Evaluating the protection status and exposure to warming of Caribbean reefs with high functional potential

Abstract

Over the past decades, climate change has emerged as a major threat to global biodiversity, negatively affecting the integrity and functioning of ecosystems and the benefits they provide to people. To mitigate these impacts, it is essential to identify climate refugia that support the persistence of the structure and function of reef ecosystems. Although most refugia analyses focus on thermal exposure to define priority areas, they often overlook other critical ecological and physical aspects of the ecosystems. We conducted the first assessment of climate refugia based on thermal exposure of a key attribute of coral reefs: their physical functionality. We used a previously published estimation of the potential for physical functionality of coral reefs in the Greater Caribbean. We used these estimates, thermal anomaly projections for 2049, and 10 Earth system models under two shared socioeconomic pathways (SSPs 3–7.0 and 5–8.5) to define priority reefs as those exhibiting high potential for physical functionality and having low exposure to climate change anomalies (refugia criteria). Forty‐six percent of the priority reefs meeting the refugia criteria were unprotected. Several Caribbean countries had high potential to increase protection for functionally important reefs, including the Bahamas, Puerto Rico, the Turks and Caicos Islands, Panama, Cuba, Belize, Haiti, Honduras, Guatemala, and Jamaica. These findings provide a valuable tool for integrating the functional aspects of ecosystems when identifying priority conservation areas in the face of climate change.

INTRODUCTION

Coral reefs support extraordinary biodiversity and provide essential provisioning, regulating, and cultural services, sustaining the livelihoods of millions of people in coastal communities (Woodhead et al., 2019). However, these ecosystems are experiencing unprecedented and rapid changes due to ocean warming, which is increasing the frequency and severity of coral bleaching events, a signal of stress in corals that can lead to sublethal effects, including declines in growth and reproduction (Briggs et al., 2024; Hughes et al., 2017; Hughes, Anderson, et al., 2018). As warming trends exceed the physiological capacity of corals to tolerate them, bleaching events more frequently result in widespread coral mortality, reduced larval recruitment, and decreased rates of coral reef carbonate production in subsequent years (Eakin et al., 2010; Hughes et al., 2019; Perry & Morgan, 2017). In addition to chronic and acute thermal stress, coral reefs face other climate change–related stressors, such as increased storm frequency and intensity, ocean acidification, sea‐level rise, and deoxygenation, all of which exacerbate reef degradation (Emanuel, 2021; Hoegh‐Guldberg et al., 2007; Hughes et al., 2020; Knowlton, 2001).

Climate change, combined with decades of local anthropogenic pressures, is causing declines in coral cover and limiting recovery due to the increasing frequency of stressful conditions. As corals differ in susceptibility to stressors, many reefs experience shifts toward coral assemblages dominated by small, encrusting species with rapid population turnover rates (e.g., Perry et al., 2015). These assemblages often lack the critical functional traits necessary to maintain the physical and ecological complexity of coral reefs (Bellwood et al., 2019; Hughes, Kerry, et al., 2018; Okazaki et al., 2017; Reverter et al., 2022). Furthermore, even when corals recover, it is often driven by species that do not build frameworks, leading to a disparity between coral cover recovery and the ecological functionality of communities (Alvarez‐Filip et al., 2013; González‐Barrios et al., 2021). The net outcome of these ecological shifts is the homogenization of coral communities and a reduction in physical functionality, which undermines the ecological stability of the system by reducing reef‐building capacity and habitat provisioning and affects essential ecosystem services, like coastal protection and fisheries productivity (Estrada‐Saldívar et al., 2019; González‐Barrios & Álvarez‐Filip, 2018; Hughes, Kerry, et al., 2018; Perry et al., 2015; Richardson et al., 2017).

