Search
Author
Title
Vol.
Issue
Year
1st Page

Journal of Animal Science - Animal Nutrition

Additive methane-mitigating effect between linseed oil and nitrate fed to cattle1

 

This article in JAS

  1. Vol. 93 No. 7, p. 3564-3577
     
    Received: June 18, 2014
    Accepted: May 05, 2015
    Published: June 26, 2015


    2 Corresponding author(s): cecile.martin@clermont.inra.fr
 View
 Download
 Share

doi:10.2527/jas.2014-8196
  1. J. Guyader*,
  2. M. Eugène*,
  3. B. Meunier*,
  4. M. Doreau*,
  5. D. P. Morgavi*,
  6. M. Silberberg*,
  7. Y. Rochette*,
  8. C. Gerard,
  9. C. Loncke and
  10. C. Martin 2*
  1. * INRA, UMR1213 Herbivores, F-63122 Saint-Genès-Champanelle, France; Clermont Université, VetAgro Sup, UMR Herbivores, BP 10448, F-63000 Clermont-Ferrand, France
     InVivo Nutrition et Santé Animales, Talhouët, F-56250 Saint Nolff, France
     INZO, Rue de l’église, BP 50019, F-02407 Chierry Cedex, France

Abstract

The objective of this study was to test the effect of linseed oil and nitrate fed alone or in combination on methane (CH4) emissions and diet digestibility in cows. The experiment was conducted as a 2 × 2 factorial design using 4 multiparous nonlactating Holstein cows (initial BW 656 ± 31 kg). Each experimental period lasted 5 wk, with measures performed in the final 3 wk (wk 3 to 5). Diets given on a DM basis were 1) control (CON; 50% natural grassland hay and 50% concentrate), 2) CON with 4% linseed oil (LIN), 3) CON with 3% calcium nitrate (NIT), and 4) CON with 4% linseed oil plus 3% calcium nitrate (LIN+NIT). Diets were offered twice daily and were formulated to deliver similar amounts (DM basis) of CP (12.2%), starch (25.5%), and NDF (39.5%). Feed offer was restricted to 90% of voluntary intake (12.4 kg DMI/d). Total tract digestibility and N balance were determined from total feces and urine collected separately for 6 d during wk 4. Daily CH4 emissions were quantified using open chambers for 4 d during wk 5. Rumen fermentation and microbial parameters were analyzed from samples taken before and 3 h after the morning feeding. Rumen concentrations of dissolved hydrogen (H2) were measured continuously up to 6 h after feeding using a H2 sensor. Compared with the CON diet linseed oil and nitrate decreased (P < 0.01) CH4 emissions (g/kg DMI) by 17 and 22%, respectively, when fed alone and by 32% when combined. The LIN diet reduced CH4 production throughout the day, increased (P = 0.02) propionate proportion, and decreased (P = 0.03) ruminal protozoa concentration compared with CON diet. The NIT diet strongly reduced CH4 production 3 h after feeding, with a simultaneous increase in rumen dissolved H2 concentration, suggesting that nitrate does not act only as an electron acceptor. As a combined effect, linseed plus nitrate also increased H2 concentrations in the rumen. Diets had no effect (P > 0.05) on total tract digestibility of nutrients, except linseed oil, which tended to reduce (P < 0.10) fiber digestibility. Nitrogen balance (% of N intake) was positive for all diets but retention was less (P = 0.03) with linseed oil. This study demonstrates an additive effect between nitrate and linseed oil for reducing methanogenesis in cows without altering diet digestibility.



INTRODUCTION

Enteric methane (CH4) from ruminants is one of the most important greenhouse gases at the farm level (Gerber et al., 2013) and represents an energy loss to the animal (2–12% of GE intake; Johnson and Johnson, 1995). Lipids and nitrate (NO3) emerged as persistent and viable dietary options for mitigating CH4 emissions from ruminants (Doreau et al., 2014a). Linseed reduced methanogenesis (–5.6% per 1% added fat; Doreau et al., 2011), but this effect was not always reported (Chung et al., 2011; Veneman et al., 2014). Linseed, rich in PUFA, may improve animal product quality (Scollan et al., 2001; Chilliard et al., 2009), but fat doses greater than 5% may lower animals’ performance (McGinn et al., 2004; Martin et al., 2008). In the diet, NO3 repeatably reduced CH4 emissions (–10% per 1% added NO3; Lee and Beauchemin, 2014), but its use as a urea substitute still requires investigations into its possible impacts on animal health, digestive parameters, and residuals in animal products for human consumption.

In the rumen, CH4 is mainly produced by methanogens using carbon dioxide (CO2) and hydrogen (H2). Both are fermentation end products, but as H2 is limiting, modulating its concentration could reduce methanogenesis (Hegarty and Gerdes, 1999). Linseed and NO3 affect the rumen H2 pool in unique ways. Linseed reduces H2 production mainly through its toxic effect against rumen protozoa, which are major H2 producers (Morgavi et al., 2010). As fat is not fermented in the rumen, substitution of rumen fermentable substrates for lipids may also reduce H2 production. To a lesser degree, PUFA can reduce H2 availability in the rumen by consuming H2 during biohydrogenation (Czerkawski, 1986). Nitrate modifies H2 consumption by reducing the number of methanogens (Van Zijderveld et al., 2010) and by acting as a H2 sink (Lewis, 1951).

As these dietary treatments share different mechanisms of action, we hypothesized that their combination would have an additive effect that leads to less net methanogenesis than when they are individually fed. However, a feeding strategy should reduce CH4 emissions without adverse effects on animals’ digestive efficiency, performance, and health. Consequently, our hypothesis was tested in an in vivo experiment with dry cows designed to evaluate the effect of linseed plus nitrate on 1) CH4 emissions and mechanisms involved in methanogenesis (rumen H2 pool and fermentation) and 2) diet digestibility and N balance.


MATERIALS AND METHODS

The experiment was conducted at the animal facilities of the experimental unit of ruminants at the Inra’s Theix Research Centre (Saint-Genès-Champanelle, France) from January to June 2013. Procedures involving animals were performed in accordance with French Ministry of Agriculture guidelines for animal research and with the applicable European Union guidelines and regulations on experiments with animals. The experiment was approved by the local Auvergne-region ethics committee on animal experimentation, approval number CE50-12.

