The mutualistic relationship between the cnidarian host and its endosymbiotic dinoflagellate algae of the family Symbiodiniaceae is vital for the persistence and continuity of coral reefs (Rowan and Knowlton, 1995; Berkelmans and Van Oppen, 2006; LaJeunesse et al., 2018). Through inorganic carbon fixation, the photosynthetic symbionts provide their host with sugars and amino acids, a crucial asset in oligotrophic seas such as the Red Sea (Muscatine and Porter, 1977). In return, the host provides the resident algae with light-rich conditions, a sheltered environment, and a supply of inorganic nutrients. Consequently, the cnidarian host and its associated dinoflagellate algae are critical partners in the diverse assemblage of eukaryotic and prokaryotic taxa that comprise the coral holobiont (Lesser et al., 2013; Aranda et al., 2016; Suggett et al., 2017).

The genetic diversity and composition of the algal symbionts has been studied using different molecular markers. Initially, Rowan and Powers (1992) applied restriction fragment length polymorphisms (RFLPs) of the nuclear small-subunit rRNA gene (SSU) to examine the phylogenetic relationships among the algal symbionts of different hosts. Subsequently, additional DNA regions were studied, including the less variable nuclear large-subunit rRNA (LSU: Lenaers et al., 1991; Rodriguez-Lanetty et al., 2001), the nuclear rRNA internal transcribed spacers (ITS1 and ITS2: LaJeunesse, 2001; Thornhill et al., 2007), the mitochondrial cytochromes b and cytochrome oxidase subunit 1 (cytb and cox1: Takabayashi et al., 2004; Sampayo et al., 2009), and the chloroplastic 23S rRNA and the gene encoding the D1 subunit of photosystem II (cp23S and psbA: Santos et al., 2002; LaJeunesse and Thornhill, 2011).

To date, 11 phylogenetic genera have been assigned to the family Symbiodiniaceae (LaJeunesse et al., 2018; Nitschke et al., 2020; Pochon and LaJeunesse, 2021). These were sampled from unrelated marine organisms such as foraminifers, ciliates, cnidarians, and mollusks, as well as from coral-reef sediment (Nitschke et al., 2022). Despite the lack of a standard molecular barcoding marker for Symbiodiniaceae identification, the ITS2 (~270 bp) is the most widely-used one to determine Symbiodiniaceae lineages at the genus level (e.g., Arif et al., 2014; Smith et al., 2020). This marker was adopted due to: (1) its large copy number (algal genomes enclose several 100 to several 1,000 copies of the rRNA cluster) (Saad et al., 2020)]; and (2) the existence of conserved flanking regions that enable the design of universal Symbiodiniaceae primers. It should also be noted that the fast rate of evolution of the ITS2 region appears to resolve the Symbiodiniaceae relationships at the genus level and even lower (Hume et al., 2018). However, the multi-copy nature of this marker creates a situation in which both intergenomic variants (i.e., polymorphism between algal individuals within a coral) and intragenomic variants (i.e., polymorphism between the ITS copies within an algal individual) may be simultaneously present within any coral sample. To identify the algal lineages present in a coral holobiont, it is critical to distinguish between these two sources of polymorphism. Specific analytical approaches have been developed to identify the confounding intragenomic diversity from the taxonomic diversity. In particular, profiling approaches, such as those implemented in SymPortal (Hume et al., 2019), enable diagnostic ITS2 type-profiles to be assigned, while examining the symbiotic algal composition at the highest possible resolution.

Coral-endosymbiotic algae interactions have been shown to be affected by heatwaves and other environmental stressors, which may result in the impairment and even breakdown of their mutualistic relationship (i.e., bleaching events; Hoegh-Guldberg and Bruno, 2010; Bahr et al., 2017; Hughes et al., 2018). In response to the increasingly recorded stressor events affecting coral reefs, research on the genetic and functional diversity of symbiodiniaceans has developed rapidly (LaJeunesse et al., 2010, 2018; Hume et al., 2019). Notably, it was found that the genetic lineages of the algal symbionts may modulate the phenotype of the holobiont in response to temperature rise or to poorly lit-environments, such as those affected by depth (Frade et al., 2008a; Cooper et al., 2011; Eckert et al., 2020; Smith et al., 2020). Consequently, an accurate determination of the Symbiodiniaceae diversity within corals is required in order to understand their hosts’ adaptive capabilities under changing environmental conditions.

