Biological Invasions

, Volume 11, Issue 4, pp 995–1003

Do introduced endosymbiotic dinoflagellates ‘take’ to new hosts?


    • Department of BiologyPennsylvania State University
  • William Loh
    • Centre for Marine StudiesUniversity of Queensland
  • Robert K. Trench
    • St Georges
Original Paper

DOI: 10.1007/s10530-008-9311-5

Cite this article as:
LaJeunesse, T.C., Loh, W. & Trench, R.K. Biol Invasions (2009) 11: 995. doi:10.1007/s10530-008-9311-5


In a recent communication by Stat and Gates (Biol Invasions 10: 579–583, 2008), discovery of a symbiotic combination involving the coral Acropora cytherea and the dinoflagellate endosymbiont, SymbiodiniumA1 (Symbiodinium microadriaticum, Freudenthal sensu stricto) in the Northwest Hawaiian Islands was interpreted to be the result of a ‘recent’ introduction. While introductions of symbiotic dinoflagellates have occurred and are occurring, the authors’ conclusion was made without sufficient information about the geographic range and host specificity exhibited by A1. The only direct genetic analysis of symbionts from the putative host vector, a jellyfish in the genus Cassiopeia sp., from Kaneohe Bay on the Island of Oahu, found that it contained a different symbiont species, A3. Furthermore, Stat and Gates (Biol Invasions 10: 579–583, 2008) did not consider the importance of host-symbiont specificity in preventing the establishment of a foreign symbiont species. In comparison to A. cytherea, A. longicyathus on the southern most Great Barrier Reef also hosts SymbiodiniumA1 and a closely related endemic, A1a. Instead of assuming that A. cytherea has an unnatural association, a practical explanation is that long-term ecological and evolutionary processes influenced by local environments underlie the unusual, but not unprecedented finding of a Pacific acroporid associating with Clade A Symbiodinium spp.




Endosymbiotic dinoflagellates in the genus Symbiodinium are among the most abundant microbial eukaryotes found in shallow tropical marine environments. Their ability to proliferate and maintain dense populations in many reef-building or reef-dwelling invertebrates such as cnidarians has resulted in the ecological dominance of their host partners. Because cnidarian-algal symbioses are sensitive to thermal stress, there is interest in understanding their ecology and evolution, and how these symbioses respond to climate change.

The diversity, ecology, and biogeography of symbiotic dinoflagellates, are comparatively easy to investigate relative to other single celled eukaryotes because their primary habitat in the tissues of sedentary hosts allows for precise spatial and temporal sampling for molecular genetic analysis. What we know is that the genus Symbiodinium comprises numerous evolutionarily divergent lineages designated as Clades A through H (Coffroth and Santos 2005). Each of these groups consists of a diversity of genetic forms with distinct ecologies and geographic ranges and probably represents different species (LaJeunesse 2001, 2002, 2004). Among these, there are species known to associate with numerous host taxa (host-generalists) and many more that are specialized to particular host species or genera (host-specialists; LaJeunesse et al. 2003). Host-generalists appear to have wide geographic ranges, whereas host-specialists often are characteristic to certain biogeographic regions and are probably endemics (LaJeunesse 2004). Therefore, given the size of the Indo-Pacific, geographic variability in host-symbiont combinations is potentially great for many species of coral associating with regionally endemic symbionts.

Current understanding of the diversity as well as the geographic and host distribution of Symbiodinium spp. remains woefully limited in many parts of the tropics. Attempts to incorporate new observations within a theoretical framework originally established from limited geographic and habitat sampling, have confounded interpretations about the ecology of these symbioses (Baker et al. 2004; but see LaJeunesse et al. 2008). Because Pacific corals in the genus Acropora spp. predominantly associate with clade C Symbiodinium spp. (Van Oppen et al. 2001; LaJeunesse et al. 2003), the unexpected identification of SymbiodiniumA1 in Acropora cytherea from the French Frigate Shoals in the Northwestern Hawaiian Island chain prompted Stat and Gates (2008) to surmise that this symbiont was introduced to this region via the up-side-down mangrove jellyfish, Cassiopeia sp. If indeed this ‘unusual’ symbiosis represents a de novo combination, it would demonstrate a greater capacity for hosts and symbionts to change partners.

