Heritabilities of the mid-infrared spectra of sheep milk throughout the lactation

Session 70, Theatre 6

International audience

Inglés. and could be considered as predictors reflecting the physiological status and performance of animal. The geneticvariability of milk FTIR spectra is well documented in cattle and goat. This study aimed to explore genetic variabilityof FTIR spectra in sheep milk by estimating heritability at each individual spectral point in the mid-infrared region fromFrench Lacaune dairy sheep’s milk obtained in SMARTER European project. The FTIR spectrum of milk providedby spectrometers contains 1,060 points, called wavenumbers (5,012 to 926 cm-1), based on transmittance. Milk wassampled during two years (2020-2021) in 8 commercial farms. After data editing, the number of records available was41,143 FTIR spectra from 5,281 ewes with a mean of 8 records per ewe for the whole testing period. 1,794 ewes weregenotyped with low-density chip then imputed to SNP50 Bead-Chip. Heritability at each wavenumber was estimated,with all records and at each lactation stage, by a single-trait animal model using AI-REML including pedigree andgenomic relationships. Parity and a vector of flock-year-stage of lactation were included as fixed effects and anadditive genetic effect on the animal, permanent environmental effect and residual as random effects. The pattern andvariability of the FTIR spectrum was similar to those estimated in dairy cattle and goat. Heritability estimates rangedfrom 0 in water absorption regions to 0.42 in regions linked to milk composition. A moderate repeatability of FTIRspectra (0 to 0.53) was observed. Two different heritability patterns were observed between the first three and lastlactation stages considered. These results will be analysed regarding the evolution of the farming conditions duringlactation. Additional analyses will allow to characterize groups of ewes with similar patterns and genetic correlationsbetween wavenumbers and milk production traits will be estimated. This information will contribute to propose lessinvasive and cheaper predictors for the characterization of adaptability of dairy ewes. Study financially supported byINRAE and Occitanie region.