Animals, Experimental Design, and Diets

Four multiparous nonlactating Holstein cows fitted with rumen cannulas (initial average BW of 656 ± 31 kg and age of 6.7 ± 1.5 yr, mean ± SD) and habituated to handling were housed in individual stalls during the experiment. The cows were randomly assigned to 4 dietary treatments in a 2 × 2 factorial design, using either calcium nitrate or linseed oil at 2 different doses (0 and 3% for calcium nitrate and 0 and 4% for linseed oil). Each experimental period lasted 5 wk, with measures performed in the final 3 wk (wk 3 to 5). The diets, given on a DM basis, were 1) control diet (CON), 2) CON with 4% linseed oil (LIN), 3) CON with 3% calcium nitrate (NIT), and 4) CON with 4% linseed oil and 3% calcium nitrate (LIN+NIT). The doses of linseed oil (Vandeputte Savonnerie et Huilerie, Mouscron, Belgium) and calcium nitrate (75% NO3 in DM; Phytosem, Pont-du-Château, France) were calculated to achieve a theoretical CH4 reduction of 20% when distributed alone (Martin et al., 2008; Van Zijderveld et al., 2011; Hulshof et al., 2012).

Ingredients and chemical composition of the experimental diets are reported in Table 1. The CON diet consisted of 50% natural grass hay (harvested in semimountainous and permanent grassland areas) and 50% concentrate (DM basis). Diets were formulated at the beginning of the experiment to meet the ME requirements for maintenance of nonlactating cows (INRA, 2010) and to provide similar levels of NDF (to avoid any risk of acidosis; Krause and Oetzel, 2006), starch (to favor protozoa development; Jouany, 1989), and CP. Diet levels of fermentable N were kept similar to assess the effect of nitrate on N output. Diets were adjusted to have the same N and Ca concentrations by including urea and calcium carbonate in the non-NIT diets (i.e., CON and LIN). Calcium carbonate was used as it has low solubility in the rumen and thus avoids the formation of calcium salts with lipids (Keyser et al., 1985). A commercial mineral–vitamin premix was added in equal amounts to all diets. Forage was distributed without further processing. All other ingredients including linseed oil or nitrate or both were pelleted in concentrates (InVivo NSA, Chierry, France).


View Full Table | Close Full ViewTable 1.

Ingredients and chemical composition of the experimental diets

 
Item Diet1
CON NIT LIN LIN+NIT
Ingredient, % of DM
    Hay 50.00 50.00 50.00 50.00
    Pelleted concentrate
        Wheat 25.23 25.23 25.23 25.23
        Corn 15.00 15.00 15.00 15.00
        Calcium nitrate2 0 3 0 3
        Linseed oil 0 0 4 4
        Calcium carbonate 1.7 0 1.7 0
        Urea 1.22 0 1.22 0
        Dehydrated beet pulp 4.08 4 0.08 0
        Molasses beet 1 1 1 1
        Binder 1 1 1 1
        Mineral–vitamin mix 0.75 0.75 0.75 0.75
        Aroma3 0.02 0.02 0.02 0.02
Chemical composition, % of DM
    OM 91.3 91.5 91.8 91.8
    CP 12.7 12.2 12.1 11.7
    NDF 40.1 40.2 38.8 38.7
    ADF 23.3 23.1 22.2 22.2
    Starch 25.4 25.7 25.7 25.3
    Ether extract 2.08 1.90 4.66 3.12
    Total fatty acids 1.61 1.24 3.53 2.05
GE, MJ/kg of DM 17.4 16.6 18.3 17.7
ME, MJ/kg of DM4 9.7 9.7 10.5 10.5
Fatty acid, % of total fatty acids
    C16:0 18.56 24.55 14.18 20.38
    C18:0 1.98 2.58 4.92 6.56
    C18:1 n-9 19.53 22.90 23.13 28.60
    C18:2 n-6 47.50 29.33 24.89 21.22
    C18:3 n-3 8.01 7.72 29.37 17.63
1CON = control; NIT = CON with 3% calcium nitrate; LIN = CON with 4% linseed oil; LIN+NIT = CON with 4% linseed oil plus 3% calcium nitrate.
25Ca(NO3)2∙NH4NO3∙10H2O; 75% NO3 in DM.
3Gusti; Nutriad Ltd., Chester, England.
4Calculated from INRA (2010).

Feeding and Management

Two weeks before starting the experiment, cows were fed the CON diet ad libitum. Then, throughout the experiment, feed offered was restricted to 90% of individual voluntary feed intakes (1.8 times ME requirements for maintenance) to ensure complete consumption. The LIN, NIT, and LIN+NIT concentrates were progressively supplied by replacing the CON concentrate. The LIN concentrate was fed at maximal dose after a 5-d transition period. The NIT and LIN+NIT concentrates were fed at their maximal dose after a 10-d transition period.

Throughout the experiment, feed was offered twice daily (66% at 0800 h and 34% at 1600 h for hay and 60% between 0800 and 0930 h in 3 equal portions and 40% between 1600 and 1630 h in 2 equal portions for concentrates). Delivery of concentrates was fractionated to reduce the risk of methemoglobinemia (metHb; Morris et al., 1958). Forage-to-concentrate ratio (50:50) was kept as close as possible to the target ratio by adjusting the amounts of hay and concentrates offered daily. Cows had free access to water throughout the experiment.

Measurements and Analyses

Feed Intake and Composition.

Feed intake was weighed and recorded daily throughout the experiment to estimate DMI. There were no refusals during the experiment. Samples of each feed (200 g of hay and concentrates) were taken on 2 d in wk 4 and 5 of each period. One subsample was used to determine DM content (103°C for 24 h) and another subsample was stored at 4°C before being pooled at the end of the experiment. These pooled samples were ground using an Ultra Centrifugal Mill (0.75-mm sieve; Retsch GmbH, Haan, Germany) and analyzed for chemical composition.

Organic matter was determined by ashing at 550°C for 6 h (method 942.05; AOAC, 2005). Total N was analyzed by combustion according to the Dumas method (method 968.06; AOAC, 2005), and CP content was calculated as N × 6.25. Fiber (NDF and ADF) was determined by sequential procedures (Van Soest et al., 1991) after pretreatment with amylase and expressed exclusive of residual ash. Starch was analyzed using an enzymatic method (Faisant et al., 1995). The GE was analyzed by isoperibolic calorimetry (model C200; IKA, Staufen, Germany). Ether extract was determined after acid hydrolysis (method 954.02; AOAC, 2005), and fatty acid composition was determined by gas chromatography of methyl esters (method 969.33; AOAC, 2005).

