FAU Institutional Repository

http://purl.fcla.edu/fau/fauir

This paper was submitted by the faculty of FAU’s Harbor Branch Oceanographic Institute.

Notice: © 1995 Queensland Museum. This manuscript is an author version with the final publication available and may be cited as: Kelly-Borges, M., & Vacelet, J. (1995). A revision of Diacarnus Burton and Negombata de Laubenfels (Demonspongiae: Latrunculiidae) with descriptions of new species from the west-central Pacific and Red Sea. Memoirs of the Queensland Museum, 38(2), 477-503.

A REVISION OF D/ACARNUSBURTON AND NEGOMBATA DE LAUBENFELS (DEMOSPONGIAE: LA TRUNCULIIDAE) WITH DESCRIPTIONS OF NEW SPECIES FROM THE WEST CENTRAL PACIFIC AND THE RED SEA

MICHELLE KELLY-BORGES AND JEAN V ACELET

Kelly-Borges, M. & Vacelet, J. 1995 12 01: A revision of Diacamus Burton and Negombata de Laubenfels (Demospongiae: Latrunculiidae) with descriptions of new species from the west central Pacific and the Red Sea. Memoirs of the Queensland Museum 38(2):477-503. Brisbane. ISSN 0079-8835.

Species previously assigned to the genus Latrunculia, which possess discate microrhabds as the microsclere complement, have been re-evaluated and assigned to Diacamus Burton and Negombata de Laubenfels, which both contain spinulate rather than discate microrhabd microscleres. The type species of Diacamus, Axos spinipoculum Carter, is redescribed, and seven new species are described: D. beilae, D. erythraeanus, D. levii, D. ardoukobae, D. bismarck.ensis, D. tubifera, and D. megaspinorhabdosa, spp. nov. The type species of Negombata, Latrunculia corticata (Carter), is redescribed and the only other known species, Latrunculia magnifica Keller, is transferred to Negombata. Diagnostic morphological char­ acters which emphasize a combination of gross morphology, spicule dimensions, microsclere morphology and disposition, are identified for Diacamus and Negombata and the affinities of these and other latrunculiids, Latrunculia and Sigmosceptrella, are compared to each other and to other demosponges. New locality and species records reveal a remarkably disjunct biogeographic distribution for Diacamus: the greatest diversity of species is found in northern Papua New Guinea and in several Micronesian atolls, but the genus extends east through New Caledonia to Fiji, and south to Port Jackson in south-eastern Australia. Two species of Diacamus are also found in the Red Sea. The genus has not been recorded in the published literature, nor in the extensive unpublished collections known for the Indo-Malay region. 0 Demospongiae, Latrunculiidae, Micronesia, Red Sea.

Michelle Kelly-Borges, Department of Zoology, The Natural History Museum, Cromwell Road, London, SW7 5BD, United Kingdom, and Coral Reef Research Foundation, P. 0. Box 70, Weno, Chuuk Atoll, Federated States of Micronesia 96942; Jean Vacelet, Centre d'Ocianologie de Marseille, Station Marine d'Endoume, Rue de Ia Batterie des Uons, Fl3007 Marseille, France; 10 June 1995.

However, it is unlikely that acan­ Sponges having discorhabds, dicasters or thomicrorhabds, which are a distinctive feature of spinulate acanthomicrorhabds as microscleres the family as presently recognised, are homolo­ have traditionally been placed within gous in all of these genera. The genus Latrunculia Bocage, Latrunculiidae Topsent. U:ltrunculia, which contains discorhabds, has re­ Distinctive microscleres of Latrunculia bear four ceived several species which are certainly atypi­ whorls of spines frequently united in a disc, on a cal. Two species from the Red Sea. Latrunculia straight axial rod, which are disposed in an erect corticata Carter, 1879 (with its claimed junior layer above the sponge ectosome. U:ltrunculia synonym L magnifica Keller. 1889), and L contains massive sponges with papillae and purpurea Carter. 1881, were isolated in a new sieve-pore areas, with skeletons of monactinal or genus Negombata by de Laubenfels (1936), a rarely diactinal spicules arranged in an irregular distinction which had been suggested long ago by reticulation of poorly defined, somewhat plu­ Ridley and Dendy (1887) and Thiele (1900). De mose, tracts. Other latrunculiids such as Laubenfels differentiated these specimens from Podospongia Bocage, Sigmosceptrella Dendy, U:ltrunculia on the morphology of the and Barbozia Dendy, contain acan­ microscleres which are spinulate sanidaster-like thomicrorhabds. The affinities of these genera acanthomicrorhabds. rather than discate as in have been diversely interpreted, resulting in allo­ U:ltrunculia. The position of Axos spinipoculum cation of the family either in the tetractinomorph (Carter, 1879), transferred to Latrunculia by Hadromerida, following Bergquist ( 1978), or in Hooper (1986), is doubtful. Burton (1934) pro­ the ceractinomorph Poecilosclerida following posed the new genus Diacamus to receive Axos Topsent (1922). spinipoculum Carter in recognition of the charac- 478 MEMOIRS OF THE QUEENSLAND MUSEUM

teristic acanthomicrorhabds which consisted of a collection program contracted to the CRRF. A straight shaft bearing four whorls of spines. complete collection of all OCDN specimens is We have found several sponges in the West located at the Smithsonian Institution (United Central Pacific and in the Red Sea which are States National Museum); Q66C, Specimen sam­ characterised by the possession of spinulate ple numbers for United States National Cancer rhabds, hereafter termed spinorhabds. These Institute shallow-water collection program pre­ microscleres are thin rods with spines or bumps viously contracted to the Australian Institute of scattered irregularly or uniformly in 4 whorls Marine Science, Townsville. This latter collec­ along the shaft. The microscleres of these new tion is now located at the Queensland Museum. sponges are clearly related to those of Axos spinipoculum Carter, Latrunculia ( =Negombata) SYSTEMATICS corticata Carter, and L ( =Negombata) magnifica Keller. These species are re-evaluated in the light Family LATRUNCULIIDAE Topsent of this new material. Diacarnus Burton, 1934 METHODS Diacamus Burton, 1934: 549 Sponges were collected using SCUBA by the authors and the Coral Reef Research Foundation TYPE SPECIES (CRRF), Micronesia, from Chuuk (Truk) State, A.xos spinipoculum Carter, 1879: 286 Micronesia, the Australian Great Barrier Reef, northern and southern Papua New Guinea, New DIAGNOSIS (emended) Caledonia, Fiji, the Philippines, the Red Sea, and Massive spherical or barrel-shaped, lobate-dig­ the Gulf of Aden (Fig. 4). On collection, samples itate, repent-branching, tubular or thickly en­ were preserved in 70% ethanol or in 10% forma­ crusting Latrunculiidae. Surface with single or lin and prepared for histological examination as multiple conules or blunt broad tubercules or described in Kelly-Borges et al. (1994). The dried mounds, microscopically smooth, slippery and holotype of Latrunculia (=Negombata) corticata rubbery. Ostia radiate in stellate formation in (Carter) was reconstituted by prolonged soaking shallow rounded depressions, oscules apical with of a fragment in dilute detergent. The colour low fleshy raised collars, frequently differentially notation for living and preserved specimens fol­ coloured cream, always highly contracted and lows the Rheinhold Colour Atlas (Kornerup & invisible in preserved specimens. Texture ex­ Wanscher, 1961). Spicule dimensions are given tremely tough but elastic. Colour usually pale as mean length (range of length measurements) purple-pink mottled with deep reddish brown and times width (range of width measurements) fol­ yellow brown, interior cream. Megascleres sub­ lowed by the number of spicule measurements tylostrongyles, with a faint subterminal swelling taken. Primary type material has been deposited at proximal end, distal end oxeote or more typi­ in the Queensland Museum, Brisbane, Australia, cally strongylote. Microscleres, two size catego­ paratypes and fragment of holotypes, schizotypes ries of spinulate acanthomicrorhabds of some authors, have been deposited in the Nat­ (spinorhabds), smallest category always present ural History Museum, London, and in the and disposed on and below scalloped ectosome­ Museum National d'Histoire Naturelle (Paris). choanosome boundary, second larger category, Registration numbers are cited in the text. disposed within the ectosome and choanosome Abbreviations used in the text: QM, Queens­ when present. Skeletal arrangement land Museum, Brisbane; BMNH, Natural History plumoreticulate. Primary fibres, conspicuous, ap­ Museum, London; MNHN, Museum National pearing as vertebrate ligaments in soft flesh. d'Histoire Naturelle, Paris; HBOM, Harbor Megascleres are packed and arranged parallel Branch Oceanographic Museum, Fort Pierce, within axis of fibre, spongin barely visible, fibre Florida; SIO, Scripps Institute of Oceanography, is often hollow. Multiple fine dendritic spicule San Diego; NSRC (UPNG), Natural Science tracts emanate from tip of primary fibre as it Resource Centre, University of Papua New passes into ectosome in an umbelliform arrange­ Guinea, Port Moresby; CRRF, Coral Reef Re­ ment, fibres radiating towards surface terminat­ search Foundation, Federated States of Microne­ ing in ectosomal megasclere brushes. Primary sia; OCDN, Specimen sample numbers for United fibres, connected by occasional short secondary States National Cancer Institute shallow-water fibres, or anastomosing with adjacent fibres. REVISION OF DlACARNUS AND NEGOMBATA 479

Ectosome, dense, rubbery, distinct from the un­ form slightly fanned brushes after narrowing to derlying choanosome, with layers of collencytes 5-lO~m or only 1-4 spicule widths. Megascleres and collagen fibrils parallel to surface. Choano­ form a palisade of brushes between dendritic tract ;ome, dense, with a few large canals surrounded brushes. Megascleres, scattered singly or in JY a dense layer of contractile cells and collagen groups of 1-3 spicules, in a vaguely radiate to fibrils. Fibres are surrounded by a sheet of elon­ confused arrangement interstitially between pri­ gate cells and bundles ofcollagen fibrils. Sponges mary fibres. Larger thicker microscleres are scat­ incubate huge, bright orange-yellow tered predominantly around edges of exhalant parenchymella larvae. canals in choanosome, around ectosomal lacu­ nae, and just below lower boundary of ectosome Diacarnus spinipoculum (Carter) comb. nov. where smaller thinner spinorhabds are also found. (Figs 1, 4A, 5A, 6A, 7A, 8A; Table 1) Ectosome, 1200-1500J.l.m deep, extremely dense with parallel collagen fibrils, easily distinguished Axos spinipoculum Carter, 1879: 286 from underlying choanosome which is dense and Diacamus spinipoculum; Burton , 1934: 549 soft. Latrunculia spinipoculum; Hooper, 1986: 181 Spicules. Megascleres. Subtylostrongyles with slight subterminal swelling: 266(232-305) x 4(2- MATERIAL EXAMINED ?)!J.ID (see Table 1). HOLOTYPE: BMNH 1846.10.14.174 from Port Jack­ Microscleres: Spinorhabds I: extremely fine son, Sydney, growing on "hard objects" (Carter 1879), with small sharp spines scattered along shaft; depth unknown, presented by J. B.Jukes, Esq (Fig. 1). 46(37-55) x

