1. Many species that are moving polewards encounter novel thermal regimes to which they have to adapt. Therefore, rapid evolution of thermal tolerance and of thermal plasticity in fitness-related traits in edge populations can be crucial for the success and speed of range expansions.2. We tested for adaptation in cold tolerance and in life history, behavioural and physiological traits and their thermal plasticity during a poleward range expansion.3. We reconstructed the thermal performance curves of life history (survival, growth and development rates), behaviour (food intake) and cold tolerance (chill coma recovery time) in the aquatic larval stage of the damselfly Ischnura elegans that is currently showing a poleward range expansion in northern Europe. We studied larvae from three edge and three core populations using a common-garden experiment.4. Consistent with the colder annual temperatures, larvae at the expansion front evolved an improved cold tolerance. The edge populations showed no overall (across temperatures) evolution of a faster life history that would improve their range-shifting ability. Moreover, consistent with damselfly edge populations from colder latitudes, edge populations evolved at the highest rearing temperature (28°C) a faster development rate, likely to better exploit the rare periods with higher temperatures. This was associated with a higher food intake and a lower metabolic rate. 5. In conclusion, our results suggest that the edge populations rapidly evolved adaptive changes in trait means and thermal plasticity to the novel thermal conditions at the edge front. Our results highlight the importance of considering besides trait plasticity and the evolution of trait means, also the evolution of trait plasticity to improve forecasts of responses to climate change.
K E Y W O R D Schill coma recovery time, evolution of plasticity, life history evolution, range expansion, thermal performance curves
| INTRODUC TI ONMany species are shifting their distributions, often in a poleward direction (Chen et al., 2011). It is becoming increasingly clear that it is crucial to study the evolution of trait changes associated with range shifts to mechanistically understand (Diamond, 2018) and forecast (Bush et al., 2016; Nadeau & Urban, 2019) range dynamics. Yet, while edge-core differentiation in phenotypic traits has been widely documented (e.g. overviews in Chuang & Peterson, 2016; Hill et al., 2011), we often lack knowledge whether the trait changes are plastic or | 1667