Introduction
Originating from the Indo-Pacific region, the ‘sun corals’ show a range of yellow and orange to reddish colour patterns attributed to the concentration of carotenoid pigments in the tissues (Lages et al., Reference Lages, Fleury, Pinto and Creed2010; Maia et al., Reference Maia, Ferreira, Costa, Lucas, Teixeira, Fleury, Edwards and de Oliveira2014; Galasso et al., Reference Galasso, Corinaldesi and Sansone2017). Tubastraea Lesson, 1830 is the most speciose genus spread across the Atlantic Ocean and classified as exotic organisms over the past nine decades (Vaughan and Wells, Reference Vaughan and Wells1943; Boschma, Reference Boschma1953; Cairns, Reference Cairns2000; Fenner, Reference Fenner2001; Fenner and Banks, Reference Fenner and Banks2004; Sammarco et al., Reference Sammarco, Porter and Cairns2010; Capel et al., Reference Capel, Creed, Kitahara, Chen and Zilberberg2019; Figueroa et al., Reference Figueroa, McClure, Figueroa and Hicks2019; Soares et al., Reference Soares, Salani, Paiva and Braga2020).
Tubastraea comprised 12 accepted species: Tubastraea coccinea Lesson, 1830; T. aurea (Quoy & Gaimard, 1833); T. micranthus (Ehrenberg, 1834); T. diaphana (Dana, 1846); T. stimpsonii (Verrill, 1866); T. faulkneri (Wells, Reference Wells1982); T. floreana (Wells, Reference Wells1982); T. tagusensis (Wells, Reference Wells1982); T. megacorallita Yiu, Chung & Qiu, Reference Yiu, Chung and Qiu2021; T. dendroida Yiu and Qiu, Reference Yiu and Qiu2022; T. chloromura Yiu & Qiu, Reference Yiu and Qiu2022; and T. violacea Yiu & Qiu, Reference Yiu and Qiu2022, and the four new Brazilian species validated in the present study: T. ramosa sp. nov., T. columnata sp. nov., T. grandidentata sp. nov., and T. megalostoma sp. nov. (Yiu et al., Reference Yiu, Chung and Qiu2021; Yiu and Qiu, Reference Yiu and Qiu2022; Serra et al., Reference Serra, Bastos, Coutinho, Schizas, Johnsson and Neves2024; Hoeksema and Cairns, Reference Hoeksema and Cairns2025). Based on the morphology of the coral skeleton, including macro and micromorphological characteristics of the corallites, the last four species were recently described from the Todos-os-Santos Bay (‘TSB’) (Bahia State, Northeastern Brazil) (Serra et al., Reference Serra, Bastos, Coutinho, Schizas, Johnsson and Neves2024).
For the first time, Tubastraea was reported outside its natural geographic range in Puerto Rico (Caribbean Sea) in the early 1940s (Vaughan and Wells, Reference Vaughan and Wells1943; Sammarco et al., Reference Sammarco, Porter and Cairns2010). In the late 1980s, colonies of sun corals were observed encrusted on an oil platform in the Campos Basin (22°S, Rio de Janeiro State, southwestern Brazil) (Castro and Pires, Reference Castro and Pires2001). Two sun coral species were later identified: T. coccinea and T. tagusensis (De Paula & Creed, Reference De Paula and Creed2004). This was the first record of Tubastraea in the Southwestern Atlantic. Previously, Laborel (Reference Laborel1974) mentioned the presence of Tubastraea in the Northeastern Atlantic, including the Gulf of Guinea, and northwards, to Gabon, Sierra Leone, and the Cape Verde Islands.
In 2008, divers alerted to the presence of sun corals in the Greek shipwreck Cavo Artemidi (13°03′.19″ S), 4.6 km from the coast, sunken between 9 and 29 m deep in the entrance of the TSB (Sampaio et al., Reference Sampaio, Miranda, Maia-Nogueira and Nunes2012). Henceforth, colonies were discovered in natural and artificial environments within the bay. At the same time, other reports along the Brazilian Northeast have been concomitantly done, being currently found from Bahia (13°S) to Ceara State (3°S), distributed across approximately 1,000 km along the Northeastern coastal zone (Sampaio et al., Reference Sampaio, Miranda, Maia-Nogueira and Nunes2012; Miranda et al., Reference Miranda, Costa, Lorders, Nunes and Barros2016a; Soares et al., Reference Soares, Davis and de Macêdo Carneiro2018).
Tubastraea colonies exhibit distinct morphologies, with morphotypes primarily classified according to two widely accepted models of corallite development for the group: plocoid (shorter corallites, e.g., T. coccinea and T. gradidentata sp. nov.) and dendroid (longer corallites, e.g., T. tagusensis and T. columnata sp. nov.). The latter may be branched (T. ramosa sp. nov.) or unbranched (T. columnata sp. nov.), while plocoid corallites may be clustered (T. coccinea) or sparsely distributed on the colony surface (T. faulkneri) (Serra et al., Reference Serra, Bastos, Coutinho, Schizas, Johnsson and Neves2024). Moreover, the calice diameter, number, and arrangement of septa have been traditionally used as taxonomic characters (Cairns, Reference Cairns2000, Reference Cairns2001; Cairns and Kitahara, Reference Cairns and Kitahara2012).
Studies focusing on the phylogenetic relationships of scleractinian corals, particularly within the family Dendrophylliidae, have been developed over the past few decades (Romano and Palumbi, Reference Romano and Palumbi1996; Fukami et al., Reference Fukami, Chen, Budd, Collins, Wallace, Chuang, Chen, Dai, Iwao, Sheppard and Knowlton2008; Kitahara et al., Reference Kitahara, Cairns, Stolarski, Blair and Miller2010; Arrigoni et al., Reference Arrigoni, Kitano, Stolarski, Hoeksema, Fukami, Stefani, Galli, Montano, Castoldi and Benzoni2014; Capel et al., Reference Capel, López, Moltó-Martín, Zilberberg, Creed, Knapp, Hernández, Forsman, Toonen and Kitahara2020; Bastos et al., Reference Bastos, Calazans, Altvater, Neves, Trujillo, Sharp, Hoffman and Coutinho2022; Yiu and Qiu, Reference Yiu and Qiu2022).
