RESUMEN
Climate change is the most significant threat to natural World Heritage (WH) sites, especially in the oceans. Warming has devastated marine faunas, including reef corals, kelp, and seagrass. Here, we project future declines in species and ecosystem functions across Australia's four WH coral reef regions. Model simulations estimating species-level abundances and probabilities of ecological persistence were combined with trait space reconstructions at "present," 2050 (+1.5°C of warming), and 2100 (+2°C) to explore biogeographical overlaps and identify key functional differences and forecast changes in function through time. Future climates varied by region, with Shark Bay projected to warm the most (>1.29°C), followed by Lord Howe, when standardized to marine park size. By 2050, ~40% of the Great Barrier Reef will exceed critical thresholds set by the warmest summer month (mean monthly maximum [MMM]), triggering mortality. Functional diversity was greatest at Ningaloo. At +1.5°C of warming, species and regions varied drastically in their functional responses, declined 20.2% in species richness (~70 extinctions) and lost functions across all reefs. At +2°C, models predicted a complete collapse of functions, consistent with IPCC forecasts. This variability suggests a bespoke management approach is needed for each region and is critical for understanding WH vulnerability to climate change, identifying thresholds, and quantifying uncertainty of impacts. This knowledge will aid in focusing management, policy and conservation actions to direct resources, rapid action, and set biodiversity targets for these reefs of global priority. As reefs reassemble into novel or different configurations, determining the winners and losers of functional space will be critical for meeting global landmark biodiversity goals.