Acclimation capacity to global warming of amphibians and freshwater fishes: Drivers, patterns, and data limitations

Amphibians and fishes play a central role in shaping the structure and function of freshwater environments. These organisms have a limited capacity to disperse across different habitats and the thermal buffer offered by freshwater systems is small. Understanding determinants and patterns of their physiological sensitivity across life history is, therefore, imperative to predicting the impacts of climate change in freshwater systems. Based on a systematic literature review including 345 experiments with 998 estimates on 96 amphibian (Anura/Caudata) and 93 freshwater fish species (Teleostei), we conducted a quantitative synthesis to explore phylogenetic, ontogenetic, and biogeographic (thermal adaptation) patterns in upper thermal tolerance (CT_max) and thermal acclimation capacity (acclimation response ratio, ARR) as well as the influence of the methodology used to assess these thermal traits using a conditional inference tree analysis. We found globally consistent patterns in CT_max and ARR, with phylogeny (taxa/order), experimental methodology, climatic origin, and life stage as significant determinants of thermal traits. The analysis demonstrated that CT_max does not primarily depend on the climatic origin but on experimental acclimation temperature and duration, and life stage. Higher acclimation temperatures and longer acclimation times led to higher CT_max values, whereby Anuran larvae revealed a higher CT_max than older life stages. The ARR of freshwater fishes was more than twice that of amphibians. Differences in ARR between life stages were not significant. In addition to phylogenetic differences, we found that ARR also depended on acclimation duration, ramping rate, and adaptation to local temperature variability. However, the amount of data on early life stages is too small, methodologically inconsistent, and phylogenetically unbalanced to identify potential life cycle bottlenecks in thermal traits. We, therefore, propose methods to improve the robustness and comparability of CT_max/ARR data across species and life stages, which is crucial for the conservation of freshwater biodiversity under climate change.