Romano and Palumbi (Reference Romano and Palumbi1996) developed one of the pioneering work on the phylogeny and evolution of Scleractinia using molecular analyses (mitochondrial 16S rDNA region), supporting two major clades: (1) ‘Robusta’, characterized by solid, heavily calcified skeletons, solid corallite walls, intratentacular budding, and more massive, plate-like colonies, and (2) ‘Complexa’, represented by less solid corals, with porous, synapticulothecate walls, extracalicinal budding, and branching, columnar, lamellar, and other colony forms.
Based on the mitochondrial COI gene, Kitahara et al. (Reference Kitahara, Cairns, Stolarski, Blair and Miller2010) attested the monophyly of the family Dendrophylliidae, pointing out that, among the azooxanthellate genera, Dendrophyllia, Tubastraea, and Balanophyllia may have diverged recently. Integrating molecular data (COI, IGR, and rDNA) with micromorphology (wall and septa projections and ornamentation) and microstructure (fibre arrangement), Arrigoni et al. (Reference Arrigoni, Kitano, Stolarski, Hoeksema, Fukami, Stefani, Galli, Montano, Castoldi and Benzoni2014) supported the monophyletic status of the Dendrophylliidae, indicating Tubastraea as the most well-supported group within the family. Nevertheless, the authors highlighted the relevance of micromorphological and microstructural analyses, emphasizing the organization of thickening deposits in Tubastraea as a robust autapomorphic character.
Adopting the integrative taxonomy (COI, IGR, and rDNA regions and morphological data), Capel et al. (Reference Capel, López, Moltó-Martín, Zilberberg, Creed, Knapp, Hernández, Forsman, Toonen and Kitahara2020) established a new genus of dendrophylliid corals for the Eastern Atlantic: Atlantia. Previously designated as Tubastraea caboverdiana, the branching Atlantia caboverdiana (Ocaña & Brito, 2015) remains restricted to the Cape Verde Archipelago, and, accordingly to the molecular reconstruction, the species is more closely related to Dendrophyllia and Leptopsammia than to Tubastraea (Capel et al., Reference Capel, López, Moltó-Martín, Zilberberg, Creed, Knapp, Hernández, Forsman, Toonen and Kitahara2020).
Using a similar approach (i.e., ITS sequences, macro-, and micromorphology), Bastos et al. (Reference Bastos, Calazans, Altvater, Neves, Trujillo, Sharp, Hoffman and Coutinho2022) identified three morphotypes of Tubastraea in Arraial do Cabo (22°S, Rio de Janeiro State, Southeastern Brazil). Morphotype I exhibits morphological patterns corresponding to T. aurea, and morphotype II to T. coccinea. However, I and II did not show significant molecular divergence with the ITS marker, which could differentiate them at the species level, and are considered morphological variations of T. coccinea. According to Serra et al. (Reference Serra, Bastos, Coutinho, Schizas, Johnsson and Neves2024), the dendroid morphotype III has morphological and molecular divergences from its congener, T. coccinea, and is related to T. columnata sp. nov.
Four new species of Tubastraea were described in Hong Kong, China Sea, between 2021 and 2022 (Yiu et al., Reference Yiu, Chung and Qiu2021; Yiu and Qiu, Reference Yiu and Qiu2022). The first one was T. megacorallita (Yiu et al., Reference Yiu, Chung and Qiu2021). Then, Yiu and Qiu (Reference Yiu and Qiu2022) published the other three new species, T. dendroida, T. chloromura, and T. violacea, also recovering the Tubastraea monophyly, and validating T. aurea as the sister group of T. diaphana.
Considering the recent integrative approaches and the description of eight new species in the last four years (2021–2025), Tubastraea has been taxonomically underestimated. Therefore, the present study provides the first molecular phylogeny of the Brazilian sun corals – T. grandidentata sp. nov., T. megalostoma sp. nov., T. columnata sp. nov., and T. ramosa sp. nov., using diagnostic new primers for Tubastraea DNA amplification. This work also compares two inference methods: phylogenetic reconstruction based on maximum likelihood (ML) and Bayesian inference (BI). Finally, the results support Serra et al. (Reference Serra, Bastos, Coutinho, Schizas, Johnsson and Neves2024), confirming the notable diversity of sun corals in the Atlantic Ocean.
Materials and methods
Study area and sampling
Sampling was carried out in three artificial environments of the TSB: Private Terminal, Salvador Capital (12°58′S, 38°30′W), Marina de Itaparica, Itaparica Is. (12°53′S, 38°41′W), and Magnetic Measurement Station, Itaparica Is. (12°52′S, 38°41′W) (Figure 1). Colonies were collected by SCUBA diving, from 0 to 10 m depth, using a hammer and spatula. Each sample was morphologically separated and placed individually in plastic bags. On board, colonies were washed in distilled water and stored in absolute ethanol. In the laboratory, tissue fragments were removed, and the colonies were bleached in 2% sodium hypochlorite for skeleton analysis. Species were identified based on their macro- and microstructures, following Serra et al. (Reference Serra, Bastos, Coutinho, Schizas, Johnsson and Neves2024). Individual data, including sample information, are available in Supplementary Table 1. The genetic variability and phylogenetic relationships were performed with 17 samples from six different species of Tubastraea: T. coccinea (n = 4), T. aurea (n = 2), T. grandidentata sp. nov. (n = 2), T. megalostoma sp. nov. (n = 3), T. columnata sp. nov. (n = 1), and T. ramosa sp. nov. (n = 5) (Supplementary Figure 1). Sequences of other Tubastraea species, A. caboverdiana and Dendrophyllia cornigera, were selected from publicly available GenBank datasets (Supplementary Table 1).
Localities of sun coral populations analysed in Todos-os-Santos Bay (TSB). (A) South America continent, indicating Brazil (light grey), the Bahia State (dark grey), and the TSB (red square). (B) TSB sampling localities: P1 = Private Terminal, Salvador Capital, P2 = Marina de Itaparica, Itaparica Is., P3 = Magnetic Measurement Station, Itaparica Is. (C) Artificial structures monitored during the research.
