| Summary: | Using a combination of computer simulations and laboratory experiments we test if the thermal sensitivity of growth rates change during ontogeny in damselfly larvae and if these changes can be predicted based on the natural progression of average temperature or thermal variability in the field. The laboratory experiment included replicated species from Southern, Central and Northern Europe. Although annual fluctuations in temperature represent a key characteristic of temperate environments, few studies of thermal performance have considered the ecological importance of the studied traits within a seasonal context. Instead, thermal performance is assumed to remain constant throughout ontogeny and reflect selection acting over the whole life cycle. The laboratory experiment revealed considerable variation among species in the strength and direction of ontogenetic performance shifts. In four species from Southern and Central Europe, reaction norms were steepest during early ontogeny, becoming less steep during later ontogenetic stages (indicative of low-temperature acclimation). In one Northern European species, the slope of reaction norms did not change during ontogeny. In the other North European species, reaction norms became steeper during ontogeny (indicative of high-temperature acclimation). We had expected high-latitude species to show strong low-temperature acclimation responses, because they have a short flight season and inhabit a strongly seasonal environment. Instead, we found the reversed pattern low-latitude species displayed strong low-temperature acclimation responses and high-latitude species displayed weak, or even reversed, acclimation responses to low temperatures. These findings suggest that low-temperature acclimation may be less beneficial and possibly more costly in habitats with rapid seasonal transitions in average temperature. We conclude that thermal performance traits are more dynamic than typically assumed and caut ion against using results from single ontogenetic stages to predict ...
|
|---|