Methane and CO2 emissions of grazing cows measured by eddy covariance technique

The eddy covariance (EC) technique has been extensively used for CO₂ and energy exchange measurements over various ecosystems. The development of fast CH₄ (and N₂O) measurement devices in the recent years opened the field to investigate CH₄ exchange with this technique. EC measurements represent a spatially integrated flux over an upwind area (footprint). Whereas for extended homogenous areas EC measurements work well, the animals in a grazing system are a challenge as they represent moving sources that create inhomogeneous conditions in space and time. Emissions from the inhomogeneous source pattern induced by the movement of the cows can be corrected if the position is known. But, data on the position is usually lacking.

We investigated the ability of EC flux measurements to reliably quantify the contribution of the grazing animals to the CH₄ and CO₂ exchange over pasture systems. For this purpose, a field experiment with a herd of twenty dairy cows in a full-day rotational grazing system was established in the Swiss central plateau. Net CH₄ and CO₂ exchange of the pasture system was measured continuously by the eddy covariance technique (Sonic Anemometer HS-50, Gill Instruments Ltd; FGGA, Los Gatos Research Inc.). To quantify the contribution of the animals to the net flux, the position of the individual cows was recorded using GPS (5 s time resolution) on each animal.

An existing footprint calculation tool (ART footprint tool) was adapted and CH₄ emissions of the cows were calculated. CH₄ emissions from cows could be used as a tracer to investigate the quality of the evaluation of the EC data, since the background exchange of CH₄ was very small. Daily mean CH₄ emissions compared well to emission values calculated based on animal weights and milk yields. Based on a corresponding quality analysis we investigated to which extent the presence of cows can be detected or missed in the CO₂ exchange measurements. For CO₂ a partitioning of the net flux was performed to separate the animal respiration flux from contributions of vegetation and soil (assimilation and respiration). The resulting animal related CO₂ emissions showed a considerable scatter but scaled with the animal density in the EC footprint.