Mesophotic coral ecosystems (MCEs) are coral-dominated communities found at a depth of 30–150 m. They act as an extension of the shallow-water reefs down to the lower edge of the photic zone (Kahng et al., 2010). MCE community structure is affected by the exponential decrease of light across depth (Kahng et al., 2019) and by the corals’ reliance on photoautotrophic carbon production via their endosymbionts (Rädecker et al., 2021). Most coral species at these depths are adapted to certain environmental conditions and can only thrive within a narrow depth range (i.e., “depth-specialists”; Liberman et al., 2022). A few corals, however, are able to thrive across a wider depth range (i.e., “depth-generalists”). The ability to host different symbiodiniaceaens across depth may benefit depth-generalists (Bongaerts et al., 2010; Ziegler et al., 2015). For example, the shallow-water population of the anthozoan Stylophora pistillata in the Gulf of Aqaba/Eilat hosts members of the genus Symbiodinium, whereas the mesophotic population is dominated by members of the genus Cladocopium (Scucchia et al., 2021). Similar shifts in Symbiodiniaceae diversity across depth exist in Madracis pharensis (Curacao: Frade et al., 2008a), Seriatopora hystrix (western Australia: Cooper et al., 2011), and Montastraea cavernosa (Belize: Eckert et al., 2020). However, changes in algal diversity across depth are not considered to be a general phenomenon, and some depth-generalist species exhibit similar Symbiodiniaceae composition across their entire depth of occurrence such as Pachyderms speciosa (Western Australia: Cooper et al., 2011), Seriatopora hystrix, and Porites spp. (Red Sea: Ziegler et al., 2015).

Despite the high abundance of zooxanthellate octocorals in MCEs (Bridge et al., 2012; Shoham and Benayahu, 2017; Benayahu et al., 2019), only a few studies have investigated the diversity of symbiodiniaceaens associated with octocorals (Goulet et al., 2019). Indeed, their diversity in MCE octocorals has not been the focus of any study to date. The genetic composition of Symbiodiniaceae inhabiting shallow-water octocorals has been examined in different biogeographical regions (Caribbean: Goulet and Coffroth, 2004; Hannes et al., 2009, Great Barrier Reef, Australia: LaJeunesse et al., 2004; Van Oppen et al., 2005; Ziegler et al., 2018, Red Sea: Barneah et al., 2004; Goulet et al., 2008). Most of the shallow-water octocoral Symbiodiniaceaen algae were studied using the RFLP or DNA fingerprint approaches (Goulet and Coffroth, 2003, LaJeunesse et al., 2004, Goulet et al., 2008, but see Van Oppen et al., 2005). These methods led to the suggestion that octocoral colonies are probably associated with members of a single genus. Recently, the algal composition of six shallow octocoral species was studied using high-throughput sequencing methods (Ziegler et al., 2018: Bayerxenia sp., Isis sp., Lobophytum sp. and Xenia sp.; Osman et al., 2020: Sarcophyton trocheliophorum and Xenia umbellata), which have also revealed members of a single Symbiodiniaceae genus associated with individual octocoral colonies.

Clearly, our current knowledge of the genetic structure of octocoral Symbiodiniaceae communities and of the more cryptic genetic identities is insufficient. Motivated by this lack of knowledge, the present study sought to (1) examine, for the first-time, the genetic lineages of Symbiodiniaceae of MCE octocorals from the Gulf of Aqaba/Eilat; and (2) compare the Symbiodiniaceae communities of depth-specialists across a depth gradient using both Sanger sequencing and high-throughput sequencing of the ITS2 with the SymPortal analysis framework. Our results revealed a pattern of conserved symbiont community along depth for most the examined species, indicating that host-symbiont specificity may drive Symbiodiniaceae diversity in the Gulf of Aqaba/Eilat octocorals.

The current study has revealed that Cladocopium (previously known as clade C) is the most abundant Symbiodiniaceae in the MCE octocorals in the Gulf of Aqaba/ Eilat (Barneah et al., 2004; Goulet et al., 2008), which is in agreement with previous studies on the algal symbionts of shallow octocorals there. Examination of these symbionts in marine organisms across the Red Sea has highlighted the distinctive diversity found in the Gulf of Aqaba/Eilat and demonstrated that members of Cladocopium are the dominant algal symbionts there. This is in contrast to the central and southern Red Sea, where members of the genus Symbiodinium are the most common algal symbionts (i.e., stony corals: Sawall et al., 2014, Terraneo et al., 2019; zooanthid: Reimer et al., 2017; giant clam: Rossbach et al., 2021). The dominance of Cladocopium over Symbiodinium in the Gulf of Aqaba/Eilat has been suggested to be driven by the prevailing cooler water temperature and high salinity there, which differ from those in the other parts of the Red Sea (Sawall et al., 2014).

