Carbon (C) and nitrogen (N) elements are essential to any biological system, as they enter into the composition of important cellular components such as proteins, genetic material or enzymes. In various living organisms, their metabolisms are therefore closely linked, and an excess or shortage of one element strongly affects the metabolism of the other (Zhang et al., 2018). In the oceans, N is often a limiting nutrient for animal and plant growth (Howarth, 1988). Therefore, the ability of organisms to efficiently use the various available N sources also affects their distribution and abundance in a given environment (Klais et al., 2017). To cope with N limitation, many benthic marine organisms form holobionts (sensus Rohwer et al., 2002), in which the animal host lives in symbiosis with a variety of microorganisms, including protists, fungi, bacteria, archaea, and viruses (Knowlton and Rohwer, 2003). Each partner makes a specific contribution to the N requirements of the holobiont. The animal host feeds heterotrophically on particulate and dissolved organic nitrogen (PON and DON respectively), found in plankton, or in the form of dissolved free amino acids and urea (Grover et al., 2006; Rix et al., 2020). Microbial symbionts such as diazotrophs can provide N to the holobiont through N fixation (Bednarz et al., 2012). Other microbes, such as photosynthetic dinoflagellates from the Symbiodininiaceae family fix dissolved inorganic carbon (DIC) and nitrogen (DIN; ammonium, nitrate) to produce organic compounds that they share with the host to meet its nutrient needs (Rädecker et al., 2015; Freeman et al., 2016). However, the contribution of each partner to the N requirements of the holobiont is species-specific and varies greatly depending on dissolved N concentrations, seawater temperature or other environmental conditions (Baker et al., 2013; Pupier et al., 2021).

Corals are one of the most widespread holobionts in coastal benthic ecosystems. They can form large marine animal forests (MAFs, sensu Rossi et al., 2017), such as coral reefs and coralligenous assemblages, which are probably among the most widespread ecosystems on Earth, and provide services of great economic value to hundreds of millions of people worldwide (Rossi and Rizzo, 2020). Nitrogen availability is a key factor in coral functioning and health, as it determines their ability to grow, reproduce and survive (Rädecker et al., 2015). However, the acquisition and allocation of N in coral holobionts is not well understood, especially in octocorals (Baker et al., 2011; Pupier et al., 2019a; Pupier et al., 2021), despite the fact that octocorals play an important structural and functional role in MAFs, and are key players in the benthic-pelagic energy fluxes through the uptake and release of nutrients (Gili and Coma, 1998; Schubert et al., 2016; Rossi and Rizzo, 2021). Coral feeding ecology has indeed mainly focused on the Hexacorallia subclass, which consists of stony corals (76% of studies; Santos et al., 2020), while only 22% of studies were dedicated to Octocorallia. Moreover, half of the studies were related to feeding behaviour (Santos et al., 2020), so the uptake rates and assimilation of nutrients such as dissolved C and N by Octocorallia are still very poorly understood (Baker et al., 2015; Pupier et al., 2019b; Pupier et al., 2021).

In temperate coastal waters of the Mediterranean Sea, gorgonian corals can reach very high densities in both shallow and deep waters (Gori et al., 2011a; Gori et al., 2011b), and contribute greatly to the biological and economic value of the ecosystems they form. Gorgonians (with an average size of 20-40 cm height) can indeed reach densities from 20 to more than 60 colonies m^-2 in some locations (Sini et al., 2015; Grinyó et al., 2016). Our understanding of their ecology, distribution patterns or reproductive strategies has increased in recent decades (Schubert et al., 2016). However, as with all Octocorallia, much less is known about their feeding strategies for dissolved organic and inorganic C and N. This lack of knowledge becomes particularly concerning in the face of climate change, which poses a serious threat to the survival of these octocorals inhabiting shallow waters (Garrabou et al., 2022). Indeed, the Mediterranean Sea has been suffering from repeated heat waves for the last decades, leading to mass mortality of gorgonians that thrive in shallow waters (Garrabou et al., 2022). In the summer of 2022, which was one of the hottest ever observed in this region, seawater temperature rose above 25°C at depths of 20-30 m, where most gorgonians thrive, reaching threshold for these organisms (Pey et al., 2013; Tignat-Perrier et al., 2022).

