Temperature variation in time and space, and its effects on insects

Nature is variable. Unfortunately, compressed representations of this variable world, like averages, are often lossy and insufficient for ecological purposes. This is particularly true for temperature variation, which organisms typically respond to in a nonlinear way. As biologists, we must therefore be careful to study temperature variation at the appropriate scales, and assess its consequences in the right biological contexts. In this dissertation, I tackle the interplay between temperature variation, seasonality, and life histories of insects, primarily focusing on Pieris butterflies. In Chapter I, I demonstrate that insect development times in naturally fluctuating settings can be accurately predicted using thermal performance curves established under constant settings. However, this accuracy is contingent upon the incorporation of environmental temperature data high-resolved in both time and space. My work in Chapter II investigates the divergent seasonal population dynamics exhibited by Pieris rapae and P. napi, two closely related and ecologically similar butterflies. The species’ differences in season-specific success correlate with distinct thermal adaptations, and delineate P. rapae and P. napi into the roles of summer and winter specialists, respectively. We hypothesize that warm-adapted summer specialists will be favored by climate warming, but that cold-tolerant winter specialists will find refuge in places with very short growth seasons. In Chapter III, a comprehensive examination spanning a 750 km latitudinal cline unveils discernible latitude-specific photoperiodic reaction norms in P. napi, yet an absence of parallel trends in their thermal responses. We argue that, in seasonal environments, the reliability of photoperiodic cues and the clear link between photoperiodism and fitness make photoperiodic responses evolve more readily than temperature responses. In Chapter IV, I integrate principles from signal processing into thermal ecology. I show that relatively sparse temperature time-series can be effectively interpolated using well-known signal processing techniques, improving the accuracy of ecological predictions. The Earth is warmer now than it was just a century ago, and will likely keep facing drastic temperature changes in the near future. This will have complex downstream effects on living organisms all over the world. As a concluding remark, I would therefore like to emphasize the importance of a nuanced perspective on the consequences of temperature variation in nature