Milk fatty acids and rumen composition as proxy measures of enteric methane

Genomic selection is a powerful tool to reduce methane emissions in ruminants. However, it requires large-scale on-farm phenotypic measures of methane. Current technologies to measure methane emissions have several limitations and may not be suitable for lactating animals. Because enteric methane is closely linked to the fermentation process in the rumen, which in turn affects milk composition, breeding for low-methane ruminants may change the rumen microbial and milk composition. Consequently, these compositions may provide proxy measures of methane for use in selective breeding of low-methane ruminants. We investigated the effect on rumen and milk composition in sheep bred for divergent methane yield and the potential for generating proxy measures of methane emissions from rumen or milk samples in lactating ewes. Four hundred genotyped lactating ewes from a sheep research flock bred specifically for high and low-methane emissions had methane measured and rumen and milk samples collected approximately 6 wk post-lambing across 4 lactation years. Rumen samples were processed to generate VFA and metagenomic profiles of the rumen microbial community, and fatty acid profiles and mid-infrared spectra were generated for the milk samples. Although no differences in total fat, protein, and lactose percentages in the milk were found, the milk fatty acid profiles differed between methane selection lines, with higher PUFA and branched-chain fatty acids levels, and lower total SFA contents in ewes from the low-methane line. Lower proportions of acetate relative to propionate were found in the rumen samples from the low-methane ewes. Predictions of methane were obtained from the rumen VFA and metagenomic profiles and the fatty acid profiles and mid-infrared spectra from milk. These predictions formed the proxy methane measures and were heritable (between 0.12 to 0.36) and correlated (between 0.29 and 0.42) with the measured methane values. The genetic correlation between proxies and measured methane was between 0.52 and 0.71. The estimated efficiency of indirect selection for methane was higher for the milk sample proxies (49%-75%) than the rumen metagenomic profiles (45%-47%) and rumen VFA profiles (12%-38%). These results suggest that milk fatty acid, MIR spectroscopic, and rumen microbial composition phenotypes have the potential to be used as proxy measures of methane in lactating ruminants, with the milk-based proxies showing greater promise. Results show that the number of animals with methane proxy measures could be increased substantially and will enable access to breeding technology in countries with limited methane measurement infrastructure.