Introduction
Coral reefs have been profoundly impacted by a range of anthropogenic pressures, particularly those related to climate change (Spalding and Brown, Reference Spalding and Brown2015; Hughes et al., Reference Hughes, Kerry, Álvarez-Noriega, Álvarez-Romero, Anderson, Baird, Babcock, Beger, Bellwood, Berkelmans, Bridge, Butler, Byrne, Cantin, Comeau, Connolly, Cumming, Dalton, Diaz-Pulido, Eakin, Figueira, Gilmour, Harrison, Heron, Hoey, Hobbs, Hoogenboom, Kennedy, Kuo, Lough, Lowe, Liu, McCulloch, Malcolm, McWilliam, Pandolfi, Pears, Pratchett, Schoepf, Simpson, Skirving, Sommer, Torda, Wachenfeld, Willis and Wilson2017; Reimer et al., Reference Reimer, Peixoto, Davies, Traylor-Knowles, Short, Cabral-Tena, Burt, Pessoa, Banaszak, Winters, Moore, Schoepf, Kaullysing, Calderon-Aguilera, Wörheide, Harding, Munbodhe, Mayfield, Ainsworth, Vardi, Eakin, Pratchett and Voolstra2024). The increasing frequency and severity of extreme weather events, combined with rising ocean temperatures (IPCC, Reference Lee and Romero2023), have led to significant declines in coral health and survival, affecting entire reef ecosystems (Hoegh-Guldberg, Reference Hoegh-Guldberg2000; Spalding and Brown, Reference Spalding and Brown2015; Mies et al., Reference Mies, Destri, Lacerda, Carvalho, Ibanhez, Güth, Luza, Campos, Cruz, Omaki, Jahn, Abreu, Amario, Andrade, Angonese, Banha, Barros, Berrettini, Bianchini, Bleuel, Braz, Brito, Calderon, Cambricoli, Campos, Capel, Castro, Comin, Conceição, Corazza, Cordeiro, Crivellaro, Dornellas, Falsarella, Faria, Felipe, Ferraz, Ferreira, Fialho, Figueiredo, Fonseca, França, Francini-Filho, Garrido, Gaspar, Gomes, Guebert, Gurgel, Ielpo, Inagaki, Jardim, Kato, Kitahara, Lima, Almeida, Lolis, Lotufo, Martello, Mendes, Miranda, Mitsuhashi, Montaño, Moraes, Mota, Nascimento, Negrão, Oliveira, Oliveira, Pacheco, Pereira, Pereira, Pinheiro, Pinto, Pires, Salaroli, Salvi, Sampaio, Santos, Santos-Filho, Segal, Silva, Silva, Sissini, Soares, Soares, Teixeira, Teschima, Toledo, Zilberberg and Longo2025). These stressors disrupt the delicate balance of coral physiology, often leading to widespread bleaching events that challenge coral tolerance and recovery (Hoegh-Guldberg, Reference Hoegh-Guldberg1999; Spalding and Brown, Reference Spalding and Brown2015; Hughes et al., Reference Hughes, Kerry, Álvarez-Noriega, Álvarez-Romero, Anderson, Baird, Babcock, Beger, Bellwood, Berkelmans, Bridge, Butler, Byrne, Cantin, Comeau, Connolly, Cumming, Dalton, Diaz-Pulido, Eakin, Figueira, Gilmour, Harrison, Heron, Hoey, Hobbs, Hoogenboom, Kennedy, Kuo, Lough, Lowe, Liu, McCulloch, Malcolm, McWilliam, Pandolfi, Pears, Pratchett, Schoepf, Simpson, Skirving, Sommer, Torda, Wachenfeld, Willis and Wilson2017; Corazza et al., Reference Corazza, Lacerda, Güth, Marcançoli, Bianchini, Calderon, Capel, Conceição, Faria, Francini-Filho, Garrido, Guebert, Kitahara, Longo, Lotufo, Salvi, Segal, Sumida, Zilberberg and Mies2024; Mies et al., Reference Mies, Destri, Lacerda, Carvalho, Ibanhez, Güth, Luza, Campos, Cruz, Omaki, Jahn, Abreu, Amario, Andrade, Angonese, Banha, Barros, Berrettini, Bianchini, Bleuel, Braz, Brito, Calderon, Cambricoli, Campos, Capel, Castro, Comin, Conceição, Corazza, Cordeiro, Crivellaro, Dornellas, Falsarella, Faria, Felipe, Ferraz, Ferreira, Fialho, Figueiredo, Fonseca, França, Francini-Filho, Garrido, Gaspar, Gomes, Guebert, Gurgel, Ielpo, Inagaki, Jardim, Kato, Kitahara, Lima, Almeida, Lolis, Lotufo, Martello, Mendes, Miranda, Mitsuhashi, Montaño, Moraes, Mota, Nascimento, Negrão, Oliveira, Oliveira, Pacheco, Pereira, Pereira, Pinheiro, Pinto, Pires, Salaroli, Salvi, Sampaio, Santos, Santos-Filho, Segal, Silva, Silva, Sissini, Soares, Soares, Teixeira, Teschima, Toledo, Zilberberg and Longo2025).
Coral bleaching is a stress response characterized by the loss of the dinoflagellate symbionts living within coral tissues and/or by the reduction of the algae’s photosynthetic pigments (Glynn, Reference Glynn1996; Douglas, Reference Douglas2003; Roberty and Plumier, Reference Roberty and Plumier2022). This phenomenon occurs when environmental conditions, such as increased water temperature, exceed the tolerance limits of the coral and its algal symbionts, resulting in the breakdown of the symbiotic relationship (Glynn, Reference Glynn1996; Hoegh-Guldberg, Reference Hoegh-Guldberg2000; Roberty and Plumier, Reference Roberty and Plumier2022). The mechanism underlying bleaching involves the photoinhibition of zooxanthellae photosystems and subsequent oxidative stress, affecting both the coral host and the algae (Lesser, Reference Lesser2006; Weis, Reference Weis2008; Roberty and Plumier, Reference Roberty and Plumier2022). Elevated temperatures can compromise the symbiont photosystem apparatus by damaging the D1 protein in the photosystem II (Warner et al., Reference Warner, Fitt and Schmidt1999), the chloroplast thylakoid membranes (Tchernov et al., Reference Tchernov, Gorbunov, De Vargas, Yadav, Milligant, Häggblom and Falkowski2004), and possibly the enzyme ribulose bisphosphate carboxylase oxygenase (Rubisco) in the Calvin–Benson cycle (Lilley et al., Reference Lilley, Ralph and Larkum2010). This dysfunction leads to the energetic uncoupling of electron transport, resulting in the overproduction of reactive oxygen species (ROS), including singlet oxygen, hydrogen peroxide, and superoxide and hydroxyl radicals (Lesser, Reference Lesser2006; Weis, Reference Weis2008). As ROS levels rise and overwhelm the antioxidant defences, oxidative damage to lipids, proteins, and/or DNA becomes increasingly harmful to algae and to the cellular homeostasis of the coral (Lesser, Reference Lesser2006). This damage triggers the host’s innate immune response to end the dysfunctional host–symbiont interaction, ultimately leading to the elimination of the symbionts (Weis, Reference Weis2008). Given the obligatory nature of the symbiosis, coral recovery from bleaching often depends on the successful re-establishment of the partnership, and failure to do so can worsen coral health and increase mortality risk (Baker, Reference Baker, Rosenberg and Loya2004).
