Emission of Gas and Dust from Livestock
EFFECT OF FEEDING STRATEGIES ON METHANE EMISSIONS OF DAIRY COWS EVALUATED BY MIR SPECTROMETRY
LESSIRE, F.1, SCOHIER, C.1, PRÉVOT, A.1, SOYEURT, H.2, DUFRASNE, I.1
1 Fundamental and applied research on animal and health, Animal Production Department, Faculty of Veterinary Medicine, University of Liège, Quartier Vallée 2, Avenue de Cureghem, 6, 4000 Liège, Belgique
2 Applied Statistics, Computer Science and Modeling Unit, AGROBIOCHEM Department, Gembloux Agro-Bio Tech, University of Liège, Passage des déportés 2, 5030 Gembloux
ABSTRACT: Reduction of methane emissions by 20% is one of the objectives of the Horizon 2020 policies of the European Commission. Yet livestock is considered responsible for 18% (dairy) and 55% (beef) of agricultural enteric methane emissions. It is thus necessary to improve the liability and the eased assessment of the cattle emissions on a large scale to determine levers of action to decrease them and quantify the impacts of these actions. The aim of this study was to estimate methane emissions of a dairy herd by MIR spectrometry and to follow up the cows on an individual basis during the winter period. Different concentrate compositions supplementing the total mixed ration (TMR) given to the herd were tested regarding their effect on predicted methane emissions. No effect was statistically noted and factors which might have influenced this result are discussed hereafter.
Keywords: methane, methane prediction, mitigation strategy, dairy cattle
INTRODUCTION: Livestock is considered to contribute significantly to green-house gases (GHG) emissions by producing enteric methane during ruminal fermentation. The dairy and beef sectors are estimated respectively responsible for 18% and 55 % of enteric methane emissions (Tubiello et al., 2014). Yet, a reduction of GHG emissions by 20% is required by the European Commission to comply with Horizon 2020 objectives. To reach this objective, it is necessary to develop methods of methane assessments liable and applicable both on an individual basis and on a large scale. The potential of mid infra-red spectra analysis to predict methane emissions has been largely highlighted (Vanlierde et al., 2015). The possibility to follow up individual cows and herds by this non-invasive and cheap method might be useful to evaluate the impacts of mitigation strategies. The aim of this study was to assess the impact of feeding strategies on methane emissions by following up the methane emissions on an individual basis of a dairy herd over a 138 d- period corresponding to the winter period. Increasing the starch content or the fat content of cows’ ration has been described as effective to lower methane emissions expressed in g.kg dry matter intake-1 (compound rich in starch – rich in fat)and in g. d
-1(compoundrich in fat). Both strategies were tested: during the first trial a concentrate rich in starch (32%) was provided to the cows and in the second one the concentrate fat content was increased from 4% to 9%. Results of the MIR predictions are presented and discussed.
Modelling 1. MATERIAL AND METHODS
1.1. The study was conducted for 138 days at the Experimental Farm of University of Liège (Liège, Belgium) on a herd of 54 Holstein dairy cows milked by an automatic milking system Lely A3.
1.2. Experimental design:The cows received a total mixed ration (TMR) whose composition was 30% maize silage, 35% grass silage, 11% beet pulp silage, 6% brewers, 7% dried forage (hay+straw) and 11% a mixed compound rich in protein (35% PB). The TMR was completed by concentrates of variable composition (Trial 1:AT2: Fat: 4%- Starch 32% - Trial 2:AT3: Starch 21.7% -Fat: 9%) supplied at milking. Zootechnical performances and methane emissions of groups receiving these 2 concentrates were compared with those of a group receiving a control concentrate (AT1: 4% Fat – 18 % starch). Thus, the herd was randomly divided into 2 groups of at least 16 cows balanced with regards to DIM, lactation number and MY. Provided concentrate (kg.cow-1.day-1) was individually calculated on basis of DIM and lactation number. Each feeding trial was divided into 2 periods. Each test period lasted for one month, including a 2-weeks transition period and milk samplings during the 3d and 4th week. During the first period (Period 1), one group received the control concentrate (AT1) and the other, the tested one (AT2 or AT3). After 1 month, the groups were switched so that the control group received the tested feed and inversely (Period 2). Each cow became thus her own control. Description of the groups at the beginning of each trial is presented on Table 1.
Table 1. Description of the groups tested during the 2 feeding trials.
