Casting a shadow: in situ shading of corals reduces hyperthermal paling

Abstract

Coral bleaching is an increasingly common and severe stressor to reefs that can result in mortality of wild and restoration colonies. One potential mitigation effort is the installation of shade structures that can minimize the synergistic UV stressor contributing to bleaching-related tissue loss. We tested the effect of temporarily installed in situ shade structures over wild colonies of two Caribbean species (Pseudodiploria clivosa and Colpophyllia natans) to determine whether they could mitigate bleaching, affect symbiont to host ratios, and alter symbiont community structure. Despite 2024 having the second highest cumulative heat stress on record at the study site, corals exhibited only moderate paling/bleaching. This is likely due to symbiont communities consisting almost exclusively of the heat-tolerant Durusdinium, a probable result of the unprecedented thermal stress the prior summer. However, even the moderate loss of color was remedied by the shade structures, preventing further paling/bleaching of experimental colonies after installation. The symbiont:host ratios exhibited moderate change through the thermal stress and recovery event, but the change was not affected by the presence of shades. We suggest that shading could be a relatively low-cost option for minimizing bleaching of high priority corals through a hyperthermal event.

1 Introduction

Coral mortality resulting from hyperthermal bleaching is recognized as one of the greatest threats to coral reefs worldwide, and is expected to become increasingly common and severe as climate change accelerates (Donner et al., 2005; Mellin et al., 2024). Coral bleaching is the breakdown of the symbiosis between the coral host and the algal endosymbiont. Without these symbionts, corals are devoid of much of their nutrition and can perish from starvation. If temperatures are high enough, the time period between bleaching and mortality can be extremely rapid as corals die from acute heat stress. Thermal stress to colonies can thus be assessed by water temperature as well as cumulative heat stress, traditionally measured as degree heating weeks (DHWs). DHWs use the degrees (°C) in excess of the bleaching threshold and the number of weeks that excess occurs, summed over the previous 12 weeks. As an example, a DHW of 4 could be driven by 1°C over the bleaching threshold for four weeks, 2°C over the bleaching threshold for two weeks, or 4°C over the bleaching threshold for one week. In most hyperthermal scenarios, DHWs better predict coral bleaching than water temperatures (Kayanne, 2017).

From 2023-2024, elevated temperatures across multiple ocean basins resulted in the Fourth Global Coral Bleaching Event (Reimer et al., 2024). In the Florida Keys, water temperatures during the summer of 2023 were record-breaking, and cumulative thermal stress was double that of any year since satellite records began in 1985 (data: NOAA Coral Reef Watch (2025), reviewed in Neely et al. (2024)). At most sites in the Florida Keys, bleaching occurred on 100% of corals. However, mortality varied substantially based on location as well as species. At most reef sites, the branching Acroporid species suffered full mortality (Manzello et al., 2025), but brain and boulder corals experienced very little mortality (Neely et al., 2024). Though Neely et al. (2024) found essentially no mortality of reef-building corals at offshore sites, one of the monitored inshore sites experienced exceptionally high temperatures and cumulative heat stress, and 55% of monitored corals experienced partial or total mortality as a result. Partial and total mortality of brain corals was particularly pronounced, with between 41% - 80% of colonies of four brain coral species losing some or all of their tissue (Neely et al., 2024). As such, we selected this site and these highly susceptible individuals to conduct an intervention experiment aimed at protecting targeted colonies during the summer of 2024.

Installation of shade structures is a potential method for mitigating thermal stress on corals. Though shading does not impact cumulative heat stress, it does reduce UV and irradiance stress which exacerbates bleaching impacts on coral colonies (Gleason and Wellington, 1993; Fitt and Warner, 1995). Shading of colonies has been conducted using a variety of methods (review in Tagliafico et al. (2022)). Various experiments in the Indo-Pacific have shown that shaded corals have delayed bleaching (Butcherine et al., 2023), have elevated chlorophyll concentrations, symbiont densities, and photosystem II proteins (Jeans et al., 2013), and have higher growth rates (Coelho et al., 2017) than unshaded controls. Trialing of shade structures in the Caribbean has been more limited.

Our project goals assessed how targeted seasonal installation of shade structures affected two reef-building Caribbean coral species (Pseudodiploria clivosa and Colpophyllia natans) that were highly susceptible to hyperthermal mortality in 2023. Specifically, we examined how shade structures affected: 1) paling/bleaching severity, 2) symbiont density (measured as symbiont to host ratios), and 3) symbiont community structure.

2 Materials & methods

2.1 Shade deployment

Shades were deployed at Newfound Harbor, a patch reef in the lower Florida Keys just south of Big Pine Key (Figure 1). The site is located less than 1 km from shore. The main reef structure is an approximately 120 m diameter shallow (< 1 m) conglomerate of mostly dead Orbicella colonies. The reef structure is surrounded by a shallow (3.5 – 5.5 m), flat hardbottom plain, colonized mostly by low-lying coral colonies and octocorals.