Several factors can affect the recovery of corals from bleaching. Many studies have focused on how genetic and physiological tolerance of individuals influence the recovery process (Van Oppen and Gates, Reference Van Oppen and Gates2006; Rodrigues and Grottoli, Reference Rodrigues and Grottoli2007; Visram and Douglas, Reference Visram and Douglas2007; Thomas and Palumbi, Reference Thomas and Palumbi2017; Corazza et al., Reference Corazza, Lacerda, Güth, Marcançoli, Bianchini, Calderon, Capel, Conceição, Faria, Francini-Filho, Garrido, Guebert, Kitahara, Longo, Lotufo, Salvi, Segal, Sumida, Zilberberg and Mies2024), while others report the role of abiotic environmental conditions. For example, water flow can mitigate temperature stress at both the reef and the colony scales (Page et al., Reference Page, Leggat, Heron, Choukroun, Lloyd and Ainsworth2019), and high nutrient enrichment is negatively correlated with reef recovery time (Robinson et al., Reference Robinson, Wilson and Graham2019), often exacerbating thermal damage to the coral holobiont, making recovery more difficult (Chumun et al., Reference Chumun, Casareto, Higuchi, Irikawa, Bhagooli, Ishikawa and Suzuki2013).
Adding another layer of complexity to this topic, interactions with associated fauna emerge as a significant ecological factor that can influence coral recovery from bleaching events (Chase et al., Reference Chase, Pratchett, Frank and Hoogenboom2018; Shantz et al., Reference Shantz, Ladd, Ezzat, Schmitt, Holbrook, Schmeltzer, Vega Thurber and Burkepile2023). These interactions, whether positive or negative, can play a crucial role in determining the outcomes of coral recovery, yet they remain underexplored. Corals often benefit from nitrogenous nutrients excreted by associated organisms, such as fishes and boring bivalves, which can enhance coral productivity and growth (Mokady et al., Reference Mokady, Loya and Lazar1998; Holbrook et al., Reference Holbrook, Brooks, Schmitt and Stewart2008; Shantz et al., Reference Shantz, Ladd, Ezzat, Schmitt, Holbrook, Schmeltzer, Vega Thurber and Burkepile2023). Associated organisms can also improve water flow and aeration, particularly through the fin movements of resident fishes that shelter within coral colonies at night (Goldshmid et al., Reference Goldshmid, Holzman, Weihs and Genin2004; Berenshtein et al., Reference Berenshtein, Reuben and Genin2015). In addition, some decapod crustaceans provide cleaning services by removing sediments and epibiotic algae (Coen, Reference Coen1988; Stewart et al., Reference Stewart, Holbrook, Schmitt and Brooks2006), while others protect colonies from predators (McKeon et al., Reference McKeon, Stier, McIlroy and Bolker2012; Stier and Osenberg, Reference Stier and Osenberg2024). Conversely, certain interactions can negatively affect coral health. For example, fish faeces can alter the coral microbiome and, when combined with thermal stress, increase coral susceptibility to dysbiosis and reduce tissue regeneration (Ezzat et al., Reference Ezzat, Lamy, Maher, Munsterman, Landfield, Schmeltzer, Gaulke, Burkepile and Thurber2019, Reference Ezzat, Merolla, Clements, Munsterman, Landfield, Stensrud, Schmeltzer, Burkepile and Vega Thurber2021). Furthermore, fishes and invertebrates can cause lesions or impose additional stress through corallivory or parasitism (Bruckner and Bruckner, Reference Bruckner, Bruckner, Woodley, Downs, Bruckner, Porter and Galloway2015; Barton et al.., Reference Barton, Bourne, Humphrey and Hutson2020), ultimately hindering recovery.
Notably, the few studies that have sought to examine the impact of associated fauna on coral recovery from thermal stress have focused on coral-dwelling fishes (Chase et al., Reference Chase, Pratchett, Frank and Hoogenboom2018; Shantz et al., Reference Shantz, Ladd, Ezzat, Schmitt, Holbrook, Schmeltzer, Vega Thurber and Burkepile2023). Nevertheless, the limited available evidence shows that under experimental conditions, these fishes can modulate coral physiological state during the recovery phase following thermal stress, as heated colonies hosting fishes exhibited higher Symbiodiniaceae density and chlorophyll a (Chl a) concentration compared to colonies without fish (Chase et al., Reference Chase, Pratchett, Frank and Hoogenboom2018). In contrast, the role of invertebrates in this context is less understood but potentially significant. Corals host a vast number of invertebrates, with high diversity in species and ecological roles (Glynn and Enochs, Reference Glynn, Enochs, Dubinsky and Stambler2011; Stella et al., Reference Stella, Pratchett, Hutchings, Jones and Gibson2011). In particular, crabs typically dominate the invertebrate fauna associated with coral colonies (Stella et al., Reference Stella, Pratchett, Hutchings, Jones and Gibson2011), and may influence coral recovery, either through beneficial interactions or stress-inducing effects. Associated crabs can contribute to coral health by removing sediments deposited on coral surfaces (Stewart et al., Reference Stewart, Holbrook, Schmitt and Brooks2006) and controlling the overgrowth of epibiotic algae through herbivory (Coen, Reference Coen1988), making their addition to degraded reef environments effective in coral restoration projects (Francis et al., Reference Francis, Filbee-Dexter, Yan and Côté2019; Spadaro and Butler, Reference Spadaro and Butler2021; Butler et al., Reference Butler, Duran, Feehan, Harborne, Hykema, Patterson, Sharp, Spadaro, Wijers and Williams2024). These functions become particularly important in a context of thermal stress because bleached corals have a reduced capacity to clear sediments themselves (Bessell-Browne et al., Reference Bessell-Browne, Negri, Fisher, Clode and Jones2017). In addition, excessive mucus production may trap large quantities of fine sediment particles and facilitate algal colonization (Diaz-Pulido and McCook, Reference Diaz-Pulido and McCook2002). Therefore, by actively scraping and removing this mucus–sediment–algal layer, crabs may prevent smothering, providing conditions that support coral recovery from bleaching. On the other hand, physical interactions with crabs can be detrimental when coral health is already compromised, making the recovery process even more challenging. Under thermal stress, mechanisms of symbiont loss, such as host–cell detachment, reduce tissue integrity and increase vulnerability to damage (Gates et al., Reference Gates, Baghdasarian and Muscatine1992; Weis, Reference Weis2008). In this weakened state, the mechanical actions performed by associated crabs (e.g., scraping, snipping, or sweeping; Kropp, Reference Kropp1986) might further harm fragile coral tissues and create lesions. Such injuries may, in turn, increase susceptibility to diseases, as bleached corals exhibit reduced regeneration capacity and immune function (Meesters and Bak, Reference Meesters and Bak1993; Fine et al., Reference Fine, Oren and Loya2002; Palmer et al., Reference Palmer, Bythell and Willis2010). Thus, associated invertebrates can act as vectors of microorganisms, potentially facilitating pathogen transmission for corals (Sussman et al., Reference Sussman, Loya, Fine and Rosenberg2003; Nicolet et al., Reference Nicolet, Hoogenboom, Gardiner, Pratchett and Willis2013). Altogether, these factors can create a stress scenario that constrains the recovery prospects of bleached corals. Investigating these complex dynamics is key to understand the ecological factors that influence coral resilience during bleaching events.
