Genetic parameters and genome-wide association studies of milk infrared spectra from Lacaune dairy ewes

Fourier-transform infrared spectroscopy is a widely used analytical tool in dairy production-either in cattle or small ruminant-for predicting milk composition and quality. Nevertheless, not all the information contained in the spectrum is used for its predictions. This study examined the genetic parameters and GWAS of all wavelengths of infrared milk spectra in Lacaune dairy ewes to explore their genetic determinism and compare it with that of milk composition traits predicted by the spectra. Data from 1,063 ewes were collected during their first lactation over 9 yr (2014-2022) at 5 milk recording controls, including infrared spectra with 1,060 wavelengths and predicted milk composition (fatty acids and proteins). The ewes were genotyped with a 54K SNP chip. Heritability estimates across the infrared spectrum ranged from 0.26 to 0.69, with higher values observed in the late lactation stages. Genome-wide association study revealed significant QTL for both spectra wavelengths and predicted milk traits, with some loci being specific to milk IR spectra. Key regions on Ovis aries (OAR) chromosome 3 (OAR3, around Socs2) and OAR11 were associated with both milk composition and spectral traits. Additionally, novel QTL for milk urea (OAR6) and certain fatty acids (OAR17) were identified, as well as QTL within spectra on 4 chromosomes (OAR1, OAR10, OAR13, and OAR20). This study demonstrates that direct GWAS on infrared spectra provides complementary insights to traditional approaches on predicted milk composition traits, uncovering new genetic regions of interest and advancing our understanding of milk composition's genetic architecture. These findings highlight the potential of milk spectra as a tool for enhancing genetic selection in dairy sheep.