If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
The objective of the present study was to investigate the effect of individual and combined use of dietary fat, nitrate, and 3-nitrooxypropanol (3-NOP) on dairy cows' enteric methane (CH4) emission and production performance. Twenty-four primiparous and 24 multiparous Danish Holstein cows (111 ± 44.6 d in milk; mean ± standard deviation) were included in an incomplete 8 × 8 Latin square design with six 21-d periods. Dietary treatments were organized in a 2 × 2 × 2 factorial arrangement aiming for 2 levels of FAT (30 or 63 g of crude fat/kg of dry matter [DM]; LF or HF, respectively), 2 levels of NITRATE (0 or 10 g of nitrate/kg of DM; UREA or NIT, respectively), and 2 levels of 3-NOP (0 or 80 mg/kg DM; BLANK or NOP, respectively). Treatments were included in ad libitum-fed partial mixed rations in bins that automatically measured feed intake and eating behavior. Additional concentrate was offered as bait in GreenFeed units used for measurement of gas emission. For total DM intake (DMI), a FAT × NITRATE interaction showed that DMI, across parities and levels of 3-NOP, was unaffected by separate fat supplementation, but reduced by nitrate with 4.6% and synergistically decreased (significant 2-way interaction) with 13.0% when fat and nitrate were combined. Additionally, 3-NOP decreased DMI by 13.4% and the combination of 3-NOP with fat and nitrate decreased DMI in an additive way (no significant 3-way interaction). The decreasing effects on DMI were more pronounced in multiparous cows than in primiparous cows. For treatments with largest reductions in DMI, eating behavior was altered toward more frequent, but smaller meals, a slower eating rate and increased attempts to visit unassigned feed bins. Energy-corrected milk (ECM) yield increased by 6.3% with fat supplementation, whereas ECM yield did not differ among diets including nitrate (FAT × NITRATE interaction). Cows supplemented with 3-NOP had 9.0% lower ECM yield than cows fed no 3-NOP. Based on three 2-way interactions including FAT, NITRATE, and 3-NOP, the combined use of the additives resulted in antagonistic effects on CH4 reduction. A 6% to 7% reduction in CH4 yield (CH4/kg of DMI) could be ascribed to the effect of fat, a 12% to 13% reduction could be ascribed to the effect of nitrate and an 18% to 23% reduction could be ascribed to the effect of 3-NOP. Hence, no combinations of additives resulted in CH4 yield-reductions that were greater than what was obtained by separate supplementation of the most potent additive within the combination. The CH4 yield reduction potential of additives was similar between parities. Increased apparent total-tract digestibility of organic matter (OM) in cows fed combinations including nitrate or 3-NOP was a result of a NITRATE × 3-NOP interaction. Apparent total-tract digestibility of OM was also increased by fat supplementation. These increases reflected observed decreases in DMI. In conclusion, combined use of fat, nitrate, and 3-NOP in all combinations did not result in CH4 reductions that were greater than separate supplementation of the most potent additive within the combination (3-NOP > nitrate > fat). Additionally, separate supplementation of some additives and combined use of all additives reduced DMI.
Globally, mitigation of enteric CH4 is of high priority because it has high effect on global warming and a short life span compared with other greenhouse gases from the dairy cattle sector. Strategies to reduce enteric emission include, among others, changes in diet formulation and the use of feed additives such as fat, nitrate, and 3-nitrooxypropanol (3-NOP).
Dietary fats have a moderate potential for CH4 reduction (
), and the potential of CH4 reduction has been evaluated in different reviews and meta-analyses. Across these, fats have a potential of achieving a 10% to 15% decline in CH4 emissions under practical conditions, depending on fat source and inclusion level (
The effect of dietary fats on methane emissions, and its other effects on digestibility, rumen fermentation and lactation performance in cattle: A meta-analysis.
. These authors claimed a CH4 yield (g CH4/kg DMI) reduction potential of nitrate of 22% and 16% in dairy cows, respectively. The inhibitor 3-NOP has in several studies been supplemented to dairy cows resulting in reductions in CH4 emissions (
3-Nitrooxypropanol decreases methane emissions and increases hydrogen emissions of early lactation dairy cows, with associated changes in nutrient digestibility and energy metabolism.
found a CH4 yield reduction potential in dairy cows of 31%, and any variation in the efficacy of 3-NOP across studies to some extent could be explained by dose of 3-NOP and diet composition (
The effect of dietary fats on methane emissions, and its other effects on digestibility, rumen fermentation and lactation performance in cattle: A meta-analysis.
The effect of dietary fats on methane emissions, and its other effects on digestibility, rumen fermentation and lactation performance in cattle: A meta-analysis.
, showed that feed intake and production performance were generally not affected by dietary inclusion of these additives, however more recent available literature have in some cases indicated reductions in DMI and milk production from cows supplemented with 3-NOP (
Fats, nitrate, and 3-NOP have different and unique modes of action for CH4 reduction. Fats replace fermentable OM in cows' diets with nonfermentable OM, resulting in less available substrate (H2) for methanogenesis. This is quantitatively the most important effect of fat for CH4 reduction. Furthermore, inclusion of fatty acids, depending on chain length and degree of saturation, inhibit the populations of protozoa and methanogens in the rumen (
suggested that the amount of H2 used for biohydrogenation is relatively small. Nitrate is a much more effective H2-sink, redirecting H2 into reduction of nitrate (NO3−) to nitrite (NO2−) and further to ammonium (NH4+). Both reducing reactions are naturally occurring in the rumen being energetically favorable for microorganisms (
The role of thermodynamics in the control of ruminal fermentation.
in: Sejrsen K. Hvelplund T. Nielsen M.O. Ruminant Physiology: Digestion, Metabolism and Impact of Nutrition on Gene Expression, Immunology and Stress. Wageningen Academic Publishers,
2006: 55-85
). The chemical inhibitor 3-NOP has a very specific mode of action, due to its blocking of methyl coenzyme M-reductase, a key enzyme in the methanogenesis (
Despite that dietary fats, alternative H2-sinks such as nitrate, and chemical inhibitors such as 3-NOP have separately been widely investigated and are suggested as potential dietary interventions for CH4 reduction by several reviews (
), their combined effects have only been investigated to a limited extent. Due to their diverse modes of action, it appears possible that combinations of these potent additives can achieve even greater CH4 reductions, than what can be achieved by separate supplementation. The suggested mode of action for such combination of additives assumes that fat is being energy source for the animal while reducing fermentable matter in the rumen, whereas nitrate, in the rumen, can limit the H2 availability for the methanogens, while 3-NOP directly inhibits the methanogenesis. Few in vivo studies have combined additives and the results of combining fat and nitrate indicated a synergistic effect on CH4 reduction in a study by
, where the effect of the combined treatment was greater than the individual effects. However, an additive effect of nitrate and fat and 3-NOP and fat was found by
Combined effects of 3-nitrooxypropanol and canola oil supplementation on methane emissions, rumen fermentation and biohydrogenation, and total tract digestibility in beef cattle.
, respectively, meaning that the combined effect equaled the sum of the 2 individual effects. To date, no studies have investigated the CH4 reduction potential and the response in dairy cows' performance, when combinations of fat, nitrate, and 3-NOP are fed to dairy cows, which therefore was the aim of the present study. Due to the well-documented effect on CH4 reduction when supplemented separately and considering the additives' unique modes of action, we hypothesized that the separate CH4 mitigating effects of fat, nitrate, and 3-NOP would be additive when combined, and that supplementing combinations of additives would not potentially compromise dairy cows' feed intake and milk production.
MATERIALS AND METHODS
The experiment was conducted from September 2020 to January 2021 and complied with the guidelines set out by the Danish Ministry of Food, Agriculture and Fisheries, The Danish Veterinary and Food Administration under act no. 474 (May 15, 2014) and executive order no. 2028 (December 14, 2020), and under consideration of the ARRIVE guidelines (
Forty-eight Danish Holstein cows were enrolled in an incomplete 8 × 8 Latin square (LS) design with 6 periods of 21 d each (126 d in total) at the experimental facilities of Aarhus University, AU Viborg–Research Centre Foulum, Denmark. In each period, the first 14 d were used for adaptation and the last 7 d were used as sampling period. Cows were assigned to 6 blocks of 8 cows each, according to parity (24 primiparous, 24 multiparous) and DIM. Primiparous cows were 114 ± 45.7 (mean ± SD) DIM, BW of 595 ± 45.8 kg, and milk production of 27.3 ± 5.53 kg and multiparous cows (13 second parity, 9 third parity, and 2 fourth parity) were 109 ± 44.3 DIM, BW of 675 ± 61.9 kg, and milk production of 39.3 ± 5.76 kg at the beginning of the experiment. Latin squares were first designed as balanced 8 × 8 LS, whereafter the last 2 periods were discarded, thus cows only received 6 of 8 dietary treatments during the experiment. Six different LS were subsequently used to balance out first order carry-over effects between treatments, so that possible carry-over effects were as balanced between treatments and changes in periods as possible. Blocks of cows were then randomly assigned to a given LS and within LS, cows were at the start of the experiment randomly assigned to receive one of 8 dietary treatments.
