A survey of the genera Sarcodictyon Forbes, 1847, and rolandia Lacaze-Duthiers, 1900, (anthozoa: octocorallia) in the North-eastern Atlantic and the Mediterranean

A survey of the genera Sarcodictyon Forbes, 1847, and Rolandia Lacaze-Duthiers, 1900, (Anthozoa: Octocorallia) in the North-eastern Atlantic and the Mediterranean O. Ocana, P .J. Lopez-Gonzalez, J. Nunez & J.C. Garcia-Gomez Ocana, 0., P.J. LOpez-Gonzalez, J. NUfiez & J.C. Garcia-Gomez. A survey of the genera Sarcodictyon Forbes, 1847, and Rolandia Lacaze-Duthiers, 1900, (Anthozoa: Octocorallia) in the North-eastern Atlantic and the Mediterranean. Zool. Med. Leiden 73 (28), 6.iii.2000: 413-426, figs 1-5.- ISSN 0024-0672. O.Ocana, Instituto de Estudios Ceutfes (CECEL-CSIC), Paseo del Revellin, 30, Apdo. 51001 Ceuta (North-Africa), Spain. J. NUflez, Departamento de Biologia Animal, Facultad de Biologia, Universidad de La Laguna, Tenerife (Canary Islands), Spain. J.C. Garcia-Gomez & P.J. LOpez-Gonzalez, Laboratorio de Biologia Marina, Departamento de Fisialogia y Biologia Animal, Facultad de Biologia, Universidad de Sevilla, Apdo. 1095, 41080 Sevilla, Spain. Key words: Stolonifera; Sarcodiclyon; Rolandia; North-eastern Atlantic; Mediterranean. Sarcodiclyon catenatum Forbes, 1847, is a common species in the Atlantic and the Mediterranean. However, often it has been mistaken for Rolandia coralloides Lacaze-Duthiers, 1900. On the basis of British and Mediterranean material we studied the main characters for distinguishing S. catenatum. We consider Rolandia coralloides Lacaze-Duthiers, 1900, to be a valid species and we assign to it all the material named Rolandia rosea by S. Weinberg (1978). Furthermore, a histological study confirmed that R. cora/loides typically forms groups of polyps embedded in a common coenenchyme. Therefore, the possibility of transferring R. cora/loides from Stolonifera to Alcyonacea is discussed. Evagorgia rosea Philippi, 1842, is here considered to be a nomen dubium. Introduction The genus Sarcodictyon Forbes (in Johnston), 1847, is based on British material of Sarcodictyon catenatum, described in the same paper. Forbes & Goodsir (1851) described a second species, Sarcodictyon agglomerata, very similar to the type species. They distinguished the species by external characters only"... a ochraceous yellow colour and in the grouping of polyps in threes and fives ...". Subsequently, Sars (1856) described Rhizoxenia filiformis, which probably represents another species of Sarcodictyon (see discussion of S. catenatum). Gosse (1858) collected and studied several colonies from the south coast of England and identified them as S. catenatum. These British papers culminated in the synthetic work of Herdman (1885), outstanding for its methodology used for the study of this genus. Ten years later, Herdman (1895) synonymized s. agglomerata with S. catenatum. Meanwhile Philippi (1842) described Evagora rosea from the Mediterranean. This species probably belongs to Sarcodictyon or Rolandia, but this remains uncertain, due to the inadequate description given by this author; it is even possible that it belongs to Scleranthelia Studer, 1878 (see Carpine, 1964). After Philippi, species with some resemblance to E. rosea were generally referred to this species. However, the situation became more complicated when Lacaze-Duthiers (1900) described another similar species from the Mediterranean, Rolandia coralloides. Thomson (1927) recorded s. catenatum from Monaco, without describing the sclerites. Later on, Berenguier (1954) 414 Ocana et a1. A survey of the genera Sarcodictyon and Rolandia. Zo01. Med. Leiden 73 (2000) cited S. catenatum for the Mediterranean and claimed that S. catenatum and R. coralloides were the same species. Pax & MUller (1962) distinguished two Adriatic species Zoantha rosea, Philippi, 1842, (possibly S. catenatum) and Zoantha corii Broch, 1935, (= R. coralloides), using sclerite characters. Weinberg (1978a) considered E. rosea and R. coralloides to be the same species, Rolandia rosea. In that paper Weinberg assumed that the sclerites in the genus Sarcodictyon are inseparably fused. This error was created by Hickson's (1921: 367) misinterpretation of Herdman's description, repeated by Oeichmann (1936), Gohar (1940) and Bayer (1956). Bayer (1981) put an end to this confusion and illustrated some sclerites of a specimen of S. catenatum from British waters (Isle of Man). In the same paper he referred Weinberg's Rolandia rosea to Sarcodictyon. Finally, Manuel (1981) synonymized S. catenatum and R. rosea as Sarcodictyon roseum. We studied material from the Atlantic as well as the Mediterranean and recognize two species: Sarcodictyon catenatum Forbes, 1847, and Rolandia coralloides LacazeDuthiers,1900. They are described and depicted below. Based on the sclerites and the anatomy we suggest to maintain R. coralloides in a genus different from Sarcodictyon. For a comparison of taxonomic characters of Sarcodictyon and Rolandia we refer to table 1. Material and Methods The colonies studied in the present work were collected by SCUBA diving during different sampling trips organized by the National Museum of Natural Sciences of Madrid along the southern Iberian Peninsula: "Fauna-I" in the Alboran Sea, "FaunaII" in the Cantabric Sea, "Fauna-ill" in the Balearic waters, and "Fauna-IV" in Balearic and Columbretes waters. Material of Rolandia coralloides from the Zoological Museum of Amsterdam (ZMA), previously studied by Weinberg, was also at our disposal. Colonies were anaesthetized with menthol crystals and preserved in 70% alcohol; specimens used for histological study were fixed in 8% formaldehyde. The general morphology and anatomy were studied by means of a stereo dissecting microscope. Anatomical and histological details were studied following the Cajal method for topographic staining (see Gabe, 1968). Sclerites were obtained by dissolving the soft tissues in concentrated sodium hypochlorite. They were examined with a light microscope and SEM. For terminology we used Bayer et al. (1983). The colonies are deposited in the Nationaal Natuurhistorisch Museum, Leiden, The Netherlands (RMNH), Museo Nacional de Ciencias Naturales de Madrid (MNCN), and Laboratorio de Biologia Marina de la Universidad de Sevilla, Spain (LBM). Key of North-eastern Atlantic and Mediterranean species of Sarcodictyon and Rolandia 1. Polyps embedded in a common coenenchyme; sometimes stolons are present. Sclerites mainly "fused quadruplets" but crosses and shuttles are also common Rolandia coralloides - Polyps growing from narrow ribbon-like stolons, which are sometimes fused to form membranous expansions. Sclerites without "fused quadruplets" 2 2. Sclerites include stellate plates Sarcodictyon catenatum Ocana et a1. A survey of the genera Sarcodictyon and Ro/andia. Z001. Med. Leiden 73 (2000) 415 No stellate plates present but six-radiates, branched spindles, crosses & rods Sarcodictyon canariensis Family Clavulariidae Hickson, 1894 Genus Sarcodictyon Forbes (in Johnston), 1847 Diagnosis.- Polyps growing from narrow ribbon-like stolons, which are sometimes fused to form membranous expansions. Mesogloea of polyp wall with many lacunae or reticular. Retractor muscle restricted or diffuse. Stolons with 2 to 5 canals, sometimes more. Sclerite types: six-radiates, stellate plates, crosses, branched spindles and shuttles. Anthocodiae with minute sclerites or none at all. Remarks.