2.1.  Coral nubbin preparation

A total of two hundred nubbins of the scleractinian coral Stylophora pistillata (~2 cm long) were obtained from three large mother colonies maintained at the Centre Scientifique de Monaco (CSM, originally collected from the Red Sea, CITES permits No. NL024463A011 & FR0808101265). These colonies have been grown at CSM and served as the source of the nubbins or fragments for these experiments; therefore, no further collection permits were required for this work. The nubbins were acclimatised and allowed to recover from the cutting lesions for three weeks in 20-L aquaria using an open system (seawater renewal at a flow rate of 8 L h^-1). The seawater supplying the aquaria was pumped from a depth of 50 m and prefiltered using a sand filter and a sock-type filtration system (100 µm and 20 µm), followed by filtration through activated carbon, and finally passed through a UV lamp for disinfection purposes before entering the aquaria. The salinity of the seawater was 37 ± 1 g kg^-1; the water temperature was maintained at 26 ± 1°C using submersible heaters (300 W Aqua Medic), light was supplied using a PAR LED module (Radiometrix LED GR2, Alpheus France) at an intensity of 200 ± 15 µmol photons m^-2 s^-1 and with a photoperiod of 12h:12h dark:light. The water heaters and the light modules were connected to an automated controlling system (Enoleo Monaco) to ensure the maintenance of a constant temperature and control of the photoperiod. Coral nubbins were fed with Artemia salina nauplii twice per week during the acclimatisation period with feeding ceased just before the beginning of the experiments to avoid interaction with Mn.

2.2.  Thermal stress experiments with manganese dosing

Coral nubbins were randomly distributed into eight aquaria (twenty-five nubbins in each 20-L aquarium), corresponding to two control and two experimental conditions (i.e., each condition in duplicate). The nubbins were exposed to two temperatures, 26°C and 31°C, corresponding to optimum-temperature and heat-stress conditions, respectively. One aquarium of each duplicate set at the two temperature conditions was supplemented with manganese, while the other set was used as the control group (i.e., one duplicate control at 26°C, and another at 31°C, with no manganese enrichment). The seawater temperature in the 31°C aquaria was increased from 26 to 31°C at a rate of 0.5°C per day. Once the temperature was reached, it was maintained for two weeks to challenge the coral nubbins to a heatwave scenario. Manganese-enriched aquaria were supplemented to a final concentration of 6.4 nM using a stock solution of 12.7 µM (as MnCl₂·4H₂O dissolved in milli-Q water). The control groups that did not receive any manganese were supplied with an equivalent volume of only milli-Q water. The solutions were added to the aquaria using a peristaltic pump dispensing at a flow rate of 4 mL hr^-1 (dripping method). This was achieved by using cycles of dispensing the Mn solution during 12 s every 14.8 min, while maintaining a constant seawater flow of 8 L h^-1 in the aquaria. Therefore, coral nubbins were exposed to a total of four conditions: controls at 26 and 31°C (Ctrl 26°C and Ctrl 31°C, respectively) and manganese-enriched at 26 and 31°C (Mn 26°C and Mn 31°C, respectively).

2.3.  Coral physiology analyses

2.4.  Coral antioxidant biomarkers

The coral nubbins that were undisturbed during the experimental period (i.e., not used for photosynthesis measurements; n = 10) were used to determine the antioxidant capacity by assessing several biomarkers as described below. Frozen coral nubbins were thawed, and a section of the tissue was obtained from the apex area (1 cm²); phosphate buffer saline was added to the tissue, and the sample was sonicated using a probe-type ultrasonic processor (20 pulses, 70%, 130 Watt, 20 kHz, Biblock Scientific Illkirch France). The extract was centrifuged (4°C, 11000 g, 10 min) and the supernatant recovered and centrifuged again under the same conditions, corresponding to the host fraction (i.e., without Symbiodiniaceae), which was used for the biomarker assays. Protein concentration was quantified in triplicate using the Bradford spectrophotometric method (λ = 595 nm) with BSA used for the standard curve. These extracts were harmonised to 1 mg protein mL^-1 and used to determine the activity of superoxide dismutase, glutathione peroxidase and the total antioxidant capacity.

2.5.  Quantification of superoxide production rates in seawater

A parallel experiment was performed to determine the power of manganese to counterbalance superoxide production in the coral ambient water. This experiment followed the same set-up as for the coral exposure previously described, except it was downscaled to 2-L aquaria. Three bottles were prepared with stock solutions of 10 µM manganese (Mn^II as MnCl₂·4H₂O, dissolved in milli-Q water), 10 µM iron (Fe^II & Fe^III as FeSO₄ & FeCl₃, respectively, using 5 µM of each species), and milli-Q water (for the control group). Iron enrichment served as a positive control to confirm the production of superoxide through the Fenton reaction. A total of twelve 2-L aquaria were set up using an open system and containing three coral nubbins each. Six aquaria were maintained at 26°C, and the other six were subjected to a temperature increase to 31°C (1°C per day). For each temperature condition, the aquaria were divided into three sets of duplicates. One duplicate set received 6.4 nM manganese (final concentration), the second set received 10 nM iron, and the third set received an equivalent volume of the other two sets but of milli-Q water (i.e., no metals, control group). The solutions were added to their corresponding aquarium using a peristaltic pump (flow rate = 0.8 mL h^-1) and were directly mixed with the incoming seawater (flow rate = 0.8 L h^-1) using a submerged aquarium pump. This experimental phase lasted eleven days, during which superoxide quantification was carried out under dark and light conditions. Superoxide production rates were quantified as a time series in the six aquaria maintained at 26°C (for 11 days), and as temperature-dependent in the six aquaria with increasing temperatures (26–31°C). For the temperature-dependent aquaria, an extra set of subsamples of the control and iron-enriched aquaria was taken, and manganese was added and incubated for 2 min before quantification of superoxide production rate to examine any rapid effect of manganese on this production rate (Mn final concentration = 6.4 nM).