Figure 1

We conducted the shading experiment on two species inhabiting the hardbottom plain that experienced particularly high bleaching-related mortality in 2023 – Colpophyllia natans and Pseudodiploria clivosa. Selected colonies were of suitable size and shape for experimental shading, with each having at least one isolate exceeding 144 cm², but not bigger than 10,000 cm². Twenty colonies of each species were selected and divided into control (non-shaded) and experimental (shaded) treatments. Per permitting requirements, shaded corals could not be in proximity to each other, and so the experimental treatment corals were intentionally selected to be geographically dispersed across the site area, but otherwise corals were haphazardly assigned into treatment groups.

Shade structures consisted of polyvinyl chloride (PVC) frames, which were filled with spray foam for flotation and then sealed against water intrusion. Frames were sized for each individual coral, such that the frame extended 5 cm on each side from maximum length and width. Light tan agricultural shade cloth (NCSNA Inc Model PS6160TAN) was attached to each frame using zip ties. The cloth was made from UV-resistant high-density polyethylene, which is durable and chemically stable to prevent degradation or toxic leaching. Based on the manufacturer’s specifications, the shade cloth reduced incoming light by 60% and did not alter the spectral composition. Using an underwater drill, we made four small (0.75 cm) holes in a rectangular arrangement around each shaded colony. These were approximately 50 cm away from the target coral and arranged so as not to affect any other benthic fauna. Into each, an 11cm eye screw was manually twisted in. Parachute cord was strung from the eye screws to the corners of each frame. Shades floated approximately 30 cm over the top of each target coral (Figure 2).

Figure 2

The experimental design called for deployment of the shades once cumulative thermal stress at the site reached 4 degree heating weeks (DHW), which is the expected onset of any paling/bleaching (Kayanne, 2017). Unfortunately, permitting delays meant that shades could not be installed until DHWs exceeded 7 (August 16-17). Additionally, we had to temporarily remove shade structures for two tropical storms (September 23 – October 1 and October 7 – October 14). Shades and all fixtures were removed when DHWs subsided to 4 on November 12. At least every two weeks, as well as when shades were reinstalled for tropical storms, shades were lightly scrubbed with a dishwashing brush to prevent fouling and supplemental shading resulting from accumulated sediment or organisms.

2.2 Field assessments

Monitoring and sampling occurred nine times, from July 16 to November 12, 2024 (Table 1).

All shaded and control colonies were monitored and sampled every two weeks, as well as whenever shades were removed or reinstalled. During each visit, the same diver visually assessed corals for percent live cover and percent mortality. They were also assessed for paling/bleaching using a Coral Watch Coral Health Card (Figure 2D). The diver was permitted to record half-steps (e.g., if the color was between a 3 and a 4, it could be recorded as 3.5). For each coral, the diver estimated the percentage of live coral that was paled/bleached and recorded the matching bleaching index number from the health card. Additionally, the percentage of the coral that was not pale/bleached and its associated color index were recorded. These two values were averaged to provide a single bleaching index number for each colony using the equation:

We analyzed the effects of treatment and time on bleaching index numbers using a Repeated Measures 2-way ANOVA. Two of the corals were omitted from these analyses because they had missing datapoints: one was a shaded C. natans colony that went missing during a tropical storm (absent sample periods 8 and 9), and one was a control P. clivosa that could not be found during sample period 7. Data were first assessed for normality (Shapiro-Wilk) and equal variance (Brown-Forsythe). Post-hoc pairwise comparisons were conducted using Holm-Sidak tests. Each species was assessed independently.

We further assessed bleaching index scores by comparing the rate of change in color between control and shaded groups for sample periods 2–5 and periods 5-9. The first interval represents the period from the application of the shades to the time of peak thermal stress. The second interval represents the recovery period, which was marked by declining temperatures but also two tropical storms as well as two removals and reinstallations of the shade structures. For each coral during each of the two intervals, we calculated the slope. The shaded C. natans that was lost after sample period 7 was excluded for the second interval since the last two data points were missing and the slope was notably greater than the others within the treatment group. The slopes during each interval were compared between treatments using a t-test when normality and equal variance passed, and a Mann-Whitney rank sum test when they did not. A Cohen’s D test assessed the difference between the means of the slopes of the two treatments (d < 0.5 represents a small to moderate difference, and d > 0.8 represents a substantial difference).

2.3 Sampling and symbiont assessments

We assessed the symbiont community of shaded and control corals using qPCR to determine the symbiont to host ratios (S:H ratio) of each symbiont clade within each sample. This metric of symbiont load estimates the ratio of the total number of symbionts to the total number of host cells within a sample. It is minimally invasive (not requiring large areal sections of tissue), allows the contributions of each symbiont clade to be identified and quantified, and is not affected by changes in the numbers of host cells (Mieog et al., 2009; Cunning and Baker, 2013, 2014).

During each monitoring period, we used cuticle clippers to take a small tissue biopsy (approximately 2 mm²) from the edge of each colony. Though edges were sometimes self-shaded and not always representative of overall colony bleaching, this sampling location was governed by permitting requirements. The location of samples from each colony was standardized by placing a nail into the substrate adjacent to the initial sampling point. The bleaching index score directly at the location of each sample was also taken during each monitoring period, with the exception of sample period 1. Each sample was put into a labelled bag and transferred into labeled tubes with DNA/RNA Shield once topside.