The documentation of microbial introductions into the marine environment has generally involved harmful bloom forming phytoplankton (Carlton and Geller 1993). The possibility that ‘symbiotic phytoplankton’ are also being distributed from ocean to ocean via ballast water and/or host vectors is important to consider, given the disruptive effect that sea surface warming is having on coral-algal symbioses (Hoegh-Guldberg 1999; Fitt et al. 2001). Partner combinations appear to have been altered during phases of major climate change (Rowan and Powers 1991; Baker 2003; LaJeunesse 2004). With present environmental conditions of the planet changing (IPCC 2007), new and potentially invasive, introductions could provide the source of novel symbiotic combinations.

Clearly, these dinoflagellates have free-living existences. The newly settled juveniles of many coral species must acquire compatible symbionts from the environment or perish. Adult host colonies release thousands to millions of symbiont cells daily (Titlyanov et al. 1996; Stimson et al. 2002). Entire coral reef communities therefore release vast numbers of Symbiodinium spp. that may be dispersed long distances by ocean currents and/or cargo ships.

The possibility that certain A. cytherea colonies in the Northwestern Hawaiian Islands associate with Symbiodinium A1 because this symbiont was introduced via a host vector (Stat and Gates 2008), is not substantiated by existing facts nor is it consistent with the specificity and stability displayed by many cnidarian-algal symbioses (Goulet 2006). Furthermore they do not consider an alternative and more pragmatic explanation. This has prompted a brief analysis of what is known about Acropora and Cassiopeia symbioses in the Pacific, and a review of the phylogeography and host specificity exhibited among Clade A Symbiodinium spp.

Materials and methods

Fragments (~6 cm2) from nine colonies of Acropora longicyathus were collected from the lagoon and fore-reef of One Tree Island on the Southern Great Barrier Reef (GBR), Australia (Fig. 1a, b). Tissue from each fragment was removed using an airgun attached to the low-pressure outlet of the first stage of an underwater dive regulator. The slurry was centrifuged (8,000 rpm, 10 min), the supernatant was discarded and the algal pellet was preserved in 20% DMSO buffer (Seutin et al. 1991) and stored at −20°C until further processing.
Fig. 1

(a) Populations of Acropora cytherea (French Frigate Shoals, Hawaii), A. longicyathus (Trunk Reef, central GBR; One Tree Island, southern GBR), and A. formosa (Tahiti) whose individuals are known to associate with Symbiodinium spp. in Clade A (b) A. longicyathus colony from One Tree Island, southern GBR (Australia) (c) ITS2-DGGE fingerprints showing colonies of A. longicyathus (One Tree Island) with populations of Symbiodinium C3, A1a, and A1, or mixtures of C3 and A1a

Nucleic acid extractions were conducted using a modified Promega Wizard genomic DNA extraction protocol (LaJeunesse et al. 2003). The internal transcribed spacer region (ITS) 2 was amplified from each extract using primers ITS2clamp and ITSintfor2 with a touch-down thermal cycle protocol used by LaJeunesse (2002). Products from these PCR reactions were electrophoresed for 10 h at 155 V on denaturing gradient gels (45–80%) using a C.B.S. Scientific system (Del Mar, CA). As described in detail by LaJeunesse (2002) and LaJeunesse et al. (2003), prominent bands characteristic of each unique fingerprint/profile were excised, re-amplified using a standard PCR profile (52°C annealing) using regular primers (ITSrev and ITSintfor2), and sequenced on an ABI 310 sequencer.

Sequences were aligned and compared against existing ITS 2 sequence datasets generated exclusively from ITS2-DGGE fingerprinting. An unrooted phylogenetic tree using maximum parsimony (with the software PAUP; Swofford 2000) was reconstructed base on sequences diagnostic of DGGE fingerprints used to characterize various Clade A Symbiodinium spp. Finally, a table of Clade A Symbiodinium diversity was assembled to describe the host and geographic origin where each DGGE fingerprint ‘type’ (i.e. barcode) has been found to occur.

Results and discussion

Probable origins of the symbiosis between A1 and Acroporacytherea

The identification of a Clade A Symbiodinium among colonies of Acropora from the Northwestern Hawaiian Islands represents an unusual, but not unprecedented observation. An unspecified Clade A Symbiodinium was identified in A. formosa colonies from a lagoon in Tahiti (Darius et al. 2000). Populations of Acropora longicyathus on the central and southern GBR consist of individuals that associate with Clade A Symbiodinium (Loh et al. 2001; Van Oppen et al. 2001), and is the only genus of scleractinian coral on the GBR known to harbor Clade A Symbiodinium (Loh et al. 1998; Van Oppen et al. 2001; LaJeunesse et al. 2003, 2004a). ITS-DGGE fingerprinting analyses of the symbionts in colonies from One Tree Island on the southern GBR found that many of them harbored SymbiodiniumC3, common to most other Acropora spp. in the region (LaJeunesse et al. 2003), but colonies also were found with SymbiodiniumA1 or A1a (Fig. 1c). The Clade A and C variation within and between colonies of A. longicyathus at high southern latitudes is conspicuously similar to the symbioses exhibited by A. cytherea at high northern latitudes (Stat and Gates 2008).

