From Marine Caves to the Deep Sea, a New Look at Caminella (Demospongiae, Geodiidae) in the Atlanto-Mediterranean Region

Zootaxa 4466 (1): 174–196 ISSN 1175-5326 (print edition) http://www.mapress.com/j/zt/ Article ZOOTAXA Copyright © 2018 Magnolia Press ISSN 1175-5334 (online edition) https://doi.org/10.11646/zootaxa.4466.1.14 http://zoobank.org/urn:lsid:zoobank.org:pub:1DDBA124-7964-4F4A-902B-4410D1E3C042

From marine caves to the deep sea, a new look at Caminella (Demospongiae, Geodiidae) in the Atlanto-Mediterranean region

PACO CÁRDENAS1,2,4, JEAN VACELET2, PIERRE CHEVALDONNÉ2, THIERRY PÉREZ2 & JOANA R. XAVIER3 1Pharmacognosy, Department of Medicinal Chemistry, Uppsala University, BMC Box 574, SE-75123 Uppsala, Sweden. 2Institut Méditerranéen de Biodiversité et d’Ecologie marine et continentale, CNRS, Aix Marseille Univ., IRD, Avignon Univ., Station Marine d’Endoume, chemin de la Batterie des Lions, 13007 Marseille, France. 3Department of Biological Sciences and K.G. Jebsen Centre for Deep-Sea Research, University of Bergen, PO Box 7803, N-5020 Bergen, Norway 4Corresponding author. E-mail: [email protected], http://orcid.org/0000-0003-4045-6718

Abstract

Caminella Lendenfeld, 1894 is a poorly known Geodiidae genus with unclear phylogenetic relationships. In order to find new lines of evidence that could shed light on the evolutionary history of Caminella, we decided to revise type material and museum material, as well as examine new material from underwater caves and deep-sea ecosystems. In doing so, we formally show that Isops maculosus Vosmaer, 1894 and Caminella loricata Lendenfeld, 1894 are junior synonyms of Caminella intuta (Topsent, 1892). We discuss different spicule morphological phenotypes in C. intuta, which may be linked to silica availability. We also discovered two new species of deep-sea Caminella: 1) from Cape Verde (Caminella caboverdensis sp. nov.) and 2) from seamounts located south of the Azores archipelago and the North of Spain (Caminella pustula sp. nov.). We reveal that Caminella sterrasters have complex surface microstructures, unique amongst the Geodi- idae, where actin tips are linked to each other. Molecular markers (COI, 28S (C1-D2) and 18S) sequenced for some specimens led to new phylogenetic analyses, which continue to suggest a close relationship of Caminella with the Erylinae and Calthropella; these affinities are discussed in light of morphological characters.

Key words: Porifera, Tetractinellida, new species, seamount, sterrasters, silica

Introduction

When Topsent (1892) described Cydonium intutum Topsent, 1892, a small, brown, shallow-water sponge, from Banyuls-sur-Mer (France), he thought he had found a new member of the Geodia group. A year later, based on a better observation of the openings (uniporal oscules and pores), he reallocated this species to a different type of Geodia, the genus Isops (Topsent 1893, 1894), to become Isops intuta. Meanwhile, Vosmaer (1894) and Lendenfeld (1894) independently described very similar specimens for which they both created new species: Caminella loricata Lendenfeld, 1894 from Croatia and Isops maculosus Vosmaer, 1894 from the deep sea around Naples. The reason for the establishment of the new genus Caminella by Lendenfeld (1894) was the presence of small cortical spherasters/spherules (called ‘microdesmen’ by Lendenfeld) that were unusual for Geodia-like species. However, this genus was not considered valid, as Topsent (1894, 1895) and Vosmaer (1933) both thought it was a junior synonym of Isops (=Geodia). It was only when Isops intuta was finally sequenced and included in a molecular phylogenetic study (Cárdenas et al. 2011) that its evolutionary relationships were revealed. Instead of grouping with Geodia species, it appeared alone on a long branch, sister taxon to Calthropella. The Erylinae Sollas, 1888 (Erylus Gray, 1867; Penares Gray, 1867, Pachymatisma Bowerbank in Johnston, 1842, Caminus Schmidt, 1862, Melophlus Thiele, 1899) and the Pachastrellidae Carter, 1875 were sister to the Geodia intuta+Calthropella clade. This unique position was the reason behind the decision to resurrect the genus Caminella by Cárdenas et al. (2011).

174 Accepted by M. Klautau: 26 Jun. 2018; published: 31 Aug. 2018 Although it is consistent with mitochondrial (COI) and nuclear (28S (C1-D2)) molecular markers, the phylogenetic position of Caminella is poorly supported. The grouping of Caminella and Calthropella is also unexpected because Caminella looks like a Geodia, and no relationship with Calthropella was ever suspected. In order to find new lines of evidence that could shed light on the evolutionary history of Caminella, we decided to revise type material and other museum material, as well as examine new samples collected from underwater caves and deep-sea ecosystems in order to better understand and define this unique genus. In doing so, we discovered two new species of deep-sea Caminella, one from Cape Verde and the other from seamounts located south of the Azores archipelago and North of Spain.

Material and methods

Abbreviations:

BMNH Natural History Museum, London, UK. MNCN Museo Nacional de Ciencias Naturales, Madrid, Spain. MNHN Muséum National d’Histoire Naturelle, Paris, France. NHMW Naturhistorisches Museum, Wien, Austria. RMNH The State Museum of Natural History, Leiden, The Netherlands. SME Station Marine d’Endoume, Marseille, France. UF Florida Museum of Natural History, University of Florida, Gainesville, FL, USA. ZMAPOR Zoological Museum in Amsterdam (collections now moved to Naturalis, Leiden, The Netherlands). ZMB Zoologisches Museum Berlin, Germany.

Sponge sampling and records. Several specimens were collected by SCUBA diving in and outside Mediterranean caves around Marseille and La Ciotat (France), as well as in the so-called ‘lithistid’ cave near Chak El Hatab (Lebanon). Underwater pictures of the specimens were taken in situ before collection. Another specimen was collected by SCUBA diving in the semi-dark cave ‘Carreiro Maldito’ in the Berlengas islands, Portugal. A deep- sea specimen was dredged during the SponGES 0617 expedition on board of the B/O Ángeles Alvariño to ‘El Cachucho’ (= Le Danois Bank) in the Bay of Biscay (Cantabrian Sea), in June 2017. Other deep-sea specimens were discovered while sorting the as yet unpublished sponge collection resulting from the ‘Seamount 2’ campaign to the seamounts south of the Azores (MNHN, 1993). More specimens and type material came from various museums or collections. These specimens, and additional Caminella records compiled from the literature, were plotted on a map using GeoMapApp 3.3 (http://geomapapp.org) (Fig. 1). Collection information on specimens examined in this study, previous records, and Genbank accession numbers, were archived in the PANGAEA data repository (https://doi.pangaea.de/10.1594/PANGAEA.891082). Morphological and molecular studies. For all specimens studied, a piece of sponge tissue was digested in bleach. The remaining spicules were washed successively with water, 50% ethanol and 100% ethanol; spicules mounts were made using EukittTM mounting medium (Sigma-Aldich, St Louis, MO, USA). Twenty-five spicules per spicule type were measured, unless otherwise stated, using a light microscope and an eyepiece micrometer (for C. intuta and the new species from Cape Verde) or using the Leica Application Suite V4.5 (for the new species from the seamounts). The spicules of most of the specimens were also examined using a Scanning Electron Microscope (SEM). For this, we digested a small piece (1 mm) with nitric acid on a glass slide over a Bunsen burner. The slide was rinsed with water, dried, and then coated with a mix of gold/palladium before being examined with a Hitachi S-570 SEM (Marine Station of Endoume). Thick sections from specimen ZMAPOR 21653 were made with a diamond wafering blade and a low speed saw using an Agar Low Viscosity Resin kit (© Agar Scientific) in accordance with the manufacturer’s mixing instructions to make a hard embedding medium. DNA was extracted using the DNeasy® Blood and Tissue kit (Qiagen, Hilden, Germany) in accordance with the manufacturer’s instructions. Polymerase chain reactions (PCRs) were made in 25 μl solutions using PuReTaqTM Ready-To-GoTM PCR beads (GE Healthcare, Little Chalfont, UK). Of all our specimens, ZMAPOR 21653 from a cave in Portugal was one of a few that had been properly preserved for molecular work, in ethanol 96% in 2005. We therefore focused on this specimen and sequenced COI, 28S (C1-D2) (Cárdenas et al., 2011) and

