A Caribbean bright spot coral reef exceeds thermal limits during the fourth global coral bleaching event

Marine heatwaves are increasing in both frequency and intensity, exposing coral reefs to prolonged periods of thermal stress and driving recurrent mass coral bleaching. Nevertheless, coral decline remains heterogeneous, with some reefs maintaining unexpectedly high coral cover despite exposure to thermal and other environmental disturbances. This study analysed benthic community change across Banco Capiro in Tela Bay and the south-side reefs of Utila, Honduras, from 2013 to 2025, spanning the 3rd and 4th Global Coral Bleaching Events. Banco Capiro represents a moderately turbid, Caribbean bright spot for live hard coral cover, whereas Utila represents a more typical contemporary Caribbean reef system. Long-term benthic monitoring data, NOAA Coral Reef Watch thermal stress metrics, Generalised Additive Mixed Models and multivariate community analyses were used to quantify changes in benthic cover and community composition. Banco Capiro experienced repeated annual thermal stress, with Degree Heating Week values reaching 12 °C-weeks or above in 2017, 2019, 2020 and 2021, and nearing 20 °C-weeks in 2023 and 2024. Before 2023, live hard coral cover was high at Banco Capiro, averaging 45 ± 3% (mean ± SE), but declined to 4 ± 1% (mean ± SE) following the 4th Global Coral Bleaching Event. Macroalgal cover increased from 26 ± 4% (mean ± SE) to 44 ± 3% (mean ± SE), while turf algae increased from 6 ± 2% (mean ± SE) to 23 ± 8% (mean ± SE). Utila also experienced further coral loss after 2023, despite entering the event from an already lower-coral state. These findings indicate that previous persistence does not guarantee survival under more extreme disturbance.

Caribbean coral reef systems have experienced widespread ecological decline due to human activities over hundreds of years (Jackson, 1997), with both the rate of decline and its systematic documentation increasing substantially from the 1970s onwards (Bellwood et al., 2004; Gardner et al., 2003). Across the Caribbean, interacting local pressures including fishing, coastal development and declining water quality have contributed to losses of coral cover and changes in reef community structure (Cramer et al., 2020; Jackson et al., 2014). These pressures are also evident throughout the Mesoamerican Reef (MAR), which extends through Mexico, Belize, Guatemala and Honduras and encompasses the reefs of the Honduran Bay Islands and northern coast (Gischler et al., 2023; Berger et al., 2022; Randazzo-Eisemann et al., 2021).

The late twentieth-century decline of structurally complex Acropora species was exacerbated by white band disease, from which populations have not recovered to their historical abundance (Cramer et al., 2020; Aronson and Precht, 2001). The Caribbean-wide mass mortality of the keystone herbivore Diadema antillarum in 1983–1984 further reduced grazing pressure on reef communities, alongside declines in herbivorous fishes associated with overfishing (Bodmer et al., 2015; Milne et al., 2022). More recently, stony coral tissue loss disease (SCTLD) has caused additional disturbance across the MAR. First detected in Honduras on Roatán in September 2020, SCTLD subsequently spread rapidly through the Bay Islands, including Utila, affecting at least 28 scleractinian species by late 2021 (Lee Hing et al., 2022; Truc et al., 2023).

Anthropogenically-driven climate change has seen a global increase in sea surface temperatures (SST) (Cheng et al., 2022; IPCC, 2023). As a result, marine heatwaves are increasing in both frequency and intensity (Frölicher et al., 2018; Oliver et al., 2021), exposing coral reefs to prolonged periods of thermal stress. Elevated temperatures can disrupt the symbiotic relationship between corals and their photosynthetic algal symbionts, resulting in coral bleaching and, when heat exposure is prolonged or severe, partial or whole-colony mortality (Hoegh-Guldberg, 1999). Such thermal stress has repeatedly affected the MAR, with widescale bleaching events documented in 1995, 1998, 2005 and 2010, followed by further bleaching during 2015-2017 (Eakin et al., 2010, Eakin et al., 2019; Spady et al., 2026; Birkart and Álvarez-Filip, 2025). Regional monitoring during this latter period recorded bleaching in the Honduran mainland and Bay Islands during the 3 rd Global Coral Bleaching Event (Muñiz-Castillo et al., 2024; Green et al., 2026).

