Random regression analysis of calving interval of Japanese black cows

Abstract

We estimated genetic parameters for the calving interval of Japanese Black cows using a random regression model and a repeatability model. We analyzed 92,019 calving interval records of 36,178 cows. Pedigree data covered 390,263 individuals. Age of cow at previous calving for each record ranged from 18 to 120 months. We used up to the second-order Legendre polynomials based on age at previous calving as sub-models for random regression analysis, and assumed a constant error variance across ages. Estimated heritability was 0.12 to 0.20 with the random regression model and 0.17 with the repeatability model. With the random regression model, the estimated genetic correlation between ages was ≥0.87, and those between 24 and 36 months, 24 and 84 months, and 36 and 84 months were 0.99, 0.95, and 0.97, respectively. Spearman’s rank correlation between breeding values of 36,178 cows with their own records estimated by the random regression model with those estimated using the repeatability model was ≥0.97, and the rank correlation was ≥0.94 for 314 sires of these cows. These results support the validity of fitting a repeatability model to the records of the calving interval of Japanese Black cows for evaluation of breeding values.