CAMINELLA IN THE ATLANTO-MEDITERRANEAN REGION Zootaxa 4466 (1) © 2018 Magnolia Press · 175 18S (this study) for this specimen. The nearly complete 18S (1690 bp) was obtained using the same protocol that worked for other Tetractinellida (Cárdenas et al. 2013; Kelly & Cárdenas 2016). We also tried to sequence COI (Folmer, 658 bp) for other specimens using primers LCO1490 and HCO2198 (Folmer et al. 1994) with the PCR program: [5 min/94 °C; 37 cycles (15 s/94 °C, 15 s/46 °C, 15 s/ 72 °C); 7 min/72 °C]. 28S (C1-D2, 813 bp) was sequenced using the primer pair C1’ASTR/D2 and the PCR protocol detailed in Cárdenas et al. (2011). For one specimen from Cape Verde (RMNH 3810) collected in 1982, the DNA was fairly degraded so we decided to amplify COI in two parts. For the first part we amplified the universal mini-barcode (130 bp) using primer pair LCO/Tetract-minibarR1 following the protocol from Cárdenas & Moore (2018). For the rest of the Folmer fragment we designed a forward primer in the mini-barcode to target Erylus species—ErylusCOI-F2 (5’- CTCCYGGATCAATGTTGGG-3’)—and coupled it with HCO2198 always using the same COI PCR program. Likewise, instead of amplifying the 28S (C1-D2) region of the Cape Verde specimen which was too long, we amplified the shorter 28S (C1-C2) region (369 bp) using the primer pair C1’-Ep3 (Chombard et al. 1998). To discuss the phylogenetic position of Caminella, we pruned the comprehensive Tetractinellida COI and 18S alignments from Kelly & Cárdenas (2016), to keep only Astrophorina species. We added additional Astrophorina COI sequences from Galapagos species (Schuster et al. 2018) and our sequences. Analyses were conducted with the CIPRES science gateway (http://www.phylo.org) (Miller et al. 2010): RAxML 8.2.10 (Stamatakis 2014) for maximum likelihood (ML) and MrBayes v. 3.2.6 (Ronquist et al. 2012) for Bayesian analyses. Bootstrap Bayesian analyses consisted of two runs of four chains, each for 5 000 000 generations and sampled every 200th tree after a 25% burn-in.

FIGURE 1. Distribution map of Caminella intuta (Topsent, 1892) in blue, Caminella caboverdensis sp. nov. in orange, Caminella pustula sp. nov. in white.

Results

Order Tetractinellida Marshall, 1876

Sub-order Astrophorina Sollas, 1888

Family Geodiidae Gray, 1867

Genus Caminella Lendenfeld, 1894

Type species. Caminella loricata Lendenfeld, 1894 (by monotypy)

Diagnosis (new). Erylinae with sterrasters with actins linked to each other through bridges (giving a brain-like surface) and cortical spherasters and/or spherules.

Caminella intuta (Topsent, 1892) (Figures 2–6, Table 1)

Synonyms. Cydonium intutum Topsent, 1892: Topsent 1892, p. XVIII. Isops intuta (Topsent, 1892): Topsent 1893, p. XLIII; Topsent 1894, p. 336, pl. XI, pl. XVI; Lendenfeld 1903, p. 95–96; Sarà 1961, p. 31; Boury-Esnault 1971, p. 296; Pouliquen 1972, Table 1; Templado et al. 1986, Table 1. Isops intutus (Topsent, 1892): Vosmaer 1933, p. 145. Isops intuta (Vosmaer, 1894): Maldonado 1992, Table 1 (mistake in authority). Caminella intuta (Topsent, 1892): Cárdenas et al. 2011; Sitjà & Maldonado 2014, Table 2. Isops maculosus Vosmaer, 1894: Vosmaer 1894, p. 273–274. Isops maculosa Vosmaer, 1894: Lendenfeld 1903, p. 96. Caminella loricata Lendenfeld, 1894: Lendenfeld 1894, p. 150–151, pl. II, pl. III, pl. VIII; Lendenfeld 1903, p. 89–90.

Not Isops intuta (Topsent, 1892): Boury-Esnault et al. 1994, p. 41 = Geodia sp. (this study)

Holotype. MNHN DT-2290 (slide with sections, in bad state), Cap l’Abeille, Banyuls, France, 25-30 m. Material examined. Caminella intuta. France: SME, collection ‘Topsent’, box 1, slide#12 labeled “Isops intuta”, Banyuls-sur-Mer (type locality), dredge; SME #90S, dry piece and slide, 17.03.1958, Le Petit Congloué, Archipel de Riou, Marseille, cliff facing NE, 40 m, coll. J. Vacelet; SME #193S, wet specimen and slide, 26.08.1998, Gameau cave, La Ciotat, 6-8 m, preserved in formalin, coll. J. Vacelet; SME PL617PC-7 (Fig. 2C, 2E) and PL617PC-10b, 18.05.2016, Cosquer cave, Calanque de la Triperie, Parc National des Calanques, coll. P. Chevaldonné, ethanol 96%; Lebanon: SME 21/09/2002-56a, 5/07/2003-1 (sac 9), 13/07/2003-2, preserved in formalin, Chak El Hatab (north of Selaata, Lebanon), "lithistid cave", in dark area, 2-3 m, colls. T. Pérez and J. Vacelet; SME field# 080524-Lb2-03, 24.05.2008, Chak El Hatab, “lithistid cave”, 2 m, coll. T. Pérez, ethanol; Alboran Sea: field# Sp175, station BV 41, 35.99362, -2.86795, 102-112 m, beam trawl, INDEMARES- ALBORAN, 21.07.2012, coll. S. Gofas, identified by Sitja & Maldonado (2014), preserved in formalin (Fig. 2F); Portugal: ZMAPOR 21653, field# B.05.09.268, Gruta do Carreiro Maldito, Berlengas, Portugal, 6-8 m, 20.09.2005, ethanol 96%, coll. A. Cunha (Fig. 2D, 3F–G). Isops maculosus, lectotype (here designated), RMNH Por 644 (not seen), Gulf of Naples, between Capri and Naples, 150-200 m, wet specimen labeled as 'Isops intutus Tops. Corallieri, 17 Mei 1884 coll. Vosmaer N.249', and 18 slides, some of which are labeled Isops maculosus, but all have the number 249. Slides are mostly histological sections, but two are spicule slides. According to GBIF records from the RMNH collection (https://www.gbif.org/ occurrence/search?taxon_key=5892880), there are at least seven other paralectotypes (here designated): RMNH Por 78, 645, 646, 647, 648, 649, 650; BMNH 1955.3.24.2 (seen), two fragments of paralectotype RMNH Por 647 (~2.5 x 1.5 cm), 12.12.1890, Gulf of Naples. Caminella loricata, holotype (not seen), ZMB Por 2423, Lesina (=Hvar), Adriatic Sea, Croatia, one small piece 1 cm x 0.5 cm, purchased in 1897, no slides; NHMW-3Zoo-EV-MP345 (Fig. 3A, 3C–E), slide from holotype with a thick section (seen high resolution pictures); NHMW-3Zoo-EV-MP346 (Fig. 3B), slide from holotype with histological sections (seen high resolution pictures). Geodia sp. SME, Balgim slide box ‘Tetractinellida’, three slides labeled ‘CP–63 (185)’ (two spicule preparations and one piece of cortex), off Morocco, 1510 m, originally identified as Isops intuta by M. J. Uriz (Boury-Esnault et al. 1994).