Samples were transported to the Rosenstiel School of Marine, Atmospheric, and Earth Science and stored in a -80°C freezer until processing. A small aliquot of each biopsy sample was preserved by adding 100μL of DNA/RNA Shield-preserved sample into 400uL of 1% sodium dodecyl sulfate (SDS) in DNA buffer (DNAB) solution and heating to 65°C for 90 minutes. Genomic DNA was extracted from the SDS lysates following Baker and Cunning (2016).

Taqman environmental master mix (Thermo Fisher Scientific) assays were used to amplify the actin gene in Symbiodinium (Winter, 2017), Breviolum (Cunning et al., 2015b), Cladocopium, and Durusdinium (Cunning and Baker, 2013). Coral hosts were amplified with SYBR-Green assays and primers that target species-specific single copy-loci. Pax-Cam (paxC) gene primers were developed and validated for both C. natans and P. clivosa (Table 2). Mixed chemistry quantitative PCR (qPCR) assays were run where the cycle threshold (Ct) values were used to estimate the S:H ratio using the StepOneR package for R (Cunning, 2018). All samples were run in duplicate for each target, and extraction and biochemical efficiencies were all checked as part of QA/QC.

In order to account for natural variability within corals, we standardized the S:H scores for each individual. We chose sample period 5 as the baseline (value = 1) because this time point had the lowest S:H scores averaged across all colonies, and because all sample runs were successful.

In addition to the three datapoints where corals could not be found, there were six samples for which qPCR data were unavailable. These were two control P. clivosa (both sample period 2: no symbiont amplification), one shaded P. clivosa (sample period 7: no sample taken), one control C. natans (sample period 9: did not pass QA/QC), and two shaded C. natans (one in sample period 1: no sample taken, and one in sample period 2: no symbiont amplification). Overall, S:H values for all time points were incomplete for four C. natans (one control, three shaded) and four P. clivosa (three control, one shaded). As such, we did not conduct repeated measures ANOVA on these data, but instead focused on the rate of change to the standardized S:H ratios across the two main intervals: sample period 2–5 and sample period 5-9. We again excluded the shaded C. natans coral that went missing from the second interval analyses since the last two datapoints were missing and the value was lower than any others within the treatment group.

3 Results

3.1 Temperature

NOAA Coral Reef Watch sea surface temperature data identified 2024 as the second highest cumulative heat stress on record at Newfound Harbor, exceeded only by the 2023 marine heat wave (Figure 3). In late September of 2024, degree heating weeks reached 10.7. The maximum temperature within the Coral Reef Watch data was 31.8°C, and from July 10 to November 15, 61 days exceeded the 30.5°C bleaching threshold (temperatures were already at 30.7°C on July 10). Three in situ temperature loggers (Hobo Pendants) were attached to the substrate (unshaded) throughout the area encompassed by the experimental colonies. These recorded hourly temperatures, but we consider only the readings from 0200 each night so as to match the nighttime comparisons from NOAA Coral Reef Watch. Two of these loggers failed in mid-September, but all three had nearly identical temperature patterns to that point. The in situ loggers recorded similar temperature trends to the satellite-derived data, but exhibited higher variability (higher high temperatures and lower low temperatures). These loggers recorded maximum temperatures of 32.6 – 32.8°C, nearly a full degree higher than the NOAA dataset (Figure 3). The logger that recorded through the end of summer documented 76 days in excess of 30.5°C after July 10, which was 16 days more than the NOAA dataset.

Figure 3

3.2 Pseudodiploria clivosa

3.2.1 Bleaching scores

Compared to controls, shaded P. clivosa colonies had significantly higher bleaching index scores (were darker) during peak thermal stress, and also had significantly different rates of change of those scores during both the thermal stress and the recovery periods. The overall trend of bleaching index scores was a decline (paling) from mid-July through mid-September as cumulative thermal stress increased, followed by an increase in index score (darkening) as temperatures declined. P. clivosa colonies were already visually pale by the time permitting allowed for shade structures to be installed in mid-August (sample period 2: Figure 4A).

Figure 4

Significant factors affecting bleaching index scores of P. clivosa were treatment (RM 2-way ANOVA: F_(1,144) = 5.8, p = 0.03), sample period (F_(8,144) = 39.7, p < 0.001), and the interaction of treatment and sample period (F_(8,144) = 2.9, p = 0.005). Shaded corals had significantly higher bleaching index scores (were darker) than unshaded during sample periods 4, 5, and 6 (p < 0.001, p = 0.002, p = 0.006 respectively).

