Fig. 5: Drivers of future coral persistence under plausible global warming.

a Factors associated with healthier reef outcomes under SSP1-2.6, SSP2-4.5, and SSP3-7.0 for both mid-century (left) and end-century (right) based on predictions of statistical linear models. The influence of each predictor (Supplementary Table 2) is measured by the change in coral cover (as percentage points) for a 10% change of the predictor within its range. The final row represents external larval supply (i.e., excluding local larval retention). CoTS crown-of-thorns starfish (Acanthaster spp.), WQ water quality. Probability (kernel) density distribution of reefs based on heat-sensitive coral cover across thermal (b, warm spots and thermal refugia) and larval supply (c, larval hubs and larval barrens) regimes, excluding reefs impacted by reduced water quality and cyclones in the preceding 5 years (resulting distributions across 200 climate projections: n = 202,570 reefs in 2050, n = 205,510 in 2100). Thermal and larval supply regimes are defined, respectively, as those reefs below the 10th or above the 90th percentile for annual DHW and annual amount of external larval supply. Reefs in thermal refugia (i.e., exposed to lowest thermal stress) and in larval hubs (i.e., exposed to highest larval supply) exhibit higher levels of coral cover compared to warm spots and larval barrens. Thermal environments and larval connectivity were calculated as the average heat stress and larval supply experienced in the preceding 5 years. Larval supply excludes local retention and only captures larvae from external source reefs. Colours of SSPs as in (a).