Geodia macandrewii, BOWERBANK, 1858

Cárdenas, Paco, Rapp, Hans Tore, Klitgaard, Anne Birgitte, Best, Megan, Thollesson, Mikael & Tendal, Ole Secher, 2013, Taxonomy, biogeography and DNA barcodes of Geodia species (Porifera, Demospongiae, Tetractinellida) in the Atlantic boreo-arctic region, Zoological Journal of the Linnean Society 169 (2), pp. 251-311 : 278-284

GEODIA MACANDREWII BOWERBANK, 1858

CYDONIUM P MACANDREWII (PhyloCode SPECIES NAME)

( FIGS 15–17 , TABLE 4)

Geodia macandrewii, Bowerbank, 1858: p. 284 ; 1872a: p. 196; Schmidt, 1866: p. 11 –12; Koltun, 1966: p. 50; Barthel et al., 1991: p. 40; Hougaard et al., 1991a: p. 225, b: p. 469; Warén & Klitgaard, 1991: p. 52; Ereskovsky, 1993: p. 23; Klitgaard 1995: p. 2; Reitner & Hoffmann, 2003: table 1; Murillo et al., 2012: p. 842; Schöttner et al., 2013: p. 2. Geodia macandrewia (misspelling), Cedhagen, 1994: p. 67. Geodia macandrewi (misspelling), Klitgaard & Tendal, 2004: p. 57; Cárdenas et al., 2010: p. 91; Cárdenas et al., 2011: table S1.

Cydonium normani, Sollas, 1888: p. 263 ; Norman, 1893: p. 347; Vosmaer, 1933: p. 102 (synonymy by this study).

Synops macandrewii, Sollas, 1888: p. 265 .

Sidonops macandrewii, Sollas, 1889: p. 277 ; von Lendenfeld, 1903: p. 100. Sidonops macandrewi (misspelling), Brøndsted, 1932: p. 5.

Geodia normani, Burton, 1930: p. 490 (synonymy by this study).

Misidentification:

Geodia megastrella, Kingston et al., 1979: p. 528 .

Type locality and deposition of holotype: Collected by Robert McAndrew. Vikna Island (formerly called Vigten or Vikten Island), North Trøndelag, Norway, 183 m. The holotype specimen was not found (E. Sherlock, BNHM, pers. comm.); only type slides were examined: BNHM 1877.5 .21.1396 (surface and two spicule preparations) and BNHM 1877.5 .21.1398 (one section, one spicule preparation) .

External morphology and cortex: Young specimens are spherical ( Fig. 15E ). The regular spherical body form persists during growth until the body reaches a size of about 10 cm in diameter; from that on the diameter is larger than the height, and our largest specimens are somewhat flattened, measuring 35–42 cm in maximum diameter and 20–24 cm in height ( Fig. 15A– C ). Some mid-sized specimens from eastern Greenland and the Flemish Cap ( Fig. 15D ) are unusually flat, the diameter measuring more than twice the height. The colour alive is whitish yellow to light grey; whitish in ethanol. The choanosome is yellowish in live specimens ( Fig. 15B ); whitish in ethanol. Small specimens have smooth surfaces; larger ones develop a fur of long spicules, essentially on the sides ( Fig. 15A ), which is frequently damaged during the catching procedure, so that large surface areas appear smooth. Cribriporal oscules ( Fig. 15F ) are regularly scattered over the upper surface; each sieve is 0.5–1 mm in diameter. Cribriporal pores ( Fig. 15G ) are scattered over the sides of the body; sieves are also 0.5–1 mm in diameter. The sterraster layer is very strong and tough, and usually 1–2 mm thick (occasionally up to 4 mm thick) ( Fig. 15H, I ). The ectocortex with spheroxyasters and microxeas is very thin (45–100 Mm thick) compared with the endocortex ( Fig. 15H, I ).

Spicules ( Fig. 16 , Table 4): Megascleres: (a) oxeas, straight or bent, length: up to 14 mm; width: 5–106 Mm. (b) Microxeas, straight or slightly bent, rarely centrotylote, length: 220–445 Mm; width: 3–13 Mm. (c) Ortho- to dichotriaenes, orthotriaenes are more common, straight rhabdome, rhabdome length: 1650–9625 Mm (maximum from Brøndsted, 1932); width: 40–165 Mm. Clads can often end with a small downward bend, orthotriaene clad length: 210– 1125 Mm; protoclad length: 220–500 Mm; deuteroclad length: 70–450 Mm. (d) Anatriaenes, common swelling on top of the cladome, rhabdome length up to 22 mm; width: 3.5–63 Mm; clad length: 24–285 Mm. (e) Meso/ protriaenes, rhabdome length: up to 12 mm; width: 5–79 Mm; clad length: 32–330 Mm; central clad length: 49–685 Mm. Microscleres: (f) sterrasters, spherical to subspherical, 124–360 Mm in diameter; thickness: 128–230 Mm; hilum diameter: 20–30 Mm. Rosettes are made of 5–7 warty rays; rosette diameter: 6–10 Mm. (g) Spheroxyasters, rough actines (difficult to see with the optical microscope), with centrum more or less developed, diameter: 4–18 Mm. (h) Oxyasters, thin rough actines, diameter: 10–88 Mm.