Cows’ BW and Methemoglobinemia.

Cows were weighed at the beginning of the experiment and at the end of each experimental period. Levels of blood metHb were measured on all cows 3 h after morning feeding (1100 h) on the day before the start of the experiment (control blood) and then at d 3 and 5 (1% calcium nitrate in the diet), d 10 (2% calcium nitrate in the diet), and d 12, 17, 19, and 22 (3% calcium nitrate in the diet) of each experimental period for cows fed the NIT and LIN+NIT diets. Blood from cows fed the CON and LIN diets was not analyzed as we assumed that there was no risk of metHb. Blood (10 mL) was sampled from the jugular vein into K2EDTA collection tubes (Venosafe; Terumo, Guyancourt, France) and packed on ice for metHb content to be determined by spectrophotometry (UV-160; Shimadzu, Marne-La-Vallée, France) within 1 h at the nearest hospital (CHU Gabriel Montpied, Clermont-Ferrand, France; method of Kaplan, 1965). A metHb threshold value was set at 30% hemoglobin. Any animal meeting this cutoff would be removed from the experiment and treated with 1% methylene blue (U.S. Pharmacopeial Convention, 2008).

Diet Digestibility and N Balance.

Total tract digestibility and N balance were determined from total and separate collection of feces and urine for 6 d during wk 4 of each experimental period. To separate urine from feces, cows were fitted with flexible pipes (Doreau et al., 2014b) connected to a 30-L flask containing 500 mL of 3 M sulfuric acid to achieve a urine pH < 3 and thus avoid N volatilization. Feces and urine were removed once daily.

Each morning, after weighing and mixing of feces, a 1% fresh aliquot was used for DM determination (103°C for 24 h) and a 0.5% fresh aliquot was pooled across days for each animal and frozen (–20°C). At the end of the experiment, pooled samples were thawed, dried (60°C for 72 h), and ground (1-mm screen) to determine OM, N, NDF, and ADF content as previously described.

Each morning, after weighing urine, a 0.5% fresh aliquot was pooled across days for each animal and frozen (–20°C). At the end of the experiment, the N content of thawed urine was determined by the Kjeldahl method (method 2001.11; AOAC, 2005) as it was impossible to apply the Dumas method on fresh urine.

Rumen Environment.

Total rumen contents were sampled (approximately 200 g) from the ventral sac through the cannula before (0745 h) and 3 h after (1100 h) the morning feed on 2 nonconsecutive days (d 3 and 5) in wk 4 of each experimental period. The samples were strained through a polyester monofilament fabric (250 μm pore size) and filtrate was subsampled for subsequent analyses. For VFA analysis, 0.8 mL of filtrate was mixed with 0.5 mL of a 0.5 M HCl solution containing 2% (wt/vol) metaphosphoric acid and 0.4% (wt/vol) crotonic acid. For ammonia–nitrogen (NH3–N) analysis, 1 mL of filtrate was mixed with 0.1 mL of 5% orthophosphoric acid. For lactate and nitrate–nitrite concentrations analysis, 3 and 20 mL of filtrate, respectively, were collected without preservative (Sar et al., 2004). All these samples were stored at –20°C until analysis. For protozoa counts, 2 mL of filtrate was mixed with 2 mL of methyl green–formalin solution and stored away from direct light until counting.

Concentrations of VFA and NH3–N were analyzed by gas chromatography with a flame ionization detector and by colorimetry, respectively (Morgavi et al., 2008). Lactate concentrations were determined by colorimetry (dl-lactic acid; BioSentec, Auzeville-Tolosane, France). Nitrate and nitrite concentrations were analyzed by colorimetry (method EPA 353.2; EPA, 1993; SmartChem 200 [Unity Scientific, Brookfield, CT]; Laboratoire Vétérinaire et Biologique, Lempdes, France). Protozoa were counted by microscopy and categorized as either small (<100 μm) or large (>100 μm) entodiniomorphs or as holotrichs (Dasytricha or Isotricha; Williams and Coleman, 1992). Data for protozoa were log10–transformed before statistical analysis.

Ruminal pH and Dissolved H2 Concentration.

Rumen pH was monitored continuously over wk 5 using commercial boluses (eBolus; eCow, Exeter, UK). One day before measurement, the boluses were calibrated using buffer solutions (pH 4 and 7; HM Digital, Culver City, CA). One bolus per cow was immersed in the ventral sac of the rumen. Data were then recorded every 15 min during 6 full days, after which the boluses were removed. At the end of each experimental period, data were uploaded by telemetry to a digital tablet before being transferred to a computer.

The dynamics of dissolved H2 concentrations in the rumen were successively measured on each cow in wk 3 (1 d per cow) with a H2 sensor (H2–500; Unisense, Aarhus, Denmark). The electrode was connected to a microsensor monometer via a 10-m wire extension (Unisense), and the monometer was connected to a portable computer running Sensor Trace Basic software (version 3.1.3; Unisense). The sensor was polarized (800 mV) once in wk 3 (8 h before the start of measurement) and calibrated daily by immersion in a water bath at 39°C bubbling with a 80% H2/20% CO2 gas mixture. The sensor and wire extension were protected using a custom-made plastic cap and tube (Fig. 1). The system was ballasted with a 1-kg weight and introduced into the cow’s ventral sac through the cannula at 30 min before the morning feed (i.e., 0730 h). The setup was fitted taking care to avoid gas and liquid leakage from the rumen. Dissolved H2 concentration readings were recorded every second for 6 h after the morning feeding. For an easier use of the sensor, it was essential to remove it when the rumen was not full (i.e., before the afternoon feeding).

Figure 1.
Figure 1.

Use of H2 sensor (Unisense, Aarhus, Denmark). A) Overall setup with sensor, monometer, and computer. B) Protection cap of the sensor.

 

Methane and CO2 Emissions.

In wk 5, animals were placed in open circuit respiration chambers (1 animal/chamber) for 4 consecutive days. Individual total CH4 and CO2 emissions were measured continuously from d 1 (0730 h) to 5 (0730 h).