Nomenclatural acts
In accordance with the International Code of Zoological Nomenclature (ICZN), the electronic version of this article constitutes a validly published work. Consequently, all new taxonomic names proposed herein are effectively published through the electronic edition alone. This work and respective nomenclatural acts have been officially registered in ZooBank, the ICZN’s online registry. The associated Life Science Identifier (LSID) can be resolved and accessed by appending it to the URL http://zoobank.org/. The LSID for this publication is as follows: urn:lsid:zoobank.org:pub:41F6915D-67AE-4E73-B545-7352D2EA3451.
DNA extraction, amplification, and sequencing
Tissue fractions of 10 mg (weighted in a precision balance, Mettler Toledo model ME204/A) were removed from the polyp oral disc and then washed twice with distilled water, proceeding to a final wash with 0.5 mL of ultrapure water by vortex and short spinning. The DNA extractions used the DNeasy – Blood and Tissue kit (QIAGEN) following the manufacturer’s recommendations, considering an adaptation for the tissue lysis step. Before the QIAGEN protocol, the coral tissue was lysed mechanically into a 1.5 mL microcentrifuge tube and crushed with a small, sterile pestle.
Specific primers for Tubastraea COI, IGR, and rDNA regions were developed to optimize DNA amplifications (Table 1). Annealing temperatures and primer specifications are detailed in Table 1. During the polymerase chain reaction (PCR) reaction, the dsDNA was denatured at 94°C/2 min, then amplified in 30 cycles of 94°C/15 s, the specific annealing temperature/25 s, and 72°C/1 min, followed by a hold stage of 72°C/5 min and 8°C/∞. The reaction was performed on a final volume of 25 µL, using PCR buffer (Invitrogen, ref. Y02026B) with two units of Platinum Taq DNA Polymerase (Invitrogen, ref. 10966030), 2 mM MgCl2 (Invitrogen, Y02016B), 0.1 µM dNTP (Invitrogen, ref. 55082), 0.4 µM oligonucleotides, 0.25 µM ultrapure BSA (Invitrogen, AM2616), and DNA 0.6 µg. PCR clean-up was obtained by mixing 20% polyethylene glycol in a 1:1 (v/v) ratio, followed by incubation at 37°C for 15 min and centrifugation at 13,000 rpm for 15 min. For purification stage, the samples were subjected to three consecutive washes with 80% ethanol, followed by centrifugation at 13,000 rpm/5 min. The samples (20 µL) were then resuspended in ultrapure water for Sanger sequencing at the Oswaldo Cruz Foundation (FIOCRUZ, Bahia State).
Information on the DNA primer sets used in the study
Ps, primer set; T, target marker; S, sequence; AT, annealing temperature; PL, product length.
Phylogenetic analyses
The DNA traces were checked for quality and end-trimmed (final length: COI = 423 bp; IGR = 447 bp; rDNA = 400 bp), followed by the alignment with reference sequences from GenBank to assemble contigs using the software Geneious Prime 2023.2.1 (https://www.geneious.com). Multiple sequence alignments were obtained in ClustalW (Thompson et al., Reference Thompson, Higgins and Gibson1994) with the default parameters as implemented in MEGA11 (Tamura et al., Reference Tamura, Stecher and Kumar2021). The GenBank accession numbers and the alignments are available in Supplementary Table 1.
Phylogenetic analysis was inferred by ML and BI. The molecular markers were analysed independently and concatenated. ML trees were constructed in raxmlGUI 2.0 (Edler et al., Reference Edler, Klein, Antonelli and Silvestro2020) based on the best substitution model (HKY+G) according to the Bayesian information criterion (BIC), with 1,000 bootstraps. BI was performed in BEAST X v10.5.0 (Suchard et al., Reference Suchard, Lemey, Baele, Ayres, Drummond and Rambaut2018) with the following settings: 10,000,000 generations, four chains, and the HKY+G model with four gamma rate categories. Dendrophyllia cornigera and A. caboverdiana were used as the out-group. The burn-in period was set to 25%. Tree topologies were visualized and refined using the Interactive Tree of Life (iTOL) platform (https://itol.embl.de/). The figure with the ML and BI trees was created using the software CorelDRAW 2019.
Results
The new Tubastraea species from Brazil, described by Serra et al. (Reference Serra, Bastos, Coutinho, Schizas, Johnsson and Neves2024, Reference Serra, Bastos, Coutinho, Schizas, Johnsson and Neves2025) and used in the present study to perform phylogenetic analysis, were described and published originally in an online-only journal without the ZooBank registration number (LISD). Although registered in the ZooBank, the species are not valid because the registration of the species and the article did not appear in the final published version. Thus, lacking formal evidence that the registration had occurred, as regulated by Art. 8.5.3 of the amended Code of the International Commission on Zoological Nomenclature (ICZN, 1999, 2012). Supporting the requirements aforementioned, this work validates the names Tubastraea ramosa, Tubastraea columnata, Tubastraea grandidentata, and Tubastraea megalostoma, fulfilling the ICZN conditions for nomenclatural availability. The LSID links are given in the Taxonomy section below, including the text modified from Serra et al. (Reference Serra, Bastos, Coutinho, Schizas, Johnsson and Neves2024, Reference Serra, Bastos, Coutinho, Schizas, Johnsson and Neves2025).
Taxonomy
Tubastraea ramosa sp. nov. Serra, Neves & Johnsson
urn:lsid:zoobank.org:act:9B5CD1C1-D2B2-42FC-B77C-B44A77667F94
Tubastraea tagusensis – de Paula & Creed 2004, 2a,b, Sampaio et al. Reference Sampaio, Miranda, Maia-Nogueira and Nunes2012: Figure 3, Figure 5, Mantelatto et al. Reference Mantelatto, Pires, Oliveira and Creed2015: Figure 1G-L, Miranda et al. Reference Miranda, Costa, Lorders, Nunes and Barros2016a: Fig. 2a, Miranda et al. Reference Miranda, Cruz and Barros2016b: Fig. 3.
Tubastraea ramosa – Serra, Neves & Johnsson, 2024: 118 Fig. 2a-d; 4a-c [unavailable].
Tubastraea ramosa – Serra, Neves & Johnsson, 2025: 73 [unavailable].
Type material: Holotype (length × height): UFBA 1436, 74.2 × 56.1 mm, 37 corallites, collected at 1 m deep, in an artificial environment. Paratypes (length × height): UFBA 1424, 63.8 × 45.5 mm, 29 corallites; UFBA 1373, 76.3 × 54.6 mm, 31 corallites; UFBA 1405, 104.6 × 55.2 mm, 90 corallites; UFBA 1434, 57.0 × 37.0 mm, 39 corallites; UFBA 1435, 59.6 × 55.5 mm, 37 corallites.