Protecting reefs exposed to low thermal stress through targeted management can help limit additional local stressors on climate refugia, thereby supporting coral persistence and recovery as global temperatures continue to rise (Beyer et al., 2018; West & Salm, 2003). Prioritizing marine protected areas (MPAs) in regions where projected warming is lower has been proposed as a long‐term buffer against environmental changes, creating refugia that remain suitable even as surrounding areas become less so (Bruno et al., 2018; Kavousi & Keppel, 2018). Focusing on reefs that are less exposed to rapid warming can maximize conservation efforts because they are more likely to retain their ecological integrity and continue providing essential ecosystem services (Hoegh‐Guldberg et al., 2023; Mumby et al., 2011). In contrast, reefs in highly exposed areas are at greater risk of irreversible degradation and loss of functionality, even with significant intervention (Hughes et al., 2017, 2023). However, current conservation planning strategies often rely exclusively on taxonomic diversity as a standard indicator in spatial prioritization, neglecting other crucial dimensions of biodiversity, such as functional or phylogenetic diversity (Devictor et al., 2010).

One approach to integrating climate change into coral reef conservation involves identifying high‐priority sites for protection by considering both acute and chronic thermal stress and the potential response of corals through acclimatization (Mumby et al., 2011). Another strategy is to identify priority sites by analyzing past and future thermal exposure in combination with environmental characteristics associated with reef persistence (Chollett et al., 2022). However, to date, most studies have primarily focused on coral bleaching, associated mortality, and coral cover loss (e.g., Babcock et al., 2021), while overlooking the effects of thermal stress on other ecological attributes. For example, the transformation of physical functionality (i.e., the capacity to create complex three‐dimensional structures) has mostly been considered only in the context of its impacts (e.g., Hughes, Kerry, et al., 2018) and has not yet been incorporated into the prioritization of conservation efforts.

As ecosystem engineers, corals act as the primary builders of the three‐dimensional structures that support and regulate ecosystem functioning and services (Graham & Nash, 2013; Mouillot et al., 2011). Crustose coralline algae are essential for binding and cementing the reef matrix, enhancing reef structural stability and complexity (Cornwall et al., 2023; Weiss & Martindale, 2017). In this context, understanding and predicting changes in the structural complexity of reefs through space and time is essential for preserving their physical functionality. The reef functional index (González‐Barrios & Álvarez‐Filip, 2018) was created based on species‐specific coral morphofunctional traits associated with their capacity to create three‐dimensional structures, such as calcification rates and morphological complexity, along with coral cover data. This index has proven to be a good approximation for comparing Mesoamerican reefs because reefs with different species configurations showed functional variation depending on the dominance patterns of key functional species. This approach offers a significant opportunity to explore changes in reef physical functionality and community dynamics, especially in the context of climate change and ecological degradation.

In this study, we utilized the results from Melo‐Merino et al. (2022), who employed the reef functional index to estimate the functional potential of reefs in the Greater Caribbean region. We assessed the protection status of reefs with high potential for physical functionality and evaluated their exposure to future temperature anomalies under climate change scenarios derived from the Coupled Model Intercomparison Project Phase 6 (CMIP6), specifically the shared socioeconomic pathways (SSP) 3–7.0 and 5–8.5 projected for the year 2049. Our goal was to identify reefs with high functional potential and low exposure to rising temperatures, highlighting them as potential climate refugia and priority sites for near‐future conservation efforts. This approach might help ensure that reefs with strong functional potential, particularly those capable of maintaining or rebuilding structural complexity, are effectively managed within MPA networks.

METHODS

Reef functional potential

To evaluate the physical functionality of Caribbean reefs, we used the reef functional potential (RFP) estimates provided by Melo‐Merino et al. (2022). The RFP represents a metric describing the potential of a reef to create complex three‐dimensional structures, accounting for environmental variability and coral species’ environmental preferences and morphofunctional trait metrics. The environmental variability and species preferences are captured through ecological niche models, generated with machine learning techniques for 49 reef‐building coral species based on occurrence records and environmental predictors. The predictors were chosen based on their contribution to limiting or shaping the distribution of shallow‐water corals, including first‐order determinants such as temperature, salinity, light availability, aragonite saturation state, and nutrients (Kleypas et al., 1999). The models produce habitat suitability maps for individual species, where higher values indicated more optimal conditions for the species (Phillips et al., 2006).