FIG. 1. Diacamus spinipoculum (Carter), preserved holotype BMNH 1846.10.14.174, Port Jackson, Sydney. Scale: Scm. REVISION OF DIACARNUS AND NEGOMBATA 481

TABLE 1. Spicule dimensions of Diacamus spinipocu/um (Carter) possession of dicastra, which form a and Diacamus be/lae sp. nov., given as mean length(range), mean single ectosomal layer of erect spic­ thickness (range), all in llffi. ules, their bases implanted in the der­ mal membrane (Fig. 9D). The Meeascleres Spinorhads I Spinorhabds II dicastra or discorhabds of Diacamus foinipoculum 277~32-305) 53(50-55) 60(53-65) Latrunculia s.s. have two central (Carter) Ho 078.1 BMNH 1846. .14.174 6(4- ) 2). The megasclere skeleton is typi­ cally a whispy loose irregular reticu- older histological section of the holotype, and so lation of indistinct tracts, megascleres the most diagnostic field characteristic of the are styles, and microscleres are discorhabds. The sponge had not been reported in recent literature. surface of Latrunculia has characteristic mammi­ The unusual and characteristic skeletal arrange­ form ostial sieve-plates. The transfer of Axos ment of the umbelliform primary fibres within the spinipoculum to Latrunculia by Hooper (1986) is ectosome, observable only with detailed serial therefore invalid, as it differs considerably in the histological sectioning, was also not possible to form and disposition of the microscleres, the form detect in the slide of the holotype, nor were the and disposition of the fibre skeleton, mesohyl two size categories of microscleres. The ectoso­ matrix construction and in surface morphology. mal megasclere brushes are paratangential in ori­ Hooper ( 1986) also suggested Axos entation in the holotype slide (Hooper, 1986) but spinipoculum and Negombo tenuistellata Dendy are vertical, as in the living sponge, where the 1905 from Ceylon are synonymous but conceded ectosome has not been squashed due to confine­ that the microscleres of the former are more reg­ ment in its container. ular than those of Negombo. Examination of N. Two specimens from East Fayu Island have tenuiste lla ta (B MNH 1907. 2.1.28) and histolog­ been assigned to Diacarnus spinipoculum (Car­ ical sections of this sponge, reveal that it is very ter) even though the spicule dimensions are different in morphology and spiculation from A slightly different from the holotype; there is con­ spinipoculum: N. tenuistellata is a series of com­ siderable variation between all specimens exam­ pressible thin-walled coalescent tubes, the skele­ ined and spicule dimensions cannot be ton of which consists oflarge gently curved styles considered alone in differentiation of species in in confused longitudinal tracts, reminiscent of the Diacarnus. Even though the spicule dimensions halichondroid organisation, with minute micro­ of these !ipecimens are slightly more similar to the spined sanidaster-like microscleres scattered in­ later described D. bellae, spicule morphology is terstitially and in a dermal membrane (Fig. 8H). very different and identical to that of D. The microscleres of N. tenuistellata are acanthose spinipoculum. microrhabds with irregular whorls of spines In a revision of Axos Gray from north-western along the shaft. and are much smaller than the Australia, Hooper ( 1986) transferred Axos microscleres of the Latrunculiidae. Axos spinipoculum to Latrunculia, considering the spinipoculum and N. tenuistellata are not synon­ general skeletal features of A. spinipoculum to be ymous and neither have microscleres typical of similar to those of Latrunculia s.s. The type spe­ Latrunculia s.s. Negombo was incorrectly placed cies of Latrunculia, L cratera Bocage 1869, has in the Spirastrellidae by Dendy ( 1905), there is no several characteristic features described clearly evidence of spiralling along the shaft of the spic­ by Bocage ( 1869), and later recognised by Ridley ule. De Laubenfels (1936: 132) used the name and Dendy ( 1887) in the descriptions of Negombo for an alleged desmoxyid sponge. This Latrunculia apicalis and L brevis (Ridley & genus name has not been used since, to our Dendy, 1886). Diagnostic features include the knowledge. 482 MEMOIRS OF THE QUEENSLAND MUSEUM

FIG. 2. Negombata corticata (Carter), dry holotype BMNH 1840.5.6.56-58, three dried fragments of the same specimen, Red Sea. Scale: 5cm. REVISION OF DIACARNUS AND NEGOMBATA 483

Diacarnus bellae sp. nov. are common within choanosome, particularly lin­ (Figs 4B, 5B, 5C, 6B, 7B, 8B; Table 1) ing canals. Thinner microscleres, found only below lower ectosomal boundary. Ectosome, lATERIAL ExAMINED 1225 to 17001.J.m deep. Choanocyte chambers, doLOTYPE: QMG305007: Anaw Wall, oceanside of approximately 25~m diameter. reef west of Anaw Channel, northwest side of Chuuk Spicules. Megascleres. (Table 1): Sub­ Atoll, Micronesia, 7"34.24'N; 151"40.19'E, 24m, col­ tylostrongyles: 273(240-300) x (3-5)1.J.m, n=50. lected by P. L. Colin, CRRF, 7 August 1993; FRAG­ Microscletes. (Table 1): Spinorhabd I: ex­ MENT OF HOLOTYPE: BMNH 1994.5.22.7. tremely fine with bumps or very short spines ADDIDONAL MATERIAL: BMNH 1994.5.22.11 (frag­ scattered irregularly along shaft: 45(36-50) x ment of OCDN 120-0): west side of Nama Island, 30 nautical miles east of Chuuk Atoll, Micronesia,

AoomONAL MATERIAL: MNHN DJV52: fore reef fibres, diverge and ramify dendritically through zone, free living on the sand and embodying a piece of ectosome, occasionally dividing just below sur­ rope, 30m, King Saud Palace, Djeddah, 21"70'N, face, giving rise to brushes (Fig. 6C). 38"80'E, collected by J. Vacelet, 8 March 1983; Megascleres, form an erect to fanned palisade BMNH 1994.5.22.17: Ras Muhammad, 2-lOm, South juxtaposed upon ectosomal brushes. Sinai, Red Sea, dry specimen collected by Loya, 1977 (Loya 11); MNHN DJV53: Ras Muhammad, South Spinorhabds, rarely present, distributed just Sinai, Red Sea, collected by Loya, 1978 (Loya 254); below the ectosome in upper choanosome, larger MNHN DJV54: Harvey Reef, lOrn, near Port Sudan, category absent. Ectosome, 600-900J.!.m deep, collected by C. Wilkinson, August 1978; BMNH distinct, rubbery. Choanocyte chambers, approx­ 1978.12.14.2: Harvey Reef, lOrn, near Port Sudan, C. imately 25J.!.m diameter. Wilkinson, August 1978. Spicules. Megascleres (Table 2). Sub­ tylostrongy les, straight, with a very faint swelling DESCRWTION of proximal end, distal end rounded: 240(190- Repent or erect branches, 20-30mm diameter 269) x (2.5-4.8)J.!.m, n=30. (Fig. 5D), variable length and branching pattern, Microscleres (Table 2). Spinorhabds 1: straight, often with slightly enlarged ends up to 50mm with occasional swellings or short spines scat­ diameter, often anastomosing to form a sprawling tered along shaft and near apices, occasional to mass up to lm diameter, frequently free living on rare: 25(24-30) x 0.8-l.3J.!.m, n= 25 (Fig. 7C). sand (Fig. 5E) or attached to debris. Texture, tough, just compressible, elastic. Surface, irregu­ GEOGRAPHIC DISTRIBUTION lar with broad blunt conules, 2 mm high, on ends South Sinai, Eilat, Harvey Reef off Port Sudan, of branches. Ostia, 40-60J.!.m diameter, localized Red Sea (Fig. 4C) in groups of about 10 in small surface depres­ sions. Oscules, 2-3mm diameter, found on upper surface of branches or at tips of enlarged ~ranch SUBSTRATE, DEPTII RANGE, ECOLOGY ends when in upright position, opening at end of Common on coral substrate or rocks, large raised cream-coloured fleshy collars. Co,our, specimens, free living or attached to small corals rose pink (9D6), mottled with oak brown (5D6) fragments in sand. Depth range, l0-30m, al­ on surface, darker brown in approximately hex­ though reported to be very common at Eilat in agonal concave depressions surrounded by cream shallow water of 2-lOm (Loya, pers. comm, ridges, more uniformly brown near end of 1978). Parenchymella larvae, 2mm diameter, branches. Interior in life and whole sponge in within choanosome of MNHN DJV52. ethanol, cream (4A2). ETYMOLOGY SKELETON The species name indicates that the sponge is Primary fibres, 500-SOOJ.!.m diameter, up to found only in the Red Sea. 1800J.!.m in centre of branch, radiate towards sponge surface, connected by sparse short sec­ REMARKS ondary fibres, 270-420J.!.m diameter. This sponge is easily recognised in the field by Megascleres, scattered interstitially in choano­ its external morphology of repent anastomosing some. Curved tracts, 30-50J.!.m diameter, less than branches, mottled surface colouration, conulose 20J.!.m diameter when crossing ectosome-choano­ surface and smooth rubbery texture, and tough some boundary, emanate from tips of primary internal fibres. The small thin spinorhabds are

FIG. 3. Comparison of the developmental and mature stages of spinorhabd microscleres of Sigmosceptrella, Diacamus and Negombata. A-D, Sigmosceptrel/afibrosa, Dendy (BMNH 1925.11.1.717), Maria Island, Tasmania: A, B, rhabd in earliest stage with strongly recurved spines similar to those of Negombata magnifica (K-L). C, the maturing spinorhabd shows signs of the fused recurved spines on the shaft, D, mature spinorhabd with a distinctly sigmoid rhabd on the right. Scale: A: 9.1~m; B: 6~m; C: !0.4~m; D: 18.2J.lm. E-G, Sigmosceptrella quadrilobata Dendy (E-100, X-371), Tulear, Madagascar. E, specimen E-1 00, sigmoid rahbd. F, specimen X-371, mature spinorhabd. G, specimen X-371, mature spinorhabd with post-fusion sigmoid rhabd. Scale: E: 9.l~m; F: 16.5~m; G: 18.2~m. H-J, Diacamus ardoulwbae sp. nov., holotype QMG305010. H, spinulate rhabd. I, maturing spinorhabd. J, mature spinorhabd. Scale: H: 9.1~m; 1: 18.8~m; J: 14.6J.liD. K-L, Negombata magnifica (Keller), MNHN DJV 56. K, rhabd with irregular sharply re-curved spines. L, mature spinorhabd. Scale: K: 5.2~m; L: 8~m. 486 MEMOIRS OF THE QUEENSLAND MUSEUM