Cows were grouped into 2 pens according to parity (primiparous or multiparous) with 24 cows in each pen. Cows were loose-housed in pens with concrete floor and cubicles with mattresses and sawdust as bedding material. Throughout the trial, each cow had access to its own individual feeding trough to receive the assigned experimental diet. All cows in each pen had access to 4 water troughs. The water troughs were refilled with approximately 36 L between visits. Individual feed intake, water intake and feeding/drinking behavior was registered automatically with the Insentec RIC system (Insentec, Marknesse, the Netherlands). Cows were milked twice daily at 0530 and 1630 h in a 2 × 12 parallel milking parlor (SAC A/S, S. A. Christensen and Co.). Milk yield was recorded automatically at every milking as was BW of individual cows when they exited the milking parlor. Lights were on between 0500 h and 2300 h and reduced during nighttime.
Diets and Feeding
Dietary treatments were organized in a 2 × 2 × 2 factorial arrangement aiming for 2 levels of fat (FAT: 30 or 63 g of crude fat/kg of DM; LF or HF, respectively), 2 levels of dietary nitrate (Silvair; Cargill Inc., the Netherlands; 5Ca[NO3]2·NH4NO3·10H2O NITRATE: 0 or 10 g of nitrate/kg of DM; UREA or NIT, respectively), and 2 levels of 3-NOP (Bovaer; DSM Nutritional Products, Switzerland; 3-NOP: 0 or 80 mg/kg DM; BLANK or NOP, respectively). The fat source was whole cracked rapeseeds, cracked on a Skiold roller (Sæby, Denmark) with roller distance <0.15 mm. Rapeseed was included in the HF diet by balancing the amount of kilograms of nonfat originating from rapeseed products (rapeseed meal and cracked rapeseed) between the 2 levels. Because the nitrate source contained nitrogen, the diets were made iso-nitrogenous with inclusion of urea in the diets without nitrate (UREA). Cows were adapted to nitrate by inclusion of 50% of the intended dose on the first day of exposure to NIT diets. The 3-NOP was included in a carrier substance (silicon dioxide + 1,2 propanediol) including 10.5% of the active compound of 3-NOP. In the diets without 3-NOP, 100% carrier compound was added as a placebo. Premixes with minerals, vitamins and sodium sulfate were composited including either 3-NOP or merely carrier substance before added to the individual partial mixed ration (PMR). Forage constituted 53% of the total PMR DM and forage consisted of 48% grass-clover silage and 52% corn silage on a DM basis. Concentrates constituted 47% of total PMR DM and was composed of NaOH treated wheat, rapeseed meal, dried beet pulp, soybean meal, and whole cracked rapeseed (for HF only). Diets were formulated by using the Nordic Feed Evaluation System (NorFor;
). Chemical composition of individual feed ingredients is presented in Table 1. Partial mixed rations were fed daily at 1030 and 2030 h for ad libitum intake aiming for 7% residues but minimum 3 kg (refilled at 0515 h if needed). Residues were removed daily before feeding at 1030 h. Additional concentrate (commercially produced at DLG A.m.b.a., Denmark; see composition in Table 1) was offered as feed bait in the GreenFeed system (maximum allowance of 1.3 kg of DM per day). Planned ingredient composition and analyzed chemical composition of the 8 diets are presented in Table 2.
OMD: in vivo OM digestibility in % calculated for grass-clover silage as 4.10 + 0.959 × in vitro OM digestibility, for corn silage as 6.73 + 0.950 × in vitro OM digestibility, and for concentrates as 5.38 + 0.867 × enzymatic in vitro OM digestibility (Åkerlind et al., 2011).
%
82.4 ± 0.48
70.8 ± 0.09
76.6 ± 0.57
91.6
78.4
86.3
82.8
90.8
83.8 ± 1.04
NO3
0.04 ± 0.006
0.05 ± 0.017
0.23 ± 0.060
NA
NA
NA
NA
NA
NA
1 Fermentation characteristics for primary growth grass-clover silage, secondary growth grass-clover silage, and corn silage, respectively: pH = 4.31 ± 0.07, 4.04 ± 0.07, and 3.66 ± 0.05; l-lactate (g/kg DM) = 28.5 ± 0.87, 34.0 ± 4.55, and 38.4 ± 1.45 (l-lactate constitute about half of total lactate;
6 OMD: in vivo OM digestibility in % calculated for grass-clover silage as 4.10 + 0.959 × in vitro OM digestibility, for corn silage as 6.73 + 0.950 × in vitro OM digestibility, and for concentrates as 5.38 + 0.867 × enzymatic in vitro OM digestibility (
The PMR diets were 2 × 2 × 2 factorially arranged aiming for 2 levels of fat (FAT: 30 or 63 g of crude fat/kg DM; LF or HF, respectively), 2 levels of dietary nitrate (NITRATE: 0 or 10 g of nitrate/kg DM; UREA or NIT, respectively), and 2 levels of 3-nitrooxypropanol (3-NOP: 0 or 80 mg/kg DM; BLANK or NOP, respectively). GreenFeed bait is not included in the analyzed chemical composition, but included in the calculated energy and protein value with 1.1 kg of DM per day.
Vilofoss Komix Type 3, analyzed or declared macro mineral composition (g/kg DM): Ca = 147, Mg = 141, Na = 116, S = 1. Added vitamins and microminerals (per kg of DM): vitamin A = 600,000.10 IU, vitamin D3 = 190,000.10 IU, vitamin E = 4,000 IU, Mn = 4,000 mg, Cu = 1,500 mg, Zn = 4,500 mg, I = 225, Co = 25 mg, Se = 50 mg.
Protein balance in the rumen at 20 kg of DMI per day, calculated according to NorFor (Åkerlind and Volden, 2011).
31
31
31
31
30
30
29
29
1 The PMR diets were 2 × 2 × 2 factorially arranged aiming for 2 levels of fat (FAT: 30 or 63 g of crude fat/kg DM; LF or HF, respectively), 2 levels of dietary nitrate (NITRATE: 0 or 10 g of nitrate/kg DM; UREA or NIT, respectively), and 2 levels of 3-nitrooxypropanol (3-NOP: 0 or 80 mg/kg DM; BLANK or NOP, respectively). GreenFeed bait is not included in the analyzed chemical composition, but included in the calculated energy and protein value with 1.1 kg of DM per day.
2 Calcium ammonium nitrate; 5Ca(NO3)2·NH4NO3·10H2O; 75% NO3 in DM (Cargill Inc., the Netherlands).
4 Vilofoss Komix Type 3, analyzed or declared macro mineral composition (g/kg DM): Ca = 147, Mg = 141, Na = 116, S = 1. Added vitamins and microminerals (per kg of DM): vitamin A = 600,000.10 IU, vitamin D3 = 190,000.10 IU, vitamin E = 4,000 IU, Mn = 4,000 mg, Cu = 1,500 mg, Zn = 4,500 mg, I = 225, Co = 25 mg, Se = 50 mg.
5 Vilofoss Suplex ADE, analyzed or declared macromineral composition (g/kg DM): Ca = 139, Mg = 91, Na = 95. Added vitamins and microminerals (per kg of DM): vitamin A = 900,000 IU, vitamin D3 = 200,000 IU, vitamin E = 2,000 IU, Se = 50 mg.
6 iNDF: indigestible NDF.
7 Net energy for lactation at 20 kg of DMI per day, calculated according to NorFor (
Feed ingredients were sampled on d 20 of each experimental period, whereas each PMR was sampled on d 20 and 21 of each experimental period. A subsample of each sample was used for DM determination. The remaining samples of each PMR were pooled across days, within period. Subsamples of forage samples and PMR representing each period were pooled for period 1 and 2, 3 and 4, and 5 and 6, respectively. Samples of concentrates (soybean meal, rapeseed cake, whole rapeseeds, wheat, and dried beet pulp) were pooled by ingredient across all periods, whereas GreenFeed bait was pooled to represent each of 3 deliveries of the GreenFeed concentrate. All samples were stored at −20°C until pooling and later chemical analysis. Milk samples were taken over the last 3 d (6 consecutive milkings) in the last week of each experimental period. These were stored at 5°C with bronopol until analysis of milk composition. Samples of feces (500 mL) of each cow were collected on d 20 (1300 h) and 21 (0700 h) in the 3 last periods (4–6). Individual fecal samples from d 20 were stored at 5°C until they were pooled with the sample from d 21. Feces consistency of the pooled samples (n = 144) was evaluated by visual scoring on a 5-point scale (1–5; 0.5 point intervals, 1 is loose and 5 is firm). On d 20 in the 4 last periods (3–6), a blood sample was taken from the tail vein, 3 h after returning from morning milking. The samples were taken in sodium heparin Vacutainers and stored on ice until analysis on whole blood.
