PROCEEDINGS OF THE 5TH INTERNATIONAL SPONGE SYMPOSIUM ^ d 4 PEZ ORIGIN & OUTLOOK 5th International Sponge Symposium 1996 BRISBANE MEMOIRS OF THE VOLUME 44 30 JUNE 1999 QUEENSLAND MUSEUM PREFACE i 5TH INTERNATIONAL SPONGE SYMPOSIUM, ‘ORIGIN £ OUTLOOK” museum In 1998 the Queensland Museum, Brisbane, hosted the 57/7 International Sponge Symposium, “Origin & Outlook’ (Sth SS) (27 June) July), including two cancurrent pre-Symposium workshops an ‘Systema Porifera” and ‘Karly Evolution ef Spanges and Biomarkers’. and a subsequent post-Symposium field excursion to Heron Island, Capncom-Bunker Group, Great Barrier Reef (5-9 July]. The Brisbane Symposium was the largest of the five international symposia to date. with 165 delegates and accompanying partiers present, from 25 countries. This was also the first occasion thia series of international symposia has been held in the southern hemisphere, with previous meetings hosted (and Proceedings published) by; the Natural Vlistary Museum. London (1968) (Fry, 1970); Muséum National d'Histoire Naturelle, Paris (1978) (Lévi & Boury-Esnault, 1979); National Museum of Natural History, Smithsonian Institution, and Peabody Museum of Natural History, Y ale University (held at Woods Hole Ocesnographic Institute, Massachi- setts, 1985) (Rützler, 1990); and the Zoölogisch Museum, University of Amsterdam (1993) {Van Soest etal., 19941, Over the past decade two smaller international meetings were also held in Berlin (1988), hosted by the Institut für Paläontologie, Freie Universitat (Reitner €: Keupp, 1991), and Lake Biwa, Japan (1996), hosted by a consortium of Japanese researchers lead by Yoko Watanabe (Ochanomizu University, Tokyo) (Watanabe de Fuselani, 1998), Additionally, over the past few decades, there have been several workshops conducted by North American and European workers, producing several substantial volumes of collective papers: Society for De- velopmental Biology, Albany, New York (1975) (Harrisan & Cowden, 1976); Sherkin Island Marine Station, Ire= land (1983) (Jones, 1987); Station Marine d'Endoume, Marseille, France (1986) (Vacelet & Boury-Esnault, 1987); Centre for Advanced Studies, Blanes, Spain (1992) (Uriz & Ratzler, 1993); and Royal Belgian Institute of Natural Sciences, Brussels, Belgium (1995) (Willenz, 1996). Including the present volume, these conference and workshop proceedings contain over 550 refereed papers, signalling an abnormally high productivity and commun- ication amongst sponge workers over a relatively short period, certainly considering the small size ofour group, The theme of the Brisbane Symposium, ‘Origin € Ourlook' — reflected in the motif depicted on the cover ("Morphology to Molecules’) — refers to the adage that scientific progress rests on knowledge of the past, and re- fects the productive interaction between palaeontology, biology, chemistry, ecology, cytology, molecular binl- ogy and other disciplines as multidisciptinary approaches to the strange but innovative world of sponges. Weare, perhaps, fortunate to work ona phylum that is important to the pharmaceutical industry, with its associated politi- cal and economic agendas, but the strength of our collective research lies in its diversity of approaches, opportu nities and collaborative-outcomes. This multidisciplinary synergy has escalated over the past decade: a trend reflected in many of the papers published here. Despite the tyranny of distance; nearly 30% of delegates at the 474 ISS in Brisbane were recent graduates or post-graduate studenls, from many countries, providing optimism for the future of sponge-related research ii an otherwise aging population ofestablished workers, The quality and diversity of their presentations, published here in 71 refereed papers and 69 additional abstracts, is testimony to this optimism. Over the five-day Symposium, 95 scientific papers and approximately 60 posters were presented in 10 sessions: one Session ofinyited papers and nine of gener al contributions. Claude Lévi was invited tà deliver the keynote ad- dress on the central (heme ‘Origin & Outlook '. Three invited plenary speakers were then presented with the chal- lenge to explore this theme through the dimension of time: the Pasr (Frangoise Debrenne), Prevent {Patricia Bergquist) and Faure (Jean Vacelet) of sponges and those who work on them. Each of these keynote and plenary speakers were assigned their general themes, bur ziven ‘carte blanche" on how to approach them. As shown here. their perspectives were very different, but their conclusions were similar. The Origin is sound. We have an exten- sive and firm scientific knowledge-base lo ‘deal’ wilh sponges in our various ways, even if many data are Still missing, we are at least becoming increasingly aware of the ‘right’ sorts of questions we should be asking ofthese dats. And, the Outloak lor sponges and spongers is a bright one. Papers presented during the nine general sessions encompassed a broad range of topics, including: the production of chemicals and the chemical ecology of sponge metabolites with pharmaceutical potential; commercial sponge fisheries and their human impacts; sponge cell behaviour and their immunological implications; the role of sponges as pollution indicators and their ecological interactions with other communities; the role of ‘living fossil" Sponges in coral reef geomorphology; advances in the origins and relationships of Porifera as evidenced by molecular biolowy; recent discoveries in biodiversity, evolution, biogeography and palacontology of the phylum; and the physiology, ultrasiructure and interactions of sponges with symbiotic microbes. 0 MEMUIRS OF THE QUEENSLAND MUSEUM It was originally intended to publish these contributions under several, broad themes (palaeantalogy, systematics, ecology, etc.), but this became nearly impossible lo organise given (hat the boundaries between many disciplines have become blurred through increasing multidisciplinary studies, and this strict arrangement is becoming less relevant Compared to the institutions that have hosted previous symposia. the Queensland Museum has na tradition of sponge research prior to 1991, However, thanks to pioneering efforts of previous generations of scientists work- ing on Australian sponges (1813-1956) — Lamarck, Bowerbank, Haeckel, Selenka, Saville-Ment, Marshall, Carter, Dendy, Ridley, Paléjaeff, Lendenfeld, Kieschnick, Kirkpatrick, Whitelegge, Hentschel, Hallmann, Row, Shaw, Burton, Guller — we know that Australian Waters contain a megadiyerse fauna of more than 1,500 pub- lished (‘valid’) species of sponges (Hooper & Wiedenmayer. 1994; with subsequent updates since 1994), mostly unique. We also now know that more than double this number of species live in these waters (Queensland Mu- seum database), representing about 30% of the world's Known sponge diversity, and of these more than 50% were collected from Queensland waters. H.was therefore appropriate forthe Queensland Museum to host the 524755, as a relatively new, active and expanding proponent for marine sponge biodiversity research and conservation. The 5t 158 provided an important forum to disseminate this new knowledge Lo the international scientific com- munity: in promoting the phylum as an important, productive target For future research; promoting the quality and di versity of our international collaborations, and to facilitate these in the Suture; and demonstrating the substantial amount of work that still remains to be done to achieve even an adequate "basic knowledge’ of this simple, but complex, phylum of animals. Itis greatly anticipated that escalated progress in many fields of sponge science will be revealed at the next sym- posium, to be hosted by the Istituto di Zoologie dell’ Universita, University of Genova, Italy, early in the new millenium. The achievements of the SrA JSS were largely made possible through the generous support of aur sponsors: the Board of the Queensland Museum; Astra Pharmaceutical (Australia) Pty Ltd and the Queensland Pharmaceutical Research Institute, Griffith University; the Commonwealth Bank of Australia; the lan Potter Foundation, and the Queensland Ciovernment Travel Centre. The United States National Science Foundation is also gratefully acknowledged for supporting the attendance of many students to participate in the Sth ISS and in sponsoring a warkshop on ‘Regional research appertunittes and career development in the sponge sciences’ for these students. I would also personally like to thank the following colleagues for providing assistance In organising and running the Symposium, workshops. field trips, providing publicity, documentation, logistic support, and in assisting with the production of this publication: Stephen Cook, Chery) Cook, John Kennedy, Sue List-Armitage, Gert Wórheide, Joachim Reitner, Sally Leys, Nelson Lauzon, Christi Adams, Andrew Mctiown, Derrick Griffin. Jenny Utz, Debra Luk, Adrian Gibb, Paul Aver, Tim Aver, and the Queensland Museum Association, Jahn NA, Hooper, Cunvenor and Editor, Sth [SS, Queensland Museum, Brisbane, Australia; 30 June 1999. LITERATURE CITED FRY, W.G. (ed.) 1970. The biology of the Porifera Symposia of the Zoological Soclety of London. No. 25 (Academic Press: London). HARRISON, FW. & COWDEN, R.R. (eds) 1976. Aspects of sponge biology. (Academic Press: New York, San Francisco, London). HOOPTH, LNA, & WILDEMMAYERR, T, 1994. Porifera. Pp, 1-624. In Wells, A. ied) Zoologicul Catalugue ot Australia, Vol. (2 (CSIRO Australia; Melboume). JONES, WC, (ed,) 1987, European contributions lo llc taxonomy of sponges. (T.ithio Press: Middleton), LEVI, €. & BOLIR F-ESNAUL I, N. (ods) 1979. Biologie des spongsgires. Sponee hioleey. Colloques Internationaux dij Centre National de ly Recherche Scientifique. No. 291 (CNRS: Paris), REITNER. J. & KEVPP, H. (eds) 1991 Fossil and Recent sponges (Sprmyer-Verlag: Berlin, Heidelberg, New York). RITZLER, K. (ed.) 1990, New perspectives in sponge hjulagy. (Smithsonian Instituton Press: Washington DOC.. London). SOES!, K.W.M. VAN, KEMPEN, TMG. VAN & BRAEKMAN, J-i (eds) 1944. Sponges In lime and space. Biology, chemistry, palognialagy, (AA, Balkema: Rotrerdam, Hrouktieldi. URIZ MJ. & RUTZLER, K (eds) 1993. Recent advanves itt ecology and systematics of sponges. Scientia Marina 57(4): 273-432, VACELET, J. & BOURY-ESNAULT, M. (eds) 1987. taxonomy of Poritera. NATO AST Workshop Proceed- ings, 513 (Springer-Verlag. Berlin, Heidelberg), WATANABE. Y & FUSETANIE N. (eds) 1998. Sponge sciences, Multidisciplinary perspectives (Springer- Verlag Tokyo, Berlin, Heidelberg, New York), WILLENZ, Ph. (cd.) 1996. Recent advances in sponge biodiversiry inventory and documentation, Bulletin de CEnstitut Royal des Sciences Naturelles de Helgnyue, Biologie 66 (suppi. 1-242. CONTENTS iii Preface. HOOPER, J.N.A. Ath International Sponge Symposium, ‘Origin & Outlook?.........cciiiiie lisse iu Invited Papers. LEVI, C. Sponge science, from origin to outlook., ... 2l. 2 eee etter eee l DEBRENNE, F. The past of sponges, sponges ofthe past... i.i ees ulii ii at 9 BERGQUIST, P.R. The present state of sponge SCIENCE... ooo o oococon rro cor 23 VACELET, J, Outlook to the future of sponges. ....¿0o0oororoorooraco ss. ela pote 27 General Papers. ADAMS, C.L., McINERNEY, J.O. & KELLY, M, Indications of relationships between poriferan classes using full-length 18s rRNA gene sequenries,ir..i.. ports rre Eet ate ra tirita bicis pudes + JOE E BAKUS, G.J. & NISHIYAMA, G.K. Sponge distribution and coral reef community structure off Mactan Island, Cebu, Philippines. otis o Hipa sie ON 45 BELL, A.H., BERGQUIST, P.R. & BATTERSHILL, C.N. Feeding biology of Polymastia croceus, ... c.c eessese me 51 BERGQUIST, P.R., SOROKIN, S. & KARUSO, P. Pushing the boundaries: à new genus and species of Dictyoceratida........0..0..... 57 BURJA, A.M., WEBSTER, N.S., MURPHY, P.T. & HILL, R.T. Microbial Symbionts of Great Barrier Reef Sponges......... llis nro 63 CALCINAI, B., CERRANO, C., BAVESTRELLO, G. & SARA, M. Biology of the massive symbiotioc sponge Cliona nigricans (Porifera: Demospongiae) in the Ligurian Sea. 2... eee m rh 77 CERRANO, C., BAVESTRELLO, G., BENATTI, U,, CATTANEO-VIETTI, R., GIOVINE, M. & SARA, M. Incorporation of inorganic matter in Chandrosia reniformis (Porifera: Demospongiae): the role of water turbulence. ....+......<.1.- A LAO ete tet 85 COLBY, A.C.C., FROST, T.M, & FISCHER, J.M. Sponge distribution and lake chemistry in northern Wisconsin lakes: Minna fewell's survey revisited. ... 2... 1oo.ooo..1.2 nn! TE TRAE A 93 CRISTOBO, T.J., RIOS, P. & URGORRI, V. Remarks on the status of Myxilla (Porifera: Poecilosclerida) on the Galician coast (NW Iberian Peninsula)....... Il o o A o UNER md Sor n= Al 101 DAVIS, A.R. £ WARD, D.W. Does the large barnacle Austrobalanus imperator (Darwin, 1854) structure benthic invertebrate communities in SE Australia? ....ooococcccooococorocc o eee 125 DESQUEYROUX-FAUNDEZ, R. Convenient genera or phylogenetic genera ? Evidence from Callyspongiidae and Niphatidae {Haplosclerida}. 20... 0... eee ee eee teens 131 DE VOOGD, N.J., SOEST, R.W.M. VAN & HOEKSEMA, B.W. Cross-shelf distribution of southwest Sulawesi recf sponges... 0.560. lise eee eee 147 DUCK WORTH, A.R., BATTERSHILL, C.N, SCHIEL, D.R, & BERGQUIST, P.R, Farming sponges for the production of bioactive metabolites. ...........0......2. 155 EVANS-ILLIDGE, E.A., BOURNE, D.J., WOLFF, C.W.W. & VASILESCU, I.M. A preliminary assessment of ‘space wars’ as a determining factor in the production of novel bioactive indoles by Jotrochata sp... oe ees 161 FAULKNER, D.J.. HARPER, M.K., SALOMON, C.E. & SCHMIDT, E.W. Localisation of bioactive metabolites in marine sponges........... 0260.6 208 eee 167 FROMONT, J. - Demosponges of the Houtman Abrolhos. ...:1.0cooocooocccionrr orcas 175 iv MEMOIRS OF THE QUEENSLAND MUSEUM FROMONT, J.. Reproduction of some demosponges in a temperate Australian shallow water habitat. . 185 FUERST,.LA., WEBB, R 1., GARSON, MLJ., HARDY, L. & REISWIG, H.M. Membrane-bounded nuclear bodies in a diverse range of microbial symbionts of Great Barrier Reef SPOñgeS.. a... nseuren esri ient mee 193 GARSON, M.J., CLARK, R.J., WEBB, R.I., FIELD, K.L., CHARAN, R.D. & MCCAFFREY, E.J. Ecological role of cytotoxic alkaloids: Haliclona sp. nov., an unusual sponge/ dinoflagellate association... 002. esl e 205 GUGEL, J. Ecological adaptions of a freshwater sponge association in the River Rhine, Germany (Porifera: Spongillidae). -usapin ce ce ee eem a m arms asas 215 HAJDU, E. Toward a phylogenetic classification of the Mycalids with anisochelae (Demospongiae; Poecilosclerida), and comments on the status of Naviculina Gray, 1867,....... 225 HILL, M.S. Morphological and genetic examination of phenotypic variability in the tropical sponge Anthosigmella varians. 6.0.0 ce hee e be eae 239 HOOPER, J.N.A., LIST-ARMITAGE, S.E., KENNEDY, J.A., COOK, S.D. & VALENTINE, CA. Sponges of the Low Isles, Great Barrier Reef: an important scientific site, or a case of mistaken identity ?..... 2.02... 6c E eo eek ee eee eee hr aca t: 249 HOOPER, J.N.A., KENNEDY, J.A, LIST-ARMITAGE, S.E., COOK, S.D. & QUINN, R. Biodiversity, species composition and distribution of marine sponges in northeast 4 AE Leu, ule A A AS A 26 ITSKOVICH, V.B., BELIKOV, S.L, EFREMOVA, S.M. & MASUDA, Y. Phylogenetic relationships between Lubomirskiidae, Spongillidae and some marine sponges according partial sequences of 188 rDNA, ooo corcr oo 275 KONIG, G.M. & WRIGHT, A.D. Cymbastela hooperi and Amphimedon terpenensis: where do they really belong? .... 281 KUBLER, B. & BARTHEL, D. A carnivorous sponge, Chondrocladia gigantea Pantene: es the ae acer sea clubsponge from the Norwegian Trench. . mt rent ....289 LAZOSKI, C., PEIXINHO, S., RUSSO, C. A.M. & SOLÉ-CAVA, A. M. Genetic confirmation of the specific status of two sponges of the genus Cinachyrella (Porifera: Demospongiae: Spirophorida) in the Southwest Atlantic. .......... 299 LEHNERT, H. & FISCHER, H, Distribution patterns of sponges and corals dawn to 107m off North Jamaica. ....... 307 LÓBO-HAJDU, G., MANSURE, J.J., SALGADO, A., HAJDU, E., MURICY, G. de ALBANO, R.M. Random amplified polymorphic DNA (RAPD) analysis can reveal intraspecific evolutionary patterns in Porifera. 2.02. eee 317 LOPEZ, J,V., McCARTHY, PJ., JANDA, R.E., WILLOUGHBY, R, & POMPONI, S.A, Molecular techniques Teveal wide phyletic diversity of heterotrophic microbes associated with Discodermia spp. (Porifera:Demospongiae). ........ sese 329 MCINERNEY, J.O,, ADAMS, C.L, & KELLY, M, Phylogenetic resolution UN CHA of 18s and 28s rRNA \ genes within the lithistid Astrophorida.. . n roe ae a ela sa ele stele E: MALDONADO, M. & URIZ, MI. A new dendroceratid sponge with reticulate skeleton. 2.0.0... 0c cere ese eee eee 353 MANCONI, R., CUBEDDU, T. & PRONZATO, R. African freshwater sponges: Makedia tanensis gen. et sp. nov. from Lake Tana, Elo tr A f $334 kai ya... 361 MOTHES, B., LERNER, C.B. & SILVA, CMM. DA Revision of Brazilian Erylus (Porifera: Astrophorida: Demospongiae) with description ofanew species... ec taad ee eee ce aaa cea ere cee it ra ras ls 369 CONTENTS y MULLER, W.E.G. & MÜLLER, I.M. Origin of the Metazoa: A review of molecular biological studies with sponges. ...... 381 MURICY, G. An evaluation of morphological and cytological data sets for the phylogeny of Homosclerophorida (Porifera: Demospongiae). ....,...,.-.--.. 1d 399 NISHIYAMA, G.K. & BAKUS, GJ. Release of ailelochemicals by three tropical sponges MAR ci de and their toxic effects on coral substrate competitors. |... isse aen All OSINGA, R., REDEKER, D., DE BEUKELAER, P.B. & WIJFFELS, R.H. Measurement of sponge growth by projected body area and underwater weight. ..... 419 PANSINI, M., CATTANEO-VIETTI, R, & SCHIAPARELLI, S. Relationship between sponges and a taxon of obligatory inquilines: the siliguariid molluscs,.... i asa corr rra rr ates telat at erase te et 427 PATTANAYAK, J.G. Annotated checklist of marine sponges of the Indian region........ oo... .... 439 PILE, A.J. Resource partitioning by Ci Caribbean coral reef sponges: is there Misc food for everyone? a etla doie goen usse nares AA R E ol HP, PISERA, A. PostPaleozoic history of the siliceous sponges with rigid skeleton. ...... 01... ..... 463 PISERA, A, Lithistid sponge Setidium obtectum Schmidt, 1879, rediscoveted.............-.... 473 PITCHER, C.R,, WASSENBERG, T.L, SMITH, G.P., CAPPO, M., HOOPER, J.N.A, & DOHERTY, P.J. Innovative new methods for measuring the natural dynamics of some rubei dominant tropical sponges and other sessile fauna.,,........-5) 04s 000s ees 479 PRONZATO, R., BAVESTRELLO, G., CERRANO, C., MAGNINO, G.. MANC ONI, R. PANTELIS, J., SARA, A. & SIDRI, M. Sponge farming in the Mediterranean Sea: new Perspectives. asp paineis: iiM: 485 ROBERTS, D.E., CUMMINS, S.P., DAVIS, A.R. & PANGWAY, C. Evidence for symbiotic algae in sponges from temperate coastal reefs in New South Wales,-Austrilia..: 2.5. scade dad cate e llo oe Ad e E es tub lets 493 REISWIG, H.M. New hexactinellid sponges from the Mendocino Ridge, Northern California, USA... . 499 RICHELLE-MAURER, E. & VAN DE VY VER, G. Expression of homeobox-containing genes in freshwater sponges. ......; pas 509 SAMAAI, T., GIBBONS, M.J. & KELLY, M. Morphological phylogenetic considerations on the relationships of /sodietya Bowerbank, ARAS rhegi crias A pas ase ssa T, SANDERS, M., DIAZ, M.C. & CREWS, P. Taxonomic evaluation of jasplakinolide- -containing sponges of the family Coppatiidae. 525 SCHÖNBERG, C.H.L. An improved method of tissue digestion for $picule mounts in Sponge taxonomy...... 533 SCHUPP, P., EDER, C., PAUL, V. & PROKSCH, P, Chemistry, ecology and biological activity of the haplosclerid sponge Oceanapia sp.; Can ecological observations and experiments give a first clue about pharmacological activity le res Sar I eee erecta rar tent mln 541 SIM, C.J. & LEE, K.J. Relationship of sand and fibre in the horny sponge, Psammrocinia. 2... 4.4 2.25.55 6. 55] SIMPSON, 3.5. & GARSON, M.J. Cyanide and thiocyanate-based biosynthesis in tropical marine sponges. ...-......- 559 SOEST, R.W.M. VAN & BRAEKMAN, J.-C. Chemosystematics of Porifera: a review. 2.2... ce cee ee menn 569 SOLE-CAVA, A, M, & BOURY-ESNAULT, N. Patterns of intra and interspecific genetic divergence in marine sponges. ........... 591 vi MEMOIRS OF THE QUEENSLAND MUSEUM TABACHNICK, KR, Abolishment of the family Caulophacidae (Porifera: Hexactinellida). ............. 603 TABACHNICK, K.R. & MENSHENINA, L.L. An approach to the phylogenetic reconstruction of Amphidiscophora (Porifera: UM AMA AA lot) ape el fe dE Rue Mmm AS 607 TURON, X., URIZ, MJ. & WILLENZ, Ph, Cuticular linings and remodelisation processes in Crambe crambe (Demospongiae: Pageilosaleiday, 2.24 topic. cx ae le lo aye o et Ine 617 VACELET, J, Planktonic armoured propagules of the excavating sponge Alectona (Porifera: Demospongiae) are larvae: evidence from Alectona wallichii and A. mesatlantica S. TQVe ra beware ey iod ERa ae als Cet elPeus (4 Sore ees bas 12... 627 VOLKMER-RIBEIRO, C., CORREIA, MME. BRENHA, S.L.A. & MENDONCA, M. A. Freshwater : sponges from a Neotropical sand dune area, 2... 20-2 ee eye ee eee ee eae 643 WEINBERG, E., ECKERT, C., MEHL, D., MUELLER, J., MASUDA, Y. & EFREMOVA, S. Extant and fossil spongiofauna from the underwater Akademician Ridge of Lake Baikal (SE GiB). in Ian set te Mbt ete trad tl etui .. 651 WEYRER, 5., RÜTZLER, K. & RIEGER, R. Serotonin in Porifera? Evidence from developing Tedania ignis, the Caribbean fire sponge (Demospongiae) 20.0... ce hme aere 659 WILKINSON, C.R., SUMMONS, R. & EVANS, E. Nitrogen fixation in symbiotic marine sponges: ecological significance and difficulties in detection A c ped iiiter v bederkidcrbepenisggég ibis 667 WILLENZ, Ph. & HARTMAN, W.D, Growth and regeneration rates of the calcareous skeleton of the Caribbean coralline sponge Ceratoporella nicholsoni: a long term Survey. a...se cesse 675 ZEA, S,, PARRA, F.J., MARTÍNEZ, A, & DUQUE, C. Production of bioactive furanosesterterpene telronic acids as possible internal chemical defense mechanism in the sponge /rcinia felix (Porifera: Demospongiae). ..... 687 Subject index Abstracts, ALVAREZ, B., CRISP, M.D., DRIVER, F., HOOPER, J.N.A & SOEST, R.W.M. VAN Approach to the phylogeny of Axinellidae (Porifera: Demospongiae) using morphological and molecular data. ...... Pygti--i:iij EI Piar haus ced 11.43 ANAKINA, R, P. Peculiarities of fertilization process in the sponge Leucosolenia complicata Montag (Calcispongiae: Calcaronea) from the Barents Sea... ooo o oco ooo 44 AUSTIN, W.C. The relationship of silicate levels to the shallaw water distribution of hexactinellids in British Columbia. .....- nuda da tasa oo tes Cab kobe: sea 44 BATTERSHILL, C.N. & ABRAHAM, R. Sponges, indicators of marine environmental health.. pha za] .-..50 BATTERSHILL, C.N., PAGE, M.J., DUCKWORTH, A.R., MILLER, K. A. BERGQUIST, P.R., BLUNT J. W., MUNRO, MH. G., NORTHCOTE, P, T. NEWMAN, DJ.& POMPONI, S.A, Discovery and sustainable supply of marine natural products as drugs, industrial compounds and agrochemicals: chemical ecology, genetics, aquaculture and cell A A A o E en ted ENE O 76 BERGBAUER, M., LANGE, R., SZEWZYK, U. & REITNER, J. Characterization of calcium-binding matrix proteins from distinct coralline ii ec pele 4 me Plueja olotito iubente RA 76 BOHM, F., DULLO, W.-C,, EISENHAUER, A., LEHNERT, H., WORHEIDE, G., REITNER, J. & JOACHIMSKI, M.M. Carbon isotope time series of coralline sponges from the Coral Sea, Philippines and Caribbean.,,,.......-.-,-- AMD ARTS (UA e rear Ea rie prine E B4 CONTENTS vil BOHM, F., DULLO, W.-C., EISENHAUER, A., JOACHIMSKI, M.M., LEHNERT, H. & REITNER, J. Carbon isotope history of Caribbean surface waters revealed by coralline sponges... . . . 91 BOND, €. Time-lapse studies of sponge motility and anatomical rearrangements. ........-..... 91 CHANAS, B. & PAWLIK, J.R. Da Caribbean sponges have physical defenses ? .... 0.2.0.6. . 20 cece eese 92 CHOMBARD, C. & BOURY-ESNAULT, N. Good congruence between morphology and molecular phylogeny of Hadromerida, or how to bother sponge taxonomists. .-. 2.2.0.0. cece eee m eee ,, 100 COOK, A, Afi overview of stromatoporoid dominated Middle Devonian reef complexes in norih Queensland... lera seca denis tut abe rete bust ants te 99 DAVY, SK. & HINDE, R. Nitrogen flux in a sponge - macroalgal symbiosis. .,.......55.0. 202.0 eiiis 124 DIAZ, M.C. & WARD, B.B. Perspectives on sponge-cyanobacterial symbioses. ~.. <... 2.222. ep ee ee ees 154 DUNLAP, M.J. & PAWLIK, LR. Polly want a sponge? : Field examination of spongivory by Caribbean parrotfishes in reef and mangrove habitats. ............. 00.2 esse e 160 ERESKOVSKY, A.V. & GONOBOBLEVA, E.L. Development of Halisarca dujardini Johnston 1842 (Porifera; Ceractinomorpha, Halisarcida) from egg to free larva... ases ases re 160 FALLON, S.J, McCULLOCH, M.T. & HOOPER, J.N.A. Trace element and stable isotope profiles from the coralline sponge (Astrosclera NUERA Pete dys bt riunioni ana Pap ett us 174 FERNÁNDEZ-BUSQUETS, X. € BURGER, M.M. Sponge cell adhesion: an evolutionary ancestor of histocompatibility systems ? .,..., 184 GRANT, A.J, HINDE, R.T. & BOROWITZKA, M.A. Evidence of transter of photosynthate from a red algal macrophyte to its symbiotic SABES o s an eat et dal a ce tila ie Pop nics a Le 204 GUERRAZZI, M.C., HAJDU, E., MORGADO DO AMARAL, E.H, & DUARTE, L.F.L. Spongivory by the Brazilian starfish Echinaster brasiliensis. |... 2. aol uus 214 HINDE, R., PIRONET, F. £ BOROWITZKA, M.A. Photosy nthesis and respiration of the A ciatis sponge, Dysidea herbaceq. ..... coro darla sela erha eiat n eh sleng peleme enbet 238 HUMPHREYS, T. Regulatory mechanisms of immune cells in sponges. .... 22. lesse 248 ILAN, M. Negombata magnifica - a magnificent (chemical) pet..........00 20.0.0... ce eee eee 248 IVANOVA, L.V, New data about morphology and feeding patterns of Barentz Sea Halichondria panicea (Pallas)a: 7. opi din ap iade aeu cd ob ohm oh bee EM AS 262 KELLY, M., McCORMACK, G.P. & McINERNEY, J.O. Morphology and molecules in lithistid taxonomy: new solutions for old problems..... 274 KNOTT, N. The replacement of natural hard substrata by artificial substrata: its effects on sponges and ascidians; sires o bet odes p daa ledit peed ead elet es bidbstesdiabes 288 KNOWLTON, A.L, & HIGHSMITH, R.C. Convergence in the time-space continuum: a predator-prey interuction.........-.... 288 KRAUTTER, M. l Remarks on the paleoecology and réef building potential of late Jurassic siliceous SDOUESS. 3. rag e o 4er erdt red Jer d e te Tuus 297 KRUSE, P.D. Distinctive Middle Cambrian sponge-calcimicrobe reefs in Iran. iiis 298 LARROUX, C. & DEGNAN, B.M. Homeobox genes expressed in the adult and reaggregating sponge. ................ 306 vill MEMOIRS OF THE QUEENSLAND MUSEUM LERNER, C., MOTHES, B., POSSUELQ, L.G., COSTA, J.C.. SCHAPOVAL, E.E.S., RECH, E., FARIAS, F.M,, HENRIQUES, A,T, & MANS, D, Antimicrobial activity of sponges from southern Brazil, Atlantic coast.............. 306 LEYS, S.P. & MACKIE, G.O, Propagated electrical impulses ina sponge... 1.2... 04. e eee eee eee ees 3M LL, S., BLUNT, J.W., DUMDEI, E. J., MUNRO, M.H.G., PANNELL, L.K. & SHIGEMATSU, N. Theonellapeptolides from the deep-water New Zealand sponge Lamellomorpha E TE NN Bem A ie rm dnt ertet n 2.2. 342 McCORMACK, G.P., McINERNEY, J.0, & KELLY, M. The phylogenetic position ‘of the sponge Spongasorités suberitoides determined by analysis of 288 rRNA gene sequence... 0.2... 062. cee sa 352 McINERNEY, J.O, & KELLY, M. Phylogeny of lithistid SpONges,.....o.o.oooomsricricr terior RR dad 352 McKENNA, S.A. & RITTER, J. Cliona lampa and disturbance on the coral reefs of Castle Harbour, Bermuda. ..-....360 MARTI, R., URIZ, M.J., BALLESTEROS, E. & TURON, X. Spatial and temporal variation of the natural toxicity in sponges of à Mediterranean cave: is thére-2 trend 7-. ui lez alle» alle JI en ell al ole d te le rh p ea e 360 MASUDA, Y., ITSKOVICH, V. B., VEINBERG, E. V. & EFREMOVA, S. M. Study on the distribution of Baikalian sponges........... seis ie 368 MATSUNO, A., KURODA, M. & MASUDA, Y. An ultrastructural study on the contractile pinacocyte of a freshwater sponge..,..... . 398 MEHL, D. The phylogenetic history of sponges in Palaeozoic times...--........--.---, os 80 MEHL, D. Towards à phylogenetic systematics of the fossil Hexactinellida. .......-....1..1.- 418 NEWBOLD, R.W., PAWLIK, J.R., JENSEN, P. & FENICAL, W. Antimicrobial activity of Caribbean sponge extracts. ......-.-..2..6.2.006 ~~ 438 PAWLIK, LR. Predation on Caribbean sponges; the importance of chemical defenses..,...,....... 426 PITCHER, C.R., BURRIDGE, C.Y., WASSENBERG, TJ. & SMITH, G.P. The impact of trawling on some tropical sponges and other sessile fauna. ....... v1.0. 455 PRONZATO, R., SIDRI, M., DORCIER, M., CAPPELLO, M. & MANCONI, R. Sponge shape as a taxonomic character: the case of Spongia officinalis and Spongia agaricing. a oi a t d> poima s iei E VOIE IRAN PAP ..456 REITNER, J., WORHEIDE, G. & HOOPER, INA. "Mud mound’ structures and coralline sponges from Osprey Reef (Queensland Plateau, Coral Sea, Australia). 130 50ocoo0coooraria casse ssa suas petietiaid 462 REITNER, J., WORHEIDE, G., ARP, G., REIMER, A. & HOOPER, J.N.A, An unusual suberitid demosponge from a marine alkaline crater lake (Satonda Island, Indonegsig) ota liuuiz sw A TT 477 REITNER, J., WÓRHEIDE, G., LANGE, R., THIEL, V., EISENHAUER, A., REIMER, A., FLIEGE, S. & BERGBAUER, M. New approaches to the biomineralization processes of calcified skeletons in coralline deimgspóngési. a seu ele pel aaa eser kd eria 492 REITNER, J.. WORIIEIDE, G. & HOOPER, J.N.À, New colonial Vaceletia-Lype sphinctozoan from the Pacific.. .......oooooooooooo.. 498 REITNER, J., SCHUMANN-KINDEL, G. & THIEL, V. Origin and early fossil record of sponges-a geobiological approach. ....... lisse 515 REITNER, J., NEUWEILER, F., SCHUMANN-KINDEL, G. & THIEL, V. Taphonomy and preservation potential of sponge tissue... ,.. 02.20.02. 005 5 MEETER 516 RITTER, F. The sponges of Paraiso nearshore fringe reef, Cozumel, Mexico....... IA 508 CONTENTS ix ROUSH, S.A. Freshwater sponges (Porifera: Spongillidae) of the Guanacaste Conservation Area, Costa Rica: a preliminaty survey... 00.0... e scene ta 540 SATOH, Y., FUJIMOTO, Y., YAMAMOTO, A, & KAMISHIMA, Y. Effects of photosynthetic activity in endosymbiotic zoochlorellae on gemmule germination of a freshwater sponge, Radiospongilla cerebellata....,.....--.. 524 SCHMAHI., G.P. Recovery and growth of the giant barrel sponge (Xesfospongia muta) following Py injury from a vessel grounding in the Florida Keys......... csse 532 SCHÓNBERG, C.H.L. The infectiveness of a bioeroding clionid sponge. «.......si eh 540 SCHÖNBERG, C.H... Wanted: the names of common bioeroding sponges of the central Great Barrier Reef. . 540 SCHUMANN-KINDEL, G., BERGBAUER, M., MANZ, W., SZEWZYK, U, & REITNER, J. Microbial influenced pyritisation of marine sponges. .........-..6.0 sese 550 SCHUMANN-KINDEL. G., BERGBAUER, M., MANZ, W.,SZEWZYK, U.& REITNER, J. Chondrosia reniformis, hitherto unknown bacteria. si 22. eee 558 SEMENOV, V.V. & IVANOVA, L.V. Regeneration abilities of spongillid larvae. ... 2.02.2... 02.0 ce eeepc 568 SOEST,R.W.M. VAN, VAN DE VY VER, G., RICHELLE-MAURER, E., WOLDRINGH, C., BRAEKMAN, J.-C. & TAVARES, R. Biology of sponge natural products,.....-..-) 0.42.2 pee dees es eee cee nns 590 TRAUTMAN, D.A. Population dynamics of a sponge/macroalgal symbiosis: possible causes for a patchy distribution at One Tree Reef, Great Barrier Reef... ..ooooomomooo poor... 602 TRAUTMAN, D.A. Photosynthesis and respiration by the symbiotic association between a coral reef sponge and its macroalgal symbiont..,.., ..ooooomocrooneraran eee eee e 606 URIZ, M.-J., MALDONADO, M., MARTÍ, R. & TURON, X. The role of early life-history stages in determining adult spatial patterns of encrusting PONLE ide chee yi tra A er htc Ea 616 WHEELER, B. Phylogenetics of the euretids (Euretidae: Hexactinellida).........)....-2.6022200. 626 WORHEIDE, G, & REITNER, J. Biogeography and taxonomy of the reef cave dwelling coralline demosponge Astrosclera willeyana throughout the Indo-Pacific... 0... anaona eer eee ee eee 650 WORHEIDE, G. & REITNER, J. Climatic changes of the last 450 years recorded in the skeleton of the coralline demosponge Astrosclera Willeyand. .....0 0.0.66 esses 658 WORHEIDE, G. & REITNER, J. Biocalcification in the Indo-Pacific coralline demosponge Astrosclera willeyana Lister - the role of basopinacoderm.. ..... 6-02. ccc ee eee tee e ee eee 666 WULFF, J,L. Rapid change and stasis in a coral reef sponge community. .. 0,-1.6... 050 saeeeae 674 WULFF, J.L. A sponge that cheats on diffuse mutualism among other sponge species. ...... AU 686 Left to right. 1" row: Alex Cook, Clare Valentine, Michelle Kelly, Jean Vacelet, CliveWilkinson, Frangoise Debrenne, Catherine Chombard, Welton Lee; 2" row: Matthias Bergbauer, Stephen Cook, Cheryl Cook, Stewart Fallon, Jerry Bakus, Pierre Kruse; 3' row: Gert Wórheide, John Kennedy, Sue List-Armitage, Adam Burja, Mary Wakeford, Libby Evans-Illidge, Shirley Sorokin, Carsten Wolff, Jane Fromont, Guilherme Muricy; 4" row: Anthony Wright, Christine Schénberg, Adrienne Grant, Donelle Trautman, Rosalind Hinde, Michael Borowitzka, Yoshihiro Fujimoto, Yoshihisa Kamishima, Jason Ritter, Malcolm Hill; 5% row: Kathleen Smith, Lisa Goudie, Ann Knowlton, Yuriko Sato, Yoshiki Masuda, Andrew Flowers, Nicole de Voogd, Jose Lopez, Nathan Knott; 6" row: Michaël Manuel, Murray Munro, Renata Manconi, Carlo Cerrano, Maurizio Pansini, Roberto Pronzato. Absent from photo: Carla da Silva, Clea ‘Lerner, Ron Quinn, Andy Davis, Peter Steinberg, Tony Carroll, Bernie Degnan, Peter Murphy, Nicole Webster, David Miller, Roland Pitcher. WndsnW GNVISNAANO AHL JO SAJONIA Left to right. 1% row: Maria Uriz, John Hooper, Nicole Boury-Esnault, Bill Austin, Janie Wulff, Patricia Bergquist, Sally Leys, Giséle Van de Vyver, Evelyn Richelle-Maurer, Ruth Martí Lluch; 2™ row: Ernie Kovacs, Michele Sarà, Claude Lévi, Steve de C. Cook, Belinda Alvarez de Glasby, Ruth Desqueyroux-Faúndez, Benjamin Wheeler, Bettina Kiibler, John Fuerst; 3" row: Eduardo Hajdu, Gisele Lóbo-Hajdu, Phillipe Willenz, Henry Reiswig, Monica Puyana, Rochelle Newbold, Matt Dunlap, Mary Garson, Alison Colby, Scott Roush, Scott Nichols, Cécile Debitus, Rick Webb; 4" row: Manuel Maldonado, Brian Chanas, Micha Ilan, Joseph Pawlik, Chung-Ja Sim, Yoko Watanabe, George Schmahl, Christi Adams, Adele Pile, Sheila McKenna, Russell Hill; 5" row: Alan Duckworth, Ronald Osinga, Peter Schupp, Michael Assman, Mary Kay Harper, John Faulkner, Calhoun Bond, Jennifer Carroll, Miranda Sanders, Cristina Diaz, Junichi Tanaka, Joachim Reitner, Klaus Ruetzler, Sven Zea, Shirley Pomponi; 6% row: Gregory Nishiyama, Herbert Peveling, Dorte Mehl, Grace McCormack, James McInerney, Toufiek Samaai, Andrzej Pisera, Manfred Krautter, Helmut Lehnert, Theo Eugeser, Cecilia Volkmer Ribeiro, Xavier Fernandez-Busquets, Jochen Gugel. SINVdIDLLaVd IX KEYNOTE ADDRESS SPONGE SCIENCE, FROM ORIGIN TO OUTLOOK CLAUDE LEVI lan Potter Foundation Keynote Address XL a Lévi, C. 1999 06 30: Sponge science, from Origin to Outlook. Memoirs of the Queensland Museum 44: 1-7. Brisbane. ISSN 0079-8835. Much ofthe important data, and most ofthe progress made on sponge biology has come from careful in vivo and in vitro studies on living populations. These techniques were used in studies conducted over a century ago, but much of this early work has been overlooked, or distorted during its transmission to the present time. This review revisits the scientific philosophy and techniques of our predecessors. It evaluates the quality of their observations, experimental prowess and originality in thought, and highlights the pivotal discoveries that have produced our present concept of *what is a sponge', underlining those fields of study, such as developmental biology, where information is incomplete or lacking altogether. O Porifera, historical review, sponge biology, animality. Claude Lévi (email; leviggmnhn.fr), Laboratoire de Biologie des Invertébrés Marins et Malacologie, Muséum National d'Histoire Naturelle 57, Rue Cuvier, 75013 Paris, Cédex í THE IAN POTTER. FOUNDATION 05, France; 19 October 1998. In 1948, as I began my career in science, | asked some of my mentors in sponge biology, Emile Topsent, Odette Tuzet, Henriette Meewis and Paul Brien, whether or not working on sponges, particularly their developmental biol- ogy, was still a worthwhile pursuit for a young scientist. Now, 50 years later, despite the tremendous progress we have made in our under- standing of the general and molecular biology of sponges, it is clear to me that we need much more work to better understand development, growth and morphogenesis of sponges. Scientific knowledge is a human collective and diachronic phenomenon, with each ofus adding our own observations, thoughts and experiences (sometimes orally transmitted, but mostly in written form and as artwork), to the accumulated body of information from the past. Reading ancient documents is often amusing for their apparent naive contents, but often the answer is not obvious without many years of experiment- ation and observation. Moreover, is the answer definitive? For this plenary address on the ‘Origin & Outlook’ of sponge biology, it is necessary to revisit the scientific philosophy and techniques available to our predecessors, in order to evaluate the quality of their observations, their experi- mental prowess and originality of thought. Indeed, personal reading of the older literature is invaluable; one can discover observations which have been forgotten or distorted by their transm- ission and subsequent interpretation, and one can also find hypotheses which more contemporary experiments subsequently invalidated or confirmed. It is essential to read original docum- ents in order to better understand the great debates on ‘animality’, individuality, diblasty and inversion of layers, origin of the phylum, origin of bathyal and abyssal fauna, cellular dif- ferentiation and re-differentiation, internal transmission of information within this multi- cellular organism, self-not-self recognition, and so forth. Studying living populations of marine animals is difficult at best, but unequivocally in vivo and in vitro observations on sponges over the centuries have contributed most to what we know about the phylum today. It was no accident that sponge science began somewhere in the eastern Mediterranean during an earlier millenium, in the province of sponge fishers, subsequently reach- ing the western Mediterranean, then the French coasts, and then the British Isles at the end of the 18th century. Discoveries made from field ob- servations have had a substantial influence on our collective thinking about the Porifera, and in some cases these discoveries have changed our perception of the phylum completely. Two such extraordinary events have occurred since the previous conference (4th International Porifera Congress, University of Amsterdam, 1993), both widely reported by the international media: |) the existence of carnivorous sponges with neither aquiferous system nor choanocytes (Vacelet & Boury-Esnault, 1995); and 2) the to presence of soft sponge elements and embryos in southern Chinese Guizhou deposits some 580 million years old (Chia-Wei Li, Ju-Yuan Chen, Tzu-En Hua, 1998). Both these discoveries have made us re-evaluate two key aspects of sponge structure and biology: the interaction between, and the respective functions of, flagellated cells and amoebocytes. First, Jean Vacelet and Nicole Boury-Esnault (1995) found that Asbestopluma lacked choano- cytes yet it could breath, eat, and reproduce successfully. They found that Asbestopluma uses microscleres embedded along long filaments that are supported by long, aligned megascleres to actively capture the prey; the prey is enveloped and ingested by epithelial cells on the filaments. This system of macrophagy replaces the micro- phagous suspension-feeding by choanocytes, unique to the Porifera. Typically, the develop- ment of the aquiferous system and differentiation of choanocytes are the final ontogenetic events in sponge development. From the pioneering work of H.V. Wilson (1932) we know that choanocytes and canals may disappear from a severely dis- tressed sponge (termed involution bodies, resting bodies or reconstitution diamorphy). Similarly, we also know from the pioneering work of J.S. Huxley (1911), that involution bodies in Calcarea are clearly made of archeo-amoebocytes that lack basal bodies. Perhaps early sponges did not require an aquiferous system? The characteristic synapomorphy of the Porif- era is the possession of an aquiferous system and choanocytes, yet this appears to be the most labile part of a sponge. By trying to understand how the aquiferous system is formed and reformed during morphogenesis, can we hope to approach the phylogenetic origin of sponges ? J.S. Huxley (1911) thought not, yet carnivorous sponges live and survive perfectly well without this characteristic synapomorphy of the phylum. It would be of great interest to know whether there is a transitional aquiferous system present during growth and development of Asbestopluma. ls Asbestopluma some kind of permanent diamorphy? The second discovery, of ancient Guizhou sponges by Chia-Wei Li, Ju-Yuan Chen and Tzu-En Hua (1998), suggests that larval flagellated cells and amoebocytes were already present in sponges as old as 580MY. Does this mean that modern species do not differ substantially from those early in the evolution of the group ? How much further does the group extend into the past? MEMOIRS OF THE QUEENSLAND MUSEUM To understand these contemporary viewpoints in the correct context, it is informative to review early historical interpretations of the structure of the aquiferous system and the anatomy of the soft parts of a sponge before modern histological techniques and descriptive embryology provided us with our current understanding of sponge morphogenesis. At the end of the 18th century Peter Pallas (1766) and John Ellis (1755, 1786) suggested that sponges were of animal nature, contrary to popular opinion in those times. Debates on the ‘animality’ of sponges continued for at least a century. Animality, as understood in those times, was assigned to organisms which were capable of voluntary movement, muscular response, and were sentient. Observers sought evidence of re- sponsiveness and of motion, which would be similar to muscular contraction, for sponges (which by nature are a fixed animal). Some investigators, like Donati (1750), observed that the contact between any object and the sponge caused the sponge to contract, whereas others found no such response. In the absence of any sensory reaction or of motion by the sponge, some philosophers suggested they could detect animality based on the nature of the smell arising during sponge decomposition: this smell depends, after all, on the ‘animal’ chemical structure or pattern of the organism. This approach is com- plicated by the fact that most naturalists at that time were only familiar with very few marine sponges (at that time assigned to the greatly misused genera Spongia, ‘Alcyonium’ and Tethya), in some cases only one species, or they worked exclusively on freshwater sponges. Freshwater spongillids are typically green when alive, which led observers to think they were plants. Convergence of viewpoints on the nature of sponges gradually emerged, especially following the works of Grant (1825-26), Dutrochet (1828), Dujardin (1838) and Laurent (1844) on fresh- water and marine sponges. Prior to these authors’ works it was clear to the casual observer that the sponge surface was perforated. This feature was the most consistent amongst the known species of sponges, which eventually led to the phylum being named Porifera (pore-bearing) by Robert Grant (1836). But few authors had any idea on the nature of these pores. Were they normal apertures produced by the sponges or were they caused by foreign organisms, such as worms or polyps, SPONGE SCIENCE, FROM ORIGIN TO OUTLOOK 3 P.S. PALLAS 1766 Animal ambiguum, torpidissimum. Stirps polymorpha, e fibris contexta, gelatina viva obvestitis. Oscula oscillantia, seu cavernae crescens, cellulae ve superficiei C.v. LINNE 1767 Foraminibus Stirps flexilis Flores. aquam. contexta, respirant radicata, pilis bibula J.ELLIS 1786 Animal fixum, flexile, polymorphum, torpidissimum, contextum vel e fibris reticulatis, vel e spinulis, gelatina viva vestitis, osculis seu foraminibus superficiei aquam respirans. FIG. |. Definitions of a sponge from the earliest literature: Pallas (1766), Linneaus (1767), Ellis & Solander (1786). burrowing into the sponge? Many, including Jean Baptiste Monet de Lamarck (1814-16), favoured the latter hypothesis. In fact Lamarck's (1814) classification of the Zoophytes was centred on the presence or absence of polyps: with sponges included in the latter group, and defined as Poly- piers empátés — made of a common substance and ... without polyps. Those who observe a living sponge, however, immediately realise that the pores are an integral part of the sponge anatomy, conducting water flow into and out ofthe body. Some authors, such as Marsigli (1711) and Ellis (1755), imagined (probably moreso than they actually observed), a double flow into and out of the same pore — a systolic-diastolic phenomenon. Others, like Grant (1825) and Dutrochet (1828), observed a unidirectional, continuous water flow exiting the sponge at constant speed. But, they wondered, if water keeps flowing out ofthe sponge, where and how does it go in? Dutrochet (1828) was convinced that the fresh- water spongillid was a plant. He perceived that it was green, it formed a membranous extension that grew through expansion at its edges, like the algae Ulva, it didn't appear to have food cavities and therefore it probably fed by absorbing solutions of water enriched with nutrients, much like a plant. To him it was a plant whose chemical structure was identical to some extent to that of animal tissues. However, Dutrochet observed cavities in sponges with a transparent membrane covering them. He noted that this membrane also covered the entire external surface; it was not sensitive when touched by a foreign object; and it was capable of creating a conical protuberance (oscules) capable of continuously expelling water through the apical part. Dutrochet dismissed the hypothesis that water was expelled from the sponge by commensal Crustacea, and instead he proposed that water was expelled from Spongilla via ‘some kind of a force’ produced by the living animal itself. He did not, however, discover the inhalant pores, but he hypothesised instead that water was drawn slowly into the sponge by adsorption over the whole surface. According to his theory, the expulsion of water from the sponge depended on endosmosis, with the continuous introduction of surrounding water into the cavities of Spongilla, which he said were filled with a denser organic fluid. Grant (1825-26) was also sure that sponges were animals. To him it was obvious that sponges had two types of orifices: 1) larger faecal orifices, through which water was forcefully expelled; and 2) numerous smaller pores through which water entered. Grant, who studied living populations of several sponge species in coastal Scotland, observed a continuous water circulation flowing through many internal canals. Although he was unable to suggest what the ‘motor’ was that drove this circulation, he was none-the-less certain that something like this existed. In fact he commented that minuscule bodies, or granules, organised along the canals might be directly involved, and that water flow is similar to that which might possibly be generated via a flagellar system. Demonstrating remarkably modern vision Grant was sure that water current was one of the vital functions of the pore-bearing animal, and that it helped the animal to feed, breath and even reproduce. In addition to the aquiferous system Grant also noted the sponge soft parts were differentiated into a general cellular substance and a body of material unifying the spicules. This cellular substance, he wrote, twenty years before the cell-theory was proposed by Virchow and others, 4 MEMOIRS OF TH" QUEENSLAND MUSEUM OBSERVATIONS SUR LA SPONGILLE RAMEUSE (SPONGILLA RAMOSA, LAMANCK, EPP DATI LACUSTHIN, LAMOUROUX | Pan M. DUTROCHET, Correspondant de l'Académie royule des Sciences- (Extral des danalos des: Sciences malurelles , yebulce ra.) Nous né connaissons point encore It véritable nature des ponge ; ves étivs , sitaés sur lo fuite qui sépare le pógne animal du règne végélal, semblent appartenii également à ces deux. règnes, On sait que ees productions singuliéres son composées d'un üs&u fihreux encroutd d'une sorte de gelée qui parait de nature animale , et dans laquelle cependant les observatenes les plus habile: n'ont jamais pu apercevoir le moindre sigue d'irritáhi- lité. Les Spongilles qui croissent dans los eaux Moats, offrent à peu près la ime organisation que les Fponges de mer. Pai observó vos Spongilles avec beaucoup de goin; elles m'ont offert des fait nouveáaux et assez curieux, FIG. 2. Definition ota sponge, from Dutrochet (1828). was mostly locuted in the spaces between the walls of internal canals, and appeared to be most obvious during the reproductive period of the sponge. Dutrochet (1828) was the first to explain shape changes of sponges through cellular movements, He focused on the transparent membrane and membranous ducts which provided a pathway for the continuous flow of water out of the sponge. Through regular observations he noticed that these ducts changed shape and length, stretching and shrinking periodically. He suggested these Movements Were not the result of sensitivity buta mechanism tor transporting substances from one part ofthe tube to another. To him the membranes were made of vesicular globules, and changes to their shape and length were caused by the Irans- port of elementary globules. These globules were not static but moved over each other without detaching, in a predelined direction, using a kind of sliding motion. Changes in shape and movement were too slow to be visually observed, much like the hands ofa clock, but shape changes could be seen over lime. ‘This fact is physiologically of the highest importance’ stated Dutrochet, ‘It is a new vital action which plays one of the main roles in stretebing the plant's size, the immediate agent responsible for the vital motion uncovered in its very nature and action mode’. Felix Dujardin (1838) was well known for his studies on Rhizopods and Amoebae, in which he had observed very slow, mitrometrical move- ments, He began studies on sponges during 1835- 1837, working on Clionu celata, Halichondria, Halisarca and Spongilla species, to try and understand sponge organization. From these studies he observed irregular globules made of a contractile substance which, when drawn 20 times at 5 minute intervals, gave 20 different profiles. These globules periodically generated round expansions and thin appendages, much like shape changes to amoebae. He subsequently demonstrated the animal nature of sponges before the French Academy of Sciences, based on the presence of these contractile expansions and by the crawling motion of these ‘packets’. Moreover, in a Spongilla from the Seine River in Paris, Dujardin also saw packets with flagellated filaments which he said ‘determined the water Now and streams in the oscules”. Dujardin and Dutrochet were clearly at odds, and hefore the French Academy, Dujardin stated ‘Mr Dutrochet, wha refuses to admit sponge contractility, explains all their shape changes by the motion of molecules, probably vesicular according to him, which make up the tissues of external membrane’, This controversy was further complicated by the ambiguity of the wards contractility, motion and movement, mandatory to the concept of animality. Whereas Dutrochet was ihe first to describe shape changes of the sponge caused by directed cell motion, and also showed the importance of the external pinacoderm, Dujardin descrihed for the first time the presence of amoebocytes with pseudopodial extensions. In 1844 Jean Laurent suggested that the presence of this external membrane was the sign of the beginning of life tor a perfect state of Spongilla species. In that same year Laurent published this classification, which has been completely forgotten today. These early, careful, time-lapse ocular observ- ations of Dutrochet and Dujardin demonstrated that the sponye and its constituents were able to SPONGE SCIENCE, PROM ORIGIN TO OUTLOOK TYRE SUMMUM DE LANIMALUEE. HOMME, TYPES VTTENMAIMAIYES A NOME FT AUS SPONGIATRIA. TYPE INFIME DE. L'ANIMALITE. Sponpittes , SPONGIAINES | Eponges alliccures, rilicepmnges. | Haléponges, - Téilicees, un. f pent Fpanges éulénires, talerponper. Y Eponge cormécr, ceratépongis. Cet essa) sur la classification du regne animal, fondee sur ce que nous connaissons en l'état actuel à Vegurd du dive- loppement complet des animaux , repose, Y^ sur la eonaide- rotion des degrés de Panimalié depuis V homme Jusqu'oux spongiaires; a” sur les notions acquiaes a l'égard den plans typiques de l'organisation animale, qui sont traduita à l'ez- terieur par les formes de l'enveloppe, par cellus du systéme solide qui protege les organes les plus Importants et surtout le systeme nerveux; 2° sur woe application dey principaux FIG. 3. Succinct classification of the animal kingdom. from Laurent (1544). move and displace, although the speed of these phenomena was visually undetectable, A century later, using time-lapse cinematography, Ankel (1965) studied cellular motion of spongillid fragments sandwiched between two glass-slides. Subsequently, Efremova (1967), Pavans de Ceccatty (1979), Rasmont (1975) and their teams produced a wealth of fascinating observations on vell motility. For Pavans de Ceccatty, who es- sentially focused on information transfer systems and integration, cell mobility and displacement were basic features of sponge organisation, For Rasmont, who was more interested in ex- perimental morphogenesis, the most striking characteristic to him was the extreme mobility of all sponge constituents. For example, using a frame-by-frame analysis he concluded that the movement of spongillid amoeboid cells during gemmulation could be statistically tested for random versus directed movement. it is strange though. that although there has been so much research on embryology, postlarval. postgemmular and postdiamorphic development, we still have so little data on growth and true morphogenesis (i.e. shape achievement), a character so important to the taxonomist but still largely speculative. We have some good models, such as Leuco- solenia, whose growth has been thoroughly studied by W, Clifford Jones (1964, 1965), and particularly from work on the spongillids — a E group intensively studied since the beginning of sponge science (e.g. Grant, Dutrachet, Dujardin. Luurent, Ankel, Rasmont. van de Vyver and inany others), Nevertheless, there are many questions still unanswered. In particular, we know little about intercellular events and chronology during the simultaneous organisation of the skeleton and the aquiferous system (se different in the three classes of Porifera), the cole of primary extemal physical factors, such as 11241 and hydrodynamics, in triggering, orienting and muntenance of cell movements; the reculatory processes influencing (he relative proportions of cell populations and the local conditions prevail- ing during their dilYerentation; the movements ni scleracyles and factors that determine which types of megaseleres are produced, where they are localised and distributed within the skeleton (such as in prismatic or cubic meshes, isotropic prismatic system, or in aseending dendritic hibres), and the consequences of their localisation within the sponge. More than a century after Grant, I also spent a great deal of time observing the littoral sponge fauna around Roscolf, including the same species studied by Grant. My curiosity led me to follow the shape changes that sponges undergo over many years. I noted that fragmentation of in- dividual specimens was frequent, as was the susequent fusion of these same sponge frag- ments. It was clear that fragmentation was not only produced by catastrophic events, such as storms or predation, but also occurred as a much more gradual process, presumably linked to adaptation of the sponge to local environmental conditions, Sometimes these processes can be observed in the aquarium, and sometimes it is possible to generate them experimentally. Slow fragmentation is the result of massive cellular movemenis, which are noi very different [rom those described by Dutrochet in Spongilla. Both fragmentation and fusion are opposite and complementary and are ‘the two fundamental tendencies of the sponge to concentrate and to isolate from the external environment’, as stated by Borojevie (1971) and Wilson (1932) before him. Through the manipulation of light and water circulation it is possible to generate the partial or complete motion of the sponge, whose cells are able to move and leave the existing skeleton to build another in a more physiologically favour- able environment. Ankel (1965), Ankel & Ko Eigenbrodt (1950) and Rasmont (1975, 1979) provided pivotal data on these processes. 6 MEMOIRS OF THE QUEENSLAND MUSEUM What other organisms are easily able to get rid of their skeleton? This phenomenon is unthinkable in more mobile organisms, and certainly does not occur within the plants, and is perhaps unique amongst the Porifera. Of course, morphological freedom associated with cellular migration has constraints and genetic limitations; but why is shape more stable in some species despite their constant local morphological re- adjustments? Even Tethya, a sponge universally characterised by it ‘golf ball” shape, can distort and move. In a compact spherical sponge that lacks cavities and has a severely localised aquif- erous system, an internal equilibrium is set up between cell populations which are using internal energy stocks, Growth under cell proliferation requires an increase in exchange between ex- ternal and internal surfaces. Growth of the exchange zone occurs through folding or multi- plication of choanocyte chambers, and by regulating inhalant cavities. Growth of the ex- ternal surface can be horizontal, polyaxial and peripheral, or, vertical, monaxial and apical, and all intermediate situations exist between the pro- late and oblate states. Recently, Jaap Kandoorp (1995) described a new fractal approach to Haliclona morphology, which in the future should be coupled to 3-dimensional analyses of the aquiferous system, following the method of corrosion casts dev- eloped by Bavestrello et al. (1988). But we also have to investigate the signals which determine the orientation of migrating cells, and we have to know the control mechanisms that determine the motile behaviour of cells of this ‘torpidissimum’ animal. We enter a new millenium with new and fan- tastic technology. Widespread use by researchers and increasing speed of nucleic acids sequencing technology, together with computerised analysis of sequences, provides more and more inform- ation on the genetic make up of sponges, and at this rate it might be reasonable to hope that by the next century we will know the complete sponge genome, a current witness of a primitive multi- cellular organism. It will be the result of teamwork which has to be carefully organised and planned. The choice of the reference model species could well be Ephydatia fluviatilis. But not everything will be explained by knowing all about the genes, although certainly it is a mandatory step. Progress in developmental biology, one of the most important scientific fields of the next century, shows that the chronology of gene expression (also an essential subject), does not yet explain the topological evolution of the blastula, a primarily spherical organism, towards increasingly complex stages possessing multiple compartments and under the constant influence of environmental factors. It be equally necessary to direct team efforts to the central theme of growth and form, following the trail pioneered such a long time ago by d'Arcy Thomson (1917). As Grant wrote, ‘This animal affords many curious and interesting subjects of inquiry to those who [like John Hooper] have leisure and opportunities of examining the more perfect species of tropical seas. Though probably the simplest ofanimal organisation, the investigation of its living habits, its structure and vital phe- nomena, and the distinguishing characters of its innumerable polymorphous species is peculiarly calculated to illuminate the most obscure part of Zoology, to exercise and investigate our intel- lectual and physical powers, and to gratify the mind with the discovery of new scenes of infinite wisdom in the economy of Nature’. LITERATURE CITED ANKEL, W.E. 1965. Der Siisswasserschwamme Ephydatia fluviatilis (Einleitende Worte zur Vorfiirung des Films) Verhandlung deutsche zoologische Gesellschaft 1964. Zoologischer Anzeiger 28(suppl.): 426-444. ANKEL, W.E. & EIGENBRODT, H. 1950. Uber die Wuchsform von Spongilla in sehr flachen Raumen. Zoologischer Anzeiger 145: 195-204. BAVASTRELLO, G., BURLANDO, B. & SARA, M. 1988. The architecture of the canal systems of Petrosia ficiformis and Chondrosia reniformis studied by corrosion casts (Porifera, Demo- spongiae). Zoomorphology 108: 161-166. BOROJEVIC, R. 1971. Le comportement des cellules d'éponge lors de processus morphogénétiques. L'Année Biologique 10(9-10): 533-545. CHIA-WEI LI, JU-YUAN CHEN & TZU-EN HUA 1998. Precambrian sponges with cellular structures. Science 279: 879-881. CLARK, H.J. 1866. Conclusive proofs of the animality of the ciliate Sponges and of their affinities with the Infusoria Flagellata. American Journal Science and Arts (2) 42: 320-324. DONATI, V. 1750. Della storia naturale marina dell Adriatico. ( Venezia). DUJARDIN, F. 1838. Observations sur les Eponges et en particulier sur la Spongille ou éponge d'eau douce. Annales des Sciences naturelles (2) 10: 5-13. DUTROCHET, R.J.H. 1828. Observations sur la Spongille rameuse (Spongilla ramosa Lamarck, Ephydatia lacustris Lamouroux). Annales des Sciences naturelles 15: 205-217. SPONGE SCIENCE, FROM ORIGIN TO OUTLOOK 7 EFREMOVA, S.M. 1967. The cell behaviour of the freshwater sponge Ephydatia fluviatilis. A time-lapse microcinematography study. Acta biologica Academia scientifica Hungarica 18(1): 37-46. ELLIS, J. 1755. An essay towards a natural history of Corallines and other marine productions of the like kind. (London). ELLIS, J. & SOLANDER, D. 1786. The Natural History of many curious and uncommon zoophytes, collected from various parts of the globe. Systematically arranged and described by the late Daniel Solander. (Benjamin White & Son: London). GRANT, R.E. 1825a. Observations and experiments on the structure and functions of the sponge. Edinburgh Philosophical Journal 13: 94-107, 333-346. 1825b. On the ova of the sponge. Edinburgh Philosophical Journal 13: 381-383. 1826a. Observations and experiments on the structure and functions of the sponges. Edinburgh Philosophical Journal 14: 113-124, 336-341. 1826b. On the structure and nature of the Spongilla friabilis. Edinburgh Philosophical Journal 14: 270-284. 1826c. Observations on structure of some siliceous sponges. Edinburgh New Philosophical Journal 1: 341-351. 1826d. Observations on the structure and functions of the sponge. Edinburgh New Philosophical Journal 2: 121-141. 1836. Animal Kingdom. Pp. 107-118, In Todd, R.B. (ed.) The encyclopedia of anatomy and physiology. Vol. 1. (Sherwood, Gilbert & Piper: London). HUXLEY, J.S. 1911. Some phenomena of regeneration in Sycon. Philosophical Transactions of the Royal Society (B) 202: 165-189. JOHNSTON, G. 1842. A history of British sponges and lithophytes. (W.H. Lizars: Edinburgh, London, Dublin). JONES, C.W. 1964. Photographic records on living oscular tubes of Leucosolenia variabilis. Parts I-II. Journal of the Marine Biological Association United Kingdom 44: 67-85, 311-331. 1965. Photographic records on living oscular tubes of Leucosolenia variabilis. Part II. Journal of the Marine Biological Association United Kingdom 45: 1-28. KAANDORP, J.A. 1995. Analysis and synthesis of radiate accretive growth in three dimensions. Journal of Theoretical Biology 175: 39-55, LAMARCK, J.B.P. de Monet 1814. Sur les polypiers empátés. Annales du Muséum d'Histoire naturelle, Paris 20: 294-312, 370-386, 432-458. 1815. Suite des polypiers empátés. Mémoirs du Muséum d’Histoire naturelle, Paris 1: 69-80, 162-168, 331-340. 1816a. Histoire naturelle des animaux sans vert- ébres. Vol. 2 (Verdiére: Paris). 1816b. Histoire naturelle des animaux sans vertébres. Vol. 3 (Verdiére: Paris). LAURENT, J.L.M. 1844. Recherches sur l'hydre et l'éponge d'eau douce. (Paris). LINNEAUS, C. VON 1767. Systema Naturae Editio duodecima, reformata. Vol. 1(2). Insecta, Vermes. (Holmiae: Amsterdam). MARSIGLI, L.F. 1711. Osservazioni naturali intorno al Mare, ed alla Grana detta Kermes. (Brieve Ristretto del Saggio Fisico intorno alla Storia del Mare scritta alla Regia Academia delle Scienze di Parigi: Venezia). PALLAS, P.S. 1766. Elenchus zoophytorum sistens generum adumbrationes generaliores. (Hagae Comitum apud Petrum van Cleef: The Hague). PAVANS DE CECCATTY, M. 1979. Cell correlations and integration in Sponges. Pp. 123-136. In Lévi, C. & Boury-Esnault, N. (eds) Biologie des Spongiaires. Colloque International du C.N.R.S. (291) (Editions du Centre National de la Recherche Scientifique: Paris). RASMONT, R. 1975. Freshwater sponges as a material for the study of cell differentiation. Pp. 141-159. Current Topics in developmental biology. Vol. 10 (Academic Press: New York). 1979, Les éponges: des métazoaires et des sociétés de cellules. Pp. 21-30. In Lévi, C. & Boury- Esnault, N. (eds) Biologie des Spongiaires. Colloque International du C.N.R.S. (291) (Editions du Centre National de la Recherche Scientifique: Paris). THOMPSON, D'ARCY W. 1917. The growth and form of Protozoa, sponges, Coelenterata, and worms. (University Press: Cambridge). VACELET, J. & BOURY-ESNAULT, N. 1995. Carnivorous sponges. Nature 373: 333-335. 1996. A new species of carnivorous sponge (Demospongiae, Cladorhiozidae) from a Mediterranean cave. Bulletin de l'Institut des sciences naturelles de Belgique, Biologie 66(suppl.): 109-115. VAN DE VYVER, G. 1975. Phenomena of cellular recognition in sponges. Current Topics in Developmental Biology 10: 123-140. WILSON, H.V. 1932. Sponges and biology. The American Naturalist 56: 159-170. PLENARY ADDRESSES THE PAST OF SPONGES — SPONGES OF THE PAST FRANCOISE DEBRENNE Astra Pharmaceuticals (Australia) Plenary Address Debrenne, F. 1999 06 30: The past of sponges, sponges of the past. Memoirs of the Queensland Museum 44: 9-21. Brisbane. ISSN 0079-8835. Fossil sponges lack many of the features seen in living sponges, with the consequence that their traditional taxonomy was nearly completely reliant on preserved skeletal architectural characteristics, producing a fossil sponge classification that had diverged considerably from that of living sponges. Subsequent discoveries of ‘living fossil’ sponges with hypercalcified basal skeletons, representing some of the groups thought to be long extinct, provided a revolutionary basis to solve some of the palaeontological enigmas and to comprehensively revise the groups themselves. Ancient groups sphinctozoans, stromatoporoids and chaetitids, with species in Recent seas, are now recognised as grades of construction rather than clades of taxa. The existence of these ‘living fossil’ sponges provided an unique opportunity to compare tissues, spicules and microstructures of the basal skeleton with well preserved fossil material; to understand the influences of biomineralisation and diagenetic alterations affecting mineral composition and microstructures in fossil sponges and to infer the systematic position of Paleozoic to Recent sponges with a calcified skeleton. Similar conclusions were reached for the archaeocyaths, with no living representative yet recorded, but with structural features consistent with the Phylum Porifera. More recent discoveries of ancient sponge tissues and larvae from Precambrian phosphorites provide even more valuable data on the early history and development of Demospongiae and Calcarea, extending the age of the latter group considerably. O Porifera, palaeontology, hypercalcified basal skeleton, sphinctozoans, stromatoporoids, chaetitids, archaeocyaths, taxonomic overview. Francoise Debrenne (email: debrenne(a)club-internet.fr), Paleontology, Muséum National ASTRA MA Astra Australia MINA. d'Histoire Naturelle, 8, Rue Buffon 75005, Paris France; 7 December 1998. We know from the old literature that living sponges have been known since Ancient Times, being familiar household items in ancient Greece and Rome. During the Middle Ages burned sponges were reputed to have therapeutic value in the treatment of various diseases, perhaps anticipating their present pharmaceutical use! Conversely, discoveries of fossil sponge-like ‘objects’ occurred much latter. These were first figured and described as ‘mushrooms’ at the end of the 16th century in the Moscardo collection, according to Zittel (1883). Other scattered examples of sponge-like objects were published later, but these authors did not know whether these forms were plants or zoophytes (Fig. 1). The first valuable observations were made in the second half of the 18th century by Guettard (1768-1783) and several other authors at the beginning of the 19th century. These authors compared their fossils to Alcyonaria or horny corals, but not to recent sponges. Goldfuss (1826) first suggested these fossil forms may be related to living horny sponges, which subsequently mineralised into silica or calcium carbonate, and they attributed known fossil forms to Recent sponge genera. With the ensuing discovery of Hexactinellida (or Hyalosponges) from deepwater dredgings, the exact position of some fossils was established (auguring the impact of the future discovery of ‘living fossil’ hypercalcified sponges or sclero- sponges). D'Orbigny (1849-1850) proposed an initial classification of fossil sponges based on external characters. He considered that these fossil sponges, the Petrospongia, a nearly extinct group, had a mainly calcareous ‘stony’ skeleton, contrary to previous interpretations whereby the horny skeleton became secondarily mineralised. De Fromentel's (1889) classification took into account the interlocking pattern of fibers, the shape of spicules and characteristics of the canal system, but it still kept separate the fossil group Spongitaria, amorphozoans with ‘testacean’ skeleton, and the extant group Spongia, amorphozoans with horny skeleton. The existence of siliceous sponges in the fossil record was confirmed by the discovery of spicules in Jurassic and Cretaceous rocks. The 10 MEMOIRS OF THE QUEENSLAND MUSEUM - Guettard (1768-1786); Animals Alcyonaria or Horny corals - Goldfuss (1826-1833): included fossils in genera of living sponges - d'Orbigny (1840-1854): Petrospongia, extinct group - de Fromentel (1859): Amorphozoan <70%) of hetertrophic bacteria, Prochlorococcus spp., Synechococcus-type cyanobacteria, pico- eukaryotes and nanoeukaryotes. These observations have implications for the distribution and abundance of sponges, and the marine ecosystems in which they exist, The high rates of feeding activity described in this liter- ature indicate that sponges are important components of benthic-pelagic coupling. Polymastia croceus is an abundant, yellow, encrusting, marine sponge, endemic to New Zealand’s coastal waters. It has recently attracted interest due to its production of a proteinaceous secondary metabolite, which has potential for use in anti-cancer and anti-HIV pharmaceuticals. As the metabolite is present in sponges in only trace levels, alternative modes of production are being examined. Wild harvest and aquaculture are options for producing the large quantities of sponge biomass required to supply sufficient metabolite for continuing research. However, itis generally considered that harvesting the required biomass directly from wild stocks is unsustainable and thus artificial supply options have to be considered. Of these options, aquaculture appears to be the one most likely to be rapidly developed (Battershill & Page, 1996). Knowledge of the feeding biology is of fundamental importance to the design and implementation of any aquaculture regime, and the assessment of the impact of removal or addition of P, croceus to the ecosystem. Previous research on P. croceus has been restricted to its reproductive ecology (Battershill & Bergquist, 1999a, in press) and taxonomy (Kelly-Borges & Bergquist, 1997). The work 32 MEMOIRS OF THE QUEENSLAND MUSEUM .---_ Leigh Marin dMenserve et = == » Cape Radney Omaha Bay Takatu Port FIG. |. Map showing the location of study sites in NE New Zealand, approximately 36^16'S, 174*48'F. reported here is based on the approaches used by Pile et al. (1996, 1997), and investigates the following hypotheses: 1) That, like the temperate sponge Mycale lingua (Pile et al, 1996), P. rroceus would efficiently consume large quantities of ultraplankton, in particular Prochloracaccus spp.; and 2) The rates at which water is processed by P. croceus would be high and relatively constant over diel periods. Polymastia croceus is a sponge capable considerable contraction, closing pores and withdrawing oscula, and for unknown reasons alternates between inflated and deflated forms. When deflated no oscula are visible and there is no apparent pumping activity taking place. In contrast, when inflated, large volumes of water appear io be turned over (confirmed visually using dye trace; Bell, 1998). MATERIALS AND METHODS Studies on diet and processing, ability were undertaken at two sites on the NE coast of New Zealand where extensive Polvmastía croceus biomass is found: Sponge Garden, within the Cape Rodney to Okakari Point (Leigh) Marine Reserve, and Takatu Point further to the south (Fig. 1). Polymastia crocens occurred between 16-18m below MLWS at both sites on sand covered base rock (Battershill & Bergquist, 1990b, in press). DIET DETERMINATION. Flow cytometry was used to determine the diet of P. eroceus. Following methods used by Pile (1997), five samples of ambient water (within 5em of the sponge), and five samples of water being exhaled from oscula were taken i situ (using Sec syringes) from each of five sponges at each site, Samples were fixed in 10% paraformaldehyde and frozen at -80°C, following the protocol described by Campbell et al. (1994). The samples were transported tu Macquarie University (Sydney, Australia) for analysis of the ultraplankton composition of cach sample using a FACScan Flow Cytometer unit (Becton Dickinson). The analysis technique was similar to that used by Marie et al. (1997). Two light scatter parameters were analysed: 1) forward light scatter, which relates to particle size; and 2) side light scatter, which relates to cell complexity. Three fluorescence parameters were also analysed: 1) green fluorescence from the SYBR Green | DNA stain (Molecular Probes Inc.); 2) orange fluorescence from the photopigment Phycoerythrin; and 3) red fluorescence from the photopigment Chlorophyll A. Each sample was run twice for all of these parameters. The first run was 100141 with autofluorescence being recorded, and the second was a ] minute run of sample that had been stained with SYBR Green | (Sul SYBR Green | to 4501 sample). The resultant data from the flow cytometry were then analysed using the custom designed Cytowin software (Vaulot, 1989), used to identify and enumerate the cells, Retention efficiency (RE) was determined by applying the following formula to each ultraplankton species: RE-100x(CCA-CCE)/CCA where CCA is mean cell count ambient water, and CCE is mean cell count exhalant water. FEEDING IN POLYMASTIA CROCEUS 53 TABLE 1. Summary of exhalant and ambient cell concentrations RESULTS (mean number of cells ml! + 1 SD) and resultant retention of ultraplankton species by Polymastia croceus at Sponge Garden. DIET DETERMINATION. The most P-values for Students t-test, a:=0.05. abundant ultraplankton (cells mI) = available to Polymastia croceus at Ultraplankton Species | Ambient (x10°) Exhalant (x10°) P-value n both sites were Synechococcus-type a = cyanobacteria, followed by Bacteria P 103.7 + 37.7 55.8 + 35.1 0.0000 46 heterotrophic bacteria, Prochlorococcus spp. | 20.5 - 244 6.8 4.1 0.0063 74 Prochl bi ococcu k spp. em ; emnes | autotrophic picoeukaryotes (Tables Symechococeus-type | 184.6 + 26.6 1174 81 0.0000 94 pw P b ( cyanobacteria a 1-2). Retention efficiencies, however, Picoeukaryotes 8.1423 10204 0.0000 88 were highest for Synechococcus-type TABLE 2. Summary of exhalant and ambient cell concentrations (mean number of cells mb! + 1 SD) and resultant retention of ultraplankton species by Polymastia croceus at Takatu Point. P-values for Students t-test, a=0.05. cyanobacteria followed by picoeukaryotes. At Sponge Garden (Table 1), retention efficiencies of Prochlorococcus spp. were next highest, followed by heterotrophic bacteria, while the opposite occurred " " i 2 Ultraplankton Species | Ambient (x10?) | Exhalant (x10?) | P-value toii d shasta bb d dis A Heterotrophic Bacteria | 88.5 + 28.7 70.8+45.1 | 0.4759 20 | ambient water differed between the Prochlorococcus spp. | 13.1+13.7 | 15.1-242 | 0.3658 | -15 | two sites, although relative Synechococcus-type proportions were constant. ' ia 117.0422.5 | 33.1+30.0 | 0.0000 72 € r fyanabacterig : Significant differences (one-way Picoeukaryotes [521223 1.8 1.7 0.0000 65 ANOVA with Bonferroni's pairwise PROCESSING RATE. To measure sponge pumping rates, a microthermistor-equipped datalogger was built for submarine use based on the design by Pile & Young (in prep.; modified from LaBarbera & Vogel, 1976). Colloquially known as a ‘Medusa’, the unit consists of a 12V battery, data logger and six microthermistor probes. The microthermistors were each placed over an osculum (five probes over oscula and one probe 20cm above the sponge in the ambient flow), and Fluorescein dye was used to visualise the outflow from the oscula to ensure that the probes were correctly in place perpendicular to the flow. The ‘Medusa’ was left in place for 24hr periods to log any changes in pumping rate over time. A Hobotemp temperature logger was also deployed, attached to the ‘Medusa’ housing, to enable the data to be calibrated for temperature variation. The logged data were down-loaded and calibrated with the temperature log data and calibration coefficients to allow conversion of voltage draw into flow rates of cm s”. All the sampled oscula were photographed and the images digitised to allow area measurements, which in turn permitted volume per unit time to be calculated. comparisons, P<0.05) occurred between sites for the ambient concentrations of Synechococcus-type cyanobacteria, which averaged 184.6x10* cells ml! at Sponge Garden but only 117.0x10? cells ml! at Takatu Point. The ambient concentrations of the other species at Takatu Point were not significantly different from those at Sponge Garden (P>0.05). The retention efficiencies of sponges differed between the two sites with the mean retention of Synechococcus-type cyano- bacteria, for example, being 94% at Sponge Garden and 72% at Takatu Point. The largest difference in retention occurred with Prochlorococcus spp.; 67% at Sponge Garden and -15% at Takatu Point. Differences between ambient and exhalant concentrations were tested (Students t-test, a=0.05) to confirm that the retention efficiencies were significant. Only heterotrophic bacteria and Prochlorococcus spp. at Takatu Point had insignificant differences. PROCESSING RATE. Due to technical difficulties only two oscula had (at the time of writing), produced reliable results over a reasonable period (Fig. 2), but it is clear that P croceus can pump at high velocities (Table 3). One oscule, for example, processed on average 26L an hour, or 304L an hour for every cm” of oscule area. The oscula showed a fairly constant pumping rate, with a period of heightened un EN TABLE 3. Summary of pumping rates (cm s-!) of two oscula measured with the ‘Medusa’ at Sponge Garden. The estimates of volume pumped (cm? s-!) were derived from the area of the oscule (8.53: m? for | and 6.96mm? for 2). Oscule Measure Ere Min Max 1 Velocity (ems?) | 84.48-4.78 | 80,15 | 103.44 2 Velocity (ems!) | 64.14+2.55 | 60.68 | 77.65 | 1 Volume (cm? st) | 7214041 6.84 | 8.82 2 Volume (em? s") | 445-0.18 422 54 activity around midday hinting at periodicity in pumping rate. The sponges from which these results were derived were not fully inflated when studied, with only a few oscula per sponge open, and many of the surrounding sponges deflated. Thus, we believe these rates are likely to be conservative as an inflated sponge is likely to have greater pumping potential. DISCUSSION The hypothesis that Polymastia croceus would consume high percentages of ultraplankton, especially Prochlorococcus spp., proved to be partially correct. High percentages of ultra- plankton were indeed consumed, but these consisted of Synechococcus-type cyanobacteria and picoeukaryotes as preferred dietary species, rather than Prochlorococcus spp., as reported in previous studies (Table 4). Heterotrophic bacteria and Prochlorococcus spp. were considerably less favoured, particularly at the Takatu Point site where their retention was statistically insignificant. The ambient samples showed that the most abundant ultraplankton species at both sites was Synechococcus-type cyanobacteria, correlating with it being the most retained species. However, picoeukaryotes, the least available ultraplankton species, had the second highest retention efficiency. This trend is most obvious at Takatu Point and suggests that a certain level of feeding selectivity by P. croceus may be present. Unselective feeding would be expected to show that those ultraplankton species occurring in higher numbers (cells ml) in the water column would also be retained in propor- tionally higher numbers simply due to the higher probability of encounter. Although Prochlorococcus spp. appeared to be in higher abundance in exhalant water at Takatu Point, this was not significant and possibly an artefact of sampling, or due to the sponges concentrating patchily distributed Prochlorococcus spp. into MEMOIRS OF THE QUEENSLAND MUSEUM 120 Ambient Flow 00-2 —— 1 GAR Ba | EAE | | n | | i n Oscula 1 80 4 = : \ E E I | Oscula 2 E \ NI I " i LS S, =p AN o 2 60 5 y i E | Q > | 40 + | 20 4 0 To oo A ss RY ry FS S ss S. SS " ss S S gre qe CEs SS PF wg Time FIG. 2. Oscula velocity (cm s!) from Polymastia croceus at Sponge Garden. Velocity readings were taken at 10 minute intervals. Flow is a record of the ambient water velocity. The large spikes in the flow record are most likely the result of fish bites. their exhalant currents, with few or no Prochlorococcus spp. cells being taken up. Synecococcus-type cyanobacteria and pico- eukaryotes are also smaller than heterotrophic bacteria or Prochlorococcus spp. and to extract them from the internal current especially at high velocities would be difficult. The general perception (e.g. Kilian, 1952; Bergquist, 1978; Pile, 1996) is that sponges are unselective filter feeders, whereas our findings that there is selectivity in dietary retention have implications for the distribution and abundance of P. croceus in the wild. The Takatu Point sponges had lower retention of ultraplankton species (both in cells ml! and percent retained), than the sponges at Sponge Garden. The reasons for this can only be postulated at this stage, but there are three primary, interrelated possibilities. 1) Sponges may have a cycle of pumping, and the time of day at which they were sampled (mid-morning for Takatu Point and mid-afternoon for Sponge Garden), may be associated with different periods of pumping activity. 2) Sponges were inflated to different degrees at the different sampling sites. Inflation and deflation occurs in P. croceus in relation to unknown environmental conditions, which in itself may suggest that different microclimates exist at the two sites. The FEEDING IN POLYMASTIA CROCEUS TABLE 4. Comparison of ultraplankton retention efficiencies (%) and mean exhalant velocities between Polymastia croceus at Sponge Garden and previously studied species. Key: Hbac, heterotrophic bacteria; Pro, Prochlorococcus spp.; Syn, Synechococcus-type cyanobacteria; Peuks, autotrophic picoeukaryotes; |, Pile et al. (1996); 2, Pile (1997); 3, Pile et al. (1997) with velocity for B. bacilifera from Savarese et al. (1997); 4, Reiswig (1971). Ln Ln flow velocities of 84.48cm s'' and 64.14cm s” are high compared to those found by Reiswig (1971), Pile et al. (1996) and Savarese et al. (1997) (Table 4). It was not possible to determine mean pumping rates for P. croceus per Sponge Species Ultraplankton Species con xhalarit unit biomass, a relationship which A AO elocity (ems) | would have allowed more - Hbac Pro Syn Peuks relevant comparisons to be made Polymastia croceus 46 74 94 87 84.48 (SD =4.78) | with data from the literature. As Ircinia felix? | 30 | 26 48 -91 the sponges were not fully inflated Ircinia strobilina? 56 52 53 32 at the time, pumping rates — Baikalospongia o" NA. "m ha: as particularly those in relation to aailifera z sponge biomass — were likely to Baikalospoygia 84 NA 66 99 have been depressed. Variations i) . AI EGOE SEE Mycale lingua | 74 93 89 86 14 (SD =9.7) in: pumping velocities -over-time ATIA “Ga ME "Em A | suggest that there may be periods A ` = : of increased velocity which are Tethya crypta | NA NA NA maid 15 independent of the ambient flow. Takatu Point sponges were observed to be less ‘open’ than those at Sponge Garden, so they may have been going into, or coming out of, a period of deflation and thus processing at a lower, less efficient rate. The lower cells mI available at Takatu Point may be linked to the suppressed pumping/retention activity, as it is possible that food availability may be a cue for inflation/ deflation (pers. observations). 3) The Takatu Point sponges, which are positioned adjacent to a constant long shore current, have some other nutrient source which can be, for example, absorbed through the pinacoderm making filtering for food less necessary. Other studies (e.g. Kilian, 1952) have shown that particles can be ingested by the pinacoderm and thus direct uptake of nutrients from the ambient water is possible, although unlikely to be significant to the sponge’s nutritional requirements. Jn situ sampling provided a realistic insight into the feeding ecology of sponges, although this insight is still limited given the lack of temporal and spatial variation in sampling. Certainly, differences between the two sites sampled suggest that there may be considerable spatial variation. The inference of selectivity shown at Takatu Point is something that has yet to be demonstrated in sponges and is an exciting find, but its verification requires considerably more work. We hypothesised that P. croceus would have relatively high pumping rates that were fairly constant over time. This cannot be verified due to the lack of data, but we can confirm the average Previous studies confirm that pumping rates can vary over time, although this appears to be species dependant. Mycale sp., for example, maintained a fairly constant level of water transport, whereas Tethya crypta had a highly changeable pumping rate, apparently determined by light intensity (Reiswig, 1971). Savarese et al. (1997) also noted high variability in pumping rates over time and with location for the freshwater sponge Baikalospongia. While the results for P. croceus are far from conclusive, it is possible that there was some variability in pumping over short periods of time. The inflation/deflation phenomenon certainly shows that over longer periods of time there is large variation in the water processing potential of any given biomass. Further work with the ‘Medusa’ would allow both short- and long-term variability to be better defined, and thus enhance the potential for making predictions on the amount of water that can be turned over in any given period of time. Determination of the variability in pumping performance with site would allow some gauge of the influence of environmental factors on pumping performance. Variation in ambient flows between sites may show the extent to which flow can assist pumping, and whether this assistance is related to differences between sites, such as the distribution and biomass of sponges: in other words, whether or not assisted flow enhances the growth and distribution of P. croceus. Ecological studies (Bell, 1998) have provided estimations of pumping rates per unit biomass during an inflation (and thus maximum 56 MEMOIRS OF THE QUEENSLAND MUSEUM pumping) event. The calculated mean oscula area per m^ at Sponge Garden, when combined with the mean velocity from oscule 2 (64.14cm 8), produced an estimate of 54ml s" m^, This estimated rate is probably conservative, but its extrapolation suggests thal (he water column directly above the sponges is turned over approx- imately every 9.3hrs at Sponge Garden. In summary, P. croceus appears to be able to process large volumes of Water over short periods. This could lead lo extremely high rates of carbon consumption hy this species, which has potentially significant. implications for ihe benthic marine ecosystem. For example, removing nr adding P. croceus To habitats such as during harvesting or aquaculture ventures, would impaet significantly on these habitats, par- ticularly in terms of the availability of primary production. LITERATURE CITED BATTERSHILL, C.N. de BERGQUIST, P:R. 19993, A novel mode of asexual reproduction in the sponge Polymastia croceus (Hadromerida, Suberitidae): Can sponge buds select settlement sites? Marine Biology (in press). 1999b. The Porifera. In Cook, S. DeC. (ed,) ‘New Zealand Coastal Invertebrates’. (University of Canterbury Press: Canterbury, New Zealand) (in press). BATTERSHILL, C.N. & PAGE, MJ. 1996. Sponge aquaculture for drug production. Aquaculture Update 16: 5-6. BELL, A.H. 1998. The feeding dynamics oFthe sponge Polymastia croceus (Porifera: Demospongiae: Hadromerida) and implications for its ecology and aquaculture. Unpublished MSc thesis (School of Biological Sciences, University of Auckland: Auckland). BERGQUIST, PR. 1978. Sponges. (Hutchinson: London). CAMPBELL, L., NOLLA, H.A, & VAULOT, D 1094, The importance of Prochloraceccus to community structure in the central North Pacific Ocean. Limnology and Oceanography 39: 954-960). KELLY-BORGES. M. & BERGQUIST. PR, 1997. Revision of Southwest Pacific Polymastiidae (Porifera: Demospongiae: Hadromerida) with deseriptions of new species of Polymastia Bowerbank. Tylexocladus Vopsent, and Acamhopolvmastia gen. nov. from New Zealand and the Norfolk Ridge, New Caledonia. New Zealand Journal of Marine and Freshwater Research 31: 367-402. KILIAN, E.F. 1952 Wasserstrommung und nahrungsaufnahme beim susswasserschammen Ephvdatia fluviatilis. Zeitschrift lür Vergleichende Physiologie 34: 407-447. LABARBERA. M. & VOGEL, S. 1976. An inexpensive thermistor flowmeter for aquatic hiology, Limnology and Oceanography 21: 750-756, MARIE, D., PARTENSKY, E, JACQUET, J. & VAULOT, D, 1997, Enumeration and cell cycle analysis of natural populations of marine picoplankton by flow eytometry using the nucleic acid stain SYBR Green I. Applied and Environmental Microbiology 63(1): 186-193. PILE, A.J. 1997, Finding Reiswig's missing carbon: Quantification of sponge feeding using dual beam Flow cytometry, In Lessios, H.A. (ed.) Proceedings of the 8th International Coral Reef Symposium, Panama, June 24-29, 1996 (Smithsonian Tropical Research Institute: Balboa, Panama). PILE, AJ., PATTERSON, M.R., SAVARESE, M., CHERNYKH, V.I. & FIALKOV, V. 1997. Tronhic effects of sponge feeding within Lake Baikal's littoral zone. IT, Sponge abundance, diet, feeding efficiency and carbon Nux Limnology and Oceanography 42(1): 178-184, PILE, A.J, PATTERSON, M.R. & WITMAN, J.D, 1996, In situ grazing on plankton <) 04m by the boreal sponge Mycale lingua. Marine Ecology Progress Series 14]: 95-102. PILE, AJ. & YOUNG, C.M, in prep. Seasonal variation in benthic-pelagic coupling of microbial food webs by ascidians. Limnology and Oceanography. REISWIG, H.M. 1971a. Particle feeding in natural populatians of three marine demosponges. Biological Bulletin 141: 568-591, 1971b. [n situ pumping activities of tropical Demospongiae. Marine Biology 9(1): 38-50. 1973. Population dynamics of three Jamaican Demospongiae. Bulletin of Marine Science 23: 191-226, 1974. Water transport, respiration and energetics of three tropical marine sponges. Journal of Experimental Marine Biology and Ecology 14: 231-249, SAVARESE, M., PATTERSON, M.R., CHERNYKH, Vi. & FIALKOV, V.A. 1997, Trophic effects of sponge feeding within Lake Baikal’s littoral zone. l. In situ pumping rates. Limnology and Oceanography 42: 171-178. VAULOT, D. 1989, Cytope: - Processing software for Now cytometric data. Signal Noise 2: 8. PUSHING THE BOUNDARIES: A NEW GENUS AND SPECIES OF DICTYOCERATIDA PATRICIA R. BERGQUIST, SHIRLEY SOROKIN AND PETER KARUSO Bergquist, P.R., Sorokin, S. & Karuso, P. 1999 06 30: Pushing the boundaries: a new genus and species of Dictyoceratida. Memoirs of the Queensland Museum 44: 57-62. Brisbane. ISSN 0079-8835. Descriptive parameters used to segregate sponges into genera, families and orders must always be subject to re-evaluation. A rare foliose sponge usually found between 20-24m depth on reefs in the central Great Barrier Reef, Australia, has occasioned such reappraisal. The distinct thick, organised sand cortex and surface characters of the oscular and poral faces, the regular, simple primary and secondary fibres and the absence of a tertiary skeleton provide the basis for the diagnosis of a new genus. The lamellate form, brilliant white colouration and regular skeletal arrangement are diagnostic of a new species. A new subfamily, Phyllospangiinae, is established within the Thorectidae, to encompass the new genus and the other foliose dictyoceratid genera Phyllospongia, Carteriospongia, Strepsichordaia and Lendenfeldia. In addition to fibre structure, chemotaxonomic characters and choanocyte chamber morphology support the establishment of the new subfamily. Members of the Phyllospongiinae contain homoscalaranes and a unique series of bishomascalaranes. While the structure of the new sponge has precipitated a subfamilial rearrangement within the Thorectidae, the task of assigning species and genera to other ihorectid subfamilies is not complete at this time. O Porifera, Dietyoceratidà, new genus, new species, Phyllaspongiinae new subfamily, chemotaxonomy, sesterterpenaids, bishomoscalaranes. Patricia R. Bergquist (email: pr bergquistajauckland.ac nz) Department of Zoalogy, School of Biological Sciences, University af Auckland, Private Bag 92019, Auckland. New Zealand; Shirley Sorokin, Museum of Tropical Queensland, 70-84 Flinders Streel, Townsville, Queensland 4810, Australia; Peter Karuso, School of Chemistry, Macquarie University, North Ryde, Sydney, NSW 2109, Australia;3 February 1999 Descripüive parameters used to segregate sponges into genera, families and orders are always subject to re-evaluation, particularly so given the fact that in large areas of the world's oceans numerous species remain undiscovered or, if collected, undescribed. This means that a taxonomist must always be prepared to confront unexpected combinations of characters in new species and to revise systematic arrangements accordingly. The sponge described in this short communication has, with its combination of structural characteristics, required that the diagnoses of two dictyoceratid families, Spongiidae and Thorectidae, be refined and thata subfamily structure be established within the latter. The revised diagnoses of all families and subfamilies will be included in a separate publication. Abbreviations: QM, Queensland Museum, Townsville branch, the Museum of Tropical Queensland; AM, Australian Museum, Sydney. SYSTEMATICS Class Demospongiae Sollas Order Dictyoceratida Minchin Family Thorectidae Bergquist Subfamily Phyllospongiinae s. lam. nov. DIAGNOSIS. Foliose, or folio-digitale Dictyoceratida in which the spongin fibres making up the anastomosing skeleton are finely laminated and contain a differentiated pith. Zones of disjunction between successive fibrous layers remain tightly adherent, producing an overall homogeneous structure with visible contiguous laminae. Pith is not sharply disjunct from the investing spongin fibre but rather merges into it. Secondary fibres show a surface striation. Skeleton 1s made up of cored primary and uncored secondary fibres to which uncored vermiform or reticulate tertiary elements are added in most genera. Choanocyte chambers are large, spherical and diplodal. Matrix is only very lightly infiltrated with collagen and appears fibrous macroscopically rather than fleshy, Cyanobacteria are always 58 MEMOIRS OF THE QUEENSLAND MUSEUM li 1 10 cm FIG. 1. Candidaspongiu flabellata. Holotype, in situ view is of the poral face and shows the verrucoses caused by barnacles, incorporated. Secondary metabolites include a series of unique bishomoscalaranes. REMARKS. In addition to the new genus described below, the subfamily contains four other genera presently classitied in the family Spongiidae: Phyllospongia Ehlers, Carteria- spongia Hyatt, Strepsichordaia Berquist, Ayling & Wilkinson and Lendenfeldia Bergquist, with the first being the type genus of the new subfamily, Lendenfeldia can now be included following the discovery of fresh specimens from the Central Pacific and Australia, which have yielded further information on their morphology. However, it is possible that some species presently assigned to Lendenfeldia, particularly those still relatively poorly known, belong in other genera, while others including the type species L. frondosa, properly belong in the Phyllospongiinae. As a result of the transfer of these genera from Spongiidae to Thorectidae (Phyllospongiinae), only seven genera now remain in Spongiidae (Coscinoderma Carter, Dactylospongía Bergquist, Hippospongia Schulze, Hvattella Lendenfeld, Leiosella Lendenfeld, Rhopaloeides Thompson, Murphy. Bergquist & Evans and Spongia Linneaus) The nominotypical thorectid subfamily, Thorectinae, contains 13 genera (Aplysinopsis Lendenfeld, Cacospongia Schmidt, Collospongia Bergquist, Cambie & Kernan, Fascaplysinopsis Bergquist, Fasciospongia Burton, FIG. 2. Candidaspongia flabellata, Holotype, view of the oscular face, Fenestraspongia Bergquist, Hyrtios Duchassaing & Michelotti, Luffariella Thiele, Petrosaspongia Bergquist, Smenospongia Wiedenmayer, Taonura Carter, Thorecta Lendenfeld and Thorectandra Lendenfeld); Phyllospongiinae contains five genera. Candidaspongia gen. nov. ETYMOLOGY. For the brilliant white colouration. DIAGNOSIS. Lamellate Phyllospongiinae with slightly roughened oscular face and finely conulose poral face. Both surfaces invested by a thick, organised sand cortex, deeper on the oscular face. The compact arrangement and uniform composition of the sand which makes up the cortex is remarkable and produces a brilliant white colour over all surface areas. Skeleton is a network of moderately cored, evenly spaced, simple primary fibres which are disposed regularly, and oriented from attachment base to the surface, Conules on the poral face are aligned along the primary fibres. Primary fibres are connected by a polygonal network of uncored secondary fibres, with a striated braided surface texture, There are no vermiform tertiary. fibre clements. Texture is compressible, firm but flexible. Choanocyte chambers are large, wide mouthed and spherical. NEW GENUS OF DICTYOCERATIDA un © FIG. regular, sand-encrusted oscular face of a flat, spreading specimen. 3. Candidaspongia flabellata. Detail of the Candidaspongia flabellata sp. nov. (Figs 1-5) MATERIAL. HOLOTYPE: QMG25081: Bowden Reef, 19?02'S, 147°56’E, 24m depth, 7.xi.1994. Coll. S. Sorokin. PARATYPES: QMG25082: Bowden Reef, 19°2’S, 147?56'E, 23m depth, 21.xi.1993, Coll. J Wolstenholme. AMZ5286: Davies Reef, 19°01’S, 148?50"E, 20m depth, 11.x.1980. Coll. C. Wilkinson. ETYMOLOGY. For the fan-like shape. DIAGNOSIS. Candidaspongia with brilliant white colouration, lamellate form, extremely regular primary skeleton and strongly developed sand cortex on both faces. DESCRIPTION. Shape. The sponge is a slightly concave, erect fan or a concave plate with a broad base of attachment (Figs 1-4). The lamella is most frequently simple and undissected but small fan-like accessory lobes are sometimes located near the base. Specimens attached under overhangs are thin flattened plates, while those attached above are erect, spreading and thicker. The body is 3-4.5mm thick, up to 25cm high and 35cm wide. Oscular and poral faces are differentiated, oscules are flush with the surface, 0.1-1.5mm diameter and scattered evenly over the entire face. The pores make up an even delicate network over the entire poral face and are 0.1-0.3mm diameter. The colour in life is brilliant white out of water and in ethanol, golden brown internally. A slight purplish tinge can develop, resulting from leaching cyanobacterial pigments. Dry specimens are cream externally, pale golden FIG. 4. Candidaspongia flabellata. Detail of the sand-encrusted but conulose poral face of the same specimen as in Fig. 3. brown internally. The texture is firm, pliant and compressible, dry to the touch with no mucus exuded in life or on collection. Surface. The oscular face is thrown into very low rounded mounds, oscules are situated on or between these. The poral surface is very finely conulose with rows of conules following the line of the underlying primary fibre tracts. Near the edge of the fan this produces a very regular pattern which is enhanced in dry specimens (Fig. 4). A well developed sandy cortex 0.3-0.6mm deep occurs on the oscular face; the poral face has a similar but thinner cortex, 0.2-0.4mm deep. Skeleton. The skeleton is a network of slender, uncored secondary fibres with irregular mesh extending between thin, cored primary fibres which are evenly spaced 1-2mm apart and which run from the attachment base to the margin of the sponge. Primary fibres 65-120um in diameter, secondaries 30-69um. As seen in SEM, the surface of the secondary fibres has a fine braided texture (Fig. 5). Soft tissue organisation. An ectosomal region is developed on both faces, and is collagen-reinforced in addition to supporting the sand cortex. The endosome is open, lightly infiltrated with collagen and with significant volume devoted to canal and choanocyte chamber space and little to matrix. Choanocyte chambers are spherical to oval, 50-90um in longest dimension, with a wide mouth. Reproduction. Sperm follicles and eggs are present in specimens collected in spring. Associations. Light microscope thin sections show that the sponge has an associated cyanobacterium. 60 MEMOIRS OF THE QUEENSLAND MUSEUM FIG. 5. Candidaspongia flabellata. Scanning electron micrograph of the primary and secondary fibre skeleton (<80). This is very similar to Oscillatoria spongeliae, but itis one third of the size of O. spongeliae as found in Dysidea herbacea and does not appear to have green pigment. The surface of the sponge is generally clean, ridges in some specimens are indicative of a polychaete burrow, and occasional surface nodules indicate the presence of a barnacle. A small (2-3mm) pale mauve nudibranch, Chromodoris sp., is often found on the sponge. Chemistry. The sponge contains a novel homo- sesterterpene, a bishomoscalarane, the structure of which will be published elsewhere, and also a cytotoxic macrolide that may have pharma- cological potential (McKee et al., 1993). The sponge can also cause severe dermatological reaction if handled. REMARKS. The brilliant colour and strongly developed, evenly dispersed sand cortex on both lamellar faces make this species easy to recognise in nature. In addition to the type material referred to above, histology has been examined for a further thirty specimens. DISTRIBUTION. Specimens of this sponge are uncommon, occurring on rock surfaces on reef slopes between 6-25m depth; most commonly between 12-25m depth. It is found on middle and outer reefs in the central region of the Great Barrier Reef. It has also been found on one reef in the Coral Sea (Wreck Reef) and on most reefs in the Capricorn-Bunker Group, S Great Barrier FIG. 6. A, Homoscalarane structure. B, Bishomoscalarane structure. C, Scalarane structure. Reef (Hooper, pers.comm.). It has also allegedly been sighted in the Philippines although this record has not yet been verified by the authors. DISCUSSION Taxonomy. Several points of interest are raised by the structure of this sponge. As a consequence it has taken time to decide on the assignment made here and, in doing so, to address references made by other workers to the integrity or otherwise of the families Spongiidae and Thorectidae. We do not regard the task of reassigning species and genera to be complete at this time. This will be done in a contribution to the forthcoming ‘Systema Porifera' (Hooper & Van Soest, in prep.). Bergquist et al. (1988) indicated clearly that the affinities of Phyllospongia, Carteriospongia, Strepsichordaia and Lendenfeldia lay with the Thorectidae, rather than the Spongiidae. The major simple point of distinction is the presence of homogeneous fibres as seen in cross section in the latter group, compared to the concentrically laminate and lightly pithed fibres of the former. Until the 1988 study was done it was not possible to properly distinguish these foliose genera and thus to determine the integrity of the group with respect to fibre structure. A subfamilial status within the Thorectidae was indicated but not formally proposed in that paper, the authors NEW GENUS OF DICTYOCERATIDA 61 preferring to wait until further material of Lendenfeldia could be examined. A subsequent paper dealing with a sponge termed, among other things, ‘Spongia’ mycofijiensis (Sanders & Van Soest, 1996) again drew attention to the difficulties in defining the boundaries between the Spongiidae and Thorectidae using the existing criterion of lam- inated versus homogeneous fibre construction. These authors did not refer to the position stated by Bergquist et al. (1988). The sponge Sanders and Van Soest were dealing with provides a perfect example of the problem we encountered in attempting to assign Candidaspongia to an appropriate taxon, There were too few estab- lished genera to accommodate the range of morphologies being discovered, and existing terminology was not sufficiently refined, or perhaps understood, to differentiate between taxa. ‘Spongia’ mycofijiensis belongs to none of the genera considered by Sanders and Van Soest or other authors reported to have expressed prior opinion on that sponge. It is a species of Petrosaspongia Bergquist (Bergquist, 1995), and is here formally assigned to that genus. In both instances, that of Candidaspongia flabellata and of Petrosaspongia mycofijiensis, new genera were required to accommodate the species. In both instances, pressure was applied to ‘give the beast a name’. Acquiescing to such pressure before certain about a generic assign- ment, merely creates confusion and obscures valid morphological or chemical patterns which may exist, One further contribution which should be referred to is that of Jaspars et al. (1997). They deal with the group of sponges which have been reported to contain scalarane sesterterpenes; this includes the Phyllospongiinae. Many of the problems in the literature that they cite stem from identifications being published without the necessary study of histology and fibre structure to verify generic identifications. In addition, the authors acknowledge that in their collections, voucher specimens are inadequate to resolve those essential morphological details. That has been confirmed by study of those vouchers. The same authors cite confusion over the reported occurrence of homosesterterpenes in Dysidea species which usually contain sesquiterpenes. In all cases this stems from the misidentification of Lendenfeldia species as Dysidea herbacea. The gross morphologies are similar, the histology, fibre structure and chemistry are not. Study of relevant vouchers confirms the above assign- ment. With regard to ‘Phyllospongia vermicularis” (Jaspars et al., 1997: ref. no. 94028) the sponge, as originally described by Lendenfeld (1889), is unrecognisable. Bergquist et al. (1988) referred to a sponge from the Great Barrier Reef which is very similar in morphology but with thicker branches. It has a different histology and skeletal arrangement and proves to be an as yet undescrib- ed species of Dysidea. The status of Lendenfeld’s species cannot be further determined without new material. The sponge from the Great Barrier Reef is a tangle of very fine branches with a grooved surface when dry, and is quite distinct from the sponge 94028 referred to by Jaspars et al. (1997). The other sponge cited by Jaspars et al. (1997; ref. no. 94515) is Carteriospongia contorta Bergquist et al. (1988). Nothing, therefore, in the contribution of Jaspers ed al. (1997) disrupts the integrity of the group of homosesterterpene containing Dictyoceratida. Sesterterpene chemistry. The terpene chemistry of the Phyllospongiinae is characterised by the presence of structures known as homoscalaranes (Fig. 6A) and bishomoscalaranes (Fig. 6B). These are scalaranes (Fig. 6C) which are alkylated at C19, or C19 and 24 respectively. Homoscalaranes are typical of species of Lendenfeldia (Kazlauskas et al., 1982; Alvi & Crews, 1992). Phyllospongia dendyi, from which these metabolites have also been recorded, is a species of Lendenfeldia. The compounds have also been reported to occur in Carteriospongia flabellifera (Schmitz & Chang, 1988). There is one record of a Red Sea sponge containing both homoscalaranes and bishomoscalaranes (Kashman & Zviely, 1979). It was identified as Dysidea herbacea but examination of vouchers shows that it is Lendenfeldia dendyi. Bishomoscalaranes are also reliably known to occur in Carteriospongia foliascens and Strepsichordaia lendenfeldi, and a second, undescribed species of this genus. All the bishomoscalaranes so far described have the ethyl group at C4 b-disposed to the ring system. The only exception is in Candidaspongia which has a-stereochemistry at C4. This is the only example ofalkylation at C25 (cf. C24 for all others), and it serves to emphasise the separation of Candidaspongia from its nearest relatives, the other bishomoscalarane or homoscalarane containing sponges. 62 MEMOIRS OF THE QUEENSLAND MUSEUM ACKNOWLEDGEMENTS Sponge collections were made by Peter Murphy, the Bioactivity Unit/NCI team, and Clive Wilkinson (Australian Institute of Marine Science) and Mary Garson, University of Queensland. Research was supported by AURGC grants to P.R. Bergquist, and MURG grants to P. Karuso. LITERATURE CITED ALVI, K.A. & CREWS, P. 1992. Homoscalarane Sesterterpenes from Lendenfeldia frondosa. Journal of Natural Products 55: 859-865. BERGQUIST, P.R., AYLING, A.M. & WILKINSON, C.R. 1988. Foliose Dictyoceratida of the Australian Great Barrier Reef. 1. Taxonomy and Phylogenetic Relationships. Publicazione Della Statzione Zoologica Napoli: Marine Ecology 9(4): 291-319. BERGQUIST, P.R. 1995. Dictyoceratida, Dendroceratida and Verongida from the New Caledonia Lagoon. Memoirs of the Queensland Museum 38(1): 1-51. HOOPER, J.N.A. & SOEST, R.W.M. VAN In prep. Systema Porifera. A guide to the supraspecific classification of the Phylum Porifera. (Plenum: New York). JASPARS, M., JACKSON, E., LOBKOVSKY, E., CLARDY, J., DIAZ, M.C. & CREWS, P. 1997. Using scalarane sesterterpenes to examine a sponge taxonomic anomaly. Journal of Natural Products 60: 556-561. KASHMAN, Y. & ZVIELY, M. 1979. New alkylated scalarins from the sponge Dysidea herbacea. Tetrahedron Letters 40: 3879-3882. KAZLAUSKAS, R., MURPHY, P.T. & WELLS, R.J. 1982. Five new C», tetracyclic terpenes from a sponge (Lendenfeldia sp.). Australian Journal of Chemistry 35: 51-59. LENDENFELD, R. VON 1889. A monograph of the Horny Sponges. (Triibuner & Co.: London). McKEE, T.C., CARDELLINA, J.H. II, BOYD, M.R., WILLIS, R.H. & MURPHY, P.T. 1993. Novel cytotoxic macrolides from a new Australian sponge. S.12. (American Society of Pharmacog- nosy 34th Annual Meeting: San Diego, July 1993). SANDERS, M.L., & SOEST, R.W.M. VAN 1996. A revised classification of Spongia mycofijiensis. Bulletin de l'Institut Royal Des Sciences Naturelles De Belgique. Biologie 66 (suppl.): 117-122. SCHMITZ, F.J. & CHANG, J.C. 1988. Sesterterpenes from a Pacific sponge Carteriospongia flabellifera. Journal of Natural Products 51: 745-748. MICROBIAL SYMBIONTS OF GREAT BARRIER REEF SPONGES ADAM M, BURJA, NICOLE S, WEBSTER, PETER T. MURPHY AND RUSSELL T. HILL Burja, A.M.. Webster, N.S., Murphy, P.T. & Hill. R.T. 1999 06 30: Microbial Symbionts of Great Barrier Reef Sponges. Memoirs ofthe Queensland Museum 44: 63-75. Brisbane, ISSN 0079-8835, Microbial symbionts of two sponges, Rhapalaeides odorabile (Dictyoceratida: Spongiidae) and a new species ‘Very White Fan’ (V WF) (Dictyoceratida: Phyllospongiidae), are being studied in detail. Bacteria isolated from R. odorabile, VWF, and the surrounding ambient seawater were characterised using morphological, biochemical, and molecular techniques. In the case of R. odorabile, a single bacterium, designated NWOOT, was found to dominate the culturable bacterial community associated with the sponge but was absent from ambient seawater samples. Strain NW00! was predominant in all individual sponges sampled (N=40) from different regions of the Great Barrier Reef, generally at more than an order of magnitude greater than the second most common bacterium (NWO002). The bacterial community associated with R. odorabile appears to be highly stable. In the case of V WF, the culturable bacterial community was more diverse and showed greater variation between individuals. This community generally comprised eight predominant bacteria, rarely isolated from water samples and constituting ca, 70% of the total culturable bacteria, Extensive biochemical testing was performed on all isolates to give data for cluster analyses to identify the major groups of bacteria present. One isolate from each sponge was characterised at the molecular level by PCR amplification and sequencing of 168 ribosomal RNA gene fragments. Analysis of sequence fram NWOD2 indicates it is a Pseudoalteramonas sp. Strain E30004315 from VWE is a microalga, with 168 rRNA sequence from its plastid closely related to that of other plastids of marine eukaryotic algae. This study produced an array of well-characterised microbes for natural products screening, in particular far important compounds known to be produced by these sponges. O Porifera, sponge, svmbiont, Diclyoceralida, Demospongíae, 168 rRNA, Rhopalaeides odorabile, microaleae, Vibrio. Adam M. Burja & Peter Murphy, Marine Biopraducts Group, Australien Institute af Marine Science, PMB No, 3, Townsville MC, OLD 4810, Australia; Nicole Webster. Faculty af Health, Life, and Molecular Sciences, James Cook University of North Queensland, Townsville. OLD 4811, Australia; Russell T. Hill (email; hillviaumbi. umd.edu), Center of Marine Biotechnology, University of Marvland Biotechnology Institute, 701 Eust Pratt Srreet, Baltimore, MD. 21202, USA; 4 April 1999. Symbiosis is considered a permanent association between organisms of different species. This term is not restricted to mutualistic associations but encompasses all associations, regardless of the type of interaction between the individuals. Mutualistic bacterial-invertebrate symbioses have been reported from many invertebrate taxa. Examples of these include cellulolytic nitrogen fixing bacteria from wood boring bivalves (Shieh & Lin, 1994), methanotrophic bacteria of bivalves (Dubilier et al., 1995) and bacterial symbionts of echinoderms (Burnett & McKenzie, 1997; Kelly & McKenzie, 1995). Although chemoautotrophic symbiosis has received the most attention, there are also many symbioses where the type of interaction between the host and their symbiont remains unknown. Most symbiotic bacteria from sponges have been located within the intercellular matrix and can occupy up to 60% of the sponge volume (Wilkinson, 1978a). The biology of bacterium-sponge associations has elicited considerable interest trom researchers investigating novel chemicals derived from sponges. The term symbiont has been broadly applied and lew investigators haye explored metabolic relationships and capabilities of the symbiont-host complex. One approach that will contribute to understanding these relationships is to isolate symbiotic bacteria and investigate their metabolic and taxonomic characteristics. Cosmopolitan microbial symbionts associated with marine sponges include heterotrophic bacteria, cyanobacteria and unicellular algae. Numerous studies have described three general classes of heterotrophic bacterial-sponge associalions (Wilkinson, 1978a), 1) ‘Small 64 populations ol cosmopolitan bacteria with a species composition similar to that found in the ambient seawater. These ure most likely utilised as a food source by the sponge. 2) Species- specific popul- ation inhabiting the mesohy! region, not found in seawater, and most likely comprising [rue symbionts. 3) Fairly ill-defined, consisting of bacteria located within the sponge cells, also likely to be true symbionts. Phenotypically related bacterial symbionts have been described fram taxonomically disparate sponges collected trom geographically remote locations, Described sponge syrmbionts have included members of the genera Pseudomonas, Alteromonas, Vibrio, deromonas, Acinetobacter, Micrococcus and Moruyella (Santavy et aL, 1990). Ji has been determined that facultative anaerobic symbionts melabalise à wide range of compounds and may be important in removing waste products whilst the sponges are not circulating water (Wilkinson, 1978321. lt has also been postulated that sticky mucoid colonies may be important contributors to sponge structural rigidity (Wilkinson, 1978c). Other functions thal have been suggested for sponge bacterial symbionts include digestion of material not available to the host sponge, direct incorporation of dissolved organic matter from the seawater, and digestion and recyclmg of insol- uble sponge collagen, Sponge symbionts are of biotechnological interes! since bioactive compounds of potential medical importance isolated from sponges may he microbial in origin (Bewley & Faulkner, 1998; Bewley et al., 1996: Stierle et al., 1988) There are several practical advantages in isolation of symbionts which produce hivactive compounds, including consistent yield and large-scale production in fermenters, obviating the need tor vollection of sponges from natural ecosystems | Zilinskas et al,, 1995). In this study, microbial symbionts were investigated in two Great Barrier Reef sponges, Rhopaloeides odorahile Thompson et al. (Dictyoceratida: Spongiidae) and a new species, termed here ‘Very White Fan’ (VWF) (see Bergquist et al., 1999, this volume). This 15 a first step towards ascertuining whether these symbionts are implicated in the production of important bioactive compounds. Rhopaloeides odorubile, common throughout the GBR, produces novel norsesterterpenes (rhopaloie acids) which exhibit potent cytoloxic activities (Ohra etal., 1996), and VWF contains the compound fanalide (P. Murphy, unpublished data), which retards the growth of several tumour cell lines. MEMOIRS OF THE QUEENSLAND MUSEUM MATERIALS AND METHODS SPONGE COLLECTION AND BACTERIAL ISOLATION. Material examined in this study was collected using SCUBA (0-30m depth). Seasonal sampling for bacterial community studies was conducted over 12 months at Davies Reef (50 nautical miles off Townsville, Queensland, Australia, 18%49,6'"5, 147%34.49'[). immediately after collection, specimens were processed for bacterial isolation. Using aseptic technique, a lcm section of sponge tissue was excised and surtace-sterilised, Sponge tissue was homogenised in sterile artificial seawater (ASW) using a mortar and pestle. Serial dilutions co, 107 and 107) were prepared in ASW and plated onto several media for isolation of microbes. Media for isolation of heterotrophic bacteria were used in this study. Difco Marine Agar 2216 as a non-selective marine medium, TCBS for enteropathogenic vibrios (Oxoid) and SBA, u selective medjum for bacterial pathogens (Oxoid Columbia Blood Agar Base), BG-11 (Stanier et al.. 1971) and MN + B12 (Waterbury & Stanier, 1978) were used for isolation of oxygenic phototrophs. Plates were incubated at 27°C fora period ranging between 2-4 weeks. Total cultur- able colony counts were obtained from Lifuo Marine Agar 2216 spread plates. Representatives of each morphotype were cultured from initial isolation and cryopreserved for further studies. Total bacterial counts were determined by fluorescent microscopic enumeration of cells stained with 4°6-diamidino-2-phenylindole (DAPI) as described by Porter & Feig (1980). SCANNING AND TRANSMISSION ELECTRON MICROSCOPY. Sponge sections were prepared using a scalpel blade to cul 1-1.5mm thick sections of sponge tissue, ensuring, that both the ectosome and choanosome were represented. Sections Were fixed in 2.5% glutaraldehyde made m 0. | M sodium cacodylate buffer foc 20hrs. Fixed samples were transferred into 0.1M sodium cacodylate and stored at 4°C until further processing, Sections of sponge tissue were placed in a 1% osmium tetroxide solution (prepared in 0.2M potassium phosphate bulter. pH 7.4) for 3.5hrs. Sections were removed and washed thoroughly in sterile distilled water, dehydrated in a graded ethanol series (15%, 35%, 55%, 15%, 85% and 95%), placed in embedding, capsules and covered with Spurr's resin. Thin sections were cul and stained with 2% uranyl acetate followed by 0.2% lead citrate, Sections were mounted on 200 mesh copper TEM grids GBR SPONGE MICROBIAL SYMBIONTS 65 TABLE 1. Type cultures used as control strains in biochemical analyses. Key: ACMM, Australian Collection of Marine Microorganisms; ATCC, American Type Culture Collection. Collection number Organism ACMM 667 Vibrio parahaemolyticus ACMM 668 V. parahaemolyticus ACMM 89 Vibrio alginolyticus ACMM 90 V. parahaemolyticus ATCC 33809 Vibrio fluvialis ATCC 33807 V. fluvialis ATCC 7966 Aeromonas hydrophila ATCC 35624 Aeromonas group 77 ATCC 33509 Vibrio ordalii coated with carbon and Formvar. Samples were visualised following standard scanning and transmission electron microscopy techniques at James Cook University and University of Queensland. BIOCHEMICAL TESTING OF BACTERIAL ISOLATES. The isolates were characterised by determining biochemical characteristics in 96-well microtitre-plates based on methods described by Hansen & Sorheim (1991). Several dye indicator tests were performed: Moller's arginine, lysine, ornithine, base; nitrate reduction, ONPG, indole, acetoin, tellurite, aesculin, alginate, acid-arabinose, arbutin, glucose, inositol, mannose, salicin, sorbitol, sucrose, and urea. The following tests were performed to determine the ability to utilise different carbon sources in assimilation broth: arabinose, cellobiose, galactose, glucose, mannose, melibiose, lactose, melizitose, sucrose, trehalose, xylose, ethanol, glycerol, propan-1-ol, sorbitol, gluconate, glucuronate, amygdalin, arbutin, citrulline, hydroxyproline, leucine, gluco- samine, hydroxybutyrate, a-ketoglutarate, succinate, base (control), adenine, aminovalerate, N-acetyl-D-glucoseamine, ethanolamine, m-erythritol, D-fructose, D-galacturonate, glutarate, inositol, malonate, maltose and valerate. The assimilation broth contained (per litre) 0.015g of yeast extract, 1.0g of ammonium chloride, 0.075g of di-potassium hydrogen ortho- phosphate, 6.1g of Tris (hydroxymethyl) aminomethane and 15g of ASW salts (pH 7.5). After autoclaving, the carbon sources were filter sterilised and added aseptically to a final concentration of 8% (wt/vol.). Inoculations were performed by suspending colony material in ASW and inoculating 100g of this suspension into each test. In addition, the following morphological characteristics were determined: colony morphology, gram stain and cell morphology, plate swarming, oxidation/ fermentation (Leifson, 1963; Lemos etal., 1985), oxidase, catalase and growth at various salt concentrations (0%, 1%, 6%, 8%). In addition several antibiotic susceptibility tests (O/129 10 & 150ug, ampicillin 10ug & polymixin B 50iu) were performed along with growth on different media (Lecithinase, DNase, TCBS, SBA). Several American Type Culture Collection (ATCC) and Australian Collection of Marine Microorganism (ACMM) strains were included as controls (Table 1). Isolates were tentatively identified to either the genus or species level by comparing their phenotypic characteristics with those of type cultures and by comparing biochemical test results, carbohydrate utilisation patterns, and cell morphologies to those of species described in Bergey's Manual of Systematic Bacteriology (Holt, 1986) and Bergey's Manual of Determinative Bacteriology (Buchanan Gibbons, 1974). DATA ANALYSIS. The levels of relatedness of the bacteria were determined from the pheno- typic data using Jaccard’s similarity index (Zar, 1984). S; =a/(a + b * c) where S; -Jaccard's similarity coefficient, a = no. species in sample A and sample B (joint occurr- ences), b — no. species in sample B but not in sample A, c — no. species in sample A but not in sample B. Phenograms were constructed by using unweighted pair group mean average (UPGMA) linkage (Sokal & Michener, 1958), Euclidean distances and the computer software package STATISTICA (StatSoft Inc., Tulsa, Oklahoma). BACTERIAL IDENTIFICATION BY 168 RIBOSOMAL RNA (rRNA) ANALYSIS. Two microbial isolates, NW002 from R. edorabile and the microalga E30004315 were identified using a molecular approach, partial sequencing of the 16S rRNA gene fragments amplified from these isolates using the polymerase chain reaction (PCR). Total DNA was prepared from strains NW002 and E30004315 using a method modified from Ausubel et al. (1987). Oligonucleo- tide primers with specificity for eubacterial 16S rRNA genes [Forward primer 8-27:5^ AGAGTTT GATCCTGGCTCAG -3' (modified from FDI) (Weisburg et al., 1991) and Reverse primer 1492:5’- GGTTACCTTGTTACGACTT-3’ (Reysenbach et 66 MEMOIRS OF THE QUEENSLAND MUSEUM al., 1992)] were used to amplily a 168 rRNA gene fragment from NW002. The cyanobacterial and plastid-specific 16S rRNA primers described by Nübel et al. (1997) were used for E30004315, since this isolate. although unialgal, may have been contaminated with low numbers of heterotrophic bacteria. PCR fragments were purified using the Microcon 30 system (Amicon, Beverly, MA), and sequenced using the PRISM Ready Reaction Kit (PE Applied BioSystems, Foster City, CA) and an ABI 310 sequencer (PE — Applied BioSystems), Sequencing data were analysed by comparison to 16S rRNA genes in the Ribosomal Database Project (Maidak et al., 1999; Maidak et al., 1997) and the Genbank database, and aligned manually using the Phydit software (Chun, 1995). Evolutionary trees were inferred using the neighbour- joining (Saitou & Nei, 1987), Fitch-Margoliash (Fitch & Margoliash, 1967) and maximum- parsimony (Kluge & Farris, 1969) algorithms in the PHYLIP package (Felsenstein, 1993). Evolutionary distance matrices for the neighbour- joining and Fitch- Margoliash methods were generated as described by Jukes & Cantor (1969), Tree topologies were evaluated by performing bootstrap analyses of the neighbour-joining data, based on 1000 re-samplings (Felsenstein, 1985), Abbreviations:AIMS, Australian Institute of Marine Science; ASW, Artificial seawater; DAPI. Diamidino-phenylindole; 16S rRNA, 168 ribosomal ribonucleic acid; SBA, Sheep FIG 1. Electron micrographs of VWF sponge sections. A, Low magnification scanning electron micrograph of sponge section. B, High magnification of sponge section showing cells identified as putative cyanobacteria on morphological criteria, C-D, Low and high magnification, respectively transmission electron micrograph of sponge mesohyl section showing location of bacteria within ‘bacteriocytes’ and putative cyanobacteria, indicated by arrows. Blood Agar; TCBS, Thiosulphate Citrate Bile electron microscopy to be present within the Salts Medium; VWF, Very White Fan. sponge VWF (Fig. 1A-D). Bacteria closely RESULTS associated with the sponge tissue, possibly embedded in a polysaccharide matrix, were presumed to be cyanobacteria based on morpho- ELECTRON MICROSCOPY, A large and logical criteria (Fig. 1B), since these cells complex bacterial community was shown by resemble filamentous cyanobacteria (e.g. genus GBR SPONGE MICROBIAL SYMBIONTS 67 Oscillatoria). Cells presumed to be other eubacteria, based on the standard morphological criteria of size, shape and membrane structure, were also in close contact with the sponge tissue and contained in cellular organelles resembling the ‘bacteriocytes’ described by Vacelet & Donadey (1977) (Fig. 1C-D). Sand grains were observed which appeared to be incorporated into the sponge externa] structure (also reported by Bergquist, et al., 1999), possibly performing the function of increasing structural integrity (Shaw, 1927). The bacterial community within A. odorabile was also large and appeared to he comprised of many different bacteria (Fig. 2). The bacteria appear to be dispersed throughout the sponge mesohyl and no bacteriocytes were evident. In contrast to VWF, cells resembling, cyanobacteria were not seen. BACTERIAL ENUMERATION. The average number of culturable bacteria from direct quam counts obtained from VWF was 3.6x 10" /ml and the average lota count observed from DAPI staining was 6.3x10 "ml. Only 0.06% of total bacteria were able to be recovered using traditional culture techniques. The range of total and culturable bacterial counts found in samples from eight individual VWF sponges are shown in Figure 3 The average percentage of culturable bacteria FIG, 2. bacterial morphologies and the density of bacterial cells within the tissue of the sponge R. odorabile. from A, odorabile was only 0.1% with a range of 0.001-0.8%. The average percentage of bacteria able to be cultured from the water column was 0.23% with a range from 0.003-0.9%. Total and culturable bacterial counts found in samples from four individual R, adorabile sponges and the ambient seawater surrounding each sponge are shown in Figure 4, BIOCHEMICAL CHARACTERISATION OF BACTERIAL ISOLATES. Morphological and biochemical daia indicated that, al least as judged from the culturable fraction, the bacterial community within VWF differed from that present in the water column. Culture results from 15 ‘Transmission electron micrograph showing the diversity of VWF individuals. from different locations revealed several similarities. Eight eubacteria, designated ABOOL to ABO008, were frequently observed as being part of the culturable bacterial community of V WF and were found to be present only in small numbers in samples from the water column. A total of 220 isolates were isolated from VWF samples collected between June 1997-May 1998 from locations between Trunk Reef and Davies Reef, Great Barrier Reef, These isolates conformed to one of eight clusters (Fig. 5). Two clusters were Gram-positive with the remainder being Gram-negative. The Gram-posilive bacteria were further sub-divided by means of cellular morphology. Approximately 40% of all bacteria (including strains ABOOA, AB007, and AB008) isolated from VWF clustered in phenons 6 and 7 (Fig. 5); these phenons contained the vibrio and acromonad representatives, re- spectively, of the Pibrionaceae type strains used in this study. In addition, approximately 30% of the total bacterial community fell in a single phenon (phenon 8), which contained the strain AB005, Of the remaining three Gram-negative 68 MEMOIRS OF THE QUEENSLAND MUSEUM FIG. 3. Total and culturable bacterial counts from tissue of eight VWF individuals collected at Davies Reef, Great Barrier Reef. clusters, phenon 3 contained pigmented bacteria, phenon 4 included strains AB003 and AB004, and phenon 5 included strains ABOO1 and AB002. Strain ABOO! appears closely related to NWO001 from R. odorabile, with both strains being representatives of the alpha-Proteobacteria (data not shown). From biochemical and morphological observations, it was apparent that the bacterial community within R. odorabile was quite distinct from the bacterial assemblages associated with the ambient water column. In general, both total and culturable counts from sponges exceeded counts from the corresponding water samples. The sponge microbiota was dominated by an organism designated NW001, whereas this isolate was completely absent in the surrounding water column. A small component of the microbial community was observed in both the sponge tissue and the ambient seawater. A total of 223 isolates were collected from 40 R. odorabile samples collected between August 1997-May 1998. These isolates conformed to one of ten major clusters (Fig. 6). Two clusters (phenons 9 & 10) were Gram-positive and these were distinguished from each other on the basis of the oxidase reaction. Two of the Gram- negative clusters (phenons 1 & 2) were oxidase- negative and showed profiles linking them to the Enterobacteriaceae. One of the Gram-negative, oxidase-positive clusters (phenon 3) was catalase- negative and the remaining five clusters were 1 O0E+10 1 ,O0E+09 + 1 N0E+08 1,008+07 CFU/ml) 1.00€+06 1.00E«05 100E«04 acterial Count Lag B: 2 m á a 21891 Sample Number 21747 tz Water Culturable Br Water total E) Sponge Culturable w Sponge tatal FIG. 4. Total and culturable bacterial counts from tissue of four R. odorabile individuals and seawater surrounding each individual collected at Davies Reef, Great Barrier Reef. catalase-positive and separated on the basis of carbon source utilisation. One of the five clusters contained an Aeromonas sp. type culture (phenon 6) and a second cluster contained the Vibrio anguillarum type culture (phenon 4). NW001 was a Gram-negative rod; oxidase, catalase, urease, VP and indole positive; utilised adenine dihydrogenase and had the ability to utilise glucose and gluconate as carbon sources. NW002 was a Gram-negative rod, oxidase, catalase, VP, indole and acid arabinose positive. It was urease-- negative and did not utilise any of the tested carbon sources. Both NW001 and NW002 clustered within phenon 5. MICROALGAL ISOLATES. In addition to the heterotrophic bacterial isolates, eight strains of oxygenic phototrophs were isolated from VWF and one of these strains, designated E300043 15 was characterised by 16S rRNA sequencing (below). A single phototroph strain was isolated from R. odorabile. Phototroph strains were not characterised by biochemical testing because of the difficulty in identification of microalgae by this means; instead 16S rRNA sequencing was used as a method for identification of strain E30004315. PHYLOGENETIC POSITIONS BASED ON 168 tRNA SEQUENCING. Phylogenetic relationships for the plastid of microalga E30004315 from VWF and heterotrophic bacterial strain NW002 from R. odorabile are shown in Figures 7 and 8, GBR SPONGE MICROBIAL SYMBIONTS 69 7 6 5 v 4 e S 5 ul A 3 oD a É 2 E l D Gram positive rods Gram positive cocci Streptococcus spp. UNIDENTIFIED GROUP Gram negative rods PIGMENTED BACTERIA Gram negative rods UNIDENTIFIED GROUP AB003 & ABODE Gram negative rods PROTEOBACTERIA ABO001 € AB002 Gram negative rods Vibrio spp. AB004, ABOO? & ABUOZ Gram negative rods Aeromonas spp. Gram negative rods Photobacterium spp. AB005 FIG. 5. Similarity dendrogram for isolates obtained from VWF. respectively. The plastid from isolate E30004315 from VWF is closely related to plastids of other marine eukaryotic algae. NW002 is a Pseudo- alteromonas sp. DISCUSSION This comprehensive biochemical and morpho- logical analysis of bacteria isolated from two Great Barrier Reef sponges further emphasises the variability in microbiota associated with marine invertebrates. Several studies have documented diverse microbial communities associated with sponges (Vacelet, 1970, 1975; Vacelet & Donadey, 1977; Wilkinson, 1978a,b,c; Santavy, 1985; Willenz & Hartman, 1989; Santavy & Colwell, 1990; Santavy et al., 1990; Lopez et al., 1999). These communities are generally comprised of large numbers of heterotrophic bacteria that often occupy up to 60% of the sponge volume (Santavy, 1985; Wilkinson, 1978a,b). 70 MEMOIRS OF THE QUEENSLAND MUSEUM 2.0 Linkage Distance 0.5 l ‘ane EDIT 0.0 | lI] 2 1 Gram negative, oxidase (-) Enterobacteria spp. VU 2 Gram negative, oxidase (-) 43 Gram negative, oxidase (+), catalase (-) 4 Gram negative, oxidase (4), catalase (+) Vibrio spp. 5 Gram negative, oxidase (4, catalase (+), CSU (-) NY7001 € NW/002 6 Gram negative, oxidase (+), catalase (+), moller’s (+) gi Gram negative, oxidase (+), acid (+) 8 Gram negative, oxidase (4), acid ©) 9 Gram positive, catalase ©), oxidase (+) Streptococcus spp. a | Gram positive, catalase (-), oxidase (- Streptococcus spp. 4| 10 FIG. 6. Similarity dendrogram for isolates obtained from R. odorabile. Phenotypic analysis of bacteria from the Caribbean evident from the present study, that the sponges sclerosponge, Ceratoporella nicholsoni revealed Rhopaloeides odorabile and VWF support significant differences in sponge and seawater taxonomically diverse microbial assemblages. phenotypes (Santavy & Colwell, 1990). It is High microbial diversity is not surprising when GBR SPONGE MICROBIAL SYMBIONTS 71 Marine clone HstpL 1 Marine clone HstpLA Strain E30004315 Rhodophyte plastid EnvOCS54 Cyanobacterinm clone LD27 Amphora delicati sama chloroplast Skeletonema costatum chloroplast Skeletonema pseudoco statum chloroplast 100, £ p SLED Marine eubacterium agg56 Odontella sinensis chloroplast 0.01 Navicula sulíricola chloroplast FIG. 7. Neighbour-joining tree for 631 bp of sequence obtained using cyanobacterial and plastid-specific primers from strain E30004315 isolated from VWF, Key: f and p indicate branches that were also found using the Fitch-Margoliash or maximum-parsimony methods, respectively. The numbers at the nodes are percentages (only values over 50% shown) indicating the level of bootstrap support, based on a neighbour-joining analysis of 1,000 re-sampled data sets. Scale bar represents 0,01 substitutions per nucleotide position. considering that sponges derive their nutrition from filtering the ambient seawater. It has been demonstrated that marine sponges are capable of discriminating between food bacteria and bacterial symbionts. The mechanisms for this recognition are not clear but it has been postulated that sponge phagocytic cells do not recognise the capsule coating of symbionts (Wilkinson et al., 1984). Eight predominant heterotrophic bacteria were evident in the culturable community isolated from VWF and these isolates were generally absent from the surrounding seawater samples. These culturable bacteria clustered in several phenons on biochemical analysis. The culturable community of VWF was more diverse than that observed in R. odorabile and showed greater fluctuations between individual sponges. One notable feature was the prevalence of cells resembling cyanobacteria within the VWF matrix, observed by microscopy. Also, eight strains of phototrophs were isolated from this sponge. It is postulated that VWF is a phototrophic sponge, deriving a component of its carbon budget from photosynthetic symbionts. Wilkinson (1992) has reported that many sponge phototrophs are morphologically flattened to increase surface area for interception of light. This is consistent with the structural morphology of VWF. In contrast, only a single phototroph was isolated from R. odorabile and cells with characteristic cyanobacterial morphology were not observed on microscopic examination of R. odorabile tissue. The culturable bacterial community of R. odorabile was dominated by strain NW001, which comprised 74% of the total culturable bacterial community in this sponge but was consistently absent from the seawater samples. This is the first report of a single bacterium comprising such a high proportion of the culturable bacteria from a sponge. Previously, a specific bacterial symbiont was found in nine of ten sponges of two classes and seven orders, and a second symbiont was specific to the sponge Verongia, but only as one component ofa large mixed bacterial community (Wilkinson et al., 1981). The relationship between NW001 and R. odorabile provides an ideal model system for investigating the relationship between this strain and its host sponge because this isolate is predominant and has a characteristic colony morphology which facilitates enumeration of strain NWOOI based solely on colony morphology. Initial indications are that this relationship persists over spatial and temporal scales and is highly stable (work in progress). Although the relationship between NW002 and R. odorabile appeared less stable, NW002 was frequently the second most predominant culturable bacterium (after NW001) present in R. odorabile, and was present in the sponge tissue at much higher concentrations than detectable in the ambient water surrounding the sponges. Similarly, strains AB001-AB008 were consistently present in the sponge V WF at higher concentrations than detected in the surrounding seawater. The mechanism whereby the sponges acquire these symbionts is a topic of current research. It appears as though the two sponges adopt different strategies for harboring their microbial communities. Rhopaloeides odorabile maintains the bacterial cells in the loose matrix of the mesohyl, whereas VWF appears to incorporate the cells into structures referred to as bacteriocytes. In VWF, the bacteria are also closely associated with the sand grains just below the cuticle. The reasons for these different approaches are uncertain but may relate to the structural composition of the sponge or the function of the bacteria within the sponge tissue. Rhopaloeides odorabile maintains 72 MEMOIRS OF THE QUEENSLAND MUSEUM Alteromonas aurantia 100 f. p Alteromonas cirea Pseudoalteremonas sp. ANG.102 Pseudoalteromonas sp. MBS-11 Hydrothermal! vent strain PVE OTU 5 87£p NW0O002 North Sea bacterium H7 Pseudoaiteromonas sp. SWOS Pseudomonas atlantica fp Pseudoaiteromonas halopianktis Pseudoalteromonas gracilis 83f Pseudoalteromonas cirea 68 f 69 Pseudoalteromenas antarctica Alteromonas elyakovit Pseudoalteromonas nigrifaciens 0.01 Alteromonas distincta FIG. 8. Neighbour-joining tree for over 1,000 bp of sequence obtained using eubacterial-specific primers from strain N W002 isolated from R. odorabile. Key: f and p indicate branches that were also found using the Fitch-Margoliash or maximum-parsimony methods, respectively. The numbers at the nodes are percentages (only values over 50% shown) indicating the level of bootstrap support, based on a neighbour-joining analysis of 1,000 re-sampled data sets. Scale bar represents 0.01 substitutions per nucleotide position. a bacterial community two orders of magnitude greater than that present in tissue of VWF. Biochemical characterisation of all culturable isolates from VWF and R. odorabile was useful in clustering each of these assemblages into distinct phenons. In some cases, the presumptive identity of isolates could be deduced by comparison with type cultures which clustered in the same phenon. However, this approach must be used with caution because of the difficulty in identifying marine isolates based on criteria generally established for readily-culturable isolates of medical significance. In addition, many isolates scored negative against almost the entire biochemical profile, as is frequently the case with marine environmental isolates. Interestingly, two phenons from each sponge comprised Gram-positive bacteria, which made up approximately 10% of the total isolates in each case. Early studies of marine microbiology found that about 95% of marine isolates were Gram- -negative (ZoBell, 1946) but recently 30% of the bacteria associated with a marine alga were found to be Gram-positive (Jensen & Fenical, 1995) and it is likely that the proportion of Gram-positive bacteria in most marine habitats has been underestimated (Jensen & Fenical, 1994). Gram-positive bacteria include actinomycetes, a group of particular importance in natural products discovery. Because of the difficulties in accurately identifying marine bacteria by conventional biochemical characterisation, molecular techniques are the most appropriate for unequivocal identif- ication of marine bacteria, A single isolate (NW002) from phenon 5 of the assemblage from R. odorabile was selected to demonstrate the utility of this approach and as a first step in the molecular identification of one isolate from each phenon. In addition, because of the difficulty in identification of marine microalgae by conventional techniques, phototroph isolate E30004315 was characterised by 165 rRNA sequence analysis of its plastid. Sequence analysis of the plastid of microalgal strain E30004315 revealed that this plastid was most closely related to sequences of clones HstpL and HstpL4, cloned from a library of the uncultured microbes associated with the seagrass Halophila stipulacea, a ubiquitous seagrass from the subtidal zone of the Gulf of Elat (Weidner et al., 1996). Another close relationship was to clone OCS54, a plastid rRNA sequence from a natural phytoplankton population collected in the Pacific Ocean, off the mouth of Yaquina Bay, Oregon (Rappe et al., 1998). The identified microalgae with plastids clustering close to E30004315 fall in the genera Odontella and Skeletonema (Fig. 7). Microscopic examination of E30004315 revealed morphology consistent with identification as a microalga in the eukaryotic phytoplanton, with cigar-shaped cells about 10um long. Strain NW002 clearly belongs to the genus Pseudoalteromonas, a genus with many marine representatives, on the basis of the close phylogenetic relationship between this isolate and sequences of strains classified in the genus Pseudoalteromonas. The 168 rRNA sequence most closely related to that of NW002 was derived from a clone derived from a microbial mat at a hydrothermal vent site, the Loihi Seamount, GBR SPONGE MICROBIAL SYMBIONTS 73 Hawair and reported as an alteromonad (Moyer et al., 1995), The percentage of culturable bacteria associated with these sponges was only 0.06% in VWF and 0.1% in R, odorabile. These percentages are considerably lower than the 3-11% culturable hacteria associated with the sclerosponge Ceratoporella nicholsani (Santavy, et al., 1990), and may indicate that a high proportion of the bacteria associated with R. odorabile and VWF are obligate symbionts, requiring a close association with the sponge tissue to grow. These results illustrate the importance ol molecular genetic techniques for total community analysis. Once more is known about the total microbial community associaled with sponges, it may be possible to use this knowledge for rational selection of culture conditions appropriate for growth of additional, presently unculturable, strains. This study has resulted in an array of well-characterised microbes for natural products screening, in particular for important compounds known to be produced by these sponges. Compounds of potential pharmaceutical import- ance from R. odorabile include diterpenes (Kazlauskas et al., 1979) and rhopaloic acid A (Ohta, et al., 1996), although it is likely that these particular compounds are not of microbial origin (Thompson elal., 1987). VWF produces a potent antitumor compound fanolide. lt is clear that marine sponges. have the potential to be a major source of microbes for natural products screening programs. ACKNOWLEDGEMENTS We are grateful to Nicholas Gudkovs and Mark Crane from the Australian Animal Health Laboratory, Fish Diseases Laboratory, Geelong, for providing the orginal methodology and identification program used in mochemical testing. Adam Reynolds and Heather Windsor of James Cook University and Richard Webb at the Centre of Microscopy and Microanalysis of the University of Queensland are thanked for assisiance wilh electran microscopy. Jacques Ravel is thanked tor assistance in phylogenetic analyses. LITERATURE CITED AUSUBEL, F.M, BRENT, R., KINGSTON, RE., MOORE, D.D., SEIDMAN, J.G., SMITH, LA. de STRUHL. K. 1987. Current protocols in molecular biology. (John Wiley and Sons: Cambridge, MA), BERGQUIST, PR.. SOROKIN, S.J. & KARUSO, P. 1999, Pushing the boundaries: a new genus and species ef Dietyoceratida. Memoirs of the Queensland Museum (ühis volume). BEWLEY, C.A. & FAULKNER, D.J, 1998, Lilhistid sponges: star performers or hosts to the stars. Angewandte Chemie International Edition 37: 2162-2178. BEWLEY, C.A; HOLLAND. N.D. & FAULKNER, DJ. 1996. Two classes of metabolites Irom Theonella swinhoei are localized in distinct popul- ations of bacterial symbioms. Experientia 52: 716-722. BUCHANAN, R.E. & GIBBONS, N.E. 1974, Bergey's Manual of Determinative Bacteriology. (Williams & Wilkins: Baltimore). BURNETT, W.J. & McKENZIE, J.D. 1997. Subcuticular bacteria from the brittle star Ophiactis balli (Echinodermata: Ophiuroidea) represent à new lineage of extracellular marine symbionts in the alpha subdivision of the class Proteobacteria. Applied and Environmental Microbiology 63; 1721-1724. CHUN, J. 1995, Computer-assisted classification and identification of actinomycetes. PhD thesis, University of Newcastle-upon-Tyne: Newcastle, DUBILIER, N., GIERE, O., DISTEL, D.L: & CAVANAUGH, C.M. 1995, Characterization of chemoautotrophic bacterial symbionts in a guiless marine worm Oligochaeta, Annelida) by phylo- genetic 165 rRNA sequence analysis and m situ hybridivation. Applied and Environmental Microbiology 61: 2346-2350. PELSENSTEIN, J. 1985. Confidence limits on phylogenies: an approach using buotstrap. Evolution 39: 753-791, 1993. PHYLIP (Phylogenetic Inference Package), Version 3.5c. (Department of Genetics, University of Washington: Seattle. WA). FITCH, W.M, & MARGOLIASH, E. 1967. Construct jon of phylogenetie trees: a method based on mulalion distances as estimated from cytochrome c sequences is of general applicahilitv. Science 155: 279-284, HANSEN, G.H, € SORHEIM, R. 1991, Improved method for phenotypical characterization of marine bacteria. Journal of Microbiological Methruls 13: 231-24}. HOLT, LG 1986, Bergey’s Manual of Systematic Bacteriology. (Williams & Wilkins: Baltimore). LEIFSON, E. 1963, Determination ol carbohydrate metabolism of marine bacteria, Journal or Bacteriology 85: 1183-1184. JENSEN. PR.: & FENICAL, W. 1994. Strategies tor the discovery of secondary metabolites trom marine bacteria: ecological perspectives. Annual Review of Microbiology 48: 559-584. 1995. The relative abundance and seawater requirements of gram-positive bacteria in mear shore tropical marine samples. Microbial Ecalogy 29: 249-257, JUKES, TH. & CANTOR, CR 1969, Evolution of protein molecules. Pp, 21-132, In Munro, HIN, 74 MEMOIRS OF THE QUEENSLAND MUSEUM (ed. Mammalian protein metabolism. (Academic Press: New York), KAZLAUSKAS, R, MURPHY, PT, WELLS, Rub, NOACK, Kau OBERHANSLI, WE. & SCHONHOLZER, P. 1979, A new series of diterpenes from Australian. Spongia species. Australian Journal of Chemistry 32; 867-880, KELLY, M.S. & McKENZIE, J.D, 1993. Survey of the occurrence and morphology of subecuticular bacteria in shelf echinoderms from the north-enst Atlantic Qeean. Marine Biology 123: 741-756, KLUGE, A.G. & FARRIS, ES. 1969, Quantitative phyletics and the evolution of annurans. Systematic Zoology 18: 1-32. LEMOS, M11... TORANZO, A.E. & BARTA, J.L. 1985. Modified medium foe the oxidation-fermentation lest in Ihe identification o marine bacteria. Applied. and Environmental Microbiology 49; 1541-1543 LOPES, J.V., MeCARTHY, PJ, JANDA, KE, WILLOUGHBY, R Æ POMPONI, S.A. 1999, Molecular techniques reyeal wide phyletic diversity of heterotrophic microbes associated with Discodermia spp, (Porifera, Demospongiae). Memoirs of the Queensland Museum (this volume). MAIDAR, B.L.. COLE, JR. PARKER, C.T. JR. GARRITY, G.M, LARSEN, N,, Ll, B,, LILBURN, TG, MeCAUGHEY, M.J., OLSEN, G.J., OVERBEEK, R., PRAMANIK, S, SCHMIDT, T.M., TIEDJE, LM. & WOESE, C.R., 1999. A new version of the RDP (Ribosomal Database Project) Nucleic Acids Research 27: 171-173. MAIDAR, B.L., OLSEN, G.J., LARSEN, N., OVERBEEK, R., MCCAUGHEY, M.J, € WOESE, C.R. 1997. The RDP (Ribosomal Database Project). Nucleic Acids Research 25: 109-111, MOYER, E.L., DOBBS, F.C, & KARL, DM. 1995 Phylogenetic diversity of the bacterial community from a microbial mal at an active, hydrothermal vent system, Loihi Scamount, Hawaii. Applied . und Environmental Microbiology 61: 1555-1562. NUBEL, U., GARCIA-PICHEL. 1, de MUYZER, G. 1997. PCR primers to amplify 165 rRNA genes Irom cyanobacteria. Applied and Environmental Microbiology 63: 3327-3332 OHTA, S, UNO, M., YOSHIMURA, M., HIRAGA, Y, & IKEGAMI.S. 1996. Rhopaloie acid A: A novel norsesterterpene from a marine sponge, Rhopaloeides sp.. which inhibits gastrulation of starfish embryos, Tetrahedron Letters 37: 2265-2266. PORTER, K.G. & FEIG, Y.S. 1980. The use of DAPI for identifying and counting aquatic microflora, Limnology and Oceanography 25: 943-948, RAPPE, M.S, SUZUKI, M.T, VERGIN, K.L. «e GIOVANNONI, S.J. 1998, Phylogenetic diversity of ultraplankton plastid small-subunit rRNA genes recovered in environmental nucleic acid samples from the Pacific and Atlantic coasts of the United States. Applied and Environmental Microbiology 64; 204-303, REYSENBACH, A.-L. GIVER, L-J., WICKHAM, G.S. & PACE, NLR, 1992, Differential amplilication ol rRNA genes by polymerase chain reaction, Applied and Environmental Microbiology 58: 3417-3418. SALTOU, N. & NEL, M. 1987. The neighbor-joining method: a new method for reconstructing phylogenctic trees. Molecular Biology and Evolution 4: 406-425, SANTAVY, D.L. 1985. The symbiotic relatioaship between a blue-pigmented bacterium and the coral ree? sponge, Terpios gramilosa, Pp, 135-140, In Harmelin Vivien, M. & Salvat, B. (eds) Proceed- ings of the Fifth International Coral Reef Congress. Vol, 5, (Antenne Muscum-Ephe: Moorea, Tahiti), SANTAVY, DL, & COLWELL, R.R. 1990. Cumpar- ison of bacterial communities associated with the Caribbean selerosponge, Ceratoporella nicholson, Marine Ecology Progress Series 67: 73-82, SANTAV Y. D.L.. WILLENZ, Ph. & COLWELL, R.R. 1990, Phenotypic study. ol bacteria associated with the Caribbean sclerosponge, Ceratoporella nicholsoni. Applied and Environmental Microbiology 56: 1750-1762. SHAW, M,E, 1927, Note on the inclusion of sand in sponges. Annals and Magazine of Natural Historv (9)19: 601-609, SHIEH, W.Y. & LIN, YM 1994, Association of heterotrophic nitrogen-fixing bacteria with à marine sponge of Halichondria sp, Bulletin of Marine Science 54: 557-564. SOKAL, RR & MICHENER. C.D, 1958, A statistical method for evaluating systematic: relationships. University of Kansas Science Bulletin 38: 1409-1438, STANIER, RY., KUNISAWA, Ra MANDEL, M. & COHEN-BAZIRE. G. 1971. Purification and properties of unicellular hlue-green algae (order Chroococcales): Bacteriological Reviews. 35: 171-205. STIERLE, AC.. CARDELLINA, J.H.l. & SINGLETON, F.L. 1988. A marine Mierococeus produces metabolites ascribed tu the sponge Tedania ignis. Experientia 44: 1021. THOMPSON, J.E.. MURPHY, P.T., BERGQUIST, PR. & EVANS, E.A. 1987. Environmentally induced variation in diterpene composition of the marine sponge Rhapalogides. odorabile. Biochemical Systematics and Ecology 15: 595-606. VACELET, J. 1970, Description dé cellules a bacteries intranuclesires chez des eponges Verongiu. Journal de Microscopie, Paris 9: 333-346, 1975, Etude en microscopie electronique de l'assoc- iation entre bacteries et spongiaires du genre Verongia (Dictyoceratida), Journal de Microscopie et de Biologie Cellulaire 23: 271-288. VACELET, J. & DONADEY, C. 1977, Electron microscope study of the association between some sponges and bacteria. Journal of Experimental Marine Biology and Ecology 30: 301-314, GBR SPONGE MICROBIAL SYMBIONTS 75 WATERBURY, J.B. & STANIER, R.Y. 1978. Patterns of growth and development in pleurocapsalean cyanobacteria. Microbiological Reviews 42: 2-44, WEIDNER, S., ARNOLD, W. & PUHLER, A. 1996. Diversity of uncultured microorganisms associated with the seagrass Halophila stipulacea estimated by restriction fragment length polymorphism analysis of PCR-amplified 16S rRNA genes. Applied and Environmental Microbiology 62: 766-71. WEISBURG, W.G., BARNS, S.M., PELLETIER, D.A. & LANE, D.J. 1991. 16S ribosomal DNA amplification for phylogenetic study. Journal of Bacteriology 173: 697-703. WILKINSON, C.R. 1978a. Microbial associations in sponges. I. Ecology, physiology and microbial populations of coral reef sponges. Marine Biology 49: 161-167. 1978b. Microbial associations in sponges. II. Numerical analysis of sponge and water bacterial populations. Marine Biology 49: 169-176. 1978c. Microbial associations in sponges. III. Ultra- structure of the in situ associations of coral reef sponges. Marine Biology 49: 177-185. 1992. Symbiotic interactions between marine sponges and algae. Pp. 112-118. In Reisser, W. (ed.) Algae and symbioses: plants, animals, fungi, viruses, interactions explored. (Biopress Ltd.: Bristol). WILKINSON, C.R., GARRONE, R. & VACELET, J. 1984. Marine sponges discriminate between food bacteria and bacterial symbionts: electron microscope, radioautography, and in situ evidence. Proceedings of the Royal Society London (B) 220: 519-528. WILKINSON, C.R., NOWAK, M., AUSTIN, B. & COLWELL, R.R. 1981. Specificity of bacterial symbionts in Mediterranean and Great Barrier Reef sponges. Microbial Ecology 7: 13-21. WILLENZ, Ph. & HARTMAN, W.D. 1989. Micromorphology and ultrastructure of Caribbean sclerosponges. I. Ceratoporella nicholsoni and Stromatospongia norae (Ceratoporellidae- Porifera). Marine Biology 103: 387-402. ZAR, J.H. 1984. Biostatistical analysis. 2nd ed. (Prentice Hall International: Englewood Cliffs, NJ). ZILINSKAS, R.A., COLWELL, R.R., LIPTON, D.W. & HILL, R.T. 1995. The global challenge of marine biotechnology. (Maryland Sea Grant: College Park, MD). ZOBELL, C.E. 1946. Marine Microbiology. (Chronica Botanica: Waltham, MA). 76 MEMOIRS OF THE QUEENSLAND MUSEUM DISCOVERY AND SUSTAINABLE SUPPLY OF MARINE NATURAL PRODUCTS AS DRUGS, INDUSTRIAL COMPOUNDS AND AGRO- CHEMICALS: CHEMICAL ECOLOGY, GENETICS, AQUACULTURE AND CELL CULTURE. Memoirs of the Queensland Museum 44: 76. 1999:- Using chemical ecological clues, it is now possible to target habitats and eco-taxonomic groups of marine organisms to increase the likelihood of discovery of species which elicit natural compounds with chemotherapeutic or industrial application. Using the same clues, combined with Geographic Information System interrogation of the benthic geomorphology and oceanography associated with target species, it is possible to identify locations allowing recollection of species of interest. The information gained from both primary collections and focused recollections, provides the basis for hypothesis-driven experiments examining sustainable supply options for extracted target metabolites where synthesis is not practicable. We describe recent results from an integrated multi- disciplinary programme designed to develop sustainable production options for a variety of marine natural products that have interesting biological activities. Three species of sponge from the genera Lissodendoryx, Mycale and Latrunculia, produce novel metabolites with anti-tumour activity. The natural abundance of each would not support a prod- uction industry based on wild harvest should their metabolites be required for drug production. Each has been successfully cultured in-sea demonstrating very good to excellent growth parameters. Each can be cultured with maintenance of target metabolite biosynthesis. In addressing the question of how to optimally produce target compounds, it has been necessary to examine a number of key biological issues pertaining to each species. These include genetic identity of populations supplying seed material, correlates with variable target metabolite biosynthesis in natural populations, origin of target metabolite biosynthesis (symbiont or sponge), and the efficacy of artificial production techniques (sea or land aquaculture or cell culture). We conclude that the guess-work can now be taken out of artificial culture of sponges with a view to produce desirable natural products. It is possible to select for a high yielding culture stock and provide techniques to enhance biosynthesis or target meta- bolites. O Porifera, marine natural products, aquaculture, genetics, cell culture. C.N. Battershill*, MJ. Page, A.R. Duckworth (email: a.duckworth@niwa.crinz) & KA. Miller**, National Institute of Water and Atmospheric Research, P.O.Box 14-901, Kilbirnie, Wellington, New Zealand; P.R. Bergquist, School of Biological Sciences, Auckland University, Private Bag, Auckland, New Zealand: J.W. Blunt, M.H.G. Munro, Chemistry Department, University of Canterbury, Private Bag, Christchurch. New Zealand; P.T. Northcote, Chemistry Department, Victoria University, Private Bag, Wellington, New Zealand; D.J. Newman, Natural Products Branch, Bldng 1052, Rm109, Box B, National Cancer Institute, Frederick, MD 21702-1201, USA; S.A. Pomponi, Harbor Branch Oceanographic Institute, 5600 Old Dixie Highway, Fort Pierce, FL 34946, USA; Present addresses: *Australian Institute of Marine Science, PMB 3, Townsville MC, Qld, 4810, Australia; **University of Wollongong, Northfields Ave, Wollongong, NSW 2500, Australia; 1 June 1998. CHARACTERIZATION OF CALCIUM- BINDING MATRIX PROTEINS FROM DISTINCT CORALLINE DEMOSPONGES. Memoirs of the Queensland Museum 44: 76. 1999:- Calcified sponges played an important role as reef building organisms during different geological time periods. Living relatives of this group investigated here, Spirastrella (Acanthochaetetes) wellsi, Astrosclera willeyana and Vaceletia n. sp., can be found in cryptic niches of indopacific coral reefs. The first known relatives of some of these sponges are known since the upper permian.The mode of biomineralization of the examined species seems to be extremely conservative, since they are phylogenetically very old and exhibit merely minor alterations in their calcareous skeletons. Each of the three species exhibits a unique type of basal skeleton with its own specific modifications of carbonate crystals. Each species was shown to have a specific array of calcium-binding macromolecules enclosed within its intraskeletal matrix, The proteins are separated by SDS polyacrylamide gel electrophoresis. A single protein was detected in S. wellsi, two proteins in A. willeyana, and four proteins in Vaceletia n. sp.. All proteins were characterized by their molecular weight and isoelectric point. The soluble matrix constituents of each species were tested for their potential to decrease precipitation of calcium and strontium carbonate, respectively, in a saturated solution, The findings strongly suggest that these soluble proteins function as the template for skeletal formation and are responsible for determining the particular type of calcium carbonate polymorphs. O Porifera, biomineralization, organic matrix, calcium-binding proteins, calcite, aragonite. Matthias | Bergbauer (email: berggaji@ mailszrz.zrz.tu-berlin.de), Robert Lange, Ulrich Szewzyk, FG Ökologie der Mikroorganismen, Franklinstr. 29, OE5, Technische Universität Berlin, 10587 Berlin, Germany; Joachim Reitner, Inst. & Museum fiir Geologie & Paláontologie, Goldschmidstr. 3, Universitat Góttingen, 37077 Gottingen, Germany; 1 June 1998 BIOLOGY OF THE MASSIVE SYMBIOTIC SPONGE CLIONA NIGRICANS (PORIFERA: DEMOSPONGIAE) IN THE LIGURIAN SEA B. CALCINAI, C. CERRANO, G. BAVESTRELLO AND M. SARA Calcinai, B., Cerrano, C., Bavestrello, G. & Sara, M. 1999 06 30: Biology of the massive symbiotic sponge Cliona nigricans (Porifera: Demospongiae) in the Ligurian Sea. Memoirs of the Queensland Museum 44: 77-83. Brisbane. ISSN 0079-8835. Cliona nigricans is a boring Atlanto-Mediterranean sponge, which on the Gallinara Island cliffs (Ligurian Sea, Italy), exhibits different growth forms: endolithic specimens bore the coralligenous cliff whereas massive specimens grow on the detritic bottom. In the latter habitat, large massive specimens of C. nigricans live partially burrowed in sediment. The sponges incorporate large amounts of sediment, selecting the greater size classes (>5mm). Several incorporated carbonatic fragments, particularly mollusc shells, are bored and crossed by canals of the aquiferous system. The distribution of the massive specimens of C. nigricans is affected by the distribution of coarser fractions in the sea bottom sediments. On the detritic bottom C. nigricans produces a large extension of secondary solid substrata, hosting a rich biocoenosis of sessile and vagile organisms. Differences in the structure of the aquiferous system between boring and massive stages are shown by corrosion casts, particularly in regard to the shape of exhalant canals. Boring forms possess cylindrical canals while those in massive specimens are moniliform. The density of the symbiotic zooxanthellae, evaluated by chlorophyll analysis of sponge papillae, is related to the seasonal solar radiation and depth. O Porifera, Cliona nigricans, growth forms, incorporation, selectivity, boring sponges, zooxanthellae, Ligurian Sea, Italy. Barbara Calcinai (email: zoologia@unige.it), Carlo Cerrano & Michele Sarà, Dipartimento per lo Studio del Territorio e delle sue Risorse dell’ Universita di Genova, Via Balbi 5, 16126 Genova, Italy; Giorgio Bavestrello, Istituto di Scienze del Mare, Via Brecce Bianche, 60131 Ancona, Italy: 16 March 1999. There are some species of boring sponges that develop different growth strategies, during their life cycles. After larval fixation, young boring sponges live endolithically with inhalant and exhalant papillae arising from the bored substratum (a form); in the following stage (B form), the papillae progressively form a thin sheet of sponge tissue; when the calcareous substratum is entirely etched away, the sponge grows into a massive form (y) (Sarà & Vacelet, 1973). The Gallinara Island (Ligurian Sea, W Mediterranean) hosts a dense population of Cliona nigricans which grows from the surface level to the detritic bottom (40-50m depth). The coralligenous cliffs are strongly eroded by C. nigricans (a. and D forms) producing large tunnels which weaken and fracture the bioherme. At the base of the island cliffs, on the detritic bottom, the x form of this species grows. The a and p forms are morphologically very different from the x form, particularly in their exhalant papillae which, in massive forms, have oscular chimneys higher than 10cm. In spite of these morphological differences, electrophoretical analysis has clearly proven that the two forms belong to the same species (Bavestrello et al., 1996a). All morphotypes of C. nigricans harbour zooxanthellae. This symbiosis is known to be present only in a small number of sponge species (Sara & Liaci, 1964; Sara, 1966; Riitzler, 1985) whose boring ability has been correlated to the presence of the symbionts (Hill, 1996; Vacelet, 1981). In this work we consider the relationships of the massive x forms with the bottom sediment and their influence on the bottom communities in providing secondary solid substrata. Moreover, we compare the symbiont density in these sponges and the anatomy of their aquiferous system to those of boring specimens. MATERIALS AND METHODS Cliona nigricans was studied at Gallinara I. (Ligurian Sea), situated about 1.5km from the coast, with underwater cliffs reaching a maximum depth of 37m on the southern side, and a Posidonia oceanica bed located between the northern side of the Island and the coast (Fig. 1). 78 MEMOIRS OF THE QUEENSLAND MUSEUM Gallinara Island FIG. 1. Density of massive specimens of Cliona nigricans (gray areas) around Gallinara I, Key: dark grey >20 specimens/10m?; light grey 5-10 specimens/10m2. The Island consists of greyish quartzitic rocks, together with pelitic layers and cretacic pudding stones (Balduzzi et al., 1994), The density of massive C. nigricans specimens was determined along the eastern, southern and western sides of the islands (Fig. 1) on the detritic bottom at 40-45m depth. Densities were evaluated directly under water by counting the specimens present in a rectangular frame of 10m?. The size of some specimens was estimated under water measuring the two main axis of their surface cleaned by sediments. Moreover, the thickness of sediments covering the sponges was measured, The granulometric features of the bottom detritus (obtained by sieving) were studied on samples collected in areas where sponge density was higher and, for comparison, in areas where massive sponges were absent. In addition, the granulometric characteristics of the bottom sediments were compared to those of the sediments incorporated by the sponges by dissolution of sponge tissues in H20; (120 vol.). The environmental sedimentation rate was estimated by placing four conical sediment traps in the area with the highest sponge density. To verify the role of C. nigricans in potentially harbouring macrobenthic organisms, specimens were photographed and collected for direct observation in the laboratory. Samples were enclosed in plastic bags and fixed directly under- water in 4% formalin. Variation in the density of symbiotic zooxanthellae population in C. nigricans were determined from fresh tissue samples, taken from peripheral portions of sponges (especially papillae). Sampling took place at different seasons (with five collections made between October 1997 to August 1998), for all morphs and along a bathymetric transect (with five replicas per specimen at 5, 10, 20, 30,37 and 42m depth). Spectrophotometric analyses of acetonic extracts of sponge tissue were conducted according to Gilbert & Allen (1973) to quantify chlorophyll-a concentrations. Anatomical differences of the aquiferous system in different growth forms were evaluated using corrosion casts (Bavestrello et al., 1995a; Burlando et al., 1990) which were studied under stereomicroscope and, ultrastructurally, by SEM. RESULTS Large, massive specimens of C. nigricans, growing on the soft detritic bottom, are cushion- shaped or lobate, with a characteristic mamillate surtace (Fig. 2C). Their maximal surface ranges from about 200-1000cm? and they are buried in the sediments up to 3-5cm deep. Their inhalant papillae are similar in size and shape to those of endolithic forms (Fig. 2A,B), whereas very high oscular chimneys (up to 10cm high) constitute the exhalant structures. In many specimens the inhalant papillae develop on the wall of the oscular chimney (Fig. 2D). Tissues of these massive forms are very rich in incorporated bottom sediments which constitute 95% of the sponge dry weight. The mechanism of incorporation appears to be non-selective regarding the origin of foreign bodies: 1.e. quartzitic or pelitic particles, rhodoliths, and organogenous detritus are collectively ingested (Fig. 3A-C). Nevertheless, the comparison between the granulometries of bottom sediment and those of sediments incorporated by the sponges, clearly indicates that C. nigricans actively selects the coarse fractions larger than 5mm diameter (Fig. 3D). Corrosion casts demonstrate that calcareous fragments incorporated in the bodies of massive sponges are bored and often crossed by the canals BIOLOGY OF CLIONA NIGRICANS FIG. 2. Cliona nigricans specimens. A, Boring a stage. B, Boring f stage. C, Massive y stage specimens on the detritic bottom characterised by high oscular chimneys and mamillate surface. D, Epibiotic bryozoan Schizobrachiella sp.). E, Epibiotic serpulid Filograna sp. (E) growing on massive specimens. 80 MEMOIRS OF THE QUEENSLAND MUSEUM pa © Sediment in Cliona e $0 = Bottom sediment Frequencies (o) <63um 63pm 125 pm 250 um 500 um 1mm 2mm >5mm Size classes FIG. 3. Incorporation of foreign material by massive specimens of Cliona nigricans. A-C, Foreign material incorporated by different specimens. D, Size frequency distribution of material incorporated by sponges compared to that occurring in the surrounding sea bed. E, Fraction of the bottom sediment with a size >Smm occurring where sponge density is highest (left), compared to the same fraction in an area without sponges (right). Scale bars in cm. of the aquiferous system (Fig. 4A). A comparison between free bottom sediments and those entrapped within the sponge reveals erosion traces in the latter (Fig. 4B-C). Coarse sediments incorporated by massive specimens derive from fragmentation ofthe over- hanging cliffs, while the thin fraction of terrigenous origin, collected by the traps, reveals an average sedimentation rate of about 10kg/m*/ year. The sponge population density is related to the amount of coarse sediment fraction present in the bottom sediments (Fig. 3E), In areas where sponges show a density greater than 20 specimens/l0m” the coarse matter represents 10-15% of the bottom sediments. By comparison, sponges are scarcer (5-10 specimens/10m*) in FIG. 4. Boring activity of massive specimens of C. nigricans. A, Portion ofa corrosion cast of'a massive specimen showing a calcareous fragment incorporated by the boring sponge and crossed by an exhalant canal. Scale bar 3mm. B-C, Comparison of the same granulometric fraction of the bottom sediments (B), with those incorporated by the sponges (C). The latter shows evidence of the perforations produced by the sponge (arrow). Scale bar = 8mm. areas where the coarse fraction is 3-7% of the total, and they are absent where the coarse fraction is less than 3%. On the soft bottoms of the Gallinara I. massive specimens of C. nigricans occupy a large surface on the soft bottom on which they constitute a secondary solid substratum, where a coralligenous-like assemblage lives. This assem- blage (Fig. 2C-E) is mainly composed of sessile organisms such as other sponges, hydroids, anthozoans, bryozoans and serpulids that, in turn, BIOLOGY OF CLIONA NIGRICANS 81 250 N o o 150 Chlorophyll-a (ug/g) 50 B 10m B5m B 20m B 30m Depth FIG. 5. Average chlorophyll-a concentrations during different periods of the year in specimens living at different depths (N = 5). Key: B = boring specimens; M — massive specimens. support a vagile fraction represented mainly by nudibranchs, polychaetes, harpacticoids, amphipods and decapods (Table 1). The concentration of symbiotic zooxanthellae does not vary significantly among massive and boring specimens, but rather exhibits a trend influenced both by season and depth distribution. In October the values are homogeneously low among the different growth forms and at different depths. In March and May these values progress- ively increase, then subsequently decrease in the following summer months. In all sampling periods a peak in values always occurs at 20m depth (Fig. 5). Corrosion casts ofthe aquiferous system reveal differences between the massive and boring specimens in the shape of their exhalant canals. These canals are cylindrical in endolithic specimens (Fig. 6A) and moniliform in massive ones (Fig. 6B). Moreover, endolithic sponges differ from massive ones in the arrangement of choanocyte chambers, which are clustered inside the boring chambers. In massive forms the choanocyte chambers are homogeneously distributed in the sponge body. DISCUSSION It is probable that the initial stages of larval development in Cliona nigricans, as in all clionids (Sarà & Vacelet, 1973), are linked to the boring activity on a suitable substratum. In coastal detritic bottoms, however, the carbonate fragments are small, and sponge size exceeds the bored fragment very precociously. From this stage, sponge growth is linked to the B37m incorporation of sediment into its tissue. It is also possible that massive sponges living on the soft detritic bottom originate from the boring specimens higher up on the cliffs which, through fragment- ation of the substratum, fall down with a portion ofthe sponge tissue. In this case, the activity of the boring specimens of C. nigricans is an agent for asexual repro- duction and spatial dispersal. Our data indicate that massive specimens of C. nigricans select larger fractions (>5mm) of sediment and that high concen- trations of these coarse sediments are necessary for successful sponge development. Cellular mechanisms which control this selection are still poorly known (Teragawa, 1986; Bavestrello et al. 1998). The selective ability of sponges to incorporate foreign matter is currently a subject of debate in the literature, with empirical support only recently available (Bavestrello et al.,1995b, 1996b). Studies in chlorophyll-a concentrations in C. nigricans give some indication ofthe quantitative changes in the symbiotic community of zooxan- thellae in relation to depth and seasonal variation. The zooxanthellae population correlates more to the seasonal cycle rather than to depth. Only in M 42m TABLE 1. List of the main phyla living on massive Cliona nigricans as sessile epibionts. Key: + occasional; ++ common; +++ present on almost each specimen. Phylum Species Abundance Porifera Oscarella lobularis * Dysidea sp. B Cnidaria Clythia hemisphaerica dH C. linearis ++ Paralcyonium sp. > Caryophyllia smithi ++ Bryozoa Smittina cervicornis + S. mammillata + Schizobrachiella sp. H+ Hippellozoon mediterraneus + Schizomavella auricolata +++ Turbicellepora avicularis +++ Polychaeta |Filograna sp. + Serpula vermicularis ++ Tunicata Halocynthia papillosa + 82 MEMOIRS OF THE QUEENSLAND MUSEUM FIG. 6. Corrosion casts made of the aquiferous system of C. nigricans. A, Cylindrical shape of boring specimens. Scale bar 250um. B, Characteristic moniliform structure of the exhalant canals of massive specimens. Scale bar = 100um. autumn, the density of zooxanthellae population in massive forms is similar to the density in all boring samples independent from depth. During spring zooxanthellae density increases in all of the sponge morphotypes reaching its maximum in May, in samples collected from 20m depth, and decreasing, subsequently, during the summer. These data indicate significant differences in the behaviour between the zooxanthellae of C. nigricans and the cyano- bacteria of Petrosia ficiformis in the same area (Bavestrello et al., 1992). Cyanobacteria density in P. ficiformis is very sensitive to light variations related to depth and, from 10 to 40m, the chlorophyll concentration decreases by about four times. In contrast, the zooxanthellae population in C. nigricans remains relatively constant, suggesting a control of the host cells on their reproduction, as in other algae-invertebrate symbioses (e.g. Cook, 1983). Rosell (1993) showed how reproductive process (sexual and asexual) can modify the density of zooxanthellae in C. viridis populations through digestion or expulsion, and how at the end of the sexual process few zooxanthellae were present. Further data are necessary to clarify how reproduction periods affect the population of symbiotic zooxanthellae in C. nigricans. Hill recently (1996) showed how symbiotic zooxanthellae are related to boring activities and growth of a tropical boring sponge (Anthosigmella varians). Vacelet (1981) also demonstrated that the most active boring sponges harbour zooxanthellae. Some authors (Hartman, 1958; Sara & Vacelet, 1973) suggested a decrease in the boring power of endolithic versus massive growth forms, whereas our data indicate that even if fragments, incorporated by massive forms, are widely bored,both endolithic and massive morphotypes have comparable amounts of zooxanthellae. Differences between the structure of the aquiferous system of boring and massive morphotypes were found through the study of their corrosion casts. The particular beaded shape of the exhalant canals in massive specimens may be determined by a system of contractile elements regularly disposed along the endo- pinacoderm of the canals. The two alternative forms of C. nigricans (endolithic and massive) are linked to different habitats (coralligenous cliffs and detritic bottoms, respectively), and may be considered in the context of developmental modulation (Smith-Gill, 1983). Morphological variability is common among many sponge species (e.g. Barthel, 1991; Bavestrello et al., 1992), and is generally thought to be linked to variations in the intensity of water movement influencing food supply, and the probability of re-inhalation of filtered waste-water (Fry, 1979). The differences we observed in the behaviour of sponges in relation to their choice of substrata, and of their pumping physiology, between endolithic and massive form of C. nigricans stress this variability. From an ecological perspective C. nigricans from Gallinara I. impacts on its environment in two alternative ways: 1) on the cliffs, where boring activity destroys the calcareous substrata, causing fragments to roll down onto the sea floor; and 2) on the detritic bottom, where this same material is gathered up by the massive speci- mens, which in turn provide a secondary hard BIOLOGY OF CLIONA NIGRICANS substratum that hosts an unusual biocoenosis, otherwise not present on the soft bottom. ACKNOWLEDGEMENTS This work was financially supported by Italian MURST funds. We wish to thank the anonymous referees who provided numerous and useful suggestions. LITERATURE CITED BALDUZZI, A., BIANCHI, C.N., CATTANEO- VIETTI, R., CERRANO, C., COCITO, S., COTTA, S., DEGL'INNOCENTI, F., DIVIACCO, G., MORGIGNI, M., MORRI, C., PANSINI, M., SALVATORI, L., SENES, L., SGORBINI, S. & TUNESI, L. 1994. Primi lineamenti di bionomia bentonica dell'Isola Gallinaria (Mar Ligure). Atti del X Congresso A.1.0.L.: 603-617. BARTHEL, D., 1991. Influence of different current regimes on the growth form of Halichondria panicea Pallas. Pp. 387-394. In Reitner, J. & Keupp, H. (eds) Fossil and recent sponges (Springer-Verlag: Berlin). BAVESTRELLO, G., ARILLO, A., BENATTI, U., CERRANO, C., CATTANEO-VIETTI, R., CORTESOGNO, L., GAGGERO, L., GIOVINE, M., TONETTI, M. & SARA, M. 1995b. Quartz dissolution by the sponge Chondrosia reniformis (Porifera, Demospongiae), Nature, London 378: 374-376. BAVESTRELLO, G., ARILLO, A., CALCINAI, B., CERRANO, C., LANZA, S., SARA, M., CATTANEO-VIETTI, R. & GAINO, E. 1998. Siliceous particles incorporation in Chondrosia reniformis (Porifera, Demospongiae). Italian Journal of Zoology 65: 343-348. BAVESTRELLO, G., BURLANDO, B. & SARA, M. 1995a. Corrosion cast reconstruction of the three- dimensional architecture of Demospongiae canal system. Body cavities: function and phylogeny. Pp 93-110. In Lanzavecchia, G. Valvassori, R. & Candia Carnevali, M.D. (eds) Selected Symposia and Monographs. Vol. 8 (Unione Zoologica Italiana, Mucchi: Modena). BAVESTRELLO, G., CALCINAI, B., CERRANO, C., PANSINI, M. & SARA, M. 1996a. The taxonomic status of some Mediterranean clionids (Porifera: Demospongiae) according to morpho- logical and genetic characters. Bulletin de l'Institut Royal des Sciences Naturelles de Belgique, Biologie 66 (Suppl.): 185-195. BAVESTRELLO, G., CERRANO, C., CATTANEO- VIETTI R., CALABRIA F., CORTESOGNO L. & SARA, M. 1996b. Selective incorporation of foreign material in Chondrosia reniformis. Italian Journal of Zoology 63: 215-220. oo 3 BAVESTRELLO, G., PANSINI, M., PRONZATO, R., CATTANEO-VIETTI, R. & SARA, M. 1992, Variazioni della concentrazione di clorofilla-a in Petrosia ficiformis (Porifera, Demospongiae) con cianobatteri simbionti. Atti del X congresso A.LO.L.: 327-331. . BURLANDO, B., BAVESTRELLO, G. & SARA, M. 1990. The aquiferous system of Spongia officinalis and Cliona viridis (Porifera) based on corrosion cast analysis. Bollettino di Zoologia 57: 233-239. COOK, K.B. 1983. Metabolic interaction in algae- invertebrate symbiosis. Pp. 177-210. In Jeon, K.W. (ed.) International Review of Cytology. Supplement 14. (Academic Press: London, New York, San Francisco). FRY, W.G. 1979. Taxonomy, the individual and the sponge. In Larwood, G. & Rosen, B.R. (eds) Biology and Systematics of colonial organism (Academic Press: London, New York, San Francisco). GILBERT, J.J. & ALLEN, L.H. 1973. Chlorophyll and primary productivity of some green fresh-water sponges. Hydrobiology 58: 633-658. HILL, M.S. 1996. Symbiotic zooxanthellae enhance boring and growth rates of the tropical sponge Anthosigmella varians forma varians. Marine Biology 125: 649-654. HARTMAN, W.D 1958. Natural history of the marine sponges of Southern New England. Bulletin Peabody Museum of Natural History, Yale University 12: 1-155. ROSELL, D. 1993. Effects of reproduction in Cliona viridis (Hadromerida) on zooxanthellae. Scientia . Marina 57: 405-413. RUTZLER, K. 1985. Association between Caribbean sponges and photosynthetic organisms. Pp. 455-466. In Rützler, K. (ed.) New perspectives in sponge biology. (Smithsonian Institution Press: Washington D.C.). SARA, M. 1966. Associazioni fra Poriferi e alghe in acque superficiali del litorale marino. Ricerca Scientifica 36: 277-282. SARA, M. & LIACI, L 1964. Symbiotic association between zooxanthellae and two marine sponges of the genus Cliona. Nature, London 203: 321. SARA, M., & VACELET, J. 1973. Ecologie des Demosponges. Pp. 462-576. In Brien, J.P. et al. (eds) Traité de Zoologie. NI. Spongiaires. Sér. Ed. Grassé, P. (Masson et Cie: Paris). SMITH-GILL, S.J. 1983. Developmental plasticity: developmental conversion versus phenotypic modulation. American Zoologist 23: 47-55. TERAGAWA, C.K. 1986. Sponge dermal membrane morphology: Histology of cell mediated particle transport during skeletal growth. Journal of Morphology 190: 335-348. VACELET, J. 1981. Algal-sponge symbioses in the coral reefs of New Caledonia: a morphological study. Proceedings of the Fourth International Coral Reef Symposium, Manila 2: 713-719. 84 MEMOIRS OF THE QUEENSLAND MUSEUM CARBON ISOTOPE TIME SERIES OF CORALLINE SPONGES FROM THE CORAL SEA, PHILIPPINES AND CARIBBEAN. Memoirs of the Queensland Museum 44; 84. 1999:- Live coralline sponges (Ceratoporella nicholsoni, Astrosclera willevana, Spirastrella (Acantho- chaetetes) wellsi) were collected from reef caves and deeper reef slopes of the Caribbean, the Visaya Sea (Philippines) and the Coral Sea (Great Barrier Reef). The specimens were dated by cither radiocarbon or uranium-thorium methods. Age ranges were from 200- 600 years. We tested the reproducibility of 3"C values measured on the aragonite of Ceratoporella nichalsani by investigating variations along single layers of a well-laminated specimen. We also compared values measured on the outermost layers of several specimens. The reproducibility for 3''C is excellent in most cases. Only few samples show depletion by up to 0.2 permil. Two parallel transects through a specimen of Astrosclera willeyana also display excellent reproducibility of à C values. All specimens show the well-known industrial decline in 6°C values starting ca. in 1850 A.D (e.g. Druffel & Benavides, 1986; Böhm etal., 1996). ln comparing the magnitude of this decline measured in our samples and in 8"C of atmospheric CO, we can estimate the local degree of isotopic equilibration between atmosphere and sea-waler. We find values range from 40% of the atmospheric change at the Great Barrier Reef and in ihe Philippines to 65% in Jamaica. For each site we compared the preindustrial 8C from total CO; (DIC) of the surface water, calculated from our sponge records, with published phosphate concentrations. The values agree with a high input of nutrient-rich subsurface water at the Philippine site and at the Great Barrier Reef. At the Great Barrier Reeflocal upwelling at the reef front has been reported. However, the measured 8"C values are much lower than expected for average phosphate concentrations. Either the upwelling is much more intense than assumed, or the Astroselera record is afTected by secondary processes and/or a vital/kinetic effect. O Porifera, coralline sponges, Philippines, Great Barrier Reef, Caribbean. Literature cited. BÖHM, F., JOACHIMSKI, M., LEHNERT, H., MORGENROTH, G., KRETSCHMER, W., VACELET, J. & DULLO, W.