OUP user menu

Microbial ecology of corals, sponges, and algae in mesophotic coral environments

Julie B. Olson, Christina A. Kellogg
DOI: http://dx.doi.org/10.1111/j.1574-6941.2010.00862.x 17-30 First published online: 1 July 2010


Mesophotic coral ecosystems that occur at depths from 30 to 200 m have historically been understudied and yet appear to support a diverse biological community. The microbiology of these systems is particularly poorly understood, especially with regard to the communities associated with corals, sponges, and algae. This lack of information is partly due to the problems associated with gaining access to these environments and poor reproducibility across sampling methods. To summarize what is known about the microbiology of these ecosystems and to highlight areas where research is urgently needed, an overview of the current state of knowledge is presented. Emphasis is placed on the characterization of microbial populations, both prokaryotic and eukaryotic, associated with corals, sponges, and algae and the factors that influence microbial community structure. In topic areas where virtually nothing is known from mesophotic environments, the knowledge pertaining to shallow-water ecosystems is summarized to provide a starting point for a discussion on what might be expected in the mesophotic zone.

  • marine
  • invertebrates
  • microorganisms
  • low-light
  • depth


Mesophotic coral ecosystems (MCEs) are deep fore reef communities that occur at intermediate depths (30–200 m) of the photic zone in which light-dependent (zooxanthellate) corals are present (Fig. 1). Also found in these low-light ecosystems are azooxanthellate scleractinian corals, macroalgae, and sponges. MCEs are typically found along island and continental slopes as well as on top of seamounts, and yet are distinct from ‘true’ deep-water azooxanthellate coral ecosystems that occur at greater depths and that are not light-dependent (Fig. 1). Recent work has demonstrated that MCEs are much more prevalent than previously thought (R. Ginsburg, pers. commun.) and manifest a higher percentage of coral cover (40–60% in MCEs compared with <20% at shallower depths; Bak et al., 2005; Menza et al., 2008), suggesting that they may play a critical role in the connectivity/continuance of coral reef environments. Because mesophotic habitats are below standard self-contained underwater breathing apparatus (SCUBA) depths, research is dependent on technical diving using mixed gases, submersibles, or remotely operated vehicles (ROVs). As a result, considerable work has been carried out both above (using conventional SCUBA) and below mesophotic depths (where ROVs are more commonly deployed), but this ‘twilight zone’ in the middle has been largely ignored, particularly in terms of microbiology. A significant knowledge gap exists for these environments.


Types of reef ecosystems and their depth ranges (after Lesser et al., 2009). Shallow reefs are characterized by multiple species of photosynthetic corals, sponges, and algae. Mesophotic reefs have a lower diversity in terms of species composition, but typically have higher percentages of coral cover. The corals and sponges may show altered morphologies (e.g. flattened to expose more surface area to light). These reefs have a combination of photosynthetic and azooxanthellate corals. Deep reefs are below the photic zone and so lack photosynthetic organisms. They are characterized by azooxanthellate corals and sponges. The gradient arrows highlight generalized trends across the depth gradient of these reef ecosystems.

MCEs have recently been attracting considerable attention as shallow-water coral reefs continue to be degraded by an assortment of factors including global climate change, overfishing, and pollution (e.g. Hoegh-Guldberg et al., 2007; Carpenter et al., 2008), while deeper reefs appear to be less impacted (Bak et al., 2005; Menza et al., 2008). Many shallow-water sessile organisms have ranges that extend to mesophotic (30–200 m) depths, making these deep reefs critical environments for research as shallow coral reefs deteriorate. Mesophotic environments may serve as ‘safe havens’ or refugia for reef species because their depth makes them less susceptible to both natural and anthropogenic impacts (Bak et al., 2005).

Because water masses are mobile and the varying topography of MCEs may or may not include sediment, this review focuses on the microbial ecology of the dominant sessile macroorganisms found in MCEs. Numerous studies have indicated that coral and algal diversity (measured as number of genera) decreases with increasing depth, which is thought to be driven by the attenuation of light and depth. Conversely, several studies have shown that sponge diversity increases with increasing depth (Liddell et al., 1997). Microbial communities associated with living organisms, such as corals, sponges, and algae, are influenced not only by physiochemical environmental factors but also by ecological interactions with their host organism. There is growing evidence that some bacterial-community associations (e.g. Rohwer et al., 2002; Hentschel et al., 2006; Longford et al., 2007) and many zooxanthellae (Goulet, 2006) are species-specific to their host, suggesting that these symbiotic microorganisms are functioning as an integral part of the macroorganism's biology.

A review of the ecology of MCEs and the factors that influence their productivity (i.e. light, nutrients) was recently published (Lesser et al., 2009). This paper seeks to extend that information to the microbial scale by summarizing the little microbial research that has been conducted on mesophotic samples, speculating on the microbial ecology of these environments based on existing knowledge of shallow- and deep-water coral ecosystems, and finally highlighting the research needs.

