Adaptation to seasonality : genetics and variation in life-history traits important in overwintering

Abstract

Adaptation to seasonal changes in environmental conditions is of crucial importance for species living at high latitudes. This involves not only tolerance to cold temperatures, but also an ability to timely predict the forthcoming cold season from gradual changes in the environment. The forecasting systems of several northern species have been found to rely on changes in photoperiodic cues, which are the most reliable seasonal cue at high latitudes. In my thesis, I examine the genetic basis of variation in one of the strongest adaptations to seasonally varying environments in insects - adult reproductive diapause - and other associated life-history traits in Drosophila montana. Study I revealed significant and predictable clinal variation in the critical day length (CDL) for diapause incidence in presence of high gene flow. Study II showed high variation in factors affecting the diapause incidence both within and between the same clinal populations and showed that while variation in females' photoperiodic responses decreases during laboratory maintenance, the mean CDLs of the strains remain the same. Study III was carried out with a "resonance" experiment, revealed no rhythmicity in the females' diapause response, which suggests that the photoperiodic timer of D. montana is based on a non-oscillatory hourglass timer or rapidly dampening circadian oscillator. In study IV we performed a QTL analysis by crossing two D. montana strains that differed in their diapause incidence, cold tolerance, egg-to-eclosion development time and body weight at a specific day length. This study revealed both unique and combined QTL for the studied traits, the most interesting result being a large-effect QTL for diapause on the X chromosome. While we did not find evidence for the involvement of a circadian clock in the photoperiodic regulation of reproductive diapause in D. montana in the above-mentioned studies, some QTL for traits regulated by different time-measuring systems were overlapping (IV).