Analytical Procedures
Subsamples of PMR representing each period were analyzed at an external laboratory for concentration of 3-NOP (DSM Nutritional Products, Kaiseraugst, Switzerland). Dry matter concentration of feed ingredients and PMR samples were determined by drying at 60°C for 48 h (
). The DM concentration of fecal samples was determined by drying for 72 h at 60°C. For chemical analyses, samples were freeze-dried and subsequently ground (Ultra Centrifugal Mill ZM 200, Verder Scientific, Hann, Germany) by using a 1.0 mm screen. Subsamples for starch determination were grinded by using a 0.5 mm screen. Ash concentration was determined in duplicates by combustion at 525°C for 6 h. In feed samples, total nitrogen (N) concentration was determined according to the Dumas method (
) by using a Vario MAX CN apparatus (Elementar Analysensysteme GmbH, Hanau, Germany). Concentration of CP was calculated as N × 6.25. Crude fat (CF) concentration was analyzed in feed samples after hydrolysis with hydrochloric acid (HCl) by using a Hydrotherm HT6 apparatus (C. Gerhardt GmbH and Co. KG) followed by Soxhlet extraction with petroleum ether using a Soxtherm SOX 416 apparatus (C. Gerhardt GmbH and Co. KG; Eurofins Steins Laboratories, Vejen, Denmark). Starch was enzymatically determined in feed samples by using a heat-stable α-amylase and amyloglucosidase, whereafter starch concentration was measured as liberated glucose (YSI 2900D Biochemistry Analyzer, YSI Inc.;
). Results were reported as ash-free NDF and iNDF. Indigestible NDF was determined as residual NDF, where samples were incubated in duplicates in F57 ANKOM bags in the rumen of 2 out of 3 used rumen-cannulated nonlactating cows for 288 h (12 d;
), before NDF analysis. Samples from the same period were incubated in the same 2 cows. The nonlactating cows were fed a standard ration at maintenance level as described by
by anaerobic incubation in rumen fluid for 48 h. The rumen fluid was collected from 3 nonlactating rumen-cannulated cows (fed as described for iNDF analysis) and filtered through 2 layers of cheesecloth before incubation. After rumen fluid incubation, the sample was incubated in a solution of HCl and pepsin for additional 48 h. Enzymatic OM digestibility of concentrates was determined by dissolving the sample in HCl and pepsin following incubation with cellulolytic enzymes (
). In vivo OM digestibility was then calculated for grass-clover silage as 4.10 + 0.959 × in vitro OM digestibility, for corn silage as 6.73 + 0.950 × in vitro OM digestibility, and for concentrates as 5.38 + 0.867 × enzymatic in vitro OM digestibility (
To analyze fermentation products and nitrate concentration in silages, water extracts were prepared by blending of 100 g of fresh silage with 1,000 mL of deionized water. The blends were subsequently centrifuged and pH, VFA, l-lactate, and ammonium (NH4) concentrations were measured in the metaphosphoric acid stabilized supernatant. Concentrations of VFA were measured by using gas chromatography (Trace 1310, Thermo Scientific). The gas chromatograph used split/splitless injector at 225°C and a flame ionization detector at 250°C with a 30 m × 0.53 mm × 1 µm HP-FFAP column using helium as carrier gas (0.3405 atm). Glucose and l-lactate was analyzed using membrane-immobilized substrate-specific oxidases (YSI 2900D Biochemistry Analyzer; YSI Inc.), whereas NH4 was determined in phosphate buffered samples by using a Cobas Mira Plus (Roche Diagnostics Systems) with a Randox Ammonia Kit-AM1015. Nitrate was determined by using spectrophotometry as outlined by
Total hemoglobin (tHb), methemoglobin (MetHb), and hematocrit (Hct) in whole blood samples were analyzed immediately after sampling by oximetry by using an acid base laboratory analyzer (Radiometer ABL90 Flex). Concentration of milk protein, fat, and lactose monohydrate were analyzed at an external laboratory (Eurofins Steins Laboratorium, Vejen, Denmark) using a Milkoscan 7 RM infrared analyzer (Foss, Hillerød, Denmark).
Gas Measurements
Measurements of CH4, CO2, and H2 were obtained from individual animals by using the GreenFeed system (C-Lock Inc., Rapid City, SD). Four GreenFeed units were used in the experiment; 2 units in each pen with 24 cows. Thus, each cow had access to 2 GreenFeed units. Concentrations of CH4, CO2, and H2 in exhaust air were measured by nondispersive near-infrared sensors (NDIR) and the individually measured concentrations were corrected for gas concentrations in ambient air and temperature and pressure according to the ideal gas law. Gas concentrations were subsequently converted to a daily flux of emission (g/d) by multiplying with the measured airflow rate. The NDIR sensors were calibrated daily by using a zero-gas mixture serving as baseline (N2 with 20% O2) and a span-gas mixture (N2 with 10 ppm H2, 500 ppm CH4, 5,000 ppm CO2, and 21% O2) provided by C-Lock Inc. Accuracy of measurements were monitored before and after the experiment and before each sampling period by releasing a gravimetrically determined amount of CO2 into the system. Average CO2 recovery for all units was 99.8% ± 1.83% (mean ± SD). If recovery tests were repeatedly biased, the airflow coefficient used for calculations of flux was adjusted by C-Lock Inc. Procedure for calibrations and CO2 recoveries are presented in a slightly modified form by
. Cows were allowed a maximum of 5 daily visits with an interval between visits of minimum 4 h and a maximum of 6 feed drops per visit (of 44.4 ± 3.30 g [mean ± SD] of each drop; see Table 1 for nutrient composition), fed with an interval of 40 s between drops. Certain criteria regarding head proximity and visit duration were set by C-Lock Inc., which filtered out unreliable gas measurements. This resulted in 30 ± 7.2 (mean ± SD) approved visits per animal per sampling week (d 15 to 21 of each period), subsequently to be included in the statistical analysis. Individual gas emission observations were averaged over the last 7 d of each period.
Calculations
In PMR samples, NEL20, AA absorbed in the small intestine (AAT20), and protein balance in the rumen (PBV20) at 20 kg DMI/d, were calculated according to NorFor (
Feed intake of PMR and GreenFeed concentrate was averaged over the last 7 d in each period (sampling week). Intake of DM was calculated based on feed intake in the sampling week and an averaged DM concentration measured over 2 of the 7 last days in each period.
Intake behavior (feed and water) was averaged over the last 7 d in each period. For eating and drinking behavior data, a new meal, or a new drinking bout was initiated if the time of termination of one visit to initiation of another visit exceeded 30 min for feed and 4 min for water (
). Total eating and drinking duration corresponded to the time the gate to the respective bin was open, whereas meal duration corresponded to the time from initiation to termination of a meal given the criteria above. Efficient eating time was calculated as the share of total eating duration out of total meal duration.
Individual BW was averaged per day and BW change over the experimental periods was calculated by linear regression by using day in period as fixed effect.
Milk yield data were averaged over the last 7 d in each period and milk composition data were obtained from yield-weighted averages of fat, protein and lactose monohydrate concentration in milk samples taken over the last 3 d in each period. Energy-corrected milk yield (3.14 MJ/kg) was calculated by using the formula ECM = milk yield (kg) × [(38.3 × fat (g/kg) + 24.2 × protein (g/kg) + 15.71 × lactose (g/kg) + 20.7)/3,140] (
For calculation of apparent total-tract digestibility of DM, OM, and NDF, concentration of iNDF in PMR was averaged for LF diets whereafter this value was subsequently corrected by a factor of 0.965 to represent the diluting effect of fat in HF diets. This procedure eliminated any possible analytical bias of iNDF determination. Concentrations of iNDF in diets, GreenFeed bait, and feces were used to calculate apparent total-tract digestibility.
Statistical Analysis
Feed intake, eating behavior, milk production, gas emission, BW change, and digestibility data were included in the statistical analysis conducted in R 4.0.2 (
). The effect of treatment on the response variables were analyzed by using the following model fitted with REML and the lmer function from the lme4 and lmeTest packages (
where Yijklpc is the dependent response variable; µ is the overall mean; A is the fixed effect of FAT (i = LF, HF); B is the fixed effect of NITRATE (j = UREA, NIT); C is the fixed effect of 3-NOP (k = BLANK, NOP); D is the fixed effect of parity (l = primiparous, multiparous); ABij, ACik, ADil, BCjk, BDjl and CDkl and ABCijk, ABDijl, BCDjkl, and ACDikl and ABCDijkl are the possible 2-, 3-, and 4-way interactions between FAT, NITRATE, 3-NOP, and parity, respectively; P is the fixed effect of period (p = 1 to 6); δ is the random effect of cow (c = 1 to 48); and εijklpc is the random residual error assumed to be normally distributed and independent with constant variance. Ten observations were discarded from the statistical analysis due to animal related issues or technical problems with equipment. Hence, 278 observations were included in the statistical analysis. Estimated marginal means (EMM) and standard error of means were reported by using the emmeans package (
) and P-values were computed by using a type 2 ANOVA. Statistical significances were declared at P ≤ 0.05. Tendencies were declared when P-values 0.05 < P ≤ 0.10. When an interaction was significant, EMM comparisons were based on Tukey's multiple comparison test and pairwise comparisons. In the tables, EMM for the 4-way interactions were presented, whereas EMM for the main effects or interactions across factors were presented in the text. Normality of residuals and homogeneity of variance were tested by using qqplots and residual plots, respectively. In case of heterogeneous variance, data were log-transformed (natural log; ln) to obtain homogeneity of variance. For such data (eating behavior, feed bin visit attempts and drinking behavior), the EMM from the model without log-transformation were presented in the tables along with the P-values associated with the log-transformed data.
RESULTS
Diet Composition
Analyzed chemical compositions of experimental diets are reported in Table 2. Crude fat concentration ranged from 26 to 30 g/kg DM in LF diets and 60 to 67 g/kg DM in HF diets. Concentration of ash, CP, starch, NDF, and iNDF was slightly lower in HF diets compared with LF diets due to a dilution caused by fat addition in HF diets. The addition of fat increased calculated energy concentration (NEL20) of HF diets compared with LF diets (Table 2). Average 3-NOP recovery in NOP diets was 95% ± 9.4% (mean ± SD) of target concentration (84 mg 3-NOP/kg of DM). This resulted in an actual dose of 76 ± 2.4 mg 3-NOP/kg of DM (mean ± SD) when corrected for intake of GreenFeed bait.
