Covariance functions and random regression models for cow weight in beef cattle

Area of Science:

• Animal Breeding and Genetics
• Quantitative Genetics
• Statistical Genetics

Background:

• Accurate estimation of genetic parameters for traits like body weight is crucial for effective selection in beef cattle.
• Traditional repeatability models may not fully capture the complex age-related changes in variance and covariance of cow weights.

Purpose of the Study:

• To evaluate cow weights using covariance function-random regression models (CF-RRM) to better understand genetic and environmental influences across different ages.
• To compare the performance of CF-RRM with traditional bivariate and repeatability models for estimating genetic parameters.

Main Methods:

• Analysis of 61,798 cow weights from Angus, Hereford, and F1 crosses using REML and CF-RRM with Legendre polynomials.
• Modeling fixed effects (age, season, year, pregnancy) and random effects (additive genetic, permanent environmental) for regression coefficients.
• Investigating different orders of polynomial fit for random regression coefficients and modeling temporary environmental effects for variance heterogeneity.

Main Results:

• Estimates of all variances increased with age, and results for older ages differed from traditional bivariate models.
• Heritability estimates ranged from 0.38 to 0.78, showing fluctuations, particularly at extreme ages.
• Genetic correlations were generally high, with the lowest estimate (0.70) between the most extreme ages.

Conclusions:

• CF-RRM provide a more flexible and accurate approach than repeatability models for analyzing cow weights across ages.
• While CF-RRM are superior, a repeatability model may serve as a practical approximation for predicting additive genetic effects in breeding programs.
• The findings highlight the importance of age-specific genetic parameter estimation for optimizing selection strategies in beef cattle.