In this research, a total of 634,949 monthly test day milk records collected from 72,523 Iranian Holstein first lactation cows calving between 1999 and 2008 were utilized to estimate heritability and repeatability using restricted maximum likelihood (REML) statistical approach. The records were obtained from 724 herds over 19provinces in the country. All monthly test day records were obtained from 3-times a day milking cows. The average monthly test day milk yield in the whole data set was 29.11 kg (SD = 7.07 kg). Arepeatability test day anin1almodel was used. In the model, fixed contemporary group (combining province-herd-year of recording-season of recording), stage of lactation (10 stages), and covariables ofHolstein gene percentage, days in milk nested in the stage of the lactation, and age of cow at recording (in month) were included. The random part of the model consisted of additive genetic, permanent environment and residual effects. The total number of animals (including daughters, sires and dams) in the pedigree was 101,601. DMU software was run in a Pentium 4 computer with 4 gigabyte memory to fit the anin1al model. Additive genetic, permanent environment and residual variance estimates were 4.352, 13.494 and 15.335 kg2 respectively. The results revealed that heritability and repeatability of monthly test day milk yield were 0.131 and 0.537 respectively. Low heritability of monthly test day milk yield obtained in this study indicates major part of the phenotypic variation for the trait under consideration is resulted from the temporary environment and model unexplained variation revealing that genetic selection based upon test day milk records may not be as cost effective as the traditional lactation model for which there is usually a higher heritability.

Keywords