Coral reefs benefit from reduced land-sea impacts under ocean warming

The study addresses how specific, actionable combinations of land- and sea-based management can enhance coral reef resistance, survival, and recovery under climate-driven marine heatwaves. Coral reefs near populated coasts face compounded pressures from land-based stressors (for example, wastewater, nutrient and sediment runoff) and sea-based stressors (for example, overfishing), alongside increasingly frequent and severe marine heatwaves that cause mass bleaching and mortality. Although traditional conservation emphasizes reducing local impacts, contemporary governance often separates terrestrial and marine management, and managers lack clear, evidence-based targets for integrated action. The authors aim to identify which local land-sea factors most effectively support coral persistence before, during, and after acute thermal disturbance, using spatially and temporally resolved data rather than coarse proxies such as population density or composite water-quality indices.

The paper situates its work within literature showing: (1) climate change is driving more frequent and severe marine heatwaves and transforming coral assemblages globally; (2) local conditions can modulate reef resistance and recovery following bleaching; (3) human population density and reef accessibility are imperfect proxies for local impacts; and (4) integrated land-sea stewardship is historically rooted in indigenous practices but is under-implemented in modern, sector-based governance. Prior studies highlight roles of fish functional groups in reef functioning, the negative effects of poor water quality (including terrestrial runoff and wastewater), and cases where even intact fish communities can be overwhelmed by extreme heat stress. The authors build on this by providing fine-scale, longitudinal measures of multiple human and environmental drivers to specify policy-relevant levers (for example, wastewater reduction, herbivore management) that improve reef outcomes.

Study region: Hawaiian Islands coastal reefs spanning broad gradients in land-sea human impacts and environmental conditions, including exposure to the most severe recorded Hawaiian marine heatwave (2015). Data: A unique, annually resolved 20-year (2000–2019), 100 m-resolution time series of local human impacts and environmental factors, including urban runoff, wastewater pollution, nutrient loading, sediment input, rainfall (annual and peak), wave exposure, ocean temperature means and variability, heat stress, irradiance, phytoplankton biomass, and fishing gear restrictions. Multiple fish biomass metrics were compiled, including total fish biomass and herbivore functional groups (scrapers, grazers, browsers). Benthic surveys: Recurring, permanently marked site-specific surveys of reef benthic communities across three periods: predisturbance (2003–2014), during and immediately following the heatwave (2014–2016), and postdisturbance (2016–2019). Sample sizes: trajectories predisturbance n=23 reefs; heatwave response n=80; four years postdisturbance n=55. Analytical approach:

• Predisturbance trajectories: Classified reefs as positive or negative trajectory if coral cover changed by more than ±3% between 2003 and 2014. Compared multivariate local conditions between trajectory groups using PERMANOVA (pseudo-F1=3.38, P=0.001) and canonical analysis of principal coordinates; evaluated allocation success; examined differences across all predictors (drop-one jackknife) including fish biomass groups and land-based impacts.
• Heatwave response (2015→2016): Used generalized additive mixed models (GAMMs; R²=0.79) to relate percentage coral cover change (accounting for starting cover) to local factors during the heatwave. Considered relative importance via summed AICc weights; handled multicollinearity by removing highly correlated predictors (|r|>0.7). Key factors retained included sediment input, scraper biomass, total fish biomass, urban runoff, phytoplankton biomass, wastewater pollution, peak rainfall, nutrient loading, grazer biomass, DHW, wave power, depth, and fishing gear restrictions.
• Postdisturbance persistence (2019): Defined reef-builder cover as hard corals + crustose coralline algae. Categorized sites into low (≤25th percentile), moderate (>25th and <75th percentile), and high (≥75th percentile) reef-builder cover. Applied ordinal logistic regression to identify significant predictors of category membership; constructed scenario analyses varying scraper biomass (sea-based management) and wastewater pollution (land-based management) to estimate probabilities of low/moderate/high cover outcomes. Management scenario ranges were anchored to observed percentiles and real-world references (for example, Kealakekua Bay protected area for scraper biomass). Validation and distributions: Predictor distributions provided in supplementary figures; extended data tables list full factor sets and model outputs. Heat stress quantified via DHW; 2015 seasonal SSTs documented. Spatial and temporal variability in drivers summarized across the coastline at 100 m resolution.