Feed Intake and Eating Behavior
Dry matter intake of PMR was affected by dietary treatments (Table 3) and total DMI of PMR varied across treatments from 15.9 to 20.2 kg/d for primiparous cows and from 16.2 to 24.3 kg/d for multiparous cows. A significant FAT × NITRATE interaction and a tendency for a FAT × 3-NOP interaction showed that PMR DMI was unaffected by fat addition alone, however when fat was combined with nitrate (PFAT × NITRATE < 0.01), or 3-NOP (PFAT × 3-NOP = 0.08) DMI of PMR was more reduced than when nitrate or 3-NOP were fed alone. Cows fed HF, NIT, or NOP had greater DMI from GreenFeed bait (all P < 0.01), than cows fed LF, UREA, or BLANK, respectively. Though the FAT × NITRATE interaction (P < 0.01) remained significant for total DMI and showed, across parities, that total DMI was similar in LF-UREA and HF-UREA diets (21.4 vs. 21.3 kg; P = 0.96), reduced by 4.6% in LF-NIT diets compared with LF-UREA diets (20.4 vs. 21.4 kg; P = 0.01), and even more reduced (−13.0%) in HF-NIT diets compared with LF-UREA diets (18.7 vs. 21.4 kg; P < 0.01). Total DMI was lower in NOP diets compared with BLANK diets (19.0 vs. 21.9 kg; −13.4%; P < 0.01). Additionally, FAT also interacted with parity (PParity × FAT = 0.02), as multiparous cows on HF reduced DMI compared with multiparous cows on LF, whereas primiparous cows were unaffected. The effects of NITRATE and 3-NOP were present in both parities, but more pronounced in multiparous cows, compared with primiparous cows (PParity × NITRATE = 0.03,PParity × 3-NOP < 0.01). Effects on DMI were likewise reflected in total intake of OM, CP, NDF, ST, and CF. A FAT × NITRATE interaction (P < 0.01) was similarly observed for BW change during the experimental periods, where cows on LF-UREA, HF-UREA, and LF-NIT had similar BW change (479, 142, and 363 g/d; all P > 0.10, but cows on HF-NIT had significantly lower BW change than LF-UREA (−697 vs. 479 g/d; P < 0.01). Additionally, across parities, cows on NOP had lower BW change than cows on BLANK (−320 vs. 463 g/d; P < 0.01).
Table 3Intake and BW change of dairy cows fed the 8 partial mixed ration (PMR) diets
The PMR diets were 2 × 2 × 2 factorially arranged aiming for 2 levels of fat (FAT: 30 or 63 g of crude fat/kg DM; LF or HF, respectively), 2 levels of dietary nitrate (NITRATE: 0 or 10 g of nitrate/kg DM; UREA or NIT, respectively), and 2 levels of 3-nitrooxypropanol (3-NOP: 0 or 80 mg/kg DM; BLANK or NOP, respectively).
Indicates significant interaction (P < 0.05) to parity group (parity; primiparous, multiparous).
0.01
0.17
Multiparous
0.70
0.58
0.66
0.61
1.62
1.19
1.42
1.14
BW change, g/d
Primiparous
823
106
804
161
660
−36.4
−16.8
−376
244.9
<0.01
<0.01
<0.01
<0.01
<0.01
0.23
0.56
0.50
Multiparous
618
370
710
−222
565
−622
−457
−1,937
1 The PMR diets were 2 × 2 × 2 factorially arranged aiming for 2 levels of fat (FAT: 30 or 63 g of crude fat/kg DM; LF or HF, respectively), 2 levels of dietary nitrate (NITRATE: 0 or 10 g of nitrate/kg DM; UREA or NIT, respectively), and 2 levels of 3-nitrooxypropanol (3-NOP: 0 or 80 mg/kg DM; BLANK or NOP, respectively).
2 Standard error of estimated marginal mean. The lowest value within the variable is reported as SEM.
* Indicates significant interaction (P < 0.05) to parity group (parity; primiparous, multiparous).
For total eating duration (min/d) a FAT × NITRATE interaction (P = 0.03) was observed and for total meal duration (min/d) a NITRATE × 3-NOP interaction (P = 0.02) was observed (Table 4). Both interactions showed that eating and meal duration was unaffected by NIT alone, but decreased even more when NIT was combined with HF or NOP, respectively, than when HF or NOP was fed separately. Relating total eating duration to meal duration (efficient eating time), a NITRATE × 3-NOP interaction (P = 0.04) showed that efficient eating time was greater for cows on UREA-NOP diets than for cows on UREA-BLANK diets (55.9% vs. 48.0%; P < 0.01), but not when NOP was combined with NIT or when NIT was fed alone. Eating rate (kg DM/min) was significantly greater in HF cows compared with LF cows (0.126 vs. 0.121 kg DM/min; PFAT < 0.01). In contrast, eating rate was lower for cows on NOP compared with BLANK (0.122 vs. 0.126 kg DM/min; P3-NOP < 0.01) and tended to be lower for cows on NIT compared with UREA (0.123 vs. 0.125 kg DM/min; PNITRATE = 0.06).
Table 4Eating behavior and visit attempts to feed bins of dairy cows fed the 8 partial mixed ration (PMR) diets
The PMR diets were 2 × 2 × 2 factorially arranged aiming for 2 levels of fat (FAT: 30 or 63 g of crude fat/kg DM; LF or HF, respectively), 2 levels of dietary nitrate (NITRATE: 0 or 10 g of nitrate/kg DM; UREA or NIT, respectively), and 2 levels of 3-nitrooxypropanol (3-NOP: 0 or 80 mg/kg DM; BLANK or NOP, respectively).
Means and SEM are from model without log-transformation, but P-values are from log-transformed model.
Primiparous
13.4
19.4
16.8
22.3
18.5
22.1
21.3
26.0
2.64
0.57
<0.01
0.02
<0.01
0.83
0.36
0.64
0.40
Multiparous
13.5
16.1
17.4
17.7
13.7
20.7
18.8
26.9
1 The PMR diets were 2 × 2 × 2 factorially arranged aiming for 2 levels of fat (FAT: 30 or 63 g of crude fat/kg DM; LF or HF, respectively), 2 levels of dietary nitrate (NITRATE: 0 or 10 g of nitrate/kg DM; UREA or NIT, respectively), and 2 levels of 3-nitrooxypropanol (3-NOP: 0 or 80 mg/kg DM; BLANK or NOP, respectively).
2 Standard error of estimated marginal mean. The lowest value within the variable is reported as SEM.
3 Duration of time where gate to Insentec bin was open and cow's head was placed inside the bin.
4 Meal initiation was defined as visits started after 30 min from the last visit.
5 Means and SEM are from model without log-transformation, but P-values are from log-transformed model.
6 Eaten DM per minute of total eating time.
* Indicates significant interaction (P < 0.05) to parity group (Parity; primiparous, multiparous).
Number of visits in the assigned feed bin (Table 4) was lower in cows on NOP compared with cows on BLANK (15.4 vs. 16.6; P < 0.01), whereas number of attempts to visit an unassigned feed bin as well as percent of visits in unassigned feed bins of total daily visit attempts increased additively by FAT, NITRATE, and 3-NOP (all P < 0.05). Daily number of meals was greater for cows on NOP compared with BLANK (8.43 vs. 8.10 meals/d; P < 0.01; Table 4), whereas meal size was lower for these cows (2.2 vs. 2.7 kg DM/meal; P < 0.01). A similar pattern for number and size of meals was observed, when NIT and HF were fed in combination (PFAT × NITRATE <0.03 and <0.01, respectively), whereas we observed no significant separate effect of HF or NIT.
Milk Production
For milk and ECM yield (Table 5) a FAT × NITRATE interaction (P ≤ 0.01) was observed. Cows on HF-UREA had 8.6% and 6.3% greater milk and ECM yield, respectively, than cows on LF-UREA (31.7 vs. 29.2 kg/d and 32.1 vs. 30.2 kg, respectively; P < 0.01), whereas for cows on LF-NIT or HF-NIT, we observed no difference in milk and ECM yield, compared with cows on a LF-UREA diet. Cows on NOP had 11.7% and 9.0% lower milk and ECM yield compared with BLANK (28.2 vs. 31.5 kg/d and 29.3 vs. 31.9 kg/d, respectively; P < 0.01), and the differences were more pronounced in multiparous (31.2 vs. 35.8 kg/d and 31.9 vs. 35.7 kg/d, respectively; P < 0.01) than in primiparous cows (25.3 vs. 27.3 kg/d and 29.3 vs. 31.9 kg/d, respectively; P < 0.01; PParity × 3-NOP < 0.01).
Table 5Milk production parameters of dairy cows fed the 8 partial mixed ration (PMR) diets
The PMR diets were 2 × 2 × 2 factorially arranged aiming for 2 levels of fat (Fat: 30 or 63 g of crude fat/kg DM; LF or HF, respectively), 2 levels of dietary nitrate (NITRATE: 0 or 10 g of nitrate/kg DM; UREA or NIT, respectively), and 2 levels of 3-nitrooxypropanol (3-NOP: 0 or 80 mg/kg DM; BLANK or NOP, respectively).
Indicates significant interaction (P < 0.05) to parity group (Parity; primiparous, multiparous).