DNA barcodes: GenBank accession nos. EU442198 , HM 592689 , HM 592696 (Folmer COI): we have sequenced COI from specimens from western and northern Norway (4), Spitsbergen (2), Davis Strait (1), and Flemish Cap (1): the Folmer COI is identical in all these specimens. No. EU552082 (28S, C1-D2 domains): we have sequenced 28S from the Bergen area (2), and Spitsbergen (1), 1-bp difference was observed between the two Bergen specimens. No. KC481225 (18S), obtained from ZMBN 89717 (Spitsbergen). Two specimens from Spitsbergen sequenced: no variation was observed .

Distribution ( Fig. 17 ): We characterize the species as northern boreal, with the ability to invade some neighbouring cold-water areas, but not penetrating into real Arctic conditions. It has been recorded at depths from 157 m (Trondheimsfjord) to 1900 m (eastern Greenland). The temperature range is –0.82 °C (north of the Faroe Islands, BIOFAR st. 901) to 8.3 °C (south-west of the Faroe Islands, BIOFAR st. 69). Most eastern records are from between 230 and 400 m, at temperatures of 5–8 °C. Occurrences at temperatures below 2 °C were essentially found in the Denmark Strait (282–467 m depth). The records on the southern flanks of the Bill Bailey and Faroe Banks came from rather deep water, 1140 and 650 m, respectively. They hardly represent the southern distribution limit of the species but could indicate that off the Shetland Isles and Scotland it can be expected to occur on the upper slope. Brøndsted (1932) and Koltun (1966) mentioned the Shetlands as part of the distribution area; although this is to be expected, there is so far no proof, the record being a mistake by Brøndsted for the Norwegian record of the type specimen. Koltun (1966) mentioned occurrences in the south-western Barents Sea and the Denmark Strait, but gave no detailed information.

Biology: We found no indications of asexual reproduction. The predator chiton H. nagelfar and the parasitic foraminiferan H. sarcophaga have been found living on G. macandrewii ( Warén & Klitgaard, 1991; Cedhagen, 1994; Todt et al., 2009). More associated fauna has been investigated by Klitgaard (1995). The chemistry (elemental analysis, amino acids, sterols, and quaternary ammonium compounds) has been investigated ( Kingston et al., 1979; Hougaard et al., 1991a, b). Note that G. macandrewii off the Labrador coast ( Canada) and from the Faroe Islands have very similar sterol composition ( Kingston et al., 1979; Hougaard et al., 1991b).

Distinctive characters: External morphology: the regularly round, almost spherical form with no conspicuous openings, the hard consistency, the uniform distribution on the top side of cribriporal oscules, and the very thick cortex. Spicules: very large sterrasters (124–360 Mm in diameter).

Remarks: It was named after Robert MacAndrew who collected this species. Spicule measurements of this species are scarce apart from Sollas (1888), Brøndsted (1932), and Koltun (1966). Although this is a fairly common species in some areas, it is rarely mentioned in the literature compared with G. barretti or G. phlegraei , perhaps because it is often confused Means are in bold; other values are ranges; N = 30 unless stated otherwise in parentheses, or unless measurements come from other studies. A dash indicates that this measurement is not given in the literature. n.f., not found.

with these species. In spicule preparations, it can be difficult to separate the largest ectocortical spheroxyasters from the smallest choanosomal oxyasters as there is no clear transition from one form to the other, so we recommend measuring them on a section, to ensure not to mix both categories (this was done for the holotype and ZMBN 77924). There is some variation within oxyasters: small (up to 25–28 Mm) with thin actines (1–2 Mm thick) usually in specimens from shallower depths (183–600 m), to very large oxyasters (40–88 Mm) with thick actines (4–7 Mm thick) in specimens living deeper than 1000 m. The large sizes are not mentioned by Koltun (1966). The large dichotriaenes with atypical forward orientated cladomes represented by Bowerbank (1872a: plate X, fig. 4) were indeed observed on the type slide number ‘Bk.1398’. However, in our opinion, they result from a contamination from another Astrophorina (probably Stelletta normani Sollas, 1880a , a common species on the Norwegian coast at these same depths).

Molecular phylogenetic studies suggest that G. macandrewii belongs to Cydoniump along with G. cydonium , G. conchilega , and Geodia papyracea Hechtel, 1965 ( Fig. 2 ). G. macandrewii shares the possession of identical cribriporal oscules and pores with G. cydonium and G. conchilega ; G. papyracea , by contrast, has uniporal oscules ( Cárdenas et al., 2009). We could nonetheless suggest that the association of cribriporal oscules and oscules could be an autapomorphy of the clade, and that the cribriporal oscules were later modified and lost in G. papyracea . These four species also share spiny euasters but spiny euasters are also present in the neighbouring Geodia clades (Depressiogeodiap and Geodiap) ( Fig. 2 ). Geodia macandrewii , G. cydonium , and G. papyracea share orthotriaenes but G. conchilega has dichotriaenes. So additional sampling is needed to better understand the Cydoniump clade and find morphological apomorphies that would support it.