This stress is commonly quantified using Degree Heating Weeks (DHW), a cumulative measure of heat exposure exceeding the local maximum monthly mean (MMM) over a 12-week period (Liu et al., 2014). As both the frequency and intensity of these thermal anomalies have increased, so too has the occurrence of mass coral bleaching (Sully et al., 2019; Eakin et al., 2019). Consequently, multiple Global Coral Bleaching Events (GCBEs), characterised by widespread, synchronous coral bleaching across large spatial scales, have occurred, causing extensive coral mortality across reef systems (Hughes et al., 2018a; Skirving et al., 2019; Eakin et al., 2026). However, bleaching exposure and ecological decline are spatially heterogeneous, with some reefs maintaining relatively high coral cover despite exposure to thermal and other environmental disturbances (Gintert et al., 2018; Browne and Bauman, 2023).

Reefs that maintain relatively high ecological state values for keystone species (e.g., total cover, abundance, biomass etc.) are often referred to as bright spots (Sully et al., 2022), positive deviants (Cinner, 2025), positive outliers or oases (Guest et al., 2018). Although there is no single accepted definition, these are commonly identified as reefs with ecological conditions above a regional or expected baseline, with some studies using standard deviations or z-scores to quantify positive outliers from expected reef condition (Cinner et al., 2016; Elahi et al., 2022). Despite growing recognition of bright spots, defined here as locations with characteristics that buffer reef communities from climate-related stressors that promote their long term persistence, and their potential importance as climate refugia, there remains limited understanding of the extent to which they can persist under repeated or intensifying thermal stress (Cinner et al., 2016; Fox et al., 2021; Mo et al., 2022; McClanahan et al., 2024).

Coral reef bright spots are not necessarily located in pristine environments (Cinner et al., 2016; Schoepf et al., 2023). Many occur in conditions that are typically considered suboptimal for coral growth, including areas exposed to high turbidity, sedimentation, or other chronic stressors (Kleypas et al., 1999; Morgan et al., 2017; Zweifler et al., 2021). Elevated concentrations of suspended particles in the water column reduce light penetration and can limit coral growth, yet some turbid reefs maintain high percentages of live hard coral cover (Guest et al., 2016a; Morgan et al., 2016; Pizarro et al., 2017). This apparent contradiction has led to increasing interest in turbid reefs as potential climate refugia, defined here as reef environments where local conditions may reduce exposure or sensitivity to climate-driven stressors (Guest et al., 2018; Kavousi and Keppel, 2018; Loiola et al., 2019). Although elevated turbidity is often considered a stressor due to its impacts on light availability and sedimentation (Fabricius, 2005; Tuttle and Donahue, 2020), reduced light penetration may moderate the effects of thermal stress by limiting irradiance during bleaching events (Bahr et al., 2016; Rosedy et al., 2023). Varadero Reef in Cartagena Bay, Colombia, provides a clear example of this paradox, with high coral cover (up to 80%) and diversity persisting despite poor water quality and high sediment loads (Pizarro et al., 2017; Lopez-Londono et al., 2021). Such systems are important because they may reveal ecological or environmental conditions that support coral persistence under disturbance, and potentially under future climate change (Cacciapaglia and van Woesik, 2016; Camp et al., 2018; Browne and Bauman, 2023).

A turbid reef system was identified at Banco Capiro in Tela Bay, Honduras, in 2012, where its high percentage cover of scleractinian coral and low percentage cover of algae (Bodmer et al., 2015; Cramp et al., 2025) highlighted it as a “bright spot” when compared to its regional counterparts, many of which had already experienced substantial coral loss following earlier bleaching events, including the 2nd GCBE (circa. 2010) (Alemu and Clement, 2014; Eakin et al., 2010). A long-term monitoring programme was established by Operation Wallacea (2013-present) in Banco Capiro, Tela Bay, mirroring the programme established in Utila, Bay Islands, Honduras in 2012. The aim of setting up surveys at both locations was to compare a typical Caribbean reef with a turbid bright spot (Cramp et al., 2025).

In the present study, benthic and community structure data were collected annually and examined from these contrasting sites between 2013 and 2025 This dataset provides a unique opportunity to examine long-term trajectories of a turbid reef system through subsequent disturbance events. This manuscript aims to do so by (i) characterising differences in coral, algal, and benthic community composition between turbid and clear-water reef systems, (ii) quantifying long-term changes in these communities through time (2013-2025), and (iii) characterising and comparing reef trajectories before and after major bleaching events across both systems.