Host-associated microorganisms

Microbiology is a key part of coral, sponge, and algal biology. These plants and invertebrates are actually holobionts: metaorganisms comprised of the macroscopic host, algal symbionts (zooxanthellae), bacteria, archaea, fungi, and viruses (e.g. Armstrong et al., 2001; Rosenberg et al., 2007a; Taylor et al., 2007). The literature regarding host-associated microbial communities tends to be very host specific, such that the similarities (e.g. complex microbial communities that are species-specific and distinct from the water column, presence of epibiotic and endosymbionts, and microorganisms that produce bioactive compounds) may be overlooked. The past decade has seen the application of molecular methods to studies of coral and sponge microbial ecology, while investigations of algae have lagged behind. In all cases, a clear sampling bias toward the convenience of shallow-water sampling (Menza et al., 2008) has resulted in a relative dearth of information regarding these organisms in mesophotic zones. Therefore, much of what will be discussed comes from studies of shallow- and deep-water host-associated microbial communities. Table 1 lists coral- and sponge-microbial ecology studies that include mesophotic depths. Table 2 lists shallow-water microbiology studies of algae that are also known to occur at mesophotic depths.

View this table:

Studies of macroorganism-associated microorganisms at or including mesophotic depths

Macro speciesMicrobial groupDepth (m)Reference
Cirrhipathes lutkeniBacteria Zooxanthellae45Santiago-Vázquez (2007)
Iciligorgia schrammiBacteria Zooxanthellae45Brück (2007)
Leptogorgia minimataBacteria Zooxanthellae45Brück (2007)
Leptoseris spp.Zooxanthellae67–100Chan (2009)
Madracis carmabiZooxanthellae5–40Frade (2008)
Madracis decactisZooxanthellae4.7–40Diekmann (2002), Frade (2008)
Madracis formosaZooxanthellae5–43Diekmann (2002), Frade (2008)
Madracis mirabilisZooxanthellae5–40Frade (2008)
Madracis pharensisZooxanthellae4.3–40Diekmann (2002), Frade (2008)
Madracis senariaZooxanthellae5–40Diekmann (2002), Frade (2008)
Montastraea cavernosaBacteria3–46Lesser (2007)
Montastraea franksiZooxanthellae>35Toller (2001)
Oculina varicosaZooxanthellae6–80Reed (2002)
Swiftia exertiaBacteria Zooxanthellae45Brück (2007)
Aplysina cavernicolaBacteria40 to 7–15Thoms (2003)
Aplysina fistularisBacteria4 to 100Maldonado & Young (1998)
Ceratoporella nicholsoniBacteria30Santavy & Colwell (1990)
Discodermia dissolutaBacteria30–60Schirmer (2005)
Ircinia felixBacteria4 to 100–200Maldonado & Young (1998)
Various speciesBacteria>33Sfanos (2005)
  • * A dash indicates a range of depths addressed in the paper; the word ‘to’ indicates transplantation between depths.

  • These are azooxanthellate corals, but were tested by molecular methods to confirm the absence of algal symbionts.

View this table:

Bacterial studies with direct relevance to mesophotic algal species

Algal speciesDeepest observed depth (m)Bacterial study referenceDepth reference
Codium spp.60Kong & Chan (1979)Leichter (2008)
Dictyota spp.70Jensen & Fenical (1995)Littler (1989)
Gracilaria spp.60Largo (1997)Leichter (2008)
Halimeda copiosa60Jensen & Fenical (1995), Jensen (1996)Leichter (2008)
Halimeda tuna70Colombo (1978)Littler (1989)
Lithothamnlum pseudosorum65Johnson (1991a, b)Johnson (1991a, b)
Lobophora variegata100Jensen & Fenical (1995), Jensen (1996)Littler (1989)
Polysiphonia spp.70Chan & McManus (1969), Kong & Chan (1979), Sieburth & Tootle (1981)Leichter (2008)
Rhipocephalus phoenix60Jensen & Fenical (1995)Leichter (2008)
Stypopodium zonale80Jensen & Fenical (1995)Littler (1989)
Udotea spp.70Colombo (1978), Jensen & Fenical (1995)Leichter (2008)
Ulva lactuca50Kong & Chan (1979)Leichter (2008)
  • Although the bacterial studies were conducted at shallower depths, these algae are also found below 30 m.


