Methane output and lactation response in Holstein cattle with monensin or unsaturated fat added to the diet.

F D Sauer, V Fellner, R Kinsman, J K Kramer, H A Jackson, A J Lee and S Chen J ANIM SCI 1998, 76:906-914.

The online version of this article, along with updated information and services, is located on the World Wide Web at: http://www.journalofanimalscience.org/content/76/3/906

www.asas.org

Downloaded from www.journalofanimalscience.org by guest on February 15, 2014

Methane Output and Lactation Response in Holstein Cattle with Monensin or Unsaturated Fat Added to the Diet1
F. D. Sauer*,2, V. Fellner*, R. Kinsman*, J.K.G. Kramer*, H. A. Jackson*,3, A. J. Lee*,3, and S. Chen
*Centre for Food and Animal Research, Research Branch, Agriculture and Agri-Food Canada, and Statistics Canada, Ottawa, ON, Canada K1A 0C6

ABSTRACT: We measured effects of continuous vs twice-daily feeding, the addition of unsaturated fat to the diet, and monensin on milk production, milk composition, feed intake, and CO2-methane production in four experiments in a herd of 88 to 109 milking Holsteins. Methane and CO2 production increased with twice-daily feeding, but the CO2:CH4 ratio remained unchanged. Soybean oil did not affect the milkfat percentages, but fatty acid composition was changed. All saturated fatty acids up to and including 16:0 decreased ( P < .01), whereas 18:0 and trans 18:1 increased ( P < .001). The 18:2 conjugated dienes also increased ( P < .01) when the cows were fed soybean oil. Monensin addition to the diet at 24 ppm decreased methane production ( P < .01); the CO2:CH4 ratios reached 15, milk production increased ( P < .01), and milkfat percentage and total milkfat output decreased ( P < .01), as did feed consumption, compared with

cows fed diets without monensin ( P < .05). Milk fatty acid composition showed evidence of depressed ruminal biohydrogenation: saturated fatty acids ( P < .05) decreased and 18:1 increased ( P < .001); most of the increase was seen in the trans 18:1 isomer. As with soybean oil feeding, addition of monensin also increased ( P < .05) the concentration of conjugated dienes. The monensin feeding trial was repeated 161 d later with 88 cows, of which 67 received monensin in the diet in the first trial and 21 cows were newly freshened and had never received monensin. Methane production again decreased ( P < .05), but this time the CO2:CH4 ratio did not change and all other monensin-related effects were absent. The ruminal microflora in the cows that had previously received monensin seemed to have undergone some adaptive changes and no longer responded as before.

Key Words: Methane, Carbon Dioxide, Ionophores, Gas Production, Dairy Cows
1998 American Society of Animal Science. All rights reserved.

J. Anim. Sci. 1998. 76:906914

Introduction
Methane production is an unavoidable by-product of ruminal fermentation and represents a significant energy loss to the host animal. Energy losses as methane can range from 2 to 15% of GE intake (Holter and Young, 1992; Johnson and Johnson, 1995). The type of carbohydrate, the addition of dietary fat, the quantity of feed ingested, and the processing of forages all affect methane output (Johnson and Johnson, 1995). Wilkerson et al. (1995) tested predictive equations for methane production in adult Holstein cows using a data file compiled from observations in their Energy Metabolism Unit. They concluded that the equation of Moe and Tyrrell

no. 2422. whom correspondence should be addressed. 3Retired. Received March 27, 1997. Accepted October 29, 1997.
2To

1Contribution

(1979) was the most accurate. The prediction equations are limited because certain assumptions must be made and conversion factors are required (Holter and Young, 1992). Therefore, methane emissions by dairy cattle may be over- or underestimated. Currently, a new technique is being tested to continuously measure the methane output from ruminants with sulfur hexafluoride ( SF6) (Johnson et al., 1994; Marik and Levin, 1996). Kinsman et al. (1995) measured methane-CO2 emissions from a 118-cow herd of lactating Holstein cows with infrared gas analyzers. Certainly, estimates of global methane emissions from domestic ruminants (Crutzen et al., 1986; Johnson and Ward, 1996) will become more reliable after data bases are established. In the present study, methane and CO2 emissions were monitored continuously in a herd of 88 to 109 lactating Holstein cows kept in a tie stall barn under a practical management routine. The gas sampling and monitoring techniques were as described previously (Kinsman et al., 1995). Our objectives were to measure methane-CO2 output by lactating Holstein 906

Downloaded from www.journalofanimalscience.org by guest on February 15, 2014

METHANE OUTPUT BY LACTATING COWS

907

Table 1. Percentage composition (DM) of concentratesa for Experiments 1, 3, and 4

Ingredient Corn Soybean meal Roasted soybeans Blood meal Barley Mono/dicalcium phosphate Urea Limestone Dynamateb Magnesium oxide Cobalt iodized salt KCI Trace mineral premixc Vitamin premixd TMRa concentrate Top dress concentrate 50.8 28.1 13.0 1.5 .4 2.45 .5 .5 1.4 .35 .9 .1 76.7 5.6 6.5 7.8 .9 .6 .5 .6 .5 .2 .1

