Near complete local extinction of iconic anemonefish and their anemone hosts following a heat stress event

Abstract

A fundamental question in modern-day ecology is: How will populations and their interactions respond to a rapidly changing climate? Documenting local collapses of ecologically and economically important populations can offer insight into broader patterns of decline. Here, we monitored anemonefish (Amphiprion bicinctus) and their host sea anemones (Radianthus magnifica) on three central Saudi Arabian Red Sea reefs from 2022 to 2024, including a 2023 marine heatwave that peaked at a Degree Heating Weeks (DHW) value of ~22 °C-weeks. Across all reefs, we observed a sequence of 100% anemone bleaching, 94.3–100% anemonefish mortality, and 66.4–94.1% anemone mortality. We compare these findings to other recent Indo-Pacific heat stress events of varying intensity, where similar declines were not observed. Our study highlights the vulnerability of mutualistic reef species to extreme heat and suggests that such events may drive local, if not regional, extinctions of ecologically important symbioses.

Introduction

The impact of heat stress on coral reefs has become more severe and more frequent over the past decades, causing massive coral losses globally^1–3. The decline in coral coverage has an obvious impact on the populations of other reef organisms that rely on these ecosystems to survive and reproduce, such as many species of fish and invertebrates^4–7. In the last 20 years, multiple “bleaching events”, where reef-building corals and other cnidarians react to heat stress through loss of their symbiotic algae, have harmed coral reefs worldwide³.

Bleaching events are frequently caused by marine heatwaves (MHWs), which are prolonged periods of abnormally high sea surface temperatures¹. These events pose a significant threat to biodiversity and can result in widespread coral bleaching, changes in species distribution, and declines in fisheries⁸. A key metric for assessing the intensity of these events is Degree Heating Weeks (DHW), which measures the cumulative thermal stress on marine organisms⁹. DHW is calculated by combining the intensity of daily temperature extremes and the total time that daily temperatures exceed the bleaching threshold over the previous 3 months¹⁰. According to NOAA, coral bleaching is likely at 4 °C-weeks, while widespread bleaching and mortality are likely at 8 °C-weeks or higher¹⁰. The impact of rising ocean temperatures has stretched to regions that were long considered thermal refugia for corals, such as the Red Sea^11–14. Notable major MHW events in the Saudi Arabian Red Sea were reported in 1998, 2010, 2015, 2020 and 2023^13,15–18.

While the impacts of these events and relevant thermal thresholds have been broadly studied for corals (e.g.,¹⁹), the sea anemones, which are also affected by bleaching^20–23 have received comparably less attention. Yet sea anemones are not simply small or soft corals, rather they exhibit distinct physiological responses, recovery trajectories, and symbiotic dependencies^24,25. Moreover, their ecological roles include hosting anemonefish, and serving as key habitat for crustaceans, cleaning mutualists, and a range of fish species^26,27. Understanding the thermal sensitivity of these cnidarians therefore adds complementary insight into reef resilience and climate vulnerability. Several studies have examined the impacts of anemone bleaching in situ (e.g.,^23,25,28,29) and in the laboratory (e.g.,²⁴), but we are unaware of any studies that have examined the short- and long-term impacts of a heat stress event as strong and as prolonged as the 2023 MHW event on sea anemones and resident anemonefish.

Like stony corals, anemones are benthic cnidarians, some of which form an obligate symbiosis with intracellular microalgae known as zooxanthellae. While the algae provide sugar and oxygen at a cellular level, the host anemones provide protection, a stable position in the water column, and other valuable products from their metabolism to the symbionts^30,31. Also like stony corals, host anemones may expel these symbionts and “bleach” during periods of stress, especially heat stress^24,25,32. Bleaching occurs when the photosynthetic algal cells (Symbiodiniaceae), located within anemones’ gastrodermal tissue, are lost from the anemone. Without these photosynthesizing cells and their pigments, the host anemone loses its color and one of its main sources of energy^33–35. If the anemone does not regain these algae, starvation occurs, and that prolonged state can lead to mortality²⁸.