TABLE 2. Spicule dimensions of Diacarnus found in small, darker surface depressions. Os­ erythraeanus sp. nov., given as mean length (range), cules, on apices of digitations, surrounded by a width, all in Jlm. white margin, 1Omm diameter in preserved spec­ imens. Surface with rounded depressions con­ Sample Megascleres Spinorhads I taining ostia in stellate arrangement. Colour in Diacamus e'lthraeanus life oak: brown (5D6), darker in ostial depres­ ~MG30500 25 3(200-285) 23.3(15-27.5) olotype (2.6-3.5) (I) sions, mottled with cream ( 4A2) in surface MNHNDN52 273(235-325) 22.5 (20-25)rare patches and around oscule margin; cream inte­ rior, uniformly white in ethanol. MNHNDN53 258(195-290) 24.7(10-30) Skeleton. Plumoreticulate arrangement with 251(200-300) 23.7 ( 12.5-30 )rare BMNH 1994.5.22.17 1(2.5-2.7) I (0.8-1.3) very thick multispicular fibres, 700-1125Jlm di­ ameter, joined by small short secondary fibres, I f:49(21~269) MNHNDN54 2.4-4.8 absent 250-375Jlffi diameter, at right angles to primary 243(211-269) fibres (Fig. 6D). Small dendritic tracts, 30-50Jlm BMNH 1978.12.14.2 (2.4-4.8) absent diameter, emanate from tip of fibre, dividing and radiating through ectosome toward surface, end­ ing in narrow brushes, surface raised in large rare and are easily overlooked in spicule prepara­ blunt conules. Megascleres form an erect palisade tions and histological sections. at surface. Interstitial megascleres, abundant. Spinorhabds, rare, found only in superficial lay­ Diacarnus levii sp. nov. ers of choanosome when present. Ectosome, (Figs 4A, 5F, 6D, 7D; Table 3) thick, 500-1 1OOJlm, composed of parallel colla­ gen fibrils in wavy bundles. Choanocyte cham­ MATERIAL EXAMINED bers, 20 to 25Jlm diameter. HOLOTYPE: QMG305009: (ORSTOM R1524), St. Spicules. Megascleres. (Table 3): Sub­ 478, lagoon, Belep, New Caledonia, 4-25m, tylostrongyles, with slight swelling at proximal 19'34.03'S, 163'42.03'E, 8 March 1990; FRAGMENT end: 258(210-300) x 2-5Jlm, n=30. OF HOLOTYPE: BMNH 1994.5.22.16; MNHN Microscleres. (Table 3): Spinorhabd 1: straight, DCL3659. with swellings or small, irregular spines usually AoomONAL MATERIAL: MNHN DCL3660: (OR­ more developed near apices, always rare, may be STOM R767), St. 206, Bane de Ia Torche, New Cale­ donia, 35m, 22'56.60'S, 167"40.00'E; MNHN absent in some specimens: 58(53-60) x 0.5- DCL3661:(0RSTOMR879).St.I84, 18-25m,lagoon 2.4Jlm, n=lO (Fig. 70). ilot Ua, New Caledonia, 22'43.00'S, 166'49.10'E, 28 June 1979; MNHN DCL3662: (ORSTOM R994), St. SUBSTRATE, DEPTH RANGE, ECOLOGY 225, 42m, Canala Pass, New Caledonia, 21'18.80'S, Found on fringing coral reef slopes down to 165'57.25'E; MNHN DCL3663: (ORSTOM R1247), 35m. Large embryos or parenchymella larvae, up St. 270, 8m, Kouare lagoon, New Caledonia, to 1.5mm diameter, white or yellow in the 22'46.50'5, 166'47.90'E, 3 March 1980; MNHN DCL3664: (ORSTOM Rl291), St. 303, 6m, fore-reef choanosome, are observed. in all specimens. zone, Cook Reef, New Caledonia, 19'45.60'5, 161'41.40'E, 23 June 1981; MNHN DCL3665: (OR­ GEOGRAPHIC DISTRIBUTION STOM R1528), St. 448, 6-15m, fore-reef zone, Abore New Caledonia; northeastern Great Barrier reef, New Caledonia, 22'20.00'S, 166'13.15'E; Reef, Australia (Fig. 4A) BMNH 1994.5.22.15 (fragment of Q66C-0248): west side of seaward vertical wall of reef, 18m, Big Broad­ hurst Reef, Great Barrier Reef, Australia, 18'56.67'S, ETYMOLOGY 147'43.86'E; BMNH 1994.5.22.8 (fragmentofQ66C- For Professor Claude Levi. 0890): Black Reef, Whitsunday Islands, Great Barrier Reef, lOrn, coiiected by M. KeJiy-Borges, 20 October REMARKS 1987. Diacamus levii appears to be fairly common ir New Caledonia, but has on! y been collected from DESCRIPTION two loc11tions on the Great Ranier Reef. In both Thick erect digitations or lobes, 30-45mm di­ locations the sponge was rare. Specimens from ameter, anastomosing in large masses (Fig. 5F). the Whitsunday Islands and Big Broadhurst Reef, Texture, tough, just compressible, elastic. Sur­ Australia, have slightly larger megascleres and face, with low rounded conules, approximately 2 the spinorhabds are slightly thicker with better mm high, 2-5mm apart. Ostia, 50Jlm diameter, developed spines, however, they are similar to the REVISION OF DlACARNUS AND NEGOMBATA 487

TABLE 3. Spicule dimensions of Diacamus levii sp. Megascleres, abundant in choanosome. Ecto­ nov., given as mean length (range), width (range), all some, generally 300!J.m deep, ranging from 120 in J.UD. to 5401J.m. Microscleres, two categories, smaller dispersed under ectosome-choanosome bound­ Sam~e Megascleres Sj)inorhads I ary, larger microscleres form a dense layer in Diacamus levii 247(190-280) 5'N5-62.5) middle of ectosome (Fig. 6E), scattered through­ Holo~ (2.4-3.0) (0. -1.0) IQMG 5009 out choanosome particularly around canals, 248(245-315) never found on the outer layer of ectosome. MNHN DCL3663 (2.5-3.0) not found Spicules. Megascleres: Subtylostrongyles: 255(220-288) x 2.4-6jlm, n=30. MNHN DCL3665 t?2(21 ~)278) {(9(58-f~?are 2.4-5.0 0.5-1.0 Microscleres. Spinorhabd I: spines are very BMNH 1994.5.22.15 245(2~288) ll(8(53-~) irregular in shape and disposition, terminal spines 2.0-5.0 0.5-1.0 occasionally bifurcate, central shaft bumpy, two 266(240-288) 57(53-60) major whorls of spines are evident, and closer to BMNH 1994.5.22.8 (2.0-5.0) (0.5-1.0) ends of spicule rather than regularly spaced along New Caledonian specimens in all other aspects; axis as in second category: 29(24-34) x 2.41J.ID Diacamus levii is differentiated from the Red Sea (Fig. 3H, 7E); Spinorhabd II: conical spines are species, D. erythraeanus, by gross morphology; typically disposed in irregular whorls along shaft. D. levii is more massive than D. erythraeanus inner two being usually a greater distance apart which is ramose and branching. The major spic­ than each are to terminal spine cap (Fig. 31, 3J, ule difference between D. levii and D. 8C). Terminal spine cap and outer whorl fre­ erythraeanus is that the spinorhabds in the former quently merge, these spicules are often shorter are almost twice as long as those of D. than those with equally spaced whorls: 54(48- erythraeanus. 67)1J.m, n=IO, total maximum width 19-3l!J.m, shaft diameter 2-:4.51J.m. Diacarnus ardoukobae sp. nov. (Figs 3H, 31, 31, 4C, 5G, 6E, 7E, 8C) GEOGRAPHIC DISTJUBUTION Known only from Djibouti, Gulf of Aden (Fig. 4C). MATERIAL EXAMINED HOLOTYPE: QMG305010, Musha island, Djibouti, Gulf of Aden, 11' 42.50'N, 43'08.40'E, fore reef zone, SUBSTRATE, DEPTH RANGE. ECOLOGY several specimens observed, some free living on sand, Specimens free-living on sand at 24m deep. 24m, collected by J. Vacelet, 5 January 1985 (M 8); Reproductive mode, possibly fragmentation and FRAGMENT OF HOWTYPE: BMNH 1994.5.22.14, release of parenchymella larvae as in other spe­ MNHNDN55. cies of Diacamus.

DESCRIPTION ETYMOLOGY Curved erect or repent branches (Fig. 5G), sep­ For the "Ardoukoba" expedition during which arate or forming a sprawling mass up to approxi­ specimens were collected. mately 50cm long, uniform L5-2.0cm diameter. Ends of branches markedly conulose, branch sur­ REMARKS faces irregularly bumpy with occasional conules, External morphology is similar to D. texture compressible, rubbery. Oscules distrib­ erythraeanus, which is also ramose and branch­ uted along branches rather than at tips of branch. ing. This latter species forms mats of anastomos­ Colour in life pale pink mottled with dark pink in ing branches which typically have enlarged ends surface depressions, uniformly cream in ethanol upon which the oscules are located. Specimens of (4A2). D. ardoukobae are separate, elongate branches, Skeleton. Primary fibres, 300-1 OOO!J.m diame­ uniform in diameter, with flush oscules along ter, radiate towards sponge surface, connected by each branch. Histologically, D. ardoukobae is abundant secondary fibres, 60-3001J.m diameter. distinct from D. erythraeanus, as it has a compar­ Six to ten short compact fibres, 30-7o,.tm diame­ atively thin ectosome and less robust fibres which ter, radiate from tip of primary fibre, occasionally form a rather tighter meshed skeleton than in D. branching to form irregular sparse surface erythraeanus. The fibres which diverge from the brushes (Fig. 6E). Megascleres form an irregular apex of the primary fibres in D. ardoukobae are and often paratangential palisade between fibres. comparatively thick and short and much reduced 488 MEMOIRS OF THE QUEENSLAND MUSEUM