The dinoflagellate Symbiodinium spp. is commonly found as a symbiont of scleractinian corals. Genetically different groups of Symbiodinium have been classified into eight zooxanthellae lineages, termed clades A–H (Rowan, 1998; Baker, 2003). Zooxanthellae have been identified from various coral species and appear adapted to different levels of irradiance and temperature, independent of their hosts' tolerances (Rowan et al., 1997). It is debated how flexible the association is between corals and zooxanthellae, because some species of corals are able to host zooxanthellae from more than one clade, while others seem to be restricted to a single clade (Baker, 2003; Goulet, 2006). Oculina varicosa has both a zooxanthellate morph in shallow water and an azooxanthellate morph in deep water; when transplanted from 6 to 80 m, the corals lost their symbionts within 4 months (Reed, 2002). Depth zonation of zooxanthellae clades has been observed in Montastraea spp., with clades A and B more common in high-irradiance environments, and clade C found in low-irradiance environments and below 30 m (Klaus et al., 2007). However, clade D replaced clade C in colonies of Montastraea franksi below 35 m (Toller et al., 2001). Recent work with ITS2 sequences from Symbiodinium inhabiting Montastraea cavernosa revealed that members of clade C were most common across a depth range of 3–91 m, but that novel subclade sequences were dominant at depths >60 m (Lesser et al., 2010). Madracis spp. were found always to contain clade B symbionts, regardless of depth (Diekmann et al., 2002), but there was variability at the subclade level based on host specificity and depth zonation (Frade et al., 2008). Even clade A symbionts have been found at mesophotic depths in the Caribbean (Baker & Rowan, 1997). Moreover, in Pacific corals, depth gradients are marked by intracladal shifts of clade C symbionts rather than shifts to a different clade (Baker, 2003; Chan et al., 2009).

Symbiodinium spp. have been reported from some species of several genera of boring sponges, including Cliona, Anthosigmella, and Spirastrella (Wilkinson, 1992 and references within). Schönberg & Loh (2005) reported distinct populations of a new subclade of G-type Symbiodinium within Cliona orientalis sponges and proposed vertical transmission of these symbionts based on their strong host specificity over a large geographic range. Granados (2008) identified three clades of Symbiodinium (A, B, and G) inhabiting Caribbean sponges of the genus Cliona. Clade G had not been reported previously in any organism from the Atlantic Ocean. Although Cliona spp. are found within mesophotic depths, no deep-water sponge samples have been examined for the presence of these dinoflagellates. The associations with other sponge-inhabiting dinoflagellates have been summarized in a recent review by Taylor (2007).

Both Symbiodinium spp. and other epiphytic dinoflagellates have been found to be associated with macroalgae. Porto (2008) reported communities of free-living Symbiodinium within clades A, B, and C from macroalgal beds at 3–30 m depths on Caribbean reefs. The macroalgae examined included Halimeda spp., Lobophora variegata, Amphiroa spp., Caulerpa spp., and Dictyota spp., many of which are seen in mesophotic environments (Caribbean: M. Slattery & M. Lesser, pers. commun.; Hawaii: Kahng & Kelley, 2007; Florida: Leichter et al., 2008). No studies have specifically examined Symbiodinium from macroalgae at mesophotic depths.

Based on the variety of clades that have thus far been observed in both shallow water and mesophotic depths, there is no trend that will allow absolute prediction of the type of zooxanthellae that will occur at a specific depth range. It is also unclear how the diversity of clades present across different host groups (i.e. corals, sponges, and macroalgae) may affect the resilience of the ecosystem.


Fungi have been known since the mid-1800s to be associated with corals and have been detected in both shallow-water (Bentis et al., 2000) and deep-sea corals (Freiwald et al., 1997; Kellogg, 2008). Most references are to unclassified endolithic fungal hyphae observed by microscopy (e.g. Freiwald et al., 1997; Bentis et al., 2000) or indirect observations based on metagenomic surveys (Wegley et al., 2007; Kimes et al., 2010). Specific groups that have been identified associated with corals include Aspergillus spp. (Smith et al., 1996; Priess et al., 2000), Basidiomycota (Domart-Coulon et al., 2004), and Ascomycota (Yarden et al., 2007). Fungi are likely present in mesophotic corals, although no studies have confirmed this as yet. Most reports from tropical coral species suggest that fungi are behaving as pathogens (e.g. Nagelkerken et al., 1997; Priess et al., 2000; Yarden et al., 2007). However, fungi have also been observed in healthy corals (Ravindran et al., 2001; Wegley et al., 2007), which may indicate that some fungi act as commensals or that some corals are able to keep these pathogens in check (Le Campion-Alsumard et al., 1995; Kim et al., 2000).

Several studies have isolated fungi from shallow-water sponges (e.g. Höller et al., 2000; Wang et al., 2008), revealing considerable diversity, including multiple genera of Ascomycota, Zygomycota, and mitosporic fungi. Some attempts have been made to differentiate between obligate marine fungi and facultative marine fungi, but still little is known. To date, only one study has used molecular genetic methods to characterize the fungal communities within two species of shallow-water sponges (Gao et al., 2008). It showed that sponges appear to support fungal communities that are distinct between species and from the ambient water column. Currently, no work has been published regarding the fungal associates of mesophotic sponges.

