Abstract #596
Section: Production, Management and the Environment
Session: Production, Management, and the Environment II
Format: Oral
Day/Time: Tuesday 3:15 PM–3:30 PM
Location: Panzacola H-1
Session: Production, Management, and the Environment II
Format: Oral
Day/Time: Tuesday 3:15 PM–3:30 PM
Location: Panzacola H-1
# 596
Evaluating extant empirical models for predicting enteric methane emissions from lactating dairy cows.
J. A. D. R. N. Appuhamy*1, E. Kebreab1, 1Department of Animal Science, University of California, Davis, CA.
Evaluating extant empirical models for predicting enteric methane emissions from lactating dairy cows.
J. A. D. R. N. Appuhamy*1, E. Kebreab1, 1Department of Animal Science, University of California, Davis, CA.
Empirical models are widely used to estimate enteric methane (CH4) emissions from dairy cows worldwide. The objective of this study was to evaluate extant models for predicting CH4 emissions from dairy cows using literature data. Thirty-nine extant models developed based on dairy cow data were evaluated using measurements from 47 studies published after 2000. The data containing dietary, production and animal information included 50, 83, and 41 enteric CH4 measurements of lactating dairy cows from North America (NA), Europe (EU), and Australia and New Zealand (AUNZ), respectively. The models were evaluated using root mean square prediction error (RMSPE), concordance correlation coefficient, and Nash-Sutcliffe efficiency statistics. An index including equally weighted statistics was used to rank the models within each region. For NA, the 5 top ranked models were those by Hristov et al. (2013), Moe and Tyrrell (1979), Ellis et al. (2007), Moraes et al. (2014), and Moate et al. (2011) [Ref 1–5, respectively, in Table 1] and had RMSPE 15–17% of the average observed value. The majority of the best performing models in NA were developed on data from NA cows. A completely different Set of models performed best on both EU (RMSPE = 11–13%) and AUNZ (RMSPE = 11–15%) data. The best performing models were those by FAO (2010), IPCC (1997), Storlien et al. (2014), Yan et al. (2000), and Nielsen et al. (2014) [Ref I–V respectively in Table 1]. Regional origin and perhaps diet type of the data on which models have been developed need to be considered when selecting a model to predict CH4 emissions successfully.
Table 1. Top ranked models for predicting enteric CH4 (MJ/cow/d) emissions1
1DMI, NSC, HC = hemicellulose, C =cellulose (all in kg/d), FA and FA%= dietary fat (g/kg of DM and % of DM), DMd = diet DM digestibility (%), GEI and MEI = gross and metabolizable energy intake (MJ/d), NDF% (% of DM), BW (kg), MilkF = milk fat %.
NA | Ref | EU and AUNZ | Ref | |
= [2.54 + 19.14 × DMI] × 0.05565 | 1 | (0.0975 – 0.0005 × DMd) × GEI | I | |
= 3.14 + 0.51 × NSC + 1.74 × HC + 2.65 × C | 2 | 0.065 × GEI | II | |
= 4.08 + 0.068 × MEI | 3 | 1.47 + 1.28 × DMI | III | |
= –9.311 + 0.042 × GEI + 0.094 × NDF% – 0.381 × FA% + 0.008 × BW + 1.621 × MilkF | 4 | 3.234 + 0.0547 × GEI | IV | |
= [exp(3.15 – 0.0035 × FA)] × DMI × 0.05565 | 5 | 1.26 × DMI | V |