Studies of random regression test day models and persistency for Iranian Holstein production traits

Thumbnail Image



Moradi Shahrbabak, Mohammad

Journal Title

Journal ISSN

Volume Title


University of Guelph


A single trait random regression animal model was applied to test day (TD) records for milk and fat yields for estimating genetic parameters and genetic evaluation. The estimated heritabilities from random regression TD models (RR-TD) were higher than those estimated from 305-d models for milk and fat yields. Genetic variance for fat yield decreased from the beginning of lactation through to the middle of lactation and then rose slightly to the end of lactation. The genetic correlation among different DIM and 305-d measurements were higher in the middle of lactation than either end of lactation for milk and fat yields. The genetic correlations for fat yield were higher in the early part of lactation than those for milk yield. Genetic evaluation using the RR-TD model based on all (TDFul) or every other TD measurements (TD11) listed the same animals as the top 10 and 20% sires as the evaluation based on 305-d model, with some reranking of sires. Correlation between EBV of sires with 20 or more daughters from TD11 and TDFul data sets were 0.991 and 0.917 for milk and fat yields, respectively. Bimonthly milk and fat testing might be an alternative testing scheme to reduce the costs of breeding programs, in particular in developing countries. Part lactation TD measurements, every three or every four, and mid-lactation period TD measurements were not as informative as all or every other TD measurements for genetic evaluation of animals. Clear differences among cows lactation curve shape and persistency measurements were found, and RR-TD could prepare the ground for accounting for those differences. Daily changes in estimated breeding values could be used as a persistency criterion. Calculated changes in milk production in the declining part of lactation based on two periods of lactation curve could represent persistency better than calculated changes based on two points of the lactation curve. Relation of persistency criteria and second part of lactation could be used as an evaluating criterion for persistency criteria. Herd effects were found to affect lactation curve shape and persistency measurements. Adjusting for herd effects changed the intercept and slopes of the animals (cows and sires) lactation curves. Lactation curve shape of cows in the same herd were also affected by adjusting for herd effects. Herd effects did not affect correlation between persistency measurements and 305-d or parts of lactation significantly.



Iranian Holstein, production traits, random regression, test day models, milk yield, fat yield