Corals exhibit a wide range of growth rates and morphologies that determine their specific roles in defining reef structure, functioning, and dynamics (González‐Barrios et al., 2021; Zawada et al., 2019). The species‐specific contribution of corals to the formation of structurally complex reefs is represented by a functional coefficient. This coefficient is built based on morphofunctional traits related to calcium carbonate precipitation and the colony morphological complexity (González‐Barrios & Álvarez‐Filip, 2018).

To generate the spatially explicit RFP estimation, the habitat suitability maps of each coral species were multiplied by their corresponding functional coefficients. Subsequently, the resulting 49 raster products were stacked based on a raster sum (Melo‐Merino et al., 2022). Reefs with high RFP are likely to be dominated by species with high calcification rates and complex morphologies, which display greater three‐dimensionality and higher functionality, enabling them to sustain diverse communities and regulate the dynamics and functioning of associated ecological processes (Graham & Nash, 2013; Pratchett et al., 2008; Richardson et al., 2017).

Protection status of RFP

To analyze the protection status of RFP of the Greater Caribbean coral reefs, we used data from the geospatial database of the Caribbean Marine Protected Area Management Network and Forum (CaMPAM, accessed March 2021) sourced from the Specially Protected Areas and Wildlife protocol of the UNEP Caribbean Environment Program. To focus our study on MPAs safeguarding coral reefs, we conducted the following spatial operations. First, we performed a spatial union between the CaMPAM layer and the RFP layer as a coarse filter. The remaining MPAs were then intersected with the shallow tropical coral reefs layer, created by the World Resources Institute for the Reefs at Risk Revisited project (Burke et al., 2011). For MPAs that intersected the RFP layer but not the reefs layer, we conducted a more detailed review to decide on exclusion. Additionally, we reviewed the MPAs that intersected the reefs layer to exclude those protecting coastal ecosystems (e.g., mangroves) but not coral reefs. When available, we used the metadata from the MPA layer to inform these decisions; otherwise, the polygon was retained.

As a second data source for MPAs, we used the World Database on Protected Areas (WDPA, accessed March 2021; https://www.protectedplanet.net/en/thematic‐areas/wdpa?tab=WDPA). Following the same procedure, we identified MPA polygons not included in CaMPAM layer and created a combined database. Among all the MPAs, those managed only for fisheries, marine mammals and deep coral reefs were excluded from the analysis. In total, our analyses encompassed 562 MPA polygons (Appendix S1). Finally, to identify the protection status of RFP cells, those intersecting the MPA polygons were considered to be protected. Although management effectiveness varied across MPAs, we assumed that all retained areas were effectively managed.

Considering that decision‐making primarily relied on national governments, we conducted the analyses at the country level. To assign each RFP cell to its respective country, we performed a spatial union between the RFP polygons and the exclusive economic zone of each country, sourced from the Marine Regions Organization database 11 (accessed March 2021, https://www.marineregions.org/about.php). Coral reefs in the United States were separated into two groups, Flower Garden Banks and the Florida reefs, because they clearly belong to different ecoregions (Spalding et al., 2007). Similarly, the coral reefs of Mexico were divided into the Gulf of Mexico and Caribbean reefs for the same reason.

We classified countries based on the number of reefs falling within MPA polygons and the relative difference in mean RFP values between protected and unprotected reefs. Four groups were identified: countries where all or nearly all reefs were protected, regardless of their RFP values, categorized as highly protected; countries where the mean RFP was higher in protected reefs than in unprotected reefs, categorized as most complex reefs protected; countries where the mean RFP was higher in unprotected reefs than in protected reefs, categorized as having an opportunity to increase protection of complex reefs; and countries where all or most reefs were unprotected, categorized as needing to implement protection.

RFP exposure to climate change

The exposure of RFP to temperature increase was assessed for two future emission scenarios from CMIP6, under the SSPs. The CMIP6 scenarios were chosen because they were based on updated versions and exhibit higher climate sensitivity compared to the previous CMIP5 scenarios (Zelinka et al., 2020). With this approximation, our projections focus on gradual climate change and therefore did not incorporate short‐term extreme events, such as tropical cyclones or acute marine heatwaves.

