Emergent increase in coral thermal tolerance reduces mass bleaching under climate change

Extreme climatic events are already causing acute impacts in marine habitats, and coral reefs are among the most vulnerable ecosystems to climate change due to marine heatwaves that trigger mass bleaching and mortality. For reefs to persist, coral assemblages must increase their thermal tolerance—the threshold above which thermal stress occurs. Although some reefs have shown reduced bleaching susceptibility after repeated events, the natural rate at which coral thermal tolerance can increase, and the ecological levels at which this occurs, remain uncertain. Quantifying historic and potential future rates of increase in thermal tolerance is crucial for forecasting bleaching trajectories and assessing whether ecological adjustments can keep pace with ocean warming.

Prior work documents recurrent mass bleaching and its drivers, with DHW (degree heating weeks) widely used to quantify accumulated heat stress; mass bleaching typically occurs beyond 4 °C-weeks and significant mortality beyond 8 °C-weeks. Observations indicate increased bleaching resistance at some reefs after repeated events, potentially via community turnover away from sensitive taxa, genetic adaptation, acclimatisation, and changes in symbiont communities. However, many future projection studies either omit adaptive mechanisms or implement fixed, linear assumptions about increased tolerance, which may be biologically unrealistic given likely nonlinearities in community and evolutionary processes. Empirical cases (e.g., Great Barrier Reef and other Indo-Pacific sites) suggest shifts in susceptibility, but the rate at which assemblage-level thermal thresholds rise has not been well quantified.

Study area and data: The analysis focuses on Palau, an isolated coral reef system. Historic heat stress was derived from NOAA Coral Reef Watch CoralTemp v3.1 satellite SST (0.05° grid, 1985–2020). DHW was computed following NOAA CRW methods using MMM and a standard 1 °C bleaching threshold. Photosynthetically available radiation (PAR) for the main heatwave years (1998, 2010, 2017) was obtained from SeaWiFS (1998) and MODIS Aqua (2010, 2017), harmonized to a common grid; interannual differences for Aug–Sep were tested using a linear mixed effects model with Tukey post hoc tests. Thermal tolerance simulations: Thirteen scenarios were created in which the thermal stress threshold (MMM + 1 °C) was increased linearly from 1988 at rates between 0.0 and 0.3 °C/decade (increments of 0.025). Thresholds were adjusted in annual steps at seasonal lows. For each rate, DHW time series were recalculated for Palau reef cells (N=152). Bleaching observations and modeling: A dataset of 237 underwater bleaching records (1998–2017) was compiled, summarizing bleaching as severity scores (0–3). For each observation, coincident DHW was extracted from the encapsulating 5 km grid. Bleaching severity was transformed to a proportion (with a standard adjustment to avoid 0/1 bounds) and modeled as a function of DHW using a spatial beta GLM fitted with INLA (R-INLA). Spatial dependence was accounted for with a Matérn correlation implemented via a Gaussian Markov random field over a 2 km Delaunay mesh (6,210 nodes). Model performance across the 13 tolerance-rate scenarios was compared using DIC (parsimony), prediction success rate, and misclassification rates (over- and under-prediction), explicitly adjusting for spatially correlated uncertainty. Future projections: Daily SST from 17 CMIP6 GCMs were downscaled to CoralTemp resolution using bilinear and nearest-neighbor regridding, seasonally adjusted to the observational baseline (1985–2010) via the delta method. Four SSPs were analyzed: SSP1-2.6, SSP2-4.5, SSP3-7.0, SSP5-8.5. For each simulated tolerance-rate (0.0, 0.1, 0.2, 0.3 °C/decade), DHW was computed and bleaching conditions identified using NOAA’s Alert Level 2 threshold (8 °C-weeks). High-frequency bleaching was defined as at least two years meeting/exceeding 8 °C-weeks within any centered 10-year window. The proportion of Palau reef cells experiencing high-frequency bleaching was calculated annually. Sensitivity/context checks: Differences in PAR among 1998, 2010, 2017 were assessed to evaluate whether light intensity could explain bleaching differences. Spatial model residuals and uncertainty patterns were examined to avoid biases from spatial non-independence.