To fill the knowledge gaps on N requirements of temperate octocorals both under normal growth conditions and heat stress, the uptake rates of DIN and DON were measured in the two most abundant gorgonian species of the western Mediterranean: the only symbiotic species Eunicella singularis (Esper, 1794), and the non-symbiotic species Paramuricea clavata (Risso, 1826) (Weinberg, 1976). In parallel, measurements of DIC and DOC uptake rates were also carried out, to link the uptake rates of C and N. Measurements were performed under normal growth temperature (17°C) and under heat stress (24°C) (Garrabou et al., 2022). To achieve our objectives, we used the techniques of stable isotopes tracers of N and C (¹⁵N and ¹³C respectively), as well as an approach based on the measurement of nutrient depletion in seawater. Our study addresses several questions. First, to what extent Mediterranean gorgonians rely on dissolved nutrients? Second, how does the existence of a symbiosis with photosynthetic dinoflagellates influence the nutrition in temperate gorgonians? Third, how do their nutritional capacities change with heat stress? We hypothesized that (1) both species can assimilate dissolved C and N; (2) the symbiotic relationship in E. singularis allows for greater dissolved N uptake through the uptake of DIN. Finally, we hypothesized that (3) heat stress had a different effect on the symbiotic than on the non-symbiotic gorgonian. By assessing how gorgonians take up dissolved organic and inorganic nutrients in different seawater temperature conditions, this comparative study will provide a better understanding of the nutritional factors, which regulate the distribution of symbiotic versus non-symbiotic gorgonians in the Mediterranean Sea. In addition, understanding the nutritional requirements of these gorgonians can help inform management practices aimed at improving their health and resilience, particularly in the face of stressors such as climate change and ocean acidification.

Understanding the nutritional requirements of Mediterranean gorgonians, which are essential in forming large Marine Animal Forests (Rossi et al., 2017), is critical for developing effective management practices that can improve gorgonian health and resilience, especially in the face of climate change stressors. Despite high concentrations of dissolved inorganic and organic nitrogen (DIN, DON) in the mesotrophic Mediterranean Sea (Coste et al., 1988; Bethoux et al., 1992), the extent to which gorgonians rely on these nutrients is unknown. The main results of the study, summarized in Table 3 and discussed below, indicate that temperate gorgonians can assimilate all forms of DIN and DON, but to a much lesser extent than tropical octocorals and scleractinian corals, despite being exposed to higher nutrient concentrations. As a result, dissolved N does not contribute significantly to their diet. In addition, at natural concentrations, they do not rely on dissolved organic matter (DOM) to feed, but they can use it when concentrations of DOM are increased. The asymbiotic gorgonian Paramuricea clavata was also able to take up relatively high amounts of ammonium (NH₄ ⁺), despite the absence of Symbiodiniaceae, which are usually the main drivers of NH₄ ⁺ uptake in many coral species (Muscatine et al., 1979; Grover et al., 2002; Grover et al., 2003; Pernice et al., 2012). Finally, an increase in temperature to 24°C for 2 weeks significantly affected the health of the gorgonians and their ability to take up nutrients.

