# Estimation of Variance Components for Growth Traits in Composite Beef Cattle Accounting for Heterosis and Recombination

## TL;DR

This study estimates genetic variance components in a composite beef cattle population in Brazil, showing that genomic models improve accuracy and support long-term breeding goals.

## Contribution

The study introduces genome-based models that account for breed composition, heterosis, and recombination in a tropical beef cattle population.

## Key findings

- Genomic models provided better fit and lower residual variances than pedigree-based models.
- Heterosis effects improved model fit and shifted variance from additive to non-additive components.
- Recombination effects had minimal impact on variance estimates but influenced selection rankings.

## Abstract

Background/Objectives: Accurate estimates of variance components are essential in breeding programs. In this context, the main objective of this study was to estimate variance components for growth traits in the Montana Composite® beef population, which was developed in Brazil by crossing various taurine and indicine breeds. After 30 years of selection, the impact of recombination, heterosis, and inbreeding may have influenced the genetic background of the population. Methods: We analyzed data of birth weight, weaning weight, post-weaning weight gain, and yearling weight using 124,255 phenotypic records, 193,129 pedigree records, and 3911 genotyped individuals. Ten single-trait animal models (M1–M10) were compared, differing in the relationship matrix (pedigree- or genome-based relationships) and the inclusion of direct/maternal breed composition, heterosis, and recombination effects. Results: Models incorporating genomic information consistently yielded better fit and lower residual variances than pedigree-based models, highlighting the advantage of genomic information in capturing Mendelian sampling and realized genetic relationships. The inclusion of heterosis effects improved model fit and led to a partial reallocation of genetic variance from additive to non-additive components. In contrast, the inclusion of recombination effects in the models minimally influenced variance component estimates. Nevertheless, more complex models affected animal rankings and altered the breed composition of top-ranked selection candidates, with selection overlap between pedigree- and genomic-based evaluations ranging from moderate to high. Conclusions: Overall, genome-based models accounting for breed composition, heterosis, and recombination provided the most robust variance component estimates and the best support for long-term selection goals in the studied tropical composite beef cattle population.