Projections of future climate change under the SSPs were based on alternative scenarios relevant to societal concerns, spanning a range of challenges to climate change mitigation and adaptation (O'Neill et al., 2016). These scenarios were combined with the representative concentration pathways (RCPs) to create integrated narratives representing the socioeconomic trends and the trajectories for different concentrations of greenhouse gas emissions indicated by different radiative forcing levels (IPCC, 2023). The chosen future emission scenarios included the high emissions SSP3–7.0 and the very high emissions SSP5–8.5, considering recent studies suggesting that CO₂ emissions were more consistent with upper‐range scenarios (Schwalm et al., 2020). Although upper‐range scenarios might be considered implausible (Hausfather & Peters, 2020), they can aid in informing conservation decision making, especially over shorter time horizons (Schwalm et al., 2020). The choice of these climate change scenarios was further supported by the warming rates observed in the Caribbean, which aligned with those estimated under business‐as‐usual scenarios (Bruno et al., 2018). The SSP3 represented a world with poor progress toward sustainability and high challenges to mitigation and adaptation (Fujimori et al., 2017; O'Neill et al., 2017). On the other hand, the SSP5 scenario was characterized by high socioeconomic challenges to adaptation and a strong dependence on fossil fuels, causing high levels of greenhouse gas emissions (Kriegler et al., 2017; O'Neill et al., 2017).

We obtained future temperature anomalies, defined as the difference between the mean temperature in the future (2020—2049) and that during the historic baseline period (1985—2014), which represented long‐term thermal conditions, for 10 Earth system models (Appendix S2) from the Physical Sciences Laboratory of the National Oceanic and Atmospheric Administration (https://psl.noaa.gov/). We opted for this nearby temporal horizon to reduce model uncertainty. To account for the inherent variability in climate models, we calculated a multimodel mean anomaly, with intermodel variability represented by the standard deviation (Appendix S3).

The spatial resolution of the temperature anomaly layers was 1° × 1° latitude–longitude (∼111 km at the equator). To standardize the spatial resolution with the RFP layer, which had a resolution of ∼9 × 9 km latitude–longitude, and reduce discontinuities in grid boundaries, we downscaled the anomaly projections with an inverse distance weighting interpolation. To compare temperature anomalies across countries, we applied a Kruskal–Wallis one‐way analysis of variance.

We evaluated the exposure of RFP cells to future ocean warming by intersecting the temperature anomalies with the RFP layer, assigning the corresponding temperature anomaly to each RFP cell. To spatially represent the exposure level, we created a bivariate choropleth map based on a discrete classification of RFP and temperature anomalies values into terciles. Each tercile corresponded to a category of exposure: low, medium, and high. Cells identified as priority for future protection were those with the highest RFP and the lowest exposure to temperature anomalies. Herein, we refer to each of the analyzed cells with a reef component as reefs.

RESULTS

Approximately half of the reefs in the Greater Caribbean were protected (48%), but there was little difference in RFP between protected and unprotected reefs (Figure 1). Protected reefs with medium to high RFP values were in the western, southwestern, and Floridian ecoregions. Several unprotected reefs with high RFP values were in the Greater Antilles and Bahamian ecoregions (Figure 1).

FIGURE 1.

For Caribbean coral reefs, (a) spatial distribution of reefs inside and outside marine protected areas and their corresponding reef functional potential (RFP) values and (b) the distribution of reef functional potential values in protected and unprotected reefs (horizontal lines, median; box ends, interquartile range; whiskers, value range). The RFP values derived from Melo‐Merino et al. (2022) are arranged by tercile.

A high proportion of countries were identified as having effective protection schemes because all their reefs were in MPAs or reefs with the highest RFP values were inside MPAs, whereas those with lower values remained unprotected (Figure 2). However, we also observed countries where reefs with the highest RFP values were outside MPAs. In this group, RFP values varied substantially, but the lack of protection for highly functional reefs was particularly evident in Belize, Honduras, and Puerto Rico, despite overall high levels of reef protection. A smaller proportion of countries were classified as needing to implement protection measures, as all or nearly all their reefs remain unprotected (Figure 2).

FIGURE 2.

