An alternative interpretation of residual feed intake by phenotypic recursive relationships in dairy cattle

Summary: Genetic evaluation on residual feed intake (RFI) often takes 2 stages. Combining these 2 modeling stages leads to one-step linear regression, eliminating the need to estimate the residuals as the RFI phenotypes specifically. However, fitting phenotypes as regressor variables in a standard linear regression is criticized because phenotypes are subject to measurement errors. Multiple-trait models have been proposed, which give the genetic values of RFI through a follow-up partial regression procedure. By rearranging the linear regression equation, we came across an alternative, causal RFI interpretation by phenotype recursiveness between DMI and energy sinks. In this technical note, we propose a Bayesian recursive structural equation model (RSEM) for directly evaluating RFI, extending its analytical capacity to multiple-trait analysis.