Each chamber was 2.2 m high, 3.6 m long, and 2.1 m wide, giving a volume of 16.6 m3. The chambers were made of steel with clear polycarbonate walls allowing sight contact between animals and with the farm personnel. Chambers had front and rear doors, with the front doors used for animal feeding and the rear doors used to enter the animals and to remove feces and urine collected in a wheeled recovery box. Front and rear doors were never simultaneously opened to avoid an air stream into the chamber. The feces and urine recovery boxes were removed each morning and immediately replaced with new ones to minimize chamber opening time (5 min per chamber, on average). When rear doors were closed, front doors were opened (5 min per chamber, on average) for morning (1 portion of hay at 0800 h and 3 portions of concentrates at 0800, 0830, and 0930 h) and afternoon (1 portion of hay at 1600 h and 2 portions of concentrates at 1600 and 1630 h) feeding.

The chambers operated at a slight negative pressure, with an air flow oscillating between 700 and 800 m3/h (approximately 45 air changes per h). Airflow entered the chamber through an aperture at the bottom of the rear door (0.42 m2) and exited through an exhaust duct situated at the top of the chamber, over the head of the animal. Airflow in the exhaust duct of each chamber was continuously measured (CP300 pressure transmitter; KIMO, Montpon-Ménestérol, France) and recorded with 1 data point every 5 min (KT-210-AO datalogger; KIMO).

Concentration of gases in the barn and in the 4 chambers was alternatively analyzed at a 0.1-Hz sample frequency for 5 min every 25 min using an infrared detector (Ultramat 6; Siemens, Karlsruhe, Germany) and recorded (Nanodac Invensys recorder/controller; Eurotherm Automation SAS, Dardilly, France). The detector was manually calibrated the day before each measurement period using pure N2 and a mixture of CH4 (650 mg/kg) and CO2 (700 mg/kg) in N2. Missing data between 2 measurement intervals were recovered by linear regression. Chamber doors were never opened during gas analysis, so no data was deleted. Real-time gas emissions in a chamber were calculated by the difference between chamber and ambient gas concentrations multiplied by the airflow corrected for temperature, relative humidity, and pressure according to the Wexler equation (Pinares-Patiño et al., 2012).

Statistical Analyses

Except for metHb, data were analyzed using the MIXED procedure of SAS (version 9.2; SAS Inst. Inc., Cary, NC). Gaseous emissions (CH4 and CO2) and rumen fermentation parameters measured during several days (n = 4 and 2 d, respectively) were averaged per period before being included in the statistical analyses. The model included the random effect of cow (n = 4) and fixed effects of period (n = 4), nitrate (CON and LIN versus NIT and LIN+NIT), linseed (CON and NIT versus LIN and LIN+NIT), and the interaction nitrate × linseed. Rumen fermentation data obtained before and after feeding (VFA, NH3–N, lactate, protozoa, nitrate, and nitrite concentrations) were analyzed using the same model and for the 2 sampling hours separately. Continuous measurements of ruminal pH, dissolved H2 concentrations, and CH4 emissions were analyzed as repeated measures. Several covariance structures were compared, and compound symmetry was selected because it resulted in the lowest values for the Akaike’s information criteria. The model included the fixed effects of period, hour, nitrate, linseed, nitrate × linseed, and the interactions between hour and dietary treatments (linseed × hour, nitrate × hour, and linseed × nitrate × hour). Differences among treatments were tested using the PDIFF option. Data were considered significant at P < 0.05, and trends were discussed at 0.05 < P ≤ 0.1. Least squares means are reported throughout.


RESULTS

Body Weight and Blood Methemoglobin

Animals gained, on average, 26.5 kg per experimental period, with a final BW at the end of the trial of 762 ± 47 kg. For diets containing nitrate (NIT and LIN+NIT), blood metHb gradually increased the first 12 d of adaptation period, but no animal exceeded 26.3% metHb (Fig. 2).

Figure 2.
Figure 2.

Box plot of blood methemoglobinemia levels of nonlactating cows fed diets containing 3% calcium nitrate with or without 4% linseed oil (n = 8). The box represents the quartiles with the median at the center and the vertical lines represent the maximum and minimum value within 1.5 interquartile range of the greater and lesser quartile, respectively. Values greater than 1.5 interquartile range are considered outliers and are identified with the symbol ǂ. Blood was analyzed during the 3-wk adaptation period; the arrow indicates the start of the measurement period.

 

Methane and CO2 Emissions

Dry matter intake of cows while in chambers was the same as outside, showing the absence of stress of animals and that CH4 determination in our experimental conditions accurately reflected emissions throughout the trial. Methane production was different among diets irrespective of the unit of expression (Table 2; P < 0.01). Compared with the CON diet, CH4 (g/d) was 18, 23, and 33% less for the LIN, NIT, and LIN+NIT diets, respectively. An additive CH4–mitigating effect between linseed and nitrate (linseed × nitrate, P > 0.05) was observed when CH4 was expressed as a function of DMI, digested DM, or digested OM or as a percentage of GE intake. When expressed per kilogram of digested NDF, CH4 emissions from cows fed nitrate-containing diets (NIT and LIN+NIT) were similar and less (P = 0.01) than emissions from cows fed other diets, showing an absence of additive effect between nitrate and linseed.


View Full Table | Close Full ViewTable 2.

Methane and carbon dioxide emissions of nonlactating cows fed diets containing linseed oil and calcium nitrate alone or in association (n = 4)1

 
Item Diet2
SEM P-value3
CON NIT LIN LIN+NIT Nitrate Linseed Linseed × nitrate
DM intake, kg/d 12.4 12.3 12.3 12.2 0.59 0.22 0.35 0.86
CH4 emissions
    g CH4/d 308.6 238.1 252.7 206.8 9.61 <0.01 <0.01 0.08
    g CH4/kg DM intake 25.0 19.4 20.7 17.0 0.70 <0.01 <0.01 0.18
    g CH4/kg digested DM 39.3 30.3 32.4 27.0 1.18 <0.01 <0.01 0.14
    g CH4/kg digested OM 36.8 28.3 30.3 25.1 1.06 <0.01 <0.01 0.12
    g CH4/kg digested NDF 55.9 43.1 47.1 43.1 2.42 0.01 0.06 0.07
    % of GE intake 7.2 5.8 5.6 4.8 0.20 <0.01 <0.01 0.24
CO2 emissions
    g CO2/d 9,191 9,323 8,988 8,789 562.1 0.84 0.06 0.35
    g CO2/kg DM intake 745 757 732 721 28.1 0.98 0.19 0.49
1Data were collected during 4 consecutive days in wk 5.
2CON = control; NIT = CON with 3% calcium nitrate; LIN = CON with 4% linseed oil; LIN+NIT = CON with 4% linseed oil plus 3% calcium nitrate.
3Linseed represents the main effect of linseed (CON and NIT versus LIN and LIN+NIT); Nitrate represents the main effect of nitrate (CON and LIN versus NIT and LIN+NIT); Linseed × nitrate represents the interaction between main effects of linseed and nitrate.

