Phenomics

Genetic correlation of methane intensity with milk Fourier-transform infrared spectra

---

Abstract

Selection indices are a method for estimating the genetic value of dairy cattle. Recently, they have been applied to evaluate the impact of including methane production in the breeding objective of Holstein Friesian dairy cattle. However, obtaining direct measurements of methane emissions from many animals is difficult. Therefore, having indirect measurement using information already available would be beneficial to implement indirect selection indices. The objective of this study was to estimate the genetic correlations between methane intensity (MI; expressed in g of methane/kg of fat and protein corrected milk) and the 1,060 frequencies (wavenumber [WN]) of the Fourier-Transform Infrared (FTIR) milk-spectra, using Single-Step Bayesian Genomic BLUP (ssGBLUP-Bayesian) with bivariate models, to study the precision that a penalized indirect selection indices (e.g., Lasso) could have for methane intensity using the milk spectra. The analyzed data were obtained from 588 Spanish Holstein Friesian cows, with pedigree and genotype information (50K SNP). The available information included data on milk production, milk composition (fat, protein, lactose percentages), spectra, and methane production (g/d) measured by sniffers. The methane intensity was calculated as: MI = (CH4 (g/d) / ([0.337 + 0.116 * milk fat (%) + 0.06 * milk protein (%)] * Daily milk yield (l/d)). The genetic correlations (rg) between methane intensity and the 1,060 WN, were estimated using bivariate mixed-effects models with a ssGBLUP-Bayesian model. For each bivariate model, we used 10,000 iterations, and a burn-in of 1,000 chains. Fixed effects included: parity number (3 levels), lactation stage (3 levels), herd/robot combination (15 levels), and year/month of calving (29 levels), while random effects were the animal's genetic effect and residual effect. The estimated heritabilities (h²) were: methane intensity (0.11 ± 0.04), milk production (0.17 ± 0.06), fat (0.21 ± 0.09), protein (0.12 ± 0.05), and lactose (0.32 ± 10). The mean of the h² for the WN was 0.14, ranging from 0.03 to 0.34. Genetic correlations between MI and WN ranged from -0.58 to 0.45. Finally, the accuracy of the indirect selection of MI using one WN at a time was calculated using the formula ACC(gyi,Ii)= |rgWN,gMI * √h2MI|. Where rgWN,gMI is the genetic correlation between methane intensity and the WN and √h2MI is the heritability of methane intensity. The accuracy values ranged between 0.00019 and 0.20. These results show the potential of milk FTIR-spectrum as an indicator trait for the indirect selection of methane intensity in genetic improvement programs.

Keywords: 2026

How to Cite:

de los Campos, G., Garcà­a-Rodrà­guez, A., González-Recio, O., Jiménez-Montero, J., Montaldo Valdenegro, H., Roman-Muà±oz, J., Ruà­z López, F., Servà­n Tome, A. & Toledo-Alvarado, H., (2026) “Genetic correlation of methane intensity with milk Fourier-transform infrared spectra”, World Congress on Genetics Applied to Livestock Production Digital Archive 2026(1): 2286912. doi: https://doi.org/10.31274/wcgalp.24161

Rights: 1