- Bayer (1981: 899) established the subfamily Sarcodictyinae, including diverse taxa such as Cyathopodium Verrill, 1868, and Tesseranthelia, Bayer, 1981, with fused plates and operculum, and Sarcodictyon and Scleranthelia, with a completely different morphology. The characters used to distinguish the subfamily are based on external morphology "Polyps short, retracting almost flush into stolons, or producing low, conical or cylindrical calices". Some species with large polyps, viz., Sarcodictyon canariensis (Ocana et al., 1992) and Scleranthelia thomsoni (Williams,i987) cannot be fitted in Sarcodictyinae. Therefore, we decided not to consider this subfamily in our present paper. The genus Scleranthelia Studer, 1878, is very similar to Sarcodictyon. It only differs by the presence of stellate plates, disposed as a tiled floor on stolon and calyx. Sarcodictyon catenatum Forbes, 1847 (figs 1-2) Sarcodictyon catenata Forbes (in Johnston), 1847: 179-180, pl. 33, figs 4-7, description external morphology and habitus, deep water, England (Loch Fine, Youghal, west coast of Scotland); Gosse, 1858: 276-280, pl. 9, description external morphology, some data of sclerites and many aquarian observations, England (Ilsam), at intertidal caverns; Herdman, 1885: 31-49, pis 1-3, description external morphology, anatomy, histology and sclerites, England (Tarbet, Loch Fyne, Sound of Sky, Aros bay, Sound of Mull, Crouling Island, Hebrides), circalitoral and occasionally in shallow waters; Herdman, 1895: 163-168, pl. 8, descripton of the yellow variety, merging Sarcodictyon agglomerata with Sarcodictyon catenatum. Sarcodictyon catena turn; Thomson, 1927: 8-9, pl. 2 fig. 17, pro parte, Dogger Bank North Atlantic; Madsen, 1944: 11, only recorded, 55 m deep; Bergen, Norway; Tixier-Durivauit & d'Hondt, 1974: 1366, record and distribution, Azores (Flores). Clavularia catenata; Hickson, 1894: 332, put in a different genus as a result of a misinterpretation of Herdman's paper. Syrnpodiurn catenaturn; Molander, 1915: 43, diagnosis and remarks; Norway, England, Scotland and Ireland. Sarcodictyon roseurn; Manuel, 1981: 36-38, fig. 8, description external morphology, sclerites, circalittoral, British coasts, South-west Europe and the Mediterranean, perhaps northward to Scandinavia? (misidentification). ?Rhizoxeniafiliformis Sars, 1856: 65-67, Pl. X, figs. 13-17, description external morphology, some habitat data, circalittoral, colour of the colony greyish-white, possibly a different species, could be a different genus. ? Evagora rosea Philippi, 1842: 36, fig. 2, brief description, Naples. Not Rolandia rosea; Weinberg, 1978: 167-173 (= Rolandia coralloides). Ocana et al. A survey of the genera Sarcodictyon and Rolandia. Zool. Med. Leiden 73 (2000) 416 Material.- Spain: (MNCN 2.04/722), Alboran Island, 15.vii.89, 35°55.95'N 3°01.56'W, two colonies, 34-44 m, circalittoral, Fauna-I cruise (sta. 33); (LBM/PI 99) Crinavis, 36°10.10'N 5°23.70' W, vii.1990, three colonies, infralittorai, 20-25 m, P.J. Lopez leg.; (RMNH Coel. 23984), Cadiz, La Atunara, 36°10.54'N 5°20.00'W, iii.1988, one colony, circalittoral 40 m, P.J. Lopez leg.; (MNCN 2.04/723) Cantabric Sea, Northern Iberian Peninsula, 43°28.62'N 2°53.77'W, 21.vi.91, one colony, 88 m, circaIittorai, Fauna-II cruise (sm. 145-A); (MNCN 2.04/724) Hormigas Islands, 37°39.52'N 0038.92'W, 21.vii.96, one colony, 25 m, infralittorai, Fauna-IV cruise (sta. 299B5); Great Britain: (MNCN 2.04/712), Portland Harbour, Dorset, ii.1977, one colony on limestone, R.L. Manuel leg.; (RMNH Coel. 24087), Millport, Island of Cumbrae, Scotland, xi.1979, one colony on shell, R.L. Manuel leg.; (RMNH Coel. 24088) Isle of Man (1 mile north of the Calf of Man), ii.1980, one colony, 30 m, attached on shell, R.L. Manuel leg. Diagnosis (for a detailed description, see Herdman, 1885; 1895).- Colonies with 2 to more than 20 polyps, joined by ribbon-like stolons, sometimes fused to form membranous expansions. Calyx length 3 to 6 mm. Anthocodiae white, anthosteles and stolons reddish to pale-yellow. Generally the anthocodiae have small sclerites: crosses, shuttles and six-radiates, 24-52 pm long and 15-32 pm wide. Stolons with six-radiates, shuttles and crosses, 70-116 pm long and 30-110 pm wide. The stellate plates are most abundant in the anthosteles; they are 56-114 pm long and 32-120 pm wide. A distinct periderm is present on the anthosteles and stolons. Mesogloea of the polyp walls with many lacunae, totally or partially covered with granular endodermal material. Endoderm of the polyps poorly developed. Ectoderm of the tentacles well developed. The sclerites are located between the ectoderm and the mesogloea. Mesenteries with a strong, restricted retractor muscle in the middle. Pharynx with a reduced mesogloea without lacunae; endoderm well developed, ciliated and glandular. Stolons with 2-4 internal canals. Ecology.