Dietary effect on the daily pattern of CH4 emissions is presented in Fig. 3. For the CON, 2 peaks of CH4 production were observed at around 2 h after feeding, with the largest peak after the morning feeding. The CH4 emissions pattern of the LIN diet was similar to the CON but emissions of the LIN diet were consistently lower throughout the day. In contrast to the CON, with the NIT and LIN+NIT diets, the peaks were not observed and CH4 emissions increased at 3 h after feeding. Contrary to CH4, CO2 emissions (g/d or g/kg DMI) were not affected by dietary treatments.

Figure 3.
Figure 3.

Daily CH4 production pattern of nonlactating cows fed diets containing linseed oil and calcium nitrate alone or in association (n = 4). Errors bars indicate SD. Treatments consisted of a control (CON) diet, the CON with 3% calcium nitrate (NIT), the CON with 4% linseed oil (LIN), and the CON with 4% linseed oil plus 3% calcium nitrate (LIN+NIT). The arrows indicate time of feeding. Symbols indicate hourly statistical comparison among treatments (linseed represents the main effect of linseed and the CON and NIT diets versus the LIN and LIN+NIT diets; nitrate represents the main effect of nitrate and the CON and LIN diets versus the NIT and LIN+NIT diets; linseed × nitrate represents the interaction between main effects of linseed and nitrate; ns = P > 0.05; *P < 0.05; **P < 0.01; ***P < 0.001).

 

Rumen Environment

Mean rumen pH was greater for nitrate-containing diets (NIT and LIN+NIT) compared with other diets (Table 3; +0.23 units, on average; P = 0.03). Compared with the CON diet, the LIN+NIT diet showed significantly greater pH values during daytime, starting 3 h after the morning feeding (Fig. 4). Mean dissolved H2 concentrations in the rumen tended (P = 0.07) to be greater for diets including nitrate compared with other diets (+89%). The H2 concentration was consistently low up to 6 h after feeding for the CON and LIN diets (3.8 μM; Fig. 5) but showed a significant jump as early as 1 h after feeding nitrate (NIT and LIN+NIT). Hydrogen concentrations started to decrease 2 h after feeding for the LIN+NIT diet and at 3 h after feeding for the NIT diet. Compared with the CON diet, H2 concentrations were, on average, 5.9 and 12.6 times greater for the LIN+NIT and NIT diets, respectively.


View Full Table | Close Full ViewTable 3.

Rumen fermentation characteristics of nonlactating cows fed diets containing linseed oil and calcium nitrate alone or in association (n = 4)1

 
Item Time after feeding, h Diet2
SEM P-value3
CON NIT LIN LIN+NIT Nitrate Linseed Linseed × nitrate
Total VFA, mM 0 73.8 72.7 69.4 71.4 6.42 0.93 0.56 0.75
3 111.9 102.6 102.6 107.7 6.52 0.74 0.74 0.28
VFA composition, mol/100 mol
    Acetate (A) 0 70.9 69.5 69.5 69.6 1.00 0.53 0.53 0.43
3 70.2 73.4 67.0 73.1 1.07 0.01 0.15 0.23
    Propionate (P) 0 15.0 15.0 17.6 16.0 0.59 0.20 0.02 0.23
3 15.8 14.8 19.4 15.4 0.95 0.01 0.02 0.06
    Butyrate (B) 0 10.3 11.4 9.0 10.4 0.71 0.08 0.11 0.81
3 10.4 8.7 10.1 8.4 1.20 0.19 0.82 0.98
    Minor VFA4 0 3.79 4.15 3.58 3.94 0.321 0.31 0.54 1.00
3 3.77 3.08 3.54 3.10 0.197 0.01 0.46 0.37
    A:P 0 4.74 4.68 3.97 4.41 0.221 0.39 0.04 0.26
3 4.48 5.03 3.52 4.79 0.233 <0.01 0.01 0.09
    (A+B):P 0 5.43 5.44 4.48 5.06 0.230 0.20 0.02 0.22
3 5.14 5.62 4.07 5.34 0.278 <0.01 0.01 0.08
    NH3–N, mM 0 5.84 6.79 4.87 6.68 0.555 0.04 0.34 0.44
3 15.11 14.34 16.15 14.35 0.932 0.22 0.59 0.60
    Total lactate, mM 0 0.56 0.65 0.57 0.65 0.039 0.06 0.81 0.97
3 0.83 0.71 0.78 0.68 0.107 0.24 0.69 0.91
    Nitrate, mg/L 0 <LoQ5 <LoQ <LoQ <LoQ
3 <LoQ <LoQ <LoQ <LoQ
    Nitrite, mg/L 0 0.12 0.58 0.12 0.83 0.246 0.07 0.66 0.66
3 0.24 0.45 0.24 0.37 0.168 0.32 0.79 0.79
    pH Mean 6.20 6.30 6.07 6.42 0.101 0.03 0.94 0.15
    Hydrogen, μM Mean 3.58 45.28 4.03 21.00 14.097 0.07 0.41 0.39
1Data were collected during 2 nonconsecutive days in wk 4.
2CON = control; NIT = CON with 3% calcium nitrate; LIN = CON with 4% linseed oil; LIN+NIT = CON with 4% linseed oil plus 3% calcium nitrate.
3Linseed represents the main effect of linseed (CON and NIT versus LIN and LIN+NIT); Nitrate represents the main effect of nitrate (CON and LIN versus NIT and LIN+NIT); Linseed × nitrate represents the interaction between main effects of linseed and nitrate.
4Minor VFA is the sum of isobutyrate, isovalerate, valerate, and caproate.
5LoQ = limit of quantification: 13.3 mg/L or 0.22 mM.
Figure 4.
Figure 4.