lected by P. L. Colin, CRRF, 6 rl Q M1crones1a November 1992; FRAGMENT OF ~ East Fayu I. HOLOTYPE: BMNH 1994.5.22.13. • ~ ® • •® Chuuk (Truk) Atoll AoomONAL MATERIAL: BMNH . Nama I. Pohnpei 1994.5.22.12 (fragment of OCDN Papua New Gumea o 0 \~(, ~- 754-C): Banban reef (uncharted), ...... ,~"l\ 20m, Kimbe Bay, West New Brit­ _, •\ Fiji ain, Papua New Guinea, 4· 55.53' v s S, 150· 55.51 'E, collected by P. L. Great ~: ®d­ RedS~~~ Barrier '-" · Colin, CRRF, 14 November 1993; '-Reef New BMNH 1993.11.5.1: oceanside of ) • Caledonia Suikin\) / Makada Reef, 15m, Duke of York Djibouti qr_:t., Islands, East New Britain, Papua P~Jackson ""-. Gulf of Aden New Guinea, 4• 6.68'S, 152· '\j 23.87'E, collected by P. L. Colin, c:Jv CRRF, 5 November 1993; BMNH A c I 1993.10.29.1: overhang on outer reef near Albatross Passage, 6- B 'II 12m, north-western New Ireland, ·:·.· Albatross Passa!Je Ojaull. Papua New Guinea, 2• 45.23'S, • s;::::~,;. o~CX..+ New Ireland 150· 43.24'E, collected by P. L. Bismarck Sea Watom ~~· Colin, CRRF, 29 October 1993; -,JJ BMNH 1993.11.1.1: south coast B9al;la9 I. ex: +1 Duke of York Is. wall near east end of Djaul Island, ol+ • ;f;. r-' I •ex: -Tabl. ·~_ombeBay western New Ireland, Papua New I , ' ..., C') Guinea, 2• 58.69'S, 150· 59.33'E, , · Planet Rock New Britain ~ collected by P. L. Colin, CRRF, 1 Papua New Guinea November I 993. I -~ DESCRIPTION I Ramose, each branch l-2cm diameter, approximately 30cm length, branches anastomosing, forming large clumps hanging off overhangs or growing erect from coral surface (Fig. 5H). Fig. 4. Geographic distribution of Diacamus spp. and Negombata spp. A. Texture, slightly compressible, Indo-West Pacific and Australia:® Diacamus spinipoculum (Carter); difficult to tear, rubbery. Sur­ n Diacamus bellae sp. nov.; V Diacamus levii sp. nov. • ;Diacarnus face, conulose but microscopi· megaspinorhabdosa sp. nov. B. Papua New Guinea: x Diacarnus cally smooth. Colour in lif bismarckensis sp. nov.; + Diacamus tubifera sp. nov.; • Diacamus cream flecked with garnet­ megaspinorlwbdosa sp. nov. C. Red Sea and Gulf of Aden: ... Diacamus brown (907) and copper red erythraeanus sp. nov; '¥ Diacamus ardoukobae sp. nov; • Negombata magnifica (Keller). (7C7), tips and undersides pale cream. Ostia, set in garnet in number compared to those of D. erythraeanus. brown stellate surface depressions, set in a paler The major difference on which these two species copper red surrounding. are separated, however, is the presence of a sec­ Skeleton. Primary fibres, 300--600!-Lm diameter, ond category of spinorhabd- a larger, more robust run parallel with branch, connected by perpendic­ and heavily spined microsclere, which is dis­ ular short secondary fibres, 120-3001-Lm diameter posed in a mid-ectosomal layer. (Fig. 6F). Dendritic fibres, 30-501-Lm diamete' radiate from tip of primary fibre, occasional: Diacarnus bismarckensis sp. nov. branching, forming surface brushes. Interstitial (Figs 4B, 5H, 6F, 7F, 80) megascleres, abundant in choanosome. Ecto­ some, 300-450!-Lm deep on sides of branches. Larger category of spinorhabds distributed in an MATERIAL EXAMINED HOLOTYPE: QMG305011: (fragment of OCDN 705- ordered layer just below surface of sponge, abun­ Y): Planet Rock, 15m, south of Madang, northern dant in choanosome and along ectosome-choano­ Papua New Guinea, 5" 15.475'S, 145. 49.12'E, col- some boundary in places (Fig. 6F). REVISION OF DIACARNUS AND NEGOMBATA 489

Spicules. Megascleres. Subtylostrongyles: Diacarnus tubifera sp. nov. 274(250-320) x 2.S-811m. n=30. (Figs 4B, 51, 51, 6G, 7G, 8E) Microscleres. Spinorhabd 1: spines are rela­ tively regular in shape, disposed in four equidis­ MATERIAL EXAMINED tant, frequently incomplete whorls: 39(33-48) x HOLOTYPE: QMG305012: oceanside reef near Pizion 2.41-lm (Fig. 7F). Spinorhabd II: sharp conical Island, Chuuk Atoll, Micronesia, 45m, 7" ll.lO'N, 151• 50.20'E, collected by P. L. Colin, CRRF, 7 Janu­ spines of equal length, disposed in regular whorls ary 1994; FRAGMENT OF HOLOTYPE: B MNH along shaft, each equidistant from the other: 1994.5.22.2 58(46-67)1!m, n=IO, total maximum width I4- ADDITIONAL MATERIAL: BMNH 1994.5.22.1: west 26!!m, shaft diameter 5-I 011m (Fig. 8D). side of East Fayu Island, 30.5m, 60nautical miles north-northwest of Chuuk Atoll, Micronesia, s· 32.6l'N, 151• 20.0l'E, collected by P. L. Colin, CRRF, 14 January 1994; BMNH 1994.5.22.10 (frag­ GEOGRAPIDC DISTRIBUTION ment of OCDN 716-M): west side of barrier reef sur­ Bismarck Sea (Fig. 4B), very common on bar­ rounding Bagabag Island, 28m, 30nautical miles rier patch reefs of western New Ireland, New offshore on northern Papua New Guinea coastline, 4" Britain, and Madang, north coast of mainland 47.15'S, 146.10.96'E,collectedbyP.L.Colin,CRRF, Papua New Guinea, particularly common around 9 November 1993; Q66C-6138: attached to vertical Djaul Island off the northwest tip of New Ireland wall. 40m, "The Grotto", southeast Watom Island, off (L. J. Bell, pers. comm. 1994). western coast of New Britain, Papua New Guinea, 4· 6.2'S, 152· 5.7'E, collected by M. Kelly-Borges, 16 October 1991; BMNH 1993.11.1.2: south coast wall near east end of Djaul Island, western New Ireland, SUBSTRATE, DEPTH RANGE, EcOLOGY Papua New Guinea, 2· 58.69'S, 150· 59.33'E, col­ lected by P. L. Colin, CRRF, I November 1993.; Found predominantly on ledges, overhangs, on BMNH 1993.10.22.1: oceanside of Eastern Fields cave walls within I 0 to 30m. Reproduction by Atoll, 9m, Coral Sea, 90 miles southwest of Port production of large yellow-pigmented larvae. Moresby, southern Papua New Guinea, 10· OO.S()'S, 145" 40.24'E, collected by P. L. Colin, CRRF;·22 October 1993; BMNH 1995.7.4.5: Tubbataha Reefs, south of Cagayan Islands, Sulu Sea, Philippines, ETYMOLOGY 8°40'N, !20°E, disturbed rubble and sand, lOrn, col­ For the Bismarck Sea. lected by L. Sharron, 21 April 1995; BMNH 1995.7.4.3: Tubbataha Reefs, south of Cagayan Is­ lands, Sulu Sea, Philippines, 8°40'N, 120°E, on verti­ cal coral wall, 37m, collected by L. Sharron, 22 April REMARKS 1995; BMNH 1995.7.4.2: Puerto Princessa, Palawan, General morphology, growth form, and col­ Philippines, 9°50'N, ll8°30'E, 12-18m, collected by ouration of this northern Papua New Guinean P.L. Colin, 15 April 1995. species is remarkably similar to that of Diacamus ardoukobae from the Gulf of Aden. Both are DESCRIPTION ramose, have a relatively thin ectosome, and pos­ Single tubular to vase-shaped, very heavy, sess a second larger category of spinorhabd dis­ thick-walled (Fig. 51), average height 80cm, I5- posed in a central to upper ectosomal layer and 30cm wide, most frequently twice as tall as wide, throughout the choanosome. The major differ­ with large lobate processes on the outer surface, ence between these two species is the spicule size surface tuberculate. Atrium, apical, typically up and morphology of the largest spinorhabds. to 15cm wide in vase~shaped sponges (Fig. SJ), Megascleres and microscleres in D. bismarcken­ typically about 8cm diameter in tubular sponges. sis are longer than in D. ardoukobae, and the Smaller specimens extremely conulose, often I 0- latter are more abundant in D. bismarckensis. In 14cm diameter, 9-IOcm high, with a Scm wide D. bismarckensis, microsclere spines are very atrium. Texture, very firm, rubbery, mesohyl ma­ regular, of equal length, and disposed equidis­ trix very dense, fleshy. Colour in life carrot red tantly along the shaft of the spicule. In D. (6B7) mottled with bright reddish brown (908), ardoukobae, spines are irregular, with the central very patchy, interior of tube bright white, beige whorl of spines being longer, and disposed closer in ethanol. to the ends of the spinorhabd. These differences Skeleton. Primary fibres, 450-6001-lm diameter, are also apparent in the smaller spinorhabds of D. sparse, run longitudinally, anastomosing along ardoukobae. axis of tubular body towards sponge apex. Large 490 MEMOIRS OF THE QUEENSLAND MUSEUM REVISION OF DlACARNUS AND NEGOMBATA 491

branches, 150-300tJ.m diameter, emerge some reproduction not observed, sponge surface fre­ distance below ectosome-choanosomal boundary quently completely infested with barnacles. -"rom upper 5-lOmm of primary fibre. Fibre ex- :nsions radiate towards sponge surface, branch­ ETYMOLOOY ng many times before entering ectosome Reflects the consistently tubular morphology. (900-2400f.lm deep; average depth 1500J.1.m, with lacunae 30-60J.l.ffi wide), forming an immense REMARKS number of very fine, long, dendritic fibres which Diacamus tubifera is conspicuous and easily meander through choanosome and ectosome ul­ recognised in the field as large long solitary or timately forming small surface brushes in outer double tubes, with an extremely heavy, rubbery ectosome (Fig. 6G). Immense umbells raise sur­ texture. Smaller specimens are also tube-shaped face into rounded lobes. Abundant megascleres and extremely conulose, also with a deep central form a palisade superimposed over the surface atrium. Histologically, the sponge is well differ­ brushes. Primary fibres anastomose, also con­ entiated from all species described thus far. The nected by short secondary fibres 120-300J.1.m di­ sponge has a very thick ectosome which is rami­ ameter. Interstitial megascleres, abundant, fied with abundant dendritic fibres which ema­ arranged in loose broad bundles between primary nate well within the choanosome, arising to form fibres. Large spinorhabds, scattered in outer ecto­ a lobed surface. The megascleres and large cate­ some, slightly more abundant in outer third of gory of spinorhabds are larger than in previously ectosome, absent from outer 50J.1.m, common described species, and the large spinorhabds are throughout choanosome. Megascleres and scattered throughout the ectosome rather than rnicroscleres abundance variable between speci­ being restricted to a distinct rnid-ectosomal band mens. Smaller category of spinorhabds dispersed such as in D. ardoukobae and D. bismarckensis. under ectosome-choanosome boundary. Spicules. Megascleres. Subtylostrongyles: Diacarnus megaspinorhabdosa sp. nov. 304(250-345) x 2.5-5J.1.m, n=30. (Figs 4B, 5K, 6H, 61, 7H, 8F) Microscleres. Spinorhabd 1: extremely fine with fine spines dispersed along shaft: 38(38-41) MATERIAL ExAMINED x 2.5J.1.m, n=lO (Fig. 7G); Spinorhabd II: distinct HOLOTYPE: QMG305013: fringing reef 150m south of whorls of spines: 60(55-67) x maximum width Motupore Island, !Om, Bootless Bay, Papua New 17-l9J.1.m, shaft width 3.5-5J.1.m, n=30 (Fig. 8E). Guinea, 9" 31.6'S, 147" 16.6'E, collected by M. Kelly­ Borges, 6 December 1985; FRAGMENTS OF HOLOTYPE: BMNH 1994.5.22.6, NSRC(UPNG) 90. GEOORAPHIC DISTRffiUTION ADDITIONAL MATERIAL: BMNH 1994.5.22.5: (frag­ Chuuk Atoll and East Fayu Island, Micronesia; ment of OCDN 673-P): lagoon side of Tab (Pig) Island, Bismarck Sea, widely and sparsely distributed on I Om, on Madang barrier reef, Papua New Guinea, 5" barrier patch reefs of western New Ireland and 10.21 'S, 145" 50.37'E, collected by P. L. Colin, CRRF, New Britain, and Madang on northern coast of 3 November 1992; HBOM 003:00925: fringing reef mainland Papua New Guinea (L. J. Bell, pers. south of Motupore Island, 10m, Bootless Bay, Papua New Guinea, 9" 31.6'S, 147" 16.6'E, collected by M. comm.); Eastern Fields Atoll, southern Papua Kelly-Borges, 24 January 1994; BMNH 1994.7.25.1 New Guinea (Fig. 4B). (fragment of OCDN 2693-M): attached to rubble, 7m, Gamao Point, Batangas, Philippines, 13" 38.39'N, 120" 56.86'E, collected by C. Arneson, CRRF, 25 July SUBSTRATE, DEPTII RANGE, EcOLOGY 1994; BMNH 1995.5.7.4.1: Zambales, West Luzon, Solitary and uncommon in any one locality, Philippines, !5°50'N, 1200E, silty reef, 18m, collected they are found on vertical walls and steep-sloped by P.L. Colin, 2 May 1995; BMNH 1995.7.4.4: oceanic fringing reefs to depths of 45m. Mode of Zambales, West Luzon, Philippines, l5°50'N, l20°E,