Superoxide radical production rates were measured using the MCLA chemiluminescence method using opaque 96-well plates [46]. Each well received 10 µL of DTPA 3 mM, 3 µL of xanthine 5 mM, 5 µL of MCLA 125 µM and 270 µL of the seawater sample from the aquaria, which was collected directly from the aquaria using a multichannel pipette and immediately transferred into the microplate containing the chemicals mentioned above to initiate and, subsequently, monitor the reaction using a microplate reader (Xenius SAFAS, Monaco). A standard curve was prepared using 5 mM xanthine and xanthine oxidase at concentrations of 0.01, 0.05, 0.25 and 1.25 U L^-1. SOD at 5 kU mL^-1 was used for blank correction.

2.6.  Manganese oxide quantification in Symbiodiniaceae ex hospite

Mn oxides were determined using the leucoberbelin blue I (LBB; Sigma 432199) assay as first described by Krumbein and Altmann [47] and adapted for quantification in microalgae as per Wang et al. [48]. Symbiodiniaceae clades A, C and D (Symbiodinium sp., Cladocopium sp., Durusdinium sp., respectively [2]), originally isolated in-house from their coral hosts, were cultured in f/2 medium [49] at manganese concentrations of 0.23, 0.46, 0.91, 1.82 and 3.64 µM (as MnCl₂·4H₂O). Triplicate cultures were used at the stationary growth phase (day 28) and harmonised to 15 × 10⁶ cells, for which cell densities were determined and the required volume taken from each culture to be centrifuged (5000 g, 10 min) to concentrate and obtain the targeted number of cells (without lysing and maintaining the whole cells). The cell pellets were resuspended in 3 mL of filtered seawater (0.45 µm) and centrifuged again, which corresponded to a washing cycle to remove any potential Mn left in the cell pellet. The supernatant was discarded, and the cell pellet containing the 15 × 10⁶ cells was resuspended in 100 µL of filtered seawater. Briefly, 500 µL of 0.004% LBB (m/v) in 0.25% acetic acid (v/v) were added and the samples incubated for 15 min in the dark (microscopic observations confirmed that the Symbiodiniaceae cells were not lysed during this process and maintained their shape). The samples were centrifuged (5000 g, 10 min) and 200 µL of the supernatant transferred into a 96-well plate and the absorbance determined at 624 nm. The blanks were f/2 medium at the Mn concentrations used for the Symbiodiniaceae cultures. The standard curve was generated using KMnO₄ at concentrations of 1.5, 5.5, 9.5, 13.5 and 17.5 µM. The final values were adjusted using a conversion factor of 2.5, as 1 M KMnO₄ oxidises 5 M LBB versus 2 M LBB oxidised by 1 M Mn^IV oxide (i.e., MnO₂) [50].

2.7.  Statistical analysis

All the data were subjected to statistical analysis using the Python programming language 3.11.11 (Python Software Foundation https://www.python.org/). Normality and homoscedasticity were initially tested in all the data; when necessary, log₁₀ or boxcox transformations were carried out to fulfill the requirements to perform parametric statistics. When normality was not achieved, non-parametric tests were performed. One-way ANOVA tests were performed to determine whether any significant differences amongst the experimental groups existed (i.e., chlorophyll and symbiont density, gross and net photosynthesis, relative electron transport rate, SOD activity, LPO, and gene expression). A two-way ANOVA test was performed to determine any significant differences in MnOx production amongst clades and MnCl₂·4H₂O culturing concentration. When the ANOVA tests returned significant differences, a post-hoc Tukey HSD test was performed. For non-normal data, Kruskal-Wallis tests were performed, followed by either the Dunn (i.e., photosynthetic efficiency of PSII, respiration, GPx activity, TAC) or Conover (superoxide production data) post-hoc tests to determine significant differences amongst groups. A significance of α = 0.05 was employed in all the statistical analyses.

3.1.  Chlorophyll, symbiont density and photosynthetic parameters

The chlorophyll a & c2 average concentrations were 4.10 and 4.32 µg chl cm^-2 in the control groups (no manganese) at 31°C and 26°C, respectively, and 3.31 and 4.36 µg chl cm^-2 for the experimental conditions with manganese enrichment at 31°C and 26°C, respectively (Fig 1A). No significant differences in chlorophyll concentrations were observed amongst the four experimental conditions (ANOVA, p = 0.4887). By contrast, the density of Symbiodiniaceae cells per coral surface area only decreased significantly with temperature increase and manganese enrichment. The coral nubbins maintained at 26°C showed significantly higher symbiont densities (Ctrl 26°C = 2.24 × 10⁶ cells cm^-2) but only when compared to the corals from the manganese-enriched treatment at 31°C (Mn 31°C = 1.50 × 10⁶ cells cm^-2; Tukey test, p = 0.0244) (Fig 1B; Table 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Summary of the effect of heat stress in the coral Stylophora pistillata. Comparison of groups with and without Mn enrichment as assessed on several biomarkers.

https://doi.org/10.1371/journal.pclm.0000897.t001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Pigment concentration and symbiont density in Stylophora pistillata challenged to heat stress.