Type locality: Private Terminal, Salvador Capital (12°58′S, 38°30′W), Todos-os-Santos Bay, Bahia State, Brazil.
Etymology: The specific epithet is attributed to the colonial growth pattern, with branches budding from the axial polyps. In Latin, ramosa = branched.
Diagnosis: Branching colonies with irregular monopodial growth; branches interspersed or in opposite distribution, with arborescent growth; dendroid corallites; extracalicinal budding; coenosarc and polyps yellow to intense orange, polyps with tentacles yellow to orange; corallite cylindrical and high (up to 39.1 mm height) with circular to elliptical calices; corallites closely spaced basally and spaced distally; synapticulotheca eventually with irregular calcification near theca edge; medium to large pores (130.0–450.0 µm) along costa and intercostal grooves; septa arranged hexamerally (21–51 septa), four cycles non-exsert (S1 > S2 > S3 > S4), S1–S3 complete, S4 rudimentary; septal faces ornamented with rounded and slightly pointed granules, septal edge regular and smooth in S1 and S2, and irregular, with laciniate projections, in S3; columella moderately developed (1.1–7.1 mm), elliptical and spongy, fossa columellar deep (4.9–12.9 mm).
Tubastraea columnata sp. nov. Serra, Neves & Johnsson
urn:lsid:zoobank.org:act:88603739-5BAB-482D-9D33-515AEA00FE1F
Tubastraea tagusensis – Luz et al. Reference Luz, Capel, Zilberberg, Flores, Migotto and Kitahara2018: Fig. 1c; Tubastraea sp. – Bastos et al. Reference Bastos, Calazans, Altvater, Neves, Trujillo, Sharp, Hoffman and Coutinho2022: Fig. 2i-l
Tubastraea columnata – Serra, Neves & Johnsson, 2024: 118 Fig. 2e-h; 4d-f [unavailable].
Tubastraea columnata – Serra, Neves & Johnsson, 2025: 73-74 [unavailable].
Type material: Holotype (length × height): UFBA 1423, yellow-orange colony, 49 × 33 mm, 28 corallites, in the artificial substrate at 1 m deep. Paratypes (length × height): UFBA 1352, 53 × 28 mm, 28 corallites, in association with T. coccinea; UFBA 1354, 51 × 37 mm, 25 corallites; UFBA 1427, 51 × 35 mm, 22 corallites; UFBA 1552, 39 × 32 mm, 24 corallites.
Type locality: Private Terminal, Salvador Capital (12°58'S, 38°30'W), Todos-os-Santos Bay, Bahia State, Brazil.
Etymology: The specific epithet is attributed to the colonial growth supported by elongated corallites, as the columns of classical Greek architecture. In Greek, columnata = columns.
Diagnosis: Massive colonies, bushy growth, flattened basis, dendroid corallites, extracalicinal budding, costate coenosteum; coenosarc and polyps yellow to light orange, polyps with yellow tentacles; corallite cylindrical and high (up to 25.4 mm height) with circular to elliptical calices, corallite base closely spaced and widely apart distally; synapticulotheca thick, small pores (80.0–138.0 µm) along the intercostal grooves; septa arranged hexamerally (23–48 septa), four cycles non-exsert (S1 ≥ S2 > S3 ≥ S4), S1–S3 complete, S4 incomplete, S3 thinner, deeply fused to S2, eventually with free axial margins, S4 rudimentary; septal faces ornamented with small slightly pointed granules, S1 and S2 with small dentate projections near to columella, S3 septal margin irregularly dentate; columella small (0.8–5.7 mm), spongy, fossa columellar deep (4.4–14.7 mm).
Tubastraea grandidentata sp. nov. Serra, Neves & Johnsson
urn:lsid:zoobank.org:act:D74C859E-E6AA-4558-9649-EF88F4D34F8E
Tubastraea grandidentata – Serra, Neves & Johnsson, 2024: 118 Fig. 2i-l; 4g-i [unavailable].
Tubastraea grandidentata – Serra, Neves & Johnsson, 2025: 74 [unavailable].
Type material: Holotype (length × height): UFBA 1361, colony with orange polyps, 28.5 × 23.9 mm, 7 corallites, collected at 3 m deep in an artificial environment. Paratypes (length × height): UFBA 1360, 29.2 × 19.0 mm, 9 corallites; UFBA 1379, 40.2 × 26.6 mm, 13 corallites; UFBA 1443, 67.1 × 30.7 mm, 19 corallites; UFBA 1437, 64.1 × 45.9 mm, 33 corallites; UFBA 1438, 61.3 × 28.9 mm, 28 corallites.
Type locality: Marina de Itaparica (12°53′S, 38°41′W), Itaparica Is., Todos-os-Santos Bay, Bahia State, Brazil.
Etymology: The epithet refers to the irregular edge of the corallite theca, which alternates deep grooves and prominent projections. Two Latin terms were used to create the specific epithet: grandis – large or great, and dentata – toothed or pointed.
Diagnosis: Massive, hemispherical to irregular encrusting colonies, plocoid corallites, extracalicinal budding, non-costate coenosteum; coenosarc and polyps bright yellow, pink or orangish-pink, polyps with yellow tentacles; cylindrical and short corallites (up to 15.0 mm height), circular to elliptical calices, corallites closer basally and spaced distally; synapticulotheca thick basally and fragile at calice margins, theca margins deeply dentate; medium-sized pores (80-0-300.0 µm) along costae and intercostal grooves; septa arranged hexamerally (25–44 septa), four cycles non-exsert (S1 > S2 > S3 > S4); S1–S3 complete, S4 rudimentary, S1 and S2 markedly thicker; septal faces strongly ornamented with small and rounded granules; columella small (1.2–6.1 mm), solid to spongy, fossa columellar moderately deep (3.5–8.4 mm).
Tubastraea megalostoma sp. nov. Serra, Neves & Johnsson
urn:lsid:zoobank.org:act:263389A4-2587-49DB-B5F2-8CD484E05578
Tubastraea megalostoma – Serra, Neves & Johnsson, 2024: 118, Fig. 2m-p; 4j-l [unavailable].