Significant differences in bleaching index scores among time periods were complex (for full values, see Supplementary Table 1). For shaded corals, bleaching index scores were significantly higher in sample period 9 (mid-November) than in all previous monitoring periods. Additionally, bleaching index scores in sample period 2 (mid-August immediately prior to shade deployment) were significantly lower than any other period except sample period 4. For unshaded corals, the primary differences were that bleaching index scores during sample period 9 were significantly higher than any other period, and scores in sample period 8 were significantly higher than those from sample periods 2 through 6. Scores in sample period 4 were also significantly lower than most other time points (except sample periods 2 and 5).

Between sample period 2 and sample period 5 (mid-August to September 23), the bleaching index scores of the control corals decreased slightly (-0.8 ± 3.3 (SE) index points *10^-3/day) while those of shaded corals increased (12.7 ± 3.7 index points *10^-3/day; t-test: p = 0.01; Cohen’s d = 1.19). During the recovery interval (sample periods 5-9), the index scores of control corals increased at a rate of 24.8 ± 2.2 index points *10^-3/day, a rate 1.4 times that of shaded corals (16.9 ± 1.6 index points *10^-3/day; t test: p = 0.01; Cohen’s d = 1.25).

The bleaching index scores taken at the point of sampling (colony edges) generally approximated the bleaching index scores for the whole colony (Supplementary Figure 1). The differences between the two scores (ΔBI) on any colony at any timepoint ranged from -0.5 to +1.0. Within each treatment group for each time point the 95% confidence limits of ΔBI almost never excluded “0”, indicating that the scores did not misrepresent the treatment groups. The one exception was control corals during sample period 3, when the 95% confidence index for sample location bleaching index scores was very slightly above 0 (the sample colors were darker than the average color scores).

No corals experienced any mortality during the experimental period.

3.2.2 Symbionts

All of the P. clivosa colonies contained only Durusdinium throughout all time points (Figure 5).

Figure 5

The rate of change in the S:H ratio did not differ between shaded and control colonies as thermal stress increased, but did differ during the recovery period. The general temporal trends of S:H ratios of both treatment groups was a decline from mid-July to late-August, variability from late August until mid-October, followed by an increase to or above pre-warming levels (Figure 4A) from mid-October to mid-November. From sample period 2 to sample period 5, S:H ratios for shaded corals increased 5.6 times faster than those of control corals (20.7 ± 5.3 standardized S:H *10^-4/day for shaded corals versus 3.7 ± 11.0 standardized S:H *10^-4/day for control corals), but this difference was not significant (Mann-Whitney: p = 0.19; Cohen’s d = 0.59). During the recovery interval (sample period 5-9), S:H ratios of shaded corals increased at a similar rate as the previous interval (16.5 ± 7.1 standardized S:H *10^-4/day) while the rate of change in control corals was 2.8 times higher than for shaded ones (46.2 ± 8.3 standardized S:H *10^-4/day; t-test: p = 0.01; Cohen’s d = 1.18).

3.3 Colpophyllia natans

3.3.1 Bleaching scores

Shaded C. natans colonies had significantly higher bleaching index scores than controls during peak thermal stress, and the rate of change of those scores also differed significantly between treatment groups during the recovery period. Visual bleaching index scores for C. natans generally declined as cumulative thermal stress rose, with subsequent increases dependent on shade treatment (Figure 4B). Like the P. clivosa colonies, C. natans corals were also visually paled before shades could be installed. However, shaded colonies generally quickly regained color once shades were deployed while control colonies paled further.

The repeated measures 2-way ANOVA identified treatment (F_(1,144) = 10.7, p = 0.004) and sample period (F_(8,144) = 6.6, p < 0.001) as significant factors affecting the bleaching index score, but not the interaction between them (F_(8,144) = 1.6, p = 0.13). Overall, shaded corals had higher bleaching index scores (were darker) than non-shaded ones, and index scores in sample period 9 were higher than periods 1-7. Shaded corals had significantly higher index scores than control ones in sample periods 4 and 5 (p = 0.003, < 0.001 respectively). For shaded corals, bleaching index scores did not differ significantly amongst any time periods. However, for unshaded corals, colonies in sample period 9 were darker than in all other time periods except 8, and colonies in sample period 8 were darker than in time periods 4 and 5 (Supplementary Table 1).

During the interval from sample period 2 to sample period 5, the bleaching index of shaded corals increased (4.9 ± 3.5 index points *10^-3/day) while the bleaching index of control corals decreased (-3.4 ± 3.3 index points *10^-3/day) but this was not significant (t-test: p = 0.07; Cohen’s d = 0.85). During the recovery interval (sample period 5 to 9), color scores of control corals increased 3.7 faster than shaded ones, a highly significant difference (33.6 ± 4.3 index points *10^-3/day for control versus 9.1 ± 4.4 index points *10^-3/day for shaded corals; t-test: p = 0.003; Cohen’s d = 1.53).

The bleaching index scores taken at the points of sampling (colony edges) were higher than the overall colony index scores for both the shaded and control groups during every time period, meaning the sample areas were generally darker than the colonies as a whole (Supplementary Figure 1). However, the 95% confidence intervals only exceeded 0 for sample period 2 (both control and shaded), sample period 4 (both control and shaded), and sample period 6 (shaded only).

