INTRODUCTION
Millepora is a hydrozoan genus of so-called ‘fire corals’, which are well known for inflicting painful stings to humans and for being important reef-builders because of their large calcareous skeletons (Lewis, Reference Lewis2006). The distribution of this genus is limited to 50 m depth in tropical seas (Lewis, Reference Lewis1989, Reference Lewis2006; Cairns et al., Reference Cairns, Hoeksema and Van der Land1999), with a clear distinction between species in the Atlantic (Amaral et al., Reference Amaral, Broadhurst, Cairns and Schlenz2002, Reference Amaral, Steiner, Broadhurst and Cairns2008) and the Indo-Pacific (Razak & Hoeksema, Reference Razak and Hoeksema2003). There are seven recognized species in the Atlantic Ocean, which can be distinguished by their growth form, pore sizes and pore patterns (Boschma, Reference Boschma1948; de Weerdt, Reference De Weerdt1984; Amaral et al., Reference Amaral, Steiner, Broadhurst and Cairns2008). Millepora alcicornis Linnaeus, 1758 is the only species recorded from across the Atlantic, while three other species are only found in the Caribbean (M. complanata Lamarck, 1816, M. squarrosa Lamarck, 1816 and M. striata Duchassaing & Michelotti, 1864) and three are endemic to Brazil (M. braziliensis Verrill, 1868, M. laboreli Amaral, Reference Amaral, Steiner, Broadhurst and Cairns2008 and M. nitida Verrill, 1868).
The existence of Millepora at Ascension Island was first mentioned by Irving (Reference Irving and Sheppard2013), who stated that both Millepora alcicornis and M. complanata occur ‘in small patches of open area of bedrock’ from 8 to 15 m depth. Although he reported that specimens were collected during the Operation Origin Expedition in 1985, he did not mention their exact locality (Irving, Reference Irving and Sheppard2013: table 1) and how they were identified. His observations took place over the course of numerous dives, down to a depth of 45 m, reported on in an earlier publication (Irving, Reference Irving1989), which at the time did not indicate the presence of Millepora. So far, no photographic records were known of Millepora at Ascension and no published reports were available on museum specimens, despite the presence of a sample in the Natural History Museum in London (BMNH 1972.9.1.2) from English Bay, which was donated by the Ascension Historical Society and identified as Millepora alcicornis by Solene Whybrow (H. Zibrowius, personal communication).
As Millepora corals are notoriously difficult to identify (Boschma, Reference Boschma1948; Stearn & Riding, Reference Stearn and Riding1973; Amaral et al., Reference Amaral, Broadhurst, Cairns and Schlenz2002; Clemente et al., Reference Clemente, Rodríguez, Brito, Ramos, Monterroso and Hernández2011; Brown & Edmunds, Reference Brown and Edmunds2013), colonies were photographed and a sample for genetic analysis was taken during a recent expedition to Ascension Island. This evidence is used in the present study to supply conclusive information on the identity of this Millepora species as part of Ascension's shallow-water fauna.
MATERIALS AND METHODS
During an expedition to Ascension Island in August/September 2012, coral specimens were photographed in situ and collected by Dr Peter Wirtz during scuba-diving in a depth range of 0 to 25 m (Zibrowius et al., Reference Zibrowius, Wirtz, Nunes, Hoeksema and Benzoni2014).
Material
RMNH Coel. 40167 from Red Rock, near English Bay (07°53.654′S 14°23.676′W), 8 m depth, 6 September 2012: 8 small fragments (~1 cm in size) in 96% ethanol. Additional material from Ascension was available for morphological study: BMNH 1972.9.1.2 from English Bay (07°53′S 14°23′W), 1971: 4 dry specimens.
DNA was extracted from specimens preserved in ethanol using the DNeasy Blood and Tissue kit (Qiagen). The 16S subunit of the ribosomal RNA was amplified by polymerase chain reaction using published primers (Cunningham & Buss, Reference Cunningham and Buss1993) and sequenced on an ABI 3130XL gene analyser. The obtained sequences were compared with 16S sequences from M. alcicornis from the North and East Atlantic (Bermuda, Colombia, Florida, Panama and Cape Verde Islands) (N = 10), M. braziliensis (N = 10), M. laboreli (N = 9) and M. nitida (N = 10) (de Souza, Reference de Souza2013), and Millepora sp. fragments from the Canary Islands, Tenerife, Poris de Abona, 6 m, 26 September 2010, don. A. Brito, RMNH Coel. 39907 (N = 1). Pairwise genetic distance was calculated between the Ascension specimen and the 40 Millepora specimens using MEGA v5.2 in order to identify the Ascension specimens to the most similar species. The sequences have been deposited in GenBank (M. alcicornis Ascension Island MA KF871426; M alcicornis Tenerife MT1 KF871427).