No spicule differences were observed between NWA and NEA specimens. Sterrasters are usually subspherical but we noted that one specimen from Davis Strait ( UPSZMC 78255 ) had some atypically shaped sterrasters, most of them ‘lemon-shaped’, but other specimens from the same locality had ‘normal’ sterrasters .

Geodia normani ( Sollas, 1888) is known only from the original description. The single specimen was collected by Rev. Canon A. M. Norman at Vikna Island, which is also the type locality for G. macandrewii . Burton (1930) worked on the Norman collection, but indicated that the single specimen was not at his disposal. Accordingly the specimen may have been untraceable even at that time. New searches in the BMNH collection for the present study have not been successful (E. Sherlock, BMNH, pers. comm., September 2012) so the type is presumably lost. We may suppose that Sollas (1888) did not examine the type specimen of G. macandrewii while describing G. normani , and misinterpreted Bowerbank’s description of G. macandrewii . Indeed, Sollas (1888) moved G. macandrewii to the genus Synops Vosmaer, 1882 and therefore did not think of comparing it with G. normani that he had placed in Cydonium Fleming, 1828 (because of the presence of cribriporal pores and oscules). However, the description of G. normani , notably concerning the fairly thick cortex, the cribriporal openings, and the spicule morphologies (large spherical sterrasters, common promesotriaenes, and anatriaenes with swollen cladomes), matches that of G. macandrewii . Geodia normani only differs from G. macandrewii by the absence of microxeas and we can suppose that these may have been overlooked or Sollas more likely thought they were a contamination. To conclude, because G. normani and G. macandrewii come from the same type locality and share so many characters (external morphology and spicules), we formally propose to consider G. normani as a junior synonym of G. macandrewii .

We examined slides (74-8-27.1) made by H. M. Reiswig from material published under the name Geodia megastrella and collected off the Labrador coast (53°24′50″N, 52°15′00″W) at 732 m depth ( Kingston et al., 1979). We observed a thick cortex (2 mm thick), orthotriaenes, common ana/protriaenes, very large sterrasters (up to 360 Mm), and spiny oxyasters with thin actines (10–30 Mm). This leaves no doubt that this is in fact G. macandrewii .

Figures

Figure 2. Molecular phylogeny of the Geodia genus, modified from Cárdenas et al. (2011: figure 2). Maximum-likelihood tree made from concatenated sequences of COI (Folmer fragment) and 28S (C1-D2 domains). Bootstrap nodal support values are given above the nodes: *, ± 75%; +, ±99% (2000 replicates).

Figure 15. External morphology and thick sections of Geodia macandrewii Bowerbank, 1858. A, specimen in situ at Traenadjupet, northern Norway, 66°58′N, 11°7′E, 293 m depth, not collected, specimen is c. 35 cm large. Notice two hermit-crabs on the top surface (picture taken by J. Schauer, Polarstern ARK-XXII/1a, 2007). B, specimen ZMBN 85207 cut in half just after collecting from the Røst reef, 312 m depth (Polarstern ARK-XXII/1a, 2007); specimen is 27 cm large. C, specimen ZMBN 89718 just after dredging from north of Spitsbergen, 215 m depth; specimen is 25 cm large (photo: A. Plotkin). D, flattened specimen UPSZMC 78254 from Flemish Cap, just after dredging at 1079 m depth (photo courtesy of NEREIDA Project). E, young specimen from the Labrador coast, SMNH-474. F, cribriporal oscules of ZMBN 85207. Scale bar: 1 mm. G, cribriporal pores of ZMBN 89717 (Spitsbergen, 215 m). H, thick section of ZMBN 77924 (Korsfjord, 200–400 m). Scale bar: 1 mm. I, close-up of cribriporal oscules (ZMBN 77924), a single sphincter (s) sits at the bottom of the oscule. Scale bars: D and E: 1 cm; F–H: 1 mm; I: 500 Mm.

Figure 16. Spicules of Geodia macandrewii Bowerbank, 1858. A–D, H and I, ZMBN 89717, Spitsbergen, 215 m depth. E–G, UPSZMC 78255, Davis Strait, 564 m. A, oxyaster. B, cortical spheroxyasters. A and B have the same scale. C, sterraster. D, microxea. E, orthotriaene. F, mesoprotriaene. G, anatriaene. H, close-up on the hilum of a sterraster. Scale bar: 10 Mm. I, close-up on the warty rosettes of a sterraster. Scale bar: 3 Mm.

Figure 17. Distribution of Geodia macandrewii Bowerbank, 1858 (map made with GeoMapApp, http://www.geomapapp.org). T, type locality.