right)\right\}\times \text{1}00\text{,}$$ (6)

in which fermentation efficiency is expressed as a percent.

Statistical Analysis

Data were tested for normality and submitted to analyses of variance with 5% significance level. The MIXED procedure of SAS (SAS Inst., Inc., Cary, NC) was used for all statistical analysis and inferences. Means for all variables were obtained by LSMEANS. The statistical model was as follows:

$${\text{Y}}_{i\left(k\right)j}{}_{\text{m}}=\mu +\beta i_{(k)}+{\tau }_{\text{j}}+{\lambda }_{\text{m}}+\tau {\lambda }_{\text{jm}}+{\text{e}}_{i\left(k\right)j}{}_{\text{m}}$$

where Y_ij(k)m represents observation _ij(k)m; µ represents the overall mean; â_i represents the random effect of ith vessel within kth trial; ô_j represents the fixed effect jth treatment and its interaction, ë_m represents the fixed effect of the mth day, and ôë _jm is the interaction between jth treatment and mth day. The residual terms ε_ij(k)m were assumed to be normally, independently, and identically distributed with variance σ²_e. The REPEATED statement was used for variables measured daily (gas variables and VFA profile). The variance components were estimated using the REML method which was selected based on the lowest Alkaine criteria. Differences of least square means were adjusted by the Tukey–Kramer test. Α level of P ≤0.05 was used to determine significance, and tendencies were associated with P values between 0.05 and 0.10.

RESULTS

DAIRY Trials (50:50 Roughage: Concentrate)

The addition of EN decreased total daily gas production when compared to control (2.47 vs. 2.54 L/d; P < 0.01; Table 3). Methane production was significantly decreased (P < 0.01) by EN (24.77 mM/d) and MON (26.34 mM/d) when compared to control (32.15 mM/d). The combination of EN + MON further decreased CH₄ production when compared to control (21.34 vs. 32.15 mM/d; P < 0.01). There were no effect of treatments on CO₂ production, however N₂O was increased by EN (41 vs. 33 (×10^–4) mM/d; P < 0.01). There were no effect of treatments on DM (%) and NDF (%) disappearance, however, EN significantly decreased CP (96.9 vs. 97.6%; P < 0.01) and starch (79.0 vs. 80.8%; P < 0.01) disappearance.

Table 3.

Effects of EN and MON on gas production and nutrient disappearance on DAIRY (50:50 roughage: concentrate) trials¹

	0 EN		1.25 EN		2.5 EN			P value1
Item	0 MON	4 MON	0 MON	4 MON	0 MON	4 MON	SEM	EN	MON	EN × MON
Dilution rate (mL, d-1)	1,523	1,526	1,576	1,545	1,518	1,514	27.91	NS	NS	NS
Gas production
Total gas (L)	2.57b	2.57b	2.48b	2.60b	2.71a	2.40c	0.20	0.01	NS	NS
CH4, mM/d	32.15a	26.34b	26.77b	23.66c	22.78c	19.03d	1.79	<0.01	<0.01	0.03
CO2, mM/d	75.32	71.64	68.11	75.91	71.58	75.85	5.06	NS	NS	NS
N2O (×10–4), mM/d	33c	37b	35b	36b	39ab	46a	3.8	<0.01	NS	NS
Nutrient disappearance (%)
DM	64.90	65.35	65.16	64.66	64.29	64.21	1.36	NS	NS	NS
NDF	27.64	26.64	28.21	28.4	28.56	26.36	1.15	NS	NS	NS
CP	97.63ab	97.91a	97.0b	97.17ab	96.85b	96.86b	0.38	<0.01	NS	NS
Starch	80.78a	81.12a	79.92ab	80.12a	78.17b	78.22b	0.63	<0.01	NS	NS

^{a– d}Within a row, means without a common superscript letter differ, P < 0.05.

