Thermal acclimation affects the repeatability of upper thermal tolerance in Atlantic killifish (Fundulus heteroclitus)

As the frequency and intensity of heat extremes increases, understanding the capacity for evolution in traits such as thermal tolerance is critical for predicting biological responses to climate change. Here, we assessed the repeatability of CT max (a common measure of acute thermal limits in ectotherms) in Atlantic killifish ( Fundulus heteroclitus ) across three acclimation temperatures (10, 18 and 26°C), because repeatability is an important determinant of a trait's ability to respond to natural selection, but the impact of thermal acclimation on CT max repeatability remains unclear. Fish were individually marked and CT max was assessed weekly at least 3 times at each acclimation temperature. Mean CT max was consistent between 18°C acclimation groups at the beginning and end of the experiment (∼6 months apart) emphasizing the reliability of CT max as an estimate of acute thermal tolerance at the group level. CT max was highly repeatable at the two lower acclimation temperatures (R = 0.74-0.81), but not at the highest acclimation temperature (R = 0.20), resulting in moderate repeatability across the entire dataset (R = 0.42). There was little correlation of individual-level CT max across acclimation temperatures, or in 18°C-acclimated fish at the beginning and end of the experiment. This pattern was associated with substantial inter-individual variation in acclimation response ratio (ARR), suggesting individual variation in thermal plasticity. Together, these results suggest that repeatability in acute thermal limits is strongly contingent on thermal history, and that low repeatability at high acclimation temperatures is likely to constrain the capacity for adaptation in CT max as our climate warms.