Sea anemones and anemonefishes form one of the most well-known mutualisms in the ocean. All ten anemone species that host anemonefish are susceptible to bleaching, and their bleaching has been well-documented across the Indo-Pacific^20–23. Given their vulnerability, anemone bleaching events not only threaten the anemones themselves but also have consequences for the anemonefishes and other associates that depend on them. The anemones protect their symbiont anemonefish and their egg clutches with their nematocyst-laden tentacles and provide nutrients to the fish³⁶. In return, anemonefish provide a range of services to their anemones, including oxygenation, predator defense, and nutrients^30,37–42. On top of affecting the fitness of the anemone itself, heat stress and associated anemone bleaching events have a range of negative consequences for the anemonefish that reside within their tentacles. Indeed, heat stress events have been shown to decrease fecundity, increase stress, and cause other behavioral changes for the anemonefish living in bleached sea anemones (e.g.,^29,43–45). Notably, recent work has even shown that heat stress events cause anemonefish to shrink in size⁴⁶. While we know that anemone bleaching can affect both anemone and anemonefish populations, the mechanisms, timing, DHW thresholds, and overall impacts of these events on the mutualism relationship between these two organisms are still understudied. As an important ecological component of reef systems (e.g., habitat providers for dozens of different taxa, including various species of cleaning crustaceans and wrasses, juvenile reef fishes, and others), and an iconic model mutualism studied across many disciplines, host sea anemone responses to heat stress events in situ are important to record (^{26,27,29,40,44}).

Here, we investigate the temperature thresholds and impacts of related heat stress and subsequent bleaching on this model marine mutualism system across three surveyed reefs in the central Red Sea, along the coast of Saudi Arabia. In the Red Sea, the endemic anemonefish Amphiprion bicinctus occupies six host sea anemones: Entacmaea quadricolor, Radianthus magnifica, Radianthus crispa, Radianthus aurora, Stichodactyla mertensii, and Stichodactyla haddoni. In this region, the magnificent sea anemone, R. magnifica, is the most common host species for A. bicinctus (reviewed in detail in ref. ⁴⁷ and ref. ⁴⁸). These populations are also important draws for ecotourism, in part due to Pixar’s Finding Nemo and subsequent presence of anemonefish in pop culture and aquaria. Our study documents severe bleaching in R. magnifica anemones, followed by a collapse in population size of anemonefishes and a subsequent collapse in population size of anemones in the central Red Sea, and places these results in the context of other studies that have investigated effects of heat stress on anemones and anemonefishes.

Results

Anemone bleaching survey—breakdown of the mutualism between anemones and zooxanthellae during the 2023 marine heatwave

Across three surveyed reefs on an inshore-offshore gradient in the central Red Sea, healthy, surveyed R. magnifica anemones bleached after the onset of the 2023 MHW, which peaked at 22.26 °C-weeks (Fig. 1). Healthy tagged anemones first bleached in early September 2023; by mid-October 2023, 100% of tagged anemones were bleached (Fig. 2A). Bleached anemones remained in a visually similar condition throughout October, November, and December (Fig. 2B, C). In January 2024, areas of pigment began to return to some tagged anemones (Fig. 2D). However, the body condition of most of these individuals did not appear healthy or typical: anemones, regardless of signs of re-pigmentation, appeared deflated and shrunken compared to the beginning of the survey.

Fig. 1. Study site locations and visual context of heat stress and bleaching.

A Map of three surveyed reefs—Coral Probiotics Village (CPV) Reef (Garcias-Bonet et al.⁸⁵), Rose Reef, and Tahala Reef—along an inshore–offshore gradient in the central Red Sea. Basemap imagery in (A) is sourced from satellite imagery, visualized using ArcGIS Online. B Heat stress visualization showing Degree Heating Weeks (DHW, °C-weeks) in the region for 2021, 2022, and 2023, as reported by NOAA from satellite-derived sea surface temperature. C Bleached Radianthus magnifica with resident Amphiprion bicinctus. Photo by Morgan Bennett-Smith. D Healthy R. magnifica with A. bicinctus. Photo by Morgan Bennett-Smith.

Fig. 2. Visual progression of bleaching and partial recovery in tagged Radianthus magnifica anemones.

A: CPV Reef, October 2023 - initial, early full bleaching in a tagged R. magnifica. B: Rose Reef, November 2023 - mid-stage bleaching with continued pale tentacles. C: Tahala Reef, December 2023 - late-stage bleached individual. D: CPV Reef, January 2024 - partial repigmentation observed, indicating early signs of visual recovery.