¹EN = encapsulated calcium nitrate (0, 1.25, and 2.5% of diet DM); MON = sodium monensin (0 and 4 mg/L).

²NS = nonsignificant; (P > 0.10).

Compared to control, total VFA production was increased by MON (35.4 vs. 38.3 mM; P < 0.01; Table 4). The primary effect of MON of VFA molar proportions was a decrease in acetate proportion (47.6 vs. 49.9 mol/100 mol; P < 0.01) and an increase in propionate proportion (26.8 vs. 22.2 mol/ 100 mol; P < 0.01). As a consequence, the A:P ratio was decreased by MON when compared to control (1.6 vs. 2.0; P < 0.01) while EN tended to increase A:P when compared to control (2.15 vs. 2.0; P < 0.1) despite no effect on acetate and propionate individually. However, butyrate was decreased by MON (16.1 vs. 19.4 mol/100 mol; P< 0.01), EN (18.2 vs. 19.4 mol/100 mol; P < 0.05) and tended to further decrease by EN + MON (15.2 vs. 19.4 mol/100 mol; P < 0.1). Nitrate significantly decreased NH₃-N concentration (17.1 vs. 21.09 mg/dL; P < 0.01) and increased both nitrate and NO₂ (P < 0.01) in rumen fluid. No effect of treatments was observed for pH (6.8 ± 0.4).

Table 4.

Effects of EN and MON on fermentation parameters on DAIRY (50:50 roughage: concentrate) trials¹

	0 EN		1.25 EN		2.5 EN			P value2
Item	0 MON	4 MON	0 MON	4 MON	0 MON	4 MON	SEM	EN	MON	EN × MON
pH	6.90	6.89	6.95	6.79	6.92	6.89	0.22	NS	NS	NS
Total VFA, mM	35.4b	38.3a	34.3b	36.7b	34.8b	37.8ab	2.08	NS	<0.01	NS
VFA composition, mol/100 mol
Acetate (A)	49.9b	47.6a	50.0b	53.7c	50.8bc	49.8b	1.17	NS	NS	<0.01
Propionate (P)	22.2b	26.8a	20.4b	27.5a	21.7b	26.1a	0.74	NS	0.01	NS
Butyrate	19.4a	16.1b	19.6a	15.1b	16.9b	15.3b	0.59	<0.05	<0.01	NS
Valerate	7.3	8.4	7.0	7.4	6.9	7.2	0.44	NS	NS	NS
Isobutyrate	2.3	2.6	2.7	4.5	2.9	2.6	0.56	NS	NS	NS
Isovalerate	4.1	3.8	3.9	4.5	3.6	3.4	0.47	NS	NS	NS
Caproate	3.6	3.1	3.8	3.2	3.5	2.8	0.32	NS	<0.05	NS
A:P	2.0b	1.6c	2.2a	1.6c	2.1ab	1.8c	0.06	<0.1	<0.01	NS
Metabolic hydrogen balance,3 µmol/mL of fermentation media
Produced	205.8a	197.3ab	208.4a	193.5c	199.3ab	197.2ab	2.9	NS	<0.01	NS
Utilized	178.5a	154.1b	151.9b	140.6bc	136.7c	125.7c	4.8	<0.01	<0.01	<0.01
Recovered, %	86.7a	78.2a	73.3a	72.0a	68.6b	63.4b	3.8	<0.01	<0.01	<0.01
Hexose fermented	32.0	35.3	32.5	31.4	36.3	32.0	1.9	NS	NS	NS
Fermentation efficiency, %	76.6bc	79.2a	76.2c	77.7b	75.8c	77.8b	1.2	NS	<0.01	NS
NH3-N (mg/dL)	21.09a	20.77a	18.10b	18.48b	16.21c	17.73b	0.72	<0.01	NS	NS
NO3−, µM	0	0	0.07	1.1	1.4	1.0	0.002	<0.01	NS	NS
NO2−, µM	0	0	0.7	3.6	4.7	1.2	0.01	<0.01	NS	NS

^a–dWithin a row, means without a common superscript letter differ, P < 0.05.