FIG. 5. Species of Diacamus in situ. See descriptions is text for size in life. A. Diacamus spinipoculum (Carter), Fiji, BMNH 1994.8.20.1. B-C. Diacamus bel/ae sp. nov., Chuuk Atoll, Micronesia: B. solitary form, BMNH 1994.5.22.9. C. coelescent form, holotype QMG305007. D-E. Diacamus erythraenus sp. nov., Djeddah, Red Sea, holotype QMG305008. F. Diacamus levii sp. nov., Belep, New Caledonia, holotype QMG305009. G. Diacamus ardoukobae sp. nov., Djibouti, holotype QMG3050IO. H. Diacamus bismarckensis sp nov., Djaul Island, Papua New Guinea, BMNH 1993.11.1.1. 1-J. Diacamus tubifera sp. nov., Djaul Island, Papua New Guinea: I. tubular form, BMNH 1993.11.1.2; J. vase-form. K. Diacamus megaspinorhabdosa sp. nov., Batangas, Philippines, BMNH 1994.7.25.1. 492 MEMOIRS OF THE QUEENSLAND MUSEUM REVISION OF DIACARNUS AND NEGOMBATA 493

43m, collected by P.L. Colin, 3 May 1995; BMNH Spicules. Megascleres. Subtylostrongyles: 1995.9.1.1, BMNH 1995.9.1.2; south side of Puerto 318(298-346) x 7-l2Jlm, approximately 20!-lm Princessa Bay, Puerto Princessa, Pallangbato, Philip­ shorter (average) in BMNH 1994.5.22.5 from pines, 9"41.62'N,l18°4.5l'E. Collected by P.L. Colin Madan g. CRRF, 15 Aprill995. Microsc/eres. Spinorhabds I: slender rods with regular whorls of small spines; 48(43-55) x 3- DESCRIPTION 5Jlm, shaft 111m wide (Fig. 7H); Spinorhabds II: Thickly encrusting, lobate, some specimens spines blunt, central shaft very thick, occasion­ with restricted base of attachment, 6-15cm diam­ ally malformed with central shaft bare or spines eter,lobes up to 4cm high (Fig. 5K). Lobe apices re-curved along shaft; 84(74-96)J.1m, n=30, max­ covered in tough sharp conules approximately 2 imum width 36(34-48), shaft width: 15(10- mm high, 2-5mm apart. Oscules, l-3mm diame­ 19)Jlm (Fig. 8F). ter, with smooth raised cream collars, scattered apically on lobes or irregularly over surface. Tex­ SUBSTRATE, DEPTH RANGE, ECOLOGY ture, barely compressible, tough, harsh to touch Found on silty, fringing coral reefs down to but microscopically velvety between conules. 1Om. Choanosome contains large conspicuous Colour in life oxblood red (9E8), mustard (5C8) embryos of parenchymella type, up to 1.5mm internally, uniformly cream (4A2) in ethanol. diameter. Easily observed in specimens in the BMNH 1994.5.22.5 from Madang was pale field, they are bright orange-yellow, contrasting brown (4B4) in life due to shading effects. Very with cream mesohyal matrix, they are abundant, large bright yellow embryos are abundant in present at any time of collection. Larvae, loosely BMNH 1994.5.22.6 (Fig. 61). embedded within the central part of body (Fig. Skeleton. Primary fibres, sparse, tough, up to 61), many contain thin megascleres dispersed 9~m diameter, raise surface into well-sepa­ without order in central region. Microscleres, ab­ rated tough, sharply pointed conules (Fig. 6H). sent from embryos. Round or ovoid inclusions, Primary fibres joined by short, very thick, sec­ 50-80J.!.m diameter, with a fibrillar content, abun­ ondary fibres, up to 5oo,mt thick. Sparse spicule dant in the choanosome, possibly spermatocysts. tracts, 60-1 OOJlm wide, diverge from primary fibre well within choanosome, branching, occa­ GEOGRAPHIC DISTRffiUTION sionally forming well separated tracts that radiate Northern (Madang), southern (Motupore Is­ within ectosome. Tracts end in dense brushes at land) Papua New Guinea; Batangas, Philippines surface where they interdigitate with a dense (Fig. 4B) loose palisade of megascleres. Interstitial megascleres, numerous in choanosome, dis­ persed without order between fibres. Larger fi­ ETYMOLOGY bres are surrounded by a sheath of bundles of Reflects large size and abundance of largest collagen fibrils, 60-80!-lm thick, which also con­ category of spinorhabds. tains loose megascleres arranged longitudinally along fibre. Large spinorhabds, abundant in ecto­ REMARKS some and disposed in an undulating band in upper The major character separating Diacamus portion of ectosome surrounding lacunae, also megaspinorhabdosa from other species of abundant above lower ectosomal boundary and Diacamus is the possession of extremely large abundant in choanosome, sometimes patchily, spinorhabds, and the largest average megasclere sometimes concentrated around canals (see Fig. dimensions. These microscleres form dense ag­ 61). Small spinorhabds, below ectosomal­ gregations in the choanosome and are disposed in choanosomal boundary. an upper and lower layer within the ectosome.

FIG. 6. A-L, Slceletal arrangements of Diacamus and Negombata spp. A, Diacamus spinipoculum (Carter), holotype BMNH 1846.10.14.174, 30x. B,Diacamusbellaesp. nov., BMNH 1994.5.22.11, 30 x. C, Diacamus erythraeanus sp. nov., MNHN DJV52, 30 x. D, Diacamus levii sp. nov., holotype QMG305009, 30 x. E, Diacamus ardoukobae sp. nov., holotype QMG305010, 30 x. F, Diacamus bismarckensis sp. nov., BMNH 1994.5.22.12, 30 x. G, Diacamus tubifera sp. nov., BMNH 1994.5.22.10, 30 x. H-1, Dicamus megaspinorhabdosa sp. nov., holotype QMG3050 13. H, skeletal arrangement, 30 x. I, larva, 120 x. J, Negombata corticata (Carter), holotype BMNH 1840.5.6.56-58, 30 x. K-L, Negombata magnifica (Keller): holotype fragment BMNH 1908.9.24.118, 30 x. L, MNHN DJV56, 120 x. 494 MEMOIRS OF THE QUEENSLAND MUSEUM

FIG. 7. A-H. Scanning electron micrographs of spinorhabd I microscleres of Diacarnus spp. A, Diacarnus spinipoculum (Carter), holotype BMNH 1846.10.14.174. Scale: 21.2~m. B, Diacarnus bellae sp. nov., BMNH 1994.5.22.11. Scale: 18.5~m. C, Diacarnuserythraeanus sp. nov., MNHN DJV52, Red Sea. Scale: Cl: 12.6~m; C2, C3: 9 . l~m . D, Diacamus levii sp. nov., DI = holotype QMG305009, D2:::: BMNH 1994.5.22.8. Scale: 21.1~m. E, Diacamus ardoukobae sp. nov., holotype QMG30501 0. Scale: 9.3~m. F, Diacarnus bismarckensis sp. nov., holotype QMG305011. Scale: 13~m. G, Diacamus tubifera sp. nov., BMNH 1994.5.22.10. Scale: 15.l~m. H, Dicamus megaspinorhabdosa sp. nov., holotype QMG305013. Scale: HI: 16.2~m; H2: 28~m. REVISION OF DIACARNUS AND NEGOMBATA 495