Chlorophyll a & c2 concentrations (A) and density of Symbiodiniaceae cells (B) with and without manganese enrichment in the ambient water. Differences in Chl concentrations amongst treatments were not significant (ANOVA; p = 0.4887). Lower case letters indicate significant differences in symbiont cell densities amongst groups (ANOVA; p = 0.0280).

https://doi.org/10.1371/journal.pclm.0000897.g001

Manganese enrichment enhanced Symbiodiniaceae photosynthetic efficiency since coral nubbins with Mn showed an increase in Fv/Fm, especially those at 31°C (Mn 31°C Fv/Fm = 0.58), which was significantly higher than the coral nubbins at 26 and 31°C without Mn (Ctrl 26–31°C, Fv/Fm = 0.35-0.40; Dunn test, p ≤ 0.0423; Fig 2A). By contrast, the relative electron transport rate (ETR λ₂₂₀) did not show any significant differences amongst all the experimental conditions (Fig 2B; ANOVA, p = 0.1192). Non-thermal stressed coral nubbins showed comparable gross photosynthesis, which was independent of manganese supplementation (0.81-0.96 µmol O₂ h^-1 cm^-2) and was significantly lower than the gross photosynthesis quantified in thermal-stressed corals (1.64-1.65 µmol O₂ h^-1 cm^-2). However, net photosynthesis was non-significantly different among treatments (0.26-0.57 µmol O₂ h^-1 cm^-2; ANOVA, p = 0.1168) (Fig 2C).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Photosynthetic parameters of Stylophora pistillata under thermal stress.

The photosynthetic efficiency was assessed as the maximum quantum yield efficiency of the photosystem II (Fv/Fm) (A) and relative electron transport rate at 220 nm (ETR λ₂₂₀) (B), and gross and net photosynthesis and respiration (C) of S. pistillata under normal and heat stress conditions with Mn addition. Significant differences amongst the experimental conditions are indicated with lower case letters (Kruskal-Wallis, p = 0.0009 for Fv/Fm and respiration; ANOVA, p = 0.0023 for gross photosynthesis).

https://doi.org/10.1371/journal.pclm.0000897.g002

3.2.  Oxidative stress biomarkers and gene expression of MnSOD

The enzyme GPx showed a significant decrease in activity with manganese enrichment at 26°C and 31°C. The Mn-enriched group at 31°C showed the lowest significant GPx activity (421.99 U mg protein^-1), especially when compared to the two control groups maintained at 26°C and 31°C (711.78 and 672.24 U mg protein^-1, respectively; Dunn test, p ≤ 0.0028; Fig 3A; Table 1). On the other hand, the enzyme SOD did not show any significant differences with temperature or manganese addition (0.15-0.18 U mg protein^-1; Tukey test, p ≥ 0.0954) (Fig 3B). Surprisingly, the gene expression of the manganese-dependent SOD, MnSOD, did show significant differences compared to the total SOD enzyme activity. Using the Ctrl 26°C as a reference, a downregulation of the MnSOD gene was observed on both groups with Mn enrichment independent of temperature (26–31°C; 0.56 fold change for both), which was also significantly lower than the control group at 31°C without Mn (0.83 fold change) (Tukey test, p ≤ 0.0206; Fig 3C).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Antioxidant enzyme quantification in thermal-stressed Stylophora pistillata with and without Mn enrichment.

Glutathione peroxidase (A), superoxide dismutase (B) and gene expression of the MnSOD enzyme (C; fold change compared to Ctrl 26°C). The white dots inside the violin plots represent the individual values of the samples (C). Significant differences amongst treatments are indicated with lower case letters (Kruskal-Wallis, p = 0.00005 for GPx; ANOVA, p = 0.0414 but Tukey test, p ≥ 0.0954, for SOD; ANOVA, p = 0.000008 for MnSOD gene expression).

https://doi.org/10.1371/journal.pclm.0000897.g003

The total antioxidant capacity of the coral nubbins was significantly lower only in the Mn-enriched group at 26°C compared to the thermal-stressed nubbins with no manganese (185.24 vs. 540.09 µM Cu reducing eq. mg protein^-1, respectively; Dunn test, p = 0.0279). The coral nubbins in the other experimental conditions showed TAC values of 324.65 & 574.01 µM Cu reducing eq. mg protein^-1 (Fig 4A), with no significant differences amongst them (Dunn test, p = 1). On the other hand, LPO values ranged from 134.56 (Ctrl 26°C) to 191.76 nmol MDA mg protein^-1 (Mn 31°C), with only the Mn 31°C group showing significant differences against the two groups at 26°C (≤148.50 nmol MDA mg protein^-1; Tukey test, p ≤ 0.0271) (Fig 4B; Table 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. Biomarkers for oxidative stress in S. pistillata.

