INTRODUCTION

The zooxanthellate-encrusting anemone genus Zoanthus is worldwide in distribution, and is commonly found on rocky and hard substrates in tropical and sub-tropical waters, and especially in coral reef environments. Uchida and Soyama (2001) list 5 species of Zoanthus as occurring in Japanese waters; Zoanthus sansibaricus (Pax and Mueller, 1957), Z. gnophodes (ibid.), Z. pacificus (first described in Walsh and Bowers, 1971), Z. erythrochloros (Pax and Mueller, 1957), and Z. vietnamensis (ibid.). For a summary of Z. pacificus, Z. sansibaricus, Z. gnophodes, and Z. erythrochloros, refer to the summary of diagnostic characteristics in Table 1 and the color photographs in Fig. 1. Historically, Zoanthus diagnostic characters for identification purposes have been oral disk color, polyp diameter, tentacle count, and mesentary count (Pax and Mueller, 1957; Walsh and Bowers, 1971; Uchida and Soyama, 2001). However, the classification and identification of the entire genus is in chaos (Burnett et al., 1995; Burnett et al., 1997). Very little work has been done on identification to the species level, although some work was done on Zoanthus species of the Great Barrier Reef and northern Australia, which found that “species” or discrete groups had considerable morphological variation (Burnett et al., 1997). Fautin's (2003) database includes 320 descriptions of Zoanthidea, but only 16 have proper species descriptions. Similarly, there were almost 120 species of Zoanthus listed in this database (last updated in July 2003), but how many of these are actually true “species” is unknown (Fautin, 2003).

Table 1

Summary of Zoanthus species morphological characteristics in previous literature and this study.

Fig. 1

Photographs of presumed species samples. All samples shown were collected from field sites in summer 2003. a) Zoanthus erythrochloros, b) Z. gnophodes, c) Z. pacificus, d) Z. sansibaricus, e) Amami Zoanthus 4. Please note that Z. pacificus was sampled from all four field sites. Diagnostic characters for identification visible in these photographs are oral disk color, polyp diameter, and tentacle number. White bar scale = 1 cm.

At Sakurajima, Japan, alone, there exists over 20 different color variations of Zoanthus at one inter-tidal site (personal observation), including the five presumed Zoanthus species mentioned above. Whether these are separate species or simply color variations of the same species is unknown.

The cytochrome oxidase subunit 1 (COI) gene has been shown in previous literature to be an accurate species-level marker (Dawson and Jacobs, 2001; Erpenbeck et al., 2002; Otranto et al., 2003). We sequenced the COI gene from collected Zoanthus samples to investigate the following question.

Are the various morphological types of Zoanthus different species as described in previous literature (for example, Uchida and Soyama, 2001), or are they conspecific?

MATERIALS AND METHODS

Sampling and DNA extraction, and PCR amplification

Samples of Zoanthus spp. (Table 2, Fig. 1) containing Symbiodinium spp. were collected from four field sites (Fig. 2) in Kagoshima Prefecture, Japan, in June∼August 2003 and stored in 100% ethanol at −20°C. As samples were collected photographs were also taken to assist in identification and for collection of diagnostic character data (oral disk/polyp diameter and tentacle count) (Table 1, Fig. 1). The number of mesentaries of samples of the four presumed species Z. pacificus, Z. sansibaricus, Z. gnophodes, and Z. erythrochloros were counted from cross-sections (Table 1). During field sampling, Palythoa samples were collected from Amami and Yakushima field sites to provide outgroup sequences for the following phylogenetic analyses.

Table 2

List of collected samples with coloration, oral disk diameter, depth, and cytochrome oxidase subunit 1 (COI) sequence type.

Fig. 2

Map of field sites. The distance from the northernmost site at Kokubu to the southernmost site at Amami is approximately 400 km, and spans from temperate inland waters in the north to sub-tropical coral reefs on the open ocean at Amami.

Usual Zoanthus tissue has a large number of ZX, and this makes it difficult to obtain uncontaminated Zoanthus DNA (personal observation, and see Maier et al., 2001 for a review of such problems). However, polyps contained zooxanthellae (ZX)-free gametes. ZX-free gametes were carefully removed from sample polyps using a dissecting microscope, and immediately processed.

A total of 29 samples were analyzed. ZX-free gamete samples weighed 0.4∼1.8 mg. DNA was extracted from the samples using a spin-column DNeasy Animal DNA Extraction protocol (Qiagen, Santa Clarita, CA, USA) (LaJeunesse and Trench, 2000).

The primers used were the universal primers HCO2198 5′-TAAACTTCAGGGTGACCAAAAAATCA-3′ and LCO1490 5′-TAAACTTCAGGGTGACCAAAAAATCA-3′ (Folmer et al., 1994).

PCR amplification was performed on the samples under the following conditions: an initial denaturing step at 95.0°C for 1 minute, followed by 35 cycles of 1 minute denature at 95.0°C, 1 minute annealing at 40.0°C, and 90 s extension at 72.0°C, followed by 7 minutes' extension at 72.0°C.

RESULTS

Zoanthus COI sequences

The alignment of our obtained sequences is shown in Fig. 3. The six different sequences were submitted to Gen-Bank (Ascension Numbers  AB128893 to  AB128898). The phylogenetic tree based on the COI gene sequences is shown in Fig. 4.

Fig. 3

Mitochondrial cytochrome oxidase subunit 1 gene sequence alignment of Zoanthus and Palythoa samples. The sequences are 649 base pairs long.

Fig. 4

Phylogenetic analysis of the COI gene sequences from collected samples. Palythoa species were used to root the tree. Numbers at the nodes refer to percentage of bootstrap support.