Negombata de Laubenfels of this sponge. The skeleton consists of a central axis of rectangular meshes formed by spongin Negombata de Laubenfels, 1936: 159 fibres, 300-6001J.m diameter, not clearly differen­ tiated into primary and secondaries (Fig. 61). TYPE SPECIES Large oxea are embedded within the fibres, and Latrunculia corticata Carter, 1879: 298 (by also occur interstitially. Megascleres are not ar­ subsequent designation, de Laubenfels, 1936: ranged uniformly within fibres, but rather, are 159) scattered singly or in groups, and oblique or oc­ casionally perpendicular to fibre axis. An ectoso­ DIAGNOSIS mal skeleton of wavy oxea is present, but this has Erect, massive, lobate to elongate, finger-form­ collapsed in the holotype. rendering the arrange­ ing, branching. Compressible, with an extremely ment difficult to determine. Microscleres, smooth surface through which large pores are densely packed on sponge surface and also in clearly visible. Megasclere skeleton consists of a choanosome, almost obscuring choanosomal fi­ uniform, elongate, to square-meshed reticulation bres. of clear spongin cored fairly irregularly with Spicules. Megascleres. I. Oxeas of main fibro­ thick, slightly curved oxea with strongylote ends, reticulation: thick, straight or slightly curved with and a clear hollow axial canal. Ectosome, with rounded strongylote ends, centrally thickened thin fusiform wavy oxeas in tracts that fan within with a distinct hollow axial canal: 346(317-384) a thick collagenous ectosome, ultimately forming x I O-l2Jlm, n=30; II. Ectosomal oxea: thin, fusi­ surface brushes. Microscleres, irregular form irregularly curved, wavy, disposed in sur­ spinorhabds in two sizes, largest very irregular, face plumose brushes: 422 (394-451) x 2.4-61J.m, thick, without a distinct shaft, smaller, with a n=30. distinct straight or curved shaft with spines of Microscleres. 1: Irregular, shaft rarely distinct, uneven length arranged roughly in two central overall shape of microsclere straight, curved whorls and two terminal bunches, terminal spines slightly or spiralled, spines of uneven length ar­ commonly double. Immature microscleres, ranged roughly in two central whorls and two straight rods in which one or more spines from terminal bunches, terminal spines commonly apices are strongly recurved, disposed predomi­ double: 26( 19-31 )Jlm, n=30; maximum width 14- nantly on very outside of ectosome, but can also 19m, shaft width 3-6Jlm, n=IO; II: Extremely be very dense within choanosome. irregular spinorhabds, almost oval when heavily centrally thickened, shaft indistinct and spins un­ Negombata corticata (Carter) evenly distributed along sides of the spicule: (Figs 2, 61) 35(26-43)1J.m, n=30, maximum width 17-24Jlm, shaft width 7-IOIJ.m, n=IO. lAtrunculia corticata Carter, 1879: 298; Keller,I889: 401 SUBSTRATE, DEPTH RANGE, ECOLOGY There is no information given in Carter (1879) MATERIAL ExAMINED on the ecology or habitat of this species. HOLOTYPE: Latrunculia corticata BMNH 1840.5.6.56-58: "from the Red Sea", three dried frag­ ments from the same specimen (Fig. 2). GEOGRAPHIC DISTRIBlTTION Carter (1879) noted that the sponge was sup­ DESCRIPTION plied by a dealer who stated that the sponge came Carter (1879) described the sponge as being from the Red Sea. "erect, solid, lobate... apparently subsessile". Type specimen consists of short, ridged, broad REMARKS fans of different lengths with a restricted base of Examination of a spicule preparation of the attachment (Fig. 2). Colour, yellowish white in holotype of Latrunculia corticata Carter (BMNH dry sponge. Surface texture of dry holotype, fi­ 1840.5.6.56-58) indicates that this sponge differs brous, roughened due to shrinkage. Carter ( 1879) considerably from species recognised within described the fresh sponge as "chondroid" and Diacarnus s.s., Sigmosceptrella s.s., or "smooth as varnish to the unassisted eye" with Latrunculia~~s. The megascleres of Negombata pores 20!J.m diameter and 7 4Jlm apart. are two forms of wavy oxea, instead of sub­ Skeleton. Reconstitution of the dry holotype ty lostrongy les as in Diacamus and styles in allowed examination of the skeletal organization Latrunculia. Examination of the reconstituted 496 MEMOIRS OF THE QUEENSLAND MUSEUM REVISION OF DIACARNUS AND NEGOMBATA 497

skeleton of the type specimen reveals without examination of the type material, it is megascleres embedded in spongin forming a rec­ impossible to speculate further on the affiliations tangular mesh from which short ectosomal tracts of this specimen. of thinner, wavy oxeas arise to form plumose De Laubenfels ( 1936) also considered the pos­ brushes within the ectosome. In Diacamus the sibility that Negombata was synonymous with fibres are clearly differentiated into very large Negombo Dendy. It is clear from examination of primary fibres, with smaller secondary fibres giv­ the type species Negombo tenuistellata Dendy ing rise to dendritic tertiary fibres in a plumose (BMNH 1907.2. 1.28) that this is not the case; the umbelliform arrangement. Although the megascleres of Negombo tenuistellata are gently microscleres of Negombata are also spinulate, curved large styles arranged in vague longitudi­ they differ from those of other latrunculiid gen­ nal to reticulated tracts within thin hollow-walled era. The largest category are so irregular they tubes, and the microscleres are different, as al­ look like spined lumps of silica, the smaller cate­ ready shown (see remarks for Diacamus gory are also irregular with dense, frequently spinipoculum). double terminal spines that are arranged only vaguely in whorls. The overall spicule is fre­ Negombata magnifica (Keller) quently curved or spiralled, with a distinct shaft (Figs 3K, 3L, 4C, 6K, 6L, 8G) only rarely visible. The smaller category of spinorhabds derive from straight rods in which Latrunculia magnifica Keller, 1889: 402 one or more spines from the apices are strongly recurved. Carter's illustrations of regular MATERIAL ExAMINED microscleres are thus misleading. FRAGMENT OF HOLOTYPE: Latruncu/ia magnifica Latrunculiapurpurea Carter, from Bass Strait, BMNH 1908.9.24.118 (fragment from the Berlin Mu­ southern Australia, was also regarded by de seum) from Suakin, Sudan, Red Sea. AoomONAL MATERIAL: MNHN DJVS6: Musha Is­ Laubenfels ( 1936) to be a species of Negombata. land, Djibouti, 24m, 11.42.50'N, 43.08.4S'E, col­ However, Latrunculia purpurea appears to be a lected by J. Vacelet, 23 January 1985 (M 3), (fragment valid species of Latrunculia. Carter ( 1881) de­ deposited as BMNH 1994.5.22.18) scribed a "flat, compressed, circular, thin, cake­ like or fungiform" sponge, brown-purple in DESCRIJYTION colour, with a "ragged and proliferous upper sur­ Keller (1889) described the sponge as having face". There is no mention, however, of a many long cylindrical digits arising from a short "chondroid dermal region" and "fibroreticulate spreading stalk. A specimen 30cm high, was de­ internal structure" described for Negombata scribed as having 25 branches of equal length conicata, rather, the internal structure is "com­ with a uniform diameter of 6-lOmm. Although pact and densely spiculous." The microscleres of the digits do not commonly divide to form Latrunculia purpurea are typical of Latrunculia branches, occasional short stumps or incipient as the two inner spined discs are disposed towards branches are seen. The texture was described as one end of the spicule, and the microscleres are compressible but firm and elastic. The surface of disposed "to the circumference on the upperside the sponge was completely smooth, slightly and darker portions, arranged perpendicularly in shiney, with a translucent surface, and punctured juxtaposition with the spinous disk of one end with regularly spaced pores 100-l50J..!.m diame­ outwards (Carter, 1881)." The megascleres, how­ ter. Oscules were uncommon, tear-shaped and ever, are described and illustrated as being oxeote raised slightly on collars. The colour in life was as in Negombata corticata and unlike the poly­ described as deep red-orange which remained in tylote or uniform styles of Latrunculia. However, the preserved specimen for a considerable time

AG. 8. Scanning electron micrographs of spinorhabd II microscleres of Diacamus spp, Negombata magnifica (Keller), and acanthose microrhabds of Negombo tenuistellata Dendy: A, Diacamus spinipoculum (Carter), holotype BMNH 1846.10.14.174. Scale: AI: 24.2Jlm; A2: 28.2Jlm. B, Diacamus be/lae sp. nov., BMNH 1994.5.22.11. Scale: 18.2Jlm. C, Diacamus ardoukobae sp. nov., holotype QMG305010. Scale: Cl: 21.3; C2: 16.6J.1m; C3: 14.6Jlm. D, Diacamus bismarckensis sp. nov., holotype QMG305011. Scale: 21.3Jlm. E, Diacamus tubifera sp. nov., BMNH 1994.5.22.10. Scale: El: 18.2Jlm; E2: 24.2Jlm. F, Diacamus megaspinorhnbdosa sp. nov. holotype QMG305013. Scale: Fl: 34Jlm; F2: 30.3Jlm; F3: 34Jlm. G, Negombata magnifica (Keller) MNHN DJV 56. Scale: Gl: 9.1Jlm; G2: 12.5Jlm. H, Negombo tenuistellata Dendy 1905, holotype BMNH 1907 .2.1.28, acanthose microrhabd microscleres. Scale: 5.2Jlm. 498 MEMOIRS OF THE QUEENSLAND MUSEUM

before eventually fading. The ectosome was more 41)J..Lm, n=30; maximum width 22-26Jlm, shaft intensely coloured than the interior of the sponge. width 7-20Jlm, n=l 0 (Fig. 8G). Embryos of the parenchymella type were present and large (825-1125J.UD diameter). SUBSTRATE, DEPTII RANGE, EcOLOGY Skeleton. Details of gross morphology, histol­ The sponge was abundant on coral reefs and ogy, arrangement of the choanosomal skeleton, vertical coral walls, and is one of the most prom­ and spiculation were accurately described and inent sponges in the Gulf ofEilat between depths illustrated by Keller (1889). However, for consis­ of 3-30m (Kashman et al., 1980). Very large tency these are re-interpreted here with details of embryos were abundant in a sponge collected in the ectosomal skeleton not evident in the bolo­ July of 1985 (Fig. 6L). It is not known whether type Negombata corticata. The skeleton consists these are pigmented bright yellow-orange as in of a central axis of thick strongylote oxeas bound Diacamus. into an irregular round-meshed reticulation with abundant stratified spongin. Interstitial GEOGRAPHIC DISTRIBtmON megascleres are common. Primary fibres cannot Suakin, Sudan, Red Sea; Djibouti, Gulf of be clearly distinguished from secondary fibres Aden (Fig. 4C) which average 125J.UD diameter (range: 60- 300J..Lm). Keller (1889) notes and illustrates the REMARKS presence of a "parasitic" algae, thought to be a Negombata magnifica is distinct from N. species of Calithamnion,living within the spon­ corticata in gross morphology, N. magnifica gin fibres. This was not observed in any of our forming long, cylindrical, whip-like branches, specimens. A distinct ectosomal skeleton is pres­ the later being a clump of broad ridged fans. ent and consists of regularly and closely spaced Histologically, the species are rather similar, but compact spicule tracts 30-70f..lm wide which di­ with discemable dimensional and morphological vide and diverge towards the ectosome where differences between the megascleres and they form fanned brushes (Fig. 6K). Megascleres microscleres. The later are extremely dense in the of the ectosomal skeleton are wavy fusiform choanosome of N. corticata. The spongin skele­ oxeas rather than the shorter thicker strongylote ton of N. corticata is more robust. oxea embedded in the spongin fibre. The tips of these spicules protrude through the dermal mem­ DISCUSSION brane in patches. Microscleres are common within the choanosome and present as a distinct Diacamus and Negombata are distinct within layer in the upper third of the ectosome. The the Latrunculiidae as they possess a fibro-reticu­ ectosome is distinct, collagenous, approximately late skeleton with an unusual umbelliform 180f..lm deep (range 240-430f..lm), and perforated "ectosomal" skeleton developed to a greater or by vertical poral canals 150-200f..lm wide which lesser degree, and a highly collagenous mesohyl lead to subdermal lacunae. matrix structure with tough fibres which superfi­ Spicules. Megascleres. I: Oxeas of main fibro­ cially resemble vertebrate ligaments in soft tis­ reticulation, thick with hollow central axis, sue. Diacarnus is remarkably homogenous with the skeletal arrangement, huge fibres and spicule straight or slightly curved, with slightly restricted dimensions of all species being similar. We have strongylote ends: 346 (298-385) x 10( 4.8-12)J..Lm, emphasised the gross morphology of the sponge, n=30; ll: Ectosomal oxea, fusiform irregularly coupled with the presence or absence of a second curved, wavy, disposed in surface plumose larger category of microsclere, and the morphol­ brushes: 435 (403-490) x 1-3Jlm, n=30. ogy of these spicules, in recognizing new species. Microscleres. I. Slender, straight, curved The disposition of these microscieres throughout slightly or spiralled, spines of even length and the sponge, and spicule and fibre dimensions are arranged in two discrete central whorls with ter­ less reliable characters for species differentiation. minal bunches, terminal spines commonly dou­ Diacarnus spinipoculum and D. tubifera are both ble, occasionally recurved, shaft clearly vasiform but the former is barrel-shaped rather distinguished between spines: 23( 19-26)J..Lm, than tubular. These two species are further sepa­ n=30; maximum width 10-12Jlm, shaft width rated on the presence of large spinorhabds in D. 3J..Lm, n=lO (Fig. 3K, 3L); II: Spicule very stout tubifera, in addition to very abundant dendritic with short scattered irregular spines, spines ab­ tertiary spicule tracts which arise from well sent in some parts of the microsclere: 33(24- within the choanosome of this species. Diacamus REVISION OF DlACARNUS AND NEGOMBATA 499