The coral Madracis decactis (Lyman, 1859) is one of the most common photosymbiotic corals along the Brazilian coast and oceanic islands and represents a dominant species in the subtropical reefs of Brazil (Grillo et al., Reference Grillo, Candido, Giglio and Longo2021). It is widely distributed throughout the tropical and subtropical Western Atlantic, from 32° N to 25° S, and is also found in the Eastern Atlantic off West Africa (Nunes et al., Reference Nunes, Norris and Knowlton2011; Pereira-Filho et al., Reference Pereira-Filho, Shintate, Kitahara, Moura, Amado-Filho, Bahia, Moraes, Neves, Francini, Gibran and Motta2019). This coral usually inhabits shallow waters (2–25 m), but it can reach a depth up to 100 m (Pereira-Filho et al., Reference Pereira-Filho, Shintate, Kitahara, Moura, Amado-Filho, Bahia, Moraes, Neves, Francini, Gibran and Motta2019; Santodomingo et al., Reference Santodomingo, Kennedy, Vermeij and Kitahara2022, accessed online 16 April 2025), occurring on both illuminated reef top substrates and shaded vertical walls (Muramatsu and da Silveira, Reference Muramatsu and da Silveira2008; Grillo et al., Reference Grillo, Candido, Giglio and Longo2021). Colonies can exhibit nodular, submassive, large-branched, laminar, or encrusting forms (Neves and Johnsson, Reference Neves and Johnsson2009), hosting a diverse fauna of fish and invertebrates, mainly polychaetes, mollusks, and crustaceans (Scott, Reference Scott1987; van der Schoot and Hoeksema, Reference van der Schoot and Hoeksema2022). Madracis decactis is an important reef-building coral, well acclimated to suboptimal environmental conditions. Namely, this species is resistant to changes in light availability along the depth gradient, contrasting hydrodynamic regimes between windward and leeward habitats, and seasonal fluctuations in temperature (Grillo et al., Reference Grillo, Candido, Giglio and Longo2021; Pereira-Filho et al., Reference Pereira-Filho, Mendes, Perry, Shintate, Niz, Sawakuchi, Bastos, Giannini, Motta, Millo, Paula-Santos and Moura2021; Santodomingo et al., Reference Santodomingo, Kennedy, Vermeij and Kitahara2022, accessed online 16 April 2025). However, despite its tolerance to variable environmental conditions, Ma. decactis is vulnerable to high sedimentation and elevated temperatures, making it moderately susceptible to diseases and bleaching (Migotto, Reference Migotto1997; Pereira-Filho et al., Reference Pereira-Filho, Shintate, Kitahara, Moura, Amado-Filho, Bahia, Moraes, Neves, Francini, Gibran and Motta2019; Santodomingo et al., Reference Santodomingo, Kennedy, Vermeij and Kitahara2022, accessed online 16 April 2025).
The small crab Mithraculus forceps (A. Milne-Edwards, 1875) is widely distributed in the Western Atlantic, ranging from North Carolina, USA, to Santa Catarina, Brazil (Melo, Reference Melo1996; Rieger and Giraldi, Reference Rieger and Giraldi1996). Along the southeastern coast of Brazil, it is the most abundant spider crab (superfamily Majoidea) in reef environments (Mantelatto et al., Reference Mantelatto, Faria, Biagi and Melo2004). The species inhabits boulder fields and different microhabitats on bedrock, from just below the intertidal zone to depths of up to 90 m, including patches dominated by macroalgae, sponges, and corals (Melo, Reference Melo1996), including Ma. decactis (Rezende-Gois et al., Reference Rezende-Gois, Nogueira, Moraes, Santos and Costa2023). Additional field observations and data from an ongoing study (Roriz et al., in prep.) indicate that this coral–crab association is relatively common. This species has been observed in a mutualistic association with the coral Oculina arbuscula, in which the crab relies on the coral for habitat and supplemental nutrition derived from coral mucus and, in return, removes seaweeds and other sessile invertebrates that would overgrow the coral colony (Stachowicz and Hay, Reference Stachowicz and Hay1999). Furthermore, Mi. forceps is also notable for its algae-cleaning capabilities and high juvenile survival rates, making it valuable in the aquarium trade (Rhyne et al., Reference Rhyne, Penha-Lopes and Lin2005).
This study aims to analyse the symbiotic and oxidative stress responses of the ten-ray star coral Ma. decactis to thermal stress, and to explore the impact of coral-associated crabs on the coral post-bleaching recovery. We expected that thermal stress would compromise the integrity of the coral–algae association, leading to increased signs of oxidative stress in the coral. Additionally, we hypothesized that the presence of crabs would influence the coral’s ability to recover, though the direction of this effect – whether beneficial or detrimental – was difficult to anticipate. Crabs could either benefit the host by removing algal epibionts and sediments (Coen, Reference Coen1988; Stewart et al., Reference Stewart, Holbrook, Schmitt and Brooks2006), but also be prejudicial by damaging coral tissue (Pratchett et al., Reference Pratchett, Graham, Sheppard and Mayes2010; Barton et al., Reference Barton, Bourne, Humphrey and Hutson2020). To test this, we conducted a laboratory experiment where corals were subjected to a thermal stress of 30°C for 21 days, followed by a 4-week recovery period. Crabs were added to control (unbleached) and recovering colonies to test whether eventual effects were general or dependent on the health condition of the coral. To determine coral bleaching and recovery, we assessed key indicators of the coral–symbiodinium association, including Chl a content and symbiont density, as well as oxidative damage measured through lipid peroxidation (LPO).
Materials and methods
Study area and sampling
The reefs of the Southwestern Atlantic are characterized by turbid waters, rich in nutrients and sediments, low coral cover, and dominance of turf and erect macroalgae (Aued et al., Reference Aued, Smith, Quimbayo, Cândido, Longo, Ferreira, Witman, Floeter and Segal2018; Mies et al., Reference Mies, Francini-Filho, Zilberberg, Garrido, Longo, Laurentino, Güth, Sumida and Banha2020; Santana et al., Reference Santana, Mies, Longo, Menezes, Aued, Luza, Bender, Segal, Floeter and Francini-Filho2023). On the northern coast of the state of São Paulo, Brazil, the São Sebastião Channel, which separates Ilhabela from the mainland, hosts relatively well-preserved rocky reefs. These reef environments are nevertheless exposed to chronic anthropogenic pressures, including pollution and sedimentation associated with port activities and coastal runoff (Zanardi et al., Reference Zanardi, Bícego, De Miranda and Weber1999; Cunha, Reference Cunha2003; Gibran and de Moura, Reference Gibran and de Moura2012).
In late March 2024, Ma. decactis patches were collected at a rocky reef located at Ponta da Sela, close to the westernmost limit of Ilhabela, Brazil (23°53′23.5″ S 45°27′41.4″ W), at a depth of 5–12 m using hammer and chisel. In this area, Ma. decactis occurs predominantly in an encrusting growth form along the rocky substrate, from which we selected flat-shaped colonies to obtain two-dimensional fragments. This approach ensured that the colonies selected in the field would (i) yield experimental units with a consistent coral tissue area for fragments of the same dimension, and (ii) minimize confounding effects arising from fragment tridimensionality, which could influence crab–coral interactions. The collected patches were approximately 10–20 cm in diameter, located at least ∼2 m apart from each other to ensure genetic variability. Considering the extent of the surface reef area (approximately 17,200 m2) and visible boundary lines between neighbouring colonies, we estimate that 8–12 different colonies were sampled. Mi. forceps crabs were collected with hand nets, at a depth of 1–12 m, from the same reef where Ma. decactis was sampled and from nearby rocky reefs within the São Sebastião Channel (Segredo Beach: 23°49′42″ S 45°25′15″ W; Beluga: 23°52′55″ S 45°26′52″ W). They were found across multiple benthic microhabitats, including on living and dead corals (such as Ma. decactis, Mussismilia hispida, and zoanthids), and under small boulders. After collection, corals and crabs were transported to the Center for Marine Biology of the University of São Paulo (CEBIMar-USP), where the experiment was conducted.
