Fig. 2: Acropora environmental genomic model for candidate heat-adaptive genotypes.

A Environmental genomic model predicting the expected frequency of candidate heat-adaptive genotypes from patterns of past heatwave exposure, described by maximal Degree Heating Week (DHW_max) measured across 5-year windows. The environmental genomic model was based on five datasets (total number of reefs N = 53). For each dataset, the figure shows the frequency of candidate heat-adaptive genotypes for each reef (left side) and the reef thermal history (right side), represented by DHW_max values aggregated into 5-year windows spanning from 25 years before sampling to the year of sampling. B Spatial predictions of the frequency of candidate heat-adaptive genotypes across the reefs of five Acropora datasets. For every dataset, predictions were made using the environmental genomic model trained on data from the other four datasets (i.e., leave-one-out cross validation). The mean absolute error (MAE) of the predicted genotype frequencies is shown in the top right corner of each map. Circles represent real genotype frequencies at sampling sites. Background maps were produced using the Global Relief Model from NOAA (ETOPO 2022⁸²) and the Global Distribution Map of Coral Reefs from the UNEP World Conservation Monitoring Center⁸³.