Tubastraea megalostoma – Serra, Neves & Johnsson, 2025: 74 [unavailable].
Type material: Holotype (length × height): UFBA 1348, 65.2 × 41.3 mm, 33 corallites, in the artificial substrate at 1 m deep. Paratypes (length × height): UFBA 1358, 84.5 × 59.2 mm, 29 corallites; UFBA 1364, 84.8 × 47.3 mm, 27 corallites; UFBA 1439, 67.3 × 53.6 mm, 21 corallites; UFBA 1440, 108.1 × 55.3 mm, 73 corallites; UFBA 1441, 98.5 × 36.7 mm, 27 corallites.
Type locality: Private Terminal, Salvador (12°58′S, 38°30′W), Todos-os-Santos Bay, Bahia State, Brazil.
Etymology: The specific epithet is attributed to the oversized calice and columella. Two Greek terms were used to create the species name: megalo – big size, stoma – mouth.
Diagnosis: Massive to encrusting colonies, plocoid corallites; extracalicinal budding; larger corallites may undergo fission, non-costate coenosteum; coenosarc and polyps dark pink to reddish, polyps with yellow-orange tentacles; corallites large, slightly conical to cylindrical (up to 15.3 mm height) with circular to elliptical calices; corallites close to moderately spaced at the base and spaced distally; synapticulotheca moderately thick and strongly granular, medium size pores (138.0–275.0 µm) along intercostal grooves; septa arranged hexamerally (32–65 septa), five cycles non-exsert (S1 = S2 ≥ S3 > S4 > S5); S1–S3 complete, S4 rarely incomplete, S5 incomplete, S1 and S2 with thicker axial margins, S3 occasionally free; septal face heavily ornamented with small, closely spaced rounded granules, S1–S3 with regular margins and poorly ornamented, S4 with smooth axial margin, with depressions towards the theca; columella well-developed (1.2–10.1 mm), spongy, fossa columellar moderately deep (4.4−11.1 mm).
Remarks
The species of Tubastraea can be branching as T. micranthus, T. dendroida, and T. ramosa sp. nov. or, alternatively, they can be non-branching, with plocoid and dendroid corallite development (Yiu et al., Reference Yiu, Chung and Qiu2021; Yiu and Qiu, Reference Yiu and Qiu2022). Despite of sharing a similar uniplanar branching pattern, T. micranthus has three septa cycles, while T. dendroida has four septa cycles arranged in Portalès Plan. Tubastraea ramosa sp. nov. is distinguishable from both congeners by the arborescent aspect, due to the irregular monopodial growth with branches interspersed or in opposite distribution.
Among non-branching species with plocoid development, small widely spaced corallites with four septa cycles are observed in T. faulkneri. On the other hand, T. megalostoma sp. nov. has larger corallites with five septa cycles and a well-developed columella. Closely spaced corallites occur in three congeners: T. grandidentata sp. nov. with thecal margins heavily dentate, and T. diaphana and T. coccinea with smooth thecal margins. Indeed, this last pair comprises very similar species, usually differentiated by the calice shape and the thickness of the theca (Arrigoni et al., Reference Arrigoni, Kitano, Stolarski, Hoeksema, Fukami, Stefani, Galli, Montano, Castoldi and Benzoni2014; Serra et al., Reference Serra, Bastos, Coutinho, Schizas, Johnsson and Neves2024, Reference Serra, Bastos, Coutinho, Schizas, Johnsson and Neves2025).
The subdendroid T. floreana shows three septa cycles, while T. aurea and T. tagusensis have four septa cycles. Moreover, T. aurea has conical corallites, while in T. tagusensis they are cylindrical. All other Tubastraea species are dendroid, but T. stimpsonii is the only one with a ceratoid shaped corallum. Tubastraea violacea and T. megacorallita share the five septa cycles arranged in Pourtalès plan, differing in the corallite shape – cylindrical in the former, and elliptical in the latter. Finally, T. chloromura and T. columnata sp. nov. are promptly differentiated by the corallite shape. The former has shorter corallites with S3 = S4, while in T. columnata sp. nov. the corallites are slender and taller with S4 rudimentary (S3 > S4) (Wells, Reference Wells1982; Yiu et al., Reference Yiu, Chung and Qiu2021; Yiu and Qiu, Reference Yiu and Qiu2022; Mehrotra et al., Reference Mehrotra, Chavanich, Monchanin, Jualaong and Hoeksema2023; Serra et al., Reference Serra, Bastos, Coutinho, Schizas, Johnsson and Neves2024, Reference Serra, Bastos, Coutinho, Schizas, Johnsson and Neves2025).
Phylogeny
Tubastraea species are endorsed by the phylogenetic analyses of the three genetic regions, indicating an important agreement between morphology and genetic data. The topology recovered by BI (Figure 2, left) and ML (Figure 2, right) consistently resolves Tubastraea as a cohesive clade, delimited by two out-groups: A. caboverdiana (green clade) and D. cornigera (purple clade).
Phylogenetic relationships between sun coral species by Bayesian inference (left) and maximum likelihood (right), using concatenated sequences of COI, IGR, and rDNA. Red clade = non-branched species, yellow clade = branched species, green and purple clades = out-groups (Atlantia and Dendrophyllia, respectively).
Tubastraea tree splits into two major clades with the lineages distributed according to two morphological patterns: (1) non-branched colonies, i.e., T. coccinea, T. aurea, T. grandidentata, T. megalostoma, and T. columnata (red clade), and (2) branched colonies represented by T. ramosa, T. chloromura*, T. diaphana*, T. micranthus, T. dendroida, T. stimpsonii, T. megacorallita*, and T. violacea* (yellow clade). Within this clade, species non-branched with subdendroid/dendroid corallites (*) are eventually designated as ‘phaceloid’ by Yiu et al. (Reference Yiu, Chung and Qiu2021) and Yiu and Qiu (Reference Yiu and Qiu2022).
ML and BI analyses result in similar tree topologies, with the latter furnishing higher branch supporting values (Figure 2, left). Concatenation of the three molecular markers increases the signal-over-noise ratio, pointing out the distinction among non-branched (red) and branched (yellow) clades, instead of plocoid and dendroid groups, as previously suggested in the literature (vide Bastos et al., Reference Bastos, Calazans, Altvater, Neves, Trujillo, Sharp, Hoffman and Coutinho2022).
