Search

We selected on knockdown temperature, the upper temperature at which insects lose the ability to cling to an inclined surface, in replicate populations of Drosophila melanogaster for 32 generations (46 generations of rearing). Knockdown temperature (Tkd) was initially bimodally distributed in both control and selected lines, and a similar pattern was found in several populations surveyed from two other continents. Within 20 generations of selection, the Up-selected lines (top 25% each generation) had lost the lower mode and the Low-selected lines (selected to fall out at approximately 37 degrees C) had largely lost the upper mode. The realized heritability of Tkd computed over the first 10 selection episodes was approximately 0.12 in the Up-selected and approximately 0.19 in the Low-selected lines. Realized heritability rose dramatically in the Low-selected lines over the first 20 generations of selection. The two modes, plus this rise in heritability, suggest that knockdown temperature is the product of one or two genes of large effect. The global polymorphism for knockdown temperature, coupled with the ease of selective removal of either mode, suggests that genetic variation for knockdown temperature may be maintained by natural selection.

We used a repeated-measures, four-factor experimental design to determine how the fecundity of Drosophila melanogaster during the first 5 days of adult life was influenced by paternal, maternal, developmental and laying temperature, with two different temperature levels (18 degrees C vs. 25 degrees C) per factor. Laying temperature had by far the largest effect on fecundity and accounted for 79 per cent of the variance in overall fecundity: flies laying at 25 degrees C began laying eggs about a day earlier and had much higher daily fecundities than did those laying at 18 degrees C. Developmental temperature had no significant effect either on overall fecundity or on the pattern of daily egg production. Dam temperature had a slight effect on the pattern of daily egg production, but not on overall fecundity. In contrast, sire temperature slightly influenced both overall fecundity and the pattern of daily egg production. Our results demonstrate that early fecundity is extraordinarily sensitive to laying temperature (360 per cent increase if laying at 25 degrees C vs. at 18 degrees C), but is relatively well buffered against developmental and cross-generational effects (maximum effect only 7 per cent, for sire temperature).