-C. 1996. Carbon isotope records from extant Caribbean and South Pacific sponges: Evolution of 8"C in surface water DIC. Earth Planetary Science Letters |39: 291-303. DRUFFEL, E.R.M. & BENAVIDES, L.M. 1986. Input of excess CO» to the surface ocean based on &"C/UC ratios in a banded Jamaican sclerosponge. Nature 321: 58-61. Florian Bohm, Wolf-Christian Dullo & Anton Eisenhauer, Geomar, Wischhofstr.1-3 D-24148 Kiel, Germany; Helmut Lehnert, Gert Warheide*, Joachim Reimer (email: jreitne(a)gwdg.de), Institut und Museum für Geologie und Paläontologie, Universitat Göttingen, Goldschmidt- Strasse 3, D-37077 Göttingen, Germany, Michael M. Joachimski, Institut für Geologie, Schloßgarten 5, D-91054 Erlangen Germany: *Present address: Queensland Museum, P.O. Box 3300, South Brisbane, Old. 4101, Australia; 1 June 1998. INCORPORATION OF INORGANIC MATTER IN CHONDROSIA RENIFORMIS (PORIFERA: DEMOSPONGIAE): THE ROLE OF WATER TURBULENCE C. CERRANO, G. BAVESTRELLO, U. BENATTI , R. CATTANEO-VIETTI, M. GIOVINE AND M. SARA Cerrano, C., Bavestrello, G., Benatti, U., Cattaneo-Vietti, R., Giovine, M. & Sara, M. 1999 06 30: Incorporation of inorganic matter in Chondrosia reniformis (Porifera: Demo- spongiae): the role of water turbulence. Memoirs of the Queensland Museum 44: 85-90, Bris- bane. ISSN 0079-8835. The role of sedimentation and sea conditions in relation to the amount of the foreign matter (sand grains and opaline sponge spicules) present in the body of the demosponge Chondrosia reniformis was evaluated monthly at two sites, each characterised by different sedimentary conditions along the rocky cliff of the Portofino Promontory (Ligurian Sea). Contrary to the process in keratose (“horny”) sponges, the mineral particles incorporated by Chondrosia are subjected to an evident turnover probably linked to its unusual ability to dissolve quartz. The quantity and size of the particles taken up by the sponge are linked to environmental sedimentation and sea conditions. These data indicate that settlement of particles on the sponge is affected by the stickiness of the sponge’s mucous surface. The large amount of quartz grains continuously incorporated and dissolved by Chondrosia, suggests a possible role played by the sponge in the local silica flux in shallow coastal waters. CJ Porifera, foreign matter, mineral selectivity, uptake, water turbulence, silica. Carlo Cerrano (email: zoologia@igecuniv.csita.unige.it), Riccardo Cattaneo-Vietti & Michele Sara, Dipartimento per lo studio del Territorio e delle sue Risorse dell’ Universita di Genova, Via Balbi 5, I-16126 Genova, Italy; Giorgio Bavestrello, Istituto di Scienze del Mare dell'Università di Ancona, Via Brecce Bianche 60131 Ancona, Italy; Umberto Benatti, Marco Giovine, Istituto Policattedra di Chimica Biologica, Viale Benedetto XV, 1-16132 Genova, Italy; 16 March 1999. Sedimentation on rocky bottoms influences the distribution of organisms, impacting significantly on larval settlement and its further development, and compromising the filtering structures of filter feeders, even with the extreme result of total exclusion from their habitat (Loosanoff & Tommers, 1948; Wilber, 1971; Rogers, 1990). Porifera, living under high sedimentation regimes, might also be subjected to both abrasion by coarse sediment particles and occlusion of inhalant pores by fine ones (Sarà & Vacelet, 1973; Verdenal & Vacelet, 1985). The filtered water volume decreases proportionally to the amount of particulate matter present in the water column; e.g. in Aplysina (=Verongia) lacunosa (Gerodette & Flechsig, 1979). Sponges can live in oligotrophic waters owing to their high filtering efficiency, but cannot survive for long periods of reduced pumping (Reiswig, 1974). Some species have developed defense mechan- isms against high sedimentation, as in the fresh-water species Ephydatia fluviatilis, where amoeboid cells of the exopinacoderm have endo- cytosis capabilities (Willenz & Van de Vyver, 1982) and can remove foreign particles (Harrison etal., 1985). Several species, such as the keratose sponge Dysidea etheria, can select sedimentary particles from their habitat, incorporating proper-sized ones in their primary fibers and removing others through the selective action of their external amoeboid cells (Teragawa, 1986a, 1986b). Chondrosia reniformis does not produce its own spicules but engulfs foreign siliceous mater- ials (i.e. siliceous sponge spicules present in the water column and sand grains), into its collag- enous ectosome. Moreover, it recognises the miner- alogical features of particulate material, dissolving quartz particles and reducing their original size (Bavestrello et al., 1995a, 1996, 1999). The aim of this study is to explore the relation- ships between the amount of allochtonous matter engulfed by C. reniformis during an annual cycle, comparing this to the different sedimentation conditions, which are closely related to the local sea conditions in two different sites of the Porto- fino Promontory (Ligurian Sea, Tigullio Gulf, Italy): Punta del Faro and Paraggi Bay (Fig. 1). These stations are well known from a bio- coenotical (Tortonese 1961; Morri et al., 1986) 86 MEMOIRS OF THE QUEENSLAND MUSEUM Partofino Promontory al "T = [r4 3 E wD, FIG. I. Schematic figure showing the main current patterns in the studied area, At Punta del Faro, the current from the Golfo Tigullio meets the main cyclonic stream of the Ligurian Sea. Paraggi Bay represents a decantation area consequent to an eddy. and a sedimentological point of view (Bavestrello et al., 1991; 1995b). The sediment- ation rate is aboul seven times higher at Paraggi Bay than at Punta del Faro, owing to differences hetween their local hydrodynamic features (Esposito & Manzella 1982; Marullo et al., 1985). In fact, Paraggi represents a decantation area, while Punta del Faro is the meeting point of two currents, one from the Ligurian Sea and the other one flowing outwards from the Tigullio Gulf (Fig. L). MATERIALS AND METHODS At Punta del Faro, where the cliff ends at 55m depth, specimens of C. reniformis were sampled monthly by SCUBA diving during March 1994- June 1995, at depths of 3m, 12m and 25m. At these last two depths two sediment traps, as desc- ribed by Bavestrello et al. (1991), were installed to collect the fraction of sediments available for sponges. At Paraggi Bay, where the cliff ends at 25m depth, sponges were collected from 3m and I 5m depths, and a sediment trap was installed at this last depth only. At both localities, a super- ficial (3m depth) sediment trap was also installed, but strong wave action prevented a sufficient continuity in data collection at this station, At each station, Lcm^ fragments of the sponge ectosome were collected monthly from six speci- mens. To analyze quantity and granulometry each fragment was dissolved in boiling hydrogen peroxide (39% weighi/volume: about 130 vol.). The dissociated foreign material was centrifuged at 5000G for 5mins, washed twice in 95% ethan- ol, resuspended in 0.5ml of 100% ethanol, and finally two subsamples of 0.1m] were mounted on two slides. All particles (sand grains and spic- ules) were counted on each slide, The main axis of 100 sand grains per slide from Punta del Faro specimens was measured using a GRAPHTEX KD 4300 digitiser connected to an IBM PC. The area of incorporated particles was expressed per square centimeter of sponge surface. Sediments collected from traps were evaluated monthly as dry-weight after combustion at 550°C for 4hrs in order to reach the inorganic fraction. Three slides were prepared from each sediment sample to collect the granulometric data. Wave height data (cm above the free sea surface) were kindly provided by the Meteor- ological Observatory of Chiavari. Measurements of wave height commenced 10 days prior to each sampling date in order to compare trends. This period was chosen after initial trials of 7, 10 and 15 days prior to sampling, as it provided the best comparison between environmental conditions and collected sediments. Additionally, investigation of the sponge ecto- some was conducted by SEM analyses to evaluate morphological relationships between sponges and settled sediments, Samples were collected and fixed underwater in 2.5% glutaraldehyde, After rinsing in artificial seawater, samples were dehydrated in an ethanol gradient, followed by critical-point drying in a CO; Pabish CPD apparatus, They were mounted on stubs with FOREIGN MATTER IN CHONDROSIA RENIFORMIS Average wave height (cm) "88858 38% Trapped sediment (g md") FMAM JJ A SONDJSFMAWM J J E E B æ Sand grains Spicules | e $ S Sand grains (N zm?) 8 3 Spicules (N cm?) o SONDJS FMAM J J 1500 7 800 C + Sand grains ] ¡+ Spicules - je a S Sa + + 8 8 Spicules (N cm?) Sand grains (N cm?) SONDJFMAMSJ J + 800 D æ Sand grains 7*- Spicules Sand grains (N cm") Spicules (N cm”) F MAMJ JASONDJFMAMJ J FIG. 2. Punta del Faro. A, Relationship between annual trend in sea conditions (histogram) and sediments collected by traps at 12 and 25m depths. B-D, Foreign matter (spicules and sand grains) incorporated by Chondrosia reniformis at 3, 12 and 25m depths, respectively. silver conducting paint, sputter-coated with gold-palladium in a Balzer Union Evaporator, and observed using a Philips EM 515 electro- probe microscope. To estimate the silica production by C. reniformis in the studied area, the sponge abund- ance and its surface were evaluated by visual census along 10 vertical 1m belt transects from the base of the cliff of the Promontory (50m depth) to the sea surface, following Hiscock’s (1987) method. RESULTS At Punta del Faro, a high energy site, the amount of sediments collected by traps was directly 87 related to the sea conditions at both depths (Fig. 2A). Sand grains and spicules (number cm”) incorporated by Chondrosia reniformis at 3m depth, peaked during periods of calm sea and declined during rough seas (Fig. 2B). Conversely, at intermediate (12m) and deep (25m depth) stations, higher quantities of sand and spicules were incorporated by the sponge during periods of rough seas, when sediment availability was higher (Fig. 2C-D). Similarly, at Paraggi Bay, a decantation site, the amount of sediments collected by traps was strongly related to the sea conditions (Fig. 3A). Both sponge stations (at 3m and 15m depths) showed the same phenomenon as did the most shallow station at Punta del Faro: high values of incorporated particles were recorded during calm periods (Fig. 3B-C), even if the available sedim- entary material was greater during periods of rough seas, as shown by our data on the trapped matter (Fig. 3A). The granulometries of the incorporated sand grains by C. reniformis at Punta del Faro showed a similar trend for all depths sampled (Fig. 4). Average values ranged between 18-51um diam- eter, with maxima occurring in July and November-December and minima occurring mainly from August to October and during winter. Comparison between these granulometries and sea conditions reveals an inverse relationship: large particles were present exclusively following periods of calm water. SEM observations on the intact sponge surface showed that numerous crystals, organised in spherical-like balls (of about 5-15um), and enveloped by a thin mucus web, emerge from the sponge ectosome (Fig. 5). Electroprobe analysis of crystals (indicating silica as the major constituent) and their shape, allowed us to conclude that these are quartz crystals. DISCUSSION Many demosponges are able to incorporate allocthonous inorganic material into their skele- tons, a mechanism that is generally considered to provide additional strength to their organic fibrous skeleton. This phenomenon occurs most widely in the ‘horny’ keratose sponges (Lendenfeld, 1889; Teragawa, 1986a; Pronzato et al., 1998), comprising the orders Dictyoceratida, Dendroceratida and Verongida. In keratose sponges the uptake seems to be irreversible, since foreign matter is cemented into primary fibers. Conversely, Chondrosia reniformis shows an evident turnover 88 50 800 € «tA +15m 1700 a E E Zz «+ e E $ al 500 = 2 4m È gay 3 fi +300 $ El + v E 25 +200 2 2 2 li E E 154 Lo FMAMJJSASONDSFEMAMJ J 3000 + 800 = -® Sand grains| E 2500 + B Spicules E is 800 7 = 2000 E = E E 15004 400 = 5 ys 2 1000 3 5 a a 200 i 500 + 0 4— I—-——L-———-n * + o F MAMJJASONDJFMAM !)J J 3000 7 800 C Sand grains| _ 2500 = Spicules | * 600 .— 5 2000 + E z z 2 1500 + +400 a $ E © 4000 2 * a 5 +200 4 9$ 500 + 04 * * + + - + +—+—_+ 0 F MA MJ JA SON DJFMAM JS J FIG. 3. Paraggi Bay. A, Relationship between annual trend in sea conditions (histogram) and sediments collected by the trap at 15m depth. B-C, Foreign matter (spicules and sand grains) incorporated by Chondrosia reniformis at 3 and 12m depths. of incorporated foreign material and a capability to discriminate amongst the incorporated particles. This finding opens new perspectives on sponge behaviour. Although influenced by environmental parameters, these phenomena suggest a continuous utilisation ofthe incorporated matter as evidenced by the quartz dissolution ability (Bavestrello et al., 1995a), and the production of quartz ‘pellets’ on the ectosome. Annual trends in the amount and size of sedi- mentary matter incorporated by C. reniformis appear to depend mainly on the local sea cond- itions and on the sponge etching. During calm periods, mainly in shallow waters, the sponge also uptakes large particles, as suggested by the inverse relationship between particle size and sea conditions. These phenomena are most evident in the shallow stations, where the highest amounts and largest sizes of incorporated foreign mater- ials are present, corresponding to periods of calm waters. Conversely, during rough periods, the sponge surface is not sticky enough to retain large particles and consequently the quantity and size of engulfed matter decrease. In deeper water, where wave disturbance is reduced and resusp- ension processes are higher, populations of C. MEMOIRS OF THE QUEENSLAND MUSEUM reniformis respond to these environmental cond- itions, incorporating higher amounts of siliceous matter. This is evident at Paraggi Bay, a more protected site than Punta del Faro, where swell conditions are frequent. In this way, it is possible to assume that sea conditions influence this phen- omenon in two ways: on one hand, rough sea conditions can limit the uptake of particles, where the effect of waves action is strong, but on the other hand, the same conditions increase the availability of sediment material, owing to resus- pension processes. This causes a higher amount of incorporated sediments in sponges living in deeper waters, where wave action is not strong enough to detach particles from the sponge surface. In C. reniformis the mechanism of incor- poration of inorganic matter involves different physical, mineralogical, and biological aspects: the settled particles are transferred, at variable speeds, to special areas of the sponge ectosome where they are quickly engulfed and, after incor- poration, the collected material remains scattered in the fibrous ectosome, where particle sizes are re-elaborated (Bavestrello et al., 1995a, 1996, 1999). Selectivity in the incorporation of foreign bodies in sponges has long been debated (Haeckel, 1872; Schulze, 1879; Lendenfeld, 1889; Sollas, 1908; Shaw, 1927; Teragawa 1986a). The uptake of particles in C. reniformis seems to be determ- ined by an active selection of the minerals (Bavestrello et al., 1998b), and a passive one regarding their size. In agreement with Schulze's hypothesis (1879), it is possible that the uptake = La on a I^] 5 8 a a 8 Average grain size (um) = E] Average wave height (cm) a 0 FMAMJJASONDJFMAM4/J J FIG. 4. Relationship between annual trends in sea conditions (histogram) and granulometries of sand grains incorporated by Chondrosia reniformis at 3m (circles), 12m (squares) and 25m (triangles) depths at Punta del Faro (Portofino Promontory). FOREIGN MATTER IN CHONDROS/A RENIFORMIS 89 mechanism is determined by the interaction between the stickiness ofthe sponge surface and the intensity of water movement, and that the biological activity of the sponge towards the quartz particles affects the granulometric trend and the amount of incorporated sediments. An important consequence of this unusual behaviour is the output of dissolved silica, thus biologically available to other organisms. Under experimental conditions (Bavestrello et al., 1995a; 1996), with excess quartz grains available on its ectosome, C. reniformis engulfs about 0.2mg cm ^ day” of quartz and produces 0.1 mg em” day” of dissolved silica. On the Portofino Promontory cliff, the average daily quartz avail- ability, evaluated with sediment traps, is 04mg em”. This suggests that quartz availability is not a limiting factor, allowing us to hypothesise that the sponge maintains the same ratio of incorp- oration and dissolution shown in laboratory experiments. Considering that the population density of C, reniformis along the Portofino Promontory 1s about 5,000cm” per meter of coast, and that the Promontory coast is about 13km long, it is poss- ible to estimate a production of dissolved biologically available silica of about 2106g yr”, Even if the most important contribution of silica to the Mediterranean Sea comes from thc Gibraltar Strait (De Master, 1981), input from rivers into the Mediterranean, although generally modest, may also be locally import- ant. In the Tigullio Gulf (Ligurian Sea), the Entella River, with an average annual flow rate of 14,8m? sec”, carries about 214x106g yr? of dissolved silica. However, the production by populations of Chondrosia at the Portofino Promontory, of about 2106g yr ', suggests that this species has a significant role in silicate turn-over, in rocky littoral areas, far removed from river input. ACKNOWLEDGEMENTS This work was financially supported by CNR 5% funds. LITERATURE CITED BAVESTRELLO, G., ARILLO, A., BENATTI, U., CERRANO, C., CATTANEO-VIETTI, R., CORTESOGNO, L., GAGGERO, L., GIOVINE, FIG. 5. SEM photographs, at three different mag- M., TONETTI, M. & SARA, M. 1995a, Quartz nifications (A-C), of spherical-like structures showing dissolution by the sponge Chondrosia reniformis quartz particles expelled by the ectosome and (Porifera, Demospongiae). Nature 378: 374-376. enveloped by a thin mucus web. Scale bars: A = | Imm; BAVESTRELLO, G., ARILLO, A., CALCINAL, B., B=%mm; C=2 mm. CERRANO, C., CATTANEO-VIETTI, R., 90 MEMOIRS OF THE QUEENSLAND MUSEUM LANZA, S., GAINO, E. & SARA, M. 1998a. Siliceous particles incorporation in Chondrosia reniformis (Porifera, Demospongiae). Italian Journal of Zoology 65: 343-348. BAVESTRELLO, G., BENATTI, U., CALCINAT, B., CATTANEO-VIETTI, R., CERRANO, C., FAVRE, A., GIOVINE, M., LANZA, S., PRONZATO, R. & SARA, M. 1998b. Body polarity and mineral selectivity in the demosponge Chondrosia reniformis. Biological Bulletin 195: 120-125, BAVESTRELLO, G., CATTANEO-VIETTI, R., DANOVARO, R. & FABIANO, M. 1991. Detritus rolling down a vertical cliff of the Ligurian Sea (Italy): the ecological role in hard bottom communities. PSZNI Marine Ecology 12: 281-292. BAVESTRELLO, G., CATTANEO-VIETTI, R., CERRANO, C., DANOVARO, R. & FABIANO, M. 1995b. Annual sedimentation rates and role of the resuspension processes along a vertical cliff (Ligurian Sea, Italy). Journal of Coastal Research 11: 690-696. BAVESTRELLO, G., CERRANO, C., CATTANEO- VIETTI, R., CORTESOGNO, L., CALABRIA, F. & SARA, M. 1996. Selective incorporation of foreign material in Chondrosia reniformis (Porifera: Demospongiae). Italian Journal of Zoology 63: 215-220. DE MASTER, D.J, 1981. The supply and accumulation of silica in the marine environment. Geochimica et Cosmochimica Acta 45(10): 1715-1732, ESPOSITO, A. & MANZELLA, G. 1982. Current circulation in the Ligurian Sea. Pp. 187-203. In Nihoul, J.C.J. (ed.) Hydrodynamics of semi- enclosed seas. (Elsevier: Amsterdam). GERRODETTE, T. & FLECHSIG, A.O. 1979. Sediment-induced reduction in the pumping rate of the tropical sponge Verongia lacunosa. Marine Biology 55: 103-110. HAECKEL, E. 1872. Die Kalkschwámme. Vols 1-3. (G. Reimer: Berlin). HARRISON, F.W., KAYE, N.W. & KAYE, G. 1985. The dermal membrane of Eunapius fragilis. Pp. 223-227. In Rützler, K. (ed.) New perspectives in sponge biology. (Smithsonian Institution Press: Washington DC). HISCOCK, K. 1987. Subtidal rock and shallow sedi- ments using diving. Pp. 198-237. In Baker, J.M. & Wolff, W.J. (eds) Biological surveys of estuaries and coasts. (Cambridge University Press: London). LENDENFELD, R. VON 1889. A monograph of the horny sponges. Royal Society of London. (Trubner and Co.: London). LOOSANOFF, V.L. & TOMMERS, F.D. 1948. Effect of suspended silt and other substances on rate of feeding of oysters. Science 107: 69-70, MARULLO, S., SALUSTI, E. & VIOLA, A. 1985. Observations of a small scale baroclinic eddy in the Ligurian Sea. Deep Sea Research 32(2): 215-222, MORRI, C., BIANCHI, C.N., DAMIANI, V., PEIRANO, A., ROMEO, G. & TUNESI, L. 1986. L'ambiente marino tra Punta della Chiappa e Sestri Levante (Mar Ligure): profilo ecotipologico e proposta di carta bionomica. Bollettino dei Musei e degli Istituti Biologici dell’ Università di Genova 52(supplement): 213-231. PRONZATO, R., BAVESTRELLO, G., & CERRANO, C. 1998. Morphofunctional adaptations of three species of Spongia (Porifera; Demospongiae), from a Mediterranean vertical cliff. Bulletin of Marine Science: 63: 317-328. REISWIG, H. 1974. Water transport, respiration and energetics of three tropical sponges. Journal of Experimental Marine Biology and Ecology 14: 231-249. ROGERS, C. S. 1990. Responses of coral reefs and reef organisms to sedimentation. Marine Ecology Progress Series 62: 185-202. SARA, M. & VACELET, J. 1973. Ecologie des démo- sponges. Pp. 462-576. In Brien, P., Lévi, C., Sarà, M., Tuzet, O. & Vacelet, J. (eds) Traité de Zoologie. Anatomie, systématique, biologie. Vol. 3(1). Spongiaires. (sér. ed. P.-P. Grassé). (Masson et Cie: Paris). SCHULZE, F.E. 1879. Untersuchungen über den Bau und die Entwicklung der Spongien. VI. Die Gattung Spongelia. Zeitschrift für Wissen- schaftliche Zoologie 32: 117-157. SHAW, M. E. 1927. Note on the inclusion of sand in sponges. Annals and Magazine of Natural History 19: 601-609. SOLLAS, I.B.J. 1908. The inclusion of foreign bodies by sponges, with a description of a new genus and species of Monaxonida. Annals and Magazine of Natural History 1; 395-401. TERAGAWA, K.C. 1986a. Particle transport and incor- poration during skeleton formation in a keratose sponge: Dysidea etheria. Biological Bulletin 170: 321-334. TERAGAWA, K.C. 1986b. Sponge dermal membrane morphology: histology of cell-mediated particle transport during skeletal growth. Journal of Morphology 190(3): 335-348. TORTONESE, E. 1961. Nuovo contributo alla conos- cenza del benthos della scogliera ligure. Archivio di Oceanografia e Limnologia 12: 163-183. VERDENAL, B. & VACELET, J. 1985. Sponge culture on vertical ropes in the Northwestern Medit- erranean Sea. Pp. 416-424. In Rützler, K. (ed.) New Perspectives in Sponge Biology. (Smith- sonian Institution Press: Washington, D.C.). WILBER, C.G. 1971. Turbidity: General aspects. Pp. 1156-1157. In Kinne, O. (ed.) Marine Ecology. Vol. 1(2). (Wiley & Sons: Chichester). WILLENZ, P. & VAN DE VYVER, G. 1982. Endoc- ytosis of latex beads by the exopinacoderm in the freshwater sponge Ephydatia fluviatilis: an in vitro and in situ study in SEM and TEM. Journal of Ultrastructure Research 79: 294-306. FOREIGN MATTER IN CHONDROSIA RENIFORMIS 91 CARBON ISOTOPE HISTORY OF CARIBBEAN SURFACE WATERS REVEALED BY CORALLINE SPONGES. Memoirs of the Queensland Museum 44: 91. 1999:- Live coralline sponges of the species Ceratoporella nicholsoni were collected from caves of north Jamaican reefs (20m depth) and from the deeper slope of Pedro Bank (125m depth). These sponges build a very dense aragonitic basal skeleton in apparent isotopic equilibrium with ambient water. Uranium-thorium dating of four specimens resulted in ages of 450-600 years. Within that timeframe, the sponge skeletons provide a continuous carbon isotope record, which starts at the end ofthe medieval warm period (1400AD) and covers the ‘Little Ice Age’ (about 1550-1850AD), as well as the industrial period (since ca. 1850AD). With a sample resolution of 0.7mm and growth rates of 0.2-0.4mm/ year the temporal resolution is about 2-4 years. The carbon isotope records show an excellent linear correlation with the atmospheric pCO, history, as recently reconstructed from Antarctic ice cores (Etheridge et al., 1996). We find no significant difference between the preindustrial and the industrial regression slopes (-0.013 permil/ppm) which agrees with a common mechanism for the observed surface water carbon isotope variations, i.e. addition/removal of isotopically ‘light’ organic carbon to/from the atmosphere-surface ocean-biosphere system. The ‘Little Ice Age’ is characterized by a slight increase of 38°C values (+0.1 permil), peaking around 1700AD During the same period, pCO, was about 6ppm lower than during the medieval warm period. Both can be explained by an increase in the terrestrial organic carbon reservoirs or in oceanic productivity. The Pedro Bank specimen, collected from the uppermost thermocline, shows only a dampened 8'°C increase during the Little Ice Age and a slightly subdued industrial 6'°C decline. This is expected because of the greater influence of deep-water at this depth. A comparison of the observed variation of marine 8°C values and 5'*C of atmospheric CO; included in Antarctic ice allows one to constrain the maximum global average cooling of the ocean surface layer during the Little Ice Age to ca. -0.7K (possible range 0 to -2K). Further comparison to the simultaneous pCO, decrease of 6ppm suggests an even smaller cooling. Alternatively, an enhanced oceanic export productivity could partly explain the observations. O Porifera, carbon isotope history, coralline sponges. Literature cited. ETHERIDGE, D.M., STEELE, L.P., LANGENFELDS, R.L., FRANCEY, R.J., BARNOLA, J.M. & MORGAN, V.I. 1996. Natural and anthropogenic changes in atmospheric CO, over the last 1000 years from air in Antarctic ice and firn. Journal of Geosphysical Research (D) 101: 4115-4128. Florian Bóhm, Wolf-Christian Dullo, Anton Eisenhauer, GEOMAR, Wischhofstr.1-3 D-24148 Kiel, Germany; Michael M. Joachimski, Institut fiir Geologie, Schlofigarten 5, D-91054 Erlangen Germany; Helmut Lehnert, Joachim Reitner (email: jreitne@gwdg.de), Institut und Museum für Geologie und Paläontologie, Universität Göttingen, Goldschmidt-Strasse 3, D-37077 Göttingen, Germany. TIME-LAPSE STUDIES OF SPONGE MOTILITY AND ANATOMICAL REARRANGE- MENTS. Memoirs of the Queensland Museum 44: 91. 1999:- Sponges have a general reputation as sessile and static animals, but this view has been contradicted by time-lapse microscope studies of live intact sponges belonging to several taxa (2 freshwater and 5 marine genera). These studies have demonstrated that adult sponges form leading margins made of crawling cells (pinacocytes and mesohyl cells), and these crawling margins appear capable of generating shape changes and locomotion of the entire sponge. These together with tracing studies have shown that sponges can move up to 1601. m/hr (4mm per day). Observed sponges also display continuous cell movements and anatomical rearrangements in their marginal regions. These rearrangements produce slow continuous changes in the spicule skeleton and in the canal systems. Both whole-sponge motility and the internal rearrangements appear to be strongly affected by factors such as substratum adhesiveness, grooves, internal tensile forces, and water flow patterns. These ongoing changes may be an important source for plasticity in a sponge's life history. O Porifera, anatomy, cells, crawling, locomotion, motility, anatomical rearrangement, time-lapse. Calhoun Bond (email: bondc@gborocollege.edu), Department of Biology, Greensboro College, 815 West Market St., Greensboro, NC 27401 USA; 1 June 1998, 92 MEMOIRS OF THE QUEENSLAND MUSEUM DO CARIBBEAN SPONGES HAVE PHYSICAL DEFENSES ? Memoirs of the Queensland Museum 44: 92. 1999:- Sponges are conspicuous members of the Caribbean marine ecosystem, but are preyed upon by a very select group of consumers called spongivores. Like other sessile reef invertebrates such as ascidians and octocorals, sponges possess a variety of novel secondary metabolites and as well as mineral and organic skeletal components. Several studies have shown that sponges possess chemical defenses that inhibit feeding by browsing generalist fish, but no study to date has demonstrated that sponge skeletal components deter predation. Sponges are soft-bodied and seem to lack an obvious physical defense, such as a mineralized shell. However, the tissues of most sponges often contain a collagen-like substance called spongin and sharp siliceous spicules in high concentrations. Spicules serve as important structural components by increasing tissue rigidity and could potentially act as a defense by irritating the mouth parts and the digestive system of predators. Calcified structures, similar in size to spicules, from octocorals and algae have been shown to reduce feeding by fish and invertebrates. Surprisingly, field and laboratory aquarium assays of sponge spicules employing predatory reef fish did not support a defensive function. Consumption by reef fish was reduced only when spicules were assayed using foods of low nutritional quality. In assessing the chem- ical defenses of Caribbean sponges, 31% of the species we studied possessed organic extracts palatable to reef fish. Interestingly, many of these undefended sponge species are abundant and consumed only by spongivores. Sponges lacking a chemical defense may be protected from generalist predators by having tissues of low nutritional value. Protein, carbohydrate, lipid, ash, and caloric content of 71 Caribbean sponge species were measured to investigate the relationship between chemical defense and nutritional value. Except for lipid content, no significant differences in nutritional quality were found between chemically defended and undefended species. Sponges lacking a chemical defense may rely on tactics other than à physical or *nutritional' defense, such as faster growth rates, to avoid predation by generalist consumers. O Porifera, chemical defenses, physical defenses, spicules, silica, nutritional quality, predatory-prey interactions, Caribbean reef ecosystems. Brian Chanas* (email: chanas(@niehs,nih.goy) & Joseph R. Pawlik, Biological Sciences, University of North Carolina at Wilmington, 601 S. College Road, Wilmington, North Carolina, USA, 28403-3297; *Present address: Laboratory of Pharmacology and Chemistry, National Institute of Environmental Health Sciences, MD C3-02, P.O. Box 12233, Research Triangle Park, North Carolina, 27709 USA; 1 June 1998. SPONGE DISTRIBUTION AND LAKE CHEMISTRY IN NORTHERN WISCONSIN LAKES: MINNA JEWELL’S SURVEY REVISITED ALISON C.C, COLBY, THOMAS M. FROST AND JANET M. FISCHER Colby, A.C.C., Frost, T.M. & Fischer, J.M. 1999 06 30: Sponge distribution and lake chemistry in northern Wisconsin lakes: Minna Jewell’s survey revisited. Memoirs of the Queensland Museum 44: 93-99, Brisbane. ISSN 0079-8835. Minna Jewell conducted an extensive survey of the regional distribution of freshwater sponges in Northern Wisconsin, USA, during the 1930”s, and examined factors that controlled the occurrence of sponges. We returned to 18 of her original 102 study lakes in 1996-97 to evaluate the long-term stability of the sponge distribution patterns that she reported. Comparisons of Jewell’s data and our recent survey reveal a decline in the distribution of Spongilla lacustris in N. Wisconsin lakes during the past 60 years. Jewell had originally reported S. /acustris present in 10 of the 18 lakes that we re-visited. As of 1996, we were unable to find S. lacustris in 5 of these 10 lakes. In addition, we observed only 1 invasion by S. lacustris in a lake that previously had not contained this species. To test how effectively four chemical variables reported by Jewell (pH, colour, conductivity and SiO») could predict the distribution of S. lacustris, we applied a discriminant model to the historical data set. Based on these four variables, we found that discriminant models poorly predicted sponge distribution patterns in Jewell’s original survey lakes and in 17 additional lakes surveyed in 1996. Our analyses indicate that S. lacustris can grow under a wide range of chemical conditions and suggest that other environmental variables are probably influencing sponge distribution in N Wisconsin lakes. CJ Porifera, ecology, freshwater sponges, fauna survey, Spongilla lacustris, Wisconsin. Alison C.C. Colby (email: accolbv@iname.com) € Thomas M. Frost, Trout Lake Station, Center for Limnology, University of Wisconsin-Madison, Madison, WI 53706, USA; Janet M. Fischer, Section of Ecology and Systematics, Corson Hall, Cornell University, Ithaca, NY 14853-2701, USA; 1 February 1999. Freshwater sponges are present in many aquatic ecosystems and may comprise a major component ofa lake’s benthic community (Frost, 1991). Currently, 27 species of freshwater sponges have been identified in North America (Jewell, 1959; Penney & Racek, 1968; Harrison, 1974; Frost, 1991; Ricciardi & Reiswig, 1993). Most of these species have been reported from the N. United States and S. Canada, and regional distribution patterns indicate that biogeographic conditions may restrict the distribution of some sponge species (Penney, 1960; Penney & Racek, 1968; Jones & Rützler, 1975; Frost, 1991). Ata more local scale, freshwater sponge distribution is influenced by environmental conditions within a particular lake or stream, however the relation- ship between the distribution of different sponge species and these environmental variables is not well understood. In a classic and unusually detailed study for the times, Minna Jewell (1935, 1939) investigated the distribution of freshwater sponges in 102 lakes in the Northern Highland Lake District of Wisconsin, as a contribution to the comparative limnological efforts of Birge, Juday, and their co-workers (Frey, 1963). Jewell identified 10 different sponge species and related their distribution to chemical variables in lakes. For each lake Jewell (1935, 1939) recorded dissolved oxygen, free- and bound- CO», pH, residue, SiO», conductivity, colour, and secchi depth. Results of her study indicated considerable variation in habitat requirements among sponge species and reported some level of correlation between abiotic environmental variables and species’ distrib- utions. We revisited Jewell’s efforts to examine the long-term stability of sponge distributions, and applied more modern analytical techniques to her original dataset. Of the ten sponge species reported by Jewell, Spongilla lacustris was by far the most prevalent. It occurred in 76 of the 102 lakes sampled, and was distributed throughout the entire range of abiotic conditions surveyed. Because of the widespread distribution of this species in Northern Wisconsin lakes (Jewell, 1935), we re-surveyed a subset of Jewell’s original study lakes to determine if S. lacustris occurred in the 94 TABLE 1. Distribution of Spongilla lacustris in 18 Northern Wisconsin lakes during Jewell’s (1935) and 1996 surveys. Water chemistry data is presented for 1996 survey. Key: 0, sponges absent; +, sponges present; * 1996 findings were different from the 1935 dataset; 2, Jewell referred to this lake as Muskelunge by Pickerel). MEMOIRS OF THE QUEENSLAND MUSEUM to expand the original dataset. Our survey techniques included shoreline and littoral zone sampling by snorkeling and boating. A small jonboat, rake, and Lake Survey | Survey | 4 pH colour (Pt, | Conductivity | DRSi net were used to complement 1935 | 1996 mgL”) | (umho em”) (gL) | specimens collected by snorkeling. Anne + 0 * | 6.14 | 2815 12 78 Water samples were collected in Bug 0 0 637 | 954 18 595 open-water regions of the lakes Crystal 0 0 6.57 | 6.04 12 14 for chemical analysis. Small Helmet 0 0 56 | 483.11 37 131 portions of sponges were brought p + 0 * | 60. | 3239 16 E back to the laboratory where they Little John Jr. 0 o op Ir ase 13 is were air-dried and stored until Little Pickerel | + + 6.85 | 15774 63 jx» | SPicule processing ^ and o identification following procedures Magie Rok a + gs | 539 H $5 | described in Frost (1991). For each Maig y Ep 7194] 456 Ls 2219 | sample, dried sponge tissue was Mary * * 6.26 | 314.05 26 961 placed in centrifuge tubes and Muskelunge? 0 0 755 | 114.11 80 7641 boiled in concentrated nitric acid Nebish + + 704 | 23.03 18 147 for one hour. The remaining Nixon " 0 * | 724 | 251.3 60 5973 spicules were rinsed in ethanol, Oswego + 0 + | 629 | 43.63 15 47 centrifuged, and slides were Street 0 0 600 | 2391 15 47 prepared for examination on a Tamarack A " 3*4 | ok 33 seo | compound light microscope. l U. Gresham 0 0 a32 | 383 254 5191 Water samples were collected in Wishow " ü "NT. 6 E polyethylene bottles and processed same habitats after 60 years, or if the distribution had shifted substantially. We focused our study on the relationship between S. /acustris and the lake chemical features, pH, colour, conductivity, and SiO», that Jewell (1935, 1939) suggested had the strongest apparent correlations with sponge distribution. Because Jewell's data were potentially limited by the analytical techniques available at that time, we applied modern statistical techniques to the original dataset to further test the degree to which a lake's chemistry could be related to sponge distribution. We applied discriminant analysis to her dataset, and used the resulting model to predict sponge distribution in a new set of 17 lakes surveyed during the summer of 1996. This provided a further test to determine how well chemical lake features are related to the occurrence of S. lacustris. MATERIALS AND METHODS Thirty five lakes in the Northern Highland Lake District of Wisconsin were surveyed for the presence of S. /acustris during the summer of 1996. Eighteen of these lakes were opportun- istically selected from those in Jewell's (1935) survey. In addition, 17 new lakes were surveyed in the laboratory. An Oakton WD-35607-10 conductivity meter and an Accumet 900 pH meter were used for analyses. A spectrophotometer was used to determine water colour following procedures described in Cuthbert (1992). Dissolved reactive silica (DRSi) concentrations were determined colourimetrically by a Technicon Segmented Flow Auto Analyzer. Discriminant analysis was used to examine the relationship between the distribution of S. lacustris and pH, SiO2, conductivity, and colour values reported by Jewell for the lakes she sampled in 1935. Our approach attempted to predict the presence or absence of S. lacustris ina study lake using an equation of the form: F= d1,Z 1 tdi;Z;t. x ddini 3 where di is the weighted discriminant coefficient, Z is the discriminating chemical variable, and Fis a categorical variable reflecting the presence or absence of a sponge (Digby & Kempton, 1994). The magnitude of the discriminant coefficient indicates the influence that the associated variable has on the distribution of 5. lacustris, We applied discriminant analysis to the entire data set on the presence or absence of S. lacustris reported by Jewell for 102 lakes. We tested the efficacy of the discriminant analysis by a cross- LAKE SPONGE DISTRIBUTION 95 TABLE 2. Performance of discriminant model fit to 99 lakes from Jewell’s (1935) survey. * = Jewell reported water chemistry data for 99 of the 102 lakes that she surveyed. Jewall's Discriminant Spongilla lacustris Analysis (1935) Results | predicted Results No. of lakes with sponges — | 73 50 No. of lakes without sponges 26 49 Total no. of lakes surveyed* 99 99 validation of the predicted results compared to the actual observed results reported by Jewell in all the lakes that she surveyed. In addition, the resulting model was applied to the 17 new lakes that we surveyed during the summer of 1996 to test whether this model predicted current distrib- ution patterns accurately. RESULTS The distribution of S. lacustris was found to be the same as reported by Jewell in 12 of the 18 lakes re-surveyed. We detected S. lacustris present in one lake in which it had not been previously recorded (Table 1). Conversely, we did not find S. lacustris in 5 of the 10 lakes in which Jewell had reported its presence. However, we found no dramatic changes in lake chemistry to account for the disappearance of S. lacustris from these lakes. Graphical analyses of the lake chemistry and sponge distribution reported by Jewell (1935), and the 35 Jakes we surveyed in 1996, showed no obvious patterns between the pH, colour, conduc- tivity, and DRSi values in relation to the presence or absence of S. lacustris (Fig. 1A-H). A comparison between our survey and that of Jewell (1935) revealed a general decline in the distribution of S. /acustris during the last 60 years (Fig. 1). Our discriminant analysis of Jewell’s dataset did not reveal any significant relationships between the pH, colour, conductivity, or SiO», and the presence or absence of S. /acustris, as reported by Jewell (Table 2). The discriminant analysis of Jewell's original data assigned discriminant coefficients to each chemical variable of 1.16 for pH, 0.46 for conductivity, 0.35 for colour, and 0.23 for DRSi. We cross-validated with Jewell's actual dataset to test the ability of these 4 coefficients to correctly predict the presence or absence of S. lacustris. We found no significant relationship between actual sponge distribution and the predicted distribution. Jewell had reported S. lacustris to be present in 73 of her study lakes and absent in 26. Using the original chemical values that Jewell reported as predictors, the cross-validation of her dataset predicted sponges to be present in 50 of the surveyed lakes and absent in 49, with an error rate of 49% (Table 2). Our more recent survey also indicated that chemical variables are ineffective predictors of the occurrence of S. /acustris. We found S. lacustris in just over half (9 of 17) ofthe lakes that we included in our new survey (Table 3). Using the discriminant model derived from Jewell's data, and the chemical data from the new survey lakes, we had predicted that 13 of the 17 new survey lakes would contain S. /acustris, with an error rate of 49% (Table 4). Furthermore, the absence of any significant relationship between the occurrence of S. /acustris and the chemical gradients that we evaluated, as illustrated by the lack of any indication of correlation in Jewell's dataset or our recent survey, strongly indicates that some factors besides the chemical variables may dictate the presence or absence of S. lacustris. DISCUSSION The notion that tolerance to a wide range of abiotic factors is a major feature of the niche of some species, is a well-recognized phenomenon (Dunson & Travis, 1991). Our research emphasises the ability of S. /acustris to tolerate a wide range of chemical conditions, setting this species apart from several other groups of aquatic organisms. Abiotic factors have been shown to limit the distribution of several fish and zooplankton species, and to directly influence aquatic macrophyte community structure (Brown & Jewell, 1926; Rahel & Magnuson, 1983; Tilman, 1988; Webster et al., 1992; Arnott & Vanni, 1993). Our results do not indicate any significant relationship between the distribution of S. lacustris and lake chemistry. Recent surveys conducted in Norway and Connecticut also note the ability of S. lacustris to tolerate a wide range of abiotic conditions (Okland & Okland, 1996; De Santo & Fell, 1996). This tolerance may be a very important adaptation for the survival of this species in freshwater habitats and may account for its reported cosmopolitan distribution. We most frequently found S. lacustris in small, sheltered regions of lakes, growing directly up from bottom sediments. In lakes with less suitable bottom substrate, smaller specimens were found 96 MEMOIRS OF THE QUEENSLAND MUSEUM N N u u M M B B E 1 E R 8375 R 100% B 0-03 31-10 14-40 4-20 74- 100+ o 0-03 .31-10 11-40 44-70 7.1-10.0 10.0+ o 10.0 F DRSi (mg/l) A F DRSi (moll) B 50 12 9 36% 40 65% d L d L 50% . A 30 : 83% A 17% 8% ga K K E 20 86% 82% E E s so, a D] 0-25 26-50 51-100 101-200 201- 0-25 26-50 51-100 101-200 201+ Ü COLOR (Pt, mg/l) C ] COLOR (Pt, mg/l) D Rap R E 60% E V o V g 94% E Y gow, * Y E 20 E D D 10 0 1 0-15 16-30 31-60 61-90 91-120 1214 1 0-15 16-30 31-60 61-90 91-120 1214 : CONDUCTIVITY (umho/cm) E 4 CONDUCTIVITY (umho/em) E 5 6 48- 60- 65- 70- 75- 80- 50-59 60-64 65-69 70-74 7.5-7.9 8.0-8.9 Za p du S4 TH e pH G pH H O# lakes @# with S. lacustris FIG. 1. Distribution of S. lacustris across chemical gradients for Jewell's (1935) survey (A,C,E,G) and 1996 survey (B,D,F,H). Lighter bars indicate the number of lakes surveyed; darker bars represent the number of lakes containing S. lacustris. Note the overall decline in percentage of lakes with S. lacustris present. A-B, Dissolved reactive silica concentrations; C-D, colour; E-F, Conductivity; G-H, pH. LAKE SPONGE DISTRIBUTION TABLE 3. Distribution of Spongilla lacustris and water chemistry in 17 Northern Wisconsin lakes surveyed in 1996. These lakes were not included in Jewell’s (1935) survey. Key: 97 typical and atypical, correlated to the morphology of spicules. Jewell defined atypical specimens as those that had 0, sponges absent; +, sponges present. aberrant forms of spined microscleres Survey colour Conductivity DRSi from l ake E woth. low silica Lake 1996 | PH | (p. mgL")| (amhocm")| (ug Lo) concentrations. We also found several S. mem Liesel "(one o3 sm | Jacustris specimens from lakes with low "Wa a | suoll ^ agat " ne | Silica concentrations to have finer, less CRUENTO a PEN. xS ud m robust spicules and smaller microscleres EFE than those specimens from lakes with Syst bog Qm] dos AN š 158 higher silica concentrations. For our Firefly 0 | 648 | 1724 17 2 analyses of Jewell's data we combined Fishtrap 0 77 40.27 93 4617 both the typical and atypical forms into Frank + | 705 | 34.11 21 100 one species classification. Goodyear spg | + | 722 18.63 73 6390 Additional observations made during Mystery + | 5.92 | 111,51 18 1348 our field survey provided some insight Oberlin + |668| 2748 15 153 into other environmental factors that Nixon creek 0 715 | 238.28 61 5712 may be influencing the distribution of S. Partridge Y 6m 66 7015 lacustris in Northern Wisconsin lakes. Rainbow fiwg | + y 128:79 20 67 We observed a slight decline in the "—' E ia "es 106 4306 presence of S. lacustris compared to its sandy beac [ek asses A 25 distribution in 1935. Many of the 12 : = = lakes that we found no change in sponge Tower CA e E a 36 gens distribution patterns are located in the Trout bog 0 | 476 | 20267 17 $0 Wisconsin State forest and have been encrusted on the underside of logs, and on the woody roots of cranberry bushes (Vaccinium spp.). Tiny specimens were found growing in very low silica and conductivity habitats, most often on the tips of aquatic macrophytes, usually Myriophyllum and Isoetes species. Spongilla lacustris appeared well-adapted to a wide range of light conditions, and depending upon the colour of the water, was found anywhere from just below the surface in Little Pickerel Lake to depths of 3m in Little Rock Lake (Frost & Elias, 1990). The trace amounts of DRSi found in some northern Wisconsin lakes do not appear to limit the occurrence of S. lacustris (Table 1). Observ- ations of freshwater sponge morphology suggest, however, that DRSi plays an important role in the growth and skeletal strength of a sponge (Jewell, 1935; Kratz et al., 1991). Limited silica availability may result in decreased strength of the spicule skeleton, causing indirect negative effects on the distribution of sponges, perhaps by providing less protection against predation (Frost, Kratz & Elias, personal communication). Jewell (1935) recognised that DRSi was an important factor in determining the degree of skeletal development in S. lacustris, and conseq- uently differentiated two different growth forms, protected from development for the past 60 years. Four of the five lakes that are now unoccupied by S. /acustris however (Anne, Joyce, Oswego and Wishow Lakes), have portions oftheir shorelines developed with privately owned cabins. Alteration of littoral habitats by develop- ment (e.g. removal of coarse woody debris; Christensen et al., 1996) may be negatively impact- ing the distribution of S. lacustris in these lakes. Both our contemporary survey and that of Jewell (1935) focused on the occurrence of S. lacustris, but not on it's biomass and prevalence, which varies substantially among habitats. It can be quite abundant in some situations (Frost et al., 1982; Frost & Elias, 1990), and nearly absent in others (Colby & Frost, personal observations). Also, while the overall distribution of S. /acustris may appear stable in some lakes, undocumented observations of significant yearly fluctuations TABLE 4. Predictions of Spongilla lacustris distribution in 17 Northern Wisconsin lakes surveyed for the first time in 1996. The discriminant model used to make these predictions was parameterised using Jewell's (1935) dataset. Spongilla lacustris pie DEM Doe No. of lakes with sponges 9 13 No. of lakes without sponges 8 4 Total no. of lakes surveyed 17 17 98 MEMOIRS OF THE QUEENSLAND MUSEUM have been observed previously (Frost, personal observation), but could not be quantified in either survey. The fact that we have not clearly linked Species occurrence patterns with lake chemistry strongly suggests that other physical or biological factors are influencing sponge distribution. These factors could include associated vegetation, available substrate, predation, dispersion, climatic conditions and disease. Apart from some interesting results reported by Jewell (1935) there is generally little information available on interactions between S. lacustris and its surrounding communities, including inter- actions with other freshwater sponge species. Competition and mutualism between different species of marine sponges has been relatively well documented (e.g. Riitzler, 1970; Sara, 1970; Wulff, 1997), and it is possible that these inter- actions occur in freshwater as well. Jewell reported nine other species that are not as common as S. lacustris in these lakes, and that we did not include in our survey. These other freshwater species may be more strongly influenced by lake chemical factors, and could also be influencing the distribution of S. /acustris. Recognition of freshwater sponges as active members of aquatic communities could lead to a better understanding of the relationships between freshwater sponges, environmental factors important to their survival, and their associated surrounding communities, ACKNOWLEDGEMENTS Financial support was provided by the National Science Foundation (deb-9527669 ) and the Chase-Noland Undergraduate Award in Zoology at the University of Wisconsin - Madison. We would like to thank the staff and summer crew at the Trout Lake Research Station, Center for Limnology, University of Wisconsin, especially S. Knight for her help collecting specimens in the field, and M. Casper for her continuous support. LITERATURE CITED ARNOTT, S.E. & VANNI, M.J. 1993. Zooplankton assemblages in fishless bog lakes: influence of biotic and abiotic factors. Ecology 74(8): 2361-2380. BROWN, H.W. & JEWELL, M.E. 1926. Further studies on the fishes of an acid lake. Transactions ofthe American Microscopical Society 45: 20-34. CHRISTENSEN, D.L., HERWIG, B.R., SCHINDLER, D.E. & CARPENTER, S.R. 1996. Impacts of lakeshore residential development on coarse woody debris in north temperate lakes. Ecological Applications 6(4): 1143-1149. CUTHBERT, 1.D. & DEL GIORGIO, P. 1992. Toward a standard method of measuring colour in freshwater. Limnology and Oceanography 37(6): 1319-1326. DE SANTO, E.M. & FELL, P.E. 1996. Distribution and ecology of freshwater sponges in Connecticut. Hydrobiologia 341: 81-89. DIGBY, P.G.N. & KEMPTON, R.A. 1994, Multivariate Analysis of Ecological Communities. (Chapman & Hall: London). DUNSON, W.A. & TRAVIS, J. 1991. The role of abiotic factors in community organization. American Naturalist 138(5): 1067-1091. FREY, D.G. 1963. Wisconsin: The Birge-Juday era. Pp. 3-54. In Frey, D.G. (ed.) Limnology in North America. (University of Wisconsin Press: Madison, WI). FROST, T.M. 1991. Porifera. Pp. 95-124. In Thorp, J.H. & Covich, A.P. (eds) Ecology and classification of North American freshwater invertebrates. (Academic Press: New York). FROST, T.M., DE NAGY, G.S. & GILBERT, J.J. 1982. Population dynamics and standing biomass of the freshwater sponge, Spongilla lacustris. Ecology 63: 1203-1210. FROST, T.M. & ELIAS, J.E. 1990, The balance of autotrophy and heterotropy in three freshwater sponges with algal symbionts. Pp. 478-484. In Riitzler, K. (ed.) New Perspectives in Sponge Biology. (Smithsonian Institution Press: Washington DC). HARRISON, F.W. 1974. Sponges (Porifera: Spongillidae). Pp. 29-66. In Hart, C.W. & Fuller, S.L.H. (eds) Pollution ecology of freshwater invertebrates. (Academic Press: New York). JEWELL, M.E. 1935. An ecological study of the fresh- water sponges of northern Wisconsin. Ecological Monographs 5: 461-504. 1939. An ecological study of the freshwater sponges of Wisconsin. I]. The influence of calcium. Ecology 20: 11-28, 1959, Porifera. Pp. 298-312. In Edmondson, W.T. (ed.) Freshwater Biology, 2nd edition. (John Wiley & Sons: New York). JONES, M.L. & RUTZLER, K. 1975. Invertebrates of the Upper Chamber, Gatún Locks, Panama Canal, with emphasis on Trochospongilla leidii (Porifera). Marine Biology 33: 57-66. KRATZ, T.K., FROST, T.M. & ELIAS, J.E. 1991. Reconstruction of a regional 12,000-yr silica decline in lakes by means of fossil sponge spic- ules. Limnology and Oceanography 36(6): 1244-1249, OKLAND, K.A. & OKLAND, J. 1996. Freshwater sponges (Porifera: Spongillidae) of Norway: dis- tribution and ecology. Hydrobiologia 330: 1-30. PENNEY, J.T. 1960. Distribution and bibliography (1892-1957) of the fresh-water sponges. Series 3. (University of South Carolina Publications). PENNEY, J.T. & RACEK, A.A. 1968, Comprehensive revision of a worldwide collection of freshwater LAKE SPONGE DISTRIBUTION 99 sponges (Porifera: Spongillidae). United States National Museum Bulletin 272: 1-184. RAHEL, F.J. & MAGNUSON, J.J. 1983. Low pH and the absence of fish species in naturally acidic Wisconsin lakes: inferences for cultural acidification. Canadian Journal of Fisheries and Aquatic Sciences 40: 3-9. RICCIARDI, A. & REISWIG, H.M. 1993. Freshwater sponges (Porifera: Spongillidae) of eastern Canada: taxonomy, distribution, and ecology. . Canadian Journal of Zoology 71: 665-682. RUTZLER, K. 1970. Spatial competition among Por- ifera: solution by epizoism. Oecologia 5: 85-95. SARA, M. 1970. Competition and cooperation in sponge populations. Pp. 273-284. In Fry, W.G. (ed.) The Biology of the Porifera. Symposium of the Zoological Society of London Number 25, (Academic Press: London). TILMAN, D. 1988. Plant strategies and the dynamics and structure of plant communities. (Princeton University Press: Princeton, N.J.). WEBSTER, K.E., FROST, T.M., WATRAS, C.J., SWENSON, W.A., GONZALEZ, M. & GARRISON, P.J. 1992. Complex biological responses to the experimental acidification of Little Rock lake, Wisconsin, USA. Environmental Pollution 78: 73-78. WULFE, J.L. 1997. Mutualisms among species of coral reef sponges. Ecology 78(1): 146-159. AN OVERVIEW OF STROMATOPOROID DOMINATED MIDDLE DEVONIAN REEF COMPLEXES IN NORTH QUEENSLAND. Memoirs of the Queensland Museum 44; 99, 1999:- Middle Devonian stromatoporoid buildups are known from the Burdekin Subprovince and the Broken River Province in the Townsville hinterland, north Queensland. Recent studies have placed these buildups within a reliable stratigraphic and sedimentologic framework. Buildups within the Burdekin Subprovince developed in a restricted near to proximal shore setting in a partially enclosed basinal setting. Those buildups within the Broken River province developed upon a more open marine shelf. Major Burdekin stromatoporoid-coral buildups were of two types: low relief extensive biostromes and associated stromatoporoid pavements, and a biohermal system of one to two metres relief from the sea floor. Additional buildups of note are small patch reefs developed within nearshore siliciclastic muddy lagoons adjacent to granitic headlands. In a number of such metre scale buildups within dominantly siliciclastic settings, assemblages of stromatoporoids and corals show repetitive growth interruption surfaces suggesting episodic stress and killing events. Storm disturbance during development the biostromal pavements was high and an important sedimentologic factor for the ‘reef? growth. Minor sponge s.s. buildups are known from the uppermost Burdekin Formation, but have not been studied. In the Broken River Province, Givetian buildups are more extensive and can be traced on the hundreds of metre scale, these have received little detailed sedimentologic study, but are of similar style to biostromal pavements from the neighbouring Burdekin Basin. Minor biohermal occurrences are found within the Papilio Mudstone, and formed on a muddy shelf, and include both stromatoporoid and sponge s.s. buildups. Stromatoporoid taxonomy has revealed the presence of eight stromatoporoid communities in the Burdekin Basin, comprising 35 taxa. Dominant stromatoporoids were dendroids Amphipora, Stachyodes and Trupetostroma, frame building. Trupetostoma, Pseudotrupetostroma, Hermatostroma, Actinostroma and Ferestromatopora. Coenostroma, Clathrocoilona, and Stromatopora were accessory to reef growth. In the Broken River detailed taxonomic work has only been partially completed. Significant overlap exists at generic level with the two adjacent provinces, but species level differences are strong suggesting distinct partitioning of open marine versus embayment faunas. This phenomenon is reflected in other faunal elements (gastropods, rugose corals). 0 Porifera, stromatoporoid, biostromes. Alex G. Cook (email: AlexC(aqm.qld. gov.au), Geology and Invertebrate Palaeontology, Queensland Museum, PO Box 3300 South Brisbane, Old 4101 Australia; 1 June 1998. 100 MEMOIRS OF THE QUEENSLAND MUSEUM GOOD CONGRUENCE BETWEEN MORPH- OLOGY AND MOLECULAR PHYLOGENY OF HADROMERIDA, OR HOW TO BOTHER SPONGE TAXONOMISTS. Memoirs of the Queensland Museum 44: 100. 1999:- Within Demospongiae, the order Hadromerida is well defined and there is a strong consensus among systematicians about its composition and validity. This order is charac- terised by the presence of tylostyles radially arranged at least in the periphery, and by microscleres, when present, of the aster type. All Hadromerida are ovip- arous and the choanocytes have a periflagellar sleeve. Ten families are without any doubt attributed to Hadro- merida, six of which with microscleres of the aster type and four of which without microscleres. The first work on molecular phylogeny of Porifera was made on the Hadromerida (Kelly-Borges, Bergquist & Bergquist, 1991). The molecule used was the 18S rRNA, which appeared to be not sufficiently informative to resolve the phylogeny at that taxonomic level. In this work we have used the 5’ end of the 28S rRNA (about 1000bp) to explore the internal phylo- geny of this order. 15 species belonging to 12 genera and 8 families were sequenced. Five outgroup species were sequenced belonging to Axinellida, Tetractinellida, and Halichondrida. Parsimony and Neighbor-Joining analyses have been done. Trees were rooted by using Tetractinellida (Cinachyrella and Discodermia) as a monophyletic outgroup. Both analyses (Parsimony and Neighbor-Joining) show that the Hadromerida are composed of four monophyletic taxa, Taxon | is comp- osed of 6 species belonging to the Spirastrellidae, Acanthochaetetidae, Clionidae, and Placospongiidae. All these families have microscleres of the spiraster- type. Taxon 2 is composed by 5 species of Timeidae and Tethyidae. These two families have microscleres of the euaster-type. Taxon 3 is composed of only one species Polymastia mamillaris belonging to the family Polymastiidae, which has no microsclere of aster type. The validity of this taxon has to be checked with other genera belonging to the Polymastiidae family. Taxon 4 is composed of three Suberitidae and an external species Halichondria panicea, which belongs to the family Halichondriidae (order Halichondrida). Neither the Suberitidae nor the Halichondriidae have microscleres of the aster type. The monophyly of each of these four taxa is well supported with high bootstrap proportions. The monophyly of the four taxa together is also well supported but the relationships between them cannot be ascertained. The monophylies of taxa | and 2 are congruent with morphology, both taxa corresponding to the hadro- merid families with spirasters and with euasters, respectively. An important and unexpected problem of classification appeared with taxon 4. The result obtained with our sequence of Halichondria panicea was confirmed with a shorter sequence of Hymeniacidon heliophila available in GenBank. When the sequence of Hymeniacidon is included, taxon 4 remains monophyletic and strongly supported by BP. From the morphological and cytological point of view there is no synapomorphy between the two groups. The Halichondrida are defined mostly by negative characters. However, we observed a fine morpho- molecular synapomorphy for taxon 4. This is the loss of a small loop of 15 bp in the secondary structure of the D2 domain, which is probably the result of only one deletion event. From the chemical point of view, there is another synapomorphy: a large amount of stanols have been described both in the Suberitidae and the Halichondrida. The best hypothesis seems to reallocate Halichondriidae to the Hadromerida. The order Hadromerida remains monophyletic. With the exception of this reallocation the classification obtained with 28S rRNA is perfectly congruent with the existing classification. All the families are monophyletic. We propose a subordinal classification : Spirastrellina, Timeina, Polymastiina and Suberitina. O Porifera, Demospongiae, molecular phylogeny, 28S rRNA, Hadromerida, Halichondrida, monophyly. Catherine Chombard (email: gdretudi(a)mnhn.fr), Service de Systematique Moleculaire (CNRS GDR 1005), Muséum National d'Histoire Naturelle, 43 rue Cuvier, 75005 Paris, France; Nicole Boury-Esnault, Centre d'Océnologie de Marseille, Station Marine d'Endoume, Université de’ Aix-Marseille 2 URA-CNRS, 41Rue de la Batterie- des-Lions, F-13007 Marseille, France; 1 June 1998. REMARKS ON THE STATUS OF MYXILLA (PORIFERA: POECILOSCLERIDA) ON THE GALICIAN COAST (NW IBERIAN PENINSULA) F.J.CRISTOBO, P. RÍOS AND V. URGORRI Cristobo, F.J., Rios, P. & Urgorri, V. 1999 06 30: Remarks on the status of Myxilla (Porifera: Poecilosclerida) on the Galician coast (NW Iberian Peninsula). Memoirs of the Queensland Museum 44: 101-123. Brisbane. ISSN 0079-8835. Myxilla Schmidt is represented on the Iberian Peninsula by six species, five of which, studied in this paper, were collected from the coast of Galicia (NW of Spain): M. incrustans, M. iotrochotina, M. macrosigma, M. rosacea and M. fimbriata, and the sixth (M. tarifensis), recently described from the Strait of Gibraltar. 188 specimens were collected from 72 stations along the coast of Galicia between 1979-1991. Illustrated descriptions of these species, their habitus, skeletal arrangement and spicules are provided, together with information on their autecology, distribution, and biometric studies of spicules. Morphological comparisons are made between these species and other Myxilla from the Atlantic region, and a taxonomic key to species of Myxilla in the NE Atlantic is provided. O Porifera, Poecilosclerida, Myxilla, Iberian Peninsula, NE Atlantic, taxonomy, ecology, key. F.J. Cristobo (email: bafjcris@usc.es), P. Rios & V. Urgorri, Laboratorio de Zooloxia Mariña, Departamento de Bioloxia Animal and Departamento de Biodiversidade e Recursos Mariños, Instituto de Acuicultura, Universidade de Santiago de Compostela, 15706 Santiago de Compostela, Spain; 1 March 1999. Only few studies have been made on Galician sponges (Solórzano & Rodriguez, 1979; Solórzano & Durán, 1982; Solórzano, 1991; Solorzano & Urgorri, 1991, 1993; Solorzano et al., 1991). Other records of sponges from the sublittoral benthos are also available in more general publications (Benito, 1976; Gili et al., 1979; Polo et al., 1979; Durán & Solórzano, 1982; Acuña et al., 1984), as well as from nudibranch - sponge dietary studies (Urgorri & Besteiro, 1984). Studies on Myxilla in the Ria de Ferrol (Cristobo, 1997) and Galician coast (this study) recorded five species: M. incrustans, M. iotrochotina, M. macrosigma, M. rosacea, and M. fimbriata. These are comprehensively described and discussed in this present study. MATERIALS AND METHODS Collections were made between 1979-1991 using direct sampling in the intertidal, and SCUBA and naturalist benthic dredge (Holme & McIntyre, 1984) in the sublittoral zones. A total of 188 specimens of Myxilla were collected from 72 stations on the Galician coast (Fig. 1). Preparation and histological methods follow Rubió (1973), Riitzler (1978), Uriz (1978, 1986) and Cristobo et al. (1993). Spicules were examined under a Hitachi S570 scanning electron microscope (SEM). Underwater photographs were taken with a Nikonos V camera and SB-102 flash. A biometric study of sponge spicules was made for specimens from the Ría de Ferrol and microscopic preparations of two paratypes of M. macrosigma (Museum National d’Histoire Naturelle, Paris (MNHN), Laboratoire de Biologie des Invertébrés Marins et Malacologie: DNBE282 from the Grotte des Calanques, and DNBE287 from Ile Grosse). All specimens were deposited in the Departamento de Bioloxía Animal in the Facultade de Bioloxia at the Universidade de Santiago de Compostela, Spain. SYSTEMATICS Order Poecilosclerida Topsent Family Myxillidae Topsent Myxilla Schmidt, 1862 Myxilla incrustans (Johnston, 1842) (Figs 2-4, 17C) MATERIAL. Stations 19, 20, 23, 43 (see Fig. 1). AUTECOLOGY. In Galicia, this sublittoral species lives in a small bathymetric zone from 8-14m depth in the outer Ria area, settling on granite rock on exposed bottoms; also found on gravel bottoms (Topsent, 1913) and as epibiont on Inachus and Cellaria (Crawshay, 1912); elsewhere it may also be found intertidally (Stephens, 1921; Kónnecker, 1973; Hoshino, 1981), in the sublittoral zone (Descatoire, 1969) 102 MEMOIRS OF THE QUEENSLAND MUSEUM 3-29 4 30 " NC 1 / H M 36 35 37 38. 39 40 41 4 ARA 43— pa 42 44 49 54 48 ^ 45 a =N> A | 5 E? Z "d 50 46 08*00'00"W 51-58 A " al p ] Em Illas Ons meshes of acanthostyles forming ascending tracts of up to 20 spicules interconnected by transverse fascicles. The ectosome is made up of tornotes in paratangential brushes which extend out in a P bouquet-like fashion. Micro- scleres are scattered throughout the sponge but anchorate chelae are more abundant in the ectosome, where they form a sub-super- ficial layer. Sigmas are dispersed within the choano- some. Megascleres: straight or slightly curved robust acantho- styles with conical spines. Dimensions: 150.3-209.0x 2.9-12.8um. Smooth, straight or 43 700'00" N 10 Km H Galicia slightly curved tornotes, with 12 27 9 20 14 19 18 22 7 13 28 2 Li tii tg p 10 24 29 8 11 73 25 25 16 15 35 4 17 5 6 asymmetrical terminations: one having a marked ellipsoidal tyle and the other with diverse irregular terminations, the most common of which is spear- shaped, in some cases bearing fine spines. Dimensions: 128.5-207.6x2.6-7.3um. Microscleres: Sigmas with the typical c- and s- shapes. Dimensions: 22.2-39.4x0.7- 1 Km Ria de Ferrol FIG. 1. Map of the study area showing location of collecting stations. and on rocky circalittoral bottoms (Vidal, 1967; Borojevic etal., 1968; Topsent, 1913) up to 170m deep (Boury-Esnault et al., 1994). DISTRIBUTION. Arctic, European Atlantic coasts, Gibraltar and Mediterranean Sea (Ackers et al., 1992); also allegedly reported from Senegal (Lévi, 1952), Japan (Hoshino, 1981), Korea (Sim, 1994) and Antarctica (Arndt, 1935), although the conspecificity of these records must be checked. In Galicia this species is known from the Ría de Ferrol, only the second record for the Iberian Peninsula, previously known from Punta Uhía, Ría de Muros (Solorzano, 1991). DESCRIPTION. An encrusting sponge, some- times massive, with a rough surface consisting of fine reticulation of spicules. Orange or yellow in colour. Skeletal arrangement: Choanosomal skeleton myxilloid with triangular and quadrangular 3.2um. Arched spatuliferous anchorate isochelae, of two different size categories: 11.3-19.2*3.5-6.3 1m and 24.1- -35.5 x10.9-16.2um. Myxilla iotrochotina (Topsent, 1892) (Figs 5-7, 17D) MATERIAL. Stations 1, 2, 7, 9, 31, 34, 36, 37, 40, 43, 45, 52, 56, 58, 66 (see Fig. 1). AUTECOLOGY. Cryptic species, occupying highly localised and well-concealed enclaves, perhaps explaining why it has been overlooked since it was first described by Topsent; in Galicia it is found in secluded places such as on the roofs of small caves and intertidal crevices in the mid-outer zone of the rias; the few references to this species describe it living in similar environments to a depth of up to 30m, such as detritic bottoms (Sara & Siribelli, 1960), and artificial breakwaters (Sara, 1961); also epibiont on other sponges such as Geodia (Ferrer- MYXILLA FROM GALICIA 103 A B AR Se A 100 um FIG. 2. Myxilla incrustans. Spicules: A, Acanthostyle; B, Tornote; C, Sigmas; D, Isochelae; E, Skeletal arrangement; F, Distribution in Galicia. MEMOIRS OF THE QUEENSLAND MUSEUM 100 um 10 um FIG. 3. Myxilla incrustans. Spicules: A, Acanthostyle; B, Tornote; C, Detail of the end of a tornote; D, Sigmas; E, Isochela. MYXILLA FROM GALICIA 140 13D 160 170 120 190 20 iD ND O 1 eR oO c D €» ex) CO FIG. 8. Myxilla macrosigma. Spicules: A, Acanthostyles; B, Tornotes; C, Sigmas; D, Isochelae; E, Skeletal arrangement; F, Distribution in Galicia. 110 MEMOIRS OF THE QUEENSLAND MUSEUM 5 um 10 um = = = = FIG. 9. Myxilla macrosigma. Spicules: A, Acanthostyle; B, Detail ofthe head of acanthostyle; C, Detail of the end of a tornote; D, Tornote; E, Sigma; F, Isochela. sigmas and isochelae, are found throughout the sponge. Spicules. Megascleres: slightly curved acanthostyles, with curvature occasionally more pronounced near the head of the spicule. Spines meshes composed of 1-5 acanthostyles. The ectosomal skeleton is made up of tornotes which are tangentially arranged to the sponge surface, sometimes forming bouquets. Microscleres, MYXILLA FROM GALICIA m N-$0 110 125 14D 143 130 133 160 163 17D 173 130 N=80 FERRO OS AMY UL X4,1-(5,9)-7.5um. C, Isochelae 18.0-(29.3)- 58.3*4.0- (8.2)-18. lium. (Topsent, 1928), on the biocenosis of Corallium rubrum between 100-200m depth (Templado et al., 1986), on bottoms of dead Madreporaria with brachiopods, tubicolous polychaetes and anthozoans at between 260-269m depth (Uriz, 1985), on Antipathes fragilis between 130-180m depth, as well as on epibathyal mud (Vacelet, 1960). DISTRIBUTION. Eastern Atlantic from the Arctic to South Africa; Pacific (Lambe, 1892); Mediterranean, (Carballo & Garcia Gómez, 1996). In Galicia it is found in à number of locations: 43°44°50"N, 08?12"W - 43°40°N, 08°55’W (Topsent, 1892), Os Feitales (Benito, 1976), Aguiño, O Grove (Rodriguez & Lorenzo, 1978), San Ciprián de Burela (Gili et al., 1979; Polo et al., 1979), Suevos, Caión, Patos (Solórzano & Rodriguez, 1979), Punta Uhía, Queixal, Insuela, Corvasa, Corasa, Centolleira, Isla de Rüa, Airós, Sálvora, Isla de Ons (Durán & Solórzano, 1982), Laxe (Solórzano & Durán, 1982), Islas Cíes (Acuna et al., 1984) and Ría de Arousa (Solórzano et al., 1991). DESCRIPTION. Morphologically variable, appearing as a massive, prominent covering on rocky substrates with osculiferous digitiform chimneys, and heart-shaped covering small-sized seaweeds. Dimensions: 2-20cm maximum diameter, 0.4- 10cm thick. Rough surface, having several characteristic crests, in some places very occasionally smooth. Soft and slightly flexible consistency; delicate ectosome and choanosome with a spongy appearance, highly perforated. The oscula may not be apparent in smaller specimens, but they are generally abundant, located in conical elevations protruding from the sponge mass from l-8cm, producing chimneys, com- monly having ascending, superficial aquiferous ducts; the osculum is circular in shape, some- times clover-shaped. Abundant ostia appear between the numerous ridges on the surface. colouration varies from various shades of orange, beige or light pink. The species frequently secretes 118 TABLE 1. Comparison between spicule dimensions of Myxilla fimbriata (Bowerbank, 1866). All measurements in um. Reference Acanthostyles | Tornotes iniri ux ane Cristobo et al. 190-260 129-200 18-25 35-60 Lundbeck, 1910 260-430 230-320 22-35 64-90 Descatoire, 1966| 210-310 160-250 25-30 60-75 mucus in formaldehyde during fixation. Skeletal arrangement: Choanosomal skeleton consists of quadrangular or triangular polyspicular meshes composed of 2-15 acanthostyles. Ectosomal tor- notes form bouquets protruding externally less than one third of the length ofthe spicule. Micro- scleres, sigmas and isochelae, are distributed throughout the sponge. Megascleres are straight or slightly curved acanthostyles, with strong conical spines highly variable in number arranged per- pendicularly to the axis of the spicules, ranging from smooth (i.e. styles) to completely bristled with spines covering the entire surface, and all intermediate gradations between. Dimensions: 89.6-162.0x2.5-12.3um. Smooth, straight tornotes, slightly fusiform, symmetrical similar extrem- ities ending in small straight spines. Dimensions: 106.2-158.6x1.5-8.3um. Microscleres: sigmas in typical c- and s- shapes in two size categories: 9.8-20.7x 0.3-1.5um and 21.3-31.7x1.6- 3.2um. Arched isochelae. Dimensions: 10.5-18.7x3.0-7.3um. Myxilla fimbriata (Bowerbank, 1866) (Figs 14-16, 17E) MATERIAL. Stations 35, 39 (see Fig. 1). AUTECOLOGY. In Galicia the species has been recorded from the biocenosis of Dendrophyllia cornigera between 50-58m depth, where it covers both the anthozoan and the brachiopod Terebratulina caputserpentis. This species is typical of the circalittoral and bathyal bottoms with bathymetric range between 50-3500m (Descatoire, 1966). It is abundant on bottoms characterised by the presence of Dendrophyllia cornigera at 60m depth, and less common in shallower waters less than 40m depth. In the sublittoral and circalittoral zones, the species is found in crevices (Descatoire, 1969). Also reported on Caryophylia clavus, Lophoelia prolifera and on rocky bottoms between 80-700m depth (Stephens, 1921). MEMOIRS OF THE QUEENSLAND MUSEUM DISTRIBUTION. North Atlantic and Arctic (Arndt, 1934). In Galicia: Laxe (Solórzano & Duran, 1982). DESCRIPTION. Encrusting sponge with a smooth or slightly crateriform surface without aquiferous orifices visibles macroscopically. Consistency is elastic; live colouration is pale yellow, brownish-ochre in alcohol. Dimensions: 23x5x2mm. Skeletal arrangement: Choanosomal skeleton is arranged in ascending tracts of acanthostyles interconnected by other transverse fascicles, with isochelae arranged around the tracts. The ectosomal skeleton consists of tornotes forming a relatively regular palisade, together with abundant isochelae. Megascleres: slightly curved acanthostyles with the distal end termin- ating in a sharp tip and irregular ornamentation with profuse spines on the head and a third of the distal region, except for the tip which is smooth. Dimensions: 133.2-264.0x6.1-12.1jum. Straight tornotes with slightly swollen and tapered distal extremities. Dimensions: 106.5-190.8x4.1-7.5um. Microscleres: spatuliferous isochelae, with two clearly differentiated size classes. Dimensions: 18.0-28.0 and 35.0-58.31m. KEY TO MYXILLA FROM THE NE ATLANTIC 1. With styles as choanosomal megascleres . ...... 2 With acanthostyles as choanosomal megascleres . . . 4 158) . With pluridentated subtylotes Lundbeck, 1905 With sharp tornotes 3. Only one class of spatuliferous anchorate chela M. pedunculata Lundbeck, 1905 Two classes of spatuliferous anchorate chela M. diversiancorata Lundbeck, 1905 4. Without isochelae . . . . . M. prouhoi (Topsent, 1892) M. pluridentata Wifhiscchelae. oe eo e EO Lais 5 5. Without sigmas. . . . M. fimbriata (Bowerbank, 1864) With sigas 4:2: bree e toga gt 6 6. With strongyles .... M.tarifensis Carballo & García Gómez, 1996 Withoutstrongyles. ......... lle 7 7. Tornotes without ends having small divergent points. 8 Tornotes with ends having small divergent points . . 9 8. Microspined tornote ends EA M. incrustans (Johnston, 1842) Smooth tornote ends . . . . M. fibrosa Levinsen, 1893 9. With tridentate chela . M. iotrochotina (Topsent, 1892) Without tridentate chela . .............. 10 10. Sigmas of two size classes OPES, av M. rosacea (Lieberkühn, 1859) Sigmas of one size class M, macrosigma Boury-Esnault, 1971 MYXILLA FROM GALICIA 119 FIG. 17. Habitus of Myxilla species from Galicia. A, M. macrosigma. B, M. rosacea. C, M. incrustans. D, M. iotrochotina. E, M. fimbriata 120 DISCUSSION Currently thirty-eight genera are assigned to Myxillidae (Hooper & Wiedenmayer, 1994), although recent revisions, such as Hajdu et al. (1994), Bergquist & Fromont (1988) and others, recognise fewer than these as correctly residing here. In Galicia species of Myxilla are amongst the more common sponges, both in terms of biomass and diversity. A key to species in the NE Atlantic is presented in Table 1. As compared to species from other latitudes, Myxilla from Galicia have certain unique characteristics in their morphology, habitat and distribution, as discussed below. Specimens of M. incrustans from the Ría de Ferrol show some differences in morphology of their tornotes as compared to specimens from other locations in Galicia. In the samples from Ferrol the tornotes have asymmetrical ends, one forming a tyle with a small elliptic head, perfectly defined by a tiny pre-capitular narrowing, and the other having a certain degree of polymorphism ranging from a spear-shaped tip to irregular shaped tips as illustrated in Figures 3-4. Other authors (e.g. Boury-Esnault et al., 1994) have described smooth tornotes with almost no spines. Myxilla iotrochotina is very similar to M. ros- acea but differs externally in its much smaller size and the fact that it forms small scales. Myxilla rosacea, on the other hand, is usually massive and frequently has considerable osculiferous chimneys, whereas the skeletal arrangement in both species is similar. The spicular composition is also similar; with acantho- styles, tornotes, and sigmas, but whereas M. rosacea has arched spatuliferous isochelae, M. iotrochotina has characteristic tridentate chelae with a straight spicular stem and three short teeth on the ends. On this basis we question the identification of Dendoryx iotrochotina from the Baleric Islands (Bibiloni, 1990), as it lacks tornotes with spiny ends and the isochelae are also different. The descriptions of M. macrosigma by Boury- Esnault (1971) and Boury-Esnault & Lopes (1985) agree with specimens described here from Galicia, highlighting one of the traits used to identify this species (viz. its mucus appearance). Acanthostyles have few spines, and the stem is practically bare; on the head spines are very scarce and may even be absent totally which gives the spicule the appearance of a true style. Comparisons between our samples from the Ria de Ferrol and paratypes collected from Banyuls- MEMOIRS OF THE QUEENSLAND MUSEUM sur-Mer revealed greatly similar lengths, widths, and maximum and minimum values of acantho- styles and tornotes between these populations. In morphological appearance, however, specimens from Ferrol have a slight thickening on the ends of many tornotes. The morphology of sigmas is comparable to the original description. Sigma sizes are, curiously enough, those that show the greatest discrepancy of the four spicular types, even though the maximum and minimum values of the two populations are found to lie within the range of dimensions originally described for the species (20-70um). The isochelae are similar in terms of size and shape. Myxilla rosacea has considerable poly- morphism in its external morphology, which may be attributed to microhabitat differences between localities (Bidder, 1923), among other factors. The most common form found in Galicia is massive, encrusting rocks and forming an irregular cushion 1-4cm thick, sometimes producing up to 20 digitiform oscular chimneys. Spicules also undergo great variations in morphology from one specimen to another, especially the acanthostyles as previously noted by Descatoire (1969), where the hispidation may be sparse, moderate or dense, with all intermediate stages. Other spicular elements (tornotes, isochelae) have a greater morphological homogeneity, whereas sigmas may be separated into two size classes. The skeletal arrangement also presents variations in terms of the geometry and arrangement of the meshes of choanosomal acanthostyles, which may be relatively slack and confused, related to the number of spicules forming skeletal meshes. The ectosomal skeleton is bouquet-shaped, a structure which has not been cited frequently in previous records of this species. Myxilla fimbriata on the coast of Galicia is typical of this species from other localities in its habitus, habitat and size although showing slight differences in spicule sizes as compared to those provided by Lundbeck (1910) and Descatoire (1966). In specimens from Galicia all types of spicules are generally smaller than those from Ingolf and Glenan, as demonstrated in Table 1. ACKNOWLEDGEMENTS We are grateful to Claude Lévi of the Museum National d'Histoire Naturelle (Paris) for providing access to these collections, to Nicole Boury-Esnault for her many positive comments, and to Chantal Bezac of the Station Marine d'Endoume (Marseille) for her competent technical MYXILLA FROM GALICIA assistance in Scanning Electron Microscopy, and to Manuel R. Solórzano of The University of Santiago de Compostela for providing the material for examination. This research was part of project number XUGA20005B95 sponsored by Xunta de Galicia. LITERATURE CITED ACKERS, R.G., MOSS, D. & PICTON, B.E. 1992. Sponges of the British Isles. A colour guide and working document. Sth Edn. (Marine Con- servation Society: Ross on Wye, Herefordshire). ACUNA, R., DURAN, C., SOLORZANO, M.R. & SANJUAN, A. 1984. Campañas de estudio del macrobentos infralitoral rocoso en el Parque Natural de las Islas Cíes (NW España). 1. Material, Metodo y Sistemática. Actas IV Simposio Ibérico de Estudio del Bentos Marino 1: 271-285. ARNDT, W. 1935. Porifera. Die Tierwelt der Nord-und Ostsee. 3* Edn. (Grimpie y Wagler: Berlin). BABIC, K. 1922. Monactinellida und Tetractinellida des Adriatischen Meeres. Zoologischen Jahrbiichen Systematik 46: 217-302. BENITO, J. 1976. Aportación al conocimiento de la fauna bentónica de la ría de Vigo (NW de España). II. Esponjas. Investigación Pesquera 40(2): 491-503. 1981. Algunas esponjas del litoral levantino español. Boletin del Instituto Español de Oceanografía 6(4): 90-102. BERGQUIST, P.R. & FROMONT, P.J., 1988. The marine fauna of New Zealand: Porifera, Demo- spongiae, part 4 (Poecilosclerida). New Zealand Oceanographic Institute Memoir 96: 1-197, BIBILONI, M.A. 1981a. Estudi faunistic del litoral de Blanes (Girona), Il. Sistemática d’esponges. Boletín Institut Catalan Historia Natural (Sección Zoología 4) 47: 5-59, 1981b. Estudio sistemático del Orden Poecilosclerida (Demospongia) de la Costa Brava (Girona). Boletín del Instituto Español Oceanografía 6: 104-154, 1990. Fauna de esponjas de las Islas Baleares. Variacion cualitativa y cuantitativas de la población de esponjas en un Gradiente Bati- métrico: Comparación Baleares-Costa catalana. Unpubl. doctoral tésis. (Universidad de Barcelona (Inédita): Barcelona). BIDDER, G.P. 1923, The relation of the form of a sponge to its currents. Quarterly Journal of Microscopical Science 67(2): 293-323. BOROJEVIC, R., CABIOCH, L. & LEVI, C. 1968. Inventaire de la faune marine de Roscoff. Spongiaires. Editions de la Station Biologique de Roscoff (1968): 1-41. BOURY-ESNAULT, N. 1971. Spongiaires de la zone rocheuse de Banyuls-sur-Mer. II.- Systématique. Vie et Milieu (B) 22(2): 287-350. BOURY-ESNAULT, N. & LOPES, T.M. 1985. Les Demosponjes littorales de l' Archipel des Açores. Annales de l'Institut Océanographique, París 61(2): 149-225. BOURY-ESNAULT, N., PANSINI, M. & URIZ, M.J. 1994, Spongiaires bathyaux de la mer d' Alboran et du golfe ibéro-marrocain. Memoires du Museum National d'Histoire Naturelle 160: 1-174. CARBALLO, J.L. 1994, Taxonomía, zoogeografia y autoecología de los Poríferos del Estrecho de Gibraltar. Unpubl. doctoral tésis. (Universidad de Sevilla: Sevilla). . . CARBALLO, J.L. & GARCIA-GOMEZ, J.C. 1996. Myxilla tarifensis sp.nov (Porifera, Poecilosclerida) with remarks on the Genus Myxilla in the Strait of Gibraltar (Iberian Peninsula). Journal of the Marine Biological Association ofthe United Kingdom 76: 875-884. CRAWSHAY, L. 1912. On the fauna of the western area of the English Channel. Journal of the Marine Biological Association of the United Kingdom 9: 292-393. CRISTOBO, F.J. 1997. Esponjas del Orden Poecilo- sclerida (Porifera, Demospongiae) de la Ria de Ferrol (NW de España). Unpubl. doctoral tésis. (Universidade de Santiago de Compostela: Santiago de Compostela). CRISTOBO, F.J., SOLORZANO, M.R. & URGORRI, V. 1992. Demosponjas de los fondos blandos de la Ria de Ferrol (Galicia, NO. de España). Nova Acta Científica Compostelana (Bioloxía) 3: 101-107. CRISTOBO, F.J., URGORRI, V., SOLORZANO, M.R. & RIOS, P. 1993. Métodos de recogida, estudio y conservación de las colecciones de poríferos. International Symposium & World Congress on Preservation and Conservation of Natural History Collections 2: 277-287. DESCATOIRE, A. 1966. Sur quelques Démosponges de l’ Archipel de Glenan. Cahiers Biologie Marine 7: 231-244. 1969. Peuplements sessiles de l'archipel de Glenan. I. Inventaire. Spongiaires. Vie et Milieu 20(1 B): 177-209. DURÁN, C. & SOLÓRZANO, M. 1982. Aportaciones al conocimiento del macrozoobentos de la zona infralitoral rocosa de Galicia mediante la utilización de la escafandra autónoma: I Desmosponjas. Trabajos Compostelanos de Biología 9: 49-67. FERRER-HERNANDEZ, F. 1918. Esponjas del litoral de Asturias. Trabajos del Museso Nacional de Ciencias Naturales 36: 1-39, GILI, C., ANADON, R., CARBONELL, J., OLIVELLA, 1. £ ROS, J. 1979: Comunidades bentónicas submarinas del litoral de Lugo. I Resultados preliminares. Actas 1 Simposio Ibérico de Estudio del Bentos Marino, San Sebastián 1: 711-749. HAJDU, E., SOEST, R.W.M. VAN & HOOPER, J.N.A., 1994. Proposal for a phylogenetic subordinal classification of Poecilosclerid sponges. Pp. 123-139, In Soest, R.W.M. van, Kempen, Th.M.G. 122 van & Braekman, J.-C. (eds) Sponges in Time and Space. (Balkema; Rotterdim). HOLME, N. & McINTYRE, A. 1984. Methods for the study of meiobenthos. (Blackwell Science Publisher: Oxford). HOOPER, J NA. & WIEDENMAYER, F. 1994, Porifera.. Pp. 1-624. la Wells, A. (ed.) Zoological Catalogue of Australia, Vol. 12. (CSIRO Australia: Melbourne). HOSHINO, T. 1981. Shallow-water demosponges of western Japan. |. Journal of Science of the Hiroshima University (Series B, Division I, Zoology) 29(1): 48-205. KONNECKER, G. 1973. Littoral and benthic investigations on the west coast of Ireland. L. (Section A: Faunistic and ecological studies). Whe sponge fauna of Kilkieran Bay and adjacent arcas, Proceedings of the Royal Irish Academy 73(B, 26): 451-472, LABATE, M. 1964. Poriferi di grotta superficiali del litoral adriatico pugliese. Annali dei Pontificio Istituto Superiori di Scienze e Lettere "8. Chiara’, Napoli 14; 319-342. LAMBE, L.M, 1892. On some Sponges from the Pacific coast of Canada and Behring Sea. Proceedings and ‘Transactions Royal Society Canada 10(4): 67-78. LEVI C. 1952, Spongiaires de la cote du Senegal Bulletin de l'Institut Francais d'Afrique Noire 14: 34-59. LUNDBECK, W, 1910. Porifera. Part Hl. Desmaci- doridae (PARS.). The Danish Ingolf Expedition 6: 1-124, PANSINL M. & PRONZATO, R. 1985. Distribution and ecology of epiphytic Porifera in two Posidonia oceanica (L,) Delile meadows of the Ligurian and ‘Tyrrhenian sea. Marine Ecology 60) 1-11, POGGIANO, G. 1965. La faune di Poriferi del Golfo di Pozzuoli, Annuario. dell’ Istituto e Museo di Zoologia della Université de Napoli 17(3); 1-19. POLO, L., OLIVELLA, L, GILL C., CARBONELL, J., ALTAMIRA, C. & ROS, J. 1979: Primera aportación a la sistemática de la flora y fauna bentónicas del litoral de San Ciprián de Burela (Lugo). Actas 1 Simposio Ibérico de Estudio del Bentos Marino, San Sebastián 1: 333375 POULIQUEN. L. 1972. Les spongiaires des grottes sous-marines de la region de Marseille: ecologic et systematique. Tethys 3(4): 717-758. PRONZATO, R. 1972. Poriferi del ‘fouling’ del porto di Génova. Bolletino dei Musei e degli Istituti Biologici dell'Università di Génova 40: 89-98, PULITZER-FINALI,G. 1978, Report on a collection of Sponges From the Bay of Naples, HI, Hadromerida, Axinellidà. Poecilosclerida, Halichondrida, Haplosclerida. Bolletino dei Musei e degli Istituti Biologici dell Università di Génova 45; 7-89, RODRIGUEZ, C. & LORENZO, J. 1978. Fauna marina de Galicia |l. Contribución al conocimiento de los poriferos del litoral Gallego. MEMOIRS OF THE QUEENSLAND MUSEUM Monografías de la Universidade de Santiago de Compostela 42; 1-68, RUBIO, M. 1973. Recolección y primera description de esponjas: fijación, conservación y preparación. Inmersión y Ciencia 5-6(3): 37-48. RUTZLER, K. 1978. Sponges on coral reef. Pp, 299-313. In Stoddart, D.R. & Johanness, R.E. (eds) Coral reefs: research methods. Vol. 5. (UNESCO: Paris). SARA, M. 1961, Poriferi di scogliera artificiale delle Golfo di Napoli. Annali dei Pontificio Istituto Superiori di Scienze e lertere *5. Chiara’, Napoli 10: 259-269. 19643. Porileri di acque. superficiali (0-3m.) del litorale italiano, Annali dei Pontificio Istituto Superior] dei Scienze et Lettere ‘S. Chiara’, Napoli 14: 299-317. 1964b. Distribuzione ed ecologia dei Poriferi in acque superficiali della riviera Ligure di Levante. Archivio Zoologico Italiano 49; 181-248, 1972. Guida dei Poriferi della fauna Italiana. Quaderni della Civica Stazione Idrobiologica di Milano 3-4: 53-98, 1974, Cutalogue des principales salissures marines. 5, Spongiaires. Centre de Recherches Etudier Oceanographiques 5: 1-42. SARÁ, M. & MELONE, N. 1963. Poriferi di acquc superficinli del litorale pugliese presso Bart. Annali dei Pontificio Istituto Superiori di Scienze e Lettere *8. Chiara’, Napoli 13: 1-28. SARA. M, & SIRIBELLI, L. 1960. La fauna di poriferi delle “seeche” del Golfo di Napoli. E La seecha della “Gaiola”, Annuario dell'Istitulo e Museo di Zoologia della Università de Napoli |2(3): 1-93 SIM, CJ. 1994, Sponges from Cheju Island, Korea. Pp. 175-181. In Soest, R. W,M. van, Kempen, Th.M.G. van & Brackman, J.-C. (eds) Sponges in Time and Space. (Balkema: Rotterdam). SOLORZANO, M.R. 1991. Inventario dos Poriferos du litoral galego (Porifera). Cadernos de Area de Ciencias Brolóxicas (Inventarios), Seminario de Estudos Galegos. Vol, 7. (0 Castro-Sada: A Coruña). . SOLORZANO, M.R. & DURAN, C. 1982. Nota preliminar sobre la fauna de poriferos asociados a Dendraphyllia cornigera (Lam., 1816) trente à las costas de Galicia. 11 Simposio Iberico de Estudio del Bentos Marino, Barcelona 3: 101-110. SOLORZANO, M.R. & RODRIGUEZ, C. 1979, Adiciones a la fauna marina de Galicia, 1. Demosponjas. Boletin del Instituto Español de Oceanogralia 5: 41-68. SOLORZANO, M.R. & URGORRI, V. 1991, Nuevas señalizaciones de Demosponjas en el Litoral Gallego (NO. de la Penisula Ibérica), Thalassas 9: 79-86. 1993. Nuevas citas de Demosponjas en la Peninsula Ibérica. Boletin de la Real Sociedad Española de Historia Natural (Seceión Biologia) 90(1-4): 61-66. MYXILLA FROM GALICIA SOLÓRZANO, M.R., CRISTOBO, F.J. & URGORRI, V. 1991. Introducción al poblamiento de poriferos de la biocenosis de maérl en la Ría de Arosa (Galicia, España) Boletin de la Real Sociedad Española de Historia Natural (Sección Biologia) 87(1-4): 175-185. STEPHENS, J. 1921. Sponges of the coast of Ireland. II. The tetraxonida (concluded). Fisheries Ireland Sciences Investigations 2: 1-75, TEMPLADO, J., GARCÍA-CARRASCOSA, M., BARATECH, L., CAPACCIONT, R., JUAN, A., LOPEZ-IBOR, A., SILVESTRE, R. & MASSO, C. 1986. Estudio preliminar de la fauna asociada a los fondos coraliferos del Mar de Alborán (SE de España). Boletin del Instituto Español de Oceanografía 3(4): 93-104. TOPSENT, E. 1892, Contribution a l'etude des spongiaires de l'Atlantique Nord. Résultats des Campagnes Scientifiques Prince Albert ler Monaco 2: 1-165, 1913. Spongiaires provenant des campagnes scientifiques de la ‘Princesse-Alice’ dans les mers du nord (1898, 1899, 1906, 1907). Résultats des Campagnes Scientifiques Prince Albert ler Monaco 45: 1-67. 1928, Spongiaires de l'Atlantique et de la Méditerranée, provenant des croisiéres du Prince Albert ler de Monaco. Résultats des Campagnes 123 Scientifiques Prince Albert ler Monaco 74: 278-280. URGORRI, V. & BESTEIRO, C. 1984. La alimentación de los moluscos nudibranquios de Galicia. Iberus 4: 51-58, URIZ, M.J. 1978. Contribución a la fauna de esponjas (Demospongia) de Catalunya. Unpublished Doctoral Tésis (Universidad de Barcelona: Barcelona). 1985. Primera aportación al conocimiento de la fauna espongicola de la plataforma de Namibia: Demosponjas de la ‘Benguela V” y de la “Valdivia I’. Simposio Internacional sobre las áreas de Afloramiento más importantes del Oeste Africano. Investigaciones Pesqueras 5(2): 639-649. 1986. Clave de identificación de las esponjas más frecuentes de la Península Ibérica. Miscelanea Zoológica 10: 7-22. 1988. Deep-water sponges from the continental shelf and slope off Namibia (Southwest Africa): Classes Hexactinellida and Demospongia. Monografías de Zoologia Marina 3: 9-157. VACELET, J. 1960. Eponges de la Mediterranée Nord Occidentale récoltées par le ‘Président-Théodore-Tissier’ (1958), Revue Travaux de l'Institut de Péches Maritimes 24(2): 257-272. VIDAL, A. 1967. Etude des fonds rocheux circalittoraux le long de la cote du Rousiillon. Vie et Milieu 18: 167-219. 124 MEMOIRS OF THE QUEENSLAND MUSEUM NITROGEN FLUX IN A SPONGE-MACRO- ALGAL SYMBIOSIS. Memoirs of the Queensland Museum 44: 124. 1999:- Nutrient cycling between corals and their zooxanthellal symbionts is important to the conservation of limiting nutrients, such as nitrogen, in oligotrophic reef waters. The tropical reef sponge Haliclona cymiformis forms an intercellular symbiosis with the red macroalga Ceratodictyon spongiosum, but it is unknown whether this association also promotes the cycling and conservation of essential nutrients. We therefore determined the potential importance of ammonium excreted by the sponge to the nitrogen- status and growth of the macroalga, using specimens collected from One Tree [sland Lagoon on the Great Barrier Reef. The association possessed the capacity to take up ammonium, nitrate and perhaps nitrite from the ambient seawater in the light, However these nutrients were commonly present at concentrations of less than 3 uM in the water at One Tree Island and so this seawater was probably only a minor source of nitrogen for the macroalga. In contrast, when the association was pre-incubated in darkness for 24hrs and its dark ammonium excretion rate measured, ammonium levels in the surrounding seawater (1 litre) increased from 0.25+ 0.1umolg dry weight to 2,2+0.6umol/g dry weight over a 6hr period; shorter dark pre-incubations resulted in lesser rates of ammonium excretion. It therefore appears that sponge waste is a major source of inorganic nitrogen for C. spongiosum, and this will be illustrated by means of a preliminary nitrogen budget. However, the enhancement of dark carbon fixation by ammonium (204M NH,CI), which increases as algae become more nitrogen limited, suggested that C. spongiosum was still nitrogen-limited in One Tree Island Lagoon, The ammonium enhancement ratio of freshly-collected C. spongiosum was 1.440,06, which compared to a ratio of 1.5+0.07 for cultured C. spongiosum when deprived of inorganic nitrogen for one week: the ratio ranged from 1.12: 0.1 for cultured C. spongiosum supplemented with a regular source of nitrogen (1004M NH,Cl) to 2.1+0.6 for cultured C. spongiosum deprived of nitrogen for 6 weeks. We therefore propose that the situation in the Haliclona-Ceratodictyon symbiosis is analogous to that in corals and other zooxanthellate invertebrates, with the animal partner being an important source of nitrogen for the alga. Furthermore, when combined with evidence for the translocation of nitrogenous compounds from the macroalga to the sponge (Grant et al., in prep), it is evident that the cycling and conserv- ation of nitrogen within the symbiosis may be an important factor in the success of this association in nutrient-poor habitats. O Porifera, sponge-macroalgal, nitrogen flux, nitrogen conservation, nitrogen budget. Simon K. Davy & Rosalind Hinde (email: rhinde@, bio,usyd.edu au), School of Biological Sciences, A12, University of Sydney. NSW 2006, Australia; 1 June 1998, DOES THE LARGE BARNACLE AUSTROBALANUS IMPERATOR (DARWIN, 1854) STRUCTURE BENTHIC INVERTEBRATE COMMUNITIES IN SE AUSTRALIA ? ANDREW R. DAVIS AND DAVID W. WARD Davis, A.R. & Ward, D.W. 1999 06 30: Does the large barnacle Austrobalanus imperator (Darwin, 1854) structure benthic invertebrate communities in SE Australia ? Memoirs of the Queensland Museum 44: 125-130. Brisbane. ISSN 0079-8835. The shallow subtidal zone of SE Australia is dominated by urchin-grazed barrens, created and maintained by a large urchin, Centrostephanus rodgersii (A. Agassiz). We sought to determine how benthic invertebrates, such as sponges and colonial ascidians, maintain space in the face of this intense grazing pressure. Our data indicate that the cover of invertebrates on vertical substrata was positively correlated with the density of a large barnacle Austrobalanus imperator and are consistent with this barnacle providing a refuge from urchin grazing. The exception was the common sponge Clathria pyramida which showed a strong negative relationship with barnacle density. We speculate that as aggregations of barnacles may represent foci for competitive interactions among sessile invertebrates, C. pyramida seeks to avoid these sites. It appears that recruitment of A. imperator is sporadic and hence the conditions which allow the establishment of high densities of this barnacle remain unclear. As our data are correlative they must be interpreted cautiously. Experimental manipulation of barnacle density will provide a much clearer indication of the role of A. imperator in structuring these communities and is the focus of current work. O Porifera, Crustacea, Echinodermata, grazing refugia, structural habitat complexity, urchin grazing, Clathria pyramida. Andrew R. Davis (email: andy. davis(a)juow.edu.au) & David W. Ward, Australian Flora and Fauna Research Program and Department of Biological Sciences, University of Wollongong, NSW, 2517, Australia; 15 March 1999. On subtidal rock walls, sessile invertebrates, such as sponges and ascidians, may cover a substantial area of the substratum. Typically in these communities habitable space is at a premium and competition for this resource is frequently intense. The loser in these spatial interactions is usually overgrown and killed. In addition to the intense spatial interactions characteristic of these communities, it is also clear that the action of predators (we use the term in its broadest sense and include grazing sea urchins) can have dramatic effects on community structure and dynamics. Urchins are capable of completely removing encrusting invertebrates and fleshy algae, thereby producing a community dominated by crustose algae (Lawrence, 1975; Vance, 1979; Ayling, 1981; Sebens, 1985; Witman, 1985). Some organisms can resist grazing by urchins and may provide refuges for less resistant species. Several studies have stressed the importance of refugia from urchin grazing in determining the structure of algal (e.g. Dayton, 1985) and invertebrate communites (e.g. Witman, 1985). For example, encrusting communities in the rocky subtidal of New England, USA, reach their greatest profusion either in the absence of urchins, or within beds of the mussel Modiolus modiolus when urchins are abundant (Sebens, 1985, 1986; Witman, 1985). This large bivalve adds significantly to the structural complexity of these rock wall habitats and serves as an important refuge for organisms against grazing urchins and predators (Witman, 1985; Ojeda & Dearborn, 1989). Structural habitat complexity may not only modify the foraging activities of predators it may also influence the settlement and recruitment of marine invertebrates (Keough & Downes, 1982). An examination of the structure of rock wall assemblages, must therefore consider the potential for grazer-resistant species (such as Modiolus) to form refuges, both from grazing and the activities of other predators. On the other hand, refuge forming species will take up space - a limiting resource - and may thus be in direct competition with those species that do not require a refuge. The expectation is then, that some species will be positively associated with refugia, while other species will be negatively associated. On the New South Wales (NSW) coast the sea urchin, Centrostephanus rodgersii, is the most 126 FIG. 1. A, The sponge Darwinella australiensis with a clump of Austrobalanus imperator as the urchin Centrostephanus rodgersii lurks nearby (Photo A. Davis). B, A high density of 4. imperator overgrown by the sponge Tedania sp. Only the opercular plates of the barnacles are visible (Photo D. Ward). conspicuous generalist grazer in the shallow rocky subtidal zone. This urchin creates and maintains extensive areas of substratum dominated by grazer-resistant encrusting coralline algae (Fletcher, 1987; Andrew, 1991). In these urchin grazed areas, sponges and colonial ascidians, reach their greatest profusion on rock walls or the vertical faces of large boulders. The large barnacle Austrobalanus imperator, which can measure over 4cm in height, also forms dense aggregations on shallow subtidal rock walls (Davis & Ward, unpublished data), thereby adding significantly to the structural complexity of these habitats (Fig. 1A). Another barnacle, Balanus trigonus, also often occurs in high densities in the subtidal zone, but is relatively small. MEMOIRS OF THE QUEENSLAND MUSEUM As part of a study to examine determinants of the structure and dynamics of natural rock surfaces in SE Australia we sought to determine how sessile invertebrates might maintain space in the face of intense grazing pressure by sea urchins, The sheer size of A. imperator combined with the high densities that occur on some tock walls should, we reasoned, represent a significant hinderance to grazing urchins and for that reason these barnacles are likely to play an instrumental role in determining the structure of subtidal rockwall communities. We initiated this study by collecting quantitative data on the relationship between invertebrate cover and the density of A. imperator, MATERIALS AND METHODS The relationship between barnacle density and invertebrate cover was investigated at three sites on the S coast of NSW, Sites were selected to encompass a length of the coast and for their convenient access. From north to south the sites were Henry Head (34°0.0°S, 151?14.2'E)at the N entrance to Botany Bay, the N end of Flinders Islet | (34?27.3'S, 150°55.7'E) near Wollongong, and Longnose Point (35°4.5'S, 150%47.0'E) within Jervis Bay (Fig. 2). These sites spanned almost 150km of coastline and we considered them representative of a much larger number of potential sites. Vertical surfaces, ranging from rock walls to the faces of large boulders were haphazardly photographed at all sites, usually between a depth of 5-15m. A camera (Nikonos V), mounted on a frame with twin strobes (Nikonos SB102 and Ikelite Ai strobes) to provide even illumination, produced photographs of an area of (0.08m?. Between 24-36 haphazard photographs were taken at each site after applying the criterion that the surface was vertical or near vertical. The total cover of invertebrates was estimated from the photographs using a transparency overlay with 100 systematically arranged dots. Only sessile inyertebrates were considered and so the anemone Anthothoe albocineta, although BENTHIC COMMUNITY STRUCTURE Henry Head r Flinders Islet * Redsands Reef * Longnose Pt. 151 00'E FIG, 2. Study sites on the S coast ol NSW, Australia, The relationship between invertebrate cover and barnacle density were assessed at Henry Head, Flinders Islet and Longnose Point. The size frequency of Austrobalanus imperator was determined at Flinders Islet and Redsands Reef. quite common in some areas was excluded owing to its mobility. The number of barnacles was also counted on each photograph. We were initially concerned that barnacles may be obscured by overgrowth, but found no evidence of this; the opercular plates of Ausirobalanus imperator were always visible (Fig. 1B), Barnacles down to a size (basal diameter) of 5mm could be reliably recorded from the photographs. Correlation coefficients (Pearson) were calculated for each site. Close examination of the photographs indicated that the pattern of distribution of one sponge, Clathria pyramida, contrasted with that of the other sessile invertebrates. To better assess the relationship between this sponge and the barnacles a series of haphazard photographs were taken of C. pyramida at Flinders Islet in September, 1995. Sponge cover was again estimated using the overlay transparency on the photographs and barnacles were also counted and recorded. In order fo get a clearer picture of the potential of A. imperator to form a refuge from urchin grazing we determined the size frequency of this barnacle at two sites; Flinders Islet and Redsands Reef (34935,7"S, 150°54.3°E, Fig. 2). The maximum diameter of the base of each barnacle was measured in the field with calipers. To ensure that basal width was a good estimator of barnacle tissue biomass we regressed barnacle tissue dry weight against basal width. Data were pooled from collections at Flinders Islet and Redsands Reef. RESULTS Clathria (Dendrocia) pyramida Lendenfeld (Porifera, Demospongiae, Poecilosclerida, Micro- cionidae). showed a significant negative relationship with the density of the barnacle 4. imperator (0.54, n=24, P<0.05, one tailed test, Fig. 3). In contrast, the total cover of invertebrates, excluding C. pyramida, was strongly positively correlated with barnacle density at all three sites. Barnacle density explained more than 30% of the variation in invertebrate cover at two of these sites, Henry Head (r=0.548, P<0.001) and Flinders Islet (170.586, P<0.001). The positive relationship between barnacle density and cover of invertebrates was not statistically significant by a one tailed students t-test at Longnose Point 100 - tn Percent Clathria Cover qu) un " e Barnacle Density FIG. 3. The cover of Clathia pyramida was negatively correlated with the density of the barnacle Austrobalanus imperator within photographic plots. Each data point represents a plot photographed on 13 September, 1995 at Flinders Islet. (r=-0.54, n=24, P<0.05, one tailed test). 128 100 a > a 2 Do E Bu? E 85 ne a E "c e i *, S 25 G e E e $5 a P. Je op" a O gdo Seo mo, * 1 10 100 1000 Log(Barnacle Density) FIG. 4. Positive correlation between total invertebrate cover (excluding Clathria pyramida) and the density of the barnacle Austrobalanus imperator. within photographic plots. Each data point represents a plot photographed at Henry Head (solid circle). Flinders Islet (open circle) and Longnose Point (square). All tests of significance were one tailed tests, Henry Head (r=0,548, P<0,001), Flinders Islet (r=0,586, P<0.001) and Longnose Point (r-0.107, P>0.05). (r=0.107, P>0.05). We present the data with the barnacle density log transformed to provide a more even spread of data on the x axis (Fig. 4). The modal size of adull barnacles at both sites was between 35-40mm, with the largest individuals having a basal width of around 55mm. Recruits of A. imperator were not recorded at either site from which we collected size frequency data. A cohort of 'sub-adults' were observed at the Redsands Reef site, but these animals were still quite large with a modal basal width of around 15mm (Fig. 5A). No such cohort was observed at the Flinders Islet site (Fig. 5B). Maximum basal width of the barnacles was an excellent estimator of animal biomass (Fig. 6). The resultant power function produced a very strong correlation (y-0.0000069" ^. 1—0.95). DISCUSSION With increasing densities of barnacles the cover of sessile invertebrates was also seen to increase. These findings are consistent with our initial suspicion that aggregations of Austrobalanus imperator form a refuge from sea urchin grazing. The rocky intertidal provides several examples of how habitat structure MEMOIRS OF THE QUEENSLAND MUSEUM influences the activities of grazers (Hawkins & Hartnoll, 1982; Creese, 1982; Dungan, 1986). For example, Creese (1982) reported that surface heterogeneity provided by bamacle shells can markedly influence the ability of grazers to feed. Limpets caged with high densities of a common intertidal barnacle starved to death (Creese, 1982). In addition to modifying patterns of invertebrate mortality, habitat structure may influence patterns of invertebrate colonisation. Bros (1987) reported a positive relationship between invertebrate recruitment and the addition of barnacle shells to glass slides in Tampa Bay Florida, although he noted that the treatments did not greatly affect the colonisation of sessile species. The responses of colonists to the modification of habitat structure on natural substrata remains unclear. Although it is tempting to ascribe our findings to an mcrease in structural heterogeneity produced by the presence ofthese large barnacles an equally plausible alternate explanation is that the barnacles simply enhance recruitment rather A. 60 50 HI Frequency B. Basal Width (mm) Frequency Basal Width (mm) FIG, 5. Size frequency of Austrobalanus imperator at. A, Flinders Islet; and B, Redsands Reef. Data were collected on a single day in October, 1994 at each site. BENTHIC COMMUNITY STRUCTURE a 2 = ap 9 0.75 2 2 A 0s o 3 E E 025 30 40 60 50 Maximum Basal Width (mm) 0 10 20 FIG. 6. Relationship between the maximum width of Austrobalanus imperator at their base with dry tissue weight (r=0.949). than reducing mortality. Changes in hydrodynamics near the rock surface or the additional surface area provided by barnacles when compared to a smooth rock surface are two potential mechanisms by which invertebrate recruitment may be enhanced. Nevertheless our data are correlative and in the absence of experimental evidence we can only speculate as to the processes which have produced the patterns we have observed. Notably, though, we have produced a conservative test of our initial hypothesis as a high density of barnacles will leave less space for other sessile invertebrates, yet we see higher invertebrate cover in the presence of high densities of barnacles. The striking negative relationship that we observed between barnacle density and the cover of C. pyramida suggests that this sponge is not reliant on the presence of barnacles to establish itself successfully and maintain space. Clathria pyramida clearly contains novel metabolites (Capon & Macleod, 1987) and field bioassays with crude solvent extracts of this sponge have revealed the presence of antifeedant natural products that dissuade urchins (Centrostephanus rodgersii) from grazing (Wright et al., 1997). The nature of the deterrent metabolites is currently unknown, but is under investigation (Davis & van Altena, work in progress). It is likely that the presence of biologically active metabolite(s) explains why this sponge does not rely on the presence of barnacles. However, the strong negative relationship we observed is consistent with the avoidance of barnacles by C. pyramida. This may be an appropriate strategy if this sponge is likely to encounter competitively superior species among aggregations of barnacles. Several studies reveal that some invertebrate larvae can detect competitive dominants and subsequently avoid sites where their survivorship is likely to be compromised (Grosberg, 1981; Davis, 1987). There is no need to invoke avoidance of barnacles or competitive dominants by the larvae of C. pyramida to explain the observed pattern; directional growth by adults could produce the same pattern. The reasons why C. pyramida avoids barnacles and the mechanisms used to do this remain speculative as nothing is known ofthe competitive ability of C. pyramida relative to other sessile species it is likely to encounter in SE Australia. It appears that the presence of A. imperator is an important contributor to the structure and dynamics of encrusting communities on vertical surfaces in SE Australia. Unfortunately, little is known of the biology of this cirripede or the determinants ofits distribution and abundance; of particular interest for example is how recruits of the barnacle withstand grazing by urchins. Our data reveal that recruitment of A. imperator is sporadic as we did not detect reliable recruitment to the barnacle population in the course of this study. There are also no data on the growth rates of this barnacle and therefore the time taken to reach a size that may interfere with the grazing activities of urchins, thereby forming a refuge. Experimental manipulation of the density of this barnacle is an important step in better understanding its role in these benthic commun- ities and is the focus of current work. ACKNOWLEDGEMENTS This work was done with financial assistance from the Australian Research Council and the Australian Flora and Fauna Research Centre, University of Wollongong. A number of people assisted us in the field, in particular we thank Martin Billingham, Carla Gannasin and Danny Roberts. We are grateful to Sue Fyfe and two anonymous reviewers whose comments improved the manuscript. This is contribution number 190 from the Ecology and Genetics Group, University of Wollongong. 130 LITERATURE CITED ANDREW, N.L. 1991. Changes in subtidal habitat following mass mortality of sea urchins in Botany Bay, New South Wales. Australian Journal of Ecology 16: 353-362. AYLING, A.M. 1981. The role of biological disturbance in temperate subtidal encrusting communities. Ecology 63: 830-847. BROS, W.E. 1987. Effects of removing or adding structure (barnacle shells) on recruitment to a fouling community in Tampa Bay, Florida. Journal of Experimental Marine Biology and Ecology 105:275-296. CAPON, R.J. & MACLEOD, J.R. 1987. 5-Thio-D-mannose from the marine sponge Clathria pyramida (Lendenfeld). the first example of a 5-thio sugar. Journal of the Chemical Society, Chemical Communication 1987:1200-1201. CREESE, D.G. 1982. Distribution and abundance of the acmaeid limpet Patelloida latistrigulata, and its interaction with barnacles. Oecologia 52: 85-96. DAVIS, A.R. 1987. Variation in recruitment of the subtidal colonial ascidian Podoclavella cylindrica (Quoy and Gaimard): The role of substratum choice and early survival. Journal of Experimental Marine Biology and Ecology 106: 57-71. DAYTON, P.K. 1985. The structure and regulation of some South American kelp communities. Ecological Monographs 55: 447-468. DUNGAN, M.L. 1986. Three-way interactions: barnacles, limpets, and algae in a sonoran desert rocky intertidal zone. American Naturalist 127: 292-316. FLETCHER, W.J. 1987. Interactions among subtidal Australian sea urchins, gastropods, and algae: effects of experimental removals. Ecological Monographs 57: 89-109. MEMOIRS OF THE QUEENSLAND MUSEUM GROSBERG, R.K. 1981. Competitive ability influences habitat choice in marine invertebrates. Nature 290: 700-702. HAWKINS, S.J. & HARTNOLL, R.G. 1982. The influence of barnacle cover on the numbers, growth, and behaviour of Patella vulgata ona vertical pier. Journal of the Marine Biological Association of the United Kingdom 62: 855-867. KEOUGH, M.J. & DOWNES, B.J. 1982. Recruitment of marine invertebrates: the role of active larval choices and early mortality. Oecologia 54: 348-352. LAWRENCE, J.M. 1975. On the relationships between marine plants and sea urchins. Oceanography Marine Biology Annual Review 13:213-286. OJEDA, F.O. & DEARBORN, J.H. 1989. Community structure of macroinvertebrates inhabiting the rocky subtidal zone in the Gulf of Maine: seasonal and bathymetric distribution. Marine Ecology Progress Series 57:147-161. SEBENS, K.P. 1985. Community ecology of vertical walls in the Gulf of Maine, USA: small scale processes and alternative community states. Pp. 346-371. In Moore, P.G. & Seed, R. (eds) The ecology of rocky coasts. (Columbia University Press: New York). 1986. Spatial relationships among encrusting marine organisms in the New England subtidal zone. Ecological Monographs 56: 73-96. VANCE, R.R. 1979. Effects of grazing by the sea urchin, Centrostephanus coronatus, on prey community composition. Ecology 60: 537-546. WITMAN, J.D. 1985. Refuges, biological disturbance, and rocky subtidal structure in New England. Ecological Monographs 55: 421-445. WRIGHT, J.T., BENKENDORFF, K. & DAVIS, A.R. 1997. Habitat associated differences in temperate sponge assemblages: the importance of chemical defence. Journal of Experimental Marine Biology and Ecology 213: 199-213. CUNVENIENT GENERA OR PHYLOGENETIC GENERA ? EVIDENCE FROM CALLYSPONGIIDAE AND NIPHATIDAE (HAPLOSCLERIDA]) RUTH DESQUEYROUX-FAUNDEZ Desqueyroux-Faúndez, R. 1999 06 30: Convenient genera or phylogenetic genera ? Evi- dence from Callyspongiidae and Niphatidae (Haplosclerida). Memoirs of the Queensland Museum 44; 131-146. Brisbane. ISSN 0079-8835. A taxonomie revision of all nominal genera in the haplosclerid Callyspongiidae and Niphatidae (Porifera; Demospongiae), is based exclusively on type species and discusses the taxonomic value of traditional characters. Of 27 available nominal genera in both families only 19 available genera are recognised. For some of them I tentatively propose subgenera (of Callyspongia), or synonymise them with other genera (e.g. —Cladochalina), Type species of 6 "chalinid' genera in the early literature of Lendenfeld 1886-1888 are comprehensively revised and their taxonomic status confirmed, as previously suggested by Burton in 1934, or changed, and some are illustrated for the first time. Morphometric characters important in defining a species group (genus or subgenus) inelude specific modifications to a specialised ectosomal skeleton, structure and distribution of ehoanosomal fibres, presence and width of the spongin sheath in the fibres, presence of foreign material and free spicules in the skeleton, and presence and amount of free spongin in the skeleton, General characters, only useful as a reference for identification of à species group, but not essential to establish their taxonomical status are also indicated. O Porifera, Demospongiae. Y, Haplosclerida, Callyspongiidae, Niphatidae, generic revision, taxonomy, morphology. Ruth Desqueyroux-Fuúndez (email: ruth faundez@mbn,ville-ge.ch), Museum d'histoire naturelle, Case postale 6434, CH-1211 Geneva 6. Switzerland; 1 February [999 The taxonomy of Haplosclerida is still controvertial as far as recognition, interpretation and definition of genera, subgenera and species groups are concerned. One reason for this is the small number of unequivocal taxonomic characters available (including secondary metabolite structures). The recognition of ‘reliable’ characters, which are discrete and consistent for a genus group, is a first condition for the construction of a classification based on natural affinities. Recent revisions of Haplosclerida provided some new insights to the taxonomy of this group. A study of West Indian Haplosclerida, notably re-examination of the Duchassaing & Michelotti collection, lead Van Soest (1980) to propose a new classification of families and genera, and to erect three new families: Niphatidae, Petrosiidae and Oceanapiidae (pro Phloeodictyidae). A phylogeny of Haplosclerida was proposed, based principally on features of the ectosomal skeleton, and nearly completely disregarding micro- scleres. A rival classification based on five descriptive criteria (colour, growth form, consistency, spicules and skeleton) was published by Bergquist (1980) and Bergquist & Warne (1980), who distinguished two orders (Haplosclerida and Nepheliospongida), and live families (Haliclonidae, Callyspongiidae, Adociidae, Nepheliospongiidae and Oceanapiidae [pro Phloeodictyidae). The monophyly of Haplosclerida was supported by de Weerdt (1985, 1986, 1989), ina study of three families of Eastern Atlantic Haplo- sclerida: Oceanapiidae (pro Phlocodictyidac), Chalinidae and Petrosiidae. De Weerdt (1985, 1986, 1989) bused ber phylogeny principally on synapomorphies of skeletal architecture, Van Soest (1990) claimed that Bergquist (1980) emphasised some of the apomorphic characters shared by Nepheliospongida and Haplosclerida, while at the same time recognising Nepheliospongida (pro Petrosida) as a distinct order, Van Soest (1990) proposed to keep Nepheliospongida (pro Petrosida) within the Haplosclerida, as suborder or superfainily. Presently, the taxonomic position of Petrosida is still unresolved. Sponge secondary metabolites also appear to be useful in characterising different groups of sponges (e.g, reviews by Bergquist & Wells, 1983; Van Soest & Braekman, 1999, this volume), and based on an analysis of sterols in Haplosclerida and Petrosida there is no chemotaxonomic support for a division of Haplosclerida into two orders (Fromont et al, 1994). A phylogeny of 132 MEMOIRS OF THE QUEENSLAND MUSEUM TABLE |. Callyspongiidae genera and their taxonomic assignments. Proposed subgenus Genus ‘Type species Original assignment | Actual assignment unen Synonymy Covachalina Carter, » 1885, Placachalina Toxochalinu. Fi " ` Lendenfeld, 1887 Callyspongia C. fallax A 7 Spinasella, 3 eit Duchéssaing & C. fallax Duchassaing & - (Caliyiprngio) Chalinapora, DA pond ly Micheloni, 1864 Michelotti 1864 filar Patulosenta [Ar Ande ae à Euplacella Lendenfeld, 1887, Chalinella Lendenfeld, 1887 Toxochalina Ridley, 1884 T. folivides Desmacidon faliores Bowerbank, (875 C, (Tusochalina) Jolivides Spinosella Tisha saroria C. (Spinoxella) Cladochalina S. sororia Duchassaing & mmi osmaer, 1885 Michelotti. 1864 soraría Schmidt, 1870 Chalinapora mi C. pipica C. (Chalinapara) Euchalina Lendenfeld, 1887 | € Pica Lendenleld, 1887 typica Lendenield, |B87 | Sclerachalina Schmidt, 1868, Siphonochalina TW M S Coriaved Siphonochalina Siphonella Schmidt, 1868 S eoriaven Schmidt, 1868 voriacea Lendenfeld, 1887. Tubulodigitus Carter, 1881 Patnloscila z P: procumbens C. (Patuloscula) Carter, [882 F- Arortimbels Carter, 1882 _procumbens Euplacella r " E. australis C. (Euplacella) Lendenteld, ]kg7 — | matrali Lendenfeld, 1887 australis Arenaselera x A. beroni Pulitzer-Finali, 1982 [4 Peron: Pulitzer-Finali, 1982 dieron Chalinapsis Dactylia a D, chalinifarmis Wer " Lendenfeld; 1886 Carter, 1885 B ehalinifürmis Carter, 1885 D: chaliniformis Chalinapsilla Lendenfeld, 1888 Haplosclerida based on molecules of the 3-alkylpiperidine alkaloid types appears incong- ruent with the phylogenetic tree of haplosclerid families, probably because of the uncertainty in identification of species analyzed and also ignorance of metabolite occurrence throughout a range of taxa (Andersen et al., 1996). Straight-chain acetylene compounds of Haplo- sclerida are efficient markers forthe whole order, and in some cases can be characteristic of certain genera and families (Van Soest et al., 1997), but these compounds are found among species of other orders, and their presence or absence is not an absolute indicator of phylogenetic relation- ships. Moreover, difficulties encountered in the application of biochemical methods can also be attributed to the diversity of other organisms living within sponges (e.g. symbiotic cyano- bacteria), making it arduous to identify the origin of ehemical compounds. in some instances (Bergquist & Wells, 1983). Consequently, chemotaxonomic data must be treated with caution, and these data are not considered in this present paper. In this paper, therefore, I am constrained to using to relatively few ‘reliable’ morphometric characters traditionally described for Haplo- sclerida, but this raises three problems. 1) Superficially at least, some of these ‘reliable’ morphological characters appear to be lacking in some taxa included in the five recognised families of Haplosclerida: Chalinidae, Niphatidae, Callyspongiidae, Phloeodictyidae and Petrosidae. 2) Conversely there are a number of nominal taxa, doubtfully available or presently considered as synonyms, which are clearly distinct from each other in their respective ‘reliable’ characters. 3) Frequently characters have been difficult to differentiate because their variability occurred within very narrow bounds, especially related to the structure of fibres. In order to clarify the status of these taxa T undertook a thorough analysis of the structural characters, especially at the skeletal level, in two families of Haplosclerida: Callyspongiidae and Niphatidae, | focused on the use of shared morphological characters as determinant tools to establish the taxonomic position of genera, TAXONOMY OF CALLYSPONGIIDAE & NIPHATIDAE 133 FIG. |. Callyspongia and Toxochalina. A, B, Callyspongia Duchassaing & Michelotti, 1864. Type species Callyspongia fallax Duchassaing & Michelotti, 1864. St. Thomas. Schizolectotype BMNH1928:11:12:5. A, Ectosomal network. Large, triangular to polygonal meshes, subdivided in smaller secondary and tertiary meshes. B, Choanosomal regular meshes, longitudinal principal and transversal connecting fibres. (Scales bars A = 200um; B = 8.2um). C, D, Toxochalina Ridley, 1884. Type species Desmacidon folioides Bowerbank. 1875, Straits of Malacca, ‘Bowerbank Collection’. Lectotype BMNH1887:5:21:2034. C, Ectosomal network, three types of meshes. D, Choanosomal network. (Scale bars C = 100um; D = 20um). subgenera and/or groups of species in Haplo- sclerida. I present here morphological evidence obtained exclusively from the study of type species of each nominal genus. MATERIALS AND METHODS Type material from the following Institutions was studied: BMNH, The Natural History Museum, London; MNHN, Muséum National d'Histoire Naturelle, Paris; IRSNB, Institut Royal de Sciences naturelles de Belgique, Brussels; ZMA, Zoological Museum, Amsterdam; MSNG, Museo de Storia Naturalle, Genova; AM, Australian Museum, Sydney; QM, Queensland Museum, Brisbane; MHNG, Muséum d'histoire naturelle. Geneva . Described characters, and taxonomic relationships based on them, reflect my own critical observations. Consequently, character analyses were completed with the inclusion of remarks from the original author’s descriptions. Specimens were studied by light and SEM micro- scopy. Lists of structural characters and character states of genera were established by comparing differences between genera. The following characters were used to compare genera: 1) variations in external morphology; 2) surface features; 3) type of ectosomal and choanosomal skeletons; 4) fibre structure and width variations of the spongin sheath; 5) presence of free spicules MEMOIRS OF THE QUEENSLAND MUSEUM FIG. 2 2. Spinosella (=Cladochalina) and Chalinopora (=Euchalina). A, C, Spinosella Vosmaer, 1885. Type species Tuba sororia Duchassaing & Michelotti, 1864. St. Thomas. Paralectotype, MUS. TORINO POR118.A, Longitudinal section through conular fascicle, surface at both sides of the figure. C, Three different sizes of triangular to polygonal ectosomal meshes and fasciculated subectosmal longitudinal fibres. (Scale bars A, C= 8.2um). B, D, Chalinopora (=Euchalina) Lendenfeld, 1887. Type species Chalinopora typica Lendenfeld, 1887. Port Jackson, NSW. Syntype BMNH1886:8:27:411 (AMG3408, slide). B, Simple ectosomal network, one size of fine triangular to rectangular meshes. Conules not visible. D, Choanosomal network, fasciculated underlying longitudinal primary fibres. Fragment of surface on top, (Scale bars B = 200m: D = 500um). in fibre meshes; and 6) amount of free spongin and type of spicules. Each character was checked for iis presence or absence; and when present, each character was scored objectively as to the expresion of the character, with at least three different states recognised per character. In the present work only the three most significant (‘reliable’) characters are presented. Other allegedly ‘inconsistent’ or *unreliable characters, deemed by previous authors to have little or no taxonomic value at the supraspecific level (e.g. external morphology), are omitted from this work, A more complete, phylogenetic analysis of these taxa will is in progress (Desqueyroux- Faúndez, in prep.). Whenever a character used in an original description was not sufficiently informative for the present work, a detailed description from a larger haplosclerid study was used (Desqueyroux- Faúndez, in prep.). In some cases, re-examin- ation of type specimens did not closely follow the published description, or there was mistaken identification of type material by the original author, with the consequence that the actual concept of some genera had to be changed (e.g. Hemigellius). Taxonomy and classification of families and genera are based on the most recently accepted classification of Porifera: Haplosclerida of Wiedenmayer, in Hooper & Wiedenmayer (1994). TAXONOMY OF CALLYSPONGIIDAE & NIPHATIDAE FIG. 3. Ceraochalina (=Chalimella) and Tubuladiginis. A, C, Ceraochalina Lendenfeld, 1887. Type species Ceraochalina typica Lendenfeld, 1887. Port Phillip, Victoria. Holotype BMNH1886:8:27:439. A, Transverse section of ectosomal network with one size of triangular meshes, subectosomal longitudinal fibres, intercalate short longitudinal fibres, small meshes and transverse fibres (arrow). C, Longitudinal section of surface, intercalate fibres and transversal fibres with echinating oxea. (Scale bars A 7 500um; C=8.21m). B, D, Tubulo- digitus Carter, 1881. Type species Tubuludigitus communis Carter, 1881. Bass Strait. Neotype BMNH- 1889:1:21:1. B, Longitudinal section through ecto and choanosomal skeleton, D, Transverse section of ectosomal skeleton, with one size of meshes. (Scale bars B, D = 20um). Modification of this classification in this work, and actual taxonomic assignments are indicated in each case (Table 1), Terminology for descriptive morphological characters is taken from Boury-Esnault & Rützler (1997). New morphometric terms were introduced and defined, when necessary. SYSTEMATICS A list of the major ‘reliable’ characters and their character states is presented as follows. CHARACTER 1. £ctosomal skeleton. 10), Tangential, regular network, three sizes of large triangular to polygonal meshes subdivided in smaller secondary and tertiary meshes (triple ectosomal network). Inconspicuous conules formed by free end of only one longitudinal primary fibre (Callyspongia, Fig. lA; Toxochalina, Fig. 1C). 1(2), Tangential regular network around a conspicuous central conule, with three sizes of large triangular to polygonal meshes subdivided in smaller secondary and tertiary meshes (triple ectosomal network). Central conspicuous conule produced by ends of the fascicle branches of longitudinal primary fibres (Spinosella (= Cladochalina), Figs 2A, C). 1(3), One size of fine triangular to rectangular meshes (simple ectosomal network) over confusely fasciculated underlying longitudinal primary fibres. Conules not visible (Chalinopora (=Euchalina), Fig. 2B, D). 1(4), One size triangular MEMOIRS OF THE QUEENSLAND MUSEUM TABLE 2. Niphatidae genera and their taxonomic assignments. Proposed subgenus. Genus Type species Original assignment | Actual assignment arrangement Synonymy ees Ridley, G. fibulata Axos fibulatus Carter, Gelliodes fibulata Microxina Topsent, | Mf eharcoti M. charcoti Topsent, Microxina charcoti 1916 : 1916 icroxina charcoti Niphates N. erecta itenim & N. erecta Duchassaing & Niphates erecta Michelotti, 1864 Michelotti, 1864 Dasychalina Ridley PS D. fragilis Ridley & Pachychalina ehali | & Dendy, 1887 D. fragilis Dendy, 1886 (Dasychalina) fragilis Dasychalina facie tie P. rustica P. inr ne Pachychalina rustica Amphimedon A, compressa : ai t Duchassaing & A, compressa Duchassaing & Amphimedon patas Bur Michelotti, 1864 Michelotti, 1864 compressa on, Hemigellius Burton, 1932 G. rudis Gellius rudis Hemigellius rudis Topsent, 1901 Cribrochalina adi C. infundibulum Cribrochalina Schmidt, 1870 O infundir Schmidt, 1870 infundibulum Haliclonissa Burton, " H. verrucosa Burton, Halielonissa 1932 Fo Verrugosá 1932 verrucosa meshes (simple ectosomal network); isolated underlying longitudinal fibres connected by 2-3 transverse fibres with echinating brushes of oxeas. Intercalate longitudinal fibres present to form small subectosomal meshes (Ceraochalina, Fig. 3A, C (= Chalinella); Siphonochalina, Fig. 4A (= Sclerochalina, Fig. 4B, D; Siphonella; Tubulodigitus, Fig. 3B, D). 1(5), One size rounded meshes (simple ectosomal network); isolated underlying subectosomal ends of longit- udinal primary fibres profusely divided to form uniform tangential network of poorly delimited unispicular fibres (Patuloscula, Fig. 5A, B). 1(6), One size rounded meshes (simple ectosomal network); ends of longitudinal primary fibres connected by three tangential successive layers of parallel fibres, echinated by numerous surface spicular brushes of oxeas. In longitudinal section subectosomal meshes appear smaller than choano- somal (peripheral condensation) (Euplacella, Fig. 5D). 1(7), Tangential irregular network of fragmentary unispicular fine fibres, one size rounded meshes (simple ectosomal TABLE 3. Presence of ectosomal and choanosomal characters and their distribution amongst genera of Callyspongiidae. network); string of foreign material (Arenosclera, Figs 6A). 1(8), Tangential irregular meshes; fine aspicular fibres Longitudinal | > iil connect finely cored by foreign material. No Gente Ectosomal | Subectosomal | intercalate | i, ectosomal distinction between primary and hieshies fibres aoa fibres secondary fibres (Dactylia, Figs 6C, D; Callyspongia three sizes isolated absent absent CRAT ica A Li CU. Toxochallina three sizes isolated absent absent rmultispicular fibres. inte ted by en AGE Spinosella three sizes | fasciculated absent absent l ongitudinal primary fibres (ramified Cerochalina one size isolated present absent spines or conules); free oxeas and sigmas Chalinopora one size fasciculated absent absent abundant (Gelliodes, Fig. 7B): 1(10), (E£uchalina) Strong free unordered oxea network, Siphonochalina interrupted by ends of longitudinal =Sclerochalina, , 5 n : iphonella DneGjze. | isolated absent absent primary fibres (brushes of spicules or ubulodigirus) strong spines), microscleres (sigma or Patuloscula one size isolated absent present microxe a) abundant or scarce Euplacella one size isolated absent present (3 layers) (Microxina, Fig. 8A; Niphates, Fig. 8C). Arenosclera one size absent absent absent 1(11), Tangential irregular network of Past Jla) | one size Eeid absent ash abundant free oxeas and fibres interrupted by strong ends of longitudinal primary TAXONOMY OF CALLYSPONGIIDAE & NIPHATIDAE FIG.4. Siphonochalina and (=Slerochalina), A, C, Siphonochalina Schmidt, 1868. Type species Siphonochalina coriacea Schmidt, 1868, La Calle, “Lacaze-Duthier’ collection. Syntype MNHN LBIM DT77. A, Dense tangential network of fine unispicular fibres with uniform meshes of only one type. C, Longitudinal section through choanosomal and ectosomal networks. Parallel, strong primary fibres with large spongin sheath, Ectosomal skeleton at the base of the figure (Scale bars A —8.2um; C =20.0um). B, D, Selerochalina Schmidt, 1868. Type species Sclerochalina asterigena Schmidt, 1868. La Calle, *Lacaze-Duthier” collection, Holotype MNHN LBIM DT 89-1 I. B. Simple ectosomal network with one size of triangular to rectangular small meshes, D, Longitudinal section through subectosomal and surface regions. Isolated parallel, strong longitudinal fibres, abundant spongin, surface below. (Scale bars B fibres (aculeations, prickles or stings) (Dasychalina, Fig. 9A). 1(12), Tangential regular fibre network with uniform rounded meshes. Ends of longitudinal primary fibres barely protruding (Amphimedon, Fig. 9C). 1(13), Perpendicular ill-defined extremely irregular network of spicule brushes in between the ends of primary fibres, riddled by aquiferous orifices (Pachychalina). (14), Tangential dense irregular network of free oxeas and sigmas forming continuous layer over ends of primary fibres (Hemigellius, Fig. 10B). 1(15), Perpendicular ends of longitudinal primary fibres expanded as strong continuous palisade of free oxeas (strongly hispid crust) (Cribrochalina, Fig. 10D). 1(16), 300um; D = 100um). Perpendicular ends of longitudinal primary fibres expanded to form wart-like elevations (verrucose surface), abundant free oxeas in between (Haliclonissa). CHARACTER 2. Choanosomal skeleton. 2(1). Regular network of longitudinal parallel strong primary fibres, regularly connected by short secondary fibres to form empty rectangular very regular meshes. Large spongin sheath present (Callyspongia Fig. 1B; Ceraochalina (= Chalinella), Fig. 3C; Siphonochalina, Fig. 4C). 2(2). Strong irregular network with large triangular to irregular meshes, poorly oriented; compact multispicular primary fibres irregularly MEMOIRS OF THE QUEENSLAND MUSEUM FIG. 5, Patuloscula and Euplacella. A, B, Patuloscula Carter, 1882. Type species Patuloscula procumbens Carter, 1882, Grenada, West Indies. Syntypes BMNH:1845:5:12:13, 15, 16. A, Longitudinal section, choanosomal region, isolated parallel subectosomal ends of primary longitudinal fibres at the base of the figure. B, Transverse section, continuous ectosomal layer, one size rounded meshes of poorly delimited unispicular fibres (simple network). (Scale bars A = 201m; B = 500um). C, D, Euplacella Lendenfeld, 1887. Type species Euplacella australis Lendenfeld, 1887. Torres Straits, Qld. Lectotype BMNH1886:8:27:591. C, Choanosomal network, paucispicular primary fibres with large spongin sheath, surface at left. D, Longitudinal section through triple ectosomal layer or ‘peripheral condensation’, with surface brushes of oxea, three layers indicated by arrows. (Scale bars C = 8.2um; D = 200um). split up to form connective fibres. Spongin sheath absent from all types of fibres, present only at fibre nodes. (Toxochalina, Fig. 1D). 2(3), Irregular confused network of multispicular fasciculated longitudinal primary fibres, irregularly split up to form short connective fibres. Empty meshes of only one type but of different sizes. All fibres with narrow spongin sheath (one 25 % of fibre diameter) (Chalinopora (= Euchalina ), Fig. 2D). 2(4), Strong network of stout longitudinal paucispicular, primary fibres and irregular short connecting fibres, irregular empty meshes. Fibres with large spongin sheath (Euplacella, Fig. 5C; Patuloscula, 5A). 2(5), Strong dense network of longitudinal primary fibres gathered to form fibrofascicles and split up to form free secondary fibres. Irregularly elongate to roundish meshes, always subdivided by tertiary finer fibres (Spinosella (= Cladochalina), Fig. 2C). 2(6), Aspicular network of longitudinal divergent primary fibres and perpendicular connecting fibres, abundantly cored by foreign material. Irregular to rectangular empty meshes (Dactylia (= Chalinopsilla ), Fig. 7C). 2(7), Intricate network of undifferentiated meshes and fibres with no preferential direction, not clearly distinguishable,abundantly cored by foreign material (Arenosclera, Fig. 6B). 2(8), Regular network of longitudinal primary fibres and short connecting fibres, isotropic to elongate TAXONOMY OF CALLYSPONGIIDAE & NIPHATIDAE 139 FIG. 6. Arenosclera and Dactylia. A, B, Arenosclera Pultizer-Finali, 1983. Type species Arenosclera heroni Pultizer-Finali, 1982. Heron Island, GBR, Qld. Holotype MSNG 46949. A, Tangential, irregular network of string of foreign material, one size rounded meshes (simple ectosomal network). B. Longitudinal section through ectosomal and choanosomal networks. (Scale bars A = 200um; B = 8.2um). C, D, Daetylia Carter, 1885. Type species Dactylia chaliniformis Carter, 1885. Port Phillip Heads, Victoria. Holotype BMNH 1886:12:15:196. C, Ectosomal network, abundant foreign material. D, Fine aspicular ectosomal fibres, finely cored by foreign material. (Scale bars: C = 20um; D = 50pm). meshes. Abundant free spicules (Cribrochalina, Fig. 10C). 2(9), Diffuse longitudinal primary fibres, tight and ill-defined meshes, abundant free spicules. Spongin abundant: free and coring the fibres (Amphimedon, Fig. 9D. 2(10), Poorly defined meshes with numerous free spicules in between the fibres; longitudinal primary fibres divergent, intermingled. Spongin inconspicuous (Haliclonissa). 2(11), Strong compact, multispicular primary fibres split up to form non connecting secondary fibres lacking orientation, irregular meshes, scattered spicules (Dasvchalina, Fig. 9B). 2(12), Open and loose, formed by compact longitudinal primary fibres and abundant free secondary fibres, oxeas and sigmas, irregular meshes, tree spicules (Gelliodes Fig. 7D). 2(13), Confused compact formed by abundant unordered strong spicules, no clear fibres or meshes. No visible spongin. Abundant sigmas (Hemigellius). 2(14), Confused, irregular lacuna with thick longitudinal primary fibres repeatedly divided to form isolated finer free ramifications, abundant free spicules, no clear meshes (Pachychalina. Fig. 10A). 2(15). Abundantly ramified with long thick non oriented fibres, large meshes and free spicules, no visible spongin, abundant microxea (Microxina, Fig. 8B). 2(16), Radiating longitudinal fasciculate primary libres, connecting secondary fibres regularly distributed to form rounded to irregular meshes. Free spicules abundant (Niphates, Fig. 8D) 140 MEMOIRS OF THE QUEENSLAND MUSEUM FIG. 7. Chalinopsilla and Gelliodes. A, C, Chalinopsilla Lendenfeld, 1888. Type species Chalinopsilla dichotoma Lendenfeld, 1886. West Coast of Australia. Lectotype BMNITI886:8:27:62. Schizolectotype: AM-G 8960 (MN HNLBIM DCL2061). A, Fine aspicular ectosomal fibres, abundantly cored by foreign material. C, Aspicular network abundantly cored by foreign material, divergent primary longitudinal fibres, and perpendicular connecting fibres. (Scale bars A =8.2um; C =20um). B, D, Gelliodes Ridley, 1884. Type species Axos fibulatus Carter, 1881. Bass Strait, Victoria. Syntype BMNH: 1882:2:23:202, B, Tangential view of surface, protruding ectosomal spines, from ends of longitudinal primary fibres (conules). D. Choanosomal skeleton of compact primary fibres and free secondary fibres. Free spicules abundant (Scale bars B 7 5001m; D — 20um). CHARACTER 3. Primary fibre structure. 3(1), Strong longitudinal, pauci to multispicular primary fibres (5-15 or more spicules), parallel to diver- gent, regular in width, isolated, not ramified or moderately ramified, not fasciculate. Spicules sparsely distributed at center of fibre. Large spongin sheath, at least 66 % of fibre diameter (Callyspongia; Patuloscula; Ceraochalina (7Chalinella). 3(2), Strong latge irregular spongin sheath cored by 3-5 spicules, some of them fused as observed by the presence of 2 or 3 central spicule rows; short, slender secondary fibres regularly split up from primaries (Euplacella). 3), Multispicular primary libres densely cored; spongin sheath absent. only with nodal spongin: secondary fibres of the same type. Unispicular tertiary fibres with very scanly spongin (Toxochalina). 3(4), Ascending parallel radially distributed primary fibres extending from internal to external sponge wall. Fibres paucispicular (3-5 spicules), not ramified. Subectosomal longitudinal intercalate fibres present (Siphonochalina (=Sclerochalina;, Siphonella; Tubulodigitus). 3(5), Aspicular to paucispicular primary fibres radiating from the sponge base to form fibrofascicles, surface conules and finer secondary and tertiary fibres. Large spongin sheath (Spinosella (=Cladochalina)). 3(6), Multi- spicular primary fibres with very narrow spongin sheath, less than 33 % of fibre diameter, or absent; TAXONOMY OF CALLYSPONGIIDAE & NIPHATIDAE 141 FIG. 8. Mieroxina and Niphates. A, B, Microxina Topsent, 1916. Type species Microxina charcoti Topsent, 1916. Antarctica. Holotype MNHN LBIM DT692, schizoholotype BMNH1926:10:26:339a. A, Ectosomal network of strong end brushes of primary fibres (strong spines). microxeas (arrows) in between. B, Strong ramified choanosomal tracts, abundant microxea. (Scale bars A = 500um; B = 20um). C, D, Niphates Duchassaing & Michelotti, 1864. Type species Niphates erecta Duchassaing & Michelotti, 1864. St. Thomas. Paralectotype MUS. TORINO PORS1; ZMA POR1633. C, Ectosomal network, tangential section. Strong free unordered oxea network, interrumpted by ends of primary fibres, brushes of spicules. D, Choanosomal network, fasciculated primary fibres, secondary fibres and rounded meshes. (Scale bars: C= 200um; D = 20um). secondary fibres of the same type (Chalinopora (=Euchalina)). 3(7), Primary and secondary fibres not clearly differentiated. Ali fibres lacking preferential direction, Spicules and foreign material variably present. Thicker fibres cored only by foreign debris, or both foreign debris and proper spicules. Thinner tracts with sparse spicules or uncored. Proper spicules absent if foreign material abundant (4renosclera). 3(8), Aspicular, primary fibres isolated longitudinal, not ramified, slightly branched to form short, fine, aspicular, amber-like, connecting fibres. All fibres with abundant foreign material (Chalinopsilla). 3(9), Multispicular, inconspic- uous, plumose, anastomosing, radially ascending primary fibres producing diffuse irregular secondary fibres (Amphimedon). 3(10), Strong, regular, well-defined multispicular and radially ascending primary fibres not ramified. Spongin sheath narrow or missing. Short interconnecting fibres of the same structure (Cribrochalina). 3(11), Very stout compact, multispicular, primary fibres split up to form paucispicular secondary fibres. No visible spongin sheath (Dasychalina). 3(12), Very stout, compact, ascending, radiating, branching and anastomosing, splitting up to produce irregularly oriented, secondary fibres. No distinct spongin sheath except at bifurcation points (Gelliodes). 3(13), Pauci to multispicular longitudinal primary fibres, divergent, diffuse, isolated and rarely ramified. Connecting fibres not defined, only abundani free spicules (Haliclonissa). 3(14), Confused, compact primary tracts, with no clear fibres, only strong abundant 142 MEMOIRS OF THE QUEENSLAND MUSEUM FIG. 9. Dasychalina and Amphimedon. A, B, Dasychalina Ridley & Dendy, 1886. Type species Dasychalina fragilis Ridley & Dendy, 1886. ‘Challenger’ Collection, Philippine Islands. Schizotype BMNH1887:2:170. A, Surface aculeations, spines or prickles, from ends of primary fibres, abundant free oxeas in between. B, Strong spicules, choanosomal fibres without distinct sheath of spongin (Scale bars A = 500um; B = 50pm). C, D, Amphimedon Duchassaing & Michelotti, 1864. Type species Amphimedon compressa Duchassaing & Michelotti, 1864, St. Thomas. Lectotype MUS.TORINO POR.35; Schizoparalectotype BMNH | 928:1 1:12:42. C, Tangential, regular ectosomal network with uniform rounded meshes, barely protruding ends of primary fibres. D, Choanosomal skeleton, plumose, anastomosing primary fibres, irregular secondary fibres, abundant spongin (Scale bars C = 50um; D = 2001m). spicules, unordered and cemented by no visible scarce spongin (Hemigellius). 3(15), Very stout, multispicular primary longitudinal, irregularly ascending, compact, large, occasionally fascic- ulated, without spongin sheath (Microxina). 3(16), Pauci to multispicular primary longitudinal fibres, diffuse, abundantly branched to form fibrofascicles, and pauci-to multispicular secondary connecting fibres. Spongin dominant between loose spicules, inside the fibres or free in meshes (Niphates). 3(17), Thick, irregular and compact multispicular primary fibres, with no preferential orientation; spongin sheath absent. Free spicules abundant. Thinner ill defined multispicular secondary fibres (Pachychalina). TAXONOMY OF CALLYSPONGIIDAE & NIPHATIDAE FIG. 10. Pachychalina, Hemigellius, Cribrochalina. A, Pachychalina Schmidt, 1868. Type species Pachychalina rustica Schmidt, 1868. La Calle, “Lacaze-Duthier” collection. Syntype MNHN LIBM DT47. Confused, irregular, lacunar choanosomal network, thick longitudinal primary fibres, repeatedly divided to form isolated finer free ramifications, abundant free spicules. (Scale bar = 20um). B, Hemigellius Topsent, 1901. Type species Hemigellius rudis Topsent, 1901. Antarctica, ‘Belgica’ Expedition. Holotype RBINSC PORO33. Ectosomal skeleton, tangential dense, unordered network of free oxeas and sigmas, no clear fibres or meshes, no visible spongin. (Scale bar = 200 um). C, D, Cribrochalina Schmidt, 1870. Type species Cribrochalina infundibulum Schmidt, 1870. West Indies. Lectotype BMNH1870:5:3:165. C, Longitudinal section through choanosomal elongate meshes, abundant free spicules. (Scale bar = 8.2um). D, Longitudinal section through ectosomal and subectosomal skeleton, ectosomal palisade (crust) at left. (Scale bar =8.2um). DISCUSSION In both Callyspongiidae and Niphatidae studied here, ‘reliable’ generic characters matched structural differences within the group: ectosomal skeleton, choanosomal skeleton structure, and fibre structure. It is therefore important to determine the presence of these characters and their stability amongst genera, in order to use them as diagnostic characters. In Haplosclerida the essential concept of ‘genus’ taxon is often misinterpreted or confused. Citations of genera as: “Petrosia Vosmaer “sensu” Ridley & Dendy, 1887”, or “Callyspongia Duchassaing & Michelotti, 1864, ‘sensu’ Burton, 1932” (Wiedenmayer, 1977), have no special validity or status in formal taxonomy. These terms have often a very different meaning than the one given by the original author. It is clear that this kind of citation should be avoided, particularly ifit is not based on personal re-examination of type material. Such citations result in a new concept of the genus, an unnecessary widening of the generic concept without corroboratory evidence from the type specimen, and where ‘sensu’ the new author gives a new subjective diagnosis of the genus - all of which tend to become ‘fixed’ in the literature with their tacit acceptance by E contemporary aulhors. 1 is true that these modifications certainly make genera more ‘convenient’ for taxonomie identification of species, particularly in cases where the original concept of the genus ts clouded or questionable. However, they omit the phylogeny and frequently lead to very heterogeneous taxa. The question is then: do we accept genera or subgenera for their ‘convenience’, or do we have to find evidence for phylogenetically valid taxa ? What ate {he characters to consider when we split or fuse a nominal genus into an actual genus? Some of the genera studied here are difficult to delimit, because af the large variability in their structural characters. For example Callyspongia, the type genus of Callyspongiidae, shares its habit with other genera of the same family: the lectotype of Callyspongia. C. fallax, has been described as massive, repent and lobate (Van Soest, 1980), but many species (including the type species) vary in habit from massive, ramose, lobate. repent to tubular. Today the genus is contams species with a great diversity and variab- ility in growth forms, such that. the concept of “shape” has little taxonomic importance at the supraspecific level im this case. Nevertheless, Callyspongia exhibits a typical skeleton with a very stable fibre structure. Hooper & Wicdenmayer (1994) include 16 nominal genera as synonyms of Callvspongia, and Wiedenmayer (1989) included 21 nominal generic synonyms of Callyspongia. This is symtomatic of the biggest problem in Haplosclerida whereby the formulation of exact definitions and delimitation of generic boundaries is nearly impossible. Large revisions of problematic genera, following a strict interpretation of the genus’ original concept, lead in some cases to the creation of genera with similar characters, or conversely to merge genera displaying some mutual characters (*splitting? versus 'lumping^). lo the present work comparisons between type species of accepted generic synonyms of Callyspongia provided a clear mandate to differentiate them, based on their ectosomal features, whereas in some cases differences in their choanosomal skeletons were so minor as to be inconsequential, Su far, differences between type species of genera considered as synonyms of Callvspongra hy Hooper & Wiedenmayer (1994) are too subtle to retain them us available genera (c.g. Ceraochalina, Chalinella; Chalinopora; Euchalina) whereas, conversely, the existence of several species showing consistent similarities in MEMOIRS OF THE QUEENSLAND MUSEUM some of their characters vonfinn the potential validity of some of their nominal species groups, for which ‘convenient’ subgenera may be appropiate and “useful! for classification (although perhaps not always phylogenetically sound taxa). A similar solution was adopted for a revision of the large family Microcionidae, with 73 nominal venera included (Hooper, 1996). Tentative conchusions summarised in Tables 1 and 2 provide a convenient, practical classif- ication, but they require confirmation from other sources using more objective methods (e.g. molecular studies). Skeletal characters analyzed here were very often difficult to objectively differentiate, because vanability occurred between very nartow limits, especially concerning the choanosomal skeleton and the structure of fibres and their variations. In these cases it 1s necessary to determine character priority in generic diagnoses as a first intent to delimit problematic genera. The next stage in this analysis, an objective interpretation of characters and the distribution of character states amongst taxa, in a phylogenetic framework, should incorporate some of the more variable morphometric characters of Haplosclerida (e.g. habit, texture, surface omamentation and aquiferous system). Authors have discarded these features as being ‘not useful" at the generic level (e.g. Van Soest, 1980), but certainly some higher taxa such as Cribrochalina (Niphatidac) can be defined by a ‘sticky texture’, reflecting consistency in both skeletal and chemical characters, whereas in others this feature is completely inconsistent and discarded. Work ts continuing in this regard (Desqueyroux- Faúndez, in prep. 1. CONCLUSIONS Grouping genera of Callyspongiidae and Niphatidae appears to be feasible using ihe structural characters defined above. Differences between genera are stable, consistent and deemed to be diagnostically important at the generic level. In Callyspongiidae (Table 3), two groups of genera are distinguished, based on comparison of the form and size of ectosomal meshes with the structure of underlying longitudinal fibres: 1) Gener with three different sizes of triangular ectosomal meshes: Callvspongia, Toxochalina and Spinosello (=Cladachalina), 2) Genera with ome size of ectosomal meshes: Chalinopora (=Euchalina), Ceraochalina (=Chalinella), Siphonochalina (=Selerochalina, Siphonella, TAXONOMY OF CALLYSPONGIIDAE & NIPHATIDAE Tubulodigitus), Patuloscula, Euplacella, Arenosclera and Dactylia (=Chalinopsilla). Furthermore, in this first group of genera this ectosomal character is associated with (linked to) the choanosomal character comprising *fasciculated or isolated subectosomal longitudinal fibres’. In contrast, the character comprising the ‘presence of transversal subectosomal connecting fibres’ appears only in some genera displaying one size of meshes: Siphonochalina (=Sclerochalina, Siphonella, Tubulodigitus); Patuloscula. Similarly, Ceraochalina (=Chalinella), with one size of meshes, presents a different modification of subectosomal isolated longitudinal fibres in the form of ‘presence of intercalate fibres and small subectosomal meshes’, whereas Euplacella has one size of meshes and isolated longitudinal fibres, but it represents a modification of this character with three successive and isolated layers of ectosomal skeleton. This is considered here to represent different degrees in the development of the ectosomal layer, similar to those observed in genera of Phloeodictyidae (Oceanapia (=Rhizochalina). Arenosclera appears to be atypical of Callyspongiidae, having an irregular and disorganised skeleton without proper fibres and fibres strongly cored by foreign material. Wiedenmayer (1989) considered that the presence/absence of foreign material in fibres is a poorly correlated feature and only occurs as a gradual transition within some species of Callyspongiidae, with no clear boundaries between present/ absent. Nevertheless, there is a precedent for recognising a subgeneric taxon with incorp- orated detritus in the skeleton in Microcionidae (Clathria (Wilsonella)) (Hooper, 1996), as this feature was consistent within the species group and corroborated by the consistent morphological features. In this regard Arenosclera might form a ‘convenient’ and phylogenetically valid subgenus in group 2 Callyspongiidae. In Niphatidae, the ectosomal skeleton appears to be a stable ans supraspecific character at the generic level, and in this regard two groups can be distinguished. 1) Genera with tangential ectosomal skeleton that may be formed by: a) a fibre network interrupted by ramified ends of longitudinal fibres (Gelliodes) or by barely protruding ends of longitudinal fibres (Amphimedon); b) a network of free oxea, interrupted by ends of longitudinal fibres, or spicule brushes with additional sigma or microxea (Microxina, Niphates); or c) both a fibre network and free oxeas interrupted by ends of longitudinal 145 fibres or aculeations (Dasychalina). 2) Genera with perpendicular ectosomal skeletons that may be formed by: a) spicule brushes and ill-defined ends of longitudinal fibres (Pachychalina); b) a spicule palisade and expanded ends of longitudinal fibres (Cribrochalina); or c) free spicules issued from the expanded ends of longitudinal fibres or wart-like conules (Haliclonissa). Although, the characters used here appear to be useful to genus groups amongst Callyspongiidae and Niphatidae, their treatment is equivocal in the absence of data about their potential interrelations. Further studies on these two families is still necessary prior to accepting these criteria for a definitive classification. Nevertheless the present analysis provide a positive beginning towards a resolution of a very difficult and slightly chaotic group of sponge taxa. ACKNOWLEDGEMENTS For many years, Professor Claude Lévi (MNHN) helped me to understand the taxonomy of Haplosclerida. John Hooper (QM) and Rob van Soest (ZMA) assisted me with their valuable time encouraging this work in taxonomic discussions. I thank many colleagues from different institutions, for providing constructive comments on this manuscript and assisting me to Obtain type material studied in this work: Philippe Willenz (IRSNB), Clare Valentine (BMNH), Valter Raineri and Enrico Borgo (MSNG). Technical work was undertaken at MHNG and I am grateful to J. Wüest (SEM), C. Ratton (photography) and I. Juriens (histology). Thanks to both reviewers who greately improved my manuscript. LITERATURE CITED ANDERSEN, R.J., SOEST, R.W.M. VAN & KONG, F.M. 1996. 3-alkilpiperidine alkaloids isolated from marine sponges in the order Haplosclerida. Pp. 301-355. In Pelletier, S.W. (ed.). Alkaloids: Chemical and Biological Perspectives. Vol. 10 (Pergamon: New York). BERGQUIST, P.R. 1980. A revision ofthe supraspecific classification of the orders Dictyoceratida, Dendroceratida and Verongida (Class Demospongiae). New Zealand Journal of Zoology 7: 443-503. BERGQUIST, P.R. & WARNE, K.P. 1980. The Marine Fauna of New Zealand: Porifera, Demospongiae, Part 3 (Haplosclerida and Nepheliospongida). New Zealand Oceanographic Institute Memoir (87): 1-77. 146 BERGQUIST, P.R. & WELLS, R.J. 1983. Chemo- taxonomy of the Porifera: The development and current status of the field. Pp. 1-46. In Scheuer, V.P.J. (ed.) Marine Natural Products: Chemical and Biological Perspectives. Vol. 5 (Plenum Press: New York). M BOURY-ESNAULT, N. & RUTZLER, K. (eds) 1997. Thesaurus of Sponge Morphology. Smithsonian Contributions to Zoology 596: 1-55. BURTON, M. 1932. Sponges. Discovery Reports 6: 237-392 (Cambridge University Press: Cambridge). 1934. Sponges. Scientific Reports ofthe Great Barrier Reef Expedition 1928-29. 4: 513-621. (British Museum (Natural History): London). FROMONT, J.P., KERR, S., KERR, R., RIDDLE, M. & MURPHY, P. 1994. Chemotaxonomic relationships within, and comparisons between, the orders Haplosclerida and Petrosida (Porifera: Demospongiae) using sterol complements. Biochemical Systematics and Ecology 22(7): 735-752. HOOPER, J.N.A. 1996. Revision of Microcionidae (Porifera: Demospongiae: Poecilosclerida), with description of Australian species. Memoirs of the Queensland Museum 40: 1-626. HOOPER, J.N.A. & WIEDENMAYER, F. 1994. Porifera. Pp. 1-620. In Wells, A. (ed.). Zoological Catalogue of Australia. Vol. 12. (CSIRO Australia: Melbourne). LENDENFELD, R.VON 1886. Studies on sponges. Two cases of mimicry in sponges. Proceedings of the Linnean Society New South Wales 10: 569-574, pls 39-44. 1887. Die Chalineen des australischen Gebietes. Zoologische Jahrbiicher Jena 2: 723-823, pls 18-27. 1888. Descriptive Catalogue of the Sponges in the Australian Museum, Sydney. (Taylor & Francis: London). MEMOIRS OF THE QUEENSLAND MUSEUM RIDLEY, S.O & DENDY, A. 1887. Report on the Monaxonida collected by H.M.S. ‘Challenger’ during the Years 1873-76. Report on the Scientific Results of the Voyage of H.M.S. ‘Challenger’ during the Years 1873-76. Vol. 20 (Her Majesty’s Stationery Office: London, Edinburgh, Dublin). SOEST, R.W.M. VAN 1980. Marine sponges from Curacao and other Carribean localities. Part II. Haplosclerida. Studies on the Fauna of Curaçao and other Caribbean Islands 62: 1-173. 1990. Toward a Phylogenetic Classification of Sponges. Pp. 344-348. In Rützler, K. (ed.) New Perspectives in sponge biology. (Smithsonian Institution Press: Washington, D.C.). SOEST, R.W.M. VAN, FUSETANI, N. & ANDERSEN, R.J. 1997. Straight-chain acetylenes as chemotaxonomic markers of the marine Haplosclerida. Pp. 3-30. In Watanabe, Y. & Fusetani, N. (eds) Sponge Sciences. Multidisc- iplinary Perspectives. (Springer-Verlag: Tokyo, Berlin, Heidelberg, New York). WEERDT, W.H. DE 1985. A systematic revision ofthe North Eastern Atlantic shallow-water Haplo- sclerida (Porifera: Demospongiae), Part I: Introduction, Oceanapiidae and Petrosiidae. Beaufortia 35(5): 61-91. 1986. A systematic revision of the North-Eastern Atlantic shallow-water Haplosclerida (Porifera: Demospongiae), Part II: Chalinidae. Beaufortia 36(6): 81-165. 1989. Phylogeny and vicariance biogeography of North Atlantic Chalinidae (Haplosclerida: Demospongiae). Beaufortia 39(3): 55-88. WIEDENMAYER, F. 1977. Shallow-water sponges of the western Bahamas. Experientia Supplementum 28: 1-287, pls 1-43 (Birkhauser: Basel). 1989. Demospongiae (Porifera) from northern Bass Strait, southern Australia. Memoirs of the Museum of Victoria 50(1): 1-242. CROSS-SHELF DISTRIBUTION OF SOUTHWEST SULAWESI REEF SPONGES N.J. DE VOOGD, R.W.M. VAN SOEST AND B.W. HOEKSEMA Voogd, N.J. de, Soest, R.W.M. van & Hoeksema, B.W. 1999 06 30: Cross-shelf distribution of southwest Sulawesi reef sponges. Memoirs of the Queensland Museum 44: 147-154. Bris- bane. ISSN 0079-8835. The quantitative distribution of open reef sponges (number of individuals per square meter) was studied across various reefs in the Spermonde Archipelago, southwest Sulawesi, Indonesia from April-July, 1997. The reefs are situated in 4 shelf zones that vary in distance offshore. Those closest to shore are subjected to freshwater inflow, nutrient input and sedimentation, whereas outer reefs are subjected to wave-action and upwelling from the Makassar Strait. Sponge individuals visible to the eye were identified and counted in 23 100% 1m? belt transects distributed over the four shelf zones at depths of 3m, 6m, 9m and 15m. Distribution patterns of reef sponges were investigated in three spatial scales: 1) distance from land; 2) depth; 3) orientation to wind direction. The highest number of sponge species (richness) and individuals (abundance) was found in a middle-shelf reef, whereas the lowest richness and abundance was found in an inner-shelf reef. Lowest species richness and abundance occur in shallow transects (3-6m), and the highest were found in deeper transects (9-15m). More sheltered sites of reefs are lower in species richness than more exposed sites. Most species appear to have a wide distribution across the Spermonde shelf, and few are restricted to specific reef zones or depths. In contrast, the number of phototrophic sponge species and individuals increases with increasing distance from shore, with the highest numbers in the reefs farthest from shore. The occurrence of these sponges seems to be related to more clear, oligotrophic waters, such as found in the open Makassar Strait. 0 Porifera, Spermonde Shelf, Indonesia, sponge distribution, species diversity, phototrophy. N.J. de Voogd (email: soest@bio.uva.nl) & R.W.M. van Soest, Institute of Systematics and Ecology (Zoological Museum), University of Amsterdam, P.O. Box 94766, 1090 GT, Amsterdam, The Netherlands; B.W. Hoeksema, National Museum of Natural History - Naturalis P.O. Box 9517, 2300 RA Leiden, The Netherlands; 5May 1998. Many marine invertebrate species appear to reach their highest diversity in the Indo-Malayan region (Briggs, 1987), and sponges are no exception (Van Soest, 1994). However, sponges in Indonesia have been largely neglected by science for a long time. The few publications on sponges from this area were descriptions of small collections picked up almost casually, published during the first half of the century. Most of our knowledge of the Indonesian sponge fauna is based on collections made by large expeditions, such as the ‘Siboga’ expedition (1899-1900), and more recently from the Indonesian-Dutch ‘Snellius II’ expedition (1984-1985). Thousands of specimens were collected and identified, but remained unpublished (Van Soest, 1989). A database of these species, including manuscript names assigned to specimens by the late Maurice Burton (BMNH collections), appear in Hooper et al. (1999), Nevertheless, published information available suggests that a high degree of dis- similarity exists between sponge faunas in various reef locations within Indonesia (Van Soest, 1989; Amir, 1992). Differences in species composition has been mostly attributed to the degree of exposure to waves and currents (Amir, 1992), but these conclusions are based on very few data. Several studies hay, een made on differences in sponge populatio 1; .elationto various physical factors on the Great Barrier Reef. Wilkinson & Cheshire (1989) studied sponge distribution patterns across a broad continental shelf, related to distance offshore and depth. They concluded that highest biomass occurred on inner-shelf reefs and decreased further away from the mainland; abundance was highest on middle-shelf reefs; and this appeared to be correlated with different environmental factors across the continental shelf. Inner-shelf reefs are influenced by terrigenous run-off, high nutrient concentrations and fresh water inflow, whereas outer-shelf reefs are susceptible to oceanic features such as wave energy, oligotrophic conditions and upwelling. Wilkinson & Trott (1985) showed that light was also a factor determining sponge distribution across a continental shelf. Light transmittance varied considerably across a longitudinal continental shelf, and thus influenced the distribution of 148 MAKASSAR STRAITS MEMOIRS OF THE QUEENSLAND MUSEUM - manna T v - = 3 Makassar > Pandang M S ¿$ > Strait SS eS a a Locality 4 MEMOIRS OF THE QUEENSLAND MUSEUM 100 4 B 1000 80 4 + 800 Ei = 604 Fe E g E : ; g 404 p40 3 E 3 A [2 < 20 4 —TI— species [ 200 —8— individuals 0 0 3 6 9 15 Depth (m) FIG. 3. Species richness (left axis) and abundance (right axis) per reef. A, localities arranged according to distance from the mainland at the capital Ujung Pandang; B, species richness and abundance per depth in metres. Indices are mean values. increasing depth (Fig. 3b). The situation at the nearshore reef at Lae-Lae differed from that on other reefs (Fig. 4b) in showing a higher richness and abundance at 6m than 9m depths, where the bottom predominantly consisted of sand and silt as compared to hard substratum on other reefs. The 15m depth habitat was absent at Lae-Lae. EXPOSED VS SHELTERED REEF SITES. Reefs at which both the exposed (west) and sheltered (east) sides were monitored, Samalona (shelf zone 2) and Kudingareng Keke (shelf zone 3), showed the lowest richness and abundance occurring on the sheltered side (Fig. 5). This difference is especially obvious at Kudingareng Keke, where species richness and abundance on the E side are about half that of the W side. Differences in both richness and abundance are most distinct between 9 and 15m depth contours. PHOTOTROPHIC SPONGE DISTRIBUTION. In contrast to the general trends observed for all sponges, abundance of phototrophic sponges continued to increase with distance offshore (Fig. 6). This trend was also reflected in species abundance, increasing from 2 at the inshore-reef, 4 at both middle-shelfreefs, to 6 at the outer-shelf reef. Conversely, abundance of individual photo- trophic species deviated from this trend: Dysidea aff. herbacea was slightly less abundant at Sama- lona (middle-shelf reef) than at Lae-Lae (inshore reef), whereas Halichondria cartilaginea was slightly more abundant at Kudingkareng Keke (outer middle-shelf reef) than at Langkai (outer-shelfreef). Averaged for all reef locations, the highest abundance of phototrophic species was found at 15m, and of individual transects sampled the highest abundance was found at the 15m transect at the outer-shelf reef of Langkai. SPONGE COMMUNITY DISTRIBUTION. Similarities in species composition and abundance (number of individuals) of all reef locations and depths were determined by means of cluster analysis (Fig. 7). Three distinct clusters were recognised. Cluster 1 contains all Lae-lae sites. This reef was generally characterised by low species diversity and abundance. The high abundance of Paratetilla bacca and Dysidea aff. herbacea was notable, but these species also occurred elsewhere. Three species were confined to Lae-Lae, but they were rare and represented by only few individuals. Cluster 2 contains all shallow sites of 3m depth, with the exception of Lae-Lae (cluster 1), and is also characterised by low species richness and abundance. Relatively high abundance of Gelliodes callista, Halichondria cartilaginea and Phyllospongia papyracea were notable, but these also occur in cluster 3. One species is confined to the 3m zone, but only a single individual was found. Cluster 3 comprises the remaining transects. Five species (Xestospong- ia ashmorica, Haliclona fascigera, Niphates olemda, Clathria vulpina and Ulosa sp.) meet the criteria to be considered as ‘characteristic and dominant species’ of this large cluster, occurring in at least 66% of the transects and an average DISTRIBUTION OF SULAWESI REEF SPONGES 15m Species richness / 100 m2 e LLL AÀ LD 0 Mahkassur Sirait Ujung Pandang -lie — Samalona K. Keke Langkai Locality I5] B 10 4 9 4 SS T \ 3 | N Bog N EE N E N 5, N 2°37 S : \ à I N N : A. Ujung Laelae Samalonà K Keke Langkal POM Purdang Locality FIG. 4. Species richness (A) and abundance (B) per reefat each depth profile, Each shelf zone is represented by a reef and arranged according to distance from the shoreline at the capital Ujung Pandang. abundance exceeding 5 % (Kaandorp, 1986). Some species only occur in subsets of this cluster, but their occurrence invariably overlaps only partially with that of other clusters, causing a dissimilarity index too low for confident recognition of distinct clusters, Some patterns are nevertheless worth noting: e.g. all transects on the east side of Kudingareng Keke are grouped together, having in common low species richness and abundance. The dendrogram clearly indicates that the first shell zone and all the shallow (3m) sites are different from the remaining sites. DISCUSSION CROSS-SHELF DISTRIBUTION PATTERNS. Sponge species richness and abundance increase with distance offshore to the third shelf zone, decreasing further offshore. The Spermonde Archipelago is affected by the NW monsoon, both geomorphologically and ecologically (de Klerk, 1983). These differences seem to be relat- ed to distance from land, and clearly suggest an environmental gradient across the shelf (Hoek- sema, 1990). The lowest number of coral species was found in the outer reef zone, but at the same time the outer-rim high energy environment provides a higher coral cover than on the inner- rim reefs (Moll, 1983). These patterns include all scleractinian corals. A study of cross-shelf dist- ribution patterns in fungiid corals revealed that the number of species increased with increasing distance from shore until the third shelf zone. The fourth reef zone exhibited a lower species richness and abundance. Iloeksema (1990) concluded that the third reef zone was the most optimal reef zone for fungiid species richness. Similar to these trends our results found that the third reef zone was also (he most optimal for sponge populations. Moll (1983) found no clear differences in 90 - 800 80 - 700 37 | 600 3 = 60 % E / x E E 400 3 | p» A 300 E $ 30 y 4 E 4 A p 20 % Hu < E 2 i0 Z 100 2 0 " 0 3 6 9 15 Depth (m) C species cast species west —1M— individuals cast —®— individuals west FIG. 5. Species richness (bars: left axis) and abundance (line: right axis). Values are mean values for transects on east and west sides of Samalona and Kudingareng Keke. 152 No. of phototrophic individuals / m2 M L 0 A $ ba Eu EPIA $ Locality d FIG. 6. Numbers of individuals of presumed phototrophic sponges in the four reel areas arranged according to distance offshore. scleractinian species richness between reef zones 2 and 3, although the third zone had slightly higher number of species. The largest differences were found between reef zones 2-3 and zone 4. In our study, we found the highest proportion of sponge species with the most restricted distrib- utions occurred at Langkai in the fourth reef zone, From earlier studies on cross-shelf distribution patterns of sponges on the Great Barrier Reef Wilkinson & Cheshire (1988, 1989) found that sponge biomass showed an inverse relationship with distance from land. They concluded that this was due to high amounts of nutrients in near- shore waters, decreasing with distance from land. Their data correlated with the hypothesis that the clear-water sponge fauna depends predominantly on nutrition from their cyanobacterial symbionts. Abundance of sponges on the Great Barrier Reef was highest in middle-shelf reefs, whereas the pattern of richness appeared more complicated. Sponge abundance on the Spermonde shelf in- creased similarly with distance from land until the midshelf reefs. Species richness also follow- ed this pattern on Spermonde reefs, whereas this was not the case on the Great Barrier Reef. However, the Great Barrier Reel constitutes a much broader area, with the outer-shelf reefs in the Townsville region located some 200km from land, whereas the inner-shelf reef is almost 20km MEMOIRS OF THE QUEENSLAND MUSEUM 3LL 6LL 2LL 3 SA east 354 JKK east 3LA 3KK 6 SA east 6KK 68A 98A 9LA 9 SA east 9KK IS KK 15 SA east 158A 6 KK east 9 KK east 15 KK east 6 LA T 15 LA l — 10 08 0.6 He == J 04 02 0 Dissimilatity FIG. 7. Dendrogram ofa hierarchical classification on species composition and sponge abundance far all transects. Transects are listed with depth in metres. Abbreviations: LL, Lae-lae; SA, Samalona; KK, Kudingareng Keke; LA, Langkai. from shore. Species compositions of Great Barrier Reef zones were not compared between zones, and it is conceivable that distinct faunal assemblages exist over the broad shelf, making ecological comparisons more complicated. The inner-shelf reefs of the Great Barrier Reef are perhaps more comparable with the second and third reef zone in the Spermonde Archipelago, where the inner-shelf reef is located only Ikm from urban Ujung Pandang. Presumably, this area is far more prone to terrestrial influences than an inner-shelfreef on the Great Barrier Reel Like on the Great Barrier Reef (Wilkinson & Cheshire, 1988; Wilkinson & Evans, 1989), the abundance of presumed phototrophic sponges in the Spermonde area appears to be related to the presence of cleat, oligotrophic waters, DEPTH DISTRIBUTION, On the Great Barrier Reef, high irradiance, high UV light and high wave energy were thought to exclude most species from shallow waters, These factors decrease with increasing depth, hence species richness and abundance increase with depth (Wilkinson & Cheshire, 1988; Wilkinson de Evans, 1989). The depth distribution of sponges at the Spermonde shelf follows this general pattern closely, and it 1s probable that these DISTRIBUTION OF SULAWESI REEF SPONGES 15 mechanisms are also responsible for differential distribution patterns in this region. WITHIN REEF ZONE VARIATION. Moll (1983) observed that species richness and ab- undance in scleractinian corals was significantly lower on the E side than at the W side of Sper- monde shelf islands. Hoeksema (1990) found that circumreef patterns in fungiid corals were mainly determined by water movement. The W (exposed) side is, in general, the most exposed to wave- energy caused by the strong NW monsoon, whereas the E (sheltered) side shows more sediment accumulation. This was evident from the low species richness and abundance on the E sides of reefs (Hoeksema, 1990). In our study we found that species richness and abundance of sponges are lower on E sides than on W sides of reefs. The slopes of E sides of the islands of Samalona and Kudingareng Keke are very steep and contain high amounts of sediment in com- parison with the W side. In conclusion, most species appear to have a wide range of distribution across the Spermonde Archipelago, and few are restricted to specific zones and depths. However, both species rich- ness and abundance increase with depth, and also with increasing distance offshore until the third reef zone. Specific measurements to correlate with these variations were not made, thus it is not possible at this moment to characterise these habitats. ACKNOWLEDGEMENTS The first author received financial support from the Dutch Royal Academy of Science and the STIR-network. We want to thank Universitas Hassanuddin for all help and especially Dr Alfian Noor for the use of his laboratory and all his other support. The research was conducted as part of the WOTRO Programme for Sustainable Man- agement of the Coastal Zone of Sulawesi, Indonesia (W01-60) through grant WK84-354 of the Netherlands Foundation for the Advance- ment of Tropical Research (WOTRO). LITERATURE CITED AMIR, I. 1992, A comparison of sponge fauna of exposed and sheltered reef flats in Eastern Indo- nesia. Marine Research in Indonesia 28: 1-12. BERGQUIST, P.R., AYLING, A.M. & WILKINSON, C.R. 1988. Foliose Dictyoceratida of the Aust- ralian Great Barrier Reef I. Taxonomy and phylogenetic relationships. Publicazionidella Stazione Zooligica di Napoli (Marine Ecology) 9(4): 291-319. I] BEST, M.B. & ZONNEVELD, J.1.S. 1989. Inter- disciplinary reef studies in Indonesia 1980-1995 (the Buginesia Programme) Pp. 35-47. In Tropical Research in development: WOTRO 1964-1989. (WOTRO: The Hague). BRIGGS, J.C. 1987. Biogeography and Plate tectonics. In Developments in Palaeontology and Strati- graphy. (Elsevier: Amsterdam). DINESEN, Z.D. 1983. Patterns in the distribution of soft coral communies across the central Great Barrier Reef. Coral Reefs 1: 229-236. HOEKSEMA, B.W. 1990. Systematics and ecology of mushroom corals (Scleratinia, Fungiidae). PhD thesis, University of Leiden, Leiden. HOOPER, J.N.A., KENNEDY, J.A. & SOEST, R.W.M. VAN 1999. Annotated checklist of the sponge from the South China Sea. The Raffles Bulletin of Zoology (in press). HUTCHINSON, D.R. 1945, Coral reefs and cays of the Makassar Straits. HQ Allied Air Forces SW Pacific Area Intelligence Memoirs 50(2): 1-30. KLERK, L.G. DE 1983. Zeespiegels, riffen en kustvlakten in Zuidwest Sulawesi, Indonesié, een morphogenetisch-bodemkundige studie. PhD thesis, University of Utrecht,Utrecht. MOLL, H. 1983. Zonation and Diversity of Scleratinia on the Reefs of S.W. Sulawesi, Indonesia. PhD thesis, University of Leiden, Leiden. PIELOU, E.C. 1975. Ecological diversity. (Wiley: New York). SHANNON, C.E. & WEAVER, A. 1949, The mathematical theory of communication. (University of Illinois Press: Urbana, Illinois). SOEST, R.W.M. VAN. 1983. Shallow-water reef sponges of Eastern Indonsia. Pp. 302-308. In Riitzler, K. (ed.) New Perspectives in sponge biology. (Smithsonian Institution Press: Washington D.C.). 1989. The Indonesian sponge fauna: A status report. Netherlands Journal of Sea Research. 23(2): 223-230. 1994. Demosponge distribution patterns. Pp, 213-223. In Soest, R.W.M. van, Kempen, T.M.G. van & Braekman, J.C. (eds) Sponges in time and space. (Balkema: Rotterdam). SOKAL, R.R. & MICHENER, C.D, 1958. A statistical method for evaluating systematic relationships. Kansas University Science Bulletin 38: 1409- 1438. VERHEU, E. 1993. Marine plants on the reefs of the Spermonde Archipelago, SW Sulawesi, Indo- nesia: Aspects of taxanomy, floristics and ecology. PhD thesis, University of Leiden, Leiden. WILKINSON, C.R. 1981. Significance of sponges with cyanobacterial symbionts on Davies Reef, Great Barrier Reef. Pp. 871-882. In Gomex, E.D. et al. (eds) Proceedings of the 4th International Coral Reef Symposium. Vol. 2. (University of the Philippines: Quezon City). 1987. Productivity and abundance of larger sponge populations on Flinders Reef flats, Coral Sea. Coral Reefs 5(4): 183-188. 1988. Foliose Dictyoceratida of the Australian Great Barrier Reef, 2. Ecology and distribution of these prevalent sponges. Pubblicazionidella Stazione Zooligica di Napoli (Marine Ecology) 9(4): 321-327, WILKINSON, C.R. & CHESHIRE, A.C. 1989, Patterns in the distribution of sponge populations across the central Great Barrier Reef. Coral Reefs 8: 127-134. WILKINSON, C.R. & EVANS, E.A. 1989. Sponge distribution across Davies Reef, Great Barrier MEMOIRS OF THE QUEENSLAND MUSEUM Reef, relative lo location, depth and water movement, Coral Reefs 8: 1-7. WILKINSON, C.R. € TROTT, L.A. 1985, Light as a factor determining the distribution of sponges across the central Great Barrier Reef. Pp. 125-129. In Harmelin Vivien, M. & Salvat, B. (eds) Proceedings of the Fifth International Coral Reef Congress. Vol. 5. (Antenne Museum-Ephe: Moorea, Tahiti). WISHART. D. 1978. CLUSTAN user manual. (Program Libary Unit, Edinburgh University: Edinburgh). PERSPECTIVES ON SPONGE-CYANO- BACTERIAL SYMBIOSES. Memoirs of ihe Queensland Museum 44: 154. 1999:- Insights on the evolution of sponge-cyanobacterial symbioses are drawn from biogeographic and molecular dala. The taxonomic and geographic distribution of sponge-cyanobactería associations is analysed after surveying their occurrence at eight localities in the Eastern and Western Tropical Pacific, and the Caribbean. Three methods - fluorescent microscopy, thin layer chromatography and transmission electron microscopy - were used to infer the existence of endosymbiotic cyanobacteria. Thirty-eight species, representing 17 families and 11 orders of Demospongiae, and one family and order of Calcarea, are added to the list of sponges involved in these associations. This number represents an increase of more than 50% over previously known occurrences of this type of metazoan-microbial association. However this increase of species numbers represents only an addition of twelve genera and two families to the taxonomic distribution of these associations. Species from 26 of the 72 recognised Demospongiae families, and 3 of the 17 recognised Calcarea families are found to harbour cyanobacterial endosymbionts. These data suggest a rather restricted taxonomic range for sponge-cyanobacterial assemblages, and invites a search for evolutionary trends among the families involved. The genera with highest number of species harbouring cyanobacteria are: Aplysina (10 spp.), Xestospongia (7 spp.), Dysidea (5 spp.), and Theonella (5 spp.). Although the updated list of' sponge- cyanobacterial assemblages shows a few biogeographic trends, the understanding of the evolution of these associations requires the study of more extensive geographic areas. The use of 168 rDNA analysis to understand the phylogenetic relationships of endosymbintic cyanobacteria is discussed. Genetic analyses promise to shed light on the understanding of the evolution and specificity of these associations. 168 rDNA gene analyses carried out so far suggest that sponge-cyanobacterial assemblages comprise diverse and complex evolutionary histories, some of which might share evolutionary pathways with other important marine symbiotic assemblages involving cyanobacteria. O Porifera, cyanobacteria endo- symbioses. biogeography, evolutionary trends, 168 ribosomal genes. M.C. Diaz (email: diuz(@cuts.uesc.edu), Institute of Marine Sciences, A316 EMS, University of California Santa Cruz, CA 95064, USA; B.B. Ward, Institute of Marine Sciences and Ocean Sciences Department, A316 EMS, University of California Santa Cruz, CA 95064, USA; 1 June 1998. FARMING SPONGES FOR THE PRODUCTION OF BIOACTIVE METABOLITES A.R. DUCK WORTH, C.N. BATTERSHILL, D.R. SCHIEL AND P.R. BERGQUIST Duckworth, A.R., Battershill, C.N., Schiel, D.R. & Bergquist, P.R. 1999 Qo 30: Farming sponges for the production of bioactive metabolites. Memoirs of the Queensland Museum 44: 155-159. Brisbane. ISSN 0079-8835. For successful aquaculture of sponges, with the aim of producing metabolites, a farming method is required that promotes sponge growth and survival, and produces high yields of target metabolites. To help develop a suitable farming method growth and survival were compared for two New Zealand sponges, Latrunculia brevis (Ridley & Dendy) and Polymastia croceus (Kelly-Borges & Bergquist), experimentally grown in a variety of ways. Explants were farmed in mesh, on rope, and with rope threaded through them. For both species of sponge, survival was greatest for explants farmed in mesh, probably because this produces little tissue damage and prevents explants from dislodging and ‘escaping’. This method also promoted highest growth of L. brevis, with some explants doubling their weight in two months. The growth of P. croceus, however, was highest in explants with rope threaded through them. Explants of both sponges farmed on rope did not attach and had poor growth and survival. These findings are a major step forward in developing a method for farming sponges in temperate waters of New Zealand. O Porifera, aquaculture, farming method. Alan R. Duckworth (email: a.duckworth(Aniwa.cri.nz), University of Canterbury/NIWA Centre of Excellence in Aquaculture and Marine Ecology, PO Box 14901, Kilbirnie, Wellington, New Zealand; Christopher N. Battershill, National Institute of Water and Atmospheric Research (NIWA), PO Box 14901, Kilbirnie, Wellington, New Zealand; David R. Schiel, Zoology Department, University of Canterbury, Private Bag 4800, Christchurch, New Zealand; Patricia R. Bergquist, Zoology Department, University of Auckland, Auckland, New Zealand; 24 March 1999, A major obstacle facing sponge aquaculture in the production of metabolites is the lack of a suitable farming method or on-growing structures (Shimizu, 1995; Osinga, 1998). To be suitable for large scale commercial use a structure must be inexpensive, have a low surface area to reduce drag and bio-fouling, and allow cost-effective and efficient harvesting. It must also promote high sponge growth and survival while also maintaining high metabolite production. Farming structures used to grow bath sponges have historically involved attaching explants to concrete discs, or threading wire through explants so that they hang in mid-water (Cotte, 1908; Moore, 1908; Crawshay, 1939). This last method was modified slightly by Verdenal & Vacelet (1990), who successfully grew commercial bath sponges by first threading plastic-coated metal wires through explants and then attaching them to vertical ropes. Development of new farming structures to grow bath sponges was constrained by market forces determining acceptable shape and size of products (Storr, 1964; Bergquist & Tizard, 1969). In contrast, explant shape has no bearing on efficient metabolite production, and consequently there is considerable flexibility in the development of new farming structures for metabolite aquaculture. We identified three general farming methods: 1) explants placed in mesh; 2) explants attached to and farmed on rope; 3) explants farmed with thin rope threaded through them (Fig. 1). The first method has already been tested with some success (Duckworth et al., 1997). For each method of farming it was necessary to test variation in structures and materials used. For example, rope thickness and composition were important considerations using methods 2 or 3 - as rope thickness increases, drag pressure as well as capital cost increases accordingly, whereas a decrease in rope thickness produces a decrease in available surface area for explant attachment, Rope composition is also important, because explant growth, survival and metabolite concentration may differ between ropes made of different materials. In this study, we tested the potential of each farming method using two New Zealand sponges: Latrunculia brevis (Ridley & Dendy, 1886), a green massive sponge found throughout New Zealand waters usually in exposed areas (Battershill & Bergquist, 1999a), and Polymastia croceus (Kelly-Borges & Bergquist, 1997), a common orange massive sponge. Both sponges contain metabolites with potential pharma- ceutical properties (Lill et al., 1995; National Cancer Institute, personal communication). The results described here are preliminary and part of a larger, ongoing experiment (October 1998). We focus here on the overall patterns of explant growth and survival between the three farming methods tested. Full results will be published after all relevant experiments are completed. MATERIALS AND METHODS For both L. brevis and P. croceus, we collected approximately forty sponges of similar size at 10-20m depth off the coast of Wellington (419215, 174°50E), situated at the southern end of the North Island of New Zealand. These sponges were cut, leaving approximately 30% of the original sponge intact to regenerate, Cut sponges left in situ had high survival and quickly healed. All collected sponges were cut under running seawater in a laboratory into cube-shaped explants, approximately 27cm’ in size and 16g in weight. All explants had at least one side uncut, with the pinacoderm intact. Three farming methods were tested for each species. Explants were: 1) placed in mesh; 2) attached directly to thick rope; 3) or had thin rope threaded through them (each method has several sub-methods, but full analysis at this stage is not yet possible given that the experiment is still in progress) (Fig. 1). Under method 2, each explant was firmly secured with cotton thread to an individual length of rope measuring 15x2.5cm. All explants in this method had their uncut side (with intact pinacoderm and oscules) facing outwards, away from the rope. Under method 3, to thread thin rope through explants, we carefully pushed a large needle, with rope attached, through each explant. Rope used in this treatment was 2-3mm thick. We used 40 explants of each species for each method. Explants were randomly selected and tied at intervals of 15cm to a rope back-line, and farmed at a depth of 12m. MEMOIRS OF THE QUEENSLAND MUSEUM FIG. 1. Schematic drawing of the 3 farming methods tested. A, explants placed in mesh; B, explants attached to rope; C, explants with thin rope threaded through them. We farmed L. brevis and P. croceus in Wellington Harbour from October 1997 to January 1998, and compared explant growth and survival. Growth was determined by wet-weighing the explants (to 0.1g) at the start and at the end of each experiment. We discovered that explants disturbed 30mins before weighing would expel all excess water, allowing us to weigh their true tissue weight. Comparisons between the different methods of farming on growth and survival in L. brevis and P. croceus were made using one-way ANOVA. RESULTS In both species growth rates were not significantly different between the three farming methods tested (F4570.24 and 0.04, N=68 and 110, P>0.05, for L. brevis and P. croceus, respectively). Conversely, survival of explants was significantly different between the methods used (Fa5731.28 and 23.79, N=120, P<0.001, respectively) (Figs 2B,D). Survival of both L. brevis and P. croceus farmed in mesh, under method 1, was excellent. Only one of the forty explants of L. brevis died and all P. croceus survived. The growth of L. brevis explants farmed in mesh was relatively good with an average weight gain of 1.2g over the 95 days of experimentation (Fig. 2A). Some of these replicates doubled their weight from 16g to over 32g during this period, a promising result given the brief time of experimentation. Many of these explants grew through the mesh, incorporating it into their tissue. In comparison, average growth of P. croceus farmed in mesh was poor, increasing only 0.1g in weight over 95 days (Fig. 2C). FARMING FOR BIOACTIVE SPONGE METABOLITES A, L. brevis growth average growth (g) (+/- 100 40 20 0 q T 1 explants placed in explants attached mesh to rope % survival rope threaded through explants 157 C. P. croceus growth 0.5 0.0 -0.5 average growth (g) (+/- S.E.) -1.5 -2.0 D. P. croceus survival 100 80 % survival rope threaded through explants explants placedin explants attached mesh to rope FIG. 2. Comparison in growth and survival of L. brevis and P. croceus between the three farming methods tested. Growth represents average explant weight gain or loss (+/- S.E.) over 95 days. Survival represents percent survival of the forty explants transplanted in each farming method. Neither species grew well on rope (method 2). On average, P. croceus lost 0.8g while L. brevis lost 1.1g over 95 days (Figs 2A,C). Under method 2, survival on rope was also poor. Polymastia croceus had 78% survival but only 1 of 40 ZL. brevis explants survived (Figs 2B,D). Under this farming method no explants of either species attached to the rope. The explant side, in contact with the rope, was similar in appearance (morphology and colour) to the other healed sides. We also observed many explants moving or growing away from the rope, ultimately becoming dislodged. Under method 3, when rope was threaded through explants, all but one P. croceus survived the 95 days experiment, whereas only 50% of L. brevis survived (Figs 2B,D). Average weight gain for both sponges was similar, approximately 0.3g (Figs 2A,C). Few explants of either species attached to the threaded rope. After 95 days, most explants had changed shape and were moving away from the rope. DISCUSSION The importance of choosing a suitable method of farming for sponge aquaculture is well demonstrated in this study. Survival of two species of sponges was greatly affected by the method used. Average growth of both species was generally low for all methods, most probably due to the short (95 day) period of experiment- ation, and factors inherent to each method mentioned below. The high survival of P. croceus and L. brevis farmed in mesh (method 1) may be explained by two factors: 1) Explants experienced the least initial damage, as they are simply placed in mesh. By comparison, explants grown under the other methods had greater disturbance, with rope either pushed through or squeezed around them, causing tissue damage and increased mortality; 2) Even in cases where mesh method is not ideal, explants were effectively trapped in mesh. We noticed many explants in the rope methods moving or growing away from the rope, ultimately becoming dislodged. For farming this is effectively the same as mortality (i.e. the sponge is lost). One disadvantage of the mesh farming method is a higher rate of fouling of mesh by sediment and sessile organisms, particularly bryozoans, reducing water flow and possibly influencing poor explant growth or even weight loss (Bakus, 1968; Duckworth et al., 1997). Restricted water movement due to fouling probably caused poor growth of P. croceus farmed in mesh. Unlike P. croceus, many explants of L. brevis quickly grew through and over the mesh, reflecting inherent species differences. This reduced the effect of fouling and, combined with low explant stress and damage, probably explains the better growth of L. brevis farmed in mesh. Harvesting sponges growing in mesh would involve cutting away tissue growth, leaving the explant behind to grow back through the mesh. Sponges farmed with rope threaded through them (method 3) were less effected by fouling because they were directly exposed to water. Whereas this may have promoted growth, mortality may have increased because of increased tissue damage. It is likely that increased tissue damage and rejection of the threaded rope caused poor survival of L. brevis. In contrast, P. croceus farmed with threaded rope survived well. Differences in growth and survival between the two sponges suggest that P. croceus is a hardier species and more amenable to different farming methods. However, given a suitable method, L. brevis achieved the best combination of growth and survival. Neither species attached well to the threaded rope, which probably caused reduced growth. Other studies have shown that only explants attached to their fastening wire or identification tag grew well (Verdenal & Vacelet, 1990). The ability of sponges to change shape (Bond & Harris, 1988; Bond, 1992) allows them to move away from unpleasant conditions and can result in loss of explants and low overall survival. This farming method will not succeed unless a rope material is found to which explants will attach. We are currently investigating this, testing explant growth, survival and attachment on threaded rope made of different natural and artificial materials. It is unlikely that this farming method will be suitable in exposed areas where high water movement can easily tear sponges away from the rope. Many studies have shown that sponges will attach well to a wide variety of natural and artificial substrata (Cotte, 1908; Moore, 1908; Crawshay, 1939; Wulff, 1984, 1985, Barthel & Theede, 1986; Bond & Harris, 1988; Rosell & Uriz, 1992). Unfortunately, both species of sponge in our study failed to attach to any of the ropes tested, perhaps a result of high substrate MEMOIRS OF THE QUEENSLAND MUSEUM selectivity shown by some sponges (Battershill & Bergquist, 1999b). Differences in growth and survival observed in the two species, L. brevis and P. croceus, probably point to inherent differences in sponge species ability to be successfully farmed. Thus, the findings of this study do not preclude the possibility of farming other New Zealand sponge species on rope. It may be possible to modify this method of farming to improve sponge attachment. For example, Battershill & Bergquist (1999b) discovered that P. croceus settles preferentially on rock chips, and it may be possible to incorporated these into the warp of a rope to promote explant attachment. Various types of rope substrate should also be tested. Many factors have to be considered in the development of a method or on-growing structure suitable for farming sponges for metabolite production. These include cost, bio-fouling, harvesting procedures, explant growth and survival, and metabolite yield. The findings of this study, which concentrated on explant growth and survival using three farming methods, will help develop a suitable on-growing structure for farming massive sponges, such as P. croceus and L. brevis. ACKNOWLEDGEMENTS We would like to thank the many divers that helped in this study, in particular Chris Woods, Pete Notman and Phil James. LITERATURE CITED BAKUS, G.J. 1968. Sedimentation and benthic invertebrates of Fanning Island, Central Pacific. Marine Geology 6: 45-51. BARTHEL, D. & THEEDE, H. 1986. A new method for the culture of marine sponges and its application for experimental studies. Ophelia 25: 75-82. BATTERSHILL, C.N. & BERGQUIST, P.R. 1990. The influence of storms on asexual reproduction, recruitment, and survival of sponges. Pp. 397-403. In Rüztler, K. (ed.) New perspectives in sponge biology. (Smithsonian Institution Press: Washington D.C.). 