Within-and between-animal variance in methane emissions in non-lactating dairy cows

Several studies on methane (CH4) emissions have focussed on selecting high and low CH4-emitting animals. One challenge faced by this work is the lack of consistency, or repeatability, in animal rankings over time. Repeatability for individual animals over time needs to be high to reliably detect high and low CH4-emitting animals. A possible explanation for the lack of repeatability is a relatively high within-animal variation in daily CH4 emissions, meaning that animals could then change their ranking when compared at different points in time. An experiment was undertaken with four non-lactating dairy cattle to assess the within-and between-animal variation in CH4 emissions over time when measured using the sulfur hexafluoride (SF6) tracer technique. Two contrasting diets were fed to the cattle at maintenance energy levels: lucerne silage (diet 1) and a cereal + lucerne + straw mixed ration diet (diet 2). Daily CH4 measurements were undertaken for 23 days on diet 1 and 30 days on diet 2. There was a significant (P < 0.001) difference between diet 1 and diet 2 in daily CH4 production, with mean (+/- s.e.) production of 124.3 (11.1) g CH4/day from diet 1 and 169.8 (+/- 11.0) g CH4/day from diet 2. Lower CH4 yield (g CH4/kg dry matter intake) was recorded on diet 1 (22.8 +/- 2.0) than diet 2 (32.0 +/- 2.0). Cows differed significantly (P < 0.05) from one another in daily CH4 yield (diet 1: cow 1 = 19.4 +/- 0.6, cow 2 = 22.2 +/- 0.8, cow 3 = 23.2 +/- 0.7, cow 4 = 25.4 +/- 0.6; diet 2: cow 1 = 26.0 +/- 0.7, cow 2 = 36.4 +/- 0.7, cow 3 = 29.3 +/- 0.7, cow 4 = 36.6 +/- 0.7). Variances for daily CH4 yield were smaller for diet 1 ( within animal = 6.91, between animals = 6.23) than for diet 2 (within animal = 10.09, between animals = 27.79). Estimates of repeatability (variation between animals/total variation) for daily CH4 yield were 47 and 73% in diet 1 and 2, respectively. Coefficients of variation in average daily CH4 emissions in this experiment ranged from 8 to 18% despite the fact that each animal received the same quantity and quality of feed each day. While further research is required, the high within-animal variability in CH4 emissions measured using the SF6 tracer technique may explain why there has been difficulty in obtaining consistent rankings in CH4 yields when animals are measured on multiple occasions. The results also suggest that the SF6 tracer technique may exaggerate apparent between animal differences in CH4 emissions.