Search
Author
Title
Vol.
Issue
Year
1st Page

Journal of Animal Science - Animal Production

Associations among methane emission traits measured in the feedlot and in respiration chambers in Angus cattle bred to vary in feed efficiency1

 

This article in JAS

  1. Vol. 94 No. 11, p. 4882-4891
     
    Received: May 07, 2016
    Accepted: Sept 07, 2016
    Published: October 27, 2016


    2 Corresponding author(s): paul.arthur@dpi.nsw.gov.au
 View
 Download
 Share

doi:10.2527/jas.2016-0613
  1. R. M. Herd*,
  2. J. I. Velazco†‡,
  3. P. F. Arthur 2§ and
  4. R. F. Hegarty
  1. * Beef Industry Centre, NSW Department of Primary Industries, Armidale, NSW 2351, Australia
     Environmental and Rural Science, University of New England, Armidale, NSW 2351 Australia
     National Institute of Agricultural Research, Treinta y Tres 33000, Uruguay
    § Elizabeth Macarthur Agricultural Institute, NSW Department of Primary Industries, Menangle, NSW 2568, Australia

Abstract

The objective of the study was to evaluate associations among animal performance and methane emission traits under feedlot conditions and in respiration chambers in Angus cattle bred to vary in residual feed intake (RFI), which is a measure of feed efficiency. Fifty-nine cattle were tested for feedlot RFI, of which 41 had methane production recorded on an ad libitum grain-based ration in the feedlot, 59 on a restricted grain-based ration in respiration chambers, and 57 on a restricted roughage ration in respiration chambers. The cattle became older and heavier as they went through the different phases of the experiment, but their feed intake (expressed as DMI) and daily emission of enteric methane (methane production rate; MPR) did not increase proportionally, as feed offered was restricted in the respiration chamber tests. Methane emissions by individual animals relative to their DMI were calculated as methane yield (MY; MPR/DMI) and as 2 measures of residual methane production (RMPJ and RMPR), which were calculated as the difference between measured MPR and that predicted from feed intake by 2 different equations. Within each test regime, MPR was positively correlated (r = 0.28 to 0.61) with DMI. Phenotypic correlations for MY, RMPJ, and RMPR between the feedlot test and the restricted grain test (r = 0.40 to 0.43) and between the restricted grain test and the restricted roughage test were moderate (r = 0.36 to 0.41) and moderate to strong between the feedlot test and the restricted roughage test (r = 0.54 to 0.58). These results indicate that the rankings of animals for methane production relative to feed consumed are relatively stable over the 3 test phases. Feedlot feed conversion ratio and RFI were not correlated with MPR in the feedlot test and grain-based chamber test but were negatively correlated with MPR in the chamber roughage test (r = −0.31 and −0.37). Both were negatively correlated with MY and RMPJ in the feedlot test (r = −0.42 to −0.54) and subsequent chamber roughage test (r = −0.27 to −0.49). Midparent estimated breeding values for RFI tended to be negatively correlated with MY and RMPJ in the feedlot test (r = −0.27 and −0.27) and were negatively correlated with MY, RMPJ, and RMPR in the chamber roughage test (r = −0.33 to −0.36). These results showed that in young growing cattle, lower RFI was associated with higher MY, RMPJ, and RMPR but had no significant association with MPR.



INTRODUCTION

Methane is a potent greenhouse gas produced by ruminants during the process of microbial fermentation of plant material mainly in the rumen. There is phenotypic and genetic variation in enteric methane emissions in beef cattle (Herd et al., 2014; Donoghue et al., 2016a). This offers the possibility of breeding cattle with lower methane emissions. However, published scientific information on recognized, sustained, and repeatable reduction in emissions in cattle across a range of diets is limited.

Given the strong positive relationship between DMI and the methane production rate (MPR; Blaxter and Clapperton, 1965; Kennedy and Charmley, 2012), it can be expected that animals with lower DMI will also have lower MPR. Variation exists among individual cattle in DMI required to maintain BW and for growth (Koch et al., 1963; Arthur et al., 2001). Residual feed intake (RFI) is a measure of this variation in feed efficiency and can be calculated as the difference between actual DMI and expected DMI for maintenance and BW gain over a test period. In beef cattle, RFI is moderately heritable (Herd et al., 2003) and EBV for RFI have been available to Australian beef cattle producers since 1999 (Herd et al., 2003).

Studies of young growing cattle that were phenotypically (Nkrumah et al., 2006) and genetically (Hegarty et al., 2007) divergent for RFI have reported lower MPR in low-RFI relative to high-RFI animals. This has led to the conclusion that breeding for low RFI could provide a breeding solution to reduce the cost of feeding growing cattle and to reduce enteric methane emissions per unit of product (Arthur et al., 2009). The advent of new measurement technologies for methane under pasture and feedlot conditions provides the opportunity to examine the RFI-methane relationship under different feeding systems. The objective of the study was to evaluate associations among animal performance and methane emission traits under feedlot and respiration chamber conditions in young growing Angus cattle bred to vary in RFI.