The two gorgonian species studied, the asymbiotic P. clavata and the symbiotic E. singularis, thrive in mesotrophic waters with high concentrations of DON (up to 14 µM, Morganti et al., 2017) and DIN (1-3 µM); that vary depending on the season (Ribes et al., 1999b; Aminot and Kérouel, 2004; Segura-Noguera et al., 2016). These values are significantly higher than those typically found in reef waters (DON: 3.8-5.6 µM (Tanaka et al., 2011), DIN: 1-2 µM). Despite being able to take up DIN, mainly in the form of NH₄ ⁺ (at rates of 60-120 nmol N g^-1 AFDW h^-1), the uptake rates of both gorgonian species were lower than those of tropical scleractinian corals (in the µmol range, Grover et al., 2002; Pupier et al., 2021) and tropical octocorals of the suborder Alcyonaceae (Pupier et al., 2021, ranging from 150 to 640 nmol N g^-1 AFDW h^-1, depending on the species). The most meaningful comparison can be made between E. singularis, which lives in symbiosis with Symbiodiniaceae dinoflagellates (LaJeunesse et al., 2018), and the tropical species mentioned above, all of which live symbiotically. On the contrary to tropical soft corals, gorgonian tissue is thin and cannot act as a barrier to N diffusion (Pupier et al., 2021). Therefore the lower uptake rates observed in E. singularis compared to tropical octocorals could be due to the lower irradiance it receives (50 – 150 µmol photons m^-2 s^-1 (Gori et al., 2012), compared to > 400 µmol photons m^-2 s^-1 for tropical species). Indeed, a positive correlation has been demonstrated between irradiance, symbiont photosynthesis, and N uptake in corals, with increased photosynthesis and N uptake under high light conditions (Grover et al., 2002). In addition, the Mediterranean seawater temperature varies between 12°C and 24°C depending on the season, compared to a minimum of 25°C for tropical species. Therefore, E. singularis may have a lower metabolism than tropical species due to the lower temperatures at which it thrives (Previati et al., 2010).

An interesting observation is that the asymbiotic gorgonian P. clavata exhibited the same rates of NH₄ ⁺ uptake into host tissue as the symbiotic E. singularis. Normally, the first site of NH₄ ⁺ uptake in corals is the Symbiodiniaceae component, which assimilates 90% of the NH₄ ⁺ taken up via the glutamine synthetase/glutamine oxoglutarate aminotransferase pathway (GS/GOGAT) (Pernice et al., 2012; Roberty et al., 2020). Although the coral host can also assimilate NH₄ ⁺ via two pathways (GS/GOGAT and glutamate dehydrogenase pathway), and has ammonium transporters (Capasso et al., 2022), direct NH₄ ⁺ uptake is generally low (Pernice et al., 2012), except under certain stress conditions (Tang et al., 2020; Xiang et al., 2020). Unlike Symbiodiniaceae-bearing corals, non-symbiotic corals are generally unable to assimilate DIN (Muscatine et al., 1979) and instead excrete NH₄ ⁺ and urea into the surrounding water (Rahav et al., 1989). Since microbes in the mucus layer accounted for only 7% of the total NH₄ ⁺ uptake of gorgonians, P. clavata must have evolved specific adaptations to assimilate NH₄ ⁺ and benefit from this additional N source. Alternatively, NH₄ ⁺ uptake may derive from the activity of microbes inside the gorgonian tissue, although the measured uptake rates are very high for such a presumed tiny microbial biomass. In addition to NH₄ ⁺ uptake, there was also very low uptake of nitrate (NO₃ ^-) in both gorgonians (although 8 to 20 times lower than NH₄ ⁺, which in turn is lower than in tropical corals). This uptake may be attributed to the microbes associated with the gorgonians, in particular from the bacterial symbiont Endozoicomonas. This bacteria is very abundant in Mediterranean gorgonians (Van De Water et al., 2017; Van De Water et al., 2018) and can take up NO₃ ^- and reduce it into NH₄ ⁺ (Neave et al., 2014; Neave et al., 2017).