bellae from Micronesia and D. levii from New The possession of large tracts of diverging Caledonia are very similar but are separated on megascleres, megasclere brushes at the surface, gross morphology and spicule dimensions. The spined acanthomicrorhabds, and large bright yel­ three ramose species D. erythraeanus, D. low larvae, in Diacamus and Negombata, is rem­ ardouk.obae, and D. bismarckensis, are separated iniscent of Sigmosceptrella Dendy. Examination geographically, and the two Red Sea species are of the type species of Sigmosceptrella, S. further separated by the presence of large quadrilobata (B MNH 25.11.1.1641 ), and spinorhabds in D. ardoukobae. Diacamus Spirastrella (=Sigmosceptrella) fibrosa Dendy megaspinorhabdosa has the largest megascleres (1897), which Dendy (1922) regarded as the of all species thus far described, and highly diag­ 'true' type species of the genus Sigmosceptrella, nostic spinorhabds. reveals several characteristics which distinguish Although there are many similarities between Sigmosceptrella from Diacamus and Nego­ Negombata and Diacamus, there are several mbata. major differences which separate them, including The largest microscleres of Sigmosceptrella are the presence of diactinal megascleres and a regu­ very similar to those of Diacamus and Nego­ lar tight-meshed "axial" fibrous skeleton in which mbata; but the spines of the two inner whorls of the fibres are only lightly cored by megascleres. the rhabd axis are longer than those of the termi­ 'f?e skeletal arrangement of Negombata superfi­ nal whorls, which are more or less in the same Cially resembles that of several myxillid, raspailid plane as the axis ofthe spicule. and they are closer and microcionid poecilosclerid genera which em­ to the terminal whorls than they are to each other, phasise an axial fibroreticulation and ectosomal imparting a dumbell appearance to the spicule brush-forming skeleton. In Diacamus, the fibres (Fig. 3D, 3F. G). The spinorhabds of Diacamus are packed regularly with spicules with no visible and Negombata are more regular with spines of spongin surrounding them. Negombata corticata equal length in whorls that are separated equidis­ and N. magnifica are differentiated primarily on tantly, in most species, along the axis of the gross morphology, spicule differences, and pos­ spicule (Fig. 3!, J, L). In all specimens of sibly the development of the ectosomal skeleton. Sigmosceptrella examined, these microscleres are packed in confusion in a dermal crust at the Sponges previously identified as Latrunculia surface (Fig. 9A,B) of the sponge, rather than (Keller) from the Red Sea uniquely magnifica scattered in a band within a broad collagenous contain compounds known as latrunculins (Nee­ detachable ectosome, as in Diacamus and Nego­ man et al., 1975; Kashman et al., 1980). With the mbata. transfer of this species to Negombata here, the latrunculins become a potentially diagnostic The developmental stages of the spinorhabds of chemical marker for sponges of the genus Nego­ these three genera also differ. The rhabds of mbata. Several latrunculin - containing voucher Sigmosceptrella are distinctly sigmoid in shape specimens from the Philippines and Indonesia (Fig. 3D, E, G). These were illustrated by Dendy supplied by M-K. Harper (Scripps Institute of (1905: Plate 18 Figure 4c; Dendy 1921: 122) who regarded these as the major diagnostic character Oceanography) and the Coral Reef Reseach for These are not to be mistaken Foundation, were examined as this report went to Sigmosceptre/la. for the "true" sigmas which can also be found in press. While these sponges are clearly related the BMNH histological slide of the holotype. morphologically to the two known species of There were also small anchorate isochelae and Negombata, there are considerable differences, tylostyles in the section and thus the lot is obvi­ requiring the examination of further material be­ ously foreign. The earliest protorhabd forms of fore a complete identification can be made. Spe­ Sigmosceptrella and Negombata are rather more cies identified as Latrunculia brevis and L similar to each other than they are to the pro­ conulosa from Australia (in Butler & Capon, torhabds of Diacarnus, which are simply spinul­ 1991; 1993), Sigmosceptrella laevis (Aibericci et ate as in the mature spicule (Fig. 3H). The earliest al., 1982), and Diacamus be/lae sp. nov. (F. J. spines of the rhabds of Negombata and Schmitz, D. J. Faulkner, pers. comm.), contain Sigmosceptrella are strongly recurved spines norsesterterpene peroxides. Confirmation of pre­ with kinked tips which eventually become the vious identifications for the specimens whose spines along the shaft. However, where the shaft chemistry is known and has been published, is of Sigmosceptrella becomes sigmoid (Fig. 3B, E) presently being carried out (Kelly-Borges and (Dendy, 1921), those of Negombata remain Mattern, in prep.). straight (Fig. 3K). 500 MEMOIRS OF THE QUEENSLA D MUSEUM

Diacanrus. NegomJxua a ud Sigmosceptrella incubate huge pare11cbymella larva, and embryos 500.Um of diameter have been found in a Mediterranean species of LaJrunculia (Vacelet, 1969), w ile all known hadromerid sponges are oviparous. However, tn ~ontrast to the obvious affinity of the~ genera wtth pocci losclerid spoflgcs which incubate larvae, Diacamr4.s and Negombata. in par­ rku lat, are similar to many had.-omerid genera such as Tethya, Spirastrella and Polymastia, which emphasise heavily collage­ llOus issue, particularly in the cor­ tex. Although there is no clear re.<>o­ Jution ofthe affinities of these gell­ eta. with other demosponges, sever-al patterns have emerged from this study of morpbologi~a1 cll.aracters. Tt 1s quite clear that Dic~eornus, Negombala and AO. 9. A-0. SJreletal arrangements ofSigmo$ceptTelJ.a arld l4trunculia. A, Sigmc~ceptrella are more closely Sigmn,Yreptrella quadrilobata Dendy, hoJotype BMNH 25.1 1.1 .1641 . In- related to each other than they are dian Oocan.JO x. B. Sigmosceptrella sp., Q66C-2779, Sisters Reef, Murray to Lat nmculia, type species of the Rwf~ 3m, Wambro Sound near Rockingham, Perth, Wcstem Australia, family Latrunculiidae . The skele- 320 2J TS 115" 41.3'E, 10 x. C-D, Latrunculia sp.. Q66C-2463, Horse- wns of these three genera, includ- llOI' R(lef Jkm WNW ofMlU'garet Brock lighthouse, 18m. CapeJafta. i ng another latruncu]jid Kutg..oon, Soul.h Australia. 36°56.8'$, 139°35.0'E. C. the skelelnn is a PI){JO pongia. are i11 complete con­ w~ispy reticulation of loose tract of sty I~ . with .a layer of ~t discale trast to that of LaJruncu/ia, all gen­ lll!Crorhahds at surfaceofsponge, 30 x. D, stnglediscorllabd wtlh tenmnal era emphasising a co mp 1ex sptnes embedded beneath 1he ectodennal membrane, above a l

iophonid poecilosclerids such as Zyzzya, as they ardoukobae in the Red Sea and D. bismarckensis share several features such as a tangential surface in the Indo-Pacific, reveal small but consistent layer of megascleres, erect surface spicules (al­ differences through geographic separation. It is though in Zyzzya these spicules are tangential likely that subtle differences will also be found tylotes or strongyles, and acanthoxeas, respec­ for the species indicated in earlier literature to tively), deep brown or green colouration, and in have a Red Sea I Indo-Pacific distribution, espe­ the possession of fistules. It seems likely from cially those genera which have few reliable char­ morphological data, also, that Diacamus, Nego­ acters to used in species separation such as mbata, Sigmosceptrella, and Podospongia, Psammaplysilla, Dysidea, and Hyrtios. It is be­ should not be included in the family coming increasingly obvious that species pre­ Latrunculiidae. Although it is clear that these viously thought to be "cosmopolitan" actually genera are similar in some respects to some consist of cryptic sibling species (Bergquist & hadromerids (collagenous tissue, radial symme­ Kelly-Borges 1991; Hooper et. al., 1992; Kelly­ try in Podospongia, etc.), the vast majority of Borges & Bergquist, 1994; Sole-Cava et al., characters are shared with poecilosclerids (plu­ 1991). mose-reticulate fibre skeleton, chemistry, repro­ duction), suggesting that this group also KEY TO SPECIES OF DIACARNUS constitutes a poecilosclerid family, but separate from the Latrunculiidae, rather than a hadromerid family. However, as there is no clear resolution Ia. Large tubular or spherical sponge with deep of this hypothesis we retain them as incertae sedis wide apical atrium ...... 2 within the Family Latrunculiidae until further 1b. Massive subspherical or digitate sponges, soli­ data is available. Secondary metabolite data and tary or coalescent ...... 3 DNA sequence data acquisition and analysis are lc. Irregular sprawling branches ...... 4 2a. Tube or vase-shaped sponge, up to 80cm long currently in progress towards this end (Kelly­ and 30cm wide. 8-15cm apical atrium, surface Borges,"unpublished data). with lobate projections, extremely abundant den­ Diaet:i"mus is predominantly a shallow-water dritic ectosomal fibres and large irregular genus with a broad latitudinal distribution within spinorhabds, found in Micronesia, common in tropical West Central Pacific marine environ­ the Bismarck Sea from Madang to Kavieng, ments (Fig. 4). Diacamus spinipoculum is very New Ireland, and rarely in southern Papua New Guinea ...... Diacamus tubifera sp. nov. widely and sparsely distributed in temperate 2b. Barrel-shaped sponge, up to 50cm high and south-eastern Australia and tropical Micronesia 40cm diameter, deep, wide, apical atrium, slen­ and Fiji. The longitudinal distribution is however, der megascleres and only very thin microscleres remarkably disjunct, as two species of Diacamus of the smallest category that can be easily over­ also occur in the Red Sea (D. erythraeanus and looked, found in Micronesia, Fiji, and in south D. ardoukobae) with the two known species of Australia. . . Diacamus spinipoculum sp. nov. Negombata. Except in the Red Sea and Gulf of 3a. Spherical sponge, 6-8cm diameter, frequently Aden, Diacamus has not been recorded further joined to adjacent sponges to form a small west than the Philippines despite extensive recent clump, oscules white-rimmed and arranged api­ sampling within the broad Indo-Pacific region. cally in a cluster or serially, slender megascleres and thin microscleres that can be easily over­ This disjunct Red Sea-Indo-Pacific distribution looked, only recorded from Micronesia so far has been indicated for several Indo-Pacific spe­ ...... Diacamus bellae sp. nov. cies of sponge, including well known species 3b. Thick erect digits or lobes, 3-4.5cm in diameter, such as Psammaplysilla purpurea, Hyrtios anastomosing to form a large sprawling mass, os­ erecta, Dysidea herbacea, Phyllospongia cules on ends of digits, slender megascleres and papyracea, and Theone/la swinhoei. Sixty per­ only very thin microscleres of the smallest cate­ cent of the Red Sea records in Uvi (1958) were gory that can be easily overlooked, found in of species known previously only in the Indo-Pa­ New Caledonia and rarely on the central Great Barrier Reef ...... Diacamus levii sp. nov. cific. A number of Red Sea coral reef fishes show 3c. Thickly encrusting to lobate, huge spinorhabds a similar level of differentiation from related spe­ present. Wide, sparse distribution - found in cies found in the Indo-West Pacific (Allen, 1979; southern Papua New Guinea, Philippines and Steene, 1977). Histological examination of the Madang coast of Papua New Guinea ...... obviously very closely related sister-species . . . Diacamus megaspinorhabdosa sp. nov. groups which contain D. erythraeanus in the Red 4a. Large regular spinorhabds present, found on Sea, and D. levii in New Caledonia, and D. Madang coast. and west coast of New Ireland, 502 MEMOIRS OF THE QUEENSLAND MUSEUM