Only one coral exhibited any mortality through the experimental period. This was the result of a tropical storm which dislodged it and rolled it approximately 30 meters across the substrate, causing substantial abrasion on the colony ridges (Figure 6).

Figure 6

3.3.2 Symbionts

Within the C. natans colonies, 10 of 21 contained only Durusdinium (Figure 5). The only other genus of endosymbiont present was Symbiodinium, and for all but three C. natans, this was only present in trace amounts (< 5%). Within coral 3443 (shaded), Symbiodinium comprised 7% of endosymbionts in sample period 2, but this dropped off to 0% by sample period 6. Coral N-61 (control) contained notable Symbiodinium during the first and last sample periods (7% and 37%), but little to none (<1%) during the intermediate sample periods. Coral 888 (control) contained 95 – 100% Symbiodinium through sample period 6, though Durusdinium was notably present in sample periods 7 and 9.

Between shaded and control C. natans, there were no significant differences in the rates of change of standardized S:H ratios during the warming nor the recovery period. The general trend was a decline in S:H ratios through early October, then an increase during the recovery period (Figure 4B). During the increasing heat stress interval (sample periods 2-5) S:H ratios of shaded corals declined 1.5 times faster (-51.5 ± 18.3 standardized S:H *10^-4/day) than control ones (-34.9 ± 15.6 standardized S:H *10^-4/day), but this difference was not significant (Mann-Whitney: p = 0.65; Cohen’s d = 0.30). During the recovery interval (sample period 5-9), S:H ratios of shaded corals increased at 1.4 times the rate (15.5 ± 6.8 standardized S:H *10^-4/day) of control colonies (10.8 ± 12.1 standardized S:H *10^-4/day), but this difference was not significant (t-test: p = 0.76; Cohen’s d = 0.14).

4 Discussion

The temporary installation of shade structures over two species of Caribbean corals significantly reduced colony paling during the 2024 summer compared to controls. This treatment effect was driven by significant differences during the most thermally stressful time periods (8–10 DHWs). For both species, shade deployment reversed paling (corals became darker) while control corals continued to become lighter until temperatures declined.

Considering that 2024 had the second highest cumulative heat stress on record (since 1985), and that the theoretical bleaching threshold of 30.5 °C was exceeded for 61 (from satellite data) to 75 (from in situ loggers) days, it was initially surprising that corals experienced only minor to moderate paling rather than partial or full bleaching and associated mortality. We suspect this is due to both control and experimental corals being overwhelmingly dominated by Durusdinium endosymbionts. Durusdinium is not a well-documented symbiont within the two coral species assessed here. Pseudodiploria clivosa from the Florida Keys have primarily associated with Breviolum symbionts (Klepac et al., 2023; Eaton et al., 2025). Colpophyllia natans from Florida and the U.S. Virgin Islands are primarily known to contain Breviolum, but have also been found harboring Cladocopium and Durusdinium (Correa et al., 2009; Beavers et al., 2023; Cunning et al., 2024). Ing (2024) found that C. natans on southeast Florida reef habitats host primarily Breviolum symbionts, but that C. natans in “urban environments” (artificial substrates inside port areas that are more prone to disturbance and temperature extremes) primarily contained Durusdinium. Corals with Durusdinium are generally known to be more heat-tolerant than those with other species of endosymbiont (Berkelmans and Van Oppen, 2006; Silverstein et al., 2015), and Durusdinium symbiosis is known to be more abundant on reefs after severe bleaching and bleaching-related mortality events (Baker et al., 2004; LaJeunesse et al., 2009; Kemp et al., 2014; Buzzoni et al., 2026).

Though we do not know the composition of the symbiont community from the corals in our experiment before the 2023 marine heatwave and severe bleaching event, it is highly likely that if they were not primarily Durusdinium before that event, that the survivors switched to that community during or afterwards. It is also likely that the tempered paling/bleaching observed in 2024 was the result of this primarily Durusdinium community providing some protection against temperature profiles that should have caused extensive bleaching and mortality. However, of interest was the single unshaded C. natans coral dominated by Breviolum, which did not exhibit more severe paling than any other unshaded coral. Nevertheless, we suggest that the extremely moderate paling/bleaching experienced by these corals in 2024 was a result of acclimation and Durusdinium colonization from the 2023 event.

We propose two hypotheses for the S:H ratios not distinguishing between the control and shaded corals as the visual bleaching index scores did:

• The location of the sampling on each coral. Sampling was only authorized via permitting on the edge of the corals; these areas exhibited less paling than the upper surfaces, particularly for C. natans. Because edges were experiencing some degree of self-shading, the impact of the artificial shades may not have been as pronounced at the sample locations. We recommend that future permitting allows and encourages sampling at the most impacted parts of colonies to more accurately assess symbiont communities during bleaching events.