Fungi are well documented to use macroalgae as hosts, and a number of potentially pathogenic species have been reported (Raghukumar et al., 1992; Correa, 1997 and references within). Invasion of fungi into algal tissues has been associated with a variety of symptoms ranging from necrotic lesions (with tissue loss) to nondecaying galls that may result in host deformation (Correa, 1997). Fungi are also found as members of epiphytic communities. Interestingly, a wide variety of macroalgae has been shown to produce antifungal compounds, which should allow the host alga to regulate the composition of the fungal populations associated with it (del Val et al., 2001; Kubanek et al., 2003). There do not appear to be any studies of algae-associated fungal communities from mesophotic depths, but fungi are expected to be present based on our knowledge of the communities associated with algal species found across a wide depth range.


Very little is known about coral-associated protists and nothing at all from mesophotic depths. A study in the Arabian Sea found thraustochytrid protists to be associated with the polyps of Porites spp., Pocillopora spp., and various Acropora spp., as well as mucus from several acroporids (Raghukumar & Balasubramanian, 1991). Enumeration using immunofluorescence detected 12–20 protists mL−1 of coral mucus (Raghukumar & Balasubramanian, 1991). Thraustochytrid protists have also been detected in the mucus of Fungia granulosa and some faviid species (Kramarsky-Winter et al., 2006; Harel et al., 2008). The protists did not seem to have a harmful effect on the corals, and it was proposed that the corals may consume the protists or indirectly benefit from protistan biochemical capabilities, such as the production of carotenoids or polyunsaturated fatty acids (Kramarsky-Winter et al., 2006). Conversely, a ciliate was determined to be the pathogen causing the skeletal eroding band in 24 species of Indo-Pacific corals (Antonius & Lipscomb, 2001).

Numerous diatoms have been reported from marine sponges, and these associations appear to be more common in polar regions (Wilkinson, 1992; Taylor et al., 2007; and references within both). Bavestrello (2000) reported the existence of parasitic diatoms in sponges collected from 100 to 120 m depths in the Ross Sea. These diatoms were shown to consume carbohydrates produced by the host sponge. Cerrano (2000) found that a hexactinellid sponge from Antarctic mesophotic depths also supported large diatom populations. Wilkinson (1992) noted two reports of cryptomonads from sponge species that can be found at mesophotic depths (Aplysina cavernicola and Stelletta spp.), but still little is known about this association.

Several studies have documented the presence of heterotrophic protists (Armstrong et al., 2000), naked amoebae (Rogerson, 1991), and diatoms (Lam et al., 2008) on shallow-water macroalgal species. No information was found regarding the presence of these organisms on mesophotic macroalgae; however, because some of these algal host species have also been documented from mesophotic depths (Table 2), a continuation of these associations is likely.


Research has shown that there are species-specific bacterial communities associated with corals (Ritchie & Smith, 1997; Rohwer et al., 2001; Rohwer et al., 2002; Bourne & Munn, 2005). Vibrionaceae and Alteromonadaceae are commonly found by culture-based studies (Rohwer et al., 2001; Bourne & Munn, 2005; Ritchie, 2006). These results are supported by molecular techniques including construction of 16S rRNA gene clone libraries and FISH, which have indicated that Gammaproteobacteria and Alphaproteobacteria dominate coral-associated communities (Rohwer et al., 2002; Bourne & Munn, 2005; Brück et al., 2007; Santiago-Vázquez et al., 2007; Lampert et al., 2008). Other bacterial taxa associated with corals include Firmicutes, Bacteroidetes, Planctomycetes, Actinomycetes, and Cyanobacteria (Rohwer et al., 2001; Frias-Lopez et al., 2002; Bourne & Munn, 2005; Lesser et al., 2007; Lampert et al., 2008). Cyanobacteria are of particular interest at mesophotic depths because decreased photosynthesis increases the ability of these bacteria to fix nitrogen (Lesser et al., 2007). However, FISH enumeration with a variety of probes to major taxonomic groups also shows that up to a quarter of the coral-associated bacterial community consists of unknown organisms (Santiago-Vázquez et al., 2007) and most studies reveal a large percentage of novel taxa (e.g. Lampert et al., 2008). Much work clearly remains to be carried out, both in shallow-water habitats and along the depth gradient into mesophotic environments, to better understand the bacterial diversity associated with corals.