Experimental procedures and set-up
In the laboratory, corals were cut into rectangular size fragments of approximately 6 cm2 each (mean = 5.71 cm2, SD = 0.88 cm2), using a Dremel® multitool with a diamond blade tip, over the 3 days following collection. During the cutting process, we made efforts to not compromise living coral tissue and ensure that the underside of fragments was cut as flat as possible to maximize further adhesion to the substrates placed in the aquaria. Fragments were then given a 6-day healing period in covered outdoor tanks supplied with open seawater circulation from the São Sebastião Channel, maintaining natural environmental characteristics (mean daily temperature = 24.3°C, SD = 1.6; mean daily maximum luminosity = 17,222 lux, SD = 3,351; estimated Photosynthetic Photon Flux Density [PPFD] ≈ 319 µmol photons m−2 s−1 within the Photosynthetic Active Radiation [PAR] range, assuming sunlight as the light source [Thimijan and Heins, Reference Thimijan and Heins1983]; reference salinity ≈ 33.5 [see Vanin, Reference Vanin2008]; and reference pH ≈ 8.1 [see Patrocinio et al., Reference Patrocinio, Lopes, Ciotti, Faria and Martins2025]). Thirty-six flat natural granite stone blocks (180 cm2) were prepared to serve as bases for the fragments, each with two PVC pipes (3 cm in diameter and 4 cm in length) previously affixed with epoxy as potential shelters for crabs. A total of 180 visually healthy fragments were then haphazardly selected and secured to the stone blocks using Aquavitro® bond (five fragments per block; Supplementary Figure S1). By doing so, we introduce natural genotypic variation within any treatment combinations, rather than considering ‘genotype’ as a factor in the experiment. The fragments were allowed an additional 6 days to recover from any handling or gluing stress in the same outdoor tanks. Subsequently, each stone block was placed at the bottom of a replicate aquarium with a capacity of approximately 2.5 L. In the laboratory, treatments were assigned to 36 aquaria at the start of the experiment, randomly positioned over bench space to avoid pseudoreplication.
Coral fragments were acclimated to indoor conditions for 14 days (Figure 1; Acclimation) at 25°C (mean = 24.6°C, SD = 0.8; Figure 2), approximating the temperature at the sampling site upon collection. Temperature in a random half of the aquaria was then gradually increased over the following 4 days, reaching 30°C (mode = 30°C). This was the target temperature for the heated set over the next 21 days (Figure 2), representing an increase of approximately 4.5°C relative to the mean summer surface temperature recorded during the 3 years preceding the experiment (2022–2024; mean = 25.5°C [Aqualink, 2026]). This exposure corresponds to a cumulative heat stress of approximately 11°C/week (i.e., degree heating weeks), imposing a high chance of coral bleaching (Kayanne, Reference Kayanne2017). After this period, the temperature was gradually reduced to ambient conditions over another 4 days (mean = 23.1°C, SD = 0.6; Figure 2). The other half of the aquaria was kept in controlled temperature conditions throughout the experiment. At the beginning of the following recovery phase (Figure 1), one pair of Mi. forceps crabs was introduced in random halves of the aquaria assigned to the heated and the control conditions for 28 days, compatible with recovery periods used in ‘moderate-duration’ thermal stress experiments (Grottoli et al., Reference Grottoli, Toonen, van Woesik, Vega Thurber, Warner, McLachlan, Price, Bahr, Baums, Castillo, Coffroth, Cunning, Dobson, Donahue, Hench, Iglesias-Prieto, Kemp, Kenkel, Kline, Kuffner, Matthews, Mayfield, Padilla-Gamiño, Palumbi, Voolstra, Weis and Wu2021). Heterosexual pairs of crabs of similar size (mean carapace width = 13.48 mm ± 0.97; weight = 1.08 g ± 0.23) were prioritized to minimize agonistic behaviour (e.g., Sainte-Marie and Lafrance, Reference Sainte-Marie and Lafrance2002). This design created four final recovery treatments, each with nine replicates: (i) unheated coral fragments without crabs (recovery control; Rc), (ii) unheated with crabs (recovery control plus crabs; Rc + cr), (iii) previously heated coral fragments without crabs (recovery post-heating; Rph), and (iv) previously heated with crabs (recovery post-heating plus crabs; Rph + cr).
Diagram illustrating the experimental design. The timeline shows the duration of each phase in days (d) – including initial preparation (collection, fragmentation, and gluing), acclimation (Ac), thermal treatment (TT), and recovery (R) – and sampling for physiological analyses (indicated by tubes). Arrows labeled Tr and Tr′ indicate the temperature ramp-up (warming) and ramp-down (cooling), respectively. Note that the thermal treatment phase consisted of 4-day heating and cooling ramps, plus 21 days at the target temperature, totalling 29 days. The treatments were designated as follows: thermal treatment control (TTc), thermal treatment heated (TTh), recovery control (Rc), recovery control with crabs (Rc + cr), recovery post-heating (Rph), and recovery post-heating with crabs (Rph + cr).
Variation in water mean temperature throughout the experiment phases (acclimation, thermal treatment, and recovery), obtained through readings of two replicate loggers. The timeline shows the duration of each phase in days (d). Note that the water was heated in only half of the aquaria during the thermal treatment phase, while the other half remained at ambient temperature. An acute temperature drop occurred in the middle of the thermal treatment period due to a 2-day power failure.
Throughout the experiment, the aquaria were maintained in a room with controlled air temperature, set at 25°C, and exposed to a continuous flow of filtered (50–60 µm) seawater, preserving the physicochemical characteristics of the natural environment, such as salinity and pH (see Vanin, Reference Vanin2008; Patrocinio et al., Reference Patrocinio, Lopes, Ciotti, Faria and Martins2025), and to natural light (mean daily maximum luminosity = 737 lux, SD = 270; estimated PPFD ≈ 14 µmol photons m−2 s−1; conversion as described above [Thimijan and Heins, Reference Thimijan and Heins1983]). This irradiance level is comparable to PAR values reported at approximately 12 m depth in the São Sebastião Channel (Foloni-Neto and Ciotti, Reference Foloni-Neto and Ciotti2024). Oxygen levels were reinforced by constant air bubbling, and air-cased i-buttons® were used to monitor water temperature (Figure 2). During the thermal stress phase, a power failure caused temperatures to drop to ambient levels in the heated aquaria for 2 days (Figure 2). Because the flow system operated independently of electricity, water circulation, and oxygenation were maintained, preventing any stress related to stagnant water. To allow coral heterotrophic feeding for supplemental nutrition (Grottoli et al., Reference Grottoli, Toonen, van Woesik, Vega Thurber, Warner, McLachlan, Price, Bahr, Baums, Castillo, Coffroth, Cunning, Dobson, Donahue, Hench, Iglesias-Prieto, Kemp, Kenkel, Kline, Kuffner, Matthews, Mayfield, Padilla-Gamiño, Palumbi, Voolstra, Weis and Wu2021), living plankton (>150 μm) was provided twice a week from the acclimation phase to the end of the experiment (mean = 0.13 g, SD = 0.08). Because the open-flow seawater system was filtered, naturally occurring zooplankton was largely excluded, making supplemental feeding necessary to avoid food limitation as an additional stressor. Fresh pieces of the red alga Dichotomaria marginata were provided as food for the crabs during the recovery phase. Coral and crab feeding sessions occurred simultaneously over 1 h, usually in the early afternoon, with the water flow closed. After that, uneaten macroalgae were promptly removed and water flow re-established. Pieces of D. marginata were also placed in the aquaria without crabs at the same time to avoid any confounding effects of the presence of algae.
Coral fragments were sampled at the end of three different periods: acclimation, thermal stress, and recovery. At each time, seven fragments were randomly selected per treatment, from different aquaria. Although the experimental design initially planned for nine fragments per treatment, some fragments were removed at the end of each phase to maintain equal numbers across replicate tanks, while others were excluded due to visible signs of disease (n = 20) to prevent contamination. As a result, seven fragments per treatment represented the maximum sample size that could be consistently collected across all treatments and experimental phases. Subsamples of each fragment were stored in a freezer at −80°C until all physiological and biochemical analyses were carried out.