Table 2. Percentage composition (DM) of concentratesa for Experiment 2

TMRa concentrate for control period 10.0 25.95 1.4 52.0 3.0 .4 4.8 .5 .5 .35 .9 .2 TMR concentrate for experimental period 10.0 25.95 1.4 42.0 3.0 .4 4.8 .5 .5 .35 .9 .2 10.0

Ingredient Corn Soybean meal Cobalt iodized salt Barley Mono/dicalcium phosphate Urea Limestone Dynamateb Magnesium oxide KCI Trace mineral premixc Vitamin premixd Soybean oil

aThe TMR consisted of corn silage, 30%; alfalfa haylage, 26%; hay, 9% and TMR concentrate, 35%, on a DM basis. bTrade name for potassium-magnesium sulfate. cTrace mineral mixture for TMR concentrate contained: copper sulfate, .58%; zinc sulfate, 1.41%; manganese sulfate, 1.29%; selenium, .02%; anhydrous sodium sulfate, 24.6%; corn, 46.3%, and the top dress concentrate contained: copper sulfate, 1.96%; zinc sulfate, 3.88%; manganese sulfate, 3.92%; magnesium oxide, 23.04%; selenium, .02%; ground corn, 37.2%. dVitamin premix contained: vitamin A, 30,000 IU/g, 16%; vitamin D3, 25,000 IU/g, 16%; vitamin E, 500 IU/g, 11.6%; corn 56.4%.

.58%; zinc sulfate, 1.41%; manganese sulfate, 1.29%; selenium, .02%; anhydrous sodium sulfate, 24.6%; corn, 46.3%. dVitamin premix contained: vitamin A, 30,000 IU/g, 16%; vitamin D3, 25,000 IU/g, 16%; vitamin E, 500 IU/g, 11.6%; corn, 56.4%.

aThe TMR was as described in Table 1. bTrade name for potassium-magnesium sulfate. cTrace mineral mixture contained: copper sulfate,

cows maintained under practical management conditions and to evaluate the effects of feeding monensin or soybean oil on methane-CO2 emissions, milk production, and milk composition as well as ruminal fermentation patterns.

a commercial laboratory (Agri-Food Laboratories, Guelph, ON). Dry matter, CP, Ca, P, and Mg analyses were done according to AOAC (1984). The ADF content of the diets was measured according to procedures of Goering and Van Soest (1970), and TDN and NE contents were estimated according to NRC (1989). The chemical composition of diets fed in all experiments is listed in Table 3.

Experimental Protocol Materials and Methods Experiments and Diets

In all experiments except Exp. 1, a continuous feeding schedule was followed. In the continuous feeding schedule, the total mixed ration ( TMR) (Table 1 ) was fed daily at 0700 and 1600, with daily feed weighbacks at 0630. In addition to the TMR, cows were fed hay at 1300 ( 1 kg/[cowd]) and concentrate top dress at 0800, 1100, and 1700 (Tables 1 and 2). In the twice-daily feeding schedule (Exp. 1), TMR was fed at 0645 and 1500, and feed weighbacks were obtained twice daily at 0915 and 1730. Milking was twice daily at 0550 and 1550. Top dress concentrate was fed to cows producing in excess of 25 L milk/d at the rate of 1 kg for every 3 L of milk over 25 L. In Exp. 3 and 4, monensin was added in the form of Rumensin (Elanco, Division of Eli Lilly, Canada), which contains monensin in a concentration of 132 g/kg, to give a final concentration of 24 ppm of monensin in the TMR and top dress on a DM basis. Representative feed samples were collected biweekly and analyzed by Gas emissions from cows were monitored continuously between the periods of October 1994 and October 1996 except for days when instruments were checked and calibrated or when fan flow rates fell <3,200 L/s (Kinsman et al., 1995). There were between 88 and 109 lactating cows used for the experiments. Four experiments were conducted between September 26, 1994, and September 30, 1996, and the experimental protocols are listed in Table 4. In Exp. 1 the effects of continuous feeding vs twicedaily feeding were recorded (Table 5). Stage 1 was the control period with a continuous feeding schedule, and stage 2 was the transition period during which the cows were gradually introduced to a twice-daily feeding schedule that was maintained during stage 3. Continuous feeding was reintroduced in stage 4. Experiment 2 evaluated the use of unsaturated fat (soybean oil) when added to the diet. Salad dressingquality soybean oil (nonhydrogenated) was purchased bec). In this from CanAmera Foods (Montreal, Que experiment, in stage 1, control concentrate without soybean oil was fed (Table 2). In stage 2, the TMR

Downloaded from www.journalofanimalscience.org by guest on February 15, 2014

908

SAUER ET AL.