• Visual paling was caused or exacerbated by the loss of chlorophyll within symbionts rather than the loss of symbionts themselves. Falkowski and Dubinsky (1981) compared zooxanthellae densities and chlorophyll concentrations in light- and shade-adapted corals. Though zooxanthellae concentrations between the two groups were nearly identical, shade-adapted corals had 3.7 times more chlorophyll a per symbiont cell than light-adapted corals. However, this response does not appear to be universal, as Tolleter et al. (2013) found that bleaching caused not only loss of symbionts, but also decreased chlorophyll a fluorescence in the remaining ones. Combining S:H ratio and chlorophyll assessment methodologies could represent a future line of research in understanding bleaching physiology.

It is possible that one or both factors resulted in the visually apparent differences in coral color between shaded and bleached corals not also being reflected in the symbiont:host ratios. We suspect that had bleaching been more pronounced (more thermal stress and/or less Durusdinium) and/or samples had been collected from the more affected areas of the colony, these trends would have matched.

These metrics suggest that deployment of shades is an effective tool to minimize or prevent paling/bleaching of Caribbean corals. We show its effectiveness during even minor paling/bleaching events, and its usefulness even if deployed after paling has already commenced. Though no bleaching-related mortality was observed on control or experimental corals, minimizing or preventing paling is likely to still have positive health benefits for corals, including lower susceptibility to disease (Muller et al., 2018) and higher reproductive capacity during the subsequent year (Levitan et al., 2014). Though these corals exhibited only minor paling/bleaching, most likely because of their Durusdinium endosymbionts, minimizing or preventing bleaching for corals hosting non-Durusdinium symbionts is likely to be even more beneficial. The documented trade-offs for corals hosting Durusdinium include reduced growth (Little et al., 2004; Jones and Berkelmans, 2010; Cunning et al., 2015a) and potentially higher disease susceptibility (Little et al., 2004). Presumably if shading could reduce the bleaching susceptibility of corals hosting other symbionts, they would be less likely to switch to Durusdinium, thus reducing the associated longer-term negative impacts.

The occurrence of severe weather during the event allowed us to opportunistically assess the recovery of a colony following severe abrasion. The two passing tropical storms were intense enough to dislodge and move two of the assessment corals. During tropical storm Helene, two of the experiment’s C. natans corals (one large, one small) were found approximately 30 m north of their previous positions. Both were resting upside down, and while the smaller coral did not suffer any tissue loss, the larger one (coral 890) had little to no tissue remaining on its ridges (Figure 6). Both were returned to their original locations. During the second tropical storm (Milton), the smaller coral again dislodged and could not be found again. Coral 890 remained in place, and our regular monitoring allowed us to track its recovery. Surprisingly, within 2 weeks of the coral being righted, the tissue remaining between the ridges had reformed to connect the polyps and create an intact, surviving tissue isolate.

From the standpoint of shade structure maintenance, the structures held up through the duration of the experiment and required minimal maintenance. Though structures were specifically removed for the two large storm events, there were other weather events through which they persisted with no damage. There was no evidence of deterioration, and all shade structures remained suitable for additional deployments in future years. All 80 eye screws remained firmly attached to the substrate, and yet were easy to remove at the end of the experiment. All structures maintained buoyancy and did not impact the targeted corals nor the surrounding reef. Fouling of the shade cloths was minimal, and the bi-weekly cleaning only required the removal of a light dusting of sediment and thus could have been less frequent. The supply costs for each shade structure were approximately $30 US.

As potential cons, deployment of shades in some locations could become hazards to fishing or boat navigation, carry low but not zero risk of entanglement of marine species, and could be considered unsightly by in-water stakeholders. Though the installation of eye screws was minimally invasive, this could be more challenging or detrimental on high coral cover reefs lacking bare substrate. The materials used (stainless steel, nylon, HDPE, and PVC) are highly inert and widely used in marine applications, but it is possible they could have unexpected impacts on marine organisms. It is also possible that shades could affect water movement around the corals while deployed, and are likely to reduce sedimentation, with potential unknown impacts. Though supply costs are relatively low, boat and personnel time for building, deployment, maintenance, and retrieval of the shades could be substantial in some locations (though significantly less than equivalent restoration costs if colonies perish; see Supplemental Discussion for an estimated cost comparison).

Our overall conclusions are that the deployment of shades was a relatively inexpensive, efficient method of minimizing paling/bleaching of corals during a hyperthermal event. We demonstrated that deployment of shades even after paling has begun can still have positive impacts on shaded corals by reversing visual paling. We expect that these impacts would have been more pronounced had shades been applied earlier, heat stress or irradiance been higher, or symbiont communities been less heat tolerant. Though deployment and retrieval of the shades was relatively simple, the labor time and costs do restrict the ability of this particular methodology to be used at a reef scale. Rather, it could be used for focused efforts on priority corals, or could be scaled up to protect larger areas. Though interventions like these do come with potential risks both known and unknown, we suggest that the rapidly declining status of reefs warrants consideration of increased management in order to prevent continued loss of survivors.

