Abstract
Upper thermal limits of ectotherms are widely used to understand and predict species' thermal responses and sensitivity to warming. These limits are often defined for species using experiments with rapid ramping temperatures that test critical thermal maxima (CTmax). However, there are issues that arise with relying on these experimental results including (1) the influence of experimental design on thermal maxima, (2) the lack of ecological realism and (3) the potential for population-level local adaptation of upper thermal limits. We addressed these issues by comparing the CTmax approach with an ecologically realistic design using slower incremental temperature ramping with diel fluctuations (ITDmax) and by applying both to evaluate local adaptation of juvenile coho salmon (Oncorhynchus kisutch). We compared populations from thermal regimes spanning 7° latitude and coastal to inland systems by testing three populations, combining results with a fourth population from a prior ITDmax study, and comparing with other studies that used CTmax experiments to test thermal maxima of juvenile coho salmon. Most notably, we found that unlike CTmax experiments, ITDmax results were not influenced by acclimation temperature. This stemmed from acclimation during the ITDmax trials, likely representing more ecologically relevant responses to longer term warming. Furthermore, local adaptation of thermal maxima, as measured by both CTmax and ITDmax, was not evident for juvenile coho salmon, with no influence of population across the nine included in the cross-study examination. The results suggest the ability to use ITDmax-based upper thermal limits across species' extents and with differing prior environmental exposure, providing a more accurate representation of responses and sensitivity to long-term warming.