- The colonies studied were collected from intertidal to circalittoral bottoms between 0 and 100 metres, on stones and shells, occasionally close to Alcyonium palmatum Pallas, 1766, or associated with ascidians. According to Manuel (1981) competition for space may determine the growth form of this species. Distribution.- Atlantic and North Sea: England, Scotland, Wales, Dogger bank; Mediterranean: Algeciras Bay, Alboran Island and Hormigas Islands (Alicante). Discussion.- Sarcodicfyon catenatum is a well known species since Herdman's description (1885, 1895). The main characters to distinguish this species are the presence of minute sclerites in the anthocodiae (especially in the aboral face of the tentacles) and the restricted retractor of the mesenteries (Herdman, 1885: pl. 2 fig 10; pl. 3 figs I, 3). However, in the tentacles of some colonies from the western Mediterranean sclerites are scarce or absent altogether. Moreover, although these colonies also have mesenteries with strong retractors these are less restricted. As these differences are gradual, we consider them infraspecific variation. However, future studies might reveal the Mediterranean population to represent a different subspecies. Sarcodictyon catenatum differs from S. canariensis in having stellate plates in the coenenchyme and small sclerites in the polyps. Also the mesogloea of the polyp wall and the retractor muscles are clearly different (see Ocana et al., 1992). Rhizoxenia filiformis (Sars, 1856) from the North Sea has some resemblance with Sarcodictyon. Sars distinguished this species from S. catenatum by differences in colour and shape: "Notre espece Norvegienne se distingue de l'espece Britannique par la Ocana et al. A survey of the genera Sarcodictyon and Rolandia. Zool. Med. Leiden 73 (2000) 417 Fig. 1. Sclerites of Sarcodictyon catenatum.I-12 anthostele; 13-25 stolon; 26-41 tentacles. Scale bar: 150 pm. Ocafta et al. A survey of the genera Sarcodictyon and Rolandia. Zool. Med. Leiden 73 (2000) 418 Fig. 2. Histology of Sarcodictyon catenatum. A, polyp wall. B, cross-section of a tentacle. C, E & F, crosssections of mesenteries, retractor muscles and filament. D, detail of the pharynx and siphonoglyph. G, cross-section of stolon, H, detailsof the canals in the stolon. Scale bars: A = 75 pm; B = 13 pm ; C = 42 pm; D = 30 pm; E = 28 pm; F = 28 pm; G = 67 pm; H = 39 pm. couleur..., par les stolons, qui sont beaucoup plus etroits, n'ayant pas la moitie de la largeur des cellules polypiferes; et par la plus grande longeur des stolons entre les cellules, ...". This taxon could be another Sarcodictyon species, but could just as well belong to Scleranthelia or even Clavularia, as some species of the latter genus can have a striking external resemblance to Sarcodictyon. This is, for instance, the case with Sarcodictyon densum and Sarcodictyon charcoti, described from the Azores by Tixier-Durivault & d'Hondt (1974: 1367-1372; fi~s 3-8). These last-named species should be 419 Ocana et al. A survey of the genera Sarcodiclyon and Rolandia. Zool. Med. Leiden 73 (2000) placed in Clavularia (see Ocana et al., 1992: 423-424). Grieg (1887) described Rhizoxenia alba from the North Sea, which also clearly belongs to Clavularia (see Grieg, 1887: 15, pl. 1; Molander, 1918: 15-19, figs 5-6). As the type material of Evagora rosea Philippi, 1842, is almost certainly lost and because of the inadequate description we consider E. rosea as a nomen dubium. Genus Rolandia Lacaze-Duthiers, 1900 Diagnosis.