The non-branched clade (red) is well supported (pp ≥ 0.94; bootstrap = 82) and comprises almost all Atlantic (Brazilian) species, which are consistently recovered as monophyletic. Tubastraea coccinea (pp = 1.00; bootstrap = 67) and T. aurea (pp = 0.72; bootstrap = 80) form a strongly supported clade (pp = 1.00; bootstrap = 90). Similarly, T. grandidentata (pp = 1.00; bootstrap = 95) and T. megalostoma (pp = 1.0; bootstrap = 91) are monophyletic and well supported (pp = 1.00; bootstrap = 99). Tubastraea columnata (pp = 0.94; bootstrap = 82) emerges as the sister group of all lineages of the clade.
The branched clade (yellow) is steadily supported by BI analysis (pp = 0.97), with species from the Atlantic Ocean (Brazil), Western Indian Ocean (Mayotte Is.), and Pacific Ocean (Thailand, Hong Kong). In this way, the major clade has five lineages. Tubastraea ramosa is monophyletic (pp = 0.94; bootstrap = 62), with the ‘phaceloid’ T. chloromura emerging as the sister group (pp = 0.71). Tubastraea diaphana is closely related to the clade T. ramosa + T. chloromura, with T. micranthus + T. dendroida (pp = 0.67) standing out as the sister group of all former lineages. The second clade comprises three lineages from the Pacific Ocean: T. stimpsonii, T. megacorallita, and T. violacea. It is supported by BI analysis (pp = 0.85), with T. violacea as the sister group of the subclade T. stimpsonii + T. megacorallita (pp = 1.00; bootstrap = 95).
Overall, BI and ML analyses are congruent regarding the phylogenetic relationships and expected monophyly for all taxa studied. Differences are restricted to support values for certain internal nodes within the clade of branched lineages, particularly for T. ramosa (pp = 0.94; bootstraps = 62). In contrast, most nodes within non-branched clade exhibit high support for both methods (pp = 1.00; bootstraps = 90–99). Finally, A. caboverdiana is recovered as monophyletic with maximal support values (pp = 1.00; bootstrap = 96).
Discussion
Tubastraea is a morphologically complex scleractinian genus. The four species of Tubastraea recently described in Brazil are genetically divergent from all previously recognized congeners, supporting their validity as independent evolutionary lineages within non-branched and branched clades.
Indeed, the diversity of Tubastraea has been trapped due to trends in the field identification protocol, widely based on the corallite development, i.e., plocoid versus dendroid, instead of a meticulous analysis of the well-defined diagnostic elements of Scleractinia, including macro- and micromorphology (Serra et al., Reference Serra, Bastos, Coutinho, Schizas, Johnsson and Neves2024, Reference Serra, Bastos, Coutinho, Schizas, Johnsson and Neves2025). Furthermore, this is the first comprehensive phylogenetic study for the Tubastraea from the Southwestern Atlantic that provides data on the congeners from the China Sea, including as out groups A. caboverdiana (Cape Verde, Eastern Atlantic) and D. cornigera (Mediterranean Sea), with a clear distinction among the three genera.
As observed for the endemic reef-building genus Mussismilia Ortmann, 1890, a small group of brain corals with variable morphologies and the morphotypes occurring sympatrically, studies have highlighted the relevance of the corallite structures, particularly the micromorphological characters (e.g., patterns of septal and costae ornamentation) (Silva et al., Reference Silva, Nogueira, Johnsson and Neves2024).
Paradoxically, data on the intraspecific variation in Tubastraea have not been published yet for any of the 16 valid species, even though it is an important aspect of the genus diversity worldwide. Our preliminary analyses indicated impressive levels of variation in Tubastraea coccinea, a ‘complex of species’. From this first scenario, two new species were described, T. megalostoma and T. grandidentata (Serra et al., Reference Serra, Bastos, Coutinho, Schizas, Johnsson and Neves2024, Reference Serra, Bastos, Coutinho, Schizas, Johnsson and Neves2025), both ratified in this study and positioned in the clade of non-branched sun corals.
Indeed, ML and BI phylogenetic trees exhibit strong phylogenetic signals and similar topologies. Despite the lower pp on the BI tree (0.51), Tubastraea clade (and its two subclades) is strongly supported by bootstrap (98), while the non-branched clade (Figure 2, red) is supported by BI pp (0.94) and a high bootstrap (82). The bootstrap values in this clade suggest that certain taxa are more readily distinguishable at the genetic level than others, although none of the bootstrap values were less than 50%. Similarly, the branched clade (Figure 2, yellow) exhibits a lower bootstrap (51), which could limit the reliability of inferences within the group if this value was inferior to 50%. However, it does not go unnoticed that the BI pp is high (0.97) with a well-defined clade of the new branched species, T. ramosa (pp = 0.94). In summary, ML and BI results are consistent and provide a robust tree for the Tubastraea phylogeny.
Tubastraea coccinea is the most dispersive exotic species in the Atlantic, with early reports from the 1940s (Vaughan and Wells, Reference Vaughan and Wells1943), and it has probably been misinterpreted in ecological studies, raising concerns about the management policies of exotic organisms in Brazil. Tubastraea coccinea and T aurea are sister groups, closely related species. Our data agree with Boschma (Reference Boschma1953), showing these are genetically and morphologically distinguishable species. The interspecific limits of both species were revised by Serra et al. (Reference Serra, Bastos, Coutinho, Schizas, Johnsson and Neves2024, Reference Serra, Bastos, Coutinho, Schizas, Johnsson and Neves2025). But, the intraspecific variation within and among populations (considering, for instance, populations from Southeastern and Northeastern Brazil) must also be assessed. Moreover, the investigation of larger DNA datasets (e.g., 1000s of genomic SNPs) would be profitable for additional confident data, considering a global phylogeny of Tubastraea.
Comparing with the Tubastraea phylogeny provided by Bastos et al. (Reference Bastos, Calazans, Altvater, Neves, Trujillo, Sharp, Hoffman and Coutinho2022), we find some differences that cannot be ignored. First, a set of universal primers was adopted by the authors for the amplification of the rDNA sequences – here, specific primers are designed for Tubastraea corals, increasing the efficiency of DNA amplification. Furthermore, in the present study, three genes (COI, IGR, and rDNA) are simultaneously applied in the BI and ML analyses, resulting in a stable topology.