19992, The Porifera. In Cook, S.D.C. (ed.), New Zealand Coastal Invertebrates. Vol. 1. (Canterbury University Press: Christchurch) (in press). 1999b. Novel asexual reproduction in the sponge Polymastia croceus: Can sponge buds select settlement sites? Marine Biology. (in press). BERGQUIST, P.R. & TIZARD, C.A. 1969. Sponge industry, Pp. 665-670. In Firth, F.E. (ed.) The FARMING FOR BIOACTIVE SPONGE METABOLITES Encyclopedia of Marine Resources. (Van Nostrand Reinhold Company: New York). BOND, C. 1992. Continuous cell movements rearrange anatomical structures in intact sponges. Journal of Experimental Zoology 263: 284-302. BOND, C. & HARRIS, A.K. 1988. Locomotion of sponges and its physical mechanism. Journal of Experimental Zoology 246: 271-284. CRAWSHAY, L.R. 1939. Studies in the market sponges 1. Growth from the planted cutting. Journal Marine Biological Association of the United Kingdom 23: 553-574. COTTE, J. 1908. Sponge culture. Bulletin of the Bureau of Fisheries 28: 567-614. DUCKWORTH, A.R., BATTERSHILL, C.N. & BERGQUIST, P.R. 1997. Influence of explant procedures and environmental factors on culture success of three sponges. Aquaculture 156: 251-267 KELLY-BORGES, M. & BERGQUIST P.R. 1997. Revision of Southwest Pacific Polymastiidae (Porifera: Demospongiae: Hadromerida) with descriptions of new species of Polymastia Bowerbank, Tylexocladus Topsent, and Acanthopolymastia gen. nov. from New Zealand and the Norfolk Ridge, New Caledonia. New Zealand Journal of Marine and Freshwater Research 31: 367-402. LILL, R.E., MAJOR, D.A., BLUNT, J.W., MUNRO, M.H.G., BATTERSHILL, C.N., MCLEAN, M.G. & BAXTER, R.L. 1995. Studies on the biosynthesis of Discorhabdin B in the New 159 Zealand sponge Latrunculia sp. B. Journal of Natural Products 58: 306-311. MOORE, H.F. 1908. A practical method of sponge culture. Bulletin of the Bureau of Fisheries 28: 545-585. OSINGA, R., TRAMPER, J. & WIJFFELS, R.H. 1998. Cultivation of marine sponges for metabolite production; process engineering aspects. Trends in Biotechnology 16:130-134. RIDLEY, S.O. & DENDY, A. 1886. Preliminary report on the Monaxonida collected by H.M.S. “Challenger”. Part II. Annals and Magazine of Natural History 18: 470-493. ROSELL, D. & URIZ, M.J. 1992. Do associated zooxanthellae and the nature of the substratum affect survival, attachment and growth of Cliona viridis (Porifera: Hadromerida)? An experimental approach. Marine Biology 114: 503-507. SHIMIZU, Y. 1995, Extracting drugs from the sea. Maritimes University of Rhode Island, Marine Programs 38: 13-16. STORR, J.F. 1964. Ecology of the Gulf of Mexico sponges and its relation to the fishery. Special Scientific Report Fisheries 466: 1-73. VERDENAL, B. & VACELET, J. 1990. Sponge culture on vertical ropes in the Northwestern Mediterranean Sea. Pp. 416-453. In Rüztler, K. (ed.) New perspectives in sponge biology. (Smithsonian Institution Press: Washington DC). WULFF, J.L. 1984. Sponge-mediated coral reef growth and rejuvenation. Coral Reefs 3: 157-163. 1985. Dispersal and survival of fragments of coral reef sponges. Proceedings of the 5th International Coral Reef Congress 2: 119-124. 160 MEMOIRS OF THE QUEENSLAND MUSEUM POLLY WANT A SPONGE? FIELD EXAMINATION OF SPONGIVORY BY CARIBBEAN PARROTFISHES IN REEF AND MANGROVE HABITATS. Memoirs of the Queensland Museum 44: 160. 1999:- Caribbean sponge species such as Xestospongia muta frequently display linear grazing scars that appear to have been made by parrotfishes, yet there are few scientific reports of parrotfish spongivory. We used a video camera to monitor 40 specimens of X. muta for a minimum of 0.5 hr/sponge to determine the frequency of parrotfish bites on this species. Ten hours of taping captured 45 bites on normally coloured sponges, and 527 bites on four bleached sponges. Also, the guts from parrotfishes collected in mangrove and reef habitats were digested in nitric acid and analysed for spicule content. Parrotfishes collected in the mangroves (Sparisoma aurofrenatum, Scarus croicensis, and Sc. taeniopterus) had a significantly greater mass of spicules in their guts than did parrotfishes collected on the reef (Sp. aurofrenatum, Sp. viride, Sp. chrysopterum, Sc. vetula, Sc. coelestinus, and Sc. taeniopterus). Up to 148mg of spicules were present in the guts of mangrove parrotfishes. The spicules of Geodia gibberosa, a sponge that is common in the mangroves but rare in exposed locations on the reef, were abundant in the gut samples. Our results suggest that some sponge species are palatable not only to specialist predators such as sea turtles and angelfishes, but also to species that are not usually recognised as sponge predators. (J Porifera, spongivory, parrotfishes, Xestospongia muta, Geodia gibberosa, Sparisoma spp., Scarus spp., spicules, ecology, predation. Matthew J. Dunlap (email: dunlapm@hotmail.com) & Joseph R. Pawlik, Department of Biological Sciences, University of North Carolina-Wilmington, 601 South College Road, Wilmington NC, 28403-3297 USA; I June 1998. DEVELOPMENT OF HALISARCA DUJARDINI JOHNSTON 1842 (PORIFERA, CERACTIN- OMORPHA: HALISARCIDA) FROM EGG TO FREE LARVA. Memoirs of the Queensland Museum 44; 160. 1999:- Embryonic development in the sexual viviparous sponge Halisarca dujardini from the White Sea (Arctic) shallow water was studied. Complete, equal, asynchronal cleavage is characterised with variability of analogous developmental stages and the lack of the strictly determined clevage spindles position, The cytoplasm is filled with numerous yolky granules with heterogenic contents, At the 16-24 cell-stage a small cavity is formed. Blastomeres and the embryo polarity are not expressed. Large nuclei containing pronucleolar bodies are situated at the central parts of the cells. From the 16-24 cell- stage, true nucleolus formation starts. The polarisation of blastomeres is expressed by the distal movement of nuclei and changes in cell form. Cleavage furrow planes obtain the similar radial pattern forming roundish coelo- blastula 130-170um in diameter with the small cavity restricted with long wedge-shaped cells. The internal layer of the larva is formed at the 100-130 cell-stage owing to the individual cells’ apolar migration out of the blastula walls. At the same time flagella are formed on the cells’ apical surfaces, yolk granules being concentrated basally. Internal cells proliferate actively, differentiating into nucleolated amoebocytes, granular cells and collencytes. The larvae are is disphaerula it consists of two flagellated sphaerae external and internal. The disphaerula is completely covered with flagella. Flagellated cells are less numerous at the posterior pole. Flagellated epithelial cells are wedge-shaped. At their apical parts they contain a drop-like nucleus with nucleolus and a flagellum embedded into a pocket-like cytoplasmic invagination. The basal 2/3 of the cell volume is filled with numerous yolk granules. Flagellated cells are connected at their apical end by outgrowths of the plasma membrane embedded into similar invaginations of the neighbouring membrane. Posterior flagellated cells are trapeziform or rectangular, and contain numerous yolk granules. The nuclei are roundish, with large nucleoli. The internal sphaera is formed by invagination of lateral cells. These sphaera are formed by a layer of cylindrical cells that have flagella inside the cavity. Their piriform nuclei contain nucleoli, and there are yolk granules in the cytoplasm. There are no specialized cell contacts between blastomeres and larval cells. The spiral symbiotic bacteria are present in the central part of the larva and in intercellular spaces. Some peculiarities of H. dujardini embryogenesis are unique among Ceractinomorpha and are a matter of principle for comparative embryological studies of Porifera. They are: 1) total equal asynchronic cleavage; 2) equal, apolar coeloblastula with a small cavity; 3) unexpressed polarity of blastomeres; 4) subsequent of the same type radial cleavage leading to the cell polarisation and coeloblastula formation; 5) formation of an internal cell mass in the embryos by multipolar cell ingression at the 100-130 cell-stage; 6) development of special larva disphaerula; 7) formation of internal sphaera by invagination. All the features mentioned can serve as additional arguments for separation of the Halisarcida as an order (Bergquist, 1996). CJ Porifera, Halisarca dujardini, embryology, cleavage, larva, cells, ultrastructure. Literature cited. BERGQUIST, P.R. 1996. The marine fauna of New Zealand: Porifera: Demospongiae. Part 5. Dendroceratida and Halisarcida. New Zealand Oceanographic Institute Memoir 107: 1-53. Alexander. V. Ereskovsky (email: ERES@sn.puru) & Elizaveta L. Gonobobleva, Biology Facultv of Saint-Petersburg State University, Universitetskaya nab. 7/9, Saint-Petersburg, 199034, Russia; 1 June 1998. A PRELIMINARY ASSESSMENT OF 'SPACU WARS’ AS A DETERMINING FACTOR IN THE PRODUCTION OF NOVEL, BIOACTIVE INDOLES BY JOTROCHOTA SP. ELIZABETH A. EVANS-ILLIDGE, DAVID J. BOURNE, CARSTEN W- W. WOLFF & IOANA M. VASILESCU Evans-Illidge, E.A., Bourne, D.J., Wolff, C. W. W. & Vasilescu, LM. 1999 (6 30: A prelimi- nary assessment of ‘space wars’ as a determining factor in the production of novel bioactive indoles by lotrochota sp. Memoirs of the Queensland Museum 44: 161-166. Brisbane, ISSN 0079-8835, lotrochota sp. from Salamander Reef, North Queensland, has yielded a plethora of at least ten bioactive indoles including mono-, di- and nou-brominated variants. Metabolite composition varies within and between individual sponges, and competition for hard substrate was suspected as a determining factor in this variability. To provide a preliminary assessment of this, profiles of six identified indoles were compared between tissue samples categorised according to neighbour contact and growth thickness. Five of these compounds contained either two indole moieties (indolyl) or one indole and one benzene ring (benzoyl), The sixth indole, by virtue ofits structure, was identified as the putative precursor of the other compounds, There were no significant differences between tissue category and abundance of either indolyl or benzoyl product, or their putative precursor, However, two predominant populations of metabolites were identified. Diminished precursor and increased indoly! product occurred in tissue from sponge edges with direct neighbour contact and thick fleshy projections, This relationship was not absolute, and some samples from these tissue categories contained increased precursor and diminished indolyl product. Tissue from thin central sponge areas and edge samples without direct neighbour contact exclusively contained chemistry in the latter group. The quantity of benzoy! product remained constant between tissue categories. These results neither clearly support nor discount the potential role of space competition in determining production of these compounds. The issues involved are more complex than those examined here, and courses for further investigation are suggested. O Porifera, chemical ecology, indole, bioactivity, competition. E. A. Evans-Illidge (email: ¢.evansillidze@aims.gov.un), D. J. Bourne & C. W. W. Wolff. Australian Institute of Marine Science, PMB 3, Townsville, 4810, Queensland, Australia; | M. Vasilescu (ioana.vasilescu@jeu.edu.au), Chemistry Depariment, James Cook University, 4811, Queensland, Australia; 13 May 1999, Porifera continue to be the most prolific source of marine derived bioactive compounds in pub- lished literature (Marmlit, 1998). Numerous authors have sought to attribute sponge second- ary metabolites to a role in the source organism. and many have supported correlations between metabolite variability and environmental param- eters such as depth, UV light. and chemical defence (Thompson et al., 1987; Kreuter et al., 1992: Pawlik, 1993; Pawlik et al., 1995; Swearingen & Pawlick, 1995). This report presents some preliminary findings on the possible tole of competition for space in vanability of some novel indoles produced by a north Queensland sponge from the genus Jotrochota. A sample of this species of /ofrochota, a black thin enerusting sponge with oecasional thick vertical fleshy projections, was initially collected from Salamander Reef, North Queensland in 1988. The sample was part of the Australian Institute of Marine Science hiodiversity collection for natural products screening, and was initially erroneously assigned to the genus /rcinia. In 1995, a crude extract of this sponge was found to be highly active in a neuronal nitrous oxide synthase inhibition bioassay (authors? unpublished data). Initial fractionation yielded a plethora of novel indoles including mono-, di- and non-brominated variants. Seven of these compounds have been isolated and identified to date (Bowden et al., 1998: this report; and authors” unpublished data), but several still await further attention. The sponge was recollected in 1996 in order to provide sufficient material for follow-up bioassay and structure elucidation. While both samples contained a similar range of novel indoles, the relative abundance of each compound varied substantially between samples, An understanding of the cause of this variability will become important if one of these TABLE 1. Characteristics (HPLC, structure and mass) of compounds examined in this study. Key: A= Brominated species showing two isotopes (79Br and 81Br); B= Nonbrominated species. f HPLC ' Peak | Figure | chromatophore Structural Ion species no. no. class [M-1] 280nm | 360nm 7 4A Yes No Precursor | 306.9/308.9^ 9 4B Yes Yes Indolyl 402.1? 14 4C Yes Yes Indolyl 356.1? 16 4E Yes Yes Benzoyl |411.0/413.0^ 19 4D Yes Yes Indolyl — | 434.0/436.0 ^ 21 4F Yes Yes Benzoyl | 395.0/397.0% compounds progresses to become a new drug candidate and there is a need to optimise its yield through manipulative culture or selective recoll- ection. Field observations indicated that comp- etition for hard substrate at Salamander Reef was fierce, yet Jotrochota sp. remained abundant. Hence ‘Space Wars’ was suspected to be a potential controlling factor in metabolite variability in this sponge. This study aims to provide a preliminary assessment of variability in the production by Totrochota sp. of six bioactive novel indoles, with respect to direct neighbour contact and tissue thickness. It also aims to create a basis for further work to develop an understanding of factors that determine indole variability in this sponge. MATERIALS AND METHODS Sponge tissue samples were collected from Salamander Reef, 19?10.91'S, 147°03.76’E, a small rocky inshore reef off Cape Cleveland near Townsville, North Queensland, in March 1998. 29 samples from 8 individual sponges where collected from 10-15m depth. Small biopsies of sponge tissue (approx 1cm?) were taken and assigned one of four categories according to the degree of direct neighbour contact and tissue thickness, as follows: 1) Edge Interaction (edge of sponge in direct contact with a neighbouring organism); 2) Edge No Inter- action (edge of sponge without direct neighbour MEMOIRS OF THE QUEENSLAND MUSEUM contact); 3) Centre Thin (centre of sponge, no fleshy projection); and 4) Centre Thick (thick fleshy projection in the centre of sponge), Tissue samples were freeze dried, and 80mg (dry weight) of tissue was extracted in 5ml of a solvent solution made up of equal parts dichloromethane and methanol, with sonication for 80mins. Extract was carefully decanted into clean vials, dried, then redissolved in 1ml meth- anol for High Performance Liquid Chromatography (HPLC) analysis, Where less than 80mg tissue was available for extraction, solvent quantities were adjusted to achieve the same extraction concentration. Extracts were analysed for brominated indoles of interest using HPLC with an Alltima C18 column ( 250x4.6mm, Alltech Australia). A linear gradient from 60-100% of methanol in water was used. UV spectra were recorded with a Shimadzu MXA diode array and absorbance monitored at 280 and 360nm. Major components of HPLC peaks were then characterised by negative-ion electrospray mass-spectrometry to confirm that common compounds could be identified between different sponge extracts. Areas under HPLC peaks were then used as a measure of relative amount of each fraction, for comparison between samples. These estimates were not suitable to compare quantities of different fractions within individuals, as a full analysis of extinction-coefficients of each compound was not undertaken. One-way analyses of variance (ANOVA) with a=0.05 was used to compare HPLC peak areas of each fraction group of interest, between sample categories. Four samples from each of the four tissue categories were selected from the available sample pool. Samples were independently selected in this way for analysis of each fraction group of interest. Where a result was non-significant, the detectable effect size (standard deviation between means) with 80% power was calculated according to Cohen (1977) and expressed as a percentage of the overall mean. TABLE 2. Power analysis results for non-significant ANOVAs on HPLC peak areas between tissue categories. Grand Mean Mean Square Within Standard Deviation between detectably different means Compound Type bi : G f NOVA . . (arbitrary units) | Groups (from A ) (Arbitrary units) (% of grand mean) Precursor 42621802 2.37143E+14 14937464.6 35 Indolyl Product 3707935 3.94114E+13 6089514.452 164 Benzoyl Product 14993786 1,08893E+13 3200897.12 21 “SPACE WARS” IN JOTROCHOTA O HN E —— HÓ A B N H N o X HN as _—— H¿CO D er N R=R¡=H R O HN — — HCO E R=H Br N F R=0H FIG. 1. Structure ofthe six indoles considered in this study: A, Putative precursor (HPLC Peak 7); B—D, Indolyl product (HPLC Peaks 9,14,19); E — F, Benzoyl product (HPLC peaks 16,21). * Assignment of substituent posi- tion not established RESULTS Six compounds of interest were separated using the HPLC system described above. Fourier Transform Mass Spectrometry and NMR con- firmed that these indoles were the major components of the peaks listed and characterised in Table | and depicted by the structures shown in Figure |. These data suggest that the low molecular weight indole in peak 7 (Fig. 1A) is the precursor of the more complex compounds in the other five peaks. These ‘product’ indoles fall into two structural classes based on whether the addi- tion to the peak 7 core contains another indole (Fig. 1B-D) (= indolyl product) or a benzene group (Fig. 1E-F) (= benzoyl product). ANOVAs on HPLC peak areas for precursor (peak 7), indolyl product (sum of peaks 9, 14 and 19) and benzoyl product (sum of peaks 16 and 21) found no significant difference in the amount of precursor or product present in tissue samples from the different tissue categories, with a=0.05. However, with 80% power, these non-significant tests were only capable of detecting differences between groups with a standard deviation between their means of 35% of the overall mean (precursor), 164% of the overall mean (indolyl product) and 21% of the overall mean (benzoyl! product) (Table 9 On the basis of HPLC, the 29 tissue sample extracts fell into two distinct groups. Figure 2 depicts a typical chromatograph of each group. When compared to Group 1, Group 2 contained more of peak 7 (Precursor) and less of peaks 9, 14 and 19 (indolyl product), while the quantity of peaks 16 and 21 (benzoyl product) were fairly consistent between the two groups. These rela- tionships are summarised and quantified further in Figure 3. Table 3 presents group membership with respect to tissue category. Only four tissue samples from three individuals had Group 1 chemistry (more indolyl product, less precursor). Two of these came from edges of direct interaction, and two 164 280n ---- 360n Absorbance (mAbs) Nt S 8 Time (mins) FIG. 2. Typical chromatograms of group 1 and group 2 samples. Peaks numbered for compounds considered in this study. came from thick fleshy projections. While there were other samples from these categories with Group 2 chemistry (less indolyl product and more precursor), samples from either central sponge tissue or edge sites without direct interaction exclusively belonged to Group 2. Most samples analysed (25 out of 29) belonged to Group 2, and only two of the ten individuals examined contributed samples to both groups. MEMOIRS OF THE QUEENSLAND MUSEUM TABLE 3: Tissue categories sampled with respect to chemistry group membership. Growth Type Group 1 Group 2 Edge Interaction 2 10 Centre Thick 2 2 Edge No Interaction 0 6 Centre Thin 0 7 DISCUSSION This work does not clearly support a direct relationship between neighbour interaction and indole chemistry in Jotrochota sp.. However, significance tests had moderate to low resolution at 80% power, and aspects of the distribution of samples containing Group 1 and Group 2 indole chemistry are consistent with an hypothesis of space competition influence. These are discussed below with respect to appropriate future direc- tions for work in this area, and are not represented as definitive conclusions. Morphological strategies are important to sessile benthic invertebrates in their struggle for substrate (Jackson, 1979; Hoppe, 1988; Vicente, 1990; Becerro et al., 1994). Becerro et al. (1994) suggested that another thin encrusting sponge, Crambe crambe, employed directional growth to either avoid stronger or confront weaker space competitors. Jackson (1979) suggested that vertical growth is another non-confrontational strategy in space competition. Whereas Jotrochota sp. is generally a thin encrusting sponge, A B C 6 13 in Legend: © 2 7 4 12 2 Group 1 D =e 6 6 Oo e Group 2 0 0 Group Membership FIG. 3. Relative amounts (area under HPLC peak) of A. Precursor, B. Indolyl product and C. Benzoy! product in the two sample groups. “SPACE WARS’ IN /OTROCHOTA 165 individuals typically sport several, thick, fleshy, vertical projections. A possible interpretation of this growth form is that the sponge utilises both confrontational and non-confrontational strategies to compete for hard substrate, whereas vertical growth in this otherwise thinly encrusting species may be a product of, or avoidance from encounters with superior space competitors at their outer growth margins. It is therefore possible that samples containing Group 1 chemistry (i.e. more indolyl product, less precursor) had assumed a space competition strategy, either through direct confrontation at their margins, or non-confrontational vertical growth. However, this trend was not consistent, where both samples at the margins of neighbour contact (1.e. confrontational samples), and thick fleshy projections (i.e. non-confrontational samples), were included in chemistry Group 2. Further investigations into patterns of indole chemistry, which address growth form with respect to different neighbour interactions and the nature of these interactions, are essential to develop appropriate hypotheses. Allelochemical interactions do not necessarily require direct contact between two individuals (Porter & Targett, 1988, Turon et al., 1996), and any non-contact interaction would be dependant on water flow. Thus, future work should also account for contact, distance and direction data (the latter with respect to currents). This species is amenable to transplantation (authors! unpub- lished data) and thus a candidate for controlled manipulative experimentation. Patterns of variability in the other indole compounds known to occur in this Jotrochoia sp. (authors! unpublished data), may also be import- ant in understanding total metabolite variability in this species. More than 40 additional compounds which can be identified tentatively as indoles on the basis of mass spectrometry evid- ence, await characterisation, structural elucid- ation, and quantification. Also, the putative precursor-product relationship proposed in this work needs to be confirmed before any strong assertions about the invocation of a secondary metabolite from its precursor can be attributed to ecological factors. ACKNOWLEDGEMENTS This work was undertaken within the Marine Bioproducts Project at the Australian Institute of Marine Science with funding support from AMRAD Pharmaceuticals. Screening was undertaken within a program led by Dr Lyndon Llewellyn. Fieldwork was conducted aboard the AIMS Research vessel ‘Titan’, and the contrib- ution of skipper Mr David James is acknowledged. Dr Diane Tapiolas assisted with Marinlit searches. The authors are also grateful to Mr John Kennedy, Queensland Museum, for identification of the sponge. This is contribution number 958 from the Australian Institute of Marine Science. LITERATURE CITED BECERRO, M.A., URIZ, M.J. & TURON, X. 1994. Trends in space occupation by the encrusting sponge Crambe crambe: variation in shape as a function of size and environment. Marine Biology 121: 301-307. BOWDEN, I.M., BOURNE, D. & MURPHY, P.T. 1998, New Metabolites from the marine sponge Ircinia sp.. P.40. In 9th International Symposium on Marine Natural Products Symposium Abstract Booklet. (James Cook University: Townsville). COHEN, J. 1977. Statistical Power Analysis for the Behavioural Sciences. (Academic Press: New York). GREEN, G., GOMEZ, P. & BAKUS, G.J. 1990. Anti- microbial and Ichthyotoxic Properties of Marine Sponges from Mexican Waters. Pp. 109-114. In Rützler, K. (ed.) New Perspectives in Sponge Biology. (Smithsonian Institution Press: Washington D.C.). HOPPE, W.F. 1988. Growth, regeneration and pred- ation in three species of large coral reef sponges. Marine Ecology Progress Series 50: 117-125. KREUTER, M.H., ROBITZKI, A., CHANG, $, STEFFEN, R., MICHAELIS, M., KLJAJIC, Z., BACHMANN, M., SCHROEDER, H.C. & MULLER, W.E.G. 1992. Production of the cyto- static agent aeroplysinin by the sponge Verongia aerophoba in in vitro culture. Comparative Bio- chemical Physiology 101C: 183-187. JACKSON, J.B.C. 1979. Morphological strategies of sessile animals. Pp. 499-555. In Larwood, G. & Rosen, B.R, (eds) Biology and Systematics of Colonial Organisms. (Academic Press: London). JUNG, J.H., SIM, C.J. & LEE, C. 1995. Cytotoxic compounds from a two-sponge association. Journal of Natural Products 58: 1722-1726. MARINLIT, 1998. A database on the literature on marine natural products. (Marine Chemistry Group, Department of Chemistry: Universityof Canterbury, New Zealand). MURICY, G., HAJDU, E., ARAUJO, F.V. & HAGLER, A.N. 1993. Antimicrobial activity of Southwestern Atlantic shallow-water marine sponges (Porifera). Scientia Marina 57(4): 427-432. PAWLIK, J.R. 1993. Marine Invertebrate Chemical Defenses. Chemical Review 93: 1911—1922. PAWLIK, J.R, CHANAS, B., TOONEN, RJ. & FENICAL, W. 1995. Defenses of Caribbean 166 sponges against predatory reef fish. I. Chemical deterrency. Marine Ecology Progress Series 127: 183-194. PORTER, J.W., & TARGETT, N.M. 1988. Allelochemical interactions between sponges and corals. Biological Bulletin 175: 230-239. SWEARINGEN, D.C. & PAWLIK, J.R. 1995. Intraspecific variability of chemical defense in a sponge. In Grassle J.P., Kelsey, A., Oates, E. & Snelgrove, P.V. (eds) 23" Benthic Ecology Meeting, 1995. MEMOIRS OF THE QUEENSLAND MUSEUM THOMPSON, J.E., MURPHY, P.T., BERQUIST, P.R. & EVANS, E.A. 1987. Environmentally induced variation in diterpene composition of the marine sponge Rhopaloeides odorabile ng, nsp.. Biochemical Systematic Ecology 15: 595-606. TURON, X., BECERRO, M.A., URIZ, M.J.,8 LLOPIS, J. 1996. Small-scale association measures in epibenthic communities as a clue for allelo- chemical interactions. Oecologia 108: 351—360. VICENTE, V.P. 1990. Overgrowth activity by the encrusting sponge Chondrilla nucula on a coral reef in Puerto Rico. Pp. 109-114. In Riitzler, K. (ed.) New Perspectives in Sponge Biology (Smith- sonian Institution Press: Washington D.C.) LOCALISATION OF BIOACTIVE METABOLITES IN MARINE SPONGES D. JOHN FAULKNER, MARY KAY HARPER, CHRISTINE E. SALOMON AND ERIC W. SCHMIDT Faulkner, D.J., Harper, M.K., Salomon, C.E. & Schmidt, E.W. 1999 06 30: Localisation of bioactive metabolites in marine sponges. Memoirs of the Queensland Museum 44: 167-173. Brisbane. ISSN 0079-8835. Marine natural product chemists have often proposed that bioactive sponge metabolites are produced by symbiotic micro-organisms. This paper discusses the rationale for these proposals, reviews the strengths and weaknesses of methods that are available to test such hypotheses and reports some experimental studies. The conclusion reached from the research to date is that it is too early to make generalisations concerning either the role of symbionts in the biosynthesis of sponge metabolites or even the best techniques for studying the cellular localisation of bioactive metabolites. CJ Porifera, bioactive metabolites, cyanobacteria, filamentous eubacteria, symbiosis, Aplysina fistularis, Dysidea herbacea, Theonella swinhoei, Oceanapia sagittaria, Jaspis splendens. * D. John Faulkner (email: jfaulkner@ucsd.edu), Mary Kay Harper, Christine E. Salomon & Eric W. Schmidt, Scripps Institution of Oceanography, University of California at San Diego, La Jolla, CA, 92093-0212, USA; 22 December 1998. Sponges are exceptionally good synthetic chemists. They can make chemicals of extreme toxicity and/or deterrent value that have undoubtedly contributed to their survival over the ages. But they may not always have acted alone. We now know that some sponges harbor populations of symbiotic micro-organisms that produce the chemicals thought to defend the sponge from competition or predation. However, it is clear that this situation is less common than the marine natural products literature would have us believe. This paper reviews the methods used to determine the cellular location of natural products in sponges and presents some recent results from our laboratory that either confirm or deny the production of ‘sponge metabolites’ by symbiotic microbes. SYMBIOSIS AS SEEN FROM THE VIEWPOINT OF CHEMISTRY. The history of natural products chemistry has been driven by the use of natural products to treat diseases. First came an interest in plant products such as digitalis and morphine, but this was superseded in the second half of this century by the discovery of a plethora of immensely important antibiotics and other drugs obtained by the fermentation of microbes. Chemists became indoctrinated with the concept that micro-organisms could provide the needs of the pharmaceutical industry, which for a long period of time was not far from the truth. Then came the discovery that marine organisms could provide many new classes of natural products that incorporated new and unexpected structural motifs. Within this group, sponges have provided not only the best source of novel compounds but also the greatest occurrence of potential pharmaceuticals (Faulkner, 1998, and references therein). However, when chemists compared the structures of sponge metabolites with those of compounds in the literature, they found many structures that were very similar to those of microbial metabolites. When chemists saw scanning electron micrographs of sponges that contained large numbers of micro-organisms, they felt justified in proposing that compounds resembling microbial metabolites were in fact of microbial origin. Furthermore, when closely related or identical compounds were found in different phyla and there was no evidence of transmission of the chemicals through the food chain, they proposed that these compounds might be produced by the same or similar micro-organisms endemic to hosts of different phyla. These hypotheses set the stage for a careful investigation of the role of symbiotic microbes in the production of ‘sponge metabolites’. LOCALIZATION OF SPONGE METABOLITES. There are two basic strategies for determining the location of specific metabolites in sponges: detection of compounds using microscopy, or cell separation followed by chemical analysis of each cell fraction. The strategy selected generally depends on the molecular properties of the compound to be investigated. Compounds that contain heavy elements such as bromine or iodine 168 Br I No eN (CHa)n (A) n = 4, aerothionin n = 5, homoaerothionin FIG. 1. Tetrabrominated metabolites. A, aerothionin. B, homoaerothionin. can be detected by using an energy dispersive spectroscopy (EDS) detector on a scanning electron microscope (SEM) or a scanning transmission electron microscope (STEM). In theory, one could use the same technique to determine the location of compounds containing chlorine or sulfur but, in practice, the levels of chloride and sulfate ions in seawater preclude its use with marine specimens. Fluorescent compounds can be conveniently detected by fluorescence microscopy and by laser scanning confocal microscopy, but this technique is susceptable to problems caused by background fluorescence due to photosynthetic pigments and general autofluorescence of cells. The method of immunostaining using polyclonal antibodies to bind to a specific compound is common in cellular biology but prior to a report at this symposium (Ilan, 1998) and one other recent paper (Perry et al., 1998) had not been applied to study sponge metabolites. Finally, there is the possibility that specific compounds may be detected in cell preparations using secondary ion mass spectrometry in conjunction with tandem mass spectrometry. The latter two methods, both of which can be fine-tuned to detect individual compounds, could offer considerable advantages over methods that rely on detecting a class of compounds. Cell separation methods take advantage of our ability to analyse the chemical content of fixed cells but suffer from the disadvantage that fractions containing only a single cell type may be difficult to prepare. Sponge tissues can be dissociated by enzymatic digestion or mechanical disruption in calcium-magnesium free seawater using squeezing, sieving, simple mincing, vigorous stirring, or even a juicer. The MEMOIRS OF THE QUEENSLAND MUSEUM dissociated cells can then be fixed, which stabilises the cells during the period between collection and analysis. It is a relatively simple matter to separate cyanobacteria using a cell sorter to distinguish fluoresecent from non-fluorescent cells but this method does not distinguish between sponge and eubacterial cells. Cell types can also be separated by density using either differential centrifugation or Ficoll or Percoll density gradients. It has been our experience that repeated centrifugation is required to produce fractions of reasonable purity and that the different sponge cell types are difficult to separate on the basis of density. Nonetheless, filamentous bacteria, mixed sponge cells and mixed unicellular bacterial cells can all be enriched to ca. 90% purity using centrifugation. To detect the compounds of interest, each cell fraction is then extracted individually and analyzed using two or more of the following techniques: mass spectrometry (MS), which can be combined with high performance liquid chromatography (HPLC) or gas chromatography (GC), HPLC using a diode array detector to measure the UV spectrum, and 'H NMR spectrometry. RESULTS AND DISCUSSION The tetrabrominated metabolites aerothionin (Fig. 1A) and homoaerothionin (Fig. 1B), which occur as a 10:1 mixture in a shallow-water specimen of Aplysina fistularis from La Jolla, were ideal candidates for study using energy dispersive spectroscopy because the molecules contain such a high concentration of bromine. Analysis of the STEM images using energy dispersive X-ray analysis revealed a 20-fold larger concentration of bromine in spherulous cells than in bacterial or other sponge cells, which were both at background levels. We therefore argued that the brominated metabolites (Fig. 1A-B) were produced and stored in spherulous cells (Thompson et al., 1983). There are two major chemotypes of Dysidea herbacea; one contains both sesquiterpenes and metabolites biosynthesised from polychlorinated amino acids, the other produces only polybrominated biphenyl ethers and lacks terpenes. Very significant populations of cyanobacteria are found in both chemotypes and in both cases, the cyanobacterium is considered to be Oscillatoria spongelliae. The fluorescent cyanobacterial cells were separated from all other non-fluorescent cells using a cell sorter and BIOACTIVE METABOLITE LOCALIZATION (B) spirodysin Br OH uy Br Br Br k^ (A)13-demethylisodysidenin (C) herbadysidolide FIG. 2. Metabolites from two major chemotypes of Dysidea herbacea. ^, 13-demethylisodysidenin. B, spirodysin. C, herbadysidolide. D, polybrominated biphenyl ether. the chemical content of each cell type was analysed by 'H NMR spectroscopy and GC-MS. In a specimen of D. herbacea from Heron Island, 13-demethylisodysidenin (Fig. 2A), a polychlorinated amino acid derivative, was extracted from the cyanobacterial cell fraction while the sesquiterpenes spirodysin (Fig. 2B) and herbadysidolide (Fig. 2C) were detected in the fraction that contained sponge and bacterial cells (Unson & Faulkner, 1993). Garson and coworkers recently separated the sponge cells using a Percoll density gradient and showed that the sesquiterpenes were located in a fraction containing archaeocytes and choanocytes (Flowers et al., 1998). A similar analysis of a specimen of D. herbacea from Palau revealed that the polybrominated biphenyl ether (Fig. 2D) was located not only in the cyanobacterial fraction but also as crystals situated just below the surface of the sponge (Unson et al., 1994). The sponge cells were considered to be devoid of brominated metabolites, although it was possible to detect a very low level of the poly- brominated biphenyl ether (Fig. 2D), which was consistent with the presence of a small number of cyanobacterial cells that remained as contaminants in the sponge cell fraction. Having shown that cyanobacteria are responsible for the halogenated chemicals in the two chemotypes of D. herbacea, there is now a need to analyse the 169 16S rRNA sequences of representative samples to determine whether the cyanobacteria represent two different strains of O. spongelliae or different cyano- bacterial species, which are two of several possible explan- ations of the chemical diversity. The diversity of chemistry assigned to Dysidea spp. may also provide a good rationale for a sponge taxonomist to re-examine the genus and particularly the chemists’ voucher specimens. The lithistid sponge Theonella swinhoei has provided chemists with an almost unequalled source of highly bioactive chemicals (Bewley & Faulkner, 1998). Our interest in this sponge was piqued by the structural similarity between swinholide A (Fig. 3A), which had previously been isolated from 7. swinhoei, and the cyanobacterial product scytophycin C (Fig. 3B) and by the fact that the cyclic peptides of 7. swinhoei, such as theopalauamide (Fig. 3C)(Schmidt et al., 1998), contain aromatic B-amino acids similar to those found in some cyanobacterial cyclic peptides (Ishibashi et al., 1986; Kitagawa et al., 1990; Bewley & Faulkner, 1998). This led to a suggestion that the prominent filamentous micro-organisms in T. swinhoei were cyanobacteria that produced both groups of compounds (e.g. Kobayashi & Ishibashi, 1993; Fusetani & Matsunaga, 1993). We had reason to suspect that this assumption was incorrect because the sponges were often found in caves, the filaments were found in the interior of the sponge, away from the light, and extracts of the endosomal tissue of the sponge did not appear to contain sufficient chlorophyll pigments. In a specimen of 7. swinhoei from Palau, there were unicellular cyanobacteria (Aphanocapsa feldmanni) in the ectosome, which was peeled away and examined separately. The ectosomal tissues were dissociated and the cyanobacteria were separated using differential centrifugation, but they did not contain the metabolites of interest. The endosomal tissues were dissociated, fixed in a mixture of formalin and glutaraldehyde in artificial seawater, and separated using (D) 170 OMe (A) swintiolide A MEMOIRS OF THE QUEENSLAND MUSEUM (B) scytophycin C aromatic B-amino acid (C) theopalauamide FIG. 3. Metabolites from the lithistid sponge Theonella swinhoei. A, swinholide A, which partially resembles scytophycin C. B, cyanobacterial product scytophycin C. C, theopalauamide. differential centrifugation into three fractions containing mixed sponge cells, a filamentous bacterium, and mixed unicellular bacteria. The cell fractions were thoroughly washed, then extracted to obtain crude extracts that were analysed by 'H NMR and HPLC. Theopalauamide (Fig. 3C) was found to be present in about 4% dry weight in the filaments, which were examined by TEM and found not to be cyanobacteria, since they lacked the thylakoid BIOACTIVE METABOLITE LOCALIZATION 171 Iz IZ (A) pyridoacridine NHCOEt (B) dercitamide VIG. 4. A, pyridoacridine skeleton. B. dercitamide. structures that house the photosynthetic apparatus of cyanobacteria (Bewley et al., 1996). We are currently characterising the eubacterial filaments using 168 rRNA analysis. Swinholide A (Fig. 3A) was extracted from the unicellular bacterial fraction, which contained many morphologically distinct bacteria. A recent re-examination of the 'H NMR spectrum of the unicellular bacterial fraction revealed the presence of the 4-methylene sterols that are typical of Theonella spp., but we need to reconfirm that result because sterols are not usually produced by cultured bacteria. Both the sponge cells and the cyanobacterium Aphanocapsa feldmanni appeared to be devaid of bioactive metabolites. The pyridoacridine alkaloids, which all possess (he same underlying tetracyclie aromatic ring system (Fig. 4A}, are examples of a class of metabolites that have been found in four different marine phyla, but predominantly in sponges and ascidians (Molinski, 1993). They have frequently been proposed to be metabolites of undesignated microbial populations that might occur as symbionts in the different phyla. We felt that there might be an alternative explanation based on the evolution of similar biosynthetic schemes in different phyla, in part because polyaromatic compounds are the most stable products that can arise from their presumed mode of biosynthesis (Steffan et al., 1993). Dercitamide (Fig. 4B) has been reported from both sponges and ascidians (Gunawardana et al., 1992; Carroll & Scheuer, 1990) — the latter authors referring to dercitamide as kuanoniamine C - and we have isolated it as the major metabolite of the sponge Oceanapia sagittaria (Salomon & Faulkner, 1996). Dercitamide is an interesting pigment that changes color from yellow in neutral or basic solution (pH > 7) to red in acidic solution (pH « 6) and has a fluorescence spectrum that is also pH dependent. Using a light microscope, one can observe a change in the color of the sponge issue when a section is acidified using trifluoroacetic acid vapor. A similar pH dependency was noted when sections were observed using fluorescence microscopy, but there was so much fluorescence from cells that were out-of-plane that it was impossible to clearly image the cells containing dercitamide. Examination of both thick sections and enriched cell fractions using a confocal microscope under both neutral and acidic pH conditions led to the conclusion that dercitamide was localised in sponge cells containing between ten and (wenty spherical inclusions, Transmission electron microscopy was employed to show that there were no intracellular bacteria thal could be responsible for the chemistry (Fig.6). The dercitamide-containing cells were characterised by TEM analysis, although they have not been classified as a partic- ilar type of sponge cell, This appears to be the first time that confocal microscopy has been employed to locate marine natural products on the basis of their autofluorescence. Research is in progress to determme the cellular location of pyridoacridine alkaloids in ascidians. Notevery study has resulted in an unambiguous localisation of metabolites. The cyclic depsipeptide jaspamide (Fig. 5) is a cytotoxic metabolite of Jaspis splendens (De Laubentels, 1954), referred to as Jaspis sp. in our earlier chemistry paper (Zabriskie et al., 1986), that has been proposed both as a chemotaxonomic marker and to be of microbial origin. It is interesting to note that jaspamide (Fig, 4) was also isolated from a completely different sponge, Auletta cf. constricta (Crews et al., 1994). The sponge was dissocialed in a juicer, a technique previously used successfully on 7. swinhoei, followed by HO Oy -O e NÍ Me O vy HM H Br jaspamide FIG. 5. Cyelic depsipeptide jaspamide from Jaspis splendens. FIG. 6. Micrograph of a dercitamide-containing inclusional sponge cell from Oceanapia sagittaria that shows the absence of intracellular symbionts. fixation and cell separation using differential centrifugation. Jaspamide was not detected in extracts of an unidentified extracellular ‘symbiont’ (Fig. 7) or associated micro- organisms, which represented a large proportion of the whole sponge biomass, but was isolated in nearly 4% yield from a fraction containing small (ca. 500nm) orange bodies and cellular debris. The identity of the orange bodies is uncertain, but we have evidence that the dissociation process may have ruptured the sponge cells with concomitant release of the orange bodies. Examination of newly acquired sponge material by light microscopy revealed the presence of numerous small orange inclusions within the sponge cells. We now believe that jaspamide (Fig. 5) is located within sponge cells and further research is in progress to test this hypothesis. CONCLUSIONS The major conclusion that we have reached during our studies of the role of symbionts in the production of sponge metabolites is that it is extremely dangerous to make any general statements about the sources of bioactive metabolites. In essence, each compound of interest requires an individual study to determine its source. The results that we and others have generated indicate that it is possible to detect specific compounds or classes of compounds in either symbiont or sponge cell fractions and that the concentrations of secondary metabolites in isolated cell fractions can be spectacularly high. However, it is often difficult to determine which sponge cell type produces the metabolite and it is nearly impossible to define a particular MEMOIRS OF THE QUEENSLAND MUSEUM FIG. 7. Micrograph of dissociated cells of Jaspis splendens. Arrows indicate the unidentified symbionts that do not contain jaspamide. unicellular bacterium as the source of a bioactive compound. The latter will undoubtedly require the culture of symbiotic micro-organisms, which is the goal of several research groups. In order to accomplish this goal, we have proposed a strategy that involves identification of the symbionts from their 16S rRNA sequence (Brantley et al., 1995) before attempting to culture them using media that are suitable for culturing their nearest relatives. While the ultimate goal is to culture symbionts that produce pharmacologically-active sponge metabolites in order to speed their pharmaceutical development, information gained from cellular localisation studies can also be useful in chemotaxonomy and the understanding of bio- synthetic pathways that may have influenced the evolution of symbioses within sponges. ACKNOWLEDGEMENTS The earlier studies reported above were taken from the thesis research of Janice E. Thompson, Mia Unson and Carole A. Bewley and have been reported in detail elsewhere. We thank the Republic of Palau and the State of Koror for permission to collect specimens and the Coral Reef Research Foundation, Koror, Palau for providing logistical support and laboratory facilities. We also thank Mary Garson for providing us with the opportunity to do research at Heron Island. This research program has been generously supported by the National Science Foundation (CHE 95-27064). BIOACTIVE METABOLITE LOCALIZATION LITERATURE CITED BEWLEY, C.A. & FAULKNER, D.J. 1998. Lithistid sponges: Star performers or hosts to the stars. Angewandte Chemie International Edition 37: 2162-2178. BEWLEY, C.A., HOLLAND, N.D. & FAULKNER, D.J. 1996. Two classes of metabolites from Theonella swinhoei are localized in distinct populations of bacterial symbionts. Experientia 52: 716-722. BRANTLEY, S.E., MOLINSKI, T.F., PRESTON, C.M. & DELONG, E.F. 1995. Brominated acetylenic fatty acids from Xestospongia sp., a marine sponge-bacterial association. Tetrahedron 51: 7667-7672. CARROLL, A.R. & SCHEUER, P.J. 1990. Kuanoni- amines A, B, C, and D: pentacyclic alkaloids from a tunicate and its prosobranch mollusk predator Chelynotus semperi. Journal of Organic Chemistry 55: 4426-4431. CREWS, P., FARIAS, J.J., EMRICH, R. & KEIFER, P.A. 1994. Milnamide A, an unusual cytotoxic tripeptide from the marine sponge Auletta cf. constricta. Journal of Organic Chemistry 59: 2932-2934. FAULKNER, D.J. 1998. Marine Natural Products. Natural Products Reports 15: 113-158. FLOWERS, A.E., GARSON, M.J., WEBB, R.I., DUMDEI, E.J. & CHARAN, R.D. 1998. Cellular origin of chlorinated diketopiperazines in the dictyoceratid sponge Dysidea herbacea (Keller). Cell & Tissue Research 292: 597-607. FUSETANI, N. & MATSUNAGA, S. 1993. Bioactive sponge peptides. Chemical Reviews 93: 1793-1806. GUNAWARDANA, G.P., KOEHN, F.E., LEE, A.Y., CLARDY, J., HE, H. & FAULKNER, D.J. 1992. Pyridoacridine alkaloids from deep water marine sponges of the family Pachastrellidae: Structure revision of dercitin and related compounds and correlation with the kuanoniamines. Journal of Organic Chemistry 57: 1523-1526. ILAN, M. 1998. Abstract. Negombata magnifica - A magnificent (chemical) pet. Memoirs of the Queensland Museum (this volume). ISHIBASHI, M., MOORE, R.E., PATTERSON, G.M.L., XU, C. & CLARDY, J. 1986. Scytophycins, cytotoxic and antimitotic agents from the cyanophyte — Scytonema pseudohofmanni. Journal of Organic Chemistry 51: 5300-5306. KITAGAWA, I., KOBAYASHI, M., KATORI, T., YAMASHITA, M., TANAKA, J., DOI, M. & ISHIDA, T. 1990. Absolute stereostructure of swinholide A, a potent cytotoxic macrolide from the Okinawan marine sponge Theonella swinhoei. Journal of the American Chemical Society 112: 3710-3712. KOBAYASHI, J. & ISHIBASHI, M. 1993 Bioactive metabolites of symbiotic marine microorganisms. Chemical Reviews 93: 1753-1769. MOLINSKI, T.F. 1993. Marine pyridoacridine alkaloids: structure, synthesis, and biological chemistry. Chemical Reviews 93: 1825-1838. PERRY, N.B., ELLIS, G., BLUNT, J.W., HAYSTEAD, T.AJ., LAKE, R.J., MUNRO, M.H.G. 1998. Okadaic acid in New Zealand sponges: detection by cytotoxicity, protein phosphatase inhibition and immunoassay techniques. Natural Product Letters 11: 305-312. SALOMON, C.E. & FAULKNER, D.J. 1996. Sagitol, a pyridoacridine alkaloid from the sponge Oceanapia sagittaria. Tetrahedron Letters 37: 9147-9148. SCHMIDT, E.W., BEWLEY, C.A. & FAULKNER, D.J. 1998. Theopalauamide, a bicyclic glycopeptide from filamentous bacterial symbionts of the lithistid sponge Theonella swinhoei from Palau and Mozambique. Journal of Organic Chemistry 63: 1254-1258. p STEFFAN, B., BRIX, K. & PUTZ, W. 1993. Biosynthesis of shermilamine B. Tetrahedron Letters 49: 6223-6228. THOMPSON, J.E., BARROW, K.D. & FAULKNER, D.J. 1983. Localization of two brominated metab- olites, aerothionin and homoaerothionin, in spherulous cells of the marine sponge Aplysina fistularis (=Verongia thiona). Acta Zoologica 64: 199-210. UNSON, M.D. & FAULKNER, D.J. 1993, Cyanobacterial symbiont biosynthesis of chlorinated metabolites from Dysidea herbacea (Porifera). Experientia 49: 349-353. UNSON, M.D., HOLLAND, N.D. & FAULKNER, D.J. 1994. A brominated secondary metabolite synthesized by the cyanobacterial symbiont of a marine sponge and accumulation of the crystalline metabolite in the sponge tissue. Marine Biology 119: 1-11. ZABRISKIE, T.M., KLOCKE, J.A., IRELAND, C.M., MARCUS, A.H., MOLINSKI, T.F., FAULKNER, D.J., XU, C. & CLARDY, J. 1986. Jaspamide, a modified peptide from a Jaspis sponge, with insecticidal and antifungal activity. Journal of the American Chemical Society 108: 3123-3124, 174 MEMOIRS OF THE QUEENSLAND MUSEUM TRACE ELEMENT AND STABLE ISOTOPE PROFILES FROM THE CORALLINE SPONGE (ASTROSCLERA WILLEYANA). Memoirs of the Queensland Museum 44: 174. 1999- Techniques developed for laser-ablation-ICP-MS analysis of corals have now been utilised for the analysis of trace elements in the coralline sponge Astrosclera willeyana. In scleractinian corals the elements B, Mg, Sr, Baand U show seasonal variations consistent with environmental parameters, predominantly sea surface temperature and variations in upwelling. We report here a preliminary investigation to determine whether elemental distributions in sclerosponges will provide meaningful proxy information about past oceanographic conditions. Samples from Taveuni, Fiji, Ruby Reef, GBR and Truk, Caroline Islands have been analysed at a sampling resolution of —40um. With current techniques and data reduction methods, sampling at this resolution produces too much variation to show any elemental correlations. When samples are filtered to ~100um resolution, longer-term (annual to several year) patterns appear, which are consistent between the B/Ca, Mg/Ca, Sr/Ca and Ba/Ca cycles. This suggests a common incorporation mechanism between these four elements. If this variation is temperature related, the method of incorporation is markedly different than corals. The boron, magnesium and barium concentrations in sclerosponges are 2-5 times lower than in corals, with concentrations of ~20ppm, ~200ppm and ~4ppm, respectively. The strontium and uranium concentrations are 1-2.5 times higher than in corals with concentrations of ~9000ppm and ~7ppm respectively. We will also present preliminary stable isotope data (8! $0 and 5 Bc) to compare with the trace element profiles. CJ Porifera, Astrosclera, Sr/Ca, Mg/Ca, Ba/Ca, laser ablation, ICP-MS, 80, sc, environmental parameters. Stewart J. Fallon (email: Stewart.Fallon@anu.edu.au) & Malcolm T. McCulloch, Research School of Earth Sciences, Australian National University, Canberra 0200, Australia; John N.A. Hooper, Queensland Museum, PO Box 3300, South Brisbane,Old, 4101, Australia; 1 June 1998. DEMOSPONGES OF THE HOUTMAN ABROLHOS JANE FROMONT Fromont, J. 1999 06 30: Demosponges of the Houtman Abrolhos, Memoirs of the Queensland Museum 44: 175-183. Brisbane. ISSN 0079-8835. The Houtman Abrolhos lie off the west coast of Australia within a biogeographic zone that has overlapping temperate and tropical components, and a significant proportion of endemic species. The islands are situated in the path of the warm, southward flowing Leeuwin current. Studies on marine biota of these islands found a dominant tropical component to the fauna. Marine sponges of the Houtman Abrolhos are poorly studied. A field program was established to collect sponges, document the biodiversity, and determine if this biota was principally tropical or temperate in origin. 77 demosponge species are reported from the two localities examined in this study, 28 of which are known to science, 14 are identified to known species but require confirmation by comparison with type material, and 35 species are probably new. Three genera are reported for the first time from Australia. This study brings the total number of demosponge species documented from the Houtman Abrolhos to 109. Preliminary assessment of tropical versus temperate affinities indicated more species of temperate than tropical origin were present. This is contrary to comparable studies on other components of the marine biota of these islands. © Porifera, Demospongiae, Houtman Abrolhos, Western Australia, biogeography. Jane Fromont (email: jane.fromont(Amuseum.wa.gov.au), Department of Aquatic Zoology, Museum of Natural Science, Western Australian Museum, Francis Street, Perth 6000, Australia; 2 December 1998. The Houtman Abrolhos (herein referred to as the Abrolhos) are 65-90km off the W coast of Australia at 28°-29°S, 113°35’-114°03’E, near the edge of the continental shelf (Wells, 1997). There are 122 islands in 4 island groups (Fig. 1). The marine habitats of these islands are unique. 1) They are the southernmost area of major coral reef development in the eastern Indian Ocean (Wells, 1997). 2) There is a co-dominance of reef building corals and macroalgae in the upper photic zone. Macroalgae (often Ecklonia) dom- inates on windward (W facing) reefs and hard coral (Acropora) on leeward slopes. In some lagoons there may be a mixture of the two com- munity types (Wells, 1997). 3) The western rock lobster, Panulirus cygnus, a species endemic to Western Australia, is commercially fished in the Abrolhos system. This is a seasonal fishery at the Abrolhos open for three months each year when the fishers and their families occupy huts on the islands (Wells, 1997). For the rest of the year the islands are largely unoccupied. 4) The islands have high conservation value. In 1994 the Mar- ine Parks and Reserves Selection Working Group acknowledged the islands as the most significant area on the WA coast, and the most worthy of reservation (Anon., 1994). These islands are considered to be within, and towards, the northern limit of the Western Over- lap Zone (Wells, 1997), a biogeographic region on the WA coastline which has temperate and tropical components, and a significant proportion of endemic species. Studies on the marine biota of the islands found a higher proportion of northern tropical species than southern temperate species, compared to the adjacent mainland coastline at Geraldton. This high proportion of tropical species in the Abrolhos is due to the southward flowing Leeuwin Current, which carries tropical water from NW Australia (Cress- well & Golding, 1980). This is a relatively warm seasonal current that flows southward most strongly in autumn and winter, hence retaining higher sea temperatures at the islands than in adjacent mainland coastal waters (Pearce, 1997). However, geographically these islands are temp- erate, hence the co-occurrence of both tropical and temperate species. The aims of this study were to document the poorly known demosponge fauna of these is- lands, to assess their biogeographic affinities, and to compare their biogeography to other marine phyla reported from there. Seven previous publications have reported on the sponge fauna of these islands, with a total of 57 demosponges described from the Abrolhos prior to this study (Table 1). 176 MEMOIRS OF THE QUEENSLAND MUSEUM HOUTMAN ABROLHOS North Island giam sou™ WALLABI GROUP East wallabi Island West Wallabi islam MATERIALS AND METHODS Field trips were made to the islands in 1996 and 1997 to examine demosponges of the Abrolhos, and to determine timing and mode of reproductive activity of species collected. The latter results will be reported elsewhere. Two islands in two island groups were visited, Rat I. in the Easter Group (December 1996), and Beacon I., Wallabi Group (March 1997) 14*00'00 onst " acon Island gm Little North Island y 2 EASTER GROUP cA anf M Leo Island Suomi Island sw] (Fig.l). Sponge species were photo- graphed in situ or soon after collection, a representative specimen of each species from each site was preserved in 7096 ethanol, and deposited in the collections of the WAM. Relative abundance of species was estimated for each dive and summarised into 3 broad categories, + = 0-5 specimens, ++ = 5-10 specimens, and +++ = 10+ specimens seen per dive. The dominant habitat type studied on these field trips was coral reef, and intertidal reef flats at Rat I. (Table 2). Abbreviations: CSIRO, Com- . Hummock Island monwealth Scientific and Industrial Research Organisation, Perth; NCI, Marine Bioproducts Group, Australian Institute of Marine Science, Townsville; QM, Queensland Museum, Brisbane; Newman Island Sf Polpaort isind UWA, University of Western Australia, Perth; WAM, Western Australian Museum, Perth. RESULTS Seventy-seven species were recorded FIG. 1. Map of the Houtman Abrolhos. (Reproduced with from the two localities examined in this permission of Fisheries Western Australia) study. Fourteen (18%) of these were common to both localities, 40 (52%) TABLE 1. Publications on Demosponges reported were reported only from Rat I. and 23 (30%) only from the Abrolhos. from Beacon I. Of these, 28 (36%) have already been described in the literature (Table 3). Of the Author ood Field collection remaining 49 (64%) species, 14 were tentatively Dendy & : assigned to a known species but require Frederick (1924) a8. __ | Palin Hooper (1989) 1 Ark One search Vessel | TABLE 2. Summary of fieldwork program undertaken at the Abrolhos for this study. USSR Research Vessel Hooper (1991) 2 Akademik Oparin, NCI Hooper & USSR Research Vessel Island/ | Subtidal Maximum | Intertidal Habitat t 1 t y Bergquist (1992) Akademik Oparin, NCI Mud SCUBA depth (m) (reef walles| (subtidal dives) Hooper & Lévi 1 USSR Research Vessel Eroup (1993) Akademik Oparin, NCI Rat I./ H TRES OY m Easter 5 18 " Coral reef slope (4), a ooper, esearch Vesse deep hole (1 Hooper (1996) 12 Akademik Oparin, WAM, NCI Group mentee Fromont (1998) WAM, this study Beacon 1./ Coral reef slope (2), Wallabi 5 25 0 deep outcrop (1), TOTAL 57 Group patch reefs (2) DEMOSPONGES OF THE HOUTMAN ABROLHOS 177 TABLE 3. Sponge species previously reported in the literature and collected during this study at the Abrolhos. * original species name that has since been synonymised with species name given in the table (Hooper & Wiedenmayer, 1994); # probable species complexes. Localities: GBR = Great Barrier Reef, NA =N Australia, NSW = New South Wales, SA =S Australia, NT = Northern Territory, IPM = Indo Pacific/Malay, Vic = Victoria, 10 = Indian Ocean, Tas = Tasmania, NWA, = NW Australia, Abrolh = Houtman Abrolhos, WA = W Australia, New Cale = New Caledonia. Abundance estimates: + = 0-5 specimens seen in 1 dive; ++ = 5-10 specimens; +++ =>10 specimens. . . Rat I. Beacon I. Other Identification WAM no. Easter Wallabi NA SA IPM IO Areas Group Group pte” E + | - | NA | wa iaa mms | + | MES ann bank. 1862) 21198, 1199 + - | GBR, NT x Red Sea Solas, 1888 Z1278, 1281, 2241 + + SA A x Cocer eio zs - | + | nwa a e [oe | EA x | x mesa ON eee zum + | - [GBRNT x | x Sese | mongya | s | os m" Prackyeladys laevispirulifer | 71185, 1186, 1187 : + sro (Henschel 191) sie | AA A | Ne à Hooper Berequist 1992 | 21176, 117, 1178 + + Abrolh TT numum dr. n o pi duis Z1191 + - NWA | Vic, Tas Gellar thee | zar | € | - WA quiu m 0| + [AW "e? (bendy Frederick, 1924) Z1158 3 + Abrolh al 21159 + |: WA, Vi Bendy 1896" zb à i Nf de Kus | aspe [oue [oe pnr ME (bendy 1992) Z1302 + " X X Abrolh Miror toria 271255, b 1257, + y. WA AO 21297 E WA cod lacte 1814) ZH» : d NSW * Hooper, poaa) selachia Z1282, 1283 + + NWA pehini bum tlathrioides Z1195 - ES NWA WA C ree Z1406 + A GBR X oper 171] aeformis 21400, 1401, 1402 , +++ | NT, GBR x x eere lege) plicata 21201, 1202, 1203 + - pl E x x 178 MEMOIRS OF THE QUEENSLAND MUSEUM TABLE 4. Undescribed species, and unconfirmed species identifications, collected during this study from the Abrolhos. Distributions: TR = tropical, ST = subtropical, TE = temperate, CO = cosmopolitan, Carrib = Carribean, Aust = Australia, NZ = New Zealand, Sth Afr = South Africa. Rat I. Beacon I. Other Identification WAM no. Easter Wallabi TR TRST |TRSTTE CO : regions Group Group Plakortis sp. 1 Z1272, 1273, 1274 E E Plakortis sp. 2 zl4, ie Cn 1271, ++ ++ Corticium cf. simplex Lendenfeld, 190 21392 d a x Corticium cf. candelabrum Schmidt, 1862 21380 "E É X Penares? cf. intermedia (Dendy, 1905) Z1397 7 5 * Stelletta cf. brevis Hentschel, 1909 Z10, 1190 pl i x Tethya cf. multistella Lendenfeld, 1888 Z1277 ij B x Anthosigmella sp. Z1303 + - X Carrib Aaptos sp. Z1165, 1166 +H+ E X Theonella sp. Z1242 aie - X Agelas cf. mauritiana E arter, 1883) Z1200 * : x Agelas sp. 1 Z1193, 1194 + - X Agelas sp. 2 Z17 - X Phycopsis sp. Z1254 - + X X Pararhapoxya sp. Z1261 + - X NZ Dragmatella? sp. Z1298 - + ? mbastela cf. vespertina Feces & Bergquist 1992 21174, 1175 s Neofibularia sp. Z1391 - + X lotrochota sp. 21291, 1292 ++ - X Tedania cf. anhelans (Lieberkühn, 1859) Z1296 i : X Ectyodoryx sp. Z1301 HER E X Bi-polar Phoriospongia sp. Z8, 1294 ++ - X Aust/ NZ Guitarra sp. Z1265, 1266, 1267 - + x e ahi Liosina sp. Z1205 + - X Clathria (Thalysias) cf. abietina [borrada 1814) Z1260 E = x Haliclona cf. toxotes Z1393, 1394, 1395, gi + x (Hentschel, 1912) 1396 Haliclona sp. 1 Z9,11, 12, 1405, 2239 + X Haliclona sp. 2 Z13 - $ X Haliclona sp. 3 21399, 1403 + X Reniera sp. 1 Z1375: - X Reniera sp. 2 Z1384 - X Reniera sp. 3 21383 ++ - X Reniera sp. 4 Z1381 dh - X Gellius sp. 1 Z1379 - + X Niphates cf. nitida Fromont, 1993 ZI 1s * = Gelliodes cf. obtusa Z1250, 1251, 1252, Hentschel, 1912 1253 e" Š m Aka sp. 1 21263, 1264 + + X DEMOSPONGES OF THE HOUTMAN ABROLHOS 179 Table 4 cont, Callyspongia sp. 1 Z1385 + x Callyspongia sp. 2 Z1386 + X Callyspongia sp. 3 Z1387 + X Callyspongia sp. 4 Z1382 + - x UM a 5 peptafo Z1388 + X pibe ni er ete, 1903) 21389 T X | Spongia sp. Z15 - + X Psammocinia sp. 21262 + X Dysidea sp. Z1299 + - X Spongionella sp. 21390 + X Dendrilla sp. Z1300 + X | Pseudoceratina sp. Z1192 * X confirmation by comparison with type material, and 35 are probably new (Table 4). Of the 28 known species reported from the Abrolhos in this study 15 (1395) had previously been reported from this locality. Two of these, Ancorina brevidens Dendy & Frederick (1924) and Megalopastas arenifibrosa Dendy & Frederick (1924), have since been synonymised with more widespread species, A. acervus (Bow- erbank, 1862) and Lendenfeldia plicata (Esper, 1806) respectively, by Hooper & Wiedenmayer (1994). Three species are apparent endemics for the Abrolhos: Cymbastela marshae Hooper & Bergquist (1992), Halichondria phakelloides Dendy & Frederick (1924) and Phorbas fic- titioides Dendy & Frederick (1924). For 2 species, Plakinastrella minor (Dendy, 1916) and Zyzza massalis (Dendy, 1922), the Abrolhos is so far their only Australian locality (Hooper & Krasochin, 1989 & this study). Two species are new records for WA; Haliclona amboinensis (Lévi, 1961) and H. cymaeformis (Esper, 1791). A further 14 species are known in the literature, but either some of their taxonomic characters were significantly different from published des- criptions, or conspecificity would have produced highly disjunct distributions. In both these cases examination of type material is required to con- firm identities, this has not yet been possible. These species are presently prefixed with ‘cf? (Table 4). Thirty five species could only be id- entified to genus and are probably new (Table 4). Generic distributions are presented as per Van Soest (1994) and in the case of the 14 uncon- firmed identifications, the known distribution of these species as reported in the literature. Three of the genera reported here represent new records for Australia. Anthosigmella has been previously reported only from the Carrib- bean (see Wiedenmayer, 1977), Guitarra from South Africa (Lévi, 1963), New Zealand (Brondsted, 1924; Dendy, 1924; Bergquist & Fromont, 1988), and E Pacific coast (Des- queyroux Faundez & Van Soest, 1997) and Liosina from Papua New Guinea (Kelly Borges & Bergquist, 1988) and Bawi Island, Zanzibar, Tanzania (Kelly-Borges, 1998). DISCUSSION This study increases the total number of demo- sponge species reported from the Abrolhos from 57 to 109. This number of species is likely to represent only a small proportion of the total sponge fauna of these islands considering only two islands in two of the four island groups were surveyed; none of the algal-dominated areas have yet been visited; and depths were restricted to less than 18 and 25m respectively. A similar style of sponge collection, but with a much larger number of sampling sites, was undertaken on the NW. Australian oceanic reefs of Ashmore, Cartier and Hibernia, from which 138 species were reported (Hooper, 1994). The 109 species so far reported from the Abrolhos therefore indicates there is a very rich sponge fauna around these islands. Fourteen of the 77 species collected during this study were common to both sites, but 40 (52%) of the remainder of species occurred only at Rat I. and 23 species (30%) at Beacon I. These dif- ferences in species compositions may indicate fundamental differences between the islands in a proportion of their sponge biota. For example, the 4 island groups are separated by channels of approximately 40m depth which may restrict movement between island groups of gametes of some species. It is also possible that there are 180 MEMOIRS OF THE QUEENSLAND MUSEUM TABLE 5. Species previously described from the Abrolhos but not recollected during this study. * original species name that has since been synonymised with the species name given in the table (Hooper & Weidenmayer, 1994); # may be a species introduced via shipping. Localities: NA = N Australia, SA = S Australia, IPM = Indo Pacific/Malay, IO = Indian Ocean, NWA = NW Australia, WA = W Australia, GBR = Great Barrier Reef, Qld = Queensland, NS W = New South Wales, NT = Northern Territory, Vic = Victoria, Tas = Tasmania, Abrolh = Abrolhos, NZ = New Zealand, Sth Afr = South Africa, Subant = Subantarctic. Species NA SA IPM 10 Other areas Stelletta debilis Thiele, 1900 x Abrol Stelletta sigmatriaena Lendenfeld, 1907 NWA Ancorina australienesis (Carter, 1883) WA Rhabdastrella rowi (Dendy, 1916) X Abrol Asteropus simplex (Carter. 1879) Qld, NWA WA, Vic X NZ, Easter I Erylus proximus Dendy, 1916 X Abrol Tethya robusta (Bowerbank, 1859) Old WA X X Red Sea * Xestospongia similis (Ridley & Dendy, 1886) NT WA, NSW X X Subant *Callyspongia mollis (Lendenfeld, 1887) NSW Abrol *Callyspongia ramosa (Gray, 1843) Qld Tas, NSW, Vic x _| NZ, Subant Oceanapia abrolhosensis (Dendy & Frederick, 1924) Abrol. * Mycale parasitica (Carter, 1885) TAS, NSW, Vic X Mycale trichophora (Dendy & Frederick, 1924) Abrol #* Mycale parishi (Bowerbank, 1875) NT, NWA WA, NSW x X Sth. Afr Biemna tubulata (Dendy, 1905) X Abrol Waldoschmittia schmidti (Ridley, 1884) WA, Tas, NSW, Vic X X Holopsamma crassa Carter, 1885 Tas, NSW, Vic Dysidea dakini (Dendy & Frederick, 1924) | Abrol Hyatella intestinalis (Lamarck, 1814) GBR, NWA WA, X X *Coscinoderma pesleonis (Lamarck, 1814) Vic, Tas, WA Echinodictyum nidulus Hentschel, 1911 NWA, NT | Clathria (Wilsonella) abrolhosensis Hooper, 1996 Abrol Clathria (Wilsonella) australiensis (Carter, 1885) NSW Clathria (Microciona) grisea (Hentschel, 1911) NWA Clathria (Dendrocia) pyramida Lendenfeld, 1888 NSW Clathria (Axociella) patula Hooper, 1996 NWA Clathria (Thalysias) aphylla Hooper, 1996 Abrol Clathria (Thalysias) cancellaria (Lamarck, 1814) NWA Clathria (Thalysias) styloprothesis Hooper, 1996 WA Antho (Antho) tuberosa (Hentschel, 1911) | NWA X Holopsamma arborea (Lendenfeld, 1888) NWA WA,NSW L Caulospongia plicata Saville Kent, 1871 NWA significant microhabitat differences between these islands, thus influencing species compos- ition of each island (cf. Hooper, 1994), but this has not been investigated. One of the aims of this study was to determine if the sponges of the Abrolhos were principally tropical or temperate in origin. For this reason a list of species previously recorded from the Abrolhos, but not recollected during this study, is included (Table 5). Inclusion of this dataset (Table 5) brings the number of species known to be endemic to the Abrolhos to a total of 8, 3 recollected during this study (noted above) and 5 others: Mycale trichophora Dendy & Frederick (1924), Dysidea dakini Dendy & Frederick (1924), Oceanapia abrolhosensis (Dendy & Frederick, 1924), Clathria (Wilsonella) abrolhosensis Hooper (1996), and C. (Thalysias) aphylla Hooper (1996). Whether these species are true endemics to the islands, or more widely distributed but not yet reported, will not be TABLE 6. Proportion of tropical, temperate and endemic species of sponges occurring at the Abrolhos. DEMOSPONGES OF THE HOUTMAN ABROLHOS 181 have benthic larvae which may not have the temporal capability 7 Trop. & Endemic | Endemic | tO Survive a migration on the | Nomb£spettós ||. Trop Temp. | Temp. Abrol WA Leeuwin current. This would In this study 9 5 T 3 6 reduce the proportion of tropical Reported in pre- E A fo 3 3 species able to recruit to the vious studies | islands. 2) The Leeuwin current is Total 17(28%) | 12 (20%) | 23(38%) | 8(13%) 13 known to have been in existence known until further work is undertaken in adjacent temperate and tropical localities in WA. In addressing tropical and temperate origins of species, those identified only to genus, or with unconfirmed identifications, were excluded (Table 4). The majority of known species (Tables 3, 5) are of temperate origin. Twenty-three species (38%) are known from temperate waters, 17 species (28%) are tropical and 12 (20%) have a more widespread tropical and temperate distribution. Thirteen species are only known so far from WA, and appear to be endemic to the State. Their distribution as either temperate or tropical, or both, is incorporated into these categ- ories in Table 6. This biogeographic analysis of sponges of the Abrolhos is considered preliminary given that a large component of the fauna is presently excluded from the assessment. However, these data on proportions of temperate versus tropical species are in marked contrast to other marine biota reported from these islands. For most phyla there is a greater component of tropical than temperate species in the fauna (Table 7); echinoderms, molluscs and fishes have similar proportions of tropical versus temperate species. In contrast, sponges have a greater temperate component; amongst other phyla only seagrasses show a temperate species dominance. Should this apparent dominance of temperate sponge species be eventually confirmed, it may be the result of: 1) The reproductive biology of the sponges, whereby some species are known to since the Eocene (40 m.y.a.), and has continued to occur in pulses since this time. Periods when the current has not flowed may have allowed for recruitment of temperate species from the adjacent coastline. In summary, this study doubles the number of demosponges reported from the Abrolhos. First indications are that the sponge fauna is relatively species rich, with a larger number of temperate than tropical species. Much work remains to fully document the fauna, including in the North and Pelsaert Island groups, algal dominated reefs, and greater depths than sampled here. Until the sponge fauna of localities both N and S ofthe Abrolhos, and on the adjacent W coast of Australia are better documented, this work remains a study in isolation. Consequently, con- clusions on species endemicity remain tentative, and the affinities of the undescribed component of the fauna are not presently known. ACKNOWLEDGEMENTS Thanks to Jane Griffith, Barry Hutchins (WAM) and Christine Hass (UWA) for field assistance, Mark Salotti (WAM) for photo- graphic, databasing and other technical help and Alex Bevan (WAM) for the use of his photo- microscopy equipment. I thank Paddy Berry, Loisette Marsh, Diana Jones, Shirley Slack-Smith, Barry Hutchins, Clay Bryce and Ken McNamara of the WAM, and Alan Pearce CSIRO, for helpful discussions. John Hooper (QM) provided unpublished information on sponges collected TABLE 7. Proportions of tropical and temperate components of marine phyla studied at the Abrolhos. * total number of different species, including 49 as yet unnamed and 60 named. Tropical Subtropical Temperate Temp. & Endemic Endemic WA e Phylum (%) (%) (%) Trop. (%) | Abrol(%) (0%) Total. speties e origine) 28 39 20 13 60 (109*) Echinoderms (Marsh. 1994) 64 15 a | 12 Molluscs 69 20 T 492 (Wells & Bryce, 1997) E " Fish (Hutchins, 1997) 67 13 20 389 Seagrasses (Brearley, 1997) 30 70 10 from the Abrolhos. Two referees provided useful comments on the manuscript. I acknowledge the assistance of Fisheries Western Australia, Ger- aldton office; Kim Nardi for logistical fieldwork support, and Randall Owens for helpful advice on diving localities and assistance at the islands. I thank the crew of the P.V. McLaughlin for transport to Rat I., the captain, Ray Howarth and crew of Island Leader for transporting equipment to Beacon I. Mr Dramsfield and the fishers and families of Beacon I. were welcoming and provided useful local advice. Thanks to L. Matz, Bruce Beaney and staff of the Histopathology Lab. of Royal Perth Hospital for allowing me use of their microtomes. The Australian Biological Resources Study (ABRS) for providing funding allowing me to participate in the 5th International Sponge Symposium, ‘Origin & Outlook’ in Brisbane, Australia, and for funding this study which was carried out at the Museum of Natural Science, Perth, Western Australia. LITERATURE CITED ANONYMOUS, 1994. Final Report. Marine Parks and Reserves Selection Working Group. (Abrolhos Islands Consultative Authority: Geraldton). BERGQUIST, P.R. & FROMONT, J. 1988. The marine fauna of New Zealand: Porifera, Demospongiae, Part 4 (Poecilosclerida). New Zealand Ocean- ographic Institute Memoir 96: 1-197. BOWERBANK, J.S. 1862. On the anatomy and physi- ology of the Spongiadae. Part 3. On the generic characters, the specific characters, and on the method of examination. Philosophical Trans- actions of the Royal Society 152: 1087-1135. BREARLEY, A. 1997. Seagrasses and isopod borers from the Wallabi Islands, Houtman Abrolhos Islands, Western Australia Pp. 64-73. In Wells, F.E. (ed.) The Marine Flora and Fauna of the Houtman Abrolhos Islands, Western Australia Volume 2. (Western Australian Museum: Perth). BRONDSTED, H.V. 1924. Papers from Dr Th. Mor- tensen’s Pacific Expeditin 1914-16. 23. Sponges from New Zealand. Part 1. Videnskabelige Meddelelser fra Dansk naturhistorisk Forening 77: 435-483. CRESSWELL, G.R. & GOLDING, T.J. 1980. Observ- ations of a south-flowing current in the southeastern Indian Ocean. Deep Sea Research 27A: 449-466. DENDY, A. 1924. Porifera. Part 1. Non-Antarctic Sponges. British Antarctic (‘Terra Nova’) Expedition, 1910. Natural History Report. 6(3): 269-392. (British Museum (Natural History), Zoology: London). DENDY, A. & FREDERICK, L.M. 1924. On a col- lection of sponges from the Abrolhos Islands, MEMOIRS OF THE QUEENSLAND MUSEUM Western Australia. Journal of the Linnean Society of London, Zoology 35: 477-519. DESQUEYROUX-FAUNDEZ, R. & SOEST, R.W.M. VAN 1997. Shallow waters Demosponges of the Galapagos Islands. Revue Suisse de Zoologie 104(2): 379-467. FROMONT, J. 1998, Revision of the marine sponge genus Caulospongia Saville Kent, 1871 (Demospongiae: Hadromerida). Part 1. Morphological and skeletal characters. Records of the Western Australian Museum 19: 65-89, HOOPER, J.N.A. 1991. Revision of the Family Raspailiidae (Porifera: Demospongiae), with description of Australian species. Invertebrate Taxonomy 5(6): 1179-1415. 1994. Coral reef sponges of the Sahul Shelf - a case study for habitat preservation. Memoirs of the Queensland Museum 36(1): 93-106. 1996. Revision of Microcionidae (Porifera: Demo- spongiae: Poecilosclerida), with description of Australian species. Memoirs of the Queensland Museum 40: 1-626. HOOPER, J.N.A. & BERGQUIST, P. R. 1992. Cym- bastela, a new genus of lamellate coral reef sponges. Memoirs of the Queensland Museum 32(1): 99-137. HOOPER, J.N.A. & KRASOCHIN, V.B. 1989. Re- description of the burrowing sponge Zyzzya massalis (Dendy) from the Seychelles and Houtman-Abrolhos Islands. The Beagle, Records of the Northern Territory Museum of Arts and Sciences 6(1): 133-140. HOOPER, J.N.A. & LEVI, C. 1993. Poecilosclerida (Porifera: Demospongiae) from the New Caledonia Lagoon. Invertebrate Taxonomy 7: 1221-1302. HOOPER, J.N.A. & WIEDENMAYER, F. 1994. Por- ifera. Pp. 1-621. In Wells, A. (ed.) Zoological Catalogue of Australia. Vol. 12. (CSIRO Australia: Melbourne). HUTCHINS, J.B. 1997. Checklist of fishes of the Houtman Abrolhos Islands, Western Australia. Pp. 239-253. In Wells, F.E. (ed.) The Marine Flora and Fauna of the Houtman Abrolhos Islands, Western Australia. Vol. 1. (Western Australian Museum: Perth). KELLY-BORGES, M. 1998. Sponges. Pp. 106-115. In Richmond, M.D. (ed.) A Guide to the Seashores of Eastern Africa and the Western Indian Ocean. (Marine Science Programme, Sarec. Ord & Vetander: Sweden). KELLY-BORGES, M. & BERGQUIST, P.R. 1988. Sponges from Motupore Island, Papua New , Guinea. Indo-Malayan Zoology 5: 121-159. LEVI, C. 1963. Spongiaires D’Afrique du Sud (1) Poecilosclerides. Transactions of the Royal Society of South Africa 37(1): 1-72. MARSH, L.M. 1994. Echinoderms of the Houtman Abrolhos Islands, Western Australia and their relationship to the Leeuwin Current. Pp. 55-61. In Guille, D.B., Feral, A. & Roux, J-P. (eds) DEMOSPONGES OF THE HOUTMAN ABROLHOS Echinoderms through Time (Balkema: Rotterdam). PEARCE, A.F. 1997. The Leeuwin Current and the Houtman Abrolhos Islands. Pp. 11-46. In Wells, F.E. (ed.) The Marine Flora and Fauna of the Houtman Abrolhos Islands, Western Australia Vol. 1. (Western Australian Museum: Perth). SOEST, R.W.M. VAN. 1994. Demosponge distribution patterns. Pp. 213-223. In Soest, R.W.M. Van, Kempen, T.M.G. van, Braekman, J.-C. (eds) Sponges in Time and Space. (Balkema: Rotterdam). WELLS, F.E. 1997. Introduction to the marine en- vironment of the Houtman Abrolhos Islands, 183 Western Australia. Pp 1-10. In Wells, F.E. (ed.) The Marine Flora and Fauna of the Houtman Abrolhos Islands, Western Australia Vol. 1. (Western Australian Museum: Perth). WELLS, F.E. & BRYCE, C.W. 1997. A Preliminary checklist of the marine macromolluscs of the Houtman Abrolhos Islands, Western Australia. Pp. 326-383. In Wells, F.E. (ed.) The Marine Flora and Fauna of the Houtman Abrolhos Islands, Western Australia. (Western Australian Museum: Perth). WIEDENMAYER, F. 1977. Shallow-water sponges of the Western Bahamas. Experimentia Sup- plementia 28: 1-287. (Birkhauser: Basel). 184 MEMOIRS OF THE QUEENSLAND MUSEUM SPONGE CELL ADHESION: AN EVOLUTIONARY ANCESTOR OF HISTO- COMPATIBILITY SYSTEMS ? Memoirs of the Queensland Museum 44: 184. 1999:- Sponges have been traditionally used as models to study cell adhesion because their rather loose and porous extracellular matrix allows a mild cell dissociation and the recovery of intercellular components in virtually native state. Species-specific cell recognition and adhesion in sponges is mediated by extracellular proteoglycan-like complexes termed aggregation factors (AFs), still not identified in higher animals. Polyvalent glycosaminoglycan interactions are involved in the species-specificity, representing one of the few known examples of a regulatory role for carbohydrates. A surprising characteristic of sponges, considering their low phylogenetic position, is that they possess an exquisitely sophisticated histocompatibility system. Any grafting between two different sponge individuals is almost invariably incompatible in the many species investigated, exhibiting a variety of transitive qualitatively and quantitatively different responses, which can only be explained by the existence of a highly polymorphic gene system regulating sponge allogeneic reactions. The development of variable-region molecules is thought to have been a crucial event in the evolution of primordial vertebrate immune systems, followed by gene rearrangement to provide more diversity. Early in the evolution of the immune system, then, a gene must have duplicated to allow such diversity to arise. Unfortunately, there is an absolute lack of protein sequence information concerning the molecules involved in invertebrate histoincompatibility reactions. Recently, we deduced from cDNA the sequence of the aggregation factor core protein from the red beard sponge, Microciona prolifera, and Southern blot analysis suggested the existence of several related genes. We have screened individual sponge cDNA libraries, identifying multiple related forms for the AF core protein (MAFp3). Northern blots show the presence in several human tissues of transcripts strongly binding a MAFp3-specific probe. We have studied tissue histocompatibility within a sponge population, finding 100% correlation between rejection behaviour and the individual-specific restriction fragment length polymorphism pattern using AF-related probes. PCR amplifications with specific primers showed that at least some of the MAFp3 forms are allelic and distribute in the population used. A pronounced polymorphism is also observed when analysing purified AF in polyacrylamide gels. Protease digestion of the polymorphic glycosaminoglycan-containing bands indicates that glycans are also responsible for the variability. The data presented reveal a high polymorphism of aggregation factor components which matches the elevated sponge alloincompatibility, suggesting an involvement of the cell adhesion system in sponge allogeneic reactions. Our present work will be discussed in the context of the evolution of histocompatibility systems and their possible divergence from primitive cell-cell interaction molecules. O Porifera, graft rejection, proteoglycans, invertebrate immunity, aggregation factors, cell adhesion, porifera genes, cDNA, histocompatibility. Xavier Fernandez-Busquets* (email: xavi@farmacia. Far.ub.es) & Max.M. Burger, Friedrich Miescher-Institut, P.O. Box 2543, CH-4002 Basel, Switzerland. *Present address: *Departament de Bioquimica i Biologia Molecular, Facultat de Farmacia, Universitat de Barcelona, Avda. Diagonal 643, E-08028 Barcelona, REPRODUCTION OF SOME DEMOSPONGES IN A TEMPERATE AUSTRALIAN SHALLOW WATER HABITAT J. FROMONT Fromont, J. 1999 06 30: Reproduction of some demosponges in a temperate Australian shal- low water habitat. Memoirs of the Queensland Museum 44: 185-192. Brisbane. ISSN 0079-8835. Species of Tethya, Chondrilla, Mycale and Echinodictyum were monitored for two years at South Mole, Fremantle, Western Australia (32°04’S, 115°45’E) to determine onset of reproductive activity, sex phenotype, and reproductive mode. Most reproductive activity occurred from late spring through summer and autumn (November-April). Most species appear to be gonochoric with both ovipary and vivipary recorded. Details of their reproductive development is reported and discussed in relation to sea temperature data. 0 Porifera, Demosponges, Fremantle, Western Australia, reproduction. Jane Fromont (email: jane.fromont@museum. wa. gov.au), Department of Aquatic Zoology, Western Australian Museum, Francis Street, Perth, WA 6000, Australia;29 January 1999. The only previous study on the reproductive biology of temperate Australian marine demosponges (Hoskins, 1992), found low numbers of oocytes present between March and September in populations of Phyllospongia sp. from Rottnest Island, Western Australia (WA). This lack of basic biological information for this region is surprising given that sponges are a dominant component of the sessile fauna in these temperate marine habitats. Consequently, the present study aimed to collect baseline reproductive data for some of the more common species of demosponges in this region, examining species of Echinodictyum, Mycale, Haliclona, Coelosphaera, Tethya and Chondrilla. Both Tethya and Chondrilla have two distinctive colour morphs at the study site, possibly indicating sympatric species, so these colour morphs were monitored as separate populations. Several studies undertaken in Europe and USA have described reproductive characteristics for temperate species belonging to some of the genera examined here. For instance, Elvin (1976) reported on the reproductive biology of Haliclona permollis in Oregon; Fell (1976) on H. loosanoffi in Connecticut; and Wapstra & Van Soest (1987) on H. oculata and H. xena in Holland. Mycale micracanthoxea from Holland (Wapstra & Van Soest, 1987) is the only temperate species of Mycale previously examined. Temperate populations of Chondrilla nucula, Tethya aurantium and T. citrina were studied by Liaci (1971a, b), whereas there are no published studies on the reproductive biology of Coelosphaera or Echinodictyum. Increasing temperature is generally accepted as a major environmental factor regulating the onset of reproductive activity in sponges occurring in regions of large seasonal change (Fell, 1983; Simpson, 1984). Only four species are presently known where gametogenesis is associated with a decrease in temperature: Halisarca dujardini (Lévi, 1956; Chen, 1976) Desmacidon fructicosum (Lévi, 1956), Tethya crypta and Aplysina gigantea (Reiswig, 1973). In this study seasonal reproductive activity is discussed in relation to sea temperature data, and results are presented on the mode of reproduction and sexual phenotype of the species examined, estimates of development time of gametes, and timing of product release. MATERIALS AND METHODS The six species of sponges investigated here live subtidally on the ocean side of South Mole, an artificial groyne that forms the southern flank of the entrance to Fremantle Harbour (32°04’S, 115%45”E). All six species at this site ranged in abundance from common to abundant (i.e. >10 specimens of each species seen during a dive of 1 hour duration). Sponges were prolific from 4-7.5m depth, occurring between the shallow seaweed (Ecklonia) fringe found at 0-4m depth and the sand flat with seagrass at the base of the groyne. Sampling was conducted over two years from October 1996 to April 1998. Sampling ceased in 186 MEMOIRS OF THE QUEENSLAND MUSEUM TABLE 1. Reproductive activity of the species studied at South Mole. Key: -, not sampled; NR, not reproductive; X, not found in the field; O, oocytes; E, embryos; L, larvae; S, sperm; (sample size). Species Monti (year of survey) A S O N D J F M A Hii ee, | - | -w || -w | we | mo | wu» | mo | xo modi 98) | X0) | (0) | NR() | NRG) | NRG) | NRG) | O@ | NRG) | NRG i dah sp. - (0) - (0) NR (2) - (0) NR (2) NR (2) NR (6) NR (2) OE (2) Goa has sp. NR (2) - (0) NR (3) NR (3) NR (3) NR (3) NR (3) NR (3) S(3) | Haliclona sp. (96-97) - (0) - (0) X (0) - (0) NR (4) NR (6) X (0) X (0) NR (2) Haliclona sp. (97-98) | EL (2 -(0) NR(2) | NR(4) | NR(4) | NR(4) | NRG) 06) S) Mycale sp. (96-97) - (0) - (0) NR (2) -(0) | OELS(5) | ES(7 | NRG) | NRQ) | NRO) Mycale sp. (97-98) NR (2) - (0) NR (4) S (6) ELS (3) | ELS(5) | NR(@) | NR) | NR(4) Chondrilla australiensis (ochre - (0) - (0) NR (2) -(0) NR (2) NR (5) O (4) NR (5) NR (3) morph)(96-97) | Chondrilla australiensis (ochre NR (2) - (0) NR (4) NR (9) NR (7) NR (7) O (6) NR (2) NR (2) morph) (97-98) Chondrilla australiensis (ataoi - (0) - (0) NR (2) - (0) NR (4) NR (2) 0(3) X (0) NR (4) morph) (96-97) Chondrilla australiensis ears NR (2 - (0) NR (2) NR (2) NR (2) NR (4) O(4) NR (3) NR (7) morph) (97-98) Tethya sp (pink - (0) - (0) NR (2) - (0) X (0) O (2) O (4) X (0) NR (2) Tebas d X (0) - (0) X (0) NR (4) O (10) NR (9) O(7) O (9) NR (8) Te sp nee - (0) - (0) X (0) -(0) NR (4) O (4) O (4) O(5) NR (2) Terva T Sg) e" NR (4) - (0) NR (7) NR (5) O(8) O (6) O(8) O (10) NR (9) the winter months of May, June and July 1997 and no sampling was possible in November 1996 and September 1997 due to bad weather and sea conditions. From October 1996 to April 1997 monthly samples of random individuals of each species were collected. With the resumption of sampling in August 1997 two regimes were adopted: sampling of random individuals as for the previous season, and sampling of known individuals of each species to monitor for sequential hermaphroditism. Two different techniques were used. Specimens of Tethya were sampled with a 0.5mm diameter cork borer, and ramose branching, encrusting, massive and fan shaped species had a small piece incised from them with a scalpel. Numbers of specimens of each species that were collected and examined by light microscopy are indicated in Table 1. After collection, individual samples were placed in labelled glass vials and on return to the laboratory were fixed in a gonad fixative, FAACC (100ml = 10ml 37-40% formaldehyde solution: 5ml glacial acetic acid: 1.3gm calcium chloride dihydrate: 85ml tap water) for <48 hours, and then transferred to 75% ethanol. Sections cut at 8um were stained with haematoxylin-eosin, mounted, and surveyed by light microscope for presence and development of eggs and sperm. Average sizes of gametes were calculated, using an ocular micrometer, by measuring gametes from each gravid individual for each sampling period. To decrease sampling variation only oocytes sectioned through the nucleus, and sperm cysts sectioned through the midline, were measured. Temperatures at 7m depth adjacent to the Fisheries Western Australia Marine Research Laboratories at Waterman were recorded twice daily (M. Rossbach, Fisheries Western Australia, pers.comm.). Monthly averages of this data, supplied by Fisheries WA, were calculated for the sampling months outlined above. REPRODUCTION IN WESTERN AUSTRALIAN DEMOSPONGES RESULTS 1. Echinodictyum clathrioides Hentschel, 1911 (Poecilosclerida: Raspailiidae). Adults of E. clathrioides are erect, fan shaped individuals that are at least 30cm in diameter. Female gametes were only found in specimens collected in February 1998, when oocytes were 45um diameter (Fig. 1A). There were no oocytes in either January or March suggesting rapid oocyte development and release of products (Table 1). Small sponges <8cm diameter were found in April 1998. 2. Coelosphaera sp. (Poecilosclerida: Coelosphaeridae). Individuals of this species are rounded mounds with apical oscules and are bright orange alive. Reproductive products were found in April of both sampling seasons (Table 1). Oocytes and embryos (Fig. 1B) were present in April 1997 and sperm in April 1998 implying that reproductive development of embryos and larvae occurs in autumn and possibly winter. 3. Haliclona sp. (Haplosclerida: Chalinidae) is a maroon, ramose branching sponge with apical oscules. Specimens of this species rarely had reproductive products during this sampling program. Large embryos and larvae (Fig. 1C) were present in August 1997, and developing oocytes were found in March and sperm in April 1998 (Table 1). Presence of gametes at this time suggest that this species is reproductively active throughout winter. This species is viviparous and individuals are either gonochoric or successive hermaphrodites. 4. Mycale sp. (Poecilosclerida: Mycalidae). This species exudes large amounts of mucous upon collection. In the field the sponge has short, thick erect lobes with prominent conules, and is irridescent mauve or vivid blue. Sampling ofthe mesohyl of this species for reproductive products was difficult as most of the mesohyl oozed away prior to the sponge being placed in the collection vial, and the mesohyl that remained was detached from the skeleton prior to fixation. However, this species has particularly obvious and abundant female gametes visible in the field as orange spheres of about 2mm diameter. Because of the mucous mesohyl, few of these products were successfully sectioned. Female gametes were found in the field during December and January in 1997 and 1998, and were absent in February of both years (Table 1). Sperm cysts were found interspersed amongst embryos and larvae (Fig. 1F). Reproductive activity occurred for a 187 minimum period of 62 days from first development noted in November to the last date when gametes were present in January. This species is viviparous and contemporaneously hermaphroditic. Two colour morphs were observed in the remaining two genera examined, Tethya and Chondrilla. Consequently, replicate specimens of each morph were monitored separately to determine if reproductive timing or sex determination differed between them. 5. Chondrilla australiensis Carter, 1873 (Hadromerida: Chondrillidae). Specimens were either ochre to brown or maroon. The ochre colour morph was the dominant morph at the study site with extensive mats, up to Im across, living in full light. Although both morphs tended to occur either in full light or shade under Ecklonia, the maroon morph occurred more frequently in shaded areas. Few reproductive products were found in either morphotype. Individuals were found with oocytes in late February 1997 (Table 1). No products were seen in the next sampling in late March. Oocytes measuring 30-40um were abundant in February 1998 and had cellular extensions between the mesohyl and the oocytes (Fig. 1E), Oocyte development in this species is rapid with 34 days elapsing between the January sampling (when no female products were visible) and the February sampling (when oocytes were 30-40um). No oocytes were present in the March sampling 21 days later. No sperm were seen in either sampling year. It is assumed therefore that spawning occurs in late February or early March, approximately 2-4 weeks earlier than in Tethya. These sponges are oviparous and probably gonochoric. Asexual reproduction by fragmentation appeared to be occurring in C. australiensis in April 1998, whereby elongated tear shaped droplets of sponge tissue were found extending from the edges of some of the adults. The tissue was thinnest at the point of attachment to the adult sponge and thickest at the furthest edge. It is likely that these droplets would detach from the adult and settle on the substrate beneath. 6. Tethya sp. (Hadromerida: Tethyidae). Specimens were either pink or orange and individuals of both colour morphs had numerous oocytes in February and March of 1997 and 1998, Oocyte development was first detected in early December when oocytes were 10-12um in diameter. Ninety eight days later, in early March, oocytes were 50-70um in diameter (Fig. 1D). 188 MEMOIRS OF THE QUEENSLAND MUSEUM FIG. |, Reproductive products of species examined in this study. A, Oocytes in Echinodictyum clathrioides 17/2/98. B, Embryo and oocytes in Coelosphaera sp. 29/4/97. C, Larvae and embryo in Haliclona sp. 26/8/97. D, Oocytes in Tethya sp. 25/3/97. E, Oocytes in Chondrilla australiensis 17/2/98. F, Larvae and sperm cysts (<) in Mycale sp. 20/12/96 (scale bar: E = 50um, A-D, F = 100m). REPRODUCTION IN WESTERN AUSTRALIAN DEMOSPONGES TABLE 2. Developmental mode of some sponge species in Western Australia. Abbreviations: X = developmental mode, developmental mode but awaits confirmation 189 European locality were reported as viviparous with contemporaneous hermaphroditism (Wapstra & Van ? = suspected Soest, 1987). In this study, Haliclona sp. is viviparous but is not contemporaneously hermaphroditic. This species is either gonochoric or a successive hermaphrodite. Species d Vivipary | Gonochorism Kis die wis ivive - Tethya sp. X ? pie a X ? Echinodictyum 9 9 clathrioides i i Mycale sp. xX X Ilan & Loya (1990) reported finding aggregations of female There were no oocytes in samples taken at the next sampling period in April, either in 1997 or 1998 (Table 1). It is therefore likely that spawning occurs in mid to late March. No sperm were found in either sampling years. These sponges are oviparous and probably gonochoric. Thin filaments were seen extending from one individual of Tethya sp. in April 1998. This individual had been sexually reproductive with oocytes visible in March. In summary, two types of reproductive mode were observed amongst these six species: ovipary in Tethya sp. and Chondrilla australiensis, and vivipary in Mycale sp., Coelosphaera sp. and Haliclona sp. (Table 2). Only one sex phenotype was determined, contemporaneous hermaphroditism was found in Mycale sp. Two opposing trends are apparent when sea temperatures are compared with timing of repro- ductive activity (Fig. 2). In Tethya sp., Chondrilla australiensis, Mycale sp. and probably Echinodictyum clathrioides reproductive activity occurs in late spring or summer when sea temperatures are increasing or reaching a maximum. Conversely, in Coelosphaera sp. and Haliclona sp. reproductive activity occurs in autumn when sea temperatures are falling. DISCUSSION REPRODUCTIVE BIOLOGY. Sponges can be either gonochoric or hermaphroditic, oviparous or viviparous. In Haplosclerida both gonochorism and hermaphroditism have been reported, although all species examined to date have been viviparous. Tropical species of Haliclona from the Great Barrier Reef (H. amboinensis and H. cymaeformis; Fromont, 1994), and temperate intertidal species from the Oregon coast (H. permollis; Elvin, 1976), and from a Connecticut estuary (H. loosanoffi; Fell, 1976) are all viviparous and gonochoric. However, two species from a temperate reproductive products in species of the haplosclerid families Chalinidae and Niphatidae, and suggest these brooding chambers may be common amongst haplosclerids. In this study Haliclona sp. did not have brood chambers but had reproductive products aligned along the midline of the branches. Mycale sp. conformed to previous reports of sex determination and reproductive mode for this genus, being viviparous and hermaphroditic. The temperate European species, Mycale micracanthoxea, the tropical Red Sea species, M. fistulifera, and Mycale sp. reported here are all viviparous and contemporaneous hermaphrodites (Wapstra & Van Soest, 1987, Meroz & Ilan, 1995, Reiswig, 1973). There are no previous reports on mode of reproduction or sex determination in species of the genus Coelosphaera or Echinodictyum. Chondrilla australiensis from South Mole was oviparous and probably gonochoric conforming to published reports of the reproductive biology of this genus. Chondrilla nucula trom Italy has been reported to be oviparous and gonochoric (Liaci et al., 1971a). Sponges possess high regenerative capacities and have been reported to reproduce asexually by budding, fragmentation and gemmulation (Fell, 1993). Fragmentation is possible because of the structural homogeneity and morphological flexibility of sponges so that even small fragments are likely to possess all essential functional elements and can readily reorganise to function as independent entities (Wulff, 1991). The droplets of tissue I found extending from the edges of some adults of C. australiensis in April 1998 appear to be a form of asexual reproduction through fragmentation. In Chondrilla nucula cellular extensions between the mesohyl and oocytes are described as long thin filipodia connecting the nurse cells surrounding the developing eggs to the egg 190 MEMOIRS OF THE QUEENSLAND MUSEUM COLOUR VARIATION australiensis Echinodictyum clathrioides Sea Temperature °C Month FIG. 2. Reproductive activity and sea temperature. A, Timing of reproductive activity, X = months when the species had reproductive products. B, Average monthly sea temperatures. surfaces, and Liaci et al. (1971a) suggested that these connections allow a direct, nonphagocytic transfer of nutrients. Similarly, the cellular extensions 1 observed surrounding eggs in Chondrilla australiensis are likely to have the same function. Tethya sp. individuals were oviparous and probably gonochoric, conforming to previous reports in the literature for both tropical species T. crypta (Reiswig, 1973), and temperate species T. aurantium and T. citrina (Liaci et al., 1971b). Asexual budding has been reported for species of Tethya. Thin filaments containing spicules extend outwards from the adult and a spherical bud forms distally. This bud detaches from the adult and can attach to the substratum (Simpson, 1984). Thin filaments were seen extending from one individual of Tethya sp. in April 1998 but distal buds were not apparent at this time. WITHIN SPECIES. Chondrilla australiensis and Tethya sp. each had two distinctive colour morphotypes at the study site. In C. australiensis the usual occurrence of the maroon morph in shaded or cave habitats suggests that its colour difference to the ochre morph, growing in full light, may be a response to reduced light conditions. The Northern hemisphere species C. nucula usually colonises illuminated bottoms and is generally brownish (Gaino et al., 1976). Arillo et al. (1993) found that its colouration is a consequence of the presence of the cyanobacteria Aphanocapsa sp. Similarly, it is speculated that colour differences between morphs at South Mole could be the result of either different cyanobacterial symbionts within the two morphs, or differing abundances of the symbionts in the two morphs. Individuals of both morphs occurred side by side with marked non-overlap zones between them. These zones were also common between specimens of the same colour morph, suggesting the occurrence of different genotypes within the same colour morphs, and that sexual reproduction is occurring to some extent in the population. In Tethya sp. individuals of each colour morph occurred side by side in full light and under Ecklonia. Therefore, colour differences between the morphs cannot be attributed to differences in light regimes. The coexistence of more than one species in a restricted area has been found previously in the genus Tethya (Sara et al., 1993), and similar analyses of genetic data and niche differentiation of the Tethya species at South Mole may find these colour morphs to also be distinctive at the species level. TIMING OF REPRODUCTIVE DEVELOPMENT. The timing of reproductive activity in sponges has previously been related to sea temperature, with many species found to REPRODUCTION IN WESTERN AUSTRALIAN DEMOSPONGES initiate activity as sea temperatures increase (Simpson, 1984 and references therein). Fewer studies have found sponges to be reproductively active as temperatures fall or are at a minimum. In the present study most sponge species were reproductive as sea temperatures were increasing or reaching their summer maximum (i.e. Tethya sp., Chondrilla australiensis, Mycale sp. and possibly Echinodictyum clathrioides). At this stage there is not enough information about the reproductive activity of E. clathrioides to say with certainty when release of reproductive products occurs. In contrast, two species commenced reproductive activity as sea temper- atures were decreasing (i.e. Coelosphaera sp. and Haliclona sp.). Haliclona sp. appears to develop embryos and larvae throughout the winter. Light regimes are another environmental factor that could influence onset of reproductive activity in sponges. Elvin (1976) found that initiation of oogenesis in the temperate intertidal species Haliclona permollis was most closely related to an increase in incident light. Ilan & Loya (1990) speculated that the reproductive activity of Niphates sp. may be related to the seasonal disappearance of algae, thereby increasing incident light to the sponges. At South Mole, biomass of the Eck/onia fringe appears to increase during the summer months when the photoperiod has increased, and may therefore increase shading of sponges. A third exogenous factor implicated in the timing of reproductive activity is food avail- ability (Sara, 1992). The occurrence of two different periods of reproductive activity of sponges at South Mole may coincide with two peaks in the abundance of ultraplankton (A. Pile, Flinders University of South Australia, pers.comm.), one peak occurring in late summer/ autumn when most of the species release their products, and a second peak in spring when reproductive products are released by Haliclona sp. Two explanations are possible to explain these opposing trends in timing of reproductive activity: 1) species are responding to different environmental cues which trigger initiation of reproductive activity, or 2) species are responding differently to the same environmental cues. SUMMARY This study shows that modes of reproduction in sponges from South Mole, southern WA, 191 conform to modes already documented in the literature for these respective genera. More work is required to unequivocally determine sex phenotype of these species, but preliminary data indicate that this aspect of their reproductive biology also appears to conform to the majority of reports in the literature. Reproductive activity in two of the species in autumn and winter is unusual, and possible reasons for this require further investigation. For the present, a baseline has been established in the timing of activity which will support future investigations on many other aspects of the reproductive biology of these species. Larval biology, diurnal timing of spawning for species that broadcast their products, and analysis of the partitioning of resources in species known to have both sexual and asexual reproduction, would all be useful studies. Species with colour morphs should be examined using genetic methods, or monitored for reproductive isolating mechanisms, to establish whether they are distinct species occurring sympatrically at the study site or if some other factors are responsible for their observed differences. ACKNOWLEDGEMENTS I am particularly grateful to Christine Hass, University of Western Australia for acting as a dive buddy throughout this study. Kelley Whittaker, University of Western Australia provided field assistance. Many thanks to Dr. L. Matz, Mr. Bruce Beaney and staff of the Histopathology Lab. of Royal Perth Hospital for allowing me to use their microtomes. The following WAM staff provided assistance: Mark Salotti provided photographic and technical assistance and Alex Bevan provided photo- microscopy equipment. Mark Rossbach of Fisheries Western Australia generously provided temperature data, and Alan Pearce CSIRO provided helpful discussions on sea temperatures in Perth coastal waters. I am very grateful to the Australian Biological Resources Study (ABRS) for providing funding that enabled me to travel to the 51h International Sponge Symposium in Brisbane, and for funding this study which was carried out at the Western Australia Museum, Perth. LITERATURE CITED ARILLO, A., BAVESTRELLO, G., BURLANDO, B. & SARA, M. 1993. Metabolic integration between symbiotic cyanobacteria and sponges: a 192 possible mechanism. Marine Biology 117: 159-162. CHEN, W.T. 1976. Reproduction and speciation in Halisarca. Pp. 113-139. In Harrison, F.W. & Cowden, R.R. (eds) Aspects of Sponge Biology. (Academic Press: New York). ELVIN, D.W. 1976. Seasonal growth and reproduction of an intertidal sponge, Haliclona permollis (Bowerbank). Biological Bulletin 151: 108-125. FELL, P.E. 1976. The reproduction of Haliclona loosanoffi and its apparent relationship to water temperature. Biological Bulletin 150: 200-210. 1983. Porifera. Pp. 1-29. In Adyodi, K.G. & Adyodi, R.G. (eds) Reproductive Biology of Invertebrates. Volume 1. Oogenesis, oviposition, and oosorption. (John Wiley & Sons: Chichester). M 1993, Porifera. Pp. 1-44. In Adyodi, K.G. & Adyodi, R.G. (eds). Reproductive Biology of Invertebrates. Vol. 6A. Asexual Propagation and Reproductive Strategies. (John Wiley & Sons: Chichester). FROMONT, J. 1994, Reproductive development and timing of tropical sponges (Order Haplosclerida) from the Great Barrier Reef, Australia. Coral Reefs 13: 127-133. GAINO, E., PANSINI, M. & PRONZATO, R. 1976. Osservazioni sull'associazione tra una cianoficea croococcale e la demospongia Chondrilla nucula. Arhives Oceanography and Limnology (Supplement 18)3: 545-552. HOSKINS, R.V. 1992. Reproduction in two species of Phyllospongia (Demospongiae). Unpublished Honours Thesis (Murdoch University: Western Australia). ILAN, M. & LOYA, Y. 1990. Sexual reproduction and settlement of the coral reef sponge Chalinula sp. . from the Red Sea. Marine Biology 105: 25-31. LEVI, C. 1956. Etude des Halisarca de Roscoff. Embryologie et systematique des demosponges. MEMOIRS OF THE QUEENSLAND MUSEUM Archives Zoologie Experiental and Generale 93: 1-181. LIACI, L. SCALERA-, SCISCIOLI, M. & MATARRESE, A. 1971a. La riproduzione sessuale di alcuni tetractinomorpha (Porifera). Atti della Societa Peloritana di Scienze Fisiche Matematiche e Naturali 17: 235-245. LIACI, L. SCALERA-, SCISCIOLI, M., PAPA, O. & LEPORE, E. 1971b. Raffronto tra i cicli sessuali di Tethya aurantium (Pallas) Gray e Tethya citrina Sarà, Melone (Porifera, Hadromerina). Atti della Societa Peloritana di Scienze Fisiche Matematiche e Naturali 17: 287-298. MEROZ, E. & ILAN, M. 1995, Life history characteristics of a coral reef sponge. Marine Biology 124: 443-451. REISWIG, H.M. 1973. Population dynamics of three Jamaican Demospongiae. Bulletin of Marine Science 23(2): 191-226. SARA, M. 1992. Porifera. Pp. 1-29. In Adyodi, K.G. & Adyodi, R.G. (eds) Reproductive Biology of Invertebrates. Volume 5. Sexual differentiation and behaviour. (John Wiley & Sons: Chichester). SARA, M., CORRIERO, G. & BAVESTRELLO, G. 1993, Tethya (Porifera: Demospongiae) species coexisting in a Maldivian coral reef lagoon: taxonomical, genetic and ecological data. P.S.Z.N.I: Marine Ecology 14(4): 341-355. SIMPSON, T.L. 1984. The Cell Biology of Sponges. (Springer-Verlag: New York). WAPSTRA, M. & SOEST R.W.M. VAN 1987. Sexual reproduction, larval morphology and behaviour in Demosponges from the Southwest of the Netherlands. Pp. 281-307. In Vacelet, J. & Boury-Esnault, N. (eds) Taxonomy of Porifera. (Springer-Verlag: Berlin, Heidelberg). WULFF, J.L. 1991. Asexual fragmentation, genotype success, and population dynamics of erect branching sponges. Journal of Experimental Marine Biology and Ecology 149: 227-247. MEMBRANE-BOUNDED NUCLEAR BODIES IN A DIVERSE RANGE OF MICROBIAL SYMBIONTS OF GREAT BARRIER REEF SPONGES JOHN A. FUERST, RICHARD I. WEBB, MARY J. GARSON, LANI HARDY AND HENRY M. REIS WIG Fuerst, J.A., Webb, R L, Garson, M.J., Hardy, L. & Reiswig, H.M. 1999 06 30: Mem- brane-bounded nuclear bodies in a diverse range of microbial symbionts of Great Barrier Reef sponges. Memoirs of the Queensland Museum 44: 193-203. Brisbane. ISSN 0079-8835. Thin sections of chemically fixed tissue of several sponge species collected from Heron Island, Great Barrier Reef, including Jaspis stellifera, Pseudoceratina crassa and Axinyssa sp., were examined to investigate the cell organisation of bacteria-like microbial symbionts present. Such symbionts have been observed in these sponges to occur as a diverse range of morphotypes based on cell shape and cell wall type. A variety of different symbiont morphotypes were found to possess a membrane-bounded nucleoid, a feature not expected in prokaryotes. These had been previously observed by us in one symbiont morphotype in the Micronesian coralline sponges Stromatospongia micronesica and Astrosclera willeyana. Several distinct microbial morphotypes containing membrane-bounded nuclear bodies were observed in Great Barrier Reef sponges, only one of which resembled the type which we have previously observed in the two Micronesian sponges. In all these forms, the fibrillar nucleoid was surrounded by a single bilayer membrane, in most morphotypes defining a compartment also containing electron-dense particles resembling ribosomes or other nucleoplasmic pre-ribosomal material; such material was sometimes less dense and sometimes more dense than the cytoplasmic particulate material. Cell wall structure of the morphotypes broadly included both Gram-negative, outer membrane-bounded types, as well as a clear subunit S-layer type structure resembling that of known Archaea including crenarcheotes. Cytoplasmic membranes can be clearly seen in some cases as distinct from nuclear body membranes, excluding plasmolysis as an explanation for membrane- boundedness of nuclear bodies. The phylogenetic relationships of these microbes may be diverse if reflecting wall type, but at least some appear to be most likely to represent members of the Domain Archaea, perhaps resembling the crenarcheote Cenarchaeum symbiosum described from North American Axinella sp. O Porifera, Bacteria, Archaea, nucleoids, membrane-bounded, sponge symbionts, electron microscopy, ultrastructure. John A. Fuerst (email: fuerst@biosci.ug.edu.au), Department of Microbiology, The University of Queensland, St Lucia 4072, Australia; Richard I. Webb & Lani Hardy, Department of Microbiology and Centre for Microscopy and Microanalysis, The University of Queensland, St Lucia 4072, Australia; Mary J. Garson, Department of Chemistry, The University of Queensland, St Lucia 4072, Australia; Henry M. Reiswig, Redpath Museum and Biology Department, McGill University, Montreal, Quebec, Canada; 7 January 1999. There are 2 known major types of cell organisation, prokaryotic where the DNA of the genome is free in the cytoplasm and not confined to a special compartment, and the eukaryotic, where the genomic DNA is confined to a double membrane-bounded organelle, the nucleus, and in addition any other of several double-membrane-bounded organelles such as mitochondria and chloroplasts may be present (but not in all eukaryotes, e.g., archezoan protozoa such as Giardia). In the prokaryote the naked genomic DNA in chemically fixed cells often appears to be folded or condensed into a fibrillar structure and this ultrastructural entity is termed the ‘nucleoid’. The prokaryotic form is characteristic of most known species within two of the three great Domains of life defined by contemporary molecular systematics, the Bacteria and the Archaea, while the eukaryotic form is known so far only within the Domain Eucarya and not in the other two Domains (Woese et al., 1990). Several questions about such a classification of cell organisation can be posed, however. Are these the only forms of cell organisation which have evolved, or might there not be intermediate forms or even more complex ones, hitherto undiscovered due to our limited knowledge of biodiversity? Related to this is a second question- are membrane-bounded nuclei or their analogues 194 exclusive to the Domain Eucarya, or might they or some analogous form of organelle occur in those two Domains of life thought to harbour only prokaryotic cells? The first indication that there might be alternative forms of cell organisation to those classical known ones was discovered ina distinct division or phylum of the Bacteria, the plancto- mycetes (Order Planctomycetales), where one species, Gemmata obscuriglobus, possesses a genome bounded by two membranes (Fuerst & Webb, 1991) while in another two, Pirellula marina and Pirellula staleyi, a single membrane separates the compartment containing the genomic DNA from the rest of the cell (Lindsay et al., 1997). We present here evidence that several distinct morphotypes of sponge symbionts (only one of which has been described by us previously; see Fuerst et al., 1998), reveal further examples of structurally novel types of cell organisation in which the genomic DNA appears compartmentalised by a single membrane from the rest of the cell cytoplasm, and that these may occur in microorganisms resembling members of the Domain Archaea, and present new data to support these findings. MATERIALS AND METHODS Stromatospongia micronesica and Astrosclera willeyana were collected from Guam (Micronesia), and Pseudoceratina crassa, Jaspis stellifera and Axinyssa sp were collected from Heron Island (Great Barrier Reef), at sites previously described in detail (Fuerst et al., 1998). Sponge tissue samples were chemically fixed with glutaraldehyde followed by osmium tetroxide before resin embedding and thin sectioning, via protocols including hydrofluoric acid treatment of either tissue blocks or Epon resin-embedded block faces, and uranyl acetate- lead citrate stained thin sections were examined via transmission electron microscopy, as described previously (Fuerst et al., 1998). For immunolabelling experiments, samples of Jaspis stellifera were fixed in 2% glutaraldehyde-4% paraformaldehyde fixative in 0.1M cacodylate buffer, and further processed as described in Fuerst et al. (1998). Immunolabelling of thin sections was via use of an anti-ds+ss DNA antibody (Boehringer-Mannheim) and goat anti- mouse [gM conjugated to either 10nm or 5nm colloidal gold, as described previously (Lindsay et al., 1997), and sections were then stained with uranyl acetate and lead citrate. Labelling of MEMOIRS OF THE QUEENSLAND MUSEUM sections to localise RNA using an RNase- colloidal gold (10nm) conjugate was performed essentially as described in Lindsay et al. (1997), followed by staining as above. In cases where double labelling was performed, RNase gold labelling was performed first, followed by anti- DNA gold immunolabelling. Goat anti-mouse IgM conjugated to 5 nm colloidal gold was used for labelling of DNA when double-labelling experiments involving both DNA and RNA labelling were performed. Voucher samples of Axinyssa sp. nov. and Pseudoceratina crassa are held at the Queensland Museum as QMG312575 and QMG304915 respectively (identified by Dr John Hooper). RESULTS AND DISCUSSION Bacteria-like symbionts of varying morpho- type were common in mesohyl of the tissue of the sponge species collected from Heron Island, including Jaspis, Pseudoceratina and Axinyssa spp. (see Fig. | A). These associates were found to include a diversity of morphotypes displaying a novel form of compartmentation, in which either the fibrillar nucleoid representing the prokaryote chromosomal DNA was surrounded by a single membrane, or , in one type, where an inverse of this compartmentation topology was displayed. In the latter type, a single membrane separates one organelle-like region of the cytoplasm from a second region of cytoplasm containing the nucleoid. The compartmented morphotype in which the nucleoid is bounded by asingle membrane has been previously described by us in sponges from Pacific Micronesia, Stromatospongia micronesica and Astrosclera willeyana (Fuerst et al., 1998). The dramatic distinction ofthis morphotype from bacteria with a classical prokaryotic ‘naked’ fibrillar nucleoid, free in the cytoplasm, is seen in Figure 1B of the mesohyl of S. micronesica. This first kind of compartmented morphotype, here seen in an actively dividing cell, displays a characteristically “butterfly” shaped nuclear body surrounded by a single membrane. The cell wall structure is of a regular subunit type consistent with membership of the Domain Archaea (see Fuerst et al., 1998, and the discussion of this wall type in other morphotypes below). This morphotype is typical ofall the nucleated symbiont morphotypes, in the consistency of features such as the cell size, the absence of membrane-bounded organelles other than the nuclear body, and cell wall structure, NUCLEAR BODIES IN SPONGE SYMBIONTS 195 FIG. 1. A, Electron micrograph of thin section of mesohyl from sponge tissue of Jaspis stellifera from the Great Barrier Reef, with a diverse range of symbiont morphotypes apparent, based on a combined consideration of cell size, shape and internal structure. For example, the large arrow indicates a Morphotype 1 cell and the small arrow indicates a Morphotype 2 cell. (Scale bar 11m). B, Electron micrograph of thin-section of sponge tissue from Stromatospongia micronesica showing two symbiont morphotypes of contrasting internal structure. One morphotype (Morphotype | in the typing scheme used in this paper), the uppermost cell in this figure, possesses a membrane-bounded nuclear body region in a dividing cell (evidence for active viable cell growth of this type in the tissue) and the second morphotype is a cell with normal bacterial (prokaryotic) ultrastructure with fibrillar nucleoid DNA free in the cytoplasm. Note in the Morphotype | cell that the nuclear body in each cell half displays an outer electron dense region as well as a central fibrillar nucleoid region. (Scale bar 1 um). C, Electron micrograph of thin-sectioned Morphotype 2 (short fat rod) symbiont from Pseudoceratina crassa from the Great Barrier Reef; note the membrane-bounded nuclear body and that this type has a relatively electron-dense cytoplasm external to the nuclear body compared with other morphotypes (e.g., Morphotype 1 seen in Fig. 1B). (Scale bar 200nm). D, Electron micrograph of thin-sectioned Morphotype 3 (D-shaped cell) symbiont from Pseudoceratina crassa. This D-shaped cell has a clear membrane-bounded nuclear region as well as radiating fibres in the cytoplasm outside the nuclear region. The cell wall displays a regular subunit periodic structure (arrow). (Scale bar 200nm). 196 with most probable phylogenetic relationships of these symbionts to non-eukaryote organisms such as Bacteria or the Archaea. This morphotype with ‘butterfly’ nuclear body has now been found also in sponges from the Great Barrier Reef, including Jaspis stellifera, Pseudoceratina crassa and Axinyssa sp. This morphotype has thus now been found to be distributed among at least 5 different sponge genera (Stromatospongia, Astrosclera, Jaspis, Pseudoceratina and Axinyssa) and in at least two different geographical locations in the Western Pacific. However, not only was this morphotype found in both Great Barrier Reef and Micronesian sponges, but it turns out to constitute only one of several different morphotypes which display membrane-bounded nuclear regions, or at least membrane-bounded compartments, separating the cell interior into a nucleoid- and non-nucleoid-containing region. These compartmented cell symbiont morpho- types could be distinguished from each other on the basis of the following criteria or combinations of such criteria; cell wall structure, cell shape, texture (fine structure) of the cytoplasm outside the nuclear body, and type of cell compartment (e.g. nucleoid-containing versus nucleoid- devoid, or butterfly-lobed versus round in outline). In most of these morphotypes, the nucleoid is surrounded by a single membrane separating the nuclear region from the rest of the cell, as exemplified most dramatically in the type with butterfly-shaped nuclear body found first in S. micronesica and A. willeyana. In one morpho- type only, the nucleoid appears external to a central single-membrane-bounded compartment effectively separating the cell into two compartments, one with nucleoid and the other without. In the morphotype classification used in this paper, Morphotype 1 is considered to be the type with a butterfly-shaped membrane-bounded nuclear region. The second type, referred to as Morphotype 2, is a short, fat, rod morphotype (Fig. 1C). In Morphotype 2, the nuclear region is bounded by a single membrane (Fig. 3A), but has a relatively electron-dense cytoplasm compared with Morphotype 1. Morphotype 3 is a D-shaped cell; this type not only shows a characteristic cell shape and a very clear membrane-bounded nuclear region but also displays radiating fibres in the cytoplasm outside this region (Fig. 1D). Most interesting of all, the cell wall displays a structure of regular subunits, compatible with possible membership of the MEMOIRS OF THE QUEENSLAND MUSEUM Domain Archaea (Fig. 1D). Members of the Archaea have been classically considered to include organisms inhabiting very hot hydrothermal and volcanically heated waters as well as methane-generating anaerobes and organisms growing in saturated salt (Woese et al, 1990). However, recently they have been reported from less extreme marine habitats (e.g., Atlantic, Pacific and Antarctic seawater; see DeLong, 1992, DeLong et al., 1994 and Fuhrman et al., 1992), including, very significantly for this context, a report of the cold water archaeon sponge symbiont Cenarchaeum symbiosum from a Californian coastal Axinella species, confirmed as an archaeon by gene probing (Preston et al., 1996). These Archaea all belong to the so-called marine group I cluster which is part of the Kingdom Crenarcheota within the Domain Archaea (DeLong, 1992). In addition to their occurrence in a sponge, such bacteria have also been recently found to occur in holothurian gut and marine fish gut (McInerney et al., 1995; Van Der Maarel et al., 1998). Another morphotype (Morphotype 4) displays a cell wall with a structure consistent with Gram- negative bacteria - those known to give anegative Gram stain reaction. The wavy nature ofthe outer cell wall membrane is very clear, consistent with Gram-negative type cell wall; in this type the DNA fibrils occupying most of the nuclear body are very obvious and enclosed within a very distinct membrane (Fig. 2B). Morphotype 5 is another morphotype with subunit cell wall consistent with membership of the Archaea, this time a rod without a D-shape bias (Fig. 2B-C). Note also the regular subunit 2-D lattice structure visible in the grazing section portion of wall (arrow in Fig. 2B). Morphotype 6 exhibits a characteristically blebbed cell wall membrane, containing a membrane-bounded internal body but in this case without a nucleoid - the nucleoid is in the other ‘cytoplasmic’ cell compartment (Fig. 2D). Thus the cell is still divided into a non-nucleoid-containing and a nucleoid-containing compartment, albeit in a reverse topological sense to that found in the other morphotypes with a membrane-bounded internal or centrally located nucleoid-containing compartment. To summarise the symbiont morphotypes described above, there are six morphotypes occurring in Great Barrier Reef sponges J. stellifera, P. crassa and Axinyssa sp. The first five are Morphotype 1 (rods with "butterfly" nuclear NUCLEAR BODIES IN SPONGE SYMBIONTS 197 FIG. 2. A, Electron micrograph of thin-sectioned Morphotype 4 (Gram-negative walled cell) symbiont from Pseudoceratina crassa displaying a wavy outer cell wall membrane similar to that in walls of Gram-negative bacteria, as well as a single-membrane-bounded nuclear region with DNA fibrils occupying most of the nuclear body. (Scale bar 100nm). B, Electron micrograph of thin sectioned Morphotype 5 (regular subunit walled normal rod) symbiont from Pseudoceratina crassa displaying a cell wall consisting of regular subunits (small arrow), especially visible as a periodic lattice in a portion in which a grazing section has occurred (large arrow). (Scale bar 200nm). C, Enlargement of thin sectioned Morphotype 5 (regular subunit walled normal rod) symbiont from Pseudoceratina crassa shown in Fig. 2B displaying a cell wall consisting of regular subunits. Note also the portion of the nuclear body and clearly displayed single membrane envelope of this body towards lower right hand side of figure. (Scale bar 100nm). D, Electron micrograph of thin sectioned Morphotype 6 symbiont from Pseudoceratina crassa displaying a characteristic blebbed cell wall outer membrane. A membrane-bounded internal body (arrow) is present but the nucleoid (N) is situated in the cell compartment external to this inner body, in contrast to all other morphotypes described in these figures. (Scale bar 200nm). 198 bodies); Morphotype 2 (short fat rods with electron-dense cytoplasm); Morphotype 3 (D-shaped cells with subunit (“archaeal”) walls); Morphotype 4 (rods with Gram-negative outer membrane walls); and Morphotype 5 (cells with large subunit (‘archaeal’) walls). All of these 5 types contain a membrane-bounded nucleoid- containing nuclear body, more or less centrally located within the cell. Note again that ‘nucleoid’ is here being used in the bacteriological sense of a fibrillar genomic DNA bundle, which is not normally membrane-bounded. Morphotype 6 differs from the first five. It consists of rods with blebbed cell wall membrane but with a membrane-bounded internal compartment without nucleoid (Fig. 3A-B). The cell is effectively divided into two compartments, however, in an analogous manner to the compart- mentalisation in the first five morphotypes, but with a reversed topology. It should be noted that every morphotype has been seen in all these Great Barrier Reef sponges examined (P. crassa, J. stellifera and Axinyssa sp.). Thus the diversity of morphotypes possess- ing membrane-bounded nuclear bodies we have seen may be a widely distributed phenomenon unrelated to host specificity. The morphotypes described here appear to be sub-types of the types E and 4, previously described in published studies (Vacelet, 1975; Wilkinson, 1978). In those studies, the appearance oftypes E and 4 was explained by the occurrence in these bacteria of an unusually large periplasm, that is, the region between cell wall and cytoplasmic membrane (Vacelet, 1975; Wilkinson, 1978). In this interpretation, the membrane-bounded nuclear body we have seen would merely represent the cell protoplast (cell cytoplasm contents) surrounded by a retracted cytoplasmic membrane, and the space between nuclear body membrane and cell wall would represent a very wide ‘periplasm’ rather than true cytoplasm. We favour an alternative inter- pretation, in which the membrane bounding the nuclear region is a true internal membrane rather than representing cytoplasmic membrane, which is closely appressed to the cell wall in the symbionts we have observed and therefore often difficult to detect. Supporting this is the clear indication of cytoplasmic membrane for at least three Morphotypes as shown in Morphotypes 2 and 3 (Fig. 3A-B) and Morphotype 1 (Fig. 2c in Fuerst et al., 1998). Also consistent with this interpretation in Morphotype 1, and several other ofthe Morphotypes, is the uniform distribution of MEMOIRS OF THE QUEENSLAND MUSEUM cytoplasm within the space between the cell wall and internal nuclear body-bounding membrane, a distribution unlikely if plasmolysis and retraction of cytoplasmic membrane were responsible for this space. To investigate this problem further, we employed immunogold labelling methods to localise DNA and RNA within the cell and thus determine the location of true cytoplasm. Figure 3C shows a nucleated symbiont Morphotype 1 cell from J. stellifera in which the DNA has been localised via immunogold labelling using mouse monoclonal antibody against single-stranded and double-stranded DNA detected via goat anti-mouse antibody conjugated to 10nm colloidal gold particles. All the cell’s DNA is localised exclusively within the membrane- bounded nuclear body, suggesting that this must be the location of the chromosomal, genomic DNA. Intracellular RNA was localised in this morphotype using the slightly different enzyme cytochemistry approach employing RNase conjugated directly to colloidal gold. By this method, RNA in these cells is located throughout the cytoplasmic region external to the membrane-bounded nuclear body (Fig. 3D), as well as being present to minor extent in the nuclear body, as would be expected if transcr- iption is to occur using a genomic DNA template. This occurrence of RNA in the cytoplasm outside the membrane-bounded nuclear region supports an interpretation of symbiont ultrastructure in which the space between the nuclear body membrane and the cell periphery is occupied by true cytoplasm and is not merely an unusually large periplasm between a retracted cytoplasmic membrane and the cell wall, and in which the nuclear body is thus a true intracytoplasmic membrane-bounded compartment of the cell. Some RNA also appears in the electron- dense-particle-rich outer zone within the nuclear body itself, as would be expected if cell RNA is transcribed from DNA in the nuclear body. Double-labelling using both DNA and RNA labelling methods with differently sized gold particles confirms the distribution of DNA and RNA relative to the nuclear body found via separate use of DNA and RNA labels (Fig. 4A) Gold labelling was also used to examine the problem posed by Morphotype 6 symbionts, where there appears to be a non-nucleoid- containing internal membrane-bounded body. A combination of RNase-gold and anti-ss and dsDNA antibody immunogold labelling demonstrated that most of the cell RNA appeared NUCLEAR BODIES IN SPONGE SYMBIONTS 199 FIG. 3. A-B, Electron micrographs of thin-sectioned Morphotypes 2 and 3. A, Morphotype 2 cell (same cell as shown in Fig. 1C) showing a clear cytoplasmic membrane (arrow) adjacent to the cell wall and external to and widely separated by electron-dense cytoplasm from the nuclear body membrane. (Scale bar 100nm). B, Morphotype 3 showing a cytoplasmic membrane (arrow) closely appressed to the cell wall. (Scale bar 100nm). C, Electron micrograph of thin-sectioned nucleated symbiont Morphotype 1 from Jaspis stellifera showing labelling of DNA only within nuclear body, via immunogold detection of mouse monoclonal antibody against single stranded and double stranded DNA (10nm colloidal gold particles). (Scale bar 200nm). D, Electron micrograph of thin-sectioned nucleated symbiont Morphotype 1 from Jaspis stellifera showing location of intracellular RNA via labelling with RNase-gold. Note the absence of labelling over the central nucleoid. (Scale bar 200nm). 200 to be confined to the membrane-bounded internal body, and that all the cell DNA was found outside that body (Fig. 13). Although there is a reverse topology to the compartmentation of DNA found in the other symbiont morphotypes, it would appear that the cell’s DNA is still restricted to a separate compartment within the cell as also occurs in the other Morphotypes, via a different compartment. The inner compartment of Morphotype 6 appears superficially to be similar structurally to the nuclear body in the other morphotypes, in the sense of being a single- membrane-bounded inner compartment, but in this case is devoid of DNA. In their possession of a membrane-bounded nucleoid, the internal organisation of the sponge symbiont Morphotypes 1-5 described here contrasts with that known for most members of Domains Bacteria and Archaca. However, it is most similar to that found previously in plancto- mycete species of the genus Pirellula, where there is enclosure of a major nucleoid-containing cell compartment, the pirellulosome, by a single membrane dividing the cell interior into two regions and where a zone of electron-dense particles around the nucleoid occurs within the pirellulosome (Lindsay et al., 1997). Also relevant may be the double membrane-bounded nucleoid compartment found in another plancto- mycete, Gemmata obscuriglobus (Fuerst & Webb, 1991). In contrast to the cell structure in Pirellula species, the sponge symbionts described here do not display any polar differen- tiation or ‘polar cap’. In natural habitat samples, the only bacterial cells appearing to show any similar ultrastructure to those described here are 0.3-0.5um diameter cells from soil, described with an internal membrane surrounding the nuclear material (Bae & Casida, 1973). There are also significant similarities, concerning cell shape and nuclear body shape during division, between the symbionts observed here and cells of Cenarchaeum symbiosum, a symbiont of the sponge Axinella sp. from the Californian coast of the Pacific Ocean determined by in situ hybridisation with oligo- nucleotide probes to be a member of the kingdom Crenarcheota of Domain Archaea (Preston et al., 1996). If the sponge symbionts exhibiting membrane-bounded nucleoids that we have observed prave to be related closely to C. symbiosum, this would be highly significant from an evolutionary perspective. This is because cell organisation in Domain Archaea is thought to be MEMOIRS OF THE QUEENSLAND MUSEUM exclusively prokaryotic, even though there are many molecular and phylogenetic similarities with eukaryotes and Domain Eucarya (Keeling et al., 1995), and because a membrane-bounded nucleus would have been demonstrated in all three Domains of Life, suggesting its possible status as an ancestral character of the last common ancestor of the three Domains retained only in some lineages of contemporary organisms. Cenarchaeum symbiosum has been determined by both 16S rRNA. sequencing and DNA polymerase sequencing (Preston et al, 1996; Schleper et al., 1997) to be a member ofthe Kingdom Crenarcheota within the Domain Archaea, and it may be relevant to the possible occurrence of nucleated cell organisation in sponge symbionts that a crenarcheote origin for the eukaryotes, or at least a crenarcheote/ eukaryote clade, has been suggested from phylogenetic analyses based on amino acid sequences from the highly conserved duplicated genes for protein synthesis elongation factors, EF-Tu and EF-G (Baldauf et al., 1996). Possible identities for the sponge symbionts with membrane-bounded nucleoids include those of a crenarcheotal Archaea member, a planctomycete member of the Bacteria, or a member of the Eucarya. If the latter, however, it must be a mitochondrion-less representative and one which has lost one membrane of the nuclear envelope. It seems most probable that at least some of the symbionts with a membrane- bounded nucleoid are members of the Archaea, since the cell wall in Morphotypes 1, 3 and 5 exhibit subunit structure consistent with an S-layer wall, the most common wall type in the Archaea (Kónig, 1994). The fluorescent probe- labelled symbionts in Axinella mexicana identified by Preston et al. (1996) are non-thermophilic members of the Crenarcheota within the Domain Archaea, and these show DNA-containing regions with similar morph- ology via fluorescence microscopy to those seen intherelevant symbionts studied here by electron microscopy. Archaeal nucleoids in the hyper- thermophilic crenarcheotes Sulfolobus acidocaldarius and Pyrodictium abyssi appear to be naked rather than membrane-bounded (Bohrmann et al., 1994; Rieger et al., 1995), but the mesophilic or even psychrophilic crenarcheotes, which have not been cultured or examined via electron microscopy, may well possess different internal structure from those hyperthermophilic representatives of the same Kingdom. Intracellular lamellar membranes NUCLEAR BODIES IN SPONGE SYMBIONTS ar ) Al FIG. 4. A, Electron micrograph of portion of a thin-sectioned nucleated symbiont Morphotype 1 from Jaspis stellifera which has been double-labelled for both DNA and RNA via use of different sizes of gold particle (10nm for RNase gold and 5nm for anti-DNA antibody labelling via goat anti-mouse IgM Ab conjugated to colloidal gold). (Scale bar 100nm). B, Electron micrograph of thin-sectioned Morphotype 6 symbiont from Jaspis stellifera labelled using RNase-gold (large 10nm gold particles) and anti-ss- and ds-DNA antibody immunogold (small dot-like 5nm gold particles) showing the occurrence of RNA but not DNA within the internal membrane-bounded body and the exclusive occurrence of DNA associated with fibrillar nucleoid (arrow) outside the inner membrane-bounded body. (Scale bar 200nm). found in Type I methanotroph-like Bacterial symbionts of deep-sea carnivorous sponges in methane-rich waters (Vacelet et al., 1996) do not appear to be similar to the membranes enclosing the nucleoids described above, which do not display membrane over-folding or multiple layering. Possible Domain membership of the symbionts can be resolved by direct probing of cells in sections using probes specific for 16S rRNA of specific Domains, or via cloning of PCR-amplified 168 rRNA genes from the symbiont community combined with hybrid- isation of sectioned or whole cells with probes designed from clone sequences. It can be predicted that the symbiont Morphotypes 1,3 and 5 with regular subunit walls should be found by such methods to be members of the Kingdom Crenarcheota within the Domain Archaea. CONCLUSIONS At least 6 morphotypes of bacteria-like symbionts in the mesohyl of the sponge genera Jaspis, Pseudoceratina, Axinyssa, found in the waters of Heron Island, Great Barrier Reef, possess membrane-bounded nuclear regions. In at least 2 of these morphotypes, cell walls composed of subunits are present, consistent with membership of the Archaeal Domain. In this context it 1s of great relevance and interest that members of the Domain Archaea belonging to the kingdom Crenarcheota have been found by direct gene probing using in situ fluorescent oligonucleotide hybridisation of whole cells to be present in sponges within genus Axinella (Preston et al., 1996). These associates or symbionts ofthis sponge have been referred to as Cenarchaeum symbiosum, and phylogenetic analysis using sequences from at least two genes from this species have confirmed membership of the Crenarcheota within the Domain Archaea (Preston et al., 1996; Schleper et al., 1997). In all but Morphotype 6, the blebbed cell wall morpho- type, all the cell DNA is in a nuclear body bounded by a single membrane, The extranuclear cytoplasm possesses most ofthe cell RNA (i.e., it does not appear to be periplasm but true cytoplasm). The diversity of morphotypes which differ in cell wall type, cell shape, cytoplasm texture and cell compartment type, yet all share compart- mentalisation of the cell into two compartments, a nucleoid-containing and a non-nucleoid- containing one, suggests either that somewhat phylogenetically diverse organisms are perhaps phylogenetically related to a common ancestor with a similar form of compartmentalisation which was retained in otherwise structurally diverse descendants, or alternatively that some environmental factor in the sponge tissue selects for or induces nuclear compartmentalisation. It is also possible that nucleated organisms, or organisms with correlated cell wall structure, are selected for by a sponge tissue factor such as a compound with antibiotic activity, for example, a compound inhibiting enzymes involved in DNA synthesis or supercoiling as performed in cells with prokaryote structure and naked chromo- somal DNA or one inhibiting synthesis of the peptidoglycan cell wall polymer found in most members of the Domain Bacteria. Classical prokaryotic Bacteria with peptidoglycan walls and naked cytoplasmic DNA may not compete efficiently with nucleated peptidoglycan-less organisms of whatever Domain. Aspects of these results are significant to our understanding of cell organisation and are fundamental to biology in general. Our results from sponge symbionts may constitute a challenge to the major structural classification of cell types based on cell organisation - that of the prokaryote and eukaryote- since at least some otherwise bacteria-like cell types appear to contain membrane-bounded DNA in a manner MEMOIRS OF THE QUEENSLAND MUSEUM analogous to eukaryote cell nuclei. This extends the challenge to that classification first revealed by the discovery of double- and single- membrane bounded nuclear bodies in the planctomycete members of the Division Bacteria. If the Archaeal nature of some of the sponge symbiont morphotypes with nuclear regions can be demonstrated by molecular sequence-based methods, then membrane- bounded nuclei may well be shown to occur at least rarely within members of all three Domains of life, Bacteria, Archaea and Eucarya, a finding which would be of fundamental significance to our understanding of how eukaryote cells may have evolved their initial defining structure. Insights from study of sponge biology may thus yet again contribute to our fundamental understanding of cell biology and evolution. ACKNOWLEDGMENTS We would liketo acknowledge funding support for the separate research programs of JAF and of MJG from the Australian Research Council, and thank Dr John Hooper for sponge identifications and voucher specimen storage for Great Barrier Reef material. LITERATURE CITED BAE, H.C. & CASIDA, L.E. JR 1973. Responses of indigenous microorganisms to incubation as viewed by transmission electron microscopy of cell thin sections. Journal of Bacteriology 113: 1462-1473. BALDAUF, S.L., PALMER, J.D. & DOOLITTLE, W.F. 1996, The root of the universal tree and the origin of eukaryotes based on elongation factor phylogeny. Proceedings ofthe National Academy of Sciences USA 93: 7749-7754. BOHRMANN, B., KELLENBERGER, E., ARNOLD- SCHULZ-GAHMEN, B., SREENIVAS, K., SURYANARAYANA, T., STROUP, D. & REEVE, J.N. 1994, Localization of histone-like proteins in thermophilic Archaea by immunogold electron microscopy. Journal of Structural Biology 112: 70-78. DE LONG, E.F. 1992, Archaea in coastal marine environments. Proceedings of the National Academy of Sciences USA 89: 5685-5689. DE LONG, E.F., YING WU, K., PREZELIN, B.B. & JOVINE, R.V.M. 1994, High abundance of Archaea in Antarctic marine picoplankton. Nature 371: 695-697. FUERST, J.A. & WEBB, R.I. 1991. Membrane- bounded nucleoid in the eubacterium Gemmata obscuriglobus. Proceedings of the National Academy of Sciences USA 88: 8184-8188. NUCLEAR BODIES IN SPONGE SYMBIONTS 20 FUERST, J.A., WEBB, R.I., GARSON, M.J., HARDY, L. & REISWIG, H.M. 1998. Membrane-bounded nucleoids in microbial symbionts of marine sponges. FEMS Microbiology Letters 166: 29-34. FUHRMAN, J.A., McCALLUM, K. & DAVIS, A.A. 1992. Novel major archaebacterial group from marine plankton. Nature 356: 148-149. KEELING, P.J. & DOOLITTLE, W.F. 1995. Archaea - narrowing the gap between prokaryotes and eukaryotes. Proceedings ofthe National Academy . OfSciences USA 92: 5761-5764. KONIG, H. 1994. Analysis of archaeal cell envelopes. Pp 85-119. In Goodfellow, M. & O'Donnell, A.G. (eds) ‘Chemical Methods in Prokaryote Systematics'. (Wiley: New York). LINDSAY, M.R., WEBB, R.I. & FUERST, J.A. 1997. Pirellulosomes: a new type of membrane- bounded cell compartment in planctomycete bacteria of the genus Pirellula. Microbiology 143:739-748. McINERNEY, J.O., WILKINSON, M., PATCHING, J.W., EMBLEY, T.M. & POWELL, R. 1995. Recovery and phylogenetic analysis of novel Archaeal rRNA sequences from a deep-sea deposit feeder. Applied and Environmental Microbiology 61:1646-1648. PRESTON, C.M., WU, K.Y., MOLINSKI, T.F. & DE LONG, E.F. 1996. A psychrophilic crenarchaeon inhabits a marine sponge: Cenarchaeum symbiosum gen.nov., sp.nov. Proceedings of the National Academy of Sciences USA 93: 6241-6246. U3 RIEGER, G., RACHEL, R., HERMANN, R. & STETTER, K.O. 1995. Ultrastructure of the hyperthermophilic archaeon Pyrodictium abyssi. Journal of Structural Biology 115: 78-87. SCHLEPER, C.A., SWANSON, R.V., MATHUR, E.J. & DE LONG, E.F. 1997. Characterization of a DNA polymerase from the uncultivated psychrophilic archaeon Cenarchaeum symbiosum. Journal of Bacteriology 179: 7803-7811. ., VACELET, J. 1975. Etude en microscopie électronique de l'association entre bactéries et spongaires du genre Verongia (Dictyoceratida). Journal de Microscopie et Biologie Cellulaire 23: 271-288. VACELET, J., FIALA-MEDIONI, A., FISHER, C.R. & BOURY-ESNAULT, N. 1996. Symbiosis between methane-oxidizing bacteria and a deep- sea carnivorous cladorhizid sponge. Marine Ecology Progress Series 145: 77-85. VAN DER MAAREL, M.J.E.C. 1998. Association of marine archaea with the digestive tracts of two marine fish species. Applied and Environmental. Microbiology 64: 2894-2898. WILKINSON, C. 1978. Microbial associations in sponges. III. Ultrastructure of the in situ associations in coral reef sponges. Marine Biology 49: 177-185. WOESE, C.R., KANDLER, O. & WHEELIS, M.L. 1990. Towards a natural system of organisms: proposal for the domains Archaea, Bacteria, and Eucarya. Proceedings ofthe National Academy of Sciences USA 87: 4576-4579. MEMOIRS OF THE QUEENSLAND MUSEUM EVIDENCE OF TRANSFER OF PHOTOSYNTHATE FROM A RED ALGAL MACROPHYTE TO ITS SYMBIOTIC SPONGE. Memoirs of the Queensland Museum 44: 204. 1999:- Symbiotic cyanobacteria are quite common in coral reef sponges providing much of the sponge’s supply of carbon. There are also several sponge species with macroalgal symbionts. In these sponges, the role of the algae is unknown. One of these symbioses is that of the sponge, Haliclona cymiformis (Haplosclerida) and the red alga, Ceratodictyon spongiosum (Rhodymeniales) which is common in the shallow tropical waters of fringing reefs of the Indo-Pacific region. The sponge tissue comprises about one third of the dry weight of the association and grows over the external surface of the alga and between the algal branchlets. In the field, the alga is dark green to purple with thick branches of tightly anastomosed (fused) branchlets. However, in culture, the branchlets are red and thin and do not fuse. Neither symbiont has been found growing separately in nature suggesting that the symbiosis is obligate. The physiological basis of this well integrated association is not yet known. The sponge obtains nutrients from the water column in the form of dissolved and particulate organic matter at rates that are similar to those of free-living sponges (Trautman, 1997). We have found that some photosynthate is transferred from the alga to the sponge, in a time-dependent manner. After 1h incubation in the light with Naz'*C0; the amount of photosynthetically fixed carbon transferred to the sponge (range 22.77- 48.3nmol carbon/mg dry wt. of sponge) represents 0.6-1.28% of the total carbon fixed by the alga during this period. When the fixed carbon in the sponge tissue is extracted using methanol/chloroform/water (24/10/4 v/v/v), to give an aqueous-soluble fraction (low molecular weight metabolites) and a chloroform-soluble fraction (lipids, sterols, chlorophyll etc.) followed by extraction in 2 M KOH (high molecular weight metabolites such as proteins, polynucleotides, polysaccharides) 75-88% of the "C -labelled carbon is found in the aqueous fraction, about 11-20% in the KOH-soluble fraction, 2-3% in the chloroform-soluble fraction and <3% in KOH- insoluble material. When the aqueous-soluble fraction is further fractionated by ion exchange chromatography into neutral (sugars), basic (amino acids), acidic (organic acids) and phosphate ester fractions, most of the fixed carbon is found in the basic (47%) and neutral (38%) fractions. Some fixed carbon is found in organic acids (14%) with very little in phosphate esters (<2%). Our data suggest that while the alga may supply the sponge with some essential nutrients, the major source of organic carbon is the particulate and dissolved organic matter in the ambient seawater. It may be that the primary role of the algal symbiont is structural rather than nutritional. CJ Porifera, symbiosis, red alga, carbon metabolism, photosynthate, translocation. Literature cited. TRAUTMAN, D. 1997. Aspects of the ecology and physiology of a tropical sponge and its macroalgal symbiont. PhD thesis (Murdoch University: Perth). A.J. Grant (email: agrant(a)bio.usyd.edu.au) & R.T. Hinde, School of Biological Sciences, University of Sydney, Sydney, N.S.W., 2006, Australia; M.A. Borowitzka, School of Biological Sciences and Biotechnology, Murdoch University, Perth, W.A., 6150, Australia; 1 June 1998. ECOLOGICAL ROLE OF CYTOTOXIC ALKALOIDS: HALICLONA N.SP., AN UN- USUAL SPONGE/ DINOFLAGELLATE ASSOCIATION MARY J. GARSON, RICHARD J. CLARK, RICHARD I. WEBB, KIM L. FIELD, ROMILA D. CHARAN AND ELIZABETH J. McCAFFREY Garson, M.J., Clark, R.J., Webb, R.I., Field, K.L., Charan, R.D. & McCaffrey, E.J. 1999 XX XX: Ecological role of cytotoxic alkaloids: Haliclona sp. nov., an unusual sponge/ dino- flagellate association. Memoirs of the Queensland Museum 44; 205-213. Brisbane. ISSN 0079-8835. Light microscopy and electron microscopy studies of the tropical marine sponge Haliclona sp. nov. (Haplosclerida; Chalinidae) from Heron Island, Great Barrier Reef, have previously revealed the characteristic presence of a dinoflagellate symbiont and nematocysts. The dinoflagellates are morphologically similar to Symbiodinium microadriaticum, the common intracellular zooxanthellar symbiont of corals. The sponge grows on coral substrates, from which it may acquire the dinoflagellates and nematocysts. Chemical investigations found the sponge contained a suite of cytotoxic alkaloids, the haliclonacyclamines. Our investigations showed that these alkaloid metabolites cause significant coral tissue necrosis at concentrations of Sppm after 160mins exposure in laboratory-based assays. At higher concentrations (10ppm and above) toxic effects were noted within 10mins exposure to the alkaloid fraction. Coral tissue necrosis was also observed after 40mins exposure to the major alkaloid component haliclonacyclamine A. In field experiments, the alkaloids were absorbed onto synthetic pads which were tied onto coral fingers. In both short term (10hrs) and long term (30hrs) experiments, coral tissue necrosis was observed at concentrations of 0.025% and above. We determined that a dose of 0.24% was equivalent to the natural exposure of coral pieces to sponge tissue, with our data indicating that haliclonacyclamines are effective toxins against coral tissue at lower than natural concentrations. When tested against natural populations of reef fish, the haliclonacyclamines were found to be potent feeding deterrents at ecologically-relevant concentrations (0.1% of sponge wet weight). O Porifera, Haliclona, Acropora, alkaloids, dinoflagellates, feeding deterrent, haliclonacyclamines, percoll density gradient fractionation, secondary metabolites, toxins, Symbiodinium microadriaticum. Mary J. Garson (email: garson@chemistry.uq.edu.au), Richard J. Clark, Kim L. Field & Romila D. Charan, Department of Chemistry, The University of Queensland, Brisbane Old 4072; Richard I. Webb, Centre for Microscopy and Microanalysis and Department of Microbiology, The University of Queensland, Brisbane Old 4072; Elizabeth J. McCaffrey; Department of Zoology, The University of Queensland, Brisbane Old 4072 Australia; 16 March 1999, Sponges are a major component of benthic fauna, representing the second largest biomass on tropical reefs after corals. The production of bioactive chemicals by marine sponges is a factor which likely enhances their competitiveness in coral reef environments. The secondary metabolites may act as chemical defenses against predation by fish, molluscs or other carnivores (reviewed in Paul, 1992), or act to prevent other marine species growing adjacent to or on top of the sponge tissue (Davis et al., 1989; Clare, 1996; Fusetani, 1997). Although sponge-derived chemicals have been implicated in allelo- chemical interactions with neighbouring corals (Sullivan et al., 1983; Porter and Targett, 1988), few rigorous ecological studies have yet been undertaken. Field studies by McCaffrey (1988) discovered a haplosclerid sponge, Haliclona sp. nov. (Haplosclerida: Chalinidae), which grows on coral substrates in the channel zones of Heron island at 10-14m depth. Although the sponge tissue was soft and easily torn, there were no feeding scars to indicate predation by fish or other scavengers, nor was its surface fouled by epiphytes; these facts suggested the presence of inhibitory chemicals. The sponge also exuded mucus upon collection. McCaffrey (1988) showed that the sponge contained antimicrobial components toxic to hydroids, corals, crustaceans and fishes, although she did not identify the 206 chemicals involved. Our subsequent research found crude organic extracts of this species exhibited potent antifungal and antimicrobial activity and an IC50 of 5ug/mL in a P388 mouse leukaemia assay. The aqueous methanol phase of a toluene:methanol (3:1) sponge extract was therefore extracted with chloroform, and the combined organic extracts processed to give a suite of novel alkaloids, the haliclonacyclamines A-D (Fig. 1) (Charan et al., 1996; Clark et al., 1998). Using Percoll gradient centrifugation of fixed cells, we demonstrated that the alkaloids are stored in, and are therefore likely biosynthetic products of, sponge cells (Garson et al., 1998). Haliclona sp. nov. has been reported to grow on coral substrate, usually Acropora nobilis, but also other corals such as Poecillopora sp. and Seriatopora hytrix and also on sand-covered coral rock (McCaffrey, 1988). When the sponge tissue was examined by light microscopy, nematocysts of mean length 12-15um were detected, as was a dinoflagellate which morpho- logically resembled Symbiodinium microadriaticum, the dinoflagellate symbiont of reef corals (McCaffrey 1988, Garson et al., 1998), Haliclona sp. nov. is a versatile sponge in that it appears to have evolved multiple defense strategies. In addition to the potential physical defense provided by mucus exudation, and the presence of nematocysts, the associated alkaloids may provide an additional chemical defence in situ. In this paper, we present some preliminary evidence on the ecological roles of the haliclonacyclamine alkaloids. MATERIALS AND METHODS CHEMICALS AND BIOCHEMICALS. Agar was purchased from Sigma Chemical Company MEMOIRS OF THE QUEENSLAND MUSEUM (MO, USA) while brine shrimp eggs and dried krill were purchased from an aquarium supply shop. Solvents used in ecological experiments and in the extraction of compounds from whole tissue or cell separation experiments were glass distilled. BIOLOGICAL MATERIALS. Samples of Haliclona sp. nov. were collected by hand using SCUBA at the Coral Gardens (10-15m depth), Heron Island (23?27'S, 151°55’E), S Great Barrier Reef, Australia, under permit numbers G96/050, G97/097, G98/037 issued jointly by the Great Barrier Reef Marine Park Authority (GBRMPA) and the Queensland National Parks and Wildlife Service; and at North Point, Lizard Island (14?39'S, 145°27°E), N Great Barrier Reef under GBRMPA permit G98/227. Sponge samples used in biological experiments were maintained in running sea water at ambient temperature and light conditions prior to use. Coral samples used for ecological studies were collected under GBRMPA permits G97/097 and G98/037. For a brief description of the sponge and the dinoflagellate symbiont, see Charan et al. (1996) and Garson et al. (1998). A voucher specimen of the sponge is accessioned in the Queensland Museum, Brisbane, collections (QM G304086). ISOLATION OF METABOLITES. A crude alkaloid extract (600mg) was prepared from frozen sponge (250g wet wt.) as described by Clark et al. (1998) and further purified by normal phase HPLC using EtOAc/hexanes/EtiN (30:65:5 or 80:15:5) to give haliclonacyclamines A (Fig. 1A; 162mg, 0.065%), B (Fig. 1B; 144mg, 0.057%), C (Fig. 1C; 26mg, 0.0012%) and D (Fig. 1D; 5mg, 0.002%). FIG. 1. Structures and stereochemistry of the haliclonacyclamines. A, Haliclonacyclamine A. B, Haliclonacyclamine B. C, Haliclonacyclamine C. D, Haliclonacyclamine D. BIOACTIVE METABOLITE LOCALIZATION ESTIMATIONS OF NATURAL CONCEN- TRATIONS. To determine the average natural concentration of metabolite in the sponge, three different samples of Haliclona sp. nov. collected from the Coral Gardens dive site were extracted, and the alkaloid content estimated. The ratio of the different haliclonacyclamines in each extract was assessed by normal phase HPLC using EtOAc/hexanes/EGN (30:65:5). For comparison, aspecimen collected at Coral Spawning dive site, 500m further along the reef, was extracted. For toxicity trials, a thin slice of frozen sponge (10mm x 20mm x Imm; approx 1g wet wt.) was extracted; this piece of sponge was equivalent in volume to a pad used for the in situ coral toxicity trials. The exudation rate of alkaloids from the sponge was assessed by aerating a 33g piece of sponge in seawater for 3hrs 45mins in ambient temperature and light under flow conditions. The sponge was carefully removed, and the residual sea water filtered through a 0.22um filter, then passed through a Cig Seppak cartridge, which was flushed with 100ml DCM to flush out organic components. Removal of the DCM solvent left a residue (3.9mg) which was analysed by TLC and NMR. MICROSCOPY STUDIES. Tissue samples were processed as described previously (Garson et al., 1998). Sections were viewed using Hitachi H-800 and Jeol 1010 transmission microscopes. Light microscopic observations of tissue or cell preparations were made on an Olympus BH-2 microscopic using Nomarski interference optics. FISH FEEDING DETERRENCY STUDIES. Agar cubes were prepared by combining 30g of agar, 2.7g of brine shrimp eggs and 2.7g of krill in one litre of Milli-Q water. This mixture was heated to 85°C and allowed to cool to approx- imately 50°C when the alkaloids were added to the agar mixture at 0.1% wt/vol (half the estimated natural concentration). The mixture was then cooled to approximately 40°C before being poured into ice cube trays. Each cube contained a lem” piece of wire gauze to which a length of dental floss was attached. Seven cubes of either treatment or control were then attached to a polypropylene rope by the dental floss with a 25cm gap between each cube. Eleven sets of paired ropes (one control and one treatment rope) were placed at the Coral Gardens field site at a depth of approximately 14m and with no more than 0.5m between the ropes. Divers stayed in the water to monitor feeding. When approximately half the cubes were eaten (approx 1hr), the ropes 207 were collected and the number of cubes eaten counted. Data was analysed with Wilcoxon’s signed rank test (Zar, 1984); two-tailed p-values are reported. The haliclonacyclamine alkaloids remained present in the agar cubes throughout the assay (TLC confirmation at the end of the experiment). CORAL TOXICITY STUDIES. Laboratory experiments. Pieces of Acropora sp. were collected and placed in an aquarium with continuous water flow for a period of 12hrs under ambient conditions of temperature and light. Pieces (approximately 2cm long) were broken off carefully and left in the aquarium for a further 10hrs to acclimatise. Treatments were prepared by dissolving the crude alkaloid extract in ethanol (at a concentration of 6mg/mL), then aliquots were dispensed into voucher jars containing 100mL of filtered sea water to give final concen- trations of 40ppm, 10ppm, and 5ppm. A fourth treatment was prepared consisting of haliclona- cyclamine A at 10ppm. A control experiment contained 666uL EtOH. There were ten replicates of each treatment and of the control. The solutions were aerated throughout the duration of the experiment. A single piece of coral was placed in each voucher jar and observed after 10mins, 20mins, 40mins, 80mins, 160mins and finally after an interval of 9.5hrs. The condition of the coral pieces was graded according to the following five point scale (Aceret et al, 1995): 1=75-100% of colony exhibiting normal polypal activity, with extended tentacles, no change in pigmentation, no mortality; 2=50-75% as above; 3=less than 50% as above; 4=tissue still evident, obvious loss of pigmentation, decreased water clarity, and no visible signs of life; 5-little or no remaining tissue evident, complete loss of pigmentation, mortality. Field experiments. Absorbent pads (Ix2cm; thickness 0.1 cm) were impregnated with alkaloid at concentrations of 0.005, 0.01, 0.025, 0.05, 0.1 and 0.4% (10 replicates at each concentration, dissolved in 2004L DCM per pad). The control consisted of a pad impregnated with 200u4L of DCM. The pads were taken underwater in plastic bags and attached to coral fingers with cable ties. The pads were left for 24hrs after which the coral was carefully detached using small pliers, taken back to the lab and left in the aquarium for 6hrs to acclimatise. The pads were removed and the condition of the coral graded (using the same scale as for the lab experiments). In a second 208 shorter term experiment using 0.01, 0.025, 0.035, 0.05 and 0.075% alkaloid, the pads were left underwater for 8hrs, then acclimatised in aquaria for 2hrs prior to grading. At the end of each experiment, the pads were extracted with dichloromethane to confirm the presence of residual alkaloids. In the 10hr experiment there were residual alkaloids present at all concentrations tested while in the 30hr experiment only the 0.4% treatment still contained residual alkaloid. RESULTS CHEMISTRY. The structures and stereo- chemistry of the haliclonacyclamine metabolites A-D are shown in Figure 1 A-D. The metabolites were characterised by 2D-NMR spectroscopy and by single crystal x-ray analysis (Charan et al., 1996, Clark et al., 1998). Haliclonacyclamines A and B (Fig.1A-B) have 5% or 57% double bonds respectively in addition to a 8^! double bond, while haliclonacyclamines C and D (Fig. 1C-D) were found to be analogous to haliclona- cyclamines A and B respectively, but lacking the ore double bond. The stereochemistry of the 6/516. 82526 or 52728 double bonds was found to be Z in all metabolites (Charan et al., 1996, Clark et al., 1998). ESTIMATION OF NATURAL CONCEN- TRATIONS. The average yield of alkaloid crude extract from the three Coral Gardens samples of Haliclona sp. nov. was 0.25% + 0.01% of sponge wet weight; the ratio of the haliclonacyaclamines estimated by HPLC was found to be consistent between all three extracts. The yields of individual alkaloids were: 1A, 0.065%; 1B, 0.057%; 1C, 0.012%; and 1D, 0.002%, giving a combined isolation yield of 0.136%. Some losses of compound are expected to occur during purification. The alkaloid yield from the specimen collected at Coral Spawning dive site was 0.24% + 0.01% of sponge wet weight; the composition of haliclonacyclamines was similar to that at Coral Gardens by HPLC. The lg piece of sponge of equivalent volume to a single pad used in the in sifu trials contained 2.4mg alkaloid. The rate of leaching of chemicals from Haliclona sp. nov. was assessed in laboratory experiments under flow conditions. An organic extract was obtained from sea water in which the sponge was immersed; TLC and 'H NMRanalysis detected the haliclonacyclamines in this extract. The amount of alkaloid detected using these analytical techniques is estimated to be 1.58mg. MEMOIRS OF THE QUEENSLAND MUSEUM Therefore if the haliclonacyclamines were present in the surrounding water, as suggested by McCaffrey (1988), then the rate of exudation was estimated to be 0.013mg/hr/g of sponge. BIOLOGY. Haliclona sp. nov. is one of the dominant sponges of the channel zone at Heron Island, where it commonly occurs at depths of 10-15m at the base of the reef slope. The sponge grows on coral substrate, and when damaged or collected, exudes abundant mucus, Its preferred substrate is Acropora nobilis, however specimens have also been observed to grow on Stylopora pistillata, Pocillopora sp., Seriatopora hytrix and on sand-covered coral rock (McCaffrey, 1988). Usually the sponge is a uniform olive-brown colour, but infrequently the tips are bleached. By light and transmission electron microscopy, dinoflagellates, usually intracellular, and nematocysts were present throughout preparations from sponge samples collected growing on acroporid substrates (Garson et al., 1998). Infrequently samples contained low populations of nematocysts, for example when collected from sand-covered coral rock. In June 1998, after an El Niño period, a bleached specimen of Haliclona sp. nov. was found growing on bleached acroporid coral at Heron Island. By microscopy, the sample was free of nematocysts, but contained some dinoflagellates; these were not healthy in appearance and were free-living rather than intra- cellular. Samples of the sponge were also found on dead coral substrate at North Point, Lizard Island; these specimens were small in size and always had bleached tips relative to the brown body colour. By microscopy, the bleached tips were free of both dinoflagellates and nemato- cysts; in contrast, the body of the sponge, which was a brown colour, contained healthy dinoflagellates but had no nematocysts. ECOLOGY. Fish deterrency. The crude alkaloid fraction of Haliclona sp. nov. deterred feeding (mean deterrency = 4.6+0.6 of 7 cubes eaten; N=11; p=0.002) by natural populations of reef fish in field assays conducted at 14m depth in the channel at Heron Island (Fig. 2). The alkaloid fraction was tested at half average natural concentration (0.1% of wet weight; 16mg per agar cube). Rabbitfish (Siganeus argenteus) were a major consumer of cubes in this experiment. Toxicity towards scleractinian corals. Figure 3 shows the results of laboratory experiments in which an alkaloid fraction from Haliclona sp. BIOACTIVE METABOLITE LOCALIZATION Mean no. of cubes eaten per rope N w > un an i Control Treatment FIG. 2. Field assays of a crude alkaloid extract (tested at half natural concentration, 0.196 of wet wt.) from Haliclona sp. nov. at Heron Island. P-values calculated using a Wilcoxon two-tailed sample test. Coral Condition MN w a Ango m on asn ngo o Time (minutes) FIG. 3. Toxicity ofan alkaloid fraction from Haliclona sp. nov. towards tips of Acropora sp. in laboratory experiments. Alkaloid concentrations ranged from 5-40ppm. Number of corals per concentration=10. nov. was added at varying concentrations to small aquaria of filtered seawater containing tips of Acropora sp. At 5ppm concentration, less than 50% of the corals were fully viable, that is exhibiting normal polypal activity, with tentacles extended and no loss of pigmentation after 160mins. At higher concentrations (10ppm and above), toxic effects were noted within 10mins of exposure to the alkaloid fraction. At the highest concentration tested (40ppm), all the corals were killed within 80mins. The major alkaloid comp- onent haliclonacyclamine A (Fig.1A) was tested at a single concentration of 10ppm. The toxic 209 effect of haliclonacyclamine A (Fig. 1A) was not as rapid as the alkaloid mixture was at 10ppm, but was equally effective in inhibiting coral and polypal activity after 80mins. Significant coral tissue necrosis was detected after 40mins exposure to this metabolite. In control experi- ments, dichloromethane solvent alone was added to the aquarium water without adverse effect to the coral pieces. Field assays were carried out using a method based on the work of Porter & Targett (1988). Alkaloid extracts were coated onto synthetic sponge pads and tied to healthy coral pieces growing in the vicinity of Haliclona sp. nov. in the channel at Heron Island. Control pads (coated with dichloromethane solvent only) produced no effects, whereas pads containing 0.025% alkaloid resulted in less than 50% of the corals tested remaining viable after 30hrs (Fig. 4). In a shorter term experiment (lOhrs duration; Fig. 5), the corals became unviable at concentrations of 0.025% alkaloid. DISCUSSION In her PhD work, McCaffrey (1988) demonstrated feeding deterrency of crude extracts from Haliclona sp. nov. by the bream, Acanthopagrus australis. Our field experiments have now demonstrated the deterrency of the haliclonacyclamine alkaloids to reef fish at ecologically-relevant concentrations. There is an increasing body of experimental evidence which demonstrates the deterrency of sponge secondary metabolites to fish predators (Rogers & Paul, 1991; Paul, 1992; Pennings et al., 1994; Pawlik et al., 1995; Chanas et al., 1996; Uriz et al., 1996). There is no obvious correlation between metabolite type and feeding deterrency since the range of structures which have been identified as feeding deterrents includes alkaloids (Chanas et al., 1996), sesterterpenes (Rogers & Paul, 1991; Duffy & Paul, 1992; Pennings et al. 1994), sesquiterpenes (Pennings et al., 1994; Uriz et al., 1996), and brominated metabolites (Paul, 1992; Pennings et al. 1994; Chanas et al., 1996). A number of these studies have considered other factors which may impact on palatability such as nutritional quality (Duffy & Paul, 1992; Chanas & Pawlik, 1995), the presence of spicules or the texture of the sponge tissue (Chanas & Pawlik, 1995; Chanas & Pawlik, 1996; Uriz et al., 1996). Some chemically-defended sponges contain spicules (Uriz et al., 1992; Chanas & Pawlik, 1995; Uriz et al., 1996). We have not yet investigated whether the spicules present in > Coral condition yn o o o 0.005 0.01 0.025 0.05 01 04 Concentration (%wt) FIG. 4. Toxicity ofan alkaloid fraction from Haliclona sp. nov. towards the periphary of Acropora sp. in field experiments. Alkaloid concentrations ranged from 0.005-0.4%. Number of corals per concen- tration-10. Length of experiment 30hrs. See text for coral toxicity gradation scale. 6 a + N Coral Condition [^] = 0.05 0 0.01 0.025 0.035 0.075 Concentration (%wt) FIG. 5. Toxicity of an alkaloid fraction from Haliclona sp. nov. towards the periphery of Acropora sp. in field experiments. Alkaloid concen- trations ranged from 0.01-0.075%. Number of corals per concentration=10. Length of experiment 10hrs. See text for coral toxicity gradation scale. Haliclona sp. nov. are an additional deterrent to fish, or whether they simply play a structural role in this fragile sponge - although it is speculated that the former is unlikely given that spicules are small, smooth, homogeneous oxeas contained completely within the choanosome (J. Hooper, pers. comm.). Further experiments are also required to assess the effect of nutritional quality on the anti-feedant properties of the haliclona- cyclamines. Marine sponges are known to release meta- bolites directly into the water column (Walker et al., 1985) or indirectly through a mucus exudate (Sullivan et al., 1983). McCaffrey (1988) investigated the exudation of biologically-active compounds from Haliclona sp. nov., but did not identify the chemicals or measure the exudation MEMOIRS OF THE QUEENSLAND MUSEUM rate. In preliminary laboratory experiments we have estimated the exudation of Haliclona meta- bolites is 0.013mg/hr/g wet weight of sponge under flow conditions. Our estimates took no account of the effect of water throughput or current or the concentration of alkaloids in the mucus exudate. Measurement of the natural leaching rate of organic extracts from marine sponges has not yet been addressed in the literature, although Henrikson et al. (1995) have measured the release of sponge chemicals from a range of artificial substrates. A detailed quanti- tative study of alkaloid leaching from Haliclona sp. nov. is in progress in our laboratory. In laboratory experiments, a mixture of the haliclonacyclamine alkaloids exhibited toxic effects towards pieces of acroporid coral and caused the corals to release mucus and to shed symbiotic zooxanthellae. When a sample of haliclonacyclamine A was tested at 10ppm, coral tissue necrosis was observed, however the purified metabolite was less toxic than the alkaloid mixture tested at the same concentration during short term exposure. These data suggest that the alkaloid mixture may be a more effective toxin than the individual chemicals. Further experiments will be required to confirm this synergistic effect. Our field results also confirmed the effective toxicity of the alkaloids. These experiments showed that the metabolites in Haliclona sp. nov. actively inhibit the metabolism and tissue survival ofadjacent acroporid corals. Our experi- ments used a range of alkaloid concentrations up to 0.075% (10hr experiment) or 0.4% (30hr experiment). If it is assumed that the alkaloids leach out of the artificial pads at the same rate as from sponge tissue, our experiment suggests that the metabolites are effective toxins at lower than natural concentrations. The field results cannot easily be related to the laboratory toxicity trials. The metabolite concentrations used in the coral pad experiment were equivalent to 250-20,000ppm, however the effective metabolite concentration that the coral may experience is much lower. Since both direct contact (using synthetic sponge pads to mimic the effects of sponge tissue), and indirect contact (addition of meta- bolites to aquarium water), resulted in toxic effects on corals, we conclude that the Haliclona alkaloids are effective allelochemicals which enable the sponge to compete successfully for space with coral substrates (Jackson & Buss, BIOACTIVE METABOLITE LOCALIZATION 1975; Wulff & Buss, 1979; Porter & Targett, 1988). Haliclona sp. nov. is an aggressive sponge which may preferentially select coral substrates as habitat, Although some studies have demon- strated the effectiveness of inhibitory substances in improving the competitiveness of sponges for space among benthic organisms (Becerro et al., 1997; McCaffrey & Endean, 1985; Thompson, 1985; Thompson etal., 1985; Walker etal., 1985; Clare, 1996; Wright et al., 1997), other studies have shown a contrasting ecological effect, for example that the presence of sponge chemicals may induce marine invertebrate larvae to settle (Bingham & Young, 1991). The ecological effectiveness of Haliclona metabolites on benthic invertebrates other than corals, or on their larvae, is yet to be tested in our laboratory. This study will enable us to determine if the haliclonacyclamines are selectively toxic or not. The Haliclona metabolites possess a liphophilic carbon backbone together with a polar amine functionality, and are therefore amphiphilic in character. The compounds, thus, have partial water solubility; for example, NMR spectra can be obtained in deuteriated water. The sponge pads placed underwater for 10hrs still retained alkaloids at the end of the experiment. In the long term exposure study, we observed loss of metabolites from the artificial pads at the lower concentrations. The ongoing release of a polar, diffusible substance by a sponge is of no value as amechanism to inhibit settlement (Becerro et al., 1997), and is also metabolically uneconomic; the most suitable chemical candidates for defense or for use as an anti-fouling agent are likely to be water-insoluble. Some recent studies have attempted to better simulate natural conditions in field trials by embedding sponge extracts onto artificial matrices which can be placed in the field for long periods. Chemicals may leach out of these gel matrix at rates which may mimic their natural release (Morse et al., 1994; Hendrikson & Pawlik, 1995), Experiments of this type are currently in progress in our laboratory. In corals, photosynthesis is performed uniquely by dinoflagellate (zooxanthellae) symbionts which supply the host with nutrients by translocation (Muscatine & Cernichiari, 1969). Cyanobacteria are the most common sponge photosynthetic symbionts, however some groups of sponges, notably the boring sponges of the order Hadromerida (e.g. Cliona spp.), have been shown to contain zooxanthellae (Vacelet, 1982; Riitzler 1990), although it is not yet known whether the dinoflagellate partners supply the hadromerid sponges with photosynthetic products. We propose that Haliclona sp. nov. may poison or kill the coral tissue on which it grows in order to acquire dinoflagellate symbionts, which provide additional metabolic benefits to the sponge, thereby enhancing its competitiveness. Sullivan et al. (1983) showed that the boring sponge Siphonodictyon sp. (=Aka, family Phloeodictyidae), uses toxin-containing mucus to kill surrounding tissue. Haliclona sp. nov. exudes abundant mucus on collection and so may perhaps use a similar process. We are currently investigating the chemistry of Haliclona sp. nov. mucus to determine if it contains the haliclonacyclamines. To our knowledge, no other marine sponge has been reported to contain nematocysts. Perhaps the nematocyst capture represents an additional serendipitous defense mechanism in Haliclona sp. nov. The high numbers of nematocysts found intracellularly in healthy sponge tissue samples are inconsistent with their casual acquisition from the surrounding water. Both nematocysts and dinoflagellates were absent in a bleached sample of Haliclona sp. nov. found growing on bleached coral at Heron Island. This sample was clearly stressed since the dinoflagellates appeared unhealthy and were free-living rather than intracellular. Partially-bleached samples of Haliclona sp. nov. collected from dead coral rock at Lizard Island were also free of nematocysts even if the body of the tissue was healthy and contained dinoflagellates. The distribution and specific source of the nematocysts present within the sponge tissue is currently under further investigation, as is their impact on predation. ACKNOWLEDGEMENTS We thank the director and staff of Heron Island Research Station, and undergraduate students from the coral reef ecology subject (ZL329) at The University of Queensland for field assistance. Dr John Hooper of the Queensland Museum identified the sponge sample for us. Our research on Haliclona sp. nov. is funded by a University of Queensland Special Program Grant and by the Australian Research Council. LITERATURE CITED ACERET, T.L., SAMMARCO, P.W. & COLL, J.C. 1995. Toxic effects of alcyonacean diterpenes on scleractinian corals. Journal of Experimental Marine Biology and Ecology 188: 63-78. BECERRO, M.A., URIZ, M.J. & TURON, X. 1997. Chemically-mediated interactions in benthic organisms: the chemical ecology of Crambe crambe (Porifera, Poecilosclerida). Hydrobiologica 356: 77-89. BINGHAM, B.L. & YOUNG, C.M. 1991. Influence of sponges on invertebrate recruitment: a field test of allelopathy. Marine Biology 109: 19-26. CHANAS, B. & PAWLIK, J.R. 1995. Defenses of Caribbean sponges against predatory reef fish. II. Spicules, tissue toughness and nutritional quality. Marine Ecology Progress Series 127: 195-211. 1996. Does the skeleton of a sponge provide a defense against predatory reef fish? Oecologia 107: 225-231. CHANAS, B., PAWLIK, J.R., LINDEL, T. & FENICAL W. 1996. Chemical defense of the Caribbean sponge Agelas clathrodes (Schmidt). Journal of Experimental Marine Biology and Ecology 208: 185-196. CHARAN, R.D., GARSON, M.J., BRERETON, I.M., WILLIS, A.C. & HOOPER, J.N.A. 1996. Haliclonacyclamines A and B, cytotoxic alkaloids from the tropical marine sponge Haliclona sp. Tetrahedron 52: 9111-9120. CLARE, A.S. 1996. Marine natural product antifoulants: status and potential. Biofouling 9: 211-229. CLARK, R.J., FIELD, K.L., CHARAN, R.D., GARSON, M.J., BRERETON, I.M. & WILLIS, A.C. 1998. The haliclonacyclamines, cytotoxic alkaloids from the tropical marine sponge Haliclona sp. Tetrahedron 54: 8811-8826. DAVIS, A.R., TARGETT, N.M., McCONNELL, O.J. & YOUNG, C.M. 1989. Epibiosis of marine algae and benthic invertebrates: Natural products chemistry and other mechanisms inhibiting settle- ment and overgrowth. Pp. 85-113 In Scheuer, P.J. (ed.) Bioorganic marine chemistry. Vol. 3. (Springer-Verlag: Berlin). DUFFY, J.E. & PAUL, V.J. 1992. Prey nutritional quality and the effectiveness of chemical defenses against tropical reef fishes. Oecologia 90: 333-339. FUSETANI, N. 1997. Marine natural products influencing larval settlement and metamorphosis of benthic invertebrates. Current Organic Chemistry 1: 127-152. GARSON, M.J., FLOWERS, A.E., WEBB, R.L, CHARAN, R.D. & McCAFFREY, E.J. 1998. A sponge dinoflagellate association in the haplosclerid sponge Haliclona sp.: cellular origin of cytotoxic alkaloids by Percoll density gradient fractionation. Cell Tissue Research 293: 365-373. HENDRIKSON, A.A. & PAWLIK, J.R. 1995. A new antifouling assay method: results from field experiments using extracts of four marine organisms. Journal of Experimental Marine Biology and Ecology 194: 157-165. JACKSON, J.B.C. & BUSS, L.W. 1975. Allelopathy and spatial competition among coral reef invertebrates. Proceedings of the National Academy of Science 72: 5160-5163. MEMOIRS OF THE QUEENSLAND MUSEUM McCAFFREY, E.J. 1988. Biologically-active compounds from marine sponges collected from Queeensland waters. PhD thesis (University of Queensland: Brisbane). McCAFFREY, E.J. & ENDEAN, R. 1985. Antimicrobial activity of tropical and subtropical sponges. Marine Biology 89 :1-8. MORSE, D.E., MORSE, A.N.C., RAIMONDI, P.T. & HOOKER, N. 1994. Morphogen-based chemical flypaper for Agaricia humulis coral larvae. Biological Bulletin 186: 172-181. MUSCATINE, L. & CERNICHIARI, E. 1969. Assimilation of photosynthetic products of zooxanthellae by a reef coral. Biological Bulletin 137: 506-523. PAUL, V.J. 1992. Chemical defenses of benthic marine invertebrates. Pp. 164-168. In Paul, V.J. (ed.) Ecological role of marine natural products. (Comstock Publishing Associates: London). PAWLIK, J.R., CHANAS, B., TOONEN, R.J. & FENICAL, W. 1995. Defenses of Caribbean sponges against predatory reef fish. I. Chemical deterrency. Marine Ecology Progress Series 127: 183-194. PENNING, S.C., PABLO, S.R., PAUL, V.J. & DUFFY, J.E. 1994. Effects of sponge secondary metabolites in different diets on feeding by three groups of consumers. Journal of Experimental Marine Biology and Ecology 180:137-149. PORTER, J.W. & TARGETT, N.M. 1988. Allelochemical interactions between sponges and corals. Biological Bulletin 175: 230-239. ROGERS, S.D. & PAUL, V.J. 1991. Chemical defenses ofthree Glossodoris nudibranchs and their dietary Hyrtios sponges. Marine Ecology Progress Series . 75221-2232. RUTZLER, K. 1990. Associations between Caribbean sponges and photosynthetic organisms. Pp. 455-471. In Rützler, K. (ed.) New perspectives in sponge biology (Smithsonian Institution Press: Washington DC). SULLIVAN, B., FAULKNER, D.J. & WEBB, L. 1983. Siphonodictidine, a metabolite of the boring sponge Siphonodictyon sp. that inhibits coral growth. Science 221: 1175-1176. THOMPSON, J.E. 1985. Exudation of biologically-active metabolites in the sponge Aplysina fistularis. 1. Biological evidence. Marine Biology 88: 23-26. THOMPSON, J.E., WALKER, R.P. & FAULKNER, D.J. 1985. Screenings and bioassays for biologically-active substances from forty marine sponge species from San Diego, California, USA. Marine Biology 88: 11-21. URIZ, M.J., MARTIN, D. & ROSELL, D. 1992. Relationships of biological and taxonomic characteristics to chemically mediated bioactivity in Mediterranean littoral sponges. Marine Biology 113: 287-297. URIZ, M.J., TURON, X., BECERRO, M.A. & GALERA, J. 1996. Feeding deterrency in sponges. BIOACTIVE METABOLITE LOCALIZATION 213 The role of toxicity, physical defenses, energetic metabolites in the sponge Aplysina fistularis. II. contents and life-history stage. Journal of Chemical evidence. Marine Biology 88: 27-32. Experimental Marine Biology and Ecology 205: WRIGHT, J.T., BENKENDORFF, K. & DAVIS, A.R. 187-204. 1997. Habitat-associated differences in temperate VACELET, J. 1982. Algal-sponge symbioses in the sponge assemblages: the importance of chemical coral reefs of New Caledonia: a morphological fei neonate eh ma Marine Biolagy study. Proceedings of the 4th International Coral WULFF LL S BUSS L.W. 1979. Do sponges help Reef Symposium 2: 713-719. hold corals reefs together? Nature 281: 474-475. WALKER, R.P., THOMPSON, J.E. & FAULKNER, ZAR, J.H. 1984. Biostatistical analysis. 2nd Edition. D.J. 1985. Exudation of biologically-active (Prentice-Hall: Englewood Cliffs, NJ). MEMOIRS OF THE QUEENSLAND MUSEUM SPONGIVORY BY THE BRAZILIAN STARFISH ECHINASTER BRASILIENSIS. Memoirs of the Queensland Museum 44: 214. 1999:- The feeding ecology of Echinaster brasiliensis has been studied on a temporal gradient (January 1995 - September 1996, 11 observation periods), along a shallow-water transect parallel to the coastline (1.5-6m depth, 2000m?) at Ponta do Baleeiro (23?49.727" S - 45°25.364’W), São Sebastião Channel (Sao Sebastião, SP, Brazil). In total, 3025 starfish were observed, 44% of which were feeding (1337/3025). Of these, 42% (557/1337) were feeding on sponges, a significantly higher proportion than the real availability of sponges in terms of area coverage by organisms. Of the 33 sponge species recognised, the most wanted prey was Mycale aff. americana, representing 40% (221/557) of the total number of observed spongivory events. Other common sponge prey items were Amphimedon sp., Haliclona sp.n., Mycale angulosa, Mycale microsigmatosa and Tedania ignis, with ca. 5% of the spongivory events each. Semiquantitative arbitrary estimations point toward these species’ high abundance in the study area. Therefore, we cannot discard the possibility of a direct link between the sponge’s abundance and apparent starfish preferences. Of the 33 sponge species eaten, at least 61% (20/33) belong to genera from which species were found (literature) to possess toxins, thus raising the question: ‘What are these toxins good for?’ The conspicuous habit of Echinaster brasiliensis suggests that it may be unpalatable to many potential predators, perhaps through the use of sequestered toxins of dietary origin, The temporal gradient studied did not reveal clear patterns, thus suggesting that inter-annual climatic oscillations may play an important role in shaping the starfish’s feeding ecology. Acknowledgement of financial support: CNPq, FAPERJ, FAPESP, FUJB-UFRJ, O Porifera, spongivory, SW Atlantic, Echinaster, Mycale, feeding ecology, chemical ecology. M.C. Guerrazzi*, Pós-Graduacdo em Zoologia, Departamento de Zoologia, Instituto de Biociéncias, Universidade Estadual Paulista - Campus de Rio Claro, Rio Claro, SP, Brazil; E. Hajdu* (email: hajduíWacd ufrj.br), Museu Nacional, Departamento de Invertebrados, Universidade Federal do Rio de Janeiro, Quinta da Boa Vista, s/n, 20940-040, Rio de Janeiro, RJ, Brazil; E.H. Morgado Do Amaral* & L.F.L. Duarte*, Departamento de Zoologia, Instituto de Biologia, Universidade Estadual de Campinas, Cidade Universitária Zeferino Vaz, Cx. Postal 6109, 13083-970, Campinas, SP, Brazil; * and Centro de Biologia Marinha, Universidade de São Paulo, São Sebastião, SP, Brazil; 1 June 1998. ECOLOGICAL ADAPTIONS OF A FRESHWATER SPONGE ASSOCIATION IN THE RIVER RHINE, GERMANY (PORIFERA: SPONGILLIDAE) JOCHEN GUGEL Gugel, J. 1999 06 30: Ecological adaptions of a freshwater sponge association in the River Rhine, Germany (Porifera: Spongillidae). Memoirs of the Queensland Museum 44: 215-224. Brisbane. ISSN 0079-8835. The species composition and autecology of freshwater sponges (Porifera, Spongillidae) were investigated in the Rhine between Karlsruhe and Bonn (Germany) between 1993 and 1995. Ephydatia fluviatilis, E. muelleri, Trochospongilla horrida, Spongilla lacustris, Eunapius fragilis and E. carteri were found. Ephydatia fluviatilis was classified as an r-strategist due to its high ability to colonise new habitats, whereas other species placed emphasis on successful establishment in more stable habitats and should therefore be classified as K-strategists (among freshwater sponges). Similarly, the production of larvae was an integral part of the life cycle only in E. fluviatilis, whereas other species put their main efforts in producing gemmules as distribution-units. Asexual vs. sexual reproductive strategies in freshwater sponges in running-water habitats is discussed in terms of their prevalence, periodicity and influence of limnological factors. O Porifera, Spongillidae, life cycle, adaptions, central Europe, running water, river Rhine. Jochen Gugel (email: jochen@post.tau.ac.il), Darmstadt University of Technology, Institute of Zoology, Schnittspahnstrasse 3, D-64287 Darmstadt, Germany. Present address: Tel Aviv University, George S. Wise Faculty of Life Sciences, Department of Zoology, Ramat Aviv, Tel Aviv 69978, Israel; 15 March 1999. A considerable number of publications on life cycles of freshwater sponges are now available (e.g. Gilbert et al., 1975, Frost et al., 1982, Courréges & Fell, 1989, Bisbee, 1992), Mostly these focus on single life cycle events, such as formation of larvae or gemmulation, and often include only a single species. Only a few papers deal with associations of several species, their colonisation strategies and spatial competition (e.g. Williamson & Williamson, 1979; Mukai, 1989; Pronzato & Manconi, 1991). The present investigation reports on five sympatric species occuring in the Rhine between Karlsruhe and Bonn (Germany), their reproductive strategies and colonisation. At the beginning of the 20th century Lauter- born distinguished 83 macrobenthic animals in the Rhine (Tittizer et al., 1990). In the 1970’s these numbers decreased to 12 species (Conrad et al., 1977), due to an extremely high level of pollution. Since that time great efforts have been undertaken to purify the water, and the number of macrobenthic animals has again risen steadily (Schóll et al., 1995). This species diversity now exceeds that of Lauterborn, whereas the species composition is not the same as in the beginning of the century (Tittizer et al., 1990), due to the huge changes in the Rhine. In recent years ongoing invasions of foreign species (Neozoa: Kinzelbach, 1995) have been taking place, and have influenced the biocoenosis considerably. Results of Franz (1992) indicate that the Rhine is a highly suitable habitat for sessile filter feeders. Not only is the nutritional situation excellent for these animals due to its eutrophic waters, but the banks are entirely covered by rocks which provide a suitable substrate. Filter feeding is not restricted to sessile animals - in particular many insects also gain their nutrition from filter feeding - and Mann et al. (1972) stated that the product- ivity of filter feeders is extremely high within waters disturbed by anthropogenic influences. Nevertheless, our knowledge about freshwater sponges in the Rhine is still fragmentary, despite regular, general studies on the macrobenthic fauna. Such studies usually include sponges, although they are often given only cursory consid- erations. This seems surprising since repeatedly high abundances of single species have attracted the attention of researchers in the past (Schón, 1957; Bartl, 1984). Large rivers are characterised by unpredictable changing water levels. This offers and destroys new habitats - a situation which seems difficult to deal with for sessile organisms. 216 MEMOIRS OF THE QUEENSLAND MUSEUM TABLE 1. Collecting sites and dates of collections Key: 1, Near the facilities of the BASF AG. 2, Outflow of the cooling water circuit of the power plant, the temperature is here up to 10°C higher than the surrounding river. 3, Slightly polluted stagnant water. *, Collection undertaken with a grap dredger on board of the research ship “Argus”. Locality 22-23.1.93 08-11.v.93 7-8.ix.93* 25-26.ix.93 24-26.viii.93* 6-7.xi.93 22.11.94 11.v.94* 8.vi.94* 30.vi.94* 15-26.vii.94* 10-11.viii.94 14-15.x.94 15.111,95 3.viii.95 8.x.95 12-14,x.95 Iffezheim Neuburg Leimersheim Sondernheim »* xxx * ojx |< x Altrip Ludwigshafen! ” x Ko Jx jx x |x x x [x jx |x |x Lampertheim-Rosen-Garten x Worms x Biblis, nuclear power plant” x Biblis, downstream n.p.pl. Worms-Rheinduerk-Heim x x Gross-Rohrheim Gross-Rohrheim, 2km downstream x x wm x Gernsheim x X x Kornsand * x x pa ES Ax [Xx [> »* ad Nierstein x x x Oxbow of Ginsheim? x Mainz-Laubenheim x x x Mainz Heidenfahrt Bingen A * xxx * Bacharach Boppard Urmitz/Kaltenenger xx |x |x |x x ojx jx jx |x »* Bad Breisig x ojx |x |x |x * Remagen Bonn-Bad Godesberg METHODS COLLECTION. Samples were collected from many sites in the Rhine (Germany) (Fig. 1), with dates of collection for each site listed in Table 1. Collections were mostly made from the banks of the river, by wading in the water and removing substrate by hand. Some collections were made with a grab dredger aboard the research ship ‘Argus’ of the federal state Hesse (indicated with an asterisk in Table 1). Sponges were removed from the substrate with a knife and preserved immediately in 70% ethanol. From each individual sponge, one microscope slide was prepared. Individual sponges growing on approximately 1.3m? available substrate at each collecting site were counted. The average number of individual sponges per m? of available substrate were calculated (Figs 4-9). Only active colonies were counted, no dead colonies or gemmules without their living mother-sponge. Methods of preparation for microscopy followed Arndt (1928), with slight modifications. Sponge identification was based RIVER RHINE SPONGE ECOLOGY Bonn-Bad Godesberg Remagen — Bad Breisig Urmitz/Kaltenenger Boppard ^, Bacharach , Bingen | Haidenfahr Karlsruhe Minz Mainz-Laubenheim ' 7 Nierstein Worms-Rheindürkheim Worms Ludwigshafen " Lake Constance Biblis Altip — Sondemheim Leimersheim Neuburg - — Mfezheim FIG. 1. Rhine collecting sites. on Arndt (1926, 1928) and Penney & Racek (1968). Slides prepared during this study are deposited inthe Senckenberg Museum of Frankfurt (SMF). EVALUATION OF THE PERIOD OF FLOODING BEFORE COLLECTION. At each collection site the actual water depth was recorded for all sponge samples. Daily inform- ation on water levels at the stations of Worms and Mainz were recorded (Fig. 2), thus for each site the period of flooding before collection could be calculated. RESULTS FAUNISTICS. Six species were found in the present study, listed according to their prevalence from abundant to rare: Trochospongilla horrida Weltner, 1893; Ephydatia fluviatilis (L., 1758); Spongilla lacustris (L., 1758); Ephydatia muelleri (Lieberkühn, 1855); Eunapius fragilis (Leidy, 1851); Eunapius carteri (Bowerbank, 1863) SUBSTRATE. Sponges were found settling on all kind of solid substrates. Within the Rhine this mainly consists of rocks placed to support the river banks; wood is rarely found. Aquatic macrophytes are almost non existent in the investigated area, only once was a small E. fluviatilis found epizootic on Fontinalis sp. (Bryophyta, Fontinalaceae). Oxbow of Ginsheim Komsand Gemsheim GroB-Rohrheim GroB-Rohrheim As a rule, the larger rocks (immovable by average currents) were more likely to be colonised with sponges. Smaller rocks (often moved by average currents) were rarely settled by sponges or other sessile organisms. These preliminary data agree with those of Riitzler (1965), who studied colonisation by marine sponges in the Mediterranean Sea. DISTRIBUTION IN RELA- TION TO DIFFERENT FLOODING REGIMES. The species-assemblages varied considerably between sites dependent on flooding events, comparing sites flooded more than six months before collecting, and those flooded only nine weeks prior to collection (Fig. 3). Ephydatia fluviatilis was the only species occurring regularly at the more recently flooded sites, whereas at sites flooded more than six months before collecting this species had the same absolute abundance in colony-counts, but colonies grew much larger. The other species, 7: horrida, S. lacustris, E. fragilis and E. muelleri, also appeared at both these categories of sites, but only in very small numbers and small size at recently flooded sites (with most having a diameter of less than lem), It was also apparent that differences in flooding events between the sites is a major factor responsible for different depth preferences of sponge species. Places recently flooded were very shallow, often dry, whereas places flooded over 6 months ago were deeper, below the levels affected during river-level fluctuations. Other factors show no depth dependent variations in the river environment. Nutrient levels should be evenly distributed within the waterbody, due to turbulent currents, and light can only penetrate about 0.75m through the water column due to high turbitidy. In Figure 3 and subsequent figures only colony-counts are given, where no distinction has been made according to the size or local abundance of colonies. SEASONAL DEVELOPMENT. The general outline of the development of sponge species associations is given in Figure 4. A generalised model ofa life-cycle ofa freshwater sponge in the Biblis nuclear power plant Lampertheim-Rosengarten MEMOIRS OF THE QUEENSLAND MUSEUM clearly visible, not the preceding states of larval development, but it was not the goal of this study to describe the life history of these sponges, only to report on the presence or absence of mature larvae, as important indicators on the ecology of species. Spongilla lacustris. This height of the water (cm) species formed thick crusts ^g e^ e" e^ o Se ue eee O > » cog AO uo ue ue e eh MI S a KS CO SC p Pu Or e Ol 4 PP PP SE FIG. is relative, without a defined zero point). Rhine is as follows (Fig. 4). In March-April young sponges hatched from their overwintering gemmules. Sponges grew until midsummer (July-August), sexually produced larvae may occur from May to July. Asexual gemmules were produced year-round, but more regularly towards the autumn. In September-October the colonies declined and desintegrated, thus producing overwintering units (gemmules). SPECIES AUTECOLOGY. Statements about the presence or absence of sexual reproduction usually require histological analysis of specimens, which was not conducted in this study. Under low magnification only fully developed larvae were more than 6 months flooded E. peters T. horrida 31% E ri al Es ra S. lacustris = 16.9 colonies/ m? ^* PPP db dodo do S quus Pu ge e e ud if FF SAS SS 2. The niveau of daily water levels at Worms from July 1993- December 1995; the vertical lines indicate the turn of the years (the scale (1-3cm thick); the outline was irregular with rounded edges. Colonies could reach a considerable size (up to 1m’). Only very few colonies showed tendencies towards branching growth forms. Colonies of S. lacustris always disintegrated in late autumn (October- November). In winter (December-February) only gemmules survived. The first young sponges hatched from gemmules in spring (beginning of April), the number of colonies then rose steadily until October (Fig. 5). The high numbers of colonies reported during the period from October-December were mainly due to these colonies being present at the beginning of October, whereas by the end of October their numbers had declined rapidly. Furthermore, the larger colonies observed in October fragmented into several smaller colonies before dying, so that counts of number of colonies rose before they dropped, and eventually disappeared completely in December-March. less than 9 weeks flooded . T. horrida 8% S. lacustris | 1% | _ E. muelleri QI j E. pi 5.7 colonies/ m FIG. 3. Species-assemblages at places with different times of flooding before collection (numbers are calculated for Im”): 26 collections were made at sites more than 6 months prior to flooding of collection sites; 28 collections took place at sites less than 9 weeks prior to flooding. RIVER RHINE SPONGE ECOLOGY TABLE 2. Life cycle data and ecological strategies of the five sympatric sponge species. Event E. fluviatilis S. lacustris E. muelleri E. fragilis T. horrida Time of greatest abun- October (autumn) October (autumn) May-June (early August-September dance June (carly summer) summer) (late summer) Overwintering units Whole colony Weakly fixed free Fixed Stee gemeñulos Attached gemmule | Attached gemmule gemmules crusts crusts Distribution units Larvae Drifting gemmules | Drifting gemmules Larvae ? Colonisation of newly Through active established habitats swimming larvae Not observed Not observed Not observed Not observed Ecological strategy r-strategy K-strategy K-strategy K-strategy K-strategy Gemmules generally appeared from August, but their appearance seemed to be less dependent on seasonality and more dependent on colony size. This was also true for other species (see Rasmont, 1962, 1963; Simpson, 1980). Colonies of S. lacustris larger than 3cm in diameter always contained at least some gemmules; smaller colonies were mostly free of gemmules, at what- ever time of the year they were encountered. Gemmules were built singly within the tissue of the mother-sponge, always within the basal parts of the colonies. The gemmulation process was more regular towards the end of the life span of intact colonies. There were often dense, single-layered carpets of gemmules, resting where they formed. The whole sponges disintegrated after death, but sheltered parts of the skeleton still remained intact so that gemmules resting in these patches of skeletal refugia were bound together and weakly fixed to the substrate. Green number of colonies / m? FE EN i | M July- September $ October- December January- March April-June Larvae Gemmulation Gemmules FIG. 4. Seasonal appearance of Spongillidae in general in the Rhine study sites. The white bar indicates when only gemmules are present; the black bar represents times of the year when active colonies are present; the time-scale of the chart corresponds with that of the bar. gemmules, due to an infestation with unicellular symbiotic algae and a gemmule-polymorphism, as described by Gilbert & Simpson (1976) and Brondsted & Brondsted (1953), were not observed in this study. Larvae were not observed in this species from the Rhine. This was very intriguing given that in other habitats colonies of S. lacustris containing larvae were regularly found (e.g. within the outflow of the ‘Steinbrücker Teich’, a eutrophic pond near Darmstadt, Germany, nearly 50% of the colonies in July 1994 contained larvae). In early autumn about 30% of colonies were bright green due to the presence of symbiotic algae (Fig. 5). Only during this part of the year were water levels low enough to provide the preferred habitats for S. lacustris (i.e. in slightly deeper, permanently flooded water, Fig. 3), with sufficient light for the successful photosynthesis of symbionts. Eunapius fragilis. This species formed low crusts (1-2cm thick), with an irregular outline and rounded edges. Rarely it exceeded a diameter of Scm. Colonies of E. fragilis usually disintegrated in summer (July-September). In winter (December-February) intact colonies were rarely found (Fig. 6). The first sponges hatched from gemmules in spring (April). Immediately after hatching the highest numbers of colonies appeared (Fig. 6). The species completed its gemmulation process up until summer (July), after which colonies began to disintegrate. Gemmules were formed in situ producing a pavement-like gemmule crust, tightly fixed to the substrate. These gemmules are virtually immovable and it is difficult to perceice how they could contribute to the dispersal within the habitat, whereas in May-June 1994 free movable larvae were found in about 10% of colonies. Colonies containing symbiotic algae were not found within the Rhine, probably because their preferred distribution was in permanently flooded, 220 m 25 BHenmsymuicuc sae e | o (without symbolic atges | = oe g 2 8 o a — p a E 3 c January-March ApnkJune July-September Octaber- December FIG. 5, Seasonal appearance and number of symbiotic colonies in Spongilla lacustris. deeper habitats, where light regimes may be insufficient for photosynthesis (Fig. 3). Ennapius carteri. This species record from the Rhine is the first time it has been encountered in central Europe (see Gugel, 1995). It was found November 1993 within the cooling-water outflow of the nuclear power plant in Biblis. A detailed deseription and discussion about its dispersal are given in Gugel (1995). Ephydatia fluviatilis. This species forms more-or-less thin encrustations (1-2em thick). Smaller colonies (less than 5cm diameter), had a circular outline, whereas larger ones (more than Jem diameter), were more irregularly shaped. Colonies of this species grew up to 20cm diameter. Ephvdatia fluviatilis was regularly seen alive in winter (December-February), in contrast to the other species. Overwintering colonies were small crusts, only 15cm diameter, in which no canal systems were visible. These probably survive in a reduced state, as suggested by Arndt (1928) and Weissenfels (1989). In early spring (April) their abundance was only slightly increased in comparison with winter (Fig. 7), probably due to hatching of gemmules (see below), whereas the number of colonies dramatically increased during June-July (Fig. 7), at which ume a large-scale production of larvae E 25 i | ü 2 2 — 2 8 i4 - - 5 Pu 1 D 2 E 05 2 al i = -Janumry-Merch ApnhJune July September Ottober- December FIG. 6. Seasonal appearance of Eunapius fragilis. MEMOIRS OF THE QUEENSLAND MUSEUM took place (June). In July these larvae had settled and built new colonies. The gemmulation process was irregular throughout the whole year and a considerable number of colonies was always devoid of gemmules. When gemmules were present their numbers were reduced: in colonies of 5em diameter not more than 10 gemmules were found. During fragmentation of sponges few gemmules were freed from the mother-sponge and these *hatched’ in spring. The highest number of colonies was encountered during October-December. As m the case of S. lacustris, the large colonies present in autumn fragmented into several smaller colonies, many of which died towards winter (December- February), the overwintering colonies were also small. In May-June about 25% of colonies produced larvae. This seemed to be the most important event in the life cycle of E. fuviatilis, as in July many very small colonies were seen in close proximity to each other, a phenomen quoted as ‘Spriihinfektion’ (spray infection) by Steuslolf (1938). As already indicated, E. fluviatilis was the only species which oceurred in higher numbers at sites flooded only a few weeks prior to collection (Fig. 3). This was probably due io the more active dispersal of larvae. Symbiotic colonies were never found. Ephydatia muelleri. Colonies of this species were mostly thickly encrusting (2-4cm thick), with irregular outline and rounded edges. The diameter was rarely more than 10cm. The first colonies of E. muelleri appeared in spring (at the beginning of April). and soon aiter haiching colonies were found in large numbers, peaking during summer (July/August). After completing gemmulation colonies died, usually from the beginning of August to October (Fig. 8). Active colonies were not observed during winter (November-March), only dead colonies with gemmules, Ephydatia muelleri often used its entire tissue for gemmule-production, whereas its skeleton remained intact for considerable period of time alier the death of the maternal sponge. Large numbers of gemmules were fixed by the skeleton to the place of production. In this way a successful recolonisation at the same site was ensured in the following year. In addition, when single gemmules became free and were no longer fixed to the substrate, they could be distributed by the current within the habitat, providing an effective mechanism for dispersal and recolonisation of adjacent habitats. Sexually produced larvae were not observed in this RIVER RHINE SPONGE ECOLOGY number of colonies/ m* al A - ] y ja A —— B IL 2 i - E y — , January-March April-June July-September October-December FIG. 7. Seasonal appearance of Ephydatia fluviatilis. species, and only a single symbiotic colony was found (from Sondernheim, in August 1994), close to symbiotic colonies of S. lacustris. Ephydatia muelleri is mainly distributed in deeper waters, below the levels affected during river-level fluctuations (Fig. 3). Trochospongilla horrida. Colonies of this species formed thin encrustations (less than 1cm thick), with a very irregular outline. Large colonies may cover an area of 0.5m^. Trochospongilla horrida began hatching from gemmules in spring (early April). The highest abundance, in both numbers and size of colonies, was reached in summer (August; Fig. 9). In autumn (September-October) colonies always disintegrated and left the gemmule crusts tightly adhered to the substrate. As in E. fragilis, gemmules remain fixed to their place of production and it is difficult to imagine that they might be dispersed within the river. Gemmulation commenced in early summer (June-July), and at this time especially T. korrida was a successful space-competitor against the otherwise dominating neozoan crustacean Corophium curvispinum (Amphipoda). When growing, small colonies tended to fuse with other colonies of the same species, thus forming larger ‘super- colonies’. Neither larvae nor colonies containing symbiotic algae were observed. The species was number of colonies; m? o Ww o | I. July-September October- December 2 January-March April-June FIG. 8. Seasonal appearance of Ephydatia muelleri. 221 mainly distributed in permanently flooded habitats (Fig. 3). Ecological strategies for each species are summarised in Table 2. DISCUSSION Freshwater sponges often display a considerable plasticity in their ecological strategies (Pronzato & Manconi, 1994a), and their life cycles are often adapted to the special requirements of their habitats (e.g. E. fluviatilis in temperate regions is usually active during summer and inactive during winter). In hot, arid regions the pattern of activity/inactivity is reversed (Harsha et al., 1983; Corriero et al., 1994). This shows that the life cycle is very adaptable to specific climatic conditions (Pronzato & Manconi, 1994b). According to Pronzato et al. (1993) the life cycle of E. fluviatilis seems to be controlled by exogenous factors in regions with strongly oscillating environmental conditions. In more stable habitats endogenous control seems to dominate. For several species the data presented in the literature differ from those presented here. For example, Bisbee (1992) reported the presence of active colonies year-round in S. /acustris from North Carolina. He observed gemmulation in late spring-early summer, sexual reproduction in April, and some sponges disappeared during summer. According to Cheathum & Harris (1953) both £. fragilis and T. horrida were active year- round in Texas. Pronzato & Manconi (1995) counted up to 324 gemmules cm” of tissue in E, fluviatilis from Sardinia. Ecological strategies of these species are given in Table 2. Ephydatia fluviatilis is considered to bean r-strategist, mainly for its ability to colonise new habitats. In contrast, the remaining species are characterised as k-strategists because they lack this ability. This general tendency is congruent with results of Pronzato & Manconi (1991), who compared E. fluviatilis and S. lacustris showing the former to be more successful in colonising new habitats, whereas the latter was more successful as a competitor. Details in the life cycle of E. fluviatilis seem to contradict this classification as an r-strategist: the dominance of sexual vs. asexual reproduction and its year-round presence; these are usually quoted as typical for k-strategists (Pianka, 1970). The successful colonisation of new habitats is here considered to be due mainly to active 331 number of colonies: av o 0 = in "Sod xw ME LR | | NE Jenuary-March Apri- lune July-September December FIG. 9, Seasonal appearance of Trochospongilla horrida: distribution of free larvae. The production of larvae in Spongillidae is mostly controlled by endogenous. factors (Gilbert et al., 1975). The seasonal appearance of larvae is confined to a few weeks in the year, and this timing is synchronous hetween various localities, Leveaux (1941, 1942) reported that production of larvae in E. fluviatilis in central Europe is confined to May-June, which «urresponds to the results presented here. Since timing in the production of larvae is severely constricted to certain weeks in the year and cannot be altered in the short term, producing larvae does not appear to be an effective strategy to react. to unpredictable environmental changes. The so-called k-strategists (T. horrida, S. lucustris, E. muelleri and E. fragilis) have the potential to produce a large number of offspring via their gemmules, traditionally an r-strategy feature. These species do produce a lot of gemmules, but these usually stay at the place of their production (see Table 2). According to Manconi & Pronzato (1991) S. lacustris follows the r-strategy in the short-term, bulis essentially a k-strategist in the long term. In many Spongillidae both strategies occur simultancously (Pronzato & Manconi 1995), even within the same structures: the gemmules serve as distnbution-units under an r-strategy, and as resting bodies under a k-strategy. Tt was surprising to see that asexual reproduction was clearly dominant over sexual reproduction, Larvae were only observed in £. fluviatilis and ta a lesser degree in E. fragilis. In the special limnological environment of running water it must be questioned whether or not gemmules are more suitable distribution- units than larvae, aside from their role as resting hedies. They are more robust than larvae, and dispersal is passive via eurrents, There is no need MEMOIRS OF THE QUEENSLAND MUSEUM or advantage in having a capacity for active dispersal. Van de Vyyer & Willenz (1975) reported that in E. fluviatilis (rom Belgium sexual reproduction is confined to overwintering colonies. In that species the development of oocyles commenced in autumn of the year prior to larval production, which occurred the following June. In Belgium, parts of colonies of E. fluviatilis survived winter as living, but reduced colonies, similar to the Rhine populations. These results are also confirmed by Weissenfels (1989), Williamson & Williamson (1979) discussed whether or not sexual reproduction was triggered by a phero- mone in Spongillidae. According to these authors sexual reproduction occurs rarely in running water because the postulated pheromone would be diluted and ineffective (in contrast to situations in stagnani water). This hypothesis would explain the occurence of many larvae in colonies of S. lacustris within the outflow of the ‘Steinbriicker Teich’ (see above), where the water flows quickly but is only 5-10cm deep. Here, the population of S. lacustris is so dense that à pheromone would not be diluted too much. In all colomes of this species at least some gemmules occurred in addition to larvae. The few reports of larvae in E. muelleriand T, horrida also originate from populations in stagnant water. The report of larvae in E. fragilis occurring in only one year (1994) suggests that sexual reproduction occurs in some years, even in species without regular sexual reproductive strategies, Many of the ditferent strategies and life-cycles mentioned above help species avoid competition or enhanee their competitive abilities. Competitive interactions among sponges, or between sponges and other organisms, are regularly observed m the field. As shown above, species have their highest abundance at different times.of the year (Figs 5-9, Table 2). Many details oT life histories can be interpreted as mechanisms tó enhance species! competitive abilities, including the fact that during periods of highest abundance of S, lecustris the proportion of colonies with symbiotic algae is also considerable (Pig. 5), These symbionts strongly enhance the growth of their host (Prost & Williamson, 1980). In T; horrida smaller colonies regularly fuse to form larger anes. Neubert & Eppler (1991) discussed wether the competitive ability was reduced, and therefore T. horrida wiks relatively rare compared to other freshwater species, However, my data show that it was the RIVER RHINE SPONGE ECOLOGY most abundant sponge in the Rhine, and is also very competitive among sponges and in competition with other organisms. It is concluded that competition for space is the species’ main challenge. ACKNOWLEDGEMENTS I thank the Hessische Landesanstalt fiir Umwelt for making it possible to use the research ship ‘Argus’, the crew of the ship is acknowledged for their good humour and helpfulness. Mr J. Wittmann kindly provided me with Spongillidae of his collections from 11.05.1994, 08.06.1994 and 30.06.1994, Dr H. Pohl helped me a lot with some figures. I am grateful to an anonymeous reviewer, who substantially improved the manuscript. This paper was written during a stay as a postdoctoral fellow in Israel financed by the DAAD. LITERATURE CITED ARNDT, W. 1926. Die Spongillidenfauna Europas. Archiv für Hydrobiologie 17: 337-365. 1928. Porifera, Schwámme, Spongien. Pp. 1-94, In Dahl, F. (ed.) Die Tierwelt Deutschlands und der angrenzenden Meeresteile nach ihren Merkmalen und ihrer Lebensweise. Vol. 4. (Porifera - Hydrozoa - Coelenterata - Echinodermata). (Gustav Fischer Verlag: Leipzig). BARTL, J. 1984. Biologische, chemische und physikalisch-chemische Bekämpfung von Ephydatia fluviatilis (L.) in Industriewasseranlagen. Inaugural-Dissertation der Naturwissenschaftlichen Fakultát. (Justus- Liebig-Universitát Gießen: Gießen). BISBEE, J.W. 1992. Life cycle, reproduction, and ecology of freshwater sponges in a South Carolina pond. I. Life cycle and reproduction of Spongilla lacustris, Transactions of the American Microscopical Society 111(2): 77-88. BRONDSTED, A. & BRONDSTED, H.V. 1953. The effect of symbiontic zoochlorellae on the germination rate of gemmules of Spongilla lacustris (L.). Videnskabeliga Meddelelser fra Dansk Naturhistorisk Forening 115: 133-144. CHEATHUM, E.P. & HARRIS, J.P. JR. 1953. Ecological observations upon the fresh-water sponges in Dallas County, Texas. Field & Laboratory 21(3): 97-103. CONRATH, W., FALKENHAGE, B. & KINZELBACH, R. 1977. Ubersicht über das Makrozoobenthon des Rheins im Jahre 1976. Gewässer und Abwässer 62/63: 63-84. CORRIERO, G., VACCARO, P., MANCONI, R. & PRONZATO, R. 1994. Life strategies of Ephydatia fluviatilis (L., 1758) in two different environments. Pp. 321-326, In Soest, R.W.M. van, Kempen, T.M.G. van & Braekman, to bo UJ J.-C.(eds) Sponges in time and space. (Balkema: Rotterdam). COURREGES, V.C, & FELL, P.E. 1989, Sexual and asexual reproduction by the freshwater sponge Anheteromeyenia ryderi, with emphasis on spermatogenic activity. Transactions of the American Microscopical Society 108(2): 127-138. FRANZ, H.W. 1992, Der Rhein und seine Besiedlung im Wandel: Schwebstoffzehrende Organismen (Hydrozoa, Kamptozoa und Bryozoa) als Indikatoren für den ókologischen Zustand eines Gewássers. Pollichia-Buch 25: 1-167. FROST, H.W. & WILLIAMSON, C. 1980. In situ determination of the effect of symbiotic algae on the growth of the freshwater sponge Spongilla lacustris. Ecology 61(6): 1361-1370. FROST, T.M., DE NAGY, G.S. & GILBERT, J.J. 1982. Population dynamics and standing biomass of the fresh-water sponge Spongilla lacustris. Ecology 63(5): 1203-1210. GILBERT, J.J. & SIMPSON, T.L. 1976. Gemmule polymorphism in the freshwater sponge Spongilla lacustris. Archiv für Hydrobiologie 78(2): 268-277. GILBERT, J.J,, SIMPSON, T.L. & DENAGY, G.S. 1975, Field experiments on egg production in the fresh-water sponge Spongilla lacustris. Hydrobiologia 46(1): 17-27. GUGEL, J. 1995. Erstnachweis von Eunapius carteri (Bowerbank 1863) (Porifera, Spongillidae) fiir Mitteleuropa. Lauterbornia 20: 103-110. HARSHA, R.E., FRANCIS, J.C. & POIRRIER, M.A. 1983. Water temperature: a factor in seasonality of two fresh-water sponge species, Ephydatia fluviatilis and Spongilla alba. Hydrobiologia 102: 145-150, KINZELBACH, R. 1995, Neozoans in european waters - exemplifying the worldwide process of invasion and species mixing. Experienta 51: 526-538, LEVAUX, M. 1941. Contribution al’ étude histologique de l'ovogénése et de la spermatogénése des Spongillidae. Annales de la Société Royale de Belgique 1941: 251-269. 1942. Contribution a l'étude histologique de Vovogénese et de la spermatogénése des Spongillidae (Suite). Annales de la Société Royale de Belgique 1942: 33-50, MANCONL R. & PRONZATO, R. 1991. Life cycle of Spongilla lacustris (Porifera, Spongillidae). A cue for environment-dependent phenotype. Hydrobiologia 220.155-160. MANN, K.H., BRITTON, R.H., KOWALCZEWSKI, A., LACK, T.J., MATHEWS, C.P. & MCDONALD, I. 1972. Productivity and energy flow at all trophic levels in the river Thames, England. Pp. 579-596. In Kajak, Z. & Hillbricht, A. l. (eds) Productivity problems of freshwaters. Proceedings ofthe IBP-UNESCO Symposium on Productivity Problems of Freshwaters Kazmierz Dolny, Poland, May 6-12, 1970 (UNESCO: Warszawa, Kraków). MUKAI, H. 1989. Growth and reproduction in four species of freshwater sponges in their natural surroundings. Science Reports of the Faculty of Education, Gumna University 38: 25-47. NEUBERT, E. & EPPLER, G. 1992. Trochospongilla horrida (Weltner 1893) - Neufunde aus dem Rhein. Lauterbornia 9: 59-63. PENNEY, J.T. & RACEK, A.A. 1968. Comprehensive revision of a worldwide collection of freshwater sponges (Porifera. Spongillidae). United States National Museum Bulletin 272: 1-184. PIANKA, E.R. 1970, On r- and K-selection. American Naturalist 104: 592-597, PRONZATO, R. & MANCONI, R. 1991. Colonization, life cycle and competition in a freshwater sponge association. Pp. 432-444. In Reitner, J. & Keupp, H.(eds) Fossil and recent sponges. (Springer Verlag: Berlin) 1994a. Adaptive strategies of sponges in inland waters. Bollettino di Zoologia 61: 395-401. 1994b. Life history of Ephydatia fluviatilis. a model for adaptive strategies in discontinuous habitats. Pp. 327-331. In Soest, R.W.M. van, Kempen, T.M.G. van & Braekman, J.-C.(eds) Sponges in time and space. (Balkema: Rotterdam). 1995. Long-term dynamics of a freshwater sponge population. Freshwater biology 35: 485-495. PRONZATO, R., MANCONI, R. & CORRIERO, G. 1993. Biorhythm and environmental control in the life history of Ephydatia fluviatilis (Demo- spongiae, Spongillidae). Bolletino di Zoologia 60. 63-67. RASMONT, R. 1962. The physiology of gemmulation in fresh-water sponges. Symposium ofthe Society for the Study of Development and Growth 20: 3-25. 1963. La róle de la taille et de la nutrition dans le determinisme de la gemmulation chez les MEMOIRS OF THE QUEENSLAND MUSEUM spongillides. Developmental Biology 8: . 243271. RUTZLER, K. 1965. Substratstabilitát im marinen Benthos als ókologischer Faktor, dargestellt am Beispiel adriatischer Porifera. Internationale Revue der gesamten Hydrobiologie 50(2): 281-292. SCHOLL, F., BECKER C. & TITTIZER, T. 1995. Das Makrozoobenthos des schiffbaren Rheins von Basel bis Emmerich 1986-1995. Lauterbornia 21: 115-137. SCHON, A. 1957. Massenentwicklung von Schwámmen in Rheinkraftwerken. Zeitschrift für Fischerei, N.F. 1(6): 85-86. SIMPSON, T.L. 1980. Reproductive processes in sponges: a critical evaluation of current data and views. International Journal of Invertebrate Reproduction 2: 251-269. M STEUSLOFF, U. 1938. Beitráge zur Okologie nordwestdeutscher Spongilliden. Archiv für Hydrobiologie 33: 309-338. TITTIZER, T., SCHOLL, F. & SCHLEUTER, M. 1990. Beitrag zur Struktur und Entwicklungsdynamik der Benthalfauna des Rheins von Basel bis Düsseldorf in den Jahren 1986 und 1987. Pp. 293-323. In Kinzelbach, R. & Friedrich, G. (eds) Limnologie aktuell 1 (Biologie des Rheins). (Gustav Fischer Verlag: Stuttgart). VAN DE VYVER, G. & WILLENZ, P. 1975. An experimental study of the life-cycle of the fresh-water sponge Ephydatia fluviatilis in its natural surroundings. Wilhelm Roux' Archiv 177: 41-52. WEISSENFELS, N. 1989. Biologie und mikroskopische Anatomie der SüDwasserschwáàmme (Spongillidae) (Gustav Fischer Verlag: Stuttgart) WILLIAMSON, C.E. & WILLIAMSON, G.L. 1979. Life-cycles of lotic populations of Spongilla lacustris and Eunapius fragilis (Porifera. Spongillidae). Freshwater Biology 9: 543-553. TOWARD A PHYLOGENETIC CLASSIFICATION OF THE MYCALIDS WITH ANISOCHELAE (DEMOSPONGIAE: POECILOSCLERIDA), AND COMMENTS ON THE STATUS OF NAVICULINA GRAY, 1867 EDUARDO HAJDU Hajdu, E. 1999 06 30: Toward a phylogenetic classification of the Mycalids with anisochelae (Demospongiae: Poecilosclerida), and comments on the status of Naviculina Gray, 1867. Memoirs of the Queensland Museum 44: 225-238. Brisbane. ISSN 0079-8835. Phylogenetic relationships for mycalids with anisochelae are revised. Several likely monophyletic species groups are included, currently assigned subgeneric rank or lower, totalling 12 groups, with special reference to Naviculina. The type species of Naviculina, N. cliftoni from SW Australia, is redescribed and its alleged relationship to Arenochalina mirabilis is contested, with more suitable affinities to 4egogropila. Its main anisochelae are termed here naviculichelae. A preliminary revision of over 230 published species names for Mycale and allied taxa with anisochelae was undertaken looking for N. cliftoni‘s kinship, yielding four likely candidates: M. cleistochela (from the W Indian Ocean and Indonesia), M. diastrophochela (from the Vema Seamount, SE Atlantic), M. obscura (from Indonesia and pan Australia), and M. peculiaris (from Papua New Guinea). A phylogeny is proposed for mycalids with anisochelae, although not fully resolved, and alternative phylogenetic classification schemes are hypothesised with discussion on the relative merits of each one. Porifera, phylogenetic classification, Mycale, Naviculina, phylogeny, Linnean hierarchy. Eduardo Hajdu (e-mail: hajdu@acd.ufrj.br), Museu Nacional, Departamento de Invertebrados, Universidade Federal do Rio de Janeiro, Quinta da Boa Vista, s/n, 20940-040, Rio de Janeiro, RJ, Brazil; Centro de Biologia Marinha, Universidade de Sáo Paulo, Sáo Sebastido, SP, Brazil; 15 March 1999. Naviculina was erected by Gray (1867) for Bowerbank's (1864: 252, pl. XXXVII, fig. 152) *naviculoid spiculum', thought by him to belong to Hymedesmia. Gray (1867) was mistaken in calling the spicule *equibianchorate', since its anisochelate condition was apparent in Bowerbank's illustration. Gray's (1867) concept of Esperiadae, to which he associated Naviculina, was essentially based on the possession of sigmas and/or chelae than on any other feature. Despite the excellent state of preservation of the type slide of N. cliftoni (Fig. 1), revealing a neat ectosomal reticulation of bundles of mycalostyles, no closer relationship was suggested by Gray (1867) between Naviculina or any other mycalid assemblage: Mycale, Aegogropila, Grapelia and Carmia. Hooper & Wiedenmayer (1994), on the contrary, considered Arenochalina Lendenfeld, 1887, a junior synonym of Naviculina, thus postulating a closer relationship between both taxa. It is the purpose of this article to explore the probable relationships of Naviculina and Aegogropila Gray, 1867 (as inferred from their sharing of a neat ectosomal reticulation), none of which is a synonym of the other in phylogenetic terms. It is postulated that Arenochalina deserves status as a valid subgeneric assemblage of Mycale, in view of its possession of a choanosomal stout quadrangular reticulation of spiculofibres (cf. Hajdu & Riitzler, 1998). Worldwide records of Mycale and allied genera are revised and a list of species which are best allocated to Mycale (Naviculina) is proposed. Hajdu & Desqueyroux-Faündez's (1994) cladistic analysis ofthe Mycalidae is reconsidered, with the inclusion of Mycale (Naviculina), and the likely monophyletic species-groups of Mycale. PHYLOGENETIC TAXONOMY. The need for a phylogenetic taxonomy has been recently stressed by de Queiroz & Gauthier (1992, and references therein), who claim that taxon names will never be explicit, universal and stable, as envisaged by the implementation of the diverse biological codes of nomenclature, if definitions continue to be assembled from lists of characters (but see Wiley, 1979). By accepting that characters may be reduced (lost), it is easily seen that groupings defined on these terms will frequently be artificial (Sundberg & Pleijel, 1994). The matching of evolution and systematics implies in the equation of species with population lineages, and of higher taxa with clades 226 (Chnstoffersen, 1995, Canto et al, 1997), De Queiroz & Gauthier [ 1992) proposed three ways of defining higher taxon names phylogenetically fas ammended by Schindler & Thollesson (1995), for the definition of taxon AB as implied by the phylogeny (€, (A,H))]; I) stem-based definitions, where taxon names are defined as the most inclusive clade that contains taxon A but not taxon C; 2) node-based definitions, where taxon names are defined as the least inclusive clade that contains taxa A and B: and 3] apomorphy-based definitions where taxon names are defined as the most inclusive clade containing some synapomorphy of AB . Biological classifications are built over categories created over 200 years ago by Linnaeus (1758). Evolutionary ranking adds retrievable information content to the classification, and is thus necessary, but Linnean categories are based un the essentialistic logic of Aristotelis, where ranks ure absolutely arbitrary, with no implied meaning across distinct taxonomic groupings (de Queiroz & Gauthier, 1992; Christoffersen, 1995), Is there any sense in comparing an Order of Denmospongiae with one such taxon of the Polychaeta? Taxa placed at the same categorical level do not represent equiyalent entities (Sundberg & Pleijel, 1994). The Linnean hierarchy has proved a constraint where diversity is considerable, hence the need for a super- subtribe, for instance. Linnean categories add no stability to the names of taxa as changes in rank imply changes in suffixes (at least). It promotes redundanoy via mandatory categories and the principle of exhaustive subsidiary taxa (e.p. Cantino. et al., 1997). Biological classification might benefit from a change in paradigm. Given the above rationale, phylogenetic vlassificatory schemes are proposed for the myoalid phylogeny where evolutionary hierarchy is preserved (retrievable) Following some guidelines revised in Amorim (1997). In two of these the Linnean ranking is preserved. In the last proposed elassification Linnean ranking is abandoned altogether. MATERIALS AND METHODS Specimens were studied under light and scanning electron microscopy, Preparations of thick sections and discuta spicules were made using procedures described clsewhere (Hooper, 1991, 1997; Hajdu, 1994), Spicule measurements are given as minimum - meun = maximun dimen- sions in micrometres. SEM study was partly MEMOIRS OF THLE QUEENSLAND MUSEUM performed in à Jeol ISM 35-L machine, under an accelerating voltage of 25KV, working distance of 15mm, and magnifications of up te 3600x; partly in a ZEISS DSM-940 machine, under accelerabing voltages between 17 and 19KV. working distance around 3mm, and magnif- ications up to 10000x. Phylogenetic analyses were performed using PAUP 3.1.1 (Swofford, 1993) with a choice for the ACCTRAN algorithm. Characters were treated as unordered and equally weighted on a first run, Subsequent weighting was applied on the basis of character’s rescaled consistency indices. The phylogenetic classificatory schemes proposed are based on the guidelines revised by Amorim (1997). The four basic rules are: 1) every taxon must be monophyletic, or alternatively, if doubt persist, it must be stated clearly; 2) all the known levels of generality must be recognisable from the classification; 3) sister-group relationships must be recognisable; and 4) it must be possible to know to which larger clade a smaller clade pertains. Nelson (1972) identified two ways of assigning: less general clades to more general ones! subordination and 'sequenciation'. In subord- inated classifications sister taxa share the same laxonomic category, and less inclusive clades are always associated to lower ranks than the more universal clades. In sequenced classifications, sister taxa need not share the same rank, but rather, successive lateral branches ate associated io the same rank, Subordination has thus the advantage of naming every clade, what may turn into a disadvantage if there are more known levels of generality than taxonomic categories are available (but sec Farris, 1976). Still, the finding of additional levels of generality (i.e. the inclusion of new taxa in (he phylogeny) implies in considerable rearrangement of categories in à subordinated system, A more basic idea concerns the use of indentation to reflect distinct levels of universality (Wiley, 1979), a strategy adopted below. In the alternative phylogenetic classifications proposed. no clade is given a new name due to the preliminarity of the exercise undertaken. In the classifications furnished below no choice was minke for ejthersubordination or "sequenciation”. as no new name js advanced and no Linnean category changed. Some procedures suggested by Wiley (1979), Amorim (1982, 1994), Christoftersen (1988) and Papavero et al. (1992), MYCALIDAE PHYLOGENY 227 TABLE 1. Rules adopted for the build-up of phylogenetic classifications from phylogenies, and their reference sources. References Rules Wiley (1979) ‘sedis mutabilis’ for taxa pertaining to politomies *group- for the more inclusive unnamed clade, which includes a less inclusive named group and its Amorim (1982) sister-group. ‘group++’ for the more inclusive unnamed clade, which includes a less inclusive | “group+” and its sister-group. Christoffersen ‘[taxon X]’ for the lower rank taxon of a monotypic redundant higher rank one, where intermediary (1988) categories are simply ommited Papavero et al. (1992) *group-1, -2, -3, ...” - clades receive the names of their oldest included genus or species (every other rank is abandoned), to which a negative index is added to indicate the number of speciation events occurred between it and the actual taxon Amorim (1994) ‘group*’ for the more inclusive unnamed clade, which includes a less inclusive named group pertainig to a polytomy, and its sister-group were used in the construction of the classifications advanced below (see Table 1). Abbreviations: BMNH, The Natural History Museum, London; INV-POR, Instituto de Investigaciones Marinas de Punta de Betin, Santa Marta, Colombia - Porifera collection; MNHN-LBIM, Muséum National d’Histoire Naturelle, Paris, Laboratoire de Biologie des Invertébrés Marins et Malacologie; MNRJ, Museu Nacional, Universidade Federal do Rio de Janeiro, Rio de Janeiro; SMF, Senckenberg Museum, Frankfurt; UERJ, Universidade do Estado do Rio de Janeiro, Rio de Janeiro; USNM, National Museum of Natural History, Smithsonian Institution, Washington D.C.; USP, Universidade de Sáo Paulo, Sáo Paulo; ZMA, Zoologisch Museum Amsterdam, Amsterdam; ZMH-S, Zoólogisch Museum Hamburg, Schwam/Sponge collection. SYSTEMATICS Class Demospongiae Sollas Order Poecilosclerida Topsent Suborder Mycalina Hajdu, Van Soest & Hooper Family Mycalidae Lundbeck Naviculina Gray, 1867 DIAGNOSIS. Mycale with an ectosomal skeleton composed of a reticulation of megasclere budles. Anisochelae include naviculichelae (complete or near fusion of both frontal alae, falx markedly expanded along the shaft, lateral alae of the head project backward and upward). Type species: N. cliftoni Gray, 1867 (by monotypy). Naviculina cliftoni Gray, 1867 (Figs 1-2) Hymedesmia sp.; Bowerbank, 1864: pl. 37, fig. 152. Naviculina cliftoni Gray, 1867: 538; Hooper & Wiedenmayer, 1994: 293. MATERIAL. HOLOTYPE: BMNH 1877.5.21.270: Freemantle, Western Australia (type slide). COMPARATIVE MATERIAL. HOLOTYPE of Mycale diastrophochela Lévi, 1969: MNHN LBIM DCL1447: Vema Seamount, SE Atlantic. HOLOTYPE of Mycale cleistochela Vacelet & Vasseur, 1971: MNHN LBIM DJV36: Tulear, Madagascar. SPECIMENS: M. cleistochela ssp. flagellifer: MNHN LBIM DJV35: det. J. Vacelet & P. Vasseur, Tulear, Madagascar. ZMA 8512: det. R.W.M. van Soest, Sumbawa, Indonesia. ZMA 8896, 8897, 8912, 8917: det. R.W.M. van Soest, Tarupa Kecil, Indonesia. ZMA 12660: det. R.W.M. van Soest, Mahé, Seychelles. HOLOTYPE of Mycale obscura (Carter, 1882): BMNH 1881.10.21.318: Torres Strait, Queensland. SPECIMENS: BMNH 1925.11.1.732: det. M.E.Shaw, Tasmania, Australia. SMF 1041: det. E. Hentschel, Aru, Indonesia. ZMA 1602: det. M. Burton, Indonesian ‘Siboga’ material. ZMH-S 1670: det. E. Hentschel, Sharks Bay, Western Australia. SPECIMENS of Mycale spp.: INV-POR 2198: det. S. Zea, Colombian Caribbean. USNM 34348: det. Mote Marine Lab., off Florida, Gulf of Mexico. USNM 41555: det. E. Hajdu, Florida, Gulf of Mexico. MNRJ 263, 362, 425, 773: det. E. Hajdu, São Sebastiáo, Brazil. REDESCRIPTION OF NAVICULINA GRAY, 1867. One single thick-section slide preparation remains. It contains a perfectly preserved fragment of the specimen's surface peel, from which it is possible to gather the whole series of spicules in Mycale. This peel contains an ectosomal skeleton characterised by a neat reticulation of megasclere bundles (2-6 spicules across) or single megascleres, forming meshes which are mostly triangular (40x70-240x350um across), and inside which pores are clearly visible (60um across). Naviculichelae abound inside the 228 MEMOIRS OF THE QUEENSLAND MUSEUM FIG. 1. Mycale (Naviculina) cliftoni (Gray, 1867). BMNH 1877.5.21.270. A, Neat ectosomal reticulation of megasclere bundles or single megascleres, most meshes triangular, pores clearly visible. Naviculichelae abound inside the meshes, specially downward expansion of the upper falx along the shaft. Dimensions: 12-/7.3-21.6um long (N=100). Sigma, slender, smooth, sharp endings. Dimensions: 14.4um long (N=1). REMARKS. The term naviculichela is proposed here for Bowerbank’s (1864) “naviculiform spiculum”. It is a type of anisocleistochela where there is complete or near fusion of both frontal alae (cf. Boury-Esnault & Riitzler, 1997), the falx is markedly expanded along the shaft (Fig. 2A), and the lateral alae of the head project backward and upward (Fig. 2B) encircling the shaft. Another common feature is the extreme narrowing of the frontal ala, in sucha way that it becomes thinner than the shaft itself (Fig. 2C). The term cleistochelae was first used by Topsent (1925) for the isochelae of a Clathria, a much simpler morphotype than that observed in Naviculina cliftoni, and related forms (e.g. Mycale cleistochela Vacelet & Vasseur, see below). In N. cliftoni over 80% of the naviculichelae are 16.8-19.2um long, and it is possible there are two | categories [possibly 12-16.8 (N=15) ' and 18-21.6um long (N=85)], but this is unclear from the distribution of spicule size categories, The origin of the single sigma observed is also dubious, possibly a contaminant. around the bundles of megascleres. B, Mycalostyles, naviculichelae and a single sigma (arrow). Scale bars 100um. meshes, especially around the bundles of megascleres (Fig. 1A). Spicules (light microscopy only, Fig. 1B). Megascleres: Mycalostyles, smooth, mostly straight, slightly fusiform, with elliptic or oval heads, and points which taper more-or-less gradually. Dimensions: 330-357.4-388um long (N=20), 4.8-8.4um thick (head, N=10), 6-9.6um thick (shaft, N=10). Microscleres: Naviculichelae, head 60-70% the total spicule length, with narrowing and complete regression of the frontal alae of the head, which may touch the one of the foot, lateral alae of the head projecting backward and slightly upward, DISCUSSION A survey of nearly 230 published descriptions of Mycale revealed there were four species bearing naviculichelae-like anisochelae. These are: M. cleistochela Vacelet & Vasseur, 1971, M. diastrophochela Lévi, 1969, M. obscura (Carter, 1882), and M. peculiaris Pulitzer-Finali, 1997. However, the status of M. cleistochela ssp. flagellifer Vacelet & Vasseur, 1971 remains uncertain. It is possibly a separate species based on its distinctive microsclere complement, but a decision is not possible until detailed morphological comparisons are made of both taxa, which is beyond the scope of this present contribution. MYCALIDAE PHYLOGENY FIG. 2. A, Naviculichelae - anisochelae characterised by the complete or near fusion of both frontal alae (a). falx (f) markedly expanded along the shaft. B, Lateral alae (la) of the head projecting backward and upward, encircling the shaft (s). C, Extreme narrowing of the frontal ala (a), in such a way thal it becomes thinner that the shaft itself, (A, C, MNRJ 773, Mycale (Nevieulina) sp. B, USNM 41555, Mycale (Naviculina) sp. All scales bars Sum. All five taxa have reat ectosomal reticulations, potentially referable lo Mycale (Aegogropila) (i.c. differentiated from other mycalids in having a neat ectosomal reticulation, but no serrated sigmas, and no isochelae). However, this act would be inconsistent with the use of M. (Paresperella), for example, used for species with serrated-sigmas, to my knowledge, all bearing an ectosomal reticulation (Hajdu & Desqueyroux-Faündez, 1994). A parallel can be made with the use af M, (Grapelia) for species with unguiferate anisochelae, all of which possess a confused tangential ectosomal architecture, typical of M. (Mycale) (Hajdu, 1995). Despite the fact that these names have been variably used as genera or subgenera in the past, it appears obvious that they comprise assemblages at distinct levels of universality. This is a common problem in the Linnean hierarchy, which has traditionally been ignored by the mere proposal of new scientific names for every assemblage recognisable an the basis of some more-or-less conspicuous trait; coupled to the acceptance of dustbin/plesiomorphic assemblages (i.e. incertae sedis). PROPOSAL OF A FRAMEWORK FOR THE CLASSIFICATION OF THE MYCALIDS WITH ANISOCHELAE. The phylogenetic analysis af the Mycalidae undertaken by Hajdu & Desqueyroux-Fatindez (1994) has been reconsidered in light of the information derived B =" from re-examination of Mycale (Naviculina). Characters and taxa included in the analysis were reevaluated in view of decisions taken in Hajdu (1995). Adveale (Anomomyvale), M. (M) immitis-group, M. (Naviculina), M. (Oxymycale), M. (Rhaphi- doteca), and M. (Zvgomyvale) have been added to the list of taxa considered here. Esperiopsis-] and -H were used as the outgroups, referring 1o those species conforming to E. villasa Carter, 1882 and E. fucorum (Esper, 1794), respectively, A list of 22 characters and their 30 states used in the cladistic analysis is given in Table 2, Taxa and their character states are tabulated in the datamatrix shown in Figure 3. Figure 4 shows the preferred tree, selected with the purpose of advancing a discussion on phylogenetic classifications in mind (see below). It is a majority-rule consensus of 81 trees (50 steps. CI-0.94, RI-0.87. RC=0.82), filtered for more-resolved topologies from 1981 most parsimonious trees generated hy PAUP's Branch and Bound exact algorithm for the datamatrix in Figure 3. Characters were treated as unordered, and multistate taxa were considered to be polymorphic, Following the suggestion by Nixon & Davis (1991), both Mycale (degogropila) and M. (Carmia) were split into terminal taxa -I and -ll, to account for presence vs. absence of micracanthoxeas, respectively. In this way, the discussion adyanced by Carballo & Hajdu (1998) on the status of micracanthoxeas within the mycalids can hopefully be refined. Ideally, this procedure would have been extended to every taxon polymorphic for one or more characters, but this would further reduce the resolution attained in Figure4, through the addition of many more terminal taxa. From this analysis neither Mycale (Aegogropila) nor M, (Carmiu) ate indicated as likely to be monophyletic. Carballo & Hajdu's (1998) hypotheses 2 and 4 appear mote probable explanations for the observed distribution of micracanthoxeas, These hypotheses. stale, respectively, that either species that possess micracanthoxeas form a monophyletic clade, and one or both subgenera are palyphyletic: or poor taxonomic resolution (and/or interpretation) TABLE 2. Morphological characters and their character-states used to build the datamatrix in Figure 3. MEMOIRS OF THE QUEENSLAND MUSEUM being the only real synapomorphies within the mycalids with Characters Character states anisochelae. . Categories of m le 1. Categories of megascleres commi 0: one, 1: two or more rare, 2: two or more A posteriori weighting of characters in Figure 3 by their 2. Main megascleres only 0: (mycalo)styles only, 1: exotyles too; 2: oxeas rescaled consistency indices does reduce the number of most 3. Three categories of chelae 0: absent, 1: present parsimonious trees to 416 (36 after 4. Basic shape of chelae 0: isochelae, 1: anisochelae filtration), but this occurs at the > Rosées categories maybe rare 0; absent, 1: one category (maybe rare), 2: two expense of resolution. The majority-rule consensus 1s similar 6. Anisochelae-I with shaft markedly curved on profile view D: absent, T; présent to Figure 4, but (Oxymycale), (Naviculina, Paresperella, 7. Anisochelae-I ratio height of the n2 ` ay 2:<2 head/total height of the spicule in % 0:>35, 1:2535,2; 25 Zygomycale), and ((Aegogropila-1, 8. Anisochelae-I unguiferate 0: absent, 1; present Carmia-L), Carmia-Il) compose a 9. Anisochelae-I shape of the foot 0: normal (falx basal), 1; with pore (falx hidden within the alae), 2: contorted and denticulated polytomy next to the mycalids with a confused tangential 10. Anisochelae-II acanthose 0: absent, 1: present ectosomal architecture. 11. Anisochelae-II 0: larger than IIL, 1: can be smaller than III It is interesting to note from the 12, Anisochelae-II and/or 111 (naviculichelae) with falx extending downward along the shaft consider- ably o : absent, 1: present present analysis that the absence of an ectosomal skeleton in Mycale (Carmia) appears as a possible 13. Anisochelae-II and/or III (naviculichelae) with frontal ala of the head extremely narrow (as thick as the falx itself) 0: absent, 1: present subsequent loss, as opposed to the findings reported by Hajdu & Desqueyroux-Faündez (1994). As argued elsewhere (e.g. Hajdu & 14, Anisochelae-II and/or III (naviculichelae) with lateral alae of the head bent backward encircling the shaft 0: absent, 1: present Van Soest, 1996), some losses are likely to be easily achieved, and conversely it could be expected 15. Anisochelae-III with a basal spur- like projection : absent, 1: present that parallel developments might also have occurred. Hajdu & Rützler (1998) reported on a M. 0 16. Micracanthoxeas 0: absent, 1: present 0 17. Serrated sigmas : absent, 1: present (Aegogropila?) which can have an 18. Toxas 0: absent, 1: present ectosomal reticulation, or may 19, Raphides F ries 0: absent or one category, 1: maybe two catego- lack any ectosomal skeleton 20. Ectosomal skeleton 0: absent, 1: reticulated; 2: confused whatsoever, thus supporting the hypothesis that such ectosomal 21. Choanosomal skeleton 0: absent, 1: stout quadrangular reticulation architectures have a low adaptive 22. Pore-grooves 0: absent, 1: present value. In other words, a careful prevents us from accessing the occurrence of micracanthoxeas in M. (Aegogropila-11) and M. (Carmia-11), which would be monophyletic instead. The strict consensus for the 81 trees selected holds the monophyly of (Aegogropila-l, Carmia-1) and of (Anomomycale, Mycale (Grapelia, immitis-group, Rhaphidoteca)). If we exclude the micracanthoxeas as potentially good synapomorphies, due to their largely underestimated occurrence, we are left with: 1) a confused tangential ectosomal skeleton, and 2) anisochelae-I markedly curved in profile view, study of species currently assigned to M. (Carmia) may indicate a more appropriate allocation in several distinct monophyletic assemblages, related to assemblages bearing ectosomal specialisations. In these cases assemblages sharing the absence/loss of ectosomal specialisation would not form a monophyletic clade, as already foreseen by the inferred relationships between Arenochalina and Carmia. PHYLOGENETIC CLASSIFICATION EXERCISES. Several proposals have been made in the specialist systematics literature, as to howa phylogenetic classification (i.e. one that reflects the relationships among taxa, should be MYCALIDAE PHYLOGENY 231 TaxalChar. 1 2 3 4 5 6 7 8 9 |10| 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 Aegogropila 0/|0/01| I 1 0/|0/[|0/[|0/|/0/,0/0/|0/09/,0/,1/0/17/0/01/|0],0 Anomomycale 2|0/0]|1 1[0/0/0/j,2/|0/0,0/,0/|0/0//0/0/|0/0/,2 0,0 Arenochalina 0|0/]|0/|1 1/06/0/[0/0/0/,0/,0/,0/0/0/0,0/0/|0]/0)]! 0 Carmia 0[0/]01| 1 1/0/0/0/0/,0/0,0/0/0/0/,1/;,01/|09/|0/0/,0 Grapelia 0/01 1] 2]1/2]1 1 1[0/|0/0/0/1/|0/,0/,0/|0/|2/,0/,0 immitis-group 2|0|lI 1 1|12|0|1/|0/|0|0|0|0|0|01,0|0|0|j01 2) 0; 0 Mycale 210/01] 1 1[0/0/0/0[]0/0/0/jJ0/,001/]|0/0/,0/01,2,;0]0l Naviculina Oo | 0/01] I 1[0/0/,0/0/0/|0./;]1 1 1/[0/]909/0]!1 0|1/0/|0 Oxymycale 020,1 1[0/0/0/,0/0/[0/0/0/0 .0/,0/0/0/,0/,1] &0/|0 Paresperella 1[0/[0,1 1[0/0/,0/0/,0/[|0/0/0/0 0/;0]|l 1[0/|1/0/|0 Rhaphidoteca 9. 1 1 1 {01| 1 0 1 0/0/|0/,0/[|[0/0/0/0/0/|01/,2/0]0 Zygomycale 0[|[0/1/01 1]0/[|0/,0/j/0,0/,0/|0/0/,0/,0,0/|0/|1/]/0/,|1 0/0 Esperiopsis-l 0[0/01/0/01/0/0/,0/,0/,0/0/[|[0/0/0/,0/,0/|0/|0/0/,0/0 0 Esperiopsis-Il 0/[0/[0/,0/,0/|0/[0/0/0/0/|/0/[0/0/0/j/0/,0:0/0/0/0/0/0 FIG. 3. Datamatrix showing the 14 mycalid taxa and their character states, for 22 characters used in the phylogenetic analysis undertaken here. (Refer to Table 2 for a list of characters and character states). constructed (e.g. Nelson, 1972; Griffiths, 1974; Wiley, 1979; de Queiroz & Gauthier, 1990, 1992; Papavero et al., 1992; Papavero & Llorente- Bousquets, 1993; Amorim, 1997). These classifications were proposed under the protocols of subordination and sequenciation, with a minor or major relation to Linnean categories. Both paleontologic, as well as biogeographic data have been variously included in several proposals, thus enhancing enormously the information content of classifications. There are at least six distinct levels of universality (for clades inferred here to be monophyletic), for mycalids with anisochelae (Fig. 4). If the terminal taxa are considered species-groups, the next five hierarchic levels (each one successively more inclusive than the preceding), could be, for instance subgenera, genera, tribes, subfamilies and families. This is the outcome of extreme commitment to Linnean hierarchy in a subordinated system, where sister taxa are always assigned similar taxonomic rank (Nelson, 1972; Amorim, 1997). In such an arrangement, Mycale (mycalids with anisochelae), would be named a family instead. Every time a new clade is found through refining phylogenetic analyses, considerable changes would have to be implemented in the Linnean hierarchies. In an extreme situation, especially applicable for speciose groups, there might be more recognised hierarchic levels than Linnean categories. Farris (1976) proposed a series of prefixes (Super-, Hiper-, Mega-, Giga-, and Sub-, Infra-, Micro-, Pico-) to allow the establishment of a nearly infinite number of categories, but is this what we seek in a pragmatic systematics? There are alternatives. Amorim (1982, 1994) suggested a coding strategy through which inclusive taxa (e.g. a ‘potential family’) would receive the name of their most basal taxon (e.g. a genus), coupled with a ‘+’ to state that the ‘potential family’ includes the mentioned genus plus its sister-group. This occurs when relationships are resolved within the inclusive taxon, whereas if they are not, the chosen name would be that of the oldest taxon coupled to an ***. Figure 5 shows the translation of the clado- gram in Figure 4 into one such classification, using the suggestion by Wiley (1979) regarding the labelling of taxa pertaining to politomies. The advantages of this system are that phylogeny is retrievable, and changes in hierarchic levels need not reflect changes in Linnean categories, thus confering, stability to names used day-by-day by non-specialists. Moreover, taxa may be kept at the hierarchic level to which they are currently assigned, and new names need not be established for every new clade. Instead, use is made of available names coupled to a symbol. The disadvantage is not exclusive. The same Linnean category may have several distinct ontologic meanings (Fig. 5). What does a subgenus mean? Subgenera represent five distinct levels of generality in this classification. They are employed for the sake of stability only. However, suggestions were made in the past to confer ontologic meaning to Linnean categories. 00 100 FIG. 4. Majority-rule consensus tree obtained from 81 trees filtered for more resolved topologies, and from 1981 most-parsimonious trees, showing the phylogenetic relationships of mycalids with anisochelae, obtained by analysing the datamatrix in Figure 3 using PAUP. Refer to Table 2 for a list of characters and states. These include an association between Linnean categories and a palaeontological age (e.g. Hennig, 1966), and the association of categories to their supposed biogeographic origin (Amorim, 1992). Untortunately, neither source of evidence is readily available for the mycalids, or for that matter for most of the Poecilosclerida. The oldest likely anisochelae-bearing mycalid (evidenced from a single palmate anisochela), is, to my knowledge, from the Early Cretaceous (Albian) (Wiedenmayer, 1994). This anisochela is only tentatively assigned to Mycale, and no finer allocation is possible. The next anisochela in the geologic scale is a curved, palmate form described by Gruber & Reitner (1991) (Lower Campanian, Cretaceous). This anisochela was assigned to a group within Mycale, the curved-assemblage of Hajdu (1995; as “sp.2”). Its position in the cladogram (Hajdu, 1995, fig. 7.4) is more basal than both M. (Grapelia) and M. (M.) immitis species-group, but it is more derived in comparison to the remaining M. (Mycale). In other words, there are subgenera of Mycale that are probably younger than the Campanian (ca. 80 Myr), and others that are probably older. Hajdu's (1995) 'sp. 3' is from the Eocene-Oligocene transition, thus younger than both records cited above, but sits in a more basal position in the cladogram. It is, therefore, pointless to assign geologic ages to categories within Mycale, on the basis of such a meager and patchy database. MEMOIRS OF THE QUEENSLAND MUSEUM Aegogropila- | An alternative methodology Carmia- | is to look for biogeographic Carmia- I| origin. According to Van Soest Naviculina (1994), most demosponge Zygomycale higher taxa (suprageneric taxa) Paresperella have notably wide Aegogropila - distributions, an indication of Oxymycale their probable early ancestry Anomomycale (e.g. Van Soest & Hajdu, 1997). Grapelia Knowledge of global tectonic immitis-group events prior to the Triassic is Rhaphidoteca fragmentary, so that no precise Mycale link would be possible between Arenochaina Clades supposed to have Esperiopsis- 1 Originated before the break-up esperiopsis- 1 Of Pangea and some likely original crustal plate. Additionally, Amorim (1997) stressed that any implementation of a classif- ication in which clades are directly linked to biogeographic categories is clearly dependent on the elaboration of a well established general area cladogram. Marine areas have been dealt with recently by many sponge specialists (Hooper & Lévi, 1994; Hajdu, 1995; Van Soest & Hajdu, 1997), and the general area cladograms generated would certainly form a framework over which to advance a classification along the lines suggested by Amorim (1992). Further speculation in this direction, however, is not possible until we draw a much clearer picture on the distribution of mycalids, as well as obtain well-supported cladograms for marine areas: all necessary prerequisites for the implementation of Amorim’s (1992) suggestions. For the time being, I propose a working hypothesis (Fig. 6), as a way of overcoming the problem of multiple significance of Linnean categories arrived at in Figure 5. This scheme takes into consideration the suggestion of Chiristoffersen (1988) on redundant taxa (Table 1), used here as an artifact to respect the hierarchic level of Linnean categories. The fundamental taxon is the genus Mycale. The terminal taxa are either subgenera or monophyletic species-groups, such as the M. (M.) immitis-group. Intermediate hierarchic levels are simply named ‘groups’, with their terminal taxa included within brackets, if monotypic. Finally, a phylogenetic classification (Fig. 7), based on the cladogram in Figure 4, was built MYCALIDAE PHYLOGENY Genus Mycale Gray, 1867 a Subgenus Arenochalina Lendenteld, 1887 group Mycaqler+ is Bro up Mycale+ EDO Subgenus Mycale ap ateei abest group Anomomycale+ Topsent, 1924 ss BSubgenus Anomomycale e group Grapelia* rar DUbgenüs Grapelia leigt group Oxymycale+ Hentschel, 1929 ranan nn DUbgenus Oxvanveale eee group. Aegogropila-I* Gray, 1867 peces slibgenus Aegograpila-ll s, m. eere perpe eget Subgenus Paresperella Dendy, 1905 s. m. TNT group Carmia-Il+ s, m, ' e Subgenus Carmia-ll MIA e eoereg group Aegogropila-1+ mam Subgenus Aegogropila-l creme a subgenus Carmia-l af nZansosseener gerd group Naviculina+ Gray, 1867 s. m. cn Subgenus Naviculina s Subgenus Zygomycale Topsent, 1930 FIG. 5. Phylogenetic classification of the mycalids with anisochelae built from the cladogram in Figure 4, by subordination, incorporating the coding strategy of Wiley (1979), and Amorim (1982). Key: ‘group+’, more inclusive taxon containing the taxon formally described with that name (the basalmost taxon, preserving its current Linnean hierarchic status) and its sister-group; ‘group**, more inclusive taxon (the oldest available name included), the relationships of its included taxa being unresolved. The priority of Rhaphidoteca Kent, 1870 over the Mycale immitis (Schmidt, 1870) species-group was decided not on the basis of knowledge of the actual dates of publication of both works, but on the fact that the species- group is bound to be named in the future, becoming then a much younger taxon: s.m, sedis mutabilis. under the provocative protocol of Papavero et al. (1992) and Papavero & Llorente-Bousquets (1993), in which Linnean categories are simply abolished. Names considered are only those of genus- and/or species-level taxa, eventually coupled to an index which indicates their hierarchic level on the phylogeny. For example, Mycale immifis (Schmidt, 1870) was used instead of M. (AZ) immitis-group of Hajdu (1995). Its coupling to a *-5' index means that the supposed ancestor of the M. (M.) immitis-growp (the terminal taxon used in the present analysis), 1s 5 hypothetic ancestral species away from the true M. immitis species (Hajdu, 1995; fig, 7.4, ‘species 10°). Thus, ancestor *-1” is the ancestor of M. immitis + its sister group ("species 6-9" of Hajdu, 1995); ancestor '-2" refers to ‘species 1-10”; ancestor *-3” to ‘species 1-12”; ancestor 4 to ‘species 1-16"; and, ancestor *-5* to the entire M. (M) immitis-group (viz. ‘species 1-17" of Hajdu, 1995). Accordingly, Mycale.; is the appropriate nomenclature in this case because Myca/e is the oldest available name within the studied clade (page and line priority considered), coupled to the observation that the root of the clade containing all the mycalids with anisochelae is three ancestors away from the terminal M. (Mycale). Ancestor *-1” is the ancestor of (Myc, (Ano, (Gra,(Rha, imm-group)))). Ancestor *-2" refers to the latter clade and its sister taxon, (Aeg-IL Par, (Nav, Zyg), ((A4eg-l, Car-I), Cur-M)). And, ancestor '-3" to the whole ingroup. PHYLOGENETIC DIAGNOSES FOR TIIE TERMINAL TAXA CONSIDERED. The clado- gram in Figure 4 is a weakly supported working hypothesis. Accordingly, there are unnamed, inore-inclusive clades remaining because relationships are bound to shuffle with the inclusion of additional terminal taxa. Nevertheless, this does not preclude the establishment of phylogenetic diagnoses for the taxa considered, as any phylogenetic hypothesis is better than no hypothesis at all. The cladogram (Fig. 4) is viewed as an improvement over the hypothesis put forward by Hajdu & Desqueyroux-Fatindez (1994) because it is a more comprehensive sample of probably monophyletic Genus Mycale Gray, 1867 - group Arenochalina [Subgenus Arenachalina Lendenteld, 1887] Tv group Mycalel! per LOU) Mycalet group Mycale [Subgenus Mycale] aces BEOU d romunmyedle* Topsent, 1924 seccenbbpounderesbt group dnomomvcale [Subgenus Anomomvcale] sonar ron Grapeliad Gray, 1867 mTM-—- cesa group Grapelia [Subgenus Grapelia| eee LOU Khaphidetevat Kem, 1870 ene oDübgenus Rhaphidoteca group Oxymycale+ Hentschel, 1929 copo group Oxpmycale [Subgenus Oxvmveale| m— group degogropila-11* Gray, 1867 eene BrOup Aegogropila-I [Subgenus Legogrupila-11] s. m. PS group Paresperella Dendy, 1905 [Subgenus Paresperelfa] s, m. "— nel group Carmia-11+ Gray, 1867 s. m. „group Carmia [Subgenus Carmia-1] group Aegogropila-IE* emnes Dübgenus Aegogropila-l A a Subgenus Curmia-l ios BroUp Navicnlinat Gray, 1867 $. m sas Subgenus Navientina borers Subgenus Zvgomycale Topsent, 1930 FIG. 6. Phylogenetic classification of the mycalids with anisochelae built from the cladogram in Figure 4, by subordination, incorporating the coding strategy of Wiley (1979), Amorim (1982) and a parallel of Christoffersen's (1988). Key: *group*', more inclusive taxon cont- 2 aining the taxon formally described with thal name (the basalmost taxon, preserving its current Linnean hierarchic status) and its sister- group; ^group*'. more inclusive taxon (the oldest available name included), the relationships of its included taxa being unresolved; s.m, sedis mutabilis. species-groups within Myeale than the earlier attempt. The proposed scheme is as follows: Subgenus Aegogropila-1 - Mycale with a reticulated tangential ectosomal skeletou and micracanthoxeas (many with toxas, and three categories of anisochelae). Subgenus degogropila-11 - Mycale with a reticulated tangential ectosomal skeleton (many with toxas, and three categories of anisochelae). Subgenus 4nomomycale - Mycale with à confused tangential ectosomal skeleton and anomochelae. «Mycale immitis (Schmidt. 1870) species-sroup MEMOIRS OF THE QUEENSLAND MUSEUM Subgenus Arenochalina - Mycale without any ectosomal skeletal specialisation, and with a stout choanosomal architecture composed of spiculofibres arranged in quadrangular meshes. Subgenus Carmia-1 - Mycale without any ectosomal skeletal specialisation and micracanthoxeas (many with toxas, and three categories of anisochelae). Subgenus Carmia-11 - Mycale without any ectosoinal skeletal specialisation (many with Loxas, and three categories of anisochelae). Subgenus Grapelia - Mycale with a vonfused tangential ectosomal skeleton, three categories of anisochelae, anisochelae-I with a eurved shafi in profile view. ratio height of the head/total height of the spicule « 25%, alae of the foot projecting downward forming a pore, and rosettes built both by anisochelac-1 and -I] (many with unguiferate anisochelae-!, acanthose anisochelae-IT, and basally-spurred anisochelae-111). Subgenus Mycale - Mycale with a confused tangential ectosomal skeleton (many with pore-groaves, three categories of anisochelae, basally-spurred anisochelae-111, and rhaphides in two categories). Mycale (Mycale) immitis-group - Mycale with a confused tangential ectosamal skeleton, anisochelac-1 with a curved shaft in profile view, ratio height of the head/total height ol the spicule > 25% and <35%, alae nf the fool projecting downward forming a pore (many with pore-grooyes, three categories of anisochelae, basally-spurred anisochelae-[11, and rhaphides in two categories). Subgenus Naviculina - Mycale with a reticulated tangential ectosomal skeleton, and naviculichelae (many with three categories of anisochelae, and toxas). Subgenus Oxymvcale- Mycale with a reticulated tangential ectosomal skeleton and inegascleres which are oxeas exclusively. Subgenus Paresperella - Mycale with a reticulated tangential ectosomal skeleton and serrated sigmas (many with loxas). MYCALIDAE PHYLOGENY I, Mycales Gray, 1867 2. Myeale 7 Gray, 1867; Arenochalina Lendenfeld, 1887 3, Myeale Gray, 1867; Jegogropila-IL5 Gray, 1867 4. Myeale Gray, 1867; Grapelia.z Gray, 1867 5. Grapelia ., Gray, 1867; Ahomomvcale Topsent, 1924 b. Grapelia Gray, 1867; Rhaphidoteca., Kent, 1870 7. Rhaphidoreca Kent. 1870; Myeale immitis.s (Schmidt, 1870) 8. degograpila-IL, Gray, 1867; Oxymveale Hentschel, 1929 9. Aegogropila- Gray, 1867: Paresperella Dendy, 1905: Aegogropila-Ly Gray, 1867; Naviculina. | Gray, 1867 10. 4egograpila-1. | Gray. 1867: Carmia-M Gray, 1867 11. Aegograpila-] Gray, 1867; Carmia-| Gray, 1867 12. Naviculina Gray, 1867; Zygomycale Topsent, 1930 FIG. 7. Phylogenetic classification of the mycalids with anisochelae built from the cladogram in Figure 4 under the protocol of Papavero, Llorente-Bousquets & Abe (1992) and Papavero & Llorente-Bousquets (1993). Linnean categories are abolished, only genus- and species-level taxa are considered, hierarchy is reirievable from a numbered sequence attributed to the oldest taxon included (priority is applied to pages and lines also), The priority of Rhaphidoteca Kent, 1870 over Mycale immitis. (Schmidt, 1870) was decided not on the basis of knowledge of the actual dates af publication of both works, bul on the fact that the species-group represented by M. immitis. is bound to be named in the future, becoming then a much younger taxon. (Refer to the text for further explanations). Subgenus Rhaphidoteca - Mycale with a confused tangential ectosomal skeleton, exotyles. and anisochelae-T with alae of the foot projecting downward forming a pore (ratio height.of the head/total height of the spicule may be > 25% and < 35%, rhaphides may be in twa categories), Subgenus Zygomycale - Mycale with a reticulated tangential ectosomal skeleton and isochelae next to anisochelae (many with three categories of anisochelae, and toxas). Phylogenetic definitions for the above taxa based on apomorphies can be obtained by referr- ing each clade to all the species sharing that clade’s synapomorphies, and those of all its descendants, Apomorphy-based definitions have been severely criticised, however, because subsequent discovery of homoplasies can lead to substantial reshuffling of clades (e.g. Schander & Thollesson, 1995). The alternative option - using node-based definitions for the terminal taxa considered above - would be premature at this stage, The definition of more-inclusive taxa is dependent upon an unambiguous understanding of the Jess-inclusive taxa it contains, Cantino et al. (1997) chose to build their node-based definitions using only species level taxa, which were selected in such a way so that the more basal genera included in the clade would be represented, These kind of data are absent, or nearly so, for most of the terminal taxa considered here. Where this information is available, node-based deliniiions can be powerful taxonomic tools (explicit, universal and stable). Hajdu (1995) published a phylogeny for the curved-assemblage of Mycale, which permits the derivation of node-based phylogenetic definitions for the immitis-group, Rhaphidoteca and Grapelia. This scheme is as follows: Subgenus Grapelia-the least inclusive clade that contains Mycale mvriasclera Lévi & Lévi, 1983 and Mycale burtoni Hajdu, 1995, Subgenus Rhaphidoteca - the least inclusive clade that contains Mycale marshallhalli (Kent, 1870) and Mycale loricata (Topsent, 1896). Mycale (Mycale) immitis-group - the least inclusive clade that contains Mycale trichela Lévi, 1963 and Mycale paschalis Desqueyroux- Faúndez. 1990, CONCLUSIONS This discussion illustrates that current poriferan classifications may be very distant from truly phylogenetic schemes. While debate 236 persists on the merits and pitfalls of retaining the Linnean hierarchy, this does not excuse any proposal based on non-phylogenetic définitions for poriferan taxa. It is imperative that taxa are always diagnosed on the basis of their synapomorphies. This makes them more likely tu be natural, and more relevant to future phylogenetic classification schemes, This is especially important when dealing with more- inclusive taxa, from which less-inclusive groups are extracted on the basis of their clearer mono- phyletic status. If, ás is the current trend, effort is made toward defining such inclusive, plesiomorphic laxa (but excluding the extracted, less-inelusive taxa), i is likely that a paraphyletic assemblage will be recognised instead. In the phylogenetic system, groups such as these are going through w metaphorical *mass-extinction episode! right now. ACKNOWLEDGEMENTS This article has benefited greatly of the comments of John N.A. Hooper (QM), Gisele Lobo-Hajdu (UERIJ), Guilherme Muricy (MNRJ) and two anonymous reviewers (who kindly told me to restart from ZERO!) The author is thankful to M, Dzwillo {ZMH}, M. GrassholT (SMF). C. Lévi (MNHN}, K. Rützler (USNM), K.P. Smith (USNM), RWM, van Soest (ZMA), C. Valentine (BMNH), J. Vermeulen (ZMA ) and S. Zea (INV) for the loan of specimens or other comparative material. Ulisses S, Pinheiro and Mariana de S. Carvalho are thanked for curatorial help as volunteers. A. Ribeiro, M.V. Cruz and E. Mattos (USP - ZEISS); and Rob W.M. van Soest and D. Platvoet (ZMA - JEOL) are thanked for the provision of SEM lacilities and technical support in SEM operation, CNPq, FAPERJ, FAPESP and FUJB (all from Brazil) provided financial support in the form of fellowships, grants and/or travel funds, which are gratefully acknowledged. A visit to the National Museum of Natural History in 1993 was made possible by a Short Term Visitor grant of the Smithsonian Institution, Washington D.C., U.S.A., which is also gratefully acknowledged. LITERATURE CITED AMORIM, D.S. 1982. Classificagao por seqüeneiacáo: uma proposta para denominação dos ramos retardados. Revista Brasileira de Zoologia 1(1); 1-9. 1992. An empirical system of ranking biological classifications using biogeographic components, MEMOIRS OF THE QUEENSLAND MUSEUM Revista brasileira de Entomologia 26¢2): 281-292, 1994, Group*: an additional artifact for phylogenetic sequenced classifications. Revista Nordestina de Biologia 8(1): 35-38. 1997. Elementos básicos de sistemática filogenética. 1-276 (Holos: Ribeirüo Preto. Brazil). F BOURY-ESNAULT, N. & RUTZLER, K, (eds) 1997, Thesaurus of sponge morphology. Smithsonian Contributions to Zoology 596; Hiv, 1-55, BOWERBANK, J.S. 1864. .A monograph of the British Spongiadae. Vol. 1. (Ray Society: London). CANTINO, P.D., OLMSTEAD, R.G. & WAGSTAFF, S.J. 1997. A comparison of phylogenetic nomenclature with the current system: a botanical case study, Systematic Biology 46; 313-331. CARBALLO, J.L, & HAJDU, E.1998, Mieromorphology in Mveale taxonomy (Mycalidae, Poecilosclerida, Demospongiae), with the description of two new micracanthoxea-bearing species. Contributions to Zoology (The Hague) 67(3): 187-195. CARTER, H.J. 1882. Some sponges from the West Indies and Acapulco in the Liverpool Free Museum, described with genera! and classificatory remarks. Annals and Magazine of Natural History (5)9: 266-301, 346-368, pls 11-12, CHRISTOFFERSEN, M.L. 1988. Genealogy and phylogenetic classification af the world Crangonidae (Crustacea, Caridea), with a new species and new records for the southwestern sven Revista Nordestina de Biologia 6(1): -59. 1995. Cladistic taxonomy, phylogenetic systemalics, und evolutionary ranking. Systematic Biology 44; 440-454, DENDY, A. 1905. Report on the sponges collected by Prof. Herdman at Ceylon in 1902. Report to the Government of Ceylon on the Pearl Oyster Pisheries of the Gulf of Manaar 111 (supplementary reports, no. xviii): 57-246. DESQUEY ROUX-FAUNDEZ, R. 1990. Spongiaires (Demospongiae) de Ple de Páques (Isla de Pascua). Revue Suisse de Zoologie 97: 373-409. ESPER, B.C, 1794, Die Pflanzenthiere, Abbildungen nach der Natur mit Farben erleuchtet, nebst Beschreibungen. (Raspe: Nürnberg). FARRIS, J.S, 1976, Phylogenetic classification of fossils with recent species. Systematic Zoology 25: 271-282. GRAY, LE. 1867. Notes on the arrangement of sponges, with the description of some new genera. Proceedings of the Zoological Society of London (1867): 492-558, pls 27-28, GRIFFITHS, G.C.T. 1974. Some fundamental problems in biological classification. Systematic Zoology 22: 338-343. GRUBER, G. & REITNER, J, 1991. Isolated porileran micro- and megascleres from the lower MYCALIDAE PHYLOGENY Campanian of Hóver (northern Germany) and remarks on the phylogeny of the taxon Geodiidae (Demospongiae). Berliner Geowissenschaftliche Abhandlungen (A)134: 107-117. HAJDU, E. 1994. A phylogenetic interpretation of hamacanthids (Demospongiae, Porifera) with a redescription of Hamacantha popana (de Laubenfels, 1935). Journal of Zoology 232: 61-77. 1995. Macroevolutionary patterns within the demosponge order Poecilosclerida. Phylogeny of the marine cosmopolitan genus Mycale, and an integrated approach to biogeography of the seas. PhD thesis (Centralle Drukkerij: Universiteit van Amsterdam, Amsterdam). HAJDU, E. & DESQUEYROUX-FAUNDEZ, R. 1994. A synopsis of South American Mycale (Mycale) (Poecilosclerida, Demospongiae), with the description of three new species and a cladistic analysis of the Mycalidae. Revue Suisse de Zoologie 101(3): 563-600. HAJDU, E. & RUTZLER, K. 1998. Sponges, genus Mycale (Poecilosclerida) from a Caribbean mangrove and comments on subgeneric classif- ication. Proceedings of the Biological Society of Washington 111(4): 737-773. HAJDU, E. & SOEST, R.W.M. VAN 1996. Choosing among poriferan morphological characters within the cladistic paradigm. Bulletin de l'Institut Royal de Sciences Naturelles de Belgique 66(supplément): 81-88. HAJDU, E., ZEA, S., KIELMAN, M. & PEIXINHO, S. 1995. Mycale escarlatei n.sp. and Mycale unguifera n.sp (Demospongiae) from the Tropical-western Atlantic. Beaufortia 45(1): 1-16. HENNIG, W. 1966. Phylogenetic Systematics. (University of Illinois Press: Urbana, Ill.). HENTSCHEL, E. 1929. Die Kiesel- und Hornschwámme des nórdlichen Eismeeres. Fauna Arctica 5(4): 852-1042, pls 1-3. HOOPER, J.N.A. 1991. Revision of the family Raspailiidae (Porifera: Demospongiae), with description of Australian species. Invertebrate Taxonomy 5: 1179-1418. 1997, Sponge Guide. (www.Qmuseum.qld.gov.au, Queensland Museum: Brisbane). HOOPER, J.N.A. & LEVI, C. 1994. Biogeography of Indo-west Pacific sponges: Microcionidae, Raspailiidae, Axinellidae. Pp. 191-212. In Soest, R.W.M. van, Kempen, Th.M.G. van & Braekman, J.C, (eds) Sponges in Time and Space (Balkema: Rotterdam). HOOPER, J.N.A. & WIEDENMAYER, F. 1994. Porifera. Pp 1-624. In Wells, A. (ed.) Zoological Catalogue of Australia. Vol. 12 (CSIRO Australia: Melbourne). KENT, W.S. 1870. On two new siliceous sponges taken in the late dredging expedition of the yacht ‘Norma’ off the coasts of Spain and Portugal. Annals and Magazine of Natural History (4)6: 217-224, PI. XV. 237 LENDENFELD, R. VON 1887. Descriptive catalogue ofthe sponges in the Australian Museum, Sydney (Taylor and Francis: London). LEVI, C. 1963. Spongiaires d'Afrique du Sud. (1) Poecilosclérides. Transactions of the Royal Society of South Africa 37(1): 1-71. 1969. Spongiaires du Vema Seamount (Atlantique Sud). Bulletin du Muséum National d'Histoire Naturelle 41: 952-973. LÉVI, C. & LÉVI, P. 1983. Démosponges bathyales récoltées par le N/O ‘Vauban’ au sud de la Nouvelle Calédonie. Bulletin du Muséum National d'Histoire Naturelle 5(4): 931-997. LINNAEUS, C. 1758. Systema naturae per regna tria naturae. (Holmiae: Stockholm). NELSON, G. 1972. Phylogenetic relationship and classification. Systematic Zoology 21: 227-231. NIXON, K.C. & DAVIS, J.I. 1991, Polymorphic taxa, missing values and cladistic analysis. Cladistics 7: 233-241. PAPAVERO, N. & LLORENTE-BOUSQUETS, J. 1993. Propuesta de un nuevo sistema de nomenclatura para la sistemática filogenética. II-VI. Publicaciones Especiales del Museu de Zoologia de la Universidad Nacional Autonoma de Mexico 6: 1-28, 29-42, 43-60. PAPAVERO, N., LLORENTE-BOUSQUETS, J. & ABE, J.M. 1992. Un nuevo sistema de nomenclatura para la sistemática filogenética. I. Publicaciones Especiales del Museu de Zoologia de la Universidad Nacional Autonoma de Mexico 5: 1-20. PULITZER-FINALI, G. 1996. Sponges from the Bismark Sea. Bollettino dei Musei e degli Istituti Biologici dell'Università di Genova 60-61: 101-138. QUEIROZ, K. DE & GAUTHIER, J. 1990. Phylogeny as a central principle in taxonomy: phylogenetic definitions of taxon names. Systematic Zoology 39: 307-322. 1992, Phylogenetic taxonomy. Annual Review of Ecology and Systematics 23: 449-480. RÜTZLER, K. 1978. Sponges in coral reefs. Pp. 299-313. In Stoddart, D.R, & Johannes, R.E. (eds) Coral reefs: research methods. Monographs on oceanographic methodology. Vol. 5 (UNESCO: Paris). SCHANDER, C. & THOLLESSON, M. 1995. Phylogenetic taxonomy - some comments. Zoologica Scripta 24: 263-268. SOEST, R.W.M. VAN 1994, Demosponge distribution patterns. Pp. 213-223. In Soest, R.W.M. van, Braekman, J.C. & Kempen, Th.M.G, (eds) Sponges in Time and Space. (Balkema: Rotterdam). SOEST, R.W.M. VAN & HAJDU, E. 1997. Marine area relationships from twenty sponge phylogenies. A comparison of methods and coding strategies. Cladistics 13: 1-20. 238 SUNDBERG, P. & PLEIJEL, F. 1994. Phylogenetic classification and the definition of taxon names. Zoologica Scripta 23(1): 19-25. SWOFFORD, D.L. 1993. PAUP. Version 3.1.1. Computer program. (Smithsonian Institution: Washington D.C. Distributed by the Illinois Natural History Survey, Champaign, Ill.). TOPSENT, E. 1896. Campagnes du Yacht Princesse Alice. Sur deux curieuses Espérellines des Acores. Bulletin de la Societé Zoologique de France 21: 147-150. 1924. Révision des Mycale de l’Europe occidentale. Annales de l'Institut Océanographique de Monaco 1(3): 77-118. MEMOIRS OF THE QUEENSLAND MUSEUM 1925. Étude des Spongiaires du Golfe de Naples. Archives de Zoologie Expérimentale et Générale 63(5): 623-725. 1930. Eponges de Lamarck conservées au Museum de Paris. Archives du Muséum National d'Histoire Naturelle (6)5: 1-56 VACELET, J. & VASSEUR, P. 1971. Éponges des récifs coralliens de Tulear (Madagascar). Tethys, supplément 1: 51-216. WIEDENMA YER, F. 1994. Contributions to the knowledge of post-Palaeozoic neritic and archibenthal sponges (Porifera). The stratigraphic record, ecology, and global distribution of intermediate and higher taxa. Schweizerische Paláontologische Abhandlungen 116: 1-147. WILEY, E.O. 1979. An annotated Linneana hierarchy, with comments on natural taxa and competing systems. Systematic Zoology 28: 308-337. PHOTOSYNTHESIS AND RESPIRATION OF THE CYANOBACTERIUM-CONTAINING SPONGE, DYSIDEA HERBACEA. Memoirs of the Queensland Museum 44: 238. 1999:- Marine sponges containing cyanobacterial endosymbionts are common in tropical waters, and the dictyoceratid sponge, Dysidea herbacea, is one ofthe most abundant sponges in the shallow lagoon at One Tree Reef, Great Barrier Reef. This sponge contains large numbers of the filamentous cyanobacterium, Oscillatoria spongeliae. The O. spongeliae trichomes are located free in the sponge mesohyl, although they are often in contact with archaeocytes. The high biomass of the cyanobacteria is illustrated by the chlorophyll a content of the association, which i is about 335ug.mL'' sponge volume, or 180.3ug.g' sponge wet weight. These values are much higher than for any other sponges so far studied. Photosynthetic and dark respiration rates were measured using an oxygen electrode in summer and winter at ambient lagoon temperatures and at saturating irradiances. The compensation point for photosynthetic O, production is reached at about 30-50umol photons.m^.sec' and photosynthesis saturates at about 300umol photons.m".sec'. No seasonal differences in the photosynthetic and respiration rates could be detected indicating that the sponge adapts to changing environmental conditions. The D. herbacea/ O. spongeliae association, does however respond to changes in temperature, with a Qio for photosynthesis of about 5. Photosynthesis and respiration rates are also sensitive to the O; concentration in the seawater. The implications of these results for the ecology of this symbiotic association will be discussed. O Porifera, Dictyoceratida, cyanobacterium, symbiosis, photosynthesis, respiration, temperature. Rosalind Hinde & F. Pironet, School of Biological Sciences, University of Sydney, NSW, 2006, Australia; M.A. Borowitzka (email: borowitz@possum.murdoch.edu.au), School of Biological Sciences & Biotechnology, Murdoch University, W.A. 6150, Australia; 1 June 1998. MORPHOLOGICAL AND GENETIC EXAMINATION OF PHENOTYPIC VARIABILITY IN THE TROPICAL SPONGE ANTHOSIGMELLA VARIANS MALCOLM $, HILL Hill, M.S. 1999 06 30: Morphological and genetic examination of phenotypic variability in the tropical sponge Anthosigmella varians. Memoirs of the Queensland Museum 44: 239-247. Brisbane. ISSN 0079-8835, Sponges commonly exhibit phenotypic variation in response to a heterogeneous environment. Determining the ecological causes and understanding the evolutionary consequences of this variation is a primary goal of biologists. Three ecotypes of the common Caribbean sponge Anthosigmella varians (Demospongiae: Hadromerida: Spirastrellidae) have been identified, and they provide a model system for exploring morphological variation. Anthosigmella varians forma incrustans is an encrusting bioeroder located on fore- and back-reef environments at depths ranging from 3->30m; A. varians forma varians is an irregularly lobate branching ecotype found in shallow («1m) water near-shore; and 4, varians forma rigida is a branching form found in sympatry with incrustans but restricted to shallower depths. In this paper, a detailed examination of ecologically important characters (e.g. tissue strength, skeletal properties, distribution) for all 3 morphs is presented. Using allozyme electrophoresis, fixed differences at two loci were discovered indicating potential reproductive isolation between branching (rigida and varians) and encrusting (incrustans) morphotypes. Results from a transplantation experiment indicate that sediment load may be an important factor in branch production. Sedimentation may also explain the competitive aggressiveness of incrustans which is often found growing over coral species (i.e. 4. varians grows over corals to gain access to CaCO; skeletons which are typically low sediment zones). It is proposed that rigida populations can exist on the reef due to the production of a spicule- and collagen-rich cortex that provides a structural defense against predators. It is suggested that wave energy may have less important effects on branch production in A. varians than either predation or sedimentation. CJ Porifera, phenotypic plasticity, spongivory, population structure, cryptic species, Anthosigmella varians. Malcolm S. Hill (email: mhill@fair] fairfield.edu), Program in Evolutionary Biology and Ecology. Department of Biology, University of Houston, Houston 77204, USA (Present Address: Biology Department, Fairfield University, Fairfield 06430); 1 February 1999. Sponges present a unique challenge to our understanding of evolutionary processes in the sea. This ancient group of metazoans has few traits that are taxonomically useful for distinguishing closely related species, and those that are employed in assigning taxonomic positions can often be modified by environmental parameters. For example, wave energy has been demonstrated to affect gross morphological characteristics of intertidal sponges (Palumbi, 1984, 1986) and changes in spicule size and shape can often reflect the influence of extrinsic factors (e.g. Jones, 1970, 1971). Thus, there is a danger of misinterpreting the evolutionary history of this metazoan group. Determining when morphological variation in sponges represents phenotypic plasticity or true speciation is a significant challenge. Several investigators have shown that morphological variants previously included within a single sponge taxon represent distinct species (Sole-Cava & Thorpe, 1986, 1990; Sole-Cava et al., 1992; Bavestrello & Sara, 1992; Barbieri et al., 1995). These studies indicate that we know little about sponge diversity, and ultimately, evolutionary processes in the ocean. These studies also highlight how poorly we understand the importance of phenotypic variation in the ecology and evolution of sponges. When faced with morphological varieties of a species, two questions must be addressed. First, are the phenotypic variants reproductively isolated? This can be inferred by examining the genetic constitution of populations. Second, if the variants are part of an interbreeding population, are environmental factors responsible for the observed variation, and 1f so, which are most important? These questions are of particular concern on tropical coral reefs since ongoing debate concerns the relative importance of equilibrial and nonequilibrial processes in the 240 maintenance and origin of diversity (Sale, 1994; Knowlton & Jackson, 1994a, b). If environmental parameters are highly predictable, then niche partitioning via habitat specialisation may be a common product of evolution on coral reefs (Knowlton & Jackson, 1994a; Palumbi, 1994). If, however, larvae settle randomly into diverse habitats, or disturbance frequencies differ among habitats, then phenotypic plasticity may evolve (e.g. Lively, 1986). Sponges present a highly tractable system for examining ecological conditions necessary for the evolution of phenotypic plasticity. They also provide a model for identifying factors important in the evolution of reproductive isolation (i.e. speciation) between morphologically divergent populations. Here, | present an analysis of morphological and genetic differences among morphotypes of the common Caribbean sponge Anthosigmella varians (Demospongiae: Hadromerida: Spirastrellidae). In addition, I discuss the roles that phenotypic plasticity and habitat specialisation may have played in the evolution of this species. NATURAL HISTORY. Anthosigmella varians (Duchassaing & Michelotti) is a common sponge of Caribbean coral reefs (Wiedenmayer, 1977; Vicente, 1978). This sponge harbors intracellular dinoflagellate zooxanthellae, bores into calcium carbonate structures (Hill, 1996a) and exhibits two distinct morphologies: branching and encrusting (Wiedenmayer, 1977). Itis included in the diets of angel fish (Randall & Hartman, 1968; Hourigan et al., 1989) and at least some parrotfish species (pers. obs.; Dunlap & Pawlik, 1996; Wulff, 1997). Taxonomists recognise two morphotypes: an encrusting growth form (forma incrustans) and an amorphous, irregularly lobate, branching growth form (forma varians), and consider these to be ecophenotypes based on their occurrence in different habitats (Wiedenmayer, 1977). Anthosigmella varians forma incrustans (hereafter referred to as incrustans) is con- spicuous on fore- and back-reefs while A. varians forma varians (hereafter referred to as varians) is typically found in shallow, lagoonal areas. In the Florida Keys, USA, varians is found close to shore on both bay and ocean sides of many islands; incrustans is only found on the reefs which run parallel to the islands approximately 8km from shore (Fig. 1). Wave energy has been proposed as the factor responsible for this distribution: varians is presumed to be unable to handle the strong currents and periodic MEMOIRS OF THE QUEENSLAND MUSEUM heavy wave action that open reefs receive (Vicente, 1978). During this study, a number of branching morphs sympatric with encrusting morphs were encountered on both Tennessee and Alligator fore-reefs in the Florida Keys (as well as Molasses Reef; J. Pawlik, pers. comm.) (Fig. 1, Table 1). This morph has lobate branches like varians but is easily distinguishable due to its much stiffer skeletal construction. It was often found less than a meter from incrustans individuals. I refer to this morph as 4. varians forma rigida (hereafter as rigida). MATERIALS AND METHODS There were three components to this study: 1) comparison of morphological and ecological characteristics of the three morphs of A. varians; 2) transplant experiments to assess the effect of sedimentation on morphological variation in varians and rigida; and 3) allozyme analysis, used to estimate genetic relatedness among 4. varians morphs. MORPHOLOGICAL CHARACTERISTICS. External. The following external features were measured in situ; surface area of attachment, number of branches, branch length, maximum height, mound height and tissue strength. Sample sizes for all parameters are listed in Table 1. Surface area of attachment was estimated using a lm* quadrat marked off in 25cm” increments. Branch length was measured as the distance from tip to node. Maximum height was measured from the highest point of the sponge perpendicular to the substratum. Mound height was measured from the substratum to the highest point on the mounding portion of branching individuals. The method for determining tissue strength was adapted from Palumbi (1984). A barbless hook, attached to a spring scale, was embedded to a depth of 0.5cm into the surface tissue of an individual. Care was taken to ensure that there was no foreign material in the sponge tissue. The spring scale was pulled perpendicularly until the hook tore free, and the maximum force required to extract the hook was recorded. Three pulls were averaged for each sponge, sample sizes for each morph are shown in Table 1. One-way ANOVA was used to compare morphotypes (Zar, 1984). Internal. Several internal characters were also compared among morphs. Cortex thickness was measured for seven individuals from each morph, 10 measurements were averaged for each individual. The cortex of a sponge is defined as VARIABILITY IN ANTHOSIGMELLA VARIANS 241 TABLE 1. Distribution and morphological characteristics of three morphotypes of Anthosigmella varians in the Florida Keys. Values represent means (+SE); sample not significantly different at P 70.05. sizes are shown. Values sharing a superscripted letter are Parameter incrustans varians rigida Distribution Reef n Bay/near island n Reef n Depth (m) 8-27 1-3 8-13 Number of branches 0a 48 2.95 (0.3) 55 1.92 (0.62) 13 Branch length (cm) n/a - 10.37 (0.51)* 161 4.36 (0.79) 25 Mound Height (cm) n/a - 3.47 (0.28) 38 3.38 (0.64)* 13 Maximum Height (cm) n/a - 12.95 (1.22 55 6.67 (1.8) 13 Area of attachment (cm?) d 48 68.1 (14.4) 32 36.3 (4.79 13 Tissue strength (N) 4.9 (0.36)? 28 2.3 (0.14)? 61 >10° 8 Cortex thickness (mm) 0.99 (0.14)* 7 n/a - 5.7 (1.33)? 7 Open Space in choanoderm 10.1% (1.8) 7 32.7% (2.1) 7 21.1% (1.9 7 Spicule conc. (mg cm?) 52(2y 7 112 (5) 8 168 (11) 5 Zooxanthella cm? (H 10°) 0.86 (0.053) 8 1.37 (0.092)? 13 1.03 (0.82)? 5 3 Wet weight (g em”) 1018 (104)? 6 833 (39) 10 759 (105)? 4 Sponge biomass (g cm?) 80.8 (8.2)* 6 100.8 (3.8)* 10 97.2(7.6)° 4 Spicule (mg cm?) 70.5 (13.9 6 132.6 (15.0)? 10 199.7 (17.4) 4 Calcareous debris (g em”) 553.9 (66.8) 6 -2 (7.3)? 10 -3 01.1)? 4 the layer of ectosome consolidated by a distinc- tive skeleton (Kelly-Borges & Pomponi, 1992). Measurements were made from the surface along the growth axis to the point where the cortex met the choanosome. Zooxanthellae densities were also compared among morphs. Densities were quantified using methods described in Hill (1996a). Spicule concentrations were estimated as follows: samples of known volume were dissolved in nitric acid until all tissue was removed; the resulting solution was washed several times with ddH^O and then dried at 60°C for 48hrs. Results were reported as mg of spicule cm? of sponge tissue. To determine if there were potential differences in pumping capabilities among morphs, I estimated the percentage of open space in the choanosome of each morph. Five estimates were taken at different locations within the choanosome of a single individual, and these were averaged to give one value per sponge. Seven individuals were measured from each morph. Percentages were arcsin transformed before they were compared using one-way ANOVA (Zar, 1984). Differences were compared among morphs in the relative content of spicules, biomass and calcareous debris. Wet weights and volumes were recorded for individuals from each morph. Sample sizes are shown in Table 1. Samples were placed in crucibles and weighed after drying for 48hrs at 60°C. Crucibles were then placed in a furnace for 6hrs at 450°C. This procedure removed sponge tissue but left behind spicules, calcareous debris and ash. Ash was washed from crucibles with ddH5;O, and crucibles were then dried for 48hrs at 60°C. After weighing, crucibles were treated with 5% HCl (which removed all CaCO;), washed with ddH;O, dried for 48hrs at 60°C and weighed. At this point, all that remained in the crucibles was spicular material. The distribution and concentration of collagen was compared among morphotypes using the Mallory-Heidenhain’s connective tissue stain (Humason, 1979). Immediately after collection, samples were fixed in Bouin’s solution. Fixed samples were run through a series of dehydration steps prior to embedding. Samples were placed in paraplast embedding media, allowed to harden, and then sectioned. After staining, sections were placed on slides for viewing. Spicule measurements were conducted following methods described in Palumbi (1986) with the following modifications. Spicules from 10 varians individuals, 6 incrustans individuals and 4 rigida individuals were cleaned of tissue using concentrated nitric acid (Kelly-Borges & Pomponi, 1992). Samples were then washed with ddH,0 and dried. Spicules were placed on glass slides and a cover slip was mounted over the preparation. Measurements of spicules were 242 y TENNESSEE REEF FIG. 1. Map of Florida’s Middle Keys; all collections for allozyme analysis were made in areas with stars as well as on Tennessee and Alligator reefs. Reef habitats (i.e. reefs between, and including, Tennessee and Alligator reefs) harbored jnerustans and rigida populations, while near-shore areas harbored varians populations. made from digitised images. Total length and head and shaft widths of subtylostyles were measured for 50 spicules from each individual from each morph. A qualitative assessment of anthosigma shape was also performed at this time. Comparisons of spicule size and shape were made using one-way ANOVA (Zar, 1984). TRANSPLANT EXPERIMENTS. To examine the effect of sedimentation on morphological variation, several varians and rigida individuals were transplanted onto platforms that were raised above heavily sedimented substrata. Transplant- ed varians individuals (n=8) were placed on the upper surface of cinder blocks at a depth of «1m. These sponges were monitored for 6 months. Morphological changes were monitored in 10 rigida individuals that were affixed to ceramic tiles (15cmx=15cm) by fishing line. The ceramic MEMOIRS OF THE QUEENSLAND MUSEUM A ALLIGATOR REEF tiles were placed above the substrata and were attached to wooden planks that had been cemeted to the reef. Both of these transplantations provided low sediment zones that represented uncolonised (i.e. competitor-free) space. ALLOZYME ANALYSIS. Samples of each morphotype were collected in July, 1996. Sympatric individuals of incrustans and rigida were collected on both Alligator and Tennessee reefs at a depth of 8m (Fig. 1). Anthosigmella varians forma varians individuals were collected from the bayside of Long Key and Lignumvitae islands at a depth of Im (Fig. 1). Small sections ( varians > incrustans, These values matched spicule concentrabons measured previously (Table 11. Collagenous fibers stained bright blue using the Mallory-Heidenhain’s connective tissue stain. Collagen was densely concentrated within the cortex of both incrustans and rigida, but was more extensive in rigida since the cortex was thicker in this morph (Table 1). Collagen content in varians was negligible and was widely scattered throughout the choanosome. There were no differences in the lengths of subtylostyles in any morph (Fig. 2), but varians had significantly wider megascleres (i.e. head and shaft; Fig. 2). There were no significant diflerences in widths between inerustans and rigida. The subtylostyles of inertistans were more curvaceous than egida or varians, Sigmata, lermed anthosigmas in this group, typically had a single bend (i.e. bow shaped) in incrustans and rigida butollen had two or more bends Ge. sigma shaped) in varians. Samples taken from the cortex and choanosome of rigida appeared to have no differences in lengths or widths. TRANSPLANT EXPERIMENTS, Both varers and rigida began to encrust the surfaces on to which they were transplanted. The manner of encrustalion was the same. A thin sheel proceeded to take over the unoccupied area spreading from the base of the branch that had been atlached to the substratum. This growth was relatively rapid and may have represented tissue reorganisation rather than true increases in Beara (for example. see Jackson & Palumbi, 1979), ALLOZYME ANALYSIS. Heterozygotes at loci Glucose-6-Phosphate Isomerase (EC 5.5.1.9; Gpi), Malate Dehydrogenase (EC 1.1,1,37; Mdh-1) and 6-Phosphogluconate Dehydrogenase (EC 1.1,1.44; 6Pedh) all showed 3-banded zymograms typical of a dimeric enzyme. Heterozygous individuals exhibited a multi- banded zymogram for Fumarate Hydratase (EC 4.2.1.2; Fum) as would be expected tor a tetrameric enzyme; individuals assumed to be heterozygous at Malic Enzyme (EC 1.1.1.40; Me) (3 suspected tetrameric enzyme) exhibited a smear. No heterozygote was detected at Arginine Kinase (EC 2.7.3.3; Ark) or Creatine Kinase (EC 244 TABLE 2, Allele frequencies for the 8 enzyme loci scored in this study categorised by morphotype of sponge. Incrustans = encrusting morph, varians = bay branching morph and rigida = ocean branching morph. N = number of individuals used for each enzyme system. Locus EC# Allele | incrustans | varians rigida Ark 2-533 1 0.11 0 0 2 0 1.0 1.0 3 0,22 0 0 4 0.67 0 0 N 9 8 9 Ck 2.7.3.2 I 0.22 L0 4-19 2 0.78 0 0 N 9 8 9 Fum 4.2.1.2 l 0.78 1.0 LO 2 0,22 0 0 N 9 8 9 Gpi 5.2.1.9 1 0 0.03 0 2 0.14 0.40 0.09 3 0.86 0.57 0,91 N 14 15 11 Me 1.1.1.40 l 0.67 0 0 2 0.33 0 0 3 0 1.0 1.0 | N 9 7 6 Mdh! 1.1.1.37 l 0.28 1.0 1.0 P: 0.72 0 0 N 9 9 Mdh2 | 1.1.1.37 l 0 0.75 0.94 2 0.11 0 0 3 0.22 0.25 0.06 | 4 0.67 0 0 N 9 8 9 6Pgdh | 1.1.1.44 1 0.33 0.5 0.25 | 2 0 0.08 0 3 0.17 0 0.67 4 | 0.28 0.25 0.08 5 0.22 0.17 0 N 9 8 9 2.7.3.2; Ck) loci. Neither varians nor incrustans populations had heterozygotes at the Mdh-2 locus. Encrusting populations appear to be genet- ically distinct from both branching morphs. There were fixed differences at Ark and Me loci (Table 2). In addition, there were large differ- ences between encrusting and branching morphs in frequencies for the following loci: Ck, Fum, Mdhl and Mdh2. However, allelic frequencies for Gpi were more similar in oceanside populations. MEMOIRS OF THE QUEENSLAND MUSEUM DISCUSSION Anthosigmella varians represents a morph- ologically diverse species with substantial phenotypic differences among the three morphotypes (Table 1). The distribution of the three morphs (Table 1) indicates that varians and rigida individuals prefer shallower and more sediment-laden habitats than incrustans. Despite significant differences among morphs, the presence of branches is the most useful diagnostic trait available in the field: rigida and varians have branches while incrustans assumes only an encrusting form. Preliminary grafting experiments demonstrat- ed that all three morphs were capable of attaching to one another, but connections between rigida and varians were strongest (unpublished results). Although sample sizes are small, genetic analysis supports this distinction since rigida and varians populations were fixed for alleles at two loci that were not present in the incrustans population. Any conclusions about reproductive isolation must be tentative, however, given the absence of heterozygotes at the loci Ark, Ck and Mdh-2. For this reason, the potential reproductive isolation in A. varians is being further explored using larger sample sizes and DNA-based molecular tech- niques which provide access to a larger number of polymorphic loci. Transplant experiments indicated that sediment may be responsible for the branching phenotype since varians and rigida individuals that were placed on sediment-free substrate (in both the Florida Bay and reef) began to encrust. Sediment- ation rates are highest in the shallow lagoonal areas where varians is found while on the reef sedimentation rates appear to decrease with depth (J. Schmerfeld, pers. com.). Although correlative at this stage, this information supports the claim that branching is a response to sediment load. There is a strong positive correlation between tissue strength and cortex thickness (Table 1); varians lacks any sort of cortex, incrustans has a well defined but relatively thin cortex, and rigida has a thick cortex. The cortex was shown to be collagen rich (with Mallory-Heidenhain stain), and it is probable that the collagen accounted for the dramatic increases in tissue strength in rigida. The observed differences among the three morphs may be due to anumber of environmental parameters. Wave energy has been hypothesised to prevent incrustans from producing branches (Vicente, 1978). Another possibility that has not been considered to date is that predation causes VARIABILITY IN ANTHOSIGMELLA VARIANS A Spicule dimensions Anthosigmella varians Spicule length (mm) o 5 o ee merusians rigida vanans B T B head width shaft width 0.019 0.005 4 Spicule widths (mm) 0.000 Morphotype FIG. 2, Measurements of total length and widths of head and shaft in subtylostyles of A. varians morphs. Lengths and widths of 50 randomly chosen spicules were averaged for each individual examined. The number of individuals used were: incrustans = 6, rigida = 4 and varians = 10. Histobars represent means (+SE); bars connected by an underline are not significantly different at P >0.05 using one-way ANOVA and Tukeys’ multiple comparisons test. increases in collagen synthesis in the cortex. Several lines of evidence suggest that predators, and not wave energy, influence tissue strength and branch production in A. varians. For instance, numerous rigida individuals were found on the high wave energy reef crest. Qualitative support for this hypothesis came when rigida and incrustans individuals were cut open to expose the choanosome. Angelfish immediately consumed large quantities of the interior portions of these individuals while completely avoiding the cortex. Caging and transplant experiments indicate that predation is probably more important than wave energy (Hill, 1996b), but the influence of wave energy on collagen production cannot be ruled out. 245 Encrusting and branching morphs play very different roles in coral reef communities. Over 40% of a surveyed incrustans population (n=48) was involved in competitive encounters while none of the branching morphs were ever ob- served growing over corals (Hill, 1996b). Furthermore, by occupying large areas of reef, incrustans indirectly affects invertebrate recruitment by usurping space that could be used for successful settlement (Table 1; Vicente, 1978). Given that incrustans appears to devote less tissue volume to pumping activities (i.e. fewer choanosomal open spaces; Table 1), it seems that rigida and varians individuals should have a larger impact on bacterioplankton communities than incrustans. Finally, incrustans penetrates deeper into carbonate structures than either varians or rigida. Neither varians nor rigida devoted as high a percentage of their biomass to boring activities (Table 1), and sponge-produced sediment is often seen on the surfaces of incrustans individuals indicating active boring. These observations indicate that incrustans plays a larger role in bioerosional processes than either rigida or varians. If the branching and encrusting morphs of A. varians are truly reproductively isolated populations, then the speciation process must be explained (see discussion of Sara, 1990). Two distinct, but not exclusive, schools of thought have been adopted by biologists to explain how the great diversity of tropical coral reefs has originated and is maintained. Many attribute observed patterns of diversity to non-equilibrial processes such as disturbance and chance (Sale, 1977, 1988; Connell, 1978, 1979; Hughes, 1989; Doherty & Fowler, 1994; Aronson & Precht, 1995). Hypotheses involving equilibrial pro- cesses, such as niche diversification, have recently received increased attention (e.g. Jackson, 1991; Knowlton & Jackson, 1994a). However, there is no clear non-equilibrial mechanism proposed to explain the origin of diversity on coral reefs (Sale, 1988). Although they are assumed to play a major role (Knowlton, 1993; Knowlton & Jackson, 1994a), it is unknown how important habitat specialisation and niche diversification have been in the origin of diversity. The results presented here suggest that niche diversification may have played a role in the separation of incrustans from rigida and varians. That is, varians and rigida populations are capable of utilising sediment laden environments that incrustans is unable to utilise. 246 Recent research has demonstrated that Por- iferan diversity may be greater than indicated by current classifications. For example, Boury-Esnault et al. (1992) measured genetic divergence among sympatric morphotypes of the Mediterranean sponge Oscarella lobularis using protein electro- phoresis. They found fixed differences at seven loci among sympatric morphotypes indicating a significant reduction (or cessation) in gene flow among populations. The results presented here, in addition to several other studies asking similar questions (e.g. Boury-Esnault et al., 1992; Stobart & Benzie, 1994; Barbieri et al., 1995), suggest that niche partitioning may be a very important diversifying process for tropical sponges. It is clear that examination of Poriferan divers- ity, especially morphologically diverse species, is essential if we are to understand evolutionary trends in the simplest metazoans. The small number of informative taxonomic traits in this Phylum has hindered interpretation of patterns and processes. Greater attention must be focused on elucidating microevolutionary processes op- erating in marine environments, and identifying barriers to gene flow should be a priority. Sponges such as A. varians provide a model system to address these questions. LITERATURE CITED ARONSON, R.B. & PRECHT, W.F. 1995. Landscape patterns of reef coral diversity: a test of the intermediate disturbance hypothesis. Journal of Experimental Marine Biology and Ecology 192: 1-14. BARBIERI, M., BAVESTRELLO, G. & SARA, M. 1995. Morphological and ecological differences in two electrophoretically detected species of Cliona (Porifera, Demospongiae). Biological Journal of the Linnean Society 54: 193-200. BAVESTRELLO, G. & SARÀ, M. 1992. Morphological and genetic differences in ecologically distinct populations of Petrosia (Porifera, Demo- spongiae). Biological Journal of the Linnean Society 47: 49-60. BOURY-ESNAULT, N., SOLE-CAVA, A.M. & THORPE, J.P. 1992. Genetic and cytological divergence between colour morphs of the Mediterranean sponge Oscarella lobularis Schmidt (Porifera, Demospongiae, Oscarellidae). Journal of Natural History 26: 271-284. CONNELL, J.H. 1978. Diversity in tropical rain forests and coral reefs. Science 199: 1302-1310. 1979, Tropical rain forests and coral reefs as open non-equilibrium systems. Pp.141-163. In Anderson, R.M., Turner, B.D. & Taylor, L.R. MEMOIRS OF THE QUEENSLAND MUSEUM (eds) Population dynamics. (Blackwell Scientific: Oxford). DOHERTY, P. & FOWLER, T. 1994. An empirical test of recruitment limitation in a coral reef fish. Science 263: 935-939. DUNLAP, M. & PAWLIK, J.R. 1996. Video-monitored predation by Caribbean reef fishes on an array of reef and mangrove sponges. Marine Biology 126: 117-123. HEBERT, P.D.N. & BEATON, M.J. 1993. Method- ologies for allozyme analysis using cellulose acetate electrophoresis. (Helena Laboratories: Beaumont, Texas). HILL, M.S. 1996a. Symbiotic zooxanthellae enhance boring and growth rates of the tropical sponge Anthosigmella varians forma varians. Marine Biology 125: 649-654. 1996b. Community consequences of sponge competitive abilities on tropical coral reefs: an examination of Anthosigmella varians. PhD thesis. (University of Houston: Texas). HOURIGAN, T.F., STANTON, F.G., MOTTA, P.J., KELLEY, C.D. & CARLSON, B. 1989. The feeding ecology of three species of Caribbean angelfishes (family Pomacanthidae). Environmental Biology of Fishes 24: 105-116. HUGHES, T.P. 1989. Community structure and diversity of coral reefs: the role of history. Ecology 70: 275-279. HUMASON, G.L. 1979. Animal Tissue Techniques. 4th edition (W.H. Freeman & Co.: San Francisco). JACKSON, J.B.C. 1991, Adaptation and diversity of reef corals. Bioscience 41: 475-482. JACKSON, J.B.C. & PALUMBI, S.R. 1979. Regener- ation and partial predation in cryptic coral reef environments: Preliminary experiments on sponges and ectoprocts. Pp. 303-308. In Lévi, C. & Boury-Esnault, N. (eds) Biologie des Spongiaires. Sponge Biology. (Editions du CNRS: Paris). JONES, W.C. 1970. The composition, development, form and orientation of calcareous sponge spic- ules. Pp. 91-123. In Fry, W.G. (ed.) The Biology of Porifera. Zoological Society of London Symposia Number 25 (Academic Press: London). 1971. Spicule formation and corrosion in recently metamorphosed Sycon ciliatum (O. Fabricius). Pp. 301-320. In Crisp, D.J. (ed.) Fourth European Marine Biology Symposium. (Cambridge University Press: Cambridge). KELLY-BORGES, M. & POMPONI, S. 1992, The simple fool’s guide to sponge taxonomy. (Harbor Branch Oceanographic Institution: Fort Pierce). KNOWLTON, N. 1993. Sibling species in the sea. Annual Review of Ecology and Systematics 24: 189-216. KNOWLTON, N. & JACKSON, J.B.C. 1994a. New taxonomy and niche partitioning on coral reefs: jack of all trades or master of some? Trends in Ecology and Evolution 9: 7-9. VARIABILITY IN ANTHOSIGMELLA VARIANS 1994b. Reply to Sale. Trends in Ecology and Evolution 9: 398. LIVELY, C.M. 1986, Predator-induced shell dimorph- ism in the acorn barnacle Chthamalus anisopoma. Evolution 40: 232-242. PALUMBI, S.R. 1984. Tactics of acclimation: morph- ological changes of sponges in an unpredictable environment, Science 225:1478-1480. 1986. How body plans limit acclimation: responses of a demosponge to wave force. Ecology 67: 208-214. 1994, Genetic divergence, reproductive isolation, and marine speciation. Annual Review of Ecology and Systematics 25: 547-572. RANDALL, J.E, & HARTMAN, W.D. 1968. Sponge- feeding fishes of the West Indies. Marine Biology 1: 216-225. RICHARDSON, B.J., BAVERSTOCK, P.R. & ADAMS, M. 1986. Allozyme electrophoresis: A handbook for animal systematics and population studies. (Academic Press: San Diego). SALE, P.F. 1977. Maintenance of high diversity in coral reef fish communities. American Naturalist 111: 337-359. 1988. What coral reefs can teach us about ecology. Proceedings of the 6th International Coral Reef Symposium 1: 19-27. 1994. Taxonomy and coral reef ecology. Trends in Ecology and Evolution 9: 398. SARA, M. 1990, Divergence between the sympatric species Tethya aurantium and Tethya citrina and 247 speciation in sponges. Pp, 338-343, In Rútzler, K. (ed.) New Perspectives in Sponge Biology. (Smithsonian Institution Press: Washington DC). SOLE-CAVA, A.M., BOURY-ESNAULT, N., VAC- ELET, J. & THORPE, J.P. 1992. Biochemical genetic divergence and systematics in sponges of the genera Corticum and Oscarella (Demospongiae: Homoscleromorpha) in the Mediterranean Sea. Marine Biology 113: 299-304. SOLE-CAVA, A.M. & THORPE, J.P. 1986. Genetic differentiation between morphotypes of the marine sponge Suberites ficus (Demospongiae: Hadromerida). Marine Biology 93: 247-253. 1990. High levels of genetic variation in marine sponges, Pp. 332-337. In Rútzler, K. (ed.) New perspectives in sponge biology. (Smithsonian Institution Press: Washington DC). STOBART, B. & BENZIE, J.A.H. 1994. Allozyme electrophoresis demonstrates that the scleractinian coral Montipora digitata is two species. Marine Biology 118: 183-190. VICENTE, V.P. 1978. An ecological evaluation of the West Indian demosponge Anthosigmella varians (Hadromerida: Spirastrellidae). Bulletin of Marine Science 28: 771-777. WIEDENMAYER, F. 1977. Shallow-water sponges of the western Bahamas. (Birkhauser Verlag: Basel). WULFF, J.L. 1997. Parrotfish predation on cryptic sponges of Caribbean coral reefs. Marine Biology 129; 41-52, ZAR, J.H. 1984. Biostatistical analysis. (Prentice Hall: Englewood Cliffs, New Jersey), 248 MEMOIRS OF THE QUEENSLAND MUSEUM REGULATORY MECHANISMS OF IMMUNE CELLS IN SPONGES. Memoirs of the Queensland Museum 44: 248, 1999:- Gray cells, large granular wandering cells present throughout the tissues of many species of sponges, have been identified as immunocytes in two species of sponges, Microciona prolifera and Callyspongia diffusa. When the tissues of two different sponge individuals are apposed, the gray cells accumulate at the boundary of contact at the time of tissue rejection. I have suggested that these cells may be viewed as the most primordial examples of evolutionary predecessors of the well-known vertebrate lymphocytes. This comparison implies that gray cells share features of vertebrate lymphocytes and I have examined this idea with studies on two prominent aspects of activation of T and B cells. The primary signalling event upon activation of a lymphocyte by recognition of an appropriate immune target is the synthesis and release of cytokines that alert and coordinate the activity of other lymphocytes in the surrounding tissue and throughout the body, In addition the activation of lymphocytes involves internal second messenger pathways converging on the transcription factor, NFkB, that are inhibited by Cyclosporin A, a drug often used medically to prevent rejection in human transplants. Using Boyden Chamber assays, the assays originally used to identify vertebrate immune system cytokines, | have succeeded in establishing in M. prolifera that contact with foreign tissue stimulates the release of cytokines activating the migration of gray cells toward the contacting tissue. Similarly, doses of Cyclosporin A commonly used to inhibit the activation of vertebrate T cells, suppresses histoincompatibility in M. prolifera and allows the healing together of tissue from two individual sponges that would normally undergo tissue rejection. These results provide further evidence that the foundations of the cellular immune system of animals were already established in the sponges and that study of gray cells will provide insight into the course of evolution of animal immunity. O Porifera, immunology, immunocytes, gray cells. Tom Humphreys (email: htom@hawaii,edu), Kewalo Marine Laboratory, University of Hawaii, 41 Ahui Street, Honolulu, HI 96813. USA; 1 June 1998. NEGOMBATA MAGNIFICA — ^ MAGNIFICENT (CHEMICAL) PET. Memoirs of the Queensland Museum 44: 248. 1999:- Negombata magnifica (Latrunculia) is a Red Sea sponge known to produce the toxin latrunculin (Lat). Since synthesis of this compound is economically non-viable, we evaluated various ways of producing it, while determining its natural mechanism of production and ecological relevance. We examined the possibility of: 1) identifying the cells which produce and harbour latrunculin; 2) establishing cell cultures; 3) forming an underwater sponge “garden”; and 4) takingadvantage of the sponge's own reproduction and larval settlement. Early in the study it became evident that N. magnifica might actually comprise two closely related species of Negombata, one of them an undescribed, new species. The work reported here refers to the original N. magnifica. 1) The location of Lat B, was studied using specific rabbit anti-Lat B antibodies. Rabbits were immunised with a conjugate of Lat B with Keyhole Limpet Hemocyanin (KLH), and the antibodies were affinity purified over a Lat B-Sepharose column.Thick and thin sections of the sponge were analysed by immuno-histochemical and immuno-gold techniques using light and transmission electron microscopy, respectively. Latrunculin B was prominently labelled in the sponge ectosome -endosome border, especially in the dense cell layer beneath the cortex. Immuno-gold localisation within the sponge revealed that Lat B resides in the sponge cells and not in its prokaryotic symbionts. The labelling density of gold particles in the archeocytes and choanocytes was significantly higher than that ofthe other sponge cell types (special cells and skeleton associated cells). The antibodies labelled primarily archeocytes and choanocytes, membrane-limited inclusions which are perhaps Lat B secretory and/or storage vesicles. The concentration of Lat B in the sponge periphery correlates with the defensive role of the toxin, since encounters with epibionts, predators and competitive neighbours take place through the surface layer. It may, therefore, be useful to isolate these cells for culture. 2) Primary cell cultures were established from adults and embryos. Mechanical dissociation of inner parts (without the external layer) proved to be superior (less contamination and more cell types) to other techniques. Primary cultures from embryos lasted significantly longer (up to 280 days) and cells survived a freezing phase. Cell lines, however, have not yet been established. 3) Initial steps were taken toward establishing an in situ ‘garden’ of N. magnifica from sponge fragments. Although growth rate of sponge fragments was superior to that of natural sponges in their vicinity, fragment survival over a year proved to depend on sponge handling, water depth and environmental conditions (currents, sedimentation etc.). 4) Negombata magnifica had a peak in sexual reproduction during the summer. Sexually produced, naturally released, larvae were settled on plates and their growth and development were followed for up to 4 months. CJ Porifera, latrunculin, natural product, localization, antibodies, reproduction, immuno- histochemical and immuno-gold techniques, cell culture, Negombata magnifica. Micha Ilan (email: milan@post.tau.ac.il), Department of Zoology, Tel Aviv University, Tel Aviv 69978, Israel; 1 June 1998. SPONGES OF THE LOW ISLES, GREAT BARRIER REEF: AN IMPORTANT SCIENTIFIC SITE, OR A CASE OF MISTAKEN IDENTITY ? JOHN N.A. HOOPER, SUSAN E. LIST-ARMITAGE, JOHN A. KENNEDY, STEPHEN D. COOK & CLARE A. VALENTINE Hooper, J.N.A., List-Armitage, S.E., Kennedy, J.A., Cook, S.D. & Valentine, C.A. 1999 06 30: Sponges of the Low Isles, Great Barrier Reef: an important scientific site, or a case of mistaken identity ? Memoirs of the Queensland Museum 44: 249-262. Brisbane. ISSN 0079-8835. Much of our early, reliable scientific knowledge on marine taxonomy, biological and other processes of coral reefs in general, and the Great Barrier Reef (GBR) in particular, comes from the 1928-29 GBR Expedition based on the Low Isles. 106 species of sponges were collected from northern reefs of the GBR Expedition and described by Burton in 1934, 36 from the Low Isles. Burton concluded that the sponge fauna contained: 'species characteristic of the Indo-Pacific’ (38% of his species); many ‘common also to the coasts of Australia’ (17%) ‘with a mixing of the Australian and Malayan sponge-faunas’; substantial cosmopolitanism (12%) with species ‘also found in the West Indies, Azores and Mediterranean’; and only few indigenous species (14% unique to the Low Isles, 19% exclusive to N Australia). Re-examination of BMNH voucher and type material found 42% of these species were misidentified, mainly concerning the so-called ‘widely distributed’ taxa. Recent collections from the Low Isles by the Queensland Museum (QM) discovered 109 species, and together with the revised Burton collection indicate a sponge fauna of 134 species (in 63 genera and 35 families). Surprisingly only 12 species (9% of the Low Isles fauna) were common to both the Burton and QM collections. Taxonomic comparisons with other provinces show several major trends for Low Isles sponges: 1) The fauna contains a generalist element comprising ‘typical GBR species’, found on virtually all reefs surveyed so far (23% of Low Isles species). 2) The fauna also contains an indigenous component of species unique to the northern GBR (48% of Low Isles species), with 32% of these not yet recorded from anywhere else, and another 16% known only from both the Low Isles and Lizard Island (200km to the north). 3) Affinities with coastal faunas are low, contrary to Burton’s hypothesis, with only 13% of Low Isles species also found on adjacent coastal regions. 4) Affinities with oceanic coral reef species are also low, with only 10% of Low Isles species found on the Coral Sea seamounts. 5) The concept of an ‘east Australian coast’ sponge fauna is not supported, contrary to both earlier collections described by Lendenfeld in 1888 and 1889, and Burton, with only 10% of Low Ises species extending southwards into more temperate Queensland waters, and only 2% extending further into southern New South Wales. 6) The concept of ‘cosmopolitan’ species is unsubstantiated. O Porifera, Low Isles, Great Barrier Reef, faunal survey, biodiversity, biogeography, taxonomy. John N.A. Hooper (email: JohnH@qm.qld.gov.au), Susan E. List-Armitage, John A. Kennedy & Stephen D. Cook, Marine Biology Laboratory, Queensland Museum, P.O. Box 3300, South Brisbane 4101, Australia; Clare A. Valentine, Department of Zoology, The Natural History Museum, Cromwell Road, South Kensington, SW7 5BD, UK; 22 December 1998. The Low Isles, Cairns Section, Great Barrier Reef (GBR), is an historically important site for coral reef research in Australasia, being the base for the 1928-29 Great Barrier Reef Expedition. These islands (16°23’S, 145°34’E) lie about 15km off the coast of far northern Queensland, 70km N of Cairns, approximately midway between the mainland and outer barrier reef (Fig. 1A), and easily accessible from both Port Douglas and Cairns. They consist of two small coral islets (Fig. 1B), one with a sand cay and the other a coral ‘shingle’ islet with marigroves, both with extensive fringing reef and connected by an expansive coralline reef flat. The geomorphology and many other aspects of these reefs have been described in detail in the Scientific Reports of the Great Barrier Reef Expedition 1928-29. Since at least the 1880s these small islands have been frequented by recreational and com- mercial fishermen, tourists and government authorities (e.g. meteorological bureau, coast- watch, and scientists). The islands owe their 250 MEMOIRS OF THE QUEENSLAND MUSEUM A "E E L^ ji y Torres Strait] ¿Shelburne Bay * Gulf of Carpentaria $ -g } Port Douglas X Cairns v / v, Wreck Reef Townsville E. te, Saumarez Reef ate Beef i Ax Swain Reefs . Flea \ def Ib E: doe / ^ 5 E BEALON y L OW l S L E S ea My å BI Loa de a uM". 1000 0 10UU Feet SHINGLE RAMPART BNULUER ZONE ` Va FIG. 1. A, Location of the Low Isles on the GBR and other localities mentioned in the text (dots indicate adjacent coastal settlements; stars indicate sites of major sponge collections undertaken by the QM). B, Low Isles (from Stephenson etal., 1931), showing collection localities ofthe 1928-29 expedition (triangles; taken from Burton's text) and 1997 QM expedition (stars). LOW ISLES SPONGES popularity largely to their close proximity to human settlement, their wide variety of habitat types (typical of the chain of about 50 low woody islets on the far northern sector of the GBR, of which the Low Isles are the most southern), in- cluding sandy beaches, a vegetated sand cay, extensive coral reef flat and lagoon, fringing reefs, and large stands of ‘uninhabitable mangrove swamp’ (Yonge, 1928), as well as a permanent settlement on the sand cay since 1878 associated with the operation and maintainance of the lighthouse — now a heritage listed building (Anon., 1993). Between the early 1880s and the early 1900s William Saville-Kent and Robert von Lendenfeld actively collected and described sponges from far northern Queensland. Unfortunately, neither author provided specific or reliable locality or habitat data, with the exception of collections made during the pearl oyster surveys off Cape York in the late 1800s (in which case the locality "Torres Strait” was usually quoted). Where locality data did exist on specimen labels it was often contradicted in the corresponding museum register and again in the published records, and therefore all of these data must be treated as suspect (Hooper & Wiedenmayer, 1994). Never- theless, it is likely that some of their material was collected from reefs in the vicinity of Cairns and Port Douglas given the close proximity of the GBR to the coast in this region, and the pop- ularity of these reefs. Their collections were deposited in both the Natural History Museum, London (BMNH), and Australian Museum, Sydney (AM), but much of this early material is dry and virtually useless for modern taxonomic determination. In 1925 the Great Barrier Reef Committee proposed a concerted program to explore the ‘origin, growth and natural resources of the Great Barrier Reef” (Yonge, 1928), with the Low Isles subsequently chosen as the site for a major ex- pedition to undertake in situ studies of coral reefs and their processes, led by C.M. Yonge. The expedition remained on the Low Isles for just over twelve months during 1928-29, During this time they surveyed most of the available habitats on and surrounding the two islands of the Low Isles (Stephenson et al., 1931). From Stephen- son’s description of sampling localities and methods, this effort was rigorous and compre- hensive, even by today’s standards. Collection of biological samples included reef-walking, dredging and diving via surface supply air (SSA) (‘tin-hat’ diving). m w The Scientific Reports of the Great Barrier Reef Expedition 1928-29 (British Museum (Natural History): London), were published in six volumes between 1928 and 1950, representing the most comprehensive study on coral reef biology, physics, chemistry and geology of the GBR system at that time, and perhaps of coral reefs in general. The sponge fauna from this expedition was published by Burton (1934), who described 36 species from the Low Isles and another 70 species from coral reefs and inter-reef habitats further north (mostly in the vicinity of Lizard 1.). Discounting the publications from the ‘Alert’ (Ridley, 1884) and ‘Challenger’ expeditions (e.g. Ridley & Dendy, 1887), which mostly concerned the coast and islands of the Torres Straits and not the GBR proper, Burton was the first author to provide accurate locality and habitat data for GBR species, unlike his predecessors Saville-Kent and Lendenfeld. It was not until 35 years later that Bergquist (1969) published the next paper on GBR sponges, and another 10 years after that with the subsequent work of Wilkinson (1978). These latter publications described only a few intertidal and shallow subtidal species, from the southern end of the GBR (Heron I.), and consequently Burton’s (1934) species have stood for over 50 years as being ‘typical’ or ‘representative’ of the entire GBR. Until this current decade his work has represented virtually the sum-total of our knowledge of the GBR sponge fauna. Burton’s (1934) species were divided into two groups: 1) ‘Common Indo-Malay’, with *Indo-Malayan' species (38% of his collections), allegedly ‘cosmopolitan’ species (12%), and ‘typical east Australian coast’ species (17%); and 2) ‘Indigenous’, with apparent ‘endemic’ species (14%), and exclusively northern Australian species (19%), described from one or only few localities. Of the former group he rarely provided descriptions or referred to any museum voucher specimen to validate his identifications; of the latter group only relatively few have been sub- sequently recollected or redescribed in the literature (e.g. de Laubenfels, 1954), some of which we suspect, or now know, are mis- identifications. It was the intention of this study, therefore, to revisit the Low Isles to: 1) ‘Rediscover’ Burton’s GBR species, locate and re-examine his voucher specimens (if they existed), of the allegedly ‘cosmopolitan’ species in particular, and ultimately to assign Burton’s species names to 252 140, O Burton 1934 E QM 1997 NUMBERS Oo o o Ng N oega No.Spp =n Genera Families FIG. 2. Comparison of species diversity and taxonomic composition between Low Isles sponges collected by the GBR Expedition 1928-29 (Burton, 1934) and col- lections of the QM in 1997, indicating the total num- ber of species collected (and species common to both expeditions), the number of new (or unnamed) spe- cies, numbers of genera and families. living populations — a theoretically simple but practically elusive task for many Australian sponge faunas. 2) Compare sponge biodiversity and species composition between the Low Isles and other reefs of the GBR from our contemp- orary collections (see Fig. 1A), to ascertain whether this fauna is indeed representative of the GBR fauna in general as has been interpreted by many contemporary authors. To achieve these aims, without having to revise the entire northern GBR fauna, we restricted this study to include only the Low Isles, ignoring for the time being those species Burton described from the more northern reefs of Eagle, Direction, Lizard, Turtle and Howick Is. MATERIALS AND METHODS All sponges were collected using SCUBA, by hand for the intertidal fauna, or a small dredge for deeper subtidal soft-bottom species. All spec- imens are housed in the permanent collections of the QM (prefix QMG). Methods of preservation, histological preparation and taxonomic identifi- cation are published elsewhere (e.g. Hooper, 1996). Abbreviations: BMNH, The Natural History Museum, London; GBRMPA, Great Barrier Reef Marine Park Authority; QM, MEMOIRS OF THE QUEENSLAND MUSEUM Queensland Museum, Brisbane; SSA, surface supplied air. RESULTS AND DISCUSSION BIODIVERSITY. The published sponge fauna of the entire Queensland region, including coast- line, Great Barrier Reef, Queensland Plateau, and the Coral Sea, so far consists of only 428 named species and subspecies (Hooper & Wiedenmayer, 1994, including literature updated since 1994). Fewer than this, perhaps 250 named species, actually belong to the GBR fauna, with the remainder restricted to coastal waters, soft sediments in the Gulf of Carpentaria, the inter-reef region in the Torres Straits, and deeper-waters off the continental shelf. Recent collections by the QM from the GBR have subsequently recorded 507 species, many of which are probably new to science (Hooper et al., 1999, this volume). Since Burton’s (1934) work there were no subsequent publications of GBR sponges until Bergquist’s (1969) description of a small inter- tidal collection from Heron I. Since Bergquist (1969), only relatively few other publications containing descriptions or redescriptions of GBR sponges have appeared, although these seem to be slowly escalating, perhaps reflecting the renewed interest in the phylum and in biodiversity in general (Wilkinson, 1978; Ayling, 1982; Pultizer-Finali, 1982; Thompson et al., 1987; Hooper, 1987, 1990, 1991, 1996; Bergquist et al., 1988; Stoddart, 1989; Wilkinson & Cheshire, 1989; Fromont, 1989, 1991, 1993, 1995; Van Soest et al., 1991; Hooper & Bergquist, 1992; Reitner, 1992; Hooper et al., 1993; Hooper & Lévi, 1993a, b, 1994; Van Soest & Hooper, 1993; Fromont et al., 1994; Bergquist, 1995; Bergquist & Kelly-Borges, 1995; Kelly-Borges & Vacelet, 1995; Reitner & Woerheide, 1995; Van Soest et al., 1996; Reitner et al., 1997). Burton (1934) recorded 36 species from the Low Isles, collected over a 12 month period by the GBR Expedition, consisting of 5 new species, 25 genera and 19 families. By comparison, collections made by the Queensland Museum in 1997 over 7 days, from similar habitats encircling the islands as described by Stephenson et al. (1931), yielded 109 species (in 59 genera and 33 families; Fig. 2), of which only 46 (42%) can be accurately assigned to a known species — i.e. the remainder are possibly new to science or perhaps belong to species described by Lendenfeld (1888, 1889) but whose identity is still a ‘mystery’ LOW ISLES SPONGES 253 (Hooper & Wiedenmayer, 1994). Surprisingly, only 12 species were common to both the Burton and QM collections from the Low Isles (although we also collected another 12 species from the Low Isles that were reported by Burton (1934) from the GBR Expedition collections made at Lizard, Turtle and Direction Islands, but not previously found on the Low Isles). In order to verify conspecificity between these two collections we undertook a search for Burton's (1934) Low Isles voucher specimens in the BMNH, of which all but three species were found (Table 1), Re-examination of this material found 15 species (42%) were misidentified, 12 belonging to completely different species than supposed by Burton (1934), and 3 split into different species (i.e. allopatric sibling species, as opposed to so called ‘widespread’ species); | is uncertain (i.e. the voucher specimen is missing and no description was provided by Burton); and 14 are transferred to other genera (based on more recent systematic revisions). Most of these 15 misidentified species were assigned by Burton (1934) to species that had ‘wide Australian distributions’ (i.e. temperate Australian, Northern Territory, and/or tropical Western Australian), ‘widespread Indo-Pacific’ (e.g. Indo-Malay archipelago, Sri Lanka and W Indian Ocean), or ‘cosmopolitan species’ (e.g. Mediterranean, Caribbean, Atlantic). These mis- identifications were detected and confirmed by comparing Burton’s samples with the type material (and/or contemporary specimens) of his named species from these other localities (QM and BMNH collections). Quantitative differences in species diversity between the GBR Expedition (36 spp.) and QM collections (109 spp.) are not surprising given the greater technological advances made in con- temporary collecting techniques (SCUBA, underwater photography), and the probable ineffectual use of generalist (non-specialist) biological collectors to undertake sponge faunal surveys, irrespective of the substantial dif- ferences between time scales of two collections (12 months versus 7 days duration, respectively). For example, Burton (1934) described Raphido- tethya enigmatica and recorded lanthella flabelliformis from more northerly reefs in the Lizard Island region (but not from the Low Isles), whereas we found both these species were relatively common on the Low Isles subtidal reefs. It is possible (but not explicit in their reports), that the GBR Expedition did not commonly use SSA and dredging around the Low Isles themselves (whereas we do know they used these techniques on the more northerly reefs), and it is likely that many or most of the Low Isles sponges were collected from the intertidal reef flat (Fig. 1B). Thus, based on the recent QM collections and the revised Burton (1934) collections, the total species diversity for the Low Isles now consists of 134 species (in 63 genera and 35 families) (Fig. 2). SPECIES COMPOSITION. The low similarity in species composition between the GBR Expedition and QM sponge collections 1s more surprising. Only 12 species or 33% of Burton’s (1934) published fauna were common to both collections, consisting mainly of widespread GBR species (e.g. Druinella purpurea, Carteriospongia foliascens, Haliclona cymaeformis, Cinachyra australiensis). Several explanations are apparent. 1) Perhaps the more recent QM collection did not find the other 66% of Burton’s (1934) species because of the shorter time-scale for collection (7 days versus 12 months), whereby these other species might represent the rare or cryptic species? This explanation is highly unlikely, however, given that we have found some of Burton’s Low Isles species elsewhere on the GBR, from collections of similar duration, and in some cases (e.g. Spirastrella inconstans, Callyspongia diffusa) these species are common. 2) It is also possible that the GBR Expedition mainly, or perhaps exclusively, targetted the easily accessible intertidal fauna, whereas QM collections were predominantly (although not exclusively) subtidal. 3) Nevertheless, there are several species (particularly some of the Haliclona and Callyspongia described by Burton) which are common on the intertidal reef flats of other reefs on the GBR, but apparently not present on the Low Isles today. It is possible that some of these species may be ‘locally extinct’ due to anthropo- genic or natural causes. Burton’s (1934) misidentifications are less easily explained. Burton had ready access to the vast BMNH collections, containing types, fragments, or representative samples of most species known at that time from the Australian, Oriental, Afrotropical, Neotropical and Palae- arctic provinces, yet 42% of his species are not conspecific with these allegedly ‘widely dis- tributed’ or ‘cosmopolitan’ species. For example, Burton recorded Jaspis stellifera from the Low Isles, noting that it did not contain asters, whereas FIG. 3. Biogeographic comparisons in sponge diversity and species composition between the Low Isles and adjacent provinces (data from Hooper et al., 1999, this volume). Square = percentage of Low Isles species that are ‘apparent endemics’; circles = percentage of Low Isles species also found in other provinces. our re-examination of his material found that it did contain asters, and moreover was not con- specific with J. stellifera, differing significantly in growth form, surface features, skeletal struc- ture, megasclere and microsclere dimensions from southern Australian populations. Burton’s specimen appears to be a new species. Other authors have also recorded similar discrepancies. Bergquist and Warne (1980) found a 25% dif- ference between Burton’s spicule measurements from the holotype of Callyspongia diffusa and their own re-examination of this specimen. Burton also appears to have overemphasised the importance of external characters in identifying some of his material, overlooking other im- portant skeletal characters. For example, his record of Haliclona camerata appears to be solely based on external features (growth form, surface features), whereas Ridley’s (1884) holotype has a multispicular skeleton with spicules 25% larger than Burton’s voucher specimen, which has a unispicular skeleton — again Burton’s specimen appears to be a new species. MEMOIRS OF THE QUEENSLAND MUSEUM BIOGEOGRAPHY. A comparison of species diversity and composition between the Low Isles (including Burton’s (1934) revised species list and our more recent QM collections), with sponge faunas of other reefs in the northern part of the GBR, indicate several patterns (Fig. 3). 1) 31 species (or 23% of the Low Isles fauna) are distributed throughout the GBR (annotated ‘3’ on Table 1). These species were recorded on virtually every reef we have surveyed so far on the GBR, and they can be defined as a ‘typical GBR sponge fauna’. Thus, the concept ofa ‘GBR sponge fauna’ is partially substantiated. Conversely, QM collections recorded several other species common throughout the GBR but notably absent from the Low Isles: Acanthella costata, Amphimedon terpenensis and another (new) species of Amphimedon, Callyspongia carens and several other Callyspongia spp., Crella calypta, Echinochalina intermedia, Hippospongia elastica, Hyrtios erecta, Phakellia flabellata, P. klethra, Phyllospongia papyracea, and several apparently undescribed species of Dysidea, Haliclona, Niphates, Pericharax, Psammoclemma, Pseudoceratina and Siphonochalina. In addition, the cryptic, cave- dwelling coral species Levinella prolifera, Astrosclera willeyana, Acanthochaetetes wellsii, Sycetta sp. and Hypograntia sp. are also absent from the Low Isles, probably because these specialised habitats are not present (e.g. Woerheide & Reitner, 1998). 2) Recent collections from Lizard Island, about 200km N ofthe Low Isles and closer to the outer barrier reef, found 176 species (Hooper et al., 1999, this volume). Of the Low Isles fauna 41 species (31%) are also found on Lizard I., with these two islands showing the highest affinities in their sponge faunas. 3) Recent collections from the adjacent north- ern coastal province (including fringing coral reefs, intertidal rock reefs, embayments and muddy reefs near the shore, extending along the Queensland coast from the Cooktown region into the Gulf of Carpentaria), found 142 species (Hooper et al., 1999, this volume). A comparison between the Low Isles sponges and this coastal fauna shows that only 17 species (13% of the Low Isles fauna) were common to both provinces (annotated ‘4’ on Table 1). Furthermore, when considered separately each of these provinces usually had an even lower similarity in species composition: Gulf of Carpentaria (8% of Low Isles species), Torres Strait (9%), Shelburne Bay LOW ISLES SPONGES including the Cockburn and Fast I. groups (20%), Turtle I. (6%) (Fig. 3). These data suggest that the Low Isles contain a greater proportion of ‘coral reef species’ than ‘inshore coastal species’, despite their closer proximity to the coast. 4) Recent collections of sponges from the coral reefs on seamounts in the Coral Sea (Osprey, Wreck, Cato and Saumerez Reefs), found 95 species (Hooper et al., 1999, this volume). A comparison between the Low Isles fauna and Coral Sea sponges shows that only 13 species (10% of the Low Isles fauna) were common to both provinces (annotated ‘5’ on Table 1). 5) Only 4 species were found in all 3 regions (Coscinoderma matthewsi, Halichondria n.sp. #1227, Myrmekioderma granulata, and Xestospongia testudinaria). 6) Recent collections from the SE Queensland fauna (extending from Hervey Bay to Moreton Bay), found 233 species (Hooper et al., 1999, this volume). Comparisons with these SE Queens- land faunas found only 14 species of Low Isles sponges (10% of the fauna) extended southward into this region: Chondrilla australiensis, Echinodictyum mesenterinum, lanthella basta, I. flabelliformis, lotrochota foveolaria, Leucetta microraphis, Myrmekioderma granulata, Pericharax heterorhaphis, Phakellia cavernosa, Pseudaxinella australis, Xestospongia testudinaria and 4 undescribed species of Dysidea, Spirastrella, Timea and Clathria (Microciona). Similarly, recent collections from N NSW (Byron Bay to Gold Coast) and S NSW (Sydney, Illawarra and Port Stephens regions) found 69 and 131 species from these regions, respectively (Hooper et al., 1999, this volume). Only 4 species living on the Low Isles also extend into S New South Wales. 7) A large number of species on the Low Isles are either ‘apparent endemics’ or have very restricted distributions here and on adjacent reefs. 43 species (32% of the Low Isles fauna) have not yet been found anywhere else, and another 22 (16%) are known only from the Low Isles and one other reef in the northern part of the GBR (mostly from Lizard Island). Thus, nearly 50% of the sponge fauna on the Low Isles is unique to this N GBR region. It is possible that this high ‘apparent species endemism' might be related to true regional endemism (such as the concept of a *northern GBR fauna"). There is some empirical support for this through comparisons with S GBR reefs: 18% of Low Isles species were recorded on Bait and Hook Reefs (central GBR); 18% from the Swain Reefs (S GBR, outer reefs); and 16% from reefs in the vicinity of Heron I. (S GBR, inner reefs) (Fig. 3). It is also probable that some of this ‘apparent endemism’ is due to the heterogeneous distributions of many coral reef sponges (Hooper, 1994), perhaps related to particular habitat requirements and local geomorphological differences between individual reef systems (such as the availability of specialised habitats on particular reef systems). COMMERCIAL ‘BATH’ SPONGES. Scientific investigation and commercial ‘exploitation’ of the Low Isles may have commenced as early as the 1890s, with the alleged introduction of commercial ‘bath’ sponge cuttings, apparently imported from the Mediterranean, seeded on the reef flat between the two islets (‘Thalamita Flat’ and ‘Mangrove Park’). Surviving remnants (or decendents) of these populations are still common in this area, with some more-or-less ‘organised’ into vague rows. Burton (1934) identified this species as the Mediterranean Spongia officinalis. Its status as a possible rem- nant of a commercial ‘sponge farm’ is supported to some extent by our 1997 observations of its ‘organised’ distribution into ‘vague rows’ on the reef flat. It is possible that Saville-Kent may have been responsible, directly or indirectly, for introducing these ‘bath’ sponges onto the Low Isles, given the popularity of ‘translocating’ exotic species during his era; he was also the Queensland Commissioner of Fisheries around this time (Harrison, 1997); and there is an anecdotal record of commercial sponge beds occuring on the Isles dating back to about the 1890s (Port Douglas Historical Society; pers. comm.). However, this evidence is inconclusive. It is more probable that these ‘bath’ sponge beds are remnants of the ‘seeding experiments’ conducted on the Low Isles and Murray Islands (Torres Strait) by Moorehouse during the GBR Expedition, and described in his report on the investigation of the potential viability of commercial sponge farming on the Great Barrier Reef (Moorhouse, 1933), Moorhouse noted that he made cuttings of wild populations of a ‘black, dome-shaped Hippo- spongia’, fitting the description of Burton’s (1934) S. officinalis, which he seeded on the reef flat using various commercial methods of his day. This suggests that these commercial ‘bath’ sponges may be native to the GBR and not intro- duced, and therefore probably not conspecific 256 with the Mediterranean S. officinalis. Re-examination of Burton’s (1934) voucher specimen of S. officinalis from the Low Isles (Table 1) showed that it belonged to Hippo- spongia (our sp. #1983), and not to Spongia. This surviving population on the Low Isles possibly represents the first attempt at sponge culture on the GBR. CONCLUSIONS Patterns in species diversity and composition of Low Isles sponges (Fig. 3) indicate a greater proportion of both ‘typical GBR species’ and “indigenous species’ (most similar to Lizard I. than other reefs); only a small proportion of species shared with adjacent coastal and oceanic provinces; and very few species shared with more southern Australian provinces. In fact Burton (1934: 513) acknowledges that ‘[although] the sponges collected by the [GBR] Expedition belong ... to species characteristic of the Indo-Pacific ... many common to the coasts of Australia ... [with] mixing of the Australian and Malayan sponge-faunas ... this broad generalization [is] in itself inconclusive and unsatisfactory, [but] is the most that can be said’. He states further that comparison between the collections of the GBR Expeditions and those of Saville-Kent (the latter comprising an over- whelming number of indigenous species, but unfortunately with no locality data), suggests that generalizations about a ‘GBR sponge fauna’ based on the Low Isles species list are probably invalid. In this conclusion he is undoubtedly correct, given the peculiar nature of the Low Isles (inshore coastal reef), as compared with outer barrier reefs of the GBR in particular. However, to some extent there does appear to be a ‘typical GBR sponge fauna’ of about 20% of regional species’ compositions, and some of these (perhaps up to 10%) are truly widely distributed throughout the Indo-west Pacific (although this latter estimate still lacks good empirical support). There is also indication that closer similarities between northern GBR reefs than with southern GBR reefs suggests the concept of a ‘typical GBR fauna’ may be too simplistic, and that the GBR itself comprises more than one province. Burton’s (1934) assumption that a significant number (12%) of GBR species may be ‘cos- mopolitan’, also found in the West Indies, Azores and Mediterranean, is rejected. His voucher specimens of all these allegedly ‘cosmopolitan’ species are misidentifications. The concept of a MEMOIRS OF THE QUEENSLAND MUSEUM generalised ‘east coast Australian sponge fauna’ (Lendenfeld, 1888, 1889; Burton, 1934) is also not supported (with the exception of 4 species). Nevertheless, despite the fact that 42% of Burton’s species were misidentified, and only relatively few species were reported from the Low Isles themselves, Burton’s (1934) report still stands as a valuable taxonomic contribution and a reasonable precis of faunal relationships of GBR sponges in general. ACKNOWLEDGEMENTS This study would not have been possible without the special permission of Great Barrier Reef Marine Park Authority (GBRMPA permit no. G96/005), to undertake extractive research from reefs of the Low Isles. Since the late 1800s the popularity of the Low Isles has led to in- evitable environmental degradation. Consequently, GBRMPA and the Department of Environment proposed a strict plan of management for the Isles (Anon., 1993). Now implemented, the plan includes designated zones for specific use, a restriction of daily visitor numbers, and (more importantly) restrictions on the types of research activities now permitted. These latter restrictions include the curtailment of manipulative and ex- tractive research (i.e. no collecting). The School of Marine Science, University of Queensland, now operates a small research station housed in some of the refurbished lighthouse buildings (Low Isles Research Station), with the intention to repeat the marine chemistry and physics ex- periments pioneered by the GBR Expedition. This present study was undertaken by the Queensland Museum in a similar spirit to revisit the pioneering work of Yonge, Burton and coworkers, and for this opportunity we are grateful to the Department of Environment and the Low Isles Preservation Society, Port Douglas. We are also grateful to the Department of Primary Industries, Fisheries, Cairns, for access to their facilities and charter of the FV ‘Gwendolyn May’. LITERATURE CITED ANONYMOUS 1993. Low Isles Draft Management Plan for Low Islets (Low Island & Woody Island) and Reef. February 1993. (Great Barrier Reef Marine Park Authority: Townsville). AYLING, A.L. 1982. A redescription of Astrosclera willeyana Lister, 1900 (Ceratoporellida, Demospongiae), a new record from the Great Barrier Reef. Memoirs of the National Museum of Victoria 43: 99-103, LOW ISLES SPONGES 257 TABLE 1. List of species collected from the Low Isles during the GBR Expedition 1928-29, described by Burton (1934), with revised nomenclature from re-examination of relevant BMNH voucher specimens, and list of species collected in 1997 by the QM. Key to codes: 1 = species collected by the QM from other reefs in the GBR but not found in our collections from the Low Isles. 2 = species reported by Burton from other more northerly reefs but not present in his Low Isles collection. 3 = species now known to be widespread throughout the Great Barrier Reef and some other Indo-west Pacific reefs. 4 = species found on both the Low Isles and the adjacent coast. 5 = species found on both the Low Isles and Coral Sea reefs. 4 — species identification presently unknown, possibly new, with unique QM species number indicated. GBR Expedition 1928-29 collection from Low Isles (Burton, 1934) BMNH voucher numbers Revised name QM 1997 collection from Low Isles CALCAREA Pericharax heteroraphis Poléjaeff, 1884 (2,3,5) Sycon gelatinosum (Blainville, 1834) (1) 1930.8.13,29a Sycon gelatinosum (Blainville, 1834) Leucetta microraphis Haeckel, 1872 (3,5) ASTROPHORIDA Jaspis stellifera (Carter, 1879) (1) 1930.8.13.23a Jaspis n.sp. (not Jaspis stellifera (Carter, EZD ý Jaspis n.sp. #2242 Jaspis n.sp. #1005 Jaspis splendens (de Laubenfels, 1954) SPIROPHORIDA Cinachyra australiensis (Carter, 1886) 1930.8.13.14a Cinachyra australiensis (Carter, 1886) Cinachyra australiensis (Carter, 1886) (3,5) Cinachyra sp. #1870 Cinachyrella sp. #2270 Raphidotethya enigmatica Burton, 1934 (2,3,5) Raphidotethya sp. #2045 HADROMERIDA Pseudosuberites andrewsi Kirkpatrick, 1900 (1) 1930.8.13.20a Pseudosuberites andrewsi Kirkpatrick, 1900 Suberites peleia (de Laubenfels, 1954) Laxosuberites proteus Hentschel, 1909 1930.8.13.11la Laxosuberites proteus Hentschel, 1909 Polymastia megasclera Burton, 1934 1930.8.13.155a Polymastia megasclera Burton, 1934 Polymastia sp. #2258 Tethya robusta Bowerbank, 1859 (1) 1930.8.13.199a Tethya robusta Bowerbank, 1859 Tethya coccinea Bergquist 8 Kelly-Borges, 1991 Tethya sp. #2249 Timea sp. #1389 Spirastrella inconstans Spirastrella inconstans (Dendy, 1887) (Dendy, 1887) (1,3) missing (some description provided) E Spirastrella aurivillii Lindgren, 1897 (1) missing ? (no description provided) = - - - Spirastrella sp. #1385 - : - Chondrilla australiensis Carter, 1873 (2,3) Chondrilla nucula Schmidt, 1862 1930.8.13.23a | Chondrilla cf. nucula Schmidt, 1862 E Chondrilla sp. #492 HAPLOSCLERIDA Haliclona camerata (Ridley, 1884) (1,3) 1930.8.13.60a Haliclona sp. ee: Haliclona camerata (Ridley, 1884)) Haliclona clathrata (Dendy, 1895) 1930.8.13.57 Reniera sp. (not Haliclona clathrata (Dendy, 1895)) Haliclona exigua (Kirkpatrick, 1900) 1930.8.13.53a Xestospongia exi: (Kirkpatrick, 1900) Xestospongia exigua (Kirkpatrick, 1900) (3) Haliclona pigmentifera (Dendy, 1408) (1) i 1930.8.13.55a Haliclona sp. TE Haliclona pigmentifera (Dendy, 1905)) 258 MEMOIRS OF THE QUEENSLAND MUSEUM Haliclona tenuispiculata Burton, 1934 1930.8.13.59a Haliclona tenuispiculata Burton, 1934 Haliclona sp. #1954 (5) Haliclona sp. #2246 Haliclona sp. #2247 Haliclona sp. #2248 Adocia fibulatus var. microsigma Dendy, 1916 missing Haliclona cymaeformis (Esper, 1791) mo description but ID probable from urton’s remarks) Haliclona cymaeformis (Esper, 1791) (3 Haliclona (Toxadocia) sp. #2253 Adocia toxius (Topsent, 1897) 1930.8.13.38a Haliclona sp. (not Haliclona toxius (Topsent, 1897) Adocia minor (Dendy, 1916) (1) 1930.8.13.62a Adocia ur (het Haliclona minor (Dendy, 1916)) Adocia pumila (Lendenfeld, 1887) (1) 1930.8.13.32a Gelliodes pumilus (Lendenfeld, 1887) Adocia sagittaria (Sollas, 1902) 1930.8.13.40a Oceanapia sagittaria (Sollas, 1902) Oceanapia sagittaria (Sollas, 1902) Aka sp. #1373 Aka sp. #2254 Aka sp. #2255 Aka sp. #2259 Gelliodes sp. #1215 Gelliodes sp. #2244 Gellius sp. #2269 Niphates sp. #2245 Call ia di Ride 1884) O 1930.8.13.47a Callyspongia (Euplacella) diffusa (Ri le. EA É i Callyspongia ridleyi Burton, 1934 (1) 1930.8.13.165a Callyspongia ridleyi Burton, 1934 Callyspongia sp. #981 istulosa , 1873) (1) Oceanapia (Bower! 1930.8.13.50a Oceanapia sp. (iot O. fistulosa (Bowerbank, 1873) Oceanapia reneiroides Burton, 1934 1930.8.13.49a Oceanapia reneiroides Burton, 1934 Oceanapia renieroides Burton, 1934 Petrosia sp. #2252 Strongylophora sp. #1580 Xestospongia testudinaria (Lamarck, 1815) (3,4,5) Xestospongia nigricans (Lindgren, 1897) Xestospongia pacifica Kelly-Borges & Bergquist, Toss (3,5) POECILOSCLERIDA Desmapsamma anchorata (Carter, 1882) (1,3) 1930.8.13.151a Ceratopsion n.sp. (Raspailiidae) P Desmapsamma sp. #1528 Totrochota purpurea (Bowerbank, 1875) 1930.8.13.90a Totrochota foveolaria (Lamarck, 1814) Totrochota foveolaria (Lamarck, 1814) (4) lotrochota sp. #377 Iotrochota sp. #2256 lotrochota sp. #2263 Clathria aculeata Ridley, 1884 1930.8.13.93a Clathria (Thalysias) abietina (Lamarck, 1814) Clathria (Thalysias) abietina (Lamarck, 1814) Tenacia coralliophila (Theile, 1903) 1930.8.13.107 Clathria (pai sias) n.sp. ne Clathria (Thalysias) coralliophila (Thiele, 1903), C karian ce sias) cervicornis (Thiele, 1903 Clathria (Thalysias) lendenfeldi Ridley & Dendy, 1886 (4) Clathria (Thalysias) tingens Hooper, Cian (Thalysias) ting P Clathria (Thalysias) vulpina (Lamarck, 1814) (2,3,4) LOW ISLES SPONGES 259 Ophlitaspongia rimosa (Ridley, 1884) (1) 1930.8.13.17a Clathria Geer eccentrica (Burton, 1934) Ophlitaspongia eccentrica Burton, [os 1930.8.13.109a Clathria veer) eccentrica (Burton, 1934) Clathria oriel eccentrica (Burton, 1934) Clathria (Microciona) n.sp. #1882 Clathria (Microciona) n.sp. #2265 Echinochalina (Echinochalina) tubulosa (Hallmann, 1912) (4) Raspailia (Raspaxilla) reticulata Hooper, 1587 Echinodictyum mesenterinum (Lamarck, 1814) Endectyon elyakovi Hooper, 1991 Raspailia (Raspaxilla) n.sp. #2264 Thrinacophora n.sp. #1993 (5) Biemna sp. #2260 Coelocarteria singaporensis (Carter, 1883) (2,3,4) Crella sp. #2243 Strongylacidon sp. #1533 Zyzzya sp. #1653 HALICHONDRIDA Acanthella n.sp. #1562 Auletta constricta Pulitzer-Finali, 1982 (3) Axinella n.sp. #2267 Axinella carteri (Dendy, 1889) (3,4) Axinyssa n.sp. #2257 Cymbastela concentrica (Lendenfeld, 1887) (3) Cymbastela coralliophila ooper & Bergquist, 1992 (3) Phakellia cavernosa (Dendy, 1921) (2,3,4) Pseudaxinella australis Bergquist, 1970 (3) Reniochalina cf. stalagmitis sp. #417 (4) Reniochalina stalagmitis Lendenfeld, 1888 (4) Leucophloeus fenestratus Ridley, 1884 (1) 1930.8.13.153a Ciocalypta fenestratus (Ridley, 1884) Ciocalypta n.sp. #2251 Halichondria sp. #1227 (4,5) Halichondria stalagmites (Hentschel, 1912) Hymeniacidon n.sp. #2261 Myrmekioderma granulata (Esper, 1830) (3,4,5) Liosina paradoxa (Thiele, 1899) (3) DICTYOCERATIDA Phyllospongia dendyi Lendenfeld, 1889 1930.8.13.199a Lendenfeldia plicata (Esper, 1806) Carteriospongia foliascens (Pallas, 1766) 1930.8.13.203a Carteriospongia foliascens (Pallas, 1766) Carteriospongia foliascens (Pallas, 1766) (3) Coscinoderma mathewsi (Lendenfeld, 1886) (3,4,5) Spongia officinalis Linnaeus, 1759 1930.8.13.188a Hippospongia sp. (not S. officinalis Linn.) Hippospongia sp. #1983 (5) Spongia cf. officinalis ecole sp. #262 (3) Dactylospongia elegans (Thiele, 1899) (3,5) Ircinia sp. #1534 Ircinia sp. #1876 (5) Ircinia sp. #2268 Ircinia cf. ramosa #1377 260 MEMOIRS OF THE QUEENSLAND MUSEUM Psammocinia sp. #487 Fascaplysinopsis sp. #2170 Fascaplysinopsis reticulata j (Hentschel, 1912) (3,5) Dysidea herbacea (Keller, 1889) 1930.8.13.175a Dysidea herbacea (Keller, 1889) Dysidea herbacea (Keller, 1889) (3) - Dysidea sp. #229 (4) - Dysidea sp. #1214 - Dysidea sp. #2250 - Dysidea sp. #2262 - Dysidea sp. #2266 VERONGIDA > Aplysinella rhax (de Laubenfels, 1954) (3) Druinella purpurea (Carter, 1880) 1930.8.13.198a Druinella purpurea (Carter, 1880) Druinella purpurea (Carter, 1880) (3) - - Pseudoceratina sp. #1565 - Pseudoceratina sp. #2196 - Pseudoceratina sp. #2399 - Tanthella basta (Pallas, 1766) (4) - lanthella cf. flabelliformis sp. #196 (4) - Tanthella flabelliformis Pallas, 1766) (2,3,4) TOTAL BURTON SPECIES = 36 spp. (5 new) 15 misidentified spp., 1 sp. uncertain, 14 revised generic assignments TOTAL QM SPECIES = 109 spp (61 spp. unnamed possibly new) Comparison between Burton & OM collections — 12 spp. in common. Total Low Isles fauna: 134 species, 63 genera, 35 families. BERGQUIST, P.R. (1969). Shallow water Demo- spongiae from Heron Island. Papers of the University of Queensland Heron Island Research Station 1(4): 63-72. 1995, Dictyoceratida, Dendroceratida and Verongida from the New Caledonia Lagoon (Porifera: Demospongiae). Memoirs of the Queensland Museum 38(1): 1-51. BERGQUIST, P.R. & WARNE, K.P. 1980. The marine fauna of New Zealand: Porifera, Demospongiae, Part 3 (Haplosclerida and Nepheliospongida). New Zealand Department of Scientific and Industrial Research Bulletin. New Zealand Oceanographic Institute Memoir 87: 1-77. BERGQUIST, P.R., AYLING, A.M. & WILKINSON, C.R. 1988. Foliose Dictyoceratida of the Aust- ralian Great Barrier Reef. 1. Taxonomy and Phylogenetic relationships. Pubblicazione della Stazione Zoologica di Napoli I: Marine Ecology 9(4): 291-319. BERGQUIST, P.R. & KELLY-BORGES, M. 1995. Systematics and biogeography of the genus lanthella (Demospongiae: Verongida: Ianthellidae) in the South-west Pacific. The Beagle, Records of the Northern Territory Museum of Arts and Sciences 12: 151-176. BURTON, M. 1934. Sponges. Scientific Reports of the Great Barrier Reef Expedition 1928-29. 4: 513-621. (British Museum (Natural History): London). FROMONT, J. 1989. Aspects of the reproductive biology of Xestospongia testudinaria (Great Barrier Reef). Pp. 685-691. In Proceedings of the 6th International Coral Reef Symposium, Australia, 1988. Vol. 2. (James Cook University of North Queensland: Townsville). 1991. Descriptions of species of the Petrosida (Porifera: Demospongiae) occurring in the tropical waters of the Great Barrier Reef. The Beagle, Records of the Northern Territory Museum of Arts and Sciences 8(1): 73-96. 1993. Descriptions of species of the Haplosclerida (Porifera: Demospongiae) occurring in the tropical waters of the Great Barrier Reef. The Beagle, Records of the Northern Territory Museum of Arts and Sciences 10: 7-40. 1995. Haplosclerida and Petrosida (Porifera: Demospongiae) from the New Caledonia Lagoon. Invertebrate Taxonomy 9: 149-180. FROMONT, J., KERR, S., KERR, R., RIDDLE, M. & MURPHY, P. 1994, Chemotaxonomic relation- ships within, and comparisons between, the orders Haplosclerida and Petrosida (Porifera: Demo- spongiae) using sterol complements. Biochemical Systematics and Ecology 22(7): 735-752. HARRISON, A.J. 1997. Savant of the Australian Seas. William Saville-Kent (1845-1908) and Australian Fisheries. (Tasmanian Historical Research Association: Hobart). HOOPER, J.N.A. 1987. New records of Acarnus Gray (Porifera: Demospongiae: Poecilosclerida) from Australia, with a synopsis of the genus. Memoirs of the Queensland Museum 25(1): 71-105. 1990. A new species of Rhabderemia Topsent (Porifera: Demospongiae) from the Great Barrier Reef. The Beagle, Records of the Northern Territory Museum of Arts and Sciences 7(1): 65-78. LOW ISLES SPONGES 1991. Revision of the family Raspailiidae (Porifera: Demospongiae), with description of Australian species. Invertebrate Taxonomy 5(6): 1179-1418. 1994, Coral reef sponges of the Sahul Shelf—a case for habitat preservation. Memoirs of the Queens- land Museum 36(1): 93-106. 1996. Revision of Microcionidae (Porifera: Poecilo- sclerida: Demospongiae), with description of Australian species. Memoirs of the Queensland Museum 40: 1-626. HOOPER, J.N.A. & BERGQUIST, P.R. 1992. Cym- bastela, a new genus of lamellate coral reef sponges. Memoirs of the Queensland Museum 32(1): 99-137. HOOPER, J.N.A., KELLY-BORGES, M. & RIDDLE, M. 1993. Oceanapia sagittaria from the Gulf of Thailand. Memoirs of the Queensland Museum 33(1): 61-72. . HOOPER, J.N.A. & LEVI, C. 1993a. Poecilosclerida (Porifera: Demospongiae) from the New Caledonian lagoon. Invertebrate Taxonomy 7(5): 1221-1302. 1993b. Axinellida (Porifera: Demospongiae) from the New Caledonian lagoon. Invertebrate Taxonomy 7(6): 1395-1472. 1994. Biogeography of Indo-west Pacific spanges: Microcionidae, Raspailiidae, Axinellidae. Pp. 191212. In Soest, R.W.M. Van, Kempen, T.M.G. van & Braekman, J.-C. (eds) Sponges in Time and Space. (Balkema: Rotterdam). HOOPER, J.N.A. & WIEDENMAYER, F. 1994. Porifera. Pp. 1-620. In Wells, A. (ed.) Zoological Catalogue of Australia, Volume 12. (CSIRO Australia: Melbourne). HOOPER, J.N.A., KENNEDY, J.A., LIST- ARMITAGE, S.E., COOK, S.D. & QUINN, R. 1999. Biodiversity, species composition and distribution of marine sponges in northeastern Australia. Memoirs of the Queensland Museum (this volume). KELLY-BORGES, M. & VACELET, J. 1995. A revision of Diacarnus 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. LAUBENFELS, M.W. DE 1954. The sponges of the West-Central Pacific. Oregon State Monographs, Zoology 7: 1-306. LENDENFELD, R. VON 1888. Descriptive Catalogue of the Sponges in the Australian Museum, Sydney. (Taylor & Francis: London). 1889. A Monograph of the Horny Sponges. (Trúbner & Co.: London). MOOREHOUSE, F.W. 1933. Commercial sponges from the Barrier Reef. Reports ofthe Great Barrier Reef Committee 4(1): 30-34. PULITZER-FINALI, G. 1982. Some new or little-known sponges from the Great Barrier Reef of Australia, Bolletino dei Musei e degli Istituti 261 Biologici dell'Università di Genova 48-49 (1980-1982): 87-141. REITNER, J. 1992. ‘Coralline Spongien’ Der Versuch einer phylogenetisch-taxonomischen Analyse. Berliner Geowissenchaftliche Abhandlungen, Reihe E, B and 1 (Selbstverlag Fachbereich Geowissenschaften, Freie Universität Berlin: Berlin). REITNER, J. & WOERHEIDE, G. 1995. New Recent sphinctozoan coralline sponge from the Osprey Reef (N’ Queensland Plateau, Australia). Fossil Cnidaria and Porifera 24(2): 70-71. REITNER, J., WOERHEIDE, G., LANGE, R. & THIEL, V. 1997. Biomineralization of calcified skeletons in three Pacific coralline demosponges - an approach to the evolution of basal skeletons. Cour.Forschungs-Institut Senckenberg 201: 371-383. RIDLEY, S.O. 1884. Spongiida, (Part 1, Australian sponges). Pp. 366-482, pls 39-43. In Reporton the Zoological Collections made in the Indo-Pacific Ocean during the voyage of H.M.S. ‘Alert’, 1881. (British Museum (Natural History): London). RIDLEY, 5.0. & DENDY, A. 1887. Report on the Monaxonida collected by H.M.S. ‘Challenger’ during the years 1873-76. Report on the Scientific Results of the Voyage of H.M.S. ‘Challenger’ during the Years 1873-76. 20(59): 1-275 (Her Majesty’s Stationary Office: London, Edinburgh, Dublin). SOEST, R.W.M. VAN, HOOPER, J.N.A. & HIEMSTRA, F. 1991. Taxonomy, phylogeny and biogeography of the marine sponge genus Acarnus (Porifera: Poecilosclerida). Beaufortia 42(3): 49-88. SOEST, R.W.M. VAN £ HOOPER, J.N.A. 1993. Taxonomy, phylogeny and biogeography of the marine sponge genus Rhabderemia Topsent, 1890 (Demospongiae: Poecilosclerida). In Uriz, M.J. & Riitzler, K. (eds). Recent Advances in Ecology and Systematics of Sponges. Scientia Marina 57(4):; 319-351. SOEST, R.W.M. VAN, DESQUEYROUX- FAUNDEZ, R., WRIGHT, A.D. & KONIG, G.M. (1996). Cymbastela hooperi sp.nov. (Hali- chondrida: Axinellidae) from the Great Barrier Reef, Australia. In Willenz, P. (ed.) Recent advances in sponge biodiversity inventory and documentation. Bulletin de l'Institut Royal des Sciences Naturelles de Belgique 66(Supplément): 103-108. STEPHENSON, T.A., STEPHENSON, A., TANDY, G. & SPENDER, M. 1931. The structure and ecology of Low Isles and other reefs. Scientific Reports of the Great Barrier Reef Expedition 1928-29 3(2): 17-122 (British Museum (Natural History): London). STODDART, J.A. 1989. Foliose Dictyoceratida of the Australian Great Barrier Reef. III. Preliminary electrophoretic systematics. Pubblicazione della 262 Stazione Zoologica di Napoli I: Marine Ecology 10(2): 167-178, THOMPSON, J.E. MURPHY, P.T., BERGQUIST. P.R. & EVANS, E.A. 1987, Environmentally induced variation in diterpene composition of the marine sponge Rhopaloeides odorabile. Biochemical Systematics and Ecology 15(5); 595-606. WILKINSON, C.R. 1978. Description of two Demo- spongiae, one being toxic, from the Great Barrier Reef. Tethys 8(3): 267-270. WILKINSON, C.R. & CHESHIRE, A.C. 1989, Patterns in the distribution of sponge populations NEW DATA ABOUT MORPHOLOGY AND FEEDING PATTERNS OF BARENTZ SEA HALICHONDRIA PANICEA (PALLAS), Memoirs of the Queensland Museum 44: 262. 1999:- Visual observations in the marine aquaria and transmission electron microscopy studies on the larvae of the intertidal sponge Halichondria punicea demonstrated individual variations in external and internal morphology, behaviour and type of metamorphosis. Parenchymulae of this species were found to possess the ability to actively feed by endocytosis (phago- and pinocytosys). The larvae crawled over the substrate and cast numerous unicellular organisms (bacteria and flagellates from 2 - Aum in size) onto the body surface by a flagellum. During this, the apical parts of the flagellated cells formed large lobopodia that served for catching and ingesting food particles, | monitored the consequent patterns of contact of the flagellates with the surface of lobopodia, their entrapment, submersion, the formation and transport of the digestive phagosomes into the basal parts of the surface cells. Each surface locomotory cell was capable of catching and ingesting food. No morphological and/or functional differences between the surface cells were found. Nevertheless, singular flagellated cells packed MEMOIRS OF THE QUEENSLAND MUSEUM across the central Great Barrier Reef. Coral Reefs 8; 127-134. WOERHEIDE, G. & REITNER, J. 1998. Bio- geography and taxonomy of the reef cave dwelling coralline demosponge Astrosclera willeyana throughout the Indo-Pacific. P. 88. In Book of Abstracts. Sth International Spnge Symposium, “Origin and Outlook”, 29 June-3 July 1998. (Queensland Museum: Brisbane). YONGE, C.M. 1928, Origin, organization and scope of ihe expedition. Scientific Reports of the Great Barrier Reef Expedition 1928-29 1: 1-11. (British Museum (Natural History): London). with the phagosomes submerged inside the larva. Here these cells could be easily distinguished by the presence of a flagellum and the typical shape of the nucleus. Later on, the submerged flagellated ceils withdrew the flagellum and acquired an amoeboid shape. Final digestion of the caught organisms occurred only inside the larva. It was suggested that endosymbionts found in the surface and inner cells of the larvae served as an additional food source for the larvae, Presence of ihe numerous pinocytosis vacuoles in the apical parts of the flagellated cells suggested that the sponge larvae are also able to absorb dissolved low-molecular matter. To conclude, parenchymula of A. panicea could be recognised as a living embodiment (a living model) of the hypothetical phagocytella of Mechnikov in which the differentiation of the body layers into kinoblast and phagocytoblast is only primordial, purely Functional and still reversible. O Porifera, intertidal, larva, Jeeding, digestion, endocytosis, digestive phagosomes, Hulichondria panicea. L.V. Ivanova (email: inna(@sokolzoo.sph.su), Pedagogical State University named after A. TL Herzen. River Moika Emb., 48, St.Petersburg, 191186, Russia; 1 June 1998. BIODIVERSITY, SPECIES COMPOSITION AND DISTRIBUTION OF MARINE SPONGES IN NORTHEAST AUSTRALIA JOHN N.A. HOOPER, JOHN A, KENNEDY, SUSAN E. LIST-ARMITAGE, STEPHEN D, COOK AND RON QUINN Hooper, J.N.A., Kennedy, J.A., List-Armitage, S.E., Cook, S.D. & Quinn, R. 1999 06 30: Biodiversity, species composition and distribution of marine sponges in northeast Australia. Memoirs of the Queensland Museum 44: 263-274, Brisbane. ISSN 0079-8835. Biodiversity, species composition and biogeographic relationships were compared between 18 regional populations of marine sponges along the NE Australian coastline (extending northwards from Byron Bay to the E Gulf of Carpentaria, including the southern fauna [rom Sydney-Illawarra as an outgroup comparison), based exclusively on samples of living populations. Much of the older literature concerning Australian sponge taxonomy is too unreliable to be used effectively as a tool to determine conspecificity and explore faunistic relationships, and consequently this literature was ignored completely in our analyses using instead recent collections made throughout the study arca, all documented in situ. Levels of biodiversity varied considerably between many regions, related in part to the size and diversity of habitats present in particular regions, but also to differences in collection effort, Several regions with apparently low sponge diversity (eg. 3 seamounts in the Coral Sea) were clearly biased by correspondingly low collection efforts, whereas in other regions these biodiversity data appear to be more realistic indicators of species richness. Faunas of the Gulf of Carpentaria and Turtle Islands were more intensively sampled but had relatively low sponge diversity, whereas those of the Swain Reefs, Capricorn-Bunker Group, Lizard Island and Moreton Bay regions had much higher species diversity with equivalent (and sometimes lower) collection effort. Five of the seven relatively highly diverse regions lay in the south (Swain Reefs, Capricorn-Bunker Group, Moreton Bay, Sydney-Illawarra, Sunshine Coast), with only two northern regions showing comparable diversity (Lizard Island, Low Isles), contrary to latitudinal irends in diversity found in some other marine phyla. Statistically these trends do hot appear to be artifacts of sampling effort but reflect true differences in provincial diversity. The number of unique (apparent endemic) species within each ofthe 18 regions had a median value of about 33%, although this value varied considerably between particular regional faunas, Species endemism was seen to he largely a function of their biogeographic isolation or proximity to other regional faunas and to ecological factors such as the possession of unique habitat types. Regions wilh highest levels of relative endemism were Sydney- lllawarra (the most southern region; 81% of species), Wreck Reef (the most isolated oceanic region; 46%), and the Gull of Carpentaria (differing substantially from all other regions in its habitat composition; 45%), Consistent discovery of about 33% af new (i.e. not previously encouniered) species trom each reef system surveyed suggests that the possible sponge biodiversity in NE Australia greatly exceeds previous estimates of about 1,500 species, Within NE Australia (ignoring the S NSW outgroup), ive provincial faunas were recognised, grouped hierarchically based on parsimony analysis, each showing grealer similarities in species composition within-regions, and fewer similarities between-regions: 1) Tweed River (Byron Bay to the Gold Coast) (with 30% provincial species endemism); 2) SE Qld (Moreton Bay to Hervey Bay) (49%); 3) GBR (Capricorn-Bunker Group to the Cockburn I5) (70%); 4) far northern region (coastal reef and islands in the vicinity of northern Cape York, extending into the Gulf of Carpentaria) (52%); 5) Coral Sea (49%) (although not yet substantially surveyed). Lendenleld's concept af a homogencous E Australian coastal fauna is rejected, and the possibility that both the GBR and Coral Sea regions each comprise more than single provinces requires further investigation. CJ Porifera, biodiversity, biogeography, fuuna survev, species distribution patterns. endemism, NE Australia, Great Barrier Reef, Coral Sea. John N.A. Hooper (email: Johittaigm.gldgovan), Susan E. List-Armituge, John A. Kennedy & Stephen D. Cook, Marine Biology Laboratory, Queensland Museum, PO. Box 3300, South Brisbane 4101, Australia; Ron Quim, Queensland Pharmaceutical Research Institute, Griffith University. Corner Dan young Road and Forest Cauri, Mt Gravatt Research Park, Nathan 4111, Australia: 2 March 1999, 264 This study examines the biodiversity, species composition and biogeographic relationships between regional populations of marine sponges along the NE Australian coastline, extending northwards from the Byron Bay region (N New South Wales (NS W)) to the E Gulf of Carpentaria (N Queensland (Qld)), with the southern fauna of Sydney-Illawarra region used as an outgroup comparison. Our study is based exclusively on our samples of living populations, documented using contemporary methods. For reasons ex- plained below we ignored the published literature completely in these analyses, Biogeographic relationships of Great Barrier Reef (GBR) sponges in particular, and of the Qld fauna in general, have been speculative ever since the pioneering studies in this region by Ridley (1884), Poléjaeff (1884a,b), Ridley & Dendy (1887), Sollas (1888), Lendenfeld (1883, 1885a,b, 1887, 1888, 1889), Thiele (1898, 1903), Schulz (1900), Kieschnick (1900) and Burton (1934). Together these earlier authors indicated that a large proportion of this fauna consisted of ‘widely distributed Indo-Malay’, ‘Indo-Pacific’, ‘cosmopolitan’ or ‘general east Australian coast- al’ species, with a much smaller proportion of indigenous species. There is some evidence from contemporary collections to support this contention in the older literature that a certain proportion of tropical and subtropical Indo-Pacific sponges extensively range in distribution from the Red Sea to the central west Pacific islands. These species are thought to comprise between 5% (Hooper & Lévi, 1994) and 15% of regional faunas (Hooper, 1994), and they mostly concern species associated with coral reefs, belonging to many different families and orders (i.e. demonstrating a diversity of reproductive strategies and mech- anisms for dispersal). They occupy a diversity of coral reef habitats, including the reef flats and lagoons (e.g. Hyrtios erecta (Keller), Carterio- spongia foliascens (Pallas)); coral rubble (e.g. lotrochota baculifera Ridley, Spirastrella (Spheciospongia) vagubunda (Ridley), Tethya robusta Bowerbank); deeper fringing reefs (e.g. Axinella carteri (Dendy), Janthella basta (Pallas)); and specialised habitats such as coral caves (e.g. Astrosclera willeyana Lister). The occurrence of these species in a particular region may be linked to the presence or absence of these habitats on each reef (e.g. Hooper, 1994), and in some cases dispersal has been assisted through anthropogenic activities (such as ship bilge water (e.g. Mycale (Zygomycale) parishii (Bower- MEMOIRS OF THE QUEENSLAND MUSEUM bank)), and oyster farming (e.g. Cliona vastifica Hancock) (e.g. Wesche et al., 1997)). Morphometrically these widely dispersed populations appear to be conspecific and in some cases they do not appear to vary morphologically across this vast geographic range. But it is still unknown to what extent these discontiguous regional populations differ genetically, their potential capabilities for interbreeding or re- hybridising, or any realistic estimates of what proportion of these species are truly widely dispersed and what proportion consist of com- plexes of closely related, but genetically distinct, species (sibling species). Increasingly, however, many of these allegedly widely distributed morphospecies are being found to consist of heterogeneous allopatric populations, with biochemical and genetic diversity not necessarily manifested at the morphological level (e.g. Solé-Cava and Thorpe, 1986, 1994; Hooper et al, 1990, 1992: Solé-Cava et al., 1991, 1992; Bavastrello & Sarà, 1992; Boury-Esnault et al., 1992; Kerr and Kelly-Borges 1994; Kelly-Borges et al. 1994; Klautau et al., 1994; Solé-Cava et al., this volume). To date only one allegedly widely distributed species, A. willeyana, has been sampled across the entire Indo-Pacific system, including populations from the GBR (Woerheide, 1997). Chemical and genetic analyses suggest that regional populations of this morphospecies may consist of several discrete sibling species, corresponding to subtle but consistent morpho- metric differences between them. No other data are yet available for other species from the GBR. The possibility that regional endemism amongst the GBR and Qld species may be higher than previously recognised (Hooper & Lévi, 1994) is supported from three sources. 1) In the more recent literature on GBR sponges local populations of so-called widely distributed species are recognised as belonging to distinct species (Wilkinson, 1978; Pulitzer-Finali, 1982; Thompson et al., 1987; Hooper, 1987, 1990, 1991, 1996; Bergquist et al., 1988, 1990; Sarà, 1990; Fromont, 1991, 1995; Van Soest et al., 1991, 1996; Hooper & Bergquist, 1992; Van Soest & Hooper, 1994; Bergquist & Kelly-Borges, 1991, 1995; Kelly-Borges & Vacelet, 1995). These contemporary studies differ from the older literature largely through their recognition that consistent (and sometimes subtle) morphometric differences between regional populations may constitute valid interspecific differences, as opposed to merely recognition of (sometimes substantial) intraspecific variability. A common BIODIVERSITY OF NORTHEASTERN AUSTRALIAN SPONGES feature of these contemporary studies is that they were largely based on living populations and not solely reliant on often antiquated, preserved or dry, museum voucher specimens (which lose most of their useful field characteristics). Some of these studies also include chemical and genetic data to support their morphological hypotheses. By comparison, very few authors of the older literature had access to living populations, with few (if any) data on living species’ character- istics. For many taxa (particularly Chalinidae, Callyspongiidae, Halichondriidae), such data are mandatory, and consequently, as stated long ago by Hallmann (1912), many of the identifications in the older literature have long been doubtful. Unfortunately, however, these species describ- ed in the contemporary literature comprise only a relatively small proportion of the published fauna of the entire GBR and Qld, with most species names established in the older literature (see Hooper & Wiedenmayer, 1994). 