Study sites

Study sites in Banco Capiro, Tela Bay and the south side of Utila Island were established by Operation Wallacea in 2013 and 2012 respectively. These locations were selected for their contrasting benthic reef systems, while individual sites within each location were selected based on their accessibility from the respective base of operations. Three sites were surveyed annually in Banco Capiro, and six sites in Utila (Figure 1; Supplementary Table S1). Years 2020, 2021 and 2022 (for Banco Capiro) were not able to b surveyed due to COVID-19 restrictions.

The Banco Capiro reefs in Tela Bay are located approximately 8 km offshore and fall within the “El Refugio de Vida Silvestre Marino de Tela,” a marine park covering 822.6km², established in 2018 (Bodmer et al., 2021). The reef crest starts at 10m depth and is characterised by high live hard coral cover and moderate turbidity (K d (490) = 0.18 ± 0.34 m -1) possibly due to its proximity to the River Ulua Estuary (approx. 30km as the crow flies) (Shi and Wang, 2010; Bodmer et al., 2015). Utila Island is situated 60km offshore and falls within the “Bay Islands National Marine Park,” established in 2010. Utila’s reefs can be considered to represent a “typical” contemporary Caribbean reef, where live hard coral cover is low (~12%) and algal cover is dominant (~ 47%) (Bodmer et al., 2015; Cramp et al., 2025; Green et al., 2026).

Survey methods

At each site, six 50 m transects were surveyed at approximately 15 m depth by SCUBA divers. Transects followed the reef contour along the 15 m depth profile. Three transects were laid consecutively to the east of the mooring buoy, and three to the west, with no gap between adjacent transects. Together, this covered a total linear reef distance of 300 m (Figure 2).

Initial transect placement was haphazard, constrained by the reef structure and the target depth contour. No permanent markers were installed; instead, transects were relocated on subsequent surveys by returning to the same mooring buoy and depth, and re-establishing transects along the same depth and direction of reef as closely as possible. This approach ensured that roughly the same reef area was surveyed annually, allowing for direct spatio-temporal comparisons.

Benthic composition data were collected by filming along transect tapes using GoPro (version variable) cameras. Videos were analysed using the point intercept method. At every 0.25 m interval, the benthic category below the point was recorded. Live hard coral and macroalgal taxa were identified to genus level, while soft corals and sponges were identified based on morphological characteristics. These data were used to assess temporal changes in (i) percentage live hard coral and macroalgal cover, (ii) benthic, live hard coral and macroalgal community composition at both locations, and identify those benthic categories or live hard coral/macroalgal genera with high contributions to the patterns (via SIMPER).

Temperature and degree heating weeks

To assess the influence of thermal stress on percentage cover of live hard coral, macro and turf algae, Sea Surface Temperature (SST) and Degree Heating Week (DHW) data were obtained from NOAA Coral Reef Watch (CRW) daily global 5km satellite dataset on the 4 th of November 2025 (NOAA, 2018).

SST was plotted for both Banco Capiro and Utila using daily mean values. Maximum Monthly Mean (MMM) SST was calculated separately for each location as the maximum of the 12 monthly mean SST climatologies. Monthly climatologies were derived from daily NOAA CRW SST data using a baseline period of 1985–1990 and 1993, consistent with the CRW heritage baseline, with 1991–1992 excluded due to known cold bias associated with the Mt Pinatubo eruption (NOAA, 2018). The bleaching threshold was defined as MMM + 1 °C, following NOAA Coral Reef Watch conventions, to indicate temperatures associated with an elevated risk of coral bleaching (Elneel et al., 2025).

Degree Heating Weeks represent the cumulative thermal stress experienced by corals, calculated as the sum of positive temperature anomalies above the local monthly maximum mean over a rolling 12-week period. The scale ranges from 0-20 °C-weeks, with a risk of bleaching starting at 4 °C-weeks. Above 16 °C-weeks a greater than 50% coral mortality is estimated and greater or equal to 20 °C-weeks, it is expected to see complete mortality (>80% coral) (NOAA, 2018).

Temporal variation of live hard coral, macro and turf algae

To examine how benthic cover changed through time and whether temporal trajectories differed between the inshore reefs of Banco Capiro and the offshore reefs of Utila, percentage cover of live hard coral, macroalgae and turf algae were modelled as a function of ‘year’ and ‘location’. Nonlinear, location specific temporal responses were explicitly allowed so that changes in benthic cover could vary independently between Banco Capiro and Utila rather than assuming a common regional trend.