Total antioxidant capacity (A) and lipid peroxidation (B), after a thermal-stress experiment with and without Mn enrichment. The letters on top of the violin plots show significant differences amongst the experimental conditions (Kruskal-Wallis, p = 0.0249 for TAC; ANOVA; p = 0.0032 for LPO). The white dots inside the plots represent the individual values of the samples.

https://doi.org/10.1371/journal.pclm.0000897.g004

3.3.  Superoxide radical quantification in seawater

Superoxide radical production rates were quantified in seawater as a function of time (11 days) while maintaining the seawater with the coral nubbins at 26°C, but also as a function of temperature (26–31°C), both under dark and light conditions with and without Mn addition. Regarding the time series measurements (Fig 5A), superoxide production rates showed no significant differences amongst the control (milli-Q only), the manganese and iron-enriched treatments over the duration of the experiment, when compared within the same conditions of either dark or light within the same day. There was however a general pattern of an increase in superoxide production rate when comparing light with dark conditions within the same day, starting from the beginning of the experiment (day 0) and progressing until day 11, potentially due to a build-up of organic matter through the duration of the experiment (e.g., exudates from the coral holobiont). For instance, the difference in the control group was significant when comparing the superoxide production rate from day 0 under dark and light conditions (423.3 vs. 783.3 pM s^-1, respectively; Conover test, p = 0.0067). Similarly, the iron-enriched group showed a significant increase in superoxide levels from dark to light conditions (e.g., day 0, 458.7 and 979.4 pM s^-1, respectively; Conover test, p = 7 × 10^-9), which corresponded to a two-fold increase in superoxide production rate after a ~ 6 h of light exposure (Fig 5A).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. Superoxide production rates in ambient water of aquaria with S. pistillata.

Coral nubbins were grown under dark and light conditions while enriched with Mn or Fe over the 11 days of incubation at 26°C (A), and as a function of a temperature increase from 26 to 31°C (B).

https://doi.org/10.1371/journal.pclm.0000897.g005

Regarding the role of temperature and light in superoxide radical production (Fig 5B), a similar pattern was observed between dark and light conditions with increasing water temperatures. A significant increase in the rate of superoxide production in light compared to dark conditions was apparent but only up to 30°C in the control and Mn-enrichment groups, and up to 29°C in the Fe-enriched groups. For example, the rate of superoxide production in the control group increased significantly from 1278.2 to 1745.1 pM s^-1 at 30°C under dark and light conditions, respectively (Conover test, p = 3 × 10^-6); however, the apparent increase from 1451.6 to 1782.8 pM s^-1 at 31°C in dark and light conditions, respectively, was not significant (Conover test, p = 0.2206) (Fig 5B). This suggests that higher concentrations of iron and manganese would be required to continue having an effect in increasing or decreasing superoxide production in ambient seawater at ≥30°C, respectively.

The effect of Mn on superoxide production rate in the ambient water of the aquaria was tested by adding Mn (Cf = 6.4 nM) to the control and Fe-rich groups, and incubating for 2 min. The steady state superoxide concentrations were significantly reduced in both groups to which Mn was added with this observed in both dark and light conditions, especially at temperatures of ≥27°C (Fig 6A).This quick action of Mn in decreasing superoxide concentrations brought the values down to levels similar to those observed at 26°C for this same control group (992.2 pM s^-1 in light conditions). A similar scenario was observed in the Fe-rich group during light conditions at 26°C with production rates of 1162.6 pM s^-1 and 983.4 pM s^-1 before and after adding Mn, respectively. This pattern is portrayed in Fig 6A; lower water temperatures and dark conditions (e.g., 26°C) showed lower superoxide radical production rates (closer to the centre of the radar plot), with a further decrease in superoxide production rates when Mn was added. The superoxide scavenging effect by Mn was higher in the control conditions, where up to 50% of the superoxide present was scavenged in the aquaria with seawater at 29°C. However, the effect of manganese addition in scavenging superoxide was weaker or remained similar at 30–31°C, in both the control and Fe-rich light conditions (33–50%; Fig 6B).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 6. Quantification of superoxide radical production rates in ambient water of aquaria with S. pistillata and scavenging action by Mn.

Superoxide levels were assessed as a function of temperature (26-31°C) with the addition of Mn to the control and Fe-rich experimental conditions 2 min before the superoxide assay. The radar plot shows the impact of Mn in a rapid decrease in superoxide, the closer to the centre, the lesser the superoxide production rates, and the further from the centre, the higher the superoxide production rates (A). The correlation of superoxide scavenging in the control and Fe-enriched seawater due to the effect of Mn addition showed a proportional pattern at seawater temperatures of 26-29°C; however, Mn oxidation by superoxide appeared to weaken at ≥30°C, where the adjustment lines started to go downward (B).

https://doi.org/10.1371/journal.pclm.0000897.g006

3.4.  Mn oxide production by Symbiodiniaceae

The three clades of Symbiodiniaceae produced Mn oxides; higher amounts were produced when cultured at increasing Mn concentrations in the f/2 culture medium, which coincided with a deposit observed at the bottom of the culture flasks, and with a blue colour that developed upon incubation with LBB, which indicated the presence of MnOx (Fig 7A). Clade A produced the lowest amount of MnOx, with no significant differences observed between cultures at 1.82 and 3.64 µM MnCl₂ (10.81-12.24 µM MnOx; Tukey test, p = 0.7213). By contrast, clade D produced the highest amount of MnOx, but similar to clade A, no significant differences were found between algae grown at 1.82 and 3.64 µM MnCl₂ (25.10-31.88 µM MnOx; Tukey test, p = 0.0702) (Fig 7B).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 7. Production of MnOx by three clades of Symbiodiniaceae grown at increasing Mn concentrations for 28 days.