• Predisturbance (2003–2014): Overall mean coral cover in 2003 was 36.9% ± 2.3 (n=23), with <3% change across the region up to 2014. However, individual reefs diverged: 44% increased, 35% decreased, remainder unchanged. Reefs on positive trajectories had substantially higher fish biomass: total and herbivorous groups (scrapers, grazers, browsers) were 24–113% greater (29–214 kg ha⁻¹ higher) than on negative-trajectory reefs. Negative-trajectory reefs had 46–80% higher wastewater pollution, nutrient loading, and urban runoff. Human population density within 15 km was paradoxically 63% higher around positive-trajectory reefs, underscoring that population density is a poor proxy for local impacts. Differences in fishing gear restrictions, depth, sediment input, ocean temperatures, phytoplankton biomass, and rainfall were minimal; wave exposure slightly higher on positive reefs but regionally low overall.
• Heatwave response (2015–2016): The 2015 heatwave peaked at 29.4 °C; reefs averaged 12 DHWs, exceeding the 8-DHW severe-bleaching threshold. Outcomes varied: 19/80 reefs lost >20% coral cover; the worst lost 49%, while 14/80 had stable or increased cover. GAMM (R²=0.79) indicated reduced coral loss at reefs with higher phytoplankton (chlorophyll-a), lower urban runoff, and to a lesser extent lower sediment input. Total fish biomass and scraper biomass were also important (high AICc weights) but had weak slopes relative to heat-driven loss. Factor importance (AICc weights): sediment input (0.99), scraper biomass (0.99), total fish biomass (0.90), urban runoff (0.60), phytoplankton (0.38), wastewater (0.28), peak rainfall (0.20), nutrient loading (0.19), grazer biomass (0.16), DHW (0.08), wave power (0.07), depth (0.06), fishing gear restrictions (0.05).
• Four years postdisturbance (2019): Reef-builder cover ranged 3.4–51.9% (mean 24.3% ± 1.7; n=55), with notable reshuffling of which reefs were high (≥75th percentile) or low (≤25th percentile) compared to pre-heatwave distributions. Ordinal logistic regression identified decreased wastewater pollution and increased scraper biomass as the most important, significant predictors (P<0.05) of higher reef-builder cover category.
• Management scenarios: With low scrapers (30 kg ha⁻¹) and high wastewater (600,000 l ha⁻¹), probability of low reef-builder cover was 0.83. Managing only sea-based (scrapers at 250 kg ha⁻¹) with high wastewater yielded 0.70 probability of moderate cover and 0.14 of high. Managing only land-based (wastewater reduced to 2,500 l ha⁻¹) with low scrapers yielded 0.83 probability of moderate cover and 0.26 of high. Integrated management (high scrapers + low wastewater) produced 0.80 probability of high reef-builder cover. This represents a three- to sixfold increase in the probability of high reef-builder cover versus single-domain management.

The findings directly address the central question of which actionable local interventions support coral persistence under ocean warming. Across distinct temporal windows, different mechanisms dominated: Before disturbance, higher fish biomass, particularly herbivores, coincided with increasing coral cover, likely via positive feedbacks where coral habitat promotes fish and herbivores limit algal competition. During the heatwave, natural energetic subsidies (higher phytoplankton) and reduced land-based pollution (urban runoff and sediment) modestly attenuated coral mortality, suggesting that both bottom-up resources and alleviation of co-stressors can buffer thermal impacts. Postdisturbance, scraper biomass emerged as critical for promoting higher reef-builder cover, consistent with scrapers facilitating substrate cleaning, crustose coralline algae, and coral recruitment. Crucially, integrated land-sea management produced synergistic benefits beyond either alone, markedly increasing the probability of high reef-builder cover four years after severe heat stress. While extreme heatwaves can overwhelm even well-protected fish communities, targeted reductions in wastewater and urban runoff combined with sustaining herbivore (scraper) populations improved persistence and recovery potential. These results provide managers with specific levers—curbing wastewater inputs and supporting herbivorous fish—aligned with indigenous mountains-to-sea stewardship principles, and argue for coupling land and sea conservation targets (for example, in 30 by 30 initiatives) to realize coastal resilience. Nonetheless, the protective capacity of local management has limits under unchecked warming, reinforcing the need for simultaneous global emissions reductions.

Simultaneous mitigation of land-based pollution (especially wastewater and urban runoff) and reinforcement of sea-based herbivory (notably scrapers) promotes coral reef persistence before, during, and after extreme marine heatwaves. The study’s high-resolution, long-term datasets identify when and how specific factors matter: broad fish biomass supports positive trajectories pre-disturbance; natural subsidies and reduced runoff lessen acute losses; and scraper biomass alongside low wastewater predicts higher reef-builder cover postdisturbance. Scenario modeling shows integrated land-sea actions yield a three- to sixfold higher probability of achieving high reef-builder cover than single-sector management. These insights translate into clear, actionable targets for managers and support a shift from siloed governance to integrated land-sea policy. Future research should test transferability across regions with differing natural productivity and governance contexts, refine thresholds for key drivers (for example, wastewater volumes, scraper biomass), and evaluate cost-effectiveness and equity of integrated interventions under varying warming scenarios, alongside global greenhouse gas mitigation.

• Generalizability and data availability: The uniquely detailed, spatially and temporally resolved datasets used here are rare; many regions lack comparable measurements and may rely on proxies, potentially limiting transferability and precise targeting.
• Thermal extremes may overwhelm local benefits: Intense or frequent future heatwaves could surpass the buffering capacity of local management, as evidenced by weak short-term fish effects during peak heat stress and documented severe mortality on protected, uninhabited reefs.
• Regional context of environmental drivers: Positive associations with phytoplankton (chlorophyll-a) in Hawai'i reflect natural productivity gradients and may not hold where chlorophyll-a indicates anthropogenic eutrophication (for example, some Great Barrier Reef settings).
• Observational inference and collinearity control: Although models accounted for multicollinearity and used robust frameworks (PERMANOVA, GAMM, ordinal logistic regression), the study remains observational; causality cannot be definitively established.
• Spatial scope: Findings are from Hawaiian reefs, which are generally low-wave exposure and have specific fisheries and wastewater profiles; effects may differ in high-energy or differently managed systems.