Coelosphaera (Coelosphaera)

Van, Rob W. M., Kaiser, Kirstie L. & Syoc, Robert Van, 2011, Sponges from Clipperton Island, East Pacific, Zootaxa 2839, pp. 1-46 : 20-22

? Coelosphaera (Coelosphaera) sp.

( Figs 8 A–D)

Material examined. MNHN DCL 4056–A, Jean-Louis Etienne Expédition Clipperton 2005, station 16, 55 m, on dead antipatharians, 01–2005, 1 specimen.

MNHN DCL 4940, Jean-Louis Etienne Expédition Clipperton 2005, station 26, on dead antipatharians, 01– 2005, 1 specimen.

Description. Thin, transparent crust, growing on black corals ( Fig. 8 A). Less than 0.5 mm in thickness and less than 1 cm in lateral expansion. In alcohol, translucent blue, black coral substratum visible through sponge. Oscules not apparent, no other distinctive surface characteristics.

Skeleton ( Fig. 8 B). Loose bundles of tylotes, 10–20 spicules in cross section, rise from the substrate and fan out at the surface without forming a continuous tangential skeleton. Microscleres scattered in the surface membrane.

Spicules. Tylotes, sigmas and trichodragmata.

Tylotes ( Fig. 8 C), thin, with prominent globular heads, entirely smooth, often slightly curved, 177– 186.1 –207 x 2.5– 3.2 –4 µm, tyles 4– 4.8 –6 µm.

Sigmas ( Fig. 8 D), of common shape, fairly robust, 18– 22.9 –29 µm.

Trichodragmas ( Fig. 8 B, arrows), usually arranged in loose groups of 4–8 dragmas, each individual dragma peculiarly shaped with raphides arranged such that the dragmas are usually slightly swollen in the mid-section, 6– 8.4 – 11 x <0.5 µm.

Ecology. Deeper reef slope, on the dead base of bushy antipatharians at 55 m.

Remarks. We refrain from naming this sponge material, as it is too flimsy and possibly displays reduced spiculation. More elaborate specimens are necessary to properly describe it as a likely new species. Assignment of this material to the myxilline family Coelosphaeridae is made on the possession of smooth tylotes and sigmas as the main spicule types present. Assignment to the genus Coelosphaera , subgenus Coelosphaera , despite its lack of the usual isochela microscleres, is made on the basis of the lack of further megasclere types and on the lack of choanosomal structure. However, assignment to Coelosphaera is tentative and based on the assumption that isochelae somehow have become lost in the differentiation of the species. Coelosphaera are also usually hollow bladder-like sponges provided with fistules, whereas the new species is thinly encrusting. Although trichodragmas are commonly reported in Coelosphaera species, the precise shape and size in our new species are quite unusual and provide the principal characteristic of the species. It is not certain that the trichodragma-like structures (arrows in Fig 7 B) are indeed siliceous as we were unable to find them in the SEM preparations. An outside possibility is that the structures are organic and are e.g. cellular inclusions of unknown composition.

There is one other species in Coelosphaera possessing, on paper, similar spiculation: the North Atlantic thinly encrusting deep sea species Coelosphaera (Coelosphaera) macrosigma ( Topsent, 1890 as Gellius ), possessing tylotes, sigmas and trichodragmas. The similarity is not very great, however, as the spicules sizes are many times those of the present new species (tylotes of 700 x 10 µm, sigmas in two size categories, the larger of which is over 400 µm, and trichodragmas of 15–65 µm). For this North Atlantic material, the genus Xytopsoocha was erected by De Laubenfels (1936a). This is considered a junior synonym of Coelosphaera , but it should perhaps be revived now that the existence of two species answering to its diagnosis have been demonstrated. Hartman (in Lewis 1965) reports a Xytopsoocha sp. from Barbados deep water, but he did not give any description so the properties of this material remain uncertain.

Possibly the species could also fit in the coelosphaerid genus Lissodendoryx Topsent, 1892 , subgenus Anomodoryx Burton, 1934 , the only megascleres of which are likewise smooth tylotes, but for that assignment our specimen should have isochelae. Species so far in that subgenus have fairly massive or elaborate growth forms unlike our material.

Strongylacidon Lendenfeld, 1897 species have simple skeletal structures of loose bundles of diactinal megascleres and are known to possess reduced microsclere complements consisting of sigmas only (e.g. Caribbean S. viride Van Soest, 1984 ), but megascleres in that genus are strongyles, not tylotes.

Figures

FIGURE 7. Clathria (Thalysias) hermicola n. sp., holotype MNHN DCL 4043 – A. A. habit, thinly encrusting on dead corals, B. choanosomal styles with smooth and microspined heads, C. subectosomal subtylostyle with microspined head, D. ectosomal subtylostyle with microspined head, E. echinating acanthostyle with relatively lightly spined upper half, F. oxhorn-like tox, G. palmate isochela.

FIGURE 8.? Coelosphaera (Coelosphaera) sp., undetermined specimen MNHN DCL 4056 – A. A. habit encrusting black corals, B. light microscopy view to show skeleton, sigmas and numerous wedge-shaped presumed trichodragmas (arrows), scalebar = 50 µm, C. tylote and details of heads, D. sigma.