An experimental comparison of selection methods

Abstract

Five methods of selection each in five replicates were examined for single-generation response in high and low 21-day pupal weight of Tribolium castaneum. The methods of selection tested were individual (ID), dam family or full (DF), sire family or half-sib (SF), progeny test (PT), and index (IX). A selection intensity, involving as parents the top 25% of the individuals on any specific phenotypic scale, was applied throughout the experiment. Each replicate included a contemporaneous, randomly mated control line. Responses attained with each method were compared with predicted gains and with results observed for other methods. Differences in responses of males and females were also assessed. Response to selection was markedly asymmetrical with greater gains observed in the downward direction. In the lines selected for low pupal weight, males and females were found to respond similarly and actual responses did not differ from those predicted. Ranking of methods by responses obtained for downward selection yielded the order IX > ID > PT > DF > SF. Selection for high pupal weight yielded differences in responses of males and females and the observed responses were significantly less than those predicted in all cases except that of PT-selected females. Observed responses for males were ordered as IX > ID > PT > DF > SF, while those for females were PT > IX > DF > SF > ID. When the data from both sexes were combined, the order of responses was PT > IX > DF > ID > SF.Key words: asymmetrical response, selection, Tribolium, pupal weight.