These patterns were quantified using Generalised Additive Mixed Models (GAMMs) with a binomial error distribution and logit link, including ‘site’ as a random effect to account for repeated measures and spatial structure (Zuur et al., 2010). Location specific smooth functions of ‘year’ were fitted for each benthic group. Model diagnostics indicated good overall fit, and dispersion assessed using Pearson residuals showed no evidence of overdispersion. For turf algae, substantial under dispersion was observed (dispersion parameter = 0.046), indicating very low residual variance following inclusion of site level random effects and nonlinear temporal structure. Values substantially below 1 indicate lower variance than expected under a binomial distribution, likely reflecting the high sampling resolution (n = 201 per transect), which produces very precise estimates of cover, combined with strong temporal structuring in the data (Zuur et al., 2009). As under dispersion leads to conservative standard errors, no correction was applied.

Spatio-temporal variation in community compositions

Spatio-temporal patterns in benthic, live hard coral, and macroalgal community composition were examined to assess differences between locations and to quantify temporal change within locations across the study period. These patterns were visualised using non-metric multidimensional scaling (nMDS) ordinations based on Bray Curtis dissimilarity matrices calculated from square root transformed percentage cover data. Differences in community composition between locations, years, and their interaction were tested using permutational multivariate analysis of variance (PERMANOVA, 999 permutations), with pairwise comparisons conducted where appropriate. Homogeneity of multivariate dispersion was assessed using betadisper (function in R package vegan) to confirm that significant PERMANOVA results reflected differences in community centroids rather than unequal within group variability. The taxa contributing most strongly to observed compositional differences between locations and years within locations were identified using similarity percentage analysis (SIMPER).

All analyses were conducted in R (version 4.1.2) using RStudio (version 2021.09.0 + 351). Generalised Additive Models were fitted using the mgcv package (Wood, 2017), and community composition was analysed using non-metric multidimensional scaling implemented in vegan (Oksanen et al., 2022). Figures were produced using ggplot2 (Wickham, 2016).

Temperature

Mean daily sea surface temperature (SST) exhibited a seasonal cycle at both Banco Capiro and Utila between 2013 and 2025, with lows of 25.5 °C to highs of 31.7 °C (Figure 3). Banco Capiro displayed higher absolute temperatures compared to Utila during the study period. SST exceeded the bleaching threshold (defined as MMM + 1 °C) in 2017, 2023 and 2024 at both locations, and this corresponded with years of Caribbean wide coral bleaching and the 3 rd and 4 th GCBEs (Wicquart et al., 2025).

Banco Capiro experienced repeated annual DHW accumulation events, with 2017, 2019, 2020 and 2021 reaching 12 °C-weeks or above (Figure 3), indicative of conditions associated with reef-wide bleaching and multi-species coral mortality (NOAA, 2018) In 2023 and 2024, Banco Caprio experienced DHW values nearing 20 °C-weeks (Figure 3), which are considered to cause total coral mortality (NOAA, 2018). In contrast, Utila generally experienced lower thermal stress, reaching 8 °C-weeks in 2017 and 12 °C-weeks in 2023 and 2024 (Figure 3).

Temporal trajectory of hard coral and macroalgae

Prior to 2019, percentage live hard coral cover in Banco Capiro remained high, with a mean of 45 ± 3% (mean ± SE) (Table 1). Between 2023 and 2025, following the 4 th GCBE, mean live hard coral cover decreased to 4± 1% (mean ± SE) (Table 1; Figure 4A). During the study period, macroalgal cover increased from a pre-bleaching mean of 26 ± 4% to a post-bleaching mean of 44 ± 3% (mean ± SE) (Table 1; Figure 4B). Turf algae percentage cover increased from a pre-bleaching mean of 6 ± 2% (mean ± SE) to a post-bleaching mean of 23 ± 8% (mean ± SE) (Table 1; Figure 4B).

Mean percentage live hard coral cover in Utila was consistent with Caribbean-wide estimates reported by the GCRMN, which indicate that mean live hard coral cover for the wider Caribbean during the same period was below 20% (Wicquart et al., 2025). Utila was therefore used as a regional benchmark for analyses of live hard coral cover trajectories in Banco Capiro. From 2013 to 2025, live hard coral cover declined from a pre-bleaching mean of 13 ± 1% to a post-bleaching mean of 7 ± 2% (mean ± SE) (Table 1; Figure 4A). Macroalgal percentage cover increased from a pre-bleaching mean of 37 ± 43% to a post-bleaching mean of 46 ± 7% (mean ± SE) (Table 1; Figure 4B). Turf algae also increased in Utila a pre-bleaching mean of 7 ± 3% to a post-bleaching mean of 5 ± 0% (mean ± SE) (Table 1; Figure 4B).

