Spiculosiphon oceana, Maldonado, Manuel, López-Acosta, María, Sitjà, Cèlia, Aguilar, Ricardo, García, Silvia & Vacelet, Jean, 2013

Maldonado, Manuel, López-Acosta, María, Sitjà, Cèlia, Aguilar, Ricardo, García, Silvia & Vacelet, Jean, 2013, A giant foraminifer that converges to the feeding strategy of carnivorous sponges: Spiculosiphon oceana sp. nov. (Foraminifera, Astrorhizida), Zootaxa 3669 (4), pp. 571-584 : 573-575

Spiculosiphon oceana sp. nov.

Etymology. The species name " oceana " is to honor the non-profit organization for ocean conservation OCEANA , which was responsible for the field collection of the type material.

Diagnosis. Non-attached, stalked and capitate test exclusively built with siliceous sponge spicules that are agglutinated together by organic cement. Undivided, hollow stalk, which slightly expands into a closed, bulb-like structure at the end contacting the substratum (proximal pole of the body). The radiating spicule tracts projecting from the central, globelike structure are always linear, never subdivided nor plumose.

Type material. The holotype (MNCN-33.11/2; fig. 1A) and the paratype (MNCN-33.11/3; fig. 1B) were complete organisms; five additional fragments, consisting of only the stalk or bearing a small, damaged portion of the globular part of the test, were stored as additional material (MNCN-33.11/4); two stalk fragments mounted on conductive transfer tape glued onto a SEM aluminum stub were also deposited in the collection (MNCN-33.11/4).

Morphology. Stalked and capitate, monothalamous test (fig. 1A). The capitate region of the test consisted of a central, hollow sub-sphere built with a large morphological variety of sponge spicules, which were loosely and irregularly agglutinated (fig. 1B–C). This globelike structure measured 1825 µm and 1250 µm in diameter in the holotype and paratype, respectively. The spicules at the globelike structure included small needle-like fragments, but also four-rayed forms (i.e., cladomes of diverse triaene types; fig. 2A–C), along with laminated fragments derived from discotriaenes and even pieces of hypersilicified desmata (not shown). Spiny spicules (acanthostyles) and microscleric spicules (e.g., diancistra of hamacanthid demosponges) were also observed when part of the globelike region of the test of a fragmentary specimen was dissolved in bleach (not shown). Irregular hollows and cavities were noticed in between the loosely agglutinated spicules that made the wall of the globelike structure (fig. 1C, 2B). These cavities appeared to result from an erratic selection and chaotic arrangement of the agglutinated sponge spicules rather than being distinct apertures elaborated by the cell.

Between 20 and 30 tracts made with agglutinated spicule fragments protruded radially from the globular centre, extending 1 to 3 mm outwards, and making the capitate region to reach about 3 to 4.5 mm in total diameter (fig. 1A–B). These radiating spiculated tracts were less than 60 µm in total thickness, built by overlapping 1 to 3, relatively thin and long, spicule fragments (fig. 2D). Unlike in the central globelike zone of the test, the spicule fragments used for the radiating tracts indicated high selectivity, being typically fragments of oxeas and styles, but also long-shafted triaenes with small cladomes. It is worth noting that whereas the cement agglutinating the spicules of the globular region was never evident under the compound microscope, the spicules of the radiating tracts were noticed to be held together by a translucent, quite evident cementing substance (fig. 2E–G), resembling the nodal spongin of demosponges.

The stalk was hollow, typically longer than 3.5 cm and averaging 500 ± 50 µm in external diameter (fig. 1A). No agglutinated material other than siliceous sponge spicules was observed to build the wall at the stalk (fig. 3A– B). There was the "quasi" exception of a small calcareous foraminifer on one of the stalk fragments. Nevertheless, it was not actually embedded or integrated in the spicule wall, but attached externally to an already formed spicule wall (fig. 1D, 3C). As the radiating tracts, the stalk was built with highly selected, needle-like

spicule fragments (occasionally including some long-shafted triaenes; fig. 3C), which were tightly agglutinated together (fig. 1E–F). The spicule fragments run almost in parallel to the direction of the major stalk axis, but slightly twisting around it (fig. 1E). Along the entire stalk, orifices or cavities were never noticed, except in those cases of accidental damage derived from collection or laboratory manipulation of the organisms (fig. 3A). SEM observations at broken portions of the stalks revealed that the test wall consists of spicules arranged in a single layer (not shown). Some of the spicules incorporated to the test showed incipient dissolution scars (pits) at their silica surface (fig. 3A). It was also noticed that the proximal end of the stalks was never attached to gravel pieces or rocks, being a closed structure that slightly expanded into a bulb-like form (figs. 1G, 3B).

Light microscopy and SEM observations revealed that some organic material (debris) flocculated on the external side of the stalks at some areas (figs. 1D; 3A, C). However, no obvious cementing material was ever noticed among the spicules when the external side of the tests was studied under light and scanning electronic microscopy (fig. 1E; 3A–C). When studying the stalk under light microscopy, discontinuous areas of brownishpurplish coloration were noticed (figs. 1D–G). SEM observations corroborated that those color tinges derived neither from mineral coatings laying at the external side of the test nor from organisms fouling the test. Rather, the irregular patchiness of coloration (figs. 1D–G) strongly suggested that the color probably derives from photosynthetic symbionts growing at the internal side of the highly translucent siliceous test (see Discussion).

X-ray microanalysis of the test. The exact nature of the cement agglutinating the sponge spicules of Spiculosiphon oceana could not be resolved in this study. It was corroborated that small portions of radiating spicule tracts, globular test, and stalk readily dissolved their cement in commercial bleach, releasing the agglutinated spicules and, hereby, confirming the organic nature of the cement. X-ray microanalyses on the external side of the stalks revealed occurrence of three detectable elements only (fig. 4A): oxygen (58.8 ± 1.3 %), carbon (27.8 ± 0.4 %), and silicon (13.4 ± 1.6 %). By considering that all the quantified silicon derived from the spicules' biogenic silica (Si O 2), it can be deduced that about 26.8 % (= 13.4 x 2) of the oxygen is also attributable to the silica, while most of the remaining oxygen (32 % = 58.8 - 26.8) would be associated to the detected carbon. That is, a bit more than half of the detected oxygen is probably combined with carbon to make the organic cement. A small carbon-oxygen percentage may also derive from occasional organic deposits formed by flocculation of debris on the external side of the stalks (fig. 3A, C). When stalk areas bearing pieces of attached debris were analyzed by EDS microanalysis (fig. 4B), it was found that, in addition to the major expected O, C, and Si elements, another two minority elements consistently occurred: Ca (2.7 ± 1.5 %) and Al (1.2 ± 0.3 %). Detection of these two elements was consistent with the presence of abundant organic calcareous debris and aluminum silicate clays at the detrital-sand bottom where the foraminifera lived. In two occasions, the analyzed debris also incorporated S (0.6 ± 0.02 %) and Ge (0.2 ± 0.03 %). More surprisingly, tellurium (Te), a quite uncommon element on earth, was once detected (0.77 %) in the attached debris. Interestingly, 5 out of 9 microanalyses conducted on the test of a calcareous foraminifer attached to a Spiculosiphon oceana stalk (figs. 1D, 3C) also revealed the presence of tellurium (3 ± 1.5 %; fig. 4C), along with O (46 ± 9 %), C (36 ± 9 %), and Ca (15 ± 5%).