One set of replicate cultures of each clade together with the supernatant and cell pellet following the incubation with LBB are shown, stronger blue colours of the supernatant indicate higher concentrations of MnOx (A). In general, increasing Mn concentration in the culture media showed higher MnOx production by Symbiodiniaceae, with clades A and D showing the lowest and highest MnOx production, respectively (B). The samples corresponding to Symbiodiniaceae cultured at 0.23 µM Mn^II are not included in the plot given that the absorbance values were below the standard curve due to the detection limit of the LBB assay using the 96-well plate version.

https://doi.org/10.1371/journal.pclm.0000897.g007

4.1.  Photosynthetic efficiency in Stylophora pistillata

Our observations did not show any significant differences in chlorophyll a & c2 content in thermal-stressed compared to non-stressed corals. While this contrasts with the findings by Biscéré et al. [21], who observed that Mn enrichment (20.7 nM, at 32°C) enhanced chlorophyll concentrations in S. pistillata, our findings are consistent with those by Tu et al. [29] on the scleractinian corals Turbinaria irregularis and Montipora mollis exposed to thermal stress with manganese enrichment (50–250 nM, at 30°C). Although we observed a significant decrease in symbiont density in thermal-stressed corals with Mn addition (6.4 nM, at 31°C), the remaining dinoflagellates (~1.5 × 10⁶ cells cm^-2) presented an increase in the chlorophyll maximum photosynthetic efficiency of the photosystem II (PSII, Fv/Fm), suggesting that they were in a better photosynthetic condition than those from the other groups without manganese (i.e., Ctrl 26°C & Ctrl 31°C). These findings are similar to those by Biscéré et al. [21] and comparable to those by Moreira et al. [27], who found an increase in Fv/Fm in S. pistillata supplemented with Mn under heat stress (75 nM, at 31°C). Similarly, Tu et al. [29] observed an increase in photosynthetic efficiency in both coral species T. irregularis and M. mollis, with effects starting at 50 nM Mn^II, although accompanied by an increase in symbiont density.

Furthermore, we observed that coral nubbins supplemented with Mn were not adversely affected by temperature in their photosynthetic efficiency as supported by the Fv/Fm, the relative electron transport rate and net photosynthesis, which did not show any significant differences compared to control corals maintained at 26°C with or without Mn. This was observed despite using a manganese concentration lower than those experimented by Biscéré et al. [21] and Tu et al. [29], who used 3 and 8–39 times higher, respectively, than the concentration used in this study. These findings suggest that symbiont dinoflagellates were able to uptake part of the manganese added to the water and use it to their advantage, especially with regards to their PSII photosynthetic efficiency. This parameter reflects the health of the photosynthetic apparatus in plants and algae, and it is thought to be the most sensitive apparatus to thermal stress and a contributor of the overall production of ROS [52]. Therefore, the healthy or still undamaged symbiont dinoflagellates that remain within the coral under high seawater temperatures are expected to tolerate a certain level of thermal stress and retain the capacity to produce ROS but also photosynthates or nutrients for the coral. This is supported by Iglesias-Prieto et al. [53] and Lesser [9] who observed that some Symbiodiniaceae exposed to temperatures and light stress at which bleaching occurs, did not completely stop photosynthesising. Moreover, Bhagooli and Hidaka [54] demonstrated that Symbiodiniaceae expelled by the coral Galaxea fascicularis under thermal stress (i.e., bleached at 30 and 32°C) were still healthy and with no damage to their photosynthetic apparatus, concluding that the coral was non-selective in the expulsion of the symbiont dinoflagellates and instead, it was mainly a response due to acute heat stress.

Manganese is an essential micronutrient for algae, cyanobacteria and plants as it acts as an important component of the water oxidising complex of photosystem II, where water splitting takes place and manganese donates electrons to enable oxygenic photosynthesis [55]. Mn^II is widely used in conventional growth medium recipes to culture microalgae in contained environments, such as the f/2 medium, which has a final manganese concentration of 0.91 µM, and it is commonly used to culture dinoflagellates such as Symbiodiniaceae ex hospite [49,56–58]. Manganese is generally needed by algal cells at trace levels, with growth-limiting concentrations reported at ≤0.1-1 nM [59,60]. Mn sustains the growth of the symbiont dinoflagellate Symbiodinium kawagutii at concentrations as low as 4.2 nM, with maximum growth observed at 42 nM [61]. This concentration is close to that reported by Tu et al. [29] who observed that the best photosynthetic efficiency in the corals M. mollis and T. irregularis occurred at 50 nM of manganese enrichment, but inhibitory effects were observed at ≥100 nM. Additionally, Moreira et al. [27] observed an improved photosynthetic efficiency in S. pistillata with Mn addition of 75 nM at 31°C.