Table 3. Chemical composition of diets

Exp. 4 Diet content Exp. 1 and 2 Exp. 3 Control With soybean oil

% DM basis Total mixed ration DM, % CP, % ADF, % TDN, % NEl, Mcal/kg NEg, Mcal/kg NEm, Mcal/kg Ca2+, % P, % Mg2+, % Ca:P Top dress concentrate DM, % CP, % ADF, % TDN, % NEl, Mcal/kg NEg, Mcal/kg NEm, Mcal/kg 43.0 14.8 23.3 74.7 1.51 1.57 2.27 .87 .53 .31 1.74 89.6 20.2 3.4 74.7 92.8 1.57 2.27 .42 .28 .46 .42 .01 .02 .03 .05 .03 .02 .08 .12 .11 .08 .42 .14 .02 .03 43.8 15.1 24.9 73.2 1.49 1.53 2.23 .84 .47 .29 1.85 90.2 19.2 3.1 92.3 1.86 1.81 2.55 .56 .22 .73 .66 .01 .02 .02 .02 .01 .01 .02 .18 .10 .16 1.24 .01 .04 .04 42.0 16.1 22.5 75.1 1.52 1.6 2.31 1.32 .60 .39 1.91 88.9 21.6 4.9 91.5 1.83 1.6 2.31 3.10 .05 .90 .80 .02 .01 .01 .37 .15 .08 .05 .02 .71 .30 .30 .01 .01 .01 38.6 14.4 23.6 74.4 1.53 1.59 2.3 1.73 .67 .41 2.57 88.1 21.6 5.2 91.2 1.82 1.59 2.3 1.9 .43 .93 .85 .02 .03 .03 .14 .04 .02 .21 .15 .30 .34 .31 .01 .01 .01

contained soybean oil and in stage 3 the cows were again fed the control diet. Monensin was added to the diet in Exp. 3 and 4. Experiments 3 and 4 were divided into eight stages according to the dates indicated (Table 4). Stage 1 was the control period and stage 2 was the transition period during which monensin was gradually introduced into the diet. In stage 2, on d 1 and 2, monensin was fed at the rate of 6 ppm, on d 3 and 4 at the rate of 12 ppm, on d 5 and 6 at the rate of 18 ppm, and on d 7 and thereafter at the rate of 24 ppm. Stage 3 was the treatment period during which monensin was fed at 24 ppm. Stages 4 to 8 were the postmonensin feeding period during which the cows were again fed the control diet and data monitoring and recording were continued.

Table 4. Experimental protocol

Experiment and stage Exp. phase Exp. 1 1 2 3 4 Exp. 2 1 2 3 Exp. 3 1 2 3 4 5 6 7 8 Exp. 4 1 2 3 4 5 6 7 8 Control, continuous feeding Transition Treatment, twice-daily feeding Control, continuous feeding Control, control diet Treatment, soybean oil Control, control diet Control Transition Treatment, monensin feeding Control Control Control Control Control Control Transition Treatment, monensin feeding Control Control Control Control Control Start End MM.DD.YY MM.DD.YY

4.17.95 5.01.95 5.09.95 6.06.95 8.19.96 9.09.96 9.30.96 7.01.95 10.23.95 10.31.95 11.21.95 12.01.95 1.01.96 2.01.96 3.01.96 2.09.96 4.22.96 4.29.96 5.20.96 6.01.96 7.01.96 8.01.96 9.01.96

4.30.95 5.08.95 6.05.95 6.25.95 9.08.96 9.29.96 10.21.96 10.22.95 10.30.95 11.20.95 11.30.95 12.31.95 1.31.96 2.29.96 3.31.96 4.21.96 4.28.96 5.19.96 5.31.96 6.30.96 7.31.96 8.31.96 9.30.96

Ruminal Volatile Fatty Acid Analysis

Ruminal samples were obtained via a stomach tube before the morning feeding and analyzed as described by Fellner et al. (1997).

Milk Analysis
Milk samples were taken from 10 randomly selected cows at the morning and evening milkings and stored at 70C for subsequent analyses. Samples were thawed at 4C, placed in a 40C water bath for 30 min, and sonicated before analysis. The morning and evening samples were analyzed separately for their fatty acid composition. Total milk lipids were extracted according to Jensen (1989) and transesteri-

Downloaded from www.journalofanimalscience.org by guest on February 15, 2014

METHANE OUTPUT BY LACTATING COWS

909

Table 5. Effects of continuous feeding, twice a day feeding, and the addition of soybean oil to the diet
Feed intake (DM), kg/(dcow) Exp. 1 1 2 3 4 1 2 3 565 c 571 c 634d 635d 602 594 589 6,352g 6,280g 6,973h 6,774h 6,240 6,542 6,609 11.28:1c 11.00:1cd 11.05:1cd 10.69:1d 10.38:1g 11.05:1h 11.28:1h 16.1gh 14.8g 16.6h 16.0gh Exp. 2 19.8c 17.3dg 16.0dh 24.9 24.9 25.5 3.66 3.71 3.57 3.03 3.11 3.14 5.78 5.76 5.80 30.7c 29.3d 26.2ef 26.6f 3.55 3.51 3.72 3.67 3.01 2.99 2.93 2.93 5.86 5.81 5.78 5.77 Milk composition Milk,b kg/(dcow) % Fat % Protein % Other solids

Stagea

CH4, L/(cowd)

CO2, L/(cowd)

Ratio of CO2:CH4

aSee Table 4 for experimental phase. bLeast squares means after adjusting for differences in feed intake. c,d,e,fWithin a column, means lacking a common superscript differ ( P g,hWithin a column, means lacking a common superscript differ ( P <

< .01). .05).

fied using anhydrous NaOCH3 (Christie, 1982). The fatty acid methyl esters were extracted with a minimum volume of hexane and analyzed directly by gas chromatography ( GC) using a 100-m fused silica capillary column (Supelco, Supelco Park, PA) as described previously (Fellner et al., 1995). For milkfat, protein, lactose, and other solids, representative composited samples were taken twice a week and analyzed by a commercial laboratory (Canadian Laboratory Services, Ottawa, ON) by infrared transmission.