¹EN = encapsulated calcium nitrate (0, 1.25, and 2.5% of diet DM); MON = sodium monensin (0 and 4 mg/L).

²NS = nonsignificant; (P > 0.10).

³Stoichiometric calculations were based on balances applied to rumen fermentation (Wolin, 1960; Demeyer and Tamminga, 1987; Demeyer, 1991).

BEEF (15:85 Roughage: Concentrate)

Total gas production when compared to control was decreased by EN (3.25 vs. 3.05 L; P < 0.01), MON (3.25 vs. 3.06 L; P < 0.01), and with further decrease by EN + MON (3.25 vs. 2.66; P = 0.06; Table 5). When compared to control, CH₄ production was decreased by EN (43.5 vs. 33.9 mM/d; P < 0.01), MON (43.5 vs. 26.7 mM/d; P < 0.01), and further decreased when additives were combined (43.5 vs. 19.3 mM/d; P = 0.01). The EN treatment decreased CO₂ production when compared to control (98.6 vs. 103.0 mM/d; P < 0.01), but no treatment effect was observed on N₂O production. Feeding EN significantly decreased disappearance of DM (73.5 vs. 74.4%; P < 0.01), NDF (29.3 vs. 32.5%; P < 0.01), and starch (88.5 vs. 90.35 %; P < 0.01) while MON significantly decreased disappearance of NDF (29.34 vs. 32.5%; P < 0.01) and increased CP (83.94 vs. 82.56%; P < 0.01).

Table 5.

Effect of EN and MON on rumen fermentation parameters of BEEF (15:85 roughage: concentrate) trials¹

	0 EN		1.25 EN		2.5 EN			P value2
Item	0 MON	4 MON	0 MON	4 MON	0 MON	4 MON	SEM	EN	MON	EN × MON
Dilution rate (mL, d−1)	1,523	1,559	1,511	1,498	1,557	1,552	35.5	NS	NS	NS
Gas production
Total gas (L)	3.25a	3.06ab	3.20ab	2.93bc	2.90c	2.39d	0.20	<0.01	<0.01	0.06
CH4, mM/d	43.57a	26.67c	35.67b	20.03d	32.20b	18.56d	1.79	<0.01	<0.01	0.01
CO2, mM/d	103.03	101.34	108.8	98.73	88.43	88.61	4.23	<0.01	NS	NS
N2O (×10−4), mM/d	34	21	22	32	30	31	3.8	NS	NS	NS
Nutrient disappearance
DM	74.38a	74.90a	74.54a	73.97ab	72.61bc	71.70c	1.36	<0.01	NS	NS
NDF	32.48a	29.34b	29.88a	25.84b	28.87ab	24.6b	1.73	<0.01	<0.01	NS
CP	82.56	83.94	82.68	83.17	81.98	82.37	0.51	NS	0.08	NS
Starch	90.35a	90.24a	89.3ab	89.4ab	87.8b	87.43b	0.64	<0.01	NS	NS

^a–dWithin a row, means without a common superscript letter differ, P < 0.05.

¹EN = encapsulated calcium nitrate (0, 1.25, and 2.5% of diet DM); MON = sodium monensin (0 and 4 mg/L).

²NS indicates P > 0.1.

Total VFA production increased by MON (50.7 vs. 48.1; P < 0.01) and by EN + MON (52.3 vs. 48.1; Table 6; P < 0.01). The molar proportion of acetate tended to increase due to EN + MON (51.2 vs. 49.5 mol/100 mol; P = 0.07) while the propionate molar proportion was increased by MON (29.6 vs. 26.2 mol/100 mol; P < 0.01). Feeding MON decreased the butyrate molar ratio when compared to control (12.6 vs. 14.9 mol/100 mol; P < 0.01) and isovalerate (1.6 vs. 2.0 mol/100 mol; P < 0.01). Because MON increased propionate the A:P ratio was decreased by MON when compared to control (1.6 vs. 1.9; P < 0.01). Nitrate significantly decreased NH₃-N concentration (6.63 vs. 8.14 mg/dL; P < 0.01) but increased both nitrate and NO₂ (P < 0.01) in rumen fluid. No effect of treatments was observed for pH (6.7 ± 0.3).