<0.01
<0.01
0.70
0.24
0.87
0.13
Multiparous
38.9
39.9
39.5
42.4
37.4
40.6
38.8
41.4
Protein
Primiparous
39.3
39.3
38.7
38.8
37.7
37.9
36.8
36.8
0.42
0.07
<0.01
<0.01
0.69
0.53
0.24
0.83
0.65
Multiparous
38.2
38.3
37.4
37.7
36.3
35.8
35.7
35.1
Lactose
Primiparous
49.7
49.2
49.4
49.1
50.7
50.4
50.1
49.7
0.28
<0.01
<0.01
<0.01
<0.01
<0.01
0.63
0.60
0.52
Multiparous
48.3
48.0
48.2
47.8
49.6
48.8
48.6
48.2
Urea, mmol/L
Primiparous
5.17
5.19
5.15
5.14
3.75
4.02
3.85
4.07
0.131
0.19
<0.01
0.79
<0.01
0.56
<0.01
0.62
0.60
Multiparous
4.94
4.92
4.93
5.14
3.56
4.00
3.43
3.90
Milk N/dietary N, %
Primiparous
26.8
27.7
27.6
27.9
28.7
30.0
30.0
31.2
0.78
0.01
<0.01
<0.01
<0.01
0.36
0.20
0.47
0.79
Multiparous
28.3
29.4
29.5
30.6
30.8
33.1
32.8
34.1
1 The PMR diets were 2 × 2 × 2 factorially arranged aiming for 2 levels of fat (Fat: 30 or 63 g of crude fat/kg DM; LF or HF, respectively), 2 levels of dietary nitrate (NITRATE: 0 or 10 g of nitrate/kg DM; UREA or NIT, respectively), and 2 levels of 3-nitrooxypropanol (3-NOP: 0 or 80 mg/kg DM; BLANK or NOP, respectively).
2 Standard error of estimated marginal mean. The lowest value within the variable is reported as SEM.
* Indicates significant interaction (P < 0.05) to parity group (Parity; primiparous, multiparous).
Milk fat concentration was lower for HF cows compared with LF cows (40.2 vs. 41.5 g/kg; P < 0.01), greater in NIT than in UREA cows (41.2 vs. 40.4 g/kg; P < 0.01), and greater in NOP than in BLANK cows (42.0 vs. 39.7 g/kg; P < 0.01). However, milk protein concentration was unaffected by 3-NOP (P = 0.69), lower in HF cows than LF cows (36.5 vs. 38.5 g/kg; P < 0.01) and lower in NIT than UREA cows (37.1 vs. 37.9 g/kg; P < 0.01). For milk lactose concentration a FAT × NITRATE interaction (P < 0.01) showed that lactose concentration was greater in HF-UREA cows compared with LF-UREA (49.9 vs. 48.8 g/kg; P < 0.01), however, we observed no difference when fed LF-NIT or HF-NIT diets, compared with LF-UREA diets. Lactose concentration was lower in cows fed NOP compared with BLANK (48.9 vs. 49.3 g/kg; P < 0.01). A FAT × 3-NOP interaction (P < 0.01) for milk urea concentration, showed that urea concentration was lower in HF-BLANK than in LF-BLANK cows (3.64 vs. 5.05 mmol/L; P < 0.01). Although milk urea concentration was similar in LF-NOP cows compared with LF-UREA cows, urea concentration was lower for HF-NOP compared with LF-NOP cows (4.00 vs. 5.10 mmol/L; P < 0.01). Although milk fat concentration decreased by fat addition, the greater milk yield in HF cows than in LF cows resulted in greater fat yield (kg/d) in HF cows, but only when fat was not combined with nitrate (PFAT × NITRATE < 0.01). Similarly, the observed effects on milk yield affected total protein and lactose yield accordingly. Milk N efficiency increased additively by FAT, NITRATE, and 3-NOP with 2.8, 1.1, and 1.2 percentage points (all P < 0.01), respectively, and was generally greater in multiparous cows than in primiparous cows (31.1% vs. 28.7%; PParity = 0.01).
Gas Emissions
Averaged across parities, CH4 production (g/d), yield (g/kg DMI), and intensity (g/kg ECM) varied from 217 to 380 g/d, 12.0 to 16.7 g/kg DMI, and 7.7 to 12.3 g/kg ECM, respectively, between treatments (Table 6). For CH4 yield, two 3-way interactions (FAT × 3-NOP × parity, P = 0.03 and NITRATE × 3-NOP × parity, P = 0.03) were found and were a result of the more pronounced decreasing effect of combinations of fat and 3-NOP, and nitrate and 3-NOP on DMI in multiparous compared with primiparous cows. Additionally, three 2-way interactions (FAT × NITRATE, FAT × 3-NOP, and NITRATE × 3-NOP; all P < 0.01) were found. Figure 1 shows the CH4 yield of the 3 interactions. The FAT × NITRATE interaction (Figure 1a) showed that cows on HF-UREA, LF-NIT, and HF-NIT had a 7.1%, 12.1%, and 8.2% lower CH4 yield than cows on LF-UREA, respectively (13.5, 12.7, and 13.3 vs. 14.5 g/kg DMI; P ≤ 0.01), whereas we observed no significant difference in the CH4 mitigating effect between the 3 diets including HF or NIT. The FAT × 3-NOP interaction (Figure 1b) revealed that CH4 yield was reduced by 6.0% for cows on HF-BLANK compared with cows on LF-BLANK (14.3 vs. 15.2 g/kg DMI; P = 0.04), and reduced by 21.0% and 18.0% for cows on LF-NOP and HF-NOP, respectively (12.0 and 12.5 vs. 15.2 g/kg DMI; P < 0.01), though we observed no difference between the 2 latter diets (P = 0.57). Similarly, the NITRATE × 3-NOP interaction (Figure 1c) resulted from that CH4 yield in cows on NIT-BLANK was 13.3% lower than in cows on UREA-BLANK (13.7 vs. 15.8 g/kg DMI; P < 0.01), whereas for cows fed UREA-NOP and NIT-NOP CH4 yield was reduced by 23.2% and 22.0%, respectively, compared with UREA-BLANK (12.1 and 12.3 vs. 15.8 g/kg DMI; P < 0.01), and we observed no difference between the 2 diets including NOP (P = 0.94).
Table 6Visits to GreenFeed and gas emissions of dairy cows fed the 8 partial mixed ration (PMR) diets
The PMR diets were 2 × 2 × 2 factorially arranged aiming for 2 levels of fat (Fat: 30 or 63 g of crude fat/kg DM; LF or HF, respectively), 2 levels of dietary nitrate (NITRATE: 0 or 10 g of nitrate/kg DM; UREA or NIT, respectively), and 2 levels of 3-nitrooxypropanol (3-NOP: 0 or 80 mg/kg DM; BLANK or NOP, respectively).
Indicates significant interaction (P < 0.05) to parity group (Parity; primiparous, multiparous).
<0.01
<0.01
0.81
0.10
<0.01
1.0
Multiparous
0.076
0.296
0.171
0.280
0.085
0.312
0.197
0.276
Intensity, g/kg ECM
Primiparous
0.065
0.232
0.135
0.217
0.065
0.189
0.121
0.182
0.010
0.10
<0.01
<0.01
<0.01
0.83
<0.01
<0.01
0.64
Multiparous
0.055
0.208
0.115
0.182
0.059
0.189
0.122
0.159
1 The PMR diets were 2 × 2 × 2 factorially arranged aiming for 2 levels of fat (Fat: 30 or 63 g of crude fat/kg DM; LF or HF, respectively), 2 levels of dietary nitrate (NITRATE: 0 or 10 g of nitrate/kg DM; UREA or NIT, respectively), and 2 levels of 3-nitrooxypropanol (3-NOP: 0 or 80 mg/kg DM; BLANK or NOP, respectively).
2 Standard error of estimated marginal mean. The lowest value within the variable is reported as SEM.
* Indicates significant interaction (P < 0.05) to parity group (Parity; primiparous, multiparous).
Figure 1Methane (CH4) yield (g/kg DMI) illustrated by treatment interactions between fat levels (FAT; LF, HF), nitrate levels (NITRATE; UREA, NIT), and 3-nitrooxypropanol levels (3-NOP; BLANK, NOP). (A) shows the FAT × NITRATE interaction (P < 0.01; averaged across levels of 3-NOP), (B) shows the FAT × 3-NOP interaction (P < 0.01; averaged across levels of NITRATE), and (C) shows the NITRATE × 3-NOP interaction (P < 0.01; averaged across levels of FAT). Orange and blue bars represent treatment estimated marginal means for primiparous and multiparous cows, respectively, and error bars represent standard error of estimated marginal means. Bars with no common letter differ (P < 0.05) within parity group (primiparous: a–c; multiparous: x–z), if parity group interacts with the 2-way interaction.
For CO2 yield (g/kg DMI), a FAT × NITRATE interaction was found. Cows on HF-NIT had greater CO2 yield than cows on LF-UREA (632 vs. 607 g/kg DMI; P < 0.01), although we observed no difference comparing HF-UREA and LF-NIT with LF-UREA (P = 0.37 and P = 0.35, respectively). Yield of CO2 increased as a response to NOP addition compared with BLANK (638 vs. 590 g/kg DMI; P < 0.01).