MATERIALS AND METHODS

Animals

This research was approved under New South Wales Department of Primary Industries (NSW DPI) Animal Research Authority ORA 13/16/004 and the University of New England (UNE) Animal Research Authorities AEC14-002 and AEC14-036. The Angus cattle were bred at the NSW DPI Agricultural Research Centre (Trangie, NSW, Australia). They were bred by AI using cows from the postweaning RFI-divergent selection lines described by Arthur et al. (2001) and stored semen from 2 sires per line that had previously been used in these selection lines. All sires and dams had EBV for postweaning RFI (RFI-EBV), dated November 2009 and calculated by the Animal Breeding and Genetics Unit, UNE (Armidale NSW). The RFI-EBV of the 2 low-RFI sires were −0.61 and −0.97 kg DM/d, and those for the 2 high-RFI sires were +0.61 and +1.42 kg DM/d. Midparent EBV for RFI (RFI-EBVmp) for each animal was calculated as the arithmetic mean of its parents’ RFI-EBV. By chance, the sires differed slightly in their EBV for weight at 400 days of age (400dWT-EBV): +40 and +68 kg for the 2 low-RFI sires and +33 and +53 kg for the 2 high-RFI sires. Male calves were castrated at 4 mo of age. At weaning on 26 February 2013, 64 calves of approximately equal numbers of steers and heifers were available for this experiment. The weaned cattle were transferred to NSW DPI Agricultural Research and Advisory Station (Glen Innes, NSW, Australia) and grown on native pastures until they reached feedlot entry weight of approximately 400 kg BW. The cattle were tested for growth, DMI, feed efficiency, and exhaled methane emissions in the feedlot and then for methane production on restricted quantities of grain and roughage diets in respiration chambers. A description of the feed type and characteristics of the 3 methane test phases are presented in Table 1.


View Full Table | Close Full ViewTable 1.

Feed and methane test phase characteristics

 
Methane test phase
Feed and traits Feedlot Chamber
grain
Chamber
roughage
Feed type Grain based Grain based Alfalfa based
Feed offered Ad libitum Restricted1 Restricted2
ME content of feed, MJ/kg DM 12.6 12.6 9.0
Methane measurement equipment3 GEM RC RC
1Restricted to 70% of feedlot DMI.
2Restricted to 1.2 times maintenance energy requirements.
3GEM and RC refer to GreenFeed emission monitors and respiration chambers, respectively.

Feedlot Test

The cattle were measured for growth rate, DMI, feed efficiency, and exhaled methane over a standard 10-wk RFI test used in Australia and described by Exton (2001). On 9 January 2014 the cattle were transported to the UNE Tullimba research feedlot, east of Armidale, NSW, and commenced the induction protocol in place at the feedlot. Over a 2-wk period, the animals were offered rations of increasing grain content fed in open bunks. They were then offered a high-grain-content finishing ration that consisted of approximately 75% grain, 10% sorghum hay, and 5% protein pellets, plus molasses and vitamin and mineral additives (as-fed basis), fed in GrowSafe feed bins (GrowSafe Systems Ltd., Airdrie, AB, Canada) for 2 wk. Feed intake by each animal was recorded using the GrowSafe system. Four of the 64 animals were removed before the start of the RFI test by the feedlot manager for failing to adapt to eating from the GrowSafe system and were excluded from the remainder of the experiment. Two of the animals removed were heifer progeny from each of the low-RFI sires, the other 2 were steer and heifer progeny from 1 of the high-RFI sires. Emissions of methane by individual cattle were determined from multiple short-term breath measurements using GreenFeed Emission Monitor (GEM; C-Lock Inc., Rapid City, South Dakota) units. The GEM units dispensed multiple small quantities of a feed pellet as an attractant. The feed pellet was designed to have the same ME content on a DM basis as the feedlot ration. The animals had previously been given a training period of 3 mo at pasture with the GEM units. Operation of the GEM system is described more fully below. Heifers and steers were kept in separate, adjacent feedlot pens, and each pen contained a GEM unit to measure exhaled methane emissions, 4 GrowSafe feed bins side by side, and a water trough (Fig. 1). The heifers and steers were swapped between feedlot pens midway through the test. Feed bins were replenished at least twice daily and only rarely were completely emptied by the cattle before replenishment.

Figure 1.
Figure 1.

Feedlot pen showing side-by-side placement of GrowSafe feed bins and GreenFeed emission monitors for recording individual animal feed intake and enteric methane production, respectively.

 

The 70-d RFI test began 3 February 2014. One animal was removed from the RFI test because it stopped eating regularly from the feed bins, and 59 animals completed the RFI test. Samples of the test ration and feed pellets from the start, midpoint, and end of the RFI test were sent to a commercial feed evaluation service (NSW DPI Feed Testing Service, Wagga Wagga, NSW, Australia). The average content of samples of the feedlot ration was 89% DM, 85% DM digestibility (DMD), 12% CP (DM basis), and ME content of 13 MJ/kg DM. The cattle were weighed at the start of the test period and then fortnightly, without fasting.