Baseline live hard coral cover was significantly lower in Utila relative to Banco Capiro (GAMM: β = −0.99 ± 0.15 SE, p < 0.001), with live hard coral cover in Banco Capiro remaining markedly higher, particularly prior to the 2023 bleaching event (Figure 4A). Additionally, higher macroalgal cover was found in Utila (GAMM: Location effect: β = 0.38 ± 0.23 SE, p = 0.10) (Figure 4B) and significantly lower turf algal cover was found at Utila relative to Banco Capiro (GAMM: β = −1.22 ± 0.19 SE, p = 1.11 × 10⁻ 8) (Figure 4B).

Spatio-temporal differences in benthic, coral and macroalgal community structures

There were significant differences in benthic, live hard coral and macroalgal community composition between locations and between pre-bleaching (2013–2019) and post-bleaching (2023–2025) periods (PERMANOVA, Table 2). The interaction between location and period was statistically significant, indicating that the temporal shift in benthic, live hard coral and macroalgal community composition occurred differently at each location (Table 2; Figure 5). These patterns were matched by the NMDS ordination, which showed a difference of pre- and post- bleaching benthic communities at both sites (Figure 5A). NMDS stress values were low (Benthic = 0.094, Hard Coral = 0.115, Macroalgae = 0.153) across all ordinations, indicating a good representation of the multivariate community data in two-dimensional space and supporting interpretation of the observed clustering patterns (Clarke, 1993).

Up to 50% of the differences in benthic community structure between Banco Capiro and Utila were primarily driven by variation in cover of live hard coral (13.6%), soft coral (12.2%), sand (11.9%), turf algae (9.9%) and macroalgae (5.9%) (Supplementary Table S3). Live hard coral cover was higher in Banco Capiro between 2013 and 2019, however, during 2023-2025 (post-bleaching) the community structure shifted towards turf algal dominance (Figure 5A). At Utila, cover of soft coral and sand was higher throughout the study, with a shift towards macroalgae and turf algae by 2023-25 (Figure 5A). This corresponds with results from the GAMM which indicated an increase in algal cover (Figure 4B).

Up to 54% of the differences in coral community structure between Banco Capiro and Utila were driven by variation in cover of Agaricia (30.5%), Diploria (6.1%), Madracis (6%), Stephanocoenia (5.9%), and Dichocoenia (5.7%), (Supplementary Table S4). Agaricia was the main driver in the differences between Banco Capiro and Utila, with it being dominant in Banco Capiro until 2023 (Supplementary Table S4; Figures 5B, 6).

Up to 47% of the differences in macroalgal community structure between Banco Capiro and Utila were primarily driven by variation in cover of Halimeda (12.6%), Dictyota (10.7%), Galaxaura (9.4%), Sargassum (7.3%), and Lobophora (6.7%) (Figure 5C; Supplementary Table S5). Halimeda was the main driver in the differences between Banco Capiro and Utila with it having a high mean cover in Utila. Cover of Dictyota was also higher in Utila, however increased in Banco Capiro post 2023 bleaching event (Supplementary Table S5; Figure 7). Galaxaura, and Sargassum had higher percentage cover in Banco Capiro, with Galaxaura demonstrating a spike post 2023 bleaching event (Figure 7). Lobophora contributed similarly to both locations, with comparable mean cover in Banco Capiro and Utila (Figures 5C, 7).

This study examined benthic community change across two contrasting Honduran reef systems, Banco Capiro in Tela Bay and the south-side reefs of Utila, from 2013 to 2025, spanning the 3rd and 4th Global Coral Bleaching Events. It is worth noting that surveys could not be conducted between 2020 and 2022 due to COVID-19 restrictions. However, given the timing of the two GBCEs relative to the available study years, the survey gap is unlikely to alter the pattens reported in this study. The results provide three main insights into how Caribbean coral bright spots may respond to increasing marine heatwave frequency and intensity. Results suggest that bright spots may temporarily maintain high coral cover under repeated thermal stress, but previous persistence does not guarantee survival under more extreme disturbance. They also show that high coral cover may conceal limited taxonomic or functional redundancy, and that distinct pre-disturbance reef communities may follow different post-bleaching trajectories rather than shifting rapidly towards a common degraded Caribbean reef state.