The Morphotype III, referred to in Bastos et al. (Reference Bastos, Calazans, Altvater, Neves, Trujillo, Sharp, Hoffman and Coutinho2022), is the dendroid T. columnata described in Serra et al. (Reference Serra, Bastos, Coutinho, Schizas, Johnsson and Neves2024, Reference Serra, Bastos, Coutinho, Schizas, Johnsson and Neves2025), which is a basal lineage closely related to the non-branched clade (Figure 2). As noted among the species in the branched clade, the elongate corallites may be a plesiomorphic condition in the genus.
We raise some questions on the phacelloid development cited in Yiu et al. (Reference Yiu, Chung and Qiu2021) and Yiu and Qiu (Reference Yiu and Qiu2022) for Tubastraea species, particularly T. megacorallita, T. chloromura, and T. violacea, since living tissue covers the entire polyp column (and the grooves between the polyps), discerning from the true phacelloid pattern observed in Mussismilia harttii (Verrill, 1868) (Silva et al., Reference Silva, Nogueira, Johnsson and Neves2024). Second, analysing the pictures of the species from the China Sea (Yiu et al., Reference Yiu, Chung and Qiu2021: Figure 1B, 4; Yiu and Qiu, Reference Yiu and Qiu2022: Figure 3D, 6, Figure 4C, 7), it is clearly observed that the extracalicinal budding pattern is related to the monopodial development of the Brazilian T. ramosa (new individuals bud from 1/3 and/or ½ the length of the column of the parental polyp). These aspects (and our phylogeny) indicate the need for a morphological review of the species from the China Sea.
Tubastraea species may have plocoid (T. coccinea, T. megalostoma), dendroid (T. columnata, T. ramosa), and subdendroid (T. aurea) corallites, with branched (T. ramosa, T. dendroida) and non-branched colonial development (T. columnata). In the tree, branched dendrophylliids, such as T. ramosa and A. caboverdiana, are not genetically related, suggesting that the pattern appears convergently within the family. Another noticeable difference between Yiu and Qiu (Reference Yiu and Qiu2022) and our phylogeny is that, in the former, T. aurea (subdendroid) is positioned as the sister group of T. diaphana (branched), while the relationship between T. violacea and Tubastraea sp.2 (from Arrigoni et al., Reference Arrigoni, Kitano, Stolarski, Hoeksema, Fukami, Stefani, Galli, Montano, Castoldi and Benzoni2014, see Figure 1D, 3) clearly indicates two branching species as sister groups.
Ecologically, branching development enables corals to inhabit environments with higher hydrodynamic energy by effectively blocking the flow, and with branch thickness varying with the flow regime (Goldshmid et al., Reference Goldshmid, Holzman, Weihs and Genin2004; Monismith, Reference Monismith2007). The spaces between coral branches also provide specialized microenvironments for other organisms, such as mutualistic fishes (Goldshmid et al., Reference Goldshmid, Holzman, Weihs and Genin2004). Most branching corals are fast-growing and undergo skeleton fragmentation as a mechanism of asexual propagation, enhancing long-life survivorship (Lirman, Reference Lirman2000). In fact, branching colonies would be advantageous for a dispersive and generalist genus (resilient to environmental changes and capable of inhabiting various habitats), with a wide bathymetric distribution and the competency to regenerate small fragments (Cairns, Reference Cairns2000; Sampaio et al., Reference Sampaio, Miranda, Maia-Nogueira and Nunes2012; Batista et al., Reference Batista, Gonçalves, Messano, Altvater, Candella, Elias, Messano, Apolinário and Coutinho2017; Luz et al., Reference Luz, Capel, Zilberberg, Flores, Migotto and Kitahara2018). This would explain the early diversification of basal-branched species within Tubastraea as suggested by our tree. Besides that, we assume that the loss of branching may represent an adaptive response to shallower areas with higher sedimentation. Non-branching encrusting colonies appear better adapted to biofouling conditions, particularly in negative positions, where reduced exposure to sedimentation and predation confers an additional advantage. Thus, we assume that subdendroid/dendroid corallites may be the plesiomorphic condition within any clade (branched and non-branched).
Here, we highlight the hidden diversity of dendrophylliid corals in geographic areas under anthropogenic pressure, such as the China Sea (Yiu et al., Reference Yiu, Chung and Qiu2021; Yiu and Qiu, Reference Yiu and Qiu2022) and the TSB (Northeast Brazilian Coast) (Serra et al., Reference Serra, Bastos, Coutinho, Schizas, Johnsson and Neves2024, Reference Serra, Bastos, Coutinho, Schizas, Johnsson and Neves2025).
Tubastraea coccinea, T. diaphana, T. faulkneri, and T. micranthus have been reported from the China Sea (Scott, Reference Scott1984; Clark, Reference Clark and Morton1998; Lam et al., Reference Lam, Morton and Hodgson2008). The region is an important oil and natural gas reserve, supporting one of the most prominent worldwide shipping trade routes. However, it is worthwhile to point out that T. faulkneri is one of Well’s (Reference Wells1982) species, with the holotype from Palau (Micronesia) and occurrence for the Tagus Cove (Galapagos Islands). Although ignored, T. faulkneri could be considered an exotic coral in the China Sea.