Table 6.

Effects of EN and MON on fermentation parameters on BEEF (15:85 roughage: concentrate) trials¹

Item	0 EN		1.25 EN		2.5 EN		SEM	P value2
	0 MON	4 MON	0 MON	4 MON	0 MON	4 MON		EN	MON	EN × MON
pH	6.7	6.8	6.7	6.9	6.8	6.7	0.22	NS	NS	NS
Total VFA, mM	48.1bc	50.7b	44.7c	55.4a	47.0bc	49.2bc	1.9	NS	<0.01	<0.01
VFA composition, mol/100 mol
Acetate (A)	49.5 ab	49.9ab	49.4ab	51.1a	48.6b	51.3a	1.17	NS	NS	0.07
Propionate (P)	26.2b	29.6a	26.0b	26.5b	27.3ab	27.2ab	0.74	NS	<0.01	NS
Butyrate	14.9b	12.6c	16.1a	13.4c	15.3ab	12.7c	0.59	<0.01	<0.01	NS
Valerate	4.6c	5.6ab	5.1bc	5.7a	4.8c	5.4ab	0.19	NS	<0.01	NS
Isobutyrate	1.03a	0.97ab	0.89ab	0.88ab	1.04a	0.79b	0.12	NS	0.02	0.03
Isovalerate	2.01a	1.61b	1.74ab	1.82ab	1.98a	1.47b	0.47	NS	<0.01	NS
Caproate	1.6	1.2	1.0	1.2	1.2	1.4	0.32	NS	NS	NS
A:P	1.9a	1.6b	1.9a	1.9a	1.8ab	1.9a	0.06	NS	<0.01	NS
Metabolic hydrogen balance,3 µmol/mL of fermentation media
Produced	196.8a	187.6b	198.4a	195.6a	195.1a	192.4ab	1.8	NS	<0.01	NS
Utilized	177.5a	103.1c	144.8b	85.5d	132.3b	82.1d	10.3	<0.01	<0.01	<0.01
Recovered, %	90.1a	54.9c	72.9b	43.4d	67.8b	42.6d	3.7	<0.01	<0.01	<0.01
Hexose fermented	41.9	41.8	39.6	43.5	39.4	42.1	1.3	NS	NS	NS
Fermentation efficiency, %	79.9	80.6	79.6	80.3	78.5	81.4	0.9	NS	<0.01	NS
NH3-N, (mg/dL)	8.14a	7.42a	7.35a	7.30a	5.92b	4.76b	0.41	<0.01	NS	NS
NO3−, µM	0	0	0.01	0.06	0.02	0.05	0.006	<0.01	NS	NS
NO2−, µM	0	0	0.1	0.17	0.11	0.16	0.008	<0.01	NS	<0.08

^a–dWithin a row, means without a common superscript letter differ, P < 0.05.

¹EN = encapsulated calcium nitrate (0, 1.25, and 2.5% of diet DM); MON = sodium monensin (0 and 4 mg/L).

²NS = non significant; (P > 0.10).

³Stoichiometric calculations were based on balances applied to rumen fermentation (Wolin, 1960; Demeyer and Tamminga, 1987; Demeyer, 1991).