2) Our re-examination of some museum voucher specimens described in the older literature has found many instances where species were misidentified, with regional populations being unjustifiably ‘lumped’ into a single widely distributed or so-called cosmopolitan taxon (e.g. Hooper, 1991, 1996; Hooper & Weidenmayer, 1994; Hooper et al.,1999, this volume). Un- fortunately, again, relatively few of these older species have yet been revised — a long and arduous process — and the status of many nominal species throughout the GBR and Qld faunas is still in doubt. Until identifications can be con- firmed, estimates of endemism are equivocal, and endemism is referred to as ‘apparent’. 3) Throughout the Indo-Pacific there are pub- lished regional faunas which have much higher levels of species endemism and relatively fewer widely distributed species than has been suggested for the GBR and Qld in the older literature. This extra-limital literature includes both earlier authors (e.g. Topsent (1897) and Thiele (1900, 1903) in describing the Ambon and Ternate faunas; de Laubenfels (1954) on the central west Pacific island and atoll faunas) and more contemporary publications (e.g. Bergquist (1968 et seq.) on the New Zealand fauna; Lévi (1967 et seq.) on the New Caledonia fauna). Theoretically, levels of species endemism amongst Qld and GBR faunas might also be expected to approach these other regions, but the existing taxonomic literature is largely unreliable to serve as a basis to analyse faunistic relation- ships of sponges in this region. 265 For these reasons, it has not been possible to develop any reliable hypothesis on the bio- geographic affinities of the GBR and Qld sponge faunas, even though 428 ‘valid’ species of sponges have already been published from this region (Hooper & Wiedenmayer, 1994; including literature published since 1994). Many of these species are still poorly known, with relatively few subsequently recorded since they were first described (particularly those of Lendenteld). Moreover, recent collections from this region now consist of >1,500 species, most documented from living populations (Queensland Museum (QM) collections), but most cannot yet be assigned reliably to a known taxon given the largely inadequate descriptions in the older literature, their lack of published data on living characteristics, the significant proportion of misidentifications amongst the so-called widely distributed species, and the inaccessibility, scattered and time-consuming task of locating and re-examining type collections. Consequently, we chose to make use of these comprehensive, but still largely unnamed QM sponge collections to explore the biogeographic affinities within the Qld regional faunas by ignoring the published literature completely. This literature, concerning the 428 described species from Qld waters, is summarised in Hooper & Wiedenmayer (1994), The QM collections were primarlily obtained from shallow coastal waters of the Qld coast, GBR and the Coral Sea (0-70m depth), with accurate GPS locality data, habitat descriptions and underwater photography. They were obtained using SCUBA and trawling, and have been identified and documented to species level (with many already known to be new to science). Our standardised method of collection and documentation provides us with the ability to unequivocally differentiate between closely related sibling species and not to rely solely on the literature to determine conspecificity and faunistics relationships. It is well beyond the scope of this paper to provide a comprehensive list of raw species data used to compare regional and provincial faunas. These raw data have been included (in tabular format) on the senior author’s personal web page at the QM web site (http://www.qmuseum.qgld.gov.au). Of these OM collections we selected 17 discrete regions within the Old fauna (i.e. ignoring some of the dispersed inter-reef regions sampled such as the collections described by Cannon et al., 1987). Together these collections consisted of approximately 800 species. As an MEMOIRS OF THE QUEENSLAND MUSEUM TABLE 1. Regional species diversity (bold numbers in the diagonal row) and similarities in species composition between sponge faunas of central and NE Australia (upper half of matrix showing the numbers of species shared between each region; lower half of matrix showing the percentage similarity between regional faunas (Greig-Smith Similarity Index; Krebs, 1978)). Key to regions: A, Sydney-Illawarra region; B, Tweed River region, from Byron Bay to the Gold Coast; C, Moreton Bay region, within the bay and outside the bay from South Stradbroke I. to Flinders Reef, N of Moreton I.; D, Sunshine Coast region, from Mooloolaba to Noosa Heads; E, Hervey Bay region, including W side of Fraser I.; F, N islands of the Capricorn-Bunker Group, S Great Barrier Reef; G, Wreck Reef, S Coral Sea; H, Cato Reef, S Coral Sea; I, Saumarez Reef, S Coral Sea; J, Swain Reefs, S Great Barrier Reef; K, Bait and Hook Reefs, Whitsunday Is region, central Great Barrier Reef; L, Lizard I. region, including the Direction Is and MacGilvray Reef, N Great Barrier Reef; M, Turtle Is region, N Great Barrier Reef (trawled fauna); N, Low Isles, N Great Barrier Reef; O, Osprey Reef, Far N Coral Sea; P, Shelburne Bay region, Far N Great Barrier Reef, including the Cockburn and Fast Is (trawled fauna); Q, Torres Strait region (trawled fauna); R, E Gulf of Carpentaria region (trawled fauna). Number of shared species Region | A B C D E F G H I J K L M N [9] P Q R A 131 | 10 10 9 2 2 0 0 3 2 2 l 1 0 5 1 1 B 10 | 69 | 33 14 5 7 1 0 0 8 2 5 4 1 0 6 3 3 Cc 6.7 | 28 | 166 | 31 1 19 6 1 3 28 8 13 3 7 2 16 4 9 | D 7.6 | 16 | 23 | 106 | 11 19 7 1 1 25 10 | 23 3 12 6 12 4 5 E 22 | 8.1 16 14 54 1l 1 0 0 11 3 7 3 6 3 8 2 4 F 0.6 | 5.5 11 13 | 91 | 187 | 21 5 7 61 22 | 44 5 25 6 23 5 4 G 19 | 13 | 49 | 75 | L5 16 | 81 7 4 25 7 20 1 12 5 5 6 1 H 0 11 | 17 0 5 15 14 2 4 1 2 0 2 0 1 0 0 I 0 0 34 | 17 0 d 8.6 | 15 12 3 A 3 0 3 0 E 0 0 J 18 | 58 15 16 | 84 | 31 17 | 3.6 | 2.7 | 208 | 28 62 6 29 13 27 11 6 K 2.1 | 3.2 | 72 | 12 | 54 | 18 10 | 28 | 87 | 21 57 | 3 5 16 4 19 4 3 L 1.3 | 41 | 7.6 | 16 | 61 | 24 16 | 2.1 | 3.2 | 32 27 | 176 | 16 | 39 16 | 32 12 9 M 1 5.8 2.5 3.4 4.8 3.9 1.3 0 0 4,3 7.9 13 70 6 0 13 6 8 N 0.8 | 11 | 49 | 11 69 | 16 12 3 4.5 18 18 26 | 63 | 134 | 6 22 9 8 [9] 0 0 2 84 | 66 | 54 | 8.5 0 0 11 8.5 15 0 7.6 | 37 5 0 1 P 43 | 71 12 12 10 16 155 | 17 | X5 17 24 | 23 15 20 | 72 | 100 | 16 16 Q ll | 5.2 | 38 | 53 4 43 | 94 0 0 87 | 78 11 10 11 0 2 46 9 R 1 47 8 6 32 | 14 0 0 45 | 51176 | 12 | 88 | 21 | 2 17 | 60 Similarity Index (%) outgroup comparison to check on species relationships throughout the Qld faunas we used recent collections of 131 species from the Sydney-Illawarra region, NSW, all of which we have documented, identified and described in the same manner as the Qld voucher material (i.e. again ignoring the published NSW fauna of Lendenteld (1884 et seq.), Whitelegge (1889 et seq.), Hallmann (1912 et seq.) and others), MATERIALS AND METHODS Species diversity, composition and distrib- utions were compared between 17 discrete regional faunas within NE Australia, extending from Byron Bay (N coast of NSW) northwards to the E Gulf of Carpentaria (Qld), including several seamounts in the Coral Sea, and comparing these with the Sydney-Illawarra region (S NSW; see Fig. 1). From the OM sessile marine invertebrate databases we retrieved 913 species collected from these regions. Some of these species have been described and recorded previously from Qld waters in the literature whereas most cannot be identified with a known taxon (i.e. probably new to science). Only species (known and unknown) for which we have collected a voucher specimen during our contemporary collections were con- sidered in this study. Other species records from the literature from the Qld fauna, for which we do not yet have a voucher specimen in QM col- lections were ignored and are not included in this study. Hence, the potential diversity of regional sponge faunas is much larger than we consider here, whereas the uncertainty still surrounding some of these taxa preclude us from using them reliably in our species' inventories. BIODIVERSITY OF NORTHEASTERN AUSTRALIAN SPONGES A ok S * E : Jj te nd An HEGEN NO SPP LAIQUE BPPUNDUE BPP LO COLL i] —L 176 50285 39 RI G0) 27 45%) 26 za 01 46) 15339 10 P101 2424* 45 | je | ae * * O 3| M39 7 * ARE H IN 134) 4432% 15 N M 70 25359 2 Y and Z are the total number of species in each region and X isthe number of shared species between regions Y and Z, expressed as à percentage). A cluster analysis was performed on all pairwise comparisons and plotted using UPGMA (Group- Average) sorting, and data were checked for consistency using non-parametric Spearman's rank correlation analysis on all pairwise correlations. A hierarchical classification based on a heuristic distance matrix was calculated G from all pairwise comparisons using PAUP 3.1.1 (Swofford, 1993) with dendrograms plotted using MacClade (Maddison & Maddison, 1992). Based on these comparisons, regions of highest similarity in species composition were then combined into six provincial faunas and reanalysed K 5 916% 5 C using these same methods. J 20 2 Together these analyses provid- ! EE E e E s AB ed information on biodiversity T sy, (total number of species within Hid 14, 4298 1 cach of the 18 regions): regional G| 8l 37 46% 12 endemism (number of unique IF 187, 71 3895 30 species within each region); sim- IE! 54| 15285 6 ilarities and differences between D 106) 35339 12 regional faunas ero a pairwise comparision of similar € | 185. En JS 45 species for each region, expressed B 69 213505 9 as a total number of shared A 131 106 BI% 30 CM FIG.1. Distribution of NE Australian regional sponge faunas, showing collecting localities grouped into regional faunas, table of the total species and a similarity index); and a biogeographic model (using cluster analysis and distance matrices producing a hierarchical classification of regional faun- istics similarities). number of species collected, the number and percentage of unique species for each region (apparent endemics), and number of collection RESULTS AND DISCUSSION stations in each region for which sponges were present (see Table 1 for key to regions). Species lists were generated for each of the regional faunas from the QM databases. Species were then tabulated as present/absent for each of the 18 regions, producing a pairwise matrix ofthe number of shared species between individual regions, and a simple index of similarity calculated for each pairwise comparison (Greig- Smith 1964, in Krebs, 1978: as (2* X/Y-Z) where SPECIES DIVERSITY. Levels of biodiversity in each of the 18 selected regions (Fig. 1) varied considerably (Table |), undoubtedly related in part to the size and diversity of habitats present in each reef system. In some cases these differences were biased by differences in the collection effort between regions, whereas in other cases they reflect more true indications of diversity (table in Fig. 1). TABLE 2. Levels of species endemism for each province (combined regions), showing total number of species, number and percentage of unique species (apparent endemics) in each province (see Table 1 for key to regions). Region No. spp Unique spp % Unique A 131 106 81 B 69 21 30 CDE 233 114 49 GHI 95 47 49 FJKLNOP 507 356 70 MOR 142 74 52 Few collections were made at Cato and Saumarez Reefs in the Coral Sea, and these reefs were also relatively homogeneous compared to other regions sampled, both factors reflecting their low sponge diversities (14 and 12 spp. respectively). In contrast, higher diversities in Moreton Bay (166 spp.), the Capricorn-Bunker Group (187 spp.) and Shelburne Bay (including the Cockburn and Fast Islands) (101 spp.) are undoubtedly related to the presence of larger and more diverse habitats in these regions, although there were also more collections undertaken from each region. By comparison, Lizard Island and the Low Isles are both relatively small reef systems but contain relatively high sponge diversities (176 and 134 spp., respectively), although the former region had over twice the collection effort of the latter. The Sunshine Coast (106 spp.) and Swain Reefs (208 spp.) had relatively fewer collections than these other regions but relatively high diversity, the latter the most diverse region yet sampled. Two inshore faunas, the Turtle Islands (70 spp.) and eastern Gulf of Carpentaria (60 spp.) were characterised by shelly and soft sediments and murky waters, yielding only few species despite relatively higher collection efforts. These trends are summarised in Figure 4. In the case of Cato Reef, Saumarez Reef, Osprey Reef, Torres Strait and Bait and Hook Reefs, apparent low biodiversity is obviously related directly to collection effort, whereas for other reef systems our collections are more valid indicators of existing sponge diversity. This is particularly evident for Lizard Island, the Capricorn-Bunker Group and the Swain Reefs in which species diversity increased despite a consecutive decrease in collection effort (Fig. 4). These differences are confirmed through one-way ANOVA, comparing MEMOIRS OF THE QUEENSLAND MUSEUM the numbers of species collected, the number of unique species, and the number of collections made (Table 2), showing significant differences between their means (P<0.001). From our data there is no evidence that species diversity increases at lower latitudes, contrary to some other phyla of marine invertebrates in which biodiversity generally increases towards the equator, especially within the GBR system (e.g. Rohde, 1979). In fact the reverse appears to be true for sponges, in which five of the seven most diverse regions lay in the south (Swain Reefs, Capricorn-Bunker Group, Moreton Bay, Sydney-lllawarra and Sunshine Coast regions), and only two northern regions had comparable sponge diversity (Lizard Island, Low Isles). SPECIES COMPOSITION. Affinities between regional faunas generally appear to be related to their proximity to each other, such that adjacent regions usually had higher proportions of similar species than did those further apart (Fig. 2). Regions containing the highest proportions of unique species (i.e. apparent endemics) were not necessarily those containing the highest bio- diversity; nor were they always artifacts of higher collection efforts (P<0.001; Fig. 4), but were those that were either more isolated or contained substantially different habitats than other regions (table in Fig. 1). The southernmost region, Sydney-Illawarra, had 81% endemic species; the most isolated oceanic coral reef, Wreck Reef, had 46%; and the Gulf of Carpentaria, with mainly soft substrata, had 45% unique species (Fig. 1). By comparison, levels of species endemism for most other regions were consistent (between 24-38%), with the exceptions of Bait and Hook Reefs in the central GBR (16%) which probably contains a more even mixture of species from both northern and southern GBR faunas. These data support previous contentions that sponge species distributions are notoriously heterogeneous, particularly in coral reef faunas, with differences in faunal composition partly attributed to differences in geomorphology between reefs (Hooper, 1994), but also with bio- geographic factors influencing composition (as indicated by the correlation between proximity and similarity in species composition). For each new reef system visited about 30% of species had not previously been encountered, with many of these possibly also new to science. Thus, our prediction of sponge biodiversity for Qld. (about 1500 species; Hooper & Lévi, 1994; Hooper & Wiedenmayer, 1994), may be a gross BIODIVERSITY OF NORTHEASTERN AUSTRALIAN SPONGES 1, TEMPERATE ROCKY REEFS 2, TEMPERATE- SUBTROPICAL ROCKY REEFS L 3. SUBTROPICAL ROCKY AND 1 CORAL REEFS AND EMBAYMENTS 2 4, GREAT BARRIER REEF CORAL REEFS 5. SOUTHERN CORAL SEA OCEANIC CORAL REEFS 6. FAR NORTHERN INSHORE AND MUDDY SUBSTRATA A. SYDNEY REGION B. BYRON BAY - GOLD COAST C. MORETON BAY REGION D. SUNSHINE COAST E. HERVEY BAY E CAPRICORN-BUNKER GP J. SWAIN REEFS L. LIZARD ISLAND AND OUTER REEFS N.LOWISLES P. SHELBURNE BAY REGION K. OUTER REEFS WHITSUNDAY REGION O. OSPREY REEF G. WRECK REEF H CATOREEF L SAUMAREZ REEF M. TURTLE ISLANDS Q. TORRES STRAIT REGION R. EASTERN GULF OF CARPENTARIA 269 REGIONAL BIOGEOGRAPHY. On the basis ofthese trends we com- bined the data for the 18 regional faunas into 6 provincial faunas, and repeated this analysis (Fig. 3, Tables 2-3). Ignoring for the time being the most southern region (Sydney- Illawarra, used as an outgroup com- parison), and the most isolated region (oceanic southern Coral Sea), similarities in species composition between the other four provinces along the NE coast ranged from only 18-25%. Highest species diversity and apparent endemism was found in the GBR provincial fauna (507 spp. and 70%, respectively). It has been sug- gested, based on more subjective criteria (Hooper et al., 1999, this volume) that recognition of a single GBR fauna may be artificial, with the possible existence of separate northern and southern GBR prov- inces. To test this we compared the two southern GBR regions (Cap- FIG, 2. Cladogram illustrating the hierarchical classification of affinities between regional sponge faunas, based on heuristic distance matrices computed using PAUP 3.1.1, indicating 6 mjaor faunistic provinces. Major provinces are: 1, Temperate rocky reefs; 2, Temperate-subtropical rocky reefs; 3, Subtropical rocky and fringing coral reefs and embayments; 4, Coral islands and reefs of the Great Barrier Reef; 5, S Coral Sea oceanic coral seamounts; 6, Far N coastal islands around Torres Strait and Gulf of Carpentaria, including fringing coral reefs and inter-reef soft substrata. underestimate, and it is conceivable that twice this number may live in this region. Parsimony analysis, producing a hierarchical classification of similarities between regions based on regional species compositions (Fig. 2), grouped the 18 regional faunas into 6 logical provinces. These were generally (but not ex- clusively) correlated with their proximity to each other, their distance from the coast (and terrestrial influences), and possession of similar habitat types in each, such that these provincial groups appear to be valid indicators of biogeographic affinities: 1) Temperate rocky reefs; 2) Temperate-subtropical rocky reefs; 3) Subtropical rocky and fringing coral reefs and embayments; 4) GBR and island coral reefs; 5) Southern Coral Sea oceanic coral reefs; 6) Far northern inshore, fringing coral reefs and inter-reef soft substrata. ricorn-Bunker Group and Swain Reefs), showing a 31% similarity in their species compositions, with the three major northern GBR regions (Low Is, Lizard I. and Cockburn and Fast Is), showing similarities between 20-26%. We then compared the combined data sets for the two southern reefs with those of the three northern reefs, discovering that there were 88 species in common, with a similarity index of 22%. Thus, between-group comparisons clearly overlap the within-group comparisons, providing no statistical support for a proposal to subdivide the GBR province. Breaking the data down even further and re-examining all the pair-wise comparisons between species similarities for each of the individual GBR regions was also uninformative (Table 1). Similarly, any biogeographic trends in sponge distributions that may be useful as a basis for subdivision may also be partially masked by the well-known heterogeneity amongst coral reef sponges (Hooper, 1994), with the two factors difficult to separate. Nevertheless, from present data we can clearly differentiate at least five provincial sponge faunas within NE Australia, each having high levels of within-group regional endemism and 270 MEMOIRS OF THE QUEENSLAND MUSEUM @ = 50 species TABLE 3 NUMBER OF SHARED SPP. FJKL mo] an s 131] 10 | 16 | 2 | 6 | oe tet (Province CDE) a ) e Byrofi Bay regign (Province B) 6] ear] 20911811102] % iat - (Province A) FIG. 3. Distribution of NE Australian provincial sponge faunas (18 regional sponge faunas amalgamated into 6 major provinces based on PAUP analysis), showing species diversity (bar graphs), percentage similarity in species composition between adjacent provincial faunas (arrows) (see Table 1 for key to regions). TABLE 3. Similarities in species composition between the 6 major provinces (upper half of matrix showing the numbers of species shared between each province; lower half of matrix showing the per- centage similarity between provincial faunas using Greig-Smith Similarity Index; Krebs, 1978); and total number of species in each province (bold numbers in the diagonal row)) (see Table | for key to regions). relatively low levels of between- group similarities in species composition: |) Tweed River region (Byron Bay-Gold Coast), with 30% of species not yet found outside this province (apparent endemic species) (Table 2). 2) SE Qld (Moreton Bay — Hervey Bay), with 49% provincial endemism. This fauna is relatively homogeneous in comparison with the other provincial faunas. 3) GBR (Capricorn-Bunker Group — Cockburn Is), with 70% provincial endemism. 4) Far northern coastal and islands region (Torres Strait — E Gulf of Carpentaria), with 52% provincial endemism. It is also likely that this province could be further subdivided, given that the combined Torres Strait — Shelburne Bay regions have only a 17% sim- ilarity in their species composition with the Gulf of Carpentaria region (Table 1). 5) Coral Sea, with 49% provincial endemism. Further col- lections from these seamounts are necessary to determine whether they contain a single homogeneous fauna or several distinct provincial faunas. The concept of an E Australian coastal sponge fauna, mentioned frequently by Lendenfeld (1888, 1889) is rejected, and the concept of homogeneous GBR and Coral Sea coral reef faunas (cf. Burton, 1934) is also questionable, although more extensive sampling of regional faunas within each of these provinces is required to further investigate any proposed biogeo- graphic subdivisions. ACKNOWLEDGEMENTS This project has been funded since 1993 by the Queensland Pharmaceutical Research Insitute, Griffith University, Brisbane, and the Queensland Museum, to collect, document and describe marine invertebrate biodiversity throughout NE Australia, and to discover new chemical structures with potential therapeutic pharmaceutical benefits BIODIVERSITY OF NORTHEASTERN AUSTRALIAN SPONGES TOTALNO: SPECIES mm» [] S, southern, O, oceanic reefs) 5 5 mm No. UNIQUE SPECIES 250 - One-way ANOVA on species diversity vs. levels of endemism vs. collecting effort. SOURCE — [MEAN VARIANCE N A Peuom 2 " 150 + (N, northern; M, mid-reef, 3 3 c 100 4 & v = 50. Zz |I HOQEKRBMGPDNACLF J REGIONAL FAUNA FIG, 4. Comparison between regional species diversity (total number of species collected in each region), endemism (number of unique species in each regional fauna), and collection effort (number of collecting stations in each region that yielded sponges), with regions arbitrarily sorted on increasing species diversity. Results of one-way ANOVA between species diversity, levels ofendemism and collecting effort are tabulated, ("bioprospecting"). Without the support of Astra Pharmaceuticals (Australia) Pty. Ltd. these extensive collections would not have been possible. For permits to collect throughout Queens- Jand waters and the Coral Sea we thank the Great Barrier Reef Marine Park Authority, Queensland Department of Primary Industries and the Low Isles Preservation Society. For collections of sponges from the New South Wales coast we thank Ms Lisa Miller (AWT EnSight, Pty Ltd), and Mr Danny Roberts (EPA Sydney, now Wyong Shire Council), LITERATURE CITED BAVASTRELLO, G. & SARA, M. 1992, Morph- ological and genetic differences in ecologically distinct populations of Petrosia (Porifera: Demospongiae). Biological Journal of the Linnean Society 47: 49-61). BERGQUIST, PR, 1968. The marine fauna of New Zealand: Porifera, Demospongiae, Part 1 (Tetractinomorpha and Lithistida). Memoirs of the New Zealand Oceanographio Institute 37: 1-98, BERGQUIST, P.R., AYLING, A.M. & WILKINSON, C.R. 1988. Faliose Dietvoceratida of the Australian Great Barrier Reef. 1, Taxonomy and phylogenetic relationships, Marine Ecology 9(4): 291-320. BERGQUIST, P.R., CAMBIE, R.C, & KERNAN, MLR, 1990, Scalarane sesterterpenes from Collo- spongia auris, a new thorectid sponge, Biochemical Systematics and Ecology 18(5): 349-357. BERGQUIST, P.R. & KELLY-BORGES, M. 1991. An evaluation of the genus Tethya (Porifera; Demospongiae: Hadromerida) with descriptions of new species from the southwest Pacific. The Beagle, Records of the Northern Territory Museum of Arts and Sciences 8(1): 37-72. 1995. Systematics and biogeography of the genus lanthella (Demospongiae: Verongida: lanthellidae) in the south-west Pacific, The Beagle, Records of the Northern Territory Museum of Arts and Sciences 12: 151-176. BOURY-ESNAULT, N., SOLE-CAVA, A.M. & THORPE, J.P. 1992, Genetic divergence between morphs of the Mediterranean sponge Oscarella lobularis Schmidt (Porifera, Demospongiae, Oscarellidae). Journal of Natural History 26: 271-284. BURTON, M. 1934, Sponges. In Scientific Reports of the Great Barrier Reef Expedition 1928-29, 4(14); 513-621 (British Museum (Natural History): Landon). CANNON, L.R.G., GOEDEN, G.B. & CAMPBELL, P. 1987. Community patterns revealed bv trawling in the inter-reef regions of the Great Barrier Reef. Memoirs of the Queensland Museum 24(1): 45-70). 155] ~ to FROMONT, J. 1991. Descriptions of species of the Petrosida (Porifera: Demospongiae) occurring in the tropical waters of the Great Barrier Reef. The Beagle, Records of the Northern Territory Museum of Arts and Sciences 8(1): 73-96. 1993, Descriptions of species of the Haplosclerida (Porifera: Demospongiae) occurring in the tropical waters of the Great Barrier Reef. The Beagle, Records of the Northern Territory Museum of Arts and Sciences 10: 7-40 HALLMANN, E.F. 1912. Report on the sponges obtained by the F.I.S. ‘Endeavour’ on the coasts of New South Wales, Victoria, South Australia, Queensland, and Tasmania, 1909-10. Part 1. Zoological Results of the Fisheries Experiments carried out by the F.LS. ‘Endeavour’ 1909-10 2: 117-300. HOOPER, J.N.A. 1987. New records of Acarnus Gray (Porifera: Demospongiae: Poecilosclerida) from Australia, with a synopsis of the genus. Memoirs of the Queensland Museum 25(1): 71-105. 1990, A new species of Rhabderemia Topsent (Porifera: Demospongiae) from the Great Barrier Reef. The Beagle, Records of the Northern Territory Museum of Arts and Sciences 7(1): 65-78. 1991. Revision of the Family Raspailiidae (Porifera: Demospongiae), with description of Australian Species. Invertebrate Taxonomy 5(6): 1179-1415. 1994. Coral reef sponges of the Sahul Shelf—a case for habitat preservation. Memoirs of the Queensland Museum 36(1): 93-106. 1996. Revision of Microcionidae (Porifera: Poecilosclerida: Demospongiae), with description of Australian Species. Memoirs of the Queensland Museum 40:1-626. HOOPER, J.N.A. & BERGQUIST, P.R. 1992. New species and genus of lamellate coral reef sponges (Demospongiae: Axinellida: Axinellidae). Memoirs of the Queensland Museum 32(1): 99-137. HOOPER, J.N.A., CAPON, R.J., KEENAN, C.P. & PARRY, D.L. 1990. Biochemical and morphometric differentiation of two sympatric sibling species of Clathria (Porifera: Demospongiae) from northern Australia, Invertebrate Taxonomy 4(1): 123-148. HOOPER, J.N.A., CAPON, R.J., KEENAN, C.P., PARRY, D.L. & SMIT, N. 1992. Chemotaxonomy of marine sponges: families Microcionidae, Raspailiidae, and Axinellidae, and their relationships with other families in the orders Poecilosclerida and Axinellida (Porifera: Demospongiae). Invertebrate Taxonomy 6(2): 261-301. HOOPER, J.N.A. & LEVI, C. 1994. Biogeography of Indo-west Pacific sponges: Microcionidae, Raspailiidae, Axinellidae. Pp. 191-212. In Soest, R.W.M. Van, Kempen, T.M.G. van & Braekman, MEMOIRS OF THE QUEENSLAND MUSEUM J.-C. (eds) Sponges in Time and Space. (Balkema: Rotterdam). HOOPER, J.N.A. & WIEDENMAYER, F. 1994. Porifera. Pp. 1-620. In Wells, A. (ed.) Zoological Catalogue of Australia, Volume 12. (CSIRO Australia: Melbourne). HOOPER, J.N.A., LIST-ARMITAGE, S.E., KENNEDY, J.A. COOK, S.D. & VALENTINE, C.A. 1999. Sponges of the Low Isles, Great Barrier Reef: an important scientific site, or a case of mistaken identity ? Memoirs ofthe Queensland Museum (this volume). KELLY-BORGES, M., ROBINSON, E.V., GUNASEKERA, S.P., GUNASEKERA, M., GULAVITA, N.K. & POMPONI, S.A. 1994, Species differentiation in the marine sponge genus Discodermia (Demospongiae: LIthistida): the utility of ethanol extract profiles as species-specific chemotaxonomic markers. Biochemical Systematics and Ecology 22(4): 353-365. KELLY-BORGES, M. & VACELET, J. 1995. A revision of Diacarnus Burton and Negombata de Laubenfels (Demospongiae: Latrunculiidae) with descriptions of new species from the west central Pacific and Red Sea. Memoirs of the Queensland Museum 38: 477-503. KERR, R.G. & KELLY-BORGES, M. 1994. Bio- chemical and morphological heterogeneity in the Caribbean sponge Xestospongia muta (Petrosida: Petrosiidae). Pp. 65-74. In Soest, R. W.M. Van, Kempen, T.M.G. van & Braekman, J.-C. (eds) Sponges in Time and Space. (Balkema: Rotterdam). KIESCHNICK, O. 1900. Kieselschwámme von Amboina. In Semon, R. (ed.) Zoologische Forschungsreisen in Australien und dem Malayischen archipel ... ausgeführt in den Jahren 1891-1893. Vol. 5(5). Denkschriften Medizinisch-Naturwissenchaftliche Gesellschaft, Jena 8: 545-582. KLAUTAU, M., SOLÉ-CAVA, A.M. & BOROJEVIC, R. 1994, Biochemical systematics of sibling sympatric species of Clathrina (Porifera: Calcarea) Biochemical Systematics and Ecology 22(4): 367-375. KREBS, C.J. 1978. Ecology. The experimental analysis of distribution and abundance. (Harper & Row: New York, Hagerstown, San Francisco, London). LAUBENFELS, M.W. DE 1954. The sponges of the West-Central Pacific. Oregon State Monographs, Zoology 7: 1-306. LENDENFELD, R. VON 1883. Über Coelenteraten der Südsee. Neue Aplysinidae. Zeitschrift fuer Wissenschaftliche Zoologie 38: 234-313. 1884. A monograph ofthe Australian sponges. Part 1. Introduction. Part 2. Morphology and physiology of the sponges. Proceedings of the Linnean Society of New South Wales 9: 121-154, 310-344. BIODIVERSITY OF NORTHEASTERN AUSTRALIAN SPONGES 271 1885a, A monograph of the Australian sponges. Part 5. Proceedings uf Ihe Linnean Society of New South Wales 10; 282-325, 1865b. A monograph of the Australian sponges. Part 6. Proceedings of the Linnean Society of New Sauth Wales 10: 481-553. 1887. Die chalineen des australischen gebietes. Zoologische Jahrbücher, Jena 2: 723-828, 1888. Descriptive Catalogue of the Sponges in Ihe Australian Museum, Sydney. (Taylor & Francis: London). 1889. A Monograph of the Horny Sponges. . (Tribner & Cu.: London). LEVI, €. 1967. Démosponges recoltées en Nouvelle- Calédonie par la mission Singer-Polignac. Expedition Francaise sur les récifs coralliens de la Nouvelle Calédonie, Fondation Singer-Polignac, Paris 2: 13-28. MADDISON, W.P & MADDISON, D.R, 1992, MacClade. Version 3.0 (Sinauer Associates, Inc. Publishers: Sunderland, Massachusetts). POLEJAEFF, N: 18844. Report on the Calcarea dredged by H.M.S. “Challenger” during the years 1873-1876. In Report on the Scientific Results of the Voyage of H,M.S. ‘Challenger’ during the years 1873-1876. 8(23): 1-76. (Her Majesty's Stationery Office: London, Edinburgh, Dublin). 18S4b, Report on the Keratosa collected by H.M.S. “Challenger” during the years 1873-1876, Jn Report on the Scientific Results ofthe Voyage of H.M.S. *Challeneer” during the years 1873-1876, 11; 1-88 (Her Majesty's Stationery Office: London, Edinburgh, Dublin). PULITZER-FINALI, G. 1982. Some new or little- known sponges from the Great Barrier Reef of Australia. Bolletino dei Musei e degli Istituti Biologici dell'Universilà di Genova 48-49(1980- | 981): 87-141. RIDLEY, 5.0. 1884. Spongiida. Pp. 366-482. 582-630. pls 39-43, 53-54. In Report on the Zoological Collections made in the Indo-Pacific Ocean during the Voyage of H.M.S. ‘Alert’ 1881-2. (British Museum (Natural History): London), RIDLEY, 8.0. & DENDY., A. 1887. Report on the Monaxonida collected by H.M,S. “Challenger” during the Years 1873-76. In Report on the Scientific Results of the Voyage of H.M.S. "Challenger? during the Years 1873-76. 20; 1-275 (Her Majesty's Stationery Office; London, Edinburgh, Dublin). ROHDE, K. 1979. A critical evaluation of intrinsic and extrinsic factors responsible for niche restriction in parasites. American Naturalist 114: 648-671. SARA, M. 1990. Australian Tethya (Porifera, Demo- spongiae) from the Great Barrier Reef with description of two new species. Bolletino di Zoologia 57: 153-157. SCHULZ, E. 1900. Die Hornschwiimme von Thursday Island und Amboina. In ‘Semon, R. (ed.) Zoologische Forschungsreisen in Australien und dem Malayischen Archipel ... ausgeführt in den Jahren 1891-1893, Volume §(5), Jenaische Denkschriften 8: 325-344. SOEST, R.W.M. VAN, HOOPER, LN.A, & HIEMSTRA, E. 1991. Taxonomy. phylogeny ana biogeography of the marine sponge genus Hearnus (Porifera: Poecilosclerida). Beaufortia 423): 49-88. SOEST, R.W.M VAN & HOOPER, J,N.A. 1994. Taxonomy, phylogeny and biogeography of the marine sponge genus Rhubderemia (Porifera: Poecilosclerida): Scientia Marina 57(4); 319-351. SOEST, R.W.M. VAN, DESQUEYROUX- FAUNDEZ, R, WRIGHT, A.D. € KOENIG, GM, 1996. Cymbastela hooperi sp,nov. (Halichondrida: Axinellidac) from the Great Barrier Reef, Australia. Bulletin de l'Institut Royal des Sciences Naturelles de Belgique 66(Supplement): 103-108. SOLE-CAVA, A.M. & THORPE, J.P. 1986. Genetic differentiation between morphotypes of the Marine Sponge Suberites ficus (Demospongiae: Hadromerida), Marine Biolugy 93(2): 247-254. 1994, Evolutionary genetics of marine sponges. Pp. 55-64, In Soest, R, W,M. Van, Kempen. T.M.G. van & Braekman, J.-C, (eds) Sponges in lime and Space. (Balkema: Rotterdam), SOLE-CAVA, AM. KLAUTAU, M. BOURY-ESNAULT, N, BOROJEVIC, R. & THORPE, J.P. 1991. Genetic evicence for eryplie speciation in allopatric populations of two cosmopolitan species of the calcareous sponge genus Clathrina, Marine Biology 111: 381-386. SOLE-CAVA, A.M., BOURY-ESNAULT, N., VACELET, J. € THORPE, J.P. 1992, Bjo- chemical genetic divergence and systematics in sponges of the genera Corticium and Oscarella (Demospongiae: Homoscleromorpha) in the Mediterranean Sea, Marine Biology 113: 209-3014. SOLLAS, W.J. 1888. Report on the Tetractinellida collected by H.M.S. “Challenger” durmg the years 1873-76, In Report on the Scientific Results of the Voyage of H,M,S, ‘Challenger’ during the years 1873-76. 25(63): 1-458 (Her Majesty's Stationery Office: London, Edinburgh, Dublin). SWOFFORD, D.L. 1993. PAUP (Phylogenetic Analysis Using Parsimony). Version 3.1.1 fur Apple Macintosh (Smithsonian Institution: Washington). THIELE, J, 1898, Studien uber pazilische Spongien. I. Zoologica 24: 1-72, 1900, Kieselschwámme von Ternate. |, Abhand- lungen der Senkenbergischen Naturforschenden Gesellschaft 25: 19-80. 1903, Kieselschwiimme von Temate. IT. Abhand- lungen der Senkenbergischen Naturforschenden Gesellschaft 25: 933-968. THOMPSON, J E., MURPHY, PL, BERGQUIST, PR. & EVANS. E.A. 1987. Environmentally induced variation in diterpene composition of the marine 274 sponge Rhopaloeides odorabile. Biochemical Systematics and Ecology 15(5): 595-606. TOPSENT, E. 1897, Spongiaires de la Baie d’ Amboine. Voyage de MM M. Bedot et C. Pictet dans |’ Archipel Malais. Revue Suisse de Zoologie 4: 421-487. WESCHE, S.J., ADLARD, R.D. & HOOPER, J.N.A. 1997. The first incidence of clionid sponges (Porifera) from the Sydney rock oyster Saccostrea commercialis (Iredale and Roughley, 1933). Aquaculture 157: 173-180. MEMOIRS OF THE QUEENSLAND MUSEUM WHITELEGGE, T. 1889. List of the marine and freshwater invertebrate fauna of Port Jackson and the neighbourhood. Journal of the Royal Society of New South Wales 23(2): 163-323. WILKINSON, C.R. 1978. Description of two Demospongiae, one being toxic from the Great Barrier Reef. Tethys 8(3): 267-270. WOERHEIDE, G. 1997. The reef cave dwelling coralline demosponge Astrosclera willeyana Lister 1900 from the Indo-Pacific. (PhD thesis, Universitaat Góttingen: Góttingen). MORPHOLOGY AND MOLECULES IN LITHISTID TAXONOMY: NEW SOLUTIONS FOR OLD PROBLEMS. Memoirs of the Queensland Museum 44; 274. 1999:- Most lithistid sponges lack an adequate range of taxonomic characters for differentiation, and in most genera these characters are extremely plastic. Consequently, the generation of morphological hypotheses in comparison with molecular phylogenies is nearly impossible due to the absence of reliable synapomorphies. Historically, lithistids have been grouped together in a single order on the basis of common possession of an interlocking siliceous skeleton. Recent morphological and palaeontological data indicate, however, that lithistid sponges are polyphyletic; several genera possess skeletal characters that suggest affinity with non-lithistid demosponges. We have found that in many cases these characters are probably non-homologous and misleading. Ongoing research on the phylogeny of lithistid sponges has revealed some interesting ‘anomalies’ of identification. Although our data collection is still incomplete, we have already found unexpected phylogenetic affinities between three lithistid species in Theonellidae and Corallistidae, comparing morphological and 28S rDNA analyses. Surprisingly, the nearest relatives of de Laubenfel’s (1954) *Plakinalopha” mirabilis are Theonella spp.; Theonella atlantica is more closely related to Corallistes spp. than to Theonella spp.; and Theonella tubulata Van Soest is more closely related to Macandrewia azorica (in the Corallistidae) than to other Theonella. What is to be done in this situation ? To what extent can molecular hypotheses be accepted over morphological hypotheses or vice versa ? We have found that rather than having to ‘accept’ one over the other, which often goes against ‘instinctual phylogeny’, molecular data makes us re-examine these problems by reciprocal illumination, through the generation of higher quality morphological research and the examination of characters that are often, not at first, obvious. With this group of lithistid sponges, triaene rhabd and clade morphology, microsclere ornamentation, and the patterns of desma zygoses, and shaft ornamentation become crucially important in differentiating taxa. Thus, for this particular group of organisms, we have found that morphological hypotheses between closely related taxa are often strongly informative and can lend crucial evidence for the acceptance of certain molecular phylogenies over others. Molecular data can clearly indicate relationships between organisms where morphological data had previously failed, and molecular data often require us to re-examine morphological characters from new perspectives, leading the discovery of new taxonomic discriminators. CJ Porifera, phylogeny, 28S rDNA, morphology, congruence, lithistid, Theonellidae, Corallistidae. Michelle Kelly* (email: m.kelly@niwa.cri.nz), Grace P. McCormack & James O. McInerney, Zoology Department, Natural History Museum, Cromwell Road, London; * Present address: Marine Ecology and Aquaculture, National Institute of Water & Atmospheric Research (NIWA), Private Bag 109-695, Newmarket, Auckland, New Zealand; 1 June 1998. PHYLOGENETIC RELATIONSHIPS BETWEEN LUBOMIRSKIIDAE, SPONGILLIDAE AND SOME MARINE SPONGES ACCORDING PARTIAL SEQUENCES OF 185 rDNA VALERIA B. ITSKOVICH, SERGEY I. BELIKOV, SOFIA M. EFREMOVA AND YOSHIKI MASUDA Itskovich, V.B., Belikov, S.I., Efremova, S.M. & Masuda, Y. 1999 06 30: Phylogenetic relationships between Lubomirskiidae, Spongillidae and some marine sponges according partial sequences of 18S rDNA. Memoirs of the Queensland Museum 44: 275-280. Brisbane. ISSN 0079-8835. Two families of Porifera are represented in Lake Baikal, Russia: cosmopolitan Spongillidae and endemic Lubomirskiidae. Systematics and phylogeny of Lubomirskiidae are still poorly known. Indeed, there is little agreement on the origin of freshwater sponges in general, and this group is considered to be polyphyletic. Latest morphological and embryological data indicate that Lubomirskiidae and Spongillidae are closely related. Using molecular data we explored the possible origins of Lubomirskiidae and determined the closest relatives of Spongillidae and Lubomirskiidae among marine sponges. Partial sequences of 18S rDNA for Halichondria japonica, Lubomirskia abietina, Swartschewskia papyracea, Spongilla lacustris and Ephydatia muelleri were compared with available sequences of 188 rDNA of other Porifera from the GenBank. Parsimony and neighbour-joining analyses gave trees of similar topology. Molecular data were in accordance with the notion of close relationships of endemic and cosmopolitan families. Some marine sponge families are assumed to be related to freshwater sponges.) Porifera, Lake Baikal, Spongillidae, Lubomirskiidae, 18S rRNA, phylogeny, freshwater sponges. Valeria B. Itskovich & Sergey I. Belikov (email: belikov@lin.irk.ru) Limnological Institute of the Siberian Branch of RAS, Irkutsk, Russia; Sofia M. Efremova, Biological Institute of St. Petersburg University, St. Petersburg, Russia; Yoshiki Masuda, Department of Biology, Kawasaki Medical School, Okayama, Japan; 1 March 1999. Three families of Porifera inhabit freshwater: Spongillidae, Lubomirskiidae and Potamolepi- dae. The problem of the origin of endemic and cosmopolitan freshwater sponges, their relation- ships with each other and with marine sponges, repeatedly attract the attention of scientists. Marshall (1885) suggested freshwater sponges were polyphyletic, with Renieridae (Haplo- sclerida) possibly being their closest marine relative. The idea of a polyphyletic origin for freshwater sponges was subsequently discussed and emphasised by many authors. In describing the genus Sterastrolepis, believed to be a Neo- tropical representative of Potamolepidae, Volkmer-Ribeiro & De Rosa-Barbosa (1978) noted that the characteristics ofits gemmoscleres were too different to assign this family to Haplo- sclerida. On the basis of gemmule structure, gemmosclere and skeleton peculiarities, they confirm Briens’ (1970) assumption about the close relationship of Potamolepidae with Hadro- merida. They also favour the hypothesis of a passive mechanism of invasion into freshwater habitats by marine sponges, noting that endemic freshwater genera (e.g. Ochridaspongia, Pachy- dictium, Lubomirskia) have been recorded from ancient lakes, remnants from past sea levels, but not from estuaries. Evidence for a hadromerid origin of some freshwater sponges (Volkmer- Ribeiro & Watanabe, 1983) is also provided by the Japanese sponge Sanidastra yokotonensis. Volkmer-Ribeiro (1990) also hypothesised that the Neotropical genus Metania may be related to the marine poecilosclerid genus Acarnus. Conversely, Racek & Harrison (1975), using palaeontologic data, suggest that Spongillidae was monophyletic having evolved from Radio- spongilla stock. The endemic family Lubomirskiidae, inhabit- ing Lake Baikal, has approximately 10 species belonging to 3 genera: Lubomirskia, Baikalo- spongia and Swartschewskia (Rezvoi, 1936). At present the systematics and phylogeny of this family is still poorly known. The history of study on the origin of Lubomirskiidae shows a number of contrary opinions. Dybowsky (1882), Swartschewsky (1902), Annandale (1913) and Rezvoi (1936) believed Lubomirskiidae was closely related to marine sponges and not to Spongillidae, owing to their considerable morph- ological differences. Later palaeontological studies (Martinson, 1940) hypothesised that 276 Lubomirskiidae were representatives of the mezolimnological fauna, originating much later than the usual palaeolimnological fauna to which Spongillidae belongs. In contrast to these beliefs, the latest comparative morphological data indicate a close relationship between Spongillidae and Lubomirskiidae (Efremova, 1981), supported by data on their loss of sexual reproduction by gemmules as an adaptive feature (Efremova, 1994). To solve contradictions in the systematic and phylogenetic in- terpretation of morphological data rDNA analysis is now widely used (e.g. Christen et al., 1991; Halanych, 1991). Although this method has been succesfully used for some marine sponge families (Lafay et al., 1992; West & Powers, 1993; Kelly-Borges & Pomponi, 1994) there are no prev- ious studies on molecular phylogeny of freshwater sponges. In this study we apply partial 188 rDNA sequence analysis, firstly to explore the origin MEMOIRS OF THE QUEENSLAND MUSEUM TABLE 1. Classification of the species used in this study. Classification oe mber References CNIDARIA: ANTHOZOA parara inea ihid) U42453 Cavalier-Smith, 1996 PORIFERA: DEMOSPONGIAE anette Aedncllidae) U43190 — | Cavalier-Smith, 1996 (Spiro eae Tetillidae) D15067 Kobayashi et al., 1993 (Poeciioscierida Microcionidae) L10825 Wainright, 1993 Halichondria Loree ondridae) AF058946 this study (aplosclerida: Lubomirskiidae) AF058947 this study a | ams | Mesa (Haplosclerida: Spongillidae) AF058949 this study 1 gposclerida: Spongillidae) AF058945 this study PORIFERA: CALCAREA ree Glathrinda: Clathrinidae) 142432 Cavalier-Smith, 1996 (ce ha Suis Syoettida: Sycettidae) L10827 Wainright, 1993 te peso Succttida: Sycettidae) D15066 Kobayashi et al., 1993 of Lubomirskiidae, and secondly to obtain new data on the origin of fresh- water sponges in general. MATERIALS AND METHODS Specimens of Lubomirskia abietina, Swarts- chewskia papyracea, Spongilla lacustris and Ephydatia muelleri were collected from Lake Baikal (Russia) and specimens of Halichondria japonica were collected from Desaki seashore (Japan) by SCUBA diving in depths between 0.5-13.5m. All specimens were photographed alive. Data on ecology, habitat and texture were recorded. Part of each sample was fixed in 70% ethanol for taxonomic identification, another part was frozen in liquid nitrogen for molecular analysis. Total genomic DNA extraction was performed with standard phenol method (Sambrook et al., 1989) and with CTAB method (Gustincich et al., 1991). PCR primer design was performed by alignment of Porifera 188 rRNA sequences available from GenBank (see Table 1). As sponges harbour a large number of symbionts, in addition to universal primers, sponge-specific primers were also designed. The primers correspond to the V4 and V5 regions of 18S rRNA: RI (5’-TAAAAAGCTCGTAGTTGGAT-3’; forward, universal, correspond to positions 629-647 in Axinella polypoides 188 rRNA (GenBank accession number U43190)); L1 (5”- GGACTACGACGGTATCTGAT-3”; reverse, universal (1008-1026)); R2 (5”-GTAGTGGC CTACCATGGTTGC-3’; forward, sponge-specific (342-361)); L2 (5”-CTAATTTTTTCAAAG TAAACGTCCCGA-3”; reverse, sponge-specitic (749- 777)). The primers were synthesised by H-phos- phonate method. Two overlapping fragments of the 18S rRNA gene (400bp each) were amplified. A 25ul PCR reaction mix contained 2.5ul of 10xPCR Buffer (Promega), 3ul of MgCl; (25mM), 0.5ul of each primer (10pmol/uD), 1101 of dNTP mix (100mM each), lul of DNA (~0, 11g), 0.211 of Taq DNA polymerase, 2511 of ddH,0. Cycle parameters were: initial denatur- ation at 94°C for 120secs, followed by 40 cycles of denaturation at 94°C for 60secs, anneling at 45?C for 60secs, and extension at 72?C for 60secs, followed by a final extension of 8mins at 72°C. About 6 tubes of each PCR reaction were purified by electrophoresis in low melting agarose. PCR fragment purification was carried out twice with equal volume of phenol, followed by precipitation by 2 volume of ethanol and 0.1 volume of 10M ammonium acetate and washing in 70% ethanol (Sambrook et al., 1989). PCR FRESHWATER SPONGE PHYLOGENY 277 Lubomirskia CGGGTGACGGAGAATTGGGGTTCGATTCCGGAGAGGGAGCCTGAGARACGGCTACCAC Svartsheyskia ....... SELECT LIST TSE TEE CT A RARA ee) Spongilla m ATRAER EbUhydétib (e th th I "T"———— —— Halichondria ...... b atei coe phew cane pba seeders bares dro ee ee ee ee Lubomirskia Suartshewskia Spongilla Ephydatia Halichondria Lubomirskia Swartshewskia . Spangilla Ephydatia Halichondria Lubomirskia SWATTSHEUSKIA «cap eect tees eee seesaw ee seneas wes ae cone Cielo Spongilla ........ AAA TERIS A AR Ephydatia eee es bib ete ea aes 23. Sate dre na PERPE: 4. es^ a ed azote 3/5 i ó in jee Halichondria ...s pees sees eeeee TE ira re ws bee Lubomirskia AGCGTATATTARAAGTTGTTGCAGTTALAAAGCTCGTAGTTGGATTTCGGGGCAGGAGG Swartshewskia ...ooooooocoomrocorcrrso rocas ¿naar Meri Spongilla ....... ooosooonoro».s — tee AA A do $9397 eee ae be eae © Ephydatia — ...onoo ooo... error cos O M PES M Halichondria ............ TO eio a Slo pne ME nl TG.CCT. Lubomirskia . CGGTCCGCCGAAAGGTAGGTACTGGACGCCAGCCCTTTTTCTCGAARGGCCCCATCTGC Swartshewskia Spongilla = T...i..e cree Ceca eher nn n Been TT Ephydatia veres canes ern se saa e Halichondria DI" .... A Wee Cal AAA e AR À... GÀ... Lubomirskia TTCACTG-AGTGGTAGGGGAGTTCGGGACGTTTACTTTGAAAAAATTAGAGTGTTCAA Swartshewskia ...... O alii roe erste jo des A 5.5 LA ee iA "Pm +. Spongilla Ephydatia Halichondria Lubomirskia Swartshewskia ........ IEEE rr rar eso e Spongilla / ........-. OD DID concerns mo. o ooaoonoos vertes Ephydatia eo soedustboewstrbésbeeveróhtersese eee ee ee cero Halichondria ..... A AS PA ee ee ee A PT Lubomirskia TTGGTTTCTGGGACCGAAGTAATGATTAAGAGGGACAGTTGGGGGCATTCGTATICAA Swartshewskia ........... TIL Ves sed. TOP d. rese ot Spongilla ....... "RD PLA Ida SITRA Eee $4 a eA apodere a Ephydatia "POI Cee naw ne pear nee PA OS pam y ta cv o "Tv Halichondria ......... À..G.. ccce eee ente T "m + TT- Lubomirskia GTCAGAGGTGAAATTCTTGGATTTATGGAAGACGAACAACTGCGARAGCATTTGCCAA Swartsheuskia ...... ce ern ÜraeeRerkb eA eer savas Pee rn oo. Spongilla bay eais pojo "T. Pra pra rra ateos doses Ephydatia E IEEE PEE O Erre Ar sees " Halichondria --............ $ osa duele EUN Ai gaip yyri TiGisssseesaver vereers Lubomirskia ATGTTTTCATT Svartshevskia ..... oe sees Spongilla cierres Ephydatia — Halichondria ........... FIG. 1. Alignment of partial 18S rDNA sequences (630 bp) obtained. Only nucleotides that differ from those of Lubomirskia abietina are indicated (identities are denoted by points and deletions by hyphens). GenBank accession numbers are: Halichondria japonica AF058946, Lubomirskia abietina AF058947, Swartschewskia papyracea AF058948, Spongilla lacustris AF058945 and Ephydatia muelleri AF058949. fragments were sequenced on both strands using fmol DNA sequencing System (Promega) according to the published protocol. Cycle parameters were: initial denaturation at 95°C for 120secs, followed by 30 cycles of denaturation at 95°C for 30secs, anneling at 42°C for 30secs, and extension at 70°C for 60 secs. The struc- tures obtained were aligned manually with the help of the GeneTools package (Resenchuk, 1991). Neighbour-joining an- alysis was derived using Treecon for Windows (Van de Peer, 1994). The distance estimation was carried out using the formula of Kimura (1980). Bootstrap values were calculated from 100 replicates. Parazoanthus axinellae was used as the outgroup. Programs SEQBOOT, DNAPARS and CONSENSE of PHYLIP 3.5c package (Felsenstein, 1995) were used to construct maximum parsimony trees. Bootstrap analyses with 100 replications were carried out. RESULTS AND DISCUSSION We obtained partial 18S rRNA gene sequences (~630bp) for five species of Porifera. GenBank accession numbers are as follows: Halichondria japonica (AF058946), Lubomirskia abietina (AF058947), Swartschewskia papyracea (AF058948), Spongilla lacustris (AF058945) and Ephydatia muelleri (AF058949). Two specimens of each species were used to obtain sequences. All structures were aligned successfully, and common length of alignment was 630bp (Fig. 1). There are a few nucleotide differences between 18S rDNA structures obtained for freshwater sponges compared to those from marine sponges. Sequences from the marine sponge H. japonica have many more transitions/ transversions events, and insertion/deletion events were observed only this species. Lubomirskia and Ephydatia show no nucleotide differences in their 18S rDNA sequences, indicating a very high level of genetic relationships between them. To study the molecular relationships between freshwater and marine sponges, sequences from MEMOIRS OF THE QUEENSLAND MUSEUM Spongilla Microciona Axinella Halichondria Tetilla Clathrina Scypha cal. Scypha cll. Parazoanthus FIG.2. Phylogenetic relationships ofthe Lubomirskiidae, Spongillidae and other Porifera based on neighbour-joining analysis of 18S rDNA (630bp). Bootstrap percentages are shown at the nodes for 100 resamplings. Parazoanthus axinellae used as the outgroup. other Porifera available from GenBank (see Table 1) were included in the alignment. Figure 2 shows a tree obtained by neighbour- joining analysis with Parazoanthus axinellae as the outgroup. High bootstrap values show that all clusters are statistically significant. Spongilla, Lubomirskia, Swartschewskia and Ephydatia form a common clade. A sister branch formed by Axinella and Microciona is the most closely situated to this clade. Parsimony analysis, performed on the basis of these sequences, provides a similar topology (not shown here). These data confirm that freshwater sponge genera form a closely related group and, except for Axinella and Microciona, Halichondria and Tetilla, also refer to the neighbouring cluster. These molecular data are in accordance with the notion of a close relationship between endemic and cosmopolitan families. They do not support the idea that Lubomirskiidae has an independent origin from Spongillidae. These data also suggest that the assumption of Racek & Harrison (1975), that endemic genera in the ancient lakes appeared independently of the cosmopolitan fauna, is invalid as far as Baikalian Lubomirskiidae is concerned. Branch length shows that divergence of Lubo- mirskiidae and Spongillidae took place much laterthan divergence oftheir common ancestor. It FRESHWATER SPONGE PHYLOGENY provides support for Efremova (1981) that Lubomirskiidae is not a relic fauna, but a flourishing group of Lake Baikal organisms. This also confirms Talievs’ (1955) opinion about the relatively fast evolution of the Lake Baikal fauna. It will be interesting to check this assumption using palaeontological studies of sponge spicules in the bottom sediments of Lake Baikal. It is possible that the scenario of Baikalian sponge fauna formation is similar to that of the Baikalian Turbellaria, which is closely related to cosmopolitan species (Timoshkin, 1995). Thus, although a part of Lake Baikal fauna really has marine origin, Baikalian sponges have a typical freshwater origin. However, as the evolution of animal 18S rDNA is non-clock-like, it is advisable to conduct investigations into the cytochrome oxidase genes whose sequences are not yet available for Porifera. This study would allow estimates to be made of divergence times between Lubo- mirskiidae and Spongillidae. Our tree also demonstrated an earlier divergence of Spongilla from the common branch of freshwater sponges. However, the few freshwater genera yet analysed, and insufficient variability of 18S rDNA, does not yet provide any unequivocal support to hypothesise relationships between certain freshwater genera. To study relationships between closely related freshwater genera, we need data from more variable regions of the gene. Work on internal transcribed spacers (ITS1 and ITS2) is currently in progress. Trochospongilla is likely to be a possible direct ancestor of Lubomirskiidae. This genus has no microscleres, and spicules have maximal mutability. Accord- ing to preliminary data, Axinella, Microciona, Halichondria and Tetilla are the most closely related to the present freshwater sponges. It is probable, however, that obtaining new data on the other marine sponge sequences, for example other Haplosclerida, will substantially change the scheme presented here. ACKNOWLEDGEMENTS This work was partly supported by Russian Foundation for Basic Research, grant # 97-04-96172. LITERATURE CITED ANNANDALE, N. 1913. Notes on some sponges from Lake Baikal in the collection of the Imperial Academy of Sciences, St. Petersburg. Annual Museum Zoology Academy Sciences of St. Peterburg 18: 18-101. 279 BRIEN, P. 1970, Les Potamolepides africaines nouvelles du Luapula et du Lac Moero. Pp. 163-186. In Fry, W.G. (ed.) Biology of the Porifera. Symposia of the Zoological Society of London. Number 25. (Academic Press: London). CAVALIER-SMITH, T., ALLSOPP, M.T.E.P., CHAO, E.E., BOURY-ESNAULT, N. & VACELET, J. 1996. Sponge phylogeny, animal and nervous system monophyly: 18S rRNA evidence. Canadian Journal of Zoology 74: 2031-2045. CHRISTEN, R., RATTO, A., BAROIN, A., PERASSO, R., GRELL, K.G. & ADOUTTE, A. 1991. An analysis of the origin of metazoans, using comparisons of partial sequences of the 28S ribosomal RNA, reveals an early emergence of triploblasts. European Molecular Biology Organization Journal 10: 499-503. DYBOWSKY, W. 1882. Studien uber die Spongien des Russischen Reiches. Memoirs of the Academy Imperial Sciences of St, Peterburg 7:1-71. EFREMOVA, S.M. 1981. The morphology and the development ofthe baikalian sponge Lubomirskia baicalensis and the phylogenetic connections of Lubomirskiidae with the other sponges. Pp. 93-107. In Korotkova, G.P. (ed.) Morphogenesis in sponges. Contributions of Biological Institute. Number 33. (in Russ.). 1994. The evolutionary pathways of baikalian sponges. P. 15. In Baikal as a Natural Laboratory for Global Change. INTAS-RAS SB Workshop, May 11-17, 1994. Abstracts. Vol. 5 (Irkutsk: Russia). FELSENSTAIN, J. 1995. PHYLIP (Phylogeny Inference Package). Version 3.5c. (University of Washington: Washington). GUSTINCICH, S., MANFIOLETTI, G., DEL SAL, G., SCHNEIDER, C. & CARNINCY, P. 1991. A fast method for high-quality genomic DNA extraction from whole human blood. Biotechniques 11: 298-300, HALANYCH, K.M. 1991. 5S ribosomal RNA sequences inappropriate for phylogenetic reconstruction. Molecular Biology and Evolution 8: 249-253, KELLY-BORGES, M. & POMPONI, S. 1994. Phylogeny and classification of lithistid sponges (Porifera, Demospongiae): a preliminary assessment using ribosomal DNA sequence comparisons. Molecular Marine Biology and Biotechnology 3: 87-103. KIMURA, M. 1980. A simple method for estimating evolutionary rate of base substitutions through comparative studies of nucleotide sequences. Journal of Molecular Evolution 16:111-120. KOBAYASHI, M., TAKAHASHI, M., WADA, H. & SATOH, N. 1993. Molecular phylogeny inferred from sequences of small subunit ribosomal DNA, supports the monophyly of the Metazoa. Zoological Science 10(5): 827-833. LAFAY, B., BOURY-ESNAULT, N., VACELET, J. & CHRISTEN R. 1992. An analysis of partial 28S ribosomal RNA sequences suggests early radiations of sponges. Biosystems 28: 139-151. MARSHALL, W. 1885. On some new siliceous sponges collected by M. Pechuel-Losche in the Kongo. Annals and Magazine of Natural History 12: 391-412. MARTINSON, G.G. 1940, Materials for the study of the fossil fauna of the Baikal area. Trudy Baikalskoi Limnolicheskoi Stantsii Academy of Sciences of the USSR 10: 425-451. (in Russ). RACEK, A.A. & HARRISON, F.W. 1975. The systematic and phylogenetic position of Palaespongilla chubutensis (Porifera: Spongillidae). Proceedings of the Linnean Society of New South Wales 99: 157-165. RESENCHUK, S.M. 1991. Gene Tools (Preparation and processing of text files containing the nucleic or amino acids sequences) Version 1.0. (Published by the author). REZVOI, P.D. 1936. Freshwater sponges of the USSR. Pp. 1-42. In Rezvoi, P.D. (ed.) The fauna of the USSR. Vol. 2 (Academy of Sciences ofthe USSR: Moskow). (in Russ.). SAMBROOK, J., FRITSCH, E.F. & MANIATIS, T. 1989, Molecular cloning: a laboratory manual. (Cold Spring Harbor: New York). SWARTSCHEWSKY, B. 1902. Materials on the sponge fauna of Lake Baikal. Memoirs of the Society of Naturalist of Kiew 17; 329-352, (In Russ.). TALIEV, D.N. 1955. Bullheads of Baikal (Cottoidei). (Academy of Sciences of the USSR: Moskow). (in Russ.) . TIMOSHKIN, O.A. 1995. Biodiversity of Baikal fauna: review of the state-of-the-art and prospects for MEMOIRS OF THE QUEENSLAND MUSEUM research. Pp, 25-52. In Timoshkin, O.A. (ed.) Guide and key to pelagic animals of Baikal. (‘Nauka’: Novosibirsk). (in Russ.). VAN DE PEER, Y. & DE WACHTER, R. 1994, TREECON for Windows: a software package for the construction and drawing ofevolutionary trees for the Microsoft Windows environment. Version 1.2. (University of Antwerp: Antwerp). VOLKMER-RIBEIRO, C, & DE ROSA-BARBOSA, R. 1978. Neotropical freshwater sponges of the family Potamolepidae Brien , 1967. Pp. 503-511. In Levi, C. & Boury Esnault, N. (eds) Biologie des Spongiaires. Colloques Internationaux du Centre National de la Recherche Scientifique. Number 291. (CNRS: Paris). VOLKMER-RIBEIRO, C. & WATANABE, Y. 1983. Sanidastra yokotonensis, n. gen. and n. sp. of freshwater Sponge from Japan. Bulletin of the National Science Museum Tokyo, Zoology 9: 151-159. VOLKMER-RIBEIRO, C. 1990. A new insight into the systematies, evolution and taxonomy of fresh- water sponges. Pp. 323-331. In Rützler, K. (ed.) New perspectives in sponge biology. (Smithsonian Institution Press: Washington DC). WAINRIGHT, P.O., HINKLE, G., SOGIN, M.L. & STICKEL, S.K. 1993. The monophyletic origins of the metazoa; an unexpected evolutionary link with fungi. Science 16, 260(5106): 340-342. WEST, L. & POWERS, D. 1993. Molecular phylo- genetic position of hexactinellid sponges in relation to Protista and Demospongiae. Molecular Marine Biology and Biotechnology 2: 71-75. CYMBASTELA HOOPERI AND AMPHIMEDON TERPENENSIS: WHERE DO THEY REALLY BELONG? GABRIELE M. KONIG AND ANTHONY D. WRIGHT König, G.M. & Wright, A.D. 1999 06 30: Cymbastela hooperi and Amphimedon terpenensis: where do they really belong? Memoirs ofthe Queensland Museum 44: 281-288. Brisbane. ISSN 0079-8835. A sponge sample identified as Cymbastela hooperi collected from Kelso Reef, the Great Barrier Reef, Australia, yielded a series of natural products, mainly diterpene isonitriles, which demonstrated significant in vitro antimalarial activity. As a result ofthese compounds being consumed in a number of bioassays it was considered desirable to have more of them so as to enable further and more detailed biological testing to be undertaken. Subsequently three Cymbastela samples (two of C. concentrica and one of C. coralliophila) and two of Amphimedon (Cymbastela) terpenensis were tested for antimalarial activity and investigated for their natural product content. The results of these investigations provided further evidence that either, Amphimedon terpenensis is more appropriately Cymbastela terpenensis, or that both C. hooperi and A. terpenensis belong to an as yet undefined genus and may perhaps be the same species. O Porifera, Cymbastela, Amphimedon, Acanthella, biological testing, malaria, cytotoxicity, taxonomy, Great Barrier Reef, diterpene isonitriles, marine natural products. Gabriele M. König € Anthony D. Wright (email: a.wright@tu-bs.de), Institute for Pharmaceutical Biology, Technical University Braunschweig, Mendelssohnstrasse 1, D 38106 Braunschweig, Germany; 2 March 1999. Since the discovery by Angerhofer et al. (19922) of the antimalarial activity of axisonitrile-3 (Fig. 1A) isolated from the sponge Acanthella klethra, much of our research activity has focused on finding further marine natural products with this biological activity. These efforts resulted in the identification of other marine derived natural products having selective antimalarial activity (Kónig et al., 1998, Wright et al., 1996), par- ticularly the compounds isolated from Cymbastela hooperi (Kónig & Wright, 1995, 1997a; Kónig et al., 1996; Linden et al., 1996; Wright et al., 1996). In order to obtain further amounts ofthese natural products it was decided to investigate some sponge samples likely to contain this class of compound. In the present paper we provide a discussion of these biologically-guided isolat- ions, the results of chemical analyses, as well as the possible taxonomic implications of these findings. MATERIALS AND METHODS General methodology follows Wright et al. (1996). Abbreviations: DCM, dichloromethane; MeOH, methanol; EtOAc, ethyl acetate; VLC, vacuum liquid chromatography; HPLC, high performance liquid chromatography; TLC, thin layer chromatography; GC, gas chromatography; GC-MS, gas chromatography coupled mass spectrometry; 'H NMR, proton detected nuclear magnetic resonance spectroscopy. MATERIAL. All sponges were collected using SCUBA from the Great Barrier Reef, in the vicin- ity of Lizard Island between 11-30m depth. All specimens were frozen, then freeze dried. Five samples of 3 sponge species were: Cymbastela coralliophila Hooper & Bergquist, 1992 (Demo- spongiae, Halichondrida, Axinellidae) (specimen CTA); C. concentrica, Lendenfeld, 1887 (Demospongiae, Halichondrida, Axinelli- dae) (specimens CTD and CTE); Amphimedon terpenensis, Fromont, 1993 (Demospongiae, Haplosclerida, Niphatidae) (specimens CTB and CTC). EXTRACTION AND ISOLATION. Initially, a small piece (—5g of freeze dried tissue) from each sample was exhaustively extracted with DCM followed by MeOH. A portion of the resultant extracts was then sent for antimalarial and cyto- toxicity testing (-2mg). 'H NMR and TLC investigations of these extracts were also made. On the basis of the results obtained from the biological testing and the 'H NMR and TLC investigations, specimens CTA, CTC and CTD were subsequently selected for bulk extraction and fractionation. 282 MEMOIRS OF THE QUEENSLAND MUSEUM TABLE 1. Antimalarial activity, towards clones D6 and W2 of Plasmodium falciparum, of the dichloromethane (D) and methanol (M) extracts from 5 sponge samples. SI = the ratio of the KB cell cytotoxicity to the Plasmodium flaciparum toxicity. * Extract was non-toxic only towards KB cells, in other cell lines it was at least 100 times more toxic. ICso (ng/ml) Clone D6 Clone W2 Sample Species KB cells ICso (ng/ml) SI ICso (ng/ml) SI CTA (D) C. coralliophila 220,000 710,000 - >10,000 CTA (M) C. coralliophila >20,000 5150 >3.9 6380 >3.1 CTB (D) A. terpenensis 720,000 4820 >4.1 >10,000 - CTB (M) A. terpenensis >20,000 3240 262 9680 2.1 CTC (D)* A. terpenensis 720,000 x41 2490 «4] 2490 CTC (M) A. terpenensis 720.000 540 237 5250 23.8 CTD (D) C. concentrica 720,000 1360 214.7 710,000 - CTD (M) C. concentrica 220,000 2730 27 1360 214.7 CTE (D) C. concentrica 720,000 710,000 - 710,000 - CTE (M) C. concentrica 720,000 8470 24 5700 23.5 1) Specimen CTA: Freeze-dried tissue (108.4g) was exhaustively extracted with DCM (2L) and MeOH (2L) to yield 15.9g (14.7%) of DCM soluble material, and 11.3g (10,4%) of MeOH/H;O solubles. VLC separation of the DCM solubles over silica, employing gradient elution from hexane to acetone to MeOH, yielded 15 fractions each of approximately 90ml. Fractions 3-14 were predominantly compounds depicted in Figure 1 C-D. The remaining fractions and the MeOH/H;O solubles were ubiquitous lipids and a number of other sterols. 2) Specimen CTC: Freeze-dried tissue (38.8g) was exhaustively extracted with DCM (1.5L) and MeOH (2L) to yield 3.0g (7.7%) of DCM soluble material. VLC separation of the DCM solubles over silica, employing gradient elution from hexane to EtOAc to MeOH, yielded 12 fractions each of approximately 100ml. Fractions 1-5 were found by GC-MS analysis to contain compounds depicted in Figure 1 E-Q. Fractions 6 and 7 were essentially the pure compound shown in Figure 1B. Fractions 10 and 11 were found to contain the compounds depicted in Figures IR-X and 2A. The remaining fractions and the MeOH/H;O solubles were ubiquitous lipids and a number of sterols. 3) Specimen CTD: Freeze-dried tissue (57.5g) was exhaustively extracted with DCM (2L) and MeOH (2L) to yield 600mg (1.1%) of DCM soluble material. VLC separation of the DCM solubles over silica, employing gradient elution from hexane to ethyl acetate (EtOAc) to MeOH, yielded 13 fractions, each of approximately 80ml. All fractions and the MeOH/H5O solubles were ubiquitous lipids and a number of sterols. One of the major components of fractions 8-12 was a sterol of the type represented by the compound shown in Figure 2A. GC SEPARATIONS. GC analyses were done ac- cording to methods previously described (Witte etal., 1993). From each of VLC fractions 1 and 2, obtained from the DCM extract of sponge specimen CTC approximately Img of material was taken and analysed my GC-MS. The results of these analyses indicated VLC fraction 1 to contain the compounds depicted in Figures 1 E-H, and VLC fraction 2 to contain the compounds depicted in Figures 11-Q. BIOLOGICAL TESTING. The antimalarial (anti- malarial activity is defined as the ability of some substance, pure or mixture, to inhibit the growth of, or be lethal to, one or other strains of Plasmodium falciparum) and cytotoxicity testing was undertaken as previously described (Anger- hofer et al., 1992b, Likhitwitayawuid et al., 1993). RESULTS A small piece (—5g of dry tissue) from each of 5 sponge samples thought likely to contain di- terpene isonitriles of the type represented by diisocyanoadociane (Fig. 1B), was exhaustively extracted with DCM, followed by MeOH, and the resultant extracts sent for antimalarial activity assessment. Out of the 10 extracts only 2 were found to have significant activity, the DCM and MeOH extracts of sample CTC (A. terpenensis) (Table 1). The only other extracts to show some promise in terms of their activity and selectivity were the DCM and MeOH extracts of sample CTD (C. concentrica), and to a lesser extent the CYMBASTELA HOOPERI AND AMPHIMEDON TERPENENSIS 283 "N-G-H Ó FIG. 1. A-X, chemical structures of secondary metabolites derived mainly from sponges of the genera Cymbastela and Amphimedon (refer to text for further information). 284 DCM and MeOH extracts of sample CTB (A. terpenensis) (Table 1). On the basis of results presented in Table 1, TLC and 'H NMR ex- aminations of the extracts, detailed investigations were made of samples CTA (C. coralliophila), CTC (A. terpenensis) and CTD (C. concentrica). For this purpose DCM and MeOH extracts were prepared from bulk material of each of the 3 samples to identify the major components present. SAMPLE CTA (C. CORALLIOPHILA). The DCM extract of this sample (CTA, C. coral- liophila) was found to contain predominantly lipids and 2 sterols (Fig. 1C-D), previously iso- lated from Pseudaxinyssa sp. The latter sample was collected from several mid-shelf reefs on the Great Barrier Reef, by Hofheinz & Oesterhelt (1979). These 2 sterols have also been isolated from another Pseudaxinyssa sp. collected from a reef fringing Pelorus Island on the Great Barrier Reef (Kónig, G. M. & Wright, A. D., unpublished data). An interesting observation concerning these compounds (Fig. 1 C-D) is, that they always seem to occur as a 1:1 mixture which is essent- ially inseparable, even by GC (Bergquist et al., 1980). The MeOH extract was composed of ubiquitous lipids and a number of other sterols (Bergquist et al., 1980). SAMPLE CTC (4. TERPENENSIS). Chromat- ographic and 'H NMR analyses indicated the MeOH extract of CTC (A. terpenensis) to contain many ofthe components to be found in the DCM extract. The MeOH extract was therefore part- itioned between water and DCM and the resulting DCM solubles combined with the DCM extract. These DCM solubles were fractionated as outlined in the experimental section. 'H NMR analysis of the resultant fractions indicated a similarity in composition to those produced by the fractionation of the DCM solubles obtained from the previously investigated C. Aooperi (König et al., 1996; König & Wright, 1997b). Based on this observation GC-MS investigations of selected VLC fractions were undertaken. These analyses indicated the sample to contain compounds shown in Figures 1 E-Q, and thus, to be almost identical in secondary metabolite content to C. hooperi (Kónig et al., 1996; Kónig & Wright, 1997b). This finding also explained the observed antimalarial activity of its DCM extract. As a result of these studies it was also observed that VLC fraction 11 contained a num- ber of resonances in the 8.0-8.3ppm region ofthe proton NMR spectrum. Comparison of this H NMR spectrum with an equivalent VLC fraction MEMOIRS OF THE QUEENSLAND MUSEUM from C. hooperi (Kónig et al., 1996; Kónig & Wright, 1997b; Wright et al., 1996) showed the two VLC fractions to be almost identical. Pur- ification of the main components from both VLC fraction 11s has resulted in the identification of 7 diterpene formamide derivatives (Fig. IR-X), a number of which are new natural products, and a mixture of peroxide containing sterols ofthe type represented by Figure 2A; the detailed results of this investigation will be presented elsewhere (Kónig et al., in preparation). SAMPLE CTD (C. CONCENTRICA). Both the DCM and MeOH extracts of this sample (CTD, C. concentrica) were found to be complex mix- tures of ubiquitous lipids and sterols. In VLC fractions 8-12, made from the DCM solubles of CTD, sterols of the type represented by Figure 2A were abundant. TLC and 'H NMR of the extracts of the two remaining sponge samples, CTB (A. terpenen- sis), and CTE (C. concentrica), clearly showed that specimen CTB is very similar in all respects to CTC (A. terpenensis) and that sample CTE shows the greatest similarity to samples CTA and CTD, particularly with respect to their 'H NMR spectra. The reduced activity of the DCM and MeOH extracts of CTB when compared to the activity of the equivalent extracts of CTC, ap- pears to be due to the relatively large amounts of lipids present in the extracts of CTB. DISCUSSION Of the 3 species of Cymbastela investigated none were shown to contain terpenoids substitut- ed with isonitrile based functionalities, This is in direct contrast to results obtained for C. hooperi (Kónig & Wright, 1995, 1997a; Kónig et al., 1996; Linden et al., 1996; Wright et al., 1996). In this respect it is of interest to note that C. hooperi is also morphologically distinguished from other Cymbastela species (Van Soest et al., 1996). The investigation of the two samples of A. terpenensis (CTB and CTC), however, led to the ident- ification of secondary metabolites identical, or closely related, to those obtained from C. hooperi. Literature relating to the secondary metabolite chemistry of sponges from Amphimedon and Cymbastela shows sponges from the former to have received the most attention. In the 40 or so publications on Amphimedon the compounds which are typically reported are: various classes of alkaloids (e.g. Fig. 2B-F; Chehade et al., 1997; Kobayashi et al., 1994a; Kobayashi et al., 1994b; Schmitz et al., 1983; Tsuda et al., 1994), long CYMBASTELA HOOPERI AND AMPHIMEDON TERPENENSIS 285 FIG. 2. A-O, chemical structures of secondary metabolites derived mainly from sponges of the genera Cymbastela and Amphimedon (refer to text for further information). chain fatty acid derivatives (e.g. Fig. 2G-H; diterpenes (e.g. Fig. 1B, J, K; Fookes et al., 1988; Carballeira & Lopez, 1989; Garson et al., 1994), Kazlauskas et al., 1980; König & Wright, 1995). glycosphingolipids (e.g. Fig. 21; Hirsh & As there have only been about 11 reports on the Kashman, 1989) and some isonitrile containing secondary metabolite chemistry of sponges from 286 Cymbastela it is possible to show most of the isolates in this contribution. From C. corallio- phila steroids of the type represented by Figure 2J were isolated (Makarieva et al., 1995). Pyraxi- nine (Fig. 2K), anovel alkaloid was isolated from C. cantharella (Mourabit et al., 1997), while from two unidentified species of Cymbastela two peptides (Fig. 2L-M; Coleman et al., 1995) and two pentacyclic bromopyrroles (agelastatins C and D; Fig. 2N-O, respectively; Hong et al., 1998) were obtained. The present authors have also published a number of works about second- ary metabolites from C. hooperi (Kónig & Wright, 1995, 1997a; Kónig et al., 1996; Linden et al., 1996; Wright et al., 1996) and the typical metabolites described are mainly isonitrile con- taining diterpenes (e.g. Fig. 1B, I-Q). When the literature is considered for Cym- bastela and Amphimedon it is evident that there is currently no class of secondary metabolite one might designate as being ‘characteristic’ for one or the other of these genera. What is evident, however, is that in both genera only one species produces isonitrile containing diterpenes; C. hooperi and A, terpenensis. Two observations can be made: A. terpenensis is positioned in an order (Haplosclerida) and family (Niphatidae) where no other sponges are known to produce secondary metabolites that have isonitrile or similar functionalities, and C. hooperi is located in an order (Axinellida or Halichondrida; see Van Soest, 1996) and family (Axinellidae) that are known to contain sponges that produce isonitrile containing secondary met- abolites. It is clear that ‘A. terpenensis” does not fit in Amphimedon as currently defined (Bergquist, Fromont, Hooper, Van Soest, pers. comm.), nor is it clearly a haplosclerid, it is possibly an axinellid close to Cymbasiela, as suggested by Van Soest et al. (1996), but it’s life characteristics do not conform well with the other species of Cymbastela (e.g. growth form, texture, mucus production, surface features and amount of spongin to spicule ratio). Cymbastela as defined by Hooper & Bergquist is a fairly homo- geneous genus, and ‘A. terpenensis” clearly disrupts that homogeneity. CONCLUSIONS These observations and the fact that C. hooperi and a specimen of A. terpenensis have been shown to have almost identical secondary met- abolite chemistry, lead to three possible conclusions concerning their current taxonomic MEMOIRS OF THE QUEENSLAND MUSEUM classification. 1) 4. terpenensis belongs to Cymbastela, as proposed by Van Soest et al. (1996); 2) both A. ferpenensis and C. hooperi belong to another, possibly new genus located in the family Axinellidae; 3) they are the same species with kooperi representing an unusual morphotype of terpenensis. The results of the current work indicate that the taxonomic clas- sification of C. hooperi and A. terpenensis needs to be clarified, particularly since sponges be- longing to both ofthese species produce so many interesting and biologically active compounds. This study also serves to further highlight the significance of secondary metabolite chemistry as an important taxonomic tool. It is also hoped that continued investigations into the biologic- ally active secondary metabolites produced by both of these sponge species will eventually lead to the development of an agent suitable for the treatment of malaria and/or some other disease. ACKNOWLEDGEMENTS We thank J.N.A Hooper, Queensland Museum, Brisbane, for identification of all sponges. C.K. Angerhofer, Department of Medicinal Chemistry and Pharmacognosy, University of Illinois at Chicago, and N. Lang-Unnasch, Division of Geographic Medicine, University of Alabama at Birmingham, provided the antimalarial and cytotoxicity data associated with this study. L. Witte, Institute for Pharmaceutical Biology, Technical University Braunschweig, Germany, performed the GC-MS analysis associated with sample CTC. All NMR measurements made were undertaken by V. Wray and his team, Gesellschaft für Biotechnologische Forschung mbH (GBF), Braunschweig. We are indebted to J. Fromont, Western Australia Museum, Perth, P. Bergquist, University of Auckland and R. Desqueyroux-Faündez, Museum d'Histoire Naturelle, Geneva, for their constructive comments concerning the possible taxonomic significance of our findings. LITERATURE CITED ANGERHOFER, C.K., KÓNIG, G.M., WRIGHT, A.D., STICHER, O. & PEZZUTO, J.M. 1992a. Antimalarial sesquiterpenes from the marine sponge Acanthella klethra. Journal of Natural Products 55: 1787-1789. ANGERHOFER, C.K., KONIG, G.M., WRIGHT, A.D., STICHER, O., MILHOUS, W.K., CORDELL, G.A., FARNSWORTH, N.R. & PEZZUTO, J.M. 1992b. Selective screening of natural products: A resource for the discovery of novel antimalarial compounds. Pp. 311-329. In CYMBASTELA HOOPERI AND AMPHIMEDON TERPENENSIS Atta-ur-Rahman (ed.) Advances in Natural Product Chemistry. (Harwood Academic Publishers: Chur, Switzerland). BERGQUIST, P.R., HOFHEINZ, W. & OESTER- HELT, G. 1980. Sterol composition and the classification of the Demospongiae. Biochemistry Systematics and Ecology 8: 423-435. CARBALLEIRA, N.M. & LOPEZ, M.R. 1989. On the isolation of 2-hydroxydocosanoic acid and 2-hydroxytricosanoic acid from the marine sponge Amphimedon compressa. Lipids 24: 89-91. CHEHADE, C.C., DIAS, R.L.A., BERLINCK, R.G.S., FERREIRA, A.G., COSTA, L.V., RANGEL, M., MALPEZZI, E.L.A., DE FREITAS, J.C. & HAJDU, E. 1997, 1,3-Dimethylisoguanine, a new purine from the marine sponge Amphimedon viridis. Journal of Natural Products 60: 729-731. COLEMAN, J.E., DE SILVA, E.D., KONG, F., ANDERSEN, R.J. & ALLEN, T.M. 1995. Cytotoxic peptides from the marine sponge Cymbastela sp. Tetrahedron 51: 10653-10662. FOOKES, C.J.R., GARSON, M.J., MACLEOD, J.K., SKELTON, B.W. & WHITE, A.H. 1988. Bio- synthesis of diisocyanoadociane, a diterpene from the marine sponge Amphimedon sp. Crystal structure of a monoamide derivative. Journal of the Chemical Society, Perkin Transactions 1: 1003-1011. GARSON, M.J., ZIMMERMANN, M.P., BATTERS- HILL, C.N., HOLDEN, J.L. & MURPHY, P.T. 1994, The distribution of brominated long-chain fatty acids in sponge and symbiont cell types from the tropical marine sponge Amphimedon terpenensis. Lipids 29: 509-516. HIRSH, S. & KASHMAN, Y. 1989. New glycosphingo- lipids from marine organisms. Tetrahedron 45: 3897-3906. HOFHEINZ, W. & OESTERHELT, G. 1979, 24-Isopropylcholesterol and 22-dehydro- 24-isopropylcholesterol; novel sterols from a sponge. Helvetica Chimica Acta 62: 1307-1309. HONG, T.M., JIMENEZ, D.R. & MOLINSKI, T.F. 1998. Agelastatins C and D, new pentacyclic bromopyrroles from the sponge Cymbastela sp., and potent arthropod toxicity of (-)-agelastatin A, Journal of Natural Products 61: 158-161. KAZLAUSKAS, R., MURPHY, P.T., WELLS, R.J. & BLOUNT, J.F. 1980. New diterpene isocyanides from a sponge. Tetrahedron Letters 21: 315-318. KOBAYASHI, J., TSUDA, M., KAWASAKI, N., MATSUMOTO, K. & ADACHI, T. 1994a. Keramaphidin B, a novel pentacyclic alkaloid from a marine sponge Amphimedon sp.: a plausible biogenetic precursor of manzamine alkaloids. Tetrahedron Letters 35: 4383-4386. KOBAYASHI, J., TSUDA, M., KAWASAKI, N., SASAKI, T. & MIKAMI, Y. 1994b. 6-Hydroxymanzamine A and 3, 4-dihydro- manzamine A, new alkaloids from the Okinawan marine sponge Amphimedon sp. Journal of Natural Products 57: 1737-1740. KONIG, G. M. & WRIGHT, A.D. 1995. 'H (300 MHz) and "C NMR spectral assignments for diisocyanoadociane. Magnetic Resonance in Chemistry 33: 694-697. 1997a. The marine environment: a source of novel antimalarial lead structures. Pp 159-168. In L. Verotta (ed). Virtual activity, real pharmacology. (Research Signpost: Trivandrum, India), 1997b, New and unusual sesquiterpenes; kelsoene, prespatane, epi-g-gurjunene and T-cadinthiol, from the tropical marine sponge Cymbastela hooperi. Journal of Organic Chemistry 62: 3837- 3840. KONIG, G.M., WRIGHT, A.D. & ANGERHOFER, C.K. 1996. Antimalarial diterpene isonitriles, isothiocyanates and isocyanates from the tropical marine sponge Cymbastela hooperi n. sp. Journal _ of Organic Chemistry 61: 3259-3267. KONIG, G.M., WRIGHT, A.D. & LINDEN, A. 1998. Antiplasmodial and cytotoxic metabolites from the Maltese sponge Agelas oroides. Planta Medica 64: 443-447, LIKHITWITAYAWUID, K., ANGERHOFER, C.K., CHAI, H., CORDELL, G. A., PEZZUTO, J.M. & RUANGRUNGSI, N. 1993. Cytotoxic and antimalarial alkaloids from the tubers of Stephania pierrei. Journal of Natural Products 56: 1468-1478. LINDEN, A., KÓNIG, G.M. & WRIGHT, A.D. 1996. X-ray crystal structure of amphilectane derivatives. Acta Crystallographica Section C 52 2601-2607. MAKARIEVA, T.N., STONIK, V.A., DMITRENOK, A.S., KRASOKHIN, V.B., SVETASHEV, V.I. & VYSOTSKII, M.V. 1995, New polar steroids from the sponges Trachyopsis halichondroides and Cymbastela coralliophila. Steroids 60: 316-320. MOURABIT, A.A., PUSSET, M., CHTOUROU, M., GAIGNE, C., AHOND, A., POUPAT, C. & POTIER, P. 1997. Pyraxinine, a novel nitrogenous compound from the marine sponge Cymbastela cantharella. Journal of Natural Products 60: 290-291, SCHMITZ, F., AGARWAL, S.K., GUNASEKERA, S.P., SCHMIDT, P.G. & SHOOLERY, J.N. 1983, Amphimedine; New aromatic alkaloid from a Pacific sponge, Amphimedon sp. Carbon connectivity determination from natural abundance '°C-'°C coupling constants. Journal of the American Chemical Society 105: 4835-4836. TSUDA, M., KAWASAKI, N. & KOBAYASHI, J. 1994. Keramaphidin C and keramamine C, plausible biogenetic precursors of manzamine C from an Okinawan marine sponge. Tetrahedron Letters 35: 4387-4388. SOEST, R.W.M. VAN, DESQUEYROUX- FAUNDEZ, R., WRIGHT, A.D. & KÓNIG, G.M. 1996. Cymbastella hooperi sp. nov. (Halichondrida: Axinellidae) from the Great Barrier Reef, Australia. Bulletin de l'Institut 288 Royal des Sciences Naturelles de Belgique, Biologie 66; 103-108. WITTE, L- RUBIOLO, P., BICCHI, C. & HART- MANN, T. 1993. Comparative analysis of pymalizidine alkaloids from natural sources by gas chromatoyraphy-mass spectrometry. Phytochemistry 32: 187-196. MEMOIRS OF THE QUEENSLAND MUSEUM WRIGHT, A.D.. KONIG, G.M.. ANGERHOFER, C.K.. GREENIDGE. Pa LINDEN, A. & DESQUEYROUX-FAUNDEZ, R. 1996. Antimalarial activity: The search for marine derived natural products which demonstrate selective antimalarial activity. Journal of Natural Products 59: 710-716, THE REPLACEMENT OF NATURAL HARD SUBSTRATA BY ARTIFICIAL SUBSTRATA; ITS EFFECTS ON SPONGES AND ASCIDIANS. Memoirs of the Queensland Museum 44: 288. 1999;- Subtidal reefs around coastal cities such as Sydney are campased ofa variety of natural and artificial substrata. Commonly these are natural rocky reefs, breakwalls. seawalls and pier pilings. These types of hard substrata differ in their structure. Most natural hard substrata consist of horizontal surfaces; most surfaces on artificial hard substrata are vertical. Therefore, replacing natural hard substrata with artificial hard substrata is likely to change the surface of substrata from predominantly horizontal to musily vertical. To understand and predict the potential effects of these changes on the assemblage af sponges and ascidians it is important to determine their distribution on harizontal and vertical surfaces, The few ecological studies on the distribution of algae and invertebrates on horizontal and vertical surfaces have reported that there are more sponges and CONVERGENCE IN THE TIME-SPACE- CONTINUUM: A PREDATOR-PREY INTERACTION. Memoirs of the Queensland Museum 44: 288. 1999:- Community structure is influenced by many biotic and abiotic factors. Predation is a key structuring mechanism for some marine communities. Prey abundances may fluctuate with strength of predator recruitment and persistence, eXceprin cases where some of the prey population has à réfuge in space or time from predation. Consistent, moderate predation levels po a predicably available prey respurce should lead to stable community structure with relatively small fluctuations in predator and prey population densities. Conversely, prey species lacking a refuge from predation are subject la major population fluctuations commensurate with strength of predator recruitment and abundance. The sponge Halichondria panicea is patchily distributed in the rocky intertidal on the south shore of Kachemak Bay. southcentral Alaska, and in certain locations is the spatial dominant. At onc site approximately 55m in horizontal length. H. panicea has dominated the mid-intertidal for al least 10 years, with low densities of potential molluscan predators such as Archidoris montereyensis: Katheriña tunicata, and Diadora aspera present. Percent cover estimates of primary space occupiers at the site were collected from 10 0.25m^ permanent quadrats established in August 1994, H. panicea averaged $3.4%0 +/-9.9% cover through August 1996. Other major cover categories were algae, 14 694 +/-6 4%, and open rock, 26.1% +/10.2%. Visits to the site in early spring of [997 revealed that the sponge colonies overwintered with ascidians on vertical than on horizontal surfaces. I1 has not been tested whether these patterns exist in the temperate waters around Sydney. Furthermore, of the studies that have examined the effects ot horizontal and vertical surfaces on tbe distribution of sponges and ascidians, none has experimentally tested the factors that cause these distributions. Here, I present results of my tesis of the hypothesis that sponges and ascidians are more abundant on vertival than horizontal surfaces in the shallow subtidal zone around Sydney. | will also discuss future manipulative experiments to determine which factors are important in creating these distributions. O Porifera, Ascidiacea, distribution, hard substrata, shallow subtidal, habitar. Nathan Knott (email: nánomablo.usvd edu.au), Special Research Centre on Ecological impacts of Coastal Cities, Marine Ecology Laboratories ALI, Universitv of Sydney, NSW, 2006 Australia; 1 June J998. few indications af major mortality events. No percent cover dala were collected at thai time, Total numbers of the nudibranch Arehidoriy monterevensis, which is a specialist predator on X. panicea, present at lhe site were recorded and ranged from 12-42 from 1994-1996, In the spring of 1997, strong recruitment resulted in an average population of 151 A. menterevensis on site from May to July. Percent cover of H. panicea declined from visual estimates of 40% in May to 15% in July. By August 1997, when the 10 permanent quadrats and 10 haphazardly placed quadrats were measured, essentially no sponge could be found at the study site. After July, the abundance af nudibranchs declined to 32 individuals commensurate with sponge reduction. By September, only one small! sponge colony and 7 predatory nudibranchs were present at the site. Even though A- panicea is abundant in the region and potential recruits should be numerous, as af April 1998, the site once dominated by H, panicea is predominantly open rock with some recruitment of annual macroalgae occurring. The predator-prey relationship of 4. montereyensis and M. panicea is an example of a chase through space and time with convergence resulting in extreme population fluctuations and an unstable community. O Porifera. predation, nudibranch, intertidal, predalar/prey interaction, community structure, Alaska, recruitment. Ana L Knowlton (email: fialicauud.edu) & Raymond C, Highsmith, Institute af Marine Science, School of Fisheries and Ocean Sciences, University af Alaska Fairbanks, P.O. Box 737220, Fairbanks, AK 99775-720. USA: 1 June 1998. A CARNIVOROUS SPONGE, CHONDROCLADIA GIGANTEA (PORIFERA: DEMOSPONGIAE: CLADORHIZIDAE), THE GIANT DEEP-SEA CLUBSPONGE FROM THE NORWEGIAN TRENCH BETTINA KUBLER AND DAGMAR BARTHEL Kiibler, B. & Barthel, D. 1999 06 30: A carnivorous sponge, Chondrocladia gigantea (Porifera: Demospongiae), the giant deep-sea clubsponge from the Norwegian Trench. Memoirs of the Queensland Museum 44: 289-298, Brisbane. ISSN 0079-8835. The ultrastructure of the deep-sea sponge Chondrocladia gigantea from the Norwegian Sea, North Atlantic, was studied for the first time. Club-shaped, erect C. gigantea has a unique form of aquiferous system, not previously observed in Porifera, consisting of rows of large choanocyte chambers running through the main axis of the sponge, which explains the numerous, normally extended water-filled spheres sitting on little stalks in the upper external part of the main body. These previously enigmatic translucent spheres serve as surface extensions of the sponge to trap prey in the food-poor, deep-sea environment. In addition, they release male reproductive cells into the water. Sexual reproduction seems to play an important role in C. gigantea, since spermatocysts were found at different stages of maturity in two out of six samples examined. No mature oocytes were encountered, leading to the assumption that this species may be hermaphroditic (probably with a seasonal reproductive cycle). The phylogenetic relationship of Chondrocladia to the other genera of the Cladorhizidae is discussed, based on the presence of an aquiferous system with choanocyte chambers as the basic ‘bauplan’ of sponges, which is lost in the other genera. O Porifera, Cladorhizidae, deep-sea, food-poor environment, adaption, aquiferous system, choanocyte chambers, macrophagy, carnivory. Bettina Kübler, (e-mail: bkuebler@ifm.uni-kiel.de), Kórnerstrafie 3, 24103 Kiel, Germany; Dagmar Barthel, Chemin du Bonnier 9, 1380 Lasne, Belgium; received 5 October 1998. Carnivory is an extremely rare feeding strategy among Recent sponges, with confirmed records so far only from the Cladorhizidae, which is a typical deep-sea family restricted to the bathyal, abyssal or even hadal zones. However, Vacelet (1999, this volume) suggests that several other poecilosclerid families are also likely to contain carnivorous species, as judged by their published descriptions. Carnivory was first discovered in Asbestopluma hypogea (Vacelet & Boury- Esnault, 1995, 1996), from a Mediterranean shallow-water cave where the habitat resembles that of the deep-sea (Vacelet et al., 1994). In adaption to its food-poor environment A. hypogea developed a carnivorous feeding strategy. The organism is no more than 15mm high, carrying long, thin filaments on its slim main axis, on which swimming prey is captured and overgrown by sponge cells within hours, Vacelet et al. (1995) also described carnivory ina Cladorhiza sp., a sponge from a mud volcano in the Barbados Trench that has developed a symbiosis with methanotrophic bacteria: ‘The sponge morphology, erect with branching pro- cesses bearing a cover of hook-like spicules, suggests that they may also feed on swimming prey ... This was supported by the presence of debris from small crustaceans on the sponges. So far, nothing is known about the feeding strategy of the third genus of the Cladorhizidae, Chondrocladia. The deep-sea sponge C. gigantea (Lundbeck, 1905) has a remarkable morphology that has always fascinated scientists. The giant clubsponge, whose skeleton is built by styli and collagen, carries spheres filled with water on little stalks. These are situated mostly on the upper part of its main body, which is slim and erect, rooted in the muddy substratum. The tallest known specimen is 600mm long and has a maximum width of 50mm. From in situ photographs thin-walled spheres are seen to be translucent, whereas when brought to the water surface ‘the spheres at the tip of the branches are shrunk into the somewhat oblong, clavate, relatively massive structures characteristic of the branches of a number of Chondrocladia species ...'(Tendal et al., 1993) (Fig. 1). Like the main body and stalk these spheres have a certain firmness attributed to the presence of collagen and styli, but are addit- ionally covered by hook-like isochelae. FIG, 1. In situ photograph of Chondrocladia sp., by H. Sahling at 4,900m depth, 54°18°N, 157?11"W. Esti- mated extended sponge diameter is approx. 50mm (reproduced with permission from Tendal & Sahling). MATERIALS AND METHODS Five specimens of Chondrocladia gigantea were dredged from 480m depth at BIOICE station 2792 in the North Atlantic (67°15.17’°N; 18*52.01' W) on 15 August 1995 (Fig. 2). Their lengths varied from 78-98mm and widths of their main axes between 3-22mm. On board samples were preserved for electron microscopy with a double fixation in 1% glutaraldehyde and 1% osmium tetroxyde according to the procedure developed by Langenbruch (1983). Afterwards, samples were desilicified with 5% hydrofluoric acid in sea water. The samples were then stored in 100% ethanol at 5°C. After two years the samples were embedded in acrylic resin (Unicryl, British Biocell Inter- national, Cardiff, UK), and cut into semi-thin (1um) and ultrathin (60-150nm) sections. The semi-thin sections were examined with a Leitz DM RB light microscope (phase contrast) after staining with toluidine blue, eosin and haema- toxylin (all stains provided by British Biocell). For transmission electron microscopy, the ultra-thin sections, cut with freshly made glass knifes on a Reichert OM U3 ultramicrotome, were placed on slot-grids and the contrast was enhanced by uranyl acetate and lead citrate MEMOIRS OF THE QUEENSLAND MUSEUM FIG. 2. Five different individuals of C. gigantea exam- ined in the present study, prior to TEM-fixation. The compact ‘bulbs’ on the main axis are deflated spheres. The largest specimen measures 20cm in length. (adapted from the method of Reynolds, 1963). The photographs were taken on a Zeiss EM9 S2 electron microscope using photo plates. The samples were also used for scanning electron microscopy on a Zeiss DSM 940 microscope with a Nikon camera system. A formalin-fixed sample of C. gigantea collected by Ole S. Tendal at BIOICE station 2085, 754m depth, 4 July 1992 was embedded in paraffin (AgarScientific Ltd, Stansted, UK), sectioned on a Leitz microtome (3-7pm) and stained with toluidine blue. RESULTS The main axis ofall specimens had an extended aquiferous system with wide canals (200-300um diameter) (Fig. 3) and oval-shaped choanocyte chambers measuring up to 100um diameter length (Fig. 4). The canals of the aquiferous system appeared to run through the whole length of the main axis of the stalk, while the choanocyte chambers were found in rows along them. The surface of the main axis seemed to carry pores CARNIVOROUS NORWEGIAN SPONGE FIG. 3. A cross section through the main axis of C. gigantea. The widths of canals measure between 200-300)1m. Spermatocysts (sc) gather in the centre of the axis, whereas choanocyte chambers (chc) are situated mostly be- tween canals (c) and surface of the sponge. Phase contrast microscopy. that were about 160m wide, narrowing to 20um diameter, Through these pores diatoms could pass, some of which were found on the walls of the canals. These inlets covered the outside in a regular order, approximately 500um apart. No choanocytes or ostia were found in the spheres, but canals ran through them and at the distal end of one sphere we noted a few openings (approximately 13) likely to be oscules. How- ever, these structures could as well be caused by deflation. The main axis and stalks contained only styli, whereas spheres carried styli as well as isochelae and, occasionally, sigmata. The outsides of the spheres were covered entirely with hook-shaped spicules, the microsclerid isochelae, standing closely together like a palisade (Fig. 8). The palisade was then underlain by styli which formed lateral layers or upward protrusions. The thickness of the palisade measured about 70pm. Because of its inflexibility the palisade was partially folded up into the deflated sphere. Many spermatocysts at different stages of maturity were found in the main axis (Figs 3, 5, 6) and spheres (Fig. 7), all enclosed in cysts. In the middle of the main axis the gametocytes were globular (Figs 5-6) but appeared to elongate when migrating towards the spheres losing their protoplasm (Fig. 7). In one case, a distinct area was visible where cells appeared to form cysts (up to 80pm diameter) and develop into spermatogonia (Fig. 5). More mature stages of FIG. 4. Oval-shaped choanocyte chamber with apopyle (a) from the main axis. Phase contrast mic- roscopy. 292 MEMOIRS OF THE QUEENSLAND MUSEUM FIG. 5. Spermatocysts from a spermatogonia region in the main axis. Within the cysts (c) the cells (sc) de- velop simultaneously, but the individual cysts are at different stages of maturity. They are enveloped in epithelial cells (e). Around the cysts are many cells with inclusions (ic) whose function is still unknown. Phase contrast microscopy. spermatocytes were found on the periphery of this area, the cysts there measuring only 60um diameter. Even in the spheres the mature Spermatozoa remained in cysts. Using semi-thin sections of mature cysts, we estimated that a single cyst contained at least 500 spermatozoa. Crustaceans in various stages of digestion were abundant only in the spheres, with up to 19 individuals per sphere (Fig. 9). No other prey organisms were noted, Due to their small size or advanced state of digestion, only two species, Calanus finmarchicus and Calanus hyperboreus, could be determined exactly and were found in greatest abundance (Fig. 10). In one sphere 16 specimens were found probably belonging to C. finmarchicus, up to 5.8mm long, and most in the copepodit stage V (Fig. 9).The largest item of prey measured 6.5mm long. FIG. 6. Spermatocyst in the main axis with spermato- cytes that could be in their reduction phase (sc2). The nucleoli (nu) are visible. Condensed heterochromatin can be seen as dark rings on the nuclei, Phase contrast microscopy. Muscle tissue was found both inside the chitin cuticule of crustaceans surrounded by archaeocytes (Fig. 11) that had migrated towards the prey, and as inclusions in cells (Fig. 12a). Whereas in the first case the tissue still appeared to be intact, showing its striation, pieces of muscle ussue within inclusion cells showed different stages of digestion. In the ones most digested striation was no longer visible, and ihe inclusions consisted of a more-or-less homogenous mass. In the outer layer of the main body globular structures of up to 5mm diameter were found being distinctly different from the surrounding sponge tissue. Consisting only of inclusions, they were enveloped in a layer of collagen with an average thickness of 301m. Bacteria were found extracellularly in the sponge tissue as well as in massive gatherings of different sizes, especlally inside the chitin cuticule of half-way digested crustaceans. where they were no longer intact. CARNIVOROUS NORWEGIAN SPONGE 295 FIG. 7. Cyst with elongated, more mature spermato- cytes inside a sphere. Phase contrast microscopy. DISCUSSION Our investigations showed that C. gigantea has developed carnivory like other members of the family, but with a major difference: whereas the other genera have apparently lost their aquiferous system. Chondrocladia still possesses it. AQUIFEROUS SYSTEM. Water is drawn into the main body through pores and then ‘pumped’ through canals into the spheres. The water current probably flows unidirectionally, from the base of the organism to the top, since the spheres are predominantly in the upper part of the sponge. It is probably not only collagen and spicules that keep this slim organism upright, but also the water pressure within the sponge. The spheres must be filled with water via the stalks. This is the reason why they appear translucent in in situ photographs and possess little biomass in relation to surface area. Water 1s assumed to be expelled through oscules on the spheres. FUNCTION OF THE AQUIFEROUS SYSTEM AND THE SPHERES. The presence of wide canals running through the main axis and rows of FIG. 8. A palisade of isochelae (isp) on the surface of a sphere. Underneath a criss-cross layer of megaseleres (ms), which are simple styli. Scanning electron mic- roscopy (SEM). choanocyte chambers allow this tall, but slim, sponge to perform an exceptional feeding mode. The spheres can be explained as structures used to catch prey. When observed in situ (Pig. 1) they are usually filled with water and are seen as extended, thin-walled, translucent spheres. It 15 likely that passing crustaceans are caught on the protruding palisade of microscleres with their many thin appendages. As suggested by Tendal & Sahling (1997), a sphere can probably collapse within tens of seconds. Through the mechanical stimulus of the prey the sphere ejects its water contents through openings at its distal end. Collapsing very quickly, the crustacean is hooked and completely surrounded by sponge tissue so it cannot escape, The crustacean is then digested. Apparently archaeocytes migrate towards the prey and ingest pieces of muscle tissue (estimated size 200um*). These archaeocytes are then inclusion cells and migrate through the sponge tissue of the spheres into the main axis. The muscle tissue was often found in rectangular shapes leaving the impression that it was dissected into distinct pieces. Aggregations of pieces of muscle tissue that were observed close to the areas where spermatogonia were formed suggest that their protein content may be utilised for the build-up of reproductive cells. REPRODUCTIVE CELLS, Only male gameto- gonia could be clearly recognised, although the presence of very young oocytes in the same 294 MEMOIRS OF THE QUEENSLAND MUSEUM FIG. 9. Contents of a sphere from a formalin-fixed sample. The largest crustacean is a Calanus species, copepodit stage V, measuring 5.8mm. specimen where spermatocytes were noted can- not be precluded (Kübler, 1998). The presence of spermatocytes in at least two out of six individ- uals and the lack of mature oocytes may indicate successive hermaphroditism. Spermatocytes were obviously produced very quickly, while oocytes seemed to take more time. Our investigations support the hypothesis that sexual reproduction plays an important role even in deep-sea organisms, as stated by Witte (1996) for other deep-sea demosponges. Finding mature stages of male reproductive cells and very young oocytes confirms the idea that seasonal reproduction also takes place in greater depths due to dependence on food availability, i.e. productivity of the surface waters, in this case in the form of crustaceans. The hypothesis that asexual reproduction in the form of budding could take place in this species (Tendal & Barthel, 1993) nevertheless cannot be precluded. Since the most mature stages of spermatocytes were found in cysts within the spheres, complete cysts may be released into the surrounding water, possibly ejected through the distal openings on the spheres concurrent with the expulsion of water. This would be a mode of reproduction similar to that described by Vacelet & Boury-Esnault (1996) for the related species Asbestopluma hypogea. According to Tuzet (1930), Fincher (1940) and others, archaeocytes constitute the source of spermatogonia in Porifera. Subsequently it was proposed that choanocyte chambers could be their predecessors (e.g. Tuzet et al., 1970; Paulus, 1989; Barthel & Detmer, 1990). Although in our FIG. 10. Calanus hyperboreus Króyer, 1838, copepodit stage V, length: 6.5mm. samples there were choanocyte chambers that apparently disintegrated to almost the same size as spermatocysts, spermatogonia could also be formed from archeocytes that contain inclusions. Theoretical considerations to calculate cell num- bers in choanocyte chambers and spermatocysts (spermatocytes 2 and spermatids) led to the result that choanocyte chambers (with about 1,500 cells) possess roughly two to three times more cells than mature spermatocysts, with the con- sequence that they probably would not originate directly from choanocyte chambers. In addition, young spermatocysts gathered in certain areas towards the centre of the main axis, whereas choanocyte chambers were regularly distributed and orientated towards the surface of the main axis (Fig. 3). Thus, choanocytes would have to migrate towards the centre to form cysts. In addition to spermatogenesis another reason for disintegration of choanocyte chambers may have been poor preservation of parts of the sponge tissue. CRUSTACEANS. Most of the crustaceans found in the spheres belonged to two species that are very common in this area of the Norwegian Sea. Almost all of these were in the last copepodite stage (V), which usually sink to the bottom of the sea in late summer to hibernate (e.g. Orr, 1934; Raymont, 1963). It can be assumed that the sponge neither selects its food nor has any food preferences. GATHERINGS OF INCLUSIONS. The function of gatherings of inclusions to build globular structures up to 5mm diameter in the main body could not be clarified. Originally, they were interpreted as embryos by Lundbeck (1905), but this could not be subsequently proven since no embryonic structures were found. Possible alternative interpretations are that these are places where by-products are deposited, or they may be CARNIVOROUS NORWEGIAN SPONGE FIG, 11, Pieces of intact muscles (striated) within an extremity of a crusta- cean found in a sphere (m, muscle; a, archaeocytes; ch, chitin cuticule). Phase contrast microscopy. depots of nutrients. The latter interpretation makes most sense given the food-poor environment in which C. gigantea lives, and the presumably vicarious seasonal food availability, but there is so farno empirical support for either hypothesis. BACTERIA. The role of bacteria in this species was not clarified from our investigations. The fact that masses of bacteria, which did not look intact, were found close to or within prey in the sponge tissue, provides two possible assump- tions: 1) either bacteria are digested by the sponge, which would mean, that this species is optionally bacterivor, or 2) bacteria facilitate digestion of prey organisms. We assume that in C. gigantea the presence of an extended aquiferous system, which in other sponges is used to filter food particles (e.g. Simpson, 1984), is modified to an elaborate mech- anism to catch prey. The spheres, which are part of this mechanism, also distribute the sponges’ sexual products into the surrounding water. Whereas closely related species like A. hypogea catch their prey (also consisting of crus- taceans) by overgrowing it, C. gigantea catches its prey by inflating its spheres (probably within tens of seconds; Tendal & Sahling, 1997). Both species react to the mechanical stimulus induced by crustaceans landing on the external surface. Living as a predator (macrophagy) instead of a filter feeder (microphagy) is a typical adaption to the deep-sea environment (Gage & Tyler, 1991). However, as noted above, feeding on bacteria 295 and/or food particles in the water column, subject to their seasonal availability, cannot be excluded. Thus, the sponge would have maintained its original feeding strategy as a filter feeder and added carniv- ory as a new, supplementary method. The peculiar morphology of this sponge may reflect its adaptation to the extreme, food-poor, deep-sea habitat in which it lives by reducing its main axis to a thin stalk that reaches into currents above the bottom and forming surface extensions through thin-walled spheres, thus providing a maximum chance to catch prey utilising minimal sponge material, i.e. body mass, as possible. Our interpretation of the presence of choano- cyte chambers and canals in C. gigantea, which are not present in other carnivorous sponges such as A. hypogea or Cladorhiza sp., is that Chondrocladia possesses the basic sponge ‘bauplan’ and thus stands at the base of the Cladorhizidae. The affiliation of Cladorhizidae within Porifera has been questioned due to its lack of the *typical' sponge feature (viz. filter feeding using an aquiferous system with choanocyte chambers creating a water current; Vacelet & Boury-Esnault, 1995). Our data show that this feature is certainly present in C. gigantea, and consequently the family certainly belongs to Porifera. Chondrocladia is a binding link between the original microvorous and the specialised carnivorous species which have lost important anatomical characteristics. In contrast to related cladorhizid species the conversion from particle-feeding to carnivory in C. gigantea is not fundamentally linked to the loss of the aquiferous system, but is a functional modifi- cation (and maybe optional use) of existing structures. ACKNOWLEDGEMENTS We are indebted to the Danish Government for financial support of the BIOICE and BIOFAR Expeditions, to Ole S. Tendal (Zoological Museum, University of Copenhagen) for pro- viding sponge material and for always being open 296 MEMOIRS OF THE QUEENSLAND MUSEUM VIG. 12. A, inclusion cell from the main axis containing at least twelve pieces of muscle tissue (m). From the bot- tom right to the top left a succession of increasingly digested pieces can be observed, While the least digested ones still possess their striation, the more digested ones are homogenous inclusions. In the middle of the cell there is a nucleus. Transmission electron microscopy (TEM). B, enlarged view of a piece of muscle least di- gested in this cell. TEM. C, enlarged view of the striation of the muscle fibre. TEM. to discussions, and to Heiko Sahling (GEOMAR, Kiel) for providing the in situ photograph of Chondrocladia sp. We are also grateful to H. Fliigel for technical advice on the TEM and R. Schmaljohann (both Institut für Meereskunde Kiel) on the SEM. Thanks also to J. Hoeg for introducing us to his methods for preparing material for electron microscopy and P.V. Jensen (both Department of Cell Biology, University of Copenhagen) for assistance in histological questions. The manuscript benefitted from the comments of two anonymous reviewers. LITERATURE CITED BARTHEL, D. & DETMER, A. 1990. The spermato- genesis of Halichondria panicea (Porifera, Demospongiae), Zoomorphology 110: 9-15. FINCHER, J.A. 1940. The origin of germ cells in Stylotella heliophilu Wilson (Tetraxonida). Journal of Morphology 67: 175-192, GAGE, J.D. & TYLER, F.A. 1991. Deep-Sea Biology: a nalural history of organisms at the deep-sea floor. (Cambridge University Press: Cambridge). KUBLER, B. 1998. Ultrastrukturelle Untersuchungen an Chondrocludia gigantea (Demospongiae, Porifera). (Diplomarbeit: Kiel). LANGENBRUCH, PF, 1983. Body structure of marine sponges. |. Arrangement of the flagellated chambers in the canal system of Reniera sp. Marine Biology 75: 319-325, LUNDBECK, W. 1905. Porifera. Part Il: Desmaci- donidae (pars.). The Danish Ingolf-Expedition 62) 1-219, ORR, A.P. 1934. On the biology of Calanus fin- marchicus. iV. Seasonal changes in the weight and chemical composition in Loch Fyne. Journal of the Marine Biological Association of the UK 19; 613-632. PAULUS, W. 1989. Ultrastructural investigation of spermatogenesis in Spongilla lacustris and Ephvdatia fluviatilis (Porifera, Spongillidae). Zoomorphology 106: 155-162. CARNIVOROUS NORWEGIAN SPONGE RAYMONT,J,E.G. 1963. Plankton and Productivity in the Oceans, (Pergamon Press: Oxford, LIK). REYNOLDS, E.S. 1963. The use of lead citrate at high pH as an electron-opaque stain in electron microscopy. Journal of Cell Biology 1 7:208-212. SIMPSON, T.L. 1984. The cell biology of sponges. (Springer Verlag: Berlin). TENDAL, 0.5. & BARTHEL, D. 1993. Chondrocladia gigantea (Demospongiae) - the giant clubsponge of the Northeast Atlantic. Deep-Sea Newsletter (20): 12-15. TENDAL, O.S., BARTHEL, D. & TABACHNIK, R.R.: 1993. An enigmatic organism disclosed —and some new enigma. Deep-Sea Newsletter (21): 11-12, TENDAL, 0.5, & SAHLING, H. 1997. Mysterious Chondrocladia(Porifera) refound and reacting to the camera. Deep-Sea Newsletter (26): 16-17. TUZET, O. 1930, Sur la spermatogenése de l'éponge Reniera simulans. Comptes Rendu Societé Biologie 103: 970-972, TUZET, O., GARRONE, R. € PAVANS DE CECATTY, M. 1970, Observations ultrastructurales sur la VACELET, J. & BOURY-ESNAULT, N. VACELET, J4 297 spermatoyenese chez. la Démosponge Aplysilla rosea Schulze (Dendroceratide): une metaplaisie exemplaire. Annales des Sciences Naturelles, Zoologie et Biologie Animale 12: 27-50, 1995. Carnivorous sponges. Nature 373(6512); 333-335. 1996. A new species. of carnivorous sponge (Demospongiae: Cladorhizidae) from a Medilerranean cave. Bulletin de l'Institut Royal des Sciences Naturelles de Belgique; Biologie 66(suppl.): 109-115. VACELET. J, BOURY-ESNAULT, N., FIALA- MEDIONI, A. & FISHER. C,R. 1995. A methano- traphic carnivorous sponge, Nature 377(1569): 296. BOURY-ESNAULT, N. & HARMELIN, J.G. 1994. Hexactinellid cave, a unique deep-sea habitai in the scuba zone. Deep-Sea Research 41(7): 965-973. WITTE, U. 1996, Seasonal reproduction in deep-sea sponges — trizgered by vertical particle flux? Marine Biology 124: 571-581. REMARKS ON THE PALEOECOLOGY AND REEF BUILDING POTENTIAL OF LATE JURASSIC SILICEQUS SPONGES. Memoirs of the Queensland Museum 44: 297. 1999:- In the early Late Jurassic (Oxfordian) siliceous sponges developed extensively. They formed a discontinous siliceous sponge reef belt extending over more than 7000km from New Foundland, Iberia, France, Switzerland, Germany, Poland, Romania to the Caucasian Mountains. Siliceous sponges are no systematic unit but belong to the different taxonomic groups Hexactinellida and the polyphyletic lithistid demosponges. Due to their different organisation and biology, the ecological demands of the different siliceous sponges groups differ remarkably. The two major groups must be carefully distinguished for paleoenvironmental interpretations, | In general, lithistid demosponges are active filter feeding organisms. They feed on nannoplancton mainly bacteria, The bathymetric distribution of demosponges corresponds to a great extent with the bathyimetrie distribution of bacteria. The fairly high preservation potential of rigid demosponges is explained by a high amount of mesohyl-dwelling bacteria, causing rapid calcification after death. Osmotrophy is an important feeding strategy of Hexactinellida. Dissolved organic matter is enriched in deeper water low-energy settings. causing the majority of Hexactinellida to dwell in such habitats, As the mesohyl of Hexactinellida consists of very thin collagenous material, there is hardly any room to harbour bacteria. Tliis easily explains why microbially induced post-mortem calcification of the sponge by microbial autmicrites occurs at a much lower rate so that fossilisation potential is much lower in comparison with rigid demosponges. | The taxonomic composition of fossil siliceous sponge populations is mainly controlled by sedimentation rate, nutrition and hydrodynamics. The dominance of major taxa is strongly influenced by bathymetry, due to changes of hydrodynamics and nutrition along a balhymetrical gradient. However, the quality of substrates, water energy or extreme oligotrophy may strongly modulate bathymetric distribution. O” Porifera, Late Jurassic, Hexactinellida, lithistid Demospongiae, palevecalogy, calcification, fossilisation. Manfred Krautier (email: mamnfred.krautter() geolowie.uni-siuttgart.de), Institut fiir Geologie und Paläontologie der Université, Hereweg 51, D- 70174 Stuttgart, Germany: 1 June 1998. MEMOIRS OF THE QUEENSLAND MUSEUM DISTINCTIVE MIDDLE CAMBRIAN SPONGE-CALCIMICROBE REEFS IN IRAN. Memoirs of the Queensland Museum 44: 298. 1999:- Following the virtual demise of archaeocyaths in the Toyonian stage and consequent collapse of the Early Cambrian archaeocyath-calcimicrobe reef consortium, Middle and Late Cambrian reefs remained generally devoid of metazoan input, being almost entirely microbial. The few exceptions in this interval generally include some minor contribution by spiculate sponges. One such spiculate sponge-calcimicrobe reef system in the Middle Cambrian of northern Iran is distinctive in that the spiculate sponges constitute a major component of the reef framework. The reefs are in units 2 and 3 of the Mila Formation in the eastern Elburz (Alborz) Mountains. The Mila Formation consists of five units, together ranging in age from Middle Cambrian to Early Ordovician. Trilobites permit correlation of unit 2 and the reef-bearing lower unit 3 with the late Middle Cambrian, and upper unit 3 with the Chinese Kushanian (terminal Middle to earliest Late Cambrian) and Changshanian (Late Cambrian) stages. The reefs are well exposed in a road section 3 km north of Shahmirzad.The reefs are constructed by a consortium of the anthaspidellid sponge Rankenella and a presumed variety of microbes including the calcimicrobe Girvanella. Rankenella is otherwise known only from the Ordian-early Templetonian stage of the Northern Territory, Australia.That stage is equivalent respectively to the late Toyonian-early Amgan and Longwangmiaoan-Maozhuangian stages of Siberia and China.Unit 2 comprises fossiliferous interbeds of grey limestone/dolostone and yellow-brown marly shale, with desiccation cracks, bidirectional ripples and probable tempestites and hardgrounds. The stratigraphically lowest known appearance of Rankenella is in upper unit 2, in a single decimetre-thick limestone bed of abundant eocrinoid ossicles. Scattered, widely conical Rankenella are preserved upright in life position, suggesting attachment to a hardground. Sponges, ossicles, trilobites and hyoliths are encrusted by Girvanella, which also forms rafts and onkoids. Texture within this biostromal bed ranges from floatstone-rudstone to Girvanella boundstone, with evidence of microbial and oxea-bearing sponge-body automicrites.The lower, reef-bearing portion of overlying unit 3 is massive, comprising pale grey stacked bioherms of similar texture and composition to the unit 2 Rankenella bed. In this interval, Rankenella adopts the entire range of co-occurring cup shapes from narrowly conical through to explanate. Clotted-peloidal biohermal mud is interpreted as automicrite. Substrate, peribiohermal and overlying sediment is commonly a bioclast rudstone rich in orthide brachiopod valves. Sponges are contributors to bioconstruction in a reef tract toward the top of lower unit 3. Component bioherms of this reef tract are constructed by ramose Rankenella encrusted by thick coatings of Girvanella to form a Rankenella-Girvanella framestone with only minor lime mud pockets. Interstices are rimmed by one to two generations of columnar cement and occluded by coarse equant cement. By comparison with Early Cambrian reefs, Rankenella and Girvanella played the roles of archaeocyaths and calcimicrobes: framework/substrate and encrusting/binding respectively. In many Early Cambrian reefs, however, lime mud represents a much greater component, while calcimicrobes were capable of building massive framework unaided by metazoans. O Porifera, Middle Cambrian, Iran, calcimicrobe, Rankenella, Girvanella, reef, automicrite. Peter D. Kruse (email: Pierre.Kruse@dme.nt.gov.au), Northern Territory Geological Survey, PO Box 2901, Darwin NT 0801, Australia; 1 June 1998. GENETIC CONFIRMATION OF THE SPECIFIC STATUS OF TWO SPONGES OF THE GENUS CINACHYRELLA (PORIFERA: DEMOSPONGIAE: SPIROPHORIDA) IN THE SOUTHWEST ATLANTIC CRISTIANO LAZOSKI, SOLANGE PEIXINHO, CLAUDIA A.M. RUSSO AND ANTONIO M. SOLE-CAVA Lazoski, C., Peixinho, S., Russo, C.A.M. & Solé-Cava, A.M. 1999 06 30: Genetic confirma- tion of the specific status of two sponges of the genus Cinachyrella (Porifera: Demo- spongiae: Spirophorida) in the Southwest Atlantic. Memoirs of the Queensland Museum 44: 299-305. Brisbane. ISSN 0079-8835. In the Caribbean Cinachyrella alloclada and C. apion are readily distinguished by their different spicule types and sizes, and by the presence of buds in the former. In contrast, in the SW Atlantic both species can reproduce by budding, and also have identical chemical profiles in lectins, fatty acids and steroids. Verification of whether C. alloclada and C. apion were different biological species or were morphotypes of a single polymorphic species on the Brazilian coast was undertaken using allozyme electrophoresis. Samples collected in the intertidal zone of Pituba beach, Salvador, Brazil, and studied independently by morphological and allozyme analyses, showed a high congruence between morphology and allozymes, and 11 (of 19) loci were diagnostic of each species. Cinachyrella apion has smooth oxeas of only one size class, protriaenes of two sizes, anatriaenes of one size, small sigmaspires and raphides. Cinachyrella alloclada has smooth oxeas of two or three size classes, protriaenes and anatriaenes with one size, and sigmaspires like those of C. apion. The unbiased genetic identity between the two species was very low (150.28), as often found for congeneric sponge species. The consistent morphological and genetic differences between the two putative species confirm that, in spite oftheir high chemical similarity, they are distinct biological species. This indicates that, at least in these species, evolutionary rates for allozymes and secondary metabolites are clearly unrelated. Porifera, Demospongiae, Spirophorida, Cinachyrella, morphology, allozymes, molecular systematics, Brazil, Cristiano Lazoski, Claudia A.M. Russo & Antonio Solé-Cava (email: sole@centroin.com.br), Departamento de Genética; Instituto de Biologia; Universidade Federal do Rio de Janeiro; Bloco A-CCS-Ilha do Funddo; 21941-490-Rio de Janeiro; Brazil; Solange Peixinho, Departamento de Zoologia; Instituto de Biologia; Universidade Federal da Bahia; Campus de Ondina; 40170-290-Salvador; Brazil; 16 March 1999. Cinachyrella apion (Uliczka, 1929) and C. alloclada (Uliczka, 1929) are common tetillid marine sponges found in the Caribbean and Brazilian coast (Mothes de Moraes, 1980; Riitzler, 1987; Riitzler & Smith, 1992). Because of their ubiquity and abundance, they have been the subject of many chemical and pharmacological studies (Atta et al., 1989; Barnathan et al., 1992a; Bergquist & Bedford, 1978; Kaul et al., 1977; Portugal, 1992; Rodriguez et al., 1997). Although the two species can be separated on the basis of reproductive and spicular characters in Caribbean populations (Rützler & Smith, 1992), their differences are less obvious on the Brazilian coast (Peixinho, unpublished results). For example, reproductive buds, reported by Rützler & Smith (1992) as occurring only in C. apion, are very common in both species in Brazil. Furthermore, samples of Cinachyrella from Brazil, identified on the basis of spiculation as C. apion or C. alloclada, had identical sterol (Rodriguez et al., 1997), lectin (Portugal, 1992) and fatty acid patterns (Jiménez, unpublished results), as well as proteases with the same chromatographic and electrophoretic profiles (Portugal, 1992). Therefore, it became important to verify whether C. apion and C. alloclada on the Brazilian coast comprised two species with a high chemical and reproductive similarity, or whether they represented the product of phenotypic polymorphism or plasticity of one single species. The method of choice for the determination of specific status of sympatric populations is the genetic interpretation of allozyme patterns (Thorpe & Solé-Cava, 1994), a complementary approach to morphology that has been used with great success to identify cryptic species in sponges (Solé-Cava & Thorpe, 1986; Solé-Cava et al., 1991a, 1991b; Bavestrello & Sarà, 1992; 300 Boury-Esnault et al., 1992; Klautau et al., 1994; Muricy et al., 1996). In this paper we compare electrophoretically sympatric populations of C. apion and C. alloclada to verify their specific status. MATERIALS AND METHODS COLLECTION. Fifteen samples each of C. alloclada and C. apion were collected in June 1995 from the intertidal zone at Pituba beach, Bahia, Brazil (13?27'S, 38°26’W). After collection, the presence of buds on each individual was verified, and the specimens were transported to the laboratory, where each one was immediately divided into two parts: one part was fixed in ethanol for morphological analysis, at the Federal University of Bahia; and the other was stored at 20*C for electrophoresis, in the Federal University of Rio de Janeiro. Both parts of each sponge were given the same code, prior to their putative identification, and the genetic and morphological analyses of the samples were performed in different laboratories. The electrophoresis laboratory, therefore, did not have any information as to the species identity of the samples. This blind-analysis helped to minimise any possible bias in the interpretation of genetic patterns, which might be critical given the supposed high similarity between the two species. To verify the consistency ofthe diagnostic loci for discriminating each species, a second collection from the same locality was made in July 1996, consisting of 33 samples of C. alloclada and 66 samples of C. apion. These samples were then analysed for 4 of the 11 diagnostic loci found in the first study. The results of the first and second experiments were merged for the final analysis. ALLOZYME ANALYSES. Horizontal 12.5% starch gel electrophoresis was carried out as previously described for sponges (Solé-Cava & Thorpe, 1986). The buffer systems used were the 0.25M Tris 0.06M citrate, pH 8.0 (Ward & Beardmore, 1977) and the discontinuous 0.03M Tris 0.005M citrate, pH 8.5 (gel), 0.30M borate, pH 8.1 (buffer tank; Poulik, 1957). Twenty enzyme systems were investigated, of which ten: acid phosphatase (Acp; E.C.3.1.3.2); catalase (Cat; E.C.1.11.1.6); esterases (Est; E.C.3.1.1.1); glutamate dehydrogenase (Gdh; E.C.1.4.1.4); hexokinase (Hk; E.C.2.7.1.1); leucine aminopeptidase (Lap; E.C.3.4.11.1); malate dehydrogenase (Mdh; E.C.1.1.1.37); MEMOIRS OF THE QUEENSLAND MUSEUM 6-phosphogluconate dehydrogenase (Pgd; E.C.1.1.1.44); phosphoglucose isomerase (Pgi; E.C.5.3.1.9); and superoxide dismutase (Sod; E.C.1.15.1.1), gave reproducible results for 19 loci. The staining of the gels followed standard procedures (Manchenko, 1994), Genotype frequency data from both species were used to estimate gene frequencies, fits to Hardy-Weinberg equilibrium, and the unbiased gene identity between them (Nei, 1978) using the BIOSYS-1 programme (Swofford & Selander, 1981). MORPHOLOGY. The overall morphology of the sponge was analysed under a binocular microscope, and the presence of porocalices and sub-ectosomal cavities (vestibules sensu Boury-Esnault & Riitzler, 1997) were visualised in histological sections of paraffin-embedded samples. Small pieces of each sponge were boiled in nitric acid to obtain clean preparations for spicule analysis. A qualitative analysis of mounted spicule preparations was made of every individual collected. Furthermore, 30 measurements of length and width of the eight types of spicule were made, using a light microscope in one individual of each putative species. Spicular and morphological nomenclature follow Boury-Esnault & Riitzler (1997). RESULTS ELECTROPHORESIS. Of the 19 gene loci observed, 11 (Acp-2, Acp-4, Cat, Est-2, Est-3, Est-5, Gdh, Lap, Mdh-2, Pgd and Sod-1; Table 1) unambiguously separated the analysed sponges into two groups, which corresponded perfectly well with the species separated by the morphological analyses. These loci were, therefore, diagnostic of each species (sensu Ayala, 1983). Levels of heterozygosity (h) within each population were high: h=0.13 in C. apion and h=0.15 in C. alloclada, as often seen in marine sponges (Solé-Cava & Thorpe, 1989, 1991). No significant deviations of genotype frequencies from Hardy-Weinberg expectations were observed at any of the loci studied (P>0.05; Fisher’s exact test, using a Bonferroni transformation for multiple tests; Lessios, 1992). The unbiased genetic identity (Nei, 1978) observed between the two species was 0.28. GENETIC CONFIRMATION OF CINACHYRELLA 301 Locus | Allele | C alloclada | N C. apion N Acp-1 1 1.00 12 1.00 15 Acp-2 I 1.00 12 0.00 15 | 2 0.00 1.00 Acp-3 l 1.00 12 1.00 15 Acp-4 1 1.00 12 0.00 15 2 0.00 1.00 Cat 1 0.00 13 0.29 14 2 0.00 0.46 3 0.00 0.25 4 1.00 0.00 Est-1 1 1,00 13 1.00 14 Est-2 1 1.00 46 0.00 80 2 0.00 1.00 Est-3 1 1.00 46 0.00 80 2 0.00 1.00 Est-4 1 0.65 13 0.00 14 2 0.00 0.86 3 0.35 0.14 Est-5 1 0.77 31 0.00 46 2 0,23 0.00 g 0.00 1.00 Gdh 1 1.00 46 0.00 80 2 0.00 1.00 Hk 1 0.00 12 0.42 13 2 0.00 0.42 3 0.25 0.12 4 0.75 0.04 Lap 1 1.00 12 0.00 15 2 0.00 1.00 Mah-1 1 0.25 10 0.03 14 2 0.50 0.04 3 0.15 0.89 4 0.10 0.00 5 0.00 0.04 | Mdh-2 1 0.50 11 0.00 14 2 0.50 0.00 3 0.00 1.00 Pgd 1 1.00 12 0.00 15 2 0.00 1.00 Pgi 1 0.00 13 0.31 13 2 0.00 0.04 3 0.46 0.61 | 4 0.00 0.04 5 0.54 0.00 Sod-1 1 0.00 13 0.96 14 2 1.00 0.00 3 0.00 0.04 Sod-2 1 1.00 6 1.00 6 TABLE 1. Cinachyrella alloclada and C. apion. Gene frequencies at the 19 loci studied. N — number of individuals analysed. MORPHOLOGY. The two species had similar overall (round) shape, yellow colour, hispid surface and firm consistency, being virtually indistinguishable on the field. Both species had porocalices, as is typical of the genus, but vestibules were only observed in C. apion. Calcareous precipitates were observed in both species. The spicular composition of C. apion consists of one size class of anatriaenes and oxeas, and two size classes of protriaenes, small sigmaspires and raphides (Table 2). Conversely, the spicular composition of C. alloclada consists of two or three size classes of smooth oxeas, one size class of protriaenes and anatriaenes, which varied in abundance from rare to abundant, and sigmaspires (Table 3). The sizes ofthe sigmaspire microscleres of the two species in Brazil were clearly not different (both had about the same average: 10um), and more like those of C. apion from the Caribbean (Table 2). Reproductive buds were observed on the surface of most specimens of both species. DISCUSSION The presence of 11 diagnostic loci and the consequent very low genetic identity, associated with the sympatry of the samples, clearly demonstrate that the Brazilian C. apion and C. alloclada are reproductively isolated, regardless of their chemical similarity. Therefore, they must be evolving independently and belong, thus, to different biological and phylogenetic species (Mayr, 1981; Cracraft, 1987). The genetic identity (770.28) found between these two species is, in fact, as small as that often found between species belonging to different genera in other invertebrate groups (Thorpe, 1982; Knowlton, 1993; Thorpe & Solé-Cava, 1994). Similarly, high levels of gene divergence have been found for some aster-bearing hadromerid genera and Oscarella (Boury-Esnault et al., 1992; Sarà et al., 1993; Barbieri et al., 1995; Boury-Esnault et al., 1999; Solé-Cava & Boury-Esnault, 1999, this volume). However, further data are necessary before overall generalisations can be made about gene divergence in sponges, and surely before decisions as to the taxonomic rank (above species level) can be directly inferred from genetic 302 MEMOIRS OF THE QUEENSLAND MUSEUM TABLE 2. Spicule micrometry of Cinachyrella apion. Measurements are given in micrometers, as minimum-mean (standard deviation)-maximum. Triaene measurements refer only to the rhabdomes. 30 spicules were measured for each type, except in the case of anatriaenes, which were rare so only 9 spicules were measured. (A = absent). Spicule type Present data us hy Oxeas | length 2217 - 3797(660) - 5478 | 3500 - 4100 - 4600 diameter 21.7 - 65.9(20.9) - 108.7 35-41-45 Oxeas 2 length A A diameter A A Oxeas 3 length A A diameter A A Protriaenes 1 length | 1587 - 3907(1045) - 5761 | 1800 - 3500 - 8000 diameter 8.6 - 16.9(5.3) - 25.9 4-83-10 Protriaenes 2 length | 588 - 1079(186) - 1400 | 400 - 1350 - 1800 diameter 3.6 - 4.3(1.4) - 7.2 1-23-4 Anatriaenes length | 2196 - 2560(222) - 2880 | 1800 - 2900 - 3500 diameter 10.8 - 72.8(1.9) - 14.4 3-46-5 Sigmaspires length 3.4 - 10.0(2.6) - 15.5 12 - /34-16 Raphides 212 - 238(14) - 259 200 - 244 - 270 identities (Solé-Cava & Boury-Esnault, 1999, this volume). Spicule sizes of the Brazilian specimens of C. apion and C. alloclada were similar to those described by Riitzler & Smith (1992) for Caribbean populations. Conversely, samples of C. alloclada were very different from those described by Mothes de Moraes (1980) (see Table 3). Sponges from the SE coast of Brazil identified as C. alloclada by Mothes de Moraes, had much smaller protriaenes and larger anatriaenes than those found by us and by other authors (Uliczka, 1929; Wiedenmayer, 1974; Rútzler & Smith, 1992). This difference might be judged large enough to warrant the creation of a new species. However, it could also be the result of phenotypic polymorphism, since spicule size can be very variable and dependent of environmental conditions (Simpson, 1978; Bavestrello et al., 1993; Schrónberg & Barthel, 1998). Further studies, possibly using genetic markers, should be done on Cinachyrella samples from the same region studied by Mothes-de-Moraes in SE Brazil, to verify their specific status. The very large genetic divergence (1=0.28) observed between C. alloclada and C. apion contrasts markedly with their overall chemical similarity (Atta et al., 1989; Portugal, 1992; Rodriguez et al., 1997). However, this is to be expected, since the evolutionary rates of the enzymes involved in the housekeeping metabolism (like those used on allozyme analyses) are not necessarily related to those of secondary metabolites. Allozyme polymorphisms are genetically based and usually neutral, since they involve small changes in areas of the protein that do not significantly affect its function (Kimura, 1991). For this reason, these polymorphisms are expected to evolve at constant rates and unlinked to environment conditions, which is very important when dealing TABLE 3. Spicule micrometry of Cinachyrella alloclada, Measurements are given in micrometers, as minimum-mean (standard deviation)-maximum. Triaene measurements refer only to the rhabdomes. 30 spicules were measured for each type. (A = absent). Spicule type Present data Riitzler & Smith, 1992 Mothes de Moraes, 1980 Oxeas 1 length 1900 - 1932/26) - 2016 1500 - 3500 - 5900 2644 - 4923 - 6256 diameter 14.4 - 18.2(1.9) - 21.6 20 - 50 - 65 17-38-56 Oxeas 2 length 1144 - 1217(78) - 1440 900 - 1800 - 2800 A diameter 10.8 - 12.8(1.8) - 14.4 1-13-20 A Oxeas 3 length 756 - 837(81) - 1008 100 - 355 - 950 74- 159-223 diameter 7.2 - 7.2(0) -7.2 2.5-54-8 4-7-9 Protriaenes 1 length 1296 - 2164(423) - 3197 2400 - 4200 - 6500 407 - 437 - 462 diameter 3.6 - 4.6(1.6) - 7.2 4-10.7 -20 3-11-23 Protriaenes 2 length A A A diameter A A A Anatriaenes length 1051 - 7225(102) - 1440 2200 - 3200 - 4000 7259 - 8289 - 9061 diameter 4.0 - 7(2.0) - 14.0 3-83-14 5-9-14 Sigmaspires length 7.0 - 10.1(1.3) - 11.2 10 - 74.3 - 23 10- 74-22 Raphides A A A GENETIC CONFIRMATION OF CINACHYRELLA 303 with problems on taxonomic status of organisms. On the other hand, studies with secondary metabolites necessarily deal with the products of enzyme lunction. which are much more constrained by natural selection and therefore are not expected to evolve in a clock-wise manner (Nei, 1987), Lectins and secondary metabolites are often involved in strong inlerspecies interactions, playing a very important role in the survival of the species (Kennedy el al, 1995; Hirabayashi & Kasai, 1998). Therefore. most of the lime they are likely to be under strong normalising or directional selecüon, which are highly dependent on environmental conditions. This may explain, for example, why steroids in three species of Cinachyrella trom W Africa are not overtly different except for small variations in their proportions (Barnathan et al., 1992b), and the presence of cholest-4-en-3-one steroids, both in the Brazilian Cinachvrella (Rodriguez, et al., 1997), and in C. tarentina (Aiello et al. 1991), A high intrageneric similarity has also been observed in sterols of the genus Aplysina (Kelecom & Kannengiesser, 1979) and in lectin composition of dxinella spp. (Bretting et àl., 1980). Structures or molecules of adaptive importance and, hence, under strong normalising selection can also present large evolutionary shifis due ta directional selection. These shifts, albeit rare, can generate homoplasy by evolutionary convergence, It is not uncommon, for example, to find species from different phyla that share a secondary metabolite which is not found in other members of the same tamily or order (Tursch et al., 1978). At the species level, therefore, secondary metabolites are not indicated for phylogenetic studies, since they can be evolutionarily too conserved at that level (i.e. plesiomorphie) or prone to homoplasy due to convergence, At this taxonomie level. thus, the use of allozymes or neutral DNA genes is more indicated (Hillis et al., 1996). The two Brazilian species of Cinachyrella can be positively identified as the Caribbean C. apien and C. alloclada, based on morphological data. However, the use of molecular systematics for the study of sponge species has unveiled a large amount of hidden biodiversity in the phylum {reviewed in Solé-Cava & Thorpe, 1994), especially when applied to so-called cosmopolitan species and many of those with very large and/or disjunct geographical distributions. In fact, all allegedly cosmopolitan sponge species studied to date by allozyme electrophoresis have been shown to consist of clusters of morphologically similar, but evolutionarily distinet species, This means that levels of endemism in marme sponges may be much larger than assumed by taxonomists using only morphometric data (Solé-Cava et al., 199 La, 1992; Klautau ct al., 1994; Klautuu et al.. in press), Ih would be very mteresting, therefore, to verily whether C^ alloolada and C. apion from the Bermudas, Senegalese coast (Barnathan et al. 1992a), Bravil (including the samples from Bahia studied here), and those from SE Brazil (Mothes de Moraes. 1980) are indeed conspecific. ACKNOWLEDGMENTS We thank Miguel Couceiro, Renata Lellis, and Paulo Vianna for help in the collection of samples and in the electrophoretic work. We are also thankful to Walter Cerqueira and Carla Stringetti for help in the morphological analyses and to Nicole Boury-Esnault for suggestions on the manuscript. This work was supported by prants from CNPq, FAPERJ, and FUJB (Brazilian Government). LITERATURE CITED AIELLO, A., FATTORUSSO, E, MAGNO, S$., MENNA, M, & PANSINI, M, 1991, Steroids of Ihe marine sponge Cinachyra tarentina: isolation ol cholesi-4-ene-3,6-dione and (24R)-24-ethylcholest-4-ene-3,6-dione. Journal of Natural Products (Lloydia) 54(1): 281-285, ATTA, A.M., BARRAL NETTO, M , PEIXINIO,S de SOUSA ATTA, M.L.B. 1989. Isolation and Functional characterization of a mitogenic lectin from the marine sponge Cinachyrella alloclada. Brazilian Journal of Medical and Biological Research 22(3): 379-385. AYALA. F.J. 1983. Enzymes as taxonomie characters. Pp, 3-26. In Oxford, G.S. & Rollinson, D. (eds) Protein polymorphism: adaptive and taxonoinic significance, (Academic Press: London). BARBIERI, M, BAVESTRELLO, G. & SARA, M. 1995. Morphological and ecological differences in two electrophoretically detected species of Cliona (Porifera, Demospongiae)- Biological Joumal of the Linnean Society 54: 193-200. BARNATHAN, C., MIRALLES, J.. GAY DOU, E.M., BOURY-ESNAULT, N: & KORNPROBST. LM. 19924 New phospholipid fatty-acids from the marine spunge Cinachyrella alloclada Uliczka. Lipids 27( 10): 779-784. BARNATHAN, G., MIRALLES, 1, NJINKOUE, J.M., MANGONI, A,, FATTORUSSO, E., DEBITUS, C., BOURY-ESNAULT, N. £ KORNPROBST, J.M. 1992b. Sterol campasition of 3 marine sponge species from Ihe genus Cinachyrella. Comparative Biochemistry and Physiology B - 304 Biochemistry and Molecular Biology 103(4): 1043-1047, i BAVESTRELLO, G., BONITO, M. & SARA, M. 1993. Silica content and spicular size variation during an annual cycle in Chondrilla nucula Schmidt (Porifera, Demospongiae) in the Ligurian sea. Pp. 421-425. In Uriz, M.-J. & Rützler, K. (eds) Recent advances in ecology and systematics of sponges. (Instituto de Ciencias del mar, C.S.I.C: Barcelona). . BAVESTRELLO, G. & SARA, M. 1992. Morphological and genetic differences in ecologically distinct populations of Petrosia (Porifera, Demospongiae). Biological Journal of the Linnean Society 47: 61-77. BERGQUIST, P.R. & BEDFORD, J.J. 1978. The incidence of antibacterial activity in marine Demospongiae: systematic and geographic considerations. Marine Biology 46: 215-221. BOURY-ESNAULT, N., KLAUTAU, M., BEZAC, C., WULFF, J. & SOLE-CAVA, A.M. 1999, Comparative study of putative conspecific sponge populations from both sides of the Isthmus of Panama. Journal of the Marine Biological Association of the United Kingdom 79:39-50. BOURY-ESNAULT, N. & RUTZLER, K. (eds) 1997. Thesaurus of sponge morphology (Smithsonian Contributions to Zoology, Smithsonian Institution Press: Washington, D.C.). BOURY-ESNAULT, N., SOLE-CAVA, A.M. & THORPE, J.P. 1992. Genetic and cytological divergence between colour morphs of the Mediterranean sponge Oscarella lobularis Schmidt (Porifera, Demospongiae, Oscarellidae). Journal of Natural History 26: 271-284. BRETTING, H., STANISLAWSKI, E., BECKER, W. & KANIGSMANN, K. 1980. A comparative study of snail galactans with the sponge lectins of Axinella polvpoides revealing some structural peculiarities of the Helix pomatia galactan. Comparative Biochemistry and Physiology 65B: 497-503. CRACRAFT, J. 1987. Species concepts and the ontology of evolution. Biology and Philosophy 2(3): 329-346. HILLIS, D.M., MORITZ, S. & MABLE, B.K. 1996. Molecular Systematics, Second edition. (Sinauer Associates: Massachusetts). HIRABAYASHI, J. & KASAI, K.-I. 1998. Evolution of animal lectins. Pp. 45-88. In Müller, W.E.G. (ed.) Molecular evolution: evidence for monophyly of Metazoa. (Springer-Verlag: Berlin). KAUL, P.N., KULKARNI, S.K., WEINHEIMER, A.J., SCHMITZ, F.J. & KARNS, T.K.B. 1977. Pharmacologically active substances from the sea. II. Various cardiovascular activities found in the extracts of marine organisms. Lloydia 40: 235-268. KELECOM, A. & KANNENGIESSER, G.J. 1979. Chemical constituents of Verongia sponges. 1 - A comparison between Brazilian and Mediterranean MEMOIRS OF THE QUEENSLAND MUSEUM species. Anais da Academia Brasileira de Ciéncias 51: 633-637. KENNEDY, J.F., PALVA, P.M.G., CORELLA, M.T.S., CAVALCANTI, M.S.M. & COELHO, L.C.B.B. 1995. Lectins, versatile proteins of recognition: a review. Carbohydrate Polymers 26: 219-230. KIMURA, M. 1991. The neutral theory of molecular evolution: a review of recent evidence. Japanese Journal of Genetics 66(131): 367-386. KLAUTAU, M., SOLE-CAVA, A.M. & BOROJEVIC, R. 1994, Biochemical systematics of sibling sympatric species of Clathrina (Porifera: Calcarea). Biochemical Systematics and Ecology 22: 367-375. KLAUTAU, M., RUSSO, C., LAZOSKI, C., BOURY-ESNAULT, N., THORPE, J.P., SOLE-CAVA, A.M. 1999, Does cosmopolitanism in morphologically simple species result from overconservative systematics? A case study using the marine sponge Chondrilla nucula. Evolution (in press). KNOWLTON, N. 1993. Sibling species in the sea. Annual Review of Ecology and Systematics 24: 189-216. LESSIOS, H.A. 1992. Testing electrophoretic data for agreement with Hardy-Weinberg expectations. Marine Biology 112: 517-523. MANCHENKO, G.P. 1994, Handbook of detection of enzymes on electrophoretic gels. (CRC Press Inc.: Ann Arbor, Michigan). MAYR, E. 1981. Biological classification: toward a synthesis of opposing methodologies. Science 214: 510-516. MOTHES DE MORAES, B. 1980. Esponjas de Sao Sebastiáo, litoral de Sáo Paulo, Brasil (Porifera, Demospongiae). Iheringia 56: 75-86. MURICY, G., SOLE-CAVA, A.M., THORPE, J.P. & BOURY-ESNAULT, N. 1996. Genetic evidence for extensive cryptic speciation in the subtidal sponge Plakina trilopha (Porifera, Demospongiae, Homoscleromorpha) from the Western Mediterranean. Marine Ecology-Progress Series 138(1-3): 181-187. NEI, M. 1978. Estimation of average heterozygosity and genetic distance from a small number of individuals. Genetics 89: 583-590. 1987. Molecular evolutionary genetics. (Columbia University Press: New York). PORTUGAL, K. 1992. Isolamento de lectinas e avaliacáo de proteases das esponjas marinhas Cinachyrella apion e Cinachyrella alloclada - uma análise comparativa. (Monograph of the Universidade Federal da Bahia Press: Salvador, Brazil). POULIK, M.D. 1957. Starch gel electrophoresis in a discontinuous system of buffers. Nature 180: 1477-1479. RODRIGUEZ, J., NUNEZ, L., PEIXINHO, S. & JIMENEZ, C. 1997. Isolation and synthesis ofthe first natural 6-hydroximino- 4-en-3- one- steroids GENETIC CONFIRMATION OF CINACHYRELLA from the sponges Cinachyrella spp. Tetrahedron Letters 38: 1833-1836. RÜTZLER, K. 1987. Tetillidae (Spirophorida, Porifera): a taxonomic revaluation. Pp. 187-203. 1n Vacelet, J. & Boury-Esnault, N. (eds) Taxonomy of Porifera from the N.E, Atlantic and Mediterranean Sea. (Springer-Verlag: London). RUTZLER, K, & SMITH, K.P. 1992, Guide to western Atlantic species of Cinachyrella (Porifera: Tetillidae). Proceedings of the Biological Society of Washington 105: 148-164. SARA, M., CORRIERO, G. & BAVESTRELLO, G. 1993. Tethya (Porifera, Demospongiae) species coexisting in a Maldivian coral reef lagoon - taxonomical, genetic and ecological data. Marine Ecology 14(4): 341-355. SCHRÓNBERG, C.H.L. & BARTHEL, D. 1998. Unreliability of Demosponge skeletal characters: the example of Halichondria paniceu. Pp. 41-54. In Watanabe, V. & Fusetani, N. (eds) Sponge Sciences. (Springer- Verlag: Tokyo). SIMPSON, T.L. 1978. The biology of the marine sponge Microciona prolifera (Elis and Solander). 3. Spicule secretion and the effect of temperature on spicule size. Journal of Experimental Marine Biology and Ecology 35(1): 31-42. SOLÉ-CAVA, A.M. & BOURY-ESNAULT. N. 1999, Patterns of intra and interspecific divergence in marine sponges. Memoirs of the Queensland Museum, this volume. SOLE-CAVA, A.M., BOURY-ESNAULT, N., VACELET, J. & THORPE, 1992. Biochemical genetic divergence and systematics in sponges of the genera Corticium and Oscarella (Demospongiae: Homoscleromorpha) in the Mediterranean Sea. Marine Biology 113(2): 299-304, SOLE-CAVA, A.M., KLAUTAU, M., BOURY-ESNAULT, N.. BOROJEVIC, R. & THORPE, LP, 199]a, Genetic evidence for cryptic speciation in allopatric populations of two cosmopolitan species of the calcareous sponge genus Clathrina. Marine Biology 111: 381-386. SOLE-CAVA, A.M, & THORPE, J.P. 1986. Genetic differentiation between morphotypes of the 305 marine sponge Suberltes ficus (Demospongiae: Hadromerida). Marine Biology 93(2): 247-253. 1989, Biochemical correlates of genetic variation in marine lower invertebrates. Biochemical Genetics 27(5-6): 303-512 1991, High levels of genetic variation in natural populations of marine lower invertebrates. Biological Journal of the Linnean Society 44: 65-80. 1994. Evolutionary genetics of marine sponges. Pp. 55-63. In Soest, R. W.M, van, Kempen, TM.G van & Braekman, J.C. (eds) Sponges in Time and Space. (Balkema: Rotterdam). SOLE-CAVA, A.M., THORPE, J.P. & MANCONI, R. 1991b. A new Mediterranean species of Axinellu detected by biochemical genetic methods. Pp. 313-321. In Reitner, J. de Keupp, H. (eds) Fossil and Recent Sponges. (Springer: Berlin). SWOFFORD, D.L. & SELANDER, R.B. 198I. BIOSYS-1, a FORTRAN programme for the comprehensive analysis of electrophoretic data in population genetics and systematics. Journal of Heredity 72: 281-283. THORPE, J.P. 1982. The molecular clock hypothesis: biochemical evolution, genetic differentiation and systematics. Annual Review of Ecology Systematic 13: 139-168. THORPE, J.P. & SOLE-CAVA, A.M. 1994. The use of allozyme electrophoresis in invertebrate systematics. Zoologica Scripta 23(1): 3-18. TURSCH, B., BRAEKMAN, J.C., DALOZE, D. & KAISIN, M. 1978. Terpenoids from coelenterates. Pp. 247-296. [n Scheuer, J.P. (ed.) Marine Natural Products. (Academic Press: Londan). ULICZKA, E. 1929. Die tetràxonen Schwámme Westindiens (auf Grund der Ergebnisse der Reise Kükenthal Hartmeyer). Zoologische Jahrbücher 16: 36-62. WARD, R.D. & BEARDMORE, J.A. 1977. Protein variation in the plaice (Pleuronectes platessa), Genetic Research 30: 45-62. WIEDENMAYER, F. 1974. Recent marine shallow water sponges of the West Indies and the problem of speciation. Verhandlungen der Naturforschenden Gesellschaft in Basel 84: 361-375. 306 MEMOIRS OF THE QUEENSLAND MUSEUM HOMEOBOX GENES EXPRESSED IN THE ADULT AND REAGGREGATING SPONGE. Memoirs of the Queensland Museum 44: 306. 1999:- Homeobox genes encode a large family of conserved transcription factors that control a range of important developmental decisions, and can be interspersed or organised in clusters (e.g. HOX genes) within metazoan genomes. A striking feature of many homeobox orthologues is how they are expressed in a similar fashion during the development of phylogenetically-disparate animals. In recent years, interest has developed in homeobox genes in the lower metazoans, such as platyhelminths, cnidarians, and sponges. Instead of studying embryological development, focus has often been on the role of homeobox genes during regeneration. Interestingly, these genes exhibit comparable expression patterns during the regeneration of lower metazoans as they do during normal embryological development of higher metazoans. In this study, homeobox genes expressed in the adult sponge (newly dissociated cells) /otrochota baculifera (Demospongiae: Poecilosclerida: Myxillidae) and during reaggregation were identified by RT-PCR with degenerate primers and sequencing. As with regeneration in other taxa, investigating the molecular genetics of reaggregation in the sponge, apart from providing many practical advantages, gives us insight into the level of conservation between embryological and other developmental processes. To confirm the poriferan origin of the homeobox genes isolated, the identical procedure was applied to another sponge and some homologous genes were obtained. As sponges appear to be monophyletic with the rest of the Metazoa and the first lineage to diverge during metazoan evolution, the study of their homeobox genes provides insight into the initial role of metazoan-specific homeobox genes in governing the multicellular state and cellular differentiation. CJ Porifera, reaggregation, development, homeobox genes, regeneration. Claire Larroux & Bernard M. Degnan (email: bdegnan@, zoology.ug.edu.au), Department of Zoology, University of Queensland, Brisbane, Old., 4072, Australia; 1 June 1998. ANTIMICROBIAL ACTIVITY OF SPONGES FROM SOUTHERN BRAZIL, ATLANTIC COAST. Memoirs of the Queensland Museum 44: 306. 1999:- Within the research program of biologically active natural products, we have investigated different species of marine sponges collected by scuba diving in the South-western Atlantic region, near the coast of South-Brazil, aiming at the evaluation of their potential as a source for new drugs. In order to achieve this goal, the antimicrobial activity of five species, Tedania ignis Duchassaing & Michelotti, Pseudaxinella reticulata (Ridley & Dendy), Polymastia janeirensis (Boury-Esnault, 1973), Batzella sp. and Petromica sp. were analysed. The sponge species found in this particular region constitute a group of shallow-water Demospongiae in which a pharmacological usage has not previously been evaluated. Aqueous extracts obtained by grinding and maceration for 30mins following freeze-drying, as well as organic solvent extracts (toluene: methanol 3:1 v/v) were prepared from organisms frozen since the harvesting. Antimicrobial activity was evaluated by the agar diffusion method on paper disks (6mm diameter). Each sponge extract was tested for growth inhibition of five bacteria species: Escherichia coli (ATCC 25922), Staphylococcus aureus (ATCC 6538P), Staphylococcus epidermidis (ATCC 12228), Baccillus subtilis (ATCC 6633), Micrococcus luteus (ATCC 9341) and two yeast species: Candida albicans (ATCC 110231) and Saccharomyces cerevisiae (ATCC 1600). None of the tested extracts analysed by this method could inhibit the growth of these microorganisms, an exception was the Petromica sp organic extract which showed an inhibitory activity for Bacillus subtilis. The same extracts were analysed for antitumour activity by in vitro inhibition of proliferation of different tumour cell lines. Some of the extracts showed promising results. The preliminary antimicrobial assay can be useful for pointing out sponge species to be further analysed. O Porifera, antimicrobial, Atlantic coast, southern Brazil, biologically active natural products, antitumor. C. Lerner (email: cblerner@portoweb.com. br), Instituto de Biociéncias da Universidade de SGo Paulo (USP) & Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP), Brazil; and Museu de Ciéncias Naturais da Fundação Zoobotánica do Rio Grande do Sul (MCN/FZB) & Fundação de Amparo a Pesquisa do Estado do Rio Grande do Sul (FAPERGS), Brazil; B. Mothes, L.G. Possuelo & J.C. Costa, Museu de Ciéncias Naturais da Fundação Zoobotánica do Rio Grande do Sul (MCN/FZB) & Fundação de Amparo a Pesquisa do Estado do Rio Grande do Sul (FAPERGS), Brazil; E.E.S. Schapoval, E.Rech, F.M. Farias & A.T. Henriques, Faculdade de Farmácia da Universidade Federal do Rio Grande do Sul (UFRGS) & Fundação de Amparo à Pesquisa do Estado do Rio Grande do Sul (FAPERGS) / Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq), Brazil; D. Mans, South-American Office for Anticancer Drug Development (SOAD), Brazil; and Hospital de Clinicas de Porto Alegre (HCPA), Brazil; 1 June 1998, DISTRIBUTION PATTERNS OF SPONGES AND CORALS DOWN TO 107 M OFF NORTH JAMAICA HELMUT LEHNERT AND HAGEN FISCHER Lehnert, H. & Fischer, H. 1999 06 30: Distribution patterns of sponges and corals down to 107m off North Jamaica. Memoirs of the Queensland Museum 44: 307-316. Brisbane. ISSN 0079-8835. Sixty species of sponges (23 new) were collected from the deep fore-reef (60-107m depth) off the North Jamaican Discovery Bay area using trimix diving. Comparison with the shallow water sponge fauna shows only 15% of shallow water sponges extend down to the deep fore-reef and 60% of deep fore-reef sponges are not found in shallow water. Mapping sponge and coral distributions around Discovery Bay to 40m depth revealed a database of 102 sites with a surveyed area of 1659m?. Multivariate analysis of this database recognizes three large scale habitats: Reef-surfaces, lagoon, and undersides of platy corals. Separate analyses of subsets indicate internal differences within habitats. Benthic colonization on reef-surfaces are continuous along depth and inclination gradients, except around river mouths. Within lagoon habitats there are subhabitats: blue hole, Thalassia seagrass-beds, ridges with freshwater outflow and protected (eastern) backreef. Zonation of Jamaican reefs appears to have changed over 34 years in comparision to data of Goreau(1959).0 Porifera, distribution patterns, depth zonation, habitat specialisation, Jamaica, coral reefs. Helmut Lehnert* (email: Helm.Lehnert@t-online.de), Institut & Museum fiir Geologie & Paldontologie, Goldschmidtstr. 3, 37077 Góttingen, Germany; Hagen Fischer, Institut für angewandte ökologische Studien, Hessestr. 4, 90443 Nürnberg, Germany; *Present address: Kónigsbrunnerstr. 9b, 86507 Oberottmarshausen, Germany; 15 May 1999. Zonation patterns of corals have been studied by several workers (Goreau, 1959; Geister, 1977). Several authors (Liddell & Ohlhorst, 1987; Rützler, 1971, 1974, Wilkinson & Cheshire, 1989; Wilkinson & Evans, 1989; Zea, 1993) mentioned the importance of sponges in coral reefs, only few (Alcolado, 1990; Alvarez et. al., 1990; Diaz et. al., 1990; Schmahl, 1990) have attempted to describe the zonation of sponges, probably due to taxonomic difficulties within this group (Rützler, 1987; Bóger, 1988, Van Soest, 1991). The first extensive study of Jamaican sponges was attempted by Hechtel (1965). He investigated the area of Port Royal on the Jamaican south coast, recording 57 species and listing the common species in each of his ten collecting areas. Few details were given on the nature ofthese localities but he did mention that a considerable number were restricted to certain habitats. This is surprising considering that his survey extended down to only 6.1m (20ft) depth. This implies sponges may have strong zonation patterns. Previous studies on zonation of coral reefs recognized more intuitive morphological differ- ences in reefs and based their zonation only in part on the sessile organisms occurring there, mainly on scleractinian corals. Geister (1977) wrote that in Caribbean reefs there was a “distinct coral zonation controlled by exposure to wave activity. Based on this zonation, six basic reef types can be distinguished, ..." But he admitted that “Influence of factors other than wave exposure, however, may considerably disturb the regular zonation pattern”. Geister (1983) gave an excellent overview about reef definitions, classifications and geological aspects of recent reefs, but worked on relatively large temporal and spatial scales. The present paper is based on mapping of species from selected sites and subsequent multi- variate analyses. The mapping of sponges and corals provides an estimate of the importance of sponges compared to corals. The aim of this study is to determine if similar species communities occur on different sites investigated, and if these similarities can be explained by environmental factors. MATERIALS AND METHODS Thirteen trimix dives to depths between 60- 107m were undertaken in Discovery Bay, Jamaica in May-June 1993 and June-July 1996, to collect and photograph sponges of the deep fore-reef. 308 MEMOIRS OF THE QUEENSLAND MUSEUM TABLE 1. Outline of classification undertaken on data, using methods proposed by Wildi (1989). Program Action Comment TRAFOA (PPS) Histogram equation (Fischer, 1994) Compensation of the extreme right-skewed distribution of population data. Species selection by variance. Only species with highest variance, with total variance of 95% of the whole data set, were selected to reduce the data set and to reduce “noise”, INIT (MULVA) Transform attribute vectors to unit length. Avoid undesired effects caused by unequal species variance RESE (MULVA) Calculate similarity matrix using similarity ratio. Recommended for sites CLTR (MULVA) Classify sites with minimum variance classification, Minimizes in-group variance and maximizes be- tween-group variance INIT (MULVA) Transform site vectors to unit length. Avoid undesired effects caused by unequal species numbers in sites RESE (MULVA) | Calculate dissimilarity matrix using chord distance. Recommended for species CLTR (MULVA) Classify species with minimum variance classifica- Minimizes in-group variance and maximizes be- tion, tween-group variance Analysis of variance (Jancey‘s F-rank see Wildi, "n “e - DIAN (MULYA) PETATA 1990) età ETA (a-194) q ! Select only significantly different species INIT (MULVA) Transform data for correspondance analysis Required for correspondence analysis RESE (MULYA) Calculate similarity retin non centered cross Is required for correspondence analysis PCAB (MULVA) Compute correspondence analysis Normal version choosen AOCL (MULVA) Analysis of concentration Ordination of species and site- groups to get meaningful sequences EDGR (MULVA) Rearrange groups reta according to correspon- | Obtain meaningful within group sequence of species and ence analysis sites TABS (MULVA) Display the ordered table > Between January-July 1993 coral reefs in the Discovery Bay area, from the mouth of the Rio Bueno in the west to the mouth of the Pear Tree River in the east, were mapped using SCUBA, using the method described by Braun-Blanquet (1964). This method, originally developed for botanical surveys, was used to estimate the abundance and percentage cover of sponge and coral species in different habitats of reefs. Site surface area was measured with a plastic tape measure, and percentage cover of sessile species was estimated using the following procedure: r=one individual specimen in the site surveyed; +=cover below 1%; l=cover below 5%; m-cover below 5%, but species abundant; a=cover 5-15%; b=cover 15-25%; 3=cover 25-50%; 4=cover 50-75%; 5=cover 75-100%. Large differences in size in both habitat and species occurrence (especially within sponges) necessitated adjustment of the size of sites according to prevailing conditions. Some habitats (e.g. undersides of platy corals), were very limited in their extent, whereas habitats like Thalassia sea-grass beds inside the lagoon, in shallow water, were far more extensive. Another factor limiting size of sites was decreasing bottom-time with increased depth. Evaluation of these data was performed using MULVA (Wildi & Orloci, 1990), CANOCO (ter Braak, 1988, 1990) and PPS (Fischer, 1994), Classification of data was made using the standard strategy for the analysis of phytosocio- logical data, suggested by Wildi (1989) with some modifications. Table 1 summarizes the analysis path, Ordination was performed with CANOCO (ter Braak 1988, 1990) based on redundancy analysis (RDA), analyzing the influence ofenvironmental factors on the fauna and providing graphical representation of the data. CANOCO offers two methods for canonical analysis: RDA and canonical correspondence analysis (CCA). CCA is preferable if the data set demonstrates large B-diversity, (i.e. if it contains several very different habitat types with very few or no species occurring in all of these types). RDA, in contrast, is applicable for small 8-diversity. Our sites were recorded from a geographically small area from similar habitats. Several species were found in most of these habitats. Consequently, RDA is the preferable method for our data set. Graphical representation of canonical ordination (RDA) depicts similarity in distance between sites based on their faunistic and ecological affinities. Metric JAMAICAN SPONGE DISTRIBUTIONS 309 TABLE 2: Jamaican corals and sponges. Columns represent site groups. The numbers are percentage frequency of the species. Site groups and species are arranged according to classification and correspondence analysis. Species groups (Sp. Gps) are indicated for each species. Site-groups are sites with similar species, co-occurring species are species-groups. 85 originally mapped species erent underwent an analysis of variance (see Table 1). The displayed 36 species have significantly di occurrence. Species not displayed are either very rare or run through all or most sites. | Site-Group | 4 3 8 a | 6 7 10 9 1 Number Of Sites 3 3 12 y 22 13 22 9 3 6 Mean Number Of Species 3 3 12 12 16 12 10 9 2 4 Id Sp. LI SPECIES NAME Gps 40 Plakortis simplex (olive) V 100 17 5 12 Clathrina primordialis 17 33 100 9 Unidentified demosponge A 100 85 Stylaster roseus 16 100 22 23 5 olive incrusting 16 100 67 | 84 Helioseris cucullata 10 33 5 28 Agelas sceptrum 10 67 75 11 Ectyoplasia ferox 9 50 32 9 11 | 61 | Agaricia agaricites var.unifaciata| 11 92 56 59 31 55 | 60 | Montastrea cavernosa 11 67 11 50 38 32 22 58 Acropora cervicornis 4 8 22 55 15 14 25 Agelas dispar 3 25 | 89 82 38 50 11 71 Siderastrea radians 3 8 33 73 31 45 11 41 Erylus formosus 2 17 11 55 8 11 81 Millepora complanata 19 8 11 14 85 9 11 57 Acropora palmata 19 23 35 lotrochota birotulata 6 8 45 15 55 33 10 Niphates erecta 6 11 27 8 45 44 | 13 | Ircinia strobilina 6 17 67 77 31 36 33 paña Diploria clivosa 20 46 9 44 43 Anthosigmella varians 20 25 18 54 27 56 var.incrustans 44 Chondrilla nucula 20 33 5 54 5 44 1 Neofibularia nolitangere 5 9 8 45 22 79 Porites furcata ] 17 89 82 69 23 44 17 78 Porites astreoides 1 58 100 TI 92 32 56 33 62 | Agaricia agaricites var.massiva | 1 33 44 64 54 14 22 59 Montastrea annularis 1 83 78 100 92 59 22 ga | Millepora alcicornis 1 17 56 59 38 18 11 15 Aiolocroia crassa 1 42 44 82 31 32 11 46 Aka coralliphaga 1 8 78 9 38 9 68 Diploria labyrinthiformis 1 32 54 72 Siderastrea siderea 12 | 17 9 8 9 78 49 Xestospongia carbonaria 18 100 | 67 |_ 54 | Haliclona coerula 18 li 50 48 Myrmekioderma rea 18 8 11 33 83 30 Amphimedon erina 18 8 5 22 100 environmental variables are displayed as arrows, displayed as points (indicating the center of the indicating the direction of average increase of occurrence of each category). The scores on the each variable, whereas categorical variables are axis represent relative distance units in the 310 RDA All sites ]. versus 2. axis lagoon Ll AS Er Ia ua ac we um aa -L. qoe "uns DJ if i Li t 4i y A > LF -t 6 u o 1 ole 2 ' numerical varmbles DE pmo omo nur CRI vmm reef. habiral FIG. 1. Redundancy Analysis (RDA) of all sites. Similarity of siles is displayed as distance. Lagoonal and reef sites are differentiated on the first two axes with lagoonal sites on the upper left and reef-sites with grey background. Note thai there is a mixing zone between lagoonal and reef-sites, indicating sites influenced by both environm Numerical variables are shown as arrows. Differences in sizes of arrows reflect different influence of environmental variables, Ordinal variables are printed as centroids without direction but influence sites nearby. Numbers refer to site-numbers, similarity matrix from the center of the data set. The following environmental variables were used in these analysis: Depth, size of site, total cover, sponge cover, coral cover, algae cover, inclination of substrate, ridges with freshwater outflow, Thalussía seagrass, backreef, bluehole, ship channel, fore-reef, deep fore-reef, pinnacle, undersides of platy corals, reel-flat, Discovery Bay, Rio Bueno, Pear Tree River, sediment cover, coral rubble. Ana priori selection of environmental variables was carried out, Only variables with p<0.05 were retained for further analysis, to ensure that only statistically significant variables influenced the analysis. 1 Ace PR ni RC mo ——À————Ó piia i 5e nt 34 MEMOIRS OF THE QUEENSLAND MUSEUM RESULTS COMPARISON BETWEEN SHALLOW WATER (0-40M) AND DEEP FORE-REEF (60-107M) SPONGES. The relatively well known Jamaican shallow-water sponge fauna consists of now 157 sponge species (Lehnert & Van Soest, 1998). 133 species (85%) are restricted to shallow water, 5 of them were new to science, and 24 species (15%) also occur in the deep fore-rect, From ihe deep fore-reef 60 sponge species were collected trom 13 trimix dives. 23 of these are new (Lehnert & Van Soest, 1996, m press), and with 13 known species a total of 36 species (60%) arc restricted to the deep fore-reef, 24 species (40%) are shared with shallow water habitats. The inventory of deep fore-reef sponges is far from being complete, with additional undescribed and described species expected. However, there are striking differences in species composition between deep water and shallow water habitats, and itis improbable that any additional deep fore-reef species will be found in the well known shallow water fauna. INTERNAL ANALYSIS OF THE SHALLOW WATER DATA SUBSETS. Classification. Table 2 shows the results of the classification abtained by thc analysis outlined in Table 1, indicating ten groups of sites attributed to three large-scale habitats. Site-groups 3 & 4 are from undersides of platy corals with the characteristic species- groups 10 & 16, Lagoonal environments are represented by site-groups 1, 2, 9 & 10. Characteristic lagoonal species include species- groups 14 & 18, restricted to lagoonal environments. Many species frequently oceur within the lagoon, but have their focal points within reef-environments, like species-groups 1. 2,3,6, 1L and 12, The remaining site-groups 5, 6, 7 and 8 are from different reef-environments. Species-group 6 is most abundant in general reef environments. Consequently these species can be used for large scale characterization of habitats only, but not for subhabitats. Ordination analysis using RDA was performed ona square-root transformation ofthe percentage substrate cover values, with centered and normalized species, The following environmental variables were statistically significant (p«0.05): Depth, algae cover, inclination of substrate, ridges with freshwater outflow, Thalassia seagrass, 4 "be" ents. JAMAICAN SPONGE DISTRIBUTIONS RDA a2 All sites ]. versus 3, axis undersides of « |platy corals — m. de 32 ED Ei N a TT oop o es TT o pitiiiitiiiy poiitispiitisiitiiiitiiii firey =2 il 0 1 2 3 4 5 FIG. 2. Redundancy Analysis (RDA) of all sites. Representation of first and third axes show that undersides of platy corals is a valid large scale habitat distinguished in the third dimension. Numbers refer to site-numbers. back-reef, Bluehole, Ship-channel, fore-reef, deep fore-reef, pinnacle, undersides of platy corals, Discovery Bay, coral rubble. A plot of sites of the first two axes (Fig. 1) shows two distinct faunistic groups. Group 1 contains all sites from lagoonal environments, whereas group 2 contains sites from reef habitats. Sites and species are printed as numbers. For species names consult Table 2. Looking at the third dimension of the sites plot (Fig. 2) a third group of sites is clearly separated from other habitats. All sites included in this group derive from undersides of platy corals where a completely different assemblage of sponge species occurs (Table 2, species groups 10 & 16). The lagoonal and reef-surface habitats were analyzed separately, whereas too few members of the ‘undersides of platy corals group’ excluded it from further investigation. Reef surface. Excluding sites from the lagoon and the undersides of platy corals, the remaining sites from reef surfaces show a more-or-less continuous ENT RDA ®© subset: reef sites 1. versus 2. axis 3 overlay: depth * 2m depth Y 33m depth “increasing depth FIG. 3. Redundancy Analysis (RDA) displaying only reef sites. No distinct site-groups are recognizable, but depth as overlay (larger symbols indicate greater depth) shows arrangement of the sites along a depth gradient. Grey arrow indicates general direction of depth gradient (increasing depth from upper left to lower right). Numbers refer to site-numbers. distribution along a gradient increasing depth, from the upper left to the lower right (Fig. 3). The sample areas 28, 32, 33 and 34 did not fit into this gradient, and have in common that they derive from reefs near river mouths. The location ‘river mouth’ is obviously different from other reef localities and therefore, these sites were omitted from further analysis. This area may be greater influenced by freshwater, sediments and turbid water although this is speculative and based on few sites only. Figure 4 shows the reef surface sites with substrate inclination overlayed, with an increasing trend towards inclination to the right indicated. The corresponding plot of the species (Fig. 5) shows preferences for species with regard to water depth and inclination. (e.g. species on the upper left are from more horizontal, shallow environments, whereas species on the lower right are from more vertical, deep environments). The two striking gradients, depth and inclination, have more-or-less the same direction, they are not 312 4 RDA subset: reef sites 1. versus 2, axis overlay: inclination 4? i E MEMOIRS OF THE QUEENSLAND MUSEUM P some freshwater outflow. Extensive Thalassia seagrass-beds occur here and the eastern part of the bay is protected by land and reef from NE trade winds. Site-groups 1, 2, 9 and 10 (Table 2) are mainly from lagoonal environments. The site-groups 9 and 10 of the classification (Table 2) deviate somewhat from the results of the ordination (Fig. 6) whereby there is mixing between some sites from shallow fore-reef habitats with sites from the lagoon. This 1s probably due to recent mipri ha T i i , a N us w 2 1 e horizontal (5 vertical FIG. 4. Redundancy Analysis (RDA) of reef sites, with inclination of substrate as overlay. Larger symbols indicate greater inclination. No site-groups can be recognized, but a continuous arrangement of sites along a gradient of inclination from horizontal sites (upper left) to vertical (lower right). Grey arrow indicates general direction of increasing inclination. Numbers refer to site-numbers. independent variables. There is undoubtedly an increase in steeper habitats with increasing depth, especially at the deep fore-reef. In these deeper waters there is a steep wall extending down to several hundred meters. However, depth alone is not sufficient to explain species' distributions because less inclined deep water habitats are settled by different species than more inclined deep water habitats. These two gradients do not produce clearly separated groups but the sites appear to be arranged along continuous gradients. Lagoon. The separation of lagoonal sample areas from reef sites is shown in Figure 1. Inside Discovery Bay several subhabitats are evident. In shallow parts of the lagoon, close to the coast, many ridges occur where freshwater flows out. Two ‘blue holes’ are the deepest parts of the lagoon (13m and 50m), with very turbid water, fine sediments on the sea bed and probably also increasing inclination hurricane destructions in which the topography of seaward lagoonal-, reef flat- and shallow fore-reef -habitats were more or less equalized, and therefore subsequently settled by similar species. For site-groups 1 (Thalassia seagrass) and 2 (freshwater ridges) classification and ordination analyses are in complete agreement. According to the ordination analysis of the lagoon subset (RDA, square root transformation, species normalized and centered), five groups of sites (a-d) are distinguishable and shown in Figure 6, 1) Blue hole: the sites 77, 78, 79, 80, 50 and 3 (site group 10) are from the large blue hole (the smaller blue hole mostly consists of bare sediments, and only a small part at the SE end is overgrown by sessile organisms). Sites 77-80 are from NE to W parts of the blue hole where seawater streams into the bay from the ship-channel. Clearly separated from these are sites 3 and 50 which are from the SE slope of the blue hole, locally known as the ‘Columbus Park’ locality. Here, influences of freshwater and pollution with bauxite are probable, the latter because of the proximity of the docking area of bauxite freightships. The remaining sample areas from group 10 are from shallow fore-reef habitats, as mentioned above. 2) Thalassia seagrass-beds: Very close to the “blue hole’ group is a group of sites from Thalassia seagrass-beds. The long leaves of the Thalassia seagrass slow down water velocity and lead to higher sedimentation comparable to the situation within the blue hole, and this is the most probable explanation for its statistical similarity to the ‘blue hole’ group. Thalassia seagrass often grows within the back-reef, and the seaward margin of the blue hole is also proximate to back-reef environments, so that these two habitats are not clearly differentiated in their faunistic composition. JAMAICAN SPONGE DISTRIBUTIONS 313 RDA 1. versus 2, axis ^" A T] Hr pP eroeenrmenmemmmeepreeteenn nne d Li 3 P x * Li s subset: reef species comparison between Hechtel’s and our results 1s not appropiate, Alcolado (1990) differentiated reef- sponge communities, which he subdivides into less than 10m depth and 10-30m a depth, mangrove-sponge communities, macrolagoons and bathyal-sponge communities (150-608m depth). He described the common species in each community and compared diversities, but obviously chose depth-classes before sampling. Alvarez et al. (1990) also -i4 as -F =l y i z 3 4 inclination depth FIG. 5. Redundancy Analysis (RDA) of reef sites, showing a complementary plot of species data from Figures 3-4. Within reef habitats there are no distinct groups of species but species are continuously arranged along gradients. Upper left; shallow reef-species; center: species from intermediate depths or with indistinct depth distribution; lower right: deep reef-species. Numbers refer to species numbers in Table 2. 3) Ridges with freshwater outflow: (group 1, Table 2). These ridges are often surrounded by Thalassia seagrass-heds but the influences of the freshwater are strong enough to promote settlement of a different fauna here. 4) Protected back-reef: Clearly separated from other lagoonal environments are two sites from the eastern back-reef. This environment is very close to the blue hole. Consequently, the water here is more turbid than in western parts of the reef, Furthermore, the eastern back-reef is in lee of prevailing waves from the NE trade winds. DISCUSSION Hechtel (1965) investigated only 10 (sometimes very small) localities on the south coast and one cannot be sure that individual random differences are responsible for the observed distributions. Because there are considerable dillerences in sponge species along the north coast, a direct focused on the importance of depth gradienis in influencing species distributions, and found that there were "a few abundant species and many uncommon ones... The most frequent species are also the most widely distributed along the depth gradient." They also conclude that “For all species, the values of density and area coverage varies along, E the transects but seems independent from depth". These data seem to contradict our results. But their investigation obviously focuses on the few abundant and doininant species, whereas our resulls are based on the analysis of the whole species composition with a data transformation avoiding dominance types. Diaz et al. (1990) also studied comm- unity structure of sponges along depth gradients, They compared species number, area coverage, density, diversity and evenness. Again, in contradiction to our results. they found that “The results of the cluster analysis confirm that the sponges in the study area lack a well defined pattern of zonation”, and the “predominance of encrusting species at almost all depths”, However, in looking for differences between transects they restricted their question to zonation based on depth alone. Schmahl (1990) investigated the distribution and abundance of sponges in southern Florida reefs at three depth zones and found that “distributional patterns of sponges may be used to identify the ecological factors that influence the communities in this area”. sl Whereas all these investigations focused on depth zonation only and comparing pre-defined regions of the reef, the present paper makes no a priori assumptions about reef zonation, irresp- ective of bathymetric or habitat bias, instead using subsets of sites with similar species composition, provided by multivariate analysis. RDA REI protected backreef 14 1,0 q TTTTTTTTTHTTTTTTTOTEÉOU- Trt +1 subset: lagoon sites 1. versus 3, axis MEMOIRS OF THE QUEENSLAND MUSEUM than 10 years have shown that the Discovery Bay reefs are representative for the island as a whole...”. Consequently we believe that comparison between their results and ours is well justified. Goreau made a more intuitive approach, naming zones either after dominant species or after striking morphological structures, while in the present paper we tried to find characteristic species from different habitats (which need not to be dominant) and their correlations to some abiotic factors. We consider it is worthwhile comparing these 64 7 H "ridg results to see what is still 1.0 * water outtlo q recognizable and what has ilaia ryj A a E O S E A yi ye changed. Goreau -t 2,5 0 E 1 1.5 2 2.5 distinguished three regions, FIG. 6. Redundancy Analysis (RDA) of lagoon sites. Several groups are distinguished, characterising different subhabitats, however, small sample sizes of group-members weakens the interpretation. Numbers refer to site numbers. Only then we began interpretation of these groups of sites with field data. The analyses made are objective and repeat- able, but we have to admit, that for our conclusions there is no objective test. We think the conclusions are well justified because in most cases there are enough members of the group, making our conclusions probable. Exceptions are the subhabitats within the lagoon. The groups have only few members and we are aware of the risks of interpreting natural variation. But the groups are easily explainable and fit very well in environmental differences, observed during dives, that we think it worth, to include them here. COMPARISON WITH GOREAU'S ZONATION OF JAMAICAN CORAL REEFS. Goreau (1959) described a reef near Ocho Rios, N Jamaica, 34 year ago, in relatively close proximity to Discovery Bay. He claimed the Ocho Rios reef to be “typical of the large fringing barrier reef communities found along the north coast of Jamaica.” and was also familiar with the Discovery Bay reefs. Furthermore, Goreau & Wells (1967) wrote “... that extensive surveys carried out in other parts of Jamaica over more back-reef, reef-crest and fore-reef, which were also divided into 7-9 different zones. These are considered separately below with remarks as to the present status of these reefs and differences between these two data sets. 1) Goreau divided the back-reef region into a shore zone, with a variety of hermatypic corals, and a lagoon zone, with less corals. The shore zone has almost disappeared. Now, only one small protected area within Discovery Bay has living corals close to shore. Acropora palmata has disappeared, Millepora complanata dominates this small spot. The upper parts of this “inshore reef” is now dominated by the green zoanthid Zoanthus sociatus, a species which is described by Goreau for the reef flat (which he called also the Zoanthus zone), but where it is barely present today. Additional to Goreau's zones the present paper distinguishes four lagoonal subhabitats (blue hole, Thalassia seagrass-beds, ridges with freshwater outflow and protected back-reef) separated on the basis of faunistic data. Obviously Goreau worked on a larger scale. Goreau's ‘inshore reefs’ were probably destroyed by the hurricanes, and he did not divide the lagoon-zone into subhabitats. 2) Goreau's reef-crest region is divided into rear-zone, reef-flat (zoanthus-zone), palmata- JAMAICAN SPONGE DISTRIBUTIONS zone (with breaker and lower palmata zone), and buttress-zone. Probably, again due to hurricane destruction, it is now not possible to recognize all of these zones. Rudiments of his rear-zone can be recognized in some parts, with still large Montastrea annularis, Diploria strigosa and Sideratrea siderea. The reef-flat or zoanthus- zone has changed very much. Zoanthus sociatus can only barely be found. There are still some Millepora, but Gorgonia and Lithothamnium are rare. Large dead coral rocks, often above sea-level, occur instead. On the sides of coral rocks some small So/enastrea sp. occur. His palmata-zone also does not exist any more, replaced by hargrounds, settled by the sponges Anthosigmella varians and Chondrilla nucula. The previously dominant Acropora palmata exists only with some scattered (sometimes large) colonies. The buttress-zone can still be easily identified but, the sediment canals, described by Goreau as “...very narrow, somewhat winding, canyons the walls of which are perpendicular or even overhanging” are now wide sediment streams, the walls less inclined. Exceptions are found in front of the mouths of the Rio Bueno and the Pear Tree River, where some buttresses come close to Goreau’s description. However, these localities seem to have suffered less destruction than any other reefs investigated. Another striking difference are the depths given by Goreau. He wrote: “At the buttress crests, the depth averages only about 2 meters whereas the canyons are between 8 and 10 meters deep.” Now the buttress crests range between 5-20m depth and the sediment areas between 7-23m. The uppermost region of the buttresses has probably been destroyed or buried and considerable amounts of the buttresses have been removed. This seems very probable because Goreau described Acropora palmata on top of the buttresses where they no longer exist, even as dead colonies. While Goreau stated the cover of living coral was 90% of the available surface, it is now about 15%, except for a few small areas at the walls of the buttresses where 90% coral cover occurs. 3) The seaward slope or fore-reef was divided by Goreau into the cervicornis-zone and the annularis-zone. Both zones have completely disappeared. He located the cervicornis-zone at ‘the uppermost region of the seaward slope’, seaward of the buttress-zone. Now there are large sandy areas between the buttress zone and the shelf break. Acropora cervicornis is now found only sporadically in the buttresses. At the shelf break there are some pinnacles, the edge, and wall of the break itself below 30m depth, covered with numerous large Montastrea annularis. This area is much deeper than Goreau described as the annularis-zone, where he gave an average depth of 15m. The ‘undersides of platy corals-habitat described here was not mentioned by Goreau, probably because he did not investigate these depths and he was also mainly interested in hermatypic corals. This habitat is a very small component of the ‘area scale’ used by Goreau, although relatively large from our faunistic approach. Some differences between our results and those of Goreau are related to our different approaches and methodologies (e.g. our lagoon subhabitats, or the undersides of platy coral habitat). Other differences, like the missing palmata zone, cervicornis-zone or the different depths of the buttress zone seem to be due to destruction by hurricanes. To summarize the changes since Goreau's investigations published in 1959, it is obvious that the fringing reefs with three distinguishable main structures (back-reef, reef crest and fore-reef) have changed into a situation where the back-reef has lost its inshore reefs and has gradually merged into a long seaward slope with nearly no living reef crest in between. Goreau's internal zonation of reef crest and seaward slope can now only be recognized in parts while several striking structures described by him have completely disappeared. Extensive growth of algae seems to inhibit recovery of the reefs. ACKNOWLEDGEMENTS HL was funded by the Deutsche Forschungsgemeinschaft (Le 822, Re 665), and thanks his dive buddies Bettina Schwarz- Lehnert, Klaus Demuth, Peter Gayle, Peter Forde, Kathy Lankester and Philip Janca. The Discovery Bay Marine Laboratory and staff supported field work in every possible way. This is contribution no. 612 of the Discovery Bay Marine Laboratory. We thank John Hooper and two anonymous reviewers for improving the manuscript. LITERATURE CITED ALCOLADO, P. M. 1990. General features of Cuban sponge communities. Pp. 351-357. In Riitzler, K. (ed.) New Perspectives in Sponge Biology. (Smithsonian Institution Press: Washington D.C.). ALVAREZ, B., DIAZ, M, C. & LAUGHLIN, R. A. 1990. The sponge fauna on a fringing coral reef in 316 Venezuela, I: Composition, distribution, and abundance. Pp. 358-366. In Riitzler, K. (ed.) New Perspectives in Sponge Biology. (Smithsonian Institution Press: Washington D.C.). BRAUN-BLANQUET, J. 1964. Pflanzensoziologie, 3. Aufl.:1-865 (Springer-Verlag: Wien). BOGER, H. 1988. Versuch tiber das phylogenetische System der Porifera. Meyniana 40: 143-154. DIAZ, M. C., ALVAREZ, B. & LAUGHLIN, R. A. 1990. The sponge fauna on a fringing coral reef in Venezuela, II: Community structure. Pp. 367-375. In Riitzler, K. (ed.) New Perspectives in Sponge Biology. (Smithsonian Institution Press: Washington D.C.). FISCHER, H. S. 1994. Simulation der ráumlichen Verteilung von Pflanzengesellschaften auf der Basis von Standortskarten. Dargestellt am Beispiel des MaB-Testgebiets Davos. Veróffentlichungen des Geobotanischen Instituts der Eidgenóssischen Technischen Hochschule, Stiftung Riibel,in Ziirich 122:1-143. GEISTER, J. 1977. The influence of wave exposure on the ecological zonation of Caribbean coral reefs. Proceedings of the Third international coral reef symposium, Miami: 23-29 1983. Holozáne westindische Korallenriffe: Geomorphologie, Okologie und Facies. Facies 9: 173-284, pls. 25-35. GOREAU, T. F. 1959. The ecology of Jamaican coral reefs. I. Species composition and zonation. Ecology 40: 67-90. GOREAU, T. F. & WELLS, J. W. 1967. The shallow water scleractinia of Jamaica: Revised list of species and their vertical distribution range. Bulletin of Marine Science 17(2): 442-453. HECHTEL, G. J. 1965. A systematic study of the Demospongiae of Port Royal, Jamaica. Bulletin of the Peabody Museum of Natural History 20: 1-103. LEHNERT, H. & SOEST, R.W.M, VAN 1996. North Jamaican Deep Fore-Reef Sponges. Beaufortia 46(4): 53-81. 1998. Shallow Water Sponges of Jamaica. Beaufortia 48(5): 71-103. 1999: More North Jamaican deep fore-reef sponges. Beaufortia, in press. MEMOIRS OF THE QUEENSLAND MUSEUM LIDDELL, W.D. & OHLHORST, S.L. 1987. Patterns of reef community structure, north Jamaica. . Bulletin of Marine Science 40(2): 311-329, RUTZLER, K. 1971. Bredin-Archbold-Smithsonian Biological Survey of Dominica: Burrowing sponges, Genus Siphonodictyon Bergquist, from the Caribbean. Smithsonian Contributions to Zoology (77): 1-17. 1974. The burrowing sponges of Bermuda. Smithsonian Contributions to Zoology (165): 1-31. 1987. Tetillidae (Spirophorida, Porifera): A taxonomic reevaluation. Pp. 187-203. In Vacelet, J. & Boury-Esnault, N. (eds) Taxonomy of Porifera. NATO ASI series. Vol. G13 (Springer- Verlag: Berlin, Heidelberg). SCHMAHL, G. P., 1990. Community structure and ecology of sponges associated with four southern Florida coral reefs. Pp. 376-383. In Riitzler, K. (ed.) New Perspectives in Sponge Biology. (Smithsonian Institution Press: Washington D.C.). TER BRAAK, C. J. F. 1988. CANOCO - a Fortran program for canonical community ordination by [partial] [detrended] [canonical] correspondance analysis, principal component analysis and redundancy analysis (version 2.1). Technical Report LWA-88-02: 1-95. (Agricultural Mathematics Group: Wagenigen). 1990. Update notes: CANOCO version 3.1. 