Our findings, together with those by Biscéré et al., Montalbetti et al., Tu et al. and Moreira et al. [21,22,27,29] suggest that coral holobionts may have a narrow range, between ~6 and 75 nM, in which Mn supplementation provides beneficial effects, especially since Tu et al. [29] observed that enrichment with manganese at 100–250 nM showed inhibitory chronic effects in photosynthetic parameters in the corals M. mollis and T. irregularis under thermal stress (14-d exposure). On the other hand, manganese sub-lethal effects observed in adult corals during acute exposures are in the µM range. Acropora millepora experienced tissue sloughing at 10 and 50% effect concentrations (EC₁₀ and EC₅₀) of 6.6 µM and 12.9 µM, respectively, during a 2-d exposure [62], with no chronic effect at ≤5.5 µM (14-d exposure). However, Acropora muricata was more sensitive, with tissue sloughing at EC₁₀ = 1.3 µM & EC₅₀ = 4.1 µM (2-d exposure; [63]). Regarding adult specimens of S. pistillata, tissue sloughing has been observed at EC₅₀ = 4.3 µM and mortality at EC₅₀ = 7.6 µM (2-d exposure, [64]). Moreover, Binet et al. [63] observed a higher sensitivity of adult corals to Mn compared to their early-stage counterparts. Therefore, Mn supplementation (dripping delivery method as in this study) in the 6.4 to 75 nM range seems to be the best approach in protecting corals under heat stress.

4.2.  Oxidative stress in thermal-stressed corals

Manganese and other metals are also utilised as cofactors by some antioxidant enzymes; hence, they are known as metalloenzymes. Common biomarkers to determine oxidative stress in living organisms include the antioxidant enzymes glutathione peroxidase and SOD, as well as biomarkers such as total antioxidant capacity and lipid peroxidation. Aquatic organisms depend on copper, zinc, iron, manganese, and nickel for an efficient activity of SOD, but the metal requirement may depend on the type of SOD present in the organism, such as Cu/Zn-SOD, Fe-SOD, Mn-SOD, or Ni-SOD. Some of these SOD types have been found in Symbiodinium kawagutii and other Symbiodiniaceae [65,66], which have shown intracellular antioxidant enzyme activity to counteract ROS production triggered by high temperatures and light [67].

Coral hosts must counteract ROS production by their symbionts to prevent oxidative damage; otherwise, this would trigger the expulsion of dinoflagellates and result in coral bleaching [68]. We observed a significant decrease in GPx activity with Mn enrichment and temperature increase, suggesting a reduced activity in breaking down hydrogen peroxide, on which GPx acts by using reduced glutathione (GSH) and converting it to oxidised glutathione (GSSG). A similar observation was reported by Montalbetti et al. (2021), who detected a significant decrease in glutathione reductase (GR) activity in thermally stressed S. pistillata (at 32°C) supplemented with Mn alone (20.7 nM) or in combination with iron. GR is essential to complete the cycle after GPx breaks down H₂O₂, as it is responsible for the regeneration of GSSG back to GSH (which must be continuously available for GPx); therefore, these two enzymes work synergistically to remove H₂O₂ [69]. Therefore, since GPx activity decreased, our observations suggest that either the concentrations of H₂O₂ decreased with Mn addition in seawater or that the coral holobiont had an alternate route to neutralise H₂O₂.

It has been hypothesised that ambient manganese enrichment may enhance MnSOD antioxidant activity; however, we did not observe any changes in total SOD activity in S. pistillata, in agreement with the observations by Montalbetti et al. [22]. Therefore, we investigated the gene expression of MnSOD to conclusively determine its role in supporting S. pistillata under heat stress and Mn enrichment. We observed a significant downregulation of the gene associated with the production of MnSOD only in the groups with manganese (at 26 & 31°C), which, together with the SOD enzyme activity, suggests that MnSOD has no role in the antioxidant defence of S. pistillata against heat stress when Mn supplementation occurs. This is potentially due to a decrease in the rate of production of superoxide radicals or an alternate non-enzymatic pathway, or both. Our findings are further supported by Montalbetti et al. (2021) who observed that the gene expression of the heat shock proteins Hsp70 and Hsp60 in S. pistillata remained unchanged after heat stress (32°C), but only when manganese alone or in combination with iron were added (20.7 nM Mn^II); by contrast, when iron or no metal enrichment were applied, Hsp70 and Hsp60 were upregulated, confirming thermal stress. Similar findings in gene expression upregulation of Hsp70, Hsp16 and MnSOD have been observed in situ in the coral Porites harrisoni under thermal and light stress, which were downregulated once the corals were removed from the stress conditions by transplanting them from shallow to deeper waters [70]. These observations suggest that Mn also provides external protection by scavenging superoxide radicals in the ambient seawater.

4.3.  Manganese role as an antioxidant

Coassin et al. demonstrated the antioxidant activity of Mn^II against hydroperoxyl radicals and its inhibitory effect on LPO in vitro using a rat liver model [35]. Mn^II can be oxidised to produce manganese oxides (e.g., MnO₂) with reversible formation of Mn^II and superoxide radicals. However, in the presence of high phosphate concentrations (0.5 M), MnO₂ could undergo reductive dissolution and form Mn^III and hydrogen peroxide [39]. Since the seawater used in our experiments did not receive any other nutrient enrichment, phosphate concentrations were not expected to be high, hence it is possible that the oxidation of Mn^II (added as MnCl₂·4H₂O) by superoxide radicals could have led to the formation of Mn oxides such as MnO₂ (Mn^IV) by Symbiodiniaceae, and not to the formation of Mn^III and H₂O₂ as suggested by Wuttig et al. and Zhang et al. who observed that MnO₂ formation and Mn^II oxidation was also promoted by increasing temperatures [71,72]. Additionally, a lack of H₂O₂ formation is also suggested by our observations of the decrease in GPx activity in corals with Mn^II supplementation.