The GEM System

Emissions of methane by individual cattle in the feedlot test were determined from multiple short-term breath measurements using GEM. The GEM unit delivered pellets (6-mm diam.) containing sorghum, wheat, and cottonseed meal (ME = 11 MJ/kg DM; CP = 14% DM) and aniseed flavor as an attractant (Fluidarom 1957, Norel, Madrid, Spain). The GEM units were programmed to deliver up to 4 drops of pellets every 45 s and to restrict too frequent access to supplement; the GEM was programmed to wait at least 3 h before a new supplement event could occur for the same animal. The average (±SD) weight of pellet dispensed in each supplement drop was 52.3 ± 0.3 g. The GEM system includes a number of firmware and software checks to avoid recording data for compromised measures, such as when animals stepped away from the shroud during measurement. “Valid” emission records were defined as those of at least 3-min duration with the animal’s head correctly in position. Not all animals volunteered to visit the GEM or to visit every day or to keep their head in the GEM for a valid visit to be recorded. This resulted in a variable number of valid records over a variable numbers of days being captured for individual animals.

For this study only data for cattle (n = 41) that had previously been trained to use the GEM, as evidenced by having had at least 1 valid GEM record (at least 3-min duration at the GEM unit while methane was being measured) in the final month at pasture, were considered for analysis. For these 41 animals, the average number of valid GEM records during the RFI test was 99 per animal, with a range from 29 to 223 records per animal. There was no evidence for a compromise in animal feed intake and growth by regular GEM usage in the feedlot, with the 41animals with GEM data analyzed having an ADG of 1.4 ± 0.2 (mean ± SD) kg/d, a DMI of 12.4 ± 1.1 kg/d, and a feed conversion ratio (FCR) of 8.9 ± 1.3 kg/kg, whereas 12 animals that failed to have a valid GEM record had an ADG of 1.3 ± 0.2 kg/d, a DMI of 11.7 ± 1.2 kg/d, and a FCR of 8.9 ± 1.6 kg/kg. Six other animals that failed to have a valid GEM record in the pasture training period did have some GEM records in the feedlot, but they are excluded from this analysis.

Animal House Test

After the feedlot test, the cattle were then moved in groups of 10 to the UNE campus (Armidale, NSW, Australia) for measurement of methane emissions on feedlot grain ration and then on a roughage diet using respiration chambers. On the campus, the animals were first held in small yards outside the animal house and fed the same feedlot finisher ration that had been offered during the feedlot test. After 3 to 5 d, they were brought into the animal house and fed in individual pens for 2 d, then weighed and placed in 1 of 10 open-circuit respiration chambers for 2 d of measurement. This weight was used as the test weight (test BW) for the animal. Each animal was to be tested on the feedlot finisher grain ration offered ad libitum like in the feedlot test. In anticipation that in respiration chambers the cattle would not eat as much as when in the feedlot, each animal was offered each day an amount of feed equal to 90% of their average DMI recorded over the final 2 wk of their feedlot RFI test. The first group of 10 cattle had large feed refusals, so the ration allocation was reduced to 70% for the next 5 groups of cattle tested. Cattle were fed between 0830 and 0900 h. Cattle remained in the respiration chambers for 2 consecutive 24-h periods, with a fresh allocation of feed offered at the start of the second 24-h period, and any feed refusals were recorded. Groups of cattle moved through the animal house and chambers over 3 wk, with groups of steers tested alternatively with groups of heifers and with progeny of each sire in each group. The animals were not given any prior training to the respiration chambers, and all 59 cattle that completed the feedlot RFI test were successfully recorded for the chamber grain test. The cattle animal house, respiration chambers, and the protocols used for cattle measurement are described in Hegarty et al. (2014) and Herd et al. (2014).

Following being measured in the chamber grain test, each group of cattle was moved to a small outside yard and fed a commercial chaff roughage diet for a minimum of 4 wk before being brought back into the animal house for a repeat of the previous routine, but this time they were offered a restricted intake of roughage ration. The ration was a commercial alfalfa and oaten hay chaff (Manuka “Blue Ribbon” chaff; Manuka Chaff Pty. Ltd., Quirindi, NSW, Australia) containing 88% DM, 67% DMD, 16% CP (DM basis), and an ME content of 9 MJ/kg DM (NSW DPI Feed Quality Service, Wagga Wagga, NSW, Australia). The cattle were individually offered feed equivalent to 1.2 times the estimated maintenance energy requirements on the basis of their test BW during their methane test on the grain ration, calculated as described in Herd et al. (2014). Two animals (a heifer from a low-RFI sire and a steer from a high-RFI sire) had large feed refusals, and data for these animals were not used, resulting in 57 of the 59 cattle that completed the feedlot RFI test being successfully recorded for the chamber roughage test.

