Multi-trait Models with Maternal Effects for Genetic Evaluation of Weaning and Ultrasound Carcass Traits in Nelore Calves

Abstract This study aimed to identify the most appropriate statistical model for the genetic evaluation of Nelore cattle and to estimate genetic parameters for adjusted weaning weight (W210) and ultrasound-measured carcass traits—longissimus muscle area (LMA), marbling (MARB), and backfat thickness (BF). The dataset comprised 2,001 calves born between 2015 and 2025 in a commercial herd in Mato Grosso do Sul, Brazil. Significant (P < 0.05) non-genetic sources of variation identified by GLM included sex, recipient breed, birth month and year, calf age (linear and quadratic), and calving. Variance components were estimated using REML under three models: M1 (direct additive), M2 (M1 + maternal additive), and M3 (M2 + maternal permanent environmental effects). Likelihood ratio tests favored M2 and M3, with maternal effects reducing residual variance. Direct heritability estimates were moderate for LMA (0.31–0.34), MARB (0.24–0.25), and W210 (0.17–0.22), and low for BF (0.08–0.11). Maternal heritability ranged from 0.08 to 0.25. Positive direct additive genetic correlations (0.22–0.70) indicated that selection for increased LMA and BF is expected to generate correlated responses in MARB and W210. Overall, maternal additive and maternal permanent environmental effects should be included in genetic evaluations of weaning and carcass traits in Nelore cattle, and ultrasound proves useful for early selection given the favorable genetic correlations.