Genetic selection for traits that have direct impact on profitability, such as productive longevity (PL), which blends cow longevity with regular reproductive performance, is fundamental for the economic success of beef cow-calf operations. The purpose of this study was to develop data screening strategy and a statistical model to predict genetic merit for PL in a multibreed beef cattle population. Pedigree (n = 1,352,765) and phenotype (n = 978,382) information were provided by Leachman Cattle of Colorado and genotypes (n = 26,342) were provided by the Zoetis commercial genotyping laboratory. A repeatability model (REP) including the systematic effects of age at first calving, year-season of progeny birth, pedigree-based retained heterosis, and parity number, as well as the random effects of additive genetic, permanent environment, contemporary group, and residual were fitted to adjust PL. In addition, a random regression model (RRM) was fitted to investigate PL considering the same effects, with the difference that random effects were regressed on parity. Estimated breeding value (EBV) were obtained by single-step GBLUP (ssGBLUP) and transformed to predict differences in number of calves through linear regression. Predictive performance was assessed in a group of 7,268 cows born in 2010. Heritability estimates for PL were relatively low, with values of 0.109 for REP and a decreasing trend for RRM with values ranging from 0.16 to 0.04. Repeatability for PL was of moderate magnitude, with values of 0.415 for REP and from 0.29 to 0.57 for RRM. Heritability estimates suggest that most of phenotypic variation was accounted for by environmental factors, but long-term genetic selection could still be effective. REP was more efficient than RRM, showing lower number of iterations and time to reach convergence with comparable solutions to RRM. Validation results showed that correlations between EBV and phenotypes (observed/pre-corrected) increased over the years ranging from 0.04 to 0.92. Repeatability values and the validation approach suggested that using a cow’s first record (second parity success or failure) is a reasonably good indicator of posterior performance for PL. Therefore, the inclusion of PL in a multibreed genetic evaluation program, incorporation into selection indexes with existing economic traits, can enable more profitable selection and breeding decisions in beef cattle herds.

中文翻译：

---

多品种肉牛种群生产寿命的遗传评价


对盈利能力有直接影响的性状进行遗传选择，例如将奶牛寿命与正常繁殖性能相结合的生产寿命 （PL），是肉牛-犊牛经营经济成功的基础。本研究的目的是开发数据筛选策略和统计模型，以预测多品种肉牛种群中 PL 的遗传价值。系谱 （n = 1,352,765） 和表型 （n = 978,382） 信息由科罗拉多州的 Leachman Cattle 提供，基因型 （n = 26,342） 由 Zoetis 商业基因分型实验室提供。拟合重复性模型 （REP） 来调整 PL，包括首次产犊年龄、后代出生年份、基于系谱的保留优势和胎次数的系统效应，以及加性遗传、永久环境、当代群体和残差的随机效应。此外，采用随机回归模型 （RRM） 来研究考虑相同效应的 PL，不同之处在于随机效应在奇偶性上回归。通过单步 GBLUP （ssGBLUP） 获得估计育种值 （EBV），并通过线性回归进行转换以预测犊牛数量的差异。对 2010 年出生的 7,268 头奶牛的预测性能进行了评估。PL 的遗传力估计相对较低，REP 的值为 0.109，RRM 的值为 0.16 至 0.04，呈下降趋势。PL 的重复性为中等幅度，REP 的值为 0.415，RRM 的值为 0.29 至 0.57。遗传力估计表明，大多数表型变异是由环境因素引起的，但长期遗传选择仍然可能有效。 REP 比 RRM 更高效，与与 RRM 相当的解决方案相比，迭代次数和达到收敛的时间更少。验证结果表明，EBV 与表型 （观察/预先校正） 之间的相关性多年来增加，范围从 0.04 到 0.92。重复性值和验证方法表明，使用奶牛的第一条记录（第二胎次成功或失败）是 PL 后验性能的合理良好指标。因此，将 PL 纳入多品种遗传评估计划，纳入具有现有经济性状的选择指标，可以在肉牛群中做出更有利可图的选择和育种决策。