Traits and Statistical Analyses

Definitions of traits analyzed and the formula used to calculate them are presented in Table 2. In the feedlot test, data for 59 animals were used to calculate RFI. Plots of BW vs. the day of test for each animal showed growth over the test was approximately linear. Start-of-test BW, midtest BW, and ADG over the test were calculated for each animal from the linear regression of its fortnightly BW against day of test. Individual animal data for weight of feed removed from the GrowSafe feed bins by day were visually inspected using a “heat map” created using a spreadsheet to assign individual animal daily feed intake data supplied by the GrowSafe system vs. the day of the RFI test and then color-coding each cell using a graded color scale from blue (0 kg/d) to red (30+ kg/d). The assigned feed disappearance (AFD) audit table supplied by the GrowSafe system was used to detect days when feed removal from a feed bin that could be assigned to animals fell below 100%. Used together, the heat map and AFD were inspected for abnormal changes in intake and feed assignment indicative of malfunction of the animal identification and/or data recording system, with data from feed bins with AFD < 90% and for animals with feed intake < 4 kg/d or > 25 kg/d receiving extra scrutiny. Individual animal data for DMI (feedlot ration plus GEM pellets) were regressed against (midtest BW)0.75 and ADG, with the residuals being RFI. For the 41 animals with regular valid GEM records in the feedlot test, MPR was calculated for each animal as the average of all its valid GEM records. In another experiment in the same research feedlot, using data from 495 Angus steers on a 70-d ad libitum feedlot diet with 46,657 GEM records showed that the simple average of multiple GEM records based on as few as 30 records is as precise an estimate of MPR as an average calculated from a much larger number (81 records) for animals that have received no prior training with the GEM (P. F. Arthur, NSW DPI, Camden, Australia, personal communication). Two forms of residual methane production (RMP) were calculated. The first, RMPJ, was calculated as the difference between the measured MPR and expected MPR calculated with a widely used prediction equation, which is MPR = 3% of GE intake of a grain ration and 6% of a roughage ration (Johnson et al., 1995). The second form, RMPR, was calculated as the residual of MPR regressed against DMI to determine whether animals were producing more or less methane than expected for their DMI.


View Full Table | Close Full ViewTable 2.

Definition of traits

 
Trait name Abbreviation Units Definition
Start weight Start BW kg BW in feedlot at start of RFI test by regression of BW on test day
Test weight Test BW kg Midtest BW for RFI test by regression of BW on test day or BW taken on morning before being put into respiration chamber
Average daily gain ADG kg/d In feedlot by regression of BW on RFI test day
Dry matter intake DMI kg/d DMI during methane test
Feed conversion ratio FCR kg/kg DMI/ADG
Residual feed intake RFI kg DM/d Difference between measured DMI and expected DMI
Methane production rate MPR g/d Methane produced
Methane intensity MI g/kg gain MPR/ADG in the feedlot
Methane yield MY g/kg DM MPR/DMI
Residual methane production, first form RMPJ g/d MPR net of expected MPR from DMI using formula of Johnson et al. (1995)
Residual methane production, second form RMPR g/d MPR was regressed against DMI, with the residuals being RMPR

Of the 60 cattle that started the feedlot RFI test, 59 completed the RFI test, and 41 had methane phenotype records for the feedlot test phase. Fifty-nine cattle had data recorded for the chamber grain test, and 57 had data recorded for the chamber roughage test. Associations among traits were examined using Pearson correlations. Simple regression analyses were conducted to evaluate the relationships between the animal performance and feed efficiency traits measured in the feedlot and the methane traits for each of the 3 methane test phases. There were moderate negative correlations between RFI-EBVmp and test BW in each of the 3 test phases, and the cattle with lower RFI-EBVmp were heavier across the duration of the experiment. To reduce the confounding effect that differences in BW could have on the magnitude of the associations between RFI-EBVmp and the methane traits, an association coefficient equivalent to an r-value was calculated as the square root of the type III sum-of-squares (SS) for RFI-EBVmp divided by the total SS from general linear models with Test BW fitted before RFI-EBVmp.


RESULTS

Descriptive statistics for animal performance, feed efficiency, and methane traits for the Angus cattle in this experiment are presented in Table 3, and individual animal data for the 3 methane tests phases are shown in Fig. 2. The cattle were older and heavier across the test phases of the experiment, but their DMI and MPR did not increase proportionally as feed offered was restricted in the respiration chamber tests. Mean methane yield (MY) was lowest in the feedlot phase under ad libitum grain feeding and increased as feed offered was restricted in the chamber grain test to 70% of that consumed individually over last 2 wk in feedlot. The MY was higher again in the restricted (to just above expected maintenance) chamber roughage test. Inspection of the relationship between MPR and DMI (Fig. 2) shows that MPR increased approximately linearly with DMI for both the feedlot grain ration and the roughage ration but increased roughly twice as quickly for the roughage ration. The rate of increase, measured as the regression coefficient (b ± SE), in the chamber roughage test was much higher (16.4 ± 2.9 g/kg) than in either the feedlot test (4.1 ± 2.2 g/kg) or the chamber grain test (8.4 ± 1.7 g/kg).


