Comparing alternative random regression models to analyse first lactation daily milk yield data in Holstein–Friesian cattle

摘要: Abstract Test-day (TD) milk yields from Spanish Holstein cows were analysed in three independent data sets (35 615, 35 209 and 27 272 TD records, respectively) with a set of random regression models. Wilmink Ali-Schaeffer lactation functions and, Legendre polynomials (RRL) varying order (up to six coefficients) on additive genetic (AG) permanent environmental (PE) effects used. The analysis the eigenvalues eigenvectors AG PE (co)variance matrices revealed possibility reducing dimension RRL submodels, particularly for effects. Lactational submodels provided largest daily variance estimates at onset lactation, as well low or even negative correlations between peripheral TD. Polynomials higher (four above) showed oscillatory patterns larger variances lower predicted extremes lactation. Model performance was assessed using broad range criteria. results strong consistency among terms models ranking. worse than same number parameters. For models, all criteria except Bayesian information criterion, favoured most complex model. This criterion selected model 2–3 coefficients 5–6