Bacterial associates of sponges are incredibly diverse, representing at least 16 recognized phyla and one candidate phylum (Poribacteria). Based on 16S rRNA gene surveys, the most common phyla recovered include the Acidobacteria, Actinobacteria, and Chloroflexi (Hentschel et al., 2006). Only a few studies have examined sponges and their bacterial associates that live between 30 and 200 m depth (Table 1), and a number of these involved transplant experiments in which sponges were either taken from the mesophotic zone and moved to shallower environments or vice versa. Thoms (2003) took A. cavernicola sponges from depths >40 m and transplanted them to shallower, more light-exposed sites between 7 and 15 m depth. After 3 months, both transmission electron microscopy (TEM) and denaturing gradient gel electrophoresis analyses indicated that the associated bacterial community remained largely unchanged and no novel Cyanobacteria had been obtained. Maldonado & Young (1998) collected Aplysina fistularis and Ircinia felix sponges from shallow (4 m) reefs and transplanted them to 100, 200, or 300 m depths. None of the sponges survived at 300 m, while 62.5% of the A. fistularis and 42.8% of the I. felix survived at 100 m for 12 months, and 28.5% of the I. felix survived at 200 m. Contrary to reports of increased sponge growth at depth (Lesser, 2006; Trussell et al., 2006), Maldonado and Young found no significant differences in growth between the shallow and the deeper-water sponges. Cyanobacterial associates were lost in transplanted I. felix individuals and the sponges altered their morphology, presumably to enhance water flow, while transplanted A. fistularis sponges maintained their cyanobacterial communities and their original shape. Recent work by X. Gao and J.B. Olson (unpublished data) with three Caribbean sponge species collected across a depth gradient (9–60 m) showed that the sponges maintained a ‘core’ group of bacterial associates (ranging from ∼20% to 50% of the associated community in this study) and that the remainder of the community may be influenced by biotic and abiotic factors.

The Harbor Branch Oceanographic Marine Microbial Culture Collection showed six major culturable bacterial clades associated with mesophotic to deep-water sponges: Alpha-, Beta-, and Gammaproteobacteria; Bacteroidetes; Actinobacteria; and Firmicutes (Sfanos et al., 2005). The affiliations of these isolates are considerably different from what is indicated from molecular studies of mesophotic (Schirmer et al., 2005), deep-water (>200 m; Olson & McCarthy, 2005; Cassler et al., 2008; Meyer & Kuever, 2008), and shallow-water (<30 m; Hentschel et al., 2006 and references within) sponge-associated bacterial communities, where Acidobacteria, Actinobacteria, and Chloroflexi dominate.

Unlike corals and sponges, there do not appear to be any studies of algae-associated bacterial communities that have been conducted at mesophotic depths. However, studies in shallow-water systems include bacterial data on many algal species whose distributions include depths >30 m (Table 2). In addition, there have been bacteriological studies of several kelp species (Mazure & Field, 1980; Corre & Prieur, 1990; Vairappan et al., 2001), which have recently been shown to exist at mesophotic depths in tropical waters (Graham et al., 2007). Most studies of algae-associated bacteria have been culture-based, and enumeration (plate counts and microscopy) revealed 106–108 bacteria g−1 (wet weight) of algal biomass (Chan & McManus, 1969; Lewis et al., 1985; Jensen et al., 1996; Largo et al., 1997). Electron microscopy counts of endophytic bacteria in Udotea petriolata and Halimeda tuna found that the numbers declined with increasing depth from 0.5 to 6 m (Colombo, 1978). Common genera from cultivation studies include Flavobacterium spp. (reviewed in Bolinches et al., 1988), Bacillus, Vibrio, Pseudomonas, and Moraxella (Chan & McManus, 1969; Lewis et al., 1985). The application of fluorescently labeled probes showed Bacteroidetes, Alpha-, Beta-, and Gammaproteobacteria, Actinomyces, Planctomyces, and 30% that could not be identified (Hempel et al., 2008). The first study to use clone libraries and sequencing to compare bacterial diversity between a red macroalga and green alga found no species in common (Longford et al., 2007). Seven bacterial phyla (79 species) were identified from the red alga and four phyla (36 species) from the green alga, all within the Alpha-, Delta-, and Gammaproteobacteria, Planctomyces, and Bacteroidetes (Longford et al., 2007).

From the research that has been carried out, a few basic trends emerge. Like sponges and corals, algae have conserved and potentially species-specific bacterial communities (Kong & Chan, 1979; Lewis et al., 1985; Johnson et al., 1991a; Longford et al., 2007) that are distinct from the surrounding water (Kong & Chan, 1979; Lewis et al., 1985; Bolinches et al., 1988; Johnson et al., 1991a). Seasonal shifts, both in bacterial numbers and in species richness, have been documented (Chan & McManus, 1969; Laycock, 1974; Mazure & Field, 1980; Sieburth & Tootle, 1981; Bolinches et al., 1988). Different bacterial communities have been found to be associated with different parts of the algae, for example, the thallus vs. the frond (Laycock, 1974; Mazure & Field, 1980; Corre & Prieur, 1990; Hempel et al., 2008).