By the final survey time point, in May 2024, 100% of remaining tagged anemones had full pigment, but, as in January 2024, were in poor overall condition. This was evidenced by a lack of inflation, shriveled tentacles, small size, and, in some cases, gaping and extrusion of internal structures through the mouth.

Anemone mortality following bleaching during the 2023 marine heatwave

As reported above, following the 2023 bleaching event, 100% of the tagged anemones included in this study were bleached. By the final survey in May 2024, cumulative anemone mortality across Red Sea reefs had reached 78.5% ± 14.2 (mean ± SD; Fig. 3A). Our first analysis focusing on anemone presence revealed a significant decline in anemone presence by 2024 (Firth logistic regression: β = –6.72 ± 1.40, p < 0.001; Supplementary Table 1). Our second analysis focusing on anemone survival revealed a significant decline in anemone survival in 2023-2024 (Firth logistic regression: β = –8.44 ± 1.91, p < 0.0001; Supplementary Table 2), and a marginally significant decline in anemone survival with increasing anemonefish group size (Firth logistic regression: β = –1.20 ± 0.69, p = 0.053; Supplementary Table 3). No significant differences in survival were found between reefs in either analysis (Tables S1–S2). The percentage of anemones remaining by reef surveyed were 25% (95% CI [4, 46]) in the CPV, 33.6 (95% CI [25, 42]) in Rose reef and 5.9% (95% CI [0, 17]) in Tahala, underscoring the severity of the 2023–2024 heat stress event in the central Red Sea. Although 21.5% ± 14.2 of individuals showed partial recovery, none of the monitored anemones were classified as fully healthy by the end of the survey period (Fig. 3A).

Fig. 3. Anemone and anemonefish decline across three surveyed reefs in the central Red Sea during and after the 2023 marine heatwave.

A Anemone health and mortality across all three surveyed reefs, 2022–2024. In December 2022, all anemones were healthy and alive. In December 2023, all were bleached but alive. By May 2024, 21.5% of the original anemones remained and all were partially recovered, while reef-specific mortality rates were 75.0% at CPV, 66.4% at Rose Reef, and 94.1% at Tahala Reef. When pooled across reefs, this equates to a total mortality of 78.5% ± 14.2% SD. B Anemonefish mortality across all three reefs, 2022–2024. The error bars represent 95% confidence intervals of the emmeans model estimates (Supplementary Table 6). In December 2022, all fish were present and healthy. In December 2023, fish per anemone declined by 87.7% on the bleached—but still living—population of tagged anemones. By May 2024, more than 96% of the original fish had disappeared. C Photographs of one group of tagged anemones on CPV Reef across survey years, showing progression from healthy (2022) to bleached (2023) to dead (2024).

Fish mortality and breakdown of the interaction between anemones and fish during the 2023 marine heatwave

Following the onset of widespread anemone bleaching in 2023, anemonefish populations collapsed across all three surveyed reefs. By December 2023, only 12.3% of the originally observed anemonefish remained, and by May 2024, just 7 individuals (4.2% of the initial 168 fish) were detected across all sites (Fig. 3B).

To quantify this decline, we first modeled changes in anemonefish group size across all tagged anemones, including anemones that had died. This mixed-effects model revealed a highly significant effect of year on group size (p < 0.001), with no significant effect of reef or a year × reef interaction (Supplementary Table 3, Supplementary Table 4). Adjusted mean group sizes declined by >80% between 2022 and 2023. In CPV, group size dropped from 1.49 (95% CI: 1.17–1.81) to 0.11 (95% CI: −0.21 to 0.43), with similar declines at Rose Reef (1.13–0.12) and Tahala Reef (1.16–0.22) (Supplementary Table 5).

To test whether group size collapse occurred even among surviving hosts, we repeated the analysis using only observations where the host anemone was alive (filtered model). This analysis again showed a strong year effect (p < 0.001), confirming that group sizes declined even within persistent anemones. No significant effect of reef or interaction was observed (Supplementary Table 6). By 2024, fish group sizes approached zero across all sites (Supplementary Table 7, Supplementary Table 8), reflecting the near-total collapse of the mutualism, even in cases where hosts remained.