CAMINELLA IN THE ATLANTO-MEDITERRANEAN REGION Zootaxa 4466 (1) © 2018 Magnolia Press · 177 External morphology and skeleton organization. (Figs. 2–3) Massive subglobular, up to 10 cm wide (Fig. 2C) with smooth, clean surface. Cave specimens from France and Portugal area are brown, to dark brown (Fig. 2A, 2C, 2D). Cave specimens from Lebanon (Fig. 2B) are pure white (small specimens), white to cream or light brown (large specimens). Deep-sea specimens are cream-colored to brown. Internal color is white to light brown. Colors are retained in ethanol and formalin. One to several uniporal oscules (0.2–2 mm in diameter) can be present, each leading into a cloaca (Fig. 3F). Cave specimens from Lebanon generally have a single oscule at the top, more rarely two or three, while other cave specimens often have several oscules. Oscules can have a slightly elevated margin in life, especially visible in underwater photographs (Fig. 2A, 2B), often flush with surface after preservation (Fig. 2E). Surface punctured by numerous uniporal pores (20–200 µm in diameter in cave specimens; 20–55 µm in deep- sea specimen), which can have elevated surrounding walls (Fig. 2E), or not (Fig. 2F). Oscules and pores can be surrounded by a conspicuous dark ring (especially in cave specimens) (Fig. 2A, 2C), or not (deep-sea specimens) (Fig. 2F). Consistency fleshy. The cortex is thin (0.2–0.5 mm thick), more or less flexible and easily detachable from the underlying choanosome. The ectocortex is composed of a dense aggregation of spherasters/spherules while the endocortex is made of sterrasters (Fig. 3D–E). Dichotriaenes support the cortex, but do not cross it (Fig. 3G). In the choanosome, a few oxeas are more or less radially organized (Fig. 3F–G). In the choanosome, oxyasters and sterrasters are common while spherasters/spherules are rare and predominantly found around the canals. Spicules. (Figs. 4–6) (Table 1) (a) oxeas, a few styloids, 900–2500 x 6-50 µm; (b) dichotriaenes (rhabdome: 370–2000 x 12–60 µm; protoclad: 60–237 µm; deuteroclad: 25–410 µm), rarely orthotriaenes; (c) sterrasters, spherical to oval, 40–84 µm, surface without clear rosettes, instead the actins build bridges between them making a brain-like surface, which is then covered with small warts; immature sterrasters have blunt actins covered with small branches that sometimes link two actins together making a honeycomb surface (similar to Placospongia selenasters). From here on we will distinguish ‘young’ (not fully grown) from ‘immature’ (fully grown but underdeveloped) sterrasters; (d) oxyasters, 3–8 actins, 10–36 µm in diameter, actins are finely acanthose and blunt at the very tip; center is more or less developed; (e) spiny spherasters to spherules, different ratios depending on the specimens, sometimes with very irregular spherasters, 3–15 µm in diameter. We observed four types of spicule phenotypes depending on the origin of the specimens. 1) Cave specimens from France and Portugal, as well as C. loricata from Croatia had immature sterrasters associated with irregular to regular spherasters, which rarely become spherules (Figs. 3E, 4). Dichotriaenes can also be irregular; orthotriaenes are occasionally found. 2) Specimens from shallow-water but deeper (25–40 m) and not in caves (specimens from the type locality identified by Topsent and from 40 m in Marseille) have immature sterrasters associated with irregular to regular spherasters, many of which are spherules. 3) Cave specimens from Lebanon (all coming from the same cave) had mature sterrasters and only spherules. Megascleres are larger and robust (Fig. 5). 4) Deep-sea specimens (Isops maculosus paralectotype and specimen from the Alboran Sea) had mature sterrasters with regular spherasters, which have on average shorter actins, and often become spherules (Fig. 6). Bathymetric range. In caves, C. intuta lives in semi-dark to dark areas, where it can be found as shallow as 2 m depth. Outside caves, it has been recorded from 25–30 m in Banyuls-sur-Mer (Topsent, 1892) to 300 m in the Alboran Sea (Templado et al. 1986). DNA barcoding. COI. ZMAPOR 21653 (HM592740) and cave specimen from Lebanon (080524-Lb2-03) had a 100% identical COI (MH477613). 28S (C1-D2). ZMAPOR 21653 (HM592804) and cave specimen from Lebanon (080524-Lb2-03, MH478114) were 100% identical. Both had a 5 bp difference with PL617PC-10b from Cosquer cave (MH478115). 18S. ZMAPOR 21653 (MH478118). ZMAPOR 21653 was submitted to the Sponge Barcoding Project (http://www.palaeontologie.geo.uni-muenchen.de/SBP/) with accession number 1777. Remarks. Suspicions of synonymy with Isops maculosus and Caminella loricata were raised early on by Topsent (1895, p. 580–581) who suggested that i) Isops maculosus might be a synonym of Isops intuta, although its colour seemed different and its sterrasters larger, ii) Caminella loricata might be a synonym of Isops intuta since the ‘microdesmen’ described by Lendenfeld (1894) were probably spherasters. However, Topsent (1895) did not conclude since he had not seen type material and Vosmaer (1894) had not given any measurements or illustrations of I. maculosus. This issue was later taken up by Vosmaer (1933), finally giving some spicule measurements of I. maculosus. He concluded that his species, I. maculosus and C. loricata, are junior synonyms of I. intuta. Once again, these conclusions were not based on the comparison of type material. The present study is the first one to examine the type material from Cydonium intutum, Isops maculosus, and Caminella loricata. We confirm that I.

178 · Zootaxa 4466 (1) © 2018 Magnolia Press CÁRDENAS ET AL. maculosus and C. loricata are junior synonyms of Caminella intuta. And yet, we do note some external morphology and spicule differences within C. intuta specimens, which will be discussed below.

FIGURE 2. Caminella intuta (Topsent, 1892). A. Specimen in Gameau cave, La Ciotat, France (not collected), note the dark ring around uniporal oscules (large openings) and pores (small openings); B. Specimens in "lithistid cave", Chak El Hatab, Lebanon; C. Large specimen in Cosquer cave, France, field# PL617PC-7; D. ZMAPOR 21653 (after preservation in ethanol) from Carreiro Maldito cave, Berlengas, Portugal; E. Oscules (large) and pores (small) from specimen PL617PC-7, Cosquer cave (fixation in ethanol); F. Oscule and minute pores from specimen Sp175, Alboran Sea, 102-112 m depth (preserved in formalin, storage in ethanol).

CAMINELLA IN THE ATLANTO-MEDITERRANEAN REGION Zootaxa 4466 (1) © 2018 Magnolia Press · 179 FIGURE 3. A. NHMW-3Zoo-EV-MP345, slide of holotype of Caminella loricata Lendenfeld, 1894; B. NHMW-3Zoo-EV- MP346, slide of holotype of C. loricata. C. Section of holotype of C. loricata showing dichotriaenes, sterrasters and numerous oxyasters, scale: 100 µm (slide 345, picture: O. Macek); D. Section of holotype of C. loricata showing uniporal pores in the cortex of sterrasters, scale: 100 µm (slide 345, picture: O. Macek); E. close-up of picture D showing a uniporal pore filled with spherasters. Note the undeveloped status of the sterrasters, scale: 50 µm (slide 345, picture: O. Macek); F. Section of Caminella intuta (ZMAPOR 21653), from Carreiro Maldito cave, Berlengas, Portugal. Note the cloaca with oscule opening. G. Same section as in F, showing dichotriaenes, oxeas and sterrasters in the choanosome, as well as the thin cortex.

180 · Zootaxa 4466 (1) © 2018 Magnolia Press CÁRDENAS ET AL. FIGURE 4. Caminella intuta (Topsent, 1892). A–B. Microscleres from specimen from Gameau cave, La Ciotat, France, 193S. A. Immature sterraster. B. Oxyasters and spherasters. C–D. Microscleres from specimen from the ‘Carreiro Maldito’ cave, 6-8 m, Berlengas, Portugal, ZMAPOR 21653. C. Young and immature sterrasters, oxyasters, spherasters (same scale as in A). D. Surface of immature sterraster with a spheraster.

CAMINELLA IN THE ATLANTO-MEDITERRANEAN REGION Zootaxa 4466 (1) © 2018 Magnolia Press · 181 FIGURE 5. Caminella intuta (Topsent, 1892), #13/07/2003-2 from “lithistid” cave, 2–3 m, Chak El Hatab, Lebanon. A. Mature sterraster, oxyasters and spherules. B. Spherules. C. Dichotriaene. Note the sterrasters at the tip of the rhabdome. D. Oxea (same scale as C).

182 · Zootaxa 4466 (1) © 2018 Magnolia Press CÁRDENAS ET AL. FIGURE 6. Caminella intuta (Topsent, 1892). A–D. BMNH 1955.3.24.2 (=RMNH Por 647), paralectotype of Isops maculosus Vosmaer (1894), Gulf of Naples, 150–200 m. A. Mature sterrasters. B. Close-up of mature sterrasters on hilum. C. Oxyasters. D. Spherasters and spherules. E–G. field# Sp175, Alboran Sea, 102–112 m. E. Mature sterrasters (same scale as A). F. Spherasters (same scale as D). G. Oxyasters (same scale as C).

