Global coral reef ecosystems exhibit declining calcification and increasing primary productivity

Coral reefs face significant threats from global climate change, including ocean acidification and increasing thermal stress. These stressors compromise reef resilience and functionality. Net ecosystem calcification (Gnet) and organic productivity (Pnet) are key indicators of reef health and longevity. Gnet reflects the balance between calcification and dissolution, while Pnet indicates the balance between photosynthesis and respiration. Pre-1975, Pnet was considered near-zero or slightly positive; however, recent observations show Pnet increasing relative to Gnet, suggesting a shift from coral- to algae-dominated ecosystems. Hydrochemical methods offer spatially and temporally specific ecosystem-scale assessments of Gnet and Pnet, integrating processes without species-level resolution, unlike community census techniques. While individual studies have identified potential influences on Gnet (depth, temperature, coral cover), a global-scale consensus on critical drivers remains elusive. Mesocosm experiments offer insights into specific factors, but may not fully represent the complexity of in situ reef systems. This meta-analysis aims to identify key global-scale drivers of coral ecosystem calcification, using hydrochemical data from numerous case studies, and to predict future trends in Gnet.

The meta-analysis included 116 unique diel-integrated calcification rates from 53 publications, encompassing 36 coral reef sites across 11 countries. Australian reefs, particularly the Great Barrier Reef, were heavily represented. The Shiraho Reef in Japan was the most extensively studied site. Mid-latitude reefs were disproportionately represented compared to their global distribution, with equatorial reefs underrepresented. Despite the availability of hydrochemical methods for over 50 years, a significant portion of the studies were conducted within the past decade. Gnet ranged from −90 to 667 mmol m−2 d−1, averaging 124.1 ± 106.6 mmol m−2 d−1, while Pnet showed a wider range and averaged 65.1 ± 254.3 mmol m−2 d−1. A substantial number of studies reported net respiratory ecosystems (negative Pnet), and some reported net ecosystem dissolution (negative Gnet). Around 25% of the studied reefs were reported as degraded due to various stressors.

The meta-analysis used data from 116 in situ hydrochemical case studies quantifying ecosystem-scale coral reef production and calcification. Data were compiled from publications found via Google Scholar using keywords like 'coral reef', 'metabolism', 'calcification', and 'productivity'. Studies were excluded if diel-integrated Gnet rates were unavailable, or if major external carbon sources (river or groundwater) were reported. Qualitative information (year, location, duration, wave action, season, methodology, nighttime productivity, reef state) and quantitative data (Gnet, Pnet, depth, calcifier cover, temperature, aragonite saturation state (Ωar), nutrients) were extracted. Linear mixed-effects regression models (LMER) were used to identify key drivers of Gnet, with year, study methodology, latitude, season, reef state, Pnet, Ωar, temperature, calcifier cover, wave action, and depth as potential predictors. A random intercept term for location was included to account for site-specific variability. A backward-selection process, using Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC), was employed to refine the model. The model's assumptions were checked, and outliers were identified and removed as needed. Additional LMER models were used to investigate seasonal Gnet and temporal trends in Gnet and Pnet at repeatedly studied sites.

The LMER analysis revealed that depth and calcifier cover were the most significant predictors of Gnet. Wave action showed some evidence of influencing Gnet. Seasonality also influenced Gnet, with warmer seasons (summer-autumn) having higher average Gnet, especially at mid-latitude sites. The relationship between temperature and Gnet was scale-dependent; while no overall global influence of temperature was detected, individual reefs showed higher Gnet at higher temperatures and in warmer seasons. Notably, the model did not identify a significant influence of aragonite saturation state (Ωar) on Gnet. Analysis of repeatedly studied reefs (seven sites) revealed a declining calcification rate of 4.3 ± 1.9% per year since the 1970s, accompanied by an increase in organic productivity of 3.0 ± 0.8 mmol m−2 d−1 yr−1. If this decline continues, global net-zero calcification may occur around 2054. Nighttime dissolution was significantly correlated with lower diel-integrated Gnet rates.

The observed decline in Gnet and increase in Pnet at a global scale supports theories of coral reef phase shifts, where coral cover loss is replaced by algal dominance. This shift reduces reef resilience, biodiversity, and ecosystem services. While loss of coral cover is a major contributor to declining Gnet, sublethal stress effects on coral calcification also play a role. The rate of Gnet decline may be underestimated, as it doesn't fully account for the potential for rapid declines associated with mass bleaching events, which can significantly reduce ecosystem calcification. Calcifier cover and depth are critical predictors of Gnet globally, reflecting the importance of benthic community structure and hydrodynamic conditions. The scale-dependent effect of temperature highlights the complex interplay between temperature-driven enhancements in calcification and the detrimental impacts of bleaching at higher temperatures. The underrepresentation of equatorial reefs and limitations in data reporting highlight the need for more comprehensive data collection and consistent reporting of parameters.

This meta-analysis reveals a significant long-term decline in global coral reef calcification and an increase in organic productivity. Depth and calcifier cover are the most important predictors of Gnet, while temperature plays a scale-dependent role. Aragonite saturation state did not emerge as a major driver of global Gnet. The continuing decline in Gnet, projected to reach net-zero around 2054, emphasizes the urgent need for effective conservation and management strategies to mitigate climate change impacts on coral reefs. Future research should focus on improving data collection and reporting, including nighttime calcification rates and data from underrepresented regions.

The study's limitations include the uneven spatial distribution of data, particularly the underrepresentation of equatorial reefs and the low number of repeatedly surveyed sites. Data limitations also hindered the inclusion of certain parameters in the statistical modeling, particularly nutrients. The reliance on previously published data introduces potential biases based on variations in sampling methods and reporting standards. The complexity of coral reef ecosystems also poses challenges in fully understanding the interplay of various factors influencing Gnet and Pnet.