Prior to 2023, Banco Capiro showed characteristics consistent with a coral bright spot, maintaining regionally high percentage live hard coral cover despite repeated thermal stress (Bodmer et al., 2021; Cramp et al., 2025). Banco Capiro also experienced repeated annual Degree Heating Week accumulation events above 12 °C-weeks in 2017, 2019, 2020 and 2021. Under the NOAA Coral Reef Watch Bleaching Alert Area scale, DHW values of 12 to <16 °C-weeks correspond to Bleaching Alert Level 3, indicating risk of multi-species coral mortality (NOAA, 2018). Importantly, the thermal exposure experienced at Banco Capiro during the third GCBE and after, occurred within a period of substantial regional bleaching rather than an absence of bleaching pressure. Bleaching was documented by Muñiz-Castillo et al. (2024) across the Mesoamerican Reef from 2015-2017, including in Tela Bay and the Honduran Bay Islands, yet, live hard coral cover remained high before 2023, making this persistence notable because it contrasts with expected outcomes under elevated thermal exposure (Hughes et al., 2018b; Skirving et al., 2019; Eakin et al., 2026).

Banco Capiro is a moderately turbid reef, indicating greater light attenuation than clearer-water reef systems (Shi and Wang, 2010). Although turbidity was not directly tested as a driver of coral persistence in this study, this environmental context is important because solar irradiance can interact with thermal stress to influence bleaching risk (Cacciapaglia and van Woesik, 2016; Fisher et al., 2019; Zweifler et al., 2024). The maintenance of high coral cover at Banco Capiro before 2023 is therefore consistent with studies showing that turbid reefs can have less severe bleaching responses than clearer-water counterparts (Morgan et al., 2017; Sully and Woesik, 2020; Rosedy et al., 2023; Zweifler et al., 2024). One possible explanation is that suspended particles attenuate solar irradiance, reducing one of the interacting stressors involved in coral bleaching alongside elevated temperature (Lucas et al., 2023; Butcherine et al., 2023). Another possibility is that long-term exposure to warm, turbid and low-light conditions has shaped coral assemblages at Banco Capiro, favouring taxa able to persist under these local conditions (Morgan et al., 2016; van Woesik and McCaffrey, 2017; Teixeira et al., 2019; Burt et al., 2020). Banco Capiro may therefore have resembled a potential thermal refuge before 2023, in the sense that high coral cover persisted despite repeated thermal stress (Woesik et al., 2012). However, because turbidity was not tested mechanistically here, it should be interpreted as a plausible contextual factor rather than a confirmed driver.

Although Banco Capiro maintained high coral cover before 2023, this should not be interpreted as equivalent to high ecological resilience (Bellwood et al., 2004; Alvarez-Filip et al., 2013; Brito-Millán et al., 2019). Its pre-disturbance coral community was strongly dominated by Agaricia, a genus associated with opportunistic growth and rapid colonisation of available reef space (Hughes and Jackson, 1985; Darling et al., 2012). Agaricia is also common in deeper and lower-light environments, suggesting that its dominance at Banco Capiro may partly reflect the reef’s moderately turbid, light-limited setting (Kahng et al., 2010; Hoeksema et al., 2017). Alternatively, Agaricia dominance may reflect an altered or legacy coral-community state, consistent with wider Caribbean evidence that disturbance can shift reefs away from historical framework-building taxa towards opportunistic or environmentally tolerant coral assemblages (Aronson and Precht, 2001; Alvarez-Filip et al., 2013; Cramer et al., 2021).This supports previous work showing that coral taxa differ in colonisation capacity, stress tolerance and competitive dominance (Darling et al., 2012; Cramer et al., 2021). However, it also highlights a limitation of assessing bright spots using percentage coral cover alone. At Banco Capiro, high cover reflected the dominance and persistence of Agaricia, rather than broad resistance across a diverse coral assemblage (Nyström, 2006; Brito-Millán et al., 2019).

This limited taxonomic redundancy may help explain the rapid post-2023 decline. Although Agaricia may have been well suited to Banco Capiro’s lower-light conditions before 2023, this did not mean it could tolerate the extreme thermal exposure recorded during the 4th GCBE. Previous studies have identified Agaricia spp. as susceptible to bleaching and mortality during severe heat stress events (Smith et al., 2013; Doherty et al., 2025). Therefore, once Agaricia declined, total live hard coral cover declined sharply across the reef. Banco Capiro therefore illustrates an important limitation of coral-cover-based bright spot assessments: apparent persistence can be associated with dominance by a narrow subset of taxa, creating vulnerability when those taxa are affected by more extreme disturbance (McWilliam et al., 2018; Brito-Millán et al., 2019).