A NITRATE × 3-NOP interaction was observed for H2 yield (g/kg DMI) and was a result of a 115% greater H2 yield in cows on NIT-BLANK compared with UREA-BLANK (0.183 vs. 0.085 g/kg DMI; P < 0.01) and a 256% and 240% greater H2 yield in cows on UREA-NOP and NIT-NOP, respectively, compared with UREA-BLANK (0.301 and 0.288 vs. 0.085 g/kg DMI; P < 0.01). However, we observed no difference in H2 yield between cows on NIT-NOP and cows on UREA-NOP (P = 0.39).
Apparent Total-Tract Digestibility
Fecal DM concentration (g/kg) and fecal score increased by FAT, NITRATE, and 3-NOP (all P < 0.05; Table 7). A NITRATE × 3-NOP interaction (P = 0.04) was observed for digestibility of OM, reflecting that both NIT and NOP increased OM digestibility, although the combination of the 2 did not lead to any further significant increase. Digestibility of OM was similar in cows on NIT-BLANK, UREA-NOP, and NIT-NOP, but all greater than in cows on UREA-BLANK (73.3%, 73.6%, and 74.2% vs. 70.7%, all P < 0.01). Additionally, OM digestibility was greater in HF compared with LF cows (73.5% vs. 72.4%; P = 0.04). Digestibility of NDF was also greater in HF cows compared with LF cows (58.3% vs. 55.6%; P < 0.01), greater in NIT compared with UREA (58.6% vs. 55.4%; P < 0.01) and greater in NOP compared with BLANK (57.8% vs. 56.2%; P = 0.03). For all digestibility measures, we observed no parity effect.
Table 7Fecal DM concentration, fecal score, and apparent total-tract digestibility of dairy cows fed the 8 partial mixed ration (PMR) diets
The PMR diets were 2 × 2 × 2 factorially arranged aiming for 2 levels of fat (Fat: 30 or 63 g of crude fat/kg DM; LF or HF, respectively), 2 levels of dietary nitrate (NITRATE: 0 or 10 g of nitrate/kg DM; UREA or NIT, respectively), and 2 levels of 3-nitrooxypropanol (3-NOP: 0 or 80 mg/kg DM; BLANK or NOP, respectively).
Standard error of estimated marginal mean. The lowest value within the variable is reported as SEM.
P-value
UREA
NIT
UREA
NIT
BLANK
NOP
BLANK
NOP
BLANK
NOP
BLANK
NOP
Parity
FAT
NITRATE
3-NOP
FAT × NITRATE
FAT × 3-NOP
NITRATE × 3-NOP
FAT × NITRATE × 3-NOP
Observations, n
Primiparous
9
9
9
9
9
8
9
9
Multiparous
9
9
9
9
9
8
7
8
Fecal DM concentration, g/kg
Primiparous
124
131
132
135
127
139
141
154
4.6
0.58
<0.01
<0.01
<0.01
0.09
0.05
0.21
0.60
Multiparous
128
129
124
137
129
138
129
152
Fecal score
Primiparous
3.0
3.2
3.3
3.2
3.0
3.4
3.4
3.7
0.14
0.64
<0.01
<0.01
<0.01
0.24
0.05
0.46
0.72
Multiparous
3.2
3.2
3.1
3.4
3.2
3.4
3.2
3.8
Digestibility, %
Ash
Primiparous
95.4
96.0
95.9
95.8
96.1
96.0
96.2
96.3
0.27
0.25
0.07
0.94
0.63
0.85
0.27
0.26
0.76
Multiparous
96.3
96.2
96.1
96.0
96.3
96.2
96.5
95.7
OM
Primiparous
69.7
73.6
72.5
72.3
70.3
74.4
73.6
75.5
0.99
0.37
0.04
<0.01
<0.01
0.52
0.60
0.04
0.86
Multiparous
71.2
72.8
73.1
74.4
71.6
73.7
74.1
74.6
NDF
Primiparous
54.1
57.5
55.1
56.1
55.2
58.2
60.7
61.6
1.46
0.41
<0.01
<0.01
0.03
0.14
0.96
0.11
0.87
Multiparous
52.0
54.6
57.8
57.6
54.6
56.9
59.7
59.9
1 The PMR diets were 2 × 2 × 2 factorially arranged aiming for 2 levels of fat (Fat: 30 or 63 g of crude fat/kg DM; LF or HF, respectively), 2 levels of dietary nitrate (NITRATE: 0 or 10 g of nitrate/kg DM; UREA or NIT, respectively), and 2 levels of 3-nitrooxypropanol (3-NOP: 0 or 80 mg/kg DM; BLANK or NOP, respectively).
2 Standard error of estimated marginal mean. The lowest value within the variable is reported as SEM.
Total Hb averaged 7.0 ± 0.07 mmol/L and Hct averaged 34.5% ± 0.65%, and we observed no treatment effects (all P > 0.10). Methemoglobin tended to be greater in NIT than in UREA supplemented cows (2.41% vs. 2.26% of tHb; P = 0.06), but no MetHb values exceeded 3.7% of tHb in NIT supplemented cows.
Supplemental Table S1: Effects of dietary fat, nitrate, and 3-nitrooxypropanol and their combinations on methane emission, feed intake, and milk production in dairy cows. Zenodo.
With nitrate supplementation, reductions in feed intake of around 5% to 8% has previously been reported in some studies with dairy cows fed similar or higher nitrate doses (
Potential roles of nitrate and live yeast culture in suppressing methane emission and influencing ruminal fermentation, digestibility, and milk production in lactating Jersey cows.
). In the present study, NIT fed separately within the FAT × NITRATE interaction decreased DMI by 4.6%, but 13% when combined with HF. A similar interaction was reported in a study by
supplementing nitrate and rapeseed oil to steers, although there was no negative effect on DMI of nitrate alone. In our study, NOP decreased DMI by 9.5% and 16.8% in primiparous and multiparous cows, respectively (13.4% across parities), when using an actual dose of 76 ± 2.4 mg 3-NOP/kg of DM. Previous literature, applying similar dose of 3-NOP mixed into feed rations for dairy cows, reported smaller (1%–5%), but statistically insignificant decreases in DMI (
reported a tendency for a linear decrease in DMI of dairy cows by increased dose of 3-NOP (−6% at doses from 100 to 200 mg/kg DM). The discrepancy between the rather substantial negative effect of the applied dose of 3-NOP in the present study compared with others is unknown, but may be related to interactions to regional differences in diet composition (
Treatments including NIT or NOP in the present study also altered cows' feeding behavior toward less time spent eating and with shorter meal duration, which is most likely directly related to the lower feed intake. However, for NOP supplemented cows, the meals appeared more efficient, with fewer interruptions of eating events, as indicated by a greater amount of meal time being used as actual eating time, although this was not the case for NIT supplemented cows. Additionally, daily eating rate was slower for these cows, summing up to only 0.6 and 0.3 kg of DM per day for NOP and NIT cows, respectively, which does not account for the observed reductions in DMI seen for these treatments (1.7 and 2.9 kg of DM per day, respectively). This indicates that other factors than slower eating rate explained cows' reduced DMI for these 2 treatments, whereas the greater eating rate in HF cows than in LF cows is merely explained by a slightly greater DM concentration of HF diets compared with LF diets. Cows on NIT and NOP were also having smaller, but more frequent meals, although meal frequency only increased significantly when NIT was combined with HF. These treatment differences were minor, but points in the same direction as what has been observed from nitrate fed beef cattle by
DMI reductions caused by nitrate, in feed restricted heifers, were more likely related to organoleptic properties of nitrate rather than physiological discomfort caused by nitrate poisoning indicated by increased MetHb levels of blood. In the present study, no NIT cows had MetHb (% of tHb) levels in blood sampled 3 h after returning from morning milking exceeding levels above 3.7%. According to
signs of toxicity appears when the proportion of MetHb of tHb exceeds 20%, indicating that DMI reductions caused by acute nitrate poisoning seems unlikely, knowing that the formation of MetHb is dependent on feed consumption rate. However,
reported that in heifers fed the same inclusion rate of nitrate as in the present study (1% of DM), MetHb never exceeded 2.73% of tHb 3 h after feeding, where 43% of the daily offered feed was consumed. For diets with 3-NOP,
Effects of 3-nitrooxypropanol on enteric methane production, rumen fermentation, and feeding behavior in beef cattle fed a high-forage or high-grain diet.
suggested that cows were unaffected by possible organoleptic properties of 3-NOP, although data from the present experiment showed that cows to a greater extent seemed to seek other feed bins, than the one they were assigned to, as attempts to visit unassigned feed bins increased for all additives. Whether this behavior was motivated by palatability or physiological discomfort of eating the assigned ration is unknown, but lower or even negative BW changes in the same cows may as well reflect less gastrointestinal tract fill, which may suggest that cows experienced hunger, and therefore sought other bins. It could furthermore have been hypothesized that dietary inclusion of the additives would have affected drinking water intake in an attempt to overcome possible discomfort, however, the observed differences in drinking water intake merely reflected the decreased DMI, as we observed no treatment differences in drinking water intake when corrected for DMI (Supplemental Table S1).