-50 -2500/ -2500/ -22 -22 -1700/ -1700/ -2000/ -2000/

25.8 13.8 15.7 1868 23-28 23-28 15-20 15-20 Oxeas Oxeas 1382 1620 7- 6- 15- 1000-2500/ (length/width) (length/width) 980- 900- 1250-

……continued on the next page -300 -280 -250/ -250/ -170/ -170/ -410/ -237/ -237/ -140/ -140/

- - 1200-1600/ 229 234 98 93 169 131 237 147 (N=5) (N=14) (N=14) 295-310 135-150/ ortho clad) 60- 25- (irregular) 80- 80- 80- 82- 175- 167- deuteroclades/ Dichotriaenes (proto/ -30 -60 -24 -580/ -900/ -890/ ies; means are presented -ies; italics; means are ranges; in other values

40 40 30 30 540 22.3 727 47.9 677 17.7 2000/ (N=3) (N=11) (N=11) 17- 32- 12- (rhabdome: (rhabdome: length/width) length/width) Dichotriaenes None measured measured None measured None measured None -25 500- -27 370- -25 400- -18

15.4 18.6 15.9 5-20 800/ 800/ 5-20 12.9 20-24 20-24 8- Oxyasters (diameter) (diameter) collected at different depths collected thedifferent bold are In across at Atlanto-Mediterranean region. 10- 10- 10- slightly spiny spiny slightly (10-12 actins), Caminella -58 -72 -46 -82/ -82/ -84

72 60 60 51.9 57.1 42.2 68.1 oval) 50-60 50-60 56- immature) immature) Sterrasters Sterrasters (spherical, (spherical, 47- 40- 40- (spherical, (spherical, (spherical, (spherical, immature) immature) immature) immature) (spherical) (spherical) 47- (spherical to to (spherical (length/width) (length/width) -8 -13 -15 -10

5.8 5.7 8.4 6.5 (length) spherules spherules) spherules) irregular) spherules) spherules) spherules) spherules) (spherules) (spherules) Spherasters/ (spheraster to (spherasters, (spherasters, (spherasters to (spheraster to (spheraster (spherasters to to (spherasters

0.5 3- 0.5 (mm) (mm) Cortex thickness m) and specimens of for thickness (mm) m) cortex

- 0.3-0.5 3-8 2 0.3-0.4 3- 2 0.3-0.4 6-8 0.3-0.4 3- 6-8 0.25-0.5 5- (m) (m) 25-30 0.2-0.4 0.2-0.4 25-30 5-7 Depth 150-200

193S Individual Individual dimensions spicule ( Material holotype holotype Lesina, Italy Italy Lesina, Gameau cave, Gameau paralectotype, paralectotype, (Topsent, 1894) Banyuls, France Banyuls, “lithistid cave” Caminella intuta Isops maculosus #080524-Lb2-03 ZMAPOR21653 La Ciotat, France Ciotat, La Caminella loricata (Lendenfeld, 1894) 1894) (Lendenfeld, (=RMNH-Por647) BMNH 1955.3.24.2 Bay of Naples, Italy Italy Naples, of Bay Berlengas, Portugal Berlengas, Portugal Carreiro Maldito Maldito cave, Carreiro Chak El Hatab, Lebanon El Hatab,Chak Lebanon Holotype, MNHN-DT2290 = not referred. TABLE 1. specimens in comespecimens between otherwise N=25otherunless measurements parentheses,from measured this stud unless stated orstudy.

-1737/ -1737/ -23 -2120/ -2120/ -1525/ -1525/ -19 -25

15 20 31 31 15.1 1502 2685/ 20-40 20-40 22-31 22-31 (N=3) (N=3) (N=2) Oxeas Oxeas (N=1) 1365 1135 > 2600/ 12- 15- 8- 1815-1947/ (wavy oxeas) (wavy (length/width) (length/width) 955- 800- 1044-

202/ 202/ -633 -790 -386 -452 -493 -296/ -200/ -239/ -239/ -150/ -150/ -230/ -115/

- 525 457 295 339 426 243 172 92 192- 182 109 165 (N=4) (N=6) (N=4) (N=19) (N=19) (N=10) (N=10) (N=13) (N=13) ortho clad) 86- 80- 50- 437- 140- 147- 108- 297- 205- 152- 169- deuteroclades/ Dichotriaenes (proto/ -50 -900/

69 69 21 21 75 75 775 37.5 43-70 43-70 (N=1) (N=2) (N=4) (N=2) (N=1) 20- (rhabdome: (rhabdome: length/width) length/width) Dichotriaenes -78 581-1100/ 581-1100/ -78 -51 1043/ -42 650--42 -51 797/-51 -55 None measured measured 70- None -55 -36 657-690/-36

39 45 23 43.5 40.8 18.9 27- 32- 12- 8- Oxyasters (diameter) (diameter) 32- 30- -88 -45 -92 -84 -107/ -73 -77 -91/ -91/ -82/ -82/ -106/ -103/

65 65 84 72 82.8 38.8 83.8 72.8 98 95 (oval) (oval) (oval) (oval) (oval) (oval) (oval) (oval) 101.5 53- 47- 70- 54- Sterrasters Sterrasters 76- 35- 69- 63- (spherical) (spherical) (spherical to to (spherical slightly oval) slightly oval) 88- (length/width) (length/width) -8 -13 -8 -10 -13 82- -13 -8.6 90- -8.6

5.1 6.6 6.4 7.0 7.5 5.9 (length) spherules spherules) spherules) spherules) spherules) (spherules) (spherules) (spherules) (spherules) Spherasters/ (spheraster to (spheraster (spherasters to to (spherasters

0.5 3.3- 0.5 (mm) (mm) Cortex thickness

75 0.15 2.5- 0.15 75 (m) (m) 545 - 545 4- 705 0.5 4- 660 - 660 5- 735 - 735 5- Depth 102-112

sp. sp.

sp. nov. sp.

nov. (Continued) Material paratype paratype paratype Holotype Cape Verde Verde Cape field# Sp175field# El Cachucho El Cachucho Alboran Sea, Sea, Alboran Cantabrian Sea Sea Cantabrian Plato Seamount Seamount Plato MNCN 1.01/1017 Hyères Seamount Hyères MNHN-IP-2008-4 Atlantis Seamount Atlantis MNHN-IP-2008-148 MNHN-IP-2008-149 Holotype, RMNHHolotype, 3810 Caminella pustula Caminella caboverdensis

TABLE 1. TABLE 1.

