Beef cattle

Genomic selection models for functional longevity in Brazilian Angus cattle

---

Abstract

Cow longevity is an economically important trait that directly impacts the profitability and sustainability of beef cattle production systems. Despite its importance, genetic selection for longevity is challenging because the trait is expressed at the end of an animal's life. This study aimed to (1) estimate genetic parameters for functional longevity (FL) in Brazilian Angus cattle, and (2) evaluate the performance of single-step genomic BLUP (ssGBLUP) using two different approaches compared to traditional BLUP for genomic predictions of FL. The FL trait was evaluated at 2 to 10 years of age, using the number of calves as the phenotype. After the phenotypic quality control, data from 48,976 cows were used to create the longevity trait. The pedigree comprised 180,330 animals, of which 19,357 were genotyped using 12 different SNP chips with densities ranging from 50k to 150k. After imputation and quality control, 71,221 SNPs remained for analysis. Single-trait random regression models (RRMs) were used to estimate variance components and to calculate traditional estimated breeding values (EBVs) and genomic EBVs (GEBVs). The GEBVs were estimated using iteration-on-data with the inversion of the full genomic matrix (GEBV_FULL) and the proven and young (GEBV_APY) algorithm. The model included type of birth (from embryo transfer or not) and systematic regressions for year-season of birth as fixed effects. Random effects included contemporary group, additive genetic, permanent environmental, and residual effects. The RRMs were fitted using orthogonal Legendre polynomials of orders 2 to 4. According to the Deviance Information Criterion, the model with four regression coefficients for each effect was the most appropriate. Genomic predictions were validated for young cows (i.e., cows born after 2018 with phenotypes recorded in the last 4 years) without phenotypes using the linear regression (LR) method. Heritability estimates were generally low, ranging from 0.02 to 0.08 across ages, with peak values observed between 5 and 9 years. Genetic correlations ranged from 0.07 to 0.99 among ages. Low genetic correlations between ages 2-3 and later ages (ranging from 0.07 to 0.39) indicated that FL at early and late life stages should be considered as genetically distinct traits. The genetic breeding values were predicted at six years of age because the correlation after this age was above 0.86, and the heritability was optimal. For young cows, the LR statistics for both genomic methods (FULL and APY) showed similar performance, with prediction accuracies of 0.41, regression coefficients of 0.89, and biases of -0.01. The predicted accuracies from ssGBLUP outperformed traditional BLUP by 17% for young cows. Despite the low heritability of FL, genomic selection can help to increase prediction accuracy, thereby enhancing the potential for genetic gain in this Angus population.

Keywords: 2026

How to Cite:

Campos, G., Junqueira, V., Oliveira, H., Lourenco, D. & Cardoso, F., (2026) “Genomic selection models for functional longevity in Brazilian Angus cattle”, World Congress on Genetics Applied to Livestock Production Digital Archive 2026(1): 2286097. doi: https://doi.org/10.31274/wcgalp.23853

Rights: 1