Statistical Analysis
Data from all experiments were analyzed using a mixed linear models approach with fixed and random effects and time series methods (SAS, 1989). Correlation analysis (results not reported) indicated that DMI, of the same day as well as the next day and day after, was correlated with gas and other measures; therefore, analysis of covariance (ANCOVA) was conducted using the following model: Yi = m + (stage effect) + bFi + b1Fi1 + b2Fi2 + errori, where m is the overall mean, stage represents the treatment term, b is the regression coefficient, Fi is the same day DMI, Fi1 is the intake of the previous day, etc. In case of DMI, an autoregressive model was used that included all the terms in the above model except bFi. Least squares mean for each experiment stage was separated using pairwise comparisons and significance was declared at P < .05. Similar models were used for weekly measures, with daily feed intake replaced by weekly aggregates. Interactions were also examined, but they were largely insignificant and therefore removed from the final model. Milk production data measured on individual cows on a weekly basis were analyzed using the following mixed analysis of covariance model: Yi = m + stage effect + cow effect + parity effect + b(DIM) + b1Fi +

b1Fi1 + error, where stage, parity, and days in milk ( DIM) were fixed covariable effects and cow was assumed to be a random effect. Parity was truncated at four and DIM at 5 wk and 50 wk were excluded. Milk production and DIM were best represented by a linear relationship that precluded the need to introduce quadratic or higher terms in the model. Ruminal VFA and milk fatty acid composition data were analyzed by including the time of sampling (before and after monensin treatment) as the main effect. In Exp. 4, specific contrasts of interest were 1 ) control (no monensin) vs once-treated and 2 ) control vs twice-treated. For Exp. 3 and 4, daily methane output measures during each of the eight stages were matched by day between the two experiments for statistical comparisons (Figure 1). This daily difference was further adjusted for differences in feed intake, both of the same and of the previous day with the following model d for each stage: di = b0 + b1Fd i + b1Fi1 + error, where b0 is the adjusted stage mean (the mean daily difference during the same stage between the two experiments if there were no difference in feed intakes). A rigorous time series analysis was performed using the SAS procedure AUTOREG (SAS, 1989) to test whether b0 = 0.

Results and Discussion Comparison of Gas Production by Lactating Cows in the Present Experiments and in Published Reports Using Different Methods
The different methods used to measure and sample methane have been reviewed by Johnson and Johnson (1995). These include open circuit indirect respiration chamber systems, face masks, isotopic and nonisotopic tracer techniques, prediction equations, mass balance techniques, and micro-meteorological methods. Not

Downloaded from www.journalofanimalscience.org by guest on February 15, 2014

910

SAUER ET AL.

quantity and type of diets fed (Johnson and Johnson, 1995). In the present experiments, the control values for methane emissions (i.e., in the absence of superimposed treatments) were 622 21 L/(cowd), similar to values obtained previously (Kinsman et al., 1995). These values apply only to mature, lactating Holsteins with body weights of 600 kg or more producing 29 1.5 kg milk/d.

Continuous vs Twice-Daily Feeding

When the cows were changed from continuous feeding to feeding twice daily (Table 5, Exp. 1 ) with a transition interval of 1 wk, methane output increased ( P < .01), as did the CO2 emissions ( P < .05). Milk production decreased ( P < .01) with twice a day feeding and stayed lower even after the cows were returned to a continuous feeding schedule in stage 4. Similar results were observed by Mu ller et al. (1980), who noted that with higher feeding frequency, lactating cows lost less energy in the urine but produced 8% more methane. Kirchgessner et al. (1993) also showed that there is a positive correlation between feed intake, body size, and methane production in dairy cattle.

Figure 1. Total herd methane emissions in the first monensin feeding trial (Exp. 3) and the repeat monensin feeding trial in Exp. 4. A to B is the transition period, and B to C is the period when 24 ppm was fed on a DM basis, followed by the postmonensin feeding period.

Effect of Adding Soybean Oil to the TMR on Milk Fatty Acid Composition and Methane Production
Long-chain fatty acids inhibit methane production in the rumen, and free fatty acids may be more potent inhibitors than triacylglycerols, although the mechanism by which this occurs is still not known (Van Nevel and Demeyer, 1996). Van Nevel and Demeyer (1996) in a summary of published data noted that free linoleic acid inhibited methane production over 50%, but lactating dairy cows fed 650 g of soybean oil per day only showed a moderate decline in methane output. In the diet used for Exp. 2 (Table 5), 3.5% soybean oil was added to the TMR on a DM basis, which, based on daily intake, is equivalent to approximately 600 g of soybean oil per cow per day. Methane production and CO2 output remained unchanged but the CO2:methane ratio increased from 10.4 to 11.3. The addition of soybean oil had no effect on milk production or the percentage milkfat. The addition of soybean oil increased ( P < .01) the relative concentration of linoleic acid (18:2[ n-6]) and all the products of biohydrogenation in the milk lipids, at the expense of the shorter-chain 6:0 to 16:0 saturated fatty acids (Table 6). The products of biohydrogenation included cis and trans 18:1 (monounsaturated fatty acids), 18: 0, the methylene interrupted isomers of 18:2, and the conjugated 18:2 isomers, particularly 9 cis, 11 trans 18:2. The ratio of cis-18:1 to trans 18:1 changed from 15:1 to 5:1 with soybean oil feeding (Table 6). The addition of soybean oil to the diet was of value with