CAMINELLA IN THE ATLANTO-MEDITERRANEAN REGION Zootaxa 4466 (1) © 2018 Magnolia Press · 185 The singularity of Caminella sterrasters was foreseen by Topsent (1894) who noted that sterrasters had complex and ornate surfaces. Later Vosmaer (1933, p. 148) very accurately observed that it “appears as if spines of neighboring actins were fused together” but concludes this is probably not the case and that “the truth being that they are only crowded close together”. The unique pattern and microstructures of the sterrasters’ surfaces is revealed here for the first time with SEM (Figs. 4–7, 9). The surface of Caminella sterrasters is indeed quite different from Geodia sterrasters which have characteristic star shaped structures called ‘rosettes’ (4–8 µm in diameter) which do not fuse with one another, and can be smooth or covered with small warts (Cárdenas et al. 2009; Cárdenas et al. 2013). In Caminella, actins produce at their tips, small perpendicular bridges (Fig. 4D) that reach towards the other actins, creating a complex fused network, which gives first a honeycomb-like structure (especially when observed with an optical microscope) then a brain-like structure when these bridges thicken. Finally, these surfaces are covered with small warts. Interestingly, all cave specimens (Fig. 4) except the ones from Lebanon have immature sterrasters, which are very similar to immature sterrasters observed in shallow-water Geodiidae in Norway: Pachymatisma normani Sollas, 1888 and Geodia barretti Bowerbank, 1858 (Cárdenas & Rapp 2013). Cárdenas & Rapp (2013) hypothesize that the lower silica concentration in shallow waters is primarily responsible for the immature sterrasters and we may contemplate a similar hypothesis for most cave specimens, which are also amongst the shallowest specimens (6–8 m). Disrupted spiculogenesis is particularly important in the cave specimen from Berlengas, Portugal, where the sterrasters are the smallest (Table 1) and dichotriaenes quite irregular. On the contrary, the Lebanese specimens have larger and fully mature sterrasters (Fig. 5) and significantly thicker spicules (Table 1), which may be linked to higher availability of dissolved silica. Interestingly, the cave where the specimens were collected was densely populated by two species of lithistids (Pérez et al. 2004), which may support this hypothesis, since lithistids have higher needs of silica than other demosponges. Spicule variations between Mediterranean caves have already been observed for the lithistid tetractinellid Discodermia polymorpha (Pisera & Vacelet 2011) but in none of those cases did it seem clear that this species was lacking silica. Shallow-water C. intuta living outside caves and slightly deeper (25–40 m) still have immature sterrasters. Only deeper specimens (>100 m) (Fig. 6), living in waters where silica availability is usually higher, share the same mature sterrasters as the cave specimens from Lebanon. This is in accordance with other Geodiidae, where mature sterrasters are usually found below 40 m (Cárdenas & Rapp 2013). The cortical spheraster morphology also seems to be affected by environmental parameters, possibly silica availability: in cave specimens with immature sterrasters, spherasters are irregular, with small centers, and therefore few spherules; deep-sea specimens have more regular spherasters with larger centers (thus hiding the actins and making spherules); Lebanese specimens reach the extreme of having only spherules, probably indicative of high silica availability. To conclude, ectocortical microsclere morphology seems to be influenced by silica availability, such as in P. normani, but unlike in G. barretti (Cárdenas & Rapp 2013). We wonder whether this means that microscleres in Erylinae (C. intuta, P. normani) and Geodinae have different origins and/or spiculogenesis mechanisms. Like G. barretti and P. normani, C. intuta may be a deep-sea species that manages to survive in shallow-waters, despite lacking silica most of the time. Actually, C. intuta, like several other Mediterranean tetractinellids (e.g. Penares helleri (Schmidt, 1864), Penares euastrum (Schmidt, 1868), Geodia cydonium (L., 1767), Caminus vulcani (Schmidt, 1862), Calthropella pathologica (Schmidt, 1868), Thrombus abyssi (Carter, 1873), Alectona millari Carter, 1879, Discodermia polymorpha Pisera & Vacelet, 2011, Neoschrammeniella bowerbanki (Johnson, 1863) and Neophrissospongia nolitangere (Schmidt, 1870)) (Pisera & Vacelet 2011; Pouliquen 1972) are typically found from marine caves to the deep sea. Currently, we do not have enough data from either population of C. intuta to assess the connectivity between shallow and deep populations. However, Vosmaer (1933) did note differences between his deeper specimens and the shallower specimens of Topsent and Lendenfeld, especially the absence of raised openings and surrounding dark rings around them. Our morphological observations confirm there are differences in external morphology: the deep specimens (from the Alboran Sea and Gulf of Naples) share the same dull color, the absence of a brown ring around the openings, and openings without walls and very small pores (20– 50 µm). The significantly different 28S from one of the Cosquer Cave specimen (5 bp difference) is rather surprising since the Lebanese and Portuguese specimens have identical 28S. These underwater caves are only 20,000 years old (before, they were above water) so we doubt it could be due to a speciation event. Either i) the population that colonized this cave originated from a very different population than the ones present in Portugal and Lebanon (maybe from the deep sea) or ii) we sequenced a variant of 28S in this species, since intragenomic 28S

186 · Zootaxa 4466 (1) © 2018 Magnolia Press CÁRDENAS ET AL. polymorphism has been observed in some demosponges (Plotkin et al. 2017). Unfortunately, we did not manage to sequence COI from the Cosquer Cave specimens, but the morphology clearly suggests it is C. intuta. More cave and deep-sea specimens need to be examined and sequenced, especially for a population genetics study that should unveil the genetic ties between the marine cave, mesophotic zone and deep populations. We noticed that the surface of the sterrasters of the ‘Isops intuta’ specimen from the Balgim expedition (Boury-Esnault et al., 1994, fig. 112e–f) was not typical of Caminella. The two tiny Balgim specimens (3 mm in diameter), dredged off Morocco at 1510 m, clearly have sterrasters with rosettes typical of Geodia species. Furthermore, we examined three slides (2 spicule preparations and 1 piece of cortex) from Balgim specimen CP63– 185. This specimen has i) slightly asymmetrical oxeas (versus symmetrical in C. intuta), ii) shorter proto+deuteroclades, iii) presence of several anatriaenes and one protriaene (overlooked by the authors), but iv) we could not find the spherasters (although they are claimed to be present), instead we found v) smaller oxyasters (8– 12 versus 10–40 in C. intuta) with a different morphology than in C. intuta (Boury-Esnault et al., 1994, fig. 112d), the oxyasters are more spiny, especially at the tip of the actins. To conclude, we are sure that the Balgim specimens have been misidentified and are a Geodia sp., probably a juvenile, which might explain the rarity of sterrasters in the cortex.

Caminella caboverdensis sp. nov. (Figure 7, Table 1)

Holotype. RMNH 3810, Cape Verde, NW of São Vincente (16.9167, -25.0333), 75 m, CANCAP VI expedition (on board HMS Tydeman), station 6.174, 22.06.1982, bottom sand, collecting gear: 1.2 m Agassiz trawl, originally identified as Isops intuta by R. van Soest (unpublished). External morphology (Fig. 7D). Similar to C. intuta. Two dark brown mottled pieces (in ethanol); it is not clear if they belong to the same specimen. Very thin cortex (150 µm). Consistency fleshy, compressible. Spicules. (Fig. 7A–C, Table 1, Supp. Mat. Appendix 1). (a) oxeas, 800–1525 x 8–23 µm, sometimes bent to double bent (‘wavy’); (b) dichotriaenes (rhabdome: 650–900 x 20–50 µm; protoclad: 80–150 µm, deuteroclad: 50– 230 µm); (c) spherical, immature and mature sterrasters, 35–45 µm; (d) oxyasters, 8–42 µm in diameter, 4–9 actins, actins are finely acanthose; center more or less developed; (e) spherasters, 2.5–8 µm in diameter, spiny, regular with large centrum, occasionally look like spherules. Bathymetric range. 75 m. DNA barcoding. COI. The holotype (MH477614) had a difference of 9 bp with the COI of C. intuta, and 4 bp with C. pustula sp. nov. 28S (C1-C2). The holotype (MH478116) has the same 3 bp difference with the cave specimen from Portugal and with C. pustula sp. nov. Submitted to the Sponge Barcoding Project with accession number 1778. Etymology. Named after its type locality, the Cape Verde Islands (‘Cabo Verde’ in Portuguese). Remarks. Although we only have one specimen, with an external and spicule morphology that closely resembles that of C. intuta, we are convinced it belongs to a new species, based on the important genetic difference found in COI (9 bp in Folmer fragment, 658 bp) and 28S (3 bp in C1-C2, 369 bp). We also noticed two spicule differences: 1) sterrasters are smaller than in C. intuta (35–45 µm versus 40–84 µm); 2) oxeas can be much more bent than in C. intuta. These genetic and morphological differences need to be confirmed with additional material. The fact that this specimen has a substantial number of immature sterrasters and few spherules suggests that it might have been living and collected in a silica-limited environment.

Caminella pustula sp. nov. (Figures 8–9, Table 1)

Holotype. MNHN-IP-2008-4, Seamount 2 expedition, St. DW184, Banc d’Hyères, 31°24’N, 28°52’W, 705 m, 16.01.1993, coll. Gofas, Métivier & Warén, barrel 2-1. Paratypes. MNHN-IP-2008-4 (four other specimens); MNHN-IP-2008-8, Seamount 2, St. CP151, Grand Banc Meteor, 30°12’N, 28°25’W, 585 m, 11.01.1993, coll. Gofas, Métivier & Warén, barrel 2-1; MNHN-IP-2008- 148 (2 specimens), Seamount 2, St. DW265, Banc Atlantis, 34°29’N, 30°36’W, 545 m, 03. 02. 1993, coll. Gofas,

CAMINELLA IN THE ATLANTO-MEDITERRANEAN REGION Zootaxa 4466 (1) © 2018 Magnolia Press · 187 FIGURE 7. Caminella caboverdensis sp. nov., RMNH 3810, holotype. A. Mature sterraster. B. Oxyaster and Spherasters. C. Mature sterrasters and spherasters. D. Holotype (2 fragments).