Phylogeographic analysis incorporating the known diversity of tropical clade A Symbiodinium characterized using ITS-DGGE identifies two distantly related ancestral host-generalists, A3 and A1 (Fig. 2). Radiating from each are lineages of host-specific and/or regionally endemic forms (Fig. 2, Table 1). Of these, A1a is unique to A. longicyathus and to the southern GBR (Table 1; Loh et al. 2001). While a comprehensive biogeographic database should establish patterns of endemism with greater confidence, the regionally unique symbiosis between A. longicyathus and these clade A Symbiodinium appear to be stable, long lived, natural associations (Gomez-Cabrera 2005). A description of the biogeographic and ecological patterns of coral-algal symbioses around the world would also help with identifying significant changes in partner combinations brought about by the invasion of a competitively dominant opportunist.
Fig. 2

An unrooted phylogenetic reconstruction based on maximum parsimony analysis of ITS 2 sequences. Each sequence corresponds to a specific DGGE fingerprint (barcode) diagnostic of an ecologically distinct Clade A Symbiodinium. Only A1 and A3 have been identified in hosts from the Indo-Pacific and Western Atlantic. Sequences published by Stat and Gates (2008) are shaded in grey and probably represent rare intragenomic rDNA variants recovered by bacterial cloning. Asterisks signify Symbiodinum spp. in culture, not yet discovered in field collected host samples. Branches are color-coded to signify regional origin, red = Red Sea; Orange = Indian Ocean; Blue = Pacific, Green = Western Atlantic

Table 1

Diversity of Clade A Symbiodinium, and their geographic, host, and depth distributions

Clade A Symbiodinium

Host taxa

Known geographic range

Depth range (m)

Accession numbers



Jellyfish (Cassiopeia xamachana,C. frondosa, Cotylorhiza tuberculata); Corals (Acropora cytherea, A. longicyathus, Pocillopora verrucosa, Stylophora pistillata)

Western Atlantic/Caribbean (Bahamas, Florida, Mexico); Mediterranean (Isreal); Red Sea (Saudi Arabia); Western Pacific (southern GBR); North Central Pacific (Hawaii)



LaJeunesse 2001, 2002; Stat and Gates 2008


Corals (Acropora longicyathus)

Western Pacific (southern GBR)



This paper


Corals (Acropora palmata, A. cervicornis, Stephoenocenia sp. Porites spp.); Hydrozoans (Millepora spp., Aglaopheonia sp.); Bivalves (Tridacna spp.); Jellyfish (Cassiopeia sp.)

Western Atlantic/Caribbean (Bahamas, Barbados, Belize, Florida, Mexico, St Croix); Western Pacific (Philippines); North Central Pacific (Hawaii); Eastern Indian Ocean (Tanzania)



Baillie et al. 2000; LaJeunesse 2001, 2002; LaJeunesse et al. 2004b


Corals (Acropora palmata)

Western Atlantic/Caribbean (Bahamas)





Corals (Porites astreoides, P. porites), Anemones (Aiptasia sp., Bartholomea sp., Condylactis gigantea), Hydrozoans (Myrionema sp.)

Western Atlantic/Caribbean (Bahamas, Barbados, Belize, Florida, Mexico, St. Croix)



LaJeunesse 2001, 2002


Corals (Porites astreoides, P. porites); Anemones (Stichodactyla sp.); Hydrozoans (Millepora alcicornis)

Western Atlantic/Caribbean (Bahamas, Barbados, Belize, Mexico)



LaJeunesse 2002


Bivalves (Tridacna spp.)

Western Pacific (Okinawa, Philippines)



Baillie et al. 2000; LaJeunesse et al. 2004a


Hydrozoans (Millepora platyphylla)

Western Pacific (southern GBR)



LaJeunesse et al. 2003


Hydrozoans (Millepora tenuis)

Western Pacific (northern GBR)





Octocorals (Stereonephthya cundabiluensis)

Red Sea (Isreal)



Barneah et al. 2007


Octocorals (Lithophyton arboreum)

Red Sea (Isreal)





Flat worms (Waminoa sp.)