Physiological traits
To evaluate coral–algae symbiosis integrity, as bleaching indicators, we assessed symbiont density and Chl a concentration (n = 7 per treatment). Each fragment was homogenized in 250 μL of filtered seawater using a sonicator (20 pulses at 36 μm amplitude; Q55 Sonicator®, QSonica). The supernatant, containing both the animal tissue and the symbiotic algae detached from the coral skeleton, was used to quantify symbiont density and Chl a. Symbiont density was estimated through cell counts in a Neubauer chamber, under light microscopy (40×), and normalized by tissue content estimated as total protein concentration (cells/mg protein; Edmunds and Gates, Reference Edmunds and Gates2002). Chl a was determined following a modified protocol from Schmidt et al. (Reference Schmidt, Heinz, Kucera and Uthicke2011), previously adapted for coral–Symbiodiniaceae systems and applied in several coral studies (e.g., Fonseca et al., Reference Fonseca, Marangoni, Marques and Bianchini2017; Glynn et al., Reference Glynn, Marangoni, Guglielmetti, Tapia, Ali, Quintero, Guerra, Yuval, Kline, Leray, Connolly and Barrett2025; Patrocinio et al., Reference Patrocinio, Lopes, Ciotti, Faria and Martins2025). Briefly, 18 μL of the supernatant of each sample was mixed with 382 μL of 95% ethanol and kept refrigerated in the dark for 24 h, vortexing four times during this period. Afterward, samples were centrifuged at 5,200 g for 5 min at 4°C, and their absorbance was measured at wavelengths of 665 and 750 nm using a spectrophotometer (Varioskan™ LUX, Thermo Fisher Scientific Inc). The Chl a content was calculated with the formula: Chl a (μg/mg protein) = 12.0 * (A665 − A750) * Ve/(protein * d) * 103; where 12.0 is the absorbance coefficient of Chl a in 95% ethanol; A665 and A750 are the absorbance values at 665 and 750 nm, respectively; Ve is the volume of ethanol used for extraction in mL; protein is expressed as mg/mL in the sample used for extraction; d is the optical path length in cm, and 103 is the correction factor for the use of a microplate. Data were normalized to protein content and expressed as μg Chl a/mg protein. Protein concentration was determined following the Bradford method (Bradford, Reference Bradford1976) with bovine serum albumin as the standard.
To evaluate the oxidative status of the corals, we measured LPO through the quantification of malondialdehyde (MDA), a byproduct of oxidative damage to membrane lipids (Lesser, Reference Lesser2006). Coral samples (n = 7 per treatment) were homogenized (1:2, w/v) in phosphate buffer (0.1 M, pH 7.0) containing PMSF 0.1 mM, using a sonicator (20 pulses at 108 μm amplitude), and then centrifuged (10,000 g) at 4°C for 10 min. After sample preparation, we quantified MDA concentration, using the Thiobarbituric Acid Reactive Substance protocol (TBARS method) with a commercial kit (Cayman Chemical® No.10009055). Absorbance of the samples was measured at a wavelength of 535 nm using a spectrophotometer (Varioskan™ LUX, Thermo Fisher Scientific Inc.). The obtained values were compared with a standard curve of MDA, normalized considering total protein content (Bradford, Reference Bradford1976), and expressed as mM MDA/mg protein.
Statistical analyses
To examine the effect of heat stress on Chl a concentration, symbiont density, and LPO in Ma. decactis, a t-test was performed to compare the TTc and TTh treatments. Additionally, a two-way ANOVA was conducted to evaluate the post-heating effects, the presence of crabs, and the interaction between these factors for the same variables, considering the four final experimental treatments established during the recovery phase (Rc; Rc + cr; Rph; Rph + cr; Figure 1). This analysis included two factors: (1) pre-heating condition, with two levels: control (Rc and Rc + cr, not subjected to heating) and thermal stress (Rph and Rph + cr, subjected to heating); and (2) crab presence, with two levels: absence (Rc and Rph) and presence of crabs (Rc + cr and Rph + cr). The ‘pre-heating condition’ factor assessed whether coral fragments subjected to thermal stress showed signs of recovery by comparing them to non-heated controls. The ‘crab presence’ factor, on the other hand, allowed us to isolate the effect of crab presence on coral physiological parameters, irrespective of whether the fragments were exposed to thermal stress or not. Finally, the interaction between these factors tested whether coral health and recovery depended on the presence of crabs. After the experimental stress period, contrasts between the heated and the control treatment served to signal the expected direction for recovering fragments. Within the factorial framework used to analyze results of the subsequent experimental period, we refer to ‘recovery’ for treatments that converged towards the undisturbed condition. To ensure homoscedasticity, Chl a and LPO data were log-transformed [ln(x)] and symbiont data were square root-transformed [(x + 1)½]. A chi-square test was also used to compare the percentage of fragments exhibiting disease in the control (TTc) and heated (TTh) conditions. Statistical analyses were performed in R (R Core Team, 2020) and RStudio (RStudio Team, 2020) using the ‘stats’ and ‘car’ packages with the following functions: leveneTest (Levene’s test), shapiro.test (Shapiro–Wilk’s test), t.test (t-test), aov (two-way ANOVA), and chisq.test (Chi-squared test).
Results
After 21 days exposed to a thermal stress of 30°C, approximately 84% of the heated fragments of Ma. decactis showed visual signs of bleaching (Figure 3), with pale and/or white polyps and, in some cases, partial tissue loss. Conversely, control fragments maintained normal coloration and tissue integrity. Heated fragments (TTh) exhibited 79.5% less Chl a concentration (t = 6.76, df = 11.20, p < 0.001; Figure 4) and 93.1% lower symbiont density compared to control corals (TTc; t = 4.75, df = 7.5, p = 0.002). The analysis of LPO revealed no significant differences between these treatments (t = −0.68, df = 10.47, p = 0.513).
Visual condition of different Ma. decactis fragments at the end of the thermal stress phase. (A) Control fragment with healthy pigmentation. (B) Heated fragment exhibiting clear visual bleaching, with fully translucent tissue exposing the white skeleton. (C) Heated fragment exhibiting visual bleaching, but showing pink coloration from endolithic algae beneath the translucent tissue, potentially confounding the assessment of bleaching severity.
The two-way ANOVA indicated a significant post-heating effect for Chl a (F 1,24 = 4.90, p < 0.001) and symbiont density (F 1,24 = 56.29, p < 0.001), with previously heated fragments (Rph and Rph + cr) exhibiting Chl a concentrations and symbiont densities that were 80.6% and 92.2% lower, respectively, compared to non-heated fragments (Rc and Rc + cr; Figure 4). However, no significant effects of crab presence were observed for these response variables (Chl a: F 1,24 = 1.21, p = 0.283; symbiont density: F 1,24 = 0.39, p = 0.539) or for the interaction between post-heating and crab presence (Chl a: F 1,24 = 1.06, p = 0.313; symbiont density: F 1,24 = 0.00, p = 0.970). Likewise, LPO values did not show significant differences for any of the factors (post-heating: F 1,24 = 1.08, p = 0.309; crab presence: F 1,24 = 0.11, p = 0.745) or their interaction (F 1,24 = 0.04, p = 0.849). Descriptive statistics for all physiological variables and treatments are summarized in Supplementary Table S1.
Chlorophyll a content (A), symbiont density (B), and lipoperoxidation levels (C) in Ma. decactis fragments after exposure to the thermal treatment and after the subsequent recovery period. Data are presented as mean ± standard error. After the thermal treatment, comparisons between groups with significant differences in t-tests are indicated by asterisks, according to the level of significance (*p < 0.05, **p < 0.01, and ***p < 0.001). Results from the two-way ANOVA, including main effects and their interaction, are described for each response variable within text boxes. TTc: thermal treatment control; TTh: heated thermal treatment; Rc: recovery control; Rc + cr: recovery control with crabs; Rph: post-heating recovery; Rph + cr: post-heating recovery with crabs.
The mortality of coral fragments during the experiment was low, with only two fragments dying completely: one from the Rph treatment and another one from the Rph + cr treatment. However, a total of 20 fragments were removed due to visible signs of disease, exhibiting whitish or blackish patches and localized tissue deterioration (Supplementary Figure S2). Although no signs of disease were observed during the acclimation phase, during the thermal treatment nearly 21% (n = 15) of the fragments exposed to heating (TTh) exhibited signs of disease, compared to only 4% (n = 3) of the fragments kept at ambient temperature (TTc) (χ2 = 7.68, p = 0.006, df = 1). Illness was only detected in two fragments during the recovery phase: one from the Rc treatment and another one from the Rph + cr treatment.
