Gene-content multiple‑trait REML improves the reliability of genetic evaluation for milk production traits in Spanish Palmera goats

Genetic evaluations of milk production traits in dairy goats typically do not explicitly account for variability in major genes such as the polymorphisms at CSN1S1 (αS1-casein), despite its substantial impact on milk protein synthesis. This study assessed the effect of incorporating CSN1S1 genotypes on the reliability of genetic evaluations in the Palmera goat breed. The dataset comprised 37,875 test-day records from 3,168 goats, 4,342 pedigree animals, and the CSN1S1 locus genotypes for 156 bucks, being sires of 1,069 phenotyped animals. Traits analyzed included milk yield (MY), and protein (PC) and fat contents (FC). A gene-content multi-trait REML model (GCMT-REML), treating gene content at the CSN1S1 locus as a correlated additional trait, was implemented and compared against a baseline univariate animal model. Heritability estimates were similar between models, ranging from 0.19 (FC) to 0.31 (PC). The incorporation of gene content information increased reliability across traits, with overall gains from 0.07% (PC) to 1.82% (MY). Improvements were markedly higher in males than females, and in genotyped (5.52% to 9.54%) compared to non-genotyped animals (≈0%). Individuals with low baseline reliability benefited the most, showing gains from 7.2% to 14.27%. These results demonstrate that including CSN1S1 genotypes in routine genetic evaluations can enhance reliability in the Palmera breed. Future work should focus on expanding the genotyped population, including females, so that dominance effects in the locus could also be accounted for.