Along with CH4 reductions caused by NIT and NOP, elevated H2 emissions were concurrently observed in the present study. It has been suggested that increased ruminal dissolved hydrogen concentration (dH2) may impair rumen fermentation, through increased ratio of the co-factors NADH to NAD+ in the rumen environment, although it is unclear if dH2 during CH4 inhibition ever reach sufficient levels to thermodynamically control reoxidation of NADH to NAD+ (
). The considerable improvements in apparent total-tract digestibility of OM in cows fed HF, NIT, or NOP in the present study, indicated that rumen fermentation was not impaired, knowing that these digestibility measures reflect total-tract digestibility and not rumen digestibility only, hence a theoretical compensatory postruminal digestion cannot be ruled out. Additionally, the fact that multiparous cows' DMI were more negatively affected by the treatments than primiparous cows, may accordingly support that the mechanism behind the reduced DMI does not relate to increased rumen fill, because increased rumen fill, as a consequence of impaired ruminal fermentation and hence slower passage rate, would be expected to limit DMI more in primiparous cows than in multiparous cows (
). Furthermore, it has been reported that increased H2 emission during CH4 inhibition was accompanied by increased ruminal concentrations of lactate, formate, and some primary alcohols such as ethanol (
). Such metabolites may facilitate systemic responses associated with discomfort or in other ways negatively affect DMI of cows. Independent on whether the mechanism behind DMI reductions in the present study was caused by palatability, physiological discomfort, or something else, it remains unclear why combinations of diets with HF (HF with NIT or HF with NOP) decreased DMI in a somewhat synergistic manner, whereas the combination of nitrate and 3-NOP additively decreased DMI.
, responses in milk production can be underestimated if changes in daily DMI of cows exceed 5 kg/d, by using crossover designs as in the present study. Total DMI in this experiment varied from 17.0 to 21.2 kg/d and from 17.3 to 25.0 kg/d for primiparous and multiparous cows, respectively. This suggest that potential responses in milk production for multiparous cows in particular may not have been fully reflected within the 3-wk periods for some treatments, although the balancing of the treatments across the periods enables evaluation of relative differences between treatments. Therefore, feed conversion efficiencies were not reported in the present study. Though, an analysis of carry-over effects showed no effect of previous period's treatment (results not shown).
Cows fed HF-UREA diets compared with LF-UREA diets responded positively in milk and ECM production (+9% and 6%, respectively), due to greater energy intake. Increased milk production by fat supplementation below 6% of DM was also concluded from a meta-analysis by
The effect of dietary fats on methane emissions, and its other effects on digestibility, rumen fermentation and lactation performance in cattle: A meta-analysis.
Enteric and manure-derived methane emissions and biogas yield of slurry from dairy cows fed grass silage or maize silage with and without supplementation of rapeseed.
also by using rapeseed as fat source. In the present study, HF supplementation also significantly increased lactose concentration of milk, which is regarded as the main osmotic driver for milk volume, hence increased milk volume most likely diluted fat and protein concentration of milk. The increase in milk production was only present if HF was not combined with NIT, which, supplemented alone, did not decrease milk production, although separate NIT supplementation negatively affected DMI (LF-NIT; −5%). Cows fed NOP, reduced milk and ECM production with 12% and 9% across parities, respectively, which again reflected decreasing effects on DMI. In general, the discrepancy between response in milk yield and ECM yield, was mainly driven by increased milk fat concentration of cows supplemented NIT or NOP, because protein and lactose concentration in milk was either decreased or unaffected by the treatments. Increased fat concentration may be a result of increased body fat mobilization due to lower DMI, as well as changes in rumen fermentation products, altering availability of precursors for mammary de novo fatty acid synthesis. However, yet unpublished data showed no or decreasing effects of treatments on ruminal concentrations of acetic acid, but increasing effects on butyric acid, from cows in the present study. Nevertheless, evaluating the sum of acetic acid and butyric acid revealed that only supplementation of NIT, alone, or in combination with NOP, resulted in increased total concentration of such precursors for milk fat, which illustrates that this cannot be the explanation for increased milk fat concentration in NOP supplemented cows. This suggests that NOP supplemented cows partly compensated for decreased DMI by increased body mobilization to maintain milk production, despite concurrent increased apparent total-tract digestibility of nutrients.
, FAT in particular had a positive effect on cows' deposition of N in milk as shown by a 2.8 percentage point increased milk N efficiency (milk N/dietary N) in HF cows compared with LF cows. Also supported by a lower milk urea concentration in fat supplemented cows, this is a result of a lower total N intake along with an unaffected daily milk protein yield (kg/d) in HF supplemented cows. Increases in N efficiency by NIT and NOP, despite lower N intake and lower N output in milk, may be due to increased mobilization of body protein or increased utilization of N in feed, also supported by an absence of a response in milk urea concentration of separate NIT or NOP supplementation.
Gas Emissions
In the present study, all 3 additives reduced absolute CH4 emission. However, with reductions in feed intake and milk production responses that may not fully reflect changes in feed intake within the course of an experimental period, only emissions related to DMI (gas yield) will be discussed in the following section.
Combining additives to reduce CH4 emission revealed interactions between the additives, why any main effects of the additives in such cases should be interpreted with caution. Therefore, as an example, the effect of separate fat addition should be evaluated based on significant interactions, wherein FAT is present. Based on the FAT × NITRATE and the FAT × 3-NOP interactions, the CH4 reduction potential of fat alone, in the present study, can be ascribed to 6% to 7%, corresponding to 1.6% to 1.9% reduction in CH4 yield per percent added fat. This reduction potential appears to be in the lower range of what previously has been reported in studies by using rapeseed as fat source (
Enteric and manure-derived methane emissions and biogas yield of slurry from dairy cows fed grass silage or maize silage with and without supplementation of rapeseed.
). Based on interactions including NITRATE, around 12% to 13% reduction in CH4 yield can be ascribed to NIT, in the present study, which is in line with reductions reported by
3-Nitrooxypropanol decreases methane emissions and increases hydrogen emissions of early lactation dairy cows, with associated changes in nutrient digestibility and energy metabolism.
; 22%). Theoretically, supplementation of 10 g of nitrate /kg of DM would yield a reduction of 2.58 g/kg DMI (stoichiometric calculations not shown), assuming a complete reduction of NO3- to NH4+. The reduction obtained by NIT in the present study was 1.95 g/kg DMI, which is equivalent to an apparent efficiency of 75.6%, even though DMI was reduced by NIT. This indicates that nitrate as CH4 mitigating additive was as effective in the present study as in other studies, although variation between studies and between animals within studies must be expected (apparent efficiency of 59%,
). Evaluating interactions including 3-NOP, the CH4 reduction potential of separate NOP supplementation was around 21% to 23% in the present study, which, similar to nitrate, appeared to be in the lower range of the elsewhere reported reduction potentials (
3-Nitrooxypropanol decreases methane emissions and increases hydrogen emissions of early lactation dairy cows, with associated changes in nutrient digestibility and energy metabolism.
). In general, these lower-than-expected reduction potentials could be a potential lack of microbial adaptation within the experimental period or a result of the negative effect on DMI, although total-tract digestibility increased, which may have given rise to compensatory rumen fermentation. This counteracted the obtained absolute effect of the additives per kg of DMI. Nevertheless, the observed treatment interaction to parity group on CH4 yield was also a result of the more pronounced effect on DMI in multiparous cows, which illustrates that the efficiency of additives per se was similar across parities.
The objective of the present study was to evaluate interactions when combining additives. Knowing that the effects on CH4 reduction of such combinations may have been different, if DMI was not decreased, it is still possible to reflect on the observed interactions. Assuming that most of the reduction potential of fat supplementation can be ascribed to the replacement of rumen fermentable OM by nonfermentable OM, it was expected that combining HF with NIT or NOP would have reduced CH4 yield in an at least additive manner. Synergistic effects on CH4 yield reduction of fat and nitrate was reported by
Combined effects of 3-nitrooxypropanol and canola oil supplementation on methane emissions, rumen fermentation and biohydrogenation, and total tract digestibility in beef cattle.
, respectively. This was not the case in the present study, as combinations of HF with either NIT or NOP as well as the combination of NIT and NOP did not lead to further significant reductions other than what could be achieved by separate supplementation of the most effective of the 2, indicating antagonistic effects of the specific combination on CH4 yield reduction due to the absence of additivity and synergistic effects. It is expected that such antagonistic effects are related to the somewhat synergistically decreasing effects on DMI in cows fed combinations of HF with NIT or NOP, or the additive decreasing effect on DMI in cows fed combinations of NIT and NOP. Moreover, as 3-NOP in the present and in previous studies (
) have resulted in increased H2 emission, one could expect that combined use of 3-NOP with either fat or nitrate would result in net-decreased H2 emission, compared with separate supplementation of 3-NOP, due to possible biohydrogenation of fatty acids or reduction of NO3- to NH4+ consuming H2 (
The role of thermodynamics in the control of ruminal fermentation.
in: Sejrsen K. Hvelplund T. Nielsen M.O. Ruminant Physiology: Digestion, Metabolism and Impact of Nutrition on Gene Expression, Immunology and Stress. Wageningen Academic Publishers,
2006: 55-85
Combined effects of 3-nitrooxypropanol and canola oil supplementation on methane emissions, rumen fermentation and biohydrogenation, and total tract digestibility in beef cattle.
, who found lower H2 emission from beef cattle fed combinations of rapeseed oil and 3-NOP compared with separate supplementation. In the present study, combined use of NIT and NOP did not change H2 emissions significantly compared with separate supplementation, although numerical reductions appeared. Similarly, FAT had no measurable effect neither. The lack of such effects may as well be related to reductions in DMI that may have affected total H2 production or a possible underestimation (or biased estimation) of H2 emission if a large amount of emitted H2 is missed if an animal does not visit the GreenFeed relatively shortly after a meal (
). Gas measurements may also have been biased for some treatments, due to relatively more visits in GreenFeed when cows were supplemented with HF, NIT, or NOP, thereby increasing the likelihood of measuring the peaks in emission.
