Genetics of milk fatty acid groups predicted during routine data recording in Holstein dairy cattle

The aim of this paper was to estimate genetic parameters for groups of milk fatty acids (FA), namely saturated (SFA), unsaturated (UFA), monounsaturated (MUFA) and polyunsaturated (PUFA), in Holstein cows. Mid-infrared spectroscopy (MIRS) was used to predict FA groups (g/100 g of milk) of 72,848 samples recorded on 17,873 cows between September 2011 and November 2012. Univariate and multivariate models were implemented in a Bayesian framework to estimate (co)variance components for SFA, UFA, MUFA, PUFA, daily milk yield, milk fat and milk protein. Statistical models included fixed effect of parity by stage of lactation, and random effects of herd-test-date, cow permanent environmental, animal additive genetic and residual. Posterior means of heritability estimates for SFA, UFA, MUFA and PUFA were 0.246, 0.069, 0.082 and 0.078, respectively. Estimates of genetic correlations between FA groups ranged from 0.405 (SFA and PUFA) to 0.952 (MUFA and UFA). The increase of fat content led to an increase of all groups of FA, in particular SFA, with undesirable effects on the healthy quality of the product. The study highlighted the existence of exploitable additive genetic variation for groups of FA routinely predicted by MIRS and thus there is potential to address the selection to healthy milk FA composition. (C) 2014 Elsevier B.V. All rights reserved.