Abbreviations: BMNH, The Natural History Museum, London, formerly British Museum (Natural History); RMNH, Naturalis Biodiversity Center, Leiden, formerly Rijksmuseum van Natuurlijke Historie.
RESULTS
Habitat
So far Millepora corals observed and sampled at Ascension (RMNH Coel 40167, Figure 1; BMNH 1972.9.1.2, Figure 2) were only found in English Bay and Red Rock at the north side of the island (for maps, see Price & John, Reference Price and John1980: figure 2; Irving, Reference Irving and Sheppard2013: figure 16.1). The underwater landscape of English Bay has been described as an area consisting of massive outcrops of bedrock with terraces, cliffs, small caves and underhangs (Irving, Reference Irving and Sheppard2013).
Morphology
The fragments collected in September 2012 consisted of tiny branch tips, which are unsuitable for identification based on pore patterns (De Weerdt, Reference De Weerdt1984). The coral colonies themselves, like the one from which the fragments were taken (Figure 1), are large encrusting plates with short, finger-like branches that may be laterally compressed at their tips, a growth pattern typical for M. alcicornis (Boschma, Reference Boschma1948; De Weerdt, Reference De Weerdt1984). A similar growth form is shown by Millepora specimens that were recently discovered at Tenerife (Clemente et al., Reference Clemente, Rodríguez, Brito, Ramos, Monterroso and Hernández2011; RMNH Coel. 39907). The BMNH specimens are 2–11 cm in length and show a low density of relatively small dactylopores (Figure 2), which is typical for M. alcicornis (see De Weerdt, Reference De Weerdt1984). Owing to their phenotypic variation, identification of milleporids based on morphological criteria is most reliable when a combination of characters is used (De Weerdt, Reference De Weerdt1984). Sizes of dactylopores (0.05–0.15 mm) and gastropores (0.15–0.25) in M. alcicornis are similar to those in M. striata but the density of dactylopores in the former (roughly 50–200 cm−2) is much lower than in the latter (300–500 cm−2) and similar to the density found in M. complanata, while the density of gastropores is less useful as a diagnostic character (De Weerdt, Reference De Weerdt1984). Specimens from Ascension (BMNH, Figure 2) show a low density (~150 cm−2) of the small dactylopores (~0.10 mm), whereas its gastropores are twice as large in diameter (~0.20 mm), which agree with data known from M. alcicornis (De Weerdt, Reference De Weerdt1984). Ampullae, organs in which Millepora medusae are kept (see Lewis, Reference Lewis1991), were not present in the material from Ascension.
Molecular genetics
The average pairwise genetic distance between the Millepora specimen collected in Ascension Island and individuals of M. alcicornis, M. braziliensis, M. nitida and M. laboreli was found to be 0.0062 ± 0.0042, 0.0786 ± 0.0015, 0.0830 ± 0 and 0.0906 ± 0.0039 respectively, while for the Millepora specimen from Canary Islands the average pairwise distances were 0.0081 ± 0.0016, 0.0759 ± 0.0019, 0.0760 ± 0, 0.0870 ± 0.0022. The genetic distance between the Millepora specimens from Ascension and Canary Islands was 0.009. Genetic distance is approximately one order of magnitude lower for M. alcicornis, compared with the other three species for both specimens from Ascension and Canary Islands. The most similar sequences (one M. alcicornis individual from Cape Verde and one from Florida) had only one mutational difference from the Ascension specimen. The molecular data confirm the identification of the Ascension and Canary Island specimens as that of M. alcicornis.
Conclusion
Although morphological characters are very variable in Millepora species, based on a combination of branch form, pore density and poor size, the Millepora species at Ascension is most likely M. alcicornis. This finding is supported by molecular evidence.
DISCUSSION
Occurrence
This second published record of Millepora from Ascension Island, after Irving's (Reference Irving and Sheppard2013), is the first with evidence from the field. Both records are remarkably recent while the BMNH specimens were already collected in 1971, confirming that natural history collections may be important as sources for earlier species records and the reconstruction of previous fauna compositions (Rainbow, Reference Rainbow2009; Luttikhuizen & Dekker, Reference Luttikhuizen and Dekker2010; Hoeksema et al., Reference Hoeksema, Van der Land, Van der Meij, Van Ofwegen, Reijnen, Van Soest and De Voogd2011; Hoeksema & Wirtz, Reference Hoeksema and Wirtz2013). An explanation for these scarce records may be that Millepora has a very restricted distribution at the northernmost tip of Ascension Island, or because most previous submarine sampling took place in shallow water, especially in the intertidal zone (compare Price & John, Reference Price and John1980; Reimer et al., Reference Reimer, Lorion, Irei, Hoeksema and Wirtz2014; Zibrowius et al., Reference Zibrowius, Wirtz, Nunes, Hoeksema and Benzoni2014). English Bay is among the most protected sites to wave exposure in Ascension (Irving, Reference Irving and Sheppard2013). It is possible that hydrodynamic turbulence and wave action in other localities prevent the settlement of Millepora in shallow water.