Discussion
Thermal stress effects: impact on the integrity of the coral–algae symbiosis
The expressive decrease in Chl a content and symbiont density in the thermally stressed fragments underscore the impact of high temperatures on the coral–algae symbiosis and the susceptibility of Ma. decactis to bleaching. In this study, the fragments experienced a 92% reduction in symbiont density – a considerable loss of zooxanthellae that typically ranges from 50% to 90% during bleaching events, indicating severe bleaching (Glynn, Reference Glynn1996). This substantial reduction indicates a disruption in the symbiotic relationship, with many zooxanthellae being eliminated from the coral tissues as a stress response, also leading to an ∼80% decrease in Chl a levels.
Our results are consistent with previous studies that demonstrated declines in both bleaching indicators under elevated temperature conditions (Kleppel et al., Reference Kleppel, Dodge and Reese1989; Bhagooli and Yakovleva, Reference Bhagooli and Yakovleva2004). Following heat stress, oxidative damage activates the host’s innate immune system (Weis, Reference Weis2008), triggering symbiont autophagy (Downs et al., Reference Downs, Kramarsky-Winter, Martinez, Kushmaro, Woodley, Loya and Ostrander2009) or the expulsion of compromised (Weis, Reference Weis2008) or even healthy algae (Fujise et al., Reference Fujise, Yamashita, Suzuki and Koike2013) as a mechanism to regulate Symbiodinium density. Consequently, a significant reduction in algal density entails a considerable loss of pigmentation.
However, the concurrent loss of symbionts and chlorophyll is not the only pathway to coral bleaching. Some studies have reported a reduction in zooxanthellae without a significant decline in chlorophyll (Hoegh-Guldberg and Smith, Reference Hoegh-Guldberg and Smith1989; W. K. Fitt and Warner, Reference Fitt and Warner1995; Bhagooli and Yakovleva, Reference Bhagooli and Yakovleva2004), while others have observed the opposite (Dove et al., Reference Dove, Ortiz, Enríquez, Fine, Fisher, Iglesias-Prieto, Thornhill and Hoegh- Guldberg2006; Rodolfo-Metalpa et al., Reference Rodolfo-Metalpa, Richard, Allemand, Bianchi, Morri and Ferrier-Pagès2006; Venn et al., Reference Venn, Wilson, Trapido-rosenthal, Keely and Douglas2006), including an increase in chlorophyll concentration in remaining symbionts, indicating an enhancement in the packing capacity of photosynthetic pigments as a compensatory response (Faria et al., submitted). This mechanism may vary according to a combination of factors. Differences in symbiont identity, for instance, may influence whether bleaching primarily involves pigment degradation, symbiont loss, or both, given that Symbiodinium clades differ in thermal sensitivity and pigment composition (Venn et al., Reference Venn, Wilson, Trapido-rosenthal, Keely and Douglas2006; Downs et al., Reference Downs, Kramarsky-Winter, Martinez, Kushmaro, Woodley, Loya and Ostrander2009; Strychar and Sammarco, Reference Strychar and Sammarco2011; Fujise et al., Reference Fujise, Yamashita, Suzuki and Koike2013). Other contributing aspects may include the intensity and duration of stress (S. R. Dunn et al., Reference Dunn, Thomason, Le Tissier and Bythell2004), environmental conditions (e.g., light and nutrients; Tanaka et al., Reference Tanaka, Inoue, Nakamura, Suzuki and Sakai2014; Rosic et al., Reference Rosic, Rémond and Mello-Athayde2020), and the coral’s physiological condition prior to stress, such as energy reserves and nutritional status (Anthony et al., Reference Anthony, Hoogenboom, Maynard, Grottoli and Middlebrook2009).
In any case, the reduction in zooxanthellae and/or Chl a content directly impacts the coral’s health by diminishing its primary source of nutrition through symbiont photosynthesis (Muscatine et al., Reference Muscatine, McCloskey and Marian1981; Grottoli et al., Reference Grottoli, Rodrigues and Palardy2006). Furthermore, depending on the mechanism of algal expulsion, whether by host–cell detachment, apoptosis, or necrosis, the coral tissue may undergo destabilization and disintegration (Weis, Reference Weis2008). Therefore, their loss can adversely affect the coral’s ability to recover and survive thermal stress (Rodrigues and Grottoli, Reference Rodrigues and Grottoli2007).
Absence of significant lipid peroxidation and challenges in oxidative stress detection
Elevated temperatures can cause damage to the symbiont photosynthetic apparatus, which continues to generate electrons that are not directed towards the production of ATP and NADPH (Weis, Reference Weis2008). This excess of excited electrons leads to the accumulation of multiple ROS in the symbiont, increasing photosynthetic dysfunction by degrading membrane lipids, proteins, and even nucleic acids, thereby creating a positive feedback loop of further damage (Tchernov et al., Reference Tchernov, Gorbunov, De Vargas, Yadav, Milligant, Häggblom and Falkowski2004; Lesser, Reference Lesser2006; Weis, Reference Weis2008). Given this mechanism, a higher level of LPO was expected in fragments exposed to thermal stress. However, no significant differences in LPO were observed between control and heat-stressed coral fragments in the present study. Similar outcomes have been reported for Mussismilia harttii under the influence of an El Niño event in the Southwest Atlantic, where coral bleaching occurred without an increase in LPO (Marangoni et al., Reference Marangoni, Dalmolin, Marques, Klein, Abrantes, Pereira, Calderon, Castro and Bianchini2019). Likewise, no changes in LPO levels were observed for Turbinaria reniformis and Galaxea fascicularis over experimental long-term exposure to increased temperature. Despite that, high oxidative stress was detected in these coral species through the antioxidant activity of enzymatic biomarkers (e.g., glutathione S-transferase [GST] and catalase [CAT]) (Dias et al., Reference Dias, Ferreira, Gouveia, Madeira, Jogee, Cabral, Diniz and Vinagre2019).
LPO analyses have been widely used as a proxy for oxidative stress in corals responding to warming (Downs et al., Reference Downs, Fauth, Halas, Dustan, Bemiss and Woodley2002; Marangoni et al., Reference Marangoni, Ferrier-Pagès, Rottier, Bianchini and Grover2020; Silva Fonseca et al., Reference Silva Fonseca, Mies, Paranhos, Taniguchi, Güth, Bícego, Marques, Fernandes de Barros Marangoni and Bianchini2021), and have even been suggested as a tool for monitoring reef degradation (Marangoni et al., Reference Marangoni, Dalmolin, Marques, Klein, Abrantes, Pereira, Calderon, Castro and Bianchini2019). However, the detection method for MDA has certain limitations due to its low stability in biological samples, since MDA has a high tendency to react with proteins, amino acids, and undergo rapid enzymatic degradation (Shin et al., Reference Shin, Huggins and Carraway1972; Siu and Draper, Reference Siu and Draper1982; Khoubnasabjafari et al., Reference Khoubnasabjafari, Ansarin and Jouyban2015). Given that thermal stress in our experiment lasted a considerable time, it is plausible that most MDA generated through LPO had been degraded or bound to other molecules by the time of detection and consequently had returned to normal levels. This limitation might explain the lack of significant results in comparison to control concentrations.