surprisingly, vastly different results have been reported. Data from over 400 Holstein cows gave mean methane emissions of 346 L/(cowd) with minimum and maximum values of 96 and 615 L/(cowd) (Moe and Tyrrell, 1979). Crutzen et al. (1986) reported daily cow methane output as 5.5% of GE intake, or 392 L/(cowd), and Gibbs and Johnson (1994) reported values of 453 L/(cowd) and 384 L/(cowd) for U.S. and European dairy cows, respectively. Wilkerson et al. (1995) reported mean methane values for lactating Holsteins (weighing 600 kg or more) of 497 L/(cowd) with minimum and maximum values of 155 to 763 L/(cowd). For lactating cows (7.5 to 26 kg milk/d), Kirchgessner et al. (1991) reported average methane yields of 420 L/(cowd) and CO2 yields of 4,940 L/(cowd) with a CO2 to methane ratio of 11.8. Kinsman et al. (1995) reported mean methane and CO2 emissions of 587 and 6,137 L/ (cowd), respectively, and a CO2 to methane ratio of 10.5 for Holstein cattle weighing 602 kg with average milk productions of 28 kg. Stored manure contributed 5.8 and 6.1%, respectively, to total CO2 and methane gas emissions. Mean control values for gas emissions in the present study were 622 L/(cowd) for methane and 6,767 L/(cowd) for CO2. These large variations in recorded methane output by mature dairy cattle may in part be related to different measurement techniques as well as to the

Downloaded from www.journalofanimalscience.org by guest on February 15, 2014

METHANE OUTPUT BY LACTATING COWS

911

Table 6. Changes in milk fatty acid composition with addition of soybean oil to the diet
Milk fatty acid Soybean oil fed group Weight, 6:0 8:0 10:0 12:0 14:0 16:0 17:0 18:0 Total 18:1 trans-18:1 cis-18:1 18:2( n-6) 18:2 Methylene interrupted 18:2 Conjugated dienes 2.4* 1.4** 2.4*** 2.4*** 10.1*** 28.0** .8** 10.3*** 26.6*** 1.7*** 24.8*** 2.2*** .8** .5** 1.9 1.0 1.5 1.4 6.9 22.5 .6 11.6 35.9 6.5 29.4 2.9 1.4 1.4 %a .06 .05 .09 .09 .19 .39 .01 .17 .71 .32 .63 .08 .05 .09

Control group

SE

aWithin a row, means lacking a common superscript differ ( * P < .05; ** P < .01; *** P < .001).

regard to improving the essential fatty acid (18: 2[n-6]) profile of the resultant milk, and there was a decrease in saturated fatty acids and an increase in conjugated dienes. The conjugated dienes have been associated with inhibition of carcinogenesis and tumorigenesis (Belury, 1995).

monensin in Exp. 3, and 21 cows had recently freshened and had never been exposed to monensin. As shown in Table 7, (Exp. 4 ) there was little detectable response to the addition of monensin to the diet. Even though methane output decreased ( P < .05) in stages 3 to 8, the drop was relatively small; similarly, the increase in CO2:methane ratio was small compared to that observed during Exp. 3. Feed consumption increased ( P < .05) and milk production decreased ( P < .05), whereas percentage fat did not change in Exp. 4. This suggests an adaptive change in the rumen microflora. Daily changes can be noted from data in Figures 1 and 2. Figure 1 compares the daily methane production by cows in Exp. 3 and Exp. 4 during the transition period ( A to B ) and the monensin feeding period ( B to C). The effect of feeding monensin after accounting for differences in feed intake was far more pronounced ( P < .01) in Exp. 3 than in Exp. 4 during periods A to C. After cessation of monensin feeding, the differences disappeared ( P < .10). When covariate analyses were performed to adjust for feed intake, DIM, and parity (results not shown), milk production increased ( P < .05) as a result of feeding monensin in Exp. 3 (stage 3 ) but returned to pretreatment levels by stage 7. However, similar results were not observed in Exp. 4. These results suggest that some adaptive changes had occurred in the rumen microflora.

Effects of Monensin on Gas Production, Feed Intake, Milk Production, and Milk Composition
The results in Exp. 3 (Table 7 ) clearly show the well-documented effects of adding monensin to a dairy cattle diet. Methane output decreased ( P < .05), feed intake decreased ( P < .05), and milk production increased ( P < .05). The percentage fat and total fat showed a temporary but significant ( P < .05) decrease. Milk protein and other solids remained essentially unchanged. The change in the ratios of CO2:methane as a result of feeding monensin to lactating dairy cows have not been previously reported. The ratio ranged from 10:1 to 11:1 before the addition of monensin and reached a maximum of 15:1 after monensin feeding. This shows that monensin had no effect on CO2 output, which remained constant, and methane output dropped significantly. This ratio may be helpful in determining whether decreased methane output is the result of inhibited production rates or simply reflects decreased feed consumption. It is of interest to note that with mixed ruminal bacteria growing in rumen fermenters, the CO2:methane ratio was approximately 3:1 (Sauer and Teather, 1987). In Exp. 4 (Table 7), monensin was added to the diet during stage 3, exactly 161 d after termination of the first monensin feeding period in Exp. 3. In Exp. 4, there were a total of 88 cows, of which 67 had received

