Comparative Analysis of Recursive and Alternative Modelling Approaches Considering Body Condition Score for Genetic...

Mature cow size, often defined by mature cow weight (MWT), height (MHT) and body condition score (BCS), is crucial to cow-calf profitability, maintenance efficiency and reproductive performance. Although MWT and MHT are often included in national cattle evaluations by many breed organisations, BCS adjustments are applied less consistently. This study investigated the impact of different modelling approaches on the estimation of breeding values for MWT, with a focus on how BCS is accounted for across models. The dataset provided by American Angus Association comprised 382,156 MWT and BCS records from 209,491 cows. Four modelling approaches were evaluated: Model 1 does not consider BCS; Model 2 treated BCS as a categorical fixed effect; Model 3 used pre-adjusted records standardised for BCS and age; and Model 4 used a recursive model to assess MWT as a genetically independent trait from BCS. Spearman correlations between breeding values predicted across models ranged from 0.79 (between Models 1 and 4) to 0.95 (between Models 1 and 2), indicating that 5%-21% of bulls could have different rankings depending on the model used. Concordance in sire selection was assessed between the top 10% of sires in each model, and model-pair comparison revealed differences ranging from 19% (between Models 1 and 2) to 40% (between Models 3 and 4). These differences highlight the potential for model choice to influence the selection outcomes. Model selection can significantly affect the sire rankings, highlighting the importance of carefully selecting the model that best aligns with the selection objectives and the underlying biology of the traits being evaluated. Although Model 4 offers theoretical advantages, this study does not allow for a definitive conclusion on its overall effectiveness, as no simulations were performed. Additional research is needed to confirm its advantages.