The TSB harbours an oil refinery, the largest petrochemical complex in the Southern Hemisphere, an industrial centre based on the metallurgical and finished products industry, with the surrounding area heavily used by agriculture and aquaculture (Tavares et al., Reference Tavares, Silva, Oliva, Sant’Anna, Lockwood, Gutierrez-Ruiz, Queiróz-Vivanco and Aguirre-Gomes2018). Threatened by multiple anthropogenic activities, the TSB has been increasingly subjected to the impact of the exotic species. Introductions may occur naturally or be influenced by human activities. The dispersion of marine organisms over natural barriers has been described in the literature and associated with some vectors – e.g., biofouling on ship hulls and oil platforms, larvae in the ballast water, and the release of aquarium organisms into nature. The dispersion is usually enhanced by the marine currents and larvae competency. Larvae or juvenile ‘rafting’, by the transport on floating garbage, has also been documented in recent years, indicating that non-biogenic substrates (e.g., styrofoam and plastic waste) can act as vectors for introducing exotic organisms in new geographical areas (DeVantier, Reference DeVantier1992; Roos and Cadée, Reference Roos and Cadée2012; Hoeksema et al., Reference Hoeksema, Pedoja and Poprawski2018). Certainly, one or more of these mechanisms are involved in the establishment of sun corals in the TSB. The literature has argued about multiple introductions and secondary dispersion of Tubastraea in the Southwestern Atlantic (Capel et al., Reference Capel, Creed, Kitahara, Chen and Zilberberg2019), which probably have facilitated the overspread of two non-native sessile cnidarians in the TSB: Briareum hamrum and the blue xeniid Sarcothelia sp. (Menezes et al., Reference Menezes, McFadden, Miranda, Nunes, Lolis, Barros, Sampaio and Pinto2022).
Other aspects regarding the new Tubastraea corals in Brazil were published by Silva et al. (Reference Silva, Alves, Dourado, Zanetti, Divino, Divino, Nepomuceno, Alcantara, Oliveira, Serra, Farias, Silva, Johnsson and Neves2025). These corals have very particular reproductive structures never observed in Scleractinia before, i.e., mucin layer surrounding the oocytes, nucleolinus inside nucleolus, and trophonema, pointing out the attention on the morphoanatomy of the polyps and the relevance of integrative studies (taxonomy, molecular phylogeny, and biology) for a better understanding of exotic marine invertebrates.
Molecular studies have been developed using similar methodologies, especially considering the same genetic markers (COI, IGR, and rDNA) and concatenation of sequences for the phylogenetic analysis (Arrigoni et al., Reference Arrigoni, Kitano, Stolarski, Hoeksema, Fukami, Stefani, Galli, Montano, Castoldi and Benzoni2014; Yiu et al., Reference Yiu, Chung and Qiu2021; Yiu and Qiu, Reference Yiu and Qiu2022; Mehotra et al., Reference Mehrotra, Chavanich, Monchanin, Jualaong and Hoeksema2023). The ambiguity of the nucleotide positions from universal primers may result in erroneous pairings, leading to non-specific amplifications (Piñol et al., Reference Piñol, San Andrés, Clare, Symondson and O’Neill2015; Cheng et al., Reference Cheng, Xu, Lei, Li, Zhang and Zhou2016). The selected genes appear more suitable for deep evolutionary analyses within this group of species, suggesting that the inclusion of additional genes may enhance the resolution of poorly supported clades and branches.
Therefore, in this work, specific primers for the Tubastraea genus were developed and standardized, reducing noise during amplification and allowing more efficient amplification reactions and reliable sequences. The consistent recovery of T. ramosa, T. columnata, T. grandidentata, and T. megalostoma as well-supported lineages not only reinforces their taxonomic status as new species but also highlights their role in refining the morphological and molecular framework for Tubastraea systematics.
The unnatural expansion of the sun corals along the Brazilian coast has raised attention about the governance of exotic organisms into coastal environments, mobilizing efforts to prevent new introductions and dispersion across invaded areas. Meanwhile, it is unclear whether the Tubastraea diversity observed in the TSB is locally restricted or extends towards other coastal areas, where the exotic dendrophylliids have been monitored in the last two decades.
Finally, after Serra et al. (Reference Serra, Bastos, Coutinho, Schizas, Johnsson and Neves2024, Reference Serra, Bastos, Coutinho, Schizas, Johnsson and Neves2025) and this study, the GenBank needs a thorough revision. It is likely to be overwhelmed by Tubastraea sequences with erroneous species identification.
Conclusion
This study contributes to the comprehension of the biodiversity of the Tubastraea species from the Southwestern Atlantic. It is based on a multigene approach, supporting and validating the new species described by Serra et al. (Reference Serra, Bastos, Coutinho, Schizas, Johnsson and Neves2024, Reference Serra, Bastos, Coutinho, Schizas, Johnsson and Neves2025), and providing a new phylogeny for selected sun corals. The tree topology suggests that colony development prevails over corallite growth. Thus, plocoid and dendroid non-branched species (T. coccinea, T. aurea, T. grandidentata, T. megalostoma, and T. columnata) are within the same clade. At the same time, a basal sister group is represented by a branched species (T. ramosa, T. chloromura, T. diaphana, T. micranthus, T. dendroida, T. stimpsonii, T. megacorallita, and T. violacea). Finally, Atlantia caboverdeana is genetically distinguishable from the branched T. ramosa.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S0025315426101210.
Acknowledgements
We are grateful to Dra. Vanessa Rodrigues Paixão Côrtes (IBIO/UFBA) and Dr. Paulo Corgosinho (UNIMONTES) for valuable molecular discussions. We are thankful to FIOCRUZ/BA for imaging support and the LABIMAR team (IBIO/UFBA) for outstanding help in the field. We are deeply indebted to the anonymous reviewers for the careful reading and insightful comments and corrections. Finally, we thank the diving school ‘Galeão Sacramento’ (Mario Bruno Rocha de Souza) for the guidance across the sea. R. Johnsson is grateful to the CNPq (Conselho Nacional de Desenvolvimento Científico e Tecnológico, Brazil) for the research fellowship provided (# PQ 304592/2020-2).
Author contributions
All authors contributed equally to the core scientific work: conception, design, and data analysis. E.G.N. and R.J. are the team leaders, handled funding acquisition and manuscript revision.
Funding
This study is part of the project ‘Assessment and research of the sun coral in the Todos-os-Santos Bay’, a cooperation agreement between UFBA and PETROBRAS (No. 5850.0107361.18.9) regulated by the R, D, and I investment clauses of the Brazilian Agency of Petroleum, Natural Gas and Biofuels (ANP Resolution 03/2015).
Competing interests
The authors declare no competing interests.
Ethical statement
Biological samples were collected following all applicable ethical regulations regarding the collection of biological samples and experimentation with animals. The investigation was performed under permit SISBIO No. 15161-1, issued by the Chico Mendes Institute for Biodiversity Conservation (ICMBio).
Data availability statement
The authors declare that the data supporting the findings of this study are available within the paper. Should any raw data files be needed in another format, they are available from the corresponding author upon reasonable request. Source data are provided with this paper.
