Introduction Many potential approaches for reducing methane (CH 4 ) emissions from cattle are at present under development. The promising ones will require verification of their efficacy at the production scale (e.g. paddock scale). We report an experiment to test whether a difference in CH 4 emissions between two grazing groups of cattle could be detected by 1) a micrometeorological technique using line-averaged concentrations up-and downwind of the cattle, and 2) by the sulfur hexafluoride (SF 6 ) tracer-ratio technique.
Material and MethodsTwo groups, of 30 one-year old Hereford x Friesian steers each, were formed with equal mean liveweight. In a flat uniform paddock composed predominantly of ryegrass (Lolium perenne), 32 rectangular strips were fenced, each 40 m by 25 m in size. Paired strips were allocated to the two groups on a daily basis, such that one group was always 65 m north of the other. First, for 6 days (Period 1) no treatment was applied, in order to test whether the emissions from the two groups were indistinguishable. For the following 10 days (Period 2), the grazing strip for the Northern group was sprayed with canola oil at a rate of 120 L ha −1 . This was expected to cause a reduction in CH 4 emissions, compared to the Southern group which did not receive any oil. For 3 days in Period 1 and for 4 days in Period 2, individual CH 4 emissions over 24 h from all steers were measured with the SF 6 tracer-ratio technique (Johnson et al., 1994). Group dry-matter intakes (DMI) were estimated from daily platemeter measurements before and after grazing. Individuals' DMI were estimated in Period 2, based on faecal outputs and invitro feed digestibility. Faecal output was estimated using titanium dioxide (TiO 2 ) as external faecal marker (Pinares-Patiño et al., 2008). CH 4 emission rates were tested for group effects with ANOVA. Concentrations of CH 4 in air were measured along lines parallel to the W and E fences of each rectangle, as 20-minute averages. For this, air was drawn into 44 m long perforated alkythene pipes mounted 0.7 m above ground, and from each of these four intake pipes via a switching manifold into a CH 4 analyser (DLT-100, Los Gatos Research, Mountain View, California, USA). The collective CH 4 emissions from each group were computed from downwind-upwind concentration differences, using a backward-Lagrangian stochastic model (WindTrax software, www.thunderbeachscientific.com; Flesch et al., 2004). Periods of unsuitable wind direction (outside ±40°of due W or due E), low wind speed or systematically turning direction were excluded from the analysis. The differences in CH 4 emissions between the groups, for each valid 20-min run, were the input data for statistical analysis with a linear mixed-effects model, using days as the random effect.
ResultsIn Period 1 (no treatment), CH 4 emissions from the two groups did not differ, according to both the micrometeorological technique (Table 1) and the SF 6 technique ( Table 2). The groups' DMI did not differ, either. By contrast, in Period...