[en] Common human health concerns and imminent needs for more sustainable nutrition patterns require from dairy industry and farmers a. o. a closer look at milk fatty acid (FA) profile. Therefore up to date calibration equations using mid-infrared (MIR) spectrometry were developed permitting the estimation of FA contents in bovine milk. The aim of this study was to estimate the variability of the major FA from data collected during the Luxembourg routine milk recording. A total of 148,296 milk samples with MIR-spectra were collected from October 2007 to January 2009 on 36,522 cows belonging to 5 breeds in 718 herds and scanned by Foss MilkoScan FT6000. The contents of saturated FA, monounsaturated FA, omega-9, short chain FA, medium chain FA, and long chain FA were obtained using Belgian MIR calibration equations. Analyzes were done by a multi-trait multi-lactation animal mixed models. Fixed effects were herd*test date, lactation stage lactation number, age*lactation number, and breed effect. Random effects were herd*year of calving, permanent environment within and across lactation, animal effect, and residual effect. Breed differences as well as lactation effects were observed. Our results showed moderate heritability values suggesting the existence of a FA genetic variability. The variability of the first Luxembourg breeding values was large enough to develop selection tools for improving the nutritional quality of bovine milk fat.