Effects of Monensin on Ruminal Volatile Fatty Acids

In Exp. 3, ruminal fluid from 10 randomly selected cows was sampled by stomach tube and analyzed for VFA (Table 8 ) 10 d before the start of monensin feedings, then resampled 11 d after the start of fulllevel monensin feeding. As expected, the molar percentage concentration of acetate and butyrate decreased, as did the acetate:propionate ratio. The propionate concentration increased. In Exp. 4, 10 cows were selected from the 21 newly freshened cows that had never received monensin previously and sampled 13 d before the start of monensin feeding, then resampled 8 d after the start of full-level monensin feeding (Table 9). The molar concentration of ruminal VFA changed as before (Table 8). Ten cows that received monensin twice were randomly selected from the 67 cows that had received monensin previously in Exp. 3. These cows were sampled 16 d after the start of full-level monensin feeding in Exp. 4. As shown in Table 9, these cows had little or no VFA response to monensin. Their VFA concentration was not different from that of the control group (i.e., the cows that never received monensin). These results indicate that even though some of the effects of feeding monensin to previously untreated cows were as expected (Sauer et al., 1989; i.e., increased milk production, decreased methane emis-

Downloaded from www.journalofanimalscience.org by guest on February 15, 2014

912

SAUER ET AL.

Table 7. Effects of adding monensin to the dairy diet

Milk composition Stagea CH4, L/(cowd) 658c 548de 517d 574ef 585ef 558e 597e 591e 633c 642cd 599de 559e 577de 572de 595d 599d CO2, L/(cowd) 7,124c 7,257cd 7,352cd 7,754de 7,890e 7,143c 7,097c 7,158c 7,330c 7,355c 7,200c 6,142de 6,367e 5,979d 6,167de 6,502e Ratio of CO2:CH4 10.78:1c 13.34:1df 14.71:1e 13.71:1d 13.77:1d 13.09:1f 11.93:1g 12.10:1g 11.60:1c 11.46:1ce 12.02:1d 10.99:1eg 11.04:1eg 10.48:1f 10.38:1f 10.90:1g Feed intake (DM), kg/(dcow) Exp. 3 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 16.1c 14.4d 14.5d 14.9de 14.6d 15.2de 15.4de 15.9ce Exp. 4 16.5c 17.0cd 17.8de 16.9ce 18.2d 16.2c 17.0ce 18.1d 31.2c 29.3d 29.1d 27.1e 27.3e 26.3e 25.2f 24.9f 3.64c 3.65cde 3.56cd 3.80d 3.77d 3.80e 3.67cd 3.70cd 1,113c 1,074d 1,033e 1,016ef 1,032e 990f 909g 921g 3.16cd 3.08cd 3.12d 3.10cd 3.09c 2.98c 3.01cd 3.08cd 5.75 5.75 5.71 5.73 5.71 5.70 5.69 5.66 27.6c 30.3d 31.5e 33.0f 34.1g 34.8g 33.3fg 32.6ef 3.61c 3.14de 2.70f 2.87df 2.87df 3.00d 3.38e 3.56e 929c 950cd 895e 921ce 985d 1,052f 1,098f 1,125f 3.04cd 3.10d 3.10d 3.06cd 3.09d 3.08d 2.98c 3.11d 5.74 5.77 5.75 5.75 5.79 5.75 5.79 5.84 Milkb, kg/(dcow) % Fat g Fat/ (dcow) % Protein % Other solids

aSee Table 4 for experimental phase. bLeast squares means after adjusting c,d,e,f,gWithin a column, means lacking

for differences in feed intake. a common superscript differ ( P < .05).

Table 8. Ruminal VFA from 10 randomly selected cows in Experiment 3 sampled before and after feeding monensin for 2 weeks
2 Weeks after start of monensin treatment Molar Acetate ( A ) Propionate ( P ) Butyrate Valerate Isobutyrate Isovalerate A:P
aWithin

VFA

Control period

SEM

%a 62.9 24.9 7.9 1.1 1.4 1.8 2.6 .88 1.06 .44 .08 .07 .13 .18

65.0 21.2* 10.3*** 1.1 1.0*** 1.4* 3.2*

a row, means lacking a common superscript differ ( * P < .05; *** P < .001).

Table 9. Ruminal VFA from 10 randomly selected cows in Experiment 4 that had never received monensin (control group), the same cows after 2 weeks on monensin treatment (treated group), and 10 randomly selected cows that received monensin in Experiment 3 and again in Experiment 4 (twice-treated group)
Contrast VFA Control group Monensin oncetreated group Monensin twicetreated group SEM Control vs treated P < Acetate ( A ) Propionate ( P ) Butyrate Valerate Isobutyrate Isovalerate A:P
aNS

Control vs twice-treated

67.4 18.1 10.4 1.0 1.3 1.9 3.8

64.8 22.0 8.1 .9 1.7 2.4 3.0

66.7 18.0 10.7 .9 1.5 2.1 3.8

.93 1.11 .46 .08 .07 .14 .19

.10 .02 .001 NSa .001 .01 .003

NS NS NS NS .02 NS NS

= not statistically significant.