Protection status and functional potential (RFP) of coral reefs across the Greater Caribbean (reef categories, protection status and the relative difference in mean RFP between protected and unprotected reef cells; highly protected, all or almost all reefs protected; most complex protected, higher RFP in protected reefs; opportunity for protection, higher RFP in unprotected reefs; need protection, all or almost all reefs unprotected; sites in groups are ordered by decreasing mean RFP in the dominant condition; numbers at left, count of reef cells in each country or territory).

The reefs in the Greater Caribbean displayed large spatial variability in heat exposure projected to 2049. The expected thermal anomalies increased from SSP3–7.0 to SSP5–8.5. In the SSP3–7.0 scenario, temperature anomalies ranged from 0.81°C to 0.98°C (mean [SD] = 0.88 [0.03]), and in the SSP5–8.5 scenario, they ranged from 0.89°C to 1.16°C (mean = 1.00 [0.03]). Both scenarios projected higher temperature change at higher latitudes in the northern Gulf of Mexico and Bermuda ecoregions. Reefs with the greatest exposure to temperature increase were in the flower garden banks (USA), potentially experiencing up to a 1.12°C increase in the SSP5–8.5 scenario (Figure 3). In contrast, countries exposed to lowest future temperature anomalies were found in the eastern Caribbean in Trinidad and Tobago and Grenada (Figure 3). The Kruskal–Wallis test showed significant differences among countries in both scenarios (Kruskal–Wallis χ² = 2449.80, df = 42, p < 0.05). Moreover, the greatest spatial variability was most evident in Mexico (Gulf of Mexico), Colombia, and Cuba. For some countries, this variability increased from SSP3–7.0 to SSP5–8.5 and was most evident in Mexico (Gulf of Mexico) and Bahamas. For others, such as Colombia, Guadeloupe, and Mexico (Caribbean), the spatial variability in temperature anomalies decreased, meaning that the reef exposure to thermal anomalies may become more homogeneous (Figure 3).

FIGURE 3.

Temperature anomalies per country in the Greater Caribbean under two future climate change shared socioeconomic pathways (SSP3–7.0 and SSP5–8.5) projected to 2049 (horizontal lines, median; bar ends, interquartile range; whiskers, value range; dots, outliers; numbers in parentheses, number of reef cells; bold, countries with the highest spatial variability in thermal anomalies).

The assessment of RFP exposure to temperature anomalies under the SSP3–7.0 scenario revealed that most reef cells with high RFP were exposed to the lowest category of temperature increase (<0.92°C) (Figure 4). The magnitude of exposure increased under SSP5–8.5, where most reefs with high and medium RFP values (40% and 38%, respectively) fell into the intermediate temperature anomaly category. Under this most severe scenario, practically no reefs with high RFP remained in the lowest exposure class (0.09%), all corresponding to Venezuela, and only a small proportion (0.45%) were projected to experience the highest warming category (>1.04°C), all located in the south coast of Cuba (Figure 5). Countries with a high percentage of reefs exposed to elevated temperature anomalies included the United States (Flower Garden Banks) and Bermuda, where 100% of their reefs were exposed to the most severe warming category under the SSP5–8.5 scenario, although they were also characterized by low RFP values (Figure 5).

FIGURE 4.

Spatial representation of the exposure of reef functional potential (RFP) to future thermal anomalies under climate change shared socioeconomic pathways (a) SSP3–7.0 and (b) SSP5–8.5 projected to 2049 (maps), percentage of reefs, and RFP values relative to the temperature anomaly according to the 1985–2014 historic baseline (graphs).

FIGURE 5.

Country‐level exposure of reef functional potential (RFP) to future thermal anomalies under climate change shared socioeconomic pathways SSP3–7.0 and SSP5–8.5 projected to 2049 (numbers, percentage of reef cells in each quadrant; bold, reefs that are a priority for conservation with high RFP and low exposure to temperature anomalies).