Gorgonians were also able to assimilate labile DON, especially dissolved free amino acids (DFAAs). However, uptake rates were relatively low, in the nanomolar range (20 to 50 nmol N g^-1 AFDW h^-1) and were at the lower end of rates measured for tropical soft corals (20 to 200 nmol N g^-1 AFDW h^-1), themselves lower than for scleractinian corals (Grover et al., 2008; Pupier et al., 2021). DON uptake rates were however higher in the host tissue of the heterotrophic gorgonian P. clavata than in the symbiotic E. singularis, suggesting that P. clavata had to compensate for lower DIN uptake due to the absence of Symbiodiniaceae (total DN uptake of P. clavata: 120 nmol N g^-1 AFDW h^-1 vs 160 for E. singularis holobiont). The uptake of nanomolar amounts of DFAAs suggests that the gorgonians either do not have a large requirement for DON or are unable to assimilate large amounts of DON. In addition, such nanomolar uptake of DFAAs cannot be measured through the incubations with natural DOM, which contains both labile and nonlabile DON at concentrations of 10 µmol N L^-1. This differs from tropical octocorals, which are able to assimilate large amounts of DON when it is present in seawater (Lange et al., 2023). Considering the low uptake rates of DIN and DON, our results suggest that Mediterranean gorgonians have a low capacity to feed on dissolved N sources. However, their low C:N ratio of 5-6 indicates that gorgonians are not limited in N. In fact, the C:N ratio is in the same range as the Redfield ratio (C:N = 6.6) and much lower than the C:N ratio of N-depleted phytoplankton or of deep-sea corals (C:N > 10, Losh et al., 2012; Ferrier-Pagès et al., 2022). Mediterranean gorgonians should therefore meet their N requirements through heterotrophic feeding on particulate organic material (plankton). To our knowledge, no previous study has assessed the amount of N supplied to gorgonians through heterotrophy. However, if we assume a C:N ratio of 6 for plankton and a heterotrophic uptake of 450 µg C g^-1 AFDW h^-1 measured for the Mediterranean gorgonian Leptogorgia sarmentosa (Ribes et al., 2003), the amount of N that can be obtained by heterotrophy may be ca. 75 µg N g^-1 AFDW h^-1 or 5.4 µmol N g^-1 AFDW h^-1, which is in fact more than 30 times the amount of N obtained by the uptake of DIN and DON.

From the measurements of the dissolved organic carbon (DOC) fluxes, it appears that, as with N, the gorgonians release more DOC than they take up from the natural DOM at ambient concentration (80 µmol C L^-1), which is consistent with a previous study on Mediterranean gorgonians (Ribes et al., 1999a; Ribes et al., 2003). The ratio between the release of DOC and the net photosynthetic rate of E. singularis (13%) is in the same range as previous measurements on tropical corals (Ferrier-Pagès et al., 1998; Tanaka et al., 2009). Nevertheless, our results show that gorgonians were able to take up significant amounts of DOC under enrichment conditions (3 and 6 x [DOM_nat]), suggesting that they require a certain threshold for DOC utilization, as previously observed in sponges (Wooster et al., 2019; Ribes et al., 2023). Such high concentrations can be observed in coastal areas receiving DOM from rivers (Cauwet et al., 1990; Ejarque et al., 2017; Sánchez-Pérez et al., 2020). Eunicella singularis can also acquire C as DIC thanks to the presence of Symbiodiniaceae in its tissues. However, although E. singularis had similar concentrations of symbionts and chlorophyll as many tropical soft corals and gorgonians (Rossi et al., 2018; Pupier et al., 2019b; Rossi et al., 2020), its C fixation rate was 10-fold lower except when compared to tropical sea rods (Rossi et al., 2018). This may be related to the polyp size of E. singularis, which is also comparable to sea rods, as a correlation between polyp size and primary productivity has been demonstrated in tropical gorgonians (Baker et al., 2015; Rossi et al., 2018). However, the amount of C supplied to E. singularis by photosynthesis under our experimental conditions (120 µmol photons m^-2 s^-1, corresponding to the summer season) is lower than the amount of C obtained by heterotrophy (Ribes et al., 1999a; Ribes et al., 2003). These measurements confirm that Mediterranean gorgonians rely more on the uptake of particulate food, such as pico- and nanoplankton, than on the uptake of dissolved material.