- Polyps embedded in a common coenenchyme. Sometimes ribbonlike stolons are present, which can fuse into membranous expansions. The mesogloea of the polyp wall is thin and without lacunae, whereas it is thick and with many lacunae in the calyx. Retractor muscles well developed and restricted. Stolons with 9-12 canals. Sclerites scattered over the surface of the calyx. Sclerite types: crosses and "fused quadruplets", shuttles and six-radiates. Polyps without sclerites. Rolandia coralloides Lacaze-Duthiers, 1900 (figs 3-4) Rolandia coralloides Lacaze-Duthiers, 1900: 424-449, pis 15-16, description external morphology, sclerites and histology, near Algeria coast; Molander, 1929: 41-42, only remarks; Rossi, 1965: 163 & 174, record and ecology in coralligenous community, Punta di Mesco (Italy). Rolandia rosea; Weinberg, 1978a: 167-173, pis 16-18, description of external morphology, sclerites, discussion and synonymy, eggs observed from May to September, Banyuls-sur-Mer, Adriatic sea, Gulf of Naples, Gulf of Marseille, and near Algeria. Although the author included Thomson's record in his synonymy apparently he does not admit that R. rosea could be present in the Atlantic Ocean; Weinberg, 1978b: 39, foto 4, record and just some ecological data; Gili, 1982: 114 & 116, figs 57a-c, 58d, brief description and ecology, Banyuls de la Marenda, Illes Medes. Rhizoxenia rosea; Marion, 1882: 985, record, on Poseidonia sp. rhizomes, 15 m, Gulf of Marseille; Thomson, 1927: Pl. 3, figs 2-4, 6, 12; brief description but sufficient for recognizing the "fused quadruplets", 845 m, Azores. Evagora corii Broch, 1935: 5-8, fig. 2, description external morphology, sclerites show clearly the "fused-quadruplets", 15-20 m, Adriatic Sea (Split). Zoantha corii; Pax & Muller, 1962: 263-264, fig. 140, brief description and some data of sclerites which are larger than those in Zoantha rosea, Adriatico (Split), 15-20 m. Sarcodictyon catenatum; Berenguier, 1954: 62-65, pis 5-71; description, two ecological growth forms, R. coralloides and S. catenatum synonymous, Gulf of Marseille. ? Evagora rosea Philippi, 1842: 36, fig. 2, brief description, Naples. Not Zoantha thalassanthos Lesson, 1826, referred to Clavularia, see Weinberg, 1978a. Not Rhizoxenia thalassanthos; Ehrenberg, 1834: 55, only record and new combination. Not Rhizoxenia rosea; von Koch, 1891: 661-663, fig. 13, confusion with an Alcyonium species, Naples. Material.- Spain: (LNM/PI 101) Crinavis, 36°10.10'N 5°23.70'W, vii.1990, three colonies, infralittoral 20-25 m, P.J. Lopez leg.; (RMNH Cael. 23985) La Atunara, 36°10.54'N 5°20.00'W, iii.1988, two colonies, circalittoral 40 m, P.J. Lopez leg.; (RMNH Cael. 23986) Gibraltar harbour, 36°08.90'N 5°21.60'W, v.1992, two colonies, circalittoral28 m, P.J. Lopez leg.; (MNCN 2.04/726) Columbretes Islands (Western Mediterranean), 39°53.56'N 0040.20'E, 16.vii.96, one colony, 25 m, infralittoral, Fauna-IV cruise (stn. 278B8); (MNCN 2.04/727) Columbretes Islands (Western Mediterranean), 39°51.27'N 0040.55'E, 18.vii.96, two colonies, 30 m, infralittoral, Fauna-IV cruise (stn. 285B22); (MNCN 2.04/728) Columbretes Islands (Western Mediterranean), 30049.77'N 0032.10'E, 2O.vii.96, one colony, 10 m, infralittoral, Fauna-IV cruise (stn. 296B7); (MNCN 2.04/731) Hormigas Islands, 37°39.52'N 0038.92'W, 21.vii.96, ." Ocana et al. A survey of the genera Sarcodictyon and Rolandia. Zool. Med. Leiden 73 (2000) 420 Fig. 3. Sclerites of Rolandia coralloides. 1-13 anthostele; 14-25 stolon. Scale bar: 75 pm. Ocana et al. A survey of the genera Sarcodictyon and Ro/andia. Zool. Med. Leiden 73 (2000) 421 Fig. 4. Histology of Ro/andia coral/oides. A, cross-section of filaments and retractor muscles. B, crosssection of a tentacle. C, two polyps embedded with a connecting (solenial) canal. D, detail of the canal. E, solenium isolated in the mesogloea, F, three polyps embedded in the mesogloea. Scale bars: A, B & E = 50 pm; C &F = 200 pm.