Increases in CO2 yield were observed in the present study, when CH4 yield was reduced. Based on significant effects only, yield of CO2 was 25 g/kg DMI greater in HF-NIT cows compared with LF-UREA cows and 48 g/kg DMI greater in cows on NOP compared with BLANK. Corresponding reductions in CH4 yield in the same groups were 1.2 and approximately 3.5 g/kg DMI for HF-NIT and NOP, respectively. Assuming that 1 mol of CO2 is used to produce 1 mol of CH4 during methanogenesis (
), reductions of 1.2 and 3.5 g of CH4/kg DMI in the present study would correspond to 3.3 and 9.6 g of CO2/kg DMI. Thus, a greater than theoretically expected CO2 yield was observed in these groups, which may suggest that additional CO2 originated from other either fermentative or metabolic processes. However, because improved total-tract digestibility could not fully compensate for decreased OM intake in these groups (see discussion of apparent total-tract digestibility), it appears likely that part of the increased CO2 yield originates from metabolic processes such as mobilization of body reserves.
Apparent Total-Tract Digestibility
The observed improvements of apparent total-tract digestibility of nutrients caused by HF, NIT, and NOP in the present study were rather substantial. Fed separately, HF, NIT (NIT-BLANK within NITRATE × 3-NOP), and NOP (UREA-NOP within NITRATE × 3-NOP) improved total-tract OM digestibility with 1.1, 2.6, and 2.9 percentage points, respectively. The improved OM digestibility increased the total amount of digested OM relatively, although it was not enough to fully compensate for the reduced OM intake. Across parities, total intake of digested OM was reduced by 4.6% and 13.4% respectively by NIT and NOP supplementation, whereas we observed no effect of FAT. Accordingly, increased digestibility of OM gave rise to 1% increase in digested OM (kg/d) by HF supplementation, but 1.0% and 10.2% decrease in digested OM by NIT and NOP, respectively (data not shown). For NOP supplemented cows compared with BLANK for example, this difference in digested OM corresponded roughly to the observed reductions in milk production.
Averaged per treatment and across parities, the obtained improvements in apparent total-tract digestibility of OM was negatively correlated with DMI (r = −0.94; data not shown), supporting the hypothesis that decreased DMI led to slower gastrointestinal tract passage rate and most likely also increased rumen retention time, as documented by
Influence of declining level of feed-intake and varying the proportion of starch in the concentrate on milk-production and whole tract digestibility in dairy-cows.
. Across parities, DMI in cows fed the combination of HF, NIT, and NOP (HF-NIT-NOP) was 5.7 kg less than in cows receiving no additives (LF-UREA-BLANK). Comparing the same groups, OM digestibility was increased by 4.6 percentage points, whereas NDF digestibility was increased by 7.7 percentage points, when DMI was decreased.
Influence of declining level of feed-intake and varying the proportion of starch in the concentrate on milk-production and whole tract digestibility in dairy-cows.
investigated the effect of feed intake on total-tract digestibility in cows during a full lactation and reported that change in digestibility of NDF depends on feed intake level and thus reported a decrease in NDF digestibility of 5.4 percentage points by an increase in DMI from 6 to 15 kg and an additional decrease of 8.3 percentage points by increasing DMI from 15 to 24 kg. This supports that the rather substantial increase in apparent total-tract digestibility obtained in the present study may be explained by decreased DMI, hence decreased passage rate. An observed higher fecal DM concentration and feces score of treatment groups with greatest reductions in DMI, further support this hypothesis. Moreover, it has been reported in a review by
The effect of dietary fats on methane emissions, and its other effects on digestibility, rumen fermentation and lactation performance in cattle: A meta-analysis.
suggested that this is dependent on fatty acid composition and the degree of unsaturation of fatty acids. Fat from rapeseed is mainly constituted of monounsaturated fatty acids and
The effect of dietary fats on methane emissions, and its other effects on digestibility, rumen fermentation and lactation performance in cattle: A meta-analysis.
suggested that the negative effect of fat on fibrolytic bacteria in the rumen were more pronounced if fat was supplemented as oils compared with oil seeds, as in the present study. However,
reported no effects on rumen digestibility nor total-tract digestibility of OM and NDF, when rapeseed were supplemented as cake, whole cracked or as oil, whereas
Combined effects of 3-nitrooxypropanol and canola oil supplementation on methane emissions, rumen fermentation and biohydrogenation, and total tract digestibility in beef cattle.
reported decreased DM and NDF total-tract digestibility from rapeseed oil supplemented to steers. Digestibility of NDF did not decrease by HF supplementation in the present study. Due to confounding effects of decreased DMI in the present study, it cannot be concluded if NIT or NOP as such affected digestibility measures, however, recent literature have not reported any negative effects, not ascribed to changes in DMI (nitrate:
Effects of calcium ammonium nitrate fed to dairy cows on nutrient intake and digestibility, milk quality, microbial protein synthesis, and ruminal fermentation parameters.
Combined effects of 3-nitrooxypropanol and canola oil supplementation on methane emissions, rumen fermentation and biohydrogenation, and total tract digestibility in beef cattle.
Observed interactions for CH4 yield (g/kg DMI) between dietary fat, nitrate, and 3-NOP illustrated antagonistic effects of combinations of additives. No combinations of treatments resulted in CH4 emission reductions that were greater than what were achieved by separate supplementation of the most effective additive within the combination. Additionally, interactions between the additives for DMI revealed that combined supplementation of HF and NIT decreased DMI in a synergistic manner, whereas NIT and NOP, alone and in combination, and the 3 in combination, decreased DMI in an additive manner. Accordingly, eating behavior data showed that cows' attempts to seek unassigned feed increased, when HF, NIT, and NOP were supplemented, and this effect was additive. Milk production was increased by HF supplementation, however, the negative effects of NIT and NOP on DMI were similarly reflected in milk production parameters, despite remarkably improved apparent total-tract digestibility. Based on the findings of the present study, it can be concluded that the separate CH4 mitigating effect of fat, nitrate, and 3-NOP were not additive, and that supplementation of combinations of these additives negatively affected production parameters.
ACKNOWLEDGMENTS
The project was funded by the Ministry of Food, Agriculture and Fisheries of Denmark under the project “Feeding and Phenotype of the Climate Efficient Dairy Cow” and was conducted at the Danish Cattle Research Center (DKC) under Department of Animal and Veterinary Sciences, AU Viborg–Research Centre Foulum, Aarhus University. The staff at DKC are acknowledged for their expert caretaking of animals, Torkild Nyholm Jakobsen, Ester Bjerregaard, Anne Svensson, Mirka Thorsteinsson, Wenji Wang, Maria Kjeldsen, Giulio Giagnoni, and Nikolaj Hansen are acknowledged for their assistance during the samplings, laboratory technicians for performing laboratory procedures, Sander van Zijderveld (Cargill Inc., the Netherlands) for supplying the nitrate source, Jørgen Eriksen and Marianne Stevenson (Department of Agroecology, AU) for nitrate analyses in forages, and Katrien Schäfer (DSM Nutritional Products, Kaiseraugst, Switzerland) for the 3-NOP analyses in feed. The authors have not stated any conflicts of interest.
REFERENCES
Åkerlind M.
Volden H.
Standard feed value.
in: Volden H. NorFor - The Nordic Feed Evaluation System. Wageningen Academic Publishers. EAAP publication no. 130,
2011: 137-139
Effects of calcium ammonium nitrate fed to dairy cows on nutrient intake and digestibility, milk quality, microbial protein synthesis, and ruminal fermentation parameters.
Enteric and manure-derived methane emissions and biogas yield of slurry from dairy cows fed grass silage or maize silage with and without supplementation of rapeseed.
Effects of 3-nitrooxypropanol on enteric methane production, rumen fermentation, and feeding behavior in beef cattle fed a high-forage or high-grain diet.
Supplemental Table S1: Effects of dietary fat, nitrate, and 3-nitrooxypropanol and their combinations on methane emission, feed intake, and milk production in dairy cows. Zenodo.
Potential roles of nitrate and live yeast culture in suppressing methane emission and influencing ruminal fermentation, digestibility, and milk production in lactating Jersey cows.
The effect of dietary fats on methane emissions, and its other effects on digestibility, rumen fermentation and lactation performance in cattle: A meta-analysis.
Influence of declining level of feed-intake and varying the proportion of starch in the concentrate on milk-production and whole tract digestibility in dairy-cows.
The role of thermodynamics in the control of ruminal fermentation.
in: Sejrsen K. Hvelplund T. Nielsen M.O. Ruminant Physiology: Digestion, Metabolism and Impact of Nutrition on Gene Expression, Immunology and Stress. Wageningen Academic Publishers,
2006: 55-85
3-Nitrooxypropanol decreases methane emissions and increases hydrogen emissions of early lactation dairy cows, with associated changes in nutrient digestibility and energy metabolism.
Combined effects of 3-nitrooxypropanol and canola oil supplementation on methane emissions, rumen fermentation and biohydrogenation, and total tract digestibility in beef cattle.
00625-2&id=gr1.jpg){kind=link}