Other reports of anemone bleaching and mortality during different marine heatwaves

In comparison to the monitored central Red Sea population, we reviewed recent literature sources to identify similar anemone bleaching and mortality events in other localities. Six previous studies were identified that included proportion or percentage of bleached anemones, DHW or an equivalent thermal stress metric, and mortality rate of the anemones under heat stress (Table 1). None of these studies reported mortality above 3%, in comparison to ~78% in the central Red Sea.

Table 1.

Anemone mortality across host anemone populations from previous studies compared to the present study

Study	Location	Year	Anemone % Bleached	Anemone Mortality	Max Degree Heat Weeks
This Study	Saudi Arabia	2023	100%	78.19%	22 °C-weeks
Versteeg et al. 2025	Papua New Guinea	2023	57%	1.92%	18 °C-weeks
Hobbs et al. 2013	Christmas Island	2010	2.6%	0%	14 °C-weeks
Hobbs et al. 2013	Papua New Guinea	2009	44.1%	1.1%	13 °C-weeks
Steinberg et al. 2022	Lord Howe, Australia	2019	100%	0%	9 °C-weeks
Hayashi and Reimer 2020	Okinawa	2017	14%	0%	3.1 °C-weeks
Hayashi and Reimer 2020	Okinawa	2016	16.7%	3%	2.42 °C-weeks

Studies are presented in declining order of peak Degree Heat Weeks values.

Discussion

Our results show severe bleaching of anemones, subsequent mortality of anemonefish, and ultimately mortality of sea anemones in the Saudi Arabian central Red Sea during and after the 2023 heating event. In other regions that also experienced major, though not quite as extreme, heating events, host anemones bleached but did not experience the same levels of DHW or mortality (e.g., Kimbe Bay, Papua New Guinea). Together with other recent reports of mass fish mortality¹⁸ our study shows that the Red Sea may not be the thermal refugium it was once pronounced to be. Current levels of climatic stress on the world’s oceans are already severe enough, at least in some regions, to lead to the breakdown of mutualisms and collapse of populations.

The Red Sea, often regarded as a thermal refugium for reef-building corals and cnidarians^11–14, is among the warmest marine environments on earth. Temperatures in the Red Sea can reach 34 °C Seawater Surface Temperature (SST) in the summer and regularly represent modeled future temperature values predicted for most tropical reefs over the next 50 years¹³. Corals and other organisms in this region are thought to have adapted to these relatively extreme thermal conditions, with bleaching events less frequent than in other localities, even during DHW events greater than 15 °C-weeks¹³. Consequently, it has been hypothesized that the impacts of warming ocean temperatures would be lessened in this region^11,49. However, the increasing frequency and severity of bleaching events, such as the ones caused by the record heatwaves in 2023 and 2024, challenge this notion. Notably, this recent event may be partly explained by the exceptionally high DHW recorded in certain parts of the Red Sea, which exceeded those of other surveyed regions, including the also threatening 18 DHW in Papua New Guinea. This suggests that even thermally adapted cnidarian communities may be vulnerable when exposed to anomalously prolonged or intense heat stress.

First, we found high levels of anemone bleaching. During other heat waves, high levels of anemone bleaching have similarly been recorded but few have observed such a high percentage, across multiple reefs. In 2019, Steinberg et al. found 100% bleaching on Entacmaea quadricolor anemones surveyed using photo quadrats at Lord Howe Island, Australia⁵⁰. In 2023, Versteeg et al. observed a population of 104 R. magnifica anemones across eight inshore reefs in Kimbe Bay, Papua New Guinea, of which 56.73% bleached⁴⁶. Beyond these studies, however, few have reported levels of anemone bleaching above ~40%. The high bleaching levels in our study, across three different parts of the reefscape, from inshore to offshore, with different depths and other biotic and abiotic factors, indicate that this event was particularly severe, reflecting the highest DHW observed across the other areas where anemones and anemonefish bleaching and mortality were surveyed.