Although LPO was not detected in our study, the occurrence of thermal bleaching itself indicated oxidative damage to cellular biomolecules, as ROS are intimately involved with the bleaching process (Lesser, Reference Lesser2006; Weis, Reference Weis2008). Such damage may be reflected in other oxidative stress biomarkers, including products of protein and nucleic acid oxidation. For instance, protein carbonyls are formed when proteins are oxidized directly or through reactions with byproducts of LPO and glycation (Dalle-Donne et al., Reference Dalle-Donne, Rossi, Giustarini, Milzani and Colombo2003). In fact, the accumulation of protein carbonyls was strongly correlated with coral bleaching in a year-long field observations of Montastraea annularis (Downs et al., Reference Downs, Fauth, Halas, Dustan, Bemiss and Woodley2002), and was also found at high levels in symbiotic dinoflagellates of Pocillopora damicornis during experimental bleaching induced by acute heat and light stress (Downs et al., Reference Downs, McDougall, Woodley, Fauth, Richmond, Kushmaro, Gibb, Loya, Ostrander and Kramarsky-Winter2013). Regarding nucleic acid damage, most studies focus on the transcriptional response to thermal stress and the activation of DNA stabilization and repair mechanisms, suggesting that oxidative stress induced by heat can cause DNA damage (i.e., Desalvo et al., Reference Desalvo, Sunagawa, Voolstra and Medina2010; Maor-Landaw and Levy, Reference Maor-Landaw and Levy2016). Nevertheless, more direct evidence of such damage was reported in an octocoral of the genus Sclerophytum exposed to short-term heat and light stress, where significant mitochondrial DNA damage in coral tissues was detected by comparing the amplification efficiency of long and short DNA fragments (Shimpi et al., Reference Shimpi, Vargas, Bentlage and Wörheide2024). Therefore, additional biomarkers should be explored to provide a more comprehensive assessment of oxidative stress and cellular damage. These biomarkers include other lipoperoxidation products (e.g., F2-isoprostanes and HNE; Tsikas, Reference Tsikas2017), antioxidant enzyme activities (e.g., superoxide dismutase, GST, and CAT; Dias et al., 2019), gene expression biomarkers (Louis et al., Reference Louis, Bhagooli, Kenkel, Baker and Dyall2017), among others (Downs et al., Reference Downs, Mueller, Phillips, Fauth and Woodley2000; Tsikas, Reference Tsikas2017).
No evidence of coral recovery: Persistent effects of thermal stress and the importance of prolonged recovery periods
The low Chl a content and symbiont density after the recovery phase of the previously heated fragments (Rph and Rph + cr) compared to non-heated fragments (Rc and Rc + cr) indicates that corals did not recover from bleaching during this experiment. This highlights the vulnerability of Ma. decactis to moderate or severe heat stress (IPCC, Reference Lee and Romero2023), even in an experimental set-up under open-flow conditions with regular feeding and a constant supply of natural seawater, which could have facilitated the uptake of new zooxanthellae. Interestingly, our laboratory outcomes contrast with field observations from the 2019 bleaching event in rocky reefs of the Southwestern Atlantic, where Ma. decactis showed relatively high resilience, with few colonies bleaching and no significant reduction in healthy coral cover 3 years after the event (Roos et al., Reference Roos, Longo, Ilieva, Motta and Pereira‐Filho2024). We speculate that this discrepancy may be explained by environmental conditions of the region, such as summer upwelling episodes (Vanin, Reference Vanin2008) and a marked thermocline that can buffer thermal stress in situ (Banha et al., Reference Banha, Capel, Kitahara, Francini-Filho, Francini, Sumida and Mies2020), which were not present in our laboratory setting.
Recovery capacity is often related to the severity of thermal stress and the extent of subsequent zooxanthellae and/or chlorophyll loss (Hueerkamp et al., Reference Hueerkamp, Glynn, D’croz, Mate and Colley2001; Visram and Douglas, Reference Visram and Douglas2007). Zooxanthellae recolonization typically begins with the reproduction of the remaining pool of healthy symbionts, already living in coral tissues. With an eventual stabilization of the symbiont population, Chl a levels gradually return to normal (Jones and Yellowlees, Reference Jones and Yellowlees1997). Given that the heated fragments in our experiment experienced a 92.2% reduction in symbiont density, it is reasonable to conclude that recovery in these corals would be challenging at best.
Furthermore, cumulative oxidative damage and dysregulation of cellular homeostasis, of both the host and their symbionts, may hamper coral recovery right after thermal stress. The excess of ROS can cause significant harm by attacking cellular components, leading to membrane LPO, enzymatic inactivation, DNA lesions, loss of organelle function, reduction in metabolic efficiency, and decreased carbon fixation (Tchernov et al., Reference Tchernov, Gorbunov, De Vargas, Yadav, Milligant, Häggblom and Falkowski2004; Lesser, Reference Lesser2006). Such damages impair the proper cellular function of both the symbiotic algae and the coral host cells (Dunn Simon et al., Reference Dunn Simon, Pernice, Green, Hoegh-Guldberg and Dove2012; Roberty and Plumier, Reference Roberty and Plumier2022), weakening coral resilience and hindering its ability to recover from thermal stress (Visram and Douglas, Reference Visram and Douglas2007). Additionally, since bleaching compromises the nutritional input from symbionts, the recovery of corals depends on their remaining energy reserves and heterotrophic behaviour (Grottoli et al., Reference Grottoli, Rodrigues and Palardy2006; Rodrigues and Grottoli, Reference Rodrigues and Grottoli2007; Anthony et al., Reference Anthony, Hoogenboom, Maynard, Grottoli and Middlebrook2009). If energy reserves are low due to prolonged stress or pre-existing conditions, and heterotrophic capacity is insufficient to compensate, the potential for recovery can be significantly constrained (Grottoli et al., Reference Grottoli, Rodrigues and Palardy2006; Rodrigues and Grottoli, Reference Rodrigues and Grottoli2007; Anthony et al., Reference Anthony, Hoogenboom, Maynard, Grottoli and Middlebrook2009).
Moreover, the 4-week recovery period in this experiment may have been insufficient for the fragments to fully recover from bleaching. For instance, Jones and Yellowlees (Reference Jones and Yellowlees1997) observed zooxanthellae densities returning to typical levels for a healthy Acropora formosa colony approximately 5 months after bleaching, Rodrigues and Grottoli (Reference Rodrigues and Grottoli2007) reported a recovery period of 8 months for P. compressa, and Fitt William et al. (Reference Fitt William, Spero, Halas, White and Porter1993) noted that this process took up to 10 months in Montastrea annularis. For Chl a concentration, recovery was achieved within 4 months in P. compressa (Rodrigues and Grottoli, Reference Rodrigues and Grottoli2007). Although no studies have documented the physiological recovery dynamics of Madracis species following thermal stress, a field study conducted 3 years after the 2019 bleaching event in the Southwestern Atlantic reported that healthy coral cover of Ma. decactis was maintained (Roos et al., Reference Roos, Longo, Ilieva, Motta and Pereira‐Filho2024). Therefore, the possibility of recovery of Ma. decactis from bleaching over a longer period cannot be ruled out, especially considering that the coral fragments had a low mortality rate throughout the experiment.
Absence of crab effects on coral fragments and bleaching recovery
Contrary to our expectations, Mi. forceps showed no detectable influence on the overall condition or recovery of Ma. decactis fragments following thermal stress. This contrasts with previous studies that highlighted the impact of associated fauna on coral resilience facing thermal stress. For example, Ezzat et al. (Reference Ezzat, Merolla, Clements, Munsterman, Landfield, Stensrud, Schmeltzer, Burkepile and Vega Thurber2021) showed that thermal stress, in combination with external microbial inputs, such as those from coral-associated fish faeces, can significantly increase the risk of dysbiosis and impair tissue regeneration in corals. Conversely, Chase et al. (Reference Chase, Pratchett, Frank and Hoogenboom2018) demonstrated in a laboratory experiment that coral-dwelling fish can mitigate the effects of bleaching and contribute to coral recovery in Pocillopora spp. colonies, with those in the presence of fish exhibiting higher photosynthetic efficiency, symbiont density, and Chl a content than those without fish. These fish likely support corals through the provision of nutrients (Roopin et al., Reference Roopin, Henry and Chadwick2008), thereby enhancing the coral’s ability to re-establish its photosynthetic apparatus (Chase et al., Reference Chase, Pratchett, Frank and Hoogenboom2018; Shantz et al., Reference Shantz, Ladd, Ezzat, Schmitt, Holbrook, Schmeltzer, Vega Thurber and Burkepile2023). However, Pocillopora species typically inhabit reefs with lower nutrient availability compared to Ma. decactis from the Southwestern Atlantic (Mies et al., Reference Mies, Francini-Filho, Zilberberg, Garrido, Longo, Laurentino, Güth, Sumida and Banha2020). Given the higher nutrient conditions in the environment of Ma. decactis, these corals may be less dependent on nutrients provided by the associated fauna, which could partly explain the lack of crab’s effects in our study. Finally, it is important to highlight that the intensity and duration of the bleaching event may have overwhelmed the natural resilience limits of Ma. decactis fragments (Hueerkamp et al., Reference Hueerkamp, Glynn, D’croz, Mate and Colley2001; Visram and Douglas, Reference Visram and Douglas2007), potentially masking any advantageous services that crabs might have provided under less severe conditions.