Papua New Guinea ...... Demospongiae: Hadromerida) with descriptions ...... Diacamus bismarckensis sp .nov. of new species from the Southwest Pacific. The 4b. Found in Red Sea ...... 5 Beagle, Records of the Northern Territory Mu­ 5a. Large spinorhabds present ...... seum of Arts and Science, 1991 8:37-72...... Diacarnus ardoulwbae sp. nov. BOCAGE, J. V. B. DU. 1869. Eponges siliceuses nou­ 5b. Large spinorhabds absent ...... velles de Portugal et de I'ile Saint-Iago (archipel ...... Diacamus erythraeanus sp. nov. de Cap vert). Jornal de Scienzas Nationale Acadernie Lisbonne 2: 159-162. ACKNOWLEDGEMENTS BURTON, M. 1934. Sponges. Scientific Report on the Great Barrier Reef Expedition 4( 14 ):513-621. We thank personnel of CRRF, Micronesia, BUTLER, M.S. & R. 1. CAPON. 1991. Norterpene AIMS, Townsville, and ORSTOM, New Caledo­ dienes from an Australian marine sponge Latrunculia brevis. Australian Journal of Chemis­ nia, for collection of samples, and the personnel try 44:77-85. of Motupore Island Research Department, Uni­ 1993. Trunculin F and Contrunculin A and B: novel versity of Papua New Guinea, for field and labo­ oxygenated norterpenes from a southern Austra­ ratory support whilst in Papua New Guinea. We lian marine sponge, Latrunculia conulosa. Aus­ thank Clive Wilkinson for specimens from the tralian Journal of Chemistry 46: 1363-1374. Red Sea, and Mary Kay Harper, SIO, La Jolla, for CARTER, H. J. 1879. On Holostrella, a fossil sponge providing a specimen from Pohnpei Island, Mi­ of the Carboniferous era, and on Hemiasterella, a cronesia. The specimens from Djibouti were col­ new genus of recent sponges. Annals and Maga­ lected during the expedition "Ardoukoba" and zine of Natural History (5)3:141-150. those from Djeddah were collected during a trip 1881. Supplementary report on specimens dredged organized by the Faculty of Marine Sciences in up from the Gulf ofManaar, together with others from the sea in the vicinity of the Basse Rocks Djeddah. We thank Klaus Kelly-Borges (prepa­ and from Bass's Straits respectively, presented to ration of material for histological examination, the Liverpool Museum by Capt. H. Cawne War­ and translation of Keller (1889) from German to ren. Annals and Magazine of Natural History English), Clare Valentine (location of speci­ (5)7:361-385. mens), Peter York and Phil Crabb (micro and DENDY, A. 1897. Catalogue of the non-calcareous macro photography), all of the NHM, London. sponges collected by J. Bracebridge Wilson in the We thank CRRF, Micronesia, for underwater neighbourhood of Port Phillip Heads. Part 3. Pro­ photography of Micronesian, Papua New Guin­ ceedings of the Royal Society of Victoria 2: 230- ean, Fiji and Philippine specimens of Diacamus, 259. and Chantal Bezac and Isabelle Mahieu, for their 1905. Report on the sponges collected by Professor Herdman, at Ceylon, in 1902. Report to the Gov­ help with SEM microphotography. Thanks to ernment of Cey Jon on the Pearl Oyster Fisheries Professor Ralph-Heiko Mattern for discussions of the Gulf ofManaar. 3(Supplernent 18):57 -246. on the chemistry of these sponges. This is Harbor 1921. The tetraxonid sponge-spicule: a study in Branch Oceanographic Institution Contribution evolution. Acta Zoologica 2: 95-152. No. 1090. 1922. Report on the Sigmatotetraxonida collected by the H.M.S."Sealark" in the Indian Ocean: In LITERATURE CITED Report of the Percy Sladen Trust Expedition to the Indian Ocean in 1905, Volume 7. Transac­ ALLEN, G. R. 1979. Butterfly and Angel fishes of the tions of the Linnean Society of London, Zoology World. Volume 2. (Wiley and Sons, New York). 18:1-164. 352pp. HOOPER, J. N. A. 1986. Revision of the marine sponge ALBERICO, M., BRAEK.MAN, J. C., DALOZE, D., genus A.xos Gray (Demospongiae: Axinellida) & TURSCH, B. 1982. Marine Invertebrates. from the north-west Australia. The Beagle, Occa­ XLV. 1be chemistry of three norsesterterpene sional papers of the Northern Territory Museum peroxides from the sponge Sigmosceptrella laevis. of Arts and Science, 1986 3:167-189. Tetrahedron 38: 1881-1890. HOOPER, J. N. A., CAPON, R. J., KEENAN, C. P., BERGQUIST, P.R. 1968. 1be Marine Fauna of New PARRY, D. L. & SMIT, N. 1992. Chemotaxon­ Zealand: Porifera, Demospongiae, Part I omy of Marine Sponges: Families Microcionidae. (Tetractinomorpha and Lithistida) New Zealand Raspailiidae and Axinellidae, and their relation­ Department Scientific and Industrial Research ships with other familes in the orders Bulletin188(New Zealand Oceanographic Institu­ Poecilosclerida and Axinellida (Porifera: tion Memoirs 37) : 105pp. Demospongiae) Invertebrate Taxonomy 6:261- 1978. Sponges. London: Hutchinson: 268pp. 301. BERGQUIST, P.R. & KELLY-BORGES, M. 1991. An KASHMAN, Y., GROWEISS, A., AND SHMUEL!, evaluation of the genus Tethya (Porifera: U. 1980. Latrunculin; a new 2-thiazolidinone REVISION OF DIACARNUS AND NEGOMBATA 503

macrolide from the marine sponge Larruncuiia RIDLEY, S. 0. & DENDY, A. 1886. Preliminary report magnifica. Tetrahedron Letters 21: 3629-3632. on the Monaxonida collected by H.M.S. KELLER, C. 1889. Die Spongienfauna des Roten 'Challenger'. Annals and Magazine of Natural Meeres I. Zeitschrift fi.ir wissenschafliche History 5(18): 325-351,470-93. Zoo1ogie 48: 311-405. 1887. Report on the Monaxonida collected by KELLY-BORGES, M., ROBINSON, E. V., H.M.S. 'Challenger' during the years 1873-1876. GUNASEKERA, S., GUNASEKERA, M., Vol 20: Pp.1-275, In: Report on the Scientific GULAVITA,N. & POMPON!, S. A. 1994. Spe­ Results of the Voyage of the H.M.S. 'Challenger' cies differentiation in the marine sponge genus during the years 1873-1876. (Her Majesty's Sta­ DiscodermU.l (Demospongiae, Lithistida): The tionary Office: London, Edinburgh, Dublin). utility of secondary metabolites as species-spe­ cific markers. Biochemical Systematics and Ecol­ SOEST, R. W. M., VAN 1984. Marine sponges from ogy 22(4):353-365. Curac;:ao and other Caribbean Localities. Part III. KELLY -BORGES, M. & BERGQUIST, P. R. 1994. A Poecilosclerida. Studies on the Fauna of Curac;:ao redescription of Aaptos aaptos (Schmidt), with and other Caribbean Islands 199: 1-167. additions to the genus Aapros in northern New SOLE-CAVA, A.M., KLAUTAU, M., BOURY-ES­ Zealand. Journal of Zoology, London 234: 301- NAULT, N., BOROJEVIC, R. & THORPE, J.P. 323. 1991. Genetic evidence for cryptic speciation in KORNERUP, A. & WANSCHER, J. H. 1961. allopatric populations of two cosmopolitan spe­ Rheinhold Color Atlas. (Reinhold Publishing cies of the calcareous sponge genus Clathrina. Corporation: New York). 224pp. Marine Biology (Berlin) Ill: 381-386. LAUBENFELS, M. W. DE 1936. A discussion of the STEENE, R. C. 1977. Butterfly and Angel Fishes of the sponge fauna of the Dry Tortugas in particular and World. Vol. 1. (John Wiley and Sons: New York). the West Indies in general, with material for the 144pp. revision of the families and orders of the Porifera. THIELE, J. 1900. K.ieselschwamme von Ternate, I. Publications of the Carnegie Institution, Papers Sonderabdruck aus den Abhandlungen der , from Tortugas Laboratory. 30 (467):1-225. Senckenbergischen naturforschenden Gesell­ LEVI, C. 1958. Spongiaires de Mer Rouge recueillis par schaft 25( 1): 19-80. Ia "Calypso" (1951-1952). Resultats scientifiques des Campagnes de Ia "Calypso" 3: 3-46 TOPSENT, E. 1922. Les megascleres polytylotes des 1973. Systematique de Ia classe des Demospongia­ Monaxonides et Ia parente des Latrunculiines. ria (Demosponges). In "Traite de Zoologie. III. Bulletin de l' Institut Oceanographique de Mo­ Spongiaires (P. -P. Grasse, ed)", Masson et Cie naco 415: 1-8. :577-631. VACELET, J. 1969. Eponges de Ia Roche du Large et NEEMAN, I., FISCHELSON, L. & KASHMAN, Y. de l' etage bathyal de Mediterranee (recoltes de Ia 1975. Isolation of a new toxin from sponge Soucoupe plongeante Cousteau et dragages). Latruncuiia magnifica in the Gulf of Aquaba. Memoires du Museum National d'Histoire Marine Biology (Berlin) 30: 293-296. Naturelle, A, Zoologie 59: 145-219.