Search

Methane (CH4) emission is an important environmental trait in dairy cows. Breeding aiming to mitigate CH4 emissions require the estimation of genetic correlations with other economically important traits and the prediction of their selection response. In this study, test-day CH4 emissions were predicted from milk mid-infrared spectra of Holstein cows. Predicted CH4 emissions (PME) and log-transformed CH4 intensity (LMI) computed as the natural logarithm of PME divided by milk yield (MY). Genetic correlations of PME and LMI with traits used currently were approximated from correlations between estimated breeding values of sires. Values were for PME with MY 0.06, fat yield (FY) 0.09, protein yield (PY) 0.13, fertility 0.17; body condition score (BCS) –0.02; udder health (UDH) 0.22; and longevity 0.22. As expected by its definition, values were negative for LMI with production traits (MY –0.61; FY –0.15 and PY –0.40) and positive with fertility (0.36); BCS (0.20); UDH (0.08) and longevity (0.06). The genetic correlations of 33 type traits with PME ranged from –0.12 to 0.25 and for LMI ranged from –0.22 to 0.18. Without selecting PME and LMI (status quo) the relative genetic change through correlated responses of other traits were in PME by 2% and in LMI by –15%, but only due to the correlated response to MY. Results showed for PME that direct selection of this environmental trait would reduce milk carbon foot print but would also affect negatively fertility. Therefore, more profound changes in current indexes will be required than simply adding environmental traits as these traits also affect the expected progress of other traits.

Methane (CH4) emission is an important environmental trait in dairy cows. Breeding aiming to mitigate CH4 emissions require the estimation of genetic correlations with other economically important traits and the prediction of their selection response. In this study, test-day CH4 emissions were predicted from milk mid-infrared spectra of Holstein cows. Predicted CH4 emissions (PME) and log-transformed CH4 intensity (LMI) computed as the natural logarithm of PME divided by milk yield (MY). Genetic correlations of PME and LMI with traits used currently were approximated from correlations between estimated breeding values of sires. Values were for PME with MY 0.06, fat yield (FY) 0.09, protein yield (PY) 0.13, fertility 0.17; body condition score (BCS) –0.02; udder health (UDH) 0.22; and longevity 0.22. As expected by its definition, values were negative for LMI with production traits (MY –0.61; FY –0.15 and PY –0.40) and positive with fertility (0.36); BCS (0.20); UDH (0.08) and longevity (0.06). The genetic correlations of 33 type traits with PME ranged from –0.12 to 0.25 and for LMI ranged from –0.22 to 0.18. Without selecting PME and LMI (status quo) the relative genetic change through correlated responses of other traits were in PME by 2% and in LMI by –15%, but only due to the correlated response to MY. Results showed for PME that direct selection of this environmental trait would reduce milk carbon foot print but would also affect negatively fertility. Therefore, more profound changes in current indexes will be required than simply adding environmental traits as these traits also affect the expected progress of other traits.

Mitigating the proportion of energy intake lost as methane could improve the sustainability and profitability of dairy production. As widespread measurement of methane emissions is precluded by current in vivo methods, the development of an easily measured proxy is desirable. An equation has been developed to predict methane from the mid-infrared (MIR) spectra of milk within routine milk-recording programs. The main goals of this study were to improve the prediction equation for methane emissions from milk MIR spectra and to illustrate its already available usefulness as a high throughput phenotypic screening tool. A total of 532 methane measurements considered as reference data (430 ± 129 g of methane/day) linked with milk MIR spectra were obtained from 165 cows using the SF6 technique. A first derivative was applied to the MIR spectra. Constant (P0), linear (P1) and quadratic (P2) modified Legendre polynomials were computed from each cows stage of lactation (days in milk), at the day of SF6 methane measurement. The calibration model was developed using a modified partial least-squares regression on first derivative MIR data points × P0, first derivative MIR data points × P1, and first derivative MIR data points × P2 as variables. The MIR-predicted methane emissions (g/day) showed a calibration coefficient of determination of 0.74, a cross-validation coefficient of determination of 0.70 and a standard error of calibration of 66 g/day. When applied to milk MIR spectra recorded in the Walloon Region of Belgium (≈2 000 000 records), this equation was useful to study lactational, annual, seasonal, and regional methane emissions. We conclude that milk MIR spectra has potential to be used to conduct high throughput screening of lactating dairy cattle for methane emissions. The data generated enable monitoring of methane emissions and production characteristics across and within herds. Milk MIR spectra could now be used for widespread screening of dairy herds in order to develop management and genetic selection tools to reduce methane emissions.