Red Sea (Isreal)



Barneah et al. 2007


Zooanthids (Zoanthus sp.)

Eastern Pacific (Mexico)



LaJeunesse et al. 2008

A13 (A1.1)

Corals (Montastraea annularis complex, Porites porites-diseased/bleached tissues)

Western Atlantic/Caribbean (Puerto Rico, Barbados)



LaJeunesse 2001


Hydrozoans (Millepora spp.)

Eastern Indian Ocean (Tanzania)


EU792884 EU792885 EU792886



Hydrozoans (Millepora spp.)

Eastern Indian Ocean (Tanzania)





Hydrozoans (Millepora spp.)

Eastern Indian Ocean (Tanzania)




Genbank accession numbers of ITS 2 sequences derived from rDNA-DGGE fingerprinting are provided

Variation in symbiotic combinations involving Cassiopeia and Symbiodinum A1

The conjecture that the symbionts in A. cytherea originate from source populations of ‘invasive Cassiopeia sp.’ is not substantiated by existing data. Direct analysis of Cassiopeia sp. from Kaneohe Bay on the island Oahu shows that it harbors A3 instead of A1 (LaJeunesse et al. 2004b). SymbiodiniumA3 was isolated in the early 1980s, from Cassiopeia ‘mertensii’ on Oahu (culture 77; LaJeunesse 2001). Two of three cultures established by Dr. Robert A. Kinzie III from Cassiopeia sp. in Kaneohe Bay also appear to be A3 (Santos et al. 2002). Stat and Gates (2008) base their argument on the chloroplast ribosomal large sub-unit (cp23S) sequence of the third isolate, which matched that of A1, but the ITS region was never sequenced for verification. Because Cassiopeia populations on Oahu appear to have originated from different regions around the world (Holland et al. 2004), molecular genetic verification that A1 is found among these populations should have been conducted to support their hypothesis.

As molecular genetic investigations of the diversity and ecology of Symbiodinium expands to new regions and environments, new symbionts species and novel partner combinations are being characterized (Darius et al. 2000; LaJeunesse and Trench 2000; Loh et al. 2001; Rodriquez-Lanetty et al. 2001, 2003; LaJeunesse 2002; LaJeunesse et al. 2004a, b; Pochon et al. 2004; Schonberg and Loh 2005; Barneah et al. 2007; LaJeunesse et al. 2008). Observations of rare or unusual partnerships expand the understanding of the extent to which symbionts and hosts can associate with each other. These data are revealing a patchwork of various host-symbiont combinations characteristic of isolated coral reef ecosystems. This geographic mosaic appears to involve stable, regionally unique, host-symbiont associations, formed by long-term ecological and evolutionary processes (LaJeunesse 2004).

Symbiodinium microadriaticum, or A1, exemplifies the geographic mosaic qualities of these associations (Thompson 1999). This dinoflagellate was originally described from cultured isolates derived from Cassiopeia xamachana (Freudenthal 1962; Trench and Blank 1987). Because cultures are not always indicative of the dominant symbiont population (Santos et al. 2001; LaJeunesse 2002), subsequent analyses of these animals in their natural habitat verified that A1 was the common symbiont of this host in the Caribbean (LaJeunesse 2002, unpubl.). As diversity studies expanded to new regions and hosts, A1 was found to occur in other hosts including the scyphozoan, Cotylorhiza tuberculata, from the Mediterranean Sea (Fine and LaJeunesse unpubl.) and in colonies of Pocillopora damicornis and Stylophorapistillata from the Red Sea (Shick 2004; Table 1). This symbiont appears to be host-specific in certain regions, while in other locations it is generalized to multiple host taxa (LaJeunesse et al. 2004a). The development and application of population genetic markers for Symbiodinium A1 would offer a finer level of genetic resolution that could detect particular clonal lines that have spread to new regions and new hosts (Santos et al. 2004; Pettay and LaJeunesse 2007).

Intragenomic rDNA variation may confound estimates of diversity

Intragenomic rDNA sequence variation exists in Symbiodinium genomes (Thornhill et al. 2007). For this reason, a combination of ITS-DGGE and direct sequencing of bands cut from these gels provides a screening step important for identifying sequence markers that are evolutionarily stable (LaJeunesse and Pinzón 2007). DGGE analysis of rDNA produces a characteristic banding pattern, or fingerprint, of the dominant sequence or sequences that serve as a means of bar-coding ecologically distinct and genetically isolated populations (i.e. species, LaJeunesse 2002; Pettay and LaJeunesse 2007). A large proportion of ribosomal sequence variants recovered by investigators sequencing from bacterially cloned PCR amplifications appear to be pseudogenes of the dominant copy (Thornhill et al. 2007). While there are potentially many sequence variants throughout the tandem cistrons of the ribosomal array, each one is rare (low in copy number), and/or may occur as a single copy in the genome.