All of our Zoanthus samples excepting Amami Zoanthus 4 (Fig. 1e) were found to be almost identical in their COI sequences (Figs. 3 and 4). In fact, all of the Zoanthus sample sequences were 100.00% identical over the 649 base pair length of the COI gene (designated “Zoanthus” in Fig. 3), excepting Sakurajima Zoanthus 1 (3 base pairs, or 0.46% different relative to the COI sequence “Zoanthus”) and Sakurajima Zoanthus 9, which differed by four base pairs (0.61%) from the “Zoanthus” sequence. Based on these results, it can be concluded that these samples are most likely intraspecific, despite their differing morphotypes. Presumed Z. pacificus sample COI sequences obtained from all 4 sites were 100.00% identical.

Amami Zoanthus 4′s sequence was different by 7 base pairs (1.07%) from the “Zoanthus” sequence, which indicates Amami Zoanthus 4 may be of a different species.

Palythoa tuberculosa sequences were shown to be 24 to 25 base pairs (3.69%∼3.85%) different from obtained Zoanthus sequences. The two sampled Palythoa (green oral disk from Amami and brown oral disk from Yakushima) differed by two base pairs from each other, indicating a relationship within the species level.

DISCUSSION

DNA sequencing was performed on the mitochondrial cytochrome c oxidase 1 (COI) subunit gene. The COI gene has been shown to be an adequate species-level marker (Dawson and Jacobs, 2001; Erpenbeck et al., 2002; Otranto et al., 2003). While investigating Aurelia (Cnidaria, Scyphozoa), Dawson and Jacobs (2001) suggested a 10% to 20% sequence difference as the benchmark of distinct species, based on their results and an examination of previous literature. In warble flies (Oestridae) Otranto et al. (2003) found intraspecific pairwise differences between 0.14%∼1.59%, while interspecific differences were 0.7%∼27%. While Dawson and Jacobs' (2001) benchmark value is much higher than the value calculated by Otranto et al. (2003), both Knowlton (2000) and Romano and Palumbi (1997) do mention that molecular evolution in cnidarians appears to be slower than in most other taxa. Supporting this hypothesis, France and Hoover (2002) found divergence levels in COI sequences among deep-sea octocorals were lower (0.4%∼10.3%) than expected.

From our present results, regardless of which COI species benchmark is used it is clear that all four sites harbor the same species of Zoanthus, despite morphological variation in oral disk color between samples. Variation in other diagnostic characters, such as polyp diameter, tentacle count, and mesentary count is much less pronounced and shows overlap in data ranges.

Only whether Amami Zoanthus 4 is a distinct and separate species or not is open to speculation, as the difference of 1.07% in its COI sequence lies in the overlapping range between intra- and interspecific values. Morphologically Amami Zoanthus 4 showed slight differences as well, as polyps were found to be individual and not clonal unlike all other samples sequenced. Sequencing of different genes and further investigation may help clarify the classification of this sample.

Zoanthus larvae are thought to be able to live for long periods; a minimum of ∼3 weeks (Burnett et al., 1995) up to several weeks before settling (Ryland, 1997) which could explain how similar genotypes could be distributed over such a wide range. The persistence of clonal genotypes and the amount of energy directed into asexual reproduction (i.e. rapid budding and changes in colony size etc.) as opposed to sexual reproduction may also contribute to similar geno-types being widespread (Ryland, 1997).

Previous literature dealing with classification and identification of Zoanthus species has been scarce, but literature up to now has identified 5 species of Zoanthus (Z. vietnamensis, Z. sansibaricus, Z. gnophodes, Z. erythrochloros, Z. pacificus) living in Japanese waters (Uchida and Soyama, 2001). While we collected Z. vietnamensis samples, we were unable to perform successful sequencing of these samples. However, the other four previously identified species samples, as well as other Zoanthus morphotypes collected appear to be conspecific genetically and morphologically, excepting oral disk color. Similar results were reported by Burnett et al. (1997) in Australia, who found that Zoanthus samples of various color morphotypes (excepting Z. vietnamensis) were all most likely conspecifiic, based on allozyme electrophoretic analyses of fourteen allozyme loci coding for eleven different enzymes. However, their study did not investigate the COI gene, so it is impossible to speculate on whether our Zoanthus and the studied Australian Zoanthus are conspecific or not.

What is clear is that current Zoanthus classification is in need of serious revision and review. The number of Zoanthus species worldwide, estimated at 120 (Fautin, 2003), may be far lower if our results are any indication.

On the other hand, if these various color morphotypes are not separate species, what accounts for their seeming morphological plasticity in oral disk color? Unlike other organisms, it appears that host tissue color is controlled in Anthozoa (including Zoanthus) solely by the green fluorescent protein (GFP) gene (Kelmanson and Matz, 2003). Kelmanson and Matz (2003) have shown that phenotypic plasticity in the great star coral Montastraea cavernosa exists in color variation. They suggest that several different loci on the (GFP) gene code for three different colors that are expressed in varying degrees, resulting in color morphs (Kelmanson and Matz, 2003). Thus, despite possessing identical GFP genotypes, color morphotypes exist in nature due to differing GFP expression. Such color variation appears to have evolved independently in several different classes of Anthozoans (Labas et al., 2002). This may be the case with Zoanthus, and investigating GFP genotypes in our samples is the next logical step in investigating the mechanism of oral disk color variation in Zoanthus. The underlying reason for such variation in color expression is not yet understood, although this may be related to the environment (depth, lighting, etc.). Further data collection is necessary.

In other work currently being conducted, it appears that Zoanthus colonies at the four study sites in this study harbor Symbiodinium zooxanthellae of different clades (Reimer et al., unpublished data). Thus, the Zoanthus-Symbiodinium “holotype” may be genetically and physiologically different from site to site despite the host Zoanthus being genetically identical, allowing this symbiosis to adapt to a variety of environments.