Integrating environmental variability to broaden the research on coral responses to future ocean conditions.

Our understanding of the response of reef-building corals to changes in their physical environment is largely based on laboratory experiments, analysis of long-term field data, and model projections. Experimental data provide unique insights into how organisms respond to variation of environmental drivers. However, an assessment of how well experimental conditions cover the breadth of environmental conditions and variability where corals live successfully is missing. Here, we compiled and analyzed a globally distributed dataset of in-situ seasonal and diurnal variability of key environmental drivers (temperature, pCO , and O ) critical for the growth and livelihood of reef-building corals. Using a meta-analysis approach, we compared the variability of environmental conditions assayed in coral experimental studies to current and projected conditions in their natural habitats. We found that annual temperature profiles projected for the end of the 21st century were characterized by distributional shifts in temperatures with warmer winters and longer warm periods in the summer, not just peak temperatures. Furthermore, short-term hourly fluctuations of temperature and pCO may regularly expose corals to conditions beyond the projected average increases for the end of the 21st century. Coral reef sites varied in the degree of coupling between temperature, pCO , and dissolved O , which warrants site-specific, differentiated experimental approaches depending on the local hydrography and influence of biological processes on the carbonate system and O availability. Our analysis highlights that a large portion of the natural environmental variability at short and long timescales is underexplored in experimental designs, which may provide a path to extend our understanding on the response of corals to global climate change.