References

• 1

  BakerA.CunningR. (2016). Bulk gDNA extraction from coral samples. Protocols.io. doi: 10.17504/protocols.io.dyg7vv

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 2

  BakerA. C.StargerC. J.McClanahanT. R.GlynnP. W. (2004). Corals' adaptive response to climate change. Nature430, 741–741. doi: 10.1038/430741a

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 3

  BeaversK. M.Van BurenE. W.RossinA. M.EmeryM. A.VegliaA. J.KarrickC. E.et al. (2023). Stony coral tissue loss disease induces transcriptional signatures of in situ degradation of dysfunctional Symbiodiniaceae. Nat. Commun.14, 2915. doi: 10.1038/s41467-023-38612-4

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 4

  BerkelmansR.Van OppenM. J. (2006). The role of zooxanthellae in the thermal tolerance of corals: a ‘nugget of hope’ for coral reefs in an era of climate change. Proc. R. Soc. B. Biol. Sci.273, 2305–2312. doi: 10.1098/rspb.2006.3567

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 5

  ButcherineP.TagliaficoA.EllisS. L.KelaherB. P.HendricksonC.HarrisonD. (2023). Intermittent shading can moderate coral bleaching on shallow reefs. Front. Mar. Sci.10, 1162896. doi: 10.3389/fmars.2023.1162896

  • CrossRef
  • Google Scholar
  • View reference in article

• 6

  BuzzoniD.Van NynattenA.CunningR.BaumJ. (2026). Persistent legacy effects of marine heatwaves on coral symbioses. Global Change Biol.32, e70818. doi: 10.1111/gcb.70818

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 7

  CoelhoV.FennerD.CarusoC.BaylesB.HuangY.BirkelandC. (2017). Shading as a mitigation tool for coral bleaching in three common Indo-Pacific species. J. Exp. Mar. Biol. Ecol.497, 152–163. doi: 10.1016/j.jembe.2017.09.016

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 8

  CorreaA.BrandtM.SmithT.ThornhillD. J.BakerA. (2009). Symbiodinium associations with diseased and healthy scleractinian corals. Coral Reefs28, 437–448.

  • Google Scholar
  • View reference in article

• 9

  CunningR. (2018). “ SteponeR: R package for importing qPCR data from StepOne™ Software,” in Zenodo.

  • Google Scholar

• 10

  CunningR.BakerA. C. (2013). Excess algal symbionts increase the susceptibility of reef corals to bleaching. Nat. Clim. Change3, 259–262. doi: 10.1038/nclimate1711

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 11

  CunningR.BakerA. C. (2014). Not just who, but how many: the importance of partner abundance in reef coral symbioses. Front. Microbiol.5. doi: 10.3389/fmicb.2014.00400

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 12

  CunningR.GilletteP.CapoT.GalvezK.BakerA. (2015a). Growth tradeoffs associated with thermotolerant symbionts in the coral Pocillopora damicornis are lost in warmer oceans. Coral Reefs34, 155–160. doi: 10.1007/s00338-014-1216-4

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 13

  CunningR.LenzE. A.EdmundsP. J. (2024). Measuring multi-year changes in the Symbiodiniaceae algae in Caribbean corals on coral-depleted reefs. PeerJ12, e17358. doi: 10.7717/peerj.17358

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 14

  CunningR.SilversteinR. N.BakerA. C. (2015b). Investigating the causes and consequences of symbiont shuffling in a multi-partner reef coral symbiosis under environmental change. Proc. R. Soc. B. Biol. Sci.282, 20141725. doi: 10.1098/rspb.2014.1725

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 15

  DonnerS. D.SkirvingW. J.LittleC. M.OppenheimerM.Hoegh‐GuldbergO. (2005). Global assessment of coral bleaching and required rates of adaptation under climate change. Global Change Biol.11, 2251–2265. doi: 10.1111/j.1365-2486.2005.01073.x

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 16

  EatonK. R.BanisterR. B.ArickL. N.HallE. R.MullerE. M. (2025). Genotype and symbiont composition rather than environment influence susceptibility to stony coral tissue loss disease in coral restoration broodstock. Sci. Rep. doi: 10.1038/s41598-025-29292-9

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 17

  FalkowskiP. G.DubinskyZ. (1981). Light-shade adaptation of Stylophora pistillata, a hermatypic coral from the Gulf of Eilat. Nature289, 172–174. doi: 10.1357/002224083788520199

  • CrossRef
  • Google Scholar

• 18

  FittW.WarnerM. (1995). Bleaching patterns of four species of Caribbean reef corals. Biol. Bull.189, 298–307. doi: 10.2307/1542147

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 19

  GleasonD. F.WellingtonG. M. (1993). Ultraviolet radiation and coral bleaching. Nature365, 836–838. doi: 10.1038/365836a0

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 20

  IngL. X. (2024). Can miami corals beat the heat? Insights into the algal symbiont assemblages of highly resilient ‘Urban corals’ in southeast Florida.