Downloaded from www.journalofanimalscience.org by guest on February 15, 2014

METHANE OUTPUT BY LACTATING COWS

913

Table 10. Changes in milk fatty acid composition from 10 randomly selected cows in Experiment 3 before and after feeding monensin
Milk fatty acid Control group Monensintreated group %c 1.2 .8 1.8 2.6 9.8 24.4 .7 7.4 34.0 11.9 22.1 4.3 .9 .08 .04 .12 .19 .41 .72 .02 .59 .95 .90 1.02 .22 .05

SE

Weight, 6:0 8:0 10:0 12:0 14:0 16:0 17:0 18:0 Total 18:1 trans-18:1 cis-18:1 18:2( n-6) 18:2 Methylene interrupted 18:2 Conjugated dienes 1.5 1.0 2.4* 3.2* 10.9 28.2** .6 11.1*** 27.3*** 4.0*** 23.3 3.4* .5*

.8*

1.3

.15

Figure 2. Milk production in response to monensin feeding in Exp. 3 and 4. A to B represents the transition period, and period B to C is when monensin was added to the diet at 24 ppm on a DM basis. At interval C, the monensin was withdrawn from the diet, and the herd was again fed the control diet without monensin.

aWithin a row, means lacking a common superscript differ ( * P < .05; ** P < .01; *** P < .001).

when monensin is added to the diet, and as Dawson and Boling (1983) noted, the percentage of monensinresistant bacteria declined in control animals.

sions, decreased feed intake, lower milkfat, altered ruminal VFA ratios, and changes in milk fatty acid composition), a repeat monensin treatment 160 d later to the same animals was largely ineffective. Even though it would be tempting to ascribe these results to an induced ruminal microbial resistance to monensin, this is by no means certain. In vitro studies have shown that Gram-positive bacteria can develop ionophore resistance (Dawson and Boling, 1983; Dennis et al., 1986), and possible explanations have been proposed. These include a modification of extracellular polysaccharides, increased ion pump activity, and futile ion cycles (Russell and Storbel, 1989). The transfer of genes encoding resistance was thought to be less likely because of the long-term (20 yr or more) effective use of monensin in the beef cattle industry. The suggestion has been made (Lana and Russell, 1996) that ionophore resistance is a fundamental aspect of growth and not the result of microbial mutation. These authors measured the concentrations of monensin required to cause halfmaximal K+ depletion ( K+ d) in ruminal bacteria and observed that approximately half the population of unadapted, mixed ruminal bacteria are resistant to high levels of monensin. These populations of resistant bacteria rapidly replace the inhibited bacteria

Effect of Monensin on Milk Fatty Acid Composition

Milk fatty acid composition was analyzed from 10 randomly selected cows in Exp. 3 before and after monensin treatment (Table 10). The fatty acid profile agrees well with that published by Wonsil et al. (1994), with slightly higher 18:1 levels in the present work. Wonsil et al. (1994) reported a ratio of 18 for cis-18:1 to trans-18:1, whereas in our work the ratio was 5:1. Jiang et al. (1996) showed a cis-18:1 to trans18:1 ratio of 9:1 in cow milk. Significant changes in milk composition were noted after cows were fed monensin in Exp. 3. Biohydrogenation apparently was decreased, as had been observed previously in mixed ruminal cultures (Fellner et al., 1997). In accordance with decreased biohydrogenation, the concentration of 10:0, 12:0, 16:0, and 18:0 was depressed ( P < .05), but total 18:1 and the trans-18:1 isomer were increased ( P < .01) in concentration. Thus, in evaluating possible benefits of altered milk composition resulting from the feeding of ionophores to dairy cattle, the decrease in saturated fatty acids should be weighed against the increase in trans-18:1 formation. The changes in milk fatty acid composition as a result of feeding monensin were similar to those observed when soybean oil was added to the diet and included the increase in the 18:2 conjugated diene concentration.

Downloaded from www.journalofanimalscience.org by guest on February 15, 2014

914

SAUER ET AL.

Implications
Monensin was shown to be highly effective when fed to lactating Holsteins in full production. Characteristically, the cows showed improved feed efficiency and milk production together with decreased milkfat, decreased methane production, and depressed biohydrogenation, as evidenced by altered milk fat composition. The herd was allowed to revert to the premonensin state after approximately 5 mo, then placed on a second 3-wk monensin treatment. This time, none of the above changes were observed. Unlike in the first trial, rumen volatile fatty acids and acetate:propionate ratios did not show the changes characteristic of ionophore feeding. Even though the results do not offer any specific insight into why the cows did not respond as effectively in the second trial, it is clear that information is needed with respect to the optimum exposure of dairy cattle to monensin as well as the advantages, if any, of rotating ionophores with other nonionophore feed additives for maximum effectiveness.