Archaea have recently been found to be commonly associated with zooxanthellate corals (Kellogg, 2004; Wegley et al., 2004, 2007; Siboni et al., 2008). They are not species-specific like coral-associated bacteria, but there are similar types of archaea on geographically distant corals of multiple species, indicating a ‘coral-specific’ group of archaea (Siboni et al., 2008). This group includes both crenarchaeotes and euryarchaeotes, and identification of archaeal amoA genes suggests that some are active in ammonia oxidation (Beman et al., 2007; Siboni et al., 2008). Attempts to detect archaea in the azooxanthellate deep-sea coral Lophelia pertusa have not been successful (Yakimov et al., 2006; Kellogg, 2008). No studies have examined coral-associated archaea in MCEs.

Most of the sponge-associated archaea, regardless of the depth at which the host sponge was found, are members of the Crenarchaeota (Taylor et al., 2007 and references within, Meyer & Kuever, 2008; Steger et al., 2008). A phylogenetically distinct group of crenarchaeal sequences has been recovered from sponge samples, suggesting the existence of specific sponge–archaeal associations (Holmes & Blanch, 2007). A few studies have shown the presence of members of the Euryarchaeota within sponges (Webster et al., 2001a; Pape et al., 2006; Holmes & Blanch, 2007).

No reference could be found to a study of algae-associated archaea. This is likely due to the fact that most microbial-ecology studies of algal surfaces have been culture-based (e.g. Bolinches et al., 1988) and archaea are difficult to cultivate. This is clearly an area ripe for research, both in mesophotic and in shallow-water ecosystems.


Viruses are known to be associated with corals (Wilson et al., 2005; Davy et al., 2006; Patten et al., 2008), sponges (Vacelet & Gallissian, 1978; Lohr et al., 2005), and macroalgae (Toth & Wilce, 1972; Clitheroe & Evans, 1974; Müller et al., 1990; Müller & Stache, 1992). Based on the morphology revealed by TEM (Wilson et al., 2005; Davy & Patten, 2007; Patten et al., 2008) and genetic similarity (Wegley et al., 2007), viruses associated with corals have multiple host targets: bacteria, archaea, zooxanthellae, fungi, and the coral animal. None of the existing marine viral literature addresses samples from mesophotic depths. That said, viruses are ubiquitous and are therefore expected to be present at some level in mesophotic hosts.

Environmental factors


The amount of irradiance has been shown to affect the microbial communities associated with both corals and sponges (Fig. 1). With increasing depth, the concentration of Symbiodinium in colonies of the coral M. cavernosa increased slightly, but the zooxanthellae had significantly higher concentrations of photosynthetic pigments (Lesser et al., 2010). Some colonies of M. cavernosa host nitrogen-fixing Cyanobacteria in their tissues (Lesser et al., 2004). Lower irradiance decreases the rate of photosynthesis, which increases the capacity for nitrogen fixation as the nitrogenase enzyme is inactivated in the presence of oxygen. As a result, the number of M. cavernosa colonies containing symbiotic Cyanobacteria increases with depth, from 0% at 3 m depth to ∼33% of colonies surveyed from mesophotic depths (30–46 m) demonstrating cyanobacterial symbionts (Lesser et al., 2007). Sponges have also been shown to support phototrophic associates whose populations are negatively affected (e.g. decrease in chlorophyll a concentrations) by a decrease in irradiance (Maldonado & Young, 1998; Becerro & Paul, 2004; Erwin & Thacker, 2008).

Based on these data, no clear trends can be identified for the effect of decreased irradiance on mesophotic species. Cyanobacterial species not involved in nitrogen fixation would likely decrease with increasing depth while the converse is expected for species involved in this process. Changes within the cyanobacterial community (and by proxy the concentration of any secondary metabolites produced by these organisms) are anticipated to have effects on the nonphotosynthetic microbial community. However, the nature of these effects is unpredictable and at this point, unstudied.

Water temperature

The role of temperature in coral bleaching has been well established, and recent studies show that bleaching results in a taxonomic and functional shift toward a Vibrio-dominated, coral-associated bacterial community (Ritchie, 2006; Bourne et al., 2008). Sponge-associated microbial communities may also be altered by changes in water temperature (Lopez-Legentil et al., 2008; Webster et al., 2008). The greater depth of mesophotic environments somewhat mitigates their exposure to surface-temperature increases, but it is expected that similar microbial community shifts would be observed in response to higher temperatures (Fig. 1).


Water masses below the thermocline possess higher concentrations of biologically available inorganic nitrogen and phosphorus than shallow warm waters (Fig. 1). Upwelling of this cold, nutrient-rich water can provide nutrients to otherwise relatively nutrient-deplete shallow-water ecosystems. The concentrations of both picoplankton and heterotrophic bacteria in the water column increase with depth (Lesser, 2006) and serve as the primary food resource for sponges. As a result, deep-water sponges demonstrate a higher growth rate than their shallow-water counterparts (Lesser, 2006; Trussell et al., 2006). The increase in sponge biomass provides an additional habitat for associated microorganisms. While less is known regarding mesophotic corals, as filter-feeding organisms, the increased availability of particulate matter is expected to also promote heterotrophic growth and subsequent maintenance of diverse microbial communities. The effects on the microorganisms associated with mesophotic algae are unknown, but the higher concentrations of available N and P in the ambient water likely promote the growth of a diverse microbial community on algal surfaces.