Second, we found extremely high levels of fish mortality following bleaching but preceding anemone mortality. There are several plausible hypotheses for this finding: (i) the anemone bleaching is the direct cause of the fish mortality; (ii) heat stress is the direct cause of the fish mortality, or (iii) a combination of these two factors is the cause of the fish mortality. To disentangle the mechanisms at play here, heat stress and anemone bleaching state would need to be experimentally manipulated to break the correlation between them. Several authors have attempted to investigate this. For example, Cortese and colleagues did something similar by experimentally manipulating bleaching state in situ and observing multiple physiological responses in symbiont anemonefish in Moorea, during a period of normal temperature on the reef⁵¹. These authors found several key pieces of supporting evidence for hypothesis (i) and/or (iii), above. First, Cortese and colleagues found that, while the standard metabolic rate of anemonefish residing in bleached anemones decreased over time compared to fish from healthy anemones, the growth rate of fish from bleached anemones was significantly lower—suggesting that animals residing in bleached hosts are at an energetic disadvantage. Second, they found that fish from bleached anemones spent more time out of their anemones, suggesting a greater need to forage in the water column (and greater risk of predation) despite being less active overall⁵¹. Somewhat in contrast to Cortese and colleagues, another study found that, in a laboratory trial, the metabolic rate increased in anemonefish occupying bleached anemones compared to those occupying healthy anemones with temperature held constant, which the authors suggested may be due to increased stress in anemonefish on bleached anemones⁴³. In both studies, anemonefish in bleached anemones faced potential physiological disadvantages compared to those in healthy anemones, with temperature held constant. However, neither study directly observed differences in anemonefish mortality as a consequence solely of anemone bleaching, so it is still not possible to rule out the impacts of heat stress or combined heat and bleaching stress on our survey population. Based on prior work and our observations, it seems likely that the anemone bleaching state is, at the very least, an extremely important component in anemonefish survivorship during a stress event. The near-total disappearance of anemonefish by May 2024 raises questions about the long-term survival of anemonefish populations if such bleaching events become more frequent and widespread.

The high levels of anemonefish mortality we observed in the Red Sea are particularly concerning considering previous findings that A. bicinctus populations in this region are already lower than those of other Amphiprion species elsewhere^52,53. Supporting this, average group sizes before the bleaching event were small across our surveyed reefs: 1.49 fish were recorded per anemone on CPV Reef, 1.13 on Rose Reef, and 1.16 on Tahala Reef. These values are markedly lower than those reported for other Amphiprion species at different sites—for example, 3.33 Amphiprion percula per anemone, 3.82 Amphiprion perideraion per anemone, 3.2 Amphiprion clarkii per anemone, and 2.14 Amphiprion chrysopterus per anemone in Madang, Papua New Guinea⁵⁴. Since successful reproduction requires at least two fish per anemone, and additional group members may enhance both reproductive success^38,55 and anemone health^56,57, it is evident that Red Sea populations were already operating below optimal social thresholds. The recent mortality observed is likely to further compromise population growth, underscoring the urgent need for targeted management and conservation strategies for both Amphiprion bicinctus and their host anemone Radianthus magnifica.

Third, we found high levels of anemone mortality following the loss of mutualistic partners. As in the case of anemonefish mortality following anemone bleaching, the direct mechanisms linking anemone mortality following anemonefish mortality during a heat stress event are difficult to disentangle. We propose, therefore, several plausible hypotheses: (i) that the mortality of anemones we observed was a consequence of the prolonged loss of their mutualist partners, both fish and endosymbiont, (ii) that heat stress directly caused their mortality, and (iii) that some combination of (i) and (ii) resulted in mortality. In prior studies, the absence of the two most important mutualists, the anemonefish and the intracellular zooxanthellae, has been shown to decrease the chances of the anemones themselves to survive. For example, the absence of resident fish makes the anemone easier prey for other organisms⁵⁶, decreases growth rate, and has been directly linked to generally greater mortality⁵⁷. Further, resident fish spending more time in the water column outside of the anemone⁵¹, has been shown to reduce anemone growth⁵⁸. Intriguingly, our data also suggests a marginally negative effect of anemonefish group size on anemone survival during this heat stress event, raising the possibility that the relationship changes from a +/+ interaction to a +/- interaction under heat stress. However, the severity of this heat event was too strong in most cases for both partners; of an original population of 143 anemones, 39 anemones survived, and of those 39 anemones, only five had fish symbionts as of the final survey count.

