Hexactinellida (Porifera) from the ANDEEP III Expedition to the Weddell Sea, Antarctica Janussen, Dorte Reiswig, Henry M. Zootaxa 2009 2136 1 20 3W2PS [151,394,151,177] Hexactinellida Rossellidae Lophocalyx Animalia Lyssacinosida 10 11 Porifera species topsenti sp. nov.  Material examined:  Lophocalyx topsenti  n. sp., Holotype, SMF10606, ANDEEP IIIExped., R.V. 'Polarstern', stn PS67/142-6, Weddell Sea,  Antarctica, 18 Mar. 2005, 62°09.93'S, 49°30.47'Wto 62°09.80'S, 49°30.59'W, 3403–3404 m, ethanol. Paratype: RSM1921.143.1389, Scottish National AntarcticExped., R.V. 'Scotia', stn 313, Weddell Sea,  Antarctica, 18 Mar. 1903, 62°10'S, 41°20'W, 3247 m, ethanol (reported as  Calycosoma validumby Topsent, 1910, 1913). Comparative material:  Calycosoma validum: holotype, USNM0 4761, U.S.F.C.S. 'Albatross', stn 2573, George's Bank, SE of Massachusetts, USA, 0 2 Sep. 1885, 40º34'18"N, 66º09'W, 3186 m, ethanol.   Description:The holotypeof the new species, when inspected in June 2006before partial dissection, was a small, intact, tubular sponge 31 mmtall and 20 mmin diameter, with a terminal osculum 9 mmin diameter, without marginalia, with wall thickness of 2 mm( Fig. 6A). Several prostal diactins project obliquely upwards up to 9 mmfrom the body surface but hypodermal spicules are not raised to form a veil; groups of broken basalia project from irregular lobes of the lower body surface. The lateral surface is generally smooth to the naked eye, but, when magnified, is irregular with low conules; entrances to inhalant canals are evident through the surface lattice. The sponge is soft and pliable; color is olive brown, the same as the bottom “green-mud” substrate. The paratype, when surveyed February – May 1996, consisted of 17 wet fragments, 0.5–3 mmin thickness, the remnants of the original six fragments reported by Topsent (1913), as probably stemming from a single large cup-shaped specimen ( Fig. 6B). Both surfaces of most fragments retain the fine spicule lattice spanning over the apertures of inhalant and exhalant canals as outlined by the choanosomal diactin bundles. Intact and broken diactins projected from one surface of most fragments, but hypodermalia were not found above the lattices of free spicules. All fragments were somewhat stiff and brittle, due to extensive but tenuous fusion of megascleres ( Fig.6D).   TABLE 3.Spicule dimensions of  Lophocalyx topsenti,  n. sp., from the Weddell Sea, Antarctica(dimensions in µm unless otherwise noted).    Holotype SMF 10606 Paratype RSM 1921.143.1389  parameter Anchorate hypodermalia mean st. dev. range n mean st. dev. range n.  tangential ray length 242 38 146–292 25 196 22 149–256 77  tangential ray width proximal ray length (mm) 21.6 5.4 12.8–32.1 25 1.62 0.59 0.86–2.47 18 27.8 3.6 16.8–33.6 30 * * * *  proximal ray width 25.0 4.8 13.9–33.9 25 28.6 4.0 16.8–37.3 56  Regular hypodermalia tangential ray length 592 33 354–1,15 25 6 636 133 457–1,534 122  tangential ray width proximal ray length (mm) 23.2 3.7 17.1–36.9 25 1.24 0.19 0.93–2.05 50 25.5 4.2 14.1–33.6 37 1.32 0.32 0.78–2.67 46  proximal ray width 24.3 3.2 18.2–34.6 25 26.3 3.9 17.9–34.2 27  Dermal pinular hexactin distal pinular ray length 107 24 71–160 25 122 25 66–185 91  ray width 9.6 1.2 6.9–11.8 25 12.3 1.8 9.9–15.8 25  tangential ray length ray width 132 17 100–162 25 7.7 1.5 5.4–11.8 25 143 1.8 99–181 131 11.9 1.9 8.2–16.3 28  proximal ray length 87 15 68–112 25 99 20 34–143 80  ray width Atrial pinular hexactin 8.0 1.3 5.9–10.9 25 11.6 1.8 8.0–16.0 25  distal pinular ray length 396 165 105–687 49 337 170 99–736 50  ray width tangential ray length 13.0 5.2 5.1–24.5 