Weight accumulation
Compared to the results for development rate, the effects of photoperiod treatment on weight were weaker, and quite population-dependent (analysis of deviance; population × treatment χ210=22.5; p=0.012). The only population that showed a significant difference in final size between the control treatments was Öland, where constant exposure to short days produced pupae that were approximately 10% heavier than did constant exposure to long days (within-population planned contrast; t567=4.01, p=0.001). This difference was not detectable at any earlier life stage. In contrast to the results for development time, changes in photoperiod regime during development did not appear to drive final size; instead, pupal weights tended to correspond to the initial photoperiod experienced (Fig. 4; Fig. S2). Females were larger than males across all populations (analysis of deviance; sex χ21=142.1; p<0.001); unlike the late effects of photoperiod, the sex difference was detectable as early as the third instar, and was increased further in later stages.