35 pp. (Agricultural Mathematics Group: Wageningen). WILDI, O. 1989, A new numerical solution to tradi- tional phytosociological tabular classification. Vegetatio 81: 95-106. WILDI, O. & ORLOCI, L. 1990. Numerical exploration of community patterns (SPB Academic Publishing bv: The Hague). WILKINSON, C.R. & CHESHIRE, A.C. 1989. Patterns in the distribution of sponge populations across the central Great Barrier Reef. Coral Reefs 8: 127-134. WILKINSON, C.R. & EVANS, E. 1989. Sponge distribution across Davies Reef, Great Barrier Reef, relative to location depth, and water movement. Coral Reefs 8: 1-7. ZEA, S. 1993, Cover of sponges and other sessile organisms in rocky and coral reef habitats of Santa Marta, Colombian Caribbean Sea. Caribbean Journal of Science 29(1-2): 75-88. RANDOM AMPLIFIED POLYMORPHIC DNA (RAPD) ANALYSIS CAN REVEAL INTRASPECIFIC EVOLUTIONARY PATTERNS IN PORIFERA GISELE LOBO-HAJDU, JOSÉ JOAO MANSURE. ADRIANA SALGADO, EDUARDO HAJDU. GUILHERME MURICY AND RODOLPHO M. ALBANG Lóbo-Hajdu, G., Mansure, J.J., Salgado, A., Hajdu, E., Muricy, G. & Albano, R,M. 1999 06 30: Random amplified polymorphic DNA (RAPD) analysis can reveal intraspecific evolutionary patterns in Porifera. Memoirs of the Queensland Museum 44: 317-328. Bris- bane, ISSN 0079-8835, Random amplified polymorphic DNA (RAPD) polymerase chain reaction (PCR) techniques were used to generate polymorphic markers to investigate evolutionary patterns in Porilera. Three primers were selected based on their amplification profiles using genomic DNA from I8 sponge species (17 Demospongiae and | Calcarea), collected from various localities along the Brazilian coast. The total number of amplified scorable bands per primer varied from 9 (primers OPS-17 and UBC-322) to 32 (primer OPG-19). The level of genetic polymorphism was measured in a population (N=9) af Hymeniacidon heliophila from Itaipu Beach (Rio de Janeiro State). The percentage of polymorphic fragments for primer UBC-322 was 31% (mean intrapopulation genetic distance=0.32) The interpapulatianal genetic diversity was calculated using two individuals from each of five different populations of H, heliaphile. The percentage of polymorphic fragments varied from 50-68% (mean genetic distance between popylations=0.53), A monomorphic band isolated from an individual of the Itaipu population of 77. heliophila was labelled and used to probe dot blots containing genomic DNA from 18 species of sponges, fruit fly, rat and 15 individuals from the 5 populations of H. heliophila. Only H, heliophila DNA was hybridised with this marker. We also generated RAPD band patterns using genomic DNA from 6 different species of the genus Mycale and primers OPG-19 and OPS-17, with an average percentage of polymorphic bands of 91 and 98%, respectively, The genetic distance between species of Mycale varied from 0.45-0,93 (mean genetic distance-0.76) for primer OPG-19 and 0.64-0,93 (mean genetic distance>0.80) for OPS-17, The RAPD technique was shown to be a useful tool to generate molecular markers and to measure polymorphism and genetic distances in Porifera. O Porifera, molecular markers, RAPD, genetic diversity, SW Atlantic, Hymeniacidon heliaphila, Mycale. Gisele Lóbo-Hajdu (email; lobohajdu(vax.apc.arg), Adriana Salgado & Rodolpho M. Albano, Departamento de Bioquimica, Instituto de Biologia Roberio Alvaritara Gomes, Universidade do Estado do Ria de Janeiro, Av, 28 de Setembro, 87 (fundos, 49 andar), 20535 1-013, Rio de Janeiro. RJ, Brazil; José Jado Mansure, Instituto de Ciéncius Biológicas e Ambientais, Universidade Santa Ursula, Rua Jornalista Orlando Damas, 59, 22231-0400, Rio de Janeiro, RJ, Brazil; Eduardo Hajdu & Guilherme Muricy, Departamento de Invertebrados, Museu Nacional, Universidade Federal da Rio de Janeiro, Quinta da Boa Vista, s/n, 20940-040, Ria de Janeiro, RJ, Brazil; Eduardo Hajdu, Centra de Biologia Marinha (CEBIMAR), Universidade de São Puulo, Sado Sebastido, SP. Brazil; 22 February 1999 Sponges (phylum Porifera) show considerable morphological variability. Ai the biochemical level this variability ts supported by hetero- geneous allozyme banding patterns reflecting high genetic heterozygosity in sponges (Solé-Cava & Thorpe, 1991, 1994). These characteristios together with the limited data on aspects of reproduction, life history, ecology and cell biology make the study of sponge genetics and systematics complicated tasks, In fact, these problems are general to many marine organisms where complexes of sibling species are known to be common (Knowlton, 1993). Very few molecular genetics studies have been initiated on sponges, the majority of them using allozyme electrophoresis or ribosomal RNA sequence coinparisons (Kelly-Borges etal., 1991; Kelly-Borges & Pompom, 1994; Solé-Cava de Thorpe, 1994; Collins, 1998). The application of molecular techniques 10 study sponge genetics and the evolution of the phylum is a difficult proposition due to the Jack of knowledge on the complexity of the sponge genome. On the other hand, fingerprinting techniques have been widely used in studying ecology and taxonomy of several marine phyla, and constitute a good integration bridge between environmental and molecular sciences (Burton, 1996). These techniques provide possibilities to reach more stable classifications by quickly assessing vast series of polymorphic molecular characters of potential phylogenetic significance. Character- isation of sponge genomes by fingerprinting techniques allow better estimation of their intra- and interspecific diversity. This estimation is essential for effective management of biological resources and sound species delimitations, par- ticulary in sibling species complexes. The study of polymorphisms in the eukaryotic genome is a way to discover the evolutionary relationships between populations or species. Among the many molecular methods currently available for genetic diversity studies, the random amplified polymorphic DNA (RAPD) polymerase chain reaction (PCR), also known as arbitrarily primed (AP) PCR (Williams et al., 1990; Welsh & McClelland, 1990), appears particularly suitable for population genetics. In RAPD analysis there is no need for a priori knowledge of DNA sequences of samples, cost and efforts are reasonable, and many individuals and loci can be efficiently assayed. In RAPD analysis a single nonspecific primer is used to randomly amplify DNA segments in the genome. The DNA sequence between inverted hybridisation sites is amplified, and, because the distance between primer sites can vary among individuals, different length fragments (or fingerprints) are generated. As the primers are small (normally 10-mers) and not specific to any given gene, many different fragments are produced in an amplification. Fragments present in one individual (or species), but not in others, are defined as polymorphic markers. The fingerprint generated is consistent for the same primer, DNA, and PCR conditions used. This fingerprint is useful in the assessment of affinities along the lower levels of biological and classificatory hierarchies, from individuals and populations to species within a genus (Hillis et al., 1996). The primary goal of this study was to generate molecular markers by RAPD analysis to dem- onstrate their usefulness in the investigation of genetic variation within, and among populations of sponges, and in the assessment of species relationships. This technique has already been MEMOIRS OF THE QUEENSLAND MUSEUM j y IAS Í à £T 235 | 44W 100 Km "looo Km. FIG. 1. Map ofthe Brazilian coast, showing collection sites for populations of Hymeniacidon heliophila (inset). 1) Cabelo Gordo Beach, São Sebastião, São Paulo State (23°48’S, 45?26' W); 2) Praia Vermelha Beach; 3) Urca Beach; Rio de Janeiro City, Rio de Janeiro State (23°0’S, 43°12’W); 4) Boa Viagem Beach; 5) Itaipu Beach, Niterói, Rio de Janeiro State (22252'S, 43°0’W). The distances between beaches are: Itaipu-Boa Viagem=20km; Boa Viagem-Urca= 10km; Urca-Praia Vermelha=3km; Praia Vermelha- Cabelo Gordo=300km. used successfully for population genetics in, for example, leguminous trees (Chalmers et al., 1992), prosobranch snails (Jacobsen et al., 1996) and weevil insects (Taberner et al., 1997). It has also been effective in phylogenetic studies of plants of the genus Brassica (Demeke et al., 1992), papaya cultivars (Stiles et al., 1993) and mangrove species (Lakshmi et al., 1997). These studies have shown that RAPD is a rapid, accurate and sensitive method to detect genetic variation, and to assist in taxonomic invest- igation at levels from populations to species. In this report we used RAPD analysis to: 1) assess the genetic diversity within and between populations of the sponge Hymeniacidon heliophila Parker, 1910; 2) generate a potential species-specific molecular marker for H. helio- phila; and 3) generate molecular markers which could be used to investigate interspecific rel- ationships within the genus Mycale Gray, 1867. MATERIALS AND METHODS SPONGE COLLECTIONS. Samples of Hymeniacidon heliophila were collected using SCUBA at Praia Vermelha Beach, Urca Beach RAPD ANALYSIS OF SPONGES 319 TABLE 1. Sponge species, classification, collection sites, and voucher numbers. Abbreviations: Ba = Bahia State, PE= Pernambuco State, RJ = Rio de Janeiro State, SP = Sáo Paulo State. MNRJ and UFRJPOR = Porifera Collections of the Museu Nacional, Universidade Federal do Rio de Janeiro. Species Classification Collection site Voucher number Aiolochroia crassa (Hyatt, 1875) Verongida, Demospongiae Fernando de Noronha, PE UFRJPOR 4766 Amphimedon viridi. : r ; Dehesa, & Michelotti, 1864 Haplosclerida, Demospongiae Praia do Sono, RJ UFRJPOR 4747 Aplysina fulva (Pallas, 1766) Verongida, Demospongiae Abrolhos, BA UFRJPOR 4699 Arenosclera sp. Haplosclerida, Demospongiae Büzios, RJ UFRJPOR 4628 Callyspongia sp. Haplosclerida, Demospongiae Fernando de Noronha, PE UFRJPOR 4783 Cer dod) alada Spirophorida, Demospongiae Parati, RJ UFRJPOR 4755 Cana fot icheloti, 1864) Poecilosclerida, Demospongiae Fernando de Noronha, PE UFRIPOR 4772 Gastrophanella sp. Lithistida, Demospongiae Fernando de Noronha, PE UFRJPOR 4773 cg aa a Halichondrida, Demospongiae Itaipu, RJ UFRJPOR 4759 rd e Halichondrida, Demospongiae Boa Viagem, RJ MNRJ 809 pomertacidon heliophila Halichondrida, Demospongiae Urca, RJ UFRJPOR 4605 a slat Y MA Halichondrida, Demospongiae Praia Vermelha, RJ UFRJPOR 4246 pen Mn heliophila Halichondrida, Demospongiae Sao Sebastiao, SP MNRJ 1307 Leucilla sp. Leucosoleniida, Calcarea Praia Vermelha, RJ UFRJPOR 4838 Mycale (degogrop: ila) aff. americana Poecilosclerida, Demospongiae São Sebastião, SP MNRJ 1584 e era R Pedal. 1995 | Poecilosclerida, Demospongiae Sáo Sebastiáo, SP UFRJPOR 4451 Mycale (Arenochalina) laxissii ; g ? m ES (Dachassatne & ed) Poecilosclerida, Demospongiae São Sebastião, SP MNRJ 1027 ene rca microsigmatosa Poecilosclerida, Demospongiae Sao Sebastião, SP MNRJ 1331 e eae arenario I | Poecilosclerida, Demospongiae Búzios, RJ UFRJPOR 2438 Mycale (Zygomycale) angulosa ` A ; : (Duchessdlos & Michelotti, 1864) Poecilosclerida, Demospongiae Parati, RJ UFRJPOR 4743 (Date eee Michelabi, 1864) Poecilosclerida, Demospongiae Sao Sebastiào, SP MNR] 429 Plakortis sp. Homosclerophorida, Demospongiae Fernando de Noronha, PE UFRJPOR 4774 end eee) Halichondrida, Demospongiae São Sebastião, SP MNRJ 289 (ae pa ophiraphidites Halichondrida, Demospongiae Fernando de Noronha, PE UFRJPOR 4779 (Rio de Janeiro City, Rio de Janeiro State), Boa Viagem Beach, Itaipu Beach (Niteró1, Rio de Janeiro State) and Cabelo Gordo Beach (Sáo Sebastiáo, Sáo Paulo State) (Fig. 1). Samples of Mycale (Mycale) arenaria were collected at Joao Fernandinho Beach (Búzios, Rio de Janeiro State), while M. (Aegogropila) aff. americana, M. (A.) escarlatei, M. (Arenochalina) laxissima, M. (Carmia) microsigmatosa and M. (Zygomycale) angulosa were collected at Cabelo Gordo Beach and surroundings (São Sebastião, São Paulo State). The other twelve species used were collected in several sites along the Brazilian coast (Table 1). Voucher specimens were deposited in the Porifera Collections ofthe Museu Nacional ofthe Universidade Federal do Rio de Janeiro (MNRJ and UFRJPOR) (Table 1). Specimens were frozen in dry ice immediately after removal from sea water and kept frozen at -20°C or -70°C prior to DNA extraction. EXTRACTION OF DNA. After careful dis- section to remove macroscopic symbionts on an ice-cooled Petri dish, specimens were ground with a glass rod in a solution of 4M guanidine hidrochloride, 0.1M Tris-HCI pH 8.0, 0.5% sarcosil and 1% f-mercaptoethanol. The suspension was incubated at 50°C for 1hr and centrifuged at 3000g/20mins. The supernatant was extracted with an equal volume of phenol: chloroform:isoamyl alcohol (25:24:1) and nucleic acids were precipitated with 1 volume of isopropanol. The pellet was washed in 70% ethanol and air dried. The dried pellet was dissolved in 0.01M Tris-HCl pH 8.0, 0.005M EDTA, 0.5% SDS, 50ug/ml proteinase K and 20pg/ml RNAse A and incubated at 50°C for 2hrs. Another extraction with phenol:chloroform: isoamyl alcohol (25:24:1) was followed by a single extraction with chloroform/0.3M sodium acetate pH 5.2 and the DNA was precipitated with two volumes of ethanol. After centrifug- ation the DNA pellet was washed in 70% ethanol, air dried and dissolved in sterile water. The DNA concentration was estimated in 1% agarose gels and, followed by dilution in water to a final concentration of Ing/ul. RAPD- PCR PROCEDURE. RAPD analysis of genomic DNA was undertaken using a set of three 10-mer random oligonucleotide primers selected on the basis of the high reproducibility of the patterns and the signal intensity of the bands from a total of seventeen decamer primers screened (Table 2). Each PCR amplification reaction mixture of 25ul contained 2ng of genomic DNA, 2.5ul of 10 x buffer (0.5M KCI, 0.1M Tris-HCI pH 8.4, 1mg/ml gelatin), 1.5ul of 50mM MgCl;, 0.5ul of SmM dNTPs (Pharmacia), 20ng of primer and 1 unit of Tag DNA polymerase (USB). The reaction mixture was overlaid with one drop of mineral oil, and amplification was carried out in a DNA thermal cycler (Perkin Elmer 480). An initial de- naturation step of 3mins at 94°C was followed by 35 cycles of 30secs at 94°C, 30secs at 37°C and 90secs at 72°C, with an additional final step of 10mins at 72°C. The amplified bands were separated by electrophoresis on 6% vertical non- denaturing polyacrylamide gels (PAGE) and visualised after silver staining following the procedure of Sanguinetti et al. (1994). The size of amplified fragments was estimated by com- parison with a standard DNA marker: A (lambda) DNA, double digested with EcoRI-Hinalll. RAPD DATA INTERPRETATION AND ANALYSIS. Relationships between RAPD profiles obtained from DNA amplifications of sponge individuals with each primer were MEMOIRS OF THE QUEENSLAND MUSEUM TABLE 2. Name/number, sequence and G+C content of the primers used. Name/ No. of primer Sequences 5-3 G+C content OPS-17/ no. 6 TGG GGA CCA C 70% OPG-19/ no. 13 GTC AGG GCA A 60& UBC-322/ no.15 GCC GCT ACT A 60% explored by comparing the presence or absence of shared bands. We scored bands using sharpness and differentiation from the background, and al- though this is an empirical technique, it produced a consistent outcome which was reproduced by at least two people in blind tests. This process of fragment selection was used throughout all experiments reported in this study. From these assays we recorded the selected markers as either present (1) or absent (0) in each DNA sample. Data recording and calculations were performed using the RAPDistance package, version 1.03 (Armstrong et al., 1994), which allowed us to construct a matrix of pairwise distances. Data from these comparisons were used to calculate similarities between pairs of samples using the Jaccard coefficient (Jaccard, 1901, 1908). This algorithm provides similarity values in the range 0-1, and the software converts these to genetic distances as (1-s). The first requirement, and a potential problem of the RAPD method, is its reproducibility. Small changes in reaction conditions are known to cause variation in the amplification pattern (Khandka et al., 1997, and references therein). These difficulties may be overcome if care is taken to ensure consistent PCR reaction con- ditions during amplification. To optimise these conditions and demonstrate RAPD reproduc- ibility we conducted multiple (four or more) PCR amplifications with the same primer and individ- ual samples (Lobo-Hajdu, unpublished data). In the conditions standardised in our laboratory the PCR-RAPD reactions were reproducible. Many studies have shown Mendelian in- heritance of RAPD markers (Lewis & Snow, 1992; Levitan & Grosberg, 1993), with a few exceptions (e.g. Riedy et al., 1992). Although RAPD analysis of DNA does provide a random sample of the genome, the banding patterns cannot be interpreted allelically because bands are dominant markers and preclude the discrim- ination of heterozygotes and homozygotes (Welsh & McClelland, 1990; Williams et al., 1990). Consequently, following Williams et al. (1990), we assumed that DNA profiles were the RAPD ANALYSIS OF SPONGES result of single alleles when computing RAPD data in this study. DOT BLOTS. Aproximately 500ng of genomic DNA from fruit fly, rat, 18 species of sponges and 15 individuals of H. heliophila were denatured with NaOH and bound to a Gene Screen plus nylon membrane (Dupont) using a Bio Dot microfiltration apparatus (Bio Rad), following the manufacturer’s recommendations. The membrane was dried in an oven at 80°C for 1hr and pre-hybridised for Ihr at 65°C in 10ml of buffer containing 50mM Tris-HCl pH 7.4, IM NaCl, 10% PEG-8000, 1% SDS and 1001g/ml of heat denatured salmon sperm DNA. A mono- morphic band obtained with primer UBC-322 in the interpopulation assay was excised after sep- aration on a 1.6% agarose gel. The fragment contained in the agarose plug was isolated using a Geneclean Kit (BIO 101) following the manuf- acturer’s recommendations, and then reamplified under standard conditions with the same primer originally employed to generate the band (UBC-322). The reamplified product was labelled with a—”P dCTP using a T7 Quick Prime Kit (Pharmacia Biotechnologies) and added to the pre-hybridisation solution. Hybrid- isation was carried out for another 12hrs. The membrane was then washed with 0.2 x SSC at 65°C and exposed overnight to an X-Ray film (Kodak) in a cassete at -70°C. RESULTS Fourteen of 17 primers used to amplify DNA of 12 specimens of Hymeniacidon heliophila from Praia Vermelha Beach, produced clear and distinct banding patterns in a standard RAPD reaction. Three primers were selected to compare banding patterns within and among different sponge species and to test if these primers would work in general for sponges. Each primer gen- erated a different number of fragments for the same species of sponge. The total number of bands per primer ranged from 9-32 (average =22). The majority of fragments scored were found in the molecular weight range between 300-2500 base pairs (bp) (Table 3). Commonly, bands with molecular weight above 2000bp are very compressed and difficult to score. DNA amplification profiles in 9 individuals of H. heliophila from Itaipu Beach, produced using primer UBC-322 (Fig. 2), yielded both mono- morphic and polymorphic bands (arrows in Fig. 2). The diversity of RAPD markers generated for this sample of a natural sponge population was 321 TABLE 3. Sponge species and average number of bands, polymorphic plus monomorphic, obtained with each primer used (range of molecular weight in base pairs). UBC-322 Sponge sample OPS-17 OPG-19 Amphimedon — 20 (400-2100) | 23 (400-2200) | 22 (100-2100) Aplysina fulva — | 22 (200-2100) | 32 (200-2500) | 13 (100-2000) Arenosclera sp. | 23 (400-2500) 22 (450-2000) 20 (500-3000) 28 (400-2500) 30 (200-2500) 21 (150-1900) 13 (500-1600) 25 (600-3000) 27 (650-2500) 09 (700-1900) 23 (400-1900) 28 (600-2500) Callvspongia sp. Ectyoplasia ferox Gastrophanella sp. Leucilla sp. 17 (700-2300) | 26 (600-3000) |19 (600-2500) Mycale (Aegogropila) 11 (600-3000) | 27 (400-2500) | 20 (300-2100) aff. americana Mycale (A egogropila) escarlatel 23 (300-2500) | 29 (400-2500) | 21 (600-2500) T | 32 (300-2500) | 24 (400-2500) A renochalina) axissima 24 (300-2500) Mycale (Carmia) microsigmatosa 23 (400-2500) | 30 (450-2500) | 10 (700-2500) nario eale) | 18 (300-2100) | 31 (200-2500) | 24 (400-2000) Mycale (Zygomycale) 21 (400-2500) | 28 (300-2500) | 25 (300-2500) angulosa Plakortis sp. 09 (400-1700) | 22 (400-3000) | 17 (600-1600) estimated by the percentage of polymorphic frag- ments which ranged between 15-42% (average =31%; Table 4). Estimated genetic distance between individuals ranged from 0.17-0.45 (average =0.32; Table 5). The percentage of polymorphic fragments obtained from RAPD amplification profiles of 2 individuals from each of the 5 different pop- ulations of H. heliophila (Fig. 3) varied from 50-68% (mean of 5 populations=61%; Table 6). Genetic distances between individuals ranged from 0.22-0.73 (average=0.53; Table 7). Genetic distances between individuals of the same pop- ulation are within the range of intrapopulation variation (0.17-0.45; Table 5): PV1-PV2=0.22; U1-U2=0.32; BV1-BV2=0.32; IT1-IT2=0.27 and SS1-SS2=0.42 (Table 7). A monomorphic band of approximately 700bp (Fig. 3, middle left arrow) isolated from one individual of A. heliophila from Itaipu Beach was purified, reamplified, labelled and used as a probe in dot blots. These blots contained genomic DNA from 18 sponge species, fruit fly, rat and from 15 individuals of the 5 populations of A. heliophila (Fig. 4). The labelled monomorphic band hybridised only with HA. heliophila DNA. 2 N N TABLE 4. Number and percentage of polymorphic fragments amplified with primer UBC-322 for each individual of Hymeniacidon heliophila shown in Fig. 2. Fragments scored in the range between 200-2000bp. Specimens of No. polymor- | Total no. epa H. heliophila phic fragments | fragments ments Specimen H1 6 17 35 | Specimen H2 3 14 21 Specimen H3 8 19 | 42 Specimen H4 5 16 31 Specimen H5 5 16 3l Specimen H6 8 19 42 Specimen H7 6 17 35 Specimen H8 2 13 15 Specimen H9 4 15 27 Mean 5 16 31 Genomic DNA from 6 species of Mycale was amplified with primers OPG-19 and OPS-17 (Fig. 5A, B). Genetic distances between Mycale species ranged from 0.45-0.93 (average=0.76; Table 8A) when using primer OPG-19 and from 0.64-0.93 (average=0.80; Table 8B) when using primer OPS-17. Genetic distances between the two Mycale (Zygomycale) angulosa from two different populations (Parati and São Sebastião) was equal to 0.65 with both primers used. Interestingly, this genetic distance was within the range obtained for interpopulation variation in Hymeniacidon (0.22-0.73; Table 7), suggesting that the range for interpopulation variation on Mycale could be similar. DISCUSSION In this study we tested the utility MEMOIRS OF THE QUEENSLAND MUSEUM and low cost. Moreover, the technique does not make use of radioactive probes and does not require large amounts of DNA or foreknowledge of the regions to be amplified. Another simple and popular way to analyse genetic variation, is to make use of allozyme electrophoresis (Shaw & Prasad, 1970; Avise, 1994; Hillis et al., 1996). This technique revealed different degrees of polymorphism than did the RAPD analysis, which can be explained by the nature of genetic markers generated by both methods. Allozymes are products of structural genes, subjected to relatively strict selective processes, and variation could also be due to post-translational modifications. In contrast, the RAPD amplification technique scans anonymous DNA sequences, whether or not they are coding regions, unique or repetitive, which are randomly distributed in the genome. RAPD markers generate a wide picture ofthe genome and are not strictly subjected to natural selection. One ofthe advantages of RAPD analysis is that RAPD patterns can be easily generated from DNA from any living organism. However, this can also be a major disadvantage, especially when the target organism harbours symbionts. Sponges are notorious for the presence of a number of different organisms living in them, such as bacteria, cyanobacteria, protozoa, fungi, algae and several invertebrates (Berquist, 1978; Wilkinson, 1978; Cheshire & Wilkinson, 1991; Preston etal., 1996; Vacelet & Donadey, 1997). A potential problem with the data and interpretations of RAPD analysis was the amplification of symbiotic DNA together with sponge DNA which could result in the presence of non-homologous, yet co-migrating bands. We of the RAPD method to estimate genetic variation in sponges and to generate molecular markers that could be used in population genetics, species delimitation and intrageneric phylogenetic studies. These markers were used to TABLE 5. Genetic distances (1-s, where s = Jaccard's similarity values in the range 0-1) between 9 individuals of Hymeniacidon heliophila from Itaipu Beach (see Fig. 2), calculated from the presence and absence of the amplification products of primer UBC-322. Complete identity is 0 and complete difference is 1. Mean genetic distance among the 9 individuals H1-H9 was 0.32. quantify the genetic diversity at Ei a 5 n a > = the intrapopulation, interpop- ulation and interspecific levels in H2 | 0.28 experiments undertaken on |3 | 036 | 035 populations of one species, H4 035 | 024 | 0.41 Hymeniacidon heliophila, and in H5 035 | 024 | 025 | 022 six species within Mycale. Hó | 0.29 | 026 | 035 | 033 | 041 Major advantages in using H7 | 038 | 028 | 029 | 0.17 | 017 | 036 RAPD techniques include their H8 | 033 | 020 | 040 | 029 | 029 | 040 | 024 simplicity, reduced running time, H9 0.40 0.29 0.30 0.45 0.28 0.38 0.40 0.35 RAPD ANALYSIS OF SPONGES TABLE 6. Number and percentage of polymorphic fragments amplified with primer UBC-322 for each individual of Hymeniacidon heliophila shown in Fig. 3. Fragments scored in the range 200-1200bp. Abbreviations: PV, Praia Vermelha Beach; U, Urca Beach; BV, Boa Viagem Beach; IT, Itaipu Beach; SS, Cabelo Gordo Beach, Sáo Sebastiáo. Specimens of H. heliophila m poly "| Total no. | % pomar / beach site fabres fragments P irl specimen 1/PV 11 17 65% specimen 2/PV 9 15 60% specimen 3/U 12 18 67% specimen 4/U 13 19 68% specimen 5/BV 12 18 67% specimen 6/BV 8 14 57% specimen 7/IT 6 12 50% specimen 8/IT 8 14 57% specimen 9/SS 8 14 57% specimen 10/88 10 16 63% Mean 10 16 61% FIG. 2. DNA amplification profiles obtained by RAPD analysis of Hymeniacidon heliophila specimens using primer UBC-322. Key: MW, molecular-weight size markers (A DNA digested with EcoRI and HindIII) in base pairs. Lanes 1-9, nine individuals from the Itaipu Beach population, lane 10, PCR control reaction without DNA. Arrows show polymorphic (upper right and lower left), and monomorphic (middle right) bands. account for this by taking into consideration that one of the first original works describing the technique (Williams et al., 1993), showed that in a competitive amplification assay using up to 460-fold molar excess of a cyanobacterial genomic DNA mixed with DNA from a soybean genome, all of the detectable amplified products were from soybean. This result suggests that the outcome of an amplification reaction is determined in part by competition for priming sites in the genome, and that a genome of high complexity (soybean) should have more target sites with a better complement to a RAPD primer, as compared to a genome of low complexity 324 MEMOIRS OF THE QUEENSLAND MUSEUM MW 1 2 3 4 5 6 7 8 9 10 11 FIG. 3. DNA amplification profiles of individuals from different populations of A. heliophila obtained by RAPD analysis with primer UBC-322. Key: MW, molecular-weight size markers in base pairs. Lanes 1-2, Praia Vermelha Beach; lanes 3-4, Urca Beach; lanes 5-6, Boa Viagem Beach; lanes 7-8, Itaipu Beach; lanes 9-10, Cabelo Gordo Beach; lane 11, PCR control reaction without DNA. The arrows show polymorphic (lower right) and monomorphic (middle left) (this band was used on dot-blot) bands. (cyanobacteria). Symbionts are d TABLE 7. Genetic distances, based on Jaccard's similarity index, commom problem for the majority among 10 individuals of Hymeniacidon heliophila from 5 populations of the molecular techniques calculated from the presence or absence of the amplification products including allozyme analysis and ofprimer UBC-322. Complete identity is 0 and complete difference is DNA sequencing. Although, 1. Mean genetic distance among the 10 individuals from the 5 beaches RAPD gels can be blotted and was 0.53. Abbreviations: PV, Praia Vermelha Beach; U, Urca Beach; markers can be checked with BV, Boa Viagem Beach; IT, Itaipu Beach; SS, Cabelo Gordo Beach, sponge derived probes, we are São Sebastião. currently approaching the ‘symbiont problem’ through the purification of sponge cells, the use SAS A A A a A E -ESSI of dissociated sponge cells in PV? | 92? RAPD amplification reactions, and | V! | 060 | 0.57 the comparison of the resulting | U2 | 0.56 | 052 | 032 RAPD pattern with the one | BVI | 0.65 | 0.73 | 0.62 | 058 obtained for the whole sponge. BV2 | 0.65 | 0.62 | 0.55 | 0.50 | 0.32 Fs APT ITI | 0.62 | 0.65 | 0.50 | 0.52 | 0.57 | 0.38 Hymeniacidon heliophila showed pestes cavo Eus l nsn esse Fran bin high levels of genetic variation _—_ ——L : -o = within and between populations ssı | 0.71 | 0.68 | 0.48 | 0.50 | 0.48 | 0.44 | 038 | 0.35 Allozyme electrophoresis studies | S82 | 0.57 | 0.65 | 0.58 | 0.48 | 045 | 0.64 | 0.60 | 0.57 | 042 RAPD ANALYSIS OF SPONGES 325 TABLE 8. Genetic distances, based on Jaccard’s similarity index, among 6 species of Mycale calculated from the presence and absence of the amplification products of primers OPG-19 (A) and OPS-17 (B). Mean genetic distance among the 6 species (considering just M. angulosa SS) was 0.76 (A) and 0.80 (B). Abbreviations: PT, Parati, Rio de Janeiro State; SS, Cabelo Gordo Beach, Sao Sebastiao. (A) OPG-19 M. arenaria poi ein M. escarlatei M. laxissima | M. microsigmatosa | M. angulosa PT M. aff. americana 0.85 M. escarlatei 0.74 0.62 M. laxissima 0.81 0.63 0.77 M. microsigmatosa 0.80 0.87 0.93 0.92 M. angulosa PT 0.75 0.67 0.71 0.78 0.76 M. angulosa SS 0.79 0.45 0.70 0.67 0.86 0.65 (B) OPS-17 M. aff. americana 0.90 M. escarlatei 0.78 1] 0.83 M. laxissima 0.81 0.86 0.88 M. microsigmatosa 0.78 0.78 0.64 0.69 M. angulosa PT 0.90 0.90 0.93 0.68 0.93 M. angulosa SS 0.89 0.85 0.84 0.71 0.71 0.65 have already shown that sponges are among the most genetically variable organisms (Solé-Cava & Thorpe, 1994, and references therein). For ex- ample, Solé-Cava & Thorpe (1994) showed that the average proportion of polymorphic loci for 21 sponge species was 0.50 (range 0.11-0.86). Our results, with an average genetic distance of 0.32 for the intrapopulation variation (Itaipu samples) and of 0.53 for the interpopulation samples, show that RAPD analysis also points to the existence of high levels of genetic variation, at least for this sponge species. 01 02 03 04 05 06 13 14 15 16 17 18 6060000 25 26 27 28 29 30 see 37 38 39 The dot-blot experiment was used to determine whether the isolated RAPD marker was specific to Hymeniacidon, or also present in other species. Our results indicate that this marker is specific to populations of Hymeniacidon within the regions studied, or perhaps to the genus Hymeniacidon. To determine more accurately the taxonomic value of this marker, we are presently expanding our sponge samples to other species of Hymeni- acidon and other genera of the Halichondriidae. Cloning and sequencing of this marker is also in progress. 07 08 09 10 11 12 19 20 21 22 23 24 ? * 9 (49.0 4 31 32 33 34 35 36 FIG. 4. Dot-Blot hybridised to a RAPD monomorphic marker (shown in Fig. 3) from Hymeniacidon heliophila. Key: Wells: 01, H. heliophila; 02, Callyspongia sp.; 03, Cinachyrella alloclada; 04, Ectyoplasia ferox; 05, Gastrophanella sp.; 06, Aiolochroia crassa; 07, Mycale (Mycale) arenaria; 08, Mycale (Aegogropila) aff. americana; 09, Mycale (Aegogropila) escarlatei; 10, Mycale (Arenochalina) laxissima; 11, Mycale (Carmia) microsigmatosa; 12, Mycale (Zygomycale) angulosa; 13, Plakortis sp.; 14, Pseudaxinella reticulata; 15, Topsentia ophiraphidites; 16, Aplysina fulva; 17, Amphimedon viridis; 18, Leucilla sp.; 19, fruit fly; 20, rat; 25-39, H. heliophila from five populations (25-27, Praia Vermelha; 28-30, Urca; 31-33, Boa Viagem; 34-36, Itaipu; 37-39, Cabelo Gordo). MEMOIRS OF THE QUEENSLAND MUSEUM FIG. 5. DNA RAPD amplifications of Mycale species. A, Amplification with primer OPG-19. B, Amplification with primer OPS-17. Key: MW, molecular-weight size markers in base pairs. Lanes: 1, Mycale (Mycale) arenaria; 2, M. (Aegogropila) aff. americana; 3, M. (A.) escarlatei; 4, M. (Arenochalina) laxissima; 5, M. (Carmia) microsigmatosa; 6, M. (Zygomycale) angulosa (Parati, Rio de Janeiro State); 7, M. (Z.) angulosa (Cabelo Gordo Beach, São Sebastião, São Paulo State); 8, Arenosclera sp. (Joao Fernandinho Beach, Búzios, Rio de Janeiro State). Markers generated by RAPD analysis have been used extensively to distinguish species (Bardakci & Skibinski, 1994; Coffroth & Mulawka, 1995; Appa Rao et al., 1996; Partis & Wells; 1996, De Bustos et al., 1998). We applied this method to a set of species of Mycale to check its utility in exploring phylogenetic affinities. Mycale is well represented at the SE Brazilian coast and phylogenetic affinities within this genus were recently generated (Hajdu & Desqueyroux-Faúndez, 1994; Hajdu & Riitzler, 1998; Hajdu, 1999, this volume). Due to the high levels of intra- and interpopulational genetic diversity obtained for Hymeniacidon, it will be necessary to survey several (5-10 at least) speci- mens of each Mycale species before deriving conclusions at the interspecific level. This work is in progress, even though the levels of inter- specific variation for the six Mycale studied here were consistently high for the two primers used. We recognise that our results are still preliminary and stress that our experiments have been conducted to explore the potential utility of RAPD-PCR techniques to assess phylogenetic relationships of congeneric species. We are cur- rently directing our efforts towards increasing the number of primers, specimens and species considered. RAPD-PCR divergence rates cannot be used as a universal clock with an invariant rate in all animals (Borowsky, 1998). However, a strong relationship seems to exist between degrees of RAPD pattern divergence and time since separation of isolated taxa (Borowsky, 1998). Data from this study show that the mean genetic distance increases from the intrapopulation (0.32) through the interpopulation (0.53) to the interspecific level (0.76 and 0.80). In fact these results support our assumption that there was no amplification of symbiotic DNA, as the presence of symbionts would not correlate well with the degree of relatedness of individuals; in other words, would not be homogeneous at all levels. Our main conclusion is that RAPD patterns show a good correlation between genetic distances and taxonomic level. Our results demonstrate that molecular markers generated by the PCR-RAPD technique provide an effective tool to assess the existing genetic polymorphism within and between populations and species of sponges. As a future outcome of the present study, specific markers for each sponge population or species can be isolated and sequenced, and population/ species-specific primers can be generated. This would allow rapid screening of a larger number RAPD ANALYSIS OF SPONGES of individuals and thus ease the genetic ident- ification of field samples, more detailed study of the genetic structure of sponge populations, and construction of phylogenies for congeneric species, ACKNOWLEDGEMENTS We thank M.L. Rosario for technical assistance, and E. Kalapothakis for donation of chemicals. J.C. de Freitas, Director of CEBIMAR-USP, is thanked for the provision of laboratory facilities during field collections at Sao Sebastião. This work was carried out with financial assistance from Fundacào de Amparo à Pesquisa do Estado do Rio de Janeiro (FAPERJ). GLH and GM were supported by a fellowship from Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq). EH is grateful to Fundagáo de Amparo à Pesquisa do Estado de Sao Paulo (FAPESP). LITERATURE CITED ARMSTRONG, J.S., GIBBS, A.J., PEAKALL, R. & WEILLER, G.F. 1994. RAPDISTANCE: random amplified polymorphic DNA analysis. Computer program, Version 1.03 The RAPDistance Package. (Australian National University: Canberra) (http://life.anu.edu.au/molecular/software/ rapd.html). APPA RAO, K.B.C., BHAT, K.V. & TOTEY, S.M. 1996. Detection of species-specific genetic markers in farm animals through random amplified polymorphic DNA (RAPD). Genetic Analysis, Biomolecular Engineering 13: 135-138. AVISE, J.C. 1994. Molecular Markers, Natural History, and Evolution. (Chapman & Hall: New York). BARDAKCI, F. & SKIBINSKI, D.O.F. 1994, Ap- plication of the RAPD technique in tilapia fish: species and subspecies identification. Heredity 73: 117-123. BERGQUIST, P.R. 1978. Sponges. (Hutchinson & Co.: London). BOROWSKY, E.L. 1998. Evolutionary divergence of AP-PCR (RAPD) patterns. Molecular Biology and Evolution 15(4): 408-414. BURTON, R.S. 1996, Molecular tools in marine ecology. Journal of Experimental Marine Biology and Ecology 200: 85-101. CHALMERS, K.J., SPRENT, J.L., SIMONS, AJ., WAUGH, R. & POWELL, W. 1992. Patterns of genetic diversity in a tropical tree legume (Gliricidia) revealed by RAPD markers. Heredity 69: 465-472. CHESHIRE, A.C. & WILKINSON, C.R 1991. Model- ling the photosynthetic production by sponges on Davies Reef, Great Barrier Reef. Marine Biology 109: 13-18. 327 COFFROTH, M.A. & MULAWKA-III, J.M. 1995. Identification of marine invertebrate larvae by means of PCR-RAPD species-specific markers. Limnology and Oceanography 40(1): 181-189. COLLINS, A.G. 1998. Evaluating multiple alternative hypotheses for the origin of Bilateria: an analysis of 18S rRNA molecular evidence. Proceedings of the National Academy of Sciences USA 95(26): 15458-63. DE BUSTOS, A., CASANOVA, C., SOLER, C. & JOUVE, N. 1998. RAPD variation in wild populations of four species of the genus Hordeum (Poaceae). Theoretical and Applied Genetics 96: 101-111. DEMEKE, T., ADAMS, R.P. & CHIBBAR, R. 1992, Potential taxonomic use of random amplified polymorphic DNA (RAPD): a case study in Brassica. Theoretical and Applied Genetics 84: 990-994. HAJDU, E. 1999, Toward a phylogenetic classification of the mycalids with anisochelae (Demo- spongiae:Poecilosclerida), and comments on the status of Naviculina Gray, 1867. Memoirs of the Queensland Museum (this volume). HAJDU, E. & DESQUEYROUX-FAUNDEZ, R. 1994. A synopsis of South American Mycale (Mycale) (Poecilosclerida, Demospongiae), with the description of three new species and a cladistic analysis of Mycalidae. Revue suisse de Zoologie 101(3): 563-600. HAJDU, E. & RUTZLER, K. 1998. Sponges, genus Mycale (Poecilosclerida: Demospongiae: Porifera), from a Caribbean mangrove and com- ments on subgeneric classification. Proceedings of the Biological Society of Washington 111(4): 737-773. HILLIS, D.M., MORITZ, C. & MABLE, B.K. (eds) 1996, Molecular Systematics. 2nd Edition. (Sinauer Associates. Sunderland, MA.). JACCARD, P. 1901. Etude comparative de la distribution florale dans une portion des Alpes et des Jura. Bulletin de la Société Vaudoise de Science Naturel 37: 547-579. 1908. Nouvelles recherches sur la distribution florale. Bulletin de la Société Vaudoise de Science Naturelle 44: 223-270. JACOBSEN, R., FORBES, V.E. & SKOVGAARD, O. 1996. Genetic population structure of the proso- branch snail Potamopyrgus antipodarum (Gray) in Denmark using PCR-RAPD fingerprints. Proceedings of the Royal Society London (B) 263: 1065-1070. KELLY-BORGES, M., BERGQUIST, P.R. & BERG- QUIST, P.L. 1991. Phylogenetic relationships within the Order Hadromerida (Porifera, Demo- spongiae, Tetractinomorpha) as indicated by ribosomal RNA sequence comparisons. Biochemical Systematics and Ecology 19(2): 117-125. KELLY-BORGES, M. & POMPONI, S.A. 1994. Phylogeny and classification of lithistid sponges (Porifera: Demospongiae): a preliminary assessment using ribosomal DNA sequence comparisons. Molecular Marine Biology and Biotechnology 3(2): 87-103. KHANDKA, D.K., TUNA, M., TAL, M., NEJIDAT, A. & GOLAN-GOLDHIRSH, A. 1997. Variability in the pattern of random amplified polymorphic DNA. Electrophoresis 18: 2852-2856. KNOWLTON, N. 1993. Sibling species in the sea. Annual Review of Ecology and Systematics 24: 189-216. LAKSHMI, M., RAJALAKSHMI, S., PARANI, M., ANURATHA, C.S. & PARIDA, A. 1997. Molecular phylogeny of mangroves I. Use of molecular markers in assessing the intraspecific genetic variability in the mangrove species Acanthus ilicifolius Linn. (Acanthaceae). Theoretical and Applied Genetics 94; 1121-1127. LEVITAN, D.R. & GROSBERG, R.K. 1993. The analysis of paternity and maternity in the marine hydrozoan Hydractinia symbiolongicarpus using randomly amplified polymorphic DNA (RAPD) markers. Molecular Ecology 2: 315-328. LEWIS, P.O. & SNOW, A.A. 1992, Deterministic pat- ernity exclusion using RAPD markers. Molecular Ecology 1: 155-160. PARTIS, L. & WELLS, R.J. 1996. Identification of fish species using random amplified polymorphic DNA (RAPD). Molecular and Cellular Probes 10: 435-441. PRESTON, C.M., WU, K.Y., MOLINSKI, T.F. & DELONG, E.F. 1996. A psychrophilic crenarchaeon inhabits a marine sponge: Cenarchaeum symbiosum gen. nov., sp. nov. Proceedings of the National Academy of Science USA 93: 6241-6246. RIEDY, M., HAMILTON, W.J. & AQUADRO, C.F. 1992. Excess of non-parental bands in offspring from known primate pedigrees assayed using RAPD PCR. Nucleic Acids Research 20: 918. SANGUINETTI, C.J., DIAS NETO, E. & SIMPSON, A.J.G. 1994. Rapid silver staining and recovery of MEMOIRS OF THE QUEENSLAND MUSEUM PCR products separated on polyacrilamide gels. Biotechniques 17(5): 915-919. SHAW, C.R. & PRASAD, R. 1970. Starch gel electro- phoresis of enzymes — a compilation of recipes. Biochemical Genetics 4: 297-320. SOLE-CAVA, A.M. & THORPE, J.P. 1991. High levels of genetic variation in natural populations of lower marine invertebrates. Biological Journal of the Linnean Society 44: 65-80. 1994, Evolutionary genetics of marine sponges. Pp. 55-63. In Soest, R.W.M. Van, Kempen, T.M.G. van & Braeckman, J.C. (eds) Sponges in time and space. (Balkema: Rotterdam). STILES, J.F., LEMME, C., SONDUR, S., MORSHIDI, M.B. & MANSHARDT, R. 1993. Using random amplified polymorphic DNA for evaluating genetic relationships among papaya cultivars. Theoretical and Applied Genetics 85: 697-701. TABERNER, A., DOPAZO, J. & CASTANERA, P. 1997. Genetic characterization of populations ofa de novo arisen sugar beet pest, Aubeonymus mariaefranciscae (Coleoptera, Curculionidae), by RAPD analysis. Journal of Molecular Evolution 45: 24-31. VACELET, J. & DONADEY, C. 1997. Electron micro- scope study of the association between some sponges and bacteria. Journal of Experimental Marine Biology and Ecology 30: 301-314. WILKINSON, C.R. 1978. Microbial associations in sponges. III. Ultrastructure of the in situ assoc- iations in coral reef sponges. Marine Biology 49: 177-185. WILLIAMS, J.G.K., HANAFEY, M.K., RAFALSKI, J.A. & TINGEY, S.V. 1993. Genetic analysis using random amplified polymorphic DNA markers. Methods in Enzymology 218: 704-740. WILLIAMS, J.G.K., KUBELIK, A.R., LIVAK, K.J., RAFALSKI, J.A. & TINGEY, S.V. 1990. DNA polymorphisms amplified by arbitrary primers are useful as genetic markers. Nucleic Acids Research 18: 6531-6535. WELSH, J. £ McCLELLAND, M. 1990. Fingerprinting genomes using PCR with arbitrary primers. Nucleic Acids Research 18: 7213-7218. MOLECULAR TECHNIQUES REVEAL WIDE PHYLETIC DIVERSITY OF HETEROTROPHIC MICROBES ASSOCIATED WITH DISCODERMIA SPP. (PORIFERA: DEMOSPONGIAE) JOSE V. LOPEZ, PETER J. McCARTHY, KATHLEEN E. JANDA, ROBIN WILLOUGHBY AND SHIRLEY A. POMPONI Lopez, J.V., McCarthy, P.J., Janda, K.E., Willoughby, R. & Pomponi, S.A. 1999 06 30: Mo- lecular techniques reveal wide phyletic diversity of heterotrophic microbes associated with Discodermia spp. (Porifera:Demospongiae). Memoirs of the Queensland Museum 44: 329-341. Brisbane. ISSN 0079-8835. Sponges are well known to harbor large numbers of heterotrophic microbes within their mesohyl. Studies to determine the diversity of these associated microbes have been attempted for only a few shallow water species. We cultured various microorganisms from several species of Discodermia collected from deep water using the ‘Johnson-Sea-Link’ manned submersibles, and characterised them by standard microbiological identification methods. Characterisation of a small proportion (ca. 10%) of the total and potential eubacterial isolate collection with molecular systematics techniques revealed a wide diversity of microbes. Phylogenetic analyses of 32 small subunit (SSU) 16S-like rRNA gene sequences from different microbes indicated high levels of taxonomic diversity associated with this genus of sponge. For example, bacteria from at least five eubacterial subdivisions — gamma, alpha, beta, Cytophaga and Gram positive — were isolated from the mesohyl of Discodermia. Several strains were unidentifiable from current sequence databases. No overlap was found between sequences of 24 isolates and 8 sequences obtained by PCR and cloning directly from sponge samples. The abundance and diversity of microbes associated with sponges such as Discodermia suggest that they may play important roles in marine microbial ecology, dispersal and evolution. O Porifera, Discodermia, microbial diversity, bacterial symbionts, in vitro culture, gene sequencing, 16S rRNA. Jose V. Lopez (email: Lopez@hboi.edu), Peter J. McCarthy, Kathleen E. Janda, Robin Willoughby & Shirley A. Pomponi, Division of Biomedical Marine Research, Harbor Branch Oceanographic Institution, 5600 US 1 North, Fort Pierce FL 34946, USA; 16 February 1999. Many marine sponges harbor numerous microbial symbionts or ‘associates’. Researchers enumerating and characterising sponge-microbial interactions have shown that bacterial biomass can reach over 50% in some marine sponges with corresponding phenomenally huge diversity (Vacelet & Donaday, 1977; Simpson, 1984; Wilkinson, 1987; Santavy et al., 1985; Fuerst et al., 1999, this volume). Although studies of sponge- bacterial associations are hampered by upper microorganism culturability limits, estimated to be below 1-10% (Austin, 1988; Eguchi & Ishida, 1990; Button et al., 1993), focus on marine microbial ecology and taxonomy is increasing. For example, novel aspects of diverse microbes and their habitats are being revealed by molecular genetics approaches (Amann et al., 1995; McInerney et al., 1995; Distel et al., 1988; Delong, 1998; R. Hill pers.comm.). An understanding of the biology of marine sponges in the polyphyletic order ‘Lithistida’ would be enhanced by characterisation of the constituent organisms coexisting with these sponges, e.g. species richness, ecological function, metabolic load, etc. Lithistids in general, and Discodermia in particular, are the source of several compounds with pharmaceutical potential (Longley et al., 1991; Kelly-Borges et al., 1994; Gunasekera et al., 1994). Identification of microbial associates in Discodermia could lead to a better under- standing ofthe ecological or physiological role of these compounds. In this study, we provide a preliminary description of eubacterial species diversity associated with species of Discodermia, using both microbiological and molecular tools. A primary goal of this exercise was the ident- ification and matching of Discodermia—associated microbe sequences to the closest rRNA relative in current sequence databases. Potentially uncultivable microorganisms were characterised using the polymerase chain reaction (PCR), cloning and DNA sequencing of cloned 16S-like small subunit (SSU) rRNA gene segments (Pace et al., 1986; Lane et al., 1991; Delong, 1998). us IE © Molecular phylogenetic analyses of these sequences highlight the diversity of lineages found within a single sponge genus. Final systematic resolution of each isolate based on rRNA sequence data alone is not attempted in this study. MATERIALS AND METHODS COLLECTION AND ISOLATION OF MICROORGANISMS FROM MARINE MACROORGANISMS. Specimens belonging to Discodermia spp. used as sources of microbial isolates characterised in this study are listed in Table 1. Isolates from several non-Discodermia sponges (Dercitus, Halichondria and Corticium spp.) were included for a cursory comparison with microbe profiles of Discodermia spp. (Table 2). All specimens were obtained in accordance with the permits and rules granted by the cooperating sovereign governments. Marine macroorganisms were collected either by SCUBA or through the use of the ‘Johnson- Sea-Link’ manned submersibles. On return to the surface, sponge samples used for microbial isolation were immediately sub-sampled and prepared for plating using the following method. The sponge was surface-sterilised by rinsing with 70% (v/v) ethanol. A cube of mesohyl (approx. lcm") was removed using aseptic technique and placed in 20ml sterile artificial sea water (ASW). The sample was then ground in a Waring blender pre-sterilised with 70% (v/v) ethanol, and the resulting suspension was serially diluted with sterile ASW. The dilution series was used as the inoculum for a series of isolation plates. Typically, 10011 of the 10°, 10° and 10° dilutions were used since these were found to bracket the range at which discrete colonies were found. Sponge cell dissociation and selective cell enrichment were performed by dissociating fresh sponge sections in calcium- and magnesium-free artificial seawater (CMF) (Pomponi & Willoughby, 1994) in a Virtis grinder for 10secs. The resulting slurry was filtered through a 70um strainer and cells were checked for viability. Cell suspensions were then placed into 1 5ml tubes and centrifuged to obtain enriched fractions of sponge cells, unicellular bacteria and filamentous bacteria. Several microbial isolates (e.g. K200, K202, K261) were obtained by fractionating dissociated sponge mesohyl through Percoll/CMF Gradients. A 10ml working solution of 15% Percoll/CMF was diluted with 5M Tris [pH 8.0] from a 90% Percoll/CMF stock solution, and placed in 15ml MEMOIRS OF THE QUEENSLAND MUSEUM centrifuge tubes at -75°C for 30mins. About I ml of sponge mesohyl (5-10g ground in 20ml of filtered sea water) was layered on top of the thawed Percoll/CMF working solutions and centrifuged at 2,000 RPM for 20mins. Tubes were punctured approximately 2.5cm from the bottom, and five 2ml fractions were collected into 24 well plates. About 100p1 of each fraction was then plated onto the appropriate isolation media (see below). Isolation media used in this study were: 1) Chitin Sea Water: Colloidal chitin (in approx. 25ml deionised H20), 0.25g dry weight/liter medium; agar, 20g; ASW, 975ml. 2) LN: Bacto Peptone (Difco), 0.5g; Yeast Extract (Difco), 0.5g; agar, l6g; 75% (v/v) ASW, 1L. 3) M3: K;HPO,, 0.466g; NaH;PO, 0.732g; KNO; 0.1g; MgSO,.7H50, 0.1g; Na propionate, 0.2g; NaCl, 0.29g; CaCO;, 20mg; FeSO,, 200mg; ZnSO,, 180mg; MnSO, 20mg; agar, 18g in 1L deionised H20. Cycloheximide (50mg) and thiamine (4mg) were added after autoclaving. 4) ISP2: Difco ISP2 prepared with 75% (v/v) ASW. 5) HSV (humic acid, sodium salt), 1g; glycerol phos- phate, 110mg; 75% (v/v) ASW, 1L; 10ml BME Vitamin Mix (Sigma) added after autoclaving. Inoculated plates were incubated at ambient temperature (approx. 25°C) for 2-4 weeks. After this period of incubation, discrete colonies were transferred to fresh plates of the isolation medium, incubated and then re-streaked until the isolate was axenic. Isolates were then transferred to Marine Agar 2216 (Difco). All strains used in this study were maintained as Marine Agar 2216 slant cultures. DNA ANALYSIS. Genomic DNA from bacterial isolates and sponge mesohyl fractions was extracted using standard purification methods (Sambrook et al., 1989), although some of this DNA required alternative purification methods or modifications which have been previously described (Pitcher et al., 1989). Segments of the 16S-like small subunit (SSU) nuclear rRNA gene amplified from bacterial cultures were sequenced directly after purification of the PCR product. Products derived from sponge mesohyl fractions were cloned before sequencing based on the following procedure. The universal eubacterial primers, Ecoli9 [5* GAG TTT GAT CAT GGC TCAG 3”] and Loop27re [5° GAC TAC CAG GGT ATC TAA TC 3”], amplify about 800bp of the 5’ end of SSU rRNA gene under standard PCR conditions (Lane, 1991). The segments encompass variable HETEROTROPHIC MICROBES IN DISCODERMIA 331 TABLE 1. Profile of Discodermia samples used for microbial characterisation. Key: 1, sample was obtained at a similar location and depth as for the samples of species not belonging to Discodermia (as indicated in Table 2). 2, Discodermia sample 20-X1-97-1-001 is listed twice because it was used to both a) isolate 51 microorganisms of which two have been analysed, and b) to derive the eight PCR-amplified products comprising the 28 and 29 clone series, which were not cultured. sno E Location collected Depth (ft.) Total di No. sequences analysed 21-111-87-3-014 Bahamas 100 5 2 18-111-87-3-001 Bahamas 515 2 2 8-X1-90-1-001 .. Bahamas 592 9 2 1-XII-92-2-001 Bahamas 110 13 1 7-XII-92-2-001 ! Bahamas 540 8 1 17-X11-92-1-005 Bahamas 535 10 1 15-1-96-2-012 Bahamas 520 1 1 27-X-96-1-003 Bahamas 543 59 7 29-X-96-4-006 Bahamas 520 — 8 1 9-X1-97-3-008 Honduras 383 65 E 16-X1-97-1-004 Honduras 440 100 2 20-X1-97-1-001 ? Honduras 415 51 2 20-X1-97-1-001 ? Honduras 415 8 Totals 331 32 regions V1—V4 of bacterial SSU rRNA (E. coli positions 9-804) and were expected to provide sufficient genetic variation for phylogenetic analyses (Lane, 1991; Liesack et al., 1991). This size also facilitated complete sequencing of both strands with a minimal number of sequencing reactions, using the above amplification primers and internal primers int-250f[5’ GAC TCC TAC GGG AGG CAG 3’ | and int-275rc [5° CAC GCG GCG TCG CTG CAT 3’]. The typical PCR amplification profile was 94°C denaturation for 45secs, 53-55°C annealing for 60secs, and 72°C extension for 60secs, repeated for 30 cycles. PCR products conforming to expected molecular weights were purified by gel isolation or Qiagen columns (Qiagen), and sequenced by the dye- terminator cycle sequencing method (Applied Biosystems Inc - ABI) run on ABI 373 automated DNA sequencers (University of Florida, ICBR, DNA Sequencing Core Lab, Gainesville FL). To identify potentially uncultivable microbes, PCR products derived from either total mesohyl preparations or enriched mesohyl fractions were ‘shotgun’ cloned into TA vectors according to manufacturers' instructions (Invitrogen). Clones derived in this manner were designated a number beginning with either 28 (enriched fraction) or 29 (total mesohyl preparation). All rRNA gene sequences were analysed with the GCG DNA analysis package (GCG, 1994) and SEQED data editor (ABI). The most conserved rRNA sequences relative to Discodermia- derived microbes were identified using queries generated by SIMILARITY- RANK in the Ribosomal Database Project (Maidak et al., 1994), or by BLAST using GenBank (Altschul et al., 1990). Preliminary alignments of sequences were made using PILEUP (GCG, 1994), followed by a manual verification for the presence of canonical rRNA secondary structures and compensatory base changes (Neefs et al., 1993; Gutell et al., 1994) (also see Fig. 1). A gap extension penalty of 1, rather than the default of 4, maximised similarity by allowing longer gaps. Maximum parsimony analysis was performed with PAUP, version 3.1.1, while neighbour-joining (NJ) and maximum likelihood (ML or DNAML) analyses were performed with PHYLIP 3.572 software, (Felsenstein, 1993; Swofford, 1993; Hillis et al., 1996). Bootstrap replications of datasets were performed a minimum of 100 times and individual taxa (or operational taxonomic units, or OTUs) which appeared to be problematic (exhibiting long branch lengths or many uninformative nucleotide substitutions) were jackknifed (Efron, 1982). Typical heuristic searches in PAUP utilised evaluations of at least 50 replications of random sequence additions, tree bisection-reconnection (TBR), and decay index assessments (Hillis et al., 1996). Subsets of the total dataset of over 40 OTU’s (including uy N MEMOIRS OF THE QUEENSLAND MUSEUM A stem loop stem 13a 13 13a’ B853 CCLAGGCGACGATTCCTAGCTGGT-CT GAGA GGATGAT-CAGCCACACTG Ri21 we ae a wn aos CTÀ.... aces aa lee wee ee am Tene dar o M465 = J ..... Cao... CTÀ.... es -T. T... moon. oo E TERET M529 = waveune [PT Bi Ga ob GA C— ene ore Ge ig cee Ge pies K146 2 a he ee ae CGGG....C.. C- + GC. .Co Gee ewe ee K275- = wt win ole wef ere han (057 AA o wess Ah d n S rfe A VEA OR 29B2 eebsucavs T.,.CGÀ.. OIE.- NN Mte DP G. 20k CCC ade ja sy Grille... sees ae Ba rues e tate 280 Me GT ow Ds. «Ll. ¿GO 41. ,GGC. Y. B stem bulge stem loop stem bulge stem 17a 17a i7b 17 17b’ 17a’ 17a” B853 TGATGCAGCCATG-CCGCGTGTATGAAGAA GGCC TTC-G GGTT GTAAA--GTACTTTC ITA at aif cae e “sews names Erin ds E a == PRA K127 wa we a be ee wee Tornanoo»o B.r..4-.4 oses sé. Zo pass acoso m TEETER E034 a te ee as la nde Git whose es feel fs a E ARE bioa == Cs tive B549 O: PA APA AAA AA Ta apos vorre =-.L.. T K121 «co Ds oe ee eu Sis tate Wt a € Eos. fe et oe ÁS. mara muaa aaa M485 Sort Eod s s E Xi v pa AG. CT atado NO ae re ==. LT rro... M529 eC Ga. Came. A APN: y 2..—— A --ACT..G.T 29B2 PROTON DIOE CREE ».«EB..Cuooo EE... An. CTICC...AÀ 26D2 CETT ae dr © Coens en Ge Toro hen ctn e Sare AE T 294 was Gee ep aul aa vid AG. Tor. ah Å=. uens osons --.CT..... FIG. 1. Representative partial alignments of SSU rRNA regions with conserved secondary structures (Neefs et al., 1993; Gutell et al., 1994), Structures corresponding to: A, Stem/Loop 13; B, Stem /Loop and Bulge 17 are shown over the nucleotide sequences. Both strands of a stem are underlined with the downstream structure marked by a (*) above the sequences. Stem 13 in (A) and stem 17 (B) begin at E. coli position 288 and 405, respectively. Stem 13a’ contains an asymmetric bulge, lengthening it relative to upstream 13a. Nucleotides identical to the first reference sequence are shown below as dots (.); gaps are indicated by (-); compensatory mutations in all stems are underlined. outgroups listed below) were analysed to verify support for each clade. For NJ, genetic distances were calculated with DNADIST, also in PHYLIP, using Kimura’s 2N parameter correct- ion for multiple substitutions and empirically derived transition/transversion ratio of 2.0. Although each maximum likelihood analysis was limited to a subset of 20 OTUs, different taxa within major clades were interchanged and substituted in separate runs of the program to monitor consistency of the consensus topology. NJ and DNAML analyses were performed on a Digital AlphaServer 8400 mainframe computer maintained at the Frederick Biomedical Super- computer Center in Frederick, Maryland. Chimeric sequences were detected by com- paring the identity of 5’ and 3” ends separately before making contiguous constructs, or by using the CHECK-CHIMERA option in RDP (Liesack et al., 1991; Maidak et al., 1994; Rheims et al., 1996). When chimeric products were found, each respective terminal sequence was analysed as a single entity up to the artificial crossover junction and included in phylogenetic analyses despite the shorter length of rRNA sequence. Sequences (with GenBank accession numbers) of the following representative eubacterial strains and genera were used as outgroups or as ref- erence sequences for the major clades observed in phylogenetic reconstructions: Bacillus firmus (X60616), Pseudomonas straminea (D84023), Capnocytophaga sp. (X97245), Alteromonas macleodii (X82145), Vibrio alginolyticus (X74690), Ridgeia piscesae (U77480), Thermotoga maritima (M21774), Actinomycetes spp. (X92705), Rhodobium marinum (M27534), Burkholderia solanacearum (U28232), Clostridium sp. (L09175, X77837), Chloroflexus (D38365), Lyngbya (AJ000714) and an unidentified marinobacter (U61848). RESULTS MICROBIAL ISOLATIONS. Phenotypic identifications of microbial isolates were made by analyses of colony morphology, microscopic observation and Gram staining. A general HETEROTROPHIC MICROBES IN DISCODERMIA TABLE 2. Comparison of isolates from different sponge species collected at similar depths, locations and times as for samples of Discodermia (listed in Table 1). The number given for Discodermia is the average for the three samples analysed. No. of Isolates No. of sponge [~~ Sponge taxon | samples used | Gram- | Gram- for analysis | Positive | negative | Fungi bacteria | bacteria Discodermia 3 $ 3 L5 | Dercitus 1 4 4 Corticium 1 8 1 Halichondria 1 3 1 3 taxonomic grouping of microbial isolates was obtained from different sponge taxa proximal to, or at similar location, depth and time as Discodermia samples (Tables 2, 3). Although major microbial groups such as eubacteria, actinomycetes and fungi were identified, archae- bacteria and protists were not cultured. Fungal isolates and their sequences will not be discussed in this paper. To investigate any possible trends or biases in microbial isolations, the number of isolates in different microbial categories obtained from different sponge taxa and different subsamples are summarised in Table 3. These values indicate that there are different profiles of microbial populations among different species of sponges, and that there is also potential variation in microbial yields among different samples ofa particular sponge species. For example, some variation in isolate yields appeared to be related to geographical differences in sample collection, whereas specimens of particular species from different depths did not exhibit any strong trends. Some variation may also be attributed to changes in the criteria used to select colonies for isolation and in the media used for certain experiments. EUBACTERIAL STRAIN IDENTIFICATION. The names of eubacterial strains resulting from highest identity scores are listed in Table 4, together with their corresponding percentage sequence identities. Relatively good agreement was observed between the two major sequence datasources, RDP and GenBank databases, with novel bacterial rRNA sequences from cultured isolates used to infer possible taxonomic placements. However, several caveats to this procedure should be emphasised. 1) Database searches revealed only the most similar sequence in the respective database, and these were not considered as an absolute identification of a particular ‘species’? or strain of bacteria, even uy U3 I] when sequence identity exceeded 99% (Fox et al., 1992). Indeed, it is probably more appropriate to make reference to a specific ‘rRNA type’ or strain than to infer that these identifications are homologous to species-level taxonomy. 2) Since it was not possible to obtain multiple operon sequences or samples of any given isolate, it was therefore not possible to assess possible intra- strain or intraspecific variation amongst the microbial taxa (Clayton et al., 1995). The wide taxonomic diversity observed in this survey was striking. For example, at least three major eubacterial divisions, or five classes (Woese, 1987; Balows et al., 1992), are represented in the microbial isolates obtained from samples of Discodermia: the gamma-, beta-, alpha- proteobacteria, cytophaga, and Gram-positive eubacteria. Of the 24 isolate sequences obtained from Discodermia the most commonly observed bacterial subdivisions were gamma-proteobacteria (9) and alpha- proteobacteria (8), followed by Gram -positive (4), beta-proteobacteria (2) and possibly a single Cytophaga-like isolate (K279) (Table 4). This Gram-negative isolate matched most closely a psychrophilic marine Cytophaga, with some regions reaching 93% correspondence in identity, although these sequences were still not fully confirmed at the time of writing. Nonetheless, detection of Cytophaga primarily from the surfaces of marine aggregate particles in marine systems has been previously described (Delong, 1998). There were only a few cases where sequence database matches appeared unequivocal: e.g. K169 showed high similarity to Pseudomonas stutzeri, a common Gram-negative microbe in RNA group 1 (Balows et al., 1992), and E034 appeared to be related to a bacterium first characterised from Pele’s hydrothermal vents in the Pacific ocean. In other instances Discodermia isolates matched sequence database entries at highly significant identity levels (97-100% similarity), but these matches were made to ‘anonymous’ strains identified only to the genus or even subdivision level. For example, the closest relatives to isolates K171 and J131 were an ‘alpha-proteobacterium MBIC3368' and Altero- monas sp., respectively. Moreover, comparison of these two isolates maintained high sequence conservation across the whole rRNA segment, in contrast to 28D which had a range of consery- ation values (77-93%) largely dependent on the region of the gene under comparison (Table 4). Since several species of Vibrio exhibited equally 334 MEMOIRS OF THE QUEENSLAND MUSEUM TABLE 3. Profile of isolates from different sponge species collected at depths and habitats similar to Discodermia. Bold face type indicates samples that were obtained at similar locations or on the same expedition, and thus reliable for more direct comparisons between samples. Number of Collection Bacteria . . Sponge taxon P inde Depth (ft.) locátios. Gram Pos. Gram Neg. Actinomycetes Fungi Halichondria 21 390 W. Barbados 1 526 Canary Islands 1 477 Bahamas 2 4 393 Bahamas 4 1 1 2 1386 Bahamas 5 1 750 Bahamas 1 1 1 1 1803 Bahamas 5 1 1 803 Bahamas 2 1 1 440 Bahamas 1 1 1 576 Bahamas 4 543 Bahamas 3 1 3 473 Bahamas 1 479 Jamaica 4 9 1 520 Jamaica 3 2 1497 Jamaica 1 410 Jamaica 1 480 Bahamas 6 3 230 Bahamas 1 149 Florida, east coast 12 33 3 5 472 Puerto Rico 3 73 2 450 Puerto Rico 1 Dercitus 10 20 Venezuela 2 3 455 Bahamas 1 4 504 Bahamas 540 Bahamas 1 1 432 Bahamas 6 1 1 700 Bahamas 6 3 1 525 Bahamas 4 4 806 Jamaica 3 1 400 Jamaica 15 1 3 384 US Virgin Islands 3 Corticium 9 581 Bahamas 1 525 Bahamas 8 1 462 Bahamas 1 2 1 443 Bahamas 4 8 1 305 Bahamas 2 2819 Bahamas 1 3 1 377 Bahamas 8 1 1585 Turks & Caicos 1 90 Puerto Rico 5 11 2 3 Discodermia 10 592 Bahamas 1 110 Bahamas 7 $: 1 540 Bahamas 4 3 1 535 Bahamas 3 2 5 520 Bahamas 8 3 1 543 Bahamas 21 31 5 2 520 Bahamas 3 2 3 383 Honduras 9 51 1 4 440 Honduras 36 61 1 2 415 Honduras 14 25 12 HETEROTROPHIC MICROBES IN DISCODERMIA 335 high scores in comparison to the isolate K261 (ca. 98%), only the genus is listed in Table 4. Although we did not deliberately attempt to detect or amplify cyanobacteria, this group is well known amongst sponges living in the photic zone (Wilkinson, 1987; Ruetzler, 1990; Diaz, 1997), and some of our samples of Discodermia were collected in or near this zone. Clone 28C exhibited strong sequence similarity to a Leucothrix, which has been described as a large- diameter, morphologically distinct, marine gliding bacteria related to cyanobacteria (Balows et al., 1992). To facilitate direct comparisons between our sample isolates and known sequences, and to monitor sampling variation, we undertook parallel rRNA sequence analysis of potentially different microbial isolates from other sponge taxa living in geographical proximity to Discodermia (Table 2). These isolates (indicated in boldface in Table 3) were derived from sponge samples collected from the same habitats and depths as Disco- dermia (indicated in bold in Table 1). These few data, although preliminary, suggest a higher fre- quency of a Bacillus in non- Discodermia sponges, which is possible circumstantial evidence for sponge-specific microbial assoc- iations (Althoff et al., 1997) (see Table 3). More extensive comparisons could be designed to determine the optimal and natural conditions of Discodermia-associated microbes, perhaps by wider sampling of proximal sponge and non- sponge habitats (e.g. sediments, seawater, etc.). PCR was also used to directly amplify rRNA sequences from a) a total sponge cell preparation, and b) a dissociated mesohyl fraction enriched for specific populations of microorganisms. Although only a small number of rRNA clones were obtained by shotgun cloning from these two sources (labelled 28A—Z for the enriched fraction and 29A-Z for the total ‘crude’ mesohyl prep- aration), their identities and overall compositions appeared to be different from those derived from cultured isolates (Table 4). In spite ofthe fact that these sequences also exhibited the highest frequency of chimeric PCR artifacts (Liesack et al., 1991), precluding analysis of the total rRNA fragment, some of these clones may represent ‘uncultivable’ microbes. Up to 99% of naturally occurring microbes may be overlooked by standard culturing techniques (Button et al., 1993; Amann et al., 1995; Hulgenhotz & Pace, 1996). For example the 5' ends of two clones in the 29 series appeared to significantly match Flectobacillus, along with some conservation (82-91%) to Ridgea and Riftia hydrothermal vent bacteria (Feldman et al., 1997). Sequences of other Discodermia isolates exhibited significant similarity to bacteria associated with hydrothermal vent habitats and organisms. Also, clone 29B2 appeared to be distantly related to Clostridia. Furthermore, many microbes have not yet been analysed or may not have been isolated from the original plates. Interestingly, the sequences of most of the PCR-derived rRNA clones exhibited identity levels below 90%, and thus it may be recommended that unidentified strains with this level of conservation to any entry in either data- base be considered strong candidates for ‘novel’ species designation. PHYLOGENY OF NOVEL ISOLATES. Phylogenies constructed from the rRNA sequences were used to characterise the diversity and relatedness of Discodermia bacterial strains, Figures 2 and 3 display dendrograms constructed with two different phylogenetic algorithms: maxi- mum parsimony and neighbour-joining (NJ), respectively. Maximum likelihood (ML) an- alyses were also performed with smaller subsets of taxa. Relationships constructed under the principal of parsimony use a criterion of mini- mum evolution (shortest tree), while NJ uses a clustering algorithm based on overall similarity or distance in a comprehensive OTUxOTU matrix of corrected pairwise distances (Hillis et al., 1996). Although NJ and ML reconstructions better compensate for rate variation among different lineages (Hasegawa & Fujiwara, 1993), ML analyses involved fewer taxa due to com- putational limitations for large datasets, and thus are not discussed further in this work. Despite relatively large differences in rRNA sequences among some of the taxa, sequence alignments appeared to be robust. Multiple invariant positions and highly conserved regions corresponding to previously described secondary structures (e.g. loop 20, loop 14-15, and stem 5) were observed by eye along the nearly 800bp of rRNA sequences. Observation of compensatory mutations in stem 13 and bulge 17, among others, in the novel bacterial rRNAs corroborate the conservation of those structures (Fig. 1). Only one highly variable region corresponding to loop 11 (Neefs et al., 1993; Gutell et al., 1994) re- quired removal due to ambiguous alignment and its effect on nucleotide site homology. For the 30 microbial taxa analysed by max- imum parsimony (Fig. 2), an estimation of tu) UJ c MEMOIRS OF THE QUEENSLAND MUSEUM TABLE 4. Discodermia-associated microbes identified by 16S-like rRNA sequences. Key: 1, the list of organisms was derived directly from the output of GenBank or RDP queries (Altshchul et al., 1990; Maidak et al., 1994). The names of the closest relative may refer to species, genus or common names. 2, percent identity was derived from BLAST scores only, and may reflect identities of different segments of a single query rRNA sequence. Thus, variable conservation of different regions of the rRNA sequence is indicated in the ranges of identity values shown. 3, since these clones were shown to be chimeric, whole contigs were not analysed and query results reflect identities for 5' termini only. 4, five microbial sequences derived from non-Discodermia sponge microbes are underlined - M234 and M196 (Corticium), M162 and M099 (Halichondria) and M119 (Dercitus). 5, clonal sequences indicated with an asterisk (*) represent the 28/29 clonal series which was derived from PCR amplification of sponge mesohyl fractions. Therefore, these sequences were not derived from cultured isolates. Most closely related genus Microbial ID % Sequence Most closely related genus or | Microbial ID % Sequence or group no. identity" group no. identity” ALPHA BETA Alpha proteobacteria K200 90 Unidentified marine K255 92-96 a proteobacterium Alpha proteobacteria MBIC3368 K202 97 . : Alcaligenes 28D2* 85-94 Alpha proteobacteria K121 88 7 : Beta proteobacteria B849 92-99 Alpha proteobacteria K126 91 I R GRAM-positive Alpha proteobacteria K275 88-96 1 : Bacillus (low GC) M680 96 Alpha proteobacteria M485 94 : . 4 Bacillus (low GC) M529 97 Alpha proteobacteria M162 97 : " Bacillus firmus M099 91 Erythrobacter E035 98 a 3 i Bacillus fusiformes 28X* 81 Alpha proteobacteria MBIC3368 K171 99 7 4 ^ Bacilllus firmus M196 97 Phodospirillum 29A* 90 K 4 Bacillus firmus M234 91 GAMMA F : n Unknown actinomycete 28D* 91 Hydrothermal vent bacterium M119 97 7 : x Nocardia, actinomycete K146 97 Vibrio B853 96 . - Nocardia, actinomycete K145 80 Alteromonas J131 100 " 7 CYTOPHAGA Hydrothermal vent bacterium E034 99 . E — Marine psychrophile K279 94 Vibrio K261 98 AMBIGUOUS GROUPING Pseudoaltermonas K127 97 7 SEA - - Flectobacillus 29W* 85 Vibrio alginolyticus K141 97 4 " 7 : Flectobacillus 29B” 82-91 Microbulbifer C724 89 ar i : Clostridia 29B2* 76-81 Unidentified gamma C723 91 Pseudomonas stutzeri K169 99 Leucothrix mucor 280? * 80-94 skewness of tree length distributions (i.e. for 10,000 random trees using the Random Trees option in PAUP), yielded a g, statistic of -0.64. This value is above the 99% significance level for the corresponding critical value of g, for more than 25 taxa, indicating a strong leftward skew- ness and high phylogenetic signal in a four-state character dataset (Hillis & Huelsenbeck, 1992). Weighting transversions over transitions by a factor of 2 shortened the overall length of MP trees. There were only two and four more trees that were one or two steps longer, respectively, than the shortest tree shown in Figure 2 using the same dataset. The clade containing beta and gamma bacteria shows the weakest support (55%), and is thus depicted as a polytomy. Low support is likely due to the uncertain placement of clones 29B, 29W and 28C. Proximal Ridgeia and Marinobacter groups are typically grouped with gamma proteobacteria. In parallel, neighbour-joining analyses of rRNA sequence data yielded very similar conclusions to parsimony (Fig. 3). Pairwise genetic distances of all OTU’s based on Kimura’s 2N parameter correction (Hillis et al., 1996), ranged from 0.04->0.70. The major differences between the NJ and MP trees were: 1) higher bootstrap sup- port for individual clades with NJ relative to MP; 2) fewer collapsed nodes and polytomies with NJ, providing clearer groupings of major proteo- bacteria subdivisions; 3) inclusion of Rhodobium and clone 29A with the cluster of Alpha eubacteria; HETEROTROPHIC MICROBES IN D/SCODERMIA ]6 JE K279 00 Cytophaga ] Cp Rhodobium Gram 28D* + Actinomycetes Chloroflexus FIG. 2. Representative maximum parsimony phylogeny of Discodermia-associated microbes. Cultured isolates are shown in standard bold font. PCR-derived/uncultivated clones are shown with asterisks (*) and in italics, while non-Discodermia microbe M119 is underlined. Numbers below each node refer to the bootstrap value after 500 iterations. Representatives of genera or families are listed in the Methods and have names written out. Preliminary heuristic searches using the same dataset, 10 random stepwise additions of taxa with tree bisection reconnection (TBR), found only 2 most parsimon- ious trees. Length of the two most parsimonious heuristic trees was 1864 steps, with a consistency index (CI) of 0.495. Sub-optimal trees that were longer by 1 or 2 steps numbered only 2 and 4, res- pectively, and retained the same basic topology of the bootstrap consensus. Morever, the heuristic trees did not collapse the beta and gamma proteobacteria clades into a single polytomy, but rather showed the beta proteobacteria as a distinct group relative to the gamma bacteria (McDonald et al., 1997). Preliminary group- ings in the proteobacteria subdivisions and Cytophaga (CP) were based on the identities obtained from BLAST sequence database searches. and 4) monophyly of all representative Gram- positive bacteria. Phylogenetic assignment of clone 29A, which appeared most closely related to the alpha subgroup, was problematic with all three algorithms (ML tree not shown). The decay index for any group that included 29A was always low (<3). Since it is not within the scope of this study to evaluate the strengths and weak- nesses of various phylogenetic methods, or to make definitive conclusions on the taxonomic to uy) - status of each novel isolate, some of these taxo- nomic placements are likely to be revised in the future. Nevertheless, the topologies of parsimony and distance trees were generally consistent in showing at least 5 major clades. Isolates K261 and K171 were at the tips ofthe MP tree and thus their omission to accelerate computation times did not have a significant effect on NJ tree topology. Several features in the present recon- structions of Discodermia microbes, such as the monophyly of all proteobacteria, strong bootstrap support for the gamma and beta subdivisions, and the outgroup status of the Gram-positive clade, are all in agreement with current bacterial taxonomy. More specifically, a long branch characterised the lineage of K279 and its strong association with Cytophaga bacteria. This branch is con- sidered by us to be a fifth major bacterial clade of the sponge. Relatively long branch lengths were prominent for several other lineages (e.g. K146 and some clones in the 28/29 series). Although uncultivated clones 28C, 29B and 29W were grouped significantly with other marine bacteria discovered from previous en- vironmental surveys (Moyer et al., 1995; Fuhrman & Davis, 1997), the database matching of 29W and 29B to Flectobacillus (a Cyto- phagales) indicates that accurate placement of these bacteria require more refined determin- ation. However, the tight clustering of taxa observed also suggests possible endemism or ecological specificity with respect to Discodermia. Consistent with the database matches to Clos- tridia, clone 29B2 was placed repeatedly near outgroup taxa at the base of all trees. The weak and unresolved positions of some taxa, such as 29B2, 29A and K146, connote another level of diversity. The inclusion of non-Discodermia isolates (M119, M169, M234, M162 or M099) in some reconstructions did not significantly alter tree topologies, nor did it suggest any evidence of taxon-specific symbioses occurring in the current dataset. Overall, these results parallel earlier des- criptions (Santavy et al., 1990) describing major bacterial groups such as Vibrio, Aeromonas, and coryneform/actinomycete (Gram-positive) strains derived from marine sponges. DISCUSSION Phenotypic, comparative DNA sequence and molecular phylogenetic analyses confirmed the presence of at least five distinct eubacterial clades of 16S-like SSU rRNA sequences from microbial Bacillus M529 K146 Actinomycetes Mi19 | E034 jis Alteromonas B853 y K169 29W* 28C* Gram 29B* Marinobacter Ridgeia B349 Burkholderia K275 Rhodobium Cytophaga Chloroflexus FIG. 3. Distance-based phylogeny reconstructed with the Neighbour-joining method. The same taxa (except K171 and K261) and annotations as that shown in Fig. 2 were analysed. isolates ofthe lithistid sponge genus Discodermia. Several of the identified Discodermia bacterial groups, such as the gamma proteobacteria, are consistent with previous characterisations of deep-sea microbes (Moyer et al., 1995; Feldman etal., 1997; Fuhrman & Davis, 1997), while other lineages (e.g. K279, 28C, 29W, 29A) appear unallied or novel. This situation may have arisen via accelerated substitution rates, while group- or strain-specific synapomorphies were maintained (Hillis et al., 1996; Peek et al., 1998). More likely, however, no close relatives exist in current prokaryotic rRNA sequence databases, which underscores possible missing links in current bacterial rRNA taxonomies. Similar to many earlier surveys of bacterial diversity from the environment (Pace et al., 1986; Giovannoni et al., 1990; Amann et al., 1995; Rheims et al., 1996), PCR and molecular methods have probably revealed unique microorganisms which are otherwise uncultivable under traditional methods. Although some variation may be attributed to differences among species of Discodermia and between individual samples of particular species, this characterisation also likely underestimates total diversity in the genus, since only about 10% of the Discodermia isolate collection has been MEMOIRS OF THE QUEENSLAND MUSEUM sequenced at the time of writing. Nevertheless, our finding that many different bacterial strains and rRNA ‘types’ stem from only one sponge genus is novel and distinguishes the present study from earlier results. Bacterial symbionts occur both intracellularly and extracellularly with respect to their sponge host mesohyl (Vacelet & Donaday, 1977; Simp- son, 1984; Wilkinson, 1987). However, without positive identification of the types of interacting organisms, elucidating symbiotic parameters such as nutrient transfer, detoxification or gene exchange will not be as meaningful as those made for well-established cnidarian-dinoflagellate associations (Trench, 1993). It is possible that some of the microbes identified here stem for- tuitously from the microbial pool derived from sponge filter-feeding activities (Reiswig, 1971; Pile et al., 1996). A long-standing question in sponge-microbial symbioses has been how do so many different symbionts coexist and seemingly thrive in the relatively inhospitable (phagoctye- filled) environment ofthe marine sponge mesohyl (Simpson, 1984; Wilkinson, 1987)? One answer may stem from the advantages of 'ecto- symbioses' and bacterial communities (Bull & Slater, 1982). Conversely, it is not unreasonable to suppose thata fraction ofthe bacterial species isolated and characterised here represent bona fide obligate symbionts of Discodermia, an expectation which has been confirmed in other sponges (Burlando et al., 1988; Althoff et al., 1998). Although not trivial, the question of determining specific microbial symbionts could be approached by probing for consistent rRNA (or other genetic) signatures among a matrix of geographically sep- arated samples of Discodermia present in our collection. Definitive conclusions on the relative abundance of a particular bacterial strain in Discodermia are precluded, since the quanti- tative recovery of several microbial types during sequencing may suggest any of the following: 1) a dominant presence in the host sponge and concomitant functional role in Discodermia physiology; 2) habitat-specific differences; or 3) experimental bias of PCR primer binding sites, genomic DNA quality, etc. (Rheims et al., 1996). It would be interesting in the future to determine whether the mode of molecular evolution in these microbes matches other observations of faster nucleotide substitution rates in symbiotic versus free-living marine bacteria, which may be a function of small population sizes of some symbiotic communities (Peek et al., 1998). HETEROTROPHIC MICROBES IN DISCODERMIA The relatively large breadth and depth of phylo- genetic diversity found among Discodermia- associated microbes, as revealed by SSU rRNA sequences, has significance in several areas. Firstly, the data reiterate previous studies show- ing that some marine sponges either maintain or tolerate high levels of microbial species richness. Consequently, this study supports claims that current numbers of catalogued bacterial species are underestimated (UNEP, 1995; Hawksworth & Colwell, 1992; Colwell, 1997). Moreover, the low frequency of duplicate rRNA sequences ob- served in this survey supports the large diversity of microbes in some sponge taxa. Lastly, these results may have ramifications for microbial and deep water ecology. Since viable deep sea habitats are generally ‘patchy’ (Grassle, 1991; Cavanaugh, 1994; Snelgrove & Grassle, 1995), sponges such as Discodermia may represent essential stepping stones for bacterial dispersal across large expanses of the seafloor bottom. Such functions are often attributed to less common and more random sinking detritus, animal carcasses (e.g. whale falls), or hydro- thermal vent habitats (Showstack, 1998; Tunnicliffe & Fowler, 1998). At the level of the organism sponges may embody oases of species richness, rather than oases of biomass, which is the perception often associated with hydrothermal vents (Snelgrove & Grassle, 1995). The detection of hydrothermal vent-like microorganisms in Discodermia, regardless of whether or not they are actual sponge symbionts, suggests a possible source of colonisers for deep water habitats. Thus, Discodermia and similar marine sponges should be re-evaluated in the context of a potent- ially pivotal role in marine microbial ecology, dispersal, and evolution. ACKNOWLEDGEMENTS We thank Julie Olson and Susan Sennett for thoughtful review of the manuscript. We acknowledge the National Cancer Institute for allocation of computing time, Gary Smythers and staff support at the Frederick Biomedical Supercomputing Center of the Frederick Cancer Research and Development Center. This manuscript is Harbor Branch Oceanographic Institution contribution HBOI #1276. LITERATURE CITED ALTSCHUL, S.F., GISH, W., MILLER, W., MYERS, E.W. & LIPMAN, D.J. 1990. Basic local alignment and search tool. Journal of Molecular Biology 215: 403-410. AMANN, R.I., LUDWIG, W. & SCHLEIFER, K.H. 1995, Phylogenetic identification and in sifu detection of individual microbial cells without cultivation. Microbiology Reviews 59: 143-169. AUSTIN, B. 1988. Marine Microbiology. (Cambridge University Press: Cambridge). ALTHOFF, K., SCHUTT, C., STEFFEN, R., BATEL, B. & MULLER, W.E.G. 1998. Evidence for a symbiosis between bacteria of the genus Rhodobacter and the marine sponge Hali- chondria panicea: harbor also for putatively toxic bacteria? Marine Biology 130: 529-536. BALOWS, A., TRUPER, H.G., DWORKIN, M., HARDER, W. & SCHLEIFER, K-H. (eds) 1992, The Prokaryotes. 2nd Edition (Springer-Verlag: New York). BEWLEY, C.A., HOLLAND, N.D. & FAULKNER, D.J. 1996. Two classes of metabolites from Theonella swinhoei are localised in distinct populations of bacterial symbionts. Experientia 52: 716-722. BULL, A.T. & SLATER, J.H. (eds) 1982. Microbial Interactions and Communities. (Academic Press: New York). BURLANDO, B., SABATINI, M.A. & GAINO, E. 1988. Association between calcareous Clathrina cerebum (Haeckel) and bacteria: electron microscope study. Journal of Experimental Marine Biology and Ecology 116: 35-42. BUTTON, D.K., SCHUT, F., QUANG, P., MARTIN, R. & ROBERTSON, B.R. 1993. Viability and isolation of marine bacteria by dilution culture: Theory, procedures and initial results. Applied Environmental Microbiology 59: 881-891. CARY, S.C., WARREN, W., ANDERSON, E. & GIOVANNONI, S.J. 1993, Identification and localization of bacterial endosymbionts in hydrothermal vent taxa with symbiont-specific polymerase chain reaction amplification and in situ hybridization techniques. Molecular Marine Biology and Biotechnology 2: 51-62. CAVANAUGH, C.M. 1994. Microbial symbiosis: Patterns of diversity in the marine environment. American Zoologist 34:79-89. CLAYTON, R.A., SUTTON, G., HINKLE, P.S., BULT, C. & FIELDS, C. 1995, Intraspecific variation in small-subunit rRNA sequences in GenBank: Why single sequences may not adequately represent prokaryotic taxa. International Journal of Systematic Bacteriology 45: 595-599. COLWELL, R.R. 1997. Microbial biodiversity and biotechnology. Pp. 279-287. In Reaka-Kudla, M.L., Wilson, D.E. & Wilson, E.O. (eds) Bio- diversity II. (Joseph Henry: Washington DC). DELONG, E.F. 1998. Molecular phylogenetics: new perspective on the ecology, evolution and biodiversity of marine organisms. Pp. 1-28. In Cooksey, K.E. (ed.) Molecular approaches to the study of the ocean. (Chapman-Hall: London). DIAZ, M.C. 1997, Molecular detection and character- ization of specific bacterial groups associated with tropical sponges. Proceedings of the 8th Inter- national Coral Reef Symposium 2: 1399-1402. DISTEL, D.L., LANE, D.J., OLSEN, G.J., GIOVAN- NONI, S.J., PACE, B., PACE, N.R., STAHL, D.A. & FELBACK, H. 1988. Sulfur-oxidizing bacterial endosymbionts: analysis of phylogeny and specificity by 16S rRNA sequences. Journal of Bacteriology 170: 2506-2510. EFRON, B. 1982. The jackknife, the bootstrap, and other resampling plans. SIAM Monograph No. 38. (Society of Industrial and Applied Mathemat- ics: Philadelphia). EGUCHI, M. & ISHIDA, Y. 1990. Oligotrophic heter- otrophic bacteria and in situ heterotrophic activity in pelagic seawaters. FEMS Microbial Ecology 73: 23-30. FELDMAN, R.A., BLACK, M.B., CARY, C.S., LUTZ, R.A. & VRIJENHOEK, R.C. 1997. Molecular phylogenetics of bacterial endosymbionts and their hosts. Molecular Marine Biology and Biotechnology 6: 268-277. FELSENSTEIN, J. 1985. Confidence limits on phylo- genies: an approach using the bootstrap. Evolution 39: 783-791. FELSENSTEIN, J. 1993. PHYLIP (Phylogenetic inference package). Version 3.5c. (Distributed by author, Department of Genetics, University of Washington: Seattle). FOX, G.E., WISOTZKEY, J.D. & JURTSHUK, P. 1992. How close is close: 16S rRNA sequence identity may not be sufficient to guarantee species identity. International Journal of Systematic Bacteriology 42: 166-170. FUHRMAN, J.A. & DAVIS, A.A. 1997. Widespread Archaea and novel bacteria from the deep sea as shown by 16S rRNA gene sequences. Marine Ecology Progress Series 150: 275-285. GENETICS COMPUTER GROUP (GCG). 1994. Program manual for Wisconsin Package. Version 8. (UWGCG: Madison WI). GIOVANNONI, S.J., BRITSCHGI, T.B., MOYER, C.L. & FIELD, K.G. 1990. Genetic diversity in Sargasso Sea bacterioplankton. Nature 345: 60-63. GRASSLE, J.F. 1991. Deep sea benthic biodiversity. BioScience 41: 464-469. GUNASEKERA, S.P., POMPONI, S.A. & McCARTHY, P.J. 1994, Discobahamins A and B, new peptides from the Bahamian deep water marine sponge Discodermia sp. Journal of Natural Products 57: 79-83. GUTELL, R.R., LARSEN, N. & WOESE, C.R. 1994. Lessons from an evolving rRNA: 16S and 23S rRNA structures from a comparative perspective. Microbioloical Reviews 58: 10-26. HASEGAWA, M. & FUJIWARA, M. 1993, Relative efficiencies of the maximum likelihood, max- imum parsimony, and the neighbor-joining methods for estimating protein phylogeny. Molecular Phylogenetics 2: 1-5. MEMOIRS OF THE QUEENSLAND MUSEUM HAWKSWORTH, D.L. & COLWELL, R.R. 1992. Biodiversity amongst microorganisms and its relevance. Biodiversity and Conservation 1: 221-345. HILLIS, D.M. & HUELSENBECK, J.P. 1992. Signal, noise and reliability in molecular phylogenetic analyses. Journal of Heredity 83: 189-195. HILLIS, D.M., MORITZ, C. & MABLE, B.K. 1996. Molecular Systematics. (Sinauer: Sunderland). HUGENHOLTZ, P. & PACE, N.R. 1996. Identifying microbial diversity in the natural environment: a molecular phylogeny approach. Trends in Biotechnology 14: 190-197. KELLY-BORGES, M., ROBINSON, E.V., GUNASEKERA, S.P., GUNASEKERA, M., GULAVITA, N.K. & POMPONI, S.A. 1994, Species differentiation in the marine sponge genus Discodermia (Demospongiae, Lithistida): the utility of ethanol extract profiles as species- specific chemotaxonomic markers. Biochemical Systematics and Ecology 22: 353-365. JENSEN, P.R. & FENICAL, W. 1994. Strategies for the discovery of secondary metabolites from marine bacteria: Ecological perspectives. Annual Review of Microbiology 48: 559-584. LANE, DJ. 1991. 168/238 rRNA sequencing. Pp. 115-148. In Stackebrandt, E. & Goodfellow, M. (eds) Nucleic acid techniques in bacterial systematics. (Wiley & Sons: New York). LIESACK, W., WEYLAND, H. & STACKEBRANDT, E. 1991, Potential risks of gene amplification by PCR as determined by 16S rDNA analysis of a mixed-culture of strict barophilic bacteria. Microbial Ecology 21: 191-198. LONGLEY, R.E., CADDIGAN, D., HARMODY, D., GUNASEKERA, M. & GUNASEKERA, S.P. 1991. Discodermolide-a new, marine-derived immunosuppressive compound. Il. In vivo studies. Transplantation 52: 656-61. MAIDAK, B.L., LARSEN, N., McCAUGHEY, M.J., OVERBEEK, R., OLSEN, G.J., FOGEL, K., BLANDY, J. & WOESE, C.R. 1994, The Ribsomal database project. Nucleic Acids Research 22: 3485-3487. McDONALD, LR, KELLY, D.P., MURRELL, J.C. & WOOD, A.P. 1997, Taxonomic relationships of Thiobacillus halophilus, T. aquaesulis and other species of Thiobacillus as determined using 16S rDNA sequencing. Archives of Microbiology 166: 394-398. McINERNEY, J.O, WILKINSON, M., PATCHING, J.W., EMBLEY, T.M. € POWELL, R. 1995. Recovery and phylogenetic analysis of novel archaeal rRNA sequences from a deep-sea deposit feeder. Applied and Environmental Microbiology 61: 1646-1648. MOYER, C.L., DOBBS, F.C. & KARL, D.M. 1995. Phylogenetic diversity of the bacterial community from a microbial mat at an active, hydrothermal vent system, Loihi Seamount, Hawaii. Applied Environmental Microbiology 61: 1555-1562. HETEROTROPHIC MICROBES IN DISCODERMIA 341 NEEFS, J-M., VAN DE PEER, Y., DE RIJK, P., CHAPELLE, S. & DE WACHTER, R. 1993. Compilation of small subunit RNA structures. Nucleic Acids Research 21: 3025-3049. PACE, N.R., STAHL, D.A., LANE, D.J. & OLSEN, G.J. 1986. The analysis of natural microbial populations by ribosomal RNA sequences. Advances in Microbial Ecology 9: 1-55. PEEK, A.S., VRIJENHOEK, R.C. & GAUT, B.S. 1998. Accelerated evolutionary rate in sulfur- oxidizing endosymbiotic bacteria associated with the mode of symbiont transmission. Molecular Biology and Evolution 15: 1514-1523. PILE, A.J., PATTERSON, M.R. & WITMAN, J.D. 1996. In situ grazing on plankton <10um by the boreal sponge Mycale lingua. Marine Ecology Progress Series 141: 95-102. PITCHER, D.G., SAUNDERS, N.A. & OWEN, RJ. 1989. Rapid extraction of bacterial genomic DNA with guanidium thiocyanate. Letters in Applied Microbiology 8: 151-156. POMPONI, S.A. & WILLOUGHBY, R. 1994. Sponge cell culture for production of bioactive met- abolites. Pp. 395-400. In Soest, R.W.M. Van, Kempen, T.M.G. van & Braekman, J.C. (eds) Sponges in Time and Space. (Balkema: Rotterdam). REISWIG, H.M. 1971. Particle feeding in natural populations of three marine demosponges. Biological Bulletin 141: 568-591. RHEIMS, H., RAINEY, F.A. & STACKEBRANDT, E. 1996. A molecular approach to the search for diversity among bacteria in the environment. Journal of Industrial Microbiology 17: 159-169. RUETZLER, K. 1990. Associations between Carib- bean sponges and photosynthetic organisms. Pp. 455-466. In Ruetzler, K. (ed.) New Perspectives in Sponge Biology. (Smithsonian Press: Washington DC). SAMBROOK, J., FRITSCH, E.F. & MANIATIS, T. 1989. Molecular Cloning: A Laboratory Manual. (Cold Spring Harbor: New York). SANTAVY, D.L., WILLENZ, P. & COLWELL, R.R. 1990. Phenotypic study of bacteria associated with the Caribbean Sclerosponge, Ceratoporella nicholsoni. Applied Environmental Microbiology 56: 1750-1762. SIMPSON, T.L. 1984. The cell biology of sponges. (Springer-Verlag: New York). SHOWSTACK, R. 1998. Whale falls may provide important stepping stone habitat between deep sea vents and seeps. Eos 79: 45. SNELGROVE, P.C.R. & GRASSLE, J.F. 1995. The deep sea: desert and rainforest. Oceanus 38: 25-29. SWOFFORD, D. 1993. PAUP (Phylogenetic Analysis Using Parsimony). Version 3.1.1. for Apple Macintosh (Smithsonian Institution: Washington DC). TRENCH, R.K. 1993. Macroalgal-invertebrate sym- bioses: A review. Endocytobiology and Cell Research 9: 135-175. TUNNICLIFFE, V. & FOWLER, M.R. 1998. Influence of sea-floor spreading on the global hydrothermal vent fauna. Nature 379: 531-533. UNITED NATIONS ENVIRONMENT PROGRAMME (UNEP) 1995. Global Biodiversity Assessment. Heywood, V.H. (ed.). Pp. 1-123. (Cambridge University Press: Cambridge). UNSON, M.D., HOLLAND, N.D. & FAULKNER, D.J. 1994. A brominated secondary metabolite synthesized by the cyanobacterial symbiont of a marine sponge and accumulation of the crystalline metabolites in the sponge tissue. Marine Biology 119: 1-11. VACELET, J. & DONADAY, C. 1977. Electron microscopic study of the association between some sponges and bacteria. Journal of Experimental Marine Biology and Ecology 30: 301-314. WILKINSON, C.R. 1984. Marine sponges discriminate between food bacteria and bacterial symbionts: electron microscope radioautography and in situ evidence. Proceedings of the Royal Society of London (B) 220: 519-528. 1987. Significance of microbial symbionts in sponge evolution and ecology. Symbiosis 4: 135-146. WOESE, C.R. 1987. Bacterial evolution. Micro- biological Reviews 51: 221-271. MEMOIRS OF THE QUEENSLAND MUSEUM PROPAGATED ELECTRICAL IMPULSES IN A SPONGE. Memoirs of the Queensland Museum 44: 342, 1999:- Previous work has shown that Rhabdocalyptus dawsoni, a hexactinellid sponge, can arrest its feeding current following mechanical or electrical stimuli. Although a propagated impulse was suspected as the signal triggering arrests, numerous attempts to record such an event failed due to the porous character of the tissue and extreme fragility of the surface membranes. Using a new approach, which involves dissociating sponge tissue, letting it reaggregate, and grafting it back on to the sponge as an autograft, we have found it possible to record propagated electrical impulses. The grafts fuse with the trabecular reticulum, a syncytial tissue that penetrates all parts of the body, including the flagellated chambers, and are eventually absorbed into the sponge. But for a while they form solid lumps that can be used for attachment of suction recording electrodes. Impulses are all-or-none events evoked by single electrical shocks that propagate diffusely through the entire preparation at 0.27 + 0.lem:s' at 10°C, presumably in the trabecular reticulum. The preparation shows an absolute refractory period of 29 s, and is relatively refractory for a further 95-100 s. Intracellular recordings have not been carried out but the wave form recorded extracellularly is suggestive of a conventional, overshooting spike. Pharmacological evidence suggests that it is calcium-based. Thus, despite its long refractory period and low conduction velocity the system is functionally equivalent to the through-conducting nerve nets and excitable epithelial conduction systems of other animals. CJ Porifera, hexactinellid, Rhabdocalyptus, conduction, electrophysiology, behaviour, pumping. Sally P. Leys* (email: leys@bit.net.au) & George O. Mackie, Department of Biology, University of Victoria, British Columbia, V8W 3N5, Canada; * Present address: Department of Zoology and Entomology, Molecular Zoology Laboratory, University of Queensland, St Lucia, Old. 4072, Australia; 1 June 1998. THEONELLAPEPTOLIDES FROM THE DEEP-WATER NEW ZEALAND SPONGE LAMELLOMORPHA STRONGYLATA. Memoirs of the Queensland Museum 44: 342. 1999:- The deep- water marine sponge, Lamellomorpha strongylata, was collected by benthic dredging at 80m on the Chatham Rise (200km offthe E Coast ofthe South Island of New Zealand). Besides the previously reported calyculins, calyculinamides and swinholide H, five new tridecapeptides, theonellapeptolides IIIa, b, c, d and e, were obtained (Fig. 1). The following strategy was used for determining the structures of the theonellapeptolides: 1) the amino acids were established by GC/MS following acid hydrolysis and derivatization; 2) methanolysis gave a linear peptide, which was sequenced by tandem mass spectrometry; 3) isobaric residues were distinguished by 2D NMR experiments; 4) detailed analysis led to the complete NMR assignment; 5) the absolute stereochemistry of Ille was determined by X-ray crystallography coupled with chiral HPLC; 6) the stereochemistry of the other peptides were established by an LC/MS method. Theonellapeptolides IIIb, c, d and e showed mild cytotoxicity against P388 cell line, but IIIa was very much less cytotoxic. This implied that the second residue from N-terminus (X) plays a key role in maintaining bioactivity. A comparison with the known theonellapeptolide Id suggested that the crystal structure of Ie is similar to that of Id although four residues are changed and the ring size is 36 in lle, not 37 as in Id. The theonellapeptolides from the I and II series have all been isolated previously from Lithistid sponges, while those from the III group are nominally from a different sponge order. A key question still to be adressed is whether or not all three groups of peptolides have a similar, or comparable, symbiont origin ? O Porifera, peptolides, amino acids, nmr spectroscopy, lc/ms, gc/ms, chiral hple, X-ray crystallography, symbionts, Lithistida, Theonella sp., Lamellomorpha strongylata, New Zealand. S. Li, J. W. Blunt, E. J. Dumdei & M. H. G. Munro (email: m.munro@chem.canterbury.ac.nz), Department of Chemistry, University of Canterbury, Christchurch, New Zealand; L. K. Pannell & N. Shigematsu, Laboratory of Analytical Chemistry, NIDDK, NIH, Bethesda, MD 20892-0805, USA; 1 June 1998. | Val— CO —CH20 CH; Theonellapeptolides Illa to e: Thr N-MefAla Y z BAla (a) X=N-MeHyMet, Y=Val, Z=N-Melle | (b) X=N-MeAla, Y=Val, Z=N-Melle (c) X=Leu, Y=N-MeAla, Z=N-Melle N-MeLeu (d) X=N-MeLeu, Y=Val, Z=N-MeVal | (e) X=N-MeLeu, Y=Val, Z-N-Melle Leu Ala— N-MeAla N -Melle lle FIG. 1. Structures of theonellapeptolides from Lamellomorpha strongylata. PHYLOGENETIC RESOLUTION POTENTIAL OF 18S AND 288 rRNA GENES WITHIN THE LITHISTID ASTROPHORIDA JAMES O. McINERNEY, CHRISTI L. ADAMS AND MICHELLE KELLY Mclnerney, J.O., Adams, C.L. & Kelly, M. 1999 06 30: Phylogenetic resolution potential of 18s and 28s rRNA genes within the lithistid Astrophorida. Memoirs of the Queensland Mu- seum 44: 343-351. Brisbane. ISSN 0079-8835. Kelly-Borges et al. (1991) and Kelly-Borges & Pomponi (1994) utilised partial 18S rRNA gene sequences to resolve relationships within hadromerid and lithistid sponges (Porifera: Demospongiae). While their results clarified several specific systematic problems, their conclusions were hampered by low levels of sequence variation. This study sought primarily to evaluate the resolution potential between regions of the 18S rRNA gene used in previous studies on sponges, and 28S rRNA genes used in more recent work. Six lithistid sponge taxa were chosen to represent a gradient of taxonomic relationships, ranging through genus, family, order and class. Approximately 1,300bp of the 18S rRNA gene and a 700bp region at the 5’ end of the 28S rRNA gene were compared with the data of Kelly-Borges & Pomponi (1994). We found that the 700bp region of the 28S rRNA gene presented the greatest potential for resolution of this group of Porifera at the genus and family level, and that the resultant molecular phylogeny was congruent with morphological hypotheses for the group. O Porifera, molecular phylogeny, evolution, Lithistida, Theonella, Discodermia, Corallistes. McInerney, J.O., *Adams, C.L., Borges, K.D. & **Kelly, M. (email: m.kelly@niwa.cri.nz), Department of Zoology, The Natural History Museum, Cromwell Road, London SW7 SBD, UK; Present address: *Queensland Museum, PO Box 3300, South Brisbane, 4101, Australia; **National Institute of Water & Atmospheric Research Ltd., Private bag 109-695, Newmarket, Auckland, New Zealand; 25 May 1999. For the past twenty years, organismal phylo- genies have been inferred from the primary sequence ofa portion of their genomes, The small subunit ribosomal RNA (SSU rRNA) or 18SrRNA gene has dominated the field as molecule of choice. The analysis of this molecule has been singularly instrumental in elucidating the phylogenetic relationships and natural history of almost all known prokaryotic species where attempts using other methods have failed (see Woese, 1987). Following the success of microbiologists adopting this approach, the systematics of eukaryotic taxa has been addressed by sequencing the 185 rRNA gene (Sogin et al., 1986). This gene is probably still the most frequently used for this purpose. It is often desirable to use more than a single gene region for the reconstruction of a phylogeny, to supply additional, potentially corroborative phylogenetic hypotheses. For example, the 5* region of the 28S rRNA gene has also been used with great effectiveness (Baroin et al. 1988; Chombard et al., 1998), as have Elongation Factor genes (Iwabe et al., 1989; Rivera & Lake, 1992), ATPases (Iwabe et al., 1989) and DNA- dependent RNA polymerases (Puhler et al. 1989), among others. There are a number of additional prerequisites for choosing a gene for the purposes of reconstructing a phylogeny. It is desirable that a constancy of function (functional orthology) is maintained throughout the evolution of the taxa of interest. Problems associated with long branches may be observed on trees where some genes have experienced a relaxation in selective pressures. The possibility of mistakenly isolating a paralogous homologue must be minimal. A suitable gene must show signs of having enough variability to discriminate between taxa at the desired taxonomic level. It must also be conserved enough to permit robust alignments and comparisons across the deepest divisions. Sponge phylogenies, in particular, have given rise to a number of contentious arguments, most of which result from a lack of suitably variable morphological characters which distinguish sponges at the species level and higher (Van Soest, 1987; Hadju et al., 1993; Kelly-Borges & Bergquist, 1997; Sandford & Kelly-Borges, 1997). The primary diagnostic morhological characters that differentiate sponge genera and species are spicule morphology and their arrangement within the sponge body. Characters often influenced by environmental factors, such as texture, surface features and colouration are less reliable as they are frequently plastic. Although the construction of poriferan phylogenetic hypotheses using molecular sequence data is still at a preliminary stage, with very few studies completed, it is likely that many future sponge systematic projects will incorp- orate a molecular moiety. For this reason, it is necessary to establish the taxonomic levels at which certain gene regions will be appropriate. For example, some stretches of DNA might be informative about phylogenetic relationships at the genus and species level, but they may be unsuitable for studies of ordinal relationships and so on. All studies so far have used different genes and gene regions for phylogenetic purposes (Kelly-Borges et al., 1991, 1994; West & Powers, 1993), preventing any useful links between these data towards the construction of larger phylo- genies. The advantage that may be gained in future by using the same gene region in all studies is therefore obvious. The primary goals of our study were to evaluate the resolution potential between various regions of sponge 18S and 28S rRNA genes, in order to determine which genes would efficiently resolve phylogenies at several taxonomic levels. Because of past difficulties in resolution using 18S rRNA (Kelly-Borges et al, 1991; Kelly-Borges & Pomponi, 1994), we took a positive approach in our more recent research to determine which gene would successfully provide resolution within a group of lithistid sponges, and thus, potentially within other taxonomic groups. In this study we evaluate the relative utility of four alignments with different gene origin, sequence length, and method of analysis. To do this, we extended the 18S rRNA gene data of Kelly-Borges & Pomponi (1994) for six species up to approximately 1,300bp, and in addition, using the same taxa, we have sequenced approx- imately 700bp of the 5’ end of the 28S rRNA gene. Taxa were selected to encompass a range of taxonomic levels including genus, family, order, and class. The criteria by which the sequence data sets were evaluated for their potential utility included counting the number of variable sites and the number of parsimony-informative sites, using maximum likelihood in order to estimate the proportion of constant sites that might be invariable, and also conducting an objective analysis of the resulting tree topologies. The resulting topologies were compared with a hypothetical MEMOIRS OF THE QUEENSLAND MUSEUM reconstruction based upon morphological characters. For the purpose of this exercise four lithistid sponges (Class Demospongiae) were selected from a much larger study on sponge phylogeny, and two hexactinellid sponges (Class Hexact- inellida) were chosen as an outgroup. Lithistid sponges represent relict forms of an ancestral fauna from which, it is thought, most demo- sponges have evolved. These sponges are characterised by the possession of a rigid siliceous skeleton made up of irregular branching desma spicules, the ends of which interlock (zygose) with neighbouring spicules (see Kelly-Borges & Pomponi, 1994, Figs 1,2). In some cases there are additional spicules present, providing clues as to their affinities with non desma-bearing sponges, and the polyphyly of at least some of these genera has been recently confirmed by Kelly-Borges & Pomponi (1994). Although very difficult to differentiate morphologically, the most reliable diagnostic characters that can be used are the morphology, ornamentation, and pattern of zygosis of the desma spicules, and the morpho- logy of the additional spicules ifthey are present. MATERIALS AND METHODS TAXA SELECTION. Sponge species were selected from a broader study of sponge phylo- geny, specifically for the purpose of examining capability of taxonomic resolution of poriferan sequence data (Table 1). All sponges were collected using Harbor Branch Oceanographic Institution’s ‘Johnson-Sea-Link II’ manned submersible, except Theonella spp. which were collected using SCUBA. Samples were identified through histological examination of skeletal structures, the procedures for which are detailed in Kelly-Borges et al. (1994). Voucher specimens have been deposited in the collections of either The Natural History Museum, London (BMNH), or the Harbor Branch Oceanographic Museum, Florida (HBOM ). Registration numbers are given in Table 1. Hexactinellid sponges Margaritella coeloptychioides and Sympagella nux (Table 1) were selected to provide outgroup sequences for the molecular phylogeny reconstruction. MORPHOLOGICAL PHYLOGENY RECON- STRUCTION. A phylogenetic analysis of morphological characters (Table 2) was carried out to examine relationships within and between the theonellid and corallistid taxa for the purpose of comparison with trees gained from recons- truction of molecular data. The analysis used the RESOLUTION POTENTIAL OF 18S AND 288 rRNA SPONGE GENES TABLE 1. Collection data and taxonomic position for sponge taxa sequenced in this study. Museum Taxon Registration Locality Class Hexactinellida Subclass Hexasterophora Order Hexactinosida Family Euretidae Margaritella coeloptychioides Schmidt, 1870 Order Lyssacinosida HBOM 003:00925 Turks and Caicos Family Caulophacidae Sympagella nux Schmidt, 1870 Class Demospongiae HBOM 003:00929 Turks and Caicos Subclass Tetractinomorpha Order Astrophorida Family Theonellidae BMNH1998.3.4.1 BMNH1998.3.4.2 BMNH1998.3.4.3 Theonella sp. 1 Belau, Micronesia Belau, Micronesia Bahamas, Caribbean Theonella sp. 2 Discodermia sp. Family Corallistidae BMNH1998.3.4.4 Corallistes typus Schmidt, 1870 TABLE 2. Morphological characters (A) and their character-states (B) for taxa in this study (see Kelly-Borges & Pomponi, 1994, for an explanation of characters). Characters indicated as * are absent from the outgroup in that they do not possess desmas. 345 Branch and Bound search option of PAUP 3.1.1, and data were unordered and unweighted. In order to obtain a directed analysis, members of two non-lithistid astrophorid families, Geodia (Family Geodiidae) and Stelletta (Family Ancor- inidae) were chosen as outgroups. The major skeletal characters that separate these lithistids from their outgroups are the possession of desmas, unique ornamented dichotriaenes, and certain types of microscleres (see Kelly-Borges & Pomponi, 1994). MOLECULAR EXPERIMENTAL PROCEDURES. Sample collection, preservation and DNA extraction have been previously described (Kelly-Borges & Pomponi, 1994). PCR primers and sequencing oligonucleotides are listed in Table 3 for both genes. PCR reactions were carried out in a 5011 reaction volume which contained a one-tenth volume of 10x PCR buffer (500mM KCl, 100mM Tris-HCl, 1% Triton X-100), dNTPs to a final concentration of 200uM, primers at a concentration of 200uM and 2.5mM MgCl. The PCR protocol began with an initial denaturation at 94°C for 5mins, followed by 35 cycles of denaturation at 94°C for I min, annealing at 55°C for 1min Aaa and extension at 72°C for 1min. This PCR ‘number Character Character atate regime was used for both genes. Following 1 Tiesmds: a, present; b, absent cycling, the success of the amplification " Desma a, tetraclonal; b, dicranoclonal; Was determined by electrophoresing on an 4 development. c, absent* ethidium bromide-stained agarose gel and a, articulated at ends of zygomes visualised by short-wave UV illumination. Zygosis (Fig. 2A,C); For each taxon, a total of 10 PCR reactions 3 N b, articulated along zygomes < architecture. (Fig. 2E); were carried out and the products were c, absent* pooled. This was an effort to reduce the " a, oxea; potential for an amplification-induced error 4 Monactinal b, tylostrongyles (Fig. 2C); 1 + $ megascleres. v ved ie (Hie, 1A in the sequences. The pooled amplification ` T products were electophoresed on a single a, short-shafted discotriaenes 1 d the band ‘sed (Figs 1C, 2B); agarose gel and the band was excise i b, short-shafted phyllotriaenes using a clean scalpel blade. The DNA 5 Triaene (Fig. 2D); megascleres, c, long-shafted ornamented dicho- and trichotriaenes (Fig. 1A,B); d, long-shafted ortho- and dichotriaenes | | B. Taxon 112]3/4|5|6 78/9 6 Euasters. a, present; b, absent Discodermia ajajajaja|bjbjaja Theonella sp. 1 alblb|b|blalb 7 Sepite a, present; b, absent Leonera Sp 212 (Fig. 2F). Theonellasp.2 | a | a |a| b |b| b |b|aj|b Small acanthose . b 8 microrhabds a, present; b, absent Corallistes ip. &| bd e A (Fig. 2B). Geodia blcjlcla|ld|aj|bj|bi|b (outgroup) Large acanthose jm 9 microrhabds a, present; b, absent telletta (Fig. 2B). (outgroup) bpeje a[d[a[b[b b 346 fragment was extracted from the agarose using the Quiaex If (Quaigen Lid, UK.) PCR purification kit. DNA sequencing reactions were carried out using the Amplitaq FS sequencing kit {Applied BioSystems, Inc.). The sequencing protocol tor both genes was carried out as per the manufacturer's instructions. Sequencing was carried out on an Applied Biosystems 373 automated sequencing apparatus and the data was analysed using the Sequence Navigator software (Applied Biosystems Inc.). We estimate that 98% of gene regions were covered by more than one sequencing read, with approximately 25% being covered by three or more sequencing overlaps. ‘The sequences have been deposited in the EMBL sequence repository under the accession numbers AJ224646-AJ224651 (demosponge 188 TENA), AJ224123-AJ224124 (hexactinellid 188 rRNA), AJO0591 |-AJ00S918 (28S rRNA), MOLECULAR PHYLOGENY RECON- STRUCTION. Nucleotide positions whose identities were not possible io establish unamb- iguously, were coded according to the International Union of Pure and Applied Chemistry (IUPAC) nomenclature, The sequences were aligned using the Genetic Data Environ- ment (Smith, 1993). In the majority of cases, the positional orthology of the nucleotides was relatively easy to establish. A conservative approach to the alignment was taken, with those positions whose homology was not possible to establish with absolute certainty, being excluded from subsequent analyses, Attempts in the most difficult cases, to relate the sequences to each other, on the basis of RNA secondary structure proved recalcitrant, In the absence of conclusive grounds for establishing homology. the sites in question were excluded. Phylogenetic hypothesis construction and sequence statistics were evaluated using PAUP*4.0d54 (Swoftford, 1993). Transition-transversion ratios were calculated from each dataset by first constructing a neighbor-joining iree from LogDet distances (Lockhart et al., 1994), This tree was considered a working hypothesis of relationships, The estimate of transition-transversion ratio might he influenced to some extent by tree topology, but this influence was not thought to be significant. sing maximum likelihood criteria, the transition-transversion ratio was chosen that yielded the highest likelihood value. The gamma shape parameter for each dataset was calculated using the optimised transition-transversion ratio, again using maximum likelihood as the MEMOIRS OF THE QUEENSLAND MUSEUM optimisation criterion. The gamma shape param- eter that yielded the highest likelihood was chosen. RESULTS MORPHOLOGICAL PHYLOGENY RECON- STRUCTION. A single minimum length tree was obtained of length 14 anda very high consist- ency index (CI) of 1.0. The phylogenetic tree hypothesises that species of Theonella are more recently derived (han Discodermia, and that these two genera form a clade more recently derived than Corallistes. This topology is identical tu one of the reconstructions derived from sequence data (Fig. 3A), and supports the current classif- ication that recognises the differentiation of Discodermia and Theonella from Corallistes in the Family Theonellidae and Family Corallist- idag, respectively (see Kelly-Borges & Pomponi, 1994), Morphological characters 2,3,7. 8 and 9 (Table 2) differentiate Discodermia and Theonella trom Coralistes, and the states of characters 4 and 5 differentiate Theonella from Discodermia, and both from Corallistes, In Discodermia and Theonella the desmas are tetracrepid (character 2a) with four clones (zygomes) which clasp (zygose) at their very ends (character 3a; Fig. 24,0). Corallistes, on the other hand, has dicrano- clonal desmas (character 2b) in which zygosis spreads along the zygomes (character 3b; Figs 1A, 2B) The desma skeleton of these genera is supplemented with monaxonal megascleres - oxea in Discodermia (character 4a), long and curved with blunt hammer-like ends (tylostrongyles) in Theonella (character 4b; Fig. 2C), and whispy roughened oxea-like spicules in Corallistes (character 4e, Fig. LA). Triaenose megascleres are found at the surface of the sponge with their head rays parallel with the surface and the rhabd perpendicular to the surface (Fig. 1A,C). These are discotriaenes in Discodermia (character 5a; Fig. |C, 2B), phyllotriaenes in Theonella (character 5b; Fig. 2D), and ornamented (character 5c; Fig. 14,B) or plain (Fig. 2F) dichotriaenes in Corallistes. Microscleres ate scattered throughout the sponge body and otten form a thick surface crust, In Discodermia there are lwo size categories of roughened microrhabds (characters 8a, 9a; Fig. 2H), whereas in Theonella there is onty one (character Ba; Fig. 2C). Corallistes lacks microrhabds but possess streptasters (Fig. 2F). RESOLUTION POTENTIAL OF 18S AND 285 rRNA SPONGE GENES TABLE 3. Oligonucleotide names and their corresponding sequences. The first and second oligonucleotides are designed to amplify a large portion of the 18S rRNA gene, whilst the third and fourth oligonucleotide sequences are designed to amplify an approximately 700bp stretch of the 28S rRNA gene. Chüginucleotde Sequence 18Sf20 TGG TAC GGT AGT GGC CTA CCA TGG 18Sr21 ACG GGC GGT GTG TAC AAA GGG CAG RD3A GAC CCG TCT TGA AAC ACG A RD5B2 ACA CAC TCC TTA GCG GA MOLECULAR PHYLOGENY RECON- STRUCTION. In total, six taxa had a portion of their 18S genes sequenced. The final alignment was in excess of 1,300bp in length. The positions whose alignment could not be determined unamb- iguously were removed. When all taxa were considered, the final alignment was 1,135 positions in length, whereas with the exclusion of the hexactinellid sequences, the number of align- able positions increased to 1,269. For the 28S rRNA gene dataset, almost 700bp were sequenced for each taxon. The final alignment lengths were 505 positions for the eight taxon dataset and 583 positions for the ingroup taxa alone. A total of eight alignments were analysed for their information content. Statistics that were evaluated included the number and percentage of variable sites, the number and percentage of parsimony-informative sites, the estimated gamma shape parameter for rate variation across sites and the transition-transversion ratio. The results of these analyses are given in Table 4. The top half of the table refers to the alignments that were used when the two hexactinellid outgroup sequences were included, whereas the bottom half ofthe table refers to the ingroup only (in this instance lithistid demosponges only). In most cases, the exclusion of the hexactinellid outgroup sequences facilitated the use of a longer gene region. This was due to the difficulty of aligning some hexactinellid regions with their equivalent location in the demosponge genes. On the other hand, removal of the hexactinellid sequences had the effect of reducing the numbers of variable and informative sites, with a consequent increase in the number of constant sites. The combined dataset always contained the highest number of variable and informative sites as would be expected, but in 347 FIG. 1. Skeletal architecture of lithistid demosponges. Transverse sections have been taken through the sponge surface, cellular material removed using hydrogen peroxide, and viewed by SEM. A, Corallistes nolitangere. Transverse section through surface of sponge showing dicranoclonal desma reticulation (D). Long-shafted ornamented dicho- triaenes (T) emerge from the desma reticulation with the shaft perpendicular to the surface and the cladomes (head) parallel with the surface. Oxeote spicules (O) resembling fine hairs can be seen in residue cellular material. Streptaster microscleres pack the surface (M). Scale=238um. B, Corallistes nolitangere. View ofthe sponge surface showing the ornamented heads ofthe dichotriaenes, with the desma reticulation visible beneath. Scale=400um. C, Discodermia sp. Transverse section through surface ofsponge showing tetracrepid desma reticulation (D) and above this a dense crust oftwo sizes of acanthose microrhabd microscleres (M). Short-shafted disco- triaenes line the surface with discs overlapping. Scale=11 1 um. neither instance did 1t contain the greatest perc- entage. One of the striking features of both kinds of analysis is the performance of the 288 rRNA gene dataset. This region had the highest 348 MEMOIRS OF THE QUEENSLAND MUSEUM FIG. 2. Desma, triaene and microsclere morphology in Discodermia, Theonella and Corallistes. A-B, Discodermia sp. A, Tetracrepid desmas with tuberculate zygoses (Z) at the ends of the zygomes. Scale=125um; B, Short-shafted discotriaene (D), large and small acanthose microrhabd microscleres (M), and large regular oxea (0). Scale=20um. C-D, Thenonella sp. C, Tetracrepid desmas with zygoses (Z) at the ends of the zygomes. Acanthose microrhabd microscleres are of one size (M) and strongyles have tylote ends (S). Scale=30um. D, Short-shafted phyllotriaene. Scale=50um. E-F, Corallistes nolitangere. E, Dicranoclonal desmas with zygoses (Z) along the zygomes. Scale=100um; F, Long-shafted dichotriaene (D), streptaster microscleres (S). Scale=50um. RESOLUTION POTENTIAL OF 18S AND 288 rRNA SPONGE GENES 349 TABLE 4. Results from analyses of eight alignments. The top half of the table refers to the alignments that were used when the two hexactinellid outgroup sequences were included, whereas the bottom half of the table refers to the ingroup only (i.e. lithistid demosponges only), where Sympagella nux and Margaritella coeloptychioides were excluded. The first column contains the gene region that was used for each particular analysis. The words “all taxa’ in parentheses indicates that all six taxa were used in that particular analysis. 18S (short) refers to the data of Kelly-Borges & Pomponi (1994). eight taxon datasets indicates that the addition of more sequences caused more variation at sites that were already variable, whilst conserved sites remained so even with the addition of more distant taxa. Of the twelve possible substitution types, there are twice as many possible transversion mutations as transition Gene region Length | Variable | Informative | Gamma | Ti/Tv mutations. For these datasets, the 18S (long) (all taxa) | 1138 | 446 (39%) | 163 (14%) | 0.31 0.98 maximum likelihood transition-trans- : = : version ratio is slightly above 1.0 in 18S (short) (all taxa) | 473 | 137(2994) | 75 (1694) 0.24 1.20 most cases (except for the 188 rRNA _ 28S (all taxa) 505 | 142(28%) | 84 (17%) 0.24 1.79 dataset with the eight taxon Total (all taxa) 1643 | 588 (36%) | 247(15%) | 0.29 1,18 alignment). This is indicative of a bias 18S (long) (ingroup) | 1269 | 214 (17%) | 37 (396) 0.54 1.16 towards transition substitutions. At 18S (short) (ingroup) | 473 | 72 (15%) 8 (2%) 0.79 1.46 very extreme genetic distances, the 28S (ingroup) 583 | 169(29%) | 40 (7%) 0.40 1.26 transition-transversion ratio will Total (ingroup) | 1852 | 383(21%)| 77(4%) 0.34 1.19 converge to 0.5. This is not apparent in percentage of parsimony-informative sites in both datasets and when outgroup taxa were removed, it also had the highest percentage of variable sites. The gamma shape parameter, which is an estimate of the rate variation across the sites in the alignment, ranged from 0.24-0.31 for alignments that included all taxa, and from 0.34-0.79 for ingroup alignments. The increased estimate of rate variation across sites (lower gamma value) in the 18S (long) Topology a Theonella sp1 Theonella sp2 Discodermia Corallistes Theonella sp1 Theonella sp2 Discodermia Corallistes Theonella sp1 Theonella sp2 Corallistes Discodermia any of the datasets in these analyses. The short 18S alignments generally produced trees with low amounts of resolution. Frequently none of the hypothesised phylogenies in Figure 3 were seen in the resulting bootstrap partition tables. The longer 18S alignments provided a greater amount of resolution, with bootstrap proportions sometimes becoming quite high. The 285 alignments were also quite well resolved and these also yielded high bootstrap values, particularly for the phylogeny that is 28S All taxa Ingrou FIG. 3. Phylogenetic reconstructions of the relationships between Discodermia, Theonella and Corallistes. The first column (topology) indicates the topology that is under consideration and the open circle indicates the internal branch whose support is being assessed. The second column indicates the type of analysis that was undertaken: LD - LogDet; ML - Maximum Likelihood; P - Parsimony. The results are in four consecutive blocks according to the gene region that was used in the analysis. Within each block, the left side indicates the bootstrap proportions for the alignment that included all taxa and the right side indicates the results when only the ingroup sequences were used. favoured by the morphological data (Fig. 3A). The combined sequence dataset alignments also displayed a reasonable amount of signal. DISCUSSION The analysis of sequence statistics showed that the 28S rRNA gene region provides the greatest amount of information per unit sequence length (Table 4). Despite the fact that the 285 alignment was hampered by the necessity of removing a large hypervariable portion, it still contained a high number and percentage of variable and parsimony-informative sites. The 18S gene behaved in a less efficient way. The combined dataset always contained the largest number of variable and parsimony-informative sites, but it did not contain the highest percentage of these sites in any of the analyses. It is curious to note the behaviour of the estimated shape of the gamma parameter, a. The gamma shape parameter is an estimation of rate variation across sites. Lower gamma parameters indicate a more severe amount of rate variation whilst higher numbers indicate that the evolut- ionary rate is more equivalent at all sites. There was an obvious difference in values between the taxon-inclusion sets. When all taxa were considered, the gamma shape parameter was always lower than when only ingroup taxa were analysed. The addition of more taxa is simply increasing the amount of variability at sites that are already free to vary. Conserved sites remain unchanged with the addition of more taxa. Although the addition of more taxa has the effect of increasing the percentage of variable and parsimony-informative sites, the gamma shape of rate variation across sites is more marked. Not all of the three phylogenetic trees were congruent with the morphological phylogeny. The shorter ofthe two 18S rRNA alignments was unable to resolve the relationships ofthe four taxa of interest with any degree of confidence. Indeed during some bootstrap replicates, some other topologies, not considered in Figure 3 were found. The main reason for this was a complete lack of variability in the dataset. The longer 18S gene region was slightly more decisive about branching order. The topology that received strong support using this region was a pairing of Discodermia and Corallistes to the exclusion of the other taxa. However, this high level of support was only achieved using parsi- mony tree reconstruction. Given the general lack of confidence using the other methods, and the MEMOIRS OF THE QUEENSLAND MUSEUM inability of parsimony to compensate for super- imposed substitutions, it is possible that these findings are a result of long branch artifacts. This clade is rooted by a particularly long branch leading to the other demosponge taxa. The topol- ogy that was observed the least number of times during bootstrapping was the pairing of Discodermia with Theonella, and this was irresp- ective of type of analysis or taxon-inclusion set. The results for the 28S gene sequences were considerably different to those seen in the 185 analyses. For this gene, the placement of Corallistes as the sister taxon to Theonella was never seen in any analysis (Fig. 3C) Of the two remaining alternative topologies, the placement of Discodermia as sister taxon to Theonella (Fig. 3A) received considerably more support than the placement of Discodermia with Corallistes (Fig 3B). Reasonably high bootstrap support was seen for Discodermia and Theonella as sister-groups using all methods of analyses irrespective of whether the outgroup taxa were used or not. The alignment combining 18S and 288 rDNA data also yielded ambiguous results. The topology that suggests a sister taxon relationship between Corallistes and Theonellais very poorly supported (Fig. 3C). The other two topologies are more strongly supported, but neither was supp- orted with any degree of confidence and the differences in the levels of support do not justify acceptance of one over the other. It is likely that the ambiguous nature of the results from the 18S rRNA gene have a detrimental effect on the combined alignment. During bootstrapping, a character can be selected from either gene region. Given that the 18S gene region is approximately 15094 larger than the 28S rRNA gene region, it will probably contribute more to each replicate on the whole. The result of this seems to be the carry-over of the ambiguous results from the separate analysis that used only the 188 rRNA gene. The topology that was most strongly supported using the 28S gene is consistent with the hypotheses of relationships deduced from morphological characters; Díscodermia and Theonella are more closely related to each other than they are to Corallistes, and they are the more derived taxa. The 28S rRNA gene generally has a higher proportion of variable and parsimony-informative sites and can provide the best possibility of resolving poriferan phylo- genetic relationships, at least at the sub-ordinal level. RESOLUTION POTENTIAL OF 188 AND 288 rRNA SPONGE GENES 3 ACKNOWLEDGEMENTS The authors thank the Division of Biomedical Marine Research, Harbor Branch Oceanographic Institution, and the crew of the ‘Johnson- Sea-Link II” manned submersible for logistic field and lab support for the collections of lithistid and hexactinellid material. We also thank the Coral Reef Research Foundation for the collection of Theonella specimens by SCUBA from the Republic of Belau, Micronesia. We are very grateful to Dr Henry Reiswig, Redpath Museum, Montreal, for identification of the hexactinellid material, and Klaus Borges, BMNH London, for SEM photography. Financial assistance through the British Airways Conservation Awards scheme is greatly apprec- iated. This research and JMcl’s postdoctoral fellowship was made possible by the Biological and Biotechnological Science Research Council, UK, and supported by The Natural History Museum, London. This is Harbor Branch Oceano- graphic Institution Contribution No. 1292, and a contribution from the Coral Reef Research Foundation. LITERATURE CITED BAROIN, A., PERASSO, R., QU, L.-H., BRUGEROLLE, G., BACHELLERIE, J.-P. & ADOUTTE, A. 1988. Partial phylogeny of the unicellular eukaryotes based on rapid sequencing of a portion of 28S ribosomal RNA. Proceedings of the National Academy of Sciences USA 85: 3474-3478, CHOMBARD, C., BOURY-ESNAULT, N. & TILLIER, S. 1998. Reassessment of homology of morphological characters in Tetractinellid sponges based on molecular data. Systematic Biology 47(3): 351-366. HADJU, E., SOEST, R.W.M. VAN & HOOPER, J.N.A. 1993. Proposal for phylogenetic subordinal classification of poecilosclerid sponges (Demo- spongiae, Porifera). Pp. 123-140. In Soest, R.W.M. van, Kempen, T.M.G. van & Braekman, J.-C. (eds) Sponges in Time and Space. (Balkema: Rotterdam). IWABE, N., KUMA, K.-I., HASEGAWA, M., OSAWA, S. & MIYATA, T. 1989. Evolutionary relationship of archaebacteria, eubacteria, and eukaryotes inferred from phylogenetic trees of duplicated genes. Proceedings of the National Academy of Sciences USA 86: 9355-9359. KELLY-BORGES, M. & BERGQUIST, P.R. 1997, Revision of South-west Pacific Polymastiidae (Porifera, Demospongia, Hadromerida) with descriptions of new species of Polymastia Bower- bank, 7y/ocladus Topsent and Acanthopolymastia nov. gen. from New Zealand and the Norfolk Un nidi Ridge, New Caledonia. New Zealand Journal of Marine and Freshwater Research 31: 367-402. KELLY-BORGES, M., BERGQUIST, P.R. & BERGQUIST, P.L. 1991, Phylogenetic relationships within the Order Hadromerida (Pori- fera, Demospongiae, Tetractinomorpha) as indicated by ribosomal RNA sequence compar- isons. Biochemical Systematics and Ecology 19(2): 117-125. KELLY-BORGES, M., ROBINSON, E.V., GUNASEKERA, S.P., GUNASEKERA, M., GULAVITA, N.K. & POMPONI, S.A. 1994. Species differentiation in the marine sponge genus Discodermia (Demospongiae: Lithistida): the utility of ethanol extract profiles as species-specific chemotaxonomic markers. Biochemical Systematics and Ecology 22(4): 353-365. LOCKHART, P.J., STEEL, M.A., HENDY, M.D. & PENNY, D. 1994. Recovering evolutionary trees under a more realistic model of sequence evol- ution. Molecular Biology and Evolution 11(4): 605-612, PUHLER, G., LEFFERS, H., GROPP, F., PALM, P., KLENK, H.-P., LOTTSPEICH, F., GARRETT, R.A. & ZILLIG, W. 1989. Archaebacterial DNA- dependent RNA polymerases testify to the evol- ution of the eukaryotic nuclear genome. Proceedings of the National Academy of Sciences USA 86: 4569-4573. RIVERA, M.C. & LAKE, J.A. 1992. Evidence that eukaryotes and eocyte prokaryotes are immediate relatives. Science 257(5066): 74-76, SANDFORD, F. & KELLY-BORGES, M. 1997. Redescription of the Atlantic hermit-crab-sponge Spongosorites suberitoides Diaz et al, (Demo- spongiae, Halichondrida, Halichondriidae). Journal of Natural History 31: 315-328. SMITH, S. 1993. GDE: The Genetic Data Environment. (Available from author at http://golgi. harvard.edu/ftp/GDE/). SOGIN, M.L., ELWOOD, H.J. & GUNDERSON, J.H. 1986. Evolutionary diversity of eukaryotic small- subunit rRNA genes. Proceedings of the National Academy of Sciences USA 83: 1383-1387. SWOFFORD, D. L. 1993. PAUP: Phylogenetic Analysis Using Parsimony. (Smithsonian Institute: Washington DC). SOEST, R.W.M. VAN 1987. Phylogenetic excercises with monophyletic groups of sponges. Pp. 227- 241. In Vacelet, J. & Boury-Esnault, N. (eds) Taxonomy of the Porifera. NATO ASI Workshop Proceedings, G13 (Springer-Verlag: Berlin, Heidelberg). WEST, L. & POWERS, D. 1993. Molecular phylo- genetic position of the hexactinellid sponges in relation to the Protista and Demospongiae. Molecular Marine Biology and Biotechnology 2(2): 71-75. WOESE, C. R. 1987. Bacterial Evolution. Microbiological Reviews 51(2): 221-271. MEMOIRS OF THE QUEENSLAND MUSEUM THE PHYLOGENETIC POSITION OF THE SPONGE SPONGOSORITES SUBERITOIDES DETERMINED BY ANALYSIS OF 288 RRNA GENE SEQUENCE, Memoirs of the Queensland Museum 44: 352. 1999:- A number of problems exist in the phylogenetic consideration of the sponge Spongosorites suberitoides that cannot be resolved on morphological grounds alone. Placing the sponge in the genus Spongosorites divides this genus into two groups; a single shallow water species and many deep-water species. Described differences between these groups include; oxea size, aerophobic colour-change and surface texture. Further, S. suberitoides shows an affinity with hadromerid sponges such as colour in life, texture, arrangement of anastomosing choanosomal tracts and the lack of an aerophobic reaction. In its association with hermit crabs and gastropods it resembles the genus Suberites. It also shows similarities to species of Aaptos and some Polymastiidae. The DNA sequence of the five prime region of the 28S ribosomal gene of S, suberitoides is compared with DNA sequences from hadromerid and halichondrid species in a phylogenetic analysis to resolve the position of this species. O Porifera, phylogeny, morphology, DNA, 28S ribosomal gene. Grace P. MeCormack (email: gimecormack@ nhm.ac.uk), James O Melnerney & Michelle Kelly*, Department of Zoology The Natural History Museum, Cromwell Rd, London, SW7 5BD, UK: Floyd R. Sanford, Biology Department, Coe College, 1220. First Ave, NE, Cedar Rapids, Towa 52402-5092, USA, * Present address: Marine Ecology and Aquaculture, National Institute of Water & Atmospheric Research (NIWA), Private Bag 109-695, Newmarket, Auckland, New Zealand; 1 June 1998. PHYLOGENY OF LITHISTID SPONGES. Memoirs of the Queensland Museum 44: 352. 1909:- Kelly-Borges and Pomponi (1994) utilised partial 185 rRNA gene sequences to resolve relationships within lithistid sponges. (Porifera; Demospongiae). While their results lent weight to the growing realisation that the Order Lithistida is polyphyletic, their conclusions were hampered by low levels of sequence variation. Our initial study sought to evaluate the resolution potential between regions of the 18S and 28S rRNA genes within a group of selected Porifera. Approximately 1,300bp of the 18S rRNA gene and a 5" region of the 288 rRNA gene were compared with the data of Kelly-Borges and Pomponi (1994). Six taxa were selected which represented a gradient of relationships, ranging through the taxonomic levels of genus, family and class. We found that the 700bp of the 288 rRNA gene presented the greatest potential for resolution of this group of porifera at the genus and family level, and that this molecular phylogeny is congruent with morphological hypotheses for the group. The study has progressed to include a number of other lithistid and non-lithistid taxa. O Porifera, Lithistida, 188 rRNA, 288 rRNA, phylogeny. James O. Melnerney (email: j.mcinerneytanhm.ac.uk) de Michelle Kelly*, Department of Zoology, The Natural History Museum, Cromwell Road, London SW? SBD, UK; * Present address: Marine Ecology and Aquaculture, National Institute of Water & Atmospheric Research (NIWA), Private Bag 109-695, Newmarket, Auckland, New Zealand; | June 1998 A NEW DENDROCERATID SPONGE WITH RETICULATE SKELETON MANUEL MALDONADO AND MARIA J, URIZ Maldonado, M. & Uriz, M.J. 1999 06 30: A new dendroceratid sponge with reticulate skele- ton. Memoirs of the Queensland Museum 44: 353-359. Brisbane. ISSN 0079-8835. A new encrusting dendroceratid sponge characterised by a skeletal network of ascending primary spongin fibers transversally interconnected by secondary fibers is described from the Alboran Sea (W Mediterranean). Primary fibers, provided with a subcircular basal plate for attachment to the substratum, are unbranched or poorly branched, fasciculate in most cases, with a pith containing foreign material, and a laminated bark. Secondary fibers are also laminated, usually lack any coring material, and may form fenestrated plates around the point where they anastomose to a primary fiber. The reticulation of the skeleton is loose and irregular, so that some primary fibers are interconnected through their basal portions only, some are interconnected along their apical portion, and some are even isolated, lacking any transversal interconnection. Although skeletal features of both the darwinellid genus Aplysilla and the dictyodendrillid genus Jgernella are recognisable in this new sponge, it cannot be taxonomically assigned to either genus, unless the current diagnosis of one of them is modified. A reanalysis of the chemical, histological and skeletal evidence available to date gives little support to the hypothesis that reticulate skeletons appeared in Dendroceratida as a single evolutionary event. Consequently, we propose expanding the diagnosis of the genus Pleraplysilla to include species with an irregular network of secondary fibers, such as Pleraplysilla reticulata sp. nov. O Porifera, Keratose sponges, Dendroceratida, Pleraplysilla, Darwinellidae, Dictyodendrillidae, new species. Manuel Maldonado (e-mail: maldonado@ceab.csic.es) & María J. Uriz, Department of Aquatic Ecology. Centro de Estudios Avanzados de Blanes (CSIC). Camino de Santa Barbara s/n. Blanes 17300. Girona. Spain; 2 March 1999. The siliceous skeleton that characterises most sponges in the class Demospongiae is replaced by a skeleton of spongin fibers in a group of sponges that comprises the orders Verongida, Dictyoceratida and Dendroceratida. To provide skeletal support to the bulk of soft sponge tissue, spongin fibers may be either anastomosed to form a network or organised as sets of diversely dendritic, unconnected structures. Dendritic skeletons occur in some Verongida and some Dendroceratida, whereas reticulate skeletons occur in all three orders. Reticulate skeletons display a wide variety of models of organisation not only within orders, but also families. Differences involve morphology, orientation and dimensions of the basic mesh, as well as several degrees of divers- ification, both in size and structure, of the anastomosing fibers (e.g. Van Soest, 1978; Bergquist, 1980, 1995, 1996; Bergquist et al., 1998). Given such a structural diversity, the acquisition of a reticulate skeleton is likely to have evolved independently in Dictyoceratida, Verongida and Dendroceratida. This idea is implicitly assumed in the historical classification of these so-called ‘fibrous’ or ‘keratose’ sponges, which are distributed in three different orders. Nevertheless, it is also assumed that the acquisition of a reticulate skeleton within the order Dendroceratida was a single, synapo- morphic evolutionary step. Consequently, dendroceratid genera with reticulate and non- reticulate skeletons are placed in two different families (e.g. Bergquist, 1980, 1995, 1996; Bergquist et al., 1998): Dictyodendrillidae Bergquist (with a reticulate skeleton made of primary fibers interconnected by secondary fibers) and Darwinellidae Merejkowsky (with exclusively dendritic skeletons made of a single type of fiber). However, some chemical and histological studies have reported unexpected affinities between members of these two families. For example, studies on the diterpenoid chemistry (Bergquist et al.,1990) suggested a relationship between the dictyodendrillid genus Igernella Topsent (with a reticulate skeleton and spongin spicules) and darwinellid genera with a typically dendritic skeleton, such as Darwinella Miiller (with spongin spicules) and Pleraplysilla Topsent (without spongin spicules). Similarly, studies on choanocyte chamber structure (Dendy, 1905; Vacelet et al., 1989; Boury-Esnault et al., U3 Un L 1990) revealed that the genus Dysidea Johnston, traditionally included in Dictyoceratida, and members of the family Darwinellidae share a remarkable feature: the presence of eurypylous chambers. At first sight, these non-skeletal affinities do not appear to be consistent with the subdivision of Dendroceratida into Darwinellidae and Dictyopleraplysillidae on the basis ofthe skeletal pattern. Rather, chemical and histological affinities suggest that reticulate skeletons may have arisen independently in Dendroceratida more than once, so that families diagnosed on the basis of this skeletal trait would not represent monophyletic groups. Never- theless, these non-skeletal features should not be considered as conclusive evidence for a new classification within Dendroceratida. Bergquist (1996) claimed that the reliability of chamber structure as a character to support a high level classification of keratose sponges is not great, since eurypylous chambers are not exclusively found in Dysidea and Darwinellidae, but they also occur in the verongid genus Janthella Gray. Furthermore, although the available body of information from the diterpenoid chemistry and histology is clearly in conflict with a taxonomic scheme based on a separation between reticulate and dendritic dendroceratid genera, it fails to reveal any robust alternative pattern of taxonomic relationships. Thus, despite the fact that skeletal pattern remains the only exclusive familial characteristic supporting a division of Dendroceratida into Dictyodendrillidae and Darwinellidae (Bergquist,1996), such a taxonomic scheme persists as the best option to date. Here we describe a new dendroceratid sponge in which the skeletal traits of the Darwinellidae and the Dictyodendrillidae are combined, suggesting that even the skeletal criterion may not be as robust as first thought to maintain the current familial scheme in Dendroceratida. Unfortunately, the contribution of the new material described here to an understanding of the relationships in Dendroceratida has been undermined by two unfortunate mishaps. First, although several individuals of this new species occurred at the collection site, we only collected one specimen because we mistook them for material belonging to the common species Darwinella muelleri Schulze, which closely resembles the new species when under water. This would not be an insurmountable problem, if the collection site (the Alboran Island) had not been a remote Mediterranean location. MEMOIRS OF THE QUEENSLAND MUSEUM Unfortunately, the Alboran Island was only accessible via a costly scientific cruise. Second, the tissue of the holotype became somewhat macerated and therefore useless for providing information on choanocyte chamber structure, since, when we realised the importance of this specimen, it had already been exposed to air during dissection to obtain its skeletal fibers. In 1996, that is eight years after this collection, a second scientific cruise (‘Fauna Ibérica - IV’) visited the same collection site, but the team of divers failed to find any individuals from this new species. Ten years after its collection and despite the problems mentioned above, we have finally decided to record this material in the scientific literature for two reasons. First, the well- preserved skeletal traits of the specimen clearly indicate that it belongs to an undescribed species. Second, this material provides skeletal inform- ation that may be crucial in retracing the path of skeletal evolution in Dendroceratida as further information is gained. MATERIALS AND METHODS The material was collected by SCUBA from the sublittoral bottom of the Alboran Island (W Mediterranean: 35?56'45"N, 3?01'38"W; 24m deep) during the ‘Ecopharm-I’ Cruise in 1988. The specimen was fixed in a 4% formalin solution and stored in 70% alcohol. The skeletal arrangement was studied under dissecting and compound microscopes after partial dissection of the specimen. Its fibers were also studied under a Hitachi S-2300 SEM, after a process of dehyd- ration in a graded series of ethanol, critical point and coating with gold palladium in an E-5000 sputtering. The holotype is deposited in the collection of the Museo Nacional de Ciencias Naturales (MNCN), Madrid, Spain. For comparative purposes, we also examined several individuals of Pleraplysilla spinifera (Schulze) from the Alboran Island and the NE coast of Spain (authors” collection), the holotype of /gernella mirabilis Lévi from Indonesia (Zoólogisch Museum Amsterdam, ZMA: POR9316), Caribbean specimens of 7gernella notabilis (Duchassaing & Michelotti) (authors' collection and ZMA POR6938 specimen), and a specimen of Dendrilla cirsioides Topsent from Banyuls (ZMA POR74). NEW DENDROCERATID SPONGE SYSTEMATICS Class Demospongiae Sollas Order Dendroceratida Minchin Familia Darwinellidae Merejkowsky Pleraplysilla Topsent, 1905 Encrusting Darwinellidae with a fiber skeleton of ascending primary fibers that may be either isolated or diversely fused to each other with or without the development of secondary fibers. Primary fibers are short (few mm), simply or partially fasciculate, poorly branched or unbranched, with a laminated bark and a pith filled with debris, and stand on the substratum on which the sponge grows, attached by means of a small basal plate. Secondary fibers, when present, are concentrically laminated, without coring material or with scarce scattered inclusions. Spongin spicules are absent (emended). Pleraplysilla reticulata sp. nov. MATERIAL. HOLOTYPE: MNCN 1.01/182. W Mediterranean, Alboran Island, 35°56’45"N, 3°01°38"W; 24m depth. ETYMOLOGY. Named for the reticulate condition of the fiber skeleton. DESCRIPTION. Encrusting, 2x3cm, 0.5cm thick; dull yellow alive, with some lemon yellow zones; soft, slippery to touch, with some mucus on the surface; surface sparsely conulose; conules 1.5-2mm high and 2-4mm apart; oscules and ostia punctiform, grouped in depressed areas located among conules. Choanosome fleshy, with tiny aquiferous channels; skeleton as a loose, irregular network made of ascending primary fibers transversally interconnected by secondary fibers (Fig. 1A-B); primary fibers non-branched or poorly branched, erect, attached to the substratum by means of a small basal plate (Fig. 2A, C). Fibers have a concentrically laminated bark and a pith contain- ing foreign material (Fig. 1C). Two or more adjacent primary fibers usually fused, yielding fibers with fasciculate appearance (Fig. 1A-C); primary fibers == > -34 - q ? 237.5-332.5/9.2 /92-12.4-16.1 | /11.5-14.8-20.7 <2. 16.1-23.5-345 | 9.2-12.7-184 450.0-550.0/ 2, 600.0-900.0/ 188.0-256.0/ 5 250.0-350.0/ ? 9.4-12,5 12.5-19.0 45.0-80.0/ ? 370-410 34-12,5 313.0-504.0-625.0/?, | 597.0-761.0-955.0 | 171.0-210.0-305.0 1 250.0-363.0-625.0/? | /7.5-164-21.3 7? 45:0-61.9-83.0/|*27.0-47:.0-646" || :8:5-12:5-16.0 361.0-522.5/ 9.2-13.8, |674.5-781.3-931.0 | 103.5-170.5-253.0 | 39.1-48.6-66.7/ n3 ë 256.5-418.0/9.2-16.1 | /9.2-12.6-18.4 | /27.6-41.7-529 | 23-35-46 | 299-443-59.8 | 6.9-10.8-13.8 DESCRIPTION. A complete description is provided by Boury-Esnault (1973), and expanded here. Megascleres (refer to Table 3 for dimensions). Orthotriaenes with short rhabd-like calthrops; rhabd hastate and mucronate on one side; cladome with clads slightly curved. Oxeas hastate or mucronate, slightly curved, sometimes straight or strongly curved; axial canal visible. Microscleres (refer to Table 3 for dimensions). Centrotylote microxeas smooth and slightly curved with acerate ends. Aspidasters elliptical- shaped, nearly regular outline, surface micro- spines stellate-shaped with 6-10 slightly conical points; developmental forms are visible with serrated margins because of stria that radiate from its central point; small hilum. Oxyasters with 10-14 microspined rays, spines more conc- entrated at the distal extremities. REMARKS ON CARIBBEAN ERYLUS The Brazilian coast is a continuity of the Caribbean biogeographic Province. Warm and shallow-water species have their southernmost limits along the coast of Santa Catarina State (27°S) (Fig.1), and some species extend up to the subtropical region of the coast of Rio Grande do Sul State (30°S) (Fig.1) and neighbouring areas (Mothes, 1996), such as £. alleni. Nine species of Erylus were listed in the Caribbean fauna by Pulitzer-Finali (1986). 1) E. goffrileri Wiedenmayer, 1977. 2) E. amphiastera Wintermann-Kilian & Kilian, 1984. 3) F. ministrongylus Hechtel, 1965. 4) E.alleni de Laubenfels, 1934, considered by Van Soest & Stentoft (1988) to be synonymous with E. transiens (Weltner, 1882), but reinstated here, for reasons described above, as a distinct species and sister species of E. transiens. 5) E. clavatus Pulitzer-Finali, 1986, also considered by Van Soest & Stentoft (1988) as a probable synonym of E. transiens, apparently differing only in the narrower width of the aspidasters; E. clavatus could also be considered as a synonym of E. formosus, however it has aspidasters (with proportion 1:3), which are not comparable with those of the latter species. 6) E. formosus Sollas, 1886. 7) E. trisphaera (de Laubenfels, 1953) (originally described in Unimia), and 8) E. bahamensis Pulitzer-Finali, 1986, both have much narrower aspidasters (with proportion 1:9) than other Caribbean species, however, E. formosus and E. trisphaera differ by the presence of oxyasters, and E. trisphaera has trilobate aspidasters. 9) E. discophorus (Schmidt, 1862) and E. euastrum (Schmidt, 1868), both originally described in Stelletta from the Adriatic, are certainly not conspecific with Caribbean species given their disjunct distributions. Stellettinopsis 378 MEMOIRS OF THE QUEENSLAND MUSEUM 5 um FIG. 6. Erylus corneus Boury-Esnault (Schizoholotype MCNPOR 2505). A, oxea extremity. B, orthotriaene. C, microxea. D, oxyaster. E, aspidaster. F, aspidaster surface. euastrum Schmidt, 1880 was cited from Grenada ministrongylus in having strongyles, by Van Soest & Stentoft (1988), but this dichotriaenes and elliptical aspidasters (with specimen may belong to £. transiens. Ofallthese proportion 1:2), although differing by the species E. diminutus sp. nov. is closest to E. presence of microstrongyles and a single BRAZILIAN ERYLUS 379 TABLE 3. Data on spicule micrometries of Erylus corneus Boury-Esnault, 1973. Holotype and Schizoholotype. Measurements given in um. Key to material of E. corneus: 1, Holotype - MNHN-NBE 973 [data from the author]; 2, Schizoholotype - MCNPOR 2505 scanning electron micrographs; to Lisandra de Moura Umpierre and Lia Gongalves Possuelo (MCN and Fundacáo de Amparo à [slides], Pesquisa do Estado do Rio Grande Material Orthotriaenes Oxeas | Aspidasters | Microxeas | Oxyasters do Sul-FAPERGS) for slides of 1 56.0-125.0 — | 546.0-673.0/ | 125.0-153.0/ | 37.0-56.0/ | 145559) Spicular dissociation, thick _actines 9.0 —19.0 69,0-84:0 1.0-3.0 - sections and preparations for SEM ads 4 0-680.0/ | 119.6-147.2/ | 27.6-57.5/ study; 10, Luciatio de Azevedo s 1962137/57-| 80-195 | 724474 | 23-33 | 92230 | Moura for the finishing of the 92 drawings. category of oxyasters. A taxonomic revision of these Caribbean species is currently in progress. DISCUSSION Lendenfeld (1910) introduced the term ‘aspidaster’ for the special spicules of Erylus, being smooth in the young stages, shield-like shape (flattened) and oval, rarely round or irregular discs. However, some species have globiform, either circular or ellipsoidal aspid- aster spicules, identical to sterrasters of Geodia. Consequently, we propose to enlarge the definition of the genus here, given that both sterr- asters and aspidasters may be found in some species of Erylus (i.e. E. polyaster, E. geodioides and E. topsenti). The proposition to enlarge the scope of the genus to include additional forms of spicules is based only on adult spicules described here, and from species previously recorded in the literature. In the present revision, a ‘provisionally endemic’ new species is described; the distrib- ution of E. formosus is enlarged; the presence of E. alleni is recorded for the first time for the Brazilian coast; and zoogeographical data on marine demosponges from Brazil, recorded by Hechtel (1976) and Mothes (1996), are expanded. ACKNOWLEDGEMENTS We are particularly grateful to Rob van Soest (ZMA) for donation of Caribbean samples and for confirmation of identifications made here. Our thanks also go to Klaus Riitzler (USNM), Clare Valentine (BMNH) and Peter Bartsch (ZMB) for lending type material; to Arno A. Lise (PUCRS), Ricardo R. Capitoli (FURG), Maria Marlúcia Ferreira Correia (UFMA) and José Audísio C. Luna (UFPE), for donating Brazilian samples; to Francisco Kiss (Universidade Federal do Rio Grande do Sul -UFRGS) and technician (MCN/FZB) who provided the Most of the material examined here was dredged under the auspices of: Diretoria de Hidrografia e Navegacáo da Marinha (DHNM); Departamento de Recursos Pesqueiros da Superintendéncia de Desenvolvimento do Nordeste (SUDENE); Pontificia Universidade Católica do Rio Grande do Sul (PUCRS); Projeto dos Recursos Vivos da Zona Económica Exclusiva (REVIZEE/N and REVIZEE/NE), supported respectively by Universidade Federal do Maranháo and Universidade Federal de Pernambuco; Projeto dos Recursos Vivos da Zona Económica Exclusiva (REVIZEE/SUL), supported by Fundação Universidade do Rio Grande (FURG); “Programa Rio Grande do Sul-I’ (PRGS-I), supported by Universidade de Sáo Paulo e Governo do Estado do Rio Grande do Sul; Projeto Talude, Fundacáo Universidade do Rio Grande (FURG), Brazil; Britannic Expedition H.M.S. “Challenger”; and French Expedition “Calypso”. LITERATURE CITED BOURY-ESNAULT, N. 1973. Campagne de la ‘Calypso’ au large des cótes atlantiques de l'Amérique du Sud (1961-1962). I, 29. Spong- iaires. Annales de l'Institut Océanographique, Paris 49 (Supplement): 263-295. DENDY, A. 1916. Report on the Homosclerophora and Astrotetraxonida collected by H.M.S. ‘Sealark’ in the Indian Ocean. In Reports of the Percy Sladen Trust Expedition to the Indian Ocean in 1905. Vol. 6. Transactions ofthe Linnean Society of London, Zoology 17: 225-271. DESQUEYROUX-FAUNDEZ, R. & SOEST, R.W.M. VAN 1997. Shallow waters Demosponges of the Galápagos Islands. Revue Suisse de Zoologie 194(2): 379-467. GRAY, J.E. 1867. Notes on the arrangement of sponges, with the description of some new genera. Proceedings of the Zoological Society of London (1867): 492-558. HECHTEL, G.J. 1965. A systematic study ofthe Demo- spongiae of Port Royal, Jamaica. Bulletin of the Peabody Museum of Natural History 20: 1-103. 1976. Zoogeography of Brazilian Marine Demo- spongiae. Pp. 237-260. In Harrison, F.W. & Cowden, R.R (eds) Aspects of Sponge Biology. (Academic Press: New York). LAUBENFELS, M.W. DE 1934. New sponges from the Puerto Rican deep. Smithsonian Miscllaneous Collections 91(17): 1-28. 1936. A discussion of the sponge fauna of the Dry Tortugas in particular, and the West Indies in general, with material for a revision of the families and orders of the Porifera. Carnegie Institute of Washington Publication.Papers of the Tortugas Laboratory 30(467): 1-225. 1953. Sponges from the Gulfof Mexico. Bulletin of Marine Science of the Gulf and Caribbean 2(3): 511-557. LENDENFELD, R. VON 1903. Tetraxonia. Pp. 1-168. In Schulze, F.E. (ed.) Das Tierreich. (Berlin). 1907. Die Tetraxonia. In Chun, C. (ed.) Wissen- schaftliche Ergebnisse der deutschen Tiefsee-Expedition auf dem Dampfer ‘ Valdivia’ 1898-1899, Jena 11: 59-374. 1910. The Erylidae. In Reports on the scientific results of the Expedition to the eastern Tropical Pacific, in charge of Alexander Agassiz, by the U.S. Fish Comission Steamer ‘Albatross’ 1904- 1905 and of other expeditions of the ‘Albatross’, 1888-1904. Memoirs of the Museum of , Comparative Zoology, Harvard 41(2): 261-324. LEVI, C. 1973. Systématique de la classe de Desmo- spongiaria (Démosponges). Pp 577-631. In Brien, P., Lévi, C., Sàra, M., Tuzet, O. & Vacelet, J. (eds) Traité de Zoologie. Anatomie, Systématique, Biologie. (Sér.ed. Grassé, P-P.) (Masson et Cie: Paris). MOTHES, B. 1996. Esponjas da Plataforma Contin- ental Norte e Nordeste (Amapá ao Maranháo) do Brasil. PhD Thesis (Universidade de Sào Paulo: Sào Paulo). MOTHES, B. & BASTIAN, M.C.K. de A. 1993. Esponjas do Arquipélago de Fernando de Noronha (Porifera, Demospongiae). Iheringia (Zoologia) (75): 15-31. MOTHES-DE-MORAES, B. 1978. Esponjas tetra- xonidas do litoral sul-brasileiro: II. Material coletado pelo N/Oc. ‘Prof. W. Besnard’ durante o Programa RS. Boletim do Instituto Oceanográfico, Sào Paulo 27(2): 57-78. PULITZER-FINALI, G. 1986. A collection of West Indian Demospongiae (Porifera) with, in appendix, a list of the Demospongiae hitherto recorded from the West Indies. Annali del Museo Civico di Storia Naturale di Genova 86: 65-216. RIDLEY, S.O. 1884. Spongiida. Report on the zoological collections made in the Indo-Pacific Ocean during the voyage of H.M.S. ‘Alert’, 1881-2. Pp. 366-482, 582-635. (British Museum (Natural History): London). MEMOIRS OF THE QUEENSLAND MUSEUM SCHMIDT, O. 1862. Die Spongien des Adriatischen Meeres. Pp. 1-48 (Wilhelm Engelmann: Leipzig). 1868. Die Spongien der Küste von Algier. Mit Nachtrügen zu den Spongien des Adriatischen Meeres (Drittes Supplement) (Wilhelm Engelmann: Leipzig). 1880. Die Spongien des Meerbusen von Mexico (und des Caraibischen Meeres). Pp. 33-90 Zweites (Schluss-) Heft (G. Fischer: Jena). SOEST, R.W.M. VAN 1994, Demosponge distribution patterns. Pp. 213-223. In Soest, R.W.M. van, Kempen, T.M.G. van & Braekman, J.-C. (eds) Sponges in Time and Space. (Balkema: Rotterdam). SOEST, R.W.M. VAN & STENTOF, N. 1988. Barbados deep water sponges. Studies on the Fauna of Curaçao and Other Caribbean Islands 70(215): 1-175. SOLE-CAVA, A.M., KELECOM, A. & KANNEN- GIESSER. G.J. 1981. Study of some sponges (Porifera Demospongiae) from the infralitoral of Guarapari, Espirito Santo, Brazil. [heringia (Zoologia) (60): 125-150. SOLLAS, W.J. 1885. A classification of the sponges. Annals and Magazine of Natural History (5)16: 395, 1886. Preliminary account of the tetractinellid sponges dredged by H.M.S. ‘Challenger’. Part I. The Choristida. Scientific Proceedings of the Royal Dublin Society 5: 177-199. 1888. Report on the Tetractinellida collected by the H.M.S. ‘Challenger’, during the years 1873-76. Pp. 1-455. In Report on the Scientific Results of the Voyage of H.M.S. ‘Challenger’ during the Years 1873-76. Vol. 25 (Her Majesty’s Stationary Office: London, Edinburg, Dublin). TOPSENT, E. 1894. Etude monographique des spong- iaires de France. 1. Tetractinellida. Archives de Zoologie Expérimentale et Générale (3)2: 259-398, pls 11-16. WELTNER, W. 1882. Beitráge zur Kenntniss der Spongien. (Inaugural-Dissertation: Univ. Freiburg). WIEDENMAYER, F. 1977. Shallow-water sponges of the Western Bahamas. Experimentia Supplementa 28: 1-28 (Birkhauser: Basel). WILSON, H.V. 1925. Silicious and horny sponges collected by U.S. Fisheries Steamer ‘Albatross’ during the Philippine Expedition, 1907-10. In Contributions to the biology of the Philippine Archipelago and adjacent regions. Bulletin of the United States National Museum 100(2, 4): 273-532. WINTERMANN - KILIAN, G. & KILIAN, E. F. 1984. Marine Sponges of the Region of Santa Marta (Colombia) - Part II. Homosclerophorida, Chor- istida, Spirophorida, Hadromerida, Axinellida, Halichondrida, Poecilosclerida. Studies on Neotropical Fauna and Environment 19(3):121-136. ORIGIN OF THE METAZOA: A REVIEW OF MOLECULAR BIOLOGICAL STUDIES WITH SPONGES WERNER E.G. MULLER AND ISABEL M. MULLER Müller, W.E.G. & Müller, I.M. 1999 06 30: Origin of the Metazoa: A review of molecular bi- ological studies with sponges. Memoirs of the Queensland Museum 44: 381-397. Brisbane. ISSN 0079-8835. The phylogenetic position of Porifera is near the base of the kingdom Metazoa. During the last few years rRNA sequences, and more importantly cDNAs/genes coding for proteins, have been isolated and characterised from sponges, especially from the marine demosponge Geodia cydonium. Analyses of their deduced amino acid sequences allowed a molecular biological approach to solve the problem of monophyly of the Metazoa. Molecules of the extracellular matrix/basal lamina, with the integrin receptor, fibronectin and galectin as prominent examples, cell-surface receptors (tyrosine kinase receptor), elements of nerve systems (crystallin, metabotropic glutamate receptor) as well as homologs/modules of an immune system (immunoglobulin-like molecules, SRCR- and SCR-repeats, Rhesus system) unequivocally classify Porifera as true Metazoa. As living fossils sponges also show pecularities not known in other metazoan phyla provided with simple, primordial molecules allowing cell-cell and cell-matrix adhesion as well as processes of signal transduction known in a more complex manner from higher Metazoa. Tissues of sponges are rich in telomerase activity, suggesting a high plasticity in the determination of cell lineages. Based on this experimental background a first successful approach to establishing a cell culture from a sponge was possible. It is concluded that molecular biological studies using sponges as models will not only help us to understand the evolution ofthe Metazoa from the Protista, but also the complex, hierarchial regulatory network of cells in higher Metazoa. O Porifera, evolution, monophyly, receptors, phylogeny, molecular biology, (Eu) Metazoa. Werner E.G. Müller (email: Wmueller(àmail. UNI-Mainz.DE), Institut für Physiologische Chemie, Universitat, Abteilung für ‘Angewandte Molekularbiologie', Duesbergweg 6, D-55099 Mainz, Germany; Isabel M. Müller, Akademie gemeinniitziger Wissenschaften zu Erfurt (Kommission: Molekulare Evolution), GotthardtstraBe 21, D-99084 Erfurt, Germany; 19 February 1999. The transition from unicellular to multicellular organisms has taken place in all five kingdoms of life; this process took place separately in Fungi (Ascomycota), Plantae (Chlorophyta) and in Metazoa. The origin of plants appears to be well established within the phylum Chlorophyta (Margulis & Schwartz, 1995), whereas the origin of Fungi, and especially of Metazoa, is perhaps the most enigmatic of all phylogenetic problems (Willmer, 1994). The origin of the Metazoa remained uncertain until a few years ago. At that time two questions were paramount: 1) what were the relationships between the different metazoan phyla in general, and between the lowest metazoan phylum, the Porifera (sponges), and those of higher inverte- brates in particular; and 2) what are the ancestor(s) of the Metazoa among the Protista ? Some authors favoured the idea that sponges had unicellular ancestors different from those of other Metazoa [polyphyly] (Margulis & Schwartz, 1995), while other scientists (e.g. Morris, 1993), believed that multicellular animals evolved only once [monophyly]. One powerful approach particularly helpful in answering questions on the presence or absence of corresponding structures in sister groups is to gather molecular data from the respective taxa. Here, a clear distinction must be made. Nucleo- tide [nt] sequence data have been gathered from genes of species in different phyla, encoding small and large ribosomal RNA. These data have been used to build phylogenetic trees to resolve deep branches. The outcome in most reports was that bootstrap statistics supporting mono- or polyphyly is low (e.g. Cavalier-Smith et al., 1996), or even not at all significant (Rodrigo et al., 1994). In a second, separate concept, amino acid [aa] sequences deduced from nt sequences from genes and proteins that are known from metazoan systems (e.g. immune, adhesion, sensory systems), have been used to obtain reliable 382 FIG. 1. Adhesion molecule ofthe sponge C. cydontum electron micrograph, A. Native form af the aggregation factor. B, Aggregation factor core structure with the circular center and the 25 radiating arms. Preparation shadowed with platinum (magnification x 140,000). insight into the branching of metazoan phyla from a potential common ancestor (Pfeifer etal., 1993). Our research group has introduced this approach to establish the phylogenetic position of Porifera, within the Metazoa. Our data are compatible with the view that all Metazoa are monophyletic in origin (Müller el al. 1994; Miiller, 1995). We also discuss evidence for evolution of cell lineages. in early Metazoa; we demonstrale the use of a new type of sponge “organotypic” cell culture in cell proliferanon and cell death: and finally we discuss evidence for the separation of the Porifera into (wo subphyla, DISCUSSION GENOME SIZE OF PORIFERA. Using the method of cytometry and DAPI staining, the DNA content of haploid cells (C-value) from Geodia cydenium and Suberites domuncula was MEMOIRS OF THE QUEENSLAND MUSEUM found to be 1.7pg, corresponding to 1.7x10'kb (Imsiecke et al, 1995). This unexpectedly high value is further exceeded by the result which came from a separate technique, the determin- ation of DNA reassociation kinetics. hia recent calculation, based on the determination of genetic complexity, a value of 3.3pg DNA/ haploid genomic set was calculated (Bartmann- Lindholm et al, 1997). In comparison, the € value for human cells is 3.4x10 kb (see Li & Graur, 1991). The number of chromosomes m the diploid state in S. domuncula is 32 (Imsiecke et al,, 1995), The recent finding that five major sub- components of DNA could be distinguished m G. cvdonium by density gradient centrifugation (Bartmann-Lindhalm etal., 1997) was surprising asit indicates a heterogeneity which has not been reported in any other metazoan. The genetic complexity within these subcomponents was determined by reassociation kinetics to vary from 2.1x10'bp to 1.4x10°bp, corresponding to a content of single copy DNA of >93% (Bartmann-Lindholm et al., 1997). The extreme heterogeneity in DNA composition of the genome of 6, cvdonium suggests that an unusually high exchange of well-defined DNA regions oceurred in this animal. METAZOAN GENES/PROTEINS IN PORIFERA. Molecules/modules of the extracellular matrix/basul lamina in G. cydonium. ll is now well acknowledged that repeated sequences or modules (Patthy, 1995) are found in proteins of the complement system, extracellular matrix and also in various cell- surface receptors. Mobile modules, according ta the defmition of Patthy (1995), are protein- coding domains that are flanked by introns of identical phase, facilitating a dispersal within the genome, It is shown here that most polypeptides deduced from cDNA sequences Irom sponges are assembled hy an unusually large variety of such modules. Two types of adhesion systems have been described in sponges, cell-cell and celi-matrix systems. With respect to the first system, the aggregation factor (AF) is considered to be the major extracellular molecular complex (Fiz. l). AFs were enriched from two sponge species. Microciona prolifera and G. cydoninm (reviewed in Miiller, 1982). The AF is a multiprotein complex which interacts with a membrane component, the aggregation receptor (AR), that has been identified but not yet purified. 384 MEMOIRS OF THE QUEENSLAND MUSEUM A Multiadhesive protein 4 100 200 300 400 500 600 700 701 aa 254 342 343 443 462 551 Module Fibronectin SRCR SCR-Sushi = HE xum FNS Group B Type itt => = ae Location extracellular matrix lymphocytes T-, B lymphocytes macrophages erythrocytes macrophages a => > Function interaction with immune reactions complement activation integrin clotting B Cc D FN3 SRCR SCR MAP_GEODIA APU_THESA Bacteria MAP_GEODIA CR2 MOUSE FAS2 SCHAM 130 HUMAN APOH, HUMAN FS21_DROME | WOT-BOVIN VB05 VACCO FINC HUMAN CD6 HUMAN HIG_DROME çi Deuterostoma LFC_TACTR FINC_RAT "m CD5 HUMAN FIG. 3. A, Scheme of the putative *multiadhesive protein' (MAP GEODIA), cloned from G. cydonium. Three modules can be identified in this protein; the fibronectin- (FN3), the a scavenger receptor cysteine-rich (SRCR)- and the a short consensus repeat (SCR; Sushi) module. B-D, Phylogenetic analyses ofthe modules. B, Unrooted phylogenetic tree composed from the deduced aa sequences of FN3 modules found in 1) Metazoa, from deuterostomes human (FINC HUMAN), and rat (FINC RAT; module 6), from protostomes FN3 from D. melanogaster (FS21 DROME) and the arthropod S. americana (FAS2 SCHAM; module one), the pseudo- coelomate C. elegans (SR13 CAEEL) and the sponge G. cydonium (MAP GEODIA), as well as in 2) Bacteria T. saccharolyticum (APU_THESA; module one). C, Phylogenetic tree computed from five SRCR scavenger molecules of group B; the modules from the sponge MAP GEODIA, the human CD6 antigen (CD6 HUMAN), human CDS surface glycoprotein (CD5 HUMAN), human M130 antigen (M130. HUMAN) and bovine WC1 surface antigen (WC1_HUMAN) were analyzed. D, Phylogenetic tree constructed from SCR modules of the following six polypetides; the module from MAP_GEODIA together with the corresponding type III SCRs from human beta-2-glycoprotein I precursor (APOH_HUMAN), mouse CR2 complement receptor type 2 precursor (CR2_MOUSE), D. melanogaster locomotion-related protein (HIG_| DROME), Limulus clotting factor C precursor from Tachypleus tridentatus (LFC_TACTR) and the host range protein precursor from the vaccinia virus strain LC16MO (VB05 VACCO). The evolutionary distance of 0.1 aa substitutions per position in the sequence is given. matrix (ECM): 1) fibronectins, 2) collagens, and Fibronectin. Fibronectins are high molecular 3) galectin. weight glycoproteins, present in most ECM and also in blood plasma. A typical fibronectin SPONGE MOLECULAR BIOLOGICAL STUDIES A 0000000000020 0590060000 000000000 140 kDa a i: ware INTEGRIN ~ n FIBRONECTIN í *o* COLLAGEN GALECTIN || _ 00000 ooo B [e] INTEGRIN INTEGRIN chains: heavy 88kDa light 18 Da 12 3 45073 1,086 aa "= E da ia ca’ go ALEFHPRYSSTOO--TLVLHTLLNGHNQTEVEVTGFPPA Peewee HAS EA e— Len og eMe PTRADS SRSA Tren Bye e FDE- — — UVIRTQ*ANNESQU* EREGEON MEE ne ***I** " FDAH*DQQAV*Y*SPQG* *G* *OREG"** AED Qr** FDABSDVNLI *C* SKKMERS G* *QRE «V * QR. C** ^» FHAHFDANTI*CT*KED*T*G* RADA **qP |*A***Qe pl MIT ilii 1i a "i Un liii FIG. 2. A, Scheme of aggregation factor (AF)-mediated cell recognition in G. cydonium. A 29-kDa aggregation receptor (AR) is inserted into the plasma membrane to which one galectin molecule binds. In the presence of Ca^ a second galectin molecule binds to the first one. Then these two molecules form a bridge between the AR and the 140 kDa polypeptide, associated with the AF. Following this scheme, the interactions between the AF and the AR involves the galectin which might bind to carbo- hydrate both at the 140 kDa polypeptide and at the AR. In addition, the sponge contains an integrin receptor which is assumed to interact with fibronectin and collagen. B, Sponge integrin receptor. Schematic presentation of the structural features in G. cydonium integrin a subunit. The heavy and light chains are indicated (the light chain is shaded). Positions of the 8 characteristic repeats of integrins are marked | to 8. Three putative Ca’’-binding sites as well as the C-terminal trans- membrane region are indicated. C, Dendrogram computed from the deduced aa sequences of integrin a subunits found to be most homologous to sponge integrin a (INTG GEODIA) with the corresponding sequences from the invertebrate species, Caenorhabditis elegans (YMA1 CAEEL) and Drosophila melanogaster (ITAP_DROME), and the vertebrate species, mouse integrin aV (ITAV MOUSE), chicken integrin a VIII (ITA8 CHICK) and human fibronectin receptor a subunit (ITAS5 HUMAN). D, Multiple alignment ofthe conserved region within the galectin carbohydrate binding domains. The asterisks (*) mark sequence identities between G. cydonium galectin-1 and -2 (GEODIA-1 and -2) and other galectins from C. elegans (CAEEL), eel (EEL), chicken (CHICK), mouse (MOUSE), rat (RAT) and human (HUMAN). The consensus sequence for galectins is given. Monoclonal antibodies have been used as tools to identify the binding domains of the AF (Wagner-Hiilsmann et al., 1996). A 140kDa polypeptide was found to participate in the AF- mediated reaggregation process. This polypeptide interacts with a galectin that links individual AF molecules to the AR at the plasma membrane, and consequently bridges two cells together (Miiller et al., 1997) (Fig. 2A). Confocal laser scanning microscopical analysis demonstrated that both the galectin and the AF are present at the rim of the cells (Wagner-Hiilsmann et al., 1997). Within the last few years, major elements characteristic of a basal lamina have also been discovered in sponges. Besides cDNAs coding for two proteins with similar complexity as those found in higher Metazoa, collagen type IV, integrin, and one fibronectin module (FN3) were identified. Integrin. One major class of extra- cellular matrix (ECM) receptors are the integrin receptors. Integrins are membrane-anchored heterodimer receptors composed of a- and p subunits. At least 16 different a- and 8 different B subunits have been identi- fied in vertebrates which yield more than 20 heterodimeric integrin receptors. We have isolated and characterised cDNA clones encoding the a subunit of an integrin from the marine sponge G. cydonium (Pancer et al., 1997a). The open reading frame encodes a 118,628Da polypeptide (Fig. 2B). Most a subunits of integrins, including the one from the sponge, contain 7-8 repeating domains (Fig. 2B). Like other a subunits of integrins the sponge molecule also contains putative divalent cationic-binding sites. A dendrogram was computed from the deduced aa sequences of integrin a subunits (Fig. 2C). The integrin receptor binds primarily molecules of the following three families, present in the extracellular SPONGE MOLECULAR BIOLOGICAL STUDIES consists of more than 10 type I-, approximately 2 type II-, and more than 15 type III (FN3) modules. Protein(s) have been isolated from G. cydonium that immunologically cross-react with human anti-fibronectin antiserum (Pahler et al., 1998). The main bands have sizes of 230 and 210kDa. In addition, a cDNA was cloned, encoding a putative ‘multiadhesive protein’ which comprises three interesting modules: 1) a fibronectin, 2) a scavenger receptor cysteine-rich (SRCR), and 3) a short consensus repeat (SCR) module (Fig. 3A). The fibronectin module of the deduced sponge protein (Pahler et al., 1998) comprises the characteristic topology and aa found in fibro- nectin type-II] (FN3) elements. Even though it remains to be proven in Porifera that this FN3 module functions as an adhesion molecule, the finding supports the immunochemical data on the presence of fibronectin-like molecules in sponges. FN3 modules have been primarily described in Metazoa; in addition they are found in a related sequence in extracellular glycohydrolases from soil bacteria (Bork & Doolittle, 1992). The unrooted phylogenetic tree (Fig. 3B) reveals that the FN3-related sequence of the bacterium branches off first from a common ancestor, while the deuterostomes (human and rat) and the protostomes (Drosophila melanogaster and Schistocerca americana) are grouped within one branch, and the acoelomate (Caenorhabditis elegans) and the sponge (G. cydonium) in another (Fig. 3B). From these data we conclude that the sponge FN3 module from G. cydonium is phylogenetically the oldest one within Metazoa. Collagen. Collagens constitute a superfamily of extracellular matrix proteins. Until recently it was accepted that collagens are present only in Metazoa. However, a new class of collagens recently identified in fungi is assumed to have arisen by convergence (Celerin et al., 1996). In sponges, primitive fibrillar collagens have been seen in several species; Chondrosia reniformis (Garrone et al., 1975), and G. cydonium (Diehl- Seifert et al., 1985a). The finding that sponges contain basement membrane collagen type IV was spectacular, as this is known to be the scaffold ofthe basal lamina (Boute et al., 1996). Galectin. As mentioned, the sponge AF interacts with the AR via galectin and the 140kDa poly- peptide (Fig. 2A). The galectin, which occurs in isoforms, was studied in detail. The purified molecules reveal forms of Mr 13 to 18kDa (Bretting et al., 1981) which bind specifically to 385 the sugars dGalNAc, dGalB1>4dGlcNAc, dGalB1—3dGIcNAc and dGalNAc. In the presence of Ca” or glycoconjugates the sponge galectins undergo conformational changes and “polymerise” to large three-dimensional clumps (Diehl-Seifert et al., 1985b). The cDNAs of two isoforms ofthe galectins from G. cydonium were cloned (Pfeifer et al., 1993; Wagner-Hülsmann et al., 1996). The predicted proteins deduced from the complete sequences display high similarity with the corresponding molecules from vertebrates and C. elegans (Hirabayashi & Kasai, 1993; Müller et al., 1997). The deduced aa sequences ofthe two isoforms feature the charac- teristic carbohydrate-recognition domain LHFNPR-G-V-N-W-E-R[H]-PF (the aas given in bold are those directly involved in binding to the carbohydrate); this domain is conserved from sponge to human (Pfeifer et al., 1993) (Fig. 2D). Based on the extent of aa substitutions the two sponge galectins were calculated to have 2,*7— HOMO LJ RATTUS E o »-P»zuom-zzm«c XENOPUS EEL ANCESTOR I N CAENO M E R T E GEODIA ! B R GEODIA 1 A - LLL. lI. IP pP A onsens LHFNPR-:-G-:-V-:-N-:-W-:-E-:-R[H]-:-PF FIG. 4. Phylogenetic tree computed from the deduced aa sequences of galectins from: 1) vertebrates - human (HOMO), rat (RATTUS), chicken (GALLUS), frog (XENOPUS), conger eel (EEL); and 2) invertebrates - nematode (CAENO) and G. cydonium (GEODIA; isoform I and II). The con- sensus sequence for galectins are given. Time scale indicates the time of divergence, based on aa substitution analysis. 386 diverged from the galectin isolated from the nematode C. elegans, 800 MYA (Hirabayashi & Kasai, 1993; Pfeifer et al., 1993) (Fig. 4). CELL-SURFACE RECEPTORS. Besides adhesion receptors, receptors involved in signal transduction, or elements of signal transduction pathways coupled to them have also been cloned from G. cydonium. The results have been recently reviewed (Müller & Müller, 1997; Müller, 19972); so only a brief summary is given here. One-transmembrane-segment receptor-receptor tyrosine kinase. Protein tyrosine kinases (PTKs) play important roles in the response of cells to different extracellular stimuli. PTKs are divided into two major groups, the receptor tyrosine kinases (RTKs), which are membrane spanning molecules with similar overall structural topologies, and the non-receptor TKs, also comp- osed of structurally similar molecules. The first RTK from lower Metazoa was identified and cloned from G. cydonium (Müller & Schácke, 1996). The putative aa sequence comprises 1) the extracellular part with a Pro/Ser/Thr-rich region, and two complete immunoglobulin (Ig)-like domains, 2) the transmembrane domain, 3) the juxtamembrane region, and 4) the catalytic tyrosine (TK)-domain. A similarity search with the G. cydonium TK-domain aa sequence showed that all RTKs fall in one branch of the tree while the non-receptor TKs are grouped in a second one; sponge RTK is placed in a separate branch, which splits-off first from the common tree of metazoan PTKs. An estimation of the time of divergence of the sponge RTK from RTKs of other metazoans was 650-665 MYA (Gamulin et al., 1997). Seven-transmembrane-segment receptors. The first seven-transmembrane-segment receptor in sponges was isolated from G. cydonium. lt is the metabotropic glutamate receptor (see description below). Most seven-spanning receptors transmit extra- cellular signals through G-proteins. G-proteins, coupled to the putative receptors, have been identified in G. cydonium. G-proteins are hetero- trimers composed of a--, B- and y subunits (Seack et al., 1998). Several secondary effector enzymes of the seven-transmembrane-segment receptor/G protein-linked receptor have been cloned from G. cydonium: the Ser/Thr kinases (STKs). These kinases are ubiquitously present in animal tissues; they express their activities in response to second messengers (e.g. Ca^ or diacylglycerol). MEMOIRS OF THE QUEENSLAND MUSEUM The STKs have been sequenced from G. cydonium (Kruse et al., 1996, 1997, 1998). A comparison of the complete structures of the sponge STKs, which are identical to those of nSTKs and cSTKs from higher Metazoa, with the structures of protozoan, plant and bacterial Ser/ Thr kinases, indicates that the metazoan STKs are different from the non-metazoan enzymes. These data imply that metazoan STKs have a universal common ancestry with the non-meta- zoan STKs with respect to the kinase domain, but they differ from them in the overall structural composition. NEURONAL-LIKE ELEMENTS IN PORIFERA. Until now no molecular evidence has been presented in demosponges for the existence ofa nervous-like cell system. Recently our group identified two elements (crystallin and a metabotropic glutamate receptor) in sponges, which are characteristic for sensory systems in higher Metazoa. Crystallins. Crystallins are categorised into two classes, the ubiquitous crystallins and the taxon- specific crystallins. No structural or functional characteristics are common to all crystallins. The a-, B- and y-crystallins are classed as ubiquitous crystallins and are found in almost all vertebrate species. The second class, the taxon-specific crystallins, includes a series of ‘enzyme erystallins*, which display catalytic functions. Until recently, no molecular sequence data was available for B -crystallins in invertebrates. The cDNA coding for the By-crystallin molecule was isolated from G. cydonium (Krasko et al., 1997). The sponge sequence comprises the four repeated motifs which compose the two domains of the By-crystallin. The peptide shows striking homologies to vertebrate By-crystallins. Each motif is composed of the four D-strands and one ‘Greek key” signature (Fig. 5A). Like in other crystallins a signal peptide is missing in the sponge sequence, suggesting that it is an intra- cellular structural protein (Krasko et al., 1997). Thus, molecules from light-sensory organs, in this case crystallins, are also present in sponges. Nerve cell receptors - presence of sensory cells: Metabotropic glutamate receptor. Sponges are (according to the literature) not provided with nerve cells. However, recently we showed that isolated cells from the marine sponge G. cydonium react to the excitatory neurotransmitter glutamate with an increase in the concentration of intracellular Ca”, (Ca?”),. This effect was also measured if the compounds L-quisqualic acid SPONGE MOLECULAR BIOLOGICAL STUDIES 387 A CRYSTALLIN B Rh-LIKE PROTEIN Domain 1 Domain 2 RH30. | HUMAN 30 kDa RH30 MACMU RH GEODIA N c 50 kDa motif 1 motif 2 motif 3 motif 4 0.1 RH50 HUMAN RH CAEEL C MGRL GC N-terminal- Transmembrane- C-terminal- Segments 129 aa 282 aa 117 aa -— | "1 2 1 1 —Ron TM: | MW it iV V vi VII mGluR4,5 mGluR1-8 MHC II GABAR acr. SP-10 D Antagonists y Agonist: Glutamate Plasma membrane 0000000 ba Uy ea C-terminus Phospholipase C: Ca M. FIG. 5. A, Crystallin: Sponge By-crystallin from G. cydonium. Schematic representation of the x -crystallin folding pattern. Two Greek key motifs form one domain. Domains 1 and 2 form the monomeric Px-crystallin. B, Rhesus-like antigen: Relationship of the sponge Rhesus(Rh)-like protein to animal Rh- and Rh-like antigens. Unrooted phylogenetic tree computed from the sponge Rh-like protein (RH_GEODIA) and Rh-related proteins: the Rh30 Ag from human (RH30 HUMAN) and rhesus monkey (RH30 MACMU), the RhD-like polypeptide from C. elegans (RH. CAEEL) and the human Rh50 Ag (RH50 HUMAN). Two clusters, comprising the ~30 000 Mr and the —50 000 Mr Rh polypeptides, are grouped. C, Metabotropic glutamate receptors (mGluRs): Schematic presentation of the sponge receptor (G. cydonium), showing its three segments. The sequences with the highest similarity to the sponge segments are listed. D, mGluRs: Scheme of the sponge mGluR, inserted by seven transmembrane segments into the cell membrane; it is coupled to G-proteins. 388 (L-QA) or L-(+)-2-amino-4- phosphono- butyric acid (L-AP-4) were used. The effects of L-QA and L-AP-4, both agonists for metabotropic gluta- mate receptors (mGluRs), could be abolished by the antagonist of group 1 mGluRs, (RS)-a- methyl-4-carboxyphenyl- glycine. These data suggest that sponge cells contain a mGluR-like protein - hence it is justified to state that sponges are provided with sensory-like cells. The demon- stration of a neuronal-like receptor in sponges also allows to challenge the question for the underlying molecules, involved in the coord- ination ofthe cells for contraction. It is interesting to note that some sponge species are known to migrate in a directed manner, e.g. the sponge Tethya spp (Fishelson, 1981) or Ephydatia fluviatilis (Bond & Harris, 1988). Using a cDNA encoding the rat mGluR subtype 1, a complete nucleotide sequence of G. cydonium cDNA coding for a 528 aa long protein (59kDa) was identified which displays high overall similarity to mGluRs as well as to GABABRs. The deduced sponge polypeptide, termed a putative mGlu/GABA-like receptor, displayed the highest similarity to the two families of metabotropic receptors within the transmembrane segment. The N-terminal part of the sponge sequence shows similarity to the mGluR4 and -5. These findings suggest again that the evolutionarily earliest metazoan phylum, Porifera, possesses a complex intercellular communication and signaling system as known from the neuronal network of higher Metazoa (Perovic et al., 1999) (Fig. 5C and D). HOMOLOGUES/MODULES OF THE IMMUNE SYSTEMS. Little is known about natural challenges to self integrity in sponges (reviewed in Pancer et al., 1996). In their extensive review Smith & Hildemann (1986) have grouped sponge alloimmune responses seen in experimental transplantations into two major rejection processes. Some species form barriers to separate themselves from non-self tissue, while others react by cytotoxic factors which destroy the transplant. However, until recently, no molecule has been identified which can be considered to be involved in self/non-self responses in sponges. Three modules are now known in deduced aa sequences of cDNAs isolated from G. cydonium, which are present also in immune molecules of higher Metazoa; 1) immunoglobulin-like domains, 2) proteins featuring scavenger MEMOIRS OF THE QUEENSLAND MUSEUM receptor cysteine-rich domains, and 3) molecules comprising short consensus repeats. Immunoglobulin-like (Ig-like) domains. To determine if the two immunoglobulin-like (Ig-like) domains of the RTK from G. cydonium display sequence polymorphism (Pancer et al., 1996), allo- and autografting experiments were performed using two grafting methods: 1) para- biotic attachment, and 2) insertion technique. Thirty-six pairs of auto- and allografts were assayed. All of the autografts fused, while only two allografts fused and 34 pairs were incomp- atibile. At the molecular level the two Ig-like domains of RTK were analyzed from two pairs of fusing and one pair of rejecting sponges (Pancer et al., 1996). High nt and aa polymorphism was recorded. Proteins featuring scavenger receptor cysteine-rich domains. Proteins featuring scavenger receptor cysteine-rich (SRCR) domains comprise a superfamily, which includes one invertebrate and several vertebrate proteins. The SRCR domain consists of a 110 aa-residue motif with conserved spacing of six to eight cysteines, which apparently participate in intra- domain disulfide bonds. Proteins of this superfamily feature 1-11 SRCR domain repeats. We identified the putative SRCR protein - belonging to group A of this family (Wijngaard et al., 1992) - from the marine sponge G. cydonium (Pancer et al., 1997b; Miiller, 1997b). Three forms of SRCR molecules were characterised, which apparently represent alternative splicing of the same transcript. The long putative SRCR protein features twelve SRCR repeats, a C-terminal transmembrane domain and a cyto- plasmic tail. The sequence of the short form is identical with the long form except that it lacks a coding region near the C-terminus, without the transmembrane domain. Homology searches revealed that the sponge putative SRCR protein shares with bovine T-cell antigen WC1 29.2% identity in 1054 aa overlap, 33.9% identity in 475 aa overlap with sea urchin speract, and 56% identity in 110 aa overlap with macrophage scavenger receptor type l. Recently, the SRCR module of the group B (Wijngaard et al., 1992) was also identified in the *multiadhesive protein’ (Pahler et al., 1998) (Fig. 3). The percentage of identity (homology) of this module of MAP GEODIA is most similar to the mammalian sequences M130 with 63% (44%) and WC1 with 50% (40%) and lower for CD5 32% (25%) and CD6 36% (25%). SPONGE MOLECULAR BIOLOGICAL STUDIES Phylogenetic analysis shows that the sponge MAP_GEODIA scavenger module branches off first from a common ancestor, whereas two other modules, the mammalian macrophage antigen M130 and the antigen WC1 (which is expressed on gamma/delta T lymphocytes) as well as those of the CD6 and the CDS antigen of lymphocytes, branch off later (Fig. 3C). Molecules comprising short consensus repeats. The short consensus repeats (SCR) also termed ‘Sushi domain’, with 11-14 conserved aa residues and four conserved cystein residues, are classified according to the consensus aa pattern into four types. In the course of seeking further splice forms of the sponge putative SRCR, a protein sequence was identified which contains the 12 SRCR repeats mentioned above plus two others that are linked to six SCR, the SRCR-SCR molecule (Pancer et al., 1997b; Müller, 1997b). The SCR modules present in this putative poly- petide belong to group II of the SCR family. The presence of an SCR module of type III ina sponge molecule, ‘multiadhesive protein’, was surprising (Fig. 3A). This module belongs to the SCR repeats, which are dominant building blocks in the complement receptor of type 1, type 2, and factor H, but also in a few non-complement proteins (e.g. D2-glycoprotein I; reviewed by Reid & Day, 1989). The phylogenetic tree built from the selected SCRs of group III revealed that the sponge SCR module forms the basis for the two related SCRs from mammals, mouse complement receptor and human beta-2-glyco- protein I precursor, and the two invertebrate sequences from the locomotion-related protein of Drosophila melanogaster and the Limulus clotting factor (Fig. 3D). The SCR from the vaccinia virus displays closest relationship to the mammalian sequences, suggesting horizontal gene transfer from host to the virus (Bishop, 1981). Rhesus-like protein. Vertebrate red blood cells display a variety of cell-surface molecules. Some, like the ABO system and the Rhesus (Rh) system of higher mammals, exhibit extensive polymorphism. Although the function of these antigens is poorly known, their role has been implicated in severe human disorders due to abnormal functioning or immunological destruction of the red blood cells (Nash & Shojania, 1987). A breakthrough in the analysis of the Rh system was marked by cloning of the Rh cDNA encoding the D antigen (Cherif-Zahar et al., 1990; Avent et al., 1990), and later also of the associated and closely related Rh50 protein 389 (Ridgwell et al., 1992; Le van Kim et al., 1992). Surpringly, a Rhesus-like protein (CDNA) of 57,000Mr was isolated from G. cydonium (Seack et al., 1997). Both the hydropathy profile of the sponge Rh-like protein and its high similarity to the aa sequence clearly show that the sponge molecule shares a common ancestor with the human and rhesus monkey Rh30 antigen, and with the —50,000Mr Rh-like polypeptides from humans and Caenorhabditis elegans (Fig. 5B). CELL LINEAGES. In contrast to higher metazoan phyla, sponges are characterised by a pronounced plasticity in the determination of cell lineages. In a first approach to elucidate the molecular mechanisms controlling the switch from the cell lineage with a putative indefinite growth capacity to senescent, somatic cells, the activity of the telomerase as an indicator for immortality has been determined. The studies were performed on two demosponges, Suberites domuncula and Geodia cydonium (Koziol et al., 1998). High activity for the telomerase in tissue of both sponges was found, reaching about 30% of that seen in telomerase-positive mammalian reference cells. In contrast, dissociated spherulous cells from G. cydonium, after an incubation period of 24hrs, contain no detectable telomerase activity. From earlier studies it is known that isolated sponge cells do not prolif- erate (Gramzow et al., 1989). From this it is assumed that the separation of the senescent sponge cell lineage from the immortal germ-/ somatic cell lineage is triggered by the loss of contact to cell adhesion factors. Preliminary evidence exists which suggests that the final progress of the senescent, telomerase-negative cells to cell death is caused by apoptosis (Fig. 6). ESTABLISHMENT OF A PRIMARY CELL CULTURE FROM A SPONGE. Despite the fact that cells from sponges contain high levels of telomerase activity, no successful approach to cultivate sponge cells has yet been described; in phyla which are evolutionary higher than sponges the somatic cells are telomerase- negative. One reason may be seen in the observation that after dissociation the cells lose their telomerase activity. In addition, no nutrients and metabolites have been identified that would stimulate sponge cells to divide. In close collaboration with the group of R. Borojevic and M.R. Custodio (Departamento de Histologia e Embriologia, Instituto de Ciécias Biomédicas, Universidade Federal, Rio de 390 MEMOIRS OF THE QUEENSLAND MUSEUM Porifera: germ- and somatic cell lin Higher Metazoa: germ cell lineage LII] Higher Porifera: Metazoa: somatic cell lineages loss of cell contact Number "somatic cells" Telomerase e Transformi galectin ran e rming even activity TRF apoptosis Precrisis cells: M-2 Crisis Replicative age FIG. 6. Hypothetical consequence of telomere loss on senescence of metazoan cells, The reduction of telomeres is given in number of loss of terminal restriction fragments (TRFs). It has been shown experimentally that the cells of the germ line in higher Metazoa and the cells of both the germ- and the somatic lineage in Porifera contain high levels of telomerase thus allowing the maintenance of stable telomeres. In higher Metazoa an early loss of telomerase activity determines the fate of the somatic cells to senescence (open box) via the two phases. 