Geographic variation in life cycles : Local adaptation and ecological genetics in a temperate butterfly

Abstract: Conditions in nature change with the seasons, necessitating seasonal adaptations that synchronize the life cycles of organisms with their surroundings. Such regulatory adaptations must vary between populations to track local variation in climate and seasonality; this local adaptation is facilitated by locally specific seasonal cues, but may be hampered by gene flow and genetic history.      For populations of temperate insects, two central features of adaptation to local climate are voltinism, the yearly number of generations; and diapause, the state of arrested development and suppressed metabolism in which most temperate insects spend winter. Delaying diapause allows for an additional generation to be produced within the same year, but this is only adaptive if the season is sufficiently long to safely accommodate such a life cycle. Hence, selection to express a locally adaptive voltinism should drive divergence between populations in diapause regulation and associated life history traits. In this thesis, I investigate variation in voltinism and life cycle regulation in a set of populations of the butterfly Pararge aegeria.      Population-level variation in seasonal plasticity was tested in two sets of experiments. The first (Paper I) focused on photoperiodic plasticity during the growing season, and revealed considerable differences between populations in diapause induction and developmental reaction norms. Mechanistic modeling based on the laboratory results indicated that differences in voltinism are actively maintained by these genetic differences. Next, I tested the idea that shorter diapause may help populations achieve higher voltinism through earlier emergence in the spring (Paper II). This idea was not supported; instead, populations differed in a manner that suggests that diapause duration is selected upon by the need to avoid premature development under warm autumn conditions.      The genetic background of seasonal adaptation in these populations was also explored. Phylogeographic structures inferred from genome-wide data put the results of the laboratory experiments into a historic context, and were used as the basis for a scan for genetic loci showing signs of differential selection (Paper III). The scan revealed novel variation in two circadian genes that have been shown to be linked to diapause control in P. aegeria, including a large deletion in the gene timeless. Finally, a test of two previously described circadian mutations (Paper IV) showed that, while these mutations may affect photoperiodic plasticity on a between-population level, they seemingly have no effect within a single population located at intermediate latitudes. Closer inspection revealed novel, locally unique mutations in the same genes, possibly compensating for the effect of diapause-delaying variants in a setting where an attempted second generation is not adaptive.      I have shown that voltinism variation in P. aegeria is enabled by population differences in seasonal plasticity, with population differences playing a greater role during some parts of the year than others. These results present voltinism as a complex trait resulting from plasticity acting at different levels of geographic specificity. Although much of the genetic variation enabling the observed local adaptation remains uncharacterized, the considerably variable circadian genes seen in these populations provide an intriguing target for future investigation.