188 · Zootaxa 4466 (1) © 2018 Magnolia Press CÁRDENAS ET AL. Métivier & Warén, barrel 2-9; MNHN-IP-2008-149, Seamount 2, St. DW248, Banc Plato, 33°14’N, 29°32’W, 735 m, 01.02.1993, coll. Gofas, Métivier & Warén, barrel 2-9. Other material. MNCN 1.01/1017, El Cachucho (=Le Danois Bank), 44º02.6999' N, 05º 06.3436' W, 660 m, ethanol 96%, SponGES 0617 expedition, station DR9, field#DR9-430, rock dredge, 16.06.2017, coll. P. Rios.

FIGURE 8. Caminella pustula sp. nov. A. MNHN-IP-2008-4, the holotype is designated with a black arrow. B. MNHN-IP- 2008-148 (tag is 7 cm long). C. Close-up of holotype showing the oscules (right) and smaller pores (bottom left). D. Right specimen in B, MNHN-IP-2008-148, specimen measured in Table 1. E. Microscleres from holotype MNHN-IP-2008-4 (optical microscope) with sterraster, oxyasters and spherules. F. Focus on the sterraster from image E to show the complex sterraster surface.

CAMINELLA IN THE ATLANTO-MEDITERRANEAN REGION Zootaxa 4466 (1) © 2018 Magnolia Press · 189 External morphology (Fig. 8). Holotype is an elongated, globular sponge, 3.5 cm long (Fig. 8A, shown with arrow, 8C). MNHN-IP-2008-8 has a more irregular shape. In ethanol, surface color is cream to brown; choanosome is cream-colored. The specimen from El Cachucho is 0.7 x 0.6 cm and cream-colored alive (slightly browner in ethanol); it was growing on a Pachastrella sp. Cortex is 1–0.5 mm thick. Specimens are slightly compressible. Uniporal oscules are 0.5 mm wide, elevated up to 1 mm, resembling pimples, with a dark brown ring. Uniporal pores sometimes with a dark ring as well, can also be elevated, but less than oscules. Most specimens are growing on coral branches. Spicules. (Fig. 9, Table 1, Supp. Mat. Appendix 1). (a) oxeas, >2600 x 20–40 µm; (b) dichotriaenes (rhabdome: 581–1100 x 43–75 µm; protoclad: 70–296 µm; deuteroclad: 108–790 µm); (c) elongated mature sterrasters, 70–107 x 53–92 µm; (d) oxyasters, 27–78 µm in diameter, 2–8 actins, the actins are finely acanthose; center usually well developed; (e) spiny spherules, 4–13 µm in diameter. Bathymetric range. 545–735 m. DNA barcoding. COI. MNCN 1.01/1017 (MH477615). There is a 6 bp difference with C. intuta and a 4 bp difference with C. caboverdensis. 28S (C1-D2). MNCN 1.01/1017 (MH478117). There is a 17 bp difference with C. intuta from Portugal and 3 bp difference with the shorter 28S (C1-C2) of C. caboverdensis. Submitted to the Sponge Barcoding Project with accession number 1780. Etymology. Named for the external surface, which is covered with ‘pimples’, pustula in Latin. Remarks. The key/diagnostic morphological character that identifies this species, is the elevated, pimple- shaped opening (either oscule or pore) with a brown tip. The spicules are overall much larger than in C. intuta and C. caboverdensis sp. nov. (Table 1, Supp. Mat. Appendix 1): i) the sterrasters are elongate (versus more or less spherical in the other two species) and much larger (70–107 µm in length versus 40–84 µm in C. intuta), ii) the oxyasters are 27–78 µm versus 5–27 µm in C. intuta and 8–42 µm in C. caboverdensis; they are sometimes reduced to only 2–4 actins versus usually >6 actins in C. intuta and C. caboverdensis sp. nov, iii) the dichotriaene rhabdomes are thicker (43–77 µm versus 17–60 µm for C. intuta and 20–50 µm in C. caboverdensis and iv) the dichotriaenes have longer clades (especially the deuteroclades). Finally, C. pustula sp. nov. lives at greater depths than the other two species: 545–735 m versus 2–300 m for C. intuta, 75 m for C. caboverdensis.

Discussion on the phylogenetic position of Caminella

Both the 28S (C1-D2) and the COI (Folmer) markers (Cárdenas et al. 2011) suggest that Caminella has evolutionary affinities with the Erylinae (Erylus, Penares, Pachymatisma, Caminus, Melophlus), and Calthropella; this is however poorly supported. Our RAxML phylogenetic analyses, on an updated COI dataset (Fig. 10), confirmed the topology found in Cárdenas et al. (2011, Figure S2). Note that the Pachastrellidae+Corallistidae clade was sister-group to the Erylinae, but this is not supported (bootstrap < 50). In MrBayes analyses of the same COI dataset Caminella and Calthropella appeared in paraphyly, with respect to the Erylinae, again with no support (post. prob. of 0.5) for the nodes in question. This underlines even more the unsupported grouping of Caminella and Calthropella. In the present study we also sequenced for the first time a partial 18S for Caminella (1690 bp). Our phylogenetic 18S analyses showed that Caminella does not group with the two Erylinae from the dataset (Penares sp. and Pachymatisma johnstonia) (data not shown). However, 18S trees are not informative in this case since there are only two Erylinae and no Calthropellidae 18S sequences. A previous 28S analysis also revealed that the lithistid Macandrewia rigida Sollas, 1888 may have phylogenetic ties with Caminella, the Erylinae and Calthropella, still with low support (Schuster et al. 2015). We noted however that C. intuta grouped with a Geodia sp. (Genbank# KC902389, UF Porifera 1673) from Moorea, French Polynesia (bootstrap of 86 and post-prob. of 0.99) sequenced by the PorToL project (Redmond et al. 2013): there is only a 2 bp difference between the two 18S sequences. Thick sections of this brown Geodia sp. (P. Cárdenas, personal collection) collected at a steep reef drop-off (7–20 m) were examined: the specimen had a very thin cortex (0.1 mm), uniporal oscules/pores, irregular triaenes (with drooping clads), cortical sphero-tylasters and very small sterrasters (20–22 µm in diameter), which all seemed immature. The close genetic distance with C. intuta along with these morphological characters convinced us that this was a Pacific species of Caminella sp. Therefore, it is clear that spherules are not a diagnostic character for Caminella since this Pacific species did not have any.

190 · Zootaxa 4466 (1) © 2018 Magnolia Press CÁRDENAS ET AL. FIGURE 9. Caminella pustula sp. nov., MNHN-IP-2008-4, holotype. A. Sterrasters. B. Young sterraster. C. Surface of mature sterraster. D. Oxyasters. E. Spherules.

CAMINELLA IN THE ATLANTO-MEDITERRANEAN REGION Zootaxa 4466 (1) © 2018 Magnolia Press · 191 FIGURE 10. RAxML phylogenetic tree based on the COI nucleotide dataset, under the generalized time-reversible Gamma— GTRGAMMA—model. Bootstrap values >50 are given at the nodes. GenBank accession numbers are given after each taxon name (we have only shown the part of the tree discussed in the article).

From a morphological standpoint, Caminella shares with the Erylinae and Calthropella several characters. With respect to the Macandrewiidae, no obvious shared morphological characters can be found with Caminella. Caminella shares with Erylus/Penares and Calthropella uniporal oscules/pores. It also shares the presence of an oscule leading to a cloaca with some Erylinae (e.g. Caminus vulcani, Penares euastrum, Penares helleri, Erylus discophorus complex, Erylus formosus) but not all. Lendenfeld’s (1903, p. 84) key to the Geodiidae families implies that he considered Caminella relatively close to Erylus species, based on 1) shared larger oscules, often with 2) a collar. There is some truth in this, as oscules in some Erylus/Penares are larger than in Geodia, but this is not always the case: it depends whether the oscule is an opening with a sphincter (as in Geodia) or an opening with a cloaca (often with a collar) with no individual sphincter (as in Caminella); each type is characteristic of different Erylus/Penares clades (Cárdenas et al. 2010). The presence of a collar is difficult to assess since it is more easily observed on live specimens. Be that as it may, our current observations suggest it is mostly present in Erylus species with a cloaca. So it appears Lendenfeld’s two characteristics stand only for some Erylus/Penares clades, not all Erylinae. More importantly, like in all Erylinae and Calthropellidae, Caminella lacks pro/anatriaenes. Also, in the Erylinae and Caminella, the triaenes and oxeas do not cross the cortex, which makes the cortex more easily detachable from the choanosome. The spherules found in north Atlantic Caminella remind us particularly of the spherules in Caminus, although in Caminus these spherules are always perfectly round and never appear as spherasters. Moreover, the Pacific Caminella mentioned earlier does not have spherules so it is not a diagnostic