The process of MnOx formation can occur in aquatic environments through abiotic reactions (e.g., photochemically), but also via biotic processes involving Mn uptake and subsequent oxidation by superoxide, as has been demonstrated in microorganisms. The green microalga Scenedesmus subspicatus was shown to produce surface-bound MnOx by Mn^II oxidation when supplemented with high Mn concentrations; otherwise, low Mn concentrations led primarily to intracellular Mn uptake [59]. Mn^II oxidation has also been shown to be enhanced by bacteria-algae interactions due to an improved generation of superoxide radicals, compared to oxidation caused by the algae alone [31]. Moreover, the formation of Mn^III/IV oxides by green microalgae, diatoms and cyanobacteria in Mn^II and phosphate-rich media has also been observed via reactions with superoxide [50]. These observations support our findings of significant amounts of MnOx in Symbiodiniaceae cultured at relatively high Mn^II concentrations (0.46-3.64 µM). To the best of our knowledge, we have demonstrated for the first time that Symbiodiniaceae ex hospite are capable of producing MnOx. Symbiodinium sp. (clade A), which was originally isolated from one of the S. pistillata colonies used in the present study, produced up to 12.24 µM MnOx when grown under MnCl₂·4H₂O supplementation (≤3.64 µM). Additionally, we observed higher levels of MnOx in Cladocopium sp. and Durusdinium sp. (clades C and D, 18.67 and 31.88 µM MnOx, respectively), which may explain why these clades have a higher tolerance to thermal stress to counteract the impact of ROS. McGinty et al. reported that clade D was the only clade that remained unaffected at elevated water temperatures, claiming that it possesses alternative mechanisms that prevent ROS upregulation [67] or, as our results suggest a potential ROS-scavenging or neutralising effect via an efficient uptake of Mn^II to produce larger amounts of MnOx with antioxidant-like properties.

Non-enzymatic manganese protection against superoxide radicals has been previously demonstrated through the accumulation of manganese (in the mM range) in organisms such as the bacterium Lactobacillus plantarum and the yeast Saccharomyces cerevisiae. Some of these organisms lack all or some types of SOD enzymes, proving the power of Mn as a ROS-scavenger and, together with its reaction products (i.e., MnOx), they have been classified as SOD-mimics as they can chemically remove superoxide radicals or other ROS via their redox properties [34,73–76]. This enzyme-mimicking role of manganese has also been demonstrated in mammalian cell lines treated with MnOx nanoparticles (Mn₃O₄, produced from MnO₂) which, upon exposure to ROS, did not show oxidative damage or changes in their endogenous antioxidant machinery, further supporting the protective, antioxidant enzyme-mimicking function of MnOx [33]. Therefore, it is plausible that MnOx formed within the coral holobiont, potentially by the symbiont dinoflagellates, mimicked its endogenous antioxidant defences to fight oxidative stress. This interpretation is supported by our observations that corals supplemented with Mn^II showed an unaltered SOD activity with a downregulated MnSOD gene, a decrease in GPx activity, as well as a decrease in the total antioxidant activity at 26°C (Mn 26°C), even when compared with the non-thermal stressed control group without Mn (Ctrl 26°C).

Our observations also suggest an antioxidant role of manganese (or Mn^II oxidation), as it reacted with superoxide radicals in seawater. These radicals may have been produced by thermal-stressed symbionts but given the short life of superoxide, they were probably generated largely through abiotic processes, including photochemical and temperature-dependent reactions with dissolved or natural organic matter, such as coral exudates [77,78]. Furthermore, ROS production by some microalgae has been proposed as a toxic mechanism in fish and other aquatic organisms given that microalgae are more photosynthetically active in the light and, hence, the production of ROS is higher, which has been demonstrated in algal cultures and in in situ studies on the Great Barrier Reef [5,46,79,80]. Our observation of increasing superoxide radicals in seawater during light conditions support these observations. Moreover, the rapid decay in superoxide radicals upon addition of Mn in the seawater (especially at temperatures ≤29°C), supports the external ROS-scavenging action of Mn^II. This has been demonstrated with other ion metals such as copper, which catalyses the decay (via disproportionation) of superoxide radicals at nM concentrations, and it was considered as the major sink for superoxide in the water column of the Southern Ocean [81]. Similarly, Wuttig et al. [82] observed that dissolved Mn^II was a significant pathway for superoxide decay in surface waters of the Eastern Tropical North Atlantic impacted by Saharan dust, which transports Mn to the mixed ocean layer. Moreover, inputs of metals such as iron, zinc, copper and manganese into the Red Sea through desert dust storms have also been reported and, when this desert dust was supplied to the corals S. pistillata and Turbinaria reniformis under heat-stress, metal uptake by the corals occurred together with an enhanced photosynthetic efficiency [83,84]. Although these experiments only looked into the metal uptake by the corals and their improved photo-physiological status, there is the possibility that formation of MnOx occurred as a mechanism against thermal stress. These observations together with those by Wuttig et al. [82] and our findings of a decrease in superoxide production rate with ambient Mn^II addition, plus the observation of MnOx production by Symbiodiniaceae, support our hypothesis of a double role of manganese as (i) an antioxidant by uptake and potential formation of MnOx within the coral holobiont as an enzyme-mimicking alternative to the endogenous antioxidant machinery (e.g., SOD, GPx), and as (ii) an environmental superoxide scavenger to protect S. pistillata under thermal and oxidative stress.

