Disruptive Selection and Assortative Mating in Tribolium castaneum

Publisher Website
Google Scholar
Abstract
Disruptive selection for 19-day pupa weight in T. castaneum was practiced for 15 generations in an attempt to produce assortative mating based on body size. Pupa weight was chosen as a character likely to have some effect on mating behavior. Selected lines (S1-S4) and control lines (C1, C2) were maintained. Lines S1 and S3 were established from the same 32 adults; similarly S2 and S4 were established from a 2nd set of 32 adults, and the controls from a 3rd set. Genetic and environmental variances increased severalfold in 3 of the 4 selected lines, but remained relatively constant in the controls. Frequency distributions of offspring pupa weight became increasingly non-normal in all selected lines, and by the end of the experiment line S3 showed an apparent deficiency of intermediate weight pupae and S4 showed clear bimodality. Progeny distributions for adults allowed mating choice (MC) compared to those under forced random mating (FRM) showed that in lines S3 and S4, the deficiencies occurred where the hybrid types were expected, suggesting that the heterogamic matings occurred less frequently in the MC samples than in the FRM samples. No such deficiencies were evident in lines S1 or S2. After generation 15, mating choice experiments were performed. There was highly significant assortative mating in lines S3 and S4, but not in S1 or S2. Assortative mating may be more readily achieved when disruptive selection is performed on a character related to mating behavior. The results also demonstrate the importance of stochastic effects on the outcome of disruptive selection experiments. In 2 instances, lines begun from the same 32 individuals and maintained identically for 15 generations gave very different responses. Such large differences among replicate lines have not been observed in directional or stabilizing selection experiments.