The sharp decline in live hard coral cover after 2023 suggests that Banco Capiro’s previous persistence under repeated heat stress had an upper limit, which was exceeded during the 4th GCBE (Mo et al., 2022; Carlson et al., 2022). The contrast between the 3 rd and 4 th GCBE is therefore particularly informative, as regional bleaching and thermal stress occurred during both events, but major live coral cover loss at Banco Capiro was only observed following the substantial elevated heat exposure in 2023-2024. To establish the upper thermal threshold, would have required frequent in-situ monitoring of colony status (e.g. weekly) throughout the event. Higher frequency monitoring of bleaching status and mortality during heat stress events should be considered as part of regular monitoring programs if the aim is to establish spatial and taxonomic variation in upper thermal limits. In 2023 and 2024, Banco Capiro experienced DHW values nearing 20 °C-weeks, among the higher thermal exposures recorded on Caribbean reefs during this event and representing extreme bleaching heat stress, consistent with the sharp post-2023 decline in live hard coral cover (Birkart and Álvarez-Filip, 2025; Reimer et al., 2024; Wicquart et al., 2025). Under the NOAA CRW Bleaching Alert Area scale, DHW values 16-20 °C-weeks indicate Bleaching Alert Level 4 and risk of severe multi-species mortality affecting more than 50% of corals (NOAA, 2018). Rather than contradicting the idea that turbid or marginal reefs can provide partial refuge, this decline suggests that such refugial effects have upper limits (Mo et al., 2022; Carlson et al., 2022). Banco Capiro persisted through earlier DHW events above 12 °C-weeks, but experienced rapid coral loss after the more extreme exposure of 2023 and 2024 (Skirving et al., 2019). This pattern suggests a possible non-linear threshold, where high coral cover was able to persist under repeated moderate to severe thermal stress, but declined sharply when exposure became more extreme, leading to substantial coral loss and benthic reorganisation (Kavousi and Keppel, 2018; Hughes et al., 2018b). This interpretation is consistent with broader work showing that potential reef refugia are not immune to bleaching or thermal stress (Smith et al., 2016; McWhorter et al., 2022).

In contrast, Utila entered the 4th GCBE from a lower-coral, more macroalgal-dominated state, consistent with many contemporary Caribbean reef assemblages (Contreras-Silva et al., 2020; Wicquart et al., 2025). Soft corals and sponges were more abundant in Utila, as expected for a more typical Caribbean reef system (de Bakker et al., 2017). Live hard coral cover remained relatively low throughout the study and was broadly consistent with regional Caribbean live hard coral cover means (Wicquart et al., 2025). However, Utila still experienced a decline in mean live hard coral cover after the 4th GCBE, showing that already lower-coral systems remain vulnerable to further coral loss under extreme thermal stress (Green et al., 2026). These findings are consistent with studies showing that many Caribbean reefs have shifted towards lower live hard coral cover and greater dominance of macroalgae, soft corals, sponges and other non-reef-building taxa, and continue to decline under thermal stress (Contreras-Silva et al., 2020; González-Rivero et al., 2011; Edmunds, 2025).

Although Banco Capiro and Utila differed strongly in percentage live hard coral cover, their coral communities were also compositionally distinct. In Utila, community dissimilarity was associated with Madracis, Stephanocoenia and Diploria, including stress-tolerant genera such as Stephanocoenia and Diploria (Darling et al., 2012; Cramer et al., 2021). Live hard coral cover in Utila remained relatively stable through the 3rd GCBE, which is notable given that bleaching was recorded during this period (Muñiz-Castillo et al., 2024; Green et al., 2026), but this stability occurred under lower thermal exposure than Banco Capiro, perhaps due to its offshore location, and within an already low-cover coral assemblage (Green et al., 2026). During the 3rd GCBE, Utila reached approximately 8 °C-weeks, corresponding to NOAA CRW Bleaching Alert Level 2, which indicates risk of reef-wide bleaching and mortality among heat-sensitive corals, but remains below the higher alert levels associated with multi-species mortality (NOAA, 2018). Utila’s muted response may therefore reflect both the lower DHW exposure and the presence of stress-tolerant remnant coral taxa (Smith et al., 2013; Gintert et al., 2018). However, this stability was not maintained during the more intense thermal exposure of the 4th GCBE. Green et al. (2026) reported that DHW intensities at some sites in Utila and across the Bay Islands during the 4th GCBE were approximately twice those recorded during the 3rd GCBE. Although Utila’s 2023 and 2024 DHW exposure remained lower than Banco Capiro’s, mean live hard coral cover still declined from 13 ± 1% (mean + SE) before bleaching to 7 ± 2% (mean + SE) after bleaching. This suggests that even more stress-tolerant coral assemblages can be vulnerable once thermal exposure intensifies beyond previous levels (Brandt, 2009; Smith et al., 2013). The Utila comparison shows that Banco Capiro was not simply the same reef state with more coral, but a distinct pre-2023, Agaricia dominated bright spot compared with Utila’s lower-coral, more macroalgal, stress-tolerant assemblage. This contrast strengthens the interpretation that Banco Capiro’s pre-2023 persistence represented a distinct bright spot state, rather than a delayed version of the Utila trajectory.