Thus, the interaction between Mi. forceps and Ma. decactis resembles a commensal-type association (Terrana et al., Reference Terrana, Caulier, Todinanahary, Lepoint and Eeckhaut2016; Barton et al., Reference Barton, Bourne, Humphrey and Hutson2020). While the crabs’ presence does not appear to affect Ma. decactis physiology, Mi. forceps may benefit from the shelter provided by the coral structure, as well as from access to food resources directly from the coral or growing on top of the colonies (Coen, Reference Coen1988; Stella et al., Reference Stella, Pratchett, Hutchings, Jones and Gibson2011), as they do in a mutualistic association with an Orculina coral in the wild (Stachowicz and Hay, Reference Stachowicz and Hay1999). Although crab behaviour was not quantified systematically, routine observations provided qualitative support for this interpretation. Individuals of Mi. forceps regularly left the PVC shelters and interacted physically with the coral fragments, often walking over them and pinching or scraping the tissue surface during feeding events (Supplementary Video S1). These behaviours are consistent with foraging activity and suggest that the crabs were indeed accessing food resources directly from the coral surface. Our observations were restricted to daytime periods and may have led to an underestimate of interaction intensity, as Mi. forceps is mostly active at night (Baroni, Reference Baroni2020), reinforcing that crabs may well benefit from the presence of corals without affecting their hosts.
However, the nature of coral–crab interactions may be better understood as part of a context-dependent symbiotic continuum, where the relationship can vary from neutral to positive or negative based on environmental conditions and ecological factors (Parmentier and Michel, Reference Parmentier and Michel2013; Terrana et al., Reference Terrana, Caulier, Todinanahary, Lepoint and Eeckhaut2016). Previous studies have shown that crabs can play significant roles in reef ecosystems by providing protective services to corals and enhancing their ability to compete with algae (Coen, Reference Coen1988; Stella et al., Reference Stella, Pratchett, Hutchings, Jones and Gibson2011; Francis et al., Reference Francis, Filbee-Dexter, Yan and Côté2019; Spadaro and Butler, Reference Spadaro and Butler2021; Stier and Osenberg, Reference Stier and Osenberg2024). For example, Spadaro and Butler (Reference Spadaro and Butler2021) demonstrated that herbivorous crabs can have an impact on coral reef ecosystems by effectively reversing the dominance of macroalgae, by grazing on the algae and maintaining open space for coral growth. On a smaller scale, crabs can directly benefit coral colonies by removing fouling algal epibionts (Coen, Reference Coen1988) or sediments (Stewart et al., Reference Stewart, Holbrook, Schmitt and Brooks2006) from their surfaces, which would otherwise constrain coral survival and growth. Additionally, crabs can actively defend their coral hosts from predators, alleviating predation pressure, and thereby allowing the corals to allocate more energy towards growth and reproduction rather than defence (Stier and Osenberg, Reference Stier and Osenberg2024). On the other hand, crabs can also have detrimental effects on corals through predation or parasitism (Simon-Blecher and Achituv, Reference Simon-Blecher and Achituv1997; Stella et al., Reference Stella, Pratchett, Hutchings, Jones and Gibson2011; Pratchett et al., Reference Pratchett, Pisapia and Sheppard2013; Terrana et al., Reference Terrana, Caulier, Todinanahary, Lepoint and Eeckhaut2016; Barton et al., Reference Barton, Bourne, Humphrey and Hutson2020). For instance, crabs can directly feed on coral tissue (Stella et al., Reference Stella, Pratchett, Hutchings, Jones and Gibson2011), and in cases of infestation, they can increase stress and mortality in vulnerable coral colonies (Pratchett et al., Reference Pratchett, Pisapia and Sheppard2013). Furthermore, some crabs create pits in coral skeletons, causing physical damage and reducing coral growth (Simon-Blecher and Achituv, Reference Simon-Blecher and Achituv1997), with gall crabs inducing coral deformities, which impair coral structure and competitiveness (Terrana et al., Reference Terrana, Caulier, Todinanahary, Lepoint and Eeckhaut2016).
Considering the multifaceted role of crabs in coral reef ecosystems, it is possible that their effects may extend beyond the simple null relationship we observed in the laboratory. Therefore, further studies are needed to elucidate whether the lack of effects is exclusive to the conditions in which the coral fragments were subjected in this experiment and whether the crabs can affect other parameters of coral physiology.
Conclusion
This study investigated the physiological response of the ten-ray star coral (Ma. decactis) to thermal stress and evaluated the role of crab presence on its recovery from bleaching. Heat stress caused significant reductions in Chl a concentration and symbiont density, resulting in persistent bleaching effects on the coral–algae symbiosis, with no signs of recovery during the experimental period. Furthermore, the presence of Mi. forceps had a negligible impact on Ma. decactis fragments, neither aiding nor hindering recovery, suggesting a commensal relationship. These findings emphasize the vulnerability of Ma. decactis to thermal stress and indicate the limited influence of associated fauna on the recovery of corals from bleaching under controlled conditions. However, the role of crabs and other symbiotic organisms may vary under more complex environmental scenarios, highlighting the importance of future studies that explore their effects across broader ecological contexts. This study represents the first attempt to specifically assess the impact of invertebrates on coral recovery after thermal bleaching, contributing to a deeper understanding of the physiological and oxidative stress responses of corals to climate change. Our results shed light on the intricate dynamics of biotic interactions and stress responses in corals, reinforcing the need for integrative approaches to study coral resilience. These insights are vital to inform future research and conservation strategies aimed at mitigating the impacts of climate change on reef ecosystems.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S0025315426101246.
Acknowledgements
We are grateful to the technical team at the Center for Marine Biology of the University of São Paulo (CEBIMar – USP), especially Joseilto Medeiros de Oliveira and José Jerônimo Pelaes Favaro, for their support in the field and assistance with setting up the experimental structure. We also thank Lisanne Küppers for her help with sample collection, experimental preparation, and for her friendly companionship throughout the experimental period.
Author contributions
PRCR conceived and designed the study, performed the experiment and data analysis, interpreted the results, wrote the manuscript, and prepared the figures. GAC assisted with the experiment and data collection. SCF contributed to study design, interpretation, and manuscript revision. AAVF contributed to study conception, design, data analysis, interpretation, manuscript revision, and supervision. All authors approved the final manuscript.
Funding
This research was supported by the National Council for Scientific and Technological Development (CNPq) through a PhD and a Productivity Fellowship (#315754/2021-4) granted to PRCR and AAVF, respectively, and by the São Paulo Research Foundation (FAPESP) through financial support (#2022/03105-7) to SCF.
Competing interests
The authors declare none.
Ethical standards
This study involved marine invertebrates (corals and crabs) that are not currently subject to regulation by the Brazilian National Council for the Control of Animal Experimentation (CONCEA), and thus ethical approval was not required. Nonetheless, all experimental procedures followed internationally recognized animal welfare principles, ensuring minimal stress and harm to the organisms.
Data
Data available on request from the authors.