192 · Zootaxa 4466 (1) © 2018 Magnolia Press CÁRDENAS ET AL. character of the genus. The sterrasters are not flattened like in most Erylus, but rather globular as in Caminus and Pachymatisma. On the other hand, Caminella has oxeas like Erylus and not strongyles like in most Caminus; Pachymatisma can have either oxeas or strongyles. The Caminella spherasters/spherules could also be related to the spherasters in Calthropella. The lumpy spherasters in Calthropella durissima (van Soest et al., 2010, Fig. 25) especially present similarities with spherules. We should mention here the New Zealand species Caminus primus Sim-Smith & Kelly, 2015 which shares most of the morphological characters of Caminella (Sim-Smith & Kelly 2015). However, several of these characters like “clusters of small, well-separated, raised oscules (0.6 mm)” are usually absent in Caminus, but very typical in all Caminella from this study. Furthermore, the sterraster surface of C. primus is very similar to what we observe in Caminella spp., and we have never seen this surface type in any other Caminus species (P. Cárdenas, unpublished data). This leads us to believe that, instead of showing a phylogenetic tie with the genus Caminus, this species should be reallocated to Caminella as Caminella prima comb. nov. To conclude, although Caminella shares some characters with Calthropella, there is at this point no synapomorphy or convincing evidence to support a sister-group relationship between Calthropella and Caminella. Molecular phylogenetic analyses also suggest that this relationship is not to be trusted, with the limited dataset at hand. There are more morphological characters that confirm affinities of Caminella with Erylinae taxa but none that could clearly tie Caminella to a particular group. We propose to temporarily place this singular genus in the Erylinae for now, until additional molecular and morphological evidence is gathered (especially from other Caminella species) to clarify its phylogenetic affinities.

Author contributions

Fieldwork: Jean Vacelet, Pierre Chevaldonné, Thierry Pérez, Joana R. Xavier Molecular work: Paco Cárdenas Morphology work: Paco Cárdenas, Jean Vacelet, Joana R. Xavier Phylogenetic analyses: Paco Cárdenas Drafting of manuscript: Paco Cárdenas Revision of the manuscript: All authors

Acknowledgments

Many thanks go to Isabelle Domart Coulon (MNHN) for making the Seamount2 collection available to P. Cárdenas and J. R. Xavier in 2010-2011 (support provided by the MNHN through an invited researcher grant to JRX and the EU SYNTHESYS program, travel grant no. FR-TAF-450), to Rob van Soest (Naturalis, Leiden) for welcoming P. Cárdenas in the collections in Amsterdam in March 2009 (support from the EU SYNTHESYS program) and to Lauren Hughes (NHM London) for welcoming P. Cárdenas in the NHM collections in December 2017 (support from an Inez Johannson grant, HT14, Uppsala University). We would also like to further acknowledge Rob van Soest for finding the type material of Isops maculosus. We are grateful to Oliver Macek (Naturhistorisches Museum, Vienna) for taking pictures of the slides of the holotype of Caminella loricata. We thank David Rees and Tone Ulvatn (University of Bergen) for their technical support. Thanks also to Pilar Ríos and Javier Cristobo (IEO, Gijón, Spain) for collecting and sharing a sample from expedition SponGES 0617; and thank you to Cèlia Sitjà and Manuel Maldonado (Center for Advanced Studies of Blanes, Spain) for sharing a specimen from the Alboran Sea. We are grateful to Ghazi Bitar, (Lebanese University, Department of Natural Sciences, Hadath, Lebanon) for the discovery and exploration of “lithistid cave” in Lebanon. Thanks are also due to Luc Vanrell, François Grosjean, Michel Olive, Xavier Delestre and Delphine Lecouvreur for the access to the Cosquer cave. P. Cárdenas and J. R. Xavier received support from the European Union's Horizon 2020 research and innovation program through the SponGES project (grant agreement No. 679849). This document reflects only the authors’ view and the Executive Agency for Small and Medium-sized Enterprises (EASME) is not responsible for any use that may be made of the information it contains.

Boury-Esnault, N. (1971) Spongiaires de la zone rocheuse de Banyuls-sur-Mer. II. Systématique. Vie et Milieu, XXII, 287–350. Boury-Esnault, N., Pansini, M. & Uriz, M.J. (1994) Spongiaires bathyaux de la mer d’Alboran et du golfe ibéro-marocain. Mémoires du Muséum national d’Histoire naturelle, 160, 1–174. Bowerbank, J.S. (1858) On the Anatomy and Physiology of the Spongiadae. Part I. On the Spicula. Philosophical Transactions of the Royal Society of London, 148, 279–332, pls XXII–XXVI. Cárdenas, P., Menegola, C., Rapp, H.T. & Díaz, M.C. (2009) Morphological description and DNA barcodes of shallow-water Tetractinellida (Porifera: Demospongiae) from Bocas del Toro, Panama, with description of a new species. Zootaxa, 2276, 1–39. Cárdenas, P. & Moore, J.A. (2018) First records of Geodia demosponges from the New England Seamounts, an opportunity to test the use of DNA mini-barcodes on museum specimens. Marine Biodiversity, First Online. 10.1007/s12526-017-0775-3 Cárdenas, P. & Rapp, H.T. (2013) Disrupted spiculogenesis in deep-water Geodiidae (Porifera, Demospongiae) growing in shallow waters. Invertebrate Biology, 132, 173–194. https://doi.org/10.1111/ivb.12027 Cárdenas, P., Rapp, H.T., Klitgaard, A.B., Best, M., Thollesson, M. & Tendal, O.S. (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, 251–311. 10.1111/zoj.12056 Cárdenas, P., Rapp, H.T., Schander, C. & Tendal, O.S. (2010) Molecular taxonomy and phylogeny of the Geodiidae (Porifera, Demospongiae, Astrophorida)—combining phylogenetic and Linnaean classification. Zoologica Scripta, 39, 89–106. 10.1111/j.1463-6409.2009.00402.x Cárdenas, P., Xavier, J. R., Reveillaud, J., Schander, C. & Rapp, H.T. (2011) Molecular phylogeny of the Astrophorida (Porifera, Demospongiae) reveals an unexpected high level of spicule homoplasy. PLoS ONE, 6, e18318. https://doi.org/10.1371/journal.pone.0018318 Carter, H.J. (1873) On two new Species of Gummineae, with special and general Observations. Annals and Magazine of Natural History, Series 4, 12 (67), 17–30, pl. I. Carter, H.J. (1875) Notes Introductory to the Study and Classification of the Spongida. Part II. Proposed Classification of the Spongida. Annals and Magazine of Natural History, Series 4, 16 (92), 126–145, 177–200. Carter, H.J. (1879) On a New Species of Excavating Sponge (Alectona Millari); and on a New Species of Rhaphidotheca (R. affinis). Journal of the Royal Microscopical Society, 2, 493–499, pls. XVII–XVIIa. Chombard, C., Boury-Esnault, N. & Tillier, S. (1998) Reassessment of homology of morphological characters in Tetractinellid sponges based on molecular data. Systematic Biology, 47, 351–366. https://doi.org/10.1080/106351598260761 Folmer, O., Black, M., Hoeh, W., Lutz, R. & Vrijenhoek, R. (1994) DNA primers for amplification of mitochondrial cytochrome c oxidase subunit I from diverse metazoan invertebrates. Molecular Marine Biology and Biotechnology, 3, 294–299. Gray, J.E. (1867) Notes on the arrangement of sponges, with the descriptions of some new genera. Proceedings of the Zoological Society of London, 2, 492–558, pls. XXVII–XXVIII. Johnson, J. Y. (1863) Description of a new siliceous sponge from the coast of Madeira. Proceedings Zoological Society London, 3, 257–259. Johnston, G. (1842) A History of British Sponges and Litophytes. W.H. Lizars, Edinburgh, XII + 264 pp. Kelly, M. & Cárdenas, P. (2016) An unprecedented new genus and family of Tetractinellida (Porifera, Demospongiae) from New Zealand’s Colville Ridge, with a new type of mitochondrial group I intron. Zoological Journal of the Linnean Society, 177, 335–352. https://doi.org/10.1111/zoj.12365 von Lendenfeld, R. (1894) Die Tetractinelliden der Adria. (Mit einem Anhange über die Lithistiden). Denkschriften der Kaiserlichen Akademie der Wissenschaften.Wien. Mathematisch-Naturwissenschaftliche Klasse, 61, 91–204, pls. I–VIII. von Lendenfeld, R. (1903) Porifera. Tetraxonia. In: Schulze, F.E. (Ed.), Das Tierreich. Friedländer, Berlin, pp. vi–xv, 1–168. Linnaeus, C. (1767) Systema naturae sive regna tria naturae, secundum classes, ordines, genera, species, cum characteribus, differentiis, synonymis, locis. 12th Edition. Vol. 1. Part 2. Laurentii Salvii, Holmiae, 1215 pp. [pp. 533–1327] Maldonado, M. (1992) Demosponges of the red coral bottoms from the Alboran Sea. Journal of Natural History, 26, 1131– 1161. https://doi.org/10.1080/00222939200770661 Marshall, W. (1876) Ideen über die Verwandtschaftsverhältnisse der Hexactinelliden. Zeitschrift für wissenschaftliche Zoologie, 27, 113–136. Miller, M. A., Pfeiffer, W. & Schwartz, T. (2010) Creating the CIPRES Science Gateway for inference of large phylogenetic trees. In: Proceedings of the Gateway Computing Environments Workshop (GCE), 14 Nov. 2010, New Orleans, LA, pp. 1– 8. https://doi.org/10.1109/GCE.2010.5676129