Literature Cited
AOAC. 1984. Official Methods of Analysis (14th Ed.). Association of Official Analytical Chemists, Washington, DC. Belury, M. A. 1995. Conjugated dienoic linoleate: A polyunsaturated fatty acid with unique chemoprotective properties. Nutr. Rev. 53:8389. Christie, W. W. 1982. Lipid Analysis (2nd Ed.). Pergamon Press, Oxford, U.K. Crutzen, P. J., I. Aselmann, and W. Seiler. 1986. Methane produced by domestic animals, wild ruminants, other herbivorous fawni and humans. Tellus Ser. B 28:271284. Dawson, K. A., and J. A. Boling. 1983. Monensin-resistant bacteria in the rumen of calves on monensin-containing and unmedicated diets. Appl. Environ. Microbiol. 46:160164. Dennis, S. M., T. G. Nagaraja, and A. D. Dayton. 1986. Effect of lasalocid, monensin and thiopeptin on rumen protozoa. Res. Vet. Sci. 46:160164. Fellner, V., F. D. Sauer, and J.K.G. Kramer. 1995. Steady-state rates of linoleic acid biohydrogenation by ruminal bacteria in continuous culture. J. Dairy Sci. 78:18151823. Fellner, V., F. D. Sauer, and J.K.G. Kramer. 1997. Effect of nigericin, monensin and tetronasin on biohydrogenation in continuous flow-through ruminal fermenters. J. Dairy Sci. 80:921928. Gibbs, M., and D. E. Johnson. 1994. Methane emissions from digestive processes of livestock. International Anthropogenic methane emissions: Estimates for 1990 EPA 230R-93-010:1. Goering, M. K., and P. J. Van Soest. 1970. Forage fiber analyses (apparatus, reagents, procedures, and some applications). Agric. Handbook No. 379. ARS, USDA, Washington, DC.

Holter, B., and A. J. Young. 1992. Methane prediction in dry and lactating Holstein cows. J. Dairy Sci. 75:21652175. Jensen, R. G. 1989. The Lipids of Human Milk. CRC Press, Boca Raton, FL. Jiang, J., L. Bjoerck, R. Fonden, and M. Emanuelson. 1996. Occurrence of conjugated cis-9, trans-11-octadecadienoic acid in bovine milk: Effect of feed and dietary regimen. J. Dairy Sci. 79: 438445. Johnson, D. E., and G. M. Ward. 1996. Estimates of animal methane emissions. Environ. Monit. Assess. 42:133141. Johnson, K. A., M. Huyler, H. Westberg, B. Lamb, and P. Zimmerman. 1994. Measurement of methane emissions from ruminant livestock using SF6 tracer technique. Environ. Sci. Technol. 28: 539362. Johnson, K. A., and D. E. Johnson. 1995. Methane emissions from cattle. J. Anim. Sci. 73:24832492. Kinsman, R., F. D. Sauer, H. A. Jackson, and M. S. Wolynetz. 1995. Methane and carbon dioxide emissions from dairy cows in full lactation monitored over a six month period. J. Dairy Sci. 78: 27602766. Kirchgessner, M., F. X. Roth, and W. Windisch. 1993. Verminderung der Stickstoff und Methanausscheidung von Schwein und Rind bers. Tierernaehr. 21:89120. durch die Fu tterrung. U Kirchgessner, M., W. Windisch, H. L. Mu ller, and M. Kreuzer. 1991. Release of methane and carbon dioxide by dairy cattle. Agrobiol. Res. 44:91102. Lana, R. P., and J. B. Russell. 1996. Use of potassium depletion to assess adaptation of ruminal bacteria to ionophores. Appl. Environ. Microbiol. 62:44994503. Marik, T., and I. Levin. 1996. A new tracer experiment to estimate the methane emissions from a dairy cow shed using sulphur hexafluoride (SF 6) . Global Biogeochem. Cycles 10:413418. Moe, P. W., and H. F. Tyrrell. 1979. Methane production in dairy cows. J. Dairy Sci. 10:15831586. Mu ller, H. L., J. Sax, and M. Kirchgessner. 1980. Energieverluste u ber Kot, Harn and Methan durch unterschiedliche Haufigkeit der Fu tterung bei nichtlaktierenden und laktierenden Ku hen. Tierphysiol. Tierernaehr. Fuetter. 44:181189. NRC. 1989. Nutrient Requirements of Dairy Cattle (6th Rev. Ed.). National Academy Press, Washington, DC. Russell, J. B., and H. J. Strobel. 1989. Effect of ionophores on ruminal fermentation. Appl. Environ. Microbiol. 55:16. SAS. 1989. SAS/STAT Users Guide (Version 6, 4th Ed.). SAS Inst. Inc., Cary, NC. Sauer, F. D., J.K.G. Kramer, and W. J. Cantwell. 1989. Antiketogenic effects of monensin in early lactation. J. Dairy Sci. 72: 436442. Sauer, F. D., and R. M. Teather. 1987. Changes in oxidation reduction potentials and volatile fatty acid production by rumen bacteria when methane synthesis is inhibited. J. Dairy Sci. 70: 18351840. Van Nevel, C. J., and D. I. Demeyer. 1996. Control of rumen methanogenesis. Environ. Monit. Assess. 42:7397. Wonsil, B. J., J. H. Herbein, and B. A. Watkins. 1994. Dietary and ruminally derived trans-18:1 fatty acids alter bovine milk lipids. J. Nutr. 124:556565. Wilkerson, V. A., D. P. Casper, and D. R. Mertens. 1995. The prediction of methane production of Holstein cows by several equations. J. Dairy Sci. 78:24022414.

Downloaded from www.journalofanimalscience.org by guest on February 15, 2014

Citations

This article has been cited by 5 HighWire-hosted articles: http://www.journalofanimalscience.org/content/76/3/906#otherarticles

Downloaded from www.journalofanimalscience.org by guest on February 15, 2014