Trial 1
Concentrate rich in starch
Trial 2 Concentrate rich in fat
AT1 AT2 AT1 AT3
Nbr of cows 16 17 17 17
Days in milk 119 ± 108 116 ± 124 97 ± 79 83 ± 54
Lactation number 2.5 ± 1.6 3.0 ± 1.8 2.6 ± 1.7 3.1 ± 1.8 Milk yield (kg.cow-1.d-1) 22.7 ± 7.0 22.9 ± 9.4 31.1 ± 6.9 33.2 ± 7.0
-1 -1 3.4 ± 1.8 3.4 ±1.8 5.3 ± 1.7 5.1 ± 1.7
Emissions of Gas and Dust from Livestock – Saint-Malo, France – May 21-24, 2017 90
methane.day-1), methane emissions.kg-1 produced milk and methane emissions.kg energy corrected milk (ECM) -1 of the groups. Fixed effects included in the model were:
group, period, period X group and DIM. The repeated statement was based on the cow identification. Descriptive statistics and the repeated models were computed using SAS 9.3 software.
2. RESULTS AND DISCUSSION: Results showed a large range of variation in MIR methane emissions among observations (448 ± 57 g.day-1; min: 246 g.day-1; max: 582 g.day-1).
Cows were then classified on basis of their emission rate, considering a cow as low emitter for methane emission < 413 g.cow-1.day-1, and high emitter for methane values >
487 g.cow-1.day-1, those values corresponded to 25% and 75% quantile of the MIR methane values distribution. It appeared that one cow being low emitter at one time could be high emitter in the following period independently from the tested concentrate (Figure 1).
2.1. Results of the feeding trials: No statistically significant decrease in methane emissions (Tables 2-3) could be detected during the feeding trials. However in the trial 2, methane.kg milk-1 and methane.ECM-1 were numerically lower inthe group receiving the concentrate rich in fat. Different factors could be invoked: the inter-individual differences and the standard error of calibration and of cross validation, respectively estimated at 66 g.day-1 and 70 g.day-1might have interfered with the effects of the tested compound (Vanlierde et al., 2016). The concentrate to roughage ratio could have been too small to influence markedly the ruminal fermentation. The length of the measurement period (138 days) induced in some cows prolonged lactations with lower MY and lower concentrate consumption that might have been insufficient to impact methane emissions. Late lactation stages caused significant differences in methane emission rates.cow-1day-1. This effect is reported and taken into account in the prediction equation (Vanlierde et al., 2015).
Table 2. Results (LSMeans ± SE) of the first feeding trial.
N= 99 AT1 AT2 Group Period DIM
***: p<0.001;**: p<0.01; *: p<0.05; trend: p=0.1; ns = not significant. Effect of interaction group X period was ns and thus not showed.
Modelling
Table 3. Results (LSMeans ± SE) of the second feeding trial.
N = 167 AT1 AT3 Group Period DIM
Milk yield (kg.cow-1.d-1) 28.1 ± 1.3 31.0 ± 1.4 trend ns ns ECM (kg.cow-1.d-1) 27.6 ± 1.3 29.8 ± 1.4 ns ns ns
Methane (g.cow-1.d-1) 461 ± 7 459 ± 7 ns ns *
Methane (g.kg milk-1) 17.4 ± 0.8 15.9 ± 0.9 ns ns ns Methane (g.kg ECM-1) 17.8 ± 0.8 16.6 ± 0.9 ns ns ns
Abbreviations: AT1: control concentrate – AT3: concentrate rich in fat. ECM: energy corrected milk; ***:
p<0.001;**: p<0.01; *: p<0.05; trend: p=0.1; ns = not significant. Effect of interaction group X period was ns and thus not showed.
Figure 1: Variations of predicted methane on the same cows throughout the trials
3. CONCLUSION: In this study, MIR spectrometry analysis was used to predict methane emissions of a herd on an individual basis over a period corresponding to the winter period. No effect of the concentrate composition was highlighted. Further study is needed to investigate the implication of the different evoked factors susceptible to have
Emissions of Gas and Dust from Livestock – Saint-Malo, France – May 21-24, 2017 92
Vanlierde A, Vanrobays M-L, Dehareng F. Froidmont, E., Soyeurt, H., McParland H, Lewis, E., 2015. Hot topic: Innovative lactation-stage-dependent prediction of methane emissions from milk mid-infrared spectra. Journal of Dairy Science, Volume 98 , Issue 8 , 5740 – 5747
Vanlierde A, Vanrobays M-L, Gengler N, Dardenne, P., Froidmont, E., Soyeurt, H., McParland S. Lewis, E., Deighton, M.H., Mathot, M., Dehareng, F., 2016. Milk mid-infrared spectra enable prediction of lactation-stage-dependent methane emissions of dairy cattle within routine population-scale milk recording schemes, Anim Prod Sci.;
56(3):258-264
Modelling
INNOVATIVE DATABASE AND ITS POTENTIAL TO REALISE LARGE SCALE STUDY TO