Experiences with joint modeling of feed efficiency and methane emission, using large-scale commercial data in Nordic Holstein.
Year: 2026
Type: conference-paper
Abstract: Enhancing feed efficiency and reducing methane (CH₄) emissions in dairy cows are of both biological interest and importance for breeding decisions, particularly to determine whether these traits are favorably or unfavorably associated. In the Nordic breeding goal, feed efficiency is represented through the Saved Feed index (residual feed intake and maintenance), and a genetic evaluation of CH₄ has recently been published. However, large-scale studies investigating the genetic relationship between feed efficiency and CH₄-particularly through joint modeling and the construction of residual traits-remain limited. The aim of this study was to explore the genetic interplay between genetic residual feed intake (gRFI) and CH₄ using joint modeling, based on commercial data from Nordic Holstein (HOL) herds. Dry matter intake (DMI) and body weight (BW) were recorded as weekly averages of daily measurements using the Cattle Feed Intake (CFIT) system from VikingGenetics. Monthly milk yield data were obtained from the official test-day scheme, and CH₄ concentrations were recorded as weekly averages of daily measurements using Guardian sniffers installed in milking robots. Across 10 herds with CFIT, 7 with sniffers, and 4 with both technologies, we collected 317,945 CFIT phenotypes from 6,344 cows, 115,752 sniffer phenotypes from 3,335 cows, and 1,641 cows with both data types within a lactation. We applied multivariate repeatability models separately for primiparous and multiparous cows. Both models included the phenotypes DMI, energy-corrected milk (ECM), BW, and CH₄, as well as fixed effects of age at first calving and lactation curve, and random effects of herd×year×week, additive genetic effect, permanent environmental effect, within-parity repeatability (multiparous cows), and residuals. We defined breeding values for lactation sums and constructed the genetic residual traits: gRFI (DMI|ECM,BW) and gCH₄ (CH₄|ECM,BW). Lactation-wise heritabilities were moderate across traits and parity groups (DMI: 0.27 & 0.33; ECM: 0.49 & 0.33; BW: 0.50 & 0.43; CH₄: 0.36 & 0.38). Genetic correlations were moderate and significantly different from zero between DMI and ECM or BW (0.48-0.69), and between ECM and CH₄ (0.25-0.34) in both parity groups. The posterior probability of additive genetic correlations were higher than 0.05, were in multiparous lactations close to unity and numerically lower in primiparous lactations (0.30 P >0.99 vs. 0.18 P > 0.88), indicating that cows genetically predisposed to consume more feed also emit more CH₄, which aligns with biological expectations. In multiparous cows, favorable genetic correlations were observed between gRFI and CH₄ (0.30) and between gRFI and gCH₄ (0.31). In summary, these findings suggest that joint selection for reduced methane emissions and improved feed efficiency is feasible using large-scale data from the CFIT and sniffer systems, as the traits are favorably correlated. Future joint multi-trait modeling is expected to enhance breeding decisions through increased accuracy of breeding values.
Institution: Aarhus University
Language: English
Subfield: Genetics
Domain: Life Sciences
SDGs: Zero hunger, Climate action
Open Access status: hybrid
OpenAlex ID: W7204741891