DISCUSSION

Fermentation Products

The rumen simulation technique (Rusitec) was developed for testing of fibrous feed (Czerkawski and Breckenridge, 1977), but the approach has been successfully used for mixed diets with different amounts of starch (Carro et al., 1992, 2009), for evaluation of feed additives on rumen metabolism (McAllister et al., 1994), and for development of CH₄ mitigation strategies (Avilla-Stagno et al., 2014; Romero-Pérez et al., 2015). The Rusitec presents several positive features for rumen studies, such as 1) sampling of fermentation products, like gas and ruminal fluid, 2) testing of different treatments, with sufficient replication, and the 3) use of high concentrations of feed additives, which could potentially be toxic to animals, all at a 4) lower cost compared to animal trials. However, caution should be taken when interpreting data and trying to extrapolate results obtained from Rusitec to in vivo situations due to microbial community differences, lack of absorption and passage rates of the liquid portion and low substrate to rumen volume ratio (Mansfield et al., 1995; Hristov et al., 2012).

In the current study, on average, EN decreased CH₄ production by 24% and 22% when compared to control, for the DAIRY and BEEF diets, respectively, confirming the potential of EN as a CH₄ mitigation option for ruminants. For DAIRY, mitigation efficacy of EN (% CH₄ reduction per 1% nitrate in the diet, assuming full reduction to NH₃) was 13.4% and 11.7% for 1.25% and 2.5% of EN in diet DM, respectively, while efficacies in BEEF were 14.5% and 10.4%. These results are similar to observed by others and demonstrate a reduction in efficacy as nitrate inclusion increases regardless of diet composition (12.2%, Van Zijderveld et al., 2010; 12.5% Lund et al., 2014; 7.3% Lee et al., 2015a; 9.2%, Veneman et al., 2015; 11.4%, Guyader et al., 2017). When evaluating the dose-effect of nitrate on CH₄ production of Holstein steers, Newbold et al. (2014) observed efficiencies ranging from 19.5% when nitrate level was 0.6% of diet DM to 9.6% when nitrate was added at 3%. The decrease in efficacy as nitrate inclusion increases can be explained by an incomplete reduction of nitrate to NH₃ and by the divergence of H₂ from propionogenesis and microbial cell synthesis rather than from methanogenesis. In animals, the absorption of nitrate and NO₂ into the bloodstream before complete reduction to NH₃ could also explain decreased efficacy as discussed by Nolan et al. (2016). However, efficacy of CH₄ mitigation can increase if nitrate intermediates reduce the number of H₂ generating microorganisms or the number of methanogens (Kluber and Conrad, 1998).

