The objective of this paper was to compare the estimates of genetic parameters obtained by REML and method <FONT FACE=Symbol>Â</FONT>. Animal model single-trait analyses were performed using a data set of Nellore cattle. Statistical models included fixed effects of contemporary group (herd, year, sex and management groups at weaning and 550 days) and age of animal (covariable-linear effect), and additive genetic and residual random effects. Heritability estimates were 0.36 and 0.39 for weight at 550 days (W550); 0.27 and 0.29 for weight gain from weaning to 550 days (WG345); 0.22 and 0.21 for conformation (CONF); 0.21 and 0.21 for precocity (PREC) and 0.22 and 0.22 for muscling (MUSC), respectively, by REML and method <FONT FACE=Symbol>Â</FONT>. For all heritability estimates, standard errors were lower than or equal to 0.02. The differences in the estimates of genetic trend obtained from the two methods were low for all traits. Spearman rank correlation between breeding values obtained in the analyses using the variance components obtained by REML and method <FONT FACE=Symbol>Â</FONT> were equal to 1 for all traits. The results suggest that REML and method <FONT FACE=Symbol>Â</FONT> provide similar results and that method <FONT FACE=Symbol>Â</FONT> can be used as an alternative to REML in single-trait analyses, mainly due to lower computational requirements.
genetic parameters; growth traits; Method <FONT FACE=Symbol>Â</FONT>; Nellore; REML; visual score