4.4.  Implications of manganese use for coral protection

Recent studies on climate change impacts emphasise both the growing demand in urgent interventions and the significant limitations associated with them. Amongst the limitations, Morrison et al. [85] argued that responding effectively to climate change impacts require “radical interventions” that can transform the social and institutional systems that drive environmental degradation. Regarding the global coral crisis, Hughes et al. [86] cautioned that restoration initiatives cannot compensate for large-scale climate impacts, as they operate at much smaller spatial scales than the disturbances affecting reef ecosystems. Expanding this discussion, Ogier et al. [87,88] highlighted substantial governance gaps as these interventions proliferate, and recommended that governance systems must evolve rapidly to match the pace of innovation in marine climate interventions. Altogether, these studies suggest that while technological and ecological interventions may play a role in climate responses, their deployment must be accompanied by systemic transformations and stronger governance frameworks. Despite these limitations, developing and implementing strategies to enhance corals’ resilience to rising seawater temperatures remain worthwhile. In this context, several efforts have been made to upscale the potential solutions for coral conservation to in situ applications. Some of these, such as probiotic supplementation, are already applied in situ in the Red Sea [89]. Additionally, supplementing corals with manganese could also be scaled up to a reef scenario in the future.

Recent laboratory developments indeed tested biopolymer composites with Mn that could be set up in reefs to allow slow release of Mn to protect corals against oxidative stress and therefore prevent bleaching [27]. We are currently testing the upscaling of Mn supplementation on other coral species as well as in industrial coral aquaculture systems, before subsequent in situ applications can take place to determine whether this intervention fulfils the criteria of low-cost and efficiency for radical conservation interventions [86,87]. Translating our work into real time in situ applications would include (i) estimating the volume of water covering the reef area to be protected, (ii) quantifying the existing concentration of Mn in the water, (iii) potential Mn competition uptake by other species, such as macroalgae and seagrasses [90,91], and (iv) targeting strategic points of Mn deployment taking into account local and regional currents and winds, as well as country legislation and permit requirements. For instance, the Australian and New Zealand indicative guideline for manganese in marine waters is 80 µg L^-1 (~404 nM if supplied as MnCl₂ ∙ 4H₂O) [92]. Should this conservation proposal prove to be efficient and viable while ensuring that the impacts in social, environmental and economic risks are zero or maintained to the minimum, its applicability in real-world events would entail designing a transformative and systematic plan of action by involving government, policymakers and community programs to create policies, monitor long-term effects and awareness of the importance of such interventions to conserve coral reefs in the face of climate change [85,87].

Overall, our findings suggest that the level of Mn employed (6.4 nM) was enough to provide protection to S. pistillata, although with a slight LPO effect observed under high temperatures (Mn 31°C group), suggesting that slightly higher concentrations of Mn could provide further environmental protection as has been previously reported. Biscéré et al. and Tu et al. [21,29] used 20.7 nM and 50–250 nM of Mn^II, respectively, and observed a delay or lack of bleaching in corals under thermal stress, although 100–250 nM of Mn showed inhibitory effects in photosynthesis in the corals M. mollis and T. irregularis [29]. It seems that the concentration used in our study of 6.4 nM Mn^II can be considered as a threshold concentration at which manganese begins to provide protection to corals, partially by the process of uptake and MnOx formation by the symbiont dinoflagellates as well as through environmental ROS-scavenging processes. Our findings, together with previous reports that also included other coral species, suggest that corals have a narrow range in which Mn supplementation provides beneficial effects, which is 6.4 to 75 nM (when added as MnCl₂ ∙ 4H₂O). It is not recommended to use higher concentrations of manganese, as this may induce metal toxicity and damage to the photosynthetic apparatus of the symbiont dinoflagellates in hospite, and hence would avoid protecting the entire holobiont under thermal stress. Furthermore, production of Mn oxides within the coral holobiont deserves further investigation given that (i) our current pioneering study has quantified MnOx in Symbiodiniaceae ex hospite grown under Mn-rich media, and (ii) the antioxidant-mimicking activity by MnOx has been observed in other biological systems including microbes and mammalian cell lines to prevent oxidative stress.

Finally, our work paves the way for further research on the production of MnOx by Symbiodiniaceae and the coral holobiont, and their role in providing protection against thermal and oxidative stress that can lead to coral bleaching. We are currently investigating the production of MnOx by different Symbiodiniaceae species grown at varying Mn concentrations, light intensities and increasing temperatures. Since Mn is an essential metal for photosynthetic processes in algae and plants [55] and production of MnOx has been observed in several algal species and bacteria [31,50], it is expected that all Symbiodiniaceae species or clades will also produce MnOx, although perhaps at varying concentrations as observed in the three clades used in the present study. Therefore, Mn supplementation could potentially benefit all or most symbiotic coral species under heat stress, which live in symbiosis with dinoflagellates, such as Symbiodiniaceae.