View Full Table | Close Full ViewTable 3.

Descriptive statistics for animal performance, feed efficiency, and methane emission traits of Angus cattle in the 3 methane test phases

 
Animal and methane test type1 Units Mean SD Minimum Maximum
Animal background information2
    Sire (n = 4) RFI-EBV kg/d 0.11 1.10 −0.97 1.42
    Accuracy of sire RFI-EBV % 68 6 59 71
    Dam RFI-EBV kg/d −0.05 0.69 −1.03 1.25
    Accuracy of dam RFI-EBV % 61 6 53 74
    Midparent RFI-EBV kg/d 0.03 0.83 −1.00 1.12
    Start BW kg 410 36 342 499
Feedlot test3
    Start of test age d 579 16 537 606
    Test BW kg 454 38 383 548
    ADG kg/d 1.40 0.20 0.97 1.95
    DMI kg/d 12.2 1.1 9.9 14.8
    FCR kg/kg 8.9 1.4 5.8 12.0
    RFI kg/d 0.0 0.8 −1.8 1.9
    MPR g/d 144 16 118 190
    MI g/kg 103 14 72 124
    MY g/kg 11.7 1.5 9.5 15.9
    RMPJ g/d 20 17 −7 66
    RMPR g/d −2 16 −21 44
Chamber grain test4
    Start of test age d 687 15 648 711
    Test BW kg 519 49 413 646
    DMI kg/d 7.2 1.5 3.6 11.6
    MPR g/d 106 23 66 177
    MY g/kg 15.0 3.1 10.4 23.3
    RMPJ g/d 34 20 2 101
    RMPR g/d 0 19 −32 68
Chamber roughage test5
    Start of test age d 722 18 684 759
    Test BW kg 544 43 460 640
    DMI kg/d 7.6 0.7 5.4 8.7
    MPR g/d 144 18 99 195
    MY g/kg 19.0 1.9 15.7 25.6
    RMPJ g/d −8 15 −35 38
    RMPR g/d 0 14 −26 44
1Test BW = test period BW; FCR = feed conversion ratio; RFI = residual feed intake; MPR = methane production rate; MI = methane intensity; MY = methane yield; RMPJ and RMPR = residual methane production with expected MPR calculated using Johnson et al. (1995) and regression of MPR on DMI, respectively.
2RFI-EBV denotes EBV for residual feed intake.
3n = 41, comprising 22 steers and 19 heifers.
4n = 59, comprising 29 steers and 30 heifers.
5n = 57, comprising 28 steers and 29 heifers.
Figure 2.
Figure 2.

Methane production rate (MPR) vs. feed DMI by Angus cattle measured first on ad libitum access to feedlot grain ration (MPR = 4.1 [93.3; SE] × DMI + 93.3; R2 = 0.08; N = 41: red circles), then restricted access to grain ration (MPR = 8.4 [45.6] × DMI + 45.5; R2 = 0.30; N = 59: blue squares), and then restricted allocation of roughage ration (MPR = 16.4 [19.2] × DMI + 19.2; R2 = 0.37; N = 57: yellow triangles).

 

Within-Test Correlations among Methane Traits

Within each of these 3 test phases the correlations between MPR and the other methane emission traits were all positive and strong (Table 4), indicating that reduction in the derived traits would be accompanied by reduction in MPR. The ranking of the magnitude of the correlation between MPR and the ratio trait MY and between MRP and the residual traits (RMPJ and RMPR) was similar within each test, with that with MY being lowest and that with RMPR being highest. The magnitude of the correlations among these 3 traits (MY, RMPJ, and RMPR) was very strong (r = 0.85 to 0.90) and similar within each test.


View Full Table | Close Full ViewTable 4.

Phenotypic correlation coefficients among methane traits within each of the 3 methane test phases1

 
Methane test and traits2 MY MI RMPJ RMPR
Feedlot test3
    MPR 0.71 0.57 0.78 0.95
    MY 0.33 0.99 0.89
    MI 0.38 0.51
    RMPJ 0.94
Chamber grain test4
    MPR 0.42 0.76 0.82
    MY 0.89 0.85
    RMPJ 0.99
Chamber roughage test5
    MPR 0.69 0.68 0.79
    MY 1.00 0.99
    RMPJ 0.99
1All the correlation coefficients are significantly (P < 0.05) different from zero.
2MPR = methane production rate; MI = methane intensity; MY = methane yield; RMPJ and RMPR = residual methane production with expected MPR calculated using Johnson et al. (1995) and regression of MPR on DMI, respectively.
3n = 41 animals.
4n = 59 animals.
5n = 57 animals.