An environmental factor that is unlikely to have an ecologically significant effect on surface-associated bacterial-community structure in mesophotic ecosystems is anthropogenic pollution (Fig. 1). Mesophotic environments exist at >30 m depth and therefore are typically distant from point sources of pollution. Although direct experiments have shown that addition of carbon and nutrients to a coral can stimulate overgrowth of the associated bacterial community (e.g. Mitchell & Chet, 1975; Kline et al., 2006), actual environmental gradients show less of an effect. Coral-associated bacterial communities were stable across a gradient from low levels of human impact to contamination with sewage, persistent organic pollutants, and trace metals (Webster & Bourne, 2007). Another study that included depth (5–20 m) and pollution gradients found that significant differences between the bacterial communities of control and polluted sites were only evident at the shallowest depth (Klaus et al., 2007). Conversely, a study by Webster (2001b) demonstrated a drastic reduction in the diversity and abundance of bacteria associated with the sponge Rhopaloeides odorabile following exposure to copper.

Biotic factors

Host organism

Most of the coral, sponge, and algal species found in MCEs are also found in shallower depths; however, it remains to be determined whether the mesophotic dwellers differ genetically or physically. There are sometimes visible morphological changes (e.g. flattening to increase the surface area exposed to low light in corals) that indicate adaptive changes. Microbial communities have been shown to shift in response to changes in the coral host's physiological state (e.g. stress, bleaching, disease; Pantos et al., 2003; Gil-Agudelo et al., 2006; Bourne et al., 2008), and so some adaptations to mesophotic depths may change holobiont-associated microbiology.

Consistent microbial community structure may be achieved by vertical transmission of associates (or ‘symbionts’) in larvae or lateral transmission (acquisition from the surrounding environment). Studies have shown evidence of the vertical transmission of both yeast (Maldonado et al., 2005) and bacteria (Enticknap et al., 2006; Schmitt et al., 2007; Sharp et al., 2007) in sponges, and compelling arguments have been made for obtaining microorganisms from the surrounding environment (e.g. Hill, 2004; Webster, 2007). There do not appear to be any systematic trends determining the mode of acquisition of bacterial symbionts in sponges. It remains to be determined whether corals and algae use similar strategies to acquire and maintain their microbial communities.

While some microorganisms are desirable, others may not be. A variety of physical antifouling strategies are used to rid macroorganisms of unwanted microbial visitors. Corals can shed their surface mucus layer, which is heavily colonized by bacteria (Ducklow & Mitchell, 1979). Similarly, some algae are able to shed their outer layer of cells and mucus continuously (Filion-Myklebust & Norton, 1981; Moss, 1982) or allow erosion of the distal end of blades (Mann, 1973; Ott, 1980).

Another mechanism that may regulate microbial community structure is the production of bioactive compounds by the host organism or its symbionts. Algae (del Val et al., 2001; Kubanek et al., 2003; Lam et al., 2008), corals (Koh, 1997; Harder et al., 2003), and sponges (Dobretsov et al., 2005; Kelly et al., 2005; Lee et al., 2006 and others) have all been shown to contain secondary metabolites that act as antimicrobials. Hosts may also produce secondary metabolites that interfere with bacterial regulatory systems, such as quorum-sensing molecules (extracellular signals that allow bacteria to communicate) such as acylated homoserine lactones (Givskov et al., 1996; Kjelleberg et al., 1997).

Bleaching and diseases

Bleaching and diseases are mechanisms that can shift the composition of a microbial community, be it coral-, sponge-, or algae-associated. Whether infection by a primary pathogen or the rise of an opportunistic pathogen, bleaching, lesions, or other symptoms are macroscopic evidence of a change in the microbial component of the holobiont (algae: reviewed by Correa, 1997, corals: reviewed by Rosenberg et al., 2007b, sponges: reviewed by Webster, 2007). If lesions are a result of a primary pathogen, the potential for spread to mesophotic ecosystems would depend on the vector's (e.g. currents, mobile fauna) incursions into MCEs. If bleaching and diseases are more of a secondary response to environmental stressors, then MCEs should be protected by their depth, which mitigates the effects of increases in water temperature, irradiance, and anthropogenic pollution (see Environmental factors). Many coral-, sponge-, and algae-associated bacteria produce antibiotics that may exclude other bacteria from colonizing the host (Kelly et al., 2005; Rao et al., 2006; Ritchie, 2006), protecting the holobiont from bleaching or disease. How prevalent these probiotic associates are in MCEs compared with shallower reefs remains to be determined.