In the rumen, nitrate is reduced to NO₂ and then to NH₃ and for every mole of nitrate reduced CH₄ production should decrease by 1 mol. Stoichiometrically, the complete reduction of nitrate to NH₃ consumes 4 mol of H₂, the same number of H₂ molecules necessary for methanogens to reduce CO₂ to CH₄. Thermodynamically, both reactions involved in nitrate reduction to NH₃ are energetically favorable compared to reduction of CO₂ by methanogens (Ungerfeld and Kohn, 2006). Further, intermediates of nitrate reduction such as NO₂ are suggested to have a toxic effect on methanogens (Lee and Beauchemin, 2014), therefore at least part of the CH₄ decrease observed in vitro (Marais et al., 1988; Capelari and Powers, 2017; Guyader et al., 2017) and in vivo (Newbold et al., 2014; Guyader et al., 2015; Lee et al., 2015a) have been linked to NO₂ effect on methanogenic archaea population. Capelari and Powers (2017) observed that nitrate effect as H₂ sink explained, on average, 67.5% of the observed CH₄ decrease (actual CH₄ mitigation in mol/theoretical mitigation potential, assuming complete nitrate reduction to NH₃). Using the same calculations in the current experiment, the effect of nitrate as a H₂ sink would explain, on average, 79% and 61% of actual CH₄ reduction observed in DAIRY and BEEF trials, respectively. Increased NO₂ levels in rumen fluid were observed when EN was fed, regardless of diet. Because samples for nitrate and NO₂ concentration analysis were collected approximately 24 h after feeding (before feed bag exchange), we can speculate higher values occurred immediately after feeding. Lee et al. (2015a) observed a significant increase in rumen NO₂ levels 3 h after feeding EN to beef heifers, but did not detect any NO₂ left in the rumen fluid 6 h after feeding. Guyader et al. (2015) observed elevated rumen dissolved H₂ up to 2 h after feeding, which was suggested to be related to a toxic effect of nitrate or its intermediates on H₂ utilizers such as methanogens, and Asanuma et al. (2015) fed nitrate to goats (6 and 9 g/d) and observed a significant reduction in methanogen copy numbers as estimated by real-time PCR. Monensin does not affect major species of nitrate reducers in the rumen, such as Selenomonas ruminantium (Chen and Wolin, 1979) and Veillonella parvula (Newbold et al., 1993). However, the combination of EN + MON numerically increased NO₂ levels in rumen fluid in DAIRY and tended to increase NO₂ in BEEF trials, suggesting a possible influence of MON on nitrate reduction in the rumen. Studies in plants have reported that nitrate transport is powered by electrochemical potential of protons across the plasma membrane (Ullrich, 1992) with high- and low-affinity transport systems demonstrating saturation kinetics (Wang and Crawford, 1996; Crawford and Glass, 1998). Monensin decreased the nitrate-induced depolarization in plant leaf cells by 85% (García-Sánchez et al., 2000). An effect of MON on nitrate reduction in the rumen would be undesirable and deserves further investigation because accumulation of NO₂ in the rumen is related to decreased diet digestibility (Marais et al., 1988) and methemoglobinemia (Lee and Beauchemin, 2014). Research efforts target reducing NO₂ accumulation in the rumen (Nolan et al., 2016; Yang et al., 2016).

As an ionophore, MON has the ability to form lipid-soluble complexes with cations and mediate their transport across lipid barrier of gram positive bacteria (Russell and Strobel, 1988). Further, MON also inhibits the growth of fungi (Stewart et al., 1987; Stewart and Richardson, 1989) and transiently inhibits the growth of ciliate protozoa (Hino, 1981; Wallace et al., 1981; Guan et al., 2006; Sylvester et al., 2009). The result is a decrease in CH₄ production (Odongo et al., 2007) resulting from an increase in propionate production to the detriment of acetate and butyrate production (Guan et al., 2006), and a protein-sparing mechanism, increasing availability of true protein (Van Nevel and Demeyer, 1977). The present study observed that MON decreased CH₄ output by 19.8% and 39.6% for DAIRY and BEEF trials, respectively. This is consistent with previous in vitro and in vivo studies (Van Nevel and Demeyer 1977; Thornton and Owens 1981; Wallace et al. 1981; Wedegaertner and Johnson 1983; Martin and Macy, 1985; Dong et al., 1999; Jenkins et al., 2003; Capelari and Powers, 2017). No sign of microbial resistance was detected indirectly through the variables analyzed during the current study (data not shown), and CH₄ decrease resulting from MON addition was maintained until day 30. An observed additive effect of EN + MON on CH₄ production suggests that additives that alter the H₂ pool in the rumen through different mechanisms might further reduce CH₄ output, such as that observed when nitrate was combined with lipids (Guyader et al., 2015), sulfate (Van Zijderveld et al., 2010), and saponin (Patra and Yu, 2014).