1) ‘Mortality Phase 1° (M-1) - cell cycle arrest - during which factors controlling life span via recognition of *damaged' DNA, e.g. the RB protein or p53 protein, are activated. 2) After transformation (downregulation of RB and/or p53) the second process, ‘Mortality Phase 2° (M-2), is initiated during which the telomeres reach in the precrisis cells a critical length from which a signal for cell death arises. In sponges it is proposed that the switch from immortal somatic cells to mortal ‘somatic’ cells occurs after a loss of adhesion factors for a given set of cells (closed box). The M-1 point is assumed to be reached after only a few rounds of cell replication. The growth arrest might be bypassed by addition of adhesion factor(s), e.g. galectin. The second phase towards cell senescence is supposed to be induced by central death signal(s), e.g. activation of the expression of the MA-3 gene, which are underthe control of either extrinsic- or an intrinsic factors. (Adapted from Harley, 1991; Harley et al., 1994). Janeiro; Brazil), we succeeded in defining the culture conditions required for the formation of multicellular aggregates of S. domuncula from dissociated single cells; these are termed ‘primmorphs’, resembling organotypic cell cultures (Custodio et al., 1998; Müller et al., 1999), These aggregates, formed in seawater supplemented with antibiotics, have a tissue-like appearance, and have been cultured for more than five months. Cross sections through (he prim- morphs revealed an organised zonation into a distinct. unicellular epithelial-like layer of pinacocytes and a central zone composed primarily of spherulous cells. After their association into primmorphs, the cells turn from the telomerase-negative state to the telomerase- positive state. Important is the finding that a major fraction of the cells in the primmorphs undergoes DNA synthesis and hence has the capacity to divide. We propose that the primmorph system developed by us is a powerful novel model system to study basic mechanisms of cell proliferation and cell death; it can also be used in SPONGE MOLECULAR BIOLOGICAL STUDIES aquaculture for the production of bioactive compounds and as bioindicator system. SYSTEMATIC CONSIDERATIONS ON THE TWO SPONGE SUBPHYLA: HEXACT- INELLIDA AND CELLULARIA. It has been proposed that Porifera should be divided into two subphyla, Cellularia (comprising Demospongiae and Calcarea), and Symplasma (containing only Hexactinellida) (Reiswig & Mackie, 1983). This classification reflects the fact that species belonging to the Cellularia are composed of uni- nuclear cells, while those in the Hexactinellida have syncytial tissues (Mackie & Singla, 1983). This fundamental structural difference raises the question whether the ancestors of the Metazoa in general, and the Porifera in particular, were colonial flagellates or syncytial ciliates. The cDNAs coding for proteins which have been used to establish the classification of subphyla within the Porifera, in particular, and the monophyly of Metazoa in general, came from two Demospongiae: Geodia cydonium (Jameson) (Demospongiae, Tetractinomorpha, Astrophorida, Geodiidae), Suberites domuncula (Olivi) (Demospongiae, Tetractinomorpha, Hadromerida, Suberitidae); one Calcarea: Sycon raphanus (Schmidt) (Calcarea, Calcaronea, Leucosoleniida, Sycettidae); and two Hexact- inellida Rhabdocalyptus dawsoni (Lambe) (Hexactinellida; Hexasterophora; Lyssacinosida; Rossellidae), Aphrocallistes vastus (Schulze) (Hexactinellida, Hexasterophora, Hexactin- osida, Aphrocallistidae). THE PHYLOGENETIC POSITION OF THE TWO SUBPHYLA OF PORIFERA. Two alternative hypotheses have been proposed to explain relationships between the major sponge classes. One groups the Porifera into the adelpho- taxa Hexactinellida and Demospongiae/Calcarea based on the gross difference in tissue structure and on differences in the structure of the flagellae, whose beating generates the feeding current through sponges (Mehl & Reiswig, 1991). The other hypothesis assumes that the Demospongiae are more closely related to Hexactinellida based on presumed larval similarities (Bóger, 1988). MOLECULAR APPROACH: PROTEIN KINASE C. In order to approach this question the cDNA encoding a protein kinase C, belonging to the C subfamily from the hexactinellid sponge R. dawsoni, has been isolated and characterised (Kruse et al., 1998). The two conserved regions, 391 the regulatory part with the pseudosubstrate site, the two zinc fingers and the C2 domain, as well as the catalytic domain were used for phylogenetic analyses. Sequence alignment and construction of a phylogenetic tree from the catalytic domains revealed that the hexactinellid R. dawsoni branches off first among the metazoan sequences; the other two classes, Calcarea (using the sequence from S. raphanus) and Demo- spongiae (using sequences from G. cydonium and S. domuncula) branch off later. The statistically robust tree also shows that the two cPKC sequences from the higher invertebrates D. melanogaster and Lytechinus pictus are most closely related to the calcareous sponge (Kruse et al., 1998) (Fig. 7A). This finding was also confirmed by comparing the regulatory part of the kinase gene. MOLECULAR APPROACH: 70KDA HEAT SHOCK PROTEIN. Previous analyses of the 70kDa heat shock protein isolated from the same sponge species (Koziol et al., 1997) justify the conclusion that: 1) within Porifera, the subphylum Hexactinellida diverged first from a common ancestor to the Calcarea and the Demo- spongiae, which both appeared later; and 2) the higher invertebrates are more closely related to the calcareous sponges (Miiller et al., 1998). ADDITIONAL SUPPORT FOR TWO SUBPHYLA: INSULIN-LIKE RECEPTORS. Further support came from the analysis of the autapomorphic character restricted to all the Metazoa including Porifera, the transmembrane receptor tyrosine kinases (RTKs). Recently, we screened for the presence of molecules grouped into one specific subfamily within the super- family of the RTKs (which includes the insulin receptors (InsR), the insulin-like growth factor 1 receptors and the InsR-related receptors), all found in vertebrates, as well as the InsR- homologue from D. melanogaster. The cDNAs, encoding the putative InsR-homologues, were isolated from the hexactinellid sponge A. vastus, the demosponge S. domuncula and the calcareous sponge $. raphanus (Skorokhod et al., submitted). Phylogenetic analyses ofthe catalytic domains of the putative RTKs showed that sponge poly- peptides have to be grouped to the putative InsR-homologues. Relationships revealed that all sponge sequences fall into one branch, while the related sequences from higher Metazoa, including the invertebrate sequences from insects and molluscs, or polypeptide(s) from one MEMOIRS OF THE QUEENSLAND MUSEUM PRC YEAST «4 YEAST Demospongiae 425MYA —* Hexactinellida 455MYA —*»- common metazoan ancestor FIG. 7. Evolutionary position of Hexactinellida, Demospongiae and Calcarea. A, Phylogenetic tree based on the alignment of the catalytic domains of the deduced PKCs from: 1) Metazoa - cPKCs from the deuterostomes Xenopus laevis (frog - cPKC XENLA) and Lytechinus pictus (sea urchin - PKC_LYTPI), from the protostomes cPKC Drosophila melanogaster (fruit fly - PKC DROME) and Aplysia californica (mollusc, PKC_APLA); 2) Sponges of the classes Demo- spongiae, Geodia cydonium (CPKC GC) and Suberites domuncula (CPKC_SD), Calcarea, Sycon raphanus (CPKC_SR), and Hexactinellida, Rhabdocalyptus dawsoni (CPKC_RD); 3) Yeast Saccharomyces cerevisiae (PKC_YEAST). B, Analysis of the insulin receptors (InsR), the insulin: like growth factor | receptors and the InsR-related receptors from vertebrates and invertebrates together with those from sponges. The deduced aa sequences of InsR homologues from the poly- peptides of the three classes of Porifera: 1) Demospongiae: S, domuncula (INR. SD); 2) Calcarea S. raphanus type 1 (INR SRI), S. raphanus type 2 (INR_SR2), S. raphanus type 3 (INR. SR3), and 3) Hexactinellida: Aphrocallistes vastus (INR. AV). All have been aligned with the related sequences for invertebrates: the insulin-like receptor precusor from the mosquito Aedes aegypti (INSR_AEDAE) and the InsR homologue from the fruit ly Drosophila melanogaster (INR_DROME) as well as the mollusc Aplysia californica InsR. (INSR_APLA), one cephalochordate: the insulin- like peptide receptor precursor from amphioxus Branchiostoma lanceolatum (ILPR_BRALA) and from selected vertebrates: the human insulin-like growth factor | receptor precursor (IGIR HUMAN), the human InsR precusor (INSR_ HUMAN), the InsR precursor from the house mouse Mus musculus (INSR MOUSE), the IGF-I-R I receptor precursor from the rat Rattus norvegicus (IGF RAT) and the InsR precusor from R. norvegicus (INSR. RAT). The RTK domain from the sponge G. cydonium (accession number X77528) was used for comparison. The rooted phylogenetic tree of the catalytic domains of these sequences is shown. C, Proposed branching order of the three classes of Porifera (Hexactinellida, Demo- spongiae and Calcarea), from a common metazoan ancestor. In addition, the separtion of the three types of the S. raphanus InsR-homologues, type 1 from type 3, and type | from type 2, are also indicated. The dates of the approximate divergence time are indicated (MY A). SPONGE MOLECULAR BIOLOGICAL STUDIES 393 VERTEBRATA INVERTEBRATA T M A e N a m [ e m p e m m h o a: a A i s t [ v b t o i e i e d Time scale a s a i a (Myrs) 0 200 400 600 CAMBRIAN EXPLOSION 4 800 mosaic proteins [| 1 I ! A autapomorphy: rec. tyrosine kinase y T i ! exon-shuffling genes for: 1 A tissue formation signal transduction transcription immune reaction ? collagen rec. tyrosine kinase homeodomain heat shock protein(s) | modules integrin Ser/Thr kinase MADS-box proteasome 4 A galectin SRCR-SCR i myosin Rhike 1 domains 1 sensory cells cell lineages — . 4,200 crystallin no determination ^ Glu-Rec FIG. 8. Phylogenetic relationship of Porifera within the animal groups based on molecular biological data, obtained from sequences of *metazoan' proteins required for tissue formation, signal transduction, transcription, immune reaction (potential) and sensory cells. The cell lineages in sponges are less determined than in higher Metazoa. Itis proposed that the Cambrian explosion of metazoan radiation became possible after the creation of the evolutionary mechanism of modularisation of distinct protein domains, thus allowing the formation of mosaic proteins by exon-shuffling; this process happened approximately 1,000MY A. It is indicated that the presence of the telomerase activity is one autapomorphic character of Porifera. cephalochordate and from selected vertebrates (human, mouse and rat) fall together into a second one. Full length clones have been isolated from S. raphanus, that in addition to having the characteristic signatures for InsR-homologues, have one complete and one incomplete calcium binding epidermal growth factor receptor (EGF)-like domain in the extracellular regions. Estimation of the rate of evolution of InsR- homologues revealed that the InsR-homologue ofthe hexactinellid sponge A. vastus is the phylo- genetically oldest one (455MYA), while the molecules from the demosponge S. domuncula (425MYA) and the calcareous sponge S. raphanus (390MYA) are phylogenetically younger (Fig. 7B-C). SPONGES AS LIVING FOSSILS. Based on the sequence data presented here it is reasonable to state that Porifera should be placed into the kingdom Animalia together with the (Eu)Metazoa (Miiller et al., 1994; Miiller, 1995, 1997a). In addition, from the analysis of these first sponge genes, especially the one coding for RTK, it is now established that modular proteins, composed by exon-shuffling, are common to all metazoan phyla; a detailed description is given elsewhere (Miiller & Miiller, 1997). This mechanism of exon-shuffling is apparently absent in plants and protists (Patthy, 1995). If this view can be accepted then the burst of ‘evolutionary creativity’ (Patthy, 1995) during the period of Cambrian explosion, which resulted in the big bang of metazoan radiation (Lipps et al., 1992), was driven by the process of modular- isation. During this process the existing domains were transformed into mobile modules, allowing the composition of mosaic proteins (Fig. 8). As an example, the Ig-like domain is not an invention of Metazoa. Molecules featuring Ig- like domains appeared early in eukaryotic evolution (e.g. they are present in yeast a-agglutinin cell wall-associated protein; Chen et al., 1995). However, their use as modules, as building blocks, for the creation of mosaic proteins only became possible after a new step in evolution was acquired which allowed exon- shuffling. The mechanism of modularisation is more universal and more versatile - it can be applied to all preexisting domains - than the process of forming new domains. Therefore, it can be assumed that during the transition from Protozoa to Metazoa, a process which lasted approximately 1,000 million years, the formation of domains with distinct folds was at the center of evolution. After having reached a critical number of domains the mechanism of modularisation allowed a rapid formation of a series of mosaic proteins by exon-shuffling. During the transition from unicellular Protista to multicellular Metazoa, the primary pattern of differentiation implies the presence of at least two different cell types, and as such the simplest multicellular organism could have consisted of one cell type specialised for feeding and the other for reproduction (Wolpert, 1990). This is in agreement with Roux-Weismann’s original concept of primary separation of somatic and germinal cell lineages (Weismann, 1892), in which the immortal germen produces a mortal soma that will sustain the growth and repro- duction of the organism but will necessarily perish. In view of the proposed monophyletic evolution of metazoans, and the position of MEMOIRS OF THE QUEENSLAND MUSEUM sponges at the base of the evolution of multi- cellularity, we have addressed the question for those molecular mechanisms which underlie the evolution of the germ cell- and somatic cell- lineages, and the potential control of their immortality or their programmed senescence and death. Sponges reproduce both asexually, by bud- and gemmule-formation, and sexually by production of gametes (reviewed in Simpson, 1984). But they lack special reproductive organs. The identi- fication of putative stem cells for primordial germ cells in sponges has not been clearly provided, and the compelling morphological evidence for the origin of gametes from the somatic fully differentiated cells, such as choan- ocytes, argues against the clear separation of the germinal and somatic cell lineages. Preliminary experimental evidence has now been presented which reveals that sponge tissue is rich in telo- merase activity, suggesting that the separation of cell lineages of somatic and germ stem cells has not been established, the determination of the fate of given sponge cells is still dynamic and might, under different physiological conditions, be reversible. It is proposed that the presence of telomerase activity is one autapomorphic character of Porifera (Fig. 8). CONCLUSION Our data show that sponges contain, as taken from deduced aa sequences, most structural elements known from higher Metazoa. It was intriguing to realise during the last three years that, while belonging to Metazoa, sponges 1) do have in some respect primitive, primordial metazoan characteristics, (e.g. simple elements of an immune system, with the ‘multiadhesive protein’ as an example), and 2) are already provided with complex and highly structurally evolved molecules not yet described from higher Metazoa such as the SRCR molecule. It is fortunate that sponges are not extinct. Assuming that Porifera were not the first metazoan phylum to evolve, they were witnesses to an evolutionary step that occurred during the maturation of the Metazoa near the Proterozoic-Phanerozoic boundary, close to 1 billion years ago. In this respect they can be considered to be living fossils. ACKNOWLEDGEMENTS This project was supported by the Deutsche Forschungsgemeinschaft (Mii 348/12-1) and the SPONGE MOLECULAR BIOLOGICAL STUDIES 395 International Human Frontier Science Program (RG-333/96-M). LITERATURE CITED AVENT, N.D., RIDGWELL, K., TANNER, M.J.A., ANSTEE, D.J. £ KUMPEL, B. 1990. cDNA cloning of a 30 kDa erythrocyte membrane protein associated with Rh (Rhesus)-blood- group-antigen expression. Biochemical! Journal 271: 821-825. BARTMANN-LINDHOLM, CH., GEISERT, M., GUNGERICH, U., MULLER, W.E.G. & WEINBLUM, D. 1997. Nuclear DNA fractions with grossly different base ratios in the genome of the marine sponge Geodia cydonium. Journal of Colloidal Polymer Science 107: 122-126. BISHOP, J.M. 1981. Enemies within: The genetics of retrovirus oncogenes. Cell 23: 5-6. BOGER, H. 1988. Versuch iiber das phylogenetische System der Porifera. Meyniana 40: 143-154. BOND. C. & HARRIS, A.H. 1988. Locomotion of sponges and its physical mechanism, Journal of Experimental Zoology 246: 271-284. BORK, P. & DOOLITTLE, R.F. 1992. Proposed acquisition of an animal protein domain by bacteria. Proceedings of the National Academy of Paes USA 89: 8990-8994, BOUTE, N., EXPOSITO, J.Y., BOURY-ESNAULT, N., VACELET, J., NORO, N., MIYAZAKI, K., YOSHIZATO, K. & GARRONE, R. 1996. Type IV collagen in sponges, the missing link in basement membrane ubiquity. Biology of the Cell 88: 37-44, BRETTING, H., PHILLIPS, S.G., KLUMPART, H.J. & KABAT, E.A. 1981, A mitogenic lactose- binding lectin from the sponge Geodia cydonium. Journal of Immunology 127: 1652-1658. CAVALIER-SMITH, T., ALLSOPP, M.T.E.P., CHAO, E.E., BOURY-ESNAULT, N. & VACELET, J. 1996. Sponge phylogeny, animal monophyly, and the origin of the nervous system: 18S rRNA evidence. Canadian Journal of Zoology 74: 2031-2045. CELERIN, M., RAY, J.M., SCHISLER, N.J., DAY, A.W., STETTER-STEVENSON, W.G. & LAUDENBACH, D.E. 1996. Fungal fimbriae are composed of collagen. EMBO Journal 15: 4445-4453. CHEN, M.H., SHEN, Z.M., BOBIN, S., KAHN, P.C. & LIPKE, P.N. 1995. Structure of Saccharomyces cerevisia a-agglutinin: Evidence for a yeast cell wall protein with multiple immunoglobulin-like domains with atypical disulfides. Journal of Biological Chemistry 270: 26168-26177. CHERIF-ZAHAR, B., BLOY, C., LE VAN KIM, C., BLANCHARD, D., BAILLY, P., HERMAND, P., SLAMON, C., CARTRON, J.P. & COLIN, Y. 1990. Molecular cloning and protein structure ofa human blood group Rh polypeptide. Proceedings of the National Academy of Sciences USA 87: 6243-6247. CUSTODIO, M.R., PROKIC, STEFFEN, R., KOZIOL, C., BOROJEVIC, E BRUMMER, F., NICKEL, M. & MULLER, W.E.G. 1998. Primmorphs generated from dissociated cells of the sponge Suberites domuncula: A model system for studies of cell proliferation and cell death. Mechanisms of Ageing & Development 105: 45-59, DIEHL-SEIFERT, B., KURELEC, B., ZAHN, R.K., DORN, A., JERECEVIC, B., UHLENBRUCK, G. & MULLER, W.E.G. 1985a. Attachment of sponge cells to collagen substrata: Effect of a collagen assembly factor. Journal of Cell Science 79: 271-285. DIEHL-SEIFERT, B., UHLENBRUCK, G., GEISERT, M., ZAHN, R.K. & MULLER, W.E.G. 1985b. Physicochemical and functional characterization of the polymerization process of the Geodia cydonium lectin. European Journal of Biochemistry 147: 517-523. FISHELSON, L. 1981. Observation on the moving colonies of the genus Tethya (Demospongiae, Porifera). Zoomorphology 98: 89-99, GAMULIN, V., SKOROKHOD, A., KAVSAN, V. MÜLLER, I.M. & MÜLLER, W.E.G. 1997. Experimental indication against blockwise evolution of metazoan protein molecules: Example, receptor tyrosine kinase gene from the sponge Geodia cydonium. Journal of Molecular Evolution 44: 242-252. GARRONE, R., HUC, A., & JUNQUA, S. 1975. Fine structure and physicochemical studies on the collagen of the marine sponge Chondrosia reniformis Nardo. Journal of Ultrastructural Research 52: 261-275. GRAMZOW, M., SCHRÖDER, H.C., FRITSCHE, U., KURELEC, B., ROBITZKI, A. ZIMMERMANN, H., FRIESE, K., KREUTER, M.H. & MULLER, W.E.G. 1989. Role of phospholipase A2 in the stimulation of sponge cell proliferation by the homologous lectin. Cell 59: 939-948. HARLEY, C.B., KIM, N.W., PROWSE, S.I WEINRICH, K.S., HIRSCH, M.D., WEST, M.D., BACCHETTI, S., HIRTE, H.W., COUNTER, C.M., GREIDER, C.W., WRIGHT, W.E. & SHAY, J.W. 1994, Telomerase, cell mortality, and cancer. Cold Spring Harbor Symposia on Quantitative Biology 59: 307-315. HARLEY, C.B. 1991. Telomere loss: Mitotic clock or genetic time bomb? Mutation Research 256: 271-282. HIRABAYASHI, J. & KASAI, K. 1993. The family of metazoan metal-independent P-galactoside- binding lectins: Structure, function and molecular evolution. Glycobiology 3: 297-304. IMSIECKE, G., CUSTODIO, M., BOROJEVIC, R., STEFFEN, R., MOUSTAFA, M.A. & MULLER, W.E.G. 1995. Genome size and chromosomes in marine sponges (Suberites domuncula, Geodia 396 eydonium). Cell Biology International 19: 995-1000. . KOZIOL, C., LEYS, S.P., MULLER, L.M. & MULLER, W.E.G. 1997. Cloning of Hsp70 genes from the marine sponges Sycon raphanus (Calcarea) and Rhabdocalyptus dawsoni (Hexact- inellida). An approach to solve the phylogeny of sponges. Biological Journal of the Linnean Society 62: 581-592. KOZIOL, C., BOROJEVIC, R., STEFFEN, R. & MULLER, W.E.G. 1998. Sponges (Porifera) model systems to study the shift from immortal- to senescent somatic cells: The telomerase activity in somatic cells. Mechanisms of Ageing and Development 100: 107-120. KRASKO, A., MULLER, LM. & MULLER, W.E.G. 1997, Evolutionary relationships of the metazoan By-crytallins, including the one from the marine sponge Geodia cydonium. Proceedings of the Royal Society London (B) 264: 1077-1084. KRUSE, M., GAMULIN, V., CETKOVIC, H., PANCER, Z., MULLER, LM. & MULLER, W.E.G. 1996. Molecular evolution of the metazoan protein kinase C multigene family. Journal of Molecular Evolution 43: 374-383. KRUSE, M., MULLER, ILM. & MULLER, W.E.G. 1997. Early evolution of metazoan serine/ threonine and tyrosine kinases: Identification of selected kinases in marine sponges. Molecular Biology and Evolution 14: 1326-1334. KRUSE, M., LEYS, S.P., MÜLLER, I.M. & MULLER, W.E.G. 1998. Phylogenetic position of the Hexactinellida within the phylum Porifera based on amino acid sequence of the protein kinase C from Rhabdocalyptus dawsoni. Journal of Molecular Evolution 46: 721-728. LE VAN KIM, C., MOURO, I., CHERIF-ZAHAR, B., RAYNAL, V. CHERRIER, C., CARTRON, J.P. & COLIN, Y. 1992. Molecular cloning and primary structure of the human blood group RhD poly- peptide. Proceedings ofthe National Academy of Sciences USA 89: 10925-10929. LI, W.H. & GRAUR, D. 1991. Fundamentals in Molecular Evolution (Sinnauer Associates: Sunderland Massachusetts). LIPPS, J.H. & SIGNOR, P.W. (eds) 1992. Origin and Early Evolution of Metazoa (Plenum Press: New York). MACKIE, G.O. & SINGLA, C.L. 1983. Studies on hexactinellid sponges. I, Histology of Rhabdocalyptus dawsoni (Lambe, 1873). Philo- sophical Transactions of the Royal Society of London (B) 301: 365-400. MARGULIS, L. & SCHWARTZ, K.V. 1995. Five Kingdoms: An Illustrated Guide to the Phyla of Life on Earth (Freeman & Co.: New York). MEHL, D. & REISWIG, H.M. 1991. The presence of flagellar vanes in choanomeres of Porifera and their possible phylogenetic implications. Zeitschrift für zoologische Systematik und Evolutions-Forschung 29: 312-319. MEMOIRS OF THE QUEENSLAND MUSEUM MORRIS, P.L. 1993. The developmental role of the extracellular matrix suggests a monophyletic origin of the kingdom Animalia. Evolution 47: . 152-165. MULLER, W.E.G. 1982. Cell membranes in sponges. International Review of Cytology 77: 129-181. 1995. Molecular phylogeny of Metazoa (animals): Monophyletic origin. Naturwissenschaften 82: 321-329. 1997a. Molecular phylogeny of Eumetazoa: Experimental evidence for monophyly of animals based on genes in sponges (Porifera). Progress in Molecular and Subcellular Biology 19: 98-132. 1997b. Origin of metazoan adhesion molecules and adhesion receptors as deduced from their cDNA analyses from the marine sponge Geodia .. cydonium. Cell & Tissue Research 289: 383-395. MULLER, W.E.G. & SCHACKE, H. 1996. Characterization of the receptor protein-tyrosine kinase gene from the marine sponge Geodia cydonium. Progress in Molecular and Subcellular . Biology 17: 183-208. MULLER, W.E.G. & MULLER, I.M. 1997, Transition from Protozoa to Metazoa: An experimental approach. Progress in Molecular and Subcellular Biology 19: 1-22. MÜLLER, W.E.G., MÜLLER, I.M. & GAMULIN, V. 1994, On the monophyletic evolution of the Metazoa. Brazilian Journal of Medical Biological . Research 27: 2083-2096. MULLER, W.E.G., BLUMBACH, B. WAGNER-HULSMANN, C. & LESSEL, U. 1997. Galectins in the phylogenetically oldest metazoa, the sponges (Porifera). Trends in Glycoscience and Glycotechnololgy 9: 123-130. MÜLLER, W.E.G., KRUSE, M., KOZIOL, C. & LEYS, S.P. 1998. Evolution of early Metazoa: Phylogenetic status of the Hexactinellida within the phylum of Porifera (sponges). Progress in .. Molecular and Subcellular Biology 21: 141-156. MULLER, W.E.G., WIENS, M., BATEL, R., STEFFEN, R., BOROJEVIC, R. & CUSTODIO, R.M. 1999, Establishment of a primary cell culture from a sponge: Primmorphs from Suberites domuncula: A review. Marine Ecology Progress Series 178: 205-219, NASH, R. & SHOJANIA, A.M. 1987. Hematological aspect of Rh deficiency syndrome: a case report and review of the literature. American Journal of Hematology 24: 267-275. PAHLER, S., BLUMBACH, B., MÜLLER, l. €: MÜLLER, W.E.G. 1998, A putative multi- adhesive basal lamina protein from the marine sponge Geodia cydonium: Cloning of the cDNA encoding a fibronectin-, an SRCR- as well as a complement control protein module. Journal of Experimental Zoology 282: 332-343. PANCER, Z., KRUSE, M., SCHACKE, H., SCHEFFER, U., STEFFEN, R., KOVÁCS, P. & MÜLLER, W.E.G. 1996. Polymorphism in the SPONGE MOLECULAR BIOLOGICAL STUDIES 397 immunoglobulin-like domains of the receptor tyrosine kinase from the sponge Geodia cydonium. Cell Adhesion & Communications 4: 327-339. PANCER, Z.. KRUSE, M., MÜLLER, I. & MÜLLER, W.E.G. 1997a. On the origin of adhesion receptors of Metazoa: Cloning of the integrin a subunit cDNA from the sponge Geodia cydonium. Molecular Biology and Evolution 14: 391-398. PANCER, Z., MÜNKNER, J., MÜLLER, I. & MÜLLER, W.E.G. 1997b. A novel member of an ancient superfamily: Sponge (Geodia cydonium, Porifera) putative protein that features scavenger receptor cysteine-rich repeats. Gene 193: 211-218. PATTHY, L. 1995. Protein Evolution by Exon-Shuffling (Springer-Verlag: New York). PEROVIC, S., PROKIC, I., KRASKO, A., MÜLLER, I.M. & MULLER. W.E.G. 1999, Origin of neuronal receptors in Metazoa: Cloning of a metabotropic glutamate/GABA-like receptor from the marine sponge Geodia cydonium. Cell & Tissue Research (in press). PFEIFER, K., HAASEMANN, M., GAMULIN, V., BRETTING, H.. FAHRENHOLZ, F. & MÜLLER, W.E.G. 1993, S-type lectins occur also in invertebrates: High conservation of the carbo- hydrate recognition domain in the lectin genes from the marine sponge Geodia cydonium. Glycobiology 3: 179-184. REID, K.B.M. & DAY, A.J. 1989. Structure-function relationship of the complement components. Immunology Today 10: 177-180. REISWIG, H.M. & MACKIE, G.O. 1983. Studies on hexactinellid sponges. III The taxonomic status of the Hexactinellida within the Porifera. Philosophical Transactions of the Royal Society London (B) 301: 419-428. RIDGWELL, K., SPURR, N.K., LAGUDA, B., MACGEOCH, C., AVENT, N.D. & TANNER, M.J.A. 1992. Isolation of cDNA clones for a 50 kDa glycoprotein of the human erythrocyte membrane associated with Rh (Rhesus) blood group antigen expression. Biochemical Journal 287: 223-228. RODRIGO, A.G., BERGQUIST, P.R., BERGQUIST, P.L. & REEVES, R.A. 1994. Are sponges animals? An investigation into the vagaries of phylogenetic interference. Pp. 47-54. In Soest, R.W.M. van, Kempen, T.M.G. van & Braekman, J.C. (eds) Sponges in Time and Space (Balkema: Rotterdam). N SEACK, J., PANCER, Z., MULLER, I.M. & MULLER, W.E.G. 1997. Molecular Cloning and primary structure of an Rh (Rhesus) antigen-like protein from the marine sponge Geodia cydonium. Immunogenetics 46: 493-498. SEACK, J., KRUSE, M. & MÜLLER, W.E.G. 1998. Evolutionary analysis ofG-proteins in early meta- zoans: Cloning of a- and f-subunits from the sponge Geodia cydonium. Biochimica et Biophysica Acta 1401: 93-103. SIMPSON, T.L. 1984. The Cell Biology of Sponges (Springer-Verlag: New York). SKOROKHOD, A., GAMULIN, V., GUNDACKER, D., KAVSAN, V., MÜLLER, I.M. & MULLER, WE.G. Origin of insulin receptor tyrosine kinase: Cloning of the cDNAs from marine Sponges. (Submitted). SMITH, L.C. & HILDEMANN, W.H. 1986. Allogeneic cell interactions during graft rejection in Callyspongia diffusa (Porifera; Demospongia); a study with monoclonal antibodies. Proceedings of the Royal Society London (B) 226: 465- de WAGNER-HULSMANN, C., BACHINSKI, DIEHL-SEIFERT, B., BLUMBACH, N STEFFEN, R., PANCER, Z. & MULLER, W.E.G. 1996. A galectin links the aggregation factor to cells in the sponge (Geodia cydonium) system. Glycobiology 6: 785-793. WEISMANN, A. 1892, Uber die Vererbung (Fischer: Jena). WIJNGAARD, P.L., METZELAAR, M.J., MACHUGH, N.D., MORRISON, W.I. & CLEVERS, H.C. 1992. Molecular charact- erization of the WCI antigen expressed specifically on bovine CD4-CD8-gamma delta T lymphocytes. Journal of Immunology 149: 3273-3277. WILLMER, P. 1994. Invertebrate Relationships (Cambridge University Press: Cambridge). WOLPERT, L. 1990, The evolution of the develop- ment. Biological Journal of the Linnean Society 39: 109-124. MEMOIRS OF THE QUEENSLAND MUSEUM AN ULTRASTRUCTURAL STUDY ON THE CONTRACTILE PINACOCYTE OF A FRESHWATER SPONGE. Memoirs of the Queensland Museum 44: 398. 1999:- Contractile cells in sponges were first observed in the oscular diaphragm of marine species, Microciona and Tedania (Bagby, 1961). They consisted of well-differentiated myocytes having myosin and actin filaments, whereas in the freshwater sponges, there are no reports of contractile apparatus or myocytes.The oscular diaphragms and body walls of freshwater sponges are contractible when stimulated. Stimulation is effective whether it is osmotic, thermal, electric, etc. According to these observations on sponges we hypothesise that pinacocytes in the oscular diaphragm and body wall must bear the contractile apparatus, and we suggested in a previous report that these cells have a network of filamentous bundles. We subsequently attempted to structurally identify the contractile filaments of the cell using the following methods: 1) Observations under light microscopy; 2) Observations under electron microscopy; 3) SDS-PAGE; 4) NBD-phallacidin staining; 5) Anti-actin gold conjugation,Pinacocytes in these sponges showed a flat and multiangular shape, measuring about Sum in diameter and 0.1,1m thick. Pinacocytes in the outer layer of the oscular diaphragms and body wall had many bundles extending radially from the central nuclear zone to the peripheral region of the cell, whereas these bundles were not observed in the inner pinacocytes. Bundles were easily stained using NBD-phallacidin. Observations of thin sections showed these bundles are composed of many thin filaments of about 4-6nm diameter, These bundles ran straight in the contracted state, and were distributed in the basal region of the cell. Thin filaments in the bundle were clearly decorated with anti-actin gold conjugation. SDS-PAGE analysis of the diaphragm revealed a protein band of 45kD. These results support the idea that thin filaments of about 4-6nm in diameter in pinacocytes are composed of actin molecules. A freshwater sponge, Ephydatia fluviatilis, has no myocytes but has contractile pinacocytes with actin bundles. O Porifera, freshwater sponges, contractile filaments, ultrastructure, actin, pinacocytes. Akira Matsuno (email: matsuno@life.shimane-u.ac.jp) & Masaaki Kuroda, Department of Biological Sciences, Faculty of Life and Environmental Science, Shimane University, Matsue 690, Japan; Yoshiki Masuda, Department of Biology, Kawasaki Medical School, Kurashiki City, Okayama 701-01, Japan; 1 June 1998. AN EVALUATION OF MORPHOLOGICAL AND CYTOLOGICAL DATA SETS FOR THE PHYLOGENY OF HOMOSCLEROPHORIDA (PORIFERA: DEMOSPONGIAE) GUILHERME MURICY Muricy, G. 1999 06 30: An evaluation of morphological and cytological data sets for the phylogeny of Homosclerophorida (Porifera: Demospongiae). Memoirs of the Queensland Museum 44: 399-409. Brisbane. ISSN 0079-8835. Congruence and resolution of two independent data sets (morphology and cytology) were compared in a phylogenetic analysis of ten Mediterranean species representing four genera of Homosclerophorida (Demospongiae). Two aspiculate genera (Oscarella, Pseudocorticium) and two with a siliceous skeleton (Plakina, Corticium) were studied to assess the validity of the traditional classification recognizing two families Plakinidae and Oscarellidae, respectively with and without a skeleton, with Discodermia polydiscus (Tetractinomorpha) used as an outgroup. Cytological data showed highest consistency but poor resolution, support, and average performance, probably due to low number of informative characters. It was the only data set that supported the monophyly of Oscarellidae. Morphological data had better congruence and support than cytological data, and indicated the non-monophyly of Oscarellidae. Combined analyses yielded a ‘total evidence’ tree topologically similar to that of morphological data, although with greater bootstrap support (average 64.7% per branch). The combined data set should be chosen for classification of Homosclerophorida because it includes all evidence available and showed the best general performance with the indices measured. Therefore, a classification including Oscarella and Pseudocorticium within Plakinidae is supported over the alternative, less parsimonious option which recognizes Oscarellidae as monophyletic. O Porifera, Homosclerophorida, phylogeny, morphology, cytology, W Mediterranean. Guilherme Muricy (e-mail: muricy@acd.ufrj.br), Departamento de Invertebrados, Museu Nacional, Universidade Federal do Rio de Janeiro. Quinta da Boa Vista, s/n°., São Cristóvdo. 20940-040 Rio de Janeiro, RJ, Brazil; 18 May 1999. Over the last few decades several new, non-morphological characters have been proposed for sponge taxonomy (e.g., embryological, cytological, biochemical and molecular data), in an effort to overcome long standing problems in traditional classification based solely on morphology. However, the relative merits of different types of characters have rarely been evaluated in a phylogenetic framework (but see Hajdu & Van Soest, 1996). The present study compares the phylogenetic information content of morphological and cytological data sets in the resolution of a particular phylogenetic problem in Porifera, viz. the monophyly of Oscarellidae Lendenfeld, 1887. Homoscleromorpha Lévi, 1973, with its single order Homosclerophorida Dendy, 1905, is widely accepted as a monophyletic group (e.g., Bergquist, 1978; Hartman, 1982; Van Soest, 1987; Diaz & Van Soest, 1994). Its synapomorphic traits include viviparous cinctoblastula larvae, a basement membrane lining the pinacoderm and choanoderm, and both exo- and endopinacocytes being flagellated (Boury-Esnault et al., 1984, 1995; Diaz & Van Soest, 1994). Its outgroup relationships are still debated; despite the fact that a relationship to Calcarea, Calcaronea, has been postulated based on shared tetractinal symmetry of spicules and possession of amphiblastula larvae (Van Soest, 1984; Grothe & Reitner, 1988), it is more generally accepted that these are homoplastic traits and Homosclerophorida is more closely related to Demospongiae, Tetractinomorpha, due to the shared presence of siliceous tetractinal calthrops spicules (Diaz & Van Soest, 1994). Homosclerophorida is usually divided into two families, Plakinidae Schulze, 1880 with five genera bearing tetractinal spicules and derivatives (Corticium Schimdt, 1862, Plakina Schulze, 1880, Plakinastrella Schulze, 1880, Plakortis Schulze, 1880, and Placinolopha Topsent, 1897; see Diaz & Van Soest, 1994), and the monotypical Oscarellidae with genus Oscarella Vosmaer, 1884 without a skeleton (Lévi, 1973; Bergquist, 1978; Hartman, 1982). The recently described genus Pseudocorticium 400 TABLE 1. Species of Homosclerophorida studied from Marseille, France. Discodermia polydiscus is an outgroup (Tetractinomorpha: MEMOIRS OF THE QUEENSLAND MUSEUM Corticium and Pseudocorticium). Discodermia polydiscus Du Bocage, Lithistida). 1869 (Tetractinomorpha: Lithistida: NT Specimens | Theonellidae), was used as outgroup _ Species Author Collection Depth (m) | Pee ined in all an alyses. Although Pseudocorticium | Boury-Esnault et Jarre Cave 15-20 25 Discodermia 1S part ofa problematic jarrei al., 1995 E = group, the lithisthid sponges (family Corticium | Schmidt, 1862 | La Vesse, Jarre 520 16 Theonellidae Lendenfeld), its close candelabrum ‘ is relationships with the Astrophorid cie oen (Schmidt, 1862) | ^ La Vesse 15-35 27 families Ancorinidae Schmidt and Geodiidae Gray have been Oscarella | (gehmidt, 1868) La Nien a RENS E demonstrated by analyses of both tniberculafa "Cave morphological data and DNA Oscarella Muricy et al., Jarre and 15-20 8 sequences (Kelly-Borges & microlobata 1996 Gameau Caves - Pomponi, 1994). The hypothesis Oscarella Muricy et al., Jarre Cave 1520 6 that ectosomal discotrienes of Viridis 1296 = Discodermia are homologous Plakina Schulze, 1880 | Grand Congloue 15 3 (although distantly related) to monolopha 7 ra . ea plakinid calthrops, both having trilopha Schulze, 1880 | — Jarre Cave 15-20 12 derived from a common tetractinal ; ancestor spicule, is implicit in the Plakina Muricy et al., ] y at rer dde endoumensis 1998 Endoume Cave 23 5 choice of this particular outgroup. i Cytological characters have been Plakina jani Muricy etal., Gameau Cave 15-20 15 y gioa ^ 1998 proposed as useful taxonomic tools Discodermia | DuBocage, | Gameau Cave 15-20 10 at the species level, both in Homo- polydiscus 1869 sclerophorida (Boury-Esnault et al., Boury-Esnault et al., 1995, although aspiculate, is more similar to Corticium than to Oscarella in its biochemical characters (allozyme patterns; Solé-Cava et al., 1992, as Corticium sp.) and in several morphological characters (presence of a cortex, leuconoid aquiferous system with diplodal chambers, cartilaginous consistency, perforated surface; Boury-Esnault et al., 1995). This led to a proposal to abandon the taxon Oscarellidae and to merge all homoscleromorph genera into a single family Plakinidae (Solé-Cava et al., 1992; Diaz & Van Soest, 1994; Boury- Esnault et al., 1995). The problem at the level of character evolution is whether the absence of skeleton in Oscarella and Pseudocorticium is a convergence, a plesiomorphy, or a synapo- morphy. Phylogenetic relationships among genera in Homosclerophorida remain unstudied, and therefore the validity of Oscarellidae, containing Oscarella and Pseudocorticium (both aspiculate), is questionable. This is currently a major phylogenetic problem in the classification of Homosclerophorida. In this study, the consistency and resolution of two data sets (morphological and cytological) were compared in a phylogenetic analysis of ten Mediterranean species in four genera of Homo- sclerophorida (viz., Oscarella, Plakina, 1984, 1992, 1995; Muricy et al., 1996, 1999) and in other sponge groups (e.g. Simpson, 1968; Pomponi, 1976; Vacelet & Donadey, 1987; Boury-Esnault et al., 1994). The present study aimed to compare the congruence and resolution of phylogenetic estimates derived from morphological and cyto- logical data sets at the genus and family levels in Homosclerophorida. The two individual and combined data sets were also compared as to their support for the monophyly of Oscarellidae. Alternative hypotheses on phylogenetic relationships between the four genera of Homo- sclerophorida studied are discussed. MATERIALS AND METHODS Samples of ten homosclerophorid species and one outgroup species (Table 1) were collected using SCUBA between 1984-1995 from caves and vertical walls at 10-30m depth along the coast of Provence, Western Mediterranean, France (5?20'W, 43°10’N). For the morphological study, samples were fixed in formalin or glutaraldehyde/ sodium cacodylate, dehydrated in an alcohol series and included in Araldite. Thick and semi-thin sections were stained with toluidine blue and studied under light microscopy (LM). Spicules were observed under LM or scanning electron microscopy PHYLOGENY OF HOMOSCLEROPHORIDA (SEM) after sputter-coating with gold- palladium. Complete descriptions of each species studied and their important morphological and cytological characters are given elsewhere (Schulze, 1880; Ridley & Dendy, 1887; Topsent, 1895; Boury-Esnault et al., 1984, 1992, 1995; Muricy et al., 1996, 1998, 1999), Vouchers were deposited at the sponge collections of the Muséum National d'Histoire Naturelle, Paris (MNHN) and the Museu Nacional of the Universidade Federal do Rio de Janeiro, Brazil (UFRJPOR). For the cytological study, specimens were fixed as described by Boury-Esnault et al. (1984): glutaraldehyde 2.5% in a mixture of 0.4M cacodylate buffer and sea water (4 vol.: 5 vol.; 1120mOsm) and postfixed in 2% osmium tetroxide in sea water. Thin sections, contrasted with uranyl acetate and lead citrate, were observed under TEM in a Hitachi Hu 600 microscope. Cytological data on Discodermia polydiscus are based on the author’s unpublished observations and those of N. Boury-Esnault (pers. comm., 1993). All parsimony analyses were carried out using PAUP 3.0 (Swofford, 1991), In all searches, the following options were used: algorithm = branch and bound; keep minimal trees only; addition sequence furthest; collapse zero-length branches; multistate taxa = polymorphisms; optimization = ACCTRAN; rooting by outgroup (=Discodermia polydiscus); all characters unordered, with equal weights. Dependence between characters was checked by searching the data matrix for characters with identical distrib- ution of states among taxa (Appendices 1,2). Although some characters were excluded by this method (e.g. presence/absence of spicules and of an ectosomal skeleton), total similarity in distribution of several characters was considered fortuitous (i.e. not artifacts of character coding), and these traits were thus kept in the phylogenetic analyses (characters 7, 18; 13, 15, 32, 33, 34; 16, 17; 19, 20; 21, 22; 23, 24; 38, 40; 41, 42). Presence/ absence of spicules was excluded from the analyses because it was a key-character in the hypotheses tested, and to avoid redundancy with characters 19-26 (spicule types). Its evolution was traced a posteriori using MacClade 3.0 (Maddison & Maddison, 1992), as well as that of all other unambiguous character-state changes. One thousand minimal random trees (MRTs) were generated and their length distribution was compared to the most parsimonious trees obtained for the real data. The observed length of the most parsimonious trees never exceeded that 401 of the shortest randomized trees, and therefore it was accepted that the original data differed significantly (P<0.001) from randomness in all data sets. The following information was recorded for each individual and combined data sets: number of characters, number of informative characters, number of most parsim- onious trees (MPTs), length of MPTs, number of extra steps (e), consistency index (CI), homo- plasy index (HI), rescaled consistency index (RC), difference in steps between MRTs and MPTs, majority-rule consensus tree, number of branches involved in polytomies, bootstrap support for each branch, and unambiguous character-state changes supporting each branch (Swofford, 1991). These parameters measure either the internal consistency (congruence) of each data matrix, the resolution of phylogenetic estimates, or the support for the phylogenies derived from the different data sets. Bootstrap support for branches in consensus trees (Felsenstein, 1985; Li & Zharkikh, 1994) was calculated by 100 replicated branch and bound searches, with the same options held constant as described above. The number of extra steps in each analysis was used to measure incongruence due to disparity within and between data sets, analogous to analysis of variance (Mikevich & Farris, 1981; Kluge, 1989). Individual and combined data sets were ranked according to their relative performances on average for all indices measured, with the exception of extra steps and rescaled CT, which were redundant with HI and CI, respectively. RESULTS Of 30 morphological characters, 24 (8094) were phylogenetically informative for parsimony analysis, which resulted in three MPTs 67 steps long, with a fully-resolved majority-rule consensus tree (Fig. 1). MPTs of morphological data set required 14 extra steps, with intermediate consistency (CI=0.79, RC=0.61), and relatively high homoplasy (HI=0.25; Table 2). The majority-rule consensus tree (Fig. 1) supported monophyly of both Plakina and Oscarella (bootstrap=82-89%), but suggested non- monophyly of Oscarellidae. The alternative clade [Plakina + Oscarella] is supported by relatively low bootstrap (49%), but also by seven synapo- morphies: small lobes, surface wrinkled, soft consistency, sylleibid canal system, eurypylous choanocyte chambers, ectosome between 5- 50mm thick, and proportion of mesohyl to chambers less than 1:1 (Fig. 1). Corticium TABLE 2. Summary of information homoplasy index; RC, rescaled consistency index. on Homosclerophorida congruence. Abbreviations: MPTs, most parsimonious trees; MRTs, minimal random trees; e, number of extra steps; CI, consistency index; HI, MEMOIRS OF THE QUEENSLAND MUSEUM candelabrum is placed as sister-group of the clade [Plakina + Oscarella], with Pseudocorticium jarrei as the most basal ingroup branch. Complete loss of spicules appeared convergently in Oscarella and Pseudocorticium. Bootstrap support per branch was intermediate among all analyses (average 62.2%). Morphological data were more informative for skeleton-bearing than for aspiculate species, as expected, since skeletal characters comprised 13 out of 30 morphological characters included in the analysis. Parsimony analysis of cytological data yielded seven MPTs, of length 25 (4 extra steps). The cytological data set showed the smallest amount of homoplasy among all analyses (CI=0.84, RC=0.65, HI=0.16), but also the smallest number of informative characters (9 of 17; Table 2). All MPTs and their consensus (Fig. 2) supported the mono- phyly of Oscarellidae (with 54% of bootstrap support), with both genera Oscarella and Information Morphology Cytology Combined Character numbers 1-30 31-47 1-47 No. of characters 30 17 47 No. of informative 24 9 33 characters y Í No. of MPTs (length) 3 (67) 7 (25) 1 (95) CI 0.79 0.84 0.78 RC 0.61 0.65 0.57 HI 0.25 0.16 0.25 e 14 4 21 No. of MRTs (length) 1 (80) 1 (28) 4(115) MRTs minus MPTs 13 3 20 No. of branches in 0 x 0 polytomies E No. of unambiguous 25 " 32 changes "i a Mean bootstrap per branch (%) 62.2 40.7 64.7 E i “ 2 2 B = E Bg Bof Bo. 3 3 8 $ got ob FS; t 3 5 3 9 à 3 8 Y $ S x E S S S 2 9 E EJ S E E a = o E 2 2 2 S E 8 E go ao o o 9 9 p S 3 8 3 S $ $ £ 5 B 85 b £ 3 g 85 8 y B Z B B B S 8 2 a a ü a Ó [9] Q Q O à a FIG. 1. Majority-rule consensus of 3 most parsimonious trees of the morphological data set. Numbers beside branches are bootstrap support for the branch, and the percentage of MPTs displaying the branch (within parenthesis). Synapomorphies (within circles) are indicated as ‘character.state’: A = 4.1; B = 5.1, 6.1, 10.4, 11.1, 12.1, 16.2, 17.2; C 2 19.1, 20.1, 27.2; D= 23.1, 24.1, 25.1; E=1.1, 6.0, 7.1, 10.3, 14.1, 16.1, 17.1, 18.1. Homoplasies (within rectangles) are: F = 5.2; G = 2.1, 4.2; S = absence of spicules. For the names of characters and states see Appendices 1-2. Plakina indicated as paraphyletic, and Corticium placed in the most basal ingroup branch. Absence of spicules appeared as a synapomorphy for Oscarella and Pseudocorticium. Support to all resolved branches was generally low (average bootstrap 40.7%, 0-2 synapomorphies per branch). This was the only data set that suggested monophyly of Oscarellidae, including Pseudocorticium jarrei within the genus Oscarella as sister-species of O. microlobata. The clade [Oscarella + Pseudocorticium] is supported by synapomorphies 31.1 (triangular apopylar cells) and 44.0 (absence of sclerocytes). Cytological characters allowed better resolution and stronger support to the aspiculate homosclerophorid species than to the skeleton-bearing Plakina and Corticium. The combined data set provided 33 informative characters (70,2%), which resulted in one MPT, 95 steps long (21 extra steps). As the combined data included all the homoplasy (incongruence) hypothesized within each data set and between data sets, consistency was slightly lower than that of any individual analyses (CI=0.78; RC=0.57; HI=0.25; Table 2). The single MPT (Fig. 3) supported the mono- phyly of both Plakina and Oscarella, with significant bootstrap (72-87%). It also supported the non-monophyly of the Oscarellidae, showing Corticium as the sister group of a clade [Plakina + Oscarella] which was supported by PHYLOGENY OF HOMOSCLEROPHORIDA TABLE 3. Rankings of data sets according to their performance in each index measured (data from Table 2) and on average. Abbreviations as for Table 2. Combined No, of characters 2 3 1 Index Morphology Cytology No. of informative characters No. of MPTs CI HI MRTs minus MPTs No. of branches in 3 1 polytomies 2 3 1 3 1 1 3 to [ty [la j ra No. of unambiguous changes ba w Mean bootstrap per branch (%) Sum of rankings 17 23 12 ba es) = Average ranking 2 3 1 arella viridis Corticium candelabrum Pseudocorticium farrei Oscarella microlobata Oscarella tuberculata Oscarella lobularis Plakina jani Plakina trilopha. Plakina monolopha Discoderrnia polydiscus Osc E 2 [s] E E c © o S x 2 a FIG. 2. Majority-rule consensus of 7 most parsi- monious trees of the cytological data set. Key as for Figure |. Synapomorphies (within circles) are indicated as ‘character.state’: A = 31.1, 44.0; B = 38.1, 40.1; S = absence of spicules. For the names of characters and states see Appendices 1-2. 66% of bootstrap and by 8 synapomorphies (small lobes, wrinkled surface, soft consistency, sylleibid canal system, eurypylous choanocyte chambers, ectosome between 5-50mm thick, proportion of mesohyl to chambers less than 1:1, and collencytes absent), Absence of spicules appeared homoplastic in Oscarella and 403 Pseudocorticium, which was placed as the most basal ingroup branch. Although the topologies suggested by the two data sets analyzed in this study were hetero- geneous, their combination slightly increased the repeatability (as measured by the mean bootstrap support per branch) of the phylogenetic estimate when compared to the individual analyses. In general, the results ofthe combined data set were very similar in topology, resolution and support to those of the morphological data set, differing only in the relationships within Oscarella. The number of extra steps in each individual and combined analyses provided a direct measure of the incongruence within and between data sets. Of a total of 21 extra steps required by the most parsimonious trees for the combined data set, 86% (14-4) were due to incompatibility between characters within each data set, leaving 3 extra steps (14%) which reflect the incongruence between the two data sets (imp Mikevich & Farris, 1981). Incongruence between sets was lower than within sets (16-20.1% in each of the individual analyses). A ranking of the average performance of individual and combined data sets (Table 3) was calculated based on their performance in each index measured (Table 2). The combined data set ranked best, followed closely by the morphological data set, and more distantly by the cytological data set. DISCUSSION Each of the two independent and combined data sets analyzed suggested a different phylogenetic reconstruction for Homosclero- phorida, with varying degrees of resolution and support. The morphological data set showed better general performance when compared to the cytological data set (Tables 2, 3). Morphological data (particularly skeletal characters) are traditionally the major source of phylogenetic information for sponge taxonomists. Skeletal morphology provides valuable information in groups with high spicule diversity, such as Astro- phorida and Poecilosclerida (e.g. Maldonado, 1993; Hajdu & Van Soest, 1996), but is of little use in skeleton-lacking species or in groups with low skeletal diversity such as Haplosclerida (e.g. Boury-Esnault et al., 1995; Muricy et al., 1996; Vacelet & Donadey, 1987; De Weerdt, 1989). Non-skeletal morphological characters (e.g., shape, surface, aquiferous system) comprised 17 out of 30 morphological characters studied, and 14 were informative for phylogenetic analysis. 404 Plakina jani (T) Plakina trilopha Plakina endoumensis Plakina monolopha Oscarelia tuberculata Oscarella lobularis Oscarella viridis Oscarella microlobata Corticium condelabrurn Pseudocorticiurm jarrel Discoderrnia polydiscus FIG. 3. Most parsimonious tree of the combined data set. Numbers beside branches are bootstrap support for each branch, Synapomorphies (within circles) are indicated as ‘character.state’: A = 4.1; B = 5.1, 6.1, 10.4, 11.1, 12.1, 16.2, 17.2, 45.0; C 2 19.1, 20.1, 27. 2; D 7 23.1, 24.1, 25.1; E7 1.1, 5.2, 6.0, 7.1, 10.3, 14.1, 16.1, 17.1, 18.1, 47.1; F = 44.0; G = 46.0; H = 5.2. Homoplasies (within rectangles) are: I 72.1, 4.2; $= absence of spicules, For the names of characters and states see Appendix 1. Such non-skeletal characters, particularly the anatomy of the aquiferous system, can also be useful for the taxonomy of Homosclerophorida and other sponge taxa (e.g. Langenbruch, 1991; Bavestrello et al., 1995), and should be taken into account more often in sponge taxonomy. The cytological data set showed poorest resolution, support, and general performance, although it had the highest consistency indices of all data sets. This was probably due to the low number of informative cytological characters. Cytology was more informative in Oscarella and Pseudocorticium than in Plakina and Corticium. This was expected since, in Homosclerophorida, the aspiculate species show greater diversity of cells with inclusions than the skeleton-bearing species (Boury-Esnault et al., 1992, 1995; Muricy et al., 1996, 1999), Cytological data also have proved useful for the systematics of other sponge groups such as Poecilosclerida, Haplo- sclerida and Hadromerida (e.g., Simpson, 1968; Pomponi, 1976; Boury-Esnault et al., 1994), and should be more frequently used in phylogenetic studies of all Porifera. However, phylogenetic interpretation of sponge cytology is presently difficult due to the scarcity of information MEMOIRS OF THE QUEENSLAND MUSEUM available on cell function and chemistry, partic- ularly of the so-called ‘cells with inclusions’ (Simpson, 1984). An increase in number of functional and cytochemical studies of sponge cells would greatly add phylogenetic information to cytological data. How to choose the ‘best’ data set upon which classification should be based? One approach is to follow the principle of maximum evidence (e.g. Kluge, 1989), according to which the results of the combined analysis should be taken as a basis for classification of Homosclerophorida, since it was built with the maximum amount of independent evidence available. This approach has only been criticized when high between-set incongruence is found, because it may increase phylogenetic ‘noise’ and therefore reduce the accuracy of the combined analysis (Shaffer et al., 1991; Chippindale & Wiens, 1994; Wiens & Chippindale, 1994). Incongruence between data sets in Homosclerophorida (1,:70.14) is relatively low when compared to that found in other organisms (Shaffer et al., 1991; Olmstead & Sweere, 1994; Omland, 1994, and references therein). Therefore, it seems advisable in this case to base the classification on a combined data set. Another logical approach is to rank the data sets by their relative performances in each index of congruence, resolution and support recorded, and the data set with the best average perform- ance would be chosen for classification. This approach is shown in Table 3, in which the sets were ranked according to data in Table 2. Itshows that, on average, the combined data set performed better than any individual data sets. This approach also supports using the combined data set as a basis for classifying the Homo- sclerophorida. The topology derived from the combined data set supported monophyly of Oscarella and Plakina and non-monophyly of Oscarellidae, with absence of spicules appearing as a homo- plastic character in Oscarella and Pseudocorticium. Therefore, with the currently avallable data, a classification of Homosclero- phorida with a single family Plakinidae Schulze (including Oscarellidae Lendenfeld and Corticiidae Vosmaer), as recently adopted by Solé-Cava et al. (1992), Diaz & Van Soest (1994), and Boury-Esnault et al. (1995), is preferred over the traditional classification with two families Plakinidae and Oscarellidae (Lévi, 1973; Bergquist, 1978; Hartman, 1982). The analysis of independent data sets PHYLOGENY OF HOMOSCLEROPHORIDA obviously help to increase the amount of indep- endent evidence upon which the phylogenetic hypotheses are based, and to reveal conflicts between data that can have strong influence in the choice of a classification. Furthermore, comparison of different data sets allows the determination of a ‘phylogenetic confidence interval’ for the phylogenetic hypothesis, which includes the topologies suggested by the individual and combined data sets (Bull et al., 1993; Huelsenbeck et al., 1994). Incongruence between data sets suggests that probably none of them has given the exact answer, and this should caution against changes in classification in face of new, conflicting data, before the congruence of the new evidence is more carefully evaluated. Therefore, the monophyly of Oscarellidae suggested by old classifications and by the cytological data set, although with relatively low support, cannot be completely ruled out at the current state of knowledge on biology of Homo- sclerophorida. It is kept as an alternative, less parsimonious hypothesis of phylogeny, together with the possibility of paraphyly of both Oscarella and Plakina suggested by cytological data, which should also be taken into account in studies on phylogeny of Homosclerophorida. ACKNOWLEDGEMENTS I am most grateful to M. Carpine (Institut Océanographique, Monaco), K. Smith (National Museum of Natural History, Washington), C. Valentine (Natural History Museum, London), R.W.M. van Soest (Zoological Museum, Amsterdam), and C. Lévi (Muséum National d'Histoire Naturelle, Paris) for the kind loans of specimens for taxonomy, and to C. Jalong for useful diving assistance. N. Boury-Esnault and J. Vacelet (Station Marine d’Endoume, Marseille) kindly assisted in morphology and cytology assessments, which were made at the Centre d’ Océanologie de Marseille, and also provided useful discussions. Critical reading by E. Hajdu (Museu Nacional, Rio de Janeiro), S. Zea (Universidad Nacional de Colombia, Santa Marta) and J.N.A. Hooper (Queensland Museum, Brisbane) greatly improved the manu- script. The author has a fellowship from CNPq. This work was partly supported by CEE MAST-2CT91-004 program, CNPq, FAPERJ and FUJB/UFRJ. LITERATURE CITED BAVESTRELLO, G., BURLANDO, B. & SARA, M. 1995. Corrosion cast reconstruction of the 405 three-dimensional architecture of Demosponge canal systems. Pp. 93-100. In Lanzavecchia, R., Valvassori, R. & Candia Carnevali, M.D. (eds) Body cavities: function and phylogeny. Selected Symposia and Monographs, U.Z.I. Vol. 8 (Mucchi: Modena). BERGQUIST, P.R. 1978, Sponges. (Hutchinson & Co.: London). . BOCAGE, J.V.B. DU 1869. Eponges siliceuses nouvelles de Portugal et de l'Ile Saint-lago (archipel du Cap Vert). Journal des Sciences Mathématiques, Physiques et Naturelles de l'Académie des Sciences de Lisbonne 2: 159-162. BOURY-ESNAULT, N., DE VOS, L., DONADEY, C. & VACELET, J. 1984. Comparative study of the choanosome of Porifera: I. The Homo- scleromorpha. Journal of Morphology 180: 3-17. BOURY-ESNAULT, N., SOLE-CAVA, A.M. & THORPE, J.P. 1992. Genetic and cytological divergence between colour morphs of the Med- iterranean sponge Oscarella lobularis (Schmidt) (Porifera, Demospongiae, Oscarellidae). Journal of Natural History 26: 271-284. BOURY-ESNAULT, N., HAJDU, E., KLAUTAU, M., CUSTODIO, M. & BOROJEVIC, R. 1994. The value of cytological criteria in distinguishing sponges at the species level: example of the genus Polymastia. Canadian Journal of Zoology 72: 795-804. BOURY-ESNAULT, N., MURICY, G., GALLISSIAN, M.F. & VACELET. J. 1995. Sponges without skeleton: a new Mediterranean genus of Homo- scleromorpha (Porifera, Demospongiae). Ophelia 43: 25-43. BULL, J.J., HUELSENBECK, J.P., CUNNINGHAM, C.W., SWOFFORD, D.L. & WADDELL, P.J. 1993, Partitioning and combining data in phylo- genetic analysis, Systematic Biology 42: 384-397. CHIPPINDALE, P.T. & WIENS, J.J. 1994. Weighting, partitioning, and combining characters in phylo- genetic analysis. Systematic Biology 43: 278-287. DENDY, A. 1905. Report on the sponges collected by Professor Herdman, at Ceylon, in 1902. Ceylon Pearl Oyster Fisheries, Supplementary Reports 18: 1-246. DIAZ, M. C. & SOEST, R. W. M. VAN 1994, The Plakinidae: a systematic review. Pp. 93-110. In Soest, R. W. M. van, Kempen, Th.M.G. van & Braekman, J.C. (eds) Sponges in time and space. (Balkema: Rotterdam). FELSENSTEIN, J. 1985. Confidence limits in phylogeny: an approach using bootstrap. Evolution 39: 783-791. GROTHE, F. & REITNER, J. 1988. Discussion on the phylogenetic position of the Homoscleromorpha (Porifera). Berliner Geowissenschaften Abhandlungen, Reihe A, Geologischen und Palaeontologischen 100: 16-17. HAJDU, E. & SOEST, R.W.M. VAN 1996. Choosing among Poriferan morphological characters within the cladistic paradigm. Bulletin de I’ Institut Royal 406 des Sciences Naturelles de Belgique, Biologie 66 (supplement): 81-88. HARTMAN, W.D. 1982. Porifera. Pp. 640-666. In Parker, S.P. (ed.) Synopsis and classification of living organisms. Vol. 1. (McGraw-Hill: New York). HUELSENBECK, J.P., SWOFFORD, D.L., CUNNINGHAM, C.W., BULL, J.J. & WADDELL, P.J. 1994, Is character weighting a panacea for the problem of data heterogeneity in phylogenetic analysis? Systematic Biology 43: 288-291. KELLY-BORGES, M. & POMPONI, S.A. 1994, Phylogeny and classification of lithistid sponges (Porifera: Demospongiae): a preliminary assess- ment using ribosomal DNA sequence comparisons. Molecular Marine Biology and Biotechnology 3: 87-103. KLUGE, A.G. 1989. A concern for evidence and a phylogenetic hypothesis of relationships among Epicrates (Boidae, Serpentes). Systematic Zoology 38: 7-25. LANGENBRUCH, P.-F. 1991, Histological indications of the phylogenesis of the Haplosclerida (Demo- spongiae, Porifera). Pp. 289-298. In Reitner, J. & Keupp, H. (eds) Fossil and Recent Sponges. (Springer-Verlag: Berlin, Heildelberg). LENDENFELD, R. VON 1887. On the systematic position and classification of sponges. Proceedings of the Zoological Society London _ (1887): 558-662. LEVI, C. 1973. Systématique de la classe des Demo- spongiaria (démosponges). Pp. 577-631. In Brien, P., Lévi, C., Sara, M., Tuzet, O. & Vacelet, J. (eds) Traité de Zoologie. Anatomie, Systématique, Biologie. (Sér. Ed. P-P. Grassé). Vol. 3. Spongiaires. (Masson et Cie.: Paris). LI, W-H. & ZHARKIKH, A. 1994, What is the boot- strap technique? Systematic Biology 43: 424-430. MADDISON, W.P. & MADDISON, D.R. 1992. MacClade: analysis of phylogeny and character evolution. Version 3.0. (Sinauer Associates: Sunderland, Mass.). MALDONADO, M. 1993. The taxonomic significance of the short-shafted mesotriaene reviewed by parsimony analysis: validation of Pachastrella ovisternata Von Lendenfeld (Demospongiae: Astrophorida). Bijdragen tot de Dierkunde 63(3): 129-148. MIKEVICH, M.F. & FARRIS, J.S. 1981. The implic- ations of congruence in Menidia. Systematic Zoology 30: 351-370. ; MURICY, G., BOURY-ESNAULT, N., BEZAC, C. & VACELET, J. 1996. Cytological evidence for cryptic speciation in Mediterranean Oscarella (Porifera, Homoscleromorpha). Canadian Journal of Zoology 74: 881-896. MURICY, G., BOURY-ESNAULT, N., BÉZAC, C. & VACELET, J, 1998. A taxonomic revision of the Mediterranean Plakina Schulze (Porifera, Demo- MEMOIRS OF THE QUEENSLAND MUSEUM spongiae, Homoscleromorpha). Zoological Journal of the Linnean Society 124(2): 169-203. MURICY, G., BOURY-ESNAULT, N., GALLISSIAN, M-F. & VACELET, J. 1999, Anatomy, cytology and symbiotic bacteria of four Mediterranean species of Plakina Schulze, 1880 (Porifera, Homosclerophorida). Journal of Natural History 33: 159-176. OLMSTEAD, R.G. & SWEERE, J.A. 1994. Combining data in phylogenetic systematics : an empirical approach using three molecular data sets in the Solanaceae. Systematic Biology 43: 467-481. OMLAND, K.E. 1994. Character congruence between a molecular and a morphological phylogeny for dabbling ducks (Anas). Systematic Biology 43: 369-386. POMPONI, S.A. 1976. A cytological study of the Haliclonidae and the Callyspongiidae (Porifera, Demospongiae, Haplosclerida). Pp. 215-235. In Harrison, F.W. & Cowden, R.R. (eds) Aspects of Sponge Biology. (Academic Press: New York). RIDLEY, S.O. & DENDY, A. 1887. Report on the Monaxonida collected by the H.M.S. ‘Challenger’ during the years 1873-1876. Reports on the Scientific Results of the Voyage of the H.M.S. ‘Challenger’ 1873-1876, Zoology 20: 1-275. (Her Majesty’s Stationery Office: London, Edinburgh, Dublin). SCHMIDT, O. 1862. Die Spongien des Adriatischen Meeres. (Wilhelm Engelmann: Leipzig). 1868. Die Spongien der Küste von Algier. Mit Nachtrágen zu den Spongien der Adriatischen Meeres (drittes Supplement). (Wilhelm Engelmann: Leipzig). SCHULZE, F. E. 1880. Untersuchungen über den Bau und die Entwicklung der Spongien. Die Plakiniden, Zeitschrift für wissenschaftliche Zoologie 34: 407-451. SHAFFER, H.B., CLARK, J.M. & KRAUS. F. 199]. When molecules and morphology clash: a phylo- genetic analysis of the north American ambystomatid salamanders (Caudata: Ambyst- omatidae). Systematic Zoology 40: 284-303. SIMPSON, T.L. 1968. The structure and function of sponge cells: new criteria for the taxonomy of poecilosclerid sponges (Demospongiae). Bulletin ofthe Peaboby Museum of Natural History, Yale University 25: 1-141. 1984. The cell biology of sponges. (Springer Verlag: New York). SOEST, R.W.M. VAN 1984. Deficient Merlia normani Kirkpatrick, 1908, from the Curacao reefs, with a discussion on the phylogenetic interpretation of Demosponges. Bijdragen tot de Dierkunde 54(2): 211-219. 1987. Phylogenetic exercises with monophyletic groups of sponges. Pp. 227-241. In Vacelet, J. & Boury-Esnault, N. (eds) Taxonomy of Porifera from the N.E. Atlantic and Mediterranean Sea. PHYLOGENY OF HOMOSCLEROPHORIDA NATO ASI Series, ser. G, Vol. 13. (Springer- Verlag: Berlin). SOLE-CAVA, A.M., BOURY-ESNAULT, N., VACELET, J. & THORPE, J.P. 1992. Biochemical genetic divergence and systematics in sponges of the genera Corticium and Oscarella (Demospongiae: Homoscleromorpha) in the Mediterranean Sea. Marine Biology 113: 299-304. SWOFFORD, D.L. 1991. PAUP: Phylogenetic analysis using parsimony, Version 3.0. (Illinois Natural History Survey: Champaign, Ill.). TOPSENT, E. 1895. Étude monographique des Spongiaires de France. II.- Carnosa. Archives de Zoologie Expérimentale et Génerale 3: 493-590. 1897. Spongiaires de la Baie d'Amboine. Revue Suisse de Zoologie 4: 421-487. 407 VACELET, J. & DONADEY, C. 1987. A new species of Halisarca (Porifera, Demospongiae) from the Caribbean, with remarks on the cytology and affinities of the genus. Pp. 5-12. In Jones, W.C. (ed.) European Contributions to the Taxonomy of Sponges. (Sherkin Island Marine Station Publications: Sherkin Island, Ireland). VOSMAER, G.C.J. 1884. Porifera. Pp. 65-176. In Bronn, H.G. (ed) Die klassen und ordnung des Thier-reichs. (Winter'sche Verlagshandlung: Leipzig). WEERDT, W.H. DE 1989. Phylogeny and vicariance biogeography of north-Atlantic Chalinidae (Haplosclerida, Demospongiae). Beaufortia 39(3): 55-88. WIENS, J.J. & CHIPPINDALE, P.T. 1994. Combining and weighting characters and the prior agreement approach revisited. Systematic Biology 43: 564-556. APPENDIX 1. Morphological and cytological characters and character-states used in phylogenetic analysis. Uninformative characters are marked with an asterisk (*). MORPHOLOGICAL CHARACTERS. . Surface: 0-smooth; 1-wrinkled; 2-perforated. oc un kW r2 8*. Oscula: 0-low; 1-with elevated rim. 11. Canal system: 0-leuconoid; 1-sylleibid. 1. Shape: 0-spherical; 1-thin spreading; 2-lobate/spreading; 3-small circular; 4-cushion; 5-lobate/pending. . Size (surface cover in cm2): 0-small (<10); 1-large (710). . Thickness (mm): 0-thin (1-5); 1-medium (5-20); 2-thick (20-50). . Fixation: 0-firm; 1- by thin filaments; 2-loose but without filaments. . Size of lobes (mm): 0-absent; 1-small (1-5); 2-medium (5-20); 3-large (20-50). . Arrangement of ostia: 0-dispersed; 1-alveolar pattern. 9. Colour: 0-white/cream; 1-light brown; 2-yellow; 3-green; 4-variable. 10. Consistency: 0-rigid; 1-cartilaginous; 2-semi-cartilaginous; 3-firm but compressible; 4-soft; 5-fragile. 408 MEMOIRS OF THE QUEENSLAND MUSEUM APPENDIX 1 (cont.). Morphological and cytological characters and character-states used in phylogenetic analysis. Uninformative characters are marked with an asterisk (*). 12. Choanocyte chambers type: 0-aphodal or diplodal; 1-eurypilous. 13*. Chamber diameter (jm): 0-small (10-30); 1-large (30-70). 14, Subdermal cayities: 0-absent; |-present. 15*. Basal cavity: 0-absent, 1-present. 16. Ectosome thickness (um): 0-greater than 100; 1-between 50 and 100; 2-between 5 and 50. 17. Proportion of mesohyl to chambers: 0-greater than 2; 1-between | and 2; 2- less than |, 18. Spicule malformations: 0-absent; |-present. 19. Diods: 0-absent; 1-present. 20. Triods: 0-absent; |-present. 21. Smooth calthrops: 0-absent; 1-present. 22. Monolophose calthrops: 0-absent; 1-present. 23. Dilophose calthrops: 0-absent; |-present. 24. Trilophose calthrops: 0-absent; 1-present. 25. Tetralophose calthrops: Ü-absent; |-present. 26*. Candelabra (heterolophose calthrops): 0-absent; 1 -present. 27. Position of first ramification in lophose actines: 0-absent; 1-proximal; 2-medial. 28*. Second ramification in lophose actines: 0-absent; 1-distal. 29. Lophose actine ends: 0-absent; 1-simple; 2-with terminal spines. 30*. Desmata, oxea, and discotriaenes: 0-absent; 1-present. CYTOLOGICAL CHARACTERS. 31. Apopylar cells: 0-absent; |-triangular; 2-ovoid, with large osmiophilic inclusions. 32*. Flagellum in pinacocytes: 0-absent; 1-present. 33*. Pinacocyte shape: ()-‘T-shaped’; 1-flat/ovoid. 34*. Basement membrane: 0-absent; |-present. 35. Vacuolar cells type A: 0-absent; 1-present. 36*. Vacuolar cells type B: 0-absent; 1-normal, sparse; 2-turgid, in groups. 37. Vacuolar cells type C: 0-absent; 1-present. 38. Paracrystallin inclusions cells: 0-absent; |-present. 39*, Homogeneous inclusions cells: 0-absent; 1 present. 40. Single inclusion cells: 0-absent; 1-present. 41*. Crescent cells: 0-absent; 1-present. 42*, Microgranular inclusions cells: 0-absent; 1-present. 43*. Spherulous/microgranular inclusions cells: 0-absent; |-present. 44. Sclerocytes: 0-absent; 1-present. 45. Collencytes: (-absent; 1-present. 46, Endosymbiont bacteria: 0-few, sparse; 1-diverse, abundant. 47. Distribution of endosymbiont bacteria in the mesohyl: 0-random/uniform; 1-patchy; 2-close to choanocytes; 3-far from choanocytes. PHYLOGENY OF HOMOSCLEROPHORIDA 409 APPENDIX 2. Data matrix of the combined data set (characters 1-24). Names of characters and character-states are listed in Appendix 1. Character Taxon i|j2|s]4]s[e[7 [s ]|» [10] 11] 12/13] 14] 15 [16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 Pseudocorticium | ¿| 1/1&| g| 3| 5| of 1| of 1| o of 1| 0| 1| 0] 0| of of 0| of 0| ol o jarrei 2 Plakinajani | 1 | 1} 1 | 2 | 2 ia lali Pe et fea E A | eel ales Plakina trilopha | 1 | 1 | 1 | 2 | 2 1|1]0[|3]|1]|1 111|1:/1]1[1]1|1/1]14|1 Plakina Boo VENA a a oe ere a rata la a JE Plaking 3]0|0|1|[1]1]0/1]0|4]|1[|1 0|1/2|2[|0|1|]1/1|1]|0]|0 monolopha | | Oscarella 1% an 2 1| | 2/ 2] 110. 1| 4| 2] 1| 1| 1,0, 1|2| 2] of 0 of of of of 0 Oscarella 18 Peek 201/72 2 2 of of 1| 4| 4| 1| 1| 110, 1 2] 2] of of 0| of of of o Oscarella |5|1'1'2|1|]1/0|1|1|4 1/1 0|1/2|2[0/0|/0|]0]0]|0|0 microlobata | ~ Oscarella viridis | 2 1 1/2] 1 1 0; 1 3.| S 1 1 0|1|[2/2.0/0/0/,0/0/0/]|0 Corticium 4|0|1]1[0|2]0|1|1]|]1/]0]|0 0|1/0/0[|[0/0]0/1|1/|0]|0 candelabrum Discodermia 9 o9 5lg|o[olololololo|o o ololo|o|jo|o|olo ollo polydiscus APPENDIX 2 (continued). Data matrix of the combined data set (characters 25-47). Names of characters and character-states are listed in Appendix 1. T Character axon 25 | 26 | 27 | 28 | 29 | 30 | 31 | 32 | 33 | 34 | 35 | 36 | 37 | 38 | 39 | 40 | 41 | 42 | 43 | 44 | 45 | 46 | 47 Pseudocorticium 696 9g|69|lo 1|1]11[0]0|1|1]11[0]0]0/0|1]1)3 jarrei Plakinajani | 1|0|2 {1 | 2 2|1]1]|1[0/|0/|1 010|0|0|0/|1]|0]|1,|1! Plakina trilopha | 1| 0| 20 | 1 211/1]1] 0 0 01010[|0|1/|/0]|1]|1/|1 Plakina oda 1, 0/1|0[|2|0|2|]1|1 1 o0|o0|0[|0|0|0|0 |o 0o 1 /0[|o0[o Plakina 0|0|[2|0|1|0|2|1|1]/1]0]0]0]0]0]0]0]0]0|1]0]|1]0 monolopha | L Oscarella olo|ololo|ol1l1l1lrilo 2 lolololololo|ololo|o lo tuberculata iL Oscarella 0|0[|[0|0]0]0|1]1]|1]1]1]1]/0/0]0]0/0|/0|]0]0]|0]|0]0 lobularis Oscarella 010/0[0]0]0|1]11|[1]1]0[0]|1]0|1|0]0|1[|0]0]|1]|0 microlobata Oscarellavirids | 0|0|0|0/0|0|1|1|1]1[0/0]0[|0,/0|]0|1]|]1]|/0[|0/0]|0]|0 Corticum 0151]/1[0/1]0]2|]1]1]1[0/0]0[|0/0]0|0|/0|0|1/|1]1]|2 candelabrum Discodermia |g8]60l0|0|1/0/0/0|0|0/.0/0|0/0/0/0|0/.0 111 110 polydiscus 410 MEMOIRS OF THE QUEENSLAND MUSEUM THE PHYLOGENETIC HISTORY OF SPONGES IN PALAEOZOIC TIMES. Memoirs of the Queensland Museum 44: 410. 1999:- According to molecular biological analysis, the origin of the phylum Porifera dates back about 800MY. The first sponges were probably aspicular, like the so far oldest definite sponge Palaeophragmodictyon from the Late Proterozoic of Ediacara (the even older Duoshantuo fossils may reverse the picture again if the sponge interpretation holds true), Early Cambrian sponge assemblages were dominated by the Hexactinellida, but by the time of the Atdabanian the Pinacophora (Demospongiae/Calcarea-taxon) were also well represented. The Archaeocyatha can be considered as a stem lineage representative of this group. Early Cambrian Calcarea comprise modern-appearing forms as well as the exclusively Palaeozoic Heteractinellida and Polyactinellida, the latter group exhibits triradiate calcitic spicules, which are probably a constituent character of the taxon Calcarea. Within the Demospongiae, the tetraxon is considered the basic spicular symmetry, from which the other spicula-types have derived. Oxyasters from the Early Cambrian, which are in the size range of megascleres and show well-developed central canals, may have evolved from tetraxone mesotriaenes, whereas the large Middle Cambrian sigmata are probably derived oxeas. This means that the differentiation in mega-and microsclerocytes known from recent demosponges may have taken place at a later stage of poriferan evolution. The first desmata-bearing demosponges (‘Lithistida’) of the group Anthaspidellidae, Orchocladina-known since the Middle Cambrian-probably originated from reticulated monaxonid precursors close to the Hazeliidae. During the Late Ordovician, the chiastoclones developed from anthaspidellid dendroclones, and the Palaeozoic groups Tricranocladina and Sphaerocladina may have derived from chiastoclonellid ancestors. Contrary to widely accepted hypotheses, there seems to be no direct phylogenetic line from the Orchocladina to the modern Tetracladina, since the origin of tetraxial desmata from anaxial chiastoclones is very unlikely. The earliest true tetraclones with definite axial canals are documented from the Permian Jereina robusta, whereas the first phyllotriaenes of the modern spirasterophoran type are known since the Late Triassic. Because of their skeletal architecture, the Palaeozoic Saccospongiidae and Orchocladina, as well as theTricranocladina and Sphaerocladina which most probably evolved from the Orchocladina, are now attributed to the Sigmatophora. The ‘megamorine’ Saccospongiidae probably originated from a monaxonid group close to the Halichondritidae, but the Palaeozoic heloclones and megaclones as well as the elongate rhizoclones of the Haplistiidae are probably not phylogenetically linked to the modern Megamorina or Rhizomorina. Mesozoic and Recent Rhizomorina are characterised by skeletons of exclusively small rhizoclone desmata with sigmaspires as microscleres. But the sigmaspire is unknown from the fossil record and almost certainly has no connection with the sigmatophoran sigmata, which are known since the Middle Cambrian. At the end of the Permian, the Palaeozoic ‘Lithistida’, maybe with the exception of the Sphaerocladina, had all become extinct. Against widely accepted ideas, there is probably no phylogenetic link from the Tricranocladina to the modern Corallistidae (Dicranocladina). The Sphaerocladina, which have recently been documented also from the Early Palaeozoic, may have lead to the Mesozoic Neosphaerocladina, but no connection can be documented between these groups and Recent genera sometimes attributed to the Sphaerocladina, such as Crambe or Vetulina. Lophocalthropses of the Plakinidae first occurred in the Early Carboniferous and are connected by transitional forms to the candelabra of the modern Homoscleromorpha, known since the Early Cretaceous. The characteristic Plakinidae spicules probably originated from the same type oftetraxones, which lead to the first dichotriaenes in the Early Carboniferous. Thus the Plakinidae/ Homoscleromorpha are probably the sister group ofthe Spirasterophora, to which most modern ‘Lithistida’ belong. O Porifera, phylogeny, Palaeozoic sponges, skeletal architecture, Calcarea, Demospongiae, Hexactinellida. Dorte Mehl (email: palaeont@zedat.fu-berlin.de), Institut für Paldontologie, Freie Universitdt Berlin, Malteserstrasse 74-100, D-12249 Berlin, Germany; I June 1998. RELEASE OF ALLELOCHEMICALS BY THREE TROPICAL SPONGES (DEMOSPONGIAE) AND THEIR TOXIC EFFECTS ON CORAL SUBSTRATE COMPETITORS GREGORY K, NISHIYAMA AND GERALD I. BAKUS Nishiyama, G.K. & Bakus, G.J. 1999 06 30; Release of allelochemicals by three tropical sponges (Demospongiac) and their toxic effects on coral substrate competitors. Memoirs of the Queensland Museum 44: 411-417. Brisbane. ISSN 0079-8835. Three sponge species (Xestospongia, Acervachalina, Plakortis spp.) from Mactan 1.. Philippines, were shown to release allelochemicals directly into the water. These allelochemicals were demonstrated to be toxic t6 one or more scleractinian coral species (deropara, Pocillapora, Porites spp.) and, for one of the sponges tested, to one hydrozoan coral (Millepora sp.), The five coral species tested (including Montipora) were beth numerically and spatially dominant organisms at the study site. Toxicity tests involved exposing corals brought into the laboratory to water that had been conditioned by the sponges, Responses of each coral species to each sponge allelochemical varied. The allelochemical from 4cervochalina was found to be highly toxic (51-75% tissue death) to both Pocillopara and Acropora, and had only a moderate effect (26-50% tissue death) on Porites. Allelochemicals of Yestospongia and Plakortis were moderately and weakly toxic (11-25% tissue death) lo Millepora, respectively. Neither sponge was toxic towards the other coral species, Montipora was not affected by allelochemicals from any of the sponges, Dead coral was noted in many positions around the sponges in the field, but mainly in the direction of the current. This might support, although not confirm, an allelopathic effect. The influence of allelochemicals on the small scale and large scale spatial structuring of coral reefs is discussed. CJ Porifera, corals, allelachemieq!, coral reef, toxieity, chemical ecology, Philippines, Xestospongia. Acervochalina, Plakarris. Gregory K, Nishiyama (email: nishivanascf use edu) de Gerald J. Bakus, Department af Biological Sciences, University of Southern California, Los Angeles, CA, 90089-0377, USA:3 February 1999, Release of allelochemicals into the surround- ing water, as a defense against benthic spatial competitors, fouling organisms, or micro- organisms, has been demonstrated in several groups of marine organisms (see Bakus el al., 1986). These organisms include sponges (Thompson, 1985: Walker et al., 1985; Porter & Targett, 1988; Targett, 1988; Bingham & Young, 1991), soft corals (Coll & Sammarco, 1983; Sammarco et al.. 1983, 1985; La Barre et al... 1986; Maida et al., 1995), an anemone (Bak & Borsboom, 1984), possibly an alga (Littler Se Litter, 1997) and hard corals (Fearon, 1997; Koh, 1997), Although only few studies have yet been undertaken, allelochemicals are believed to play a role in structuring particular marine habitats (Jackson & Buss, 1975; Davis et al., 1989: Turon et al.. 1996: Thacker et al,, 1998). whereas the magnitude of this role is still uncertain, La Barre et al. (1986) demonstrated that when three species of soft corals were relocated in pairs under contact and non-contact conditions, tissue necrosis was observed in all contact pairs. Only one species o['soft coral produced tissue necrosis when placed near, but not in contact with, the other two species of soft coral, They suggested that avoidance behavior, such as those caused by allelopathy, may contribute to the dispersion patterns of plants and animals in space. In another study on soft corals, Sammarco et al. (1985) found that scleractinian corals varied in their susceptibility when exposed to soft corals in both contact and non-contact conditions. Porter & Targett (1988) demonstrated that the sponge Plakortis halichondroides was capable of damaging or destroying tissue in all coral species examined, with almost half the corals growing naturally in contact with, or near to, these sponges experiencing bleaching or tissue necrosis. It was suggested that by creating dead zones on the corals P halichondroides was subsequently able to overgrow them. In a more recent study, Turon et al. (1996) suggested that Crambe crambe could have an impact on adjacent organisms by possibly releasing allelochemicals 412 into the surrounding water. The allelochemical effect of this sponge was at the small scale level (centimeters) (Turon et al., 1996), where patterns such as an increase in the amount of dead coral or tolerant species found adjacent to the producer of allelochemicals became evident. Turon et al. (1996) suggested that at different scales different patterns might be recognised. On previous trips to the Philippines we observed similar bleached areas or dead zones up to lcm in width in areas where some of the species of sponges came in contact, or close contact, with coral species. The purpose of the present study was to: 1) Determine ifthree tropical sponges were releasing allelochemicals potentially toxic to five hard corals on a coral reef; 2) Identify patterns due to allelochemical effects, such as the presence or absence of dead space adjacent to sponges (given that sponges releasing toxic allelochemicals would have a preponderance for dead coral, or tolerant species adjacent to them, and these might be predominantly located in the direction where the allelochemicals were most concentrated), and 3) Determine ifthe allelochemicals were toxic to common substrate competitors. MATERIALS AND METHODS Our study site was located on a limestone reef approximately 0.25km offthe Tambuli Resort on Mactan I., Philippines. The study site was chosen on the basis of its high marine diversity and close proximity to the Maribago Marine Station, operated by the University of San Carlos. Depths Plastic Container \ \ MEMOIRS OF THE QUEENSLAND MUSEUM in the vicinity of the study site ranged from 5m, where a seagrass bed began, to 15m, which bordered the start ofa steep slope. However, most of the experiments conducted in this study were located between 8-11m depth. The study was conducted in May and June, 1996. (For a detailed description and map of the site see Bakus & Nishiyama, 1999, this volume). Species of sponges included: Xestospongia sp. (Haplosclerida: Petrosiidae), Acervochalina sp. (Haplosclerida: Chalinidae), and Plakortis sp. (Homosclerophorida: Plakinidae), and the corals: Acropora sp. (Scleractinia: Acroporidae), Millepora sp. (Milleporina: Milleporidae), Montipora sp. (Scleractinia: Acroporidae), Pocillopora sp. (Scleractinia: Pocilloporidae) and Porites sp. (Scleractinia: Poritidae). ALLELOCHEMICAL DETECTION AND ISOLATION. To determine if sponges were releasing chemicals into the water, an allelo- chemical collecting apparatus was constructed from a battery operated bilge pump and SEP paks (Fig. 1). This apparatus was a modified version of one used by Coll et al. (1982), and later by Schulte et al. (1991). The battery operated bilge pump was fitted at its outflow hose with a step- down tubing connector (2cm to 0.7cm), which allowed two plastic screw valves to be connected (diameter 0.7cm). C18 SEP paks were initially conditioned by passing 10ml of Etoh followed by 10ml of deionised water through each SEP pak using a plastic pipette. Conditioning SEP paks before use was critical, otherwise flow would be Outflow Hose , SEP pak S Sponge N Bilge Pump FIG. 1. Allelochemical isolating apparatus used to isolate allelochemicals from sponges. TROPICAL SPONGE ALLELOCHEMICALS interrupted. /n situ, one conditioned SEP pak was inserted into each end of the plastic screw valve. At the inflow end of the pump, a plastic collapsible container (20x30x50cm) was attached which had a large hole cut at the bottom (diameters 20x30cm; Fig. 1). The purpose of the container was to assist in concentrating allelochemicals to increase probability of detection. Once the SEP paks were attached, the apparatus was placed over the sponge, with the sponge protruding into, but not touching, the plastic container. Although flow rate was determined both at sea level and 10m depth by allowing the output water to flow into an empty container for 30mins, no flow meter continuously measured flow rates in this initial model. Water surrounding eight unidentified sponges was sampled for l.5hrs, coinciding with the maximum continuous run-time for the pumps. A control apparatus was set-up approximately 0.5m upstream from each apparatus covering a sponge. A newer model is currently being constructed with a flow meter and an added battery included for longer run times and continuous monitoring. Eight different sponge species were sampled at the study site using this apparatus to detect the release of any allelochemicals. After each run, the SEP paks and the whole sponge specimens being sampled were immediately sealed separately in small plastic bags, and upon arrival to the laboratory (not more than 20mins later), were placed in a freezer (-5°C). Upon departure from the Philippines the SEP paks and sponge specimens were placed in dry ice, and upon arrival in the USA the items were placed into a deep freezer (-20°C). Chemicals were initially eluted out by passing 20ml of dichloromethane, followed by 20ml of Etoh through SEP paks. Extractions were air evaporated for approximately 24hrs. It was subsequently discovered that extraction with 50ml acetone produced higher extraction yields. Thereafter, only acetone extractions were used. Thin layer chromatographies (TLC) were conducted using acetone as the mobile phase throughout the extraction process to insure that chemicals were not lost during the evaporation . TLCs were visualised using both long and short UV light. There was no noticeable changes in chemical composition of the extract within this period. TLCs were run on SEP pak extracts, control SEP paks, and whole sponge extracts (5gm samples of each sponge species extracted in acetone). SMALL SCALE PATTERNS AROUND SPONGES. Patterns of organisms found adjacent to, or within approximately 5cm from, sponges were investigated by first taking photographs of between 29-39 individuals of each species and their adjacent organisms in situ with a Nikonos V camera. A scale bar with a waterproof compass attached to it was laid parallel to the direction of the current and placed next to each sponge before photographs were taken. Sponges in each photo- graph were divided into four quadrants and the presence or absence of dead coral in each quad- rant was recorded. The widths of the dead zones on corals ranged from 1-10mm. Seven categories were devised to quantify the positions of dead coral: 1) Horizontal: dead coral only on two quadrants that faced currents (roughly in the E and W positions); 2)Vertical: dead coral in two quadrants perpendicular to the current (roughly in the N and § directions); 3) Dead coral found more in horizontal than vertical positions; 4) Half-half: dead coral found equally in horizontal and vertical positions; 5) More dead coral in vertical than horizontal positions; 6) Dead coral found completely around the sponge; and 7) Only live organisms completely surrounding the sponge. TOXICITY OF ALLELOCHEMICALS ON HARD CORALS. Fifty pieces (approximately 3-4cm long) from each of two individuals of the five coral species were obtained at the study site using a geological pick, and placed in separate plastic bags for each coral species. These plastic bags were then placed into buckets to ensure minimal mechanical stress during transport. Sponges were removed from the substrate by chiselling around each sponge and carefully removing them. Sponges were then positioned and left on dead coral for one week to allow recuperation following their removal. When ready for use, two whole sponges of each of the three sponge species were placed in separate plastic bags. Care was taken to minimise damage when removing and transporting corals and sponges. Both were returned to the laboratory and sponges were immediately placed in a 1L beaker filled with filtered seawater obtained from the study site (ambient water, gravity filtered with a#1 Whatman filter). Corals were placed in separate plastic buckets with unfiltered seawater from the site until ready for use. Sponges were allowed to condition the water for 1hr before the water was passed through a Whatman #1 filter and gravity filtered. This filtration process required less than 30mins. Two sponge individuals from each species were allowed to condition water separately to test for individual variability 414 Xestospongia sp. Plakartis sp, No. of Sponges Acervachalina sp Hori. Mathair More Verks Valid Dead Space Positions AN Dea Al Lave FIG. 2. Positions of dead space found around sponges from Mactan Island, Philippines. (See text for detailed descriptions of categories). towards allelochemical toxicity. While this filtration process was carried out, corals were placed in glass finger bowls along with 100ml of ambient water. When sponge-conditioned water was ready, water in finger bowls was decanted off and replaced with an equal volume of sponge-conditioned water. A control treatment of IL ofnon-conditioned, filtered seawater was also initially set aside for 1hr. The condition of corals and water temperature were noted. Corals were exposed to conditioned and control water experiments for 24hrs. A change in temperature from 25°C to 22°C was recorded for the water in the bowls. The ambient water temperature at the study site during the experiment was 27°C. After this 24hr period, corals were returned to their original site of collection and placed under a metal cage (5x30x100cm; mesh size approx. lcm). The condition of coral fragments was monitored for 4 days thereafter. Following this period, close-up photographs were taken, both in MEMOIRS OF THE QUEENSLAND MUSEUM situ and after the corals were returned to the laboratory. These photographs were enlarged (approximately 200%), and the area of the exposed surface of each piece of coral, as well as the area of dead tissue, traced onto acetate. The percentage of dead tissue was determined after estimating the total area of each coral fragment and the area of dead tissue, using an image analysis program (SigmaScan [SPSS Inc], 1998). Detailed notes and sketches of each piece of coral in situ were made earlier and were compared to the calculated dead tissue areas. In all instances both estimates were comparable. Coral tissue was considered to be dead or dying if there were signs of cell death or, as in most cases, actual detachment of tissue from the skeletal base. Loss of coloration was also observed in all dead tissue. One-way ANOVA and post hoc Tukey tests were conducted on the percentage tissue death of each coral species, with the allelochemical and control treatments being the main factor. RESULTS Of the eight unidentified sponges sampled for allelochemical release at our study site, three were shown to be releasing allelochemicals. These species were subsequently identified as Acervochalina, Plakortis, and Xestospongia spp. TLC comparisons between water immediately surrounding these three sponges, control water samples, and whole sponge extracts confirmed that allelochemicals were found within the respective sponges but not in the water column upstream from the sponges. The presence and position of dead space adjacent to the the species of sponges occurred predominantly in the direction facing (either partially or completely) the current flow (Fig. 2). Since the direction of water flow was reversed periodically, depending on whether the tide was rising or falling, dead space occurred on both sides of the sponge facing these currents. In Xestospongia dead space occurred in horizontal positions (i.e. facing currents) and positions mostly facing current flows in 71% of specimens (Fig. 2). Dead space was located in these positions in 73% of Plakortis, and 66% of Acervochalina. In no instance for any of the three species of sponges was dead space found only in the vertical position (i.e. not facing current flow). Water conditioned individually by each of the three species of sponges was toxic to one or more of the five coral species tested (Fig. 3). Responses of each coral species towards each sponge TROPICAL SPONGE ALLELOCHEMICALS oh Puritén Ap Percentage Tissue Denth nil Millepora sp. wi ee FIG 3. Average percentage tissue death of five coral species exposed to water conditioned by three sponge species (Key: ]=Xeslospongia; 2=Plakortis; 3=Acervochalina; C= control treatments, representing corals exposed to filtered seawater). (Bar represents 1 S.E.). allelochemical varied. Control corals exper- ienced minimal tissue death, but not always less death than all the other treatments (Fig. 3). Forall coral species, except Montipora, significant differences existed among the ireatments (one- way ANOVA). Results of the Tukey tests and average percentage coral tissue death show that when compared to controls, the allelochemical of Acervochalina was highly toxic (51-75% tissue death) to Pocillopora (P<0.05) and Acropora (P<0.05), and had a moderate effect (26-50% tissue death) towards Porites (P<0.05) The sponge was not toxic towards Millepora. Xestospongia and Plakortis were moderately (26-50% tissue death; P<0.05) and weakly toxic (11-25% tissue death) (P=0.05) to Millepora, respectively. Neither sponge was toxic towards the other coral species. Montipera was not affected by allelochemicals from any of the sponges (P>0,05). oy Montipori Nj. DISCUSSION Results show that filtered sea- water conditioned by all three species of sponges, Xestospongia, Plakortis and Acervochalina, were toxic to al least one species of hard coral (Porites, Pocillopora, Acropora and Millepora). Responses of corals towards sponge-condit- ioned water varied, as expected. Susceptibility of corals in contact with, or in proximity to sponges suggests the possibility that sponge allelochemicals may influence patierns of distributions of organ- isms adjacent to sponges. In other words, certain tolerant species of corals may grow adjacent to sponges, whereas others may never or rarely be found in close proximity. In a parallel study we conducted at our study site, whole sponges were transplanted and placed into direct contact with corals. We determined that several species of corals were highly susceptible to (i.e. damaged by) several species of sponges. Under natural conditions, these corals were rarely found growing in direct contact with these sponges (Nishiyama & Bakus, unpublished information). In rare cases where they did grow naturally adjacent to these sponges, a zone of dead or bleached tissue was often noted at the site of contact. Furthermore, if a species of sponge was not found to be deleterious to a species of coral, incidences of growth with direct contact between the two species were more often noted. This corresponds to field observations made by Porter & Targett (1988) where almost half the corals growing nextto Plakortis halichondroides experienced bleaching or tissue necrosis. Both these studies support our hypothesis that water borne allelochemicals may deter particular substrate competitors from growing in direct contact with sponges. It should be cautioned, however, that laboratory bioassays using sponge- conditioned water were conducted on corals in still water, whereas in nature currents probably have a major influence in dissipating allelochemicals, and thus, their physiological impact may not be as extensive as those observed in the laboratory. 416 Although particular sponges may have an impact on adult corals, these toxic effects may also have a greater impact in preventing coral larvae settling adjacent to sponges. The occur- rence of dead coral space around sponges, mainly in the direction facing currents, may reflect the directions that highest allelochemical concen- trations occur given that toxins are carried away from sponges. Although only suggestive, this supports the notion that toxic allelochemicals were being released by sponges. Maida et al. (1995) provided evidence to suggest that a soft coral could influence the direction from which recruitment occurred. An alternative hypothesis is that some hydrodynamic effect prevents settlement of larvae in areas facing the current. It may also be advantageous for a sponge to encourage adjacent settlement and growth of coral species which are susceptible to the sponges’ allelochemicals. This would ensure that the sponge would be able to grow and expand into the adjacent area. Through their release of allelochemicals sponges may stop the growth of adjacent, adult substrate competitors, such as corals, However, this effect may operate only on a local or small scale, at least for the three species of sponges investigated here, because zones of dead or bleached tissue on corals growing adjacent to the three species of sponges extended for only 1em at most. Acervochalina, however, may also possibly overgrow corals, with an observed high density in comparison to the other two species, generally having the highest toxicity towards corals, and being thinly encrusting with possibly greater lateral growth. Acervochalina was also observed growing on branches of corals that were dead at the bases (where the sponges occupied), yet alive at the tips (where the sponge had not yet extended). Allelochemicals of Acervochalina may operate to stop corals from overgrowing it and as a mech- anism to kill coral tissue, to open up space for its own growth. In a study conducted by Bakus & Nishiyama (1999, this volume), data from transects at the study site show that no apparent sequence existed where a particular substratum type (i.e. live hard coral, sponges, coral rubble, etc.) was found next to sponges; that is, the succession of organisms and substrata were independent of each other. In that study, however, individual species of both corals and sponges were not differentiated, and therefore, specific species pairs may exist. Using transect line data techniques, Turon et al. (1996) investigated the possibility of allelochemicals MEMOIRS OF THE QUEENSLAND MUSEUM being released by sponges. They determined that Crambe crambe had toxic effects up to 1cm from particular coral substrate competitors, corresp- onding to the effective distance of allelochemicals suggested in the present study. Turon et al. (1996) also suggested that these small scale effects might only be detected by sampling at a small scale (centimeters), whereas sampling at 3cm intervals produced a different outcome. Where allelochemicals have effective distances at a scale of less than 1 em, observations made at 1cm intervals may not suffice in detecting these chemical interactions. In the present study, extreme care was taken to accurately determine dead space and organisms adjacent to sponges, and in most cases, measurements were deter- mined to the nearest millimeter. The chemical nature of allelochemicals released by sponges, as measured by SEP paks, are currently being ascertained, as are the toxic- ities of these isolated chemicals towards the five species of corals. Although SEP paks retained chemicals released by sponges, this does not necessarily confirm any toxicity by the sponge. Only isolation of the active chemicals from sponges and verification of their toxicity towards corals would confirm that these are allelo- chemicals deleterious to corals. Data from another study conducted by the authors are currently being analyzed (Nishiyama & Bakus, unpublished data), involving the deterrence of settlement of substrate competitors by sponges placed next to plates kept in the water at the site for approx- imately one month. After hard corals, sponges were the dominant organisms at Mactan I. (Bakus & Nishiyama, 1999, this volume), also showing relatively high diversity (Bakus & Nishiyama, unpublished data). Of eight sponges investigated, three released allelochemicals into the water column, suggest- ing that many more sponges, not investigated here, may release allelochemicals. Thus, sponge allelochemicals may play an important role in structuring the coral reef community at a small scale, local level. More work is needed, however, to determine the extent of this role. ACKNOWLEDGMENTS We greatly appreciate the help of Dr Filapina B. Sotto, Head of Marine Biology Section, University of San Carlos, Cebu City, who made our studies possible. We would also like to thank Jason Young, Jonathan Apurado and Ben Pangatungan for their help throughout the project, TROPICAL SPONGE ALLELOCHEMICALS both in the laboratory and in the field. We appreciate all the assistance given by many of the members of the Marine Biology Section at the University of San Carlos. We are also grateful to Cynthia Kay who assisted us in all aspects of the project. Finally, we thank Domingo Ochavillo, Chona Sister and Cynthia Kay for reviewing this manuscript. Rob van Soest and John Hooper assisted with sponge identifications. LITERATURE CITED BAK, R.P.M. & BORSBOOM, J.L.A. 1984. Allelo- pathic interaction between a reef coelenterate and benthic algae. Oecologia 63: 194-198. BAKUS, J.G., TARGETT, N.M. & SCHULTE, B. 1986. Chemical ecology of marine organisms: An overview. Journal of Chemical Ecology 12(5): 951-987, BINGHAM, B.L. & YOUNG, C.M.. 1991. Influence of sponges on invertebrate recruitment: a field test of allelopathy. Marine Biology 109: 19-26. COLL, J.C., BOWDEN, B.F. & TAPIOLAS, D.M. 1982. In situ isolation of allelochemicals released from soft corals (Coelenterata: Octocorallia): A totally submersible sampling apparatus. Journal of Experimental Marine Biology and Ecology 60: 293-299, COLL, J.C. & SAMMARCO, P.W. 1983. Terpenoid toxins of soft corals (Cnidaria, Octocorallia): Their nature, toxicity, and ecological significance. Toxicon (Supplement) 3:69-72. DAVIS, A.R., TARGETT, N.M., MCCONNELL, O.J. & YOUNG, C.M. 1998. Epibiosis of marine algae and benthic invertebrates: natural products chemistry and other mechanisms inhibiting settle- ment and overgrowth. Pp. 85-114. In Scheuer, P.J. (ed.) Bioorganic Marine Chemistry. Vol. 3. (Springer-Verlag: Berlin). FEARON, R.J. 1997. Toxins of Hermatypic Corals. Unpublished Ph.D. thesis (University of Queensland: Brisbane) JACKSON, J.B.C. & BUSS, L. 1975. Allelopathy and spatial competition among coral reef invert- ebrates. Proceedings of the National Academy of Sciences USA 72(12): 5160-5163. KOH, E.G.L. 1997. Secretion of bioactive compounds by a scleractinian coral. Proceedings of the 8" International Coral Reef Symposium 2: 1263-1266. LA BARRE, S.C., COLL, J.C. & SAMMARCO, P.W. 1986. Competitive strategies of soft corals (Coelenterata: Octocorallia): III. Spacing and 417 aggressive interactions between alcyonaceans. Marine Ecology Progress Series 28: 147-156. LITTLER, M.M. & LITTLER, D.S, 1997. Epizoic red alga allelopathic (?) to a Caribbean coral. Coral Reefs 16:168. MAIDA, M., SAMMARCO, P.W. & COLL, J.C. 1995. Preliminary evidence for directional allelopathic effects of the soft coral Sinularia flexibilis (Alcyonacea: Octocorallia) on scleractinian coral recruitment. Bulletin of Marine Science 56(1): 303-311. PORTER, J.W. & TARGETT, N.M. 1988. Allelochemical interactions between sponges and corals. Biological Bulletin 175: 230-239, SAMMARCO, P.W., COLL, J.C. & LA BARRE, S. 1985. Competitive strategies of soft corals (Coelenterata: Octocorallia). II. Variable defensive responses and susceptibility to sclera- ctinian corals. Journal of Experimental Marine Biology and Ecology 91: 199-215, SAMMARCO, P.W., COLL, J.C., LA BARRE, S. & WILLIS, B. 1983. Competitive strategies of soft corals (Coelenterata: Octocorallia): Allelopathic effects on selected scleractinian corals. Coral Reefs 1: 173-178. SCHULTE, B.A., DE NYS, R., BAKUS, G.J., CREWS, P., EID, C., NAYLOR, S. & MANES, L.V. 1991. A modified allomone collecting apparatus. Journal of Chemical Ecology 17(7): 1327-1332. TARGETT, N.M. 1988. Allelochemistry in Marine Organisms: Chemical fouling and antifouling strategies. Pp. 609-617. In Thompson, M., Sarojini, R. & Nagabhanam, R. (eds) Marine Bio- deterioration. (Oxford and IBH publishing Co.: New Delhi). THACKER, R.W., BECERRO, M.A., LUMBANG, W.A. & PAUL, VJ. 1998. Allelopathic inter- actions between sponges on a tropical reef. Ecology 79:1740-1750. THOMPSON, J.E. 1985. Exudation of biologically- active metabolites in the sponge Aplysina Jistularis. I. Biological evidence. Marine Biology 88:23-26. TURON, X., BECERRO, M.A., URIZ, M.J. & LLOPIS, J. 1996. Small-scale association measures in epibenthic communities as a clue for allelochemical interactions. Oecologia 108: 351-360. WALKER, R.P., THOMPSON, J.E. & FAULKNER, D.J. 1985. Exudation of biologically-active metabolites in the sponge Aplysina fistularis. Il. Chemical evidence. Marine Biology 88: 27-32. 418 MEMOIRS OF THE QUEENSLAND MUSEUM TOWARDS A PHYLOGENETIC SYSTEM- ATICS OF THE FOSSIL HEXACTINELLIDA. Memoirs of the Queensland Museum 44: 418. 1999:- The vast majority of all fossil hexactinellid taxa has been described from the Mesozoic. This is due to the rich occurrence of Mesozoic hexactinellids, especially in the well-exposed Jurassic and Cretaceous strata of Europe, and to the generally larger preservation potential of the rigid Mesozoic hexactinellids compared to the predominantly non-rigid Palaeozoic ones. Nevertheless, most of the main hexactinellid taxa can be traced back to the Early Palaeozoic. Isolated hexasters of the Hexasterophora occur in the Early Ordovician, and the first hexactinosans are known from the Late Devonian, whereas the earliest definite amphidiscophorans are documented from the Late Silurian. However, the bulk of the Palaeozoic hexactinellid sponges, although well established as monophyletic groups, cannot definitely be attributed to any recent taxon and require an exclusively fossil-based systematics. The Early Palaeozoic Protospongiidae and Hintzespongiidae are derived from areticulate hexactine-bearing ancestors, probably close to the Mattaspongia-Microstaura-group, which can be regarded as adelphotaxon of the Hexactinosans. The Dictyospongiidae (s.str.), which are hexasterophorans, probably also originated from the Mattaspongia-stem lineage, as did the modern Sceptrulophora (Clavularia-Scopularia-taxon), which recently have been traced back to the Early Palaeozoic through the documentation of Ordovician scopules. The Brachiospongiidae, including the Stiodermatidae, may be attributed to the amphidiscophorans, because of the great similarity in skeletal architecture between Strobilospongia and the modern Hyalonematidae. However, the systematic affinity of many Palaeozoic lyssacine hexactinellids which appear (or were in fact) primitive, including most Early Cambrian genera such as Quadrolaminiella, Solactiniella and Hyalosinica, is still uncertain, and these taxa have to be classified within the probably non-monophyletic grouping *Rossellimorpha'. At the end ofthe Permian, all major Palaeozoic hexactinellid groups had become extinct, and from the Mesozoic onwards, the Hexactinellida are represented by modern forms, mainly Hexactinosans and Lychniscosans. ‘Lyssacinosa’, which comprise the majority of Recent hexactinellid taxa, are not commonly found in Mesozoic strata, but nevertheless there are some important occurrences, from which recent genera can be identified. Regadrella of the Euplectellidae is known with several species from the Cretaceous, and the first species of the Hyalonematidae, Hyalonema cretacea, has been described from the Campanian. But more, new, Late Cretaceous representatives of these groups and also of the Rosselliidae from the section of Arnager (Bornholm, Denmark) are still to be described. The earliest definite lychniscosans are known since the Middle Jurassic, and the group reached its maximal diversity during Late Cretaceous time. Probably, this group did not arise from the hexactinosans, but it is the adelphotaxon of some lyssacine group, maybe the Euplectellidae. Today the Lychniscosans have become almost extinct, so the exact systematic attribution ofthe Mesozoic families and genera to recent ones is problematic and in many cases impossible. The same thing is true to many Mesozoic hexactinosans, although many Recent genera have now been identified from the Late Cretaceous, and this allows an approach ofthe zoological systematics at least for Late Mesozoic and Tertiary sponge fossils. However, still many Cretaceous and most Jurassic hexactinosans classified by Schrammen in the grouping ‘Inermia’, such as the Casearia-Porospongia-group, cannot be definitely attributed to any taxa within the Recent systematics, but have to be subject to a phylogenetic-systematic approach based on fossil representatives only. O Porifera, phylogeny, systematics, Hexactinellida, fossils, Mesozoic. Dorte Mehl (email: palaeont(a) zedat,fu-berlin.de), Institut für Paläontologie, Freie Universität Berlin, Malteserstrasse 74-100, D-12249 Berlin, Germany: 1 June 1998, MEASUREMENT OF SPONGE GROWTH BY PROJECTED BODY AREA AND UNDERWATER WEIGHT RONALD OSINGA, DAVID REDEKER, PETER B. DE BEUKELAER AND RENÉ H. WIJFFELS Osinga, R., Redeker, D., De Beukelaer, P.B. & Wijffels, R.H. 1999 06 30: Measurement of sponge growth by projected body area and underwater weight. Memoirs of the Queensland Museum 44: 419-426. Brisbane. ISSN 0079-8835. In vitro growth rates of the Indo-Pacific demosponge Pseudosuberites andrewsi (Kirkpatrick) were measured using two alternative techniques to estimate biomass: determination of projected body area, and determination of underwater weight. Four small explants of P andrewsi were fed regularly with the microalgae Rhodomonas sp. and Chlorella sorokiniana, and growth was monitored over a period of 24 days. Three explants showed considerable increase in both projected body area and underwater weight, but the growth pattern was irregular. Although the observed trends in growth were similar for both methods, the absolute values were not in general agreement, which may be due to the fact that photographic data were two-dimensional. It was concluded that determination of underwater weight is a promising method for measuring growth of sponges if the size of the explants used is sufficiently large. Measuring projected body area has a higher precision when explants are smaller than 10mg and is a preferred method when small explants are used. O Porifera, Pseudosuberites andrewsi, growth monitoring, projected body area, underwater weight. Ronald Osinga (e-mail: ronald.osinga@algemeen.pk.wau.nl), David Redeker, Peter B. de Beukelaer & René H. Wijffels, Wageningen Agricultural University, Food and Bioprocess Engineering Group, PO Box 8129, 6700 EV Wageningen, The Netherlands; 17 March 1999. Due to the rich potential of marine sponges as producers of interesting natural compounds, there is a growing need for methods to produce large amounts of sponge biomass (Munro et al., 1994; Osinga et al., 19982). /n vitro cultivation of sponges in bioreactors may be an interesting option for mass production of sponge metabolites, because such systems can easily be manipulated and optimised. An essential pre- requisite for studying and optimising in vitro growth of sponges, is to have a good method to monitor this growth. Sponge growth can be monitored by measuring increases in biomass. Thus, to detect slight changes in growth rates, a method is required to precisely measure sponge biomass. This method should not negatively affect the survival or growth of sponges, and in this respect it is important to keep sponges continuously under- water. Although some sponge species can tolerate short exposure to air, it is generally assumed that exposure to air can be harmful to living sponge tissue. Air entering the aquiferous system can irreversibly damage choanocyte chambers (Fossa & Nilsen, 1996). Therefore, determination of sponge volume (by water replacement), wet weight, or drip dry wet weight, although often applied (e.g. Barthel, 1986; Thomassen & Riisgard, 1995), are not preferred methods to measure living sponge biomass and to simultaneously maintain viable experimental populations. In some studies, growth rates are determined by measuring the area of two-dimensionally projected images of the sponge body. Ayling (1983) used this technique to measure in situ growth and regeneration rates of several encrusting sponge species in the coastal water of New Zealand. Series of photographs were taken underwater over a period of time and the images were projected on graph paper. Poirrier et al. (1981) used similar methods to measure in vitro growth of the freshwater sponges Ephydatia fluviatilis and Spongilla alba. Although these methods may be suitable to measure growth in almost two-dimensionally growing encrusting species, or for regularly shaped species such as the sphere-shaped Cinachyrella spp. and Tethya spp., problems may occur when the two- dimensional surface area data are converted into three-dimensional volumes. Especially with more irregularly shaped species, increases in surface area can easily under- or overestimate increases in body volume, especially in more irregularly shaped species. 420 FIG. 1. A 200dn air-lift bioreactor for maintaining the culture of P. andrewsi. In this study, we introduce a new, three- dimensional measure of sponge biomass: under- water weight. Determination of underwater weight is used to measure in vitro growth rates of the Indo-Pacific demosponge Pseudosuberites andrewsi (Kirkpatrick), These results are compared with two-dimensional growth rates obtained from projected body areas. The value of underwater weight as a measure of sponge biomass is further evaluated by correlating these data to other biomass parameters (volume, wet weight, dry weight and ash-free dry weight). MATERIALS AND METHODS SPONGES. On the basis of previous results (Osinga et al., 1998b), P. andrewsi was selected as a model species for further experiments to improve the methodology for in vitro sponge culture. Living material of P. andrewsi was obtained from Blijdorp Zoo (Rotterdam, The MEMOIRS OF THE QUEENSLAND MUSEUM Netherlands), where it was growing in a large, shallow basin, in which a strong water current was generated to simulate an intertidal environ- ment. We are uncertain about the location were these sponges originally came from. They had been introduced in the zoo coincidentally on so called ‘living stones’, which were presumably collected from Indonesian coastal waters. In our laboratory, we have been able to maintain small colonies of this species for more than a year under the conditions described below. Sponges were held in a 200dm' airlift bio- reactor (Fig. |) containing artificial seawater (using Instant Ocean Reef Crystals artificial sea salt) with a salinity of ~32%o. This water was replaced continuously (D=0.033d"'). The temperature in the bioreactor varied between 25-29°C. In order to provide the sponges with a source of silica, 0.25mM Na ,O3Si 9H,0 was added to the artificial seawater. Measurements of the silica concentration in outflow water showed that this addition was sufficient to cope with sponge demands. Non-axenic batch-cultures of two species of microalgae were regularly added as a food source for the sponges. Twice a week, 1dm' of a culture of the freshwater alga Chlorella sorokiniana (Chlorophyceae, average size -3um) was added, containing -1x10? cells cm”. In addition, Idm? of a culture of marine Rhodomonas sp. (Cryptophyceae, average size ~6um), containing ~1x10° cells cm”, was added weekly. The algae were cultured at a temperature varying between 17-20°C. A light-dark cycle of 14hrs light and 10hrs darkness was applied. The growth media for the algae are given in Table 1. When the cultures were added to the sponge reactor, the algae were usually near the end of their logarithmic growth phase. The choice to use these two algae in current experiments was based on the literature. Additions of Chlorella sorokiniana were used successfully to enhance the growth of the temperate sponge Halichondria panicea in semi-controlled cultures (Barthel & Theede, 1986). Rhodomonas sp. was used by Thomassen & Riisgárd (1995) to feed in vitro cultures of H, panicea. Growth experiment. Comparative growth rate measurements were performed on four explants colonies of P. andrewsi. Explants were prepared using razor-sharp knifes. Pieces of sponge tissue were tied onto perspex slides with nylon fishing- line. Explants were placed in temperature controlled, 1.58dm’ bioreactors, equipped with a SPONGE GROWTH MEASUREMENT TABLE 1. Growth media for algae (a freshwater medium for Chlorella sorokiniana and a seawater medium for Rhodomonas sp.). The freshwater medium was based on the A9 medium described by Lee & Pirt (1981). Concentrations are given in mM, unless indicated otherwise. F reshwater Seawater medium Component oak goer teen Concentration NaHCO; 10.0 5.00 KNO; 1.00 0.50 NaH;PO, 0.10 0.05 Instant Ocean Reef Crystals artificial ~33¢g dm? seasalt MgSO,7H;0 4.99 CaCl-2H,0 0.272 EDTANa;2H;0 0.391 FeCl; 0.148 Na;B40710H;O 4.72 - 107 ZnSO4.7H,0 3.13 - 107 CuSO,-5H,0 3.20 - 107 MnSO,.H,0 3,59. 107 Na;MoO,:2H;0 2.07 - 10? NiSO,:6H;O 2.85 - 107 | NaVO; 2.85.10 thyamin-HCI 5.93 - 10? cyanocobalamin 5.90 - 10% biotin 1.64 - 10% sparger for air-supply and a magnetic stirrer to keep food particles in suspension. Sponges were fed with C. sorokiniana (twice a week, 50cm?) and Rhodomonas sp. (once a week, 50cm^) using material from batch-cultures described in the previous section. Temperature and salinity in the bioreactors were kept constant at 25°C and 33%o, respectively. The experiment was run for a period of 24 days. Monitoring ofthe growth of the explants was performed according to the procedures described below. Determination of projected body area. During the experiment, several photographs of the explants were made to determine changes in the projected body area. To take photographs, explants were removed from the bioreactor (kept underwater, in a beaker glass) and placed onto a rack (also underwater) on which black dots were painted to indicate a known distance (Fig. 2). Photographs were taken under a straight angle with a digital camera (Hewlett Packard Photo- Smart Model C5340A). Digital images were printed, the areas of sponges were cut out with scissors and these cuttings were weighed. FIG. 2. Photographic method. A, Overview of the system. B, Detail of the explants laying on the perspex rack. The black dots on the rack indicate a known distance. Weights ofthese cuttings were converted to areas by comparing them to the weight ofa cutting ofa known area. These values were converted to real body areas using the marked distances on the rack as areference. Photographs were taken at Days 1, 7, 10, 20 and 22. Determination of underwater weight. Underwater weight was measured using a A&D HR300 analytical balance (weighing range: 0.0001-300g) equipped with an underweighing- possibility. A hanger was connected to the balance, in which the slides with the explants could be placed. This balance was placed over a small basin filled with artificial seawater in such a manner that the part of the hanger containing the explant would remain underwater (Fig. 3). It is important to keep the level and salinity of the seawater in the basin constant. Changes in salinity will change the density of the seawater. Since sponge tissue is not much denser than seawater, a slight change in salinity will affect the underwater weight of sponges. The salinity of the seawater in the basin was always maintained at 33%o. 422 FIG. 3. Underwater weight measurement. Detail of an explant placed in the hanger under the balance. The underwater weight was calculated by subtracting the weight of the carrier slide from the combined weight of the explant + carrier slide. It was therefore important that the weight of the carrier slide remained constant throughout the experiment, especially since the slides in this study were much heavier than the explants. Therefore, an inert material was required to be used as carrier, and consequently we chose to use perspex slides instead of the commonly used glass slides (e.g. Simpson, 1963; Poirrier et al., 1981; Vethaak et al., 1982), because glass was found to dissolute slowly in seawater, causing a slow, but steady decrease of the underwater weight of the carrier slide. Explants were transported from the bioreactors to the weighing basin underwater in a beaker glass. Measurements were performed on Days 1, 2, 5, 9, 12, 22 and 24. Determination of volume, wet weight, dry weight and organic carbon and nitrogen content. In MEMOIRS OF THE QUEENSLAND MUSEUM order to evaluate the utility of underwater weight as a measure of biomass, underwater weight data were correlated to other non-destructive biomass parameters, volume and wet weight (WW). To determine volume and WW, sponge- explants were removed from the water and firmly shaken until they no longer dripped. Volume was then determined by putting an explant into a graded cylinder filled to a certain reference level with artificial seawater. After addition of the explant, all water in excess of this reference level was removed with a syringe and transferred into a lcm? glass pipette, in which the volume of excess water could be determined. In this way, sponge volumes of about 0.1cm? could be determined with reasonable precision (the methodological error was less than 10%). The corresponding WW of explants was measured on an analytical balance. For these measurements, it was imperative that explants were not attached to carrier materials, and hence, volume and WW determinations of experimental explants were undertaken immediately prior to growth experiments. Some additional explants were measured to obtain more reliable correlations/ conversion factors. Dry weight (DW) organic carbon content and organic nitrogen content of sponge tissue were also determined, but only for a single sample, since these measurements are destructive and only limited amounts of sponge material were available. For the determination of DW, pieces of sponge were dried for 24hrs in an oven at 80°C and weighed. The dried material was ground and analysed for organic carbon and nitrogen on a Fisons EA 1108 Elemental Analyser. RESULTS AND DISCUSSION GROWTH RATES AND KINETICS. Results of growth experiments are presented in Figure 4, showing changes in surface area and underwater weight. Three of the four explants showed growth during the experimental period, both when measured with two-dimensional photo- graphy and with the underwater weighing technique. The fourth explant did not show obvious changes in projected body area or under- water weight. This explant failed to attach to the perspex slide, while the other three explants firmly attached within a few days. Explants used for the experiment were made shortly before the experiment started. In future work, only healthy (attached) explants should be used, demon- strating viability after a period of acclimatisation SPONGE GROWTH MEASUREMENT explant 1 UW (mg) o — b t La PBA fem“) o e 5 a ho > by D explant 2 UW (mg) a o ~ HH [^1 A PBA forn“ [zl a o a t > be a 14 explant 3 UW (mg) oF fh Oc «d = 25 o — hm a> + PBA (cm*) D 5 10 15 20 25 days 14 explant 4 m 12 m n. E 8 \ pE Es E & > a 4 1 2 Ü 0 [tU 5 10 15 20 25 days — Underwater Weight (UW) — Projected Body Grea (PBA) FIG. 4. Results of the growth experiment. Changes in underwater weight (mg, open symbols) and projected body area (dm?, black symbols) of the four explants. Error bars for underwater weight data indicate the standard deviation of two replicate measurements. 423 in aquaria. However, some explants (e.g. explants in Fig. 3) attach to the glass only at one point, and start to form lateral processes that are not attached. These explants do grow, but this growth is difficult to quantify using photographic techniques. As these processes can easily break off from the explant, such explants are also not very suitable for growth experiments using determination of underwater weight or other weight parameters. It is not easy to deduce general statements about the kinetics of growth in P. andrewsi from the data presented in Figure 4. Explants 1 and 2 seem to exhibit a kind of lag-phase, followed by a period of exponential growth after Day 12. The lag-phase may be some kind of response to cutting sponge tissue: the tissue must rearrange and attach to the substratum before growth can start. This process may have also caused the observed decrease in underwater weight of explants 1, 2 and 4 that occurred around Day 10- 12. Rearrangement of the body into a functional, pumping sponge will cost energy that is probably obtained from respiring sponge tissue. More data are needed to give a reliable description of sponge growth in this species. It is difficult to estimate a specific growth rate for P. andrewsi under the given experimental conditions, due to the strong variability in our data. We calculated specific growth rates only for explants | and 2, based on data measured after the lag-phase. These data tend to show exponential growth, which justifies calculation of a specific growth rate u according to the formula: w=t! « In CyC, where Co is sponge biomass at the start of the exponential growth, C, is sponge biomass at the end ofthe experiment, and t is the number of days between the start of the exponential growth and the end ofthe experiment. The calculated specific growth rates (Table 2) were between 0.08-0. 10d"! for data of projected body area and 0.16d' for underwater weight. These values are consid- erably higher than previously reported growth TABLE 2. Specific growth rates (d-!) for explants 1 and 2 during the period of exponential growth. Calculations are performed with data for projected body area (PBA) and underwater weight (UW). Explant | Period used for u(PBA) WUW) 1 Days 7-22 0.08 1 Days 12-24 0.16 2 Days 7-22 0.10 2 Days 9-24 0.16 424 rates, which range from 0.01-0.058 (see Table 2, in Thomassen & Riisgard, 1995). Hence, P. andrewsi is able to grow faster under the applied conditions. PROJECTED BODY AREA VS. UNDER- WATER WEIGHT. Although general trends in results are similar between the two methods, some differences are apparent, The calculated specific growth rates for explants 1 and 2 (Table 2) are almost twice as high when underwater weight is compared to projected body area. Furthermore, data of underwater weight show a more irregular pattern than data of projected body area. The steep decrease in underwater weight for explants 1, 2 and 4 around Day 12 was not reflected in surface area. Shrinking of more massive body parts may not be reflected in changes in projected body area. Finally, the absolute growth after 22 days of the explants in projected body area is different from the growth in underwater weight (Table 3). These differences could be caused by the projection of three-dimensional growth onto two dimensional body areas. Explants | and 2 may have spread out horizontally without a corresponding increase in body mass, leading to an overestimation of actual growth. In contrast, explant 4 may have formed vertical outgrows that are difficult to quantify as increase in body area on a two-dimensional image, thus leading to an underestimation of growth. A possible improvement for the photography method would be to use so-called ‘sandwich- cultures’, flat sponge tissue cultures growing ina narrow space between a glass slide and a cover slip. This method, introduced by Ankel & Eigenbrodt (1950) to study development of freshwater sponges, was successfully applied to seawater sponges by Langenbruch (1983) and Sanchez-Moreno (1984). Sandwich-cultures can be viewed as forced two-dimensional explants, TABLE 3. Growth of the sponge explants after 22 days (projected body area and underwater weight) and 24 days (underwater weight), Growth is defined as the newly formed sponge biomass, expressed as a percentage of the initial projected body area (PBA) or underwater weight (UW). MEMOIRS OF THE QUEENSLAND MUSEUM and may thus be very suitable for growth rate measurements based on changes in projected body area. However, growth of sandwich- cultures may not mimic that of normal explants, which could be a major drawback when using this type of culture. Despite the precision of 0.1mg provided by the analytical balance, the methodological error in the weighing technique (expressed in Fig. 4 as the standard deviation of two replicate measure- ments) is usually around Img. Hence, the precision of the weighing method for determining growth rates decreases when small explants are used. A better stabilised weighing device could probably improve this precision, but it is probably more practical to work with bigger explants with an underwater weight of at least 10mg. The methodological error in photographic measurements is not shown in Figure 4. A previous study in our lab (D. Redeker, unpublished data), set up to develop the photo- graphic method, showed that this error is generally less than 10%, even when small explants are used. Images can be easily enlarged without losing too much contrast, which makes the photographic method more favourable over the weighing method when small explants are used. VOLUME AND WEIGHT PARAMETERS. Volume (V), wet weight (WW) and underwater weight were compared in order to determine conversion factors for these parameters and to evaluate the utility of underwater weight as a measure of sponge biomass. Correlations are shown in Figure 5, and the corresponding conversion factors can be found in Table 4. Both V and WW of P. andrewsi showed a moderate positive correlation (Fig. 5A; r-0.78), that is highly significant (n=11, a=0.001). In a study on Halichondria panicea, Barthel (1986) also found that the correlation between V and WW was not very strong, probably due to variability in the water- and spicule-content of sponge tissue. In contrast, we found considerably stronger TABLE 4. Wet Weight (WW), Underwater Weight (UW) and Dry Weight (DW) of lem tissue of P Increase in | crease in UW | Increase in UW andrewsi, and the percentages of Organic Carbon Explant m^ after after 22 days after 24 days Content (OCC) and Organic Nitrogen Content (ONC) SL: in the dried sponge material. Key: 1, Not significant; 1 215% - 30% 63 % 2, Reliability unknown (one sample only), 2 500 % 105 % 175 % OCC(% of | ONCA of 3 -37% 796 -21% WW(ng) | UW(mg) | DWimg) | Dw) DW) 4 605 % 745 96 735 % 0.68 0.044! 0.017 13.9 3.15 SPONGE GROWTH MEASUREMENT 0 0.05 0.1 0.15 0.2 volume (ml) UW (mg) 100 120 UW (mg) O — WM Q P O0 O - o0 0.04 0.06 0.08 0.1 Volume (ml) FIG. 5. Comparison between the three techniques for measuring biomass. A, correlation between WW and V. B, correlation between WW and underwater weight (UW). C, correlation between V and UW. correlations for Axinella polycapella (0.98) and Cinachyrella apion (0.99) (R. Osinga & E. Planas Muela, unpublished results), and for these species conversion factors are much more reliable. To evaluate the use of underwater weight as a measure of sponge biomass, the underwater weight data were compared to corresponding measurements of WW and V (Fig. 5B-C). Here, WW showed a significant (r=0.73; n=8; 070.025) correlation with underwater weight. Hence, our underwater weight data may be converted to WW, using the conversion factor given in Table 4, and underwater weight therefore seems to be an acceptable measure of 425 sponge biomass. However, no significant relation between underwater weight and V could be detected, despite the weak, but significant correlation found for volume and WW. This indicates that tissue of P. andrewsi might be subject to a large intraspecific variation in density, which implies that the other data in Table 4 (DW, organic carbon content, and organic nitrogen) must be seen as a first indication only. CONCLUSIONS We found that under the applied food regimen (batches of the microalgae Rhodomonas sp. and Chlorella sorokiniana), the sponge Pseudosuberites andrewsiis able to grow rapidly. However, we have not yet succeeded in creating artificial circumstances that enable a constant growth rate; fluctuations in time and intraspecific differences between explants were large. In further studies, this may be improved by using only those explants that have already shown the ability to grow and by adding food particles continuously using continuous cultures of algae. The two methods used in this study to determine growth have both proven their value in studying sponges. Photography of the body area is the most suitable technique when the avail- ability of sponge material is limited (i.e. when small explants are used). Determination of underwater weight is a promising alternative for photography. Underwater weight has the advantage of being a direct measure of biomass, and therefore, the accuracy of this method to measure growth may be better. The method has a detection limit of ~lmg, which makes it less suitable for small explants. More data are needed to provide a reliable picture of the relation between volume and weight parameters for tissue of P. andrewsi, as this species seems to exhibit a high variability in tissue density. ACKNOWLEDGEMENTS We thank Michael Laterveer (Blijdorp Zoo, Rotterdam) for providing us with sponge material and Dr Rob W.M. van Soest (University of Amsterdam) for identifying sponge material. Ellen M. Meijer and Boudewijn van Veen are acknowledged for their technical support and Marcel Janssen for his valuable suggestion to use underwater weighing as a method to determine sponge biomass. 426 LITERATURE CITED ANKEL, W.E. & EIGENBRODT, H. 1950. Uber die Wuchsform von Spongilla in sehr flachen Ráumen. Zoologischen Anzeiger 145: 195-204. AYLING, A.L. 1983. Growth and regeneration rates in thinly encrusting Demospongiae from temperate waters. Biologial Bulletin 165: 343-352. BARTHEL, D. 1986. On the ecophysiology of the sponge Halichondria panicea in Kiel Bight. I. Substrate specificity, growth and reproduction. Marine Ecology Progress Series 32: 291-298. BARTHEL, D. & THEEDE, H. 1986. A new method for the culture of marine sponges and its application for experimental studies. Ophelia 25: 75-82. FOSSA, S.A. & NILSEN, A.J. 1996. Korallenriff-Aquarium, Band 5. Einzellige Organismen, Schwámme, marine Würmer und Weichtiere im Korallenriff und für das Korallenriff-Aquarium. (Birgit Schmettkamp Verlag: Bornheim). LANGENBRUCH, P.-F. 1983. Untersuchungen zum Kórperbau von Meeresschwámmen. II. Das wasserleitungssystem von Halichondria panicea. Helgolander Meeresuntersuchungen 36: 337-346. LEE, Y.-K. & PIRT, S.J. 1981. Energetics of photosynthetic algal growth: influence of inter- mittent illumination in short (40 s) cycles. Journal of General Microbiology 124: 43-52. MUNRO, M.H.G., BLUNT, J.W., LAKE, R.J., LITAUDON, M., BATTERSHILL, C.N., PAGE, M.J. 1994. From seabed to sickbed: what are the prospects? Pp. 473-484. In Soest, R.W.M. van, Kempen, T.M.G. van & Braekman, J.C. (eds), MEMOIRS OF THE QUEENSLAND MUSEUM Sponges in time and space. (Balkema: Rotterdam). OSINGA, R., TRAMPER, J. & WIJFFELS, R.H. 1998a. Cultivation of marine sponges for metab- olite production: applications for biotechnology? Trends in Biotechnology 16:130-134. OSINGA, R., PLANAS MUELA, E., TRAMPER, J. & WIJFFELS, R.H. 1998b. /n vitro cultivation of four marine sponge species. Determination of the nutritional demands. Pp. 121-127. In LeGal, Y. & Muller-Feuga, A. (eds) Marine microorganisms for industry. Actes des colloques 21. (IFREMER: Plouzané). POIRRIER, M.A., FRANCIS, J.C. & LABICHE, R.A. 1981. A continuous-flow system for growing fresh-water sponges in the laboratory. Hydrobiologia 79: 255-259. SANCHEZ-MORENO, H. 1984. Cultivo experimental de dos esponjas marinas en condiciones de laboratorio. Anales Instituto de Investigaciones Marinas Punta de Betin 14: 17-28. SIMPSON, T.L. 1963. The biology of the marine sponge Microciona prolifera (Elis and Solander) I. A study of cellular function and differentiation. Journal of experimental Zoology 154: 135-152. THOMASSEN, S. & RIISGARD, H.U. 1995. Growth and energetics of the sponge Halichondria panicea. Marine Ecology Progress Series 128: 239-246. VETHAAK, A.D., CRONIE, R.J.A. & SOEST, R.W.M. VAN 1982. Ecology and distribution of two sympatric, closely related sponge species, Halichondria panicea (Pallas, 1766) and H. bowerbanki (Burton, 1930)(Porifera, Demospongiae), with remarks on their speciation. Bijdragen tot de Dierkunde 52: 82-102. PREDATION ON CARIBBEAN SPONGES: THE IMPORTANCE OF CHEMICAL DEFENSES. Memoirs of the Queensland Museum 44: 426. 1999:- The conventional view has been that the impact of predation on Caribbean reef sponges is minimal: generalist predatory fishes are deterred by sponge spicules and chemistry, while the few spongivorous fishes are *smorgasbord' feeders that circumvent chemistry by eating small amounts of many different sponge species. New data suggest that this traditional view needs to be re-examined. Generalist predatory fishes are deterred by chemistry, but not by structural elements, toughness, or nutritional quality of sponge tissue. Spongivorous fishes are not smorgasbord feeders, but instead choose to eat chemically undefended sponge species. Transplantation experiments reveal that the grazing activity of spongivorous fishes restricts certain sponge species to refugia, including cryptic habitats on the reef and mangrove and grassbed environments, where these fish are absent. Chemical defense plays an important role in the ecology of sponges on Caribbean reefs. 0 Porifera, chemical defense, predation, Caribbean, ecology. Joseph R. Pawlik (email: pawlikj@unewil.edu ), Biological Sciences, UNC-Wilmington, Wilmington, NC 28403-3297 USA; 1 June 1998. RELATIONSHIP BETWEEN SPONGES AND A TAXON OF OBLIGATORY INQUILINES: THE SELIQUARIID MOLLUSCS MAURIZIO PANSINI, RICCARDO CATTANEO-VIETTI AND STEFANO SCHIAPARELLI Pansini, M., Cattaneo-Vietti, R. & Schiaparelli, S. 1999 06 30: Relationship between sponges and a taxon of obligatory inquilines: the siliquariid molluscs. Memoirs of the Queensland Museum 44: 427-438. Brisbane. ISSN 0079-8835. Some coenogastropod molluses are adapted to living embedded in a matrix of sediment, coral or sponge tissue. In the latter case the siliquariid molluscs are obligatory inhabitants of the sponge hosts. Siliquariidae is a small family with three extant genera with circum-tropical and temperate distribution. Information on theirassociation with Porifera is so far limited to L5 records, The present study analyses 35 sponge species hosting siliquariids from the Mediterranean Sea, E. Atlantic, New Zealand, Philippines and New Caledonia, living in a water depth between 10-440 m. A close species-specific association was not found, although only a restricted number of sponge families host siliquariid molluscs, From these data itis apparent that siliquariids prefer hosts with a compact and rigid sponge skeletal structure, produced by a radial organisation and/or high spicule density. Commensal siliquariids show different growth rates. When their larvae settle on the sponge surface larval shells (protoconchs) are partially overgrown by the host sponge. As soon as the mollusc begins development it opens a slit along its entire length, hence commencing close interactions with the sponge. The mollusc is able to modify the shape of the longitudinal slit, adapting it to ihe sponge aquiferous system by transforming the slit into a series of contiguous holes that communicate with the sponge's excurrent canals, Based on the trend that there is a successively decreasing diameter of these canals, it seems evident that the siliquariid conveys selfdrained water into ihe sponge incurrent canal system. This behaviour was studied using x-ray photography and casts obtained from resin injections into the aquiferous system. It is clear that the mollusc obtains most benefit from this association, achieving: protection against predators, defence from sediment clogging, and increased feeding etticiency. Minor benefits are obtained by the sponge host: increased water inflow with an energy saving, and a secondary source of food from the mollusc's expelled water. The sponge does not seem to be negatively affected hy the siliquariid presence and is able to maintain, through its plasticity, its original skeletal structure, This form of strict and integrated association between filler-feeders may well be interpreted as commensalism and probably as facultative mutualism. O Porifera, siliquariid molluses, association, commensalism, symbiosis, adaptations, behaviour. Maurizio Pansini (email: zoologia(aigecuniv cisi.unige.il), Istituto di Zoologia, University af Genova, Via Balbi, 5, 1-16126, Genova, Italy; Riccardo Cattaneo-Vietti & Stefana Sehiaparelli, Istituto di Scienze Ambientali Marine, University af Genoa, Viale Benedetta XV 5, 1-106132 Genova, laly; 18 January 1999, Symbiotic associations between sponges and other organisms involve a diversity of taxa, from bacteria to large crustaceans (see review by Sard et al, 1998), whereas other associations, less integrated or intimate, involve other sponge dwellers which, according to cases, may be regarded as commensals or inquilines. Most studies on these latter associations (e.g. Pearse, 1932; Pansini, 1970; Bacescu, 1971; Rützler, 1975; Koukouras, 1996) have examined ihe descriptive aspects of the association whereas only a few have focused on the interaction between host and commensal (Forbes, 1966; Connes et al., 1971; Uriz et al., 1992). Concerning molluscan commensals, natural suspension feeders such as bivalves are most frequently associated with sponges (e.g. Chlamys varia in Halichondria panicea (Porester, 1979), Hyatella artica in Geodia cydonium (Santucci, 1922), Ostrea permollis in Stelletta grubii (Forbes. 1966)), whereas associations between sponges and gastropods are rare and restricted only to sessile gastropods with a filter feeding strategy that requires important shell adaptations (Seilacher & Gunji, 1993; Savazzi, 1996). The shells of gastropods living embedded in sponges do not exhibita fixed geometrical constraint but a ‘heteromorph’ growth pattern (Morton, 1951, 1955; Gould, 1966; Savazzi, 1996; Schiaparelli 428 et al., 1998), as seen in several Vermiculariinae and in all slit-bearing Siliquariidae (genera Tenagodus and Pyxipoma). The obligatory association of these molluscs with sponges is supported by the absence of any scar on the shells from attachment to a substrate (Deshayes, 1864; Savazzi, 1996). Slit-bearing Siliquariidae represent a separate unit from other uncoiled gastropods (as Vermetidae, Vermiculariinae and Stephopoma, the third siliquariid genus), due to the presence of the shell’s longitudinal slit that drains the incoming water-flow from the shell aperture. In the adult mollusc the slit is only partially open and func- tional, with partial closure due to secondary carbonate deposition. During its growth the living mollusc shifts its position along the shell aligning the mantle cavity opening with the func- tional apertures of the slit. As for most ciliary feeders the water incurrent flow is produced by means of cilia on numerous filamentous gills; these gills also retain food particles (Morton, 1951). According to Bieler (1992, 1996) three genera are presently included in the family Siliquariidae: Pyxipoma March, 1861, with a short longitudinal slit and a smooth shell; Tenagodus Guettard, 1770, with a longer slit and either a smooth or a spiny shell; and Stephopoma Mirch, 1861, which is devoid of slit and shows a vermetid-like ecology. Siliquariid molluscs have a wide range of shell coiling patterns with whorls developing either on asingle or on several planes. In addition, some Tenagodus species are able to produce a series of transversal cracks in their smooth shells, thus allowing adjustment to the curvature of shell coils (Savazzi, 1996). This capacity to modify adult shell shape is unique amongst molluscs (Savazzi, 1996). Little is known about the biology, ecology and geographical distribution of these obligate sponge dwellers because most reports in the liter- ature concern descriptions of empty shells or shell fragments (Bieler & Hadfield, 1990). Virtually nothing is known about their repro- duction, larval longevity and dispersal capacity. In addition, siliquariids are rather rare and live mainly at considerable depths. For these reasons, their associations with host sponges have never been extensively studied: only Morton (1955) remarked on the association between the mollusc slit and the sponge aquiferous system, and Savazzi (1996) hypothesised the existence of a MEMOIRS OF THE QUEENSLAND MUSEUM unidirectional water outflow from the mollusc into the sponge aquiferous system. This study aims to clarify the different aspects of the sponge-siliquariid association, notwith- standing the lack of access we have to living material available for study. More specific topics involving the functional morphology of siliquariid molluscs are treated in a separate paper (Schiaparelli et al., in preparation). MATERIALS AND METHODS To date only 35 sponge specimens associated with siliquariid molluscs were studied, collected from the W Mediterranean, E Atlantic and the Pacific area including Philippines, Japan, New Caledonia and N New Zealand. Most of this material was collected by the Muséum National d'Histoire Naturelle, Paris from several deep- water expeditions, and kindly trusted us for study. Several small lots of specimens were also obtained from other sources (Museo di Zoologia of Bologna and private collections). Most of the studied material comes from relatively deep waters (10-550m depth). All the massive sponge specimens with siliquariids embedded in their bodies were carefully studied in toto. X-ray photographs of some Penares and Spongosorites specimens were made using health diagnostic X-ray facilities, to ascertain the distribution and align- ment of molluscs within the sponge body. Casts of the water flow routes were made by injecting a setting resin into both the shell’s main aperture and the sponge’s oscule in alcohol preserved specimens and, after the compound had set, by dissolving the sponge body by soaking it in HCI (Bavestrello et al., 1988). Much better results would have been obtained through in situ applic- ation of resin into living specimens, but this has not yet been possible due to the restricted material available to us. Sub-samples of the two associated organisms were then separated for species identification. Spicules were prepared by dissolving pieces of sponge in nitric acid in a vial, then dehydrated and mounted either on slides with Eukitt resin or directly on stubs. The skeletal arrangement was studied by hand cut (tangential and transversal) sponge sections. The abundant siliquariid material available, allowed us to leave specimens intact to study in toto and to dissect and prepare the main diag- nostic parts for ultramicroscopy (protoconchs, SPONGE COMMENSAL SILIQUARIID MOLLUSCS 429 TABLE 1. Literature of sponges associated with siliquarid molluscs. References refer to papers reporting whole animals, not just empty shells. Sponge Species Mollusc Species Locality Depth Reference tw amorphus Burton, unidentified Siliquariid South Africa - Burton, 1926 Epulus burtoni Lévi & Lévi, | identified Siliquariid New Caledonia 425-430m Lévi & Lévi, 1983b Erylus carteri Sollas, 1888 unidentified Siliquariid Gulf of Manaar - Lévi & Lévi, 1983b Row ge dioides Burton & unidentified Siliquariid Mergui Archipelago 119m Burton & Rao, 1932 Erylus nigra Bergquist, 1968 | — unidentified Siliquariid New Zealand 129m Bergquist, 1978 Erylus proximus Dendy, unidentified Siliquariid Cargados 55m Dendy, 1916 CN schulzei (Dendy, unidentified Siliquariid New Caledonia, Ceylon 182-430m Dendy, 1905 Penares sp. Pyxip e T MD New Zealand - Morton & Miller, 1968 Racodiscula sceptrellifera Tenagodus cumingii A (Carter, 1881) (Mörch, 1861) Indian Ocean 27-55m Annandale, 1911 Racodiscula sceptrellifera Tenagodus trochlearis ; (Carter, 1881) Mörch, 1861 Indian Ocean - Annandale, 1911 Sili : EN . " d ingii Mórch, . : Hog ino n BE J pone Md uad iRel si SN Japan Intertidal Hoshino, 1981 Spongosorites topsenti Tenagodus muricatus (Born Indian Oc 55-69 Annandale. 1911 endy, 1905 1778) an Ocean m andale, ME ede ruetzleri (Van | unidentified Siliquariid Barbados 108-153m Wan Soret dk Stedt, arte «ili M Van unidentified Siliquariid Barbados, Jamaica 108-170m Van i Sentot, Unidentified (?) sponge Fray aie o did Dall, Bermuda - Dall, 1881 ; : Tenagodus obtusus i 2 " Unidentified (?) sponge (Schumacher, 1817) South Africa Barnard, 1963 Unidentified sponge forcpama fecus Lament, Indian Ocean x Morch, 1860 Unidentified sponge Pyxipoma weldii (Tennison New Zealand - Morton, 1951 Woods, 1876) 7 Unidentified sponge RS Ta Philippines 2-3m Savazzi, 1996 denn Tenagodus armatus 1 Unidentified sponge Kuroda et al., 1971 Japan 50-100m Kuroda et al., 1971 Unidentified sponge Tenagodus bernardi Morch ? Senegal - Gould, 1966 Unidentified sponge Tenagodus chuni Thiele, South Africa 40-155m Barnard, 1963 Unidentified sponge Tenagodus cumingii Motch, Philippines - Morch, 1860 Unidentified sponge Tenagodus cumingii Morch, Western Pacific 10-100m Kuroda et al., 1971 ; ; Tenagodus obtusus : ae Unidentified sponge (Schumacher, 1817) Mediterranean - Philippi, 1836 : R Tenagodus squamatus Unidentified sponge Blainville, 1827 Bermuda | 549-732m Gould, 1966 : ; Tenagodus squamatus Unidentified sponge Blainville, 1827 Bermuda 732m Abbott, 1974 : ; Tenagodus wilmanae ; Unidentified sponge Tomlin, 1918 South Africa 150m Kenseley, 1973 : : Tenagodus wilmanae : Unidentified sponge Tomlin, 1918 South Africa - Barnard, 1963 430 opercula and radulae). A Philips 515 microscope was used for SEM observations. RESULTS Twenty nine records of sponges associated with siliquariids have been recorded in the liter- ature (Table 1), but the identification of both partners was complete only in five cases. Never- theless, from these data, 13 sponge species in total, belonging to 6 genera and 5 families (Ancorinidae, Coppatiidae, Geodiidae, Halich- ondriidae, Theonellidae) have been identified (Table 1). By comparison, in the present study, preliminary identifications of 35 sponge specimens associated with siliquariids differ- entiated 19 sponge species belonging to the same 5 families cited above (Table 2), in addition to a fragment of an unidentified horny ‘keratose’ sponge. Siliquariids studied belonged to 6 species of the genus Tenagodus and to a single species of Pyxipoma (Table 2). The taxonomic part of the study, including the description of several new species of both sponges and siliquariids, will be the object of future papers. Geographic and bathymetric distributions of material showed that 5 specimens were from temperate and 30 from tropical regions, and 34 specimens out of 35 were collected at more than 50m depth (Table 2). A similar trend is shown in literature, with 11 temperate and 18 tropical records and 12 specimens out of 17, with known depths of collection, coming from waters deeper than 50m (Table 1). In all cases but one the sponge specimens hosted a variable number of molluscs belonging to a single species. The exception is a sponge specimen from New Caledonia, collected around 230m depth, and tentatively attributed to the genus Epipolasis, that hosted two species of spiny Tenagodus that were also recorded in association with other sponge species, indicating that their association is facultative. Some siliq- uariids may be associated with as many as six different sponge species, as the case of the two spiny Tenagodus (Tenagodus sp.5 and T. cf. anguinus) (Table 2). Different specimens of Holoxea furtiva Topsent from distant localities hosted slightly different siliquariid species. Two Mediterranean specimens from Sardinia and Tunisia were assoc- iated with Tenagodus obtusus, whereas a specimen from Cabo Verde hosted 7. senegalensis. Different specimens of the same sponge species collected in the same area (e.g. MEMOIRS OF THE QUEENSLAND MUSEUM Spongosorites cf. solomonensis, Spongosorites sp.3 and Topsentia sp.1) may host different siliq- uariid species (Table 2). Considering the five families of sponges that host these siliquariids, three types of skeletal patterns were distinguished: a radial structure in Ancorinidae, Coppatiidae and Geodiidae; a disordered structure in Halichondriidae, and the usual articulated, solid ‘lithistid’ structure in Theonellidae. Analyzing the distribution of siliq- uariids belonging to the genus Tenagodus, which has species with either smooth or spiny shells, we found a remarkable correlation with the sponge skeletal architecture. 1) Smooth Tenagodus species were always associated with sponges that had radial structure (Table 2). These molluscs were completely embedded in the sponge body with only the shell apertures protruding from the sponge surface (Fig. 1A). X-ray photographs showed that the direction of shell growth is straight, determined largely by the radial pattern of the sponge skeleton (Fig. 1B). According to the position of the shell apertures, which are almost flush with the sponge surface, it may be inferred that the growth rate of the associated organisms is nearly the same. This behaviour was observed only in small and medium sized siliquariid species with smooth shells. 2) Conversely, spiny Tenagodus species were always associated with sponges having disorderly arranged skeletons (Table 2). In these cases the molluscs were not completely embedded in the sponge body because part of the shell laid on the sponge surface (Fig. 1C). X-ray photographs showed that spiny siliquariids shorten as much as possible the ray of curvature of their first coils, and that a precise direction of shell growth cannot be defined (Fig. 1D). Shell uncoiling is more accentuated towards the sponge surface. The mollusc growth rate certainly exceeded that of the sponge when the shell develops on the host surface. The same behaviour was observed in large size smooth Tenagodus specimens (Fig. 2B), which, instead of laying on the sponge surface, raise the terminal part of their shells (Schiaparelli et al., in prep.). Siliquariids associated with Theonellidae were completely entrapped among desmas. Here they are so constrained by the rigid skeletal structure that they are unable to transversally crack their shells, varying their shape as described by Savazzi (1996), and consequently obliged to grow very irregularly. Siliquariid protoconchs (larval shells which SPONGE COMMENSAL SILIQUARIID MOLLUSCS 43] er eres FIG. 1. Associations between siliquariids and sponges. A, Specimen of Penares intermedia (Dendy, 1905): the shell apertures are indicated by arrows, whereas the oscules are marked by stars. B, X-ray photograph of the same specimen of P. intermedia showing the associated siliquariids (Tenagodus sp. 4): the arrows mark the axes of two coiled shells. C, Specimen of Spongosorites sp.1 associated with Tenagodus cf. anguinus. D, X-ray photograph of Spongosorites sp.1. E-F, Two aspects of an aquiferous system cast of a Topsentia sp.1 specimen associated with Tenagodus cf. anguinus. 432 TABLE 2. List of sponge species associated with siliquarid molluscs. Different shades of grey refer to the sponge MEMOIRS OF THE QUEENSLAND MUSEUM skeletal structure and to the external morphology of the shells. Specimen Sponge Family Sponge Species Mollusc Species Locality Depth Discodermia cf T Lus Me] SI 23/30 Theonellidae laevidiscus Carter, | — Tenagodus sp. 4 Philippines 92-97m 1880 Va E SI 13/31 Erylus sp. nov. Tenagi odus sp. 4 A New Caledonia 234-242m S125 Erylus sp. nov. Tenagodus sp. 4 = Philippines 92-97m Erylus nigra BT Se a S1 33 Bergquist, 1968 Tenagodus uae | New Caledonia 415m Geodia cf. parasitica Tenagodus F sis Bowerbank, 1873 | senegalensis Senegal Holoxea furtiva ; ; : " SET Topsent, 1892 | Tenagodus obtusus | Italy SI 1/2 Holoxea furtiva | Tenagodus obtusus Tunisia 9; 32m Holoxea furtiva | Tenagodus o x S13 Topsent, 1892 LU es Cabo Verde 55-60m S197 Jaspis sp. seria (?) Australia - Penares intermedia : 81 12/34 E (Dendy, 1905) E Tenagodus sp.4 New Caledonia 430m SI 19/26 Penares sp. nov. Tenagodussp.4 | New Caledonia 270-300m S120 Penares sp. 1 Tena, «4 Philippines 92-97m S135 Penares sp. 2 ia weldii New Zealand - SI 8 Halichondriidae (?) Epipolasis sp. New Caledonia 234-242m -T- Spongosorites cf. SI 15 Halichondriidae | salomonensis Dendy, New Caledonia 243m ou. tt 1921 i Spongosorites cf. SI 32 Halichondriidae | salomonensis Dendy, New Caledonia 440m ^ 1921 817 Halichondriidae | SPongosories sp. New Caledonia 242m S116 Halichondriidae | Spongosorites sp. New Caledonia 300m SI9 Halichondriidae Spongosorites sp. 1 | New Caledonia 270-300m SL17 Halichondriidae Spongosorites sp. 1 New Caledonia 260m S127 Halichondriidae | Spongosorites sp. 1 New Caledonia 237-550m SI 18 Halichondriidae | Spongosorites sp. 2 New Caledonia 397-439m SI 24 Halichondriidae | Spongosorites sp.3 Philippines 92-97m S129 Halichondriidae — | Spongosorites sp. 3 Philippines 183-187m SIII . Halichondriidae Topsentia sp. nov. New Caledonia 233m SI 14 Halichondriidae | Topsentia sp. 1 Philippines 186-187m SI 21 . Halichondriidae Topsentia sp. | Philippines 92-97m ^ i S122 -Halichondriidae Topsentia sp: 1 Philippines 92-97m SI28 Halichondriidae | Topsentia sp. 2 New Caledonia 410-440m S16 3 Ragiieutota horny i Tenagodus maoria New Zealand - ` sponge, dark violet | ' i Sponges with Sponges with Siliquariids Siliquariids radial skeletal disordered with smooth with spiny growth skeletons shells shells SPONGE COMMENSAL SILIQUARIID MOLLUSCS 433 are separated by a boundary (concave septum) from the adult shells, called teloconchs), have been observed on the surface of several sponge specimens. According to characteristics of their coils (number and size), they belong both to planctotrophic and lecitotrophic species (Schetelma, 1978). Planctotrophic larvae have been observed in Tenagodus senegalensis to settle preferentially near the mollusc slit, where they probably find the most suitable water- movement conditions. Recruits are covered by the host sponge and develop in the remaining space between the adult siliquariids. The functional associations between sponges and associated molluscs were also ascertained by the study of casts. Resin injected into the oscule of a specimen of Topsentia sp.l containing Tenagodus cf. anguinus came out from the main aperture of the shell and vice versa (Fig. 1F). Casts show that the water pushed by the mollusc ciliary movement seeps through the slit (Fig. 1E) and enters the sponge aquiferous system. There is a reciprocal morphological adaptation of the two associated organisms because the sponge moulds its aquiferous system on the continuous slit aperture and then the mollusc divides this simple slit into a series of holes (Fig. 2AD). Spicule tracts correspond to the carbonate pillars separ- ating the holes (Fig. 2CD). The wide aquiferous system canals (0.6mm diameter) conveying water from the mollusc into the sponge, fit perf- ectly with the slit holes (Fig. 1E). Thereafter these canals divide either dichotomously or by emitting transverse branches (Fig. 1 F). Their size decreases continuously up to a minimum detect- able diameter of 0.1 mm. Casts, however, are interpreted as single moments of a continuous growth process which involves both the partners in the association. Particularly important is the shifting ofthe living mollusc, as far as it grows, towards the shell opening, which causes the moulding of a new part of a functional slit. The growth process also determines a rapid closure by carbonate depos- ition of the non-functional slit apertures behind the mollusc body: holes in Tenagodus (Fig. 2E) and a continuous slit in Pyxipoma (Fig. 2F). It was also observed that whenever an open part of the slit accidentally lost its sponge covering it was immediately closed by the mollusc. DISCUSSION A siliquariid mollusc living within a sponge has three primary needs: 1) to be at least partially covered by the sponge in order to get support and protection; 2) to have the water-outflow drained through the sponge body; and 3) to maintain the shell opening free for the water-inflow, laying on the sponge surface or variously raised. Such requirements may be fulfilled only by sponges with peculiar characteristics, as demonstrated by the restricted number of sponge taxa currently known to host siliquariids. Some of these charac- teristics may be tentatively identified as: a massive growth form, assuring an adequate volume to host the molluscs; and a solid structure, generally bound to a high spicule content, contributing to maintain a constant space ratio between the associated organisms, in order to guarantee a plain water outflow. A soft, elastic sponge which continually moves is probably less adapted to maintain a steady association — involving the aquiferous system - with a host dwelling in a rigid shell. The host molluscs, however, display a remarkable adaptive capacity to different situations as demonstrated by the fact that two species of siliquariids were found assoc- iated with six different sponge species. As a rule each siliquariid species colonises a single species of sponge (with the exception of the New Caledonian sponge mentioned above), with the number of successful mollusc recruits related to sponge size. Siliquariid recruitment is either through lecithotrophic larvae, which develop in situ, or planktotrophic larvae that are released into deep, relatively still waters, and probably do not disperse over large areas. Several factors may favour the recruitment of young siliquariids in this association. One of these is the combined pumping activity of the sponge and associated molluscs, that produces a water current from the surroundings towards the sponge surface which may attract the swimming larvae. According to our observations it seems probable that the association between sponges and siliquariids is not species specific. This hypo- thesis is supported by the behaviour of Holoxea furtiva, a sponge with a wide geographic distrib- ution that hosts two siliquariid species in geographically distant localities. Similarly, different specimens of the same sponge species in the same geographic locality host different siliq- uariids (e.g. Spongosorites cf. salomonensis, Spongosorites sp. 3 and Topsentia sp. 1), also support this contention. From present knowledge the association between sponges and siliquariids seems to be relatively frequent in the tropics and in deeper waters — where the latter taxon is more abundant — MEMOIRS OF THE QUEENSLAND MUSEUM FIG. 2. Associations between siliquariids and sponges. A, Specimen of Tenagodus with a slit divided into holes. B, Large, smooth specimens of Tenagodus senegalensis with a continuous slit and the terminal part of the shell uncoiled. C, Portion ofa Jaspis sp. specimen overgrowing a siliquariid slit. D, Formation of holes in the slit ofa Tenagodus specimen by carbonate denticulation. E, Closure of a slit hole in a Tenagodus specimen by a carbonate lunula during the molluscan growth. F, Continuous slit closed in a specimen of Pyxipoma weldii. SPONGE COMMENSAL SILIQUARIID MOLLUSCS but these conclusions are based on restricted samples, and future surveys of the shallow-water areas might alter this presumed distribution pattern. The relationship between the shell morphology (smooth or spiny) and the sponge skeletal archi- tecture (radial or disordered) is particularly strong. Of possible hypotheses to explain why smooth siliquariids are associated with radially structured sponges, and spiny ones with dis- orderly arranged skeletons, the most consistent seems to be that the choanosomal space is so reduced due to high spicule density, and the physical constraints so strict due to the presence of radial spicule tracts, that only smooth shells can adapt to the radial structure. Smooth siliq- uariids, in fact, which have transversally cracked shells, may adapt their shells to extremely confined spaces by changing the curvature of their coils. However, when smooth Tenagodus specimens are tightly entrapped into an articul- ated 'lithistid' desma reticulation (e.g. Discodermia), they are unable to modify their shell shape and must change their normal growth habit becoming uncoiled (Schiaparelli et al., in prep.). Spiny siliquariids, on the contrary, which very rarely show shell cracks, cannot modify their shape in order to adapt to very hard sponges and are associated with disorderly arranged skeletal structures (such as those found in Halichondriidae), where the available space inside the skeleton is certainly wider. Since the main factor that forces all siliquariids to live permanently within sponges is the demand for protection against predators (Vermej, 1987), the production of spines by these molluscs may be interpreted as a reaction against an inadequate protection from the host sponge. The fact that the last coils of spiny species lay uncovered on the sponge surface is due to the mechanical protec- tion offered by the spines against muricid molluscs which, according to the shape of perfor- ations (Carriker & Jockelson, 1968), seem to be the most common siliquariid predators. Smooth species, on the contrary, are much more vulner- able, as demonstrated by the high number of unsuccessful muricid holes (in the uncovered shell portions) and by the attitude to close, by means of a calcareous lamina, every portion of their slit accidentally left uncovered by the host sponge. The prompt responses shown by siliquariids to new situations, together with the ability to shift their position along the shell during growth, determine a continuous and complex variation of the slit morphology (Schiaparelli et al., in prep.). Casts show that close relationships are estab- lished with the sponge aquiferous system to obtain an effective drain of water pumped by the mollusc. Water entering the shell aperture is pushed by ciliary beating through the slit towards the sponge canals, thus obtaining an obligate flow direction. The sponge does not try to clog the slit but, on the contrary, seems to mould its skeletal structure on it. The mollusc, on the other hand, is able to modify the slit shape by forming holes that correspond exactly to the sponge canals. The dichotomous branching and ever- decreasing diameter of canals, even if negligible as an absolute figure given its variation between specimens (Bavestrello et al., 1988), are typical of the sponge incurrent system (Bavestrello et al., 1990, 1995). Therefore the sponge receives from each associated siliquariid a continuous water flow. CONCLUSIONS Associations between sponges and siliquariids are examined in terms of benefits and disadv- antages for either partner. A sponge associated with siliquariids may obtain two major benefits: 1) a considerable energy-saving for the pumping activity, due to the water flow pushed by the mollusc; 2) an additional food supply coming from the fine edible particles that a gastropod ctenidium is unable to hold. The presence of shells, on the contrary, could be an obstacle to the formation of the normal skeletal frame, but the sponge plasticity seems to easily overcome this constraint, Siliquariids may obtain a greater amount of benefits from their association with sponges: 1) an effective defence from predators, which is certainly mechanical and possibly chemical (in the latter case, however, the mollusc should have developed a form of resistance to the sponge bioactive products, with the assumption that the association between both partners has a signi- ficant evolutionary history); 2) in terms of space the sponge primarily offers the mollusc a steady platform of support, even on unstable, detritic bottoms, and secondly a raised position, less disturbed by the sediment, which may confer trophic advantages to the filter-feeder; and 3) the sponge pumping activity, determining a contin- uous water flow towards the sponge surface, certainly brings food particles that the siliquariid can consume and, possibly, even attracts molluscan larvae. 436 ‘Theoret ically, the main potentia! disadvantage for a siliquariid associated with a sponge 1s the risk of being killed by the host growth over- whelming its shell apertures, However, even if some sponges, under special conditions, are able to increase their growth rates several-fold (Ayling, 1983), such cases of overwhelming by the host would probably occur very rarely, because the growth of the terminal parl of the shell, bearing the aperture, is probably very rapid, especially in spiny species. Finally, siliquariid molluscs have developed such basic structural adaptations to live in assoc- iation with sponges (sce also Schiaparelli et al, in prep.), that the obligatory nature of their relation- ship is a logical and predictable consequence, On the contrary, sponges may or may not react negatively to the ‘invasion’ or colonisation of their body by siliquariids, but they are surely getting remarkable benetits in terms of food, and are obviously tree to live without associated siliq- uariids. The association, therefore. may be viewed as a form of commensalism and probably even of facultative (for the sponges at least) mutualism, LITERATURE CITED AYLING, A.L. 1983. Growth and regeneration Rates in thinly encrusting Demospongiae from temperate waters, Biological Bulletin 165; 343-352. ABBOTT, R. T. 1974, American seashells: the marine Mollusca of the Atlantic and Pacific coasts of North America. 2nd edition (Van Nostrand Reinhold Co.: New York). ANNANDALE, N. 1911, Some sponges associated with gregarious Molluscs of the family Vermetidae. Records of the Indian Museum 6; 47-55. BACESCU, M. 1971. Les Spongiaires, un des plus intéressants biotopes benthiques marins, Rapports de la Commission internationale pour l'explor- ation scientifique de la Mer Méditerranée 20( 39); 239-241. BARNARD, K. H. 1963, Contributions to the knowledge of South African marine Mollusca. Part II. Gastropoda: Prosobranchiata: Taenio- alossáta, Annals of the South African Museum 47: 144-155, . BAVESTRELLO, G., BURLANDO, B. & SARA, M. 1988, The architecture of the canal systems of Petrosia ficiformis and Chondrosia reniformis studied by corrosion cast (Poritera, Demo- spongiae), Zoomorphology 108: 161-166, 1990. The aquiferous system of Spongia officinalis and Cliona viridis (Porifera) based on corrosion cast analysis. Bollettino di Zoologia 57; 233-239. 1995. Corrosion cast reconstruction of the thred- dimensional architecture od demosponge canal MEMOIRS OF THE QUEENSLAND MUSEUM systems. Pp. 93-110, In Lanzavecchia, G., Valvassori, R. & Candia Carnevali, M.D. (eds) Body cavities: function and phylogeny. Selected Symposia and Monographs U.Z.I. Vol. & (Mucchi: Modena). BAVESTRELLO, G., CERRANO, C., CORRIERO, G. & SARÀ, M. 1998. Three-dimensional archi- tecture of the vanal system of some Hadromerids (Porifera, Demospongiae). Pp. 235-247. In Watanabe, Y. & Fusetani, N, (eds) Sponge Sciences. Multidisciplinary Perspectives. (Springer Verlag; Tokyo). BERGQUIST, P.R. 1978. Sponges. (Hutchinson: London). BIELER, R. 1992. Tenagodus or Siliquaria? Unraveling taxonomic confusion in marine "Worm-Snails" (Cerithioidea: Siliquariidac). The Nautilus 106(1): 15-20. 1996, Mürch's worm snail taxa (Caenogastropoda, Vermetidae, Siliquariidae, Turritellidae). American Malacological Bulletin 13( 1-2): 23-35. BIELER, R. & HADFIELD, G. 1990. Reproductive biology of the sessile Gastropod Vermicularia spirata (Cerithioidea: Turritellidae), Journal of Molluscan Studies 56: 205-219. BURTON, M. 1926, Descriptions of South African sponges collected in the South African marine survey. Part I. Myxospongida and Astro- tetraxonida. South Africa Fisheries and Marine Biological Survey 4; 1-29, BURTON, M. & RAO, HLS. 1932. Report on the shallow water marine sponges in the collection of the Indian Museum. Part I. Records of the Indian Museum 34: 299-356, CARRIKER, M. R. & YOCHELSON, E.L. 1968. Recent gastropod boreholes and Ordovician cylindrical borings. U.S. Geological Survey Professional Papers 593-B: B1-B23. CONNES, R., PARIS, J. & SUBE, J. 1971. Réactions tissulaires de quelques Dérnosponges vis-à-vis de leurs commensaux et parasites. Naturaliste Canadien 98: 923-935, DALL, W. H. 1881-1882. Preliminary report on the Mollusca. (Blake Reports, No 15.). Bulletin of the Museum of Comparative Zoology 9/29): 30, DENDY, A. 1905, Report on the ponies collected by Prolessor Herdman. at Ceylon in 1902. Pearl Oyster Fisheries, Supplementary Report 18: 57-246. 1916. On the Homosclerophora and Astro- tetraxonida collected by H.M.S. Sealark in the Indian Ocean. Transactions of the Linnean Society of London Zoology 17: 225-271. DESHAYES, G. P. 1864, Description des Animaux sans Verlebres Découverts dans le Bassin de Parts. Pp. 278-299. Tome Deuxième (Texte et Atlas) (Bailliere J.B, et Fils: Paris). FORBES, MLL, 1966, Life cycle of Ostreg permollis and ils relationship to the host sponge, Stelletta SPONGE COMMENSAL SILIQUARIID MOLLUSCS grubii. Bulletin of Marine Science Gulf Caribbean 16: 273-301. FORESTER, A. J. 1979. The association between the sponge Halichondria panicea (Pallas) and the scallop Chlamys varia (L.): a commensal- protective mutualism. Journal of Experimental Marine Biology and Ecology 36: 1-10. GOULD, S.J. 1966. Notes on shell morphology and classification of the Siliquariidae (Gastropoda). The protoconch and slit of Siliquaria squamata Blainville. American Museum Novitates 2263: 1-13. HOSHINO, T. 1981. Shallow-water Demosponges of Western Japan, II. Journal of Science of the Hiroshima University (B, 1, Zoology) 29(2): 256-257. KENSLEY, B. 1973. Seashells of Southern Africa. Gastropods. (Maskew Miller Limited: Cape Town). KOUKOURAS, A., RUSSO, A., VOULTSIADOU- KOUKOURA, E., ARVANITIDIS, C. & STEFANIDOU, D. 1996. Macrofauna associated with sponge species of different morphology. Marine Ecology 17(4): 569-582. KURODA, T., HABE, T. & OYAMA, K. 1971. The seashells of Sagama Bay. (Biological Laboratory _ Imperial Household, Marunzen: Tokyo). LEVI, C. & LEVI P. 1983a. Démosponges bathyales récoltées par le N/O ‘Vauban’ au sud de la Nouvelle-Calédonie. Bulletin du Muséum National d'Histoire Naturelle, Paris (4) 5(A, 4): 931-997. 1983b. Eponges Tétractinellids et Lithistides bathyales de Nouvelle-Calédonie. Bulletin du Muséum National d'Histoire Naturelle, Paris (4) _ X(A, 1): 101-168. MORCH, O.A.L. 1860. Review ofthe genus Tenagodus Guettard. Proceedings of the Zoological Society of London (28): 400-415. MORTON, J.E. 1951. The structure and adaptations of the New Zealand Vermetidae. Part II. The genera Stephopoma and Pyxipoma. Transactions of the Royal Society of New Zealand 79: 1-51. 1955. The evolution of Vermetid Gastropods. Pacific Sciences 9: 3-15. MORTON, J. & MILLER, M. 1968. The New Zealand Sea Shore. (Collins: London, Auckland). PANSINI, M. 1970. Inquilinismo in Spongia officinalis, Ircinia fasciculata e Petrosia ficiformis della riviera ligure di levante. Bollettino dei Musei e degli Istituti Biologici dell'Università di Genova 38: 5-17. 437 PEARSE, A. S. 1932. Inhabitants of certain sponges at Dry Tortugas. Papers from Tortuga Laboratories. Carnegie Institution of Washington 28(77): 117-124. PHILIPPI, R.A. 1836. Enumeratio Molluscorum Siciliae cum viventium tum in tellure tertiaria Fossilium quae in itinere suo observavit. (Berolini. Sumptibus Simonis Schroppi et . Sociorum: Berlin). RUTZLER, K. 1975. Ecology of Tunisian commercial sponges. Tethys 7(2-3): 249-264. SANTUCCI, R. 1922. La Geodia cydonium come centro di associazione biologica. Memorie del Regio Comitato Talassografico Italiano 103: 1-20. SARA, M., BAVESTRELLO, G., CATTANEO- VIETTI, R. & CERRANO, C. 1998. Endosymbiosis in sponges: relevance for epigenesis and evolution. Symbiosis 25: 57-70. SAVAZZI, E. 1996. Adaptations of Vermetid and Siliquariid Gastropods. Palaeontology 39(1): 157-177. SCHETELMA, R.S. 1978. On the relationship between dispersal of pelagic veliger larvae and the evol- ution of marine prosobranch gastropods. Pp. 303-322. In Battaglia, B. & Beardmore, J.A. (eds) Marine Organisms. (Plenum: New York). SCHIAPARELLI, S., CATTANEO-VIETTI, R. & PANSINI, M. 1998. Relationship between Siliquariidae (Mollusca: Caenogastropoda) and sponges. P. 295. In Bieler, R. & Mikkelsen, P.M. (eds) 1998 World Congress of Malacology, Book of Abstracts (Unitas Malacologica: Washington D.C.). (in prep.). Slit-bearing Siliquariidae (Mollusca: Caenogastropoda) as obligate sponge dwelling guests: morphological constraints and adap- tations. Paleobiology, to be submitted. SEILACHER, A. &. GUNJI, P. Y. 1993. Morpho- genetic countdowns in heteromorph shells. Neues Jahrbuch für Geologie und Paläontologie Abhandlungen 190(2-3): 237-265. SOEST, R.W.M. VAN & STENTOFT, N. 1988. Barbados deep-water sponges. Studies on the fauna of Curacao and other Caribbean Islands 70(215): 1-175. URIZ, M. J., ROSELL, D. & MALDONADO, M. 1992. Parasitism, commensalism or mutualism? The case of Seyphozoa (Coronatae) and horny sponges. Marine Ecology Progress Series 81: 247-255, VERMEIJ, G.J. 1987. Evolution and Escalation. An ecological history of life. (Princeton University Press: New Jersey). 438 MEMOIRS OF THE QUEENSLAND MUSEUM ANTIMICROBIAL ACTIVITY OF CARIBBEAN SPONGE EXTRACTS. Memoirs of the Queensland Museum 44: 438. 1999:- Marine sponges produce a diversity of unusual chemical compounds, but the ecological functions of these metabolites remain largely unknown. Organic extracts from 33 Caribbean sponges were assayed against a panel of 8 marine bacterial strains to determine if sponge secondary metabolites have ecologically significant antimicrobial effects. The test panel was comprised of an opportunistic pathogen (Vibrio parahaemolyticus), a common fouling bacterium (Deleya marina), and strains isolated from seawater and healthy and necrotic Caribbean sponges. Extracts were tested for antibiotic activity at concentrations that were volumetrically equivalent to those found in sponge tissues (i.e., whole-tissue concentrations). Bioassay results revealed that 16 species extracts (48% of those tested) exhibited antibiotic activity against at least one bacterial isolate and that the necrotic sponge isolates were the most sensitive test strains (inhibited by 40% of the extracts). Extracts from Amphimedon compressa, Amphimedon erina, Aplysina lacunosa, Ptilocaulis spiculifera and Axinella corrugata inhibited the largest numbers of test strains and exhibited the most potent antibiotic activities with values frequently exceeding that of the control antibiotic (Gentamicin). The pattern of antimicrobial activity was different for 15 of the 16 active species indicating that diverse taxa do not produce similar antibacterial metabolites. In total only 23% of the extracts/bacterial interactions tested generated antimicrobial activity indicating that conspicuous members of the Caribbean sponge community do not generally produce broad-spectrum antibacterial metabolites. All the extracts from species that exhibited antibacterial activity also deterred feeding by reef fish in a previous study, suggesting that some secondary metabolites may have evolved with multiple functions. Stevensine, a compound from Axinella corrugata known to deter feeding by predatory reef fishes, exhibited weak antibacterial activity, suggesting that this potent feeding deterrent is not solely responsible for the antimicrobial activity detected in the crude sponge extract. O Porifera, antimicrobial, antibiotic, secondary metabolites, chemical defense, ecology. Rochelle W. Newbold (email: rwnewbold@ hotmail.com) & Joseph R. Pawlik, Biological Sciences and Center for Marine Science Research, University of North Carolina at Wilmington, Wilmington, North Carolina 28403-3297, USA; Paul Jensen & William Fenical, University of California, San Diego, Scripps Institution of Oceanography, La Jolla, California 92093-0236, USA; 1 June 1998. ANNOTATED CHECKLIST OF MARINE SPONGES OF THE INDIAN REGION J.G. PATTANAYAK Pattanayak, J.G. 1999 06 30: Annotated checklist of marine sponges of the Indian region. Memoirs of the Queensland Museum 44: 439-455. Brisbane. ISSN 0079-8835. An annotated checklist of marine sponges from the Indian region was compiled primarily from: identified sponges in the holdings of Zoological Survey of India; published descriptions of species from this region; and personal correspondence with experts in the field both from European and Oriental regions. 451 species are recorded from the 3 classes, 17 orders, 64 families and 168 genera. Few taxonomic studies on sponges have been undertaken in the Indian region over the past two decades, but there is a large early literature commencing in 1873. Consequently most of the literature and species descriptions are relatively old, sometimes too brief, and sometimes entirely inadequate to differentiate between related species. It is also well known that sponges are notoriously difficult to identify and misidentifications are common. Unfortunately re-examination of many type or voucher specimens was not possible given that they are scattered throughout many different museums and institutions. The preparation of this checklist therefore relies heavily on the published literature, with only relatively few species yet checked from original material. O Porifera, taxonomic checklist, Indian sponges, sponge identification. J.G. Pattanayak, Zoological Survey of India, Estuarine Biological Station, Hillpatna, Berhampur (Ganjam) - 760 005, Orissa, India; 28 October 1998. Published literature on the Indian region sponge fauna has a rich and productive history, com- mencing with the work of Bowerbank (1873), Carter (1880-1887), Dendy (1887-1916) and Schulze (1894-1904) followed by the present century workers Annandale (1911-1915), Kumar (1924-1925), Dendy & Burton (1926), Burton (1928-1937), Burton & Rao (1932), Rao (1941), Ali (1956), Thomas (1968-1993) and Pattanayak (1995-1998), with some revisions of this fauna also made by Hooper (1996). With the exception of several papers by Thomas few taxonomic studies have been undertaken in the Indian region over the past two decades. Consequently most of the literature and species descriptions are relatively old, sometimes too brief, and sometimes entirely inadequate to differentiate between related species. It is also well known that sponges are notoriously difficult to identify and misidentifications are com- mon. Unfortunately re-examination of many type or voucher specimens was not possible given that they are scattered throughout many museums and institutions. Consequently, this annotated checklist of marine sponges from the Indian region is based primarily on: a) identification of sponges in the holdings of Zoological Survey of India by several workers, including the author; b) cross-referencing ofall available literature pertaining to this region; and c) personal correspondence with experts in the field both from European and Oriental regions. This heavy reliance on the published literature is presently unavoidable, with only relatively few species yet checked from original material, although it is anticipated that this checklist will be both expanded (with current work being undertaken by the Zoological Survey of India) and revised (as type material becomes available to check species’ identities, and as contemporary authors increase their published revisions). For convenience the Indian region is divided into 9 zones (see Fig. 1). In the species checklist each published record is accompanied by one or more literature citation, the locality and region of collection, as defined above in brackets follow- ing the citation. For brevity, only the Indian literature is cited in this paper. Literature pertain- ing to the higher taxa can be found in Hooper & Wiedenmayer (1994), SYSTEMATIC CHECKLIST OF SPONGES Phylum Porifera Class Demospongiae Subclass Homoscleromorpha Order Homosclerophorida Family Plakinidae Schulze Genus Corticium Schmidt. C. acanthastrum Thomas, 1968e: 260, fig. la-b; Thomas, 1985: 353, pl. VIII, fig. 21 (Palk Bay) (5). C. candela- brum Schmidt; Thomas, 1968e: 261, fig. 2a-b; 440 MEMOIRS OF THE QUEENSLAND MUSEUM aa aaa] FIG. 1. Study zones of the Indian region. 1, Arabian sea - off-shore waters; 2, Laccadive Islands; 3, northern W coast - Gujarat and Maharashtra; 4, southern W coast - Karnataka and Kerala; 5, SE coast - Tamil Nadu and adjacent islands in the Gulf of Mannar; 6, NE coast - Andhra Pradesh, Orissa and West Bengal; 7, Andaman Islands; 8, Nicobar Islands; 9, Bay of Bengal - off-shore waters. Thomas, 1985: 352, pl. VIIL fig. 20 (Gulf of Mannar and Palk Bay) (5). Genus Plakina Schulze. P. acantholopha Thomas, 1970c: 53, fig. 3; Thomas, 1985: 354, pl. IX, fig. 3 (Palk Bay) (5). P. monolopha Schulze; Thomas, 1970c: 208, fig. 10; Thomas, 1970c: 52, fig. 1; Thomas, 1985: 353, pl. VIII, fig. 22 (Palk Bay) (5). P. trilopha Schulze; Thomas, 1970c: 53, fig. 2; Thomas, 1985: 353, pl. IX, fig. 1 (Gulf of Mannar) (5). Genus Plakinastrella Schulze. P. ceylonica (Dendy, 1905); Thomas, 1985: 351, pl. VIIL fig. 17 (Gulf of Mannar, as Dercitopsis ceylonica) (5). P. minor (Dendy, 1905); Thomas 1985: 351, pl. VIII, fig. 18 (Gulf of Mannar and Palk Bay, as Dercitopsis minor) (5). Subclass Tetractinomorpha Order Spirophorida Family Tetillidae Sollas Genus Cinachyra Sollas. C. arabica (Carter); Burton & Rao, 1932: 326; Pattanayak, 1998 (Camorta I., Nicobars, Krusadai I., Andamans) (5,7). C. australiensis (Carter); Burton & Rao, 1932; Pattanayak, 1998 (Andamans) (7). C. cavernosa (Lamarck); Burton, 1937: 12; Rao, 1941: 425, (Gulf of Mannar and Palk Bay, as Chrotella australiensis); Thomas, 1985: 341, pl. VII, fig. 21 (Gulf of Mannar and Palk Bay); Thomas, 1984: 101, fig. Ih (SE coast of India) (5,9). C. hirsuta (Dendy); Burton, 1930: 666 (Gulf of Mannar) (5). Genus Paratetilla Dendy. P. bacca (Selenka); Kumar, 1925: 218; Burton & Rao, 1932: 325 (Kilakarai, Ramnad Dt.); Burton & Rao, 1932: 325, Pattanayak, 1998 (Andamans); Thomas, 1980b: 16 (Minicoy I.); Thomas, 1985: 342, pl. VII, fig. 22 (Gulf of Mannar and Palk Bay)) (2,5,7). Genus Samus Gray. S. anonyma, Gray; Carter 1880: 59; Thomas, 1985: 354, pl. IX, fig. 3 (Gulf of Mannar) (5). Genus Tetilla Schmidt. T. barodensis Dendy, 1916b: 105, pl. 1, fig. 3a-d (Off Dwarka, Okha- mandal) (3). T. cranium (Müller); Burton & Rao, 1932: 326 (Invisible bank, Andamans; 2-3 miles W of Cape Comorin) (5,7). T. dactyloidea (Carter); Annandale, 1915c: 53, pl. V, fig. 4 (Chilka Lake, Orissa, as T. dactyloidea var. lingua); Dendy, 1916b: 102, pl. II. fig. 10a-c (Balapur, Okha- mandal); Burton & Rao, 1932: 326; Pattanayak, 1998 (Port Blair, Andamans) (3,6,7). T. hirstua (Dendy): Dendy, 1916b: 104 (Vamiani Point, off Okhamandal) (3). 