Under both climate scenarios, 40% of reef cells in the Greater Caribbean met the criteria for being priority reefs, having the highest RFP and the lowest exposure to temperature anomalies. However, the definition of lowest exposure was scenario specific. In SSP3–7.0, these reefs fell within the lowest exposure category. In contrast, under SSP5–8.5, only 0.09% of reefs remained in that category. Of the total identified priority reefs, only 54% were covered by an MPA under the current scheme of the Caribbean MPA network. In both scenarios, the countries with the highest proportion of priority reefs were Belize and Honduras (≥80%) (Figure 6). Our assessment also identified countries with the greatest opportunity to increase reef protection under climate change scenarios based on the percentage of unprotected priority reefs (Figure 6).

FIGURE 6.

Proportion of reefs classified as a priority for future protection in the 10 countries with the highest number of unprotected priority reefs (numbers, percentage of priority reefs estimated under the climate change shared socioeconomic pathway SSP3–7.0 projected to 2049; countries ordered by decreasing number of unprotected priority reefs).

Finally, to illustrate the identification of priority reefs, we used Cuba because it was the only country with reefs exposed to the highest thermal anomalies category. We observed that most priority reefs were in the southern coast of the island. The number of priority reefs decreased from 56% in the SSP3–7.0 scenario to 54% in the SSP5–8.5 scenario due to the impact of elevated temperature anomalies in the Ana Maria Gulf. From the identified reefs, only 30% were protected by the MPA network. Priority reefs lacking MPA coverage were mainly located in the Guacanayabo Gulf, on the south coast, and in the Guanahacabibes Gulf, on the west coast (Figure 7).

FIGURE 7.

Reefs identified as a priority for future protection in Cuba under climate change shared socioeconomic pathways SSP3–7.0 and SSP5–8.5 projected to 2049.

DISCUSSION

Over the past decades, ocean warming caused by climate change has profoundly affected the ecology and function of coral reefs (Hughes, Kerry, et al., 2018). This highlights the urgency of enhancing conservation measures to minimize associated losses and maximize protection. Additionally, there is a critical need to establish conservation goals aimed at maintaining the ability of reefs to sustain high biological diversity and ecosystem services, especially in the face of increasing climatic perturbations. In this study, we integrated information on the capacity of reefs to build complex three‐dimensional structures with projections of expected thermal anomalies to offer decision support for identifying reef systems that might be conservation priorities in the present and under future climate change scenarios. These sites may serve as climate refugia for coral reefs with a high potential for physical functionality (i.e., the ability to build complex three‐dimensional structures), making them less susceptible to the effects of rising temperatures. Our findings revealed that almost half of the reefs meeting the criteria to be considered climate refugia with high potential for physical functionality remain unprotected.

With the increasing pressures on coral reefs, MPAs have become extremely valuable conservation tools. However, to be effective in the long term, they must account for the imminent effects of climate change (Chollett et al., 2022). In this regard, the identification, protection, and management of climate refugia have gained prominence as adaptive measures to support the persistence of species, communities, ecosystems, and sociocultural values (Morelli et al., 2020). Nevertheless, recent studies emphasize the need to consider factors beyond exposure in refugia analyses, recommending the inclusion of ecological complexity, such as species traits and the physical characteristics of ecosystems (Miatta et al., 2021; Michalak et al., 2020; Morelli et al., 2020).

Our findings identified reef areas with a high potential for physical functionality that currently lie outside MPAs, highlighting the importance of incorporating functional aspects into the spatial prioritization of Caribbean reefs. Similar spatial mismatches have been observed in other taxonomic groups when functional traits are considered in the analysis. For instance, an assessment of reef fish biodiversity distribution, comparing taxonomic and trait metrics, revealed new hotspots of functional diversity in areas not considered for conservation (Stuart‐Smith et al., 2013). Although we recognize that no single parameter can fully capture the functioning of an entire ecosystem (Rees et al., 2012), our integration of reefs’ physical functionality estimates with climate change exposure represents a promising approach for preserving one of the most important attributes of coral reefs: their complex three‐dimensional structure.

