Milk productivity of Holstein cows depending on the genomic evaluation of breeding bulls

Abstract

The study was conducted to determine the accuracy of the implementation of foreign genomic evaluation indicators for bulls in the actual milk productivity of daughters in the conditions of the Northern Trans-Urals to determine the possibility of using such criteria in further selection. The object of the study was 769 Holstein cows (age difference ≤12 months) – daughters of 11 sires (an average of 70 daughters per bull), bred in the Tyumen region. The relationship between the complex breeding value index (TPI), the index of lifetime profit (NM$), the predicted transmitting ability (PTA) for milk yield (PTAM), the mass fraction of fat (%PTAF), the mass fraction of protein (%PTAP), milk fat (PTAF), and milk protein (PTAP) with the milk productivity of the bulls' daughters was examined by conducting a correlation analysis and quartile distribution by NM$ and PTAM of the fathers. A comparative analysis of the breeding value of bulls was performed using genomic assessment data from 2020 and 2025, as well as progeny quality assessments in the Russian Federation and on the farm. The NM$ and PTAM genomic indices in the Northern Trans-Urals region weakly correlate with milk yield and milk component yield (r=0.13-0.23; R2<0.060). Grouping the bulls' daughters by the NM$ index revealed a nonlinear dependence of milk yield on quartile (the superiority of Q3 and Q4 over Q1 was 640...642 kg (P≥0.999), while the differences between Q3 and Q4 were insignificant). The relationship between PTAM and milk productivity turned out to be more predictable: the milk yield of the daughters of Q2-Q4 bulls was significantly higher by 270-777 kg (P≥0.95-0.999) compared to Q1. Genomic assessments of bulls across years are characterised by a close relationship (r = 0.91-0.98; R2 = 0.828-0.960). However, their correlations with actual offspring performance under farm conditions were weaker (r = 0.13-0.57; R2 = 0.017-0.325). This indicates partial implementation of foreign genomic predictions and justifies the need to develop domestic assessment models calibrated using local population data, taking into account regional feeding and management characteristics.