V. V. Gart, S. G. Kulikova, E. V. Kamaldinov, A. F. Petrov, K. N. Narozhnykh
Novosibirsk State Agrarian University, Novosibirsk, Russia
E-mail: This email address is being protected from spambots. You need JavaScript enabled to view it.
Volume 25 No. 9
Date of paper submission: 10.07.2025, date of review: 28.07.2025, date of acceptance: 12.08.2025.
Published: 09/30/2025
Abstract. The purpose of this study was to validate a method for the early assessment of milk productivity in Holstein first-calf heifers. This method is based on a composite score calculated by sequentially summing the adjusted scores of linear type traits, ranked in descending order of their individual correlation with a total productivity indicator. Methods. The model quality was assessed using cross-validation on primary data, which included the composite score, 305-day milk yield, and fat yield. The predictive ability of the composite score was evaluated at two levels: for individual animals (by calculating correlation) and for groups of animals with the same exterior composite score (by applying cross-validation to polynomial models based on group mean values). This comprehensive approach allowed for a thorough assessment of the potential of using the exterior composite score for the indirect evaluation of milk productivity in practical breeding work. Results. It was established that polynomial models developed for the mean productivity values of animal groups, aggregated by their composite score, demonstrate high predictive ability. Specifically, the cross-validated R² for milk yield was 0.7566, with statistical errors RMSE of 257.87 kg and MAE of 177.84 kg. For fat yield, the cross-validated R2 reached 0.9688, with RMSE and MAE of 3.61 kg and 2.78 kg, respectively. Post-hoc power analysis revealed that the statistical power of the third-degree polynomial models for mean productivity values was very high (100 % for both milk and fat yield) and statistically significant (p < 0.001). The method of indirect assessment and prediction of the average milk productivity level for groups of first-calf heifers, unified by their composite score, can be practically applied to evaluate the potential of replacement heifer batches and to make general management decisions at the herd level. The scientific novelty of this research lies in the validation of a new method for the early prediction of productivity in Holstein first-calf heifers based on a complex of exterior indicators.
Keywords: Holstein breed, first-calf heifers, productivity prediction, milk yield, milk fat yield, Western Siberia, conformation, linear type traits, composite score, cross-validation, regression model
Acknowledgements. The study was carried out on the topic of the State Assignment No. FESF-2023-0016.
For citation: Gart V. V., Kulikova S. G., Kamaldinov E. V., Petrov A. F., Narozhnykh K. N. Validation of the first method for predicting milk productivity in first-calf heifers based on a composite score of linear type traits. Agrarian Bulletin of the Urals. 2025; 25 (09): 1372‒1383. https://doi.org/10.32417/1997-4868-2025-25-09-1372-1383 (In Russ.)
Download the full text of the article