Three main factors have been suggested to affect the Symbiodiniaceae-type present in a coral holobiont: the cnidarian host species, light intensity (i.e., water depth) and seawater temperature (Cooper et al., 2011; Tonk et al., 2013; Reimer et al., 2017). Our results indicate that the host’s taxon plays a significant role in determining the octocoral-associated Symbiodiniaceae, while depth was found to significantly affect only the depth-generalist S. eilatensis. The degree of algal-specificity among cnidarian holobionts has also been suggested to be associated with the mode of symbiont acquisition, with strong specificity associated with vertical rather than horizontal transmission (Swain et al., 2018). Among octocorals, the relationships between algal-specificity and the mode of symbiont transmission is poorly understood, since to date it has only been recorded at the symbiont genus level (Barneah et al., 2004; Van Oppen et al., 2005; Goulet et al., 2019). In this work, we mainly investigated horizontal transmitters. The only vertical transmitter representatives are the members of the genus Litophyton (Benayahu et al., 1992; Goulet et al., 2008). Our results have revealed that distinct major Symbiodinium types are exclusively associated with each of the two Litophyton congeners (Figures 1, 2, 6; Supplementary Table S2). Out of the 37 Symbiodinium types identified in the two Litophyton congeners, only two minor ones were shared by both (i.e., A1iv and A1iW). In contrast, in the horizontal transmitters several Cladocopium types present in one host species are often found also in another species. As a case in point the dominant type C1 is not only present in congeners (e.g., S. leptoclados S. vrijmoethi., S. eilatensis MCE, and S. mesophotica) but also in species from different genera (i.e., Ovabunda, Sinularia, and Rhytisma). The same is true for C3, which is also present in high frequency in the members of both Sinularia and Rhytisma (Figure 5; Supplementary Table S1). Our results, obtained using high-throughput sequencing of the ITS2, thus provide for the first time evidence of a strong algal specificity in octocoral vertical transmitters, not only at the algal genus level but also at the ITS2 type profile resolution. C1 and C3 are ancestral sequence variants for major Cladocopium radiations which likely diversified more rapidly than the rate at which ITS2 accumulates mutations in this genus (Thornhill et al., 2014). Thus, even though the ITS2 sequence is shared across a variety of host species, a higher resolution marker such as psbA might indicate that in fact, different C1 lineages associate with different octocoral hosts (LaJeunesse and Thornhill, 2011). Furthermore, future research should attempt to understand the influence of members of the genus Symbiodinium on the dispersal, distribution, and photo-physiology of both Litophyton host species across a broad depth gradient.

Despite being dominant on many Red Sea and Indo-Pacific reefs, octocoral-symbiont associations have been poorly studied. Nonetheless, several comparisons between the current study and previous ones can be made. First, as suggested by Goulet et al. (2008), the current ITS2-type profile analysis indicated that octocorals mostly host a single population of symbionts. However, some exceptions were found, such as in most of the Ovabunda sp. colonies and in one colony of the MCE S. vrijmoethi, both of which were found to host an additional background population (Figure 5). Second, Osman et al. (2020) showed that ITS2 C65 variants associate with Sarcophyton trocheliophorum in the Gulf of Aqaba/Eilat as well as in other reefs of the northern Red Sea. This corresponds to our finding that eight species, including two Sarcophyton ones, associate with ITS2 type C65 or variants that are closely related, such as ITS2 C71 (Figures 3, 5). Interestingly, ITS2 C65 types have been previously reported to associate only with octocoral species (LaJeunesse et al., 2004; Goulet et al., 2008). In addition, ITS2 C71 has also only been found in Sarcophyton sp. from the South China Sea (Sanqiang et al., 2018). Furthermore, the dominant Symbiodinium ITS2 types found in the two Litophyton species in this study were found to be closely related to ITS2 sequences identified in other Litophyton colonies in the southern Red Sea (A9 EU449028 and A10 EU792882; Figure 2). These observations suggest that the association between octocorals and Symbiodiniaceae is not only at the genus level (see Goulet et al., 2008) but also at the lineage-level within a Symbiodiniaceae genus. The finding of specific octocoral-Symbiodiniaceae associations may be attributed to the different ‘micro-environment’ conditions in this group of organisms compared to those of scleractinians, such as reduced light amplification and dispersion by the skeleton (Kramer et al., 2022), a thick coenenchyma that limits gas and nutrient exchanges through the epidermal tissue, and the ability of octocorals to contract their tissue, thus preventing the symbionts from obtaining optimal exposure to light and nutrients (Fabricius and Klumpp, 1995; Ferrier-Pagès et al., 2021).