The lack of endosymbiosis—the “bleached” condition—also has been shown to impact the host resistance and resilience in several ways. Bleached anemones may be forced to adjust the composition of their symbionts, at unknown cost, and prior work with other species has indicated that the bleached anemone condition results in a significant decrease in glucose levels and a significant increase in glycerol, suggesting that bleached anemones degrade lipids to compensate for the loss of symbionts^59,60. The thresholds for thermal collapse remain poorly defined and may be influenced by multiple factors, including symbiont retention, energy reserves, and interactions with anemonefish. Future laboratory studies are needed to determine when and how these processes drive mortality. While anemonefish may enhance host survivorship through multiple mechanisms under normal conditions, the endosymbiotic algae remain a vital source of organic carbon. Heat stress alone can therefore be sufficient to cause mortality, though additional mechanisms may also play a role.

Of growing interest, it has been reported that the structure of microbial communities associated with sea anemones, such as Anthopleura elegantissima, is tightly linked to their symbiotic state rather than to other factors⁶¹. The same authors found significantly higher species richness values for microbial communities associated with anemones hosting Elliptochloris marina. Similar to the shifts observed in other Cnidaria and across different hosts^62,63, the switch from a stable, beneficial-dominated microbiome associated with a healthy host to an unhealthy/pathogenic microbial community (i.e., dysbiosis)^64,65 associated with sick hosts might contribute to the stress susceptibility of the anemones. This dysbiotic process associated with the bleached state can even be the ultimate cause of mortality, although deeper research investigations are required to test this hypothesis. If confirmed, microbial therapies already successfully applied to protect other cnidarian hosts during heat stress^66–69, including in situ trials⁷⁰ that promoted the beneficial restructuring of coral-associated fish⁷¹, represent a promising strategy to be tested on anemones (and fish) as well.

Finally, we show that high levels of anemone mortality occur at very high DHW thresholds. According to a widely cited NOAA standard, coral bleaching is likely at 4 °C-weeks and widespread bleaching and mortality are likely at 8 °C-weeks or greater (e.g.,^10,72). These thresholds, however, were developed based on coral responses, and may not reliably capture stress responses in other symbiotic cnidarians such as sea anemones. Furthermore, a recent Nature study revealed major inconsistencies in how heat stress is calculated across studies and modeling efforts, particularly between DHW and Degree Heating Months (DHM)⁷³. These findings underscore the uncertainty surrounding standardized thermal stress metrics and highlight the risk of over- or underestimating biological impacts when extending coral-derived thresholds to other taxa. At least for this species and at these sites, mortality occurred at considerably higher DHW values than the conventional thresholds used for reef-building corals, suggesting anemones may possess different thermal tolerances or failure points that remain poorly defined.

While both the Red Sea and regions such as Kimbe Bay in Papua New Guinea experienced elevated thermal stress during the 2023 marine heatwave, the ecological outcomes for anemone populations were markedly different. A key explanatory factor may be the magnitude and duration of heat exposure reaching some critical threshold. Red Sea sites experienced thermal anomalies with DHW values exceeding 21, far above commonly recognized bleaching thresholds. For corals, nonlinear responses to heat stress have been well documented—for example, on the Great Barrier Reef, coral cover loss was minimal at 0–3 °C-weeks but increased to 40% at 4 °C-weeks, 66% at 8 °C-weeks, and over 80% at 9 °C-weeks or more⁷⁴. Considering the duration of heating, our study population of anemones remained bleached for 5–6 months before partial recovery, longer than any previous bleaching event we have observed in this region. This extended duration of bleaching during this heat stress event in the Red Sea may play a critical role in explaining the observed mortality, potentially compounding the physiological stress imposed by high DHW values. Considering the magnitude of heating, even seemingly small differences in thermal exposure, such as between 18 and 22 °C-weeks, may represent biologically critical tipping points, beyond which physiological impacts may compound and shift organisms from bleaching toward mortality. Collectively, the exceptional intensity and prolonged duration of heat stress in the Red Sea likely drove the widespread anemone mortality observed in this study. More investigations should be performed to further test the importance of heating duration and magnitude and to determine refined DHW values specific to anemone stress and mortality.

