Genetic parameters of measures and population-wide infrared predictions of 92 traits describing the fine composition and technological properties of milk in Italian Simmental cattle

The objectives of this study were to estimate, for the Italian Simmental cattle population, genetic parameters for 92 traits and their infrared predictions (IP) and to investigate the genetic relationship between measured traits (MT) and IP. Data for milk fat fatty acid composition (n = 1,040), detailed protein composition (n = 3,337), lactoferrin (n = 558), pH (n = 3,438), coagulation properties (n = 3,266), curd yield and composition obtained by a micro-cheese making procedure (n = 1,177), and content of Ca, P, Mg, and K (n = 689) were obtained using reference laboratory analysis. Infrared prediction for all the investigated traits was performed using 143,198 spectra records belonging to 17,619 Italian Simmental cows. (Co)variance components for MT and their IP were estimated in a set of bivariate animal model REML analyses and genetic correlations between MT and IP were estimated using all IP obtained at the population level. A significant positive relationship was observed between the coefficient of determination of the infrared prediction models and the phenotypic and genetic variation of the IP. The decrease in the estimated genetic variance of IP compared with MT was on average 64%. For traits exhibiting calibration models with coefficients of determination in cross-validation (R²_CV) greater than 0.9, the decrease in the genetic variance ranged from approximately 20 to 50%. Most traits (88 out of 92) exhibited lower heritability estimates for IP than for the corresponding MT. The estimated genetic correlations between IP and MT (r_a) were in general very high. A positive relationship (r = 0.57) between R²_CV of calibration models and the estimated r_a has been detected. For calibration models exhibiting R²_CV higher than 0.75, r_a were greater than 0.9. The variability in the estimated correlations increased when R²_CV decreased, and for calibration models of moderate predictive ability, estimates of r_a ranged from 0.2 to 1. Genetic parameter estimates suggested that IP can be used as indicator traits in breeding programs for the enhancement of fine composition and technological properties of milk. The genetic gain achievable selecting for IP is expected to be high for fatty acid composition, minerals, and for technological properties of milk, whereas it will be low for casein and whey protein composition and for the content of lactoferrin.