Evaluation of selection indices under various parameter combinations in simulated genetic populations.

In a simulation study, the efficiency of some selection indices was tested under various parameter combinations. For the purpose of such comparison a criterion, named as 'evaluation coefficient' was developed. In the evaluation coefficient both the total genetic gain and the relative gains obtained in the individual traits, as a measure of harmony, were taken into consideration.The results have indicated that the genotypically constructed selection indices were more efficient than the phenotypically constructed selection index of Elston (1963). Furthermore, the effectiveness of selection indices, irrespective of the model used, was found to be decreased in case of tight linkage as compared to that of loose linkage. It was, however, observed that the genotypically constructed indices were more sensitive to linkage and the reduction in their performance was relatively greater if characters included in the index had varying heritability coefficients. The efficiency of the indices was also reduced when the characters incorporated in the index possessed different economic weights. The results, however, suggested that relatively faster rate of improvement in the character having low heritability could be obtained by assigning higher economic weights to that character. This would mean that, manipulation of economic weights may help in bringing about the harmonical improvement in all the characters to be selected simultaneously.