Daily pattern of rumen pH of nonlactating cows fed diets containing linseed oil and calcium nitrate alone or in association (n = 4). Errors bars indicate SD. Treatments consisted of control (CON) diet, the CON with 3% calcium nitrate (NIT), the CON with 4% linseed oil (LIN), and the CON with 4% linseed oil plus 3% calcium nitrate (LIN+NIT). The arrows indicate time of feeding. Symbols indicate hourly statistical comparison among treatments (linseed represents the main effect of linseed and the CON and NIT diets versus the LIN and LIN+NIT diets; nitrate represents the main effect of nitrate and the CON and LIN diets versus the NIT and LIN+NIT diets; linseed × nitrate represents the interaction between main effects of linseed and nitrate; ns = P > 0.05; *P < 0.05; **P < 0.01; ***P < 0.001).

 
Figure 5.
Figure 5.

Rumen dissolved hydrogen concentrations up to 6 h after feeding nonlactating cows with diets containing linseed oil and calcium nitrate alone or in association (n = 4). Treatments consisted of control (CON) diet, the CON with 3% calcium nitrate (NIT), the CON with 4% linseed oil (LIN) and the CON with 4% linseed oil plus 3% calcium nitrate (LIN+NIT). The arrow indicates time of morning feeding. Symbols indicate hourly statistical comparison among treatments (linseed represents the main effect of linseed and the CON and NIT diets versus the LIN and LIN+NIT diets; nitrate represents the main effect of nitrate and the CON and LIN diets versus the NIT and LIN+NIT diets; linseed × nitrate represents the interaction between main effects of linseed and nitrate; ns = P > 0.05; *P < 0.05; **P < 0.01; ***P < 0.001).

 

Concentrations of total VFA were similar among diets before and after feeding. Linseed-containing diets increased propionate proportions before and after feeding (P = 0.02), leading to lower acetate:propionate and (acetate + butyrate):propionate ratios compared with the other diets. Nitrate-containing diets modified VFA profiles only after feeding (P = 0.01), with greater acetate and lower propionate proportions, inducing greater acetate:propionate and (acetate + butyrate):propionate ratios compared with the other diets. Nitrate-containing diets increased NH3–N (+20%; P = 0.04) concentrations before feeding. Nitrate concentrations in the rumen were less than the limit of quantification (13.3 mg/L or 0.22 mM).

Before feeding, the LIN diet decreased (P = 0.03) total protozoa concentration in the rumen whereas the NIT diet did not affect this population. The toxic effect of linseed toward protozoa was not observed when associated with nitrate (P = 0.02; Table 4). Compared with the CON diet, the LIN diet reduced total protozoa concentration specifically by acting on entodiniomorphs (–52%). Inversely, the NIT diet tended to increase (P = 0.09) large entodiniomorphs and increased (P = 0.02) Isotricha before feeding.


View Full Table | Close Full ViewTable 4.

Rumen protozoa populations of nonlactating cows fed diets containing linseed oil and calcium nitrate alone or in association (n = 4)1

 
Item Time after feeding, h Diet2
SEM P-value3
CON NIT LIN LIN+NIT Nitrate Linseed Linseed × nitrate
Total protozoa, log10/mL 0 5.87 5.71 5.55 5.73 0.060 0.91 0.03 0.02
3 5.71 5.49 5.37 5.58 0.080 0.95 0.14 0.03
Entodiniomorphs, log10/mL
    Small (<100 μm) 0 5.86 5.68 5.54 5.71 0.057 0.95 0.03 0.02
3 5.69 5.46 5.36 5.56 0.080 0.86 0.16 0.03
    Large (>100 μm) 0 4.09 4.18 3.66 4.01 0.110 0.09 0.03 0.29
3 3.97 4.00 3.62 3.97 0.109 0.14 0.13 0.18
Holotrichs, log10/mL
    Dasytricha (<100 μm) 0 3.51 3.65 2.67 3.58 0.497 0.29 0.35 0.42
3 3.49 3.78 2.75 3.69 0.521 0.23 0.40 0.51
    Isotricha (>100 μm) 0 1.90 3.19 2.29 3.11 0.484 0.02 0.63 0.47
3 2.88 3.25 2.53 2.89 0.494 0.42 0.42 1.00
1Data were collected during 2 nonconsecutive days in wk 4.
2CON = control; NIT = CON with 3% calcium nitrate; LIN = CON with 4% linseed oil; LIN+NIT = CON with 4% linseed oil plus 3% calcium nitrate.
3Linseed represents the main effect of linseed (CON and NIT versus LIN and LIN+NIT); Nitrate represents the main effect of nitrate (CON and LIN versus NIT and LIN+NIT); Linseed × nitrate represents the interaction between main effects of linseed and nitrate.

Diet Digestibility and N Balance

Daily DM and OM intake were not affected by treatments and averaged 12.4 kg DMI/d (Table 5). Fiber intake was reduced with linseed-containing diets (LIN and LIN+NIT; P < 0.01) compared with the other diets. Total tract digestibility of DM and OM was not affected by diets, and linseed supplemented alone or in association with nitrate tended to reduce (P < 0.10) fiber digestibility.


View Full Table | Close Full ViewTable 5.

Daily nutrient intake and total tract digestibility of nonlactating cows fed diets containing linseed oil and calcium nitrate alone or in association (n = 4)1

 
Item Diet2
SEM P-value3
CON NIT LIN LIN+NIT Nitrate Linseed Linseed × nitrate
Daily nutrient intake, kg/d
    DM 12.4 12.3 12.5 12.3 0.55 0.09 0.73 0.51
    OM 11.4 11.3 11.4 11.3 0.51 0.14 0.74 0.45
    NDF 5.0 5.0 4.8 4.7 0.22 0.08 <0.01 0.41
    ADF 2.9 2.9 2.8 2.7 0.13 0.05 <0.01 0.76
GE intake, MJ/d 216.8 205.1 228.5 217.2 9.67 <0.01 <0.01 0.88
Total tract digestibility, %
    DM 63.7 64.1 64.0 63.3 0.77 0.85 0.65 0.43
    OM 68.1 68.5 68.3 67.9 0.64 0.98 0.76 0.50
    NDF 44.8 45.2 44.2 40.1 1.58 0.22 0.07 0.14
    ADF 44.5 45.1 42.9 38.4 2.11 0.31 0.06 0.20
1Data were collected during 6 consecutive days in wk 4.
2CON = control; NIT = CON with 3% calcium nitrate; LIN = CON with 4% linseed oil; LIN+NIT = CON with 4% linseed oil plus 3% calcium nitrate.
3Linseed represents the main effect of linseed (CON and NIT versus LIN and LIN+NIT); Nitrate represents the main effect of nitrate (CON and LIN versus NIT and LIN+NIT); Linseed × nitrate represents the interaction between main effects of linseed and nitrate.