Intragenomic variation probably explains the minor sequence variants recovered by Stat and Gates (2008) that cluster within one base change of A1 but do not match any of the other dominant marker sequences diagnostic of ecologically distinct Symbiodinium spp. (Fig. 2). The most commonly recovered sequence from bacterial cloning corresponds to the main band of the A1 fingerprint (Fig. 1c). When bacterial cloning of ITS sequences was applied to cultured iso-clonal cell lines, a similar sequence relationship involving the common sequence variant and rare singletons was recovered (Thornhill et al. 2007). Based on these observations, it is inferred that the Clade A sequence diversity reported by Stat and Gates are intragenomic variants of A1 (Fig. 2).

The potential for biological invasion of an opportunistic symbiont

The duration for which a Symbiodinium cell can remain and even divide in the external environment depends on many factors and probably differs from species to species. Stat and Gates (2008) detected Symbiodinium D in the ballast water of a shipping vessel in Honolulu harbor. The ITS sequence they report does not correspond to any known marker sequences for members of this Symbiodinium clade. Their use of bacterial cloning complicates the determination of how this ballast water resident relates to other clade D Symbiodinium because the sequence may represent a rare intragenomic variant that is not representative of the dominant copy in the genome (Fig. 2; Thornhill et al. 2007). Their sequence was one base change different from the diagnostic sequence of SymbiodiniumD1a-f (Genbank # EU812742) that occurs in Montipora capitata and M. patula from Hawaii (conservatively listed as D1a in LaJeunesse et al. 2004b). This brings up the possibility that they cloned a variant sequence from the genome of a local Symbiodinium D species that is the resident of two common species of coral in Hawaii.

A particular Clade D Symbiodinium, D1a, has been identified in a variety of hosts from the GBR, Thailand, Zanzibar, Persian Gulf, and Western Atlantic (LaJeunesse 2002, LaJeunesse et al. 2003, 2004a; Mostafavi et al. 2007; unpubl.). Perhaps this species has the physiological qualities that allow it to tolerate long distance dispersal, which would explain its wide geographic range. Limited ecological analyses conducted on D1a in the Caribbean suggest that its physiology allows it to grow and persist in hosts from stressful environments where other symbiont species cannot survive (cf. Toller et al. 2001). If and when conditions improve, Dla populations are typically displaced by symbionts that are more specific (better adapted?) to the host (Thornhill et al. 2006b). Of all the Symbiodinium spp. examined ecologically, invasion of D1a into a coral community exposed to severe stress is a possibility.

Micro-organisms are continually being introduced by human activity to new regions and habitats. While successful invasions have had serious economic and ecological impacts (Ruiz et al. 1997; Pimentel 2002), it is unknown whether introductions of mutualistic symbionts will have any effect on native symbioses. It is possible that an introduced host-generalist symbiont could spread through a host population by the infection of their offspring (Coffroth et al. 2001) or by invading and displacing the native symbionts of stressed adult colonies. Introductions of symbiotic Indo-Pacific cnidarians have occurred to the Caribbean and include the mushroom coral, Fungia scutaria (LaJeunesse et al. 2005) and the jellyfish, Phyllorhiza punctata, (Graham et al. 2003). Recent analyses of F. scutaria introduced to Jamaica 35 years ago found that it still contained an Indo-Pacific species of Symbiodinium (LaJeunesse et al. 2005). Despite its exposure to diverse and abundant sources of Caribbean Symbiodinium, the symbiont populations in this introduced coral remained stable. The long-term fidelity exhibited by colonies of various species of coral, analyzed for many years (Goulet and Coffroth 2003; LaJeunesse et al. 2004b; Thornhill et al. 2006a,b) suggests that most introductions of foreign Symbiodinium will have no effect on these symbioses.


Thanks to D. Tye Pettay (Penn State) for bringing the Stat and Gates paper to our attention and to Maria del Carmen Gomez-Cabrera (University of Queensland) for sending her dissertation on Acropora longicyathus symbioses. Scott Santos (Auburn University) provided helpful comments. This work was funded by Pennsylvania State University.

Copyright information

© Springer Science+Business Media B.V. 2008