Across-Test Correlations among Methane Traits

Correlation coefficients for the methane emission traits across the 3 test phases of the experiment are presented in Table 5. It should be noted that across the methane test phases the feed type, amount of feed, and digestibility of feed offered were different. For MPR the correlations across tests ranged from r = 0.27 to 0.47. For the methane emission traits adjusted for DMI, the correlations for MY, RMPJ, and RMPR between the feedlot test and the chamber grain test (r = 0.40 to 0.43) and between the chamber grain test and the chamber roughage test were moderate (r = 0.36 to 0.41), and they were moderate to strong between the feedlot test and the chamber roughage test (r = 0.54 to 0.58).


View Full Table | Close Full ViewTable 5.

Phenotypic correlation coefficients for methane traits across the 3 methane test phases

 
Chamber grain test
Chamber roughage test
Trait1 MPR MY RMPJ RMPR MPR MY RMPJ RMPR
Feedlot test
    MPR 0.27a,2 0.46b,3
    MY 0.40b 0.54b
    RMPJ 0.43b 0.58b
    RMPR 0.43b 0.58b
Chamber grain test
    MPR 0.47b,4
    MY 0.36b
    RMPJ 0.38b
    RMPR 0.41b
aCorrelation coefficients are significantly different from zero at P < 0.10.
bCorrelation coefficients are significantly different from zero at P < 0.05.
1MPR = methane production rate; MY = methane yield; RMPJ and RMPR = residual methane production with expected MPR calculated using Johnson et al. (1995) and regression of MPR on DMI, respectively.
2n = 41 animals.
3n = 40 animals.
4n = 57 animals.

Associations with Feed Efficiency

There was substantial variation in feed efficiency between animals in the feedlot test, with a greater than 2-fold range in FCR and over a 3 kg/d difference in RFI observed between animals (Table 3). The associations between animal performance and feed efficiency traits are presented in Table 6. Both measures of efficiency were positively correlated with each other, so the same animals tended to be superior for both traits, but the correlation was moderate (r = 0.41), implying the 2 traits were not identical. Residual feed intake was independent of both ADG and test BW, whereas FCR was strongly correlated with ADG but independent of only test BW. As a measure of the genetic variation in RFI, RFI-EBVmp was positively correlated with both feedlot RFI and FCR. However, unlike RFI, the RFI-EBVmp was correlated with both ADG and test BW, indicating that the genetically more efficient cattle (those with lower EBVs) had heavier test BW across the duration of the experiment. Genetically efficient cattle also tended (P < 0.10) to have lower DMI.


View Full Table | Close Full ViewTable 6.

Phenotypic correlation coefficients between animal performance and feed efficiency traits in the feedlot test

 
Trait1 Test BW DMI RFI FCR RFI-EBVmp2
ADG 0.41b 0.17 0.00 −0.83b −0.37b
Test BW 0.67b 0.00 −0.03 −0.30b
DMI 0.73b 0.39b 0.24a
RFI 0.41b 0.56b
FCR 0.49b
aCorrelation coefficient is significantly different from zero at P < 0.10. n = 59 animals.
bCorrelation coefficient is significantly different from zero at P < 0.05. n = 59 animals.
1Test BW = test period BW; RFI = residual feed intake; FCR = feed conversion ratio.
2Denotes midparent EBV for residual feed intake.

The associations between the methane traits and the animal performance and feed efficiency traits across the 3 test phases of the study are presented in Table 7. In the feedlot test MPR was correlated with only test BW, likely because of the moderate to strong correlation (Table 6) between DMI and test BW. The moderate to strong correlation (r = −0.69 and 0.68) of methane intensity (MI) with ADG and FCR in the feedlot was expected as ADG is a component trait of the 2 ratio traits (MI and FCR). Both MY and RMPJ had similar moderate correlations (r= −0.39 to −0.54) with DMI, FCR, and RFI in the feedlot, whereas RMPR was not related to DMI or FCR and was only weakly (P < 0.10) related to RFI. In the chamber grain test, DMI was positively correlated with MPR and negatively correlated with MY. The correlations of the other methane traits with the animal performance and feed efficiency traits were not significant. In the chamber roughage test the animal performance and feed efficiency traits were negatively and weakly to moderately correlated (r = −0.27 to −0.49) with all the methane traits, except for DMI, which was positively and moderately to strongly correlated (r = 0.61) with MPR and not significantly correlated with MY and the RMP traits.


View Full Table | Close Full ViewTable 7.

