OUTDOOR MEASUREMENT OF CATTLE METHANE EMISSIONS USING THE EDDY-COVARIANCE TECHNIQUE IN COMBINATION WITH GEOLOCALIZATION DEVICES

Methane emissions account for 8% of the EU-15 GHG emissions and livestock generates approximately half of these emissions [1]. In order to improve emissions reporting and properly test mitigation options, techniques for measuring methane emissions from cattle must be developed and adapted to each management system. Among available micrometeorological methods, the use of eddy-covariance is still in its infancy [2] and its relevance and robustness for cattle flux estimation has still to be proved. On one hand, it is well adapted to seasonal grazing systems, is non-invasive, needs little animal handling and allows detection of daily emission patterns. On the other hand, it has the drawback of requiring cattle geo-localization and long periods of measurements (typically one month). In this study, we combined measured CH4 fluxes with a footprint model [3] and cattle positions (GPS devices) over several one-month campaigns at key periods in the grazing season in order to obtain CH4 emissions per cow at herd scale. Accelerometers were also added to the system for behaviour detection, opening the possibility of linking emissions to feeding behaviour. Measurements were performed and are still ongoing at the Dorinne Terrestrial Observatory in 2014/2015. The first campaign provided a mean emission per cow of 65±6 kg CH4.LSU-1.year-1. Cattle emission pattern was tightly linked with behaviour pattern, emissions being higher during and shortly after grazing (i.e. at dawn and dusk). Uncertainties linked to the method will be discussed and quantified (footprint model validity, geo-localization precision, eddy covariance corrections and filtering specificities linked to CH4 measurements). Compilation of data from multiple campaigns will allow quantification of the effects of forage quality, animal weight and lactating state on emissions per cow