Treatments did not alter CO₂ production in DAIRY while EN decreased CO₂ production in BEEF. The production of CO₂ in the rumen is mainly associated with the fermentative process, and more specifically acetate and butyrate-producing pathways which generates 2 and 1.5 mol of CO₂ per mole of glucose fermented, respectively (Ungerfeld and Kohn, 2006). Because total DM disappearance did not differ in DAIRY while it was reduced by EN in BEEF, it was expected that CO₂ production would follow this pattern, even though CO₂ concentration in the rumen also respects an equilibrium and can be influenced by other factors such as pH (Allen, 1997). The production of N₂O was detected in this trial supporting previous studies that also observed N₂O production in the rumen (Wang, 2012; de Raphelis-Soissan et al., 2014; Capelari and Powers, 2017). However, while EN increased N₂O in DAIRY trials, no treatment effect was observed in BEEF. Because N₂O generating microorganisms likely occupy a small niche in the rumen, it is possible that under difference physicochemical conditions during BEEF trials (e.g., lower pH), the maintenance of such species in the system was compromised. Despite the increase in N₂O in DAIRY trials when EN was present, the reduction in CH₄ production, a gas produced in significantly greater volume than N₂O, the benefit of the strategy in reducing GHG production was still maintained.

Consistent across the DAIRY and BEEF trials, the addition of EN significantly decreased NH₃ concentration in fermentation media in a dose–response manner although diets were iso-nitrogenous. The effect has been reported in other in vitro (Lin et al., 2011; Guyader et al., 2017) and in vivo (El-Zaiat et al., 2014; Lee et al., 2015a) trials. Guyader et al. (2017), in an attempt to track the fate of H₂ when CH₄ was inhibited with different feed additives, reported concomitant reduction in NH₃ and increased microbial protein synthesis in Rusitec fermenters fed EN. Increased efficiency of microbial protein synthesis could explain the decreased NH₃ observed in the current study as more NH₃ would be incorporated into microbial biomass (Nocek and Russell, 1988). More studies measuring microbial protein synthesis in animals fed nitrate could clarify the mechanism for the observed effect of nitrate on rumen N metabolism. Alternatively, the form of nitrate fed could influence the NH₃ concentration. The nitrate form used in the trials was encapsulated, where nitrate is protected by a matrix that regulates the release rate in rumen fluid (Lee et al., 2017). During the experiment, it was possible to visually identify and count intact capsules after 48 h of incubation (approximately 5% of total number of EN capsules initially added; data not shown). Because treatments not receiving EN were iso-nitrogenous through addition of urea (not encapsulated) part of the effect on NH₃ concentration could be explained by incomplete release of nitrate content. Finally, reduction of nitrate to NH₃ is a slower process than the immediate conversion of urea to NH₃ in the rumen, which is confirmed by the significant levels of nitrate and NO₂ observed in rumen fluid 24 h post feeding EN.

Fermentation Balance

Stoichiometric calculations are helpful in describing alterations in fermentation, but they must be interpreted with caution because manipulating fermentation may invoke reactions not considered in the calculations and fermentation uncoupled from cell growth may occur (Chalupa, 1977). Reductions in metabolic H₂ production were observed when MON was added to both DAIRY and BEEF feeds as a consequence of decreased acetate and butyrate concentrations in support of increased propionate. Acetate and butyrate formations release 4 mol of H₂ per mole of glucose fermented. Propionate formation, on the other hand, requires a net input of H₂ (Ungerfeld and Kohn, 2006). Methane is the primary H₂ sink in the rumen, and as consequence of decreased CH₄ formation when additives were present, utilization and recovery of metabolic H₂ were significantly decreased by the treatments. Even though gaseous H₂ was not measured, it is predicted that a part of the H₂ spared from methanogenesis when EN was fed exit the rumen as gaseous H₂ as previously documented in vivo (Van Zijderveld et al., 2011; Guyader et al., 2015; Olijhoek et al., 2016) and in vitro (Guyader et al., 2017).

In rumen studies, fermentation efficiency can be defined as a measure of the energy present in a fermentable substrate (i.e., hexoses and AA) that is recovered in usable forms, such as VFA and microbial protein, to the host animal. From an energy stand point, production of propionate from hexose fermentation is more efficient than production of other VFA because less energy is lost as H₂ and C. Therefore, shifts in rumen fermentation caused when MON in both DAIRY and BEEF diets were fed resulted in significant increases in overall fermentation efficiency due to greater propionate concentration. However, it should be stressed that the profile of VFA produced in the rumen also has consequences on the host animal metabolism. Excess propionate can lead to undesirable consequences such as reduction in voluntary feed intake (Oba and Allen, 2003; Allen et al., 2009) and low milk fat content and fat syndrome in lambs (Ørskov and Ryle, 1990).