The depth-related zonation patterns of Symbiodiniaceae on coral reefs have been discussed in several studies examining stony-coral symbiont associations, highlighting certain aspects of the holobiont ecology and its response to changing environmental conditions (Frade et al., 2008a; Bongaerts et al., 2011, 2015; Cooper et al., 2011; Byler et al., 2013; Ziegler et al., 2015; Eckert et al., 2020). Our study is the first to examine the host-symbiont associations in ‘depth-generalist’ octocorals across their entire depth of occurrence. Specifically, our findings indicate that the shift in symbiont community composition across depth is species related and cannot be considered as a general phenomenon. The ‘depth-generalist’ S. eilatensis was found to harbor different ITS2-types at shallower depths compared to in the MCEs, whereas S. vrijmoethi displayed similar ITS2 sequences across depth (S. eilatensis:12–17 m: C65, C65a, C107, and C3hy; 60–67 m: C1, C1b, C39, and C41; S. vrijmoethi: C65, C65a, C107, and C3hy, Figure 5). Previous studies have suggested that while a shift in the symbiont community across depth may functionally benefit certain coral holobionts, this may incur metabolic costs (Frade et al., 2008a,b; Cooper et al., 2011). However, a recent study examining the physiological properties of three depth-generalist octocorals on Eilat’s reefs (i.e., S. eilatensis, S. vrijmoethi, and S. leptoclados), showed that all three exhibit a similar trend of lower photo-physiological capabilities and nitrogen assimilation with increased depth (Ferrier-Pagès et al., 2021). Therefore, it seems that, as in the case of S. eilatensis, the shift in symbiont community across depth might not impact the functional physiology of the colonies with increased depth, relative to the congener S. vrijmoethi. It is possible, therefore, that the change in symbiont community structure of S. eilatensis has little or no functional significance and is, alternatively, driven by the host’s distribution and factors such as species lineage and reproduction (Frade et al., 2008a). Nonetheless, in light of the small sample size, caution must be applied, as the findings might not relate evenly to all the studied species. For example, although S. leptoclados colonies were found to host different ITS2-profile across depth (Figure 5), due to the small sample size (n = 2 for each depth habitat) this result was not found to be significant and awaits confirmation in future studies. In addition, three of the host species reported here were only found once during our collection surveys (Table 1), and therefore, although the information on their algal symbionts adds to the total diversity, further confirmation regarding their endosymbionts requires a larger sample size.

The current finding of Durusdinium type D5 in K. utinomii from both shallow and mesophotic colonies (Figure 4) is noteworthy. Symbionts of the genus Durusdinium (previously known as clade D) have been recorded in the warmer waters of the central and southern Red Sea (Sawall et al., 2014; Terraneo et al., 2019). It is commonly considered as ‘thermally tolerant’ (e.g., Durusdinium trenchii; Kennedy et al., 2015; Silverstein et al., 2017; Wham et al., 2017; but see Smith et al., 2017). However, several other Durusdinium types were found in scleractinian coral species from marginal habitats over a large latitudinal gradient while demonstrating high algal-specificity. The presence of Durusdinium in these species was suggested to be related to the vertical transmission of photosymbionts in these host corals (Lien et al., 2013; Lajeunesse et al., 2014). Future studies should examine the unique association between the Klyxum host and its Durusdinium symbionts, along with its possible selective advantages for the holobiont association complex.

In summary, despite being abundant in many reef ecosystems, Symbiodiniaceae associations with octocorals are still poorly understood, especially in regard to their diversity at the type-level in MCE habitats. The current study has revealed novel ITS2-types and profiles, previously not recorded in any other cnidarian-hosts. Additionally, we demonstrate that the diversity of these associations across depth, at the symbiont-lineage level, is likely to be driven by a strong algal-specificity. Such distinctive associations probably represent a co-diversification between the octocoral hosts and their micro-algae and, therefore, a thorough examination of these associations, using an NGS approach, may well yield further insights into the evolution and ecology of octocoral symbiosis with their photosymbionts.

The Sanger sequences of the ITS2 and cp23s markers are deposited in the GenBank database (National Center for Biotechnology Information, NCBI), under accession numbers OP120823–OP120848 and OP135806-OP135828. The Illumina sequences of the ITS2 marker are deposited under the biosample accessions SAMN30076298-SAMN30076348 at NCBI. The raw data, the sequence alignments, and the R scripts used in this work can be found at https://figshare.com/articles/dataset/CE-Octocoral-microalgae/20407596/1.

RL led the study, collected the samples, performed the laboratory work, analyzed the data, and wrote the first draft. DH supported and facilitated the laboratory work, and supervised the data analysis. YB and DH co-supervised the study, provided the resources, and edited the manuscript. All authors contributed to the article and approved the submitted version.

This research was supported by Horizon 2020 EU grant GA n° 634674 Project Acronym: TASCMAR- Tools and strategies to access to original bioactive compounds from cultivated marine invertebrates and associated symbionts. DH acknowledges the support of Israel Science Foundation (Grant Number 652/20).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.