Effects on reef assemblages

Perhaps the most interesting role that algae-associated microorganisms may have is in affecting community structure at the reef/ecosystem level. Settlement and metamorphosis of the larvae from many macroscopic reef invertebrates (e.g. corals, sea urchins) have been shown to be induced by the bacterial communities on the surfaces of both macroalgae and crustose coralline algae (Johnson et al., 1991b; Johnson & Sutton, 1994; Negri et al., 2001; Dworjanyn & Pirozzi, 2008). The composition of these bacterial biofilms determines on a local level where corals recruit, the population size of grazing urchins, and whether an invasion of predatory Acanthaster planci will reduce coral cover. Could the higher coral cover, but lower species richness that has been observed in MCEs, be due in part to differences in the microbial biofilms on the available hard bottom?


Microbial ecology of shallow-water reef environments has developed into a recognized field of study only within the last decade. Mesophotic coral reef environments are a natural extension of that field and hold great promise for research, both from a biotechnological and from a management perspective. For example, a wide variety of biologically active molecules have been isolated from marine sponges and, in many cases, their associated microorganisms have been credited with producing these compounds (reviewed in Taylor et al., 2007). Collections of sponges showed that 50% more species from mesophotic depths demonstrated bioactivity than species from shallow-water environments (M. Slattery, pers. commun.). Additionally, for those sponge species that occurred over a depth range, individuals from deeper depths displayed higher concentrations of bioactive molecules than their shallow-water counterparts (M. Slattery, pers. commun.). In conjunction with their biotechnological potential, an understanding of the microbial component of the macroorganisms that live in MCEs is necessary for determining their capacity to serve as refugia for impacted shallow-water species. While these ‘twilight reefs’ may seem far removed from the coastal anthropogenic pressures that impact shallow reefs (e.g. sewage), many human activities (e.g. trawling, drilling, dredging, anchoring) still have the potential to impact MCEs negatively. As such, appropriate management practices may need to be introduced to protect these ecosystems, such as declaring areas ‘habitats of particular concern’ or marine protected areas. These management decisions rely on scientific data to characterize the uniqueness and resilience of these mesophotic reefs, and microbiology is a key underlying factor.

In shallow-water systems, molecular methods have been used to characterize the microbial associates of corals and sponges: 16S rRNA gene clone libraries (Webster et al., 2001a; Frias-Lopez et al., 2002; Hentschel et al., 2002; Rohwer et al., 2002; Kellogg, 2004; Wegley et al., 2004; Bourne & Munn, 2005; Webster & Bourne, 2007; Kennedy et al., 2008; Lampert et al., 2008), functional genes (Beman et al., 2007; Bayer et al., 2008; Mohamed et al., 2008; Siboni et al., 2008), FISH (Webster et al., 2001a; Wegley et al., 2004; Enticknap et al., 2006), and metagenomics (Yokouchi et al., 2006; Wegley et al., 2007). These techniques are just beginning to be applied to algae (Longford et al., 2007; Hempel et al., 2008). This suite of tools needs to be used over the entire depth range of photosynthetic coral environments in order to better understand the microbial diversity and functional ecology of microorganisms in MCEs.

As a whole, MCEs are a poorly understood, but potentially critical component of coral reefs. Even less is known about the microorganisms associated with corals, sponges, and algae found at mesophotic depths. For example, while studies of zooxanthellar clades in corals have included some mesophotic depths (Table 1), no investigation has been carried out on these dinoflagellates in association with sponges or algae. Bacterial studies have been conducted on a few mesophotic sponges (Table 1), but other than one study on coral-associated Cyanobacteria (Lesser et al., 2007), the only coral-associated studies from mesophotic depths involve azooxanthellate corals rather than photosynthetic scleractinians (Table 1). No mesophotic data exist for coral-, sponge-, or algae-associated archaea, fungi, viruses, or protists. These environments provide a wealth of opportunities to examine the connectivity between shallow- and deeper-water ecosystems and to begin to understand the role that microorganisms play in the success (or failure) of their host.


The authors thank H.L. Spaulding for recommending mesophotic algal references; Drs Slattery, Lesser, and Gochfeld for sharing data and offering editorial advice; and our anonymous reviewers for providing helpful suggestions. Thanks are also due to the other participants at NOAA's Mesophotic Coral Ecosystems Workshop for thought-provoking discussions. Funding for this work was provided to J.B.O. by NOAA-NIUST (grant number NA16RU1496) and NOAA-Ocean Exploration and to C.A.K. by USGS Coastal and Marine Geology Program and USGS Terrestrial, Freshwater, and Marine Ecosystems Program.


  • Editor: Ian Head


View Abstract