Total N losses (% of N intake) were greater for diets including linseed compared with other diets (P = 0.03), leading to less N retention for the LIN and LIN+NIT diets (P = 0.03; Table 6). This was not related to differences in daily fecal N losses between diets but to numerically greater urinary N losses with linseed-containing diets (P = 0.08).


View Full Table | Close Full ViewTable 6.

Nitrogen balance of nonlactating cows fed diets containing linseed oil and calcium nitrate alone or in association (n = 4)1

 
Item Diet2
SEM P-value3
CON NIT LIN LIN+NIT Nitrate Linseed Linseed × nitrate
N intake, g/d 252.5 242.5 242.5 227.5 11.59 <0.01 <0.01 0.13
Fecal N losses
    g/d 101.6 95.4 94.8 96.8 4.28 0.47 0.37 0.18
    As % of N intake 40.1 39.4 39.5 42.5 1.18 0.27 0.25 0.10
Urinary N losses
    g/d 133.1 117.7 135.8 120.2 6.13 0.02 0.61 0.99
    As % of N intake 52.5 48.8 56.5 52.7 1.82 0.09 0.08 0.97
Total N losses
    g/d 234.7 213.0 230.6 217.0 9.12 0.01 0.99 0.45
    As % of N intake 92.6 88.3 96.0 95.2 2.25 0.20 0.03 0.35
N retained
    g/d 18.5 28.3 10.7 11.8 5.82 0.26 0.03 0.36
    As % of N intake 7.4 11.8 4.0 4.8 2.25 0.20 0.03 0.35
1Data were collected during 6 consecutive days in wk 4.
2CON = control; NIT = CON with 3% calcium nitrate; LIN = CON with 4% linseed oil; LIN+NIT = CON with 4% linseed oil plus 3% calcium nitrate.
3Linseed represents the main effect of linseed (CON and NIT versus LIN and LIN+NIT); Nitrate represents the main effect of nitrate (CON and LIN versus NIT and LIN+NIT); Linseed × nitrate represents the interaction between main effects of linseed and nitrate.


DISCUSSION

Nitrate Toxicity

In the rumen, nitrate is converted to nitrite and then ammonia. Although nitrate is nontoxic, nitrite can be poisonous for the animal. If nitrite accumulates in the rumen, it can pass through the rumen wall into the blood and convert hemoglobin to metHb, which cannot then transport oxygen to the tissues (Lewis, 1951). The level of blood metHb determines the severity of symptoms, which are brown mucous membrane discoloration, depressed feed intake and animal performance, and even coma and death in extreme cases (Bruning-Fann and Kaneene, 1993). Throughout this experiment, animals were unaffected by nitrate supplementation, as shown by the BW gain, the constant intake, and the low rumen concentrations of nitrate and nitrite and blood metHb. Nitrate feeding requires precise management of its distribution and careful control of animal health status. To deal with these issues, the use of slow-release encapsulated nitrate was shown to be effective at mitigating CH4 emissions of lambs (3.4% nitrate in DM, inducing a 9.7% CH4 reduction per percent added nitrate; El-Zaiat et al., 2014) or beef heifers (2.3% nitrate in DM, inducing a 8.0% CH4 reduction per percent added nitrate; Lee et al., 2014a,b) without raising blood metHb levels.

Methane Emissions

Supplying 2.6% added fat from linseed oil reduced CH4 (g/kg DMI) by 17%, corresponding to a 6.5% reduction in CH4 per percentage unit of added lipids from linseed. This result is in the range of previous meta-analysis data reporting that CH4 (g/kg DMI) is reduced by 4.4% per percentage unit of fat (irrespective of lipid source) added to diet (Grainger and Beauchemin, 2011) or by 5.6% per percentage unit of linolenic acid from linseed (Doreau et al., 2011). Conversely, Veneman et al. (2013) did not explain the absence of any CH4–mitigative effect (g/kg DMI and g/kg milk) of a similar level of linseed oil in lactating cows.

Nitrate fed alone reduced CH4 (g/kg DMI) by 22%, corresponding to a 9.8% reduction per percentage unit of nitrate fed. This result is in the range of previous experimental data reporting a CH4 (g/kg DMI) reduction of between 7.9 and 12.2% per percentage unit of added nitrate in the diet of sheep (Nolan et al., 2010; Van Zijderveld et al., 2010) or cattle (Van Zijderveld et al., 2011; Hulshof et al., 2012; Veneman et al., 2013). The CH4–mitigating effect of nitrate is consequently greatly repeatable whatever the diet and the ruminant species.

The association of nitrate and linseed oil reduced CH4 (g/kg DMI) by 32%. This result showed for the first time that there is a positive and additive effect between nitrate and linseed oil on methanogenesis. Theoretically, as these dietary strategies have different mechanisms of action, CH4 reduction should reach 39% for a fully additive effect. Several reasons may explain the difference between theoretical and observed CH4 reduction. First, we suggest that linseed reduced H2 production and that nitrate acted only on this reduced H2 pool. Then, according to stoichiometry and considering that control CH4 emissions is equal to 100, CH4 emissions corrected for the CH4–mitigating effect of the LIN diet (17%) would be 100 – 100 × 0.17 = 83. These CH4 emissions corrected for the CH4–mitigating effect of the NIT diet (22%) would be 83 – 83 × 0.22 = 65. In total, this corresponds to an expected CH4 reduction of 35% with the LIN+NIT diet, which is close to the observed level of CH4 reduction. In addition, the LIN+NIT diet had less fatty acid content compared with the LIN diet, which may be linked to unnoticed pellets manufacturing issues. Knowing that 1% added fat from linseed reduced CH4 by 6.5%, the difference in fatty acid content between the LIN+NIT (1.0% added fat) and LIN (2.6% added fat) diets corresponded to a CH4 mitigation potential of 10.4%, suggesting a fully additive effect between linseed oil and nitrate. At least, the formation of calcium salts via the reaction between lipids and soluble calcium from calcium nitrate may reduce the additive effects of the LIN+NIT diet (Keyser et al., 1985).