194 · Zootaxa 4466 (1) © 2018 Magnolia Press CÁRDENAS ET AL. Pérez, T., Vacelet, J., Bitar, G. & Zibrowius, H. (2004) Two new lithistids (Porifera: Demospongiae) from a shallow eastern Mediterranean cave (Lebanon). Journal of the Marine Biological Association of the United Kingdom, 84, 15–24. https://doi.org/10.1017/S0025315404008859h Pisera, A. & Vacelet, J. (2011) Lithistid sponges from submarine caves in the Mediterranean: taxonomy and affinities. Scientia Marina, 75, 17–40. https://doi.org/10.3989/scsimar.2011.75n1017 Plotkin, A., Voigt, O., Willassen, E. & Rapp, H.T. (2017) Molecular phylogenies challenge the classification of Polymastiidae (Porifera, Demospongiae) based on morphology. Organisms Diversity & Evolution, 17, 45–66. https://doi.org/10.1007/s13127-016-0301-7 Pouliquen, L. (1972) Les spongiaires des grottes sous-marines de la région de Marseille: écologie et systématique. Tethys, 3 (4), 717–758. Redmond, N.E., Morrow, C.C., Thacker, R.W., Diaz, M.C., Boury-Esnault, N., Cárdenas, P., Hajdu, E., Lôbo-Hajdu, G., Picton, B.E., Pomponi, S.A., Kayal, E. & Collins, A.G. (2013) Phylogeny and Systematics of Demospongiae in Light of New Small Subunit Ribosomal DNA (18S) Sequences. Integrative and Comparative Biology, 53, 388–415. https://doi.org/10.1093/icb/ict078 Ronquist, F., Teslenko, M., van der Mark, P., Ayres, D.L., Darling, A., Höhna, S., Larget, B., Liu, L., Suchard, M.A. & Huelsenbeck, J.P. (2012) MrBayes 3.2: Efficient Bayesian Phylogenetic Inference and Model Choice Across a Large Model Space. Systematic Biology, 61, 539–542. https://doi.org/10.1093/sysbio/sys029 Sarà, M. (1961) La fauna di Poriferi delle grotte della isole Tremiti. Studio ecologico e sistematico. Archivio Zoologico Italiano, 46, 1–59. Schmidt, O. (1862) Die Spongien des adriatischen Meeres. Wilhelm Engelmann, Leipzig, viii + 88 pp., pls. 81–87. Schmidt, O. (1864) Supplement der Spongien des adriatischen Meeres. In: Enthaltend die Histologie und systematische Ergänzungen. (Wilhelm Engelmann: Leipzig), pp. i–vi + 1–48, pls. 41–44. Schmidt, O. (1868) Die Spongien der Küste von Algier. Mit Nachträgen zu den Spongien des Adriatischen Meeres (Drittes Supplement). Wilhelm Engelmann, Leipzig, iv + 44 pp., V pls. Schmidt, O. (1870) Grundzüge einer Spongien-Fauna des atlantischen Gebietes. Wilhelm Engelmann, Leipzig, 2 (iii–iv) + 88 pp., VI pls. Schuster, A., Cárdenas, P., Pisera, A., Pomponi, S.A., Kelly, M., Wörheide, G. & Erpenbeck, D. (2018) Seven new deep-water Tetractinellida (Porifera: Demospongiae) from the Galápagos Islands—morphological descriptions and DNA barcodes. Zoological Journal of the Linnaean Society. https://doi.org/10.1093/zoolinnean/zlx110 Schuster, A., Erpenbeck, D., Pisera, A., Hooper, J.N.A., Bryce, M., Fromont, J. & Wörheide, G. (2015) Deceptive desmas: molecular phylogenetics suggests a new classification and uncovers convergent evolution of lithistid demosponges. PLoS ONE, 10 (1), e116038. https://doi.org/10.1371/journal.pone.0116038 Sim-Smith, C. & Kelly, M. (2015) The Marine Fauna of New Zealand: Sponges in the family Geodiidae (Demospongiae: Astrophorina). NIWA Biodiversity Memoir, 128, 1–102. Sitjà, C. & Maldonado, M. (2014) New and rare sponges from the deep shelf of the Alboran Island (Alboran Sea, Western Mediterranean). Zootaxa, 3760 (2), 141–179. https://doi.org/10.11646/zootaxa.3760.2.2 van Soest, R.W.M., Beglinger, E.J. & de Voogd, N.J. (2010) Skeletons in confusion: a review of astrophorid sponges with (dicho-)calthrops as structural megascleres (Porifera: Demospongiae: Astrophorida). Zookeys, 68, 1–88. Sollas, W.J. (1888) Report on the Tetractinellida collected by H.M.S. Challenger, during the years 1873–1876. Report on the Scientific Results of the Voyage of H.M.S. Challenger, 1873–1876. Zoology, 25, 1–458, pls. I–XLIV. Stamatakis, A. (2014) RAxML Version 8: A tool for Phylogenetic Analysis and Post-Analysis of Large Phylogenies. Bioinformatics, 30, 1312–1313. https://doi.org/10.1093/bioinformatics/btu033 Templado, J., Garcia-Carrascosa, M., Baratech, L., Capaccioni, R., Juan, A., Lopez-Ibor, A., Silvestre, R. & Masso, C. (1986) Estudio preliminar de la fauna asociada a los fondos coraliferos del mar de Alboran (SE de Espana). Boletin Instituto Espanol de Oceanografia, 3, 93–104. Thiele, J. (1899) Studien über pazifische Spongien. II. Ueber einige Spongien von Celebes. Zoologica. Original-Abhandlungen aus dem Gesamtgebiete der Zoologie, Stuttgart, 24, 1–33, pls. I–V. Topsent, E. (1892) Diagnoses d’éponges nouvelles de la Méditerranée et plus particulièrement de Banyuls. Archives de Zoologie Expérimentale et Générale, 2 (Notes et Revue 6), xvii–xxviii. Topsent, E. (1893) Nouvelle série de diagnoses d’éponges de Roscoff et de Banyuls. Archives de Zoologie Expérimentale et Générale, Series 3, 1 (Notes et Revue 10), xxxiii–xliii. Topsent, E. (1894) Étude monographique des spongiaires de France I. Tetractinellida. Archives de Zoologie Expérimentale et Générale, 3, 259–400, pls. XI–XVI. Topsent, E. (1895) Étude monographique des spongiaires de France II. Carnosa. Archives de Zoologie Expérimentale et Générale, 3, 493–500, pls. XXI–XXIII.

CAMINELLA IN THE ATLANTO-MEDITERRANEAN REGION Zootaxa 4466 (1) © 2018 Magnolia Press · 195 Vosmaer, G.C.J. (1894) Preliminary notes on some tetractinellids of the Bay of Naples. Tijdschrift der Nederlandsche Dierkundige Vereeniging, 2, 269–286. Vosmaer, G.C.J. (1933) The sponges of the Bay of Naples, Porifera Incalcaria. With analyses of genera and studies in the variations of species. Vol. I. Martinus Nijhoff, The Hague, 456 pp.

Supplementary material. Appendix 1. Morphometric analyses of three spicules types (sterrasters, oxyasters and spherasters/spherules) from the individual samples examined in this study.