Exploring flexible microbiome-driven breeding strategies to mitigate methane emissions in beef cattle

Microbiome-driven breeding (MDB), which utilises the abundances of the most informative ruminal microbial genes that are host-genetically correlated with methane emissions, has been proposed as a cost-effective strategy to breed cattle for reduced methane yield. In this study, we developed an MDB approach for methane production (CH4P) and analysed different strategies for integrating this methodology into the genetic evaluation of production traits. A Bayesian analysis using host genomics and whole metagenomic data was applied to a deeply phenotyped beef population. The genomic analysis of the data revealed that CH4P was strongly and unfavourably correlated with daily feed intake, average daily gain (ADG), and age-adjusted carcass weight. To mitigate these unfavourable correlations, we developed an MDB strategy that considered only microbial genes favourably correlated with both reduced CH4P and improved ADG (MDB41, where the number denotes the most informative microbial genes identified) and compared it to a scenario that included microbial genes targeting only CH4P (MDB43). Selection of the 10% best animals based on genomic estimated breeding values (GEBVs) using MDB43 resulted in a reduction in CH4P by 18% of its mean (GEBV accuracy = 0.72, HPD95%: -22% to -13%) and a 7% decrease in ADG (0.65, -3% to -10%). In contrast, equivalent intense selection using MDB41 led to a lower CH4P reduction of 14% (0.61, -19% to -10%) but a slight increase in ADG of 1% (0.58, -3% to 6%). We also integrated MDB into a multi-trait model alongside production traits and analysed the impact of selection on each of these traits on the correlated responses of the others. Based on these results, we developed a new selection criterion of residual methane emissions that did not negatively affect production traits. Our findings demonstrate that MDB can be flexibly incorporated into existing genetic evaluation systems to reduce methane emissions while simultaneously avoiding reduction in production traits