Nutrient Digestion

The disappearance of DM and other chemical fractions for both the DAIRY and BEEF trials were within expected values for relatively good quality diet substrates and were similar to those observed by Dong et al. (1999) and Avila-Stagno et al. (2014) when testing the effect of additives and different levels of glycerol in fermenters fed high-quality hay and corn silage–based diets. Nitrate decreased CP and starch disappearance in DAIRY trials and NDF and starch disappearance in BEEF trials. Results suggest that nitrate or a reduction product temporarily present in the media can affect utilization of some chemical components of the diets (through solubilization in ruminal fluid and/or degradation by microorganisms enzymes). Marais et al. (1988) reported suppression of in vitro digestion of diet DM when NO₂ was present, but not when nitrate was added, thereby proposing that NO₂ was the primary factor affecting digestibility in the rumen. In the same study, NO₂ was shown to reduce solubilization of structural components of the diet, such as cellulose and hemicellulose, confirmed by a reduction in the cellulolytic and xylanolytic microbial populations (Ruminococcus flavefaciens, Ruminococcus albus, and Butyrivibrio fibrisolvens) with concomitant reduction in cellulase and xylanase activity. Reduction of ADF digestibility has been reported in beef heifers fed increasing levels of nitrate on a barley silage and ground corn–based diet despite an increase in total-tract DM digestibility, likely due to increased digestibility of starch (Lee et al., 2015a). Asanuma et al. (2015) reported increased starch-utilizing organisms when goats were fed 9 g/d nitrate. Encapsulation of nitrate, especially in BEEF diets, may limit the synchronization between readily available carbohydrates (nonforage carbohydrates [NFC] in DAIRY = 46.6% and BEEF = 64.4% of diet DM; Table 1) and N sources to rumen starch utilizers, therefore reducing the growth of these microorganisms. In addition, the Rusitec is a limited system in terms of hosting the complete rumen microbiome (Hristov et al., 2012; Martínez-Fernández et al., 2015) thus preventing the observation of potential benefits on specific groups of microorganisms.

The effect of MON on nutrient digestibility has been inconsistent in both in vitro and in vivo studies. In the current study, MON did not affect nutrient and DM disappearance in DAIRY trials but decreased NDF disappearance in BEEF trials by 10% when compared to control. Wallace et al. (1981) reported a reduction in cellulose and hemicellulose digestion when the Rusitec was fed 10 mg/L of MON daily, a dose more than two times that tested in the current trials. Dong et al. (1999) found a reduction in cellulose (30%) and hemicellulose (21%) digestion due to MON when the Rusitec received 20.5 µM/d of MON. In various in vitro and in vivo studies, MON has depressed fiber digestion (Poos et al., 1979; Mir, 1989), affected fiber digestion minimally (Ricke et al., 1984; Faulkner et al., 1985) or not at all (Kone and Galyean, 1990; Duff et al., 1995), and even increased fiber digestibility (Wedegaertner and Johnson, 1983). Potential negative effects of MON on ruminal fiber degradation are likely related to the sensitivity of the three primary cellulolytic species, R. albus, R. flavefaciens, and B. fibrisolvens, to MON (Chen and Wolin, 1979). Factors such as the dosage used, diet chemical composition, and retention time of digesta may account for variability in results.

In summary, the combination of EN and MON can serve as a potential mitigation strategy for enteric CH₄ abatement. Caution should be taken with dosage because the digestion of feed components was impaired when additives were combined at higher dosages independent of diet offered. Animal studies to evaluate the combination of nitrate and MON on rumen and metabolic parameters as well as animal performance and health are needed before recommending this dietary strategy under field conditions.