7. pilula Dendy, 1916b: 104, pl. 1, fig. 2a-c (Okhamandal) (3). MARINE SPONGES OF THE INDIAN REGION Family Scleritodermidae Sollas Genus Aciculites Schmidt, A. orientalis Dendy, 1905; Thomas, 1985: 323, pl. VI, fig. 26 (Gulf of Mannar) (5). Genus Amphibleptula Schmidt. A. herdmani (Dendy, 1905); Thomas, 1985: 255, pl. II, fig. 28 (Gulf of Mannar, as Taprobane herdmani). Order Astrophorida Family Ancorinidae Schmidt Genus Ancorina Schmidt. A. simplex (Lenden- feld); Kumar, 1925: 213 (Ramnad Dist., S India) (5). Genus Asteropus Sollas. A. simplex (Carter); Dendy, 1916b: 98 (Okhamandal); Kumar, 1925: 212 (Orissa Coast); Thomas, 1985: 326, pl. VI, fig. 30 (Gulf of Mannar and Palk Bay, as Stelletti- nopsis simplex) (3,5,6). Genus Ecionemia Bowerbank. E. acervus Bowerbank, 1863; Burton & Rao, 1932: 318 (Andaman and Nicobar Is, as E. carteri; Kilakarai); Burton, 1937: 5 (Krusadai I., as E. bacilifera); Thomas, 1980b: 15 (Minicoy I.) ; Thomas, 1985: 334, pl. VII, fig. 6 (Gulf of Man- nar) (2,5,7,8). E. laviniensis Dendy, 1905; Thomas, 1985: 335, pl. VU, fig. 7 (Gulf of Mannar) (5). E. thielei Thomas, 1980b: 15, fig. 2c (Minicoy 1.) (2). Genus Myriastra Sollas. M. clavosa (Ridley); Dendy & Burton, 1926: 246; Burton & Rao, 1932: 331 (off Cinque I., Andamans, as Srelletta clavosa); Thomas, 1985: 336, pl. VII, fig. 9 (Gulf of Mannar) (5,7). M. purpurea (Ridley); Burton & Rao, 1932: 310 (Tuticorin, Nicobar L, as Stelletta purpurea); Burton, 1937: 4 (Krusadai L., as Stelletta purpurea); Rao, 1941: 418; Thomas, 1985: 336, pl. VII, fig. 8 (Gulf of Mannar); Pattanayak, 1998 (Andaman and Nicobar Is) (5,7,8). Genus Penares Gray. P. intermedia (Dendy, 1905); Thomas, 1984: 100, fig. 1j,k (SE coast of India); Thomas, 1985: 334, pl. VII, fig. 5 (Gulf of Mannar (5,9). Genus Rhabdastrella Thiele. R. globostellata (Carter); Burton & Rao, 1932: 317 (Aberdeen Reef, Andaman, as Aurora globostellatta); Thomas, 1984: 100 (SE coast of India, as Aurora globostelletta); Thomas, 1985: 336, pl. VII, fig. 10 (Palk Bay, as Aurora globostelletta) (5,7,9). R. providentiae (Dendy); Thomas, 1985: 337, pl. VII, fig. 11 (Gulf of Mannar and Palk Bay, as Aurora providentiae) (5). R. rowi Dendy, 1916b; 441 Burton & Rao, 1932: 317 (Ganjam Coast, as Aurora rowi) (6). Genus Stelletta Schmidt. S. aruensis Hentschel, 1912; Burton & Rao, 1932: 312 (Ganjam coast) (6). S. cavernosa (Dendy, 1916b); Burton & Rao, 1932: 311 (Nicobar I.) (8). S. haeckeli (Sollas); Dendy, 1916b: 97 (Okhamandal, as Myriastra (Pilochrota) haeckeli); Dendy & Burton, 1926: 246 (Interview I., Andamans (3,7). S. herdmani Dendy, 1905; Thomas, 1985: 338, pl. VII, fig. 13 (Gulf of Mannar) (5). S. orientalis Thiele; Burton and Rao, 1932:311 (Andamans) (7). S. tethyopsis Carter, 1880: 137, pl. 6, figs 39,40; Thomas, 1985: 337, pl. VII, fig. 12 (Gulf of Mannar) (5). S. validissima Thiele; Burton and Rao, 1932: 310; Dendy & Burton, 1926: 241; Pattanayak, 1998 (Andamans) (7). S. vestigium Dendy, 1905; Thomas, 1985: 338, pl. VII, fig. 14 (Gulf of Mannar) (5). Family Coppatiidae Topsent Genus Cryptotethya Dendy. C. agglutinans Dendy, 1905; Thomas, 1985: 328, pl. VI, fig. 36 (Gulf of Mannar) (5). Genus Jaspis Gray. J. bouilloni Thomas, 1973: 65, pl. 3, fig. 15; Thomas, 1985: 327, pl. VI, fig. 37 (Gulf of Mannar) (5). J. investigatrix (Annan- dale, 1915b: 460, pl. 34, figs 1,2) (Gulf of Mannar, as Coppatius investigatrix); Thomas, 1985: 327, pl. VI, fig. 33 (Gulf of Mannar) (5). J. penetrans (Carter, 1880: 141, pl. 7, fig. 44) (as Tisiphonia penetrans); Thomas, 1972: 352, pl. I, fig. 6A,B; Thomas, 1985: 326, pl. VI, fig. 32 (Gulf of Mannar) (5). J. reptans (Dendy, 1905); Thomas, 1985: 327, pl. VL fig. 31 (Gulf of Mannar); Dendy, 1916b: 97 (Okhamandal) (3,5). Genus Zaplethea de Laubenfels. Z. diagnoxea var. diastra Vacelet and Vasseur; Thomas, 1985: 328, pl. VI. fig. 35 (Gulf of Mannar and Palk Bay) (5). Family Geodiidae Gray Genus Erylus Gray. E. carteri Sollas; Carter, 1880: 15, pl. 7, fig. 41 (as Stelletta euastrum); Thomas, 1985: 341, pl. VII, fig. 20 (Gulf of Mannar) (5). E. lendenfeldi Sollas; Burton & Rao, 1932: 320; Pattanayak, 1998 (Andamans) (7). Genus Geodia Lamarck. G. areolata Carter, 1880: 133, pl. 6, figs 36-37; Thomas, 1985: 339, pl. IX, fig. 8 (Gulf of Mannar); Burton, 1937: 8, pl. 1, fig. 3 (Krusadai I.) (5). G. globostellifera Carter, 1880: 134, pl. 6, fig. 38; Thomas, 1985: 340, pl. VII, fig. 18 (Gulf of Mannar) (50). G. 442 inconspicua (Bowerbank); Burton & Rao, 1932: 322 (Kilakarai, Romnad Dist.; Ganjam coast; Gulf of Mannar); Thomas, 1985: 339, pl. VII, fig. 16 (Gulf of Mannar) (5,6), G. lindereni (Lenden- feld); Thomas, 1980b: 16 (Minocoy I.); Thomas, 1985: 340, pl. VII, fig. la (Gulf of Mannar and Palk Bay) (2,5). G. perarmata Bowerbank; Carter, 1880: 131, pl. 6, figs 32-35; Thomas, 1985: 339, pl. VII, fig. 15 (Gulf of Mannar) (5). G. picteti (Topsent); Burton, 1937: 9 (Krusadai I.) (5). G. ramodigitata Carter, 1880: 133, pl. 34, fig. 31; Thomas, 1985: 340, pl. VII, fig. 17 (Gulf of Mannar) (5). G. variospiculosa Thiele; Dendy, 1916b: 99 (Okhamandal) (3). Family Pachastrellidae Carter Genus Dercitus Gray. D. simplex (Carter); Burton & Rao, 1932: 309 (Invisible Bank, Andamans, as D. plicatus var. simplex). Genus Halina Bowerbank. H. extensa (Dendy, 1905); Thomas, 1985: 349, pl. VIII, fig. 13 (Gulf of Mannar) (5). H. plicata (Schmidt); Carter, 1880: 60 (as Samus simplex); Thomas, 1970a: 207, fig. 9; Thomas, 1985: 349, pl. VIII, fig. 12 (Gulf of Mannar) (5). Genus Pachamphilla Lendenfeld. P. dendyi Hent- schel; Thomas, 1977: 119, fig. 1D,E; Thomas, 1985:351, pl. VII, fig. 19 (Gulf of Mannar) (5). Genus Pachastrella Schmidt. P. parasitica Carter, 1880: 60 (as Samus (Pachastrella) parasitica); Thomas, 1985: 350, pl. VIII, fig. 15 (Gulf of Mannar) (5). P. nana (Carter, 1880: 138, pl. 7, fig. 4) (as Tisphonia nana); Thomas, 1985: 350, pl. VII, fig. 16 (as Nethea nana) (Gulf of Mannar) (5). Genus Poecillastra Sollas. P. eccentrica Dendy & Burton, 1926: 238; Pattanayak, 1998 (Anda- mans) (7). P. schulzei Sollas; Thomas, 1970: 208, fig. 8, 8a, 8b; Thomas, 1985: 355, pl. IX, fig. 4 (Gulf of Mannar) (5). P. tenuilaminaris Sollas; Dendy and Burton, 1926: 238; Burton & Rao, 1932: 309; Pattanayak, 1998 (Andamans) (7). Genus Sphinctrella Schmidt. S. annulata (Carter, 1880: 140, pl. 5, fig. 28) (as Tisiphonia annulata); Thomas 1985: 349, pl. VIII, fig. 14 (Gulf of Mannar)(5). Family Theneidae Sollas Genus Thenea Gray. T. andamanensis Dendy & Burton, 1926: 235; Pattanayak, 1998 (Andamans) (7). MEMOIRS OF THE QUEENSLAND MUSEUM Order Hadromerida Family Chondrillidae Gray Genus Chondrilla Schmidt. C. agglutinans Dendy, 19162: 102, pl. I, fig. 1a,b (Okhamandal) (3). C. australiensis Carter; Dendy, 1916b: 101 (Okhamandal); Burton and Rao, 1932: 325 (Gan- jam coast, Krusadai I.); Thomas, 1985: 355, pl. IX, fig. 5 (Gulf of Mannar) (3,5,6). C. kilakaria Kumar, 1925: 214, fig. 1 (Kilakarai, Ramnad Dt.); Thomas, 1985: 356, pl. IX, fig. 7 (Gulf of Mannar) (5). C. sacciformis Carter; Thomas, 1985: 356, pl. IX, fig. 6 (Palk Bay) (5). Genus Chondrosia Nardo. C. reniformis Nardo; Burton & Rao, 1932: 324 (Pearl Oyster Bank, Tuticorin); Burton, 1937: 10 (Krusadai I.); Rao, 1941: 424; Thomas, 1985: 358 (Gulf of Mannar) (5). Family Clionidae Gray Genus Annandalea Thomas. A. laeviaster (An- nandale, 1915: 462, fig. 2); Thomas, 1979: 179, fig. 5F (Bay of Bengal) (6). Genus Cliona Grant. C. anulifera Annandale, 1915; Thomas, 1979a: 178, fig. IN, 4G, 5LJ,K, pl. 2, fig. 9 (Gulf of Mannar) (5). C. carpentari Hancock; Thomas, 1975: 122 (Zuari and Man- dovi Estuary, Goa); Thomas, 1979a: 175 (Port Blair, Andamans); Thomas, 1985: 319, pl. VI, fig. 17 (Gulf of Mannar and Palk Bay); Thomas, Ramadoss & Vincent, 1993: 145 (Coast of Kerala and Tamil Nadu) (3,4,5,7). C. celata Grant; Thomas, 1975: 120 (Zuari and Mandovi Estuary, Goa); Thomas: 11 (Minicoy I.); Thomas, 1985: 317, pl. VI, fig. 12 (Gulf of Mannar and Palk Bay); Thomas, Ramadoss & Vincent, 1993: 145 (Coast of Kerala and Tamil Nadu) (2,3,4,5). C. coronaria (Carter); Dendy, 1916b: 132 (Okha- mandal) (3). C. ensifera Sollas; Thomas, 1979a: 176; Thomas, 1985: 320, pl. VI, fig. 18 (Gulf of Mannar); Pattanayak, 1998 (Andamans) (5,7). C. kempi Annandale, 1915b: 462, fig. 2; Thomas, 1979a: 179, fig. 5E (Andamans) (7). C. lobata Hancock; Burton, 1937: 16; Thomas, 1979a: 172, fig. 2N; Thomas, 1985: 318, pl. VI, fig. 14 (Gulf of Mannar); Thomas, Ramadoss & Vin- cent, 1993: 145 (Coast of Kerala and Tamil Nadu); Pattanayak, 1998 (Andamans) (4,5,7). C. margaritifera Dendy, 1905; Annandale, 1915b: 9; Thomas, 1972: 348, pl. II, fig. 1; Thomas, 1985: 321, pl. VI, fig. 21 (Palk Bay); Thomas, 1975: 122 (Zuan and Mandovi estuary, Goa); Thomas, Ramadoss & Vincent, 1993: 145 (Coast of Kerala and Tamil Nadu) (3,4,5). C. mucronata Sollas; Thomas, 1972: 347, pl. 1, fig. SA,B,C,D; MARINE SPONGES OF THE INDIAN REGION Thomas, 1979a: 175, fig. 32, fig. 41,J; Thomas, 1985: 320, pl. VLfig. 19 (Gulf of Mannar); Pattanayak, 1998 (Andamans) (5,7). C. orientalis Thiele; Thomas, 1972: 347, pl. II, fig. 2A,B; Thomas, 1979a: 177, fig. 1M, fig. 4F, pl. IV, fig. 1; Thomas 1985: 321, pl. VI, fig. 20 (Gulf of Mannar) (5). C. quadrata Hancock; Carter, 1880: 370, pl. 18, fig. 6; Thomas, 1972: 349, pl. III, fig. 1; Thomas, 1979a: 174, fig. 5,D; Thomas, 1985: 319, pl. VI, fig. 15 (Gulf of Mannar); Pattanayak 1998 (Andamans ) (5,7). C. vastifica Hancock; Annandale, 1915b: 37, pl. IV, fig. 7 (Chilka Lake, Orissa); Thomas, 1972: 345, pl. I, fig. 3,3A,3B; Thomas, 1985: 318, pl. VI, fig. 13 (Gulf of Mannar and Palk Bay); Thomas, 1975: 121; Thomas and Thanapati, 1980: 54 (Zuari and Man- dovi Estuary, Goa); Thomas, 1980b: 12 (Minicoy I.); Thomas, Ramadoss & Vincent, 1993: 145 (Coast of Kerala and Tamil Nadu); Pattanayak, 1998 (Andamans) (2,3,4,5,6,7). C. viridis Schmidt; Thomas, 1972: 349, pl. IL, fig. 1; Thomas, 19792: 174, fig. 1F; Thomas, 1985:319, pl. VL fig. 16 (Gulf of Mannar and Palk Bay); Kumar, 1925: 228 (Kilakarai, Ramnad Dt.) (5). Genus Delectona de Laubenfels. D. higgini (Car- ter, 1880: 58, pl. 5, fig. 25); Thomas, 1972: 351, pl. II, fig. 4; Thomas, 1985: 322, pl. VI, fig. 24 (Gulf of Mannar) (5). Genus Donotella Dendy. D. acustella (Annan- dale, 1915b: 14, fig. 2) (Ganjam coast) (6). Genus Dotona Carter. D. pulchella Carter, 1880: 57, pl. 5, fig. 24; Thomas, 1979a: 180, fig. 5f; Thomas, 1985: 323, pl. VI, fig. 25 (Gulf of Man- nar) (5). Genus Thooce de Laubenfels. 7. socialis (Carter, 1880: 56, pl. 5, fig. 23) (as Thoosa socialis); Thomas, 1972: 350, pl. 3, fig. 5; Thomas, 1985: 322, pl. VI, fig. 23 (Gulf of Mannar) (5). Genus Thoosa Hancock. 7. (Cliothosa) armata (Topsent); Thomas, 1979a: 181, figs 2A, 3B (Bay of Bengal) (9). T. (Cliothosa) fischeri (Topsent); Thomas, 1985: 321 (Gulf of Mannar and Palk Bay) (5). T. (Clithosa) hancocki (Topsent); Thomas, 1979a: 180, fig. 1p (Bay of Bengal); Pattanayak, 1998 (Andamans) (7,9). T. (Clio- thosa) investigatoris (Annandale, 1915b); Thomas, 1985: 322, pl. VI, fig. 22 (Gulf of Man- nar and Palk Bay) (5). Family Latrunculiidae Topsent Genus Latrunculia Bocage. L. tenuistellata (Dendy, 1905); Thomas, 1985: 299, pl. V, fig. 12 (Gulf of Mannar) (5). 443 Family Placospongiidae Gray Genus Placospongia Gray. P. corinata (Bowerbank); Dendy, 1916b: 132 (Okhmandal); Burton, 1937: 16; Thomas, 1985: 314, pl. VI, fig. 8 (Gulf of Mannar); Thomas, 1980: 11 (Minicoy I.) (2,3,5). P. melobesioides Gray; Thomas, 1985: 313, pl. VI, fig. 7 (Gulf of Mannar) (5). Family Polymastiidae Gray Genus Polymastia Bowerbank. P. gemmipara Dendy, 1916b: 135, pl. I, fig. 9a,b (Okhamandal) 3). Family Spirastrellidae Ridley & Dendy Genus Spirastrella Schmidt. S. andamanensis Pattanayak, 1998 (Andamans) (7). S. aurivilli Lindgren; Thomas, 1972: 340, pl. 1, fig. 4A-C; Thomas, 1979a: 185, figs 10,4L; Thomas, 1985: 306, pl. V, fig. 25 (Gulf of Mannar) (5). S. coccinea (Duch & Mich.); Thomas, 1985: 304, pl. V, fig. 21 (Gulf of Mannar) (5). S. cuspidifera (Lamarck); Thomas, 1972: 338; Thomas, 1979a: 183; Thomas, 1985: 305, pl. V, fig. 22 (Gulf of Mannar and Palk Bay); Thomas, 1980b: 9 (Mini- coy 1.) (2,5). 5. florida Lendenfeld; Kumar, 1925: 227 (Kilakarai and Gulf of Mannar) (5). S. in- constans (Dendy); Burton, 1937: 14; Thomas, 1972: 339; Thomas, 1979a: 183, fig. 1F, 4M; Thomas, 1985: 305, pl. V, fig. 23 (Gulf of Mannar); Thomas, 1980b: 10 (Minicoy 1.); Pat- tanayak, 1998 (Andaman and Nicobar I.) (2,5,7). S. pachyspira Levi; Thomas, 1968f: 265; Thomas, 1985: 306, pl. V, fig. 24 (Gulf of Mannar) (5). S. punctulata Ridley; Kumar, 1925: 228 (Kilakarai and Gulf of Mannar) (5). S. vaga- bunda var. tubulodigitata Dendy, 1916b: 132 (Okhamandal) (3). Family Suberitidae Schmidt Genus Aaptos Gray. A. aaptos (Schmidt, 1864); Dendy, 1916b: 101 (Okhamandal, as Tuberella aaptos); Burton, 1937: 13; Thomas, 1985, 313, pl. VL fig. 5 (Gulf of Mannar); Thomas, 1980b: 10 (Minicoy I.) (2,3,5). A. unispiculus (Carter, 1880: 45, pl. 4, fig. 8); Thomas, 1985: 313, pl. VI, fig. 6 (Gulf of Mannar) (5). Genus Laxosuberites Topsent. L. aquaedulcioris (Annandale); Annandale, 1915c: 42, pl. iv, figs 5,6 (Chilka Lake, Orissa) (6). L. conulosus Bur- ton, 1930: 669; Thomas, 1985: 311 (Gulf of Mannar) (5). L. cruciatus (Dendy, 1905); Dendy, 1916: 135 (Okhamandal as Suberites cruciatus); Kumar, 1925: 229 (Waltair, as Suberites cruciatus var. depressa); Burton, 1937: 14; E Thomas, 1985: 310, pl. VI, fig. 1 (Gulf of Man- nar); Thomas, 1980b: 10 (Minicoy I.) (2,3,5,6). L. lacustris Annandale, 1915: 45, pl. V, figs 2,3 (Chilka Lake, Orissa); Burton, 1937: 14; Thomas, 1985: 311, pl. VL fig. 2 (Gulf of Man- nar) (5,6). L. proteus (Hentschel); Burton, 1930: 669 (Gulf of Mannar) (5). Genus Pseudosuberites Topsent. P. andrewsi Kirkpatrick; Burton, 1937: 14; Rao, 1941: 425; Thomas, 1985: 312, pl. VI, fig. 3 (Gulf of Man- nar) (5). Genus Suberites Nardo. S. carnosus (Johnston); Dendy, 1916b: 134 (Okhamandal); Rao, 1941: 426; Thomas, 1985: 310, pl. V, fig. 35 (Gulf of Mannar); Thomas, 1980b: 10 (Minicoy I.) (2,3,5). S. flabellatus Carter; Dendy, 1916b: 135 (Okhamandal) (3). S. sericeus Thiele; Annan- dale, 1915c: 36, pl. IV, fig. 4 (Chilka lake, Orissa) (6). S. tylobtusa Lévi; Thomas, 1985: 310, pl. V, fig. 36 (Gulf of Mannar) (5). Genus Terpios Duchassaing & Michelotti. 7. fugax Duch. & Mich.; Burton, 1937: 15; Thomas, 1985: 312, pl. VI, fig. 4 (Gulf of Mannar) (5). Family Tethyidae Gray Genus Tethya Lamarck. T. andamanensis (Dendy & Burton, 1926: 248) (as Donatia anda- manensis); Pattanayak, 1998 (Andamans) (7). T. diploderma Schmidt; Burton, 1937: 12; Thomas, 1985: 331, pl. VIL, fig. 2 (Gulf of Mannar); Pattanayak, 1998 (Andaman and Nicobar Is) (5,7,8). T. japonica Sollas; Burton, 1937: 13; Thomas, 1985: 331, pl. VIL, fig. 3 (Gulf of Mannar); Thomas, 1980b: 14 (Minicoy I.) (2,5). T. repens (Schmidt); Dendy & Burton, 1926: 247 (as Donatia repens); Pattanayak, 1998 (Anda- mans); Burton, 1937: 12; Thomas, 1985: 332, pl. VIL, fig. 4 (Gulf of Mannar, as Tethytimea repens); Thomas, 1980b: 14 (Minicoy I.) (2,5,7). T. robusta Bowerbank; Burton, 1937: 13; Thom- as, 1985: 330, pl. VII, fig. 1 (Gulf of Mannar); Thomas, 1980b: 12 (Minicoy I.), Pattanayak, 1998 (Andamans) (2,5,7). T. seychellensis (Wright); Dendy, 1916b: 100 (Okhamandal, as Donatia seychellensis) (3). Genus Xenospongia Gray. X. patelliformis Gray; Thomas, 1985: 309, pl. V, fig. 34 (Gulf of Man- nar) (5). Family Timeidae Topsent Genus Kotimea de Laubenfels. K. moorei (Carter, 1880: 50, pl. 4, fig. 11); Thomas, 1985: 309, pl. V, fig. 33 (Gulf of Mannar) (5). MEMOIRS OF THE QUEENSLAND MUSEUM Genus Timea Gray. T. capitatostellifera (Carter, 1880: 51, pl. 4, fig. 12); Thomas, 1985: 307, pl. V, fig. 30 (Gulf of Mannar) (5). T. curvistellifera (Dendy, 1905); Thomas, 1985: 308, pl. V, fig. 31 (Gulf of Mannar) (5). T. spinatostellifera (Carter, 1880: 51, pl. 4, fig. 13); Thomas, 1985: 307, pl. V, fig. 29 (Gulf of Mannar) (5). T. stellata (Bow- erbank); Burton, 1937: 15; Thomas, 1985: 307, pl. V, fig. 26 (Gulf of Mannar) (5). 7. stelligera (Carter); Thomas, 1985: 307, pl. V, fig. 27 (Gulf of Mannar and Palk Bay) (5). T. stellivarians (Carter, 1880: 50, pl. 4, fig. 10a-c); Thomas, 1985, pl. v, fig. 32 (Gulf of Mannar) (5). Order Lithistida Suborder Triaenosina Family Corallistidae Sollas Genus Corallistes Schmidt. C. aculeata Carter, 1880: 143, pl. 7, fig. 45; Thomas, 1985: 343, pl. VIIL fig. 1 (Gulf of Mannar) (5). C. elegant- issima Carter, 1880: 144, pl. 7, fig. 47; Thomas, 1985:343, pl. VIII, fig. 2 (Gulf of Mannar) (5). C. verucosa Carter, 1880: 144, pl. 4, fig. 46; Thom- as, 1985: 343, pl. VIII, fig. 3 (Gulf of Mannar) (5). Family Theonellidae Lendenfeld Genus Discodermia du Bocage. D. enigmatica Dendy, 1905; Thomas, 1985: 345, pl. VIII, fig. 9 (Gulf of Mannar) (5). D. gorgonoides Burton, 1928: 109; Pattanayak, 1998 (Andamans) (7). D. interspersa Kumar, 1925: 215 (Ganjam coast) (6). D. laevidiscus Carter, 1880: 149, pl. 8, fig. 51; Thomas, 1985: 344, pl. VIII, fig. 6 (Gulf of Mannar) (5). D. papillata Carter, 1880: 146, pl. 8, fig. 48; Thomas, 1985: 344, pl. VIII, fig. 4 (Gulf of Mannar); Burton & Rao, 1932: 305; Pattanayak, 1998 (Andamans) (5,7). D. spinispirulifera Carter, 1880: 148, pl. 8, fig. 50; Thomas, 1985: 344, pl. VIII, fig. 5 (Gulf of Mannar) (5). D. sceptrellifera Carter, 1881: 372, pl. 18, fig. 2; Thomas, 1985: 345, pl. VIIL fig. 8 (Gulf of Mannar); Burton & Rao, 1932: 307 (Ganjam coast) (5,6). D. sinuosa Carter, 1887: 372, pl. 18, fig. 1; Thomas, 1985: 345, pl. VIII, fig. 7 (Gulf of Mannar) (5). Genus Theonella Gray. T. coupla Burton, 1928: 110 (Laccadive I.) (2). 7. swinhoei Gray; Burton & Rao, 1932: 308; Thomas, 1985: 346, pl. VIII, fig. 10 (Gulf of Mannar); Burton & Rao, 1932: 308; Pattanayak (Andamans) (5,7,8). MARINE SPONGES OF THE INDIAN REGION Suborder Anoplina Family Azoricidae Sollas Genus Leiodermatium Schmidt. L. pfefferae; Burton, 1928 (Andamans, as Azorica pfefferae) (7). Family Desmanthidae Topsent Genus Lophocanthus Hentschel. L. rhab- dophorus Hentschel; Burton, 1937: 11; Thomas, 1985: 348, pl. VII, fig. 11 (Gulf of Mannar); Thomas, 1980b: 16, fig. 2f (Minicoy I.) (2,5). Subclass Ceractinomorpha Order Agelasida Family Agelasidae Verril Genus Agelas Duchassing & Michelotti. A. cey- lonica Dendy, 1905; Thomas, 1980: 3 (Minicoy L); Thomas, 1985: 256, pl. II, fig. 30 (Gulf of Mannar) (2,5). A. mauritiana (Carter); Thomas, 1980a: 2 (Minicoy 1.); Thomas, 1985: 256, pl. Il, fig. 29 (Gulf of Mannar) (2,5). Incertae Sedis Genus Acanthostylotella Burton & Rao. A. cornuta (Topsent); Thomas, 1985: 257, pl. I, fig. 31 (Gulf of Mannar) (5). Order Poecilosclerida Suborder Microcionina Family lophonidae Burton Genus Acarnus Gray. A. souriei (Levi); Thomas, 1970b: 46, figs 1,2a-h; Thomas, 1985: 263, pl. Ill, fig. 4 (Gulf of Mannar, as Acanthacarnus souriei) (5). A. ternatus Ridley; Thomas, 1985: 262, pl. III, fig. (Gulf of Mannar) (5). A. thielei Lévi; Thomas, 1970b: 44, figs 3a-g, 4; Thomas, 1985: 263, pl. III, fig. 3 (Gulf of Mannar) (5). A. tortilis Topsent; Dendy, 1916: 130 (Okhamandal) Genus Cornulum Carter. C. vesiculatum (Dendy, 1905); Thomas, 1968e: 260, fig. la-b; Thomas, 1985: 238, pl. I, fig. 33 (Gulf of Mannar) (5). Genus Damiria Keller. D. simplex Keller; Thomas, 1985: 243, pl. II, fig. 8 (Gulf of Mannar) (5). D. fistulatus (Carter, 1880: 53, pl. 15, fig. 22); Thomas, 1985: 236, pl. I, fig. 31 (Gulf of Mannar, as Xytopsene fistulatus) (5). Genus Zyzzya de Laubenfels. Z. papillata (Thomas, 1968e: 252, fig. 6-8); Thomas, 1985: 243, pl. II, fig. 12 (Gulf of Mannar, as Damirina papillata) (50). Family Microcionidae Carter Genus Antho Gray. Subgenus Antho (Antho) Gray. A. (A.) mannarensis (Carter, 1880: 37) (as 445 Dictyocylindrus mannarensis); Thomas, 1985: 257, pl. II, fig. 32 (as Plocamilla mannarensis) (Gulf of Mannar); Burton & Rao, 1932: 255 (Laccadive sea, as Plocamia mannarensis); Bur- ton & Rao, 1932: 355 (off Mangalore, off Karwar) (2,4,5). A. (A.) tuberosa (Hentschel); Burton & Rao, 1932: 341 (Ganjam coast, E coast of India) (6). Genus Artemisina Vosmaer. A. indica (Thomas, 1974: 312; Thomas, 1985: 287, pl. IV, fig. 19 (Gulf of Mannar, as Qasimella indica) (5). Genus Clathria Schmidt. Subgenus Clathria (Clathria) Schmidt. C. (C.) decumbens (Ridley); Burton, 1937: 29; Thomas, 1985: 278, pl. IV, fig. 3 (Gulf of Mannar) (5). C. (C.) indica Dendy, 1905; Burton & Rao, 1932: 336; Thomas, 1985: 278, pl. IV, fig. 4 (Gulf of Mannar). C. (C.) maeandrina Dendy, 1905; Burton, 1930: 668 (Gulf of Mannar) (5). C. (C.) prolifera (Verrill); Burton & Rao, 1932: 344 (Mangalore, off Kar- war, Arabian sea) (1,4). Subgenus Clathria (Dendrocia) Hallman. C. (D.) antvaja Burton & Rao, 1932: 348 (Gulf of Mannar, as Dendrocia antyaja) (5). Subgenus Clathria (Microciona) Bowerbank. C. (M.) aceratoobtusa Carter; Thomas, 1985: 273, pl. IH, fig. 20 (Gulf of Mannar) (5). C. (M.) affinis Carter, 1880: 41, pl. 4, fig. 15; Thomas, 1985: 273, pl. HL, fig. 19 (Gulf of Mannar) (5). C. (M.) atrasanguinea Bowerbank; Burton, 1937: 30 (Krusadai 1.); Burton & Rao, 1932: 344 (off Karwar, Arabian sea; Travancore, Kilakarai, Puri coast, Andamans); Thomas, 1985: 273, pl. III, fig. 18 (Gulf of Mannar); Pattanayak, 1998 (Andamans) (1,5,7). C. (M.) fascispiculifera (Carter, 1880: 44, pl. 4, fig. 7a-g); Thomas, 1985: 275, pl. HI, fig. 24 (Gulf of Mannar) (5). C. (M.) rhopalophora (Hentschel); Thomas, 1970a: 206, fig. 7; Thomas, 1985: 274, pl. III, fig. 23 (Gulf of Mannar) (5). Subgenus Clathria (Thalysias) Duch & Mich. C. (T.) amiranteiensis Burton, 1937: 29, pl. 3, fig. 20; Thomas, 1985: 276, pl. III, fig. 26 (Gulf of Mannar, as Colloclathria ramosa) (5). C. (T) encrusta Kumar, 1925: 221 (Orissa coast) (6). C. (T.) lendenfeldi Ridley & Dendy, 1886; Burton & Rao, 1932: 334 (Tuticorin, Ganjam Coast) (5,6). C. (T.) longitoxa (Hentschel); Burton, 1937: 30; Thomas, 1985: 274, pl. III, fig. 22 (Gulf of Mannar) (5). C. (T) micropunctata Burton & Rao, 1932: 340 (off Tuticorin) (5). C. (T.) procera (Ridely); Burton & Rao, 1932: 340 (Tuticorin, Cape Comorin, as Tenacia procera); Burton, 1937: 28; Thomas, 1985: 277, pl. IV, fig. 2 (Gulf 446 of Mannar) (5), € (T.) procera var. tessellata (Dendy, 1905); Burton, 1937: 29 (Gulf of Mannar) (5). C. (T) reinwardtii (Vosmaer): Dendy, 1916b: 128 (Okhamandal) (3). C. (T.) spiculosus Dendy, 1889: Dendy, 19/6b: 128 (Okhamandal) (3). C. (T.) toxifera (Heantschel); Burton, 1937: 31; Thomas, 1985: 274. pl. M, fig. 21 (Gulf of Mannar). C. (T.) vulpina (Lamark); Dendy, 1916b; 128 (Okhamandal, as Clathrica corallitincra); Burton & Rao, 1932: 337 (Tuticorin, Kilakarai, Andamans. as Tenacia frondifera), Burton, 1937: 27; Thomas, 1985: 277, pl. IV, fig. 1 (Gulf of Mannar); Pattanayak, 1998 (Andamans) (3.5.7). Genus Echinochalina Thiele. Subgenus Echino- chalina (Echinochalina) Thiele. E. (E.) barba (Lamarck); Thomas, 1977: 115; Hooper, 1996: 521 (Andaman sea) (7). Genus Echinoclathria Carter. E. rimosa {Ridley ); Thomas, 1985: 276, pl. IH, fig. 27 (Gulf of Mannar) (5). Genus Holopsumima Carter, 1885. H. erassa Carter; Thomas, 1985: 268 (Gulf of Mannar and Palk Bay) (5) Family Raspailiidae Hentschel Genus Aulosporigus Norman, A, sessilis (Carter, 1880: 38) (as Dicivocylindrus sessilis); Thomas, 1985: 270, pl. IH, fig. 11 (Gulf of Mannar) (5). 4. tubulatus (Bowerbank); Burton & Rao, 1932: 347 (Tuticorin); Burton, 1937: 32; Thomas, 1985; 269. pl. UL, fig. 10 (Gulf of Mannar) (5). Genus Cyamon Gray. C. quadriradiata Carter, 1880: 42, pl. 4. fig. 4; Thomas, 1985, pl. If, fig. 33 (Gulf of Mannar) (3). C. quinquer üdiata (Carter, 1880: 43, pl. 4, fig. 5); Thomas, 1985: 258, pl. I, fig. 34 (Gulf of Munnar) (5) C. vickersil (Bowerbank) Burton & Rao, 1932: 355 (Manga- lore) (4). Genus Echinodictyum Radley. E. asperum Ridley & Dendy; Burton & Rao, 1932: 348; Pattanayak, 1998 (Andamans) (7). E. clathrum Dendy, 1905; Burton, 1937: 31; Rao, 1941: 451; Thomas, 1985, pl. IL, fig. 18 (Gulf of Mannar) (5). E. gorgon- aides Dendy, 1916b; 129 (Okhamandal); Burton & Rav, 1932: 348 (Tuticorin); Thomas, 1985: 251, pl. II, fig. 19 (Gulf of Mannar) (3,5). £. longishiluin Thomas, 1968b; 246. pl. 2, fig. A,B; Thomas, 1985: 251, pl. Hi, fig. 20 (Gulf af Mannar); Thomas, 1980a: 1 (Minicoy 1.) (2,5). E. nervosum Ridley, 1881; Burton & Rao, 1932: 348 (Ganjam coast) (6). Genus Endectyon Topsent. E. fruticosa (Dendy, 1905); Thomas, 1985: 271, pl. IN, fig, 15 (Gulfof MEMOIRS OF THE QUEENSLAND MUSEUM Mannar) (5). E. lamellosa Thomas, 19764: 170, pl. L figs | A,B,C, figs 1a-d; Thomas, 1985; 272, pl. M, fig. 16 (Gulf of Mannar) (3). E, thursioni (Dendy); Burton & Rao, 1932: 347 (Tuticorin, as Hemectyon thurstoni); Burton, 1937: 34; Thomas, 1985: 272, pl, HI, fig. 17 (Gulf of Mannar) (5). Genus Eurypon Gray. E, clavatum (Bowerbank); Thomas, 1985; 275, pl. IIl, fig. 25 (Gulf of Mannar) (5). Genus Raspailia Nardo. Subgenus Raspailia (Raspailia) Nardo. R. (R.) anastomosa Kumar, 1925: 223 (Ganjam coast) (6). R. (R.) fruticosu Dendy; Kumar, 1925: 224 (Waltair) (6). R. (RJ frulicosa var. tenniramasa Dendy, 1905; Dendy, 1916b: 130 (Okhamandal) (3). R. (R) hornelli Dendy, 1905; Burton, 1937: 33; Thomas, 1985: 269, pl. iil, fig. 9 (Gulf of Mannar) (5). R. (RJ viminalis (Schmidt); Burton & Rao, 1932: 342 (Puri coast, Andamans); Pattanayak, 1998 (An- damans) (6,7). Genus Thrinacophara Ridley. T. cervicornis Ridley & Dendy; Dendy, 1916: 117 (off Dwarka) (3). Family Rhabdermiidae Topsent Genus Rhabderemia Topsent. R, acanthostvla Thomas, 1968b; 247, pl. 2, figs 4,5; Thomas, 1985: 271, pl. UL fig. 14 (Gulf of Mannar) (5). R. indica Dendy, 1905; Burton, 1937: 33; Thomas, 1985: 270, pl. HII, fig. 12 (Gulf of Mannar) (5). R. prolifera Annadale, 1915b; 464, pl. 34, fig. 3 (Gulf of Mannar, Andamans, E coast of India): Thomas, 1985: 271, pl. MI, fig. 13 (Gulf of Mannar) Pattanayak, 1998 (Andamans) (5,7,9). Suborder Myxillina Family Anchinoidae Topsent Genus Ectyobalzella Burton £ Rao. E. enig- matica Burton & Rao, 1932: 332, pl. XVIU, fig. 6; Pattanayak (1998) (Nicobar 1.) (8). Genus Kirkpatrickiu Topsent. K. spiculaphila Burton & Rao, 1932: 332, pl. XVII, figs 5,5a; Pattanayak, 1998 (Andamans) (7). Genus Phorbas Duchaissaing & Michelotti: P dubia Burton; Thomas, 1985: 252, pl. II, fig. 21 (Gulf of Mannar, as Anchinoe dubia) (5). Family Coelosphaeridae Hentschel Genus Coelosphaera Thomson. C. encrusta (Kumar, 1925; 220, fig. 3) (Kilakarai, as Histoderma encrusta), Thomas, 1985; 254, pl. IT, fig, 24 (Gulf of Mannar) (5). C. navicelligera MARINE SPONGES OF THE INDIAN REGION (Ridley & Dendy); Thomas, 1985: 253 (Gulf of Mannar, as Siderodermella navicelligera) (5). Genus Ectyodoryx Lundbeck. E. lissostyla Thomas, 1970a, p. 203, figs 1-3; Thomas, 1985: 260, pl. IL fig. 36 (Gulf of Mannar) (5). Genus Lissodendoryx Topsent. L. balanophilus Annandale; Thomas, 1985: 266, pl. II, fig. 5 (Gulf of Mannar) (5). L. similis Thiele; Ali, 1956: 293 (Madras) (5). L. sinensis Brondsted; Burton, 1937: 26; Thomas, 1985: 266, pl. HI, fig. 6 (Gulf of Mannar) (5). L. ternatensis (Thiele); Burton & Rao, 1932: 332 (Madras coast) (5). Genus Waldoschmittia de Laubenfels. W. schmidti (Ridley); Thomas, 1980a: 2 (Minicoy L); Thomas, 1985: 252, pl. II, fig. 22 (Gulf of Mannar) (2,5). Family Crambeidae Lévi Genus Psammochela Dendy. P. elegans Dendy, 1916: 126, pl. I, fig. 6, pl. III, fig. 22 (Okhamandal); Pattanayak, 1998 (Andamans) (3,7). P. fibrosa (Ridley); Thomas, 1985: 268, pl. Ill, fig. 8 (Gulf of Mannar) (5). Family Hymedesmiidae Topsent Genus Hymedesmia Bowerbank. H. dendyi Bur- ton; Burton & Rao, 1932: 350 (Andamans) (7). H. mannarensis Thomas, 1970a: 204; Thomas, 1985: 261, pl. II, fig. 38 (Gulf of Mannar) (5). H. mertoni Hentschel; Thomas, 1968f: 264; Thom- as, 1985: 260, pl. II, fig. 37 (Gulf of Mannar) (5). H. stylophora Thomas, 1970a: 204; Thomas, 1985: 261, pl. II, fig. 39 (Gulf of Mannar) (5). H. tenuissima (Dendy); Thomas, 1985: 261, pl. II, fig. 40 (Gulf of Mannar) (5). Family Myxillidae Topsent Genus Damiriopsis Burton. D. brondstedi Bur- ton, 1928: 124 (Andamans) (7). Genus Desmapsamma Burton. D. anchorata (Carter); Thomas, 1985: 267, pl. III, fig. 7 (Gulf of Mannar) (5). Genus lotrochota Ridley. 1. baculifera Ridley; Dendy, 1916b: 123 (Okhamandal); Burton & Rao, 1932: 353 (Nicobars); Thomas, 1985: 235, pl. L fig. 29 (Gulf of Mannar) (3,5,8). Genus Myxilla Schmidt. Subgenus Myxilla (Myxilla) Schmidt. M. (M.) arenaria Dendy, 1905; Dendy, 1916b: 127 (Okhamandal); Thomas, 1985: 259, pl. II, fig. 35 (Gulf of Mannar) (3,5). 447 Family Phoriospongiidae Lendenfeld Genus Chondropsis Carter. C. kirkii (Carter); Dendy, 1916b: 127 (Okhamandal) (3). Genus Strongylacidon Lendenfeld. S. stelliderma (Carter); Rao, 1941: 449; Thomas, 1985: 237(Gulf of Mannar) (5). Family Tedaniidae Ridley & Dendy Genus Tedania Gray. T. (T) anhelans (Liber- kühn); Burton & Rao, 1932: 353 (Andamans, Gulf of Mannar); Burton, 1937: 27; Ali, 1956: 293 (Madras); Thomas, 1980a: 4 (Minicoy I.); Thomas, 1985: 261, pl. III, fig. 1 (Gulf of Mannar); Pattanayak, 1998 (Andamans) (2,5,7). Suborder Mycalina Family Guitarridae Burton Genus Guitarra Carter. G. indica Dendy, 1916b: 124, pl. I, fig. a-b, pl. III, fig. 21 (Okhamandal) (3). Family Desmacellidae Ridley & Dendy Genus Biemna Gray. B. annexa (Schmidt); Bur- ton, 1928: 120 (Laccadive sea) (2). B. fistulosa Topsent, Burton, 1937: 25; Thomas, 1985: 288, pl. IV, fig. 22 (Gulf of Mannar) (5). B. fortis (Topsent); Burton & Rao, 1930: 327 (Puri coast); Thomas, 1980a: 6 (Minicoy I.); Thomas, 1985: 288, pl. IV, fig. 21 (Gulf of Mannar) (2,5,6). B. lipsosigma Burton, 1928: 120; Pattanayak, 1998 (Andamans) (7). B. microstyla Thomas, 1984: 95, fig. 1b (Bay of Bengal, SE coast of India) (9). B. tubulata (Dendy, 1905); Burton & Rao, 1932, p.327 (Andamans, Nicobars, Tuticorin); Thomas, 1985: 287, pl. IV, fig. 20 (Gulf of Man- nar, as Toxemna tubulata) (5,7). Genus Desmacella Schmidt. D. tubulata Dendy, 1905; Dendy, 1916b: 116 (Okhamandal) (3). Family Mycalidae Lundbeck Genus Mycale Gray. Subgenus Mycale (Carmia) Gray. M. (C.) madraspatna Annandale; Burton, 1937; 24 (Krusadai L.); Ali, 1956: 295 (Madras); Thomas, 1985: 285, pl. IV, fig. 15 (Gulf of Mannar, as Carmia madraspatna) (5). M. (C.) monanchorata (Burton & Rao, 1932: 329 (Kilakarai); Rao, 1941: 447; Thomas, 1985: 284, pl. IV, fig. 13 (Gulf of Mannar) (5). M. (C.) sulevoidea Sollas; Thomas, 1968d: 256; Thomas, 1985: 284; pl. IV, fig. 14 (Gulf of Mannar and Palk Bay) (5). 448 Subgenus Mycale (Mycale) Gray. M. (M.) aegagropila (Johnston); Rao, 1941: 445 (Gulf of Mannar) (5). M. (M.) aegagropila var. militaris (Annandale); Ali, 1956: 295 (Madras) (5). M. (M.) coronata Dendy; Burton, 1928: 121 (Andamans) (7). M. (M.) crassissima (Dendy); Thomas, 1985: 282, pl. IV, fig. 10 (Gulf of Mannar) (5). M. (M.) grandis (Gray); Burton, 1937: 23; Thomas, 1985: 279, pl. IV, fig. 5 (Gulf of Mannar); Thomas, 1980a: 6 (Minicoy I.) (2,5). M. (M.) gravelyi Burton, 1937: 24; Thomas, 1985: 283, pl. IV, fig. 9 (Gulf of Mannar) (5). M. (M.) indica (Carter, 1887); Burton Rao, 1932: 328; Pattanayak, 1998 (Andamans); Rao, 1941: 445 (Pamban, Gulf of Mannar) (5,7). M. (M.) mannarensis Thomas, 1968a: 255, fig. 1-2; Thomas, 1985: 283, pl. IV, fig. 12 (Gulf of Mannar) (5). M. (M.) mytilorum Annandale; Bur- ton, 1937: 24; Thomas, 1985: 282 (Gulf of Mannar); Ali, 1956: 294 (Madras) (5). M. (M.) plumosa (Carter); Dendy, 1916b: 121 (Okha- mandal) (3). M. (M.) spongiosa (Dendy); Burton, 1928: 119; Thomas, 1985: 279, pl. IV, fig. 6 (Gulf of Mannar) (5). M. (M.) tenuispiculata (Dendy, 1905); Burton, 1937: 23; Thomas, 1985: 280, pl. IV, fig. 7 (Gulf of Mannar) (5). M. (M.) tricomal- iensis Rao, 1941: 447, pl. 12, fig. 19; Thomas, 1985: 283, pl. IV, fig. 11 (Gulf of Mannar) (5). Genus Paresperella Dendy, 1905. P. bidentata Dendy, 1905; Burton, 1937: 26; Thomas, 1985: 286, pl. IV, fig. 18 (Gulf of Mannar) (5). P serratohamata (Carter, 1880: 49, pl. 5, fig. 20a-d); Thomas, 1985: 286, pl. IV, fig. 17 (Gulf of Mannar) (5). Genus Zygomycale Topsent. Z. parishii (Bower- bank); Burton & Rao, 1932: 328; Thomas, 1985: 285, pl. IV, fig. 16 (Gulf of Mannar and Palk Bay); Thomas, 1980: 6 (Minicoy I.) (2,5). Order Halichondrida Family Axinellidae Carter Genus Acanthella Schmidt. A. cavernosa Dendy; Burton, 1937: 36 (Gulf of Mannar); Thomas, 1980b: 9 (Minicoy 1.); Thomas, 1984: 97 (Bay of Bengal, SE coast of India) (2,5,9). 4. mega- spicula Thomas, 1984: 99 (Bay of Bengal, SE coast of India) (9). A. ramosa Kumar, 1925: 224 (Ganjam coast) (6). Genus Auletta Schmidt. A. andamanensis Pattanayak, 1998 (Andamans) (7). A. elongata Dendy, 1905; Burton, 1937: 37; Thomas, 1985: 302, pl. V, fig. 19 (Gulf of Mannar) (5). A. elongata var. fruticosa Dendy, 1916: 119 ( MEMOIRS OF THE QUEENSLAND MUSEUM Okhamandal) (3). A. lyrata (Esper); Dendy, 1889: 92 (as A. aurantiacea); Burton, 1937: 34 (as Axinella lyrata); Thomas, 1985: 303 (as Acanthella lyrata) (Gulf of Mannar) (5). A. lyrata var. glomerata Dendy, 1905; Dendy, 1916b: 119 (Okhamandal) (3). Genus Axinella Schmidt. A. acanthelloides Pattanayak, 1998 (Andamans) (7). A. agarici- formis (Dendy, 1905); Thomas, 1985: 291, pl. IV, fig. 27 (Gulf of Mannar) (5). A. bubarinoides Dendy; Burton, 1937: 36; Thomas, 1985: 293, pl. V, fig. 3 (Gulf of Mannar) (5). A. carteri (Dendy, 1889); Burton, 1937: 35; Thomas, 1985: 290, pl. IV, fig. 24 (Gulf of Mannar) (5). A. ceylonensis (Dendy, 1905); Burton, 1937: 35; Thomas, 1985: 293, (Gulf of Mannar) (5). A. conulosa Dendy; Burton, 1937: 36 (Kruasadai L, Gulf of Mannar) (5). A. crassistylifera (Dendy, 1905); Thomas, 1985: 293, pl. V, fig. 4 (Gulf of Mannar) (5). A, donnani (Bowerbank); Dendy, 1916b: 118 (Ok- hamandal, as Phakellia donnani); Burton, 1937: 35, Thomas, 1970a: 207; Thomas, 1985: 289 (Gulf of Mannar) (3,5). A. durissima (Dendy, 1905); Thomas, 1985: 292 (Gulf of Mannar) (5). A. halichondroides Dendy, 1905; Thomas, 1985; 293, pl. V, fig. 1 (Gulf of Mannar) (5). A. laby- rinthica Dendy, 1889: 88; Thomas, 1985: 290 (Gulf of Mannar) (5). A. lamellata (Dendy, 1905); Thomas, 1985: 291 (Gulf of Mannar) (5). A, manus Dendy, 1905; Thomas, 1985: 292, pl. IV, fig. 29 (Gulf of Mannar) (5). A. symmetrica Dendy, 1905; Thomas, 1985: 291, pl. IV, fig. 26 (Gulf of Mannar) (5). A. tenuidigitata Dendy, 1905; Thomas, 1985: 290, pl. IV, fig. 25 (Gulf of Mannar) (5). A. virgultosa Carter, 1887; Dendy, 1916b: 118 (off Dwarka) (3). Genus Bubaris Gray. B. columnata Burton, 1928: 130 (Andaman sea) (7). B. gorgonoides Thomas, 1984: 96, fig. 1 fg, pl. 1A (Bay of Bengal, SE coast of India) (9). B. radiata Dendy, 1916b: 131, pl. L fig. 8a,b, pl. IV, figs 24a,b (Okhamandal). B. vermiculata (Bowerbank); Carter, 1880: 46 (as Hymerhaphia vermiculata var. erecta); Thomas, 1985: 296, pl. V, fig. 6 (Gulf of Mannar) (5). Genus Monocrepidium Topsent. M. eruca (Carter, 1880: 46, pl. 4, fig. 9a-c); Thomas, 1985: 297, pl. V, fig. 8 (Gulf of Mannar) (5). Genus Phakettia Bowerbank. P. ridleyi (Dendy); Thomas, 1985: 294, pl. V, fig. 5 (Gulf of Mannar) (5). Genus Rhabdoploca Topsent. R. curvispiculifera (Carter, 1880: 43, pl. 4, fig. 6) (as Microciona curvispiculifera); Thomas, 1985: 296, pl. V, fig. 7 (Gulf of Mannar) (5). MARINE SPONGES OF THE INDIAN REGION Family Desmoxyidae Hallmann Genus Higginsia Higgin. H. higgini Dendy; Thomas, 1985: 298, pl. V, fig. 10 (Gulf of Man- nar) (5). H. mixta (Hentschel); Thomas, 1977: 116 (Bay of Bengal); Thomas, 1985: 297, pl. V, fig. 9 (Gulf of Mannar) (5). Genus Myrmekioderma Ehlers. M. granulata (Esper); Burton, 1930: 668 (as Acanthoxifer ceylonensis); Burton, 1937: 39; Thomas, 1985: 298, pl. V, fig. 11 (Gulf of Mannar); Thomas, 1980b: 8 (Minicoy I.) (2,5). Family Dictyonellidae Van Soest, Diaz & Pomponi Genus Liosina Thiele. L. paradoxa Thiele; Bur- ton, 1937; 39; Thomas, 1985: 236, pl. I, fig. 30 (Gulf of Mannar) (5). Family Halichondriidae Vosmaer Genus Amorphinopsis Carter. A. arcotti (Ali, 1956: 296) (Rayapuram coast, Madras, as Pro- stylissa arcotti) (5). A. excavans Carter; Thomas, 1972: 341, pl. 2, fig. 3; Thomas, 1979a: 168; Thomas, 1985: 316, pl. VI, fig. 9 (Gulf of Mannar) (5). A. excavans var. digitifera Annan- dale, 1915; Kumar, 1925: 225 (Waltair coast) (6). A. foetida (Dendy); Burton, 1937: 37; Thomas, 1985: 324, pl. VI, fig. 27 (Gulf of Mannar); Thomas, 1980b: 12 (Minicoy I.); Thomas, 1984: 99, fig. 1a (Bay of Bengal, SE coast of India); Pattanayak, 1998 (Andamans) (2,5,7,9). A. kempi Kumar, 1925: 226 (Waltair) (6). A. oculata (Kie- schnick); Burton, 1937: 38; Thomas, 1985: 325, pl. VI, fig. 28 (Gulf of Mannar, as Prostylyssa oculata) (5). Genus Axinyssa Lendenfeld. A. flabelliformis (Keller); Burton, 1937: 35; Thomas, 1985; 309, pl. VI, fig. 39 (Gulf of Mannar) (5). Genus Ciocalypta Bowerbank. C. dichotoma, Dendy, 1916: 119, pl. HI, fig. 18 (Okhamandal) (3). C. penicillus Bowerbank; Burton, 1937: 38 (as Prostylissa heterostyla); Thomas, 1985: 301, pl. V, fig. 17 (Gulf of Mannar) (5). Ciocalypta polymastia (Lendenfeld); Thomas, 1973b: 441; Thomas, 1980b: 8 (Minicoy I.) (5). Genus Collocalypta Dendy. C. digitata Dendy, 1905; Thomas, 1985: 303, pl. V, fig. 20 (Gulf of Mannar) (5). Genus Epipolasis de Laubenfels. E. lapidiformis (Dendy, 1905); Thomas, 1985: 329, pl. VI, fig. 38 (Gulf of Mannar) (5). E. topsenti (Dendy, 1905); Thomas, 1985: 329, pl. VI, fig. 37 (Gulf of Mannar) (5). 449 Genus Halichondria Fleming. H. glabrata Kel- ler, 1891; Burton, 1937: 37; Thomas, 1985: 300, pl. V, fig. 14 (Gulf of Mannar and Palk Bay) (5). H. panicea Johnston; Dendy, 1916b: 112 (Ok- hamandal); Thomas, 1985: 300, pl. V, fig. 13 (Gulf of Mannar) (3,5). H. reticulata Baer; Dendy, 1916b: 113, pl. II, fig. 14a,b (Okhmandal) (3). Genus Hymeniacidon Bowerbank. H. petrosioides Dendy, 1905; Thomas, 1985: 302, pl. V, fig. 18 (Gulf of Mannar) (5). Genus Petromica Topsent. P. massalis Dendy, 1905; Thomas, 1985: 304 (Gulf of Mannar); Bur- ton, 1928: 110; Pattanayak, 1998 (Andamans) (5,7). Genus Spongosorites Topsent. S. aplysinoides (Dendy); Burton, 1937: 38; Thomas, 1985: 301, pl. V, fig. 16 (Gulf of Mannar as Trachyopsis aplysinoides) (5). S. andamanensis Pattanayak, 1998 (Andamans) (7). S. cavernosa (Topsent, 1896); Burton, 1937: 38 (Gulf of Mannar) (5). S. halichondrioides Dendy, 1905; Burton, 1928: 118; Pattanayak, 1998 (Andamans); Thomas, 1985: 300, pl. V, fig. 15 (Gulf of Mannar, as Trachyopsis halichondroides) (5,7). S. solida Topsent; Burton, 1937: 38 (Gulf of Mannar) (5). Genus Topsentia Berg. T. nigrocutis (Carter); Thomas, 1985: 325, pl. VI, fig. 29 (Gulf of Man- nar) (5). Order Haplosclerida Family Callyspongiidae de Laubenfels Genus Callyspongia Duchassaing & Michelotti. C. barodensis Burton; Thomas, 1985: 250 (Gulf of Mannar) (5). C. ceylonica (Dendy, 1905); Thomas, 1985: 249, pl. II, fig. 17 (Gulf of Man- nar) (5). C. clathrata (Dendy, 1905); Thomas, 1985: 249 (Gulf of Mannar) (5). C. diffusa (Rid- ley); Burton, 1930: 20; Rao, 1941: 432; Thomas, 1985: 247, pl. II, fig. 13 (Gulf of Mannar); Ali, 1956: 292 (Madras); Thomas, 1979b: 15 (Mini- coy L) (2,5). C. fibrosa (Ridley & Dendy); Burton, 1930: 669; Burton, 1937: 21; Thomas, 1985: 248, pl. II, fig. 14 (Gulf of Mannar) (5). C. fistularis (Topsent); Burton, 1937: 21; Thomas, 1985; 248 (Gulf of Mannar) (5). C. pambanensis Rao, 1941: 441; Thomas, 1985: 249, pl. IL, fig. 16 (Gulf of Mannar) (5). C. spinosissima (Dendy, 1887); Burton, 1937: 21; Thomas, 1985: 248, pl. II, fig. 15 (Gulf of Mannar) (5). Genus Siphonochalina Schmidt. S. crassifibra Dendy, 1889; Dendy, 1916b: 114 (Okhamandal) (3). S. minor Dendy, 1916b: 115, pl. IL, fig. 15 (Okhamandal) (3). 1 450 Family Chalinidae Gray Genus Adocia Gray. A. carnosa (Dendy, 1889); Burton, 1937: 19 (Gulf of Mannar) (5). A. semi- fibrosa (Dendy, 1916b: 110) Okhamandal, as Reniera semifibrosa); Burton, 1937: 19 (Gulf of Mannar) (3,5). Genus Gellius Gray. G. flagellifer Ridley & Dendy; Burton, 1928: 114; Pattanayak, 1998 (Anda- mans) (7). G. fibulatus (Schmidt); Kumar, 1925: 218 (Waltair) (6). G. fibulatus var. microsigma Dendy, 1916b: 107 (Okhamandal) (3). G. mega- stoma Burton, 1928: 115; Pattanayak, 1998 (Andamans) (7). G. ridleyi Hentschel; Dendy, 1916b: 107 (Okhamandal) (3). G. toxius Kumar, 1925: 219 (Waltair) (6). Genus Haliclona Grant. H. camerata (Ridley); Burton, 1937: 17; Thomas, 1985: 233, pl. 1, fig. 23 (Gulf of Mannar) (5). H. implexa (Schmidt); Thomas, 1985: 233, pl. I, fig. 22 (Gulf of Mannar) (5). H. madrepora (Dendy, 1889); Burton, 1937: 17; Thomas, 1985: 233, pl. 1, fig. 24 (Gulf of Mannar) (5). H. obtusispiculifera (Dendy, 1905); Burton, 1937: 17; Thomas, 1985: 234, pl. I, fig. 27 (Gulf of Mannar) (5). H. occulata (Linnaeus); Rao, 1941: 429; Thomas, 1985: 232, pl. I, fig. 20 (Gulf of Mannar) (5). H. pigmentifera (Dendy, 1905); Burton, 1937: 19 (as Adocia pigmenti- Jera); Thomas, 1985: 234, pl. I, fig. 26 (Gulf of Mannar) (5). H. tenuiramosa (Burton, 1930: 666) (as Chalina tenuiramosa); Burton, 1937; 17; Thomas, 1985: 235, pl. I, fig. 28 (Gulf of Mannar) (5). H. viridis (Duch. & Mich.); Burton, 1937: 18; Thomas, 1985: 232, pl. I, fig. 21 (Gulf of Mannar) (5). Genus Reniera Nardo. R. delicatula Ali, 1956: 290, figs 1,4 (Madras) (5). R. fibroreticulata Den- dy, 1916b: 110 (Okhamandal) (3). R. hornelli Dendy, 1916b: 110 (Okhamandal) (3). R. permollis (Bowerbank); Dendy, 1916b: 109 (Okhamandal) (3). R. topsenti Thiele; Dendy, 1916b: 109 (Okha- mandal) (3). R. tuberosa (Dendy); Kumar, 1925: 220 (Ganjam coast) (6). Genus Sigmadocia de Laubenfels. S. carnosa (Dendy, 1905); Thomas, 1985: 240, pl. II, fig. 4 (Gulf of Mannar) (5). S. fibulata (Schmidt); Car- ter, 1880: 48; Thomas, 1985: 239, pl. II, fig. 1 (Gulf of Mannar); Thomas, 1979b: 15 (Minicoy I.) (2,5). S. petrosioides (Dendy, 1905); Thomas, 1985: 240, pl. IL, fig. 3 (Gulf of Mannar) (5). S. pumila (Lendenfeld); Burton, 1937: 20, Rao, 1941: 431; Thomas, 1985: 240, pl. II, fig. 2 (Gulf of Mannar); Thomas, 1979: 15 (Minicoy I.) (2,5). Genus Joxadocia de Laubenfels. T. dendyi (Bur- ton); Thomas, 1985: 241 (Gulf of Mannar) (5). T. ridleyi (Dendy); Thomas, 1985: 241 (Gulf of MEMOIRS OF THE QUEENSLAND MUSEUM Mannar) (5). T. toxius (Topsent, 1897); Thomas, 1985: 241, pl. II, fig. 5 (Gulf of Mannar) (5). Family Niphatidae Van Soest Genus Aka de Laubenfels. 4. diagnoxea Thomas, 1968c: 250; Thomas, 19792: 168; Thomas, 1985: 317, pl. VI, fig. 11 (Gulf of Mannar) (5). A. minuta Thomas, 1972: 343, pl. 2, figs 4,4a; Thom- as, 1979a: 170; Thomas, 1985: 317, pl. VI, fig. 10 (Gulf of Mannar) (5). Genus Amphimedon Duch. & Mich. A. multi- formis (Dendy); Ali, 1956: 290 (Madras, as Pachychalina multiformis var. mannarensis) (5). Genus Gelliodes Ridley. G. cellaria (Rao, 1941: 39) (as Callyspongia cellaria var. fusca); Thom- as, 1985: 238, pl. I, fig. 32 (Gulf of Mannar) (5). G. fibrosa Dendy, 1905; Dendy, 1916b: 108 (Ok- hamandal); Thomas, 1985: 237 (Gulf of Mannar) (3,5). G. fibulatus (Carter); Burton, 1928: 115, Pattanayak, 1998 (Andamans) (7). G. incrustans Dendy, 1905; Thomas, 1985: 238 (Gulf of Man- nar) (5). Family Phloeodictyidae Carter Genus Calyx Vosmaer. C. clavata Burton, 1928: 117; Pattanayak, 1998 (Andamans) (7). Genus Oceanapia Norman. O. arenosa Rao, 1941: 443; Thomas, 1985: 255 (Gulf of Mannar); Ali, 1956: 292 (Madras) (5). O. fistulosa (Bower- bank); Carter, 1880: 37 (as Desmacidon Jeffreysii); Rao, 1941: 443; Thomas, 1985: 254, pl. IL, fig. 25 (Gulf of Mannar) (5). O. media (Thiele); Burton, 1937: 22; Thomas, 1985: 254, pl. I, fig. 26 (Gulf of Mannar) (5). O. putridosa Burton, 1928: 118 (Orissa coast, as Phloeo- dictyon putridosa) (6). O. sagittaria (Sollas); Thomas, 1985: 242, pl. IL, fig. 6 (Gulf of Mannar) (5). O. zoologica (Dendy); Thomas, 1985: 254, pl. IT, fig. 27 (Gulf of Mannar) (5). Family Petrosiidae Van Soest Genus Petrosia Vosmaer. P. nigricans Lindgren; Thomas, 1985: 246, pl. II, fig. 11 (Gulf of Man- nar) (5). P. similis Ridley & Dendy; Burton, 1930: 666 (as Chalina similis); Thomas, 1985: 246, pl. II, fig. 10 (Gulf of Mannar) (5). Genus Strongylophora Dendy. S. durissima Den- dy, 1905; Thomas, 1985: 242, pl. II, fig. 7 (Gulf of Mannar) (5). Genus Xestospongia de Laubenfels. X. exigua (Kirkpatrick); Burton, 1937: 17; Thomas, 1985: 234, pl. I, fig. 25 (Gulf of Mannar) (5). X. testu- dinaria (Lamarck); Burton, 1937: 22; Thomas, MARINE SPONGES OF THE INDIAN REGION 1985: 246, pl. Il, fig. 9 (Gulf of Mannar, as Petrosia testudinaria); Pattanayak, 1998 (Anda- mans) (5,7). Order Dictyoceratida Family Irciniidae Gray Genus /rcinia Nardo. I. aruensis (Hentschel); Burton, 1937: 40 (Gulf of Mannar, as Hircinia aruensis) (5). I. cactiformis Rao, 1941: 459 (Gulf of Mannar) (5). /. fusca (Carter, 1880: 36); Bur- ton, 1937: 40; Rao, 1941: 458 (Gulf of Mannar, as Hircinia fusca); Thomas, 1985: 224, pl. I, fig. 6 (Gulf of Mannar) (5). /. ramodigitata Burton; Rao, 1941: 458 (Gulf of Mannar) (5). I. ramosa (Keller); Burton, 1937: 40 (as Hircinia ramosa); Thomas, 1985: 224, pl. I, fig. 7 (Gulf of Mannar) (5). I. tuberculata (Poléjaeff); Thomas, 1985: 225, pl. L fig. 8 (Gulf of Mannar) (5). Family Thorectidae Bergquist Genus Cacospongia Schmidt. C. mollior Schmidt; Thomas, 1985: 226, pl. I, fig. 11 (Gulf of Mannar) (5). C. scalaris Schmidt, 1862; Thom- as, 1985: 227, pl. I, fig. 12 (GulfofMannar) (5). Genus Fasciospongia Burton. F. anomala (Dendy, 1905); Thomas, 1985: 228, pl. I, fig. 14 (Gulf of Mannar) (5). E cavernosa (Schmidt); Burton, 1937: 42 (as Aplysinopsis reticulata) Rao, 1941; 465 (as Aplysinopsis reticulata); Thomas, 1985: 227, pl. I, fig. 13 (Gulf of Mannar) (5). Genus Hyrtios Duch. & Mich. H. erecta Keller; Burton, 1937: 43 (as Duriella nigra); Thomas, 1985: 220, pl. I, fig. 2 (Gulf of Mannar) (5). Family Spongiidae Gray Genus Hyattella Lendenfeld. H. cribriformis (Hyatt); Dendy, 1916b: 141 (Okhamandal, as Hippospongia clathrata); Burton, 1937: 41 (as Luffariospongia clathrata), Rao, 1941: 464 (as Luffariospongia clathrata); Thomas, 1985: 221, pl. I, fig. 4 (Gulf of Mannar); Thomas, 1979b: 13 (Minicoy I.) (2,3,5). H. intestinalis (Lamarck); Thomas, 1985: 221, pl. 1, fig. 3 (Gulf of Mannar) (5). Genus Phyllospongia Ehlers. P. dendyi Lenden- feld; Thomas, 1973b: 440; Thomas, 1979: 14 (Minicoy I.). P foliascens (Pallas); Thomas, 1979b: 14 (Minicoy I.); Pattanayak, 1998 (Andamans) (2,7). P. papyracea (Esper); Rao, 1941: 454 (Gulf of Mannar) (5). P. papyracea var. polyphylla de Laubenfels; Thomas, 1985: 222, pl. I, fig. 5 (Gulf of Mannar) (5). Genus Spongia Linneaus. S. hispida Lamark; Thomas, 1985: 220 (Gulf of Mannar) (5). S. officinalis var. ceylonensis Dendy, 1905; Burton, 1937: 39; Thomas, 1985: 219, pl. I, fig. 1 (Gulfof Mannar); Thomas, 1979: 12 (Minicoy I.) (2,5). S. officinalis var. fenestrata Rao, 1941: 455 (Gulf of Mannar) (5). Order Dendroceratida Family Dysideidae Gray Genus Dysidea Johnston. D. cineria (Keller); Dendy, 1916b: 140 (Okhamandal as Spongelia cinerea) (3). D. elegans Nardo; Dendy, 1916b: 140 (Okhamandal as Spongellia elegans) (3). D. fragilis (Montagu); Burton, 1937: 41; Rao, 1941: 463; Thomas, 1985: 228, pl. I, fig. 15 (Gulf of Mannar); Thomas, 1979b: 14 (Minicoy I.) (2,5). D. fragilis var. ramosa (Schulze, 1879); Dendy, 1916: 139 (Okhamandal) (3). D, herbacea (Kel- ler); Thomas, 1985: 329, pl. I, fig. 16 (Gulf of Mannar); Thomas, 1979b: 14 (Minicoy I.) (2,5). Genus Spongionella Bowerbank. S. nigra Dendy, 1889; Burton, 1937: 42 (Gulf of Mannar) (5). S. tuberosa Burton, 1937; 42, pl. 9, fig. 58; Thomas, 1985: 301, pl. V, fig. 16 (Gulf of Mannar) (5). Family Darwinellidae Merekowsky Genus Darwinella Müller. D. australiensis Carter; Dendy, 1916b: 139 (Okhamandal) (3). D. mulleri Schulze; Thomas, 1985: 230, pl. I, fig. 19 (Gulf of Mannar) (5). Genus Dendrilla Lendenfeld. D. cactos (Selen- ka); Burton, 1937: 42 (as Spongionella pulvilla); Thomas, 1985: 229, pl. I, fig. 17 (Gulf of Mannar) (5). D. membranosa (Pallas); Burton, 1937: 43 (Gulf of Mannar) (5). D. nigra (Dendy, 1889); Burton, 1937: 43; Thomas, 1985: 230, pl. I, fig. 18 (Gulf of Mannar) (5). Genus Hexadella Topsent. H. purpurea Burton, 1937: 43; Thomas, 1985: 231 (Gulf of Mannar) (5). Family Dictyodendrillidae Bergquist Genus Dictyodendrilla Bergquist. D. retiaria (Den- dy, 1915: 137, pl. IV, fig. 27) (Okhamandal) (3). Order Verongida Family Aplysinidae Carter Genus Aplysina Nardo. A. lacunosa (Lamarck); Thomas, 1985: 225, pl. I, fig. 9 (Gulf of Mannar) (5). Family Druinellidae Lendenfeld Genus Druinella Lendeteld. D. purpurea (Carter, 1880: 36) (as Aplysilla purpurea); Thomas, 1985: 231 (Gulf of Mannar as Prammaplysilla purpurea) (5). Class Calearea Subclass Calcinea Order Clathrinida Family Clathrinidae Minchin Genus Clathrina Gray. C, coriacea (Montagu): Pattanayak, 1998 (Andamans) (7). Family Leucettidae de Laubentels Genus Pericharax Poléjacif. P. heteroraphis Poléjaeff; Burton & Rao, 1932: 304; Pattanayak, 1998 (Andamans) (7). Subclass Calearonea Order Leucosolenida Family Grantiidae Dendy Genus Leucandra Haeckel. L. donani var. tenuiradiata Dendy, 1916a: 86, pl. 1, fig. 4a,b (Okhamandal) (3). L. dwarkaensis Dendy, 1916a: 86, pl. L fig. 6. pl. II, fig. 10 a-e (Ok- hamandal) (3). L. wasinensis (Jenkin); Dendy, 1916a: 87, pl. f, fig. 5 (off Dwarka) (3). Genus Ute Schmidt. U, syconoides (Carter); Bur- ton & Rao, 1932: 305 (Tuticorin) (5). Family Heteropiidae Dendy Genus Grantessa Lendenteld. G. hastifera (Row); Dendy, 1916a: 81, pl. I, fig. 2,2a, pl. I, fig, 7a-f (off Dwarka) (3). Genus Heteropia Carter. H. glomerosa (Bower- bank); Dendy, 19163: 83, pl. I, fig. 3a-b, pl. I, fig. Sa-g (Okhamandal) (3). Family Sycettidae Dendy Genus Syeor Risso. S. grantioides Dendy. 1916a: 79. pl, L fig. 1 (off Dwarka) (3). Class Hexactinellida Suborder Amphidiscophora Order Amphidiscosida Family Hyalonematidae Gray Genus Hyalonema Gray, H, uculeatum Schulze; Schulze, 1902: 31, pl. II, figs 1-14; Pattanayak, 1998 (Andamans) (7). H. affine Marshall; Schulze, 1902: 27, pl. VIL; Pattanayak, 1998 (Andamans) (7). H. aleoeki Schulze, 1895; Schulze, 1902: 23, pl. VI, figs 1-8 (Laccadive MEMOIRS OF THE QUEENSLAND MUSEUM Sea) (2). H. indicum Schulze, 1895; Schulze, 1902- 10, pl. HL figs 1-13, pl. IV, figs 1-14 (Andamans, Laccadives); Pattanayak; 1998 (An- damans) (2,7). H. lamella Schulze, 1900; Schulze, 1902: 15, pl. XIX (Cape Comorin); Pattanayak, 1998 (Andamans) (5,7). A. martabanense Schulze, 1900; Pattanayak, 1902 (Andamans) (7). H. masoni Schulze, 1895; Schulze, 1902: 13, pl. Vi Pattanayak. 1998 (Andamans) (7). H. nicobaricum Schulze, 1904 (Nicobars) (8), H. rapa Schulze, 1900; Schulze, 1902: 18, pl. XVII (Bay of Bengal); Pattanayak, |998 (Andamans) (7.9). H. weltneri Schulze, 1900; Schulze, 1902: 36, pl. IV, figs 15-24 (Laccadive Sea) (2). Genus Lophophysema Schulze. L. inflatum Schulze, 1902: 38, pl. XX, XXI; Pattanayak, 1998 (Andamans) (7). Family Pheronematidae Gray Genus Pheronema Leidy. P. raphanus Schulze, 1895, Schulze, 1902: 5, pl. 1; Pattanayak, 1998 (Andamans) (7). Genus Semperella Gray. S. cucumis Schulze, 1895; Schulze. 1902: 41, p. VIII; Pattanayak, 1998 (Andamans) (7). Subclass Hexasterophora Order Hexactinosida Family Aphrocallistidae Gray Genus Aphrocallistes Gray. A. beatrix Gray; Schulze, 1902: 87, pl. XV, figs 1-13; Burton & Rao, 1932: 302; Dendy & Burton, 1926: 227, Pattanayak, 1998 (Andamans) (7). A. bocagei Wright; Schulze, 1902: 93, pl. XVI (Andamans, Cape Comorin); Pattanayak, 1998 (Andamans) (5.7). A. ramosus Schulze; Schulze, 1902: 97, pl. XV, fig. 14; Pattanayak, 1998 (Andamans) (5). Family Farreidae Gray Genus Farrea Bowerbank. E occa Bowerbank; Schulze, 1902: 86 (Andamans); Dendy & Bur- ton, 1926: 226 (Cape Comorin, Andamans) (5,7). Family Tretodictyidae Schulze Genus Hexactinella Carter. H. minor Dendy & Burton, 1926: 227; Pattanayak, 1998 (Andamans) (7). MARINE SPONGES OF THE INDIAN REGION 45 Order Lyssacinosida Family Euplectellidae Gray Sub family Corbitellinae ljima Genus Dictyaulus Schulze. D. elegans Schulze, 1895; Schulze, 1902: 70, pl. XII (Laccadive, Cape Comorin) (2,5). Genus Regadrella Schmidt. R. decora Schulze, 1900; Schulze, 1902: 67, pl. XXII, figs 10-18 (Cape Comorin) (5). Subfamily Euplectellinae Ijima Genus Euplectella Owen. E. aspera Schulze, 1895; Schulze, 1902: 59, pl. XI (Laccadive Sea) (2). E. aspergillum Owen; Burton & Rao, 1932: 302 (Andamans) (7). E. regalis Schulze, 1900; Schulze, 1902; 61, pl. XXII, figs 1-9; Pattanayak, 1998 (Andamans). E. simplex Schulze, 1895; Schulze, 1902: 51, pl. X; Pattanayak, 1998 (Andamans) (7). Family Rossellidae Gray Sub family Lanuginellinae Schulze Genus Lophocalyx Schulze. L. spinosa Schulze, 1900; Schulze, 1902: 82, pl. XXIII; Pattanayak, 1998 (Andamans). INCERTAE SEDIS Cryptospongia enigmatica Burton, 1928: 133, pl. 2, fig. 5; Pattanayak, 1998 (Andamans) (7). Protoschmidtia cerebrum Burton, 1928: 116, pl. 1, fig. 2; Pattanayak, 1998 (Andamans) (7). ACKNOWLEDGEMENTS I thank the Director, Zoological Survey of In- dia for facilities and permission to examine collections in the general Non-Chordate section of the ZSI, and the Officer-in-charge, Estuarine Biological Station of ZSI, Berhampur for encour- agement to work on sponges. Several courtesies have been extended by colleagues and friends, to whom I am thankful. LITERATURE CITED ALI, M.A. 1956. Addition to the sponge fauna of Ma- dras. Journal of the Madras University B 26(2): 289-301. ANNANDALE, N. 1911. Some sponges associated with gregarious molluscs of the family Ver- metidae. Records of the Indian Museum 6: 47-55. 1915a. Indian boring sponges of the family Clion- idae. Records of the Indian Museum 11: 1-24, 1915b. Some sponges parasitic on Clionidae with further notes on that family. Records of the In- dian Museum 11: 457-478. uy 1915c. Fauna of Chilka Lake. Memoirs of the Indian Museum 5: 21-54. BOWERBANK, J.S. 1873. Report on a collection of sponges found at Ceylon by E.W.H. Holdsworth Esq. Proceedings of the Zoological Society of London 1873: 25-31, pls 5-7. BURTON, M. 1928. Report on some deep sea sponges from the Indian Museum collected by the R.I.M.S. ‘Investigator’ Part IL Tetraxonida (concluded) and Euceratosa. Records of the Indian Museum 30(1): 109-138. 1930. Addition to the sponge fauna of Gulf of Mannar. Annals and Magazine of Natural His- tory (5)10: 665-676. 1937. Supplement to the littoral fauna of Krusadai Island. Bulletin of the Madras Government Mu- seum, Madras 1(2) pt. 4: 1-58, pls1-IX. BURTON, M. & RAO, H.S. 1932. Report on the shal- low-water marine sponges in the collection of the Indian Museum. Records of the Indian Museum 34(3): 299-356. CARTER, H.J. 1880. Report an specimens dredged up from the Gulf of Mannar and presented to the Liv- erpool Free Museum by Capt. W.H. Cawne Warren. Annals and Magazine of Natural History (5)5: 437-457, pls 18-19; Series (5)6: 35-61; 129- 156, pls 4-8. 1881. Supplementary report on specimens dredged up from the Gulf of Mannar together with others from the sea in the vicinity of the Basse Rocks & from Bass’s Straits respectively, presented to the Liverpool Free Museum by Capt. W.H. Cawne Warren, Annals and Magazine of Natural History (5)7: 361-385, pls 18. 1887. Report of Marine sponges, collected from King Island in the Mergui Archipelago, collected for the Trustees of the Indian Museum, Calcutta by Dr. John Anderson. Journal of the Linnaean Society London 21: 61- 84, pls 5-7. DENDY, A. 1887. The sponge fauna of Madras. A re- port on a collection of sponges obtained in the neighbourhood of Madras by Edgar Thurston Esq. Annals and Magazine of Natural History (5)20: 153-165, pls 9-12. 1889. Report on a second collection of sponges from the Gulf of Mannar. Annals and Magazine of Natural History (6)3: 73-99, pls 3-5. 1905. Report on the sponges collected by Prof. Herdman, at Ceylon, in 1902. Report to the Gov- ernment of Ceylon on the Pearl Oyster Fisheries of the Gulf of Mannar, Royal Society, London, Supplement 3(18): 57-246. 1916a. Report on the calcareous sponges collected by Mr. James Hornell at Okhamandal in Kattiawar in 1905-1906. Report of Govt. of Baroda on the Marine Zoology of Okhamandal (2) 17: 79-91. 1916b. Report on the non-calcareous sponges col- lected by Mr. James Hornell at Okhamandal in Kattiawar in 1905- 1906. Report of Government 454 MEMOIRS OF THE QUEENSLAND MUSEUM of Baroda on the Marine Zoology of Okhamandal (2) 17: 93-146, pls 1-4. DENDY, A. & BURTON, M. 1926. Report on some deep sea sponges from the Indian Museum col- lected by the R.L.M.S. ‘Investigator’. Part-I. Hexactinellida & Tetraxonida (Pars.). Records of the Indian Museum 28: 225-248. HOOPER, J.N.A. 1996: Revision of Microcionidae (Porifera: Poecilosclerida: Demospongiae), with description of Australian species. Memoirs of the Queensland Museum 40: 1-626. HOOPER, J.N.A. & WIEDENMAYER, F. 1994, Porifera. Pp 1-624. In Wells, A. (ed.) Zoological Catalogue of Australia. Vol. 12. (CSIRO Austra- lia: Melbourne). KUMAR, A. 1924. On some Tetraxonid sponges in the collection of the Indian Museum. Calcutta. Pro- ceedings of the 11th Indian Science Congress (1924): 111. 1925. Report on some Tetraxonid sponges in the collection of Indian Museum. Records of the In- dian Museum 27; 211-229, PATTANAYAK, J.G. 1995. Fauna of Chilka Lake. Porifera. In Director Zoological Survey of India, Calcutta (ed.) Wetland Ecosystem series I: 221-226. 1998. Diversity of sponges in Andaman and Nicobar Islands, India. (PhD Thesis, University of Calcutta:Calcutta). RAO, H.S. 1941. Indian and Ceylon sponges in the Naturalhistoriska Riksmuseet, Stockholm, col- lected by K. Fristedt. Records of the Indian Museum 43: 417-496. SCHULZE, F.E. 1894, Hexactinelliden des Indischen Ocean. 1. Die hyalonematiden. Aus Der Abhand- lungen Der Konig! Preuss Akademie Der Wissenschaften Zu Berlin Vom Jahre (1894): 1- 60. 1895. Hexactinelliden des Indischen Ocean. 2. Die Hexasterophora. Aus Der Abhandlungen Der Konig] Preuss Akademie Der Wissenschaften Zu Berlin Vom Jahre (1895): 1-92. 1900. Hexactinelliden des Indischen Oceanes. Aus Der Abhandlungen Der Konig] Preuss Akademie Der Wissenschaften Zu Berlin Vom Jahre (1900): 1-92. 1902. An account of the Indian Triaxonia collected by the Royal Indian Marine Survey Ship ‘Investigator’ translated into English by R. von Lendenfeld. Pp 1-113. (Trustees of the Indian Museum: Calcutta). 1904. Die Hexactinellia. Wiss Ergebn dt. Tiefsee-Exped. ‘Valdivia’ 1898-1899 4: 1-265. THOMAS, P.A. 1968a. The sponge resources of India. Pp. 31-32. In Symposium on the living resources of the seas around India. (Central Marine Fish- eries Research Institute: Mandapam Camp). 1968b. Studies on Indian sponges -I. Two new spe- cies of sponges belonging to the genera Echinodictyum Ridley and Rhabderemia Topsent (class Demospongiae Sollas, order Poecilosclerida Topsent). Journal of the Marine Biological Association of India 10(2): 245-249. 1968c. Studies on Indian sponges-II. Two new spe- cies of silicious sponges belonging to the genera Aka de Laubenfels and Damirina Burton. Journal of the Marine Biological Association of India 10(2): 250-254. 1968d. Studies on Indian sponnges-II]. Two species of the siliceous sponges of the family Ophlitaspongiidae de Laubenfels (class Demospongiae, order Poecilosclerida Topsent). Journal of the Marine Biological Association of India 10(2): 255-259. 1968e. Studies on Indian sponges-IV. Additions to the genus Corticium Schmidt with notes on the distribution of Corticium candelabrum Schmidt. Journal of the Marine Biological Association of India 10(2): 260-263. 1968f. Studies on Indian sponges-V. Two new re- cords of siliceous sponges belonging to the families Myxillidae Hentschel and Spira- strellidae Hentschel from the Indian region. Journal of the Marine Biological Association of India 10(2): 264-268. 1969a. Boring sponges of the reefs of Gulf of Mannar and Palk Bay. P. 19. In Symposium on corals and coral reefs. (Marine Biological Asso- ciation of India: Mandapam Camp). 1969b. Catalogue of sponges in the reference col- lection ofthe Central Marine Fisheries Research Institute, Mandapam Camp. Bulletin Central Marine Fisheries Research Institute (7): 13-21. 19702. On some deep sea sponges from the Gulf of Mannar with descriptions of three new species. Journal of the Marine Biological Association of India 12(1&2): 202-209, 1970b. Studies on Indian sponges - VI. Two new re- cords of siliceous sponges (Poecilosclerida: Tedaniidae) from the Indian region. Journal of the Marine Biological Association of India 12(1&2): 43-50. 1970c. Studies of Indian sponges-VII. Two new re- cords and a new species of the genus Plakina Schulze from the Indian region. Journal of the Marine Biological Association of India 12(1&2): 51-56. 1972. Boring sponges of the reefs of Gulf of Mannar and Palk Bay. Pp. 333-362. In Proceedings of a Symposium on Corals and Coral Reefs, 1969. (Marine Biological Association of India: Cal- cutta). 1973a. The sponge resources of India. Pp. 693-699. In Proceedings of a Symposium on Living Re- sources of India. (Marine Biological Association of India: Calcutta). 1973b. Two new records of Demospongiae from the Indian Ocean. Journal of the Marine Biological Association of India 15(1): 430-442. 1974. A new genus and species (Qasimella indica) of Demospongiae from Indian seas. Journal of MARINE SPONGES OF THE INDIAN REGION the Marine Biological Association of India 16(1): 311-313. 1975. Boring sponges of Zuari and Mandovi esturies. Bulletin of the Department of Marine Science, University of Cochin 7(1): 117-126. 1976a. Endectyon lamellosa n. sp. (Demospongiae: Poecilosclerida, Raspailiidae) from the Indian seas and a revised key to the Indian species of Endectyon Topsent. Journal of the Marine Bio- logical Association of India 18(1) : 169-172. 1976b. History of spongiology ofthe Indian Ocean. Journal of the Marine Biological Association of India 18(3): 610-625. 1977. Studies on Indian sponges VIII. Four new re- cords of siliceous sponges Echinochalina glabra (Ridley & Dendy), Higginsia mixta (Hentschel), Geodia lindgreni (Lendenfeld) and Pacham- philla dendyi Hentschel from the Indian Ocean. Journal of the Marine Biological Association of India 19(1): 115. 1979a. Boring sponges destructive to economically important molluscan beds and coral reefs in the Indian seas. Indian Journal of Fisheries 26(1&2): 163-200. 1979b. Demospongiae of Minicoy Island (Indian Ocean) Part-I, orders Keratosida and Haplo- sclerida. Journal of the Marine Biological Association of India 21(1&2): 10-16. 1980a. Demospongiae of Minicoy Island (Indian Ocean) Part-II order Poecilosclerida. Journal of the Marine Biological Association of India 22(1&2): 1-7. 1980b. Demospongiae of Minicoy Island (Indian Ocean) Part-II]. Order Halichondrida and Hadromerida, Epipolasida and Choristida. Jour- nal of the Marine Biological Association of India 22(1&2): 8-20. 1983. Distribution and affinities ofthe sponge fauna of the Indian region. Journal of the Marine Bio- logical Association of India 25(1&2): 7-16. 1984. Sponges collected aboard R.V. ‘Skipjack’ from the south east coast of India. Journal of the Marine Biological Association of India 26(1&2): 95-102. 1985. Demospongiae of the Gulf of Mannar and Palk Bay. Pp 205-365. In James, P.S.B.R. (ed.) Recent Advance in Marine Biology (Today To- morrow's Printers and Publishers: New Delhi). THOMAS, P.A., RAMADOSS, K. & VINCENT, S.G. 1993, Invasion of Cliona margaritifera Dendy and C. lobata Hancock on the molluscan beds along the Indian coast. Journal of the Marine Bio- logical Association of India 35(1&2): 145-156. THOMAS, P.A. & THANAPATI, V. 1980. An ancient window pane oyster bed in Goa with comparative notes on the Oyster in an extent bed. Indian Jour- nal of Fisheries 27(1&2): 54-60. THE IMPACT OF TRAWLING ON SOME TROPICAL SPONGES AND OTHER SESSILE FAUNA. Memoirs of the Queensland Museum 44: 455. 1999:- Following a replicated repeated-trawl depletion experiment, which removed 70-90% of the initial biomass of sessile fauna, non-destructive methods have been used to measure recovery of the structurally dominant species. To date, three post-impact surveys have been completed (1 month, 1 year, 2 years) on 6 trawled swathes and 6 controls. These involved quantitative video observations from a tow-sled and from an ROV with positioning precision of 2m. More than 80 taxa of sessile fauna were recorded and about 15 taxa, including sponges, gorgonians, and alcyonarians and hard corals, were abundant enough for analysis. The attributes measured for each organism were species, position, size (height, width, and area), and condition (proportion intact, dead, or encrusted). These data enabled analyses of density, size, condition and species composition, with high precision. We have demonstrated statistically that the methods are very powerful for detecting recovery on the trawled tracks, but the time series is, as yet, too short to identify any recovery. Nevertheless, in treatment vs control or before vs after comparisons, all taxa analysed for which sample size was adequate, showed significant impact due to trawling, for at least one or more ofthe measured attributes. For all benthos taxa combined, the rate of decrease in benthos density with trawl intensity corresponded to ~10% per trawl, so that in areas trawled 13 times, the benthos density was only ~25% of that in un-trawled areas. This conclusion was very similar to the overall estimate of removal obtained from depletion-regression analysis of the catch obtained during the repeat-trawl experiment. The results also indicated the nature in which trawl impact was manifested for different organisms with different structure and morphology. The sea whips were most resilient to removal, though they could be damaged. Sponges and soft corals were relatively easily removed and hard corals were easily broken. The gorgonians were intermediate and variable in resilience. The interaction between these differences and trawl impact caused a marked change in, and degradation of, the community composition of the seabed habitat. O Porifera, sessile fauna, impact, condition, recovery, trawling, ROV, video, resilience. C.R. Pitcher (email: roland pitcher(a)marine.csiro.au ), C.Y. Burridge, TJ. Wassenberg & G.P. Smith, CSIRO Division of Marine Research, PO Box 120, Cleveland, Qld. 4163, Australia; 1 June 1998. 456 MEMOIRS OF THE QUEENSLAND MUSEUM SPONGE SHAPE AS A TAXONOMIC CHARACTER: THE CASE OF SPONGIA OFFICINALIS AND SPONGIA AGARICINA. Memoirs of the Queensland Museum 44: 456. 1999:- The extreme phenotypical plasticity is one of the main characterizing traits of Porifera at all organizational levels (West Heberard, 1986, 1989; Gaino et al., 1995). Body shape is highly variable both in time and space particularly among demosponges living in shallow waters with high selective pressures exerted by fluctuations of water movement and light, by substrate stability and shape and by spatial competition (Bidder, 1923; Hartman, 1950; Reiswig, 1973; Wilkinson & Vacelet, 1979; Palumbi, 1986; Barthel, 1986, 1991; Pansini & Pronzato, 1990; Gaino etal., 1991; Pronzato & Pansini, 1994). Moreover external morph can be influenced, in some cases, by the age of the sponge and therefore by its life cycle (Barthel, 1986; Manconi & Pronzato, 1991). Spongia officinalis is reported as being very variable both in colour and body shape not only by spongologists (de Laubenfels, 1948; Storr, 1976; Pronzato et al., in press), but also by fishermen that encounter difficulties distinguishing it from other species as S. zimocca or Ircinia spinosula and Cacospongia scalaris. On the other hand S. agaricina displays such a peculiar shape that it is easily identified also by students of zoology. Such a different degree of body shape variation in these two close species is an intriguing case. The present paper aims to investigate the temporal and spatial evolution of body shape of Spongia officinalis and Spongia agaricina in order to ascertain if and what external morphological traits can have a diagnostic value to clarify their taxonomic status. The trait of body shape could considered as the result of several associated sub-traits as growth in height; growth in width; number and shape of oscules; distribution of oscules; presence and distribution of lobes; differentiation of inhalant and exalant area; presence and distribution of conules. A comparative analysis was performed between Eastern and Western Mediterranean populations of S. officinalis. The rarity of S. agaricina in the Ligurian Sea meant it was not possible to carry out a comparison with the studied Aegean population. The following material was considered: S. officinalis: 56 specimens collected at 15m depth on hard bottoms by diving at Portofino; 63 specimens at 5-10m depth on hard bottom around the island of Crete; 50 living specimens from Portofino: 25425 sponges settled, respectively, at 8-10 and 20-25m depth. S. agaricina: 37 specimens collected, at 50- 100m depth, on sand bottoms by gangava around the island of Kalymnos. Living sponges were monitored by under-water photography from 1994-1995 in September, July and November to follow the temporal evolution of body shape. To avoid morphological variations linked to pumping activity and to rhythmic contraction/expansion processes, cleaned skeletons were studied at the laboratory. The identification of specimens was performed by SEM at the level of the skeletal net. Results highlight that S. agaricina displays a constant body plan in spite of a wide size variation within the considered population. In the case of S. officinalis a constant body plan is displayed by living sponges within the same population at different depth. Spongia agaricina: body shape is relatively constant; cup-shaped with distal margin ondulati, lateral profile trapezio-like; height and max diameter (at the cup aperture) show an isometric growth; diameter at the base seems to be allometric with respect to the other growth axis and constrained by the substrate morphology; the fan-like shape is absent, with the exception of two specimens, in this population; the distal margin is constantly elliptic with a low difference among the two axis; the irregular growth of the distal margin seems not linked to sponge size. It is possible to hypothesize a shape shifting during ageing from an opened cup toward a tronco-conic shape (un po' tirata). C1 Porifera, bath sponges, body shape, variability, Spongia officinalis, Spongia agaricina. R. Pronzato (email: zoologia(a)igecuniv cisi.unige.it), M. Sidri, M. Dorcier, M Cappello, Istituto di Zoologia dell'Università, via Balbi 5, 1-16126 Genova, Italy; R. Manconi, Dipartimento di Zoologia e Antropologia Biologica dell Universita, via Muroni 25, -07 100 Sassari, Italy; 1 June 1998. RESOURCE PARTITIONING BY CARIBBEAN CORAL REEF SPONGES: IS THERE ENOUGH FOOD FOR EVERYONE ? ADELE J, PILE Pile, A.J. 1999 06 30: Resource partitioning by Caribbean coral reef sponges: is there enough food for everyone? Memoirs of the Queensland Museum 44: 457-461. Brisbane. ISSN 0079-8835. Sponges are known to graze primarily on the ultraplankton fraction (plankton < 5um) of the water column community and haye been implicated as primary coral reef consumers of ultraplankton, but it is unknown if there is inter- or intraspecies competition for food resources. I characterised diet and retention efficiency of three co-occuring species of sponge at Chub Cay Reef, Bahamas (25722782"N, 77°51°93"W), The erect tube sponge Callyspongia vaginalis, the mounding sponge Spongia tubulifera, and small Aplysina Jistularis were conspicuous and common members of the benthic community, and had mean heights above the substrate of 22.5, 7.0, and 1.2cm, respectively. Ambient and exhalent current water samples were collected by snorkelers and analyzed for ultraplankton using flow cytometry. Callyspongia vaginalis retained only Synechocaceus-lype cyanobacteria with an efficiency of 90%. In contrast, the diets of S. tubulifera and A. fistularis were more reflective of the overall water column community consisting of heterotrophic bacteria, Prochlorococens, SynechococcusAype cyanobacteria and autotrophic picoeucaryotes. Spongia tubulifera had retention efficiencies of 41, 29, and 86% for heterotrophic bacteria, Prochlorococcus, and Synechococcus-type cyanobacteria, respectively. Retention efficiencies were highest for 4. fistularis, the smallest sponge, with 96% for heterotrophic bacteria, 95% for Prochlorococeus, 99% for Synechococcus-type cyanobacteria and 100% for autotrophic picoeucaryotes. Food availability increased closer to the benthos such that an order of magnitude more ultraplankton cells were available to S. tubulifera and A. fistularis. Overall low abundance of food particles (<105 cells mh?!) 22cm above the benthos may prevent effective capture by choanocytes. Competition for food resources between phylla is most likely the cause of the resource partitioning found at this location rather than competition between sponges, O Porifera, feeding, ultraplankton, Caribbean, coral reef. competition. Adele J, Pile (email-Adele.Pile@flinders.edi.au), The Flinders University of South Australia, Department of Biological Sciences, GPO 2100, Adelaide 5001, SA, Australia: 29 January 1999, Given that oligotrophic conditions inherently characterise coral reefs, it is not surprising that they are net sinks for all types of planktonic foods, such as zooplankton (Glynn, 1973), nanoplankton (Glynn, 1973), and picoplankton (Buss & Jackson, 1981; Ayukai, 1995; Charpy & Blanchot, 1998), consumed by sessile benthic organisms. However, difficulties in measuring food availability at a scale relevant to these organisms themselves has restricted our understanding of the role of competition for food by benthic suspension feeders, This primarily has been limited by large sample sizes required to quantify naturally occurring. low densities of many food types. Ultraplankton (plankton = Sum; Murphy & Haugen, 1985) is the most abundant food source on coral reefs both numerically and in terms of total carbon ( Ayukai, 1992, 1995; Pile, 1997; Charpy & Blanchot, 1998), and has recently been found to be the major component of the diet of sponges (Reiswig, 1971; Pile, 1997), ascidians (Pile & Young, in review), and soft corals (Fabricius et aL, 19953, b; Ribes et al., 1998) common to coral reels. Considering that the potential guild of active and passive suspension feeders that will graze on ultraplankton is quite large it is reasonable to suspect that competition for food resources could limit the distribution of some organisms. Sponges are known to graze primarily on the ultraplankton fraction of the water column community (Pile et al., 1996, 1997; Pile, 1997), and have been implicated as the primary coral reef consumers of ultraplankton (Reiswig, 1971; Pile, 1997; Charpy & Blanchot, 1998) . On Pacific reefs, 90% of the ultraplankton is removed from water that passes over a reef and it 458 TABLE 1. Mean ultraplankton availability (103 cells ml-! + sd, n=10) at Chub Cay, Bahamas. Mean height above bottom (cm) Type of ultraplankton 22 7 1.4 Procaryotes Heterotrophic bacteria | 4.99 (1.37) | 77.1 (48.9) | 116.0 (49.2) Prochlorococcus 2.00 (0.44) | 45.8 (26.6) | 37.2 (17.9) Synechococcus-type ' uyanobaciária 8.59 (12.2) | 93.8(81.9) | 177.0 (15.3) Eucaryotes Autotrophic eucaryotes - " s . .2 (9. f 3 (<3um) 0.10 (0.11) | 16.2(9.37) | 63.9 (48.3) has been suggested that this is the result of grazing by the benthos (Ayukai, 1995). In the Caribbean, sponges are the dominant benthic invertebrate, contributing up to 2.5kgm of the benthic biomass (Wilkinson, 1987). Concurrent with this high biomass the sponge community is very diverse with morphologies ranging from encrusting to massive (Wilkinson, 1987). High abundance and species diversity of sponges coupled with oligotrophic conditions common to coral reefs could require partitioning of food resources between sponges or with other members of the guild of primary consumers of ultraplankton which is not found in more eutrophic ecosystems (Stuart & Klumpp, 1984; Lesser et al., 1992). Abelson et al. (1993) hypothesised that the morphology of coral reef organisms modifies the flow patterns around them such that it predisposes their diets. In their model, organisms with a high slenderness ratio (the ratio between body height and downmost width 7 1) will graze on fine particulate matter whereas organisms with a low slenderness ratio (« 1) will feed primarily on bed load particles. Upright tubular sponges, gorgonians and other soft corals all have high slenderness ratios and it is highly likely that they will utilise the same food resources. Small mounding, massive, and encrusting sponges all have low slenderness ratios and would be able to exploit an unoccupied niche by grazing on ultraplankton 1f all other low slenderness ratio organisms (i.e. flattened types of corals, solitary fungiid coral species, and bryozans) grazed primarily on bed load particles. Therefore, in this study I quantified the food availability and diet of three co-occuring species of demosponges on a coral reef with varying slenderness ratios to determine if there was greater competition for MEMOIRS OF THE QUEENSLAND MUSEUM food resources for species with high slendemess ratios. MATERIALS AND METHODS Diets and retention efficiencies were measured for three co-occuring species of sponge at Chub Cay Reef, Bahamas (25?22'82"N, 77°51793"W). Chub Cay Reef is a patch reef that has a maximum depth of 5m. The erect tube sponge Callyspongia vaginalis, the mounding sponge Spongia tubulifera, and very small Aplysina fistularis were conspicuous and common members of the benthic community and had mean heights (n=10) above the substrate of 22.5 (+ 3.8 sd), 7.0 (+ 1.3 sd), and 1.2 (+ 0.4 sd)cm respectively. Retention of ultraplankton was quantified from 1ml water samples collected using 5cc syringes from 10 individuals of each species while snorkeling to a depth of no greater than 3m. Samples were taken from water adjacent to the sponge and from the exhalent current of each individual and preserved for flow cytometry using standard protocols (Campbell et al., 1994). Ultraplankton populations were quantified using an Epic Elite flow cytometer (Coulter Electronics Corporation, Hialeah, Florida) at Harbor Branch Oceanographic Institution, following the techniques of Marie et al. (1996). Orange fluor- escence (from phycoerythrin), red fluorescence (from chlorophyll), and green fluorescence (from DNA stained with SYBR Green) were collected through band pass interference filters at 575, 680, and 450nm, respectively. The five measured parameters (forward- and right-angle light scatter (FALS and RALS), orange, red, and green fluorescence) were recorded on 3-decade logarithmic scales, sorted in list mode, and analyzed with a custom-designed software (CYTOWIN; Vaulot, 1989). Ultraplankton populations were identified to general cell types of heterotrophic bacteria (HBac), Prochlorococeus (Pro), Synechocaccus-type cyanobacteria (Syn), and autotrophic eucaryotes «3um (Peuc), visually confirmed (except for Prochlorococcus), and mean cell diameter measured (n=50) using epifluorescence microscopy. Differences between cell counts from ambient and exhalent current water of each type of ultraplankton were analyzed using two tailed t-tests for each species of sponge with a Bonferroni-transformed experimentwise ; of 0.00625 to determine the effects of sponges on RESOURCE PARTITIONING IN CARIBBEAN DEMOSPONGES 459 TABLE 2. Mean 103 cells ml-! (+ sd, n=10) in the exhalent current demonstrating the effect of each sponge on the four types of ultraplankton. concentrations (Table 1). * p < 0.00625. Individual t-tests comparing mean cell concentrations to ambient cell Species of Sponge E ger liic Prochlorococcus p do ei boul Callyspongia vaginalis 22,5 5.08 (0.96) | 2.12 (0.58) 0.90* (0.23) 0.04 (0.05) Spongia tubulifera 7.0 45.8 (0.49) 32.3 (0.49) 13.5* (0.29) 9.09 (0.21) Aplysina fistularis 1.2 4.2* (0.42) 1.70* (0.24) 2.03* (1.43) 0.17* (0.06) ultraplankton (Zar, 1984). The mean retention efficiency for each sponge was calculated as ((mean cell count ambient - mean cell count exhalent)/mean cell count ambient)x 100 for each 35 — Callyspongia vaginalis 30 gzzzza amb mean = exh mean 25 20 | 15 * 10 | A LZA EM Spongia tubulifera 10? cells mI" 10 | JPN AN 35 - — Aplysina fistularis 30 10% cells mI” a 10^ cells mi’ a HBac B FIG. 1. Effect of each sponge on ultraplankton populations. Concentration of each type of ultraplankton in ambient water and water from the exhalent currents of each sponge. Stippled bars are for ambient water and black bars for water from the exhalent current. Abbreviations: Hbac- heterotrophic bacteria, Syn= Synechococcus-type cyanobacteria, and Peuk- autotrophic eukaryotes <3um. Note that the y axis is an order of magnitude less for C. vaginalis. * Cell concentrations between ambient water and exhalent current water which are significantly different (paired t-test with a Bonferroni transformed experimentalwise <<0.00625). type of ultraplankton. Student t tests, one tailed, were used to determine if the retention efficiency for each type of ultraplankton was significantly >0 employing a Bonferronni transformed exper- imentwise error of «c=0.0001, p=0.00625. RESULTS Ultraplankton abundance decreased with height above the benthos (Table 1). Abundance at all three heights followed the pattern of Synechococcus-type cyanobacteria as the most abundant cell type followed by heterotrophic bacteria, Prochlorococcus, and autotrophic eucaryotes < 311m were the least abundant. Ultra; plankton abundance increased from RE 57x10* cells mI! at 22cm to 29.1x10* cells mI at 1.4cm from the benthos. Callyspongia vaginalis retained only Synechococcus-type cyanobacteria (Table 2, Fig. 1) with an efficiency of 90% (Fig. 2). In contrast, the diets of S. tubulifera and A. fistularis were more reflective of the overall water column community consisting of heterotrophic bacteria, Prochlorococcus, Synechococcus-type cyano- bacteria and autotrophic picoeucaryotes (Table 2, Fig. 1). Spongia tubulifera had retention efficiencies of 41, 29, and 86% for heterotrophic bacteria, Prochlorococcus, and Synecho- coccus-type cyanobacteria respectively (Fig. 2). Retention efficiencies were highest for A. fistularis, the smallest sponge, with 96% for heterotrophic bacteria, 95% for Prochlor- ococcus, 99% for Synechococcus-type cyanobacteria and 100% for autotrophic pico- eucaryotes (Fig. 2). DISCUSSION Typical of other demosponges all three species grazed primarily on the ultraplankton fraction of the water column community (Reiswig, 1971; Pile et al., 1996, 1997; Pile, 1997). Retention efficiencies by C. vaginalis and S. tubulifera were substantially lower than those previously reported for demosponges and this may be related 460 MEMOIRS OF THE QUEENSLAND MUSEUM 100 4 Si 4 SZ <> XXX SOK 80 4 XK KKM OOOO SS Y, 2 60 + CX Ox <> XX Ses 5 ee x 40 } CX L "V: SS Retention efficiency xx e SxS S 20 4 0 Ys < A.fistularis X © IS CX (905 CX KO e S. tubulifera Species of sponge FIG. 2. Retention efficiency (x + sd, n = 10) for each species of sponge for each type of ultraplankton. Abbreviations: Hbac — heterotrophic bacteria, Syn — Synechococcus-type cyanobacteria, and Peuk = autotrophic eukaryotes <3um to the low abundance of cells available to the sponges (Reiswig, 1971; Pile et al., 1996, 1997; Pile, 1997). When the abundance of ultra- plankton approached those normally found on coral reefs (Ayukai, 1995; Pile, 1997), such as those in water surrounding A. fistularis, retention efficiencies are similar to those previously observed (Reiswig, 1971; Pile, 1997). It should be noted that at Chub Cay Reef Synecho- coccus-type cyanobacteria was the most abundant food source, which is unusual in that bacteria are normally the most abundant food source on coral reefs (Ayukai, 1995; Pile, 1997). Increasing ultraplankton availability nearer to the benthos opposes the pattern of ultraplankton community structure in shallow waters found in the Red Sea (Yahel et al., 1998) and Lake Baikal (Pile et al., 1997) where abundance decreases closer to the benthos. As predicted by the model of Abelson et al. (1993) ultraplankton availability increased closer to the benthos and this trend is most likely due to decreasing competition for it as a food source, Ultraplankton abundance was extremely low (< 10° cells ml’) 22cm above the bottom and availability increased closer to the benthos such that an order of magnitude more ultraplankton cells were available to S. tubulifera and 4. fistularis. Overall low abundance of food particles 22cm above the benthos may be E HBac Pro EXX] Syn KO EN Peuk preventing effective capture by the choano- cytes and merits further investigation. Competition between phylla for food resources is most likely the cause of the resource partitioning found at this reef rather than competition between sponges. The other major benthic organisms at Chub Cay Reef are gorgonian corals Gorgonia flabellum, G. ventalina, Plexaura flexuosa, and P. porosa. Recently, soft corals have been found to signif- C. vaginallis icantly impact ultraplankton communities. In the Caribbean Plexaura flexuosa and P. porosa graze on the ultraplankton fraction >3um (Ribes et al., 1998) while in the Red Sea the soft corals Dendronephthya hemprichi, D. sinaiensis, and Scleronephthya corymbosa and the gorgonian Acabaria sp. have been found to graze on plankton down to Synechococcus-type cyanobacteria (typically 1.2-1.8pm) (Fabricius et al., 1995b). Soft coral biomass is considerable in some communities were sponges are also prolific (Kinzie, 1973) and may be a significant competitor for ultra- plankton. Since soft corals and gorgonians typically have a higher s/r ratio they will most likely impact a zone of water that is higher from the benthos than sponges with a low s/r ratio. Most other organisms with low s/r ratios, such as hard corals, bryozans, and bivalves are typically bed load feeders (e.g. Abelson et al., 1993; Jorgensen, 1996; Riisgárd & Manriquez, 1997). Sponges with a low s/r ratio may be the only group of organisms to graze on ultraplankton. If this is true, then they have cornered a niche which has allowed for their success in benthic commun- ities. ACKNOWLEDGEMENTS I wish to thank my snorkeling buddies Tracy Griffen and Mike Temkin for all their assistance and Ross Longley and the Division of Bio- medical Marine Research of the Harbor Branch RESOURCE PARTITIONING IN CARIBBEAN DEMOSPONGES Oceanographic Institution for graciously providing access to the flow cytometer. This manuscript was improved through the thoughtful comments from two anonymous reviewers and John Hooper. The research was funded through a post doctoral fellowship from the Harbor Branch Institution. LITERATURE CITED ABELSON, A., MILOH, T. & LOYA, Y, 1993. Flow patterns induced by substrata and body morphologies of benthic organisms, and their roles in determining availability of food particles. Limnology and Oceanography 38: 1116-1124. AYUKAI, T. 1992. Picoplankton dynamics in Davies Reef lagoon, the Great Barrier Reef, Australia. Journal of Plankton Research 14: 1593-1606. 1995, Retention of phytoplankton and planktonic microbes on coral reefs within the Great Barrier Reef, Australia. Coral Reefs 14: 141-147. BUSS, L. & JACKSON, J.B.C. 1981. Planktonic food availability and suspension-feeder abundance: evidence of in situ depletion. Journal of Experimental Marine Biology and Ecology 49: 151-161. CAMPBELL, L., NOLLA, H.A. & VAULOT, D. 1994. The importance of Prochlorococcus to community structure in the central North Pacific Ocean. Limnology and Oceanography 39: 954-960. CHARPY, L. & BLANCHOT, J. 1998. Photosynthetic picoplankton in French Polynesian atoll lagoons: estimation of taxa contribution to biomass and production by flow cytometry. Marine Ecology Progress Series 162: 57-70. FABRICIUS, K.E., BENAYAHU, Y. & GENIN, A. 1995a. Herbivory in asymbiotic soft corals. Science 268: 90-92, FABRICIUS, K.E. GENIN, A. & BENAYAHU, Y. 1995b. Flow-dependent herbivory and growth in zooxanthellae-free soft corals. Limnology and Oceanography 40: 1290-1301. GLYNN, P.W. 1973. Ecology of a Caribbean coral reef. The Porites reef-flat biotype: Part II. Plankton community with evidence for depletion. Marine Biology 22: 1-21. JORGENSEN, C.B. 1996. Bivalve filter feeding revisited, Marine Ecology Progress Series 142: 287-302. KINZIE, R.A. 1973. The zonation of West Indian gorgonians. Bulletin of Marine Science 23: 93-155. LESSER, M.P., SHUMWAY, S.E., CUCCI, T. & SMITH J. 1992. Impact of fouling organisms on mussel rope culture: interspecific competition for food among suspension-feeding invertebrates. 461 Journal of Experimental Marine Biology and Ecology 165: 91-102. MARIE, D., PARTENSKY, F., JACQUET, S. & VAULOT, D. 1996. Enumeration and cell cycle analysis of natural populations of marine picoplankton by flow cytometry using the nucleic acid stain SYBR Green I. Applied and Environmental Microbiology 63: 186-193. MURPHY, L.S. & HAUGEN, E.M. 1985. The distribution and abundance of phototrophic ultraplankton in the North Atlantic. Limnology and Oceanography 30: 47-58. PILE, A.J. 1997. Finding Reiswig’s missing carbon: Quantification of sponge feeding using dual-beam flow cytometry. Pp. 1403-1410. In Proceedings of the 8th International Coral Reef Symposium Vol. 2 (Smithsonian Tropical Research Institute: Balboa, Panama). PILE, AJ. & YOUNG, C.M. in review. Seasonal variation in benthic-pelagic coupling of the microbial food web by ascidians. Limnology and Oceanography. PILE, A.J., PATTERSON, M.R. & WITMAN, J.D. 1996. In situ grazing on plankton < 10 mm by the boreal sponge Mycale lingua. Marine Ecology Progress Series 141: 95-102. PILE, A.J., PATTERSON, M.R., SAVARESE, M., CHERNYKH, V.I. & FIALKOV, V.A. 1997. Trophic effects of sponge feeding within Lake Baikal's littoral zone. 2. Sponge abundance, diet, feeding efficiency, and carbon flux. Limnology and Oceanography 42: 178-184. REISWIG, H.M. 1971. Particle feeding in natural populations of three marine demosponges. Biological Bulletin 141: 568-591. RIBES, M., COMA, R. & GILL, J.-M. 1998. Hetero- trophic feeding in symbiotic gorgonian corals. Limnology and Oceanography 43: 1170-1179. RUSGARD, H.U. & MANRIQUEZ, P. 1997. Filter- feeding in fifteen marine ectoprocts (Bryozoa): particle capture and water pumping. Marine Ecology Progress Series 154: 223-239. STUART, V. & KLUMPP, D.W. 1984. Evidence for food-resource partitioning by kelp-bed filter feeders. Marine Ecology Progress Series 16: 27-37. VAULOT, D. 1989. CYTOPC: processing software for flow cytometric data. Signal Noise 2: 8. WILKINSON, C.R. 1987. Interocean differences in size and nutrition of coral reef sponge populations. Science 236: 1654-1657. YAHEL, G., POST, A.F., FABRICIUS, K., MARIE, D., VAULOT, D. & GENIN, A. 1998. Phytoplankton distribution and grazing near coral reefs. Limnology and Oceanography 43: 551-563. ZAR, J.H. 1984. Biostatistical Analysis. (Prentice Hall: New York). 462 MEMOIRS OF THE QUEENSLAND MUSEUM ‘MUD MOUND’ STRUCTURES AND CORALLINE SPONGES FROM OSPREY REEF (QUEENSLAND PLATEAU, CORAL SEA, AUSTRALIA), Memoirs of the Queensland Museum 34: 462. 1999:- Osprey Reetis located atthe NW tip of the Queensland Plateau, This reef represents an open oceanic reef platform on a drowned carbonate platform (Queensland Plateau). The metamorphic basement was drilled to a depth of about 450m below the sea floor (Betzler etal., 1995), The reef caves of Osprey Reef were studied during two one-week expeditions in 1995 and 1996 using the DPI RV “Gwendoline May’. At Osprey Reef a very patchy distribution of coralline sponges was observed. At some dive sites caves and reef internal surfaces (RIS) were free of this fauna, whereas a few hundreds of meters away, the walls of caves and the RIS were covered with coralline sponges. The reasons for this very patchy distribution are not presently clear. The structure of the cave community is the same as found on ihe GBR, The caves with abundant coralline sponges are mainly located between 15-20m depth. At all sites, 4strosclera was the most dominant sponge, and similar to populations on the outer barrier reefs of the GBR, it sometimes lives in semi-dark conditions and is colored red or green. Its size never exceeds 5em. Spirastrella (Acanthochaetetes) was rare al Osprey Reef compared lo the GBR, We found two new species of colonial variations of the 'sphinetozoan' sponge laceleria at Osprey Reef which still remain unnamed and largely undescribed, although preliminary notices of their occurrence and brief descriptions are provided by Reitner & Wórheide (1995) and Wórheide & Reitner (1996). These new *“sphinctozoans” are common in most of the caves and are mostly associated with Vace/etia crypta. They occur in only the darkest parts of the caves, and from their large biomass and ‘insinuating habit’ (i.e. growing between dead coral), they appear to be important components of reef structure, This type of coralline sponge shows similarities to Permo-Triassic reef building sphinctozoans. At 250-300m depth aggregates of medium sized mound structures were observed, These structures are located at the NW steep escarpment of Osprey Reefand are ca. 10-20m long, 1-2m wide, and approximately 0.5-2m high. The surface is rigid and sometimes overgrown with sponges and gorgonians. Between the elongated mounds groove systems are developed where reef sediments are transported. Little sediment ‘snow’ is fixed by the mound surfaces. These mound structures are comparable with micritic sponge reefs known from Mesozoic reef sites. O Porifera, coralline Sponges, mud-mounds, Vacelelia, reef-building sphinctozoans, Osprey Reef, Coral Sea. Literature cited. BETZLER, CHR.. BRACHERT, T. & KROON, D. 1995. Role of climate in partial drowning of the Queensland Plateau carbonat platform (northeastern Australia). MarineGeology 123: 11-32, REITNER, J, & WORHEIDE, G. 1995, New Recent sphinetozoan coralline sponge from the Osprey Reef (N'Queensland Plateau, Australia). Fossil . Cnidaria & Porifera 24(2, B): 70-72. WORHEIDE, G. & RELTNER, J. 1996. ‘Living fossil’ spinctozoan coralline sponge colonies in the shallow water caves of the Osprey Reef (Coral See) and the Astrolabe Reefs (Fiji Islands), Pp. 145-148. In Reitner, J., Neuweiler, F., & Gunkel, F. (eds) Globale und regionale Steuerungsfaktoren biogencr Sedimentation. Göttinger Arbeiten zur Geologie und Palüontologie SB2 (University of Göttingen: Germany). Joachim Reimer (email: jreitne(agwde.de) & Gert Worheide*, Institut und Museum fiir Geologie und Paläontologie, Universität Göttingen, Goldschmidt- Strasse 3, D-37077 Göttingen, Germany: John NA. Hooper *Queensland Museum, PO. Box 3300, South Brisbane, Old, 4101, Australia; | June 1998. POSTPALEOZOIC HISTORY OF THE SILICEOUS SPONGES WITH RIGID SKELETON ANDRZEJ PISERA Pisera, A. 1999 06 30: PostPaleozoic history of the siliceous sponges with rigid skeleton. Memoirs of the Queensland Museum 44: 463-472. Brisbane. ISSN 0079-8835, Most of the Mesozoic groups of siliceous sponges with rigid skeletons (Hexactinosa, Lychniscosa and ‘Lithistida’) have Paleozoic roots, except the Lychniscosa known from the Middle Jurassic to Recent, and the ‘lithistid’ Didymorina known only from the Jurassic. The Triassic record of these groups is poor, and all become common only in the Middle - Late Jurassic, but probably reach maximum diversity and frequency during the Late Cretaceous. The Tertiary record of all these groups is much poorer than for the Mesozoic. Hexactinosa and ‘Lithistida’ are common elements of Recent deeper water faunas, while Lychniscosa, which were very common during the Mesozoic, are very rare and of low diversity in modern seas. Known and newly discovered Tertiary faunas show many affinities with Cretaceous ones indicating lesser susceptibility of these sponges to K/T boundary disturbances, than seen in other organisms. Large faunas of siliceous sponges with rigid skeletons occur in the fossil record in a punctuated manner, and are correlated with high sea level stands. O Porifera, Hexactinosa, Lychniscosa, Lithistida, PostPaleozoic history, K/T boundary. Andrzej Pisera (email: apis@twarda.pan.pl), Instytut Paleobiologii PAN, ul. Twarda 51/55, 00-818 Warszawa, Poland;6 January 1999. Siliceous sponges with rigid skeletons are virtually the only ones which have sufficient fossil records to allow some generalisations to be deduced about their history and patterns of distribution during the Mesozoic. Here belong the Hexactinosa, Lychniscosa and the polyphyletic assemblage we currently know as 'Lithistida'. Even these records, however, are very punctuated. Sponges with skeletons consisting of loose spicules that fell apart after death, such as lyssacinosan hexactinellids and soft demosponges, usually have much poorer records. What is worse, these records are nearly exclusively represented by mixtures of spicules belonging to various taxa (see for example Hinde, 1880; Reif, 1967; Pisera, 1997, and literature therein). This does not mean, of course, that such sponges are not preserved intact at all. There are numerous examples of both demosponges (Propachastrella from the Upper Cretaceous; Schrammen, 1910-1912), or lyssacinosan hexactinellids having loose spicules (Stauractinella from the Upper Jurassic; Mehl, 1992; Pisera,1997) whose preservation as body fossils was facilitated either by very early cementation and/or collagen cement. Another example is a rosselid sponge from the Upper Cretaceous of Denmark (Mehl, 1992). In the Eocene of Catalonia intact demosponges and hexactinellids with loose spiculation (Busquets et al., 1997; Pisera unpublished) are preserved, most probably due to catastrophic burial. All these, however, are relatively rare cases showing only how much information is missing in the fossil record. Some lyssacinosan hexactinellids may have entirely, or partly fused skeletons and such sponges are quite common among fossils, examples can include Triassic Cypellospongia from the USA (Rigby & Gosney, 1983), Hexactinoderma and Silesiaspongia (Fig. 1C) from Poland (Pisera & Bodzioch, 1991), Cretaceous Proeuplectella from France (Moret, 1926) and Regadrella from Germany (Salomon, 1990) and Tertiary (Brimaud & Vachard, 1986b). The fusion in these sponges, however, is of irregular type (at points of contact and without formation of dictyonal strands; Fig.1C), and happens rather late in ontogeny, in opposition to Hexactinosa and Lychniscosa where the fusion is an early ontogenetical phenomenon. I will not refer to these unusual sponges any further. STRATIGRAPHICAL DISTRIBUTION OF HEXACTINOSA, LYCHNISCOSA AND LITHISTIDA HEXACTINOSA. .Hexactinosa are hexasterophoran hexactinellids having skeletons fused in a regular way, i.e., particular hexactines form longitudinal dictyonal strands (Reid, 1963) (Fig. 1B, D) (fusion of hexactines happens by 464 MEMOIRS OF THE QUEENSLAND MUSEUM FIG. 1. Skeletons of hexactinellid sponges. A, Lychniscosan choanosomal skeleton, ?Rhogastomium sp., Late Jurassic, Germany, (scale bar 200um). B, Hexactinosan choanosomal skeleton (note that hexactines are fused together with the help of silica cement), Sphenaulax sp., Late Jurassic, Germany, (scale bar 500um). C, Surface of lyssacinosan choanosomal skeleton (note the irregular fusion of spicules which are mostly diactines), Silesiaspongia rimosa Pisera & Bodzioch, Middle Triassic, Poland, (scale bar Imm). D, Hexactinosan choanosomal skeleton, Dactvlacalyx sp., Recent, (scale bar ] mm). joining two rays into a common silica envelope). Fusion between particular strands may be less regular. For a long time Hexactinosa were regarded as a typical Mesozoic group (see Rigby, 1983); during the last several years, however, numerous reports appeared about the presence of this group in Devonian rocks (Rigby etal., 1981). Much earlier reports of their presence in the Paleozoic (Fraipont, 1911; Mayr, 1930) have been mostly overlooked. The best and the richest fauna of these sponges | know so far is from the Late Devonian (Frasnian) of the Holy Cross Mountains in Poland (Rigby et al., 1981, and in prep.). These sponges are already very close to Jurassic and Cretaceous representatives of Hexactinosa, which suggests that they have an even longer, but unfortunately unknown Paleozoic history. ( There is a chance that some sponges described in the literature from the Ordovician as belonging to sphaerocladine lithistids are in fact Hexactinosan sponges. It is their extremely regular structure, typical of Hexactinosa, but unknown in lithistids, which suggest such a possibility. We know of both, fossil and Recent hexactinosan sponges, that have enlarged sphaerical nodes. This problem needs further study), Smaller faunas of Devonian hexactinosans were reported from Germany, Belgium and Australia. There is a large gap between this late Devonian fauna and the next undoubted hexactinosans that occur only in the Mesozoic. Undoubted Triassic hexactinosan sponges occur in Sichuan, China, Caucasus and in the Alps (Keupp et al., 1989; Wend et al., 1989; Boiko, 1990; Rigby et al., 1998). It is interesting that some of them belong to genera established from the Upper Jurassic. Among them are characteristic genera Casearia and Sphenaulax. Hexactinosa become very diversified and common starting from the Late Jurassic (Schrammen, 1936-37; Trammer, 1982, 1989; Mehl, 1992; Pisera, 1997, and literature therein). POSTPALEOZOIC RIGID-SKELETON SPONGES For example, a Late Jurassic fauna of Hexactinosa from the Swabian Alb is composed of at least of 23 genera and 48 species (Pisera, 1997). Rich hexactinosan faunas are known from the Early Cretaceous (Lagneau-Hérenger, 1962) and Late Cretaceous of Europe (Schrammen, 1910-1912; Moret, 1926). Tertiary hexactinosans are infrequent and reported mostly in recent times (Rigby, 1981; Rigby & Jenkins, 1983; Brimaud & Vachard, 1986b; Busquets et al., 1997). Today hexactinosans form important and diversified elements of deep-water sponge faunas mainly in tropical areas, and include about 40 genera and 135 species (Reid, 1968, and literature therein). LYCHNISCOSA. Lychniscosan sponges also display development of dictyonal strands and fusion of spicules, but they have lychnisc (octahedral) nodes (Fig. 1A), in contrast to the solid nodes of hexactines in Hexactinosa. Their history is quite different. There is no trace of this group before the Middle Jurassic. Earlier reports of Triassic lychniscosan sponges (Vinassa de Regny, 1911; Keupp et al., 1989; Wendt et al., 1989; Wu Xi Chun, 1990) were proven to be erroneous (Mostler, 1990; Pisera & Bodzioch, 1991; Mehl, 1992). Also the report of Early Jurassic lychniscosans (Broglio Loriga et al., 1991) seems very doubtful. The oldest known bodily preserved lychnis- cosan sponge (Pisera, 1993, and in prep.) belongs most probably to the Late Jurassic genus Pachyteichisma and occurs in the uppermost Bajocian of the Meések Mountains in southern Hungary. This genus is common in the Callovian of Kutch, India (Mehl & Fiirsich, 1997, referred to as Sporadopyle) and it occurred also during the Late Jurassic (Pisera, 1997). Lychniscosan sponges are an important part of the large Late Jurassic fauna from the Swabian Alb (34 species and 15 genera; Pisera, 1997) and Upper Cretaceous faunas of Northern Germany (81 species and 34 genera; Schrammen, 1912). Even if there is an ‘oversplitting’ of taxa this diversity is impressive, especially when compared with Recent diversity of this group. For a long time very little was known about Tertiary lychniscosan sponges, and only one genus, Manzonia from the Miocene of Italy and Spain, was known. To this Pomel’s genus Pachychlaenium (=Tremabolites) was added by Mehl (1992). More recent discoveries in the Eocene of Catalonia (Pisera in prep.) show that their diversity was lower than during the Late 465 Cretaceous, but was still higher than today, for they are represented there by at least 5-6 species and 5 genera, which are quite prevalent. Today, lychniscosan sponges are a relict group represented only by 3 species and 2 genera (Mehl, 1992) and are rare. The reason why Lychniscosa developed during the Cenozoic differently than the Hexactinosa remains unknown. *LITHISTIDA'. Lithistids are demosponges characterised in having choanosomal skeletons composed of desmas joined by articulation without cementation by silica (Fig. 2A). There is no doubt that they are a polyphyletic group, and when we speak about ‘lithistids’ we should split them into smaller units that have the same geometry of desmas, with a greater probability of being monophyletic (i.e., Tetracladina Zittel, 1878, Rhizomorina Zittel, 1878, Dicranocladina Schrammen, 1924 (=Corallistidae Sollas, 1888), Sphaerocladina Schrammen, 1910, Megamorina Zittel, 1878 and Didymorina Zittel, 1878). Lithistids as a group are known from the Lower Palaeozoic (Rigby, 1983), and some Mesozoic groups, 1.e., Rhizomorina (if the Paleozoic rhizomorines are the same lineage as Mesozoic one), Sphaerocladina and Megamorina have their Palaeozoic representatives. The Dicranocladina are most probably closely related to Paleozoic hindiids (Finks, 1971). Diversity of fossil and Recent lithistids, as a whole group, is probably comparable because in the Upper Jurassic of the Swabian Alb about 42 species have been found by me, whereas a Recent fauna of lithistids from the New Caledonia region is composed of 23 species according to Lévi & Lévi (1983). Taking into account that the Recent fauna represents only one time plane, the diversity may be regarded as similar. Diversity of particular groups, however, differs considerably. Tertiary lithistids, as a whole, are rather poorly known (Moret, 1924; Brimaud & Vachard, 1986a). For excellent and more detailed review of lithsitids distribution than presented below see Wiedenmayer (1994). The oldest bodily preserved Rhizomorina Zittel, 1878, characterised by irregular, usually thorny desmas called rhizoclones (Fig. 2B-E), based on monaxons, are known from the Early Jurassic from Georgia (Nutsubidze, 1965), although rhizoclone spicules are known from the Triassic (Wiedenmayer, 1994). Rhizomorine sponges are common only from the Late Jurassic on (Schrammen, 1910-1912, 1937; Moret, 1924; Brimaud & Vachard, 1986a; Pisera, 1997). They are probably the most common group of 466 MEMOIRS OF THE QUEENSLAND MUSEUM FIG. 2. Lithistid sponge spiculation; A, Articulation of desmas (note that desmas are joined only by articulation without cementation), Recent tetractinellid lithistid, Caribbean, (scale bar 100jm). B, Choanosomal skeleton of Recent rhizomorine sponge Setidium sp., (scale bar 200m). C-E, Desmas (rhizoclones) of the Late Jurassic rhizomorine sponges, Germany, (scale bars 100um). F-G, Desmas of the fossil sphaerocladine sponges, Late Jurassic, Germany, (scale bars 100m). H, Choanosomal skeleton of the Recent sphaerocladine lithistid Vetulina sp., Caribbean, (scale bar 100um). L, Desma (megaclone) of fossil megamorine sponge, Late Jurassic, Germany, (scale bar 100um). J-K, Desmas (didymoclones) of fossil didymorine lithistid Cylindrophyma sp., Late Jurassic, Germany, (scale bar 100um). POSTPALEOZOIC RIGID-SKELETON SPONGES Mesozoic lithistids known in Europe. Rhizomorines are quite common in the Tertiary (Moret, 1924), and today they are represented by 8-10 genera. Bodily preserved Tetracladina Zittel, 1878, characterised by usually regular desmas based on a tetraxon called tetraclone (Fig. 3F-G), already existin the Triassic (known from the thin sections only; Keupp et al., 1989), but they are rare. They become common only in the Late Jurassic (5 genera and 6 species), with maximum fossil diversity in the Late Cretaceous where Moret (1926) cited 26 genera and 62 species from France alone. In Recent seas they are also common and represented by about 130 described species worldwide (Wiedenmayer, 1994 and the literature therein). In more resticted areas like New Caledonia 23 species and 16 genera were reported (Lévi, 1991). The earliest Dicranocladina Schrammen, 1924, characterised by regular strongly tuberculated, usually tripodal or tetrapodial, desmas (Fig. 3A-E), based on monaxons, are known from 5 species and 3 genera of the Late Jurassic. They are distinguished with difficulty from Recent ones. Dicranocladina diversity is high in the Late Cretaceous (Moret, 1926, lists 8 genera and 15 species). Tertiary Dicranocladina are rarely reported (Rigby, 1981; Brimaud & Vachard, 19862), but they are quite common in Recent seas (Lendenfeld, 1903; Levi & Levi, 1983; Levi, 1991). The next group of lithistids, i.e., Megamorina Zittel, 1878 (which corresponds to Recent Pleromatidae Sollas, 1888), are characterised by desmas called megaclones (Fig. 21). They appear as early as the Ordovician but their Mesozoic record starts in the Middle Triassic (Wiedenmayer, 1994, and literature therein). They become common and diversified in the Jurassic and Cretaceous (Schrammen, 1910-1912, 1937; | Moret, 1926; Lagneau-Hérenger, 1962; Pisera, 1997). Their Tertiary record is poor and only one bodily preserved, recently discovered, but still undescribed species, has been found in the Eocene. Today Megamorina is also a small group with only 2 genera (Wiedenmayer, 1994). Similar is the history of the Sphaerocladina Schrammen, 1910, which are characterised by desmas called sphaeroclones or astroclones (Fig. 2F-H). They first appear in the Paleozoic (Wiedenmayer, 1994, and references therein) (if astylospongids are included here, and that may be 467 questioned), but their Mesozoic record starts in the Late Jurassic (3 genera and 5 species). Sphaerocladina have the same diversity in the Cretaceous (Moret, 1926, lists 3 genera, 5 species). They have a poor Tertiary record, and in Recent seas are represented most probably only by Vetulina stalactites Schmidt (Fig. 2H). This species has been suggested to be the rhizomorine (Gruber, 1993), but more recent studies of the holotype suggests that it is not. The last group to be considered is Didymorina Zittel, 1878, with desmas called didymoclones (Fig. 2J-K). They are extinct and undoubted representatives occur only in the Middle and Late Jurassic - 2 genera and 3 species. Similar loose spicules of uncertain affinity were reported by Mostler (1976) from the Triassic. PATTERNS IN THE HISTORY OF SILICEOUS SPONGES WITH RIGID SKELETON. Longevity of sponge genera. It has been known long that numerous Recent hexactinellid genera are long ranging and occur even in Upper Cretaceous rocks. For example Mehl (1992) found that of 31 genera from the Late Cretaceous Hexactinosa, 11 survive in Recent seas. The hexactinosan genus Laocoetis (=Craticularia) ranges most probably from the Early Jurassic (Nutsubidze, 1965) until today (Lévi, 1986) - a duration of nearly 200 million years. The Recent genus Dactylocalyx has been reported by Trammer (1989) from the Late Jurassic. Recently discovered sponge faunas strengthen this pattern by showing the presence of the genera known from the Upper Jurassic also in the Triassic (e.g., the hexactinosan Casearia and Sphenaulax; Rigby et al., 1998), and other Cretaceous genera in the Tertiary (e.g., the lychniscosan Becksia, Sporadoscinia, Brachiolites in newly discovered Eocene faunas from Spain; Pisera, unpublished), thus pointing to the extremely conservative nature of these sponges at the generic level. Lithistids also include several long ranging genera. Cnemidiastrum, for example, is typical of the Upper Jurassic but has been recently recorded in the Miocene, while the Cretaceous genus Aulaxinia has been discovered by Lévi & Lévi (1988) in Recent waters around New Caledonia. The Recent rhizomorine genus Amphibleptula from the Atlantic has been recognised in the Late Jurassic (Pisera, 1997). Numerous lithistids of the Late Cretaceous (especially rhizomorine sponges) show no recognisable differences when compared with Late Jurassic forms. It seems that different names given to these different faunas 468 MEMOIRS OF THE QUEENSLAND MUSEUM E A d FIG. 3. Lithistid sponge spiculation; A, Recent Corallistes sp., choanosomal skeleton composed of strongly tuberculated dicranoclones, and dermal dichotriaenes, Gulf of Mexico, (scale bar 100um); B-E, Fossil Dicranoclonella sp., Late Jurassic, Germany. B, Upper surface showing dermal dichotriaenes and rhizoclone-like modified dicranoclones between them, (scale bar 2004 m); C, Choanosomal skeleton composed of strongly tuberculated dicranoclones, (scale bar 200um); D, Isolated typical dicranoclone, (scale bar 2001m); E, Fragment of choanosomal skeleton, (scale bar 500um). F, Choanosomal skeleton of Recent tetracladine sponge, Carribean, (scale bar 500um); G, Choanosomal skeleton of fossil tetracladine sponge, Oligocene, the Ukraine, (scale bar 500m). stem from the philosophy that large age differences are enough to justify establishing a new genus. This approach is questionable. It appears that lithistids are also rather conservative and slowly evolving. Sponges and K/T boundary. The Cretaceous- Tertiary (K/T) boundary was a time when most fossil groups of marine organisms were severely decimated. The pattern of distribution of siliceous sponges with rigid skeleton across this boundary is interesting. There are no sponge POSTPALEOZOIC RIGID-SKELETON SPONGES Meters above present sea level (in hundreds) 65 43 2 1 0 Large sponge faunas 469 meandrispongids (Brachiolites, Plocoscyphia), and Sporadoscinia, which are characteristic of the Late PLIO-PLEISTOCENE T I-ULU 4 n 1 Cretaceous, dominate the MIOCENE Eocene faunas (Pisera, OLIGOCENE PALEOC. & EOCENE unpublished data). A similar fauna has been reported also CRETACEOUS from the Eocene of the USA (Rigby, 1981; Finks, 1983, 1986). Miocene faunas from Algeria, described by Moret JURASSIC (1924), and from Spain described by Brimaud & Vachard (1986a, b) have many Mesozoic elements. Recent TRIASSIC faunas of sponges with rigid skeleton are also considered by PERMIAN CARBONIFEROUS DEVONIAN Reid (1967) as of the Tethyan (Mesozoic) origin. All this indicates that sponges were less strongly affected by K/T boundary disturbances than other organisms. How to explain such behaviour of the siliceous SILURIAN ORDOVICIAN <= Present sea level sponges discussed here? It follows because they are (and were) rather deep-water creatures. They were at least probably protected by a water column from disturbances CAMBRIAN | KEST - H7] 6.5 4372 111 FIG. 4. Sea level curve for the Phanerozoic and the distribution of large faunas of siliceous sponges (sea level curve from Hallam, 1992). faunas known directly above the boundary, but when one considers the newly discovered Eocene faunas from the Pyrenees (Busquets et al., 1997) composed of both lithistids (mostly tetractinellids, one megamorine and some rhizomorine sponges have been also found; Pisera, unpublished data) and hexactinellids with rigid skeletons, including both hexactinosans and lychniscosans. Such characteristic forms as Guettardiscyphia, a typical Cretaceous hexactinosan, are important elements (rock forming in places) of these Eocene faunas. Among lychniscosan sponges the so called occurring at the surface. The rather simple character of sponges, which fed on colloidal matter and bacteria, may have also played a role, forthey were less influenced by supposed disturbances of the food chain during the K/T event (whatever its cause). On the other hand, it is difficult even to speculate at the moment, upon what differences, other than chance, caused different behaviour of particular groups of sponges (like Hexactinosa and Lychniscosa) in relation to K/T boundary event. Large sponge faunas. So far I have been concentrated on the stratigraphic ranges of particular groups or lineages of sponges, but there are some interesting patterns in distribution of large sponge faunas as such. In this context large sponge faunas refer to faunas that are diverse, have wide geographical distribution, and in which sponges occur in profusion, where 470 sponges are usually a rock forming element. During the Late Jurassic, for example, they formed biostromal or reefal structures, and the sponge facies extends across the whole Europe from Portugal to Romania (Trammer, 1982, 1991; Leinfelder et al., 1994; Krautter, 1997; Pisera, 1997, and literature therein). Distribution of large faunas of bodily preserved (intact) fossil sponges during the Meso-Cenozoic is rather punctuated and limited to certain periods of time. The largest such faunas known for a long time are associated with the Upper Jurassic, Upper Cretaceous, and the Miocene rocks (Fig. 4). Because of large gaps separating these faunas, they appear at the genus level to be composed of very different taxa, which makes interpretation of evolution of these sponges difficult. Such a pattern of distribution has been interpreted, at least for lithistids, by Rigby (1983) as the result of “selective preservation and discovery, not one of original limited diversity and density”. This interpretation has partly found support in more recent important discoveries from the Late Triassic (China - Wendt et al., 1989; Wu Xi Chun, 1990), Middle Jurassic (Spain - Scheer, 1988; Hungary - Pisera, 1993; India - Mehl & Fiirsich, 1997), Eocene (Spain - Busquets et al., 1997) and Miocene (Spain - Brimaud & Vachard, 1986a, b) and a smaller one from the Oligocene (Antigua - Wiedenmayer, 1994; Ukraine - Pisera, unpublished). Generally, however, if we look at the distribution of large sponge faunas, the pattern of punctuated record is preserved. The largest are of Late Jurassic and Late Cretaceous age (across the whole of Europe); smaller ones occur in the Upper Triassic (The Alps, Sichuan), Middle Jurassic (Spain, France, Hungary, Kutch in India), Eocene (Spain, Italy, Turkey) and the Miocene (Algeria, Spain, Italy), When one compares all these occurrences with the sea level history (Fig. 4), a clear correlation appears: the occurrence of the large siliceous sponge faunas is correlated with the high sea level times during the Mesozoic and Tertiary (and it seems that this pattern is valid also for the Paleozoic in the case of Hexactinosa in the Frasnian/Famenian). It points to the importance of sea level in controlling distribution of large sponge faunas, This relationship, however, is mostly of environmental character, although some evolution, took place especially at the species level. The sponges considered here are deep-water creatures and their widespread development may be interpreted as a possibility to colonise new, vast, relatively deep-water areas. MEMOIRS OF THE QUEENSLAND MUSEUM These areas were not available during lower sea level periods, and when sponges of these groups existed only in relatively narrow refugia along continental and island slopes, as it is in many cases today. Such new areas of relatively deep-water were at a distance from shore, had low sedimentation rates and low hydrodynamic energy, and thus were suitable for sponge colonisation. LITERATURE CITED BOIKO, E.V. 1990, O mnogoobrazii sklietenykh struktur u kamiernykh gubok. Pp. 119-129. In Sokolov, B.S. & Zhuravleva, I.T. (eds) “Iskopajemyje probliematiki SSSR’. Trudy Instituta Gieologii i Gieofiziki AN SSSR, Sibirskoje Otdielenie (783). BRIMAUD, C. & VACHARD, D.1986a. Les Spongiaires siliceux du Tortonien des Bétiques (Miocéne de l'Espagne du Sud): espéces nouvelles ou peu connues I. Choristides et Lithistides. Bulletin Muséum National d'Histoire Naturelle, Paris (8,C) 3: 293-341. 1986b. Les Spongiaires siliceux du Tortonien des Bétiques (Miocene de l'Espagne du Sud): espèces nouvelles ou peu connues II. Hexactinellides. Bulletin Muséum National d'Histoire Naturelle, Paris (8,C) 4: 415-445. BROGLIO LORIGA, C., MASETI, D., FORASTIERI, S. & TREVISANI, E. 1991. Comunita a Poriferi nei Calcari Grigi delle Vette Feltrine. Annali dell’ Universita di Ferrara (Nuova Serie), Scienze della Terra 3: 51-81. BUSQUETS, P., PISERA, A., REGUANT, S. & SERR-KIEL, J. 1997. Biofacies of the outer continental shelf in the Bartonian part of the Ebro Basin (NE Spain). Boletin de la Real Sociedad Espafíola de Historia Natural. Seccion Geologica 92: 249-256. FINKS, R.M. 1971. A new Permian eutaxicladine demosponge, mosaic evolution, and the origin of the Dicranocladina. Journal of Paleontology 45: 977-997, 1983. Fossil Hexactinellida. Pp. 101-115. In Rigby, J.K. & Stearns, C.W. (eds) ‘Sponges and Spongiomorphs. Notes for a short course’. University of Tennessee Department of Geological Sciences Studies in Geology (7). 1986. The Castle Hayne sponge fauna (Eocene) and the history of Tertiary sponges. A15. (Fourth North American Paleontological Convention: Boulder, Colorado). FRAIPONT, C. 1911. Une Hexactinellide nouvelle du Dévonien belge (Calcaire Frasnien), Pseudopemmatites formarieri, g. et sp.n. Annales de la Société Géologique de Belgique 38:197-206. GRUBER, G. 1993, Mesozoische und rezente desmentragende Demospongiae (Porifera, “Lithistidae”) (Paláobiologie, Phylogenie und POSTPALEOZOIC RIGID-SKELETON SPONGES Taxonomie). Berliner Geowissenschaftliche Abhandlungen (E) 10: 1-73. HALLAM, A. 1992. Phanerozoic sea-level changes. (Columbia Univeristy Press: New York). HINDE, G.J. 1880. Fossil sponge spicules from the Upper Chalk found in the interior of a single flint-stone from Horstead in Norfolk. (Inaugural- Dissertation Erlangung der akademischen Doctorwurde bei der philosophischen Facultát der Ludwig-Maximilians-Universitat: Miinchen). KEUPP, H., REITNER, J. & SALOMON, D. 1989. Kieselschwámme (Hexactinellida und *Lithistida") aus den Cipit-Kalken der Cassianer Schichten (Karn, Südtirol). Berliner Geowissenschaftlische Abhandlungen (A) 106: 221-241. KRAUTTER, M. 1997. Aspekte zur Paláókologie postpaláozoischer Kieselschwámme. Profil 11: 199-324. LAGNEAU-HERENGER, L. 1962. Contribution a l'étude des spongiaires siliceux du Cretacé inférieur. Mémoires de la Société Géologique de France 95; 1-252. LEINFELDER, R.R., KRAUTTER, M., LATERNSER, R., NOSE, M., SCHMID, D.U., SCHWEIGERT, G., WERNER, W., KEUPP, H., BRUGGER, H., HERMANN, R, REHFELD-KIEFER, U., SCHROEDER, J.H., REINHOLD, C., KOCH, R., ZEISS, A., SCHWEIZER, V., CHRISTMANN, H., MENGES, G, & LUTERBACHER, H. 1994. The origin of Jurassic reefs: current research _ developments and results. Facies 31: 1-56. LEVI, C. 1986. Laocoetis perion nov. sp., Spongiaire Hexactinellide Craticulariidae de l'océan Indien. Bulletin Muséum National d'Histoire Naturelle Paris (4, 8, A) 3: 437-442. 1991. Lithistid sponges from the Norfolk Rise. Recent and Mesozoic Genera. Pp. 72-82. In Reitner, J. & Keupp, H. (eds) ‘Fossil and Recent _ Sponges’ (Springer Verlag: Berlin). LEVI, C. & LEVI, P. 1983. Eponges tetractinellides et Lithistides bathyaux à affinités crétacées de la Nouvelle Calédonie. Bulletin Muséum National d'Histoire Naturelle Paris (4, 5, A) 1: 101-168. 1988. Nouveaux Spongiaires Lithistides bathyaux à affinités crétacées de la Nouvelle Calédonie. Bulletin Muséum National d'Histoire Naturelle Paris (4, 10, A) 2: 241-263. LENDENFELD, R. VON, 1903. Tetraxonia. Das Tierreich 19: 1-168. MAYR, J. 1930. Gabki dewonskie Gór Swietokrzyskich. Sprawozdania Towarzystwa Naukowego Lwów 9 (1929): 246. MEHL, D. 1992, Die Entwicklung der Hexactinellida seit dem Mesozoikum. Paláobiologie, Phylogenie und evolutionsókologie. Berliner Geowissen- schaftlische Abhandlungen (E) 2: 1-164. MEHL, D. & FURSICH, F.T. 1997. Middle Jurassic Poriferas from Kutch, western India. Paláontologische Zeitschrift 71(1/2): 19-33. 471 MORET, L. 1924. Contribution à l'étude des spongiaires siliceux du Miocéne de l'Algérie. Mémoires de la Société Géologique de France (Nouvelle Série) 1: 1-27. 1926. Contribution à l'étude des spongiaires siliceux du Cretacé supérieur français. Mémoires de la Société Géologique de France (4) 3: 121-334. MOSTLER, H. 1976. Poriferenspiculae der Alpinen Trias. Geologisch-Paláontologische Mitteilungen Innsbruck 6: 1-42. 1986. Beitrag zur stratigraphischen verbreitung und phylogenetischen Stellung der Amphidiscophora und Hexasterophora (Hexactinellida, Porifera). Mitteilungen Osterreichischen Geologischen Gesselschaft 78: 319-359. 1990. Hexactinellide Poriferen aus pelagischen Kieselkalken (Unterlias, Nórdlische Kalkalpen). Geologisch-Paláontologische Mitteilungen Innsbruck 17:143-178. NUTSUBIDZE, K.SH. 1965. Liasovyje gubki Dzirulskovo masiva. Trudy Geologitsheskovo Instituta, serija gelogitscheskaja 14(19, 1964): 5-35. PISERA, A. 1993. Siliceous sponges from the Middle Jurassic of the Meések Mountains (southern Hungary). In *4th International Porifera Congress. Sponges in Time and Space. Book of Abstracts’ (University of Amsterdam: Amsterdam). 1997, Upper Jurassic siliceous sponges from the Swabian Alb- their taxonomy, environmental setting, and history. Palaeontologia Polonica 57: 1-216. PISERA, A. & BODZIOCH, A. 1991. Middle Triassic lyssacinosan sponges from Silesia (southern Poland), and history of hexactinosan and lychniscosan sponges. Acta Geologica Polonica 41; 193-207. REID, R.E.H. 1963. Dictyonal structure in Hexactinosa and Lychniscosa. Journal of Paleontology 37: 212-217. 1967. Tethys and the zoogeography of some modern and Mesozoic Porifera. Pp. 171-181. In Adams, C.G. & Ager, D.V. (eds) *Aspects of Tethyan biogeography’. (Systematic Association: London). 1968. Bathymetric distribution of Calcarea and Hexactinellida in the present and in the past. Geological Magazine 105: 546-559, REIF, W.-E. 1967. Schwammspicula aus dem Weisen Jura Zeta von Nattheim (Schwäbische Alb). Palaeontographica (A ) 127: 85-102 RIGBY, J.K. 1981. The sponge fauna of the Eocene Castle Hayne Limestone from East-Central North Carolina. Tulane Studies in Geology 16(4): 123-144. 1983. Fossil Demospongia. Pp, 12-39. In Rigby, JK. & Stearns, C.W. (eds) ‘Sponges and Spongiomorphs. Notes for a short course’, 472 University of Tennessee Department of Geological Sciences Studies in Geology (7). RIGBY, J.K. & GOSNEY, T.C. 1983, First reported Triassic lyssakid sponges from North America. Journal of Paleontology 57: 787-794. RIGBY, J.K., RACKI, G. & WRZOLEK, T. 1981. Occurrence of dictyid hexactinellid sponges in the Upper Devonian of the Holy Cross Mts. Acta Geologica Polonica 31; 163-168. RIGBY, J.K. & JENKINS, D.F. 1983. The Teriary sponges Aphrocallistes and Enrete from western Washington and Oregon. Contribution in Science Natural History Museum of Los Angeles County 344; 1-13. RIGBY, S.K., WU XICHUN & FAN JLASONG. (1998) Triassic hexactinellid sponges from patch reefs in North Central Sichuan, People's Republic of China. Brigham Young University Geology Studies 43: 119-165. SALOMON, D. 1990. Ein neuer lyssakiner Kieselschwamm, Regadrella leptotoichiva (Hexasterophora, Hexactinellida) aus dem Untercenoman von Baddeckensted (Nordwestdeutschland). Neues Jahrbuch fiir Geologie und Paläontologie, Monatshefte 6: 342-352, SCHEER, U, 1988. Influence of the paleogeographic position and sea-level changes on spongiolitic limestones in the Lower Bajocian (discites-to humphresianum-zone) Middle Jurassic in Northern Spain. Berliner Geowissenschafiliche Abhandlungen (A) 100: 36. SCHRAMMEN, A. 1910. Die Kieselspongien der oberen Kreide von Nordwest-Deutschland, [. Teil, Tetraxonia, Monaxonia und Silicea incertae sedis. Palaeontographica Supplement 5: 1-175. MEMOIRS OF THE QUEENSLAND MUSEUM 1912. Die Kieselspongien der oberen Kreide von Nordwestdeutschland. Palaeontographica Supplement 5: 176-385. 1924. Zur Revision der Jura-Spongien von Süddeutschlands. Jahresberichte und Mitteilungen Oberrheinischen Geologischen Vereins (1924): 125-154. 1936. Die Kieselspongien des Oberen Jura von Süddeutschland. Palaeontographica 84: 149-194, 1937. Die Kieselspongien des Oberen Jura von Süddeutschland. Besonderer — Teil. Palaeontographica 85: 1-114, TRAMMER, J. 1982. Lower to Middle Oxfordian sponges of the Polish Jura. Acta Geologica Polonica 32; 1-39. 1989, Middle to Upper Oxfordian sponges of the Polish Jura. Acta Geologica Polonica 39: 49-91, 1991. Ecologic history of the Oxfordian sponge assemblage in the Polish Jura Chain. Pp. 506-515. In Reitner, J. & Keupp, H. (eds) ‘Fossil and Recent sponges’. (Springer Verlag: Berlin) VINASSA DE REGNY, P. 1901. Neue Schwámme, Tabulaten und Hydrozoen aus dem Bakony. Resultate Wissenschaftlischen Erforschung Balatonsee. Palaeontologischer Anhang 1: 1-17, WENDT, J., WU, XI-CHUN & REINHARDT, J.W. 1989. Deep water hexactinellid sponge mounds from the Upper Triassic of northern Sichuan (China). Palaeogeography, Palaeoclimatology, Palaeoecology 76: 17-29. WIEDENMAYER, F. 1994. Contribution to the knowledge of post-Paleozoic neritic and archibenthal sponges (Porifera). Schweizerische Paliiontologische Abhandlungen 116: 1-147, WU, XI CHUN 1990, Late Triassic lychniscose fauna in northwestern Sichuan, Acta Palaeontologica Sinica 29: 357-363 Note added in proof. Eocene sponge faunas should be supplemented with a lithistid fauna from the southern Western Australia (Pickett, 1983, and references therein), of which | was earlier unaware. After preliminary examination of a new collection of sponges from this region (thanks to Dr, P. Gammon, Canada) it seems to be the largest, the most diversified, and the best preserved lithistid fauna of the Tertiary. Literature cited, PICKETT, J.W. 1983, An annotated bibliography and review of Australian fossil sponges, Memoir of the Association of Australasian Palaeontologists 1: 93-120 LITHISTID SPONGE SETIDIUM OBTECTUM SCHMIDT, 1879, REDISCOVERED ANDRZEJ PISERA Pisera, A. 1999 06 30: Lithistid sponge Setidium obtectum Schmidt, 1879, rediscovered. Memoirs of the Queensland Museum 44: 473-477. Brisbane. ISSN 0079-8835. The poorly known genus and species Setidium obtectum Schmidt, 1879 was revised based on the holotype and newly collected material from the Caribbean, and referred to the family Scleritodermidae Sollas. O Porifera, Rhizomorine lithistids, Setidium, systematic position. Andrzej Pisera (email: apis@twarda.pan.pl), Instytut Paleobiologii PAN, ul. Twarda 51/55, 00-818 Warszawa, Poland; 6 January 1999. The rhizomorine lithistid genus and species Setidium obtectum was established by Schmidt (1879) based on a unique specimen dredged off Havana from 234-439m depth. The occurrence of remarkable, numerous bundles of very long oxeas protruding from the inner (upper) surface ofthe sponge confers upon it a very characteristic form. As a result of its incomplete description and single previous record its systematic position remained obscure. Recently, new rich material of this sponge was discovered in the collections of the Marine Invertebrates Museum, University of Miami, Florida (MIM-RSMAS), dredged from several localities in the Caribbean (Fig. 1). Investigation of this new material and re-examination of the holotype in Schmidt's collection at the Museum of Comparative Zoology, Harvard University (MCZ), permitted determination of its characteristics and establishment of its taxonomic position. SYSTEMATICS SYSTEMATIC POSITION. Lendenfeld (1903) regarded Setidium a synonym of Leiodermatium Schmidt, and placed it within Leiodermatidae Lendenfeld, which encompasses rhizomorine lithistids without microscleres. Van Soest & Stentoft (1988) placed Setidium obtectum among rhizomorine lithistids of the family Siphonidiidae Sollas, which is characterised by an ectosomal skeleton composed of desmas without zygosis, and absence of microscleres as well. The present investigation found that choanosomal desmas of Setidium obtectum are typical thorny rhizoclones, and ectosomal spicules are mostly amphioxeas that form a dense tangential layer, showing all transitional forms to rhizoclones. Numerous thorny sigmaspire microscleres were found both in the holotype and the new material. As a result, Setidium obtectum Schmidt, 1879, should be placed in Scleritodermidae Sollas, 1888. The presence of sigmaspires, rhizoclone type, as well as the ectosomal skeleton of amphioxeas, make this genus very close to the genus Scleritoderma Schmidt, 1879. Family Scleritodermidae Sollas, 1888 Setidium Schmidt, 1879 DIAGNOSIS. Rhizomorine sponges bearing on the surface (mostly the upper one) bundles of long oxeas protruding from the choanosome; ectosomal spicules are amphioxeas that show a complete transition to rhizoclone desmas. Sigmaspire microscleres concentrated around oscules. Choanosomal skeleton very dense, confused and composed of strongly “thorned” rhizoclones. REMARKS. This is so far a monotypic genus. It is close to Scleritoderma, but differs in having smooth ectosomal spicules that are amphioxeas and their derivatives while Scleritoderma has acanthose microstrongyles (Van Soest & Stentoft, 1988). Setidium obtectum Schmidt, 1879 (Figs 2-4) Setidium obtectum Schmidt, 1879: 30-31, Pl. 1, fig. 9, Pl. 2, fig. 14; Lendenfeld, 1903: 145-148; Van Soest & Stentoft, 1988: 74, MATERIAL. HOLOTYPE MCZ6462: off Havana, 234-441m depth, collected Blake Expedition. MIM-RSMAS G688: off Miami, 26°53’N, 78°16’W, 492m depth (two specimens); G13 12: off Miami, 26°38’N, 79°02’ W, 516m depth (one deciduous specimen); P1141: off Great Inagua, 20°51’N, 73°16’W, 429m depth (4 specimens). 474 MEMOIRS OF THE QUEENSLAND MUSEUM | (ND GRAND BAHAMA Ss A bo THERA | Tues IS. ZS 2 Ea gn que. yes ISLAND J a N o C GREAT INAGUA FIG. 1. Distribution of the known specimens of Setidium obtectum Schmidt. H=holotype, NI, N2=new material. FIG. 2. Morphology of Setidium obtectum Schmidt (scale bars 1cm). A-B, Holotype, lateral and upper side views, MCZ 6462, off Havana. C-D, Large deciduous specimen, lateral and upper side views, G1312, off Miami. E-F, Living specimen with ectosomal spicules, lateral and upper side views, P1141/2, off Great Inagua. DESCRIPTION. Shape and structure of the 1997) about 5cm wide, 3cm high, with a wall skeleton. The holotype is irregular vase-shaped about lcm thick and rounded margin. The base (or turbinate - see Boury-Esnault & Rützler, showsa very short peduncle. The new specimens SETIDIUM OBTECTUM REDISCOVERED 475 FIG. 3. Setidium obtectum Schmidt. A, Upper surface, bundles of oxeas protruding from the choanosome, ?oscula and ectosomal spiculation, P1141/1, off Great Inagua (scale bar 2mm). B, Details of upper surface showing ?oscules and tangentially arranged ectosomal spicules, P1141/1, off Great Inagua (scale bar 500um). C, Upper surface of the choanosomal skeleton (HNO; preparation) showing ?oscula and bundles of oxeas (broken) protruding from choanosomal skeleton, P1141/1, off Great Inagua (scale bar 1mm). D, Rhizoclones on the upper surface of choanosomal skeleton (HNO; preparation), note partly incorporated young desma, P1141/1, off Great Inagua (scale bar 200um). E, Lower (outer) surface showing ostia protected by a tepee-like organised oxeas, P1141/1, off Great Inagua (scale bar Imm). F, Close-up of ?osculum with numerous sigmaspires, P1141/1, off Great Inagua (scale bar 20m). G, Upper surface showing ?osculum and ectosomal spicules, holotype, MCZ. 6462, off Havana (scale bar 2004m). H, Close-up of osculum with sigmaspires, the holotype, MCZ 6462, off Havana (scale bar 20um). 476 | | MEMOIRS OF THE QUEENSLAND MUSEUM FIG. 4. Setidium obtectum Schmidt, P1141/1, off Great Inagua. A-C, Sigmaspires (scale bar 511m). D, Young desma (scale bar 100um). E-J, ectosomal spicules showing transition to desma (scale bars 501m). range in shape from deep vase or turbinate (especially when young), to shallow widely open vase-shaped with very thick walls, about 1-1.5cm thick, and with a short massive peduncle. The margin of the vase is more-or-less square (rounded in the holotype) and its surface is nearly smooth or bearing low, wide elevations from which protrude bundles of long oxeas (now all broken). Both surfaces are rugose as a result of numerous irregularly to slightly radially distributed conules, 2-5mm high, from which protrude bundles of long oxeas, up to 20 spicules each, that are at least 20mm long (now all broken). Because the conules are slightly oblique to the surface, especially in old individuals, they may form kind of short, radially aligned ridges. Except for the tips of conules both surfaces are covered with small openings - presumably ostia and oscules. These are less densely distributed on the upper surface and are located in shallow depressions, sometimes surrounded by a slightly rised rim of ectosomal spicules. On the lower surface ostia are very densely distributed, practically touching each other, and placed on tops of small elevations; these elevations are not visible on the surface of choanosomal skeleton and thus they are probably formed only by ectosomal spiculation. Ostia are surrounded and covered by a tepee-like structure formed by delicate oxeas up to 0.5mm long (length of visible part only) protruding from the surface. These openings on the upper surface measure 160-220um (when ectosomal spicules are present) and corresponding canal openings in the choanosomal skeleton are 230-300um. Those on SETIDIUM OBTECTUM REDISCOVERED the lower surface are slightly larger and measure 200-260um (when ectosomal spiculation preserved) and corresponding canal openings in the choanosomal skeleton are 230-280um. Ectosomal spicules are tangentially arranged amphioxeas. No other structures are present on the surface of the choanosomal skeleton beyond these canal openings that lead to ostia and oscula. On the surface of the margin slightly sinuous and partly open canals, 0.20-0.25mm wide, run close to the surface of the skeleton from the lower to the upper surface of the sponge. Similar canals pierce the wall in cross section. Close to the outer (lower) surface of the choanosomal skeleton there are numerous large connected lacunes 0.4-0.5mm diameter. In cross section are also visible bundles of oxeas that protrude from the conules on the surface and enter deeply into the choanosomal skeleton. Colour. White or yellow when dry or in alcohol, white when devoid of ectosomal spicules. Desmas. These are strongly branched and thorny thizoclones measuring 200-400um long; rhizoclones around canals are only tuberculated on the canal side. Other megascleres. Large oxeas protrude from conules on both surfaces (but are more developed on the upper one) and deep in the choanosome. They are invariably broken so total length is unknown but they are at least 1cm long and 15- 23m thick. Ectosomal spicules are tangentially arranged, smooth, regular amphioxeas that are 120-210um long and 10-14um thick. They show all transitions to desmas. Derivatives with shapes modified toward rhizoclone desmas may be 477 270um long and 40m thick or more. Ectosomal spicules may be concentrically arranged around oscula. Small oxeas protect ostia on the lower surface forming tepee-like structures around the ostia. These spicules are usually broken but they appear to be 300-500um long and 3.5-5um diameter. Microscleres. Thorny sigmaspires measure 8- 13um long and up to 1.7m thick. ACKNOWLEDGEMENTS The material was kindly loaned for study by Prof. Nancy A. Voss (Marine Invertebrates Museum, RSMAS, University of Miami, Miami, Florida). All SEM photos were made using a Philips XL-20 scanning microscope at the Institute of Paleobiology, Polish Academy of Sciences. Special thanks are given to Ardis Johnston (Museum of Comparative Zoology, Harvard University) for loan of Schmidt's material. LITERATURE CITED BOURY-ESNAULT, N. & RUTZLER, K. 1997. Thesaurus of sponge morphology. Smithsonian Contributions to Zoology 596: 1-55. LENDENFELD, R. VON 1903. Tetraxonia. Das Tierreich 19: 1-168. SCHMIDT, O. 1879. Die Spongien des Meerbusen von Mexico (Verlag von Gustav Fisher: Jena). SOEST, R.W.M. VAN & STENTOFT, N. 1988. Barbados deep-water sponges. Studies on the Fauna of Curagao and other Caribbean Islands 70(215): 1-175. AN UNUSUAL SUBERITID DEMOSPONGE FROM A MARINE ALKALINE CRATER LAKE (SATONDA ISLAND, INDONESIA). Memoirs of the Queensland Museum 44: 477-478. 1999:- To date, the Satonda crater lake, Indonesia, is the only ‘marine’ lake known to have an increased alkalinity compared to seawater (Kempe & Kazmierczak, 1990, 1993; Kempe et al., 1997). The lake was originally filled with freshwater, as evidenced by the presence of fossil peat deposits (3,150 "C-yrs BP). Later, the lake was rapidly filled with seawater, as indicated by the settlement of a marine fauna. Today the lake is divided into three water bodies with differing salinities, separated by two pycnoclines (Kempe & Kazmierczak, 1990, 1993; Kempe et al., 1997). Both bottom water layers are anaerobic due to intensive oxygen consuming bacterial decomposition processes, linked to the large input of organic matter from the surrounding vegetation. As a result, an intense sulfate reduction occurs in both bottom water bodies, producing high amounts of bicarbonate ions. As a result of seasonal mixings events, waters from the upper layers of these high-alkalinity bottom waters are transferred to the well oxygenated mixolimnion, causing a slight rise in alkalinity to 4-5 meq/I in the brackish (32 %o salinity) surface waters (‘alkalinity pump’; Kempe, 1990; Kempe & Kazmierczak, 1993; Kempe et al., 1997). The pH values of mixolimnion waters range between 8.3-8.6. As a consequence of raised carbonate alkalinities, the lake generally contains a decreased amount of Ca”, (Cont. over) 478 The red-algal-microbialite reefs exhibit a vertical development which started with a serpulid framework, followed by microstromatolites encrusting former green algal filaments, and loose crusts largely composed of the aragonitic squamariacean red alga Peyssonnelia (Kempe & Kazmierczak, 1990, 1993; Kempe etal., 1997). The uppermost calcareous crust is formed by a framework of Lithoporella, Peyssonnelia and intercalated micrite layers of presumed microbial origin. The living reef community is located on top of this layer. The special hydrochemical situation might be responsible for the very specific and endemic development of the biota. Cyanobacteria and heterotrophic microbes exhibit large diversities in contrastto just one sponge taxon (Laxosuberites n. sp.). Common are cyanobacteria of the morphological taxa Phormidium, Calothrix, Pleurocapsa, Hyella, and Spirulina, in addition to unidentified taxa (Arp et al., 1996) The steep slopes of the red algae reefs are entirely covered by a dense curtain of Cladophora tufts extending to 15-16m depth. Sponges grow underneath this curtain, where light regimes are between 200-300lux. In 1993, the depth limit of sponges was recorded at 20m, i.e. slightly above the upper pyenocline. The sponge fauna is represented by different morphotypes of the hadromerid taxon Suberites, which is characterized by tylostyle megascleres only. The dermal layer of the sponge is constructed of plumose bundles of short tylostyles (150-200um long), the choanosomal spicules are randomly orientated and much larger than the dermal ones (300-5001). Most of the sponges observed exhibit a lateral, encrusting growth habit and therefore show a well developed exhalant canal system. The exhalant system is differentiated into star-shaped units Castrorhizae”-pattern). [n each unit the main exhalant canals conjugate in one large osculum. A second sponge morphotype exhibits a more-or-less erect growth habit and does not show any star-shaped outer exhalant system. This sponge forms tubes with a central osculum, reminiscent of Polymastia and which is phylogenetically closely related to Suberites. The new species is referred to Laxosuberites, based on its spicule inventory, geometry and arrangement. This species shows many color variations: from dark green, brown, yellow/brown to yellow. Different colors are related to microorganisms within the soft tissue. The dark green color is restricted to specimens in extremely shallow water (20-50cm depth), produced by living unicellular green algae enclosed within the mesohyle of the sponge. These algae are part of the plankton and filtered by the sponge. The brownish color variation is restricted to few specimens from deeper water (18-20m depth), with coloration related to the presence of large populations ofa still-unidentified mesohyle bacterium. The presumably symbiotic, native bacterial flora of the sponge is rare and very small (less then lum — ‘nanobacteria’). The size and abundance of these bacteria are comparable to those observed in the MEMOIRS OF THE QUEENSLAND MUSEUM marine hadromerid coralline sponge Acantho- chaetetes. In many cases the encrusting sponges form very thin films (ca, 50um thick), growing within interspaces of dead red algal knobbs. The sponges occupy large spaces between the dead portions of the algal reef surface. Apparently, they prefer light protected areas, except forthe algaebearing specimens. Theoretically these sponges are particle feeders. Within vacuoles of archaeocytes, for example, the remains of diatoms were observed. However, the sponges also have abundant ostia in their basopinacoderm, that is, in all observed cases, growing on active heterotrophic biofilms. This may suggest a close relationship between the biofilms and the sponge. We assume that the biofilms release metabolic products consumed by the sponge. This behavior may also explain the enormous lateral growth of thin sponge sheets. Of further significance is the ability of these sponges to build resting bodies, which are located in small protected cryptic niches between coralline algae or in small caverns 200-500um diameter. The resting bodies are hemispherical or sack-shaped, and filled with archaeocytes/ thesocytes.The sponge fauna seems perfectly adapted to this extreme environment. Pending additional ultrastructural studies, we assume that these sponges are new and restricted to this special environment. O Porifera, suberitids, alkaline crater lake, Indonesia. Literature cited. ARP, G., REITNER, J., WÓHRHEIDE, G. & LANDMANN, G. 1996. New data on microbial communities and related sponge fauna from the alkaline Satonda crater lake. Góttinger Arbeiten zur Geologie und Paläontologie, Sonderband 2: 1-7. KEMPE, S. 1990. Alkalinity: The link between anaerobic basins and shallow water carbonates ? Naturwissenschaften 77: 426-427. KEMPE, S. & KAZMIERCZAK, J. 1990. Chemistry and stromatolites of the sea-linked Satonda Crater Lake, Indonesia: A recent model for the Precambrian sea ? Chemical Geology 81: 299-310. 1993. Satonda Crater Lake, Indonesia: Hydrogeo- chemistry and Bicarbonates. Facies 28: 1-32. KEMPE, S., KAZMIERCZAK, J., REIMER, A., LANDMANN, G. & REITNER, J. 1997. Satonda: a porthole view into the oceanic past. Pp. 156-166. In Tomascik, T., Mah, A.J., Nontji, A. & Moosa, M.K, (eds) The Ecology of Indonesian Seas (Periplus Editions: Jakarta). Joachim Reitner (email: jreitne@gwdg.de), Gert Worheide*, Gernot Arp & Andreas Reimer, Institut und Museum für Geologie und Paläontologie, Universität GóttingenGoldschmidt-Strasse 3, D-37077 Göttingen, Germany; John N.A. Hooper, *Queensland Museum, P.O. Box 3300, South Brisbane. Qld 4101 Australia; Stephan Kempe, Darmstadt, Germany. INNOVATIVE NEW METHODS FOR MEASURING THE NATURAL DYNAMICS OF SOME STRUCTURALLY DOMINANT TROPICAL SPONGES AND OTHER SESSILE FAUNA C.R. PITCHER, T.J. WASSENBERG, G.P. SMITH, M. CAPPO, J.N.A. HOOPER AND P.J. DOHERTY Pitcher, C.R., Wassenberg, T.J., Smith, G.P., Cappo, M., Hooper, J.N.A. & Doherty, P.J. 1999 06 30: Innovative new methods for measuring the natural dynamics of some structurally dominant tropical sponges and other sessile fauna. Memoirs of the Queensland Museum 44: 479-484. Brisbane. ISSN 0079-8835. Population dynamics (growth, mortality, recruitment, and reproduction) of large sessile fauna (including sponges, gorgonians, and corals), that dominate and provide structure on patches of seabed habitat in open shelf waters, 20-50m depth, is currently under study at several sites in the Great Barrier Reef region. Sites, chosen to contain representative benthos habitat fauna, are being surveyed using an ROV to document dynamics, including: mapping the large sessile fauna at each site; tagging a size range of individuals of each of several dominant species; measuring growth and mortality rates through time; observing the occurrence of new small individuals for measurements of recruitment; taking samples to confirm taxonomy and for histological examination in the laboratory to determine reproductive strategies. By tagging a full size-range of individuals of each study species, we aim to estimate lifetime growth curves in three years. From these data we will construct models of the population dynamics of sessile fauna, which can be used to estimate how fast the seabed habitat might recover in new reserve areas, This study will also document the usage of living seabed habitat by key fish species. CJ Porifera, sessile fauna, dynamics, growth, mortality, recruitment, ROV, video, tagging. C.R. Pitcher (email roland. pitcher(Amarine.csiro.au), T.J. Wassenberg, G.P. Smith, CSIRO Marine Research, PO Box 120, Cleveland, Old. 4163; M. Cappo, P.J. Doherty, Australian Institute of Marine Science, PMB No 3, Townsville MC, Old. 4810; J.N.A. Hooper, Queensland Museum, PO Box 3300, South Brisbane, Old. 4101, Australia; 26 May 1999. Communities of large sessile epibenthic fauna (such as sponges, gorgonians, alcyonarians and corals), provide structural complexity to the seabed - an important component of habitat for a myriad of other species - and also contribute to the biodiversity of the marine environment (Van Dolah et al., 1987; Hutchings, 1990; Pitcher, 1997). Further, they form the basis of “bioprospecting’ to discover natural products of pharmaceutical promise (Hooper et al., 1998). However, mega- benthos communities are vulnerable to damage by sedimentation, dredging and extensive trawling on the seabed (Pitcher et al., 1997; Sainsbury et al., 1997). A recently completed project (Poiner et al., 1998) has demonstrated the significance of impacts of prawn trawling on tropical seabed habitats. One method of managing these impacts for ecological sustainability will be spatial closures (Sainsbury et al., 1997), e.g. establishing refuge areas to preserve representative seabed habitats. Predicting the response of megabenthos to the establishment of refuge areas, and acquiring an understanding of the ecological interactions between trawled and refuge areas, are both essential steps in the design of effective refuges for fisheries habitat and the stocks and biodiversity they support. To achieve these goals, it is first necessary to obtain information on the recovery rates of habitat and the processes that link trawled areas and refuges. Estimation of recovery rates requires inform- ation on population dynamics, but these are virtually unknown for large sessile epibenthic fauna (Hutchings, 1990). We are investigating the fundamental population dynamics (recruitment, growth, mortality, reproduction) of structurally dominant megabenthos habitat fauna and documenting the relationship between benthic habitat and ecological usage by some commer- cially important finfish species. MATERIALS AND METHODS STUDY AREA. Seabed areas located in the Great Barrier Reef off Townsville, Australia, were surveyed for suitable study sites in August 1997, during a voyage of the AIMS vessel RV ‘Lady Basten’. Towed video cameras were used in grided surveys of four areas, each about 4,000km?, in the main lagoon and among the mid-shelf reef matrix. Four sites were chosen in an inshore area around the Palm Islands (18.7°S, 146.5°E), in depths ranging from 20-30 m. Another four siles were chosen in a mid-shelf area in the vicinity of the Slashers Reefs (18.5°S, 147.1°E), in depths ranging from 30-50 m. There is evidence that inshore areas have higher productivity of food organisms for filter feeders, than offShore areas (Adele Pile, pers, comm.) — consequently dynamics can be expected to differ between these sites. Each site was chosen to contain benthos habitat with species representative of those found on the lypes of seabed that are trawled for prawns or fin-fishes. DEMOGRAPHIC MEASUREMENTS OF MEGABENTHOS. The population dynamics of sessile fauna that provide structural habitat is being documented by mapping the dominant fauna at each site; tagging several dominant species of sponges, gorgonians, and alcyonarians to identify individuals; making video measurements of indivi- dual growth and mortality rates through time; observing the occurrence of new small individuals in quadrats for measurements of recruitment; taking samples to confirm taxonomy; and histological examination in the laboratory to determine repro- ductive strategies. At cach site, a 4x3m quadrat was established to measure recruitment. Initially, all individuals of all species of sessile fauna within each quadrat were tagged so that the appearance of new tndivi- duals can be detected, mapped and tagged. Typically, 10-20 individuals were present and 106 80 Sa Sue pen] 40 20 a 20 0 5] #0 100 Size tyear- MEMOIRS OF THE QUEENSLAND MUSEUM tagged in the quadrats. The settlement ofany new individuals on 0.25m* concrete marker blocks placed at one corner of each quadrat will also be recorded. To estimate lifetime growth curves in three years, we tagged across the full size-range of individuals of several species common in the area, and for the next three years we will measure tagged individuals every six months. The dominant species included sponges (Xestospongia, lanthella, Cymbastela, Ircinia), gorgonians (Clenocella, Subergorgia. Semperina, Menella, Junceella, Muricella, Echinogorgia) and the hard coral Turbinaria. We attempted to cover the size spectrum available at each site by tagging 3-5 individuals, varying in size from small to large, of each species. The absolute size range depended on the species. After tagging individuals in the recruitment quadrat, the size spectrum was completed by choosing animals within a 20-30m radius of the quadrat. Typically 35-50 individuals were tagged at each site. Large and/or old sessile fauna may grow very slowly and, in a three-year study, their growth may not be measured as precisely as small or young fauna. To counter this, benthos are being measured as accurately as possible, using laser scaling equipment and video image capture and analysis techniques (see below), The latitude/longitude position on the seabed of each tagged individual was recorded with an underwater tracking system and differential GPS. The positions of tagged individuals were mapped so they could be found easily on subsequent 100 80 bd Sire [emi an > a E] qu 15 w Age (years) FIG. |. A, Example ofa Ford-Walford regression plot of size (yr) vs, size (yr) to estimate the parameters growth rate K (trom slope=e*) and asymptotic size Le (from intersection of regression with Y=X) for a hypothetical species of sessile benthic fauna. B, The corresponding estimated von Bertalanffy growth curve for the species that may reach a size of 100cm after more than 20 years. NEW METHODS FOR MEASURING DYNAMICS OF SPONGES Li —-DGPS & Radio-link SU — Clump Winch ( dH y — am Tracking Receiver Umbilicat—| | \ Responder /| \ | \ / ROV Clump Weight _ Co — Pa UN (Cmm Ny FIG. 2. Schematic diagram showing the method of anchoring a 20 m vessel over the study sites and deployment of the ROV on nylon rope. occasions, for measurement. Growth of tagged individuals will be estimated by measuring increm- ents in linear and/or areal dimensions seasonally. Growth curves of the von Bertalanffy form will be parameterised by statistical analysis of Ford- Walford Plots (e.g. Fig. 1; Gulland, 1983). Mortality will be estimated by the disappearance oftagged individuals. Mortality, when not directly observed from skeletal or decayed remains, can be separated from tag loss by cross-checking any apparent losses with accurate position information and the photographic record. Every six months, separate specimens of the same suite of species are being collected for histological studies of reproductive condition. The taxonomy, identification and reproductive studies of the sessile fauna are being undertaken atthe Queensland Museum. We have concentrated on relatively few species of structurally dominant fauna and, even if we cannot assign scientific names to each, we will be able to separate differ- ent species and determine which different forms belong to the same species. LOGISTICS OF RESEARCH. Tagging in the marine environment is typically troubled by fouling and grazing by fish, which lead to difficulties with tag reading and tag loss and associated ambig- uities. To minimise these problems and facilitate identification, the tags used in the study were radio-frequency identification tags in 23x4mm glass capsule form (Texas Instruments RI-TRP- RRHP), that were read automatically by an induction transceiver (Texas Instruments TIRIS Series 2000 module) mounted in an underwater housing. The 481 tags were attached to sessile epifauna by cable-ties, or inserted into sponges with a large needle, or moulded into epoxy pucks placed at the base of the target animal. A small remotely operated vehicle (ROV — Hydrovision ‘Offshore HYBALL’) is being used to conduct most of the underwater tasks. SCUBA divers assist to a maximum depth of 30 m, by setting up quadrats and tagging benthos in shallow sites. The ROV conducted these tasks on deeper sites, where it can operate for virtually unlimited periods. An acoustic underwater tracking system and differential GPS navigation enables accurate (+1m) latitude/longitude position fixing and location of tagged fauna for measurement at each sampling time. The ROV telemetry link also allows data such as tag numbers to be auto- matically acquired in real time, displayed, logged to database along with corresponding position, video frame numbers and captured image file- names. A pair of parallel lasers fitted in the ROV provide a 100mm scale on the video images of megabenthos for measurements. A manipulator on the ROV is used to apply tags and take samples. Deployment of the ROV involves anchoring the vessel precisely over the study sites with a 800kg weight as an anchor on a 25mm plaited nylon rope that can absorb up to 30% rise and fall of the vessel on the waves (Fig. 2). The ROV umbilical is clipped onto the rope to minimize the drag due to currents. This method is simple and effectively enables repeated, accurate positioning of the vessel over the study sites. Integrated data acquisition, storage and retrieval are central to the logistics of the field operations and analysis. Custom software controls this integr- ation of data for vessel position and orientation, ROV tracking, video frame, and tag numbers (Fig. 3). It also provides a navigation system that gives accurate coordinate positions of the vehicle, which are overlaid on the video tape record and displayed as an ROV track on a plotter window. The positions of tagged fauna are shown as way- points to facilitate their location (Fig. 4). When a tagged animal is detected, previous images of that subject are displayed for confirmation and to enable the same image orientation and perspective to be captured. The laser scaled images of fauna recorded from the ROV’s video camera are captured live or from tape. The lasers are calibrated by projecting onto scaled grids to check accuracy and precision of measurements of size through time. Image analyses are achieved efficiently by using a custom software application to control, link and synchronise the field 482 MEMOIRS OF THE QUEENSLAND MUSEUM Ships Gyro Ships Ploter CJ | WO. Computer Monitor receiver Video a O VER L Sonar u ms - Video dT ROTH |] Camera | Qua! Tag Interrogator FIG. 3. System diagram showing integration of components necessary for automated tracking of the ROV and synchronous logging of position, tracking, tag numbers and video data, to facilitate post-analysis and measurement of sessile benthic fauna. DGPS: differential global positioning system, VCR: video recorder, PC: logging computer, SCU: surface control unit for ROV, TXD: tracking system transducers. FIG. 4. User interface of the custom acquisition, tracking and logging software, showing vessel-relative ROV tracker (right window) and ROV navigation plotter (left window) with waypoints (e.g. 0070 to 0074), ROV track (irregular black line), vessel position (arrowhead), ROV position (pale circle off starboard bow near 0073). The lower window shows data acquired and status. NEW METHODS FOR MEASURING DYNAMICS OF SPONGES databases (of tracking, positioning, and tag numbers) with the video images and execute macros on the Optimas® image analysis software. An operator digitises the lines for the laser points (100 mm scale), height, width, and area of profile as appropriate for the growth form (Fig. 5) and Wallie Ch Deep HW) 000 ain SYST 024 Ae i FIG. 5. Captured image of a computer screen showing results of the macro run on Optimas by the custom control software, to measure benthic fauna — Xestospongia testudinaria in this example. The tag interrogator antenna is visible in the lower left of the image. 483 chooses species and condition information from a select list. The software transfers the measure- ment data to a database along with the image filename and corresponding field data. This provides a semi-automated method for extracting the required quantitative data in the form of date/ time, site, tag-number, species, position, morpho- metrics and condition. RESULTS The first tagging fieldwork was conducted in March 1998. Eleven of the most abundant species were targeted for tagging and a total of 174 sessile fauna were successfully tagged, including 26 putat- ive species. The available size-spectrum of each species was successfully covered in most cases (Table 1). Specimens of these species were collec- ted for taxonomic confirmation. The second tagging field trip is currently under- way (December 1998). In the four study sites around the Palm Islands, 95% of tagged individ- uals have been re-measured, 9 individuals were not re-located, 5 incidences of mortality were confirmed, and 13 new recruits were observed in the quadrats. To date, results have demonstrated that the logistics of the project are working as planned. Tagged indiv- iduals can be re-located successfully, tags can be re-interrogated and cross-referenced in the database, captured images can be meas- TABLE 1. List of species targeted for tagging during fieldwork in March 1998, with statistics of each species for height (cm) of tagged specimens, count of numbers tagged by size-categories in classes of one standard deviation (sd) unit relative to the mean, and total count. ured with mm accuracy, recruit- ment can be observed and mortal- ities confirmed. The method- ological protocols that have been established for ROV Species Min | Mean | Max | Small | Med-small | Med-large | Large | Total deployment and for benthos Height | Height | Height | <-1 sd | -10 0] >1 sd | Count measurement will be used for Ctenocella | 196 | 341 | 699 | 4 6 8 4 | m | future repeated visits to indivi- pectinata y - q dual tagged fauna to provide a Xestospongia 9.8 2312 42.1 3 5 6 4 18 istent . f testudinaria Ere iS 3 consistent series of measure- Menella 50 | 180 | 289 | 2 3 10 1 16 | ments. Cymbastela eei | SoA Joar] A35 3 4 4 2 13 DISCUSSION Subergorgia reticulata | 146 | 43.9 | 1081 | 2 | 4 4 M. Our study has developed Turbinaria | 20 | 244 | 62.0 1 5 6 1 13 | techniques for in situ invest- Semperina | 100 | 346 | 533 i y J » igations of the types of large brunei : i sessile fauna that provide Tänihella i68 T | 467 | 4 1 à 2 T structural habitat in deeper asa areas, where access by divers ox 90 | 275 | 412 2 2 7 0 1 is limited orimpossible using rere? fet eae? ll aod 7 : i A * conventional breathing ud — 2 : = equipment. These methods “hi j ? B Echinogorgia | 20.4 33.6 45.8 1 2 2 1 6 open the opportunity for under- Others ?7 | standing the poorly known 484 ecology of these faunas, and will lead to greater appreciation of their role and importance. The study will also document usage of living habitat by key fin-fish species, in terms of species micro-distribution, shelter requirements, and food chain links. The growth, mortality and recruitment rates estimated by this study will be used For develop- ing population dynamics models of the large sessile fauna, The structure of these models will be of a size-based matrix form (c.g. Hughes, 1984). Basically, for each species, the models will have several size categories, the number of which will depend on life history characteristics; recruitment to the smallest size-category will be the probability of settlement of larvae; each larger category will receive recruitment due to growth from lower categories; and individuals within cach size category will have a probability of dying or growing to the next size category (the possibility of negative growth will also be included if required, Other factors to be included in the models include the possible effects of density of the same and other benthos taxa, and the repro- ductive potential and proximity of source popul- ations On supply of larvae. Results of this study can be used to examine a number ot issues, including: the establishment of refuge areas on the seabed, trawling strategies, habitat restoration, stakeholder conflicts, and cons- ervation, These issues revolve around the impact of awling on seabed habilat and associated stocks, and the rate of recovery of habitat if areas were reserved. In particular, models will contribute to the development of management strategies to improve the environmental sustainability of trawl- ing, by simulating the interaction ofthe dynamics of habitat fauna with the dynamics of trawl impacts and estimating levels of trawl effort that do not cause continuing degradation. Information of this kind will become increas- ingly important as the requirement for ecologic- ally sustainable fisheries management is implemented in trawl fisheries from the temperate zone to the tropics, The lessons learned from this study in the form of knowledge of habitat dynamics, and methods tor monitoring habitats and the commercial stocks they support, will contribute to a rational balance between ecolog- ically sustainable fishing and biodiversity conservation when ESD related management objectives are implemented in those Australian fisheries dependent on seabed habitat. MEMOIRS OF THE QUEENSLAND MUSEUM ACKNOWLEDGEMENTS The Fishing Industry Research and Develop- ment Corporation. CSIRO. AIMS and the Queensland Museum are funding this project. B. Hilland D, Vance provided constructive criticism on the manuscript. LITERATURE CITED GULLAND, J.A, 1983. Fish Stock Assessment: a manual us ir methods, (FAO/Wiley Inter-Science: New rark). HOOPER, JNA. QUINN, RJ. & MURPHY, ET. 1998. Bioprospecting for marine invertebrates Pp. 109-112. In Proceedings of the Bioprospecting, Biotechnology & Biabusiness Conference, Decem- ber 1998 (University of Westem Australia: Perth). HUGHES, TP. 1984. Population dynamics based on individual size rather than age: a general model with a coral reef example. American Naturalist 123: 778-795. HUTCHINGS, P. 1990. Review of the effects ol trawl- ing on macrobenthic epifaunal communities. Australian Journal of Marine and Freshwater Research 41:111-120 PITCHER, C.R (997, Status of Inter-Reefal Benthos in the GBR World Heritage Area, Pp. 323—334. In State of the GBR World Heritage Area 1995. Technical Workshop Proceedings, November 1995, GBRMPA Workshop Series #23. (Great Barrier Reef Marine Park Authority: Townsville). PITCHER, CR., BURRIDGE, CY, WASSENBERG, T.J. & POUNER, LR. 1997. The effects of prawn trawl fisheries on GBR seabed habitats. Pp. 107-123. In The Great Barrier Reef, science, use and management: a national conference: Proc- eedings, Vol. 1, (Great Barrier Reef Marine Park Authority: Townsville). POINER, LR., GLAISTER, J., PITCHER, C.R. BURRIDGE, C., WASSENBERG, T., GRIBBLE, N., HILL, B., BLABER. S..M., MILTON, D,M., BREWER, D. & ELLIS, N, 1998. The environ- mental effects of prawn trawling in the far northern section of the Great Barrier Ree! 1991-96. Final Report to GBRMPA and FRDC. CSIRO Division of Marine Research — Queensland Department of Primary Industries Report. (CSIRO: Brisbane). SAINSBURY, K.J., CAMPBELL, R.A.. LINDHOLM, R. & WHITELAW, W. 1997. Experimental manage- meot af an Australian multispecies fishery: Examining the possibility of trawl-induced habitat modification. Pp 107-112 In Pikiteh, E.K., Huppert, D.D. & Sissenwine, M.E (eds) Global Trends: Fisheries Management. (American Fisheries Society: Maryland), VAN DOLAH, RF.. WENDT, P.H. & NICHOLSON, N. 1987, Effects of a research trawl on a hard bottom assemblage of sponges and corals. Fisheries Research 5:39-54 SPONGE FARMING IN THE MEDITERRANEAN SEA: NEW PERSPECTIVES ROBERTO PRONZATO, GIORGIO BAVESTRELLO, CARLO CERRANO, GIUSEPPE MAGNINO, RENATA MANCONI, JANNIS PANTELIS, ANTONIO SARA AND MARZIA SIDRI Pronzato, R., Bavestrello, G., Cerrano, C., Magnino, G., Manconi, R., Pantelis, J., Sara, A. & Sidri, M. 1999 06 30: Sponge farming in the Mediterranean Sea: new perspectives. Memoirs of the Queensland Museum 44: 485-491. Brisbane. ISSN 0079-8835. Some Mediterranean species of Spongia and Hippospongia are characterised by a soft and absorbent skeleton and usually harvested for commercial purposes. Recently, the synergetic effect of a widespread epidemic, together with overfishing, has strongly reduced their density, leading local populations of these species to the brink of extinction. Recovery of populations takes a long time and even now, after several years, sponge density is still very low. A simple solution to this problem is sponge-farming. Sponges are sessile filter feeding organisms and through their pumping activity they are able to retain bacteria and suspended organic matter from the entire water-column in littoral marine environments. This ability provides the basis for an integrated aquaculture of sponges and fish in coastal areas: floating-cage fish farms release a lot of organic wastes that can be recycled as a rich source of food for surrounding intensive commercial sponge cultures. Moreover, the interest shown by chemists and pharmacologists in regard to natural products extracted from sponges creates new possibilities for sponge farming. O Porifera, aquaculture, organic pollution, overfishing, bath sponges, natural products, cicatrisation. Roberto Pronzato (email: zoologia@igecuniv.cisi.unige.it), Carlo Cerrano, Giuseppe Magnino, Antonio Sara & Marzia Sidri, Dipartimento per lo studio del Territorio e delle sue Risorse, Universita, via Balbi 5, 16126 Genova, Italy; Renata Manconi, Dipartimento di Zoologia ed Antropologia Biologica dell'Università, V. Muroni 25, I 07100 Sassari, Italy; Jannis Pantelis, Secretariat of Fishery, Municipality of Kalymnos, GR 85200, Greece; Giorgio Bavestrello, Istituto di Scienze del Mare dell'Università, Via Brecce Bianche 1, 68031 Ancona, Italy; 29 April 1999. Commercial ‘horny’ or Keratose sponges have been harvested and utilised as bath sponges since ancient times. Phoenicians and Egyptians collected sponges stranded along the seashore, while the millenary history of sponge fishery takes root in the ancient Greek civilisation. Trad- itionally fishermen used a heavy stone as ballast to easily reach the sea bottom and gather sponges into anet basket. At the end of the last century this harvesting system was replaced by ‘hard hat’ diving-suits. The introduction of this device rapidly increased fishing effort, although many divers died from decompression sickness, and as a consequence many Governments banned this technique. Modern developments in hyperbaric medicine and diving equipment have since solved both legal and medical problems assoc- iated with commercial diving activities, but have created a new suite of problems for sponge fisheries. In recent times sponge population density has continually decreased, both through overfishing and from the so-called sponge disease. Older professional divers relate the existence of an incredible abundance of commercial species during the 1930s, along the coasts of Cyprus, Crete and Sardinia, consisting of more than 200- 300 specimens/100m?. Prior to the sponge disease epidemic, unexploited commercial sponge banks contained sponge densities of about 100 specimens/ 100m”, whereas at present, the mean density is often less than 50 specimens/ 100m? (Pronzato et al., 1996, 1999; Pronzato, 1999), Sponge diseases do not occur frequently, but have been recorded in populations from both the Mediterranean and Caribbean Seas. Between 1985-1988 commercial sponges practically disappeared in many of these areas, especially in the eastern Mediterranean Basin, with conseq- uent heavy economic losses. Sick sponges are easily recognisable through exposure of their internal skeleton. Sponge disease is caused by invasive pathogenic micro-organisms: first they destroy the sponge’s external fibrous layer, then proceed rapidly into the sponge body, destroying living tissues. Fibres become fragile and flake off, losing their characteristic durability and 486 TABLE 1. Percentage of survival of monitored species farmed in Kalymnos and Paraggi. Hippospongia communis, Petrosia ficiformis and Cacospongia mollior are perfectly suited, while Axinella damicornis and Ircinia variabilis are unsuitable. Abbreviation: N=no. of transplanted fragments. MEMOIRS OF THE QUEENSLAND MUSEUM Due to depletion of sponge populations from 1960-1990 many commercial producers have gone out of business, with exports from many Mediterranean countries diminishing substantially. The decrease in catch has produced a sharp increase in price for Mediterranean sponges, with the consequence that lower quality, cheaper Caribbean and Pacific stocks have invaded the market (Verdenal & Vacelet, 1990). Recovery of the sponge banks is a long term process (Rizzitello et al, 1997). Ten years after the onset of Mediterranean sponge disease, commercial sponges are still rare in many sites we examined during our experiments (this paper). In the last few years chemical Species N verdes after Survival after hours (%) | 2 months (%) Kalymnos (Dodecanese, Greece) Spongia officinalis 75 69.4 69.4 Hippospongia communis 252 100 100 Paraggi (Ligurian Sea, Italy) Agelas oroides 46 45.8 44 Axinella damicornis 50 0 0 Cacospongia mollior 60 83.3 83.3 Ircinia variabilis 50 2 1 Petrosia ficiformis 40 100 98 Spongia officinalis 50 66 66 Spongia agaricina 49 42.8 40 researchers have also shown an interest in softness (Gaino & Pronzato, 1989; Pronzato & Gaino, 1991; Gaino et al., 1992). There is undoubtedly a synergetic effect between over- fishing and sponge disease in reducing populations, given that overexploitation may lower the sponge’s self-defence mechanisms, increasing the risk of environmental aggravation (Pronzato, 1999). Moreover, it is also known that types of pollution are responsible for decreasing biodiversity amongst Porifera (Pansini & Pronzato, 1975; Carballo et al., 1996). FIG. 1. The sponge experimental plant in Kalymnos: arrows indicate the structures for sponge farming, moored on the bottom. sponge culture, owing to the presence of natural products useful in pharmacology. Metabolites extracted from Porifera are providing promising results in the prevention and treatment of tumours (De Flora et al., 1995), antiphlogistic compounds (De Rosa et al., 1995) and other properties (Uriz et al., 1991). More- over, sponge extracts appear in catalogues of laboratory products at very high prices. Our goal is to satisfy the market request for these products without reducing natural populations. This paper reports on the preliminary results from two different experiments in sponge farming. The first aimed to reconvert sponge fishery toward a more profitable and environmentally sustainable activity, located in Kalymnos Island (Dodecanese, Greece), during March 1998. The two target species were Spongia officinalis and Hippospongia communis, the most common commercial species in this area. The second experimental sponge culture was directed towards pharmacology, located in the W Mediterranean (Paraggi, Ligurian Sea), testing the survival of different non-commercial sponge species under farming conditions. MATERIALS AND METHODS The study site in Kalymnos (Dodecanese, Greece: 36°58’N, 27°02’E), was situated in the sheltered Bay of Vathi where there was a fish farming plant hosting 30 floating cages. During March 1998 four metallic SPONGE FARMING IN THE MEDITERRANEAN FIG, 2. The horizontal structures moored on the bottom with sponge fragments fixed onto lines. horizontal structures were moored on a flat bottom 200-500m away from the floating fish cages at 15m depth (Figs 1-2). Specimens of S. officinalis var. adriatica and H. communis were cut into 4x4cm fragments and threaded onto a nylon line, separated by plastic tube spacers (7x0.6cm) (Figs 2, 3A,B). In total, 350 fragments were attached to lines. A team of operators monitored the plant of Kalymnos daily for the first week, and subsequently every ten days for the following two months. Using the same method, the plant at Paraggi (Ligurian Sea, Italy: 44°18°N, 9%09"E), was situated on a flat bottom, at a depth of 25m. During May 1998 fifty fragments were obtained from each of the following species: Agelas oroides, Axinella damicornis, Cacospongia mollior, Chondrosia reniformis, Ircinia variabilis, Petrosia ficiformis, S. agaricina and S. officinalis var. adriatica. These were fixed onto horizontal structures using the methods described above. These species are the most common sponges living on the rocky cliffs ofthe Ligurian Sea. During the first week of experiments samples from the cut surfaces were collected daily from both plants (Paraggi and Kalymnos), fixed in Glutaraldehyde 2.5% in ASW, dehydrated in a graded series of ethanol, critical-point dried using a Pabish CPD 750 drier coated with gold using a Balzers SCD 004 coater, and observed under a Philips EM 515 scanning electron micro- scope. RESULTS AND DISCUSSION Many experimental approaches have been applied to investigate the problem of producing a profitable sponge culture since the beginning of the century, and at present, some data are well- grounded (see Pronzato, 1999, for a review). From these data, on average, over two years sponges increase their volume by 100-200%. 488 FIG. 3. A, Fragments of Hippospongia communis, B, Spongia officinalis just after transplantation: portions of the dark original external fibrous layer are maintained. C, Details of the thin cell laver in H. communis after 24hrs recovery. D, Thin cell layer recovering in S. officinalis. E, Spongia officinalis three weeks after transplantation: the charact- eristic dark pigmentation and the rounded shape have been restored. F, A dead specimen of S. officinalis. Death occurs mainly after the first 48hrs of transplantation. (Scale bars-lem). Generally, smallest fragments show the highest growth rates (Verdenal & Vacelet, 1985). KALYMNOS PLANT. Mean mortality, less than 20%, was limited to the first 48hrs after trans- plantation, and A. communis seemed to be more resistant than S. officinalis. In fact, S. officinalis showed a survival rate of 69.4% whereas H. communis gave excellent results with a survival of 100% (Table 1). Mortality may be due to high sedimentation rates which favours bacterial MEMOIRS OF THE QUEENSLAND MUSEUM proliferation: only naked skeletons of dead fragments remained on the ropes (Fig. 3F). The regenerative process. starts immed- iately after transplantation. Within 2-3 days sponges rebuild their external protect- ive layer; after 24hrs a thin transparent cell layer covers the cut surfaces (Fig. 3C,D). After one week the characteristic dark external pigmentation of the sponge was restored. Afler one month sponge frag- ments assumed a rounded shape, with the external fibrous layer and the aquiferous system of the cut surfaces completely reorganised (Fig. 3E). Among all the phyla of filter feeders, sponges play a remarkable role in the auto-epurative processes of the sea (Sara. 1973). In accordance with modern integ- rated aquaculture systems, the association of sponge culture with floating cage fish farms have the potential to reduce enyiron- mental impact on coastal areas due to pollution produced by intensive fish farming (Manconi et al., 1998; Pronzato et al., 1998). The major impact occurs on the sea bed, under floating cages, where a rain of particles falls on benthic organisms causing rapid eutrophication (i.e. decrease in dissolved oxygen and increase in nutnent levels) (Wu, 1995), Food wastes and faecal pellets released by captive fish are rapidly colonised and degraded by bacteria (Honjo & Roman, 1977). Filtering activity of sponges has the potential to contribute to reduce this pollution within the precinct of fish farms, Sponges can retain about 80% of organic particulate material suspended in water, and about 70% of bacteria (Reiswig, 1971, 1975), with sponges filtering the entire water column in a single day (Reiswig, 1974). This integrated aquaculture provides a double bonus: purified water and commer- cial bath sponges. Following our first attempt the Munic- ipality of Kalymnos is presently planning to farm many thousands of sponges within the boundaries of floating fish cages along the island's coast. PARAGGI PLANT. Petrosia ficiformis was the most productive species at this site. The cut surface produced a new pinacoderm within 4-5 days, and survival of fragments was close to 100% (Table 1). SPONGE FARMING IN THE MEDITERRANEAN FIG. 4. The cicatrisating process of some investigated species. A, The cut surface on the day of transplantation in Agelas oroides. B, Cut surface of Agelas oroides after three days when the exopinacoderm is completely restored. C, Cut surface of Axinella damicornis after the first day. D, Cut surface of Axinella damicornis after the third day, where the external layer has not yet rebuilt. E, Cut surface of Petrosia ficiformis immediately after cutting. F, Cut surface of Petrosia ficiformis after three days, where the external cell layer is perfectly reconstructed. (Scale bars: A, D, E, F=100um; B=10um; C=1mm). Percentage survival rates of S. officinalis and C. mollior were satisfactory (about 60-80% over two months), and data on S. officinalis were concordant between the Kalymnos and Paraggi plants (Table 1). It is important to underline that environmental conditions, the state of health of the mother sponge, and techniques used in trans- plantation all influence survival and growth rate of farmed specimens, as also noted by Verdenal & Vacelet (1990). For instance, in their Marseilles farm, Verdenal & Vacelet (1990) found that S. agaricina showed a survival of 100% whereas we observed a mortality rate of about 60%. 489 Axinella damicornis and I. variabilis showed a high mortality rate, probably due to damage incurred during the cutting process. In fact, A. damicornis is very fragile and must be handled with care, whereas /rcinia is so compact that it is difficult to cut without squeezing and poten- tially damaging tissues (Table 1). Of all species tested, C. reniformis was completely unsuitable for our experi- mental conditions. The collagen matrix cut itself on the thread and in 1-2 days the sponge ‘dripped’ down from the thread. This behaviour, reminescent of the variable structure of Echinoderms (Candia Carnevali et al., 1990), is very interesting and the subject of a recent study (Bavestrello et al., 1998). Recovery of the exposed choanosome starts from the borders of the cut and increases concentrically. The reconstruction process differs between species, depending on whether the external layer is areal pinacoderm, or, as for bath sponges (Spongia and Hippospongia), it is a fibrous layer without cells. Our experience shows that in A. oroides (Fig. 4A,B) and P ficiformis (Fig. 4E,F) the restoration of the exopina- coderm is complete in 2-3 days, whereas in 4. damicornis (Fig. 4C,D) this process does not occur at all. Commercial bath sponges show the lowest mortality rate, probably due to the possession ofa fibrous layer in which the recovery process is different from other species. Spherical cells, with long pseudopodia, travel along the cut edges producing collagen fibrils and completing the cicatrisation process after only a few days (Fig. 5). The rapidity of reconstitution of the new external layer on cut portions of sponges varies 44) MEMOIRS OF THE QUEENSLAND MUSEUM FIG. 5. A, The regenerative process of Spongia agaricina shows recovery of a new fibrous layer on exposed spongin fibers, B, Many globous mobile cells occur on the sponge surface. C, Elongated pseudopodia actively produce the collagen deposited on sponge fibers. D, After a few days the superficial fibrous layer is more-or-less completely restored. (Scale bars: A= Imm, B, C=1004m; D=10um). between species (e.g. 2-3 days in A. oroides and P. ficifarmis; to about 10 days 1n C. mollior). CONCLUSIONS Commercial bath sponges have practically dis- appeared from the E Mediterranean Sea, owing to both overfishing and sponge disease. Sponge aquaculture has the potential to decrease fishing pressure, thus facilitating the natural repop- ulation of these affected areas. Pharmacological research on marine natural products extracted from sponges are providing promising results, opening new perspectives in the exploitation of new species of Porifera. However, there are virtually no data on the dens- ities of wild sponge banks for most species, and the risk of overfishing of potentially commer- cially valuable species could become real. An evaluation of the adaptability of the most common Mediterranean species to farming cond- itions could provide a valuable resource to any future exploitation of these species for pharmacological or other products. ACKNOWLEDGEMENTS Special thanks lo Mr Emanuele Bruzzone and Mr Carlo Grattarola for their technical support. LITERATURE CITED BAVESTRELLO, G., BENATTI, U., CALCINAI, B., CATTANEO-VIETTI, R, CERRANO, C., FAVRE, A., GIOVINE, M., LANZA, S., PRONZATO, R. & SARA, M. 1998. Body polarity and mineral selectivity in the Demo- sponge Chondrosia reniformis. Biological Bulletin 195: 120-125. CANDIA CARNEVALI, M.D.. BONASSORO, F., WILKIE, I.C., ANDRIETTI, F. & MELONE, G. 1990. Functional morphology of the peristomial membrane of regular sea-urchins: structural organisation and mechanical properties in Paracentrotus lividus, Pp. 207-216. In De Ridder, SPONGE FARMING IN THE MEDITERRANEAN C., Dubois, P., Lahaye, M.C. & Jangoux, M. (eds) Echinoderm Research. (Balkema: Rotterdam). CARBALLO, J.L., NARANJO, S.A. & GARCIA- GOMEZ, J.C. 1996. Use of marine sponges as stress indicators in marine ecosystems at Algeciras Bay (Southern Iberian Peninsula). Marine Ecology Progress Series 135: 109-122. DE FLORA, S., BAGNASCO, M., BENNICELLI, C., CAMOIRANO, A., BOJNEMIRSKI, A. & KURLEC, B. 1995. Biotransformation of geno- toxic agents in marine sponges. Mechanisms and modulation. Mutagenesis 10(4): 357-364. DE ROSA, S., CRISPINO, A., DE GIULIO, A., IODICE, C., PRONZATO, R. & ZAVODNIK, N. 1995. Cacospongionolide B, a new sesterterpene from the sponge Fasciospongia cavernosa. Journal of Natural Products 58: 1776-1780. GAINO, E. & PRONZATO, R. 1989. Ultrastructural evidence of bacterial damage to Spongia officinalis fibres (Porifera, Demospongiae). Diseases of Aquatic Organisms 6: 67-74. GAINO, E., PRONZATO, R., CORRIERO, G. & BUFFA, P. 1992. Mortality of commercial sponges: incidence in two Mediterranean areas. Bollettino di Zoologia 59: 79-85. HONJO, S. & ROMAN, M.R. 1978. Marine copepod faecal pellets: production, preservation and sedi- mentation. Journal of Marine Research 36(1): 45-57. MANCONI, R., CUBEDDU, T., CORRIERO, G. & PRONZATO, R. 1998. Commercial sponge farming as natural control of floating cages pollu- tion. P. 17. In Abstracts of the International Symposium on New Species for Mediterranean Aquaculture. (Istituto Sperimentale Italiano L. Spallanzani: Milano). PANSINI, M. & PRONZATO, R. 1975. Analisi preliminare sulla distribuzione dei Poriferi in aree sottoposte a differenti tipi di inquinamento. Bollettino dei Musei degli Istituti di Biologia dell’ Università di Genova 43: 21-32. PRONZATO, R. 1999, State ofthe art of sponge fishery, disease and farming in the Mediterranean Sea. Aquatic Conservation: in press. PRONZATO, R. & GAINO, E. 1991. La malattia delle spugne commerciali: considerazioni storico- economiche. Bollettino dei Musei degli Istituti di Biologia dell' Università di Genova 55: 17-25. PRONZATO, R., RIZZELLO, R., DESSY, E., CORRIERO, G. & SCALERA LIACI, L. 1996. Distribuzione e pesca di Spongia officinalis lungo il litorale lonico Pugliese. Bollettino dei Musei degli Istituti di Biologia dell'Università di Genova 60-61: 79-89. 491 PRONZATO, R., BAVESTRELLO, G. & CERRANO, C. 1998. Morphofunctional adaptations of three species of Spongia (Porifera, Demospongiae) from a Mediterranean vertical cliff. Bulletin of Marine Science 63(2): 317-328. PRONZATO, R., CERRANO, C., CUBEDDU, T., LANZA, S., MAGNINO, G., MANCONI, R., PANTELIS, J., SARA, A. & SIDRI, M. 1998. Sustainable development in coastal areas: role of sponge farming in integrated aquaculture. Pp. 231-232. In Grizel, H. & Kesmont, P. (eds) Abstract ofthe International Conference on Aqua- culture Europe *98, Aquaculture and water: fish culture, shellfish culture and Water usage. Vol. 26. (European Aquaculture Society special public- ation: Oostende). PRONZATO, R., CERRANO, C., CUBEDDU, T., MAGNINO, G., MANCONI, R., SARA, A. & SIDRI, M. 1999, Commercial sponges: fishing, farming and integrated aquaculture. Biologia Marina Mediterranea: in press. REISWIG, H.M. 1971. Particle feeding in natural populations of three marine Demosponges. Biological Bulletin 141: 568-591. 1974. Water transport, respiration and energetic of three tropical marine sponges. Journal of Experimental Marine Biology and Ecology 14: 231-249. 1975. Bacteria as food for temperate water marine sponges. Canadian Journal of Zoology 53(5): 582-589. RIZZITELLO, R., CORRIERO, G., SCALERA LIACI, L. & PRONZATO, R. 1997. Estinzione e ricolonizzazione di Spongia officinalis nello Stagnone di Marsala. Biologia Marina Mediterranea 4(1): 443-444. SARA, M. 1973. Animali filtratori ed autodepurazione nel mare: il ruolo dei Poriferi. Pp. 35-51. In Atti IL Simposio sulla Conservazione della natura. Vol. 1. (Cacucci Editore: Bari). URIZ, M.J., MARTIN, D., TURON, X. BALLESTEROS, E., HUGHES, R. & ACEBAL, C. 1991. An approach to the ecological signif- icance of chemically mediated bioactivity in Mediterranean benthic communities. Marine Ecology Progress Series 70: 175-188. VERDENAL, B. & VACELET, J. 1985. Sponge culture on vertical ropes in the north-western Medit- erranean sea. Pp. 416-424. In Rützler, K. (ed.) New Perspectives in Sponge Biology. (Smith- sonian Institution Press: Washington D.C). WU, R.S.S. 1995. The environmental impact of marine fish culture: towards a sustainable future. Marine Pollution Bulletin 31: 159-166. 492 MEMOIRS OF THE QUEENSLAND MUSEUM NEW APPROACHES TO THE BIO- MINERALIZATION PROCESSES OF CALCIFIED SKELETONS IN CORALLINE DEMOSPONGES, Memoirs of ihe Queensland Museum 44: 492. 1999:- Biomineralization of calcareous basal skeletons in coralline sponges is a strongly phylogenetically convergent character (Reitner, 1992), However, the basie mineralization process is ancestral and exhibits similarities with mineralization processes known in bacterial biofilms and organo-mineralization via controlled taphonomy (Reitner et al, 1997). The main biocalcification events in the phylogenetically distinct taxa Maceletia sp., Astroselera willeyana, Cerataporella nicholsoni, and Spirastrella (Acanthochaetetes) wellsi are discussed. Vacelelia, a demosponge with a thalamid basal skeleton, exhibits the most ancestral mode to build a calcareous skeleton via controlled taphonomy. The stromatoporoid Astrosclera willeyana intracellularly forms egg-shaped aragonitic asters ina first step which crow together via an epitactical process. The chaetetid S. (Acanihochaetetes) wellsi, phylogenetically the most evolved coralline sponge taxon, forms its unique high-Mg calcitic skeleton in extracellular acidic organic mucilages in the presence of collagen. In all cases the mineralization is controlled by acidic matrix proteins, All aragonitic basal skeletons are characterized by high amounts of Sr and U. In Ceratoporella nicholsoni an increase of Mg, Sr and U in the old skeletal parts is observed using ICP-MS based geochemical analyses. The decrease of P concentrations is probably linked to the collapse of the intracrystalline matrix proteins. In Ceratoporella two distinct Ca^ - binding matrix proleins are observed which are enriched in the amino acids asp (20 molo) and glu (15 mol%) (18kd, >100kd). The uppermost growing zones exhibit relatively light 8^C 3,8 and. "O —).4 values on average. Astrosclera differs in many aspects from the phylogenetically closely related Ceratoporella, The newly formed aragonite asters are depleted in "C (5C 3.5) in comparison with the mature basal skeleton, The spherulites are enriched in Sr, P, Li and Mo. In the youngest cementing area of the spherulites the content af ^C increases to 8" C 4.03. Astrosclera exhibits five acidic matrix proteins, the smaller ones (17kd, 30kd) controlling the spherulite growth, and a large one (120kd) which probably controlls the cementation process, The young portion of the basal skeleton of Vaceletia n.sp. differs from the aragonitic basal skeletons of the other coralline sponges, Mg and P is extremly enriched and the carbon isotope composition is relatively light (86°C 3.8), Vaceleria exhibits 10 acidic matrix proteins. Acanthochaetetes has the most evolved basal skeleton which exhibit five acidic matrix protein. The small ones (ca, 20kd) control the initial calcification, The basal skeleton is a Me rich calcite (19-20mol% MgC0»). The uppermost growing zones exhibit light 8"C values (2.65). The simultaneously growing inner cements have 6°C 3.03. Based on the measured geochemical and isotopic data a vital effect during the early formation of the basal skeletons is very probable. The old or mature basal skeleton portions exhibit in all cases signals of a mineralization in equilibrium with the ambient sea water. O Porifera, coralline sponges, Vaceletia, reef-building | sphinctozoans, Acanthochaetetes. Literature cited. REITNER, J. 1992. Coralline Spongien. Der Versuch einer phylogenetisch-taxonomischen Analyse. Berliner geowissenschaftlichen Abhandlungen (E) 1: 1-352. REITNER, J., WÓRHEIDE, G., LANGE, R. & THIEL, V. 1997. Biomineralization of calcified skeletons in three Pacific coralline demosponges - an approach to the evolution of basal skeleton. Courier Forschungsinstitut Senckenberg 201: 371-383, Joachim Reitner (email: jreitne(aigwde.de), Gert Würheide*, Robert Lange, Volker Thiel, Anton Eisenhauer, Andreas Reimer & Stephanie Fliege, Institut und Museum fiir Geologie imul. Paläontologie, Universitit Göttingen, Goldschmidtstr 3, D-37077. Göttingen, Germany: Matthias Bergbauer, Fachgebiet Oekologie der Mikroorganismen, OF 5, Technische Universitaat Berlin, Franklins: 29, D-10587 Berlin, Germany; * Present address: Queensland Museum, PO. Box 3300, South Brisbane