Identity
Based on morphological and genetic analyses, all specimens in the present study were identified as Millepora alcicornis. The earlier record of M. complanata (Irving, Reference Irving and Sheppard2013) could not be confirmed. In the field, specimens of both species may be hard to separate (Stearn & Riding, Reference Stearn and Riding1973; Brown & Edmunds, Reference Brown and Edmunds2013). Evidence suggests that they may form a ‘species complex’ (Ruiz-Ramos et al., Reference Ruiz-Ramos, Weil and Schizas2014).
Morphology
The growth form of Millepora alcicornis at Ascension predominantly consists of large crusts with small branches. Although M. alcicornis is generally known to show much ecophenotypic variation, this encrusting shape is relatively most common at shallow depths and on vertical rock surfaces, where it helps to resist wave action (Stearn & Riding, Reference Stearn and Riding1973; De Weerdt, Reference De Weerdt1981; Edmunds, Reference Edmunds1999).
Distribution range
Millepora alcicornis is the only Millepora species with an amphi-Atlantic distribution range. Millepora spp. are generally known as infrequent breeders and may apply fragmentation as an asexual reproduction strategy (Lewis, Reference Lewis1989, Reference Lewis2006; Edmunds, Reference Edmunds1999), favouring short-distance dispersal, which may explain why most Atlantic Millepora species have limited, western Atlantic ranges or very local distributions, as shown by Brazilian endemics (Amaral et al., Reference Amaral, Steiner, Broadhurst and Cairns2008). In contrast, most Indo-Pacific Millepora species have wide ranges (Boschma, Reference Boschma1948; Razak & Hoeksema, Reference Razak and Hoeksema2003). It is unclear why M. alcicornis has such a wide range, including not only Ascension but also the Cape Verde Islands (Boekschoten & Best, Reference Boekschoten and Best1988) and the Canary Islands (present results), from where it was reported as recently introduced (Clemente et al., Reference Clemente, Rodríguez, Brito, Ramos, Monterroso and Hernández2011). Although M. alcicornis has been recorded from very shallow depths (Lewis, Reference Lewis2006), it may be rare in the intertidal zone (Stearn & Riding, Reference Stearn and Riding1973). Since Millepora corals are able to settle as crusts on various substrates (De Weerdt, Reference De Weerdt1981; Lewis, Reference Lewis2006), species that would be able to survive well in shallow water could also be successful as rafters (Hoeksema et al., Reference Hoeksema, Roos and Cadée2012). Comparative studies on the reproduction, dispersal, settlement and phylogeography of the various Millepora species are required (cf. Nunes et al., Reference Nunes, Norris and Knowlton2011) in order to clarify the occurrence of M. alcicornis as a single Millepora species at Ascension and islands in the eastern Atlantic.
ACKNOWLEDGEMENTS
This paper was made possible by participation by Dr Peter Wirtz in the August–September 2012 expedition to Ascension Island on invitation by Dr Paul Brickle. We thank him for donating the coral fragments, the in situ photographs and locality information. The funding for this work came from a grant to the Shallow Marine Surveys Group from the Darwin Initiative (EIDCF012). Work by F.L.D. Nunes was supported by a grant from the Regional Council of Brittany, from the European Funds (ERDF) and supported by the ‘Laboratoire d'Excellence’ LabexMER (ANR-10-LABX-19) and co-funded by a grant from the French government under the program ‘Investissements d'Avenir’. We are grateful to the Shallow Marine Surveys Group and the South Atlantic Environmental Research Institute for organizing the expedition. We are also very grateful to the Ascension Island Government and the members of staff at the Conservation Centre and Ascension Island Dive Club for their cooperation, accommodation and hospitality. We are grateful to British Forces South Atlantic Islands for their logistical support. Dr Alberto Brito donated Millepora fragments from Tenerife, Canary Islands, to Naturalis for identification and biogeographic comparison. Dr Helmut Zibrowius provided photographs of BMNH specimens. Constructive remarks from two anonymous reviewers helped to improve the text.