Methane emission of dairy cows reflected by sensor measurements

Reduction of methane emission by dairy cows through breeding requires large-scale collection of phenotypic data on methane emission. Such large amounts of data can be obtained by using methane sensors in automatic milking systems. Measurements recorded by these sensors can be used to generate different traits that reflect methane emission. The aim of this study was to compare these different traits with methane emission recorded in climate respiration chambers (CRC). Data was obtained from 20 dairy cows that were housed individually in one of two CRC for five days each. Inlet and exhaust air of each CRC was quantified and sampled for methane (CH4) and carbon dioxide (CO2) concentrations using a non-dispersive infrared method at 10-min intervals. Each CRC was equipped with a Fourier transform infrared sensor in the upper part of the feeding trough. The same sensors have also been installed in automatic milking systems for collection of methane data on commercial dairy farms. These sensors recorded CH4 and CO2 twice per second. Preliminary analyses suggest correlations of 0.8 between CH4 obtained from sensor measurements and CH4 obtained from CRC measurements, and correlations of 0.6 between CH4/CO2 obtained from sensor measurements and CH4 obtained from CRC measurements. These results indicate that a substantial part of methane emission by dairy cows can be measured with sensors that can be installed in automatic milking systems. This enables large-scale collection of methane emission data. Further analysis is in progress, focussing on different traits that reflect methane emission, the influence of several fixed factors, and the impact of recording data only during a few short timeframes per day.