The association of nitrate and linseed oil appears interesting: this same level of CH4 reduction with linseed oil or nitrate fed individually could not be achieved without greater risks of metHb for nitrate or lower diet digestibility for linseed oil. Other kinds of antimethanogenic combinations have shown various interactions. Tea saponin and soybean oil reduced CH4 (g/kg DMI) from lambs by 27 and 14%, respectively, when distributed alone and by 19% when fed in association (Mao et al., 2010). In lambs, CH4 (g/kg DMI) was reduced by 25% by chestnut tannin, 14% by coconut oil, and 33% by the association chestnut tannin plus coconut oil (Liu et al., 2011). A fully additive effect was observed with 2 H2–sink products fed to lambs, with a CH4 reduction of 32% with nitrate, 16% with sulfate, and 47% with nitrate plus sulfate (Van Zijderveld et al., 2010).

Rumen Environment

The reduction in CH4 emissions observed in this trial did not cause a rumen dysfunction, as VFA concentration was not affected by diet and pH was only marginally modified. Two factors may explain the CH4–mitigating effect of linseed oil. On the one hand, lipids from linseed oil reduced the rumen concentration of protozoa, although not as strongly as in previous experiments testing similar levels of lipids (–82% in a silage-based diet [Chung et al., 2011] and –84% in a concentrate-rich hay-based diet [Ueda et al., 2003]). The antiprotozoal effect of linseed combined with nitrate was less evident, probably because of the lower fat content in the LIN+NIT diet compared with the LIN diet. Protozoa are known to be important H2 producers via their hydrogenosomes (Morgavi et al., 2012) and their reduction is often associated with a decrease in methanogenesis (Guyader et al., 2014). Consequently, in this study, linseed supplementation reduced H2 production, but as dissolved H2 concentrations in the rumen were not affected by lipids, we assume that methanogens also used less H2. On the other hand, linseed oil increased propionate proportion, which is a H2–consuming pathway competing with methanogenesis (Newbold et al., 2005). Most literature reports do not show an effect of linseed on rumen VFA composition (Chung et al., 2011; Doreau et al., 2009; Martin et al., 2011). To a minor extent, H2 may have been consumed during PUFA biohydrogenation, but this pathway would deviate only 1 to 2.6% of ruminal H2 (Czerkawski, 1986). The lower CH4 emissions throughout the day from LIN cows compared with the CON cows indicated that linseed oil continuously modified rumen fermentation and microbial parameters.

Nitrate is an electron acceptor in several anaerobic environments. Its CH4–mitigating effect is assumed to be related to a reduction of H2 availability for methanogens due to its reduction to nitrite and ammonia (Ungerfeld and Kohn, 2006). To our knowledge, ours is the first study to report a postfeeding pattern of dissolved H2 concentrations in the rumen. The CON and LIN diets presented stable and low rumen H2 concentrations (3.8 μM, on average), which are in the range of concentrations (0.1 to 50 μM) given by a literature review (Janssen, 2010). However, adding nitrate to the diet with or without linseed oil induced a peak in rumen dissolved H2 concentrations up to 2 h after feeding (up to 88 μM, on average), coinciding with a drop in CH4 emissions and a rise of gaseous H2 (measured in wk 5 of the last 2 experimental periods; data not shown) as reported by Van Zijderveld et al. (2011). In presence of nitrate, the excess of dissolved H2 further released in belched gas means that H2 was produced at a greater rate than it was utilized. This may result from a toxic effect of nitrate (Van Zijderveld et al., 2010) or nitrite (Iwamoto et al., 2001) on H2 users such as methanogens. This putative action is transient, lasting for 3 h after feeding, as shown by the increase in CH4 emissions from nitrate-fed cows up to levels similar to control-diet-fed cows.

Diet Digestibility and N Balance

Supplying diets with linseed oil (2.6% added fat) did not affect total tract digestibility of DM and OM but tended to reduce total tract fiber digestibility to a same extent when fed alone or in association with nitrate. This result is not consistent with a previous study on lambs supplemented with crude linseed (2.4% added fat; Machmüller et al., 2000). These different results may be explained by the forms of linseed that affect availability of lipids supply: linseed oil would have a more negative effect on total tract digestibility than extruded and crude linseed (Martin et al., 2008). Adding 3% calcium nitrate as a substitute for urea did not reduce total tract digestibility, confirming previous experiments on sheep fed hay and 4% potassium nitrate (Nolan et al., 2010) and on dairy cows fed maize silage and 2.8% calcium nitrate (Van Zijderveld et al., 2011). Nitrate affected neither N retention nor the distribution of N losses between urine and feces. Similar results were obtained with dairy cows (2.6% nitrate; Van Zijderveld et al., 2011), steers (2.3% nitrate; Lee et al., 2014a), and lambs (2.3% nitrate; Li et al., 2012) fed isonitrogenous diets, showing that nitrate can substitute urea as a source of NPN.

The association of nitrate and linseed oil is an efficient strategy to decrease CH4 yields in nonlactating cows without altering diet digestibility. Linseed oil supplementation reduced CH4 emissions throughout the day, whereas nitrate had a transient but marked action from when fed up to 3 h after feeding. Methane production was further reduced when both linseed and nitrate were fed in association. Linseed oil reduced H2 producers such as protozoa, whereas nitrate acted as a H2 sink and may have inhibited rumen H2 users, as suggested by the rise of dissolved H2 concentrations with this dietary treatment. Further work to characterize the quantity, activity, and diversity of rumen microbiota should clarify the mechanisms behind the effects of these dietary treatments. In addition, it will be necessary to assess the long-term CH4–mitigative effect of linseed oil associated with nitrate on ruminants in production. Finally, additional research is required to check the absence of nitrate residues in end products from nitrate-fed ruminants.

 

References

Footnotes


Comments
Be the first to comment.



Please log in to post a comment.
*Society members, certified professionals, and authors are permitted to comment.