Sustainability & efficiency

Genetic insights from sniffer methane data: contrasting approaches to background definitions

---

Abstract

Reducing global greenhouse gas emissions can be achieved by selecting dairy cattle with lower methane emissions. To effectively identify and select low methane-emitting cows, methane measurements must be recorded across thousands of animals. In dairy cattle, methane concentration is commonly measured using breath-analyzers, often referred to as "sniffers", installed in automatic milking systems (AMS). Currently, background corrections for environmental methane concentrations have been proposed but not thoroughly evaluated. This study genetically contrasts five background correction methods by estimating heritabilities and genetic correlations of methane concentration phenotypes after background subtraction and the uncorrected phenotype. The background definitions fall into two categories based on their calculation period: (1) during milking, when the animal is present, and (2) during the idle period before the milking, when no animal is in the AMS. The five approaches were: 1) Milking period-based: a) Average of the three lowest values; b) Average of the five lowest values; and c) 0.001th quantile. 2) Idle period-based: a) First quantile of each plateau; and b) Moving average and variance, adjusted for date and event duration using least-square means. Final phenotypes were derived by subtracting each background estimate from the AMS visit average methane concentration. The dataset comprised 123,201 weekly averages of visit methane concentration records from 7,760 Dutch Holstein cows across 64 farms in the Netherlands, collected between 2019 and 2023. Average methane concentrations for the corrected phenotypes ranged from 395 ppm (5 lowest values) to 499 ppm (0.001th quantile) and 514.7 ppm for the uncorrected methane concentration. Genetic parameters were estimated using a bivariate repeatability model with fixed effects including herd-year-season interaction (n=346), Holstein fraction (6/8, 7/8, 1), and parity (1, 2, 3+) nested within age of cow at calving. Random effects included additive genetic, permanent environmental, and residual components. Pedigree included 147,548 individuals. Heritability estimates were consistent across all corrected phenotypes (0.14-0.16, SE=0.02). Genetic correlations among the corrected phenotypes were highly positive, as were correlations with the uncorrected phenotype (0.98-1.00 SE≤0.01). Phenotypic correlations ranged from 0.91 to 0.99, indicating strong agreement across background correction methods. These results suggest that average methane concentration is minimally influenced by the choice of background correction method, and omitting background correction yields similar heritability estimates. However, when converting methane concentration to grams per day (g/d), background correction may influence the quantification of emissions. Future work will extend background correction to additional traits like the average of maximum eructation peaks and the area under the methane concentration curve.

Keywords: 2026

How to Cite:

de Haas, Y., Gredler-Grandl, B., Manzanilla-Pech, C., Sharif-Islam, M., van Breukelen, A. & Veerkamp, R., (2026) “Genetic insights from sniffer methane data: contrasting approaches to background definitions”, World Congress on Genetics Applied to Livestock Production Digital Archive 2026(1): 2286327. doi: https://doi.org/10.31274/wcgalp.23929

Rights: 1