Test day-milk yields variance component estimation using repeatability or random regression models in the Rendena breed

This study has aimed to compare Repeatability (RP-TDm) and Random-Regression Test Day models (RR-TDm) in genetic evaluations of milk (M), fat (F) and protein (P) yields in Rendena breed. Variance estimations for Milk (M), Fat (F) and Protein (P) were obtained on a sample of 43,842 TD belonging to 2,692 animals controlled over 15 years (1990-2005). RP-TDm estimates of h2 were of 0.21 for M and 0.17 for both F and P, whereas RR-TDM provided a trend of h2 ranging from 0.15-0.34 for M, 0.15-0.31 for F and 0.10-0.24 for P. Both RP-TDm and RR-TDm results agreed with literature, even though RR-TDm provided a pattern of h2 along the lactation different from other studies, with the lowest h2 at the beginning and at the end of lactation. PSB, MAD and -2Log L parameters revealed lower power of RP-TDm as compare with the RR-TDm.