Performance of embryos reared under fluctuating conditions does not conform to predictions based on performance at constant thermal conditions

Little is known about the impacts of fluctuating temperatures on early development in fishes, as most experiments on the effects of temperature use constant temperature exposures. Here, we asked whether thermal performance curves (TPCs) for development generated at constant temperatures can be used to make predictions about performance at fluctuating temperatures. We incubated embryos of a topminnow (Fundulus heteroclitus) to the same mean temperature with differing extents of diel fluctuation (26±0°C, 26±3°C, 26±5°C, 26±7°C). Based on TPCs from constant temperatures, we predicted that developmental rate and survival would decrease with increasing fluctuation. Consistent with our prediction, embryos incubated at 26±7°C had lower survival, but inconsistent with our prediction, they developed more rapidly than all other groups. In addition, fish in the highest fluctuation regime were longer and had a larger yolk-sac volume at hatch, suggesting that greater fluctuations result in more efficient energy utilization. At the mRNA level, embryos incubated at 26±7°C had higher expression levels of an inducible heat shock protein, hsp70.2, suggesting thermal stress. Once hatched, larvae were raised at a common constant temperature of 26°C to test for persistent effects. Embryos exposed to 26±7°C during development were larger than those reared under constant conditions 1 week post-hatch, and several genes involved in the heat-shock response and DNA methylation exhibited altered mRNA levels. Our data demonstrate that embryos raised under constant and fluctuating temperatures have different phenotypic responses, which highlights the need to incorporate variable thermal regimes into developmental studies.