<![CDATA[Heterosis is often utilized as a success indicator in a crossbreeding program. It is defined as the deviation of the crossbred means relative to their parental breeds. Heterosis mechanism is highly dependent on the genetic factors and thus, we incorporated genetic information in its estimation. The objective of this article was to compare heterosis estimated with conventional and mixed model approaches. In total, phenotypes of 3804 individuals were recorded. Data were obtained from a crossbreeding experiment involving Boer bucks and Jawarandu does. Observed traits were birth weight, weaning weight and average daily gain. Conventional and mixed model methods were used to estimate the heterosis. The heterosis values (%) between B×B vs B×J, estimated with conventional method were -11.38, -10.51 and -10.39; with mixed model were -6.23, -9.27 and -9.68 for BW, WW and ADG respectively. Heterosis values in B×(B×J) relative to B×B, estimated with conventional method were -6.16, -10.35 and -11.69; whereas with mixed model were -8.01, -10.82 and -9.14 for BW, WW and ADG respectively. Conventional method tends to underestimate the means phenotype with lower standard errors compared to mixed model analysis results in all traits. Conventional method also introduces biased heterosis estimates compared to the mixed model. Conventional method ignores any potential effects in the estimation procedures; whereas mixed model approach incorporates all the systematic and random effect including family relationship information. Thus, mixed model produced more reliable results in genetic parameters estimation. We recommend employing mixed model analysis in estimating heterosis.]]>