In general, sea anemones appear to tolerate much higher DHW values before mortality occurs compared to corals. In our review of literature sources that include temperature, bleaching, and mortality data for anemonefish-hosting sea anemones in situ, anemones experienced DHW values of 9, 13, 14, and even 18 °C-weeks without mortality above 3%. This suggests that their thermal thresholds may be considerably higher. One potential explanation for this resilience is the ability of many anemones to rely more on heterotrophic feeding. This has been demonstrated in some non-host anemone species^34,35 and likely applies to host species as well. If anemones can maintain sufficient energy intake during bleaching, then the duration of the thermal stress becomes a key determinant of survival, with mortality occurring only when the bleaching persists longer than their ability to compensate heterotrophically.

The regional dynamics documented for coral bleaching, including differences in heat tolerance, recovery trajectories, and environmental buffering (e.g.,^75–78), almost certainly apply to sea anemones as well. Yet anemones remain significantly less studied than corals, and species- and location-specific responses are still poorly understood. While we did not directly assess local environmental factors such as hydrodynamics and historical bleaching patterns, these may be important contributors to variation in bleaching severity and recovery potential.

Sea anemones are important components of coral reefs in the Red Sea and elsewhere. They provide protection, foraging grounds, cleaning stations, breeding locations, and more, for different reef organisms^{30,37–42,79}. As with coral associates, the mortality of these important species has a cascading effect on the reef. Even though our results in the central Red Sea were consistent across three reefs on an inshore-offshore gradient, a broader survey in other regions of the Red Sea should be conducted to evaluate if this is a broad Red Sea pattern or if there are other more localized stress factors at play. If Saudi Arabian Red Sea anemonefish populations are already lower than in other places, the additive stress of host anemone bleaching could constitute a major threat to the conservation of these species^52,53. More studies are needed to evaluate these important populations and re-evaluate their conservation status at large, expanding the surveys worldwide. As DHW continues to increase globally, ecosystem conservation and restoration initiatives may need to be considered to preserve these populations in affected areas^1,18,19.

Methods

Study population

In December 2022, we tagged 143 visually healthy host anemones (Radianthus magnifica) on three reefs (Coral Probiotic Village (CPV), n = 16; Rose Reef, n = 110; Tahala Reef, n = 17) in the central Red Sea. Our goal was to tag every anemone on the selected reef sites, but this was likely not feasible on larger reefs. For example, CPV Reef is part of a larger reef complex called Al Fahal Reef that stretches for several kilometers; we tagged the entire population of anemones within one part of this reef that is known as the Coral Probiotics Village, but other anemones likely existed on different, un-surveyed parts of the reef complex. On Rose Reef, we were able to census the entire population from a depth of ~18 m to the surface, because Rose Reef is smaller and can be circled in one SCUBA dive. At Tahala Reef, we censused a haphazardly-selected area on the southern end of the reef. These reefs are located on an inshore-offshore gradient, ranging from a few hundred meters from shore (Tahala Reef) to several kilometers offshore (Rose Reef; Fig. 1A). We monitored these anemones for two years, from 2022 to 2024. We conducted periodic visual surveys, noting bleaching conditions in particular. Specifically, surveys were conducted in December 2022, September 2023, October 2023, December 2023, January 2024, and May 2024.

Temperature

We used the DHW metric to assess the accumulated heat stress in our surveyed sites from 2021 to 2023. Specifically, we downloaded the NOAA time series datasets reporting daily SST and DHW metric for the “Al Madinah and Makkah” region, corresponding to the Saudi Arabian central Red Sea coast (Fig. 1B; NOAA Coral Reef Watch Version 3.1 Daily Global 5 km Satellite Coral Bleaching Degree Heating Week Product, 2023-2024. College Park, Maryland, USA: NOAA Coral Reef Watch. Data accessed 2024-07-01 at the following links: https://coralreefwatch.noaa.gov/product/vs/gauges/madinah_makkah.php;

Anemone monitoring

Anemones (initial n = 143) were censused and individually photographed to assess anemone health state (e.g., Fig. 2A–C). To confirm the bleaching state, University of Queensland Coral Watch Coral Health Charts⁸⁰ were used and deployed next to each encountered anemone. Anemones were considered bleached if their coloration most closely matched a “2” or lower on the Health Chart.

Fish monitoring

Anemonefish associated with each tagged anemone host were counted and photographed. Anemonefish per anemone of ~12 mm in length were counted on surveys. Estimated <12 mm fish were omitted to prevent bias in surveyor skill—these small fish are ephemeral, hard to detect, and may have been missed by some observers but not others. Initial total number of anemonefish = 168; initial mean group size = 1.17 ± 0.08.