  • Google Scholar

• 21

  JeansJ.CampbellD. A.HoogenboomM. O. (2013). Increased reliance upon photosystem II repair following acclimation to high-light by coral-dinoflagellate symbioses. Photosynth. Res.118, 219–229. doi: 10.1007/s11120-013-9918-y

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 22

  JonesA.BerkelmansR. (2010). Potential costs of acclimatization to a warmer climate: growth of a reef coral with heat tolerant vs. sensitive symbiont types. PloS One5, e10437. doi: 10.1371/journal.pone.0010437

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 23

  KayanneH. (2017). Validation of degree heating weeks as a coral bleaching index in the northwestern Pacific. Coral Reefs36, 63–70. doi: 10.1007/s00338-016-1524-y

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 24

  KempD. W.Hernandez-PechX.Iglesias-PrietoR.FittW. K.SchmidtG. W. (2014). Community dynamics and physiology of Symbiodinium spp. before, during, and after a coral bleaching event. Limnol. Oceanogr.59, 788–797. doi: 10.4319/lo.2014.59.3.0788

  • CrossRef
  • Google Scholar

• 25

  KlepacC. N.EatonK. R.PetrikC. G.ArickL. N.HallE. R.MullerE. M. (2023). Symbiont composition and coral genotype determines massive coral species performance under end-of-century climate scenarios. Front. Mar. Sci.10, 1026426. doi: 10.3389/fmars.2023.1026426

  • CrossRef
  • Google Scholar

• 26

  LaJeunesseT. C.SmithR. T.FinneyJ.OxenfordH. (2009). Outbreak and persistence of opportunistic symbiotic dinoflagellates during the 2005 Caribbean mass coral ‘bleaching’ event. Proc. R. Soc. B. Biol. Sci.276, 4139–4148. doi: 10.1098/rspb.2009.1405

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 27

  LevitanD. R.BoudreauW.JaraJ.KnowltonN. (2014). Long-term reduced spawning in Orbicella coral species due to temperature stress. Mar. Ecol. Prog. Ser.515, 1–10. doi: 10.3354/meps11063

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 28

  LittleA. F.Van OppenM. J.WillisB. L. (2004). Flexibility in algal endosymbioses shapes growth in reef corals. Science304, 1492–1494. doi: 10.1126/science.1095733

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 29

  ManzelloD. P.CunningR.KarpR. F.BakerA. C.BartelsE.BonhagR.et al. (2025). Heat-driven functional extinction of Caribbean Acropora corals from Florida’s Coral Reef. Science390, 361–366. doi: 10.1126/science.adx7825

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 30

  MellinC.BrownS.CantinN.Klein-SalasE.MouillotD.HeronS. F.et al. (2024). Cumulative risk of future bleaching for the world’s coral reefs. Sci. Adv.10, eadn9660. doi: 10.1126/sciadv.adn9660

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 31

  MieogJ. C.van OppenM. J.BerkelmansR.StamW. T.OlsenJ. L. (2009). Quantification of algal endosymbionts (Symbiodinium) in coral tissue using real‐time PCR. Mol. Ecol. Resour.9, 74–82. doi: 10.1111/j.1755-0998.2008.02222.x

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 32

  MullerE. M.BartelsE.BaumsI. B. (2018). Bleaching causes loss of disease resistance within the threatened coral species Acropora cervicornis. eLife7, e35066. doi: 10.7554/elife.35066

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 33

  NeelyK. L.NowickiR. J.DoblerM. A.ChaparroA. A.MillerS. M.TothK. A. (2024). Too hot to handle? The impact of the 2023 marine heatwave on Florida Keys coral. Front. Mar. Sci.11, 1489273. doi: 10.1101/2024.08.31.610635

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 34

  NOAA Coral Reef Watch (2025). Florida 5 km single-pixel virtual station time series graphs. Available online at: https://coralreefwatch.noaa.gov/product/vs_single_pixel_exp/florida_keys.php (Accessed October 19, 2025).

  • Google Scholar

• 35

  ReimerJ. D.PeixotoR. S.DaviesS. W.Traylor-KnowlesN.ShortM. L.Cabral-TenaR. A.et al. (2024). The fourth global coral bleaching event: where do we go from here? Coral Reefs43, 1121–1125. doi: 10.1007/s00338-024-02504-w

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 36

  SilversteinR. N.CunningR.BakerA. C. (2015). Change in algal symbiont communities after bleaching, not prior heat exposure, increases heat tolerance of reef corals. Global Change Biol.21, 236–249. doi: 10.1111/gcb.12706

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 37

  TagliaficoA.BakerP.KelaherB.EllisS.HarrisonD. (2022). The effects of shade and light on corals in the context of coral bleaching and shading technologies. Front. Mar. Sci.9, 919382. doi: 10.3389/fmars.2022.919382

  • CrossRef
  • Google Scholar

• 38

  TolleterD.SenecaF. O.DeNofrioJ. C.KredietC. J.PalumbiS. R.PringleJ. R.et al. (2013). Coral bleaching independent of photosynthetic activity. Curr. Biol.23, 1782–1786. doi: 10.1016/j.cub.2013.07.041

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 39

  WinterR. N. (2017). Environmental controls on the reassembly of Symbiodinium communities in reef corals following perturbation: implications for reef futures under climate change. (doctoral dissertation). University of Miami.

  • Google Scholar