In this setting, identifying the places where conservation efforts should focus is crucial to providing adequate protection and optimizing financial expenditure. Our results suggest there are abundant opportunities to strengthen protection for structurally complex reefs within existing MPAs and to pinpoint sites for establishing new MPAs in countries with low protection coverage (Figure 6). However, it must be acknowledged that biodiversity conservation is context dependent. A country's ability to study and manage its biodiversity is shaped by the interplay of scientific capacity, wealth, political will, and governance (Lira‐Noriega & Soberón, 2015). The Caribbean, a complex geopolitical system comprising developing countries and overseas territories under the jurisdiction of wealthy developed nations (Vallès et al., 2021), has experienced varying levels of investment in biodiversity conservation. Nevertheless, wealth alone does not guarantee a country's ability to effectively conserve biodiversity. We found some developing countries, like Honduras, have robust protection schemes for potential climate refugia of functionally important reefs, where 80% of the reefs were identified as a priority for protection, with 66% already within MPA (Figure 6).

In the case of Cuba, the priority areas aligned with those identified in other studies that consider climate change effects and other stressors (Chollett et al., 2022). Additionally, the same reef areas were identified as having the potential to support reefs with high coral diversity when considering diversity models and environmental characteristics (Selmoni et al., 2023). These consistencies underscore the utility of our approach, demonstrating its potential to strengthen decision‐making. The areas identified as important for conservation based on physical functionality also supported high biological diversity and larval connectivity and, aside from temperature increases, were less exposed to other climate change–related impacts, such as hurricanes (Chollett et al., 2022; Selmoni et al., 2023).

As observed in various studies of climate refugia for coral reefs, when temperatures rise to higher values, the number and extent of coral refugia decrease, or even disappear. A consistent finding in our study was the reduction in the percentage of priority reefs with refugia potential in Cuba, which decreased from 56% to 54% in the high emission scenario. Dixon et al. (2022) found that under a 1.5°C warming scenario, only 0.2% of climate refugia would remain, whereas under a 2.0°C warming scenario, no climate refugia for coral reefs would persist. We defined priority reefs as those experiencing the lowest exposure to temperature anomalies, without applying an explicit thermal threshold. If, instead, we had set a threshold based on the minimum projected thermal anomalies, excluding all reefs exceeding that value, the proportion of reefs considered climate refuges would be reduced to 0.09% throughout the Caribbean. Additionally, Caribbean reefs are more likely to cross their critical coral reef bleaching thresholds by midcentury (Lawman et al., 2022) or even sooner, considering the catastrophic bleaching event of 2023 caused by unprecedented heat stress conditions in the Caribbean basin (Hoegh‐Guldberg et al., 2023).

With our approach, we identified priority sites based solely on heat exposure; however, a more comprehensive view of vulnerability to climate change should be considered when conducting climate refugia analysis. Differences in how species respond to climate change (sensitivity) and their ability to adapt will influence the location, duration, and success of climate refugia (Chollett et al., 2022; Morelli et al., 2020). Corals can adapt to temperature variation by switching symbionts, changing protein expression, or acclimatizing to warmer temperature anomalies (Boulotte et al., 2016; Palumbi et al., 2014; Scharfenstein et al., 2022; Sully et al., 2019). Although large‐scale information on these adaptive processes is practically nonexistent (but see McClanahan et al., 2020), it should be incorporated into conservation strategies as it becomes available.

The multimodel mean anomalies we used project chronic thermal exposure; however, they do not capture the potential effects of climate extremes (i.e., acute thermal exposure), which can have major ecological consequences for coral reefs. In recent years, extreme climatic events linked to ocean warming have caused profound changes in coral reefs. Widespread coral bleaching events have led to significant coral mortality on many reefs, accelerating the degradation of the reef physical structure and altering community composition (Eakin et al., 2019; Hughes, Kerry, et al., 2018). The differential susceptibility of coral species and their varying functional traits are driving these changes. Unfortunately, the species most affected are often those that play a critical role as reef builders. For instance, the 2016 bleaching event in the Great Barrier Reef caused population declines of up to 75% in fast growing and structurally complex staghorn and tabular corals, significantly altering both the structure and functioning of many reefs due to shifts in the functional trait composition (Hughes, Kerry, et al., 2018).