Data visualization, statistics and reproducibility

All analyses were conducted using R (R Core Team) with a sample size or replication of at least 3. The biological replicates included in each analysis are included in the respective figures and tables legend. Data was visualized in R Studio (RStudio Team (2023). RStudio: Integrated Development Environment for R. RStudio, PBC, Boston, MA URL http://www.rstudio.com/) and JMP Pro, Version 18.0.2. (SAS Institute Inc., Cary, NC, 1989–2024).

Anemone data analysis

To evaluate variation in anemone presence across time and reef locations, we used Firth logistic regression implemented via the R package logistf⁸¹. Anemone presence was 100% across all reefs in both 2022 and 2023, followed by a sharp decline in 2024. Because standard logistic regression assumes sufficient outcome variability across combinations of predictor levels, our data exhibited quasi-complete separation, which caused generalized linear models (glm()) to produce unstable coefficient estimates and uninterpretable p-values. To address this, we applied Firth’s bias-reduced penalized likelihood method⁸². In this model, anemone presence (0/1) was the response variable, with “Year” and “Reef” included as predictors. To assess whether the number of anemonefish in a given year (t) predicted host anemone survival to the following year (t + 1), we constructed a logistic regression model with binomial error using the R package lme4⁸³. The response variable was survival from t to t + 1, while the number of anemonefish observed in year t, reef, and time interval were included as predictors. Due to the presence of separation in this model as well, we additionally ran a Firth logistic regression including the same predictors to obtain more reliable coefficient estimates.

Fish data analysis

We conducted two complementary analyses to assess changes in anemonefish population size and group size over time. First, to evaluate changes in group size across all observed anemones (regardless of subsequent survival), we used a linear mixed-effects model implemented with the lmer function from the R package lme4⁸³. In this model, anemonefish group size was the response variable, with “year” and “reef” as fixed-effect predictors and “anemone identity” as a random intercept to account for repeated measurements. This model captures overall temporal trends in population size decline, including the contribution of anemones that eventually died.

Second, to isolate changes in group size among only surviving anemones, we re-fitted the model using a filtered dataset restricted to observations where the host anemone was alive. This model tests whether group sizes collapsed even within persistent hosts.

For both models, we assessed fixed effects using Type III ANOVA (with Satterthwaite’s approximation for degrees of freedom) and evaluated the significance of the random effect using a random-effects ANOVA (ranova). Pairwise comparisons across years and reefs were conducted using the emmeans package⁸⁴, with the Kenward–Roger method applied for small-sample correction of standard errors and degrees of freedom.

Literature review methods—placing our results in context

We conducted a structured literature search using Web of Science and Google Scholar to identify studies examining thermal stress responses in sea anemones. We used combinations of the following keywords: “anemone bleaching”, “bleaching sea anemones”, “anemonefish bleaching”, “heat stress on sea anemones”, “sea anemone thermal tolerance”, “symbiotic cnidarian bleaching”, and “clownfish bleaching”. We downloaded all relevant studies and reviewed their content in detail. However, only those that reported all three of the following parameters were included in our analysis:

1. The proportion or percentage of bleached anemones.

2. The DHW or an equivalent thermal stress metric.

3. The mortality rate of the anemones under heat stress. Data from qualifying studies were extracted and compiled for comparative analysis.

Author contributions

M.B.S., R.P., and H.V. conceptualized the study with support from P.M.B., M.L.B., and S.C. R.P. and P.M.B. funded the study. M.B.S., H.V., M.S., V.P., and N.G. conducted the field work. M.B.S., H.V., N.G., and F.C.G. conducted the data analysis. T.R. and M.V. contributed to the contextualization of results and literature review. M.B.S. wrote the original draft of the manuscript, and all authors contributed to reviewing and editing.

Data availability

The complete dataset used in the analysis is included in the supplementary file with this submission. All statistical analyses and visualizations were conducted in R (v4.3.1) using the following packages: dplyr, tidyr, ggplot2, scales, RColorBrewer, lme4, lmerTest, emmeans, logistf, sjPlot, and performance. The full R code used in this manuscript is available upon request.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s44185-025-00107-4.