Phenotypic correlation coefficients between methane traits and animal performance and feed efficiency traits

 
Methane test and traits1 MPR MI MY RMPJ RMPR
Feedlot test2
    ADG 0.19 −0.69b 0.24 0.23 0.22
    Test BW 0.47b −0.08 −0.05 0.05 0.28a
    DMI 0.28a 0.26 −0.48b −0.39b −0.04
    FCR −0.01 0.68b −0.47b −0.42b −0.22
    RFI −0.05 0.16 −0.54b −0.52b −0.29a
    Midparent RFI-EBV3 −0.09 0.21 −0.22 −0.25 −0.18
    Midparent RFI-EBV4 0.08 0.30a −0.27a −0.27a −0.09
Chamber grain test5
    DMI 0.55b −0.50b −0.13 −0.02
    Feedlot FCR −0.20 −0.12 −0.21 −021
    Feedlot RFI −0.05 −0.19 −0.23 −0.21
    Midparent RFI-EBV3 −0.04 −0.10 −0.13 −0.12
    Midparent RFI-EBV4 0.01 −0.10 −0.11 −0.01
Chamber roughage test6
    DMI 0.61b −0.14 −0.17 0.00
    Feedlot FCR −0.31b −0.29b −0.27b −0.30b
    Feedlot RFI −0.37b −0.48b −0.49b −0.49b
    Midparent RFI-EBV3 −0.35b −0.39b −0.39b −0.40b
    Midparent RFI-EBV4 −0.15 −0.35b −0.36b −0.33b
aCorrelation coefficient is significantly different from zero at P < 0.10.
bCorrelation coefficient is significantly different from zero at P < 0.05.
1Test BW = test period BW; FCR = feed conversion ratio; RFI = residual feed intake; MPR = methane production rate; MI = methane intensity; MY = methane yield; RMPJ and RMPR = residual methane production with expected MPR calculated using Johnson et al. (1995) and regression of MPR on DMI, respectively.
2n = 41 animals.
3Denotes midparent EBV for residual feed intake.
4Coefficents recalculated after first fitting Test BW in the linear models; see Materials and Methods for further explanation.
5n = 59 animals.
6n = 57 animals.

Midparent EBV for RFI was not associated with MPR or MI under ad libitum feeding conditions in the feedlot test or with MPR in the chamber grain test but was significantly negatively associated with MPR in the chamber roughage test (Table 7). There were moderate negative correlations between RFI-EBVmp and test BW in each of the 3 test regimes, and the cattle with lower RFI-EBVmp were heavier across the duration of the experiment. When the associations for methane traits with RFI-EBVmp were recalculated after first fitting Test BW in the linear models, there was no significant association between RFI-EBVmp and MPR in any of the 3 test phases and a trend toward a positive association with MI, that is, for lower RFI-EBVmp to be associated with lower methane intensity. Feedlot RFI is phenotypically independent of Test BW and ADG by definition and was not associated with MPR or MI in the feedlot test or with MPR in the chamber grain test but was significantly negatively associated with MPR 4 mo later in the chamber roughage test. For methane emission traits calculated relative to feed intake, the association for RFI-EBVmp, recalculated after first fitting Test BW, trended to be positive with MY and RMPJ in the feedlot test and was positive with MY, RMPJ, and RMPR in the chamber roughage test. Lower feedlot RFI was associated with higher methane production relative to feed intake (MY, RMPJ, and RMPR) in both the feedlot test and the chamber roughage test.


DISCUSSION

This experiment provided individual animal enteric methane emission data under feedlot and chamber test conditions with which to determine the correlations among methane emission traits across a range of feeding regimes and 2 methane measurement technologies. In terms of total daily methane emission, the expectation from previous studies (Blaxter and Clapperton, 1965; Johnson et al., 1995) is that bigger, faster-growing cattle will generally eat more and produce more enteric methane than smaller, slower-growing cattle under the same feeding regime. This expectation was confirmed in this experiment, in which in all 3 test phases MPR was strongly positively correlated with DMI. Feeding or animal breeding strategies to directly lower MPR, although desirable from a greenhouse gas emissions perspective, risk being associated with lower feed intake, which may be undesirable from an animal production perspective. In the animal house chamber tests DMI during the restricted grain test (7.2 kg) was similar to that during the restricted roughage test (7.6 kg); however, the difference in the slope of the regression of MPR on DMI for the lower-digestibility restricted roughage diet (16.4 ± 2.9) was almost 2-fold that for the restricted grain diet (8.4 ± 1.7). This result is in agreement with expectations from the methane production prediction equations of Blaxter and Clapperton (1965), which take into account the digestibility of the feed.

In evaluating the strength of the correlations among methane traits across the 3 methane test phases, the test phase differences in age and BW of the animals, diet, and the measurement methodologies (GEM vs. respiration chamber) should be borne in mind. In the feedlot test BW data over 70 d were available, allowing ADG and MI to be calculated, but ADG could not be determined over the short duration of the respiration chamber tests, so the repeatability of MI across the 3 tests could not be determined. In the feedlot and chamber test phases, DMI data were available, allowing MY and the RMP traits (RMPJ and RMPR) to be calculated. The correlations for the latter 3 traits describing methane production relative to DMI between the feedlot test and the chamber grain test and between the chamber grain test and the chamber roughage test were moderate (r = 0.36 to 0.43) but were moderate to strong between the feedlot test and the chamber roughage test (r = 0.54 to 0.58). These results show that the same animals are statistically more likely to have higher or lower emissions relative to their feed intake over time and across a range of nutritional regimes, from a feedlot grain ration to a roughage diet.