The health of gorgonian species was significantly affected when the temperature was raised to 24°C for two weeks. A temperature of 24°C - 25°C is a threshold for many Mediterranean corals (Rodolfo-Metalpa et al., 2006; Tignat-Perrier et al., 2022), which was not often reached except at very shallow depths before 2011, but since then has been increasingly measured down to 40 m depth (Cerrano et al., 2000; Garrabou et al., 2009; Garrabou et al., 2022). After several weeks (5 to 8 weeks) at 25°C, gorgonians generally show tissue necrosis and microbial diseases (Tignat-Perrier et al., 2022). In this study, we did not aim to reach this final health state, but only to induce heat stress with reversible effects on gorgonian health. In E. singularis, bleaching and increased respiration rate resulted in a significant loss of cellular C, N and P content, as observed previously (Baker et al., 2018). In P. clavata, the increased respiration rates resulted in a significant decrease in lipid content and thus cellular C content, consistent with Tignat-Perrier et al. (2022). To compensate for the higher respiration rates and energy costs, E. singularis significantly increased its heterotrophic diet on DOC. Indeed, feeding on 3x and 6x [DOM_nat] allowed E. singularis to meet more than 100% of its respiration requirements, consistent with previous findings that some corals can rely on DOC to compensate for loss of autotrophy (Levas et al., 2016). However, DOC uptake was not accompanied by major increase of DON or DIN uptake in either gorgonian. Overall, these changes resulted in a significant increase in the C:N ratio (and C:P ratio) of symbionts and host tissues of E. singularis, suggesting that this species may become N and P limited in the long term under heat stress. Paramuricea clavata also lost N in the form of DON. This release may be due to either increased release of mucoid exudates or increased release of coral cellular material (e.g., detachment of host cells) following heat stress (Tanaka et al., 2009). The last hypothesis seems to be the most likely, since it has indeed been observed that P. clavata is subjected to severe oxidative stress at high temperatures, leading to tissue necrosis after several weeks of stress (Tignat-Perrier et al., 2022). Overall, these observations strongly indicate that gorgonians are highly vulnerable to starvation during marine heat waves (MHWs). During the summer months, the stratification of the water column leads to a significant decline in zooplankton abundance in surface waters (Roemmich and McGowan, 1995), which directly impacts filter feeders like gorgonians by depriving them of a vital food source. Unlike tropical corals, Mediterranean gorgonians lack the ability to rely on the uptake of dissolved inorganic and organic nutrients, exacerbating their susceptibility to food scarcity.

The diet of Mediterranean gorgonians appears to rely more on the capture of plankton prey than on dissolved inorganic and organic nutrients. Previous studies has shown that P. clavata predominantly ingests particulate detritus (Ribes et al., 1999a) and captures a large number of nauplii and small prey items (< 200 µm) during periods of high prey abundance in autumn and spring (Coma et al., 1994). In E. singularis, symbionts can however bring a significant amount of C to the symbiosis, especially during summer, when light is maximal, and plankton abundance is minimal. We studied here the only symbiotic octocoral gorgonian of the Mediterranean Sea (Eunicella singularis) and one widespread asymbiotic gorgonian, Paramuricea clavata. As the response may differ across species, there may be other species-specific differences aside from symbiosis vs non-symbiosis that remains to be investigated. In addition, in tropical reefs, sponges are known to play a key role in processing DOM before returning it to corals in the form of particulate organic matter (the “sponge loop”, de Goeij et al., 2013). Given the high abundance of sponges in the coralligenous assemblages of the Mediterranean Sea, it would be valuable to investigate the potential importance of the sponge loop for gorgonian’s nutrition.

The original contributions presented in the study are included in the article/ Supplementary Material . Further inquiries can be directed to the corresponding author.

The manuscript presents research on animals that do not require ethical approval for their study.

KL, SR and CF-P conceived and designed the experiment. KL performed the experiments. J-FM performed the mass spectrometry analyses. KL analyzed the data. All authors contributed to the article and approved the submitted version.