Latitudinal patterns in butterfly life history and host plant choice

Abstract: The central subject in this thesis is adaptation of insect populations to different environments along latitudinal gradients. When latitude decreases, the longer favorable season causes an increase in time available for development. However, predictions are complicated by shifts in voltinism. When a species uses the summer for two generations (bivoltine lifestyle) instead of one, the time for each generation is virtually cut in two halves. For this reason models predict a saw-tooth pattern in development time and size. In herbivorous insects, a related model predicts specialization on plants permitting fast growth when time available for development is short (the voltinism-suitability hypothesis). The hypotheses apply to insects with discrete number of generations per year and I have studied butterflies. Manuscript I is an investigation of size response in the butterfly Polyommatus icarus over a latitudinal shift from univoltinism to bivoltinism. The investigation of wild-caught specimens is supplemented with a laboratory experiment on tree populations originating from above, inside and below this shift. The study confirms the predicted saw-tooth pattern size response to available time in this species. Additionally, the laboratory results reveal so-called counter-gradient variation. The second manuscript focuses on the voltinism-suitability hypothesis, and the model organism is now another butterfly, Polygonia c-album, also the subject for the last two studies. The main result of the second study is that female host plant preference and larval survival show much variation between areas, whereas larval development time and growth rate show little variation. Further, the variation in larval performance and adult preference between the two studied host plants (Salix caprea and Urtica dioica) seems to be best explained by host plant phenology. The study also reveals a possible trade-off in performance on the two studied host plants. The third study is an investigation of the genetics of host plant preference in one population via breeding and selection experiments; and the fourth a study of the genetics of preference differences between geographically separated populations via cross-breeding experiments. The variation in host plant preference within the population seems caused by non-additive autosomal inheritance, perhaps with a small effect of the y-chromosome. The variation between populations is primarily caused by one or more genes on the x-chromosome.