56 401 152 124–759 66 13.6 4.4 8.2–27.3 32 312 151 120–613 71  ray width 14.5 5.5 5.8–28.8 64 14.1 4.6 7.5–24.0 43  proximal ray length ray width 320 261 70–865 50 12.7 4.5 4.8–21.5 55 214 290 57–1,042 50 12.3 4.5 5.9–23.9 47  Large principal hexactin  ray length ray width 709 165 502–911 7 25.9 4.1 21.2–32.0 7 1,098 424 261–2,824 108 33.2 4.2 21.2–41.2 30  Thick diactin length (mm) 9.8 2.5 7.9–18.5 16 15.9 6.1 6.6–33.0 28  width Thin diactin length (mm) 39.5 14.1 24.8–76.4 28 3.1 1.1 1.2–5.5 27 75 33 23–140 134 4.9 2.2 1.7–9.4 30  width 12.6 3.7 6.5–20.9 25 13.8 3.8 6.7–24.0 51  Oxyhexaster diameter primary ray length 170 12 152–197 25 5.3 0.9 3.3–6.7 25 151 18 64–193 100 6.4 1.2 3.9–10.0 32  secondary ray length 80.0 5.8 70.1–94.9 25 71.5 6.7 55.0–84.6 33  Plumicome diameter primary ray length 71.6 7.6 51.3–82.7 25 13.2 1.5 10.2–15.8 25 65.3 8.5 43.2–82.8 110 14.0 1.1 11.2–18.2 101  secondary ray length 27.5 3.8 16.7–32.0 25 23.8 4.1 13.0–33.6 101 * no intact spicules of this typefound Megascleres (dimensions given in Table 3) consist of hypodermal anchorate pentactins, hypodermal regular pentactins, pinular hexactine dermalia, pinular hexactine atrialia, large choanosomal hexactins and thick choanosomal diactins as principalia (and as lateral prostalia for the diactins), and thin choanosomal diactins. Anchorate hypodermalia ( Figs. 6C, 7A) are generally smooth with the crucially-arrayed tangential rays bent abruptly back 70–90° near their mid-points; ray tips are parabolic-rounded or bullet-shape and only the proximal ray tip is sometimes roughened near its end. Regular hypodermal pentactins ( Fig. 7B) have straight, crucially-arrayed, tangential rays, smooth except for roughening at ray ends; neither these nor the anchorate forms are erected above the lateral surfaces. Occasional hypodermalia ( Fig. 7B) are mixtures of the regular and anchorate types, with some straight rays and some shorter bent rays. Dermalia ( Fig. 7C) are almost exclusively pinular rough hexactins with the distal ray differing from the others in thickness and density and size of its proclined spines; the pinular ray is generally shorter than the tangentials but longer than the proximal ray. The pinular ray tip is cylindric with a blunt tip while the other rays are tapered and end in sharply pointed tips. Atrialia ( Fig. 7D) are rough pinular hexactins of much larger size range and slighter spination than the dermalia; in smaller forms the distal pinular ray (extending into atrial cavity) is the longest ray; in larger spicules of this typethe proximal ray may be increased disproportionately and become larger than all other rays, while spination of the entire spicule is reduced or entirely lost. Large choanosomal hexactins ( Fig. 7E) with straight rays sometimes have one very short and distally heavily spined ray. Thick diactins ( Fig. 7F) serve as primary choanosomal and prostal spicules; they are slightly curved, and smooth except for fine spination just before the smooth rounded or bullet-shape spicule tips; they have no swelling at the spicule center. Occasional orthodiactins or L-form thick diactins occur. Thin choanosomal diactins ( Fig. 7G) occur in bundles; they are mostly smooth, with rounded to parabolic and occasionally inflated roughened tips and usually a slight swelling at the spicular center. Microscleres (dimensions given in Table 3) consist of oxyhexasters, hemioxyhexasters, rare oxyhexactins, and strobiloplumicomes. Oxyhexasters and hemioxyhexasters ( Figs. 6E, 7H) are the most abundant microscleres; their short primary rays each bear 1–3 thin straight terminals; the thick primary