Effect of residual variance heterocedasticity on milk yield and age at first calving genetic parameter estimates

• Main Author: Araújo Neto, F. R. [UNESP]
• Publication Date: 2010
• Other Authors: Aspilcueta-Borquis, R. [UNESP], Baldi, F. [UNESP], Tonhati, H. [UNESP]
• Format: Article
• Language: eng
• Source: Repositório Institucional da UNESP
• Download full: http://hdl.handle.net/11449/227817
• topic: Bayesian inference; Gibbs sampling; Heterocedasticity; Milking buffaloes

Genetic evaluation programs assumed homocedasticity of residual variance estimates for different milk yield levels. The objective of this study was to estimate genetic parameters for 305-d milk yield (MY305) and age at first calving (AFC) considering residual homogeneity or heterogeneity variances for different milk yield classes (according with milk yield level). The covariance components were estimated using mult i-trait analysis by Bayesian inference method, applying the GIBBS2F90 and GIBBS3F90 software (MISZTAL, 2007). The model included the fixed effects of contemporary groups (herd -year and calving season), number of milking (two levels) and age of cow at calving as covariable (quadratic and linear effect). The additive genetic and residual effects were considered as random effects in the model. The heritability estimates for MY305 and AFC were 0.25 and 0.09 when residual variance was homogeneity. Considering residual variance heterocedasticity, the heritability estimates were 0.17 and 0.25 for MY305 and, 0.13 and 0.10 for AFC, at low and high milk yield level, respectively. Thus, the results of this study shown that is need consider residual variance heterocedasticity for the genetic evaluation of MY305 and AFC traits in milking buffaloes.