- Historic events and PAR: Palau experienced mass bleaching conditions in 1998 and 2010 with maximum DHW of 7.1 and 9.0 °C-weeks (reef-cell averages 5.6 and 6.4, respectively). In 2017, maximum DHW was 7.8 °C-weeks (average 6.5), yet little to no bleaching was recorded. PAR during Aug–Sep was similar among years (1998: 47 ± 3; 2010: 44 ± 2; 2017: 44 ± 2 mol m⁻² day⁻¹), with 1998 slightly higher (~3 mol m⁻² day⁻¹; P < 0.001), suggesting light differences do not explain 2017’s reduced bleaching. - Most-likely tolerance increase: Among 13 simulated rates (0.0–0.3 °C/decade), 0.1 °C/decade produced the most parsimonious model (lowest DIC), highest prediction success (>65%), and lowest misclassification (≈15% over-predictions, 19% under-predictions). Faster simulated rates caused unrealistic low DHW in later years and higher misclassification. - Sensitivity of DHW to tolerance rate: For the 2010 heatwave, maximum DHW was 6.6 °C-weeks with no enhancement, but only 1.7 and 0.3 °C-weeks under 0.1 and 0.2 °C/decade, respectively; at 0.3 °C/decade DHW did not accumulate. - Future projections: Without tolerance increases, high-frequency bleaching (≥2 events per decade with DHW >8) occurs by ~2040 across SSPs. Maintaining 0.1 °C/decade substantially reduces bleaching trajectories. Under SSP1-2.6, high-frequency bleaching peaks around 2050 (≈50% of reefs) and declines to ≈25% by 2100. Under SSP2-4.5, SSP3-7.0, and SSP5-8.5, 0.1 °C/decade delays high-frequency bleaching onset by ~10–20 years relative to no enhancement, but most (SSP2) to all reefs (SSP3, SSP5) face high-frequency bleaching by 2100. Faster increases (0.2–0.3 °C/decade) can avoid high-frequency bleaching for most reefs under SSP1-2.6 and SSP2-4.5, partly avoid under SSP3-7.0 (only 0.3 °C/decade), and only delay impacts under SSP5-8.5.

The study addresses whether natural increases in coral thermal tolerance can keep pace with ocean warming by quantifying an emergent, assemblage-level increase in Palau of approximately 0.1 °C/decade since the late 1980s. The absence of mass bleaching in 2017 despite DHW comparable to earlier mass-bleaching events, coupled with similar PAR, points to biological mechanisms—community composition shifts, genetic adaptation, acclimatisation, and symbiont changes—rather than environmental differences alone. These findings imply some ecological resilience and indicate that incorporating dynamic thermal tolerance into projections can meaningfully alter expected bleaching trajectories. However, under middle-to-high emissions scenarios, even continued increases at historic rates predominantly delay rather than prevent high-frequency bleaching, underscoring the critical importance of ambitious emissions reductions alongside potential management interventions (e.g., assisted evolution, protection and reseeding from resistant reefs).

This work introduces a simulation and spatial statistical framework that infers historical rates of increase in coral thermal tolerance using observed bleaching records and heat stress. For Palau, an emergent increase of about 0.1 °C/decade best explains reduced bleaching severity in recent events. Applying dynamic tolerance in climate projections shows substantial mitigation of bleaching risk under low-to-mid emissions scenarios if such increases continue, but only delays under high emissions. The study contributes quantitative evidence for rising thermal tolerance, highlights likely biological underpinnings, and supports integrating adaptive capacity into forecasts. Future research should resolve species- and symbiont-level mechanisms, potential physiological limits and trade-offs, and nonlinear dynamics in tolerance shifts. Meeting Paris Agreement targets remains essential to preserve coral reefs, with management actions potentially enhancing natural tolerance gains.

- The model treats thermal tolerance as a gradually and linearly increasing threshold starting in 1988; real-world processes may be nonlinear and punctuated by selective events. - Physiological upper limits to thermal tolerance, potential trade-offs with other fitness traits, and changing responses near thermal limits are not captured. - Analyses are at the assemblage level; species- and within-species variability in thresholds are not explicitly modeled. - The approach cannot assign probabilities to future rates of tolerance increase or fully disentangle mechanisms (composition turnover, genetic adaptation, acclimatisation, symbiont shifts). - Environmental variables beyond SST and PAR (e.g., water quality, fine-scale temperature variability, thermal priming) may influence bleaching but were not fully incorporated.