rays are smooth while the terminal rays are covered with very fine recurved thorns detectable in LM; rare oxyhexactins of the same size occur among this spicule category ( Figs. 6F, 7I). Strobiloplumicomes ( Figs. 6G–I, 7J) are very common only in dermal and atrial surface tissues; primary rays bear a subterminal spherical inflation from which originate 3–5 whorls of sickle-shaped terminal rays of a series of different lengths, each of which is ornamented with two rows of spines on their distal concave (inner) surface; the pegs of the primary rays extending beyond the swelling are smooth and terminate in rounded tips.   Etymology:The species is named in honor of the eminent spongologist, Professor Emile Topsent, who originally described the paratype.   Remarks: Topsent (1910, 1913) was very clear in his original description of the 'Scotia' specimen, that it differed from Schulze's  Calycosoma validumin several aspects, the most notable being the presence of anchorate hypodermalia (absent in  C. validum) and the absence of pentactine dermalia (abundant in  C. validum). He nonetheless attributed these features to malformations and individualistic variation, and assigned the Antarcticspecimen to  C. validum. Barthel and Tendal (1994)accepted these obvious discrepancies and rediagnosed  Calycosomaand  C. validumto include features of both the N Atlantic and Antarcticspecimens. Tabachnick (2002b)was the first reviewer to act in moving Topsent's Antarcticspecimen from  Calycosomato  Lophocalyx, but suggested no species name. Menshenina et al. (2007), in their re-diagnoses of  Calycosomaand  Lophocalyx, noted once again the clear differences between Schulze's description of  C. validumand Topsent's description of the Antarcticspecimen and again stated that Topsent's specimen should be transferred to  Lophocalyxpending re-analysis of the  C. validum typespecimen. We have here examined both Schulze's typespecimen of  C. validumand Topsent's original Antarcticspecimen, as well as a new specimen of  Lophocalyxfrom the Weddell Sea. We verify that the differences between  C. validumand Topsent's specimen are as repeatedly noted. There are no material differences between the new ANDEEP III specimen and Topsent's original Antarcticspecimen, except for megasclere fusion which is absent in the former and extensive, but tenuous, in the latter. We chose to designate the new specimen as holotypeof the new species because it was complete and intact, although the smaller of the two, and designate Topsent's specimen, although the larger specimen, as paratypebecause it was fragmentary and severely damaged. No single feature distinguishes the new species,  L. topsenti,from the other members of the genus. Its pinular dermal hexactins separate it from the group with all or most dermalia as stauractins,  L. philippinensis(Gray, 1872),  L. spinosaSchulze, 1900, and  L. sululanusIjima, 1927, and from  L. moscaleviaTabachnick, 1988, which has partly dermal pentactins. It differs from  L. biogassiand  L. oregoni, both of Menshenina et al., 2007, in having no pentactin atrialia. It differs from  L. pseudovalida Menshenina et al., 2007in the much larger size of its choanosomal hexactins, 0.26–2.84 mmversus 0.14–0.53 mmin the latter. It differs from  L. atlantiensisand  L. brasiliensis, both Menshenina et al., 2007, in form of its main oxyhexaster microscleres and the strongly reflected tangential rays of the anchorate hypodermalia which are only slightly or gradually recurved in those species. Finally, it clearly differs from  L. profundum, described above, in having hexactine rather than diactine dermalia.