Projections from model simulations, based on changes in species abundances under climate change scenarios, suggest a loss or collapse of functions due to declining species richness and its associated functional traits in Australia (Quigley & Baird, 2024). As the return time of marine heatwaves is expected to shorten and temperature anomalies are expected to become more extreme, coral recovery capacity will likely diminish, leaving reefs more vulnerable to thermal stress, with even the most resistant corals potentially succumbing to future warming events, thereby leading to a gradual decline in reef functions over time (Bellwood et al., 2019; Mellin et al., 2024; Quigley & Baird, 2024). Moreover, the stability of the reef matrix may be further compromised as crustose coralline algae, which are key calcifying organisms supporting the reef framework, are also susceptible to ocean warming and, particularly, to ocean acidification (Cornwall et al., 2022; Krieger et al., 2023).

We consider our approach a valuable tool for marine conservation and the protection of coral reefs because it incorporates the potential to form complex three‐dimensional structures and considers widely used metrics, such as coral cover, in the conceptual framework of the decision‐making process. Yet, it is essential to acknowledge some limitations of our approach that managers implementing conservation actions or researchers interested in further developing should consider.

A central consideration is the variability in trait expression resulting from phenotypic plasticity because traits may differ across environmental contexts. For example, coral morphological traits may vary in response to differences in light availability and hydrodynamic forces. Such variation can lead to differences in the physical functionality of reef assemblages, even when the same species are present. Additionally, the spatial resolution of the study limits the ability to capture fine‐scale processes that influence species assemblages and trait composition, potentially affecting the accuracy of functionality estimates. In this regard, a limitation of this study lies in the assumption that all modeled species are present in all the reef cells and contribute more highly to physical functionality in areas of high habitat suitability. However, ecological factors, such as dispersal limitations, local extinctions, and small‐scale community dynamics, may prevent certain species from occurring in some locations. For example, some framework‐building species like Acropora palmata (Lamarck, 1816) or Dendrogyra cylindrus (Ehrenberg, 1834), which have experienced severe population declines due to diseases, are no longer present in areas where they once were or might present severely reduced populations, negatively affecting reef physical functionality (Alvarez‐Filip et al., 2022; Cramer et al., 2020). Another opportunity to further develop our method lies in the fact that we considered only primary reef builders. As a result, we may have underestimated the net local complexity of the system. Other taxonomic groups, such as crustose coralline algae, which also contribute to carbonate production, and soft corals and sponges, which play an important role in enhancing reef complexity and increasing habitat diversity, may also help shape the ecological space used by other organisms and influence the reef‐building processes (Coppock et al., 2024; Cornwall et al., 2023; Epstein & Kingsford, 2019).

We identified a significant opportunity to enhance protection for reefs with high potential for physical functionality in the Greater Caribbean, which might serve as climate refugia. However, to effectively achieve conservation goals, local authorities and decision makers must also consider the effects of local stressors, such as coastal development and marine pollution (Kavousi & Keppel, 2018). For example, a study evaluating the combined effects of ocean warming, acidification, and local eutrophication on habitat availability for coral reefs showed that, when accounting for these three stressors, only 6% of the world's coral reefs may act as temporary refugia (Guan et al., 2020). Furthermore, our results may be conservative, considering that some MPAs lack proper reinforcement or face financial resource limitations to operate, potentially providing a false sense of protection. There is no doubt that climate change is already transforming and will continue to threaten the structural integrity of coral reefs and the essential services they provide. Although defining climate refugia is a complex task, it is a critical step toward safeguarding the ecological and societal value of these ecosystems. Our results highlight priority areas for protection by integrating key aspects of ecosystem functioning into conservation planning. Incorporating additional climate‐related stressors, such as storm exposure, ocean acidification, and deoxygenation, in future analyses will enhance the ability to evaluate, delineate, and maintain climate refugia more effectively.

Supporting information

Supporting information

Supporting information

Supporting information

Melo‐Merino, S. M. , Lira‐Noriega, A. , Martínez‐Meyer, E. , Reyes‐Bonilla, H. , & Álvarez‐Filip, L. (2026). Evaluating the protection status and exposure to warming of Caribbean reefs with high functional potential. Conservation Biology, 40, e70302. 10.1111/cobi.70302