Estimation of genotype-by-milking system interaction for daily milk yield in Holstein cattle

Milk production recorded by different milking systems are commonly combined for genetic evaluations. However, there could be genotype-by-milking system interactions for daily milk yield that could affect animal ranking and the effectiveness of dairy cattle breeding programs. Therefore, we estimated (co)variance components and genetic parameters for daily milk yield (DMY) measured by milking robots or conventional milking parlors, considered as different traits, based on a 2-trait repeatability model. The systematic effects fitted in the model included calving year and season, days in milk as a linear covariate, and calving age as linear and quadratic covariates. The random effects included additive genetic, permanent environment, and residual effects. Phenotypic records of 11,788 genotyped cows with 60,277 SNP (after quality control) were used for the analyses. The traits have substantial additive genetic variance and low to moderate heritability estimates (from 0.10 ± 0.01 to 0.14 ± 0.01 for DMY measured by milking parlors, and from 0.20 ± 0.02 to 0.21 ± 0.02 for DMY from milking robots). The genetic correlation estimates for DMY based on milking system (milking robots x milking parlors) were 0.88 ± 0.03 in parity 1 and 0.92 ± 0.05 in parity 2. No evidence of strong genotype-by-milking system interaction was found between DMY measured by milking parlors or milking robots, indicating no differential responses to selection for DMY across milking systems.