To date, the most comprehensive sets of genetic parameters for methane emission traits in beef cattle (Donoghue et al., 2016a) and sheep (Pinares-Patiño et al., 2013) are based on animals tested for methane under a restricted roughage ration in respiration chambers. It is very likely that in practice the sources of methane emission data will be from animals measured in respiration chambers as well as those measured in feedlots. The moderate positive correlations (r = 0.54 to 0.58) between the feedlot test and chamber roughage test for MY, RMPJ, and RMPR provide evidence that measurements under the 2 systems (GEM vs. chamber) are related. This result is encouraging in light of the fact that in beef cattle the repeatability of respiration chamber methane tests conducted within 60 d of the first test was moderate to strong (r = 0.59 to 0.91), with corresponding moderate (r = 0.30 to 0.44) phenotypic correlations (Donoghue et al., 2016b). In the current study the chamber roughage test was conducted 73 d after the end of the feedlot grain methane test. Similar repeatability results were obtained for respiration chamber methane tests in sheep (Pinares-Patiño et al., 2013). Also, genetic correlations are usually higher than phenotypic correlations. For example, in the review by Koots et al. (1994) across multiple large experiments, the correlation between BW at weaning and mature BW was 0.37, whereas the underlying genetic correlation was 0.66. Hence, the expectation is that the genetic correlation between the methane emission traits measured in the feedlot and those in the chamber roughage test will be higher than the 0.54 to 0.58 phenotypic correlations. Further, genetic improvement based on testing sires or their progeny in a feedlot test will lead to improvement in animals tested on roughage diets and vice versa, and genetic improvement based on testing sires or their progeny on a roughage diet will lead to improvement in animals tested on feedlot ration. More research is required to estimate these genetic correlations.

From the perspective of knowledge of correlations between methane traits within test phases or of knowledge of associations between methane traits across test phases, the ratio trait MY could be replaced by the residual traits RMPJ and RMPR without loss of information. The use of ratio traits is generally avoided in animal breeding, and RMPJ or RMPR could be used in place of MY to avoid the use of the ratio traits, a conclusion also reached for MY by Herd et al. (2014) on the basis of a much larger data set. Although the 2 RMP traits were calculated by different formulas, they were highly correlated, and either could be used, but RMPR calculated using the relationship between MPR and DMI from within the test data generally had higher correlations with the other traits than RMPJ, calculated using a more general prediction formula.

A review of breeding strategies to reduce methane emissions in beef cattle concluded that low-RFI (feed-efficient) cattle produce less MPR per unit of product under ad libitum feeding (as reviewed by Arthur et al., 2009). In this experiment midparent EBV for RFI was not associated with MPR or MI under ad libitum feeding conditions in the feedlot test or with MPR in the chamber grain test but was significantly negatively associated with MPR in the chamber roughage test. There were moderate negative correlations between RFI-EBVmp and test BW in each of the 3 test phases, and the cattle with lower RFI-EBVmp were heavier across the duration of the experiment. By chance, there was a small difference in 400dWT-EBV between the low-RFI and high-RFI sires used to produce the animals in this experiment, which may explain the observed association of higher test BW with lower RFI-EBVmp. When the associations for methane traits with RFI-EBVmp were recalculated after first fitting Test BW in the linear models, there was no significant association between RFI-EBVmp and MPR in any of the 3 test phases, and there was a trend toward a positive association with MI, that is, for lower RFI-EBVmp to be associated with lower methane intensity. For methane emission traits calculated relative to feed intake, the association with RFI-EBVmp, recalculated after first fitting Test BW, tended to be positive for MY and RMPJ in the feedlot test and was positive for MY, RMPJ, and RMPR in the chamber roughage test. Lower feedlot RFI was associated with higher methane production relative to feed intake (MY, RMPJ, and RMPR) in both the feedlot test and the chamber roughage test. These results indicate that in young growing cattle lower RFI is not always associated with methane production and methane intensity, but as also reported by Freetly and Brown-Brandl (2013), it can be associated with higher methane production per unit feed intake.

This experiment has shown that the same animals are statistically more likely to have higher or lower methane production relative to their feed intake over time and across a range of nutritional regimes, from a feedlot grain ration to a roughage diet. The phenotypic correlations for these traits (methane yield and residual methane production) were strong across time (approximately 4 mo between the middle of the feedlot test and the chamber roughage test), across different feeding regimes (ad libitum and restricted), and across different ration types (feedlot grain and roughage). The underlying genetic correlations are likely to be stronger, so that genetic improvement based on testing sires or their progeny in a feedlot test will lead to improvement in animals tested on roughage diets and vice versa, and genetic improvement based on testing sires or their progeny in a roughage diet will lead to improvement in animals tested on feedlot ration. The utility of breeding strategies to reduce methane emissions in beef cattle by selection for low RFI requires further research into its effectiveness in the cattle production system cattle to which it might be applied.

 

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.