Following the 4th GCBE, the benthic community in Banco Capiro shifted towards a turf algal and macroalgal-dominated community (Ostrander et al., 2000; Hughes et al., 2018b; Elma et al., 2023). Turf and macroalgal cover followed cyclical patterns in relation to the 2017 and 2023 bleaching events at both study locations. This is consistent with post-disturbance algal dynamics, where turf algae rapidly colonise space opened by coral mortality, potentially inhibiting coral recruitment before being replaced or outcompeted by larger macroalgae (O’Brien and Scheibling, 2018; Roth et al., 2018; Pessarrodona et al., 2023). However, the ecological implications of algal expansion depend not only on total algal cover, but also on algal identity and traits (Box and Mumby, 2007; Nieder et al., 2024; Lee et al., 2026). At Banco Capiro, Galaxaura was the second most dominant algal genus, after Dictyota, and increased in percentage cover after the 4th GCBE. This is important because Galaxaura can have allelopathic effects on coral and induce bleaching in close proximity and inhibit coral recruitment (Nieder et al., 2024; Bonaldo et al., 2017). It may also be less palatable to herbivorous fishes than other macroalgae such as Dictyota and has been associated with reduced herbivore feeding (Lee et al., 2026; Spiers and Frazer, 2023). Increasing Galaxaura cover may therefore contribute to the persistence of an algal-dominated state by limiting coral recruitment and reducing effective herbivore control (Rasher and Hay, 2010; Nieder et al., 2019).

In the two years following the 4th GCBE, Banco Capiro did not show early convergence with the lower-coral, macroalgal-dominated state observed in Utila. Instead, it shifted from a high coral, Agaricia dominated assemblage towards a turf and macroalgal dominated community with increasing Galaxaura cover, while Utila remained characterised by, higher macroalgal cover, dominated by Dictyota and Halimeda, soft corals, sponges and low cover of stress-tolerant remnant coral genera. This suggests that early post-disturbance trajectories may differ among reef systems affected by the same regional bleaching event, particularly where DHW exposure, local environmental conditions and pre-disturbance community composition vary (Donovan et al., 2021; Zweifler et al., 2024). Reduced light availability may continue to filter benthic community composition after disturbance, influencing which coral and algal taxa persist or expand (Guest et al., 2016b; Loiola et al., 2019; Cartwright et al., 2023). This interpretation should remain cautious because the post-bleaching period captured here is short. Surveys in 2024 and 2025 represent only one and two years after the 2023 bleaching event, which is unlikely to capture the full trajectory of coral recovery, delayed mortality, recruitment or longer-term community stabilisation (Robinson et al., 2019; Gouezo et al., 2019). Banco Capiro may still converge towards a more typical contemporary Caribbean reef state over time, or it may continue along a distinct post-disturbance trajectory.

Future assessments should move beyond percentage live hard coral cover alone and consider coral community composition, dominance structure, taxonomic and functional redundancy, algal identity and post-disturbance recovery potential (Brito-Millán et al., 2019). Future monitoring should also incorporate direct in situ measurements of turbidity, as the present study relied on satellite-derived K d (490) as a proxy and therefore could not capture reef-scale variation in turbidity. Continued long-term monitoring will be essential for determining whether Banco Capiro’s post-2023 assemblage represents a transient early response or a more persistent shift in benthic community composition.

This study reports on how a Caribbean coral bright spot responded to escalating thermal stress across the 3rd and 4th Global Coral Bleaching Events. Overall, these findings suggest that coral reef bright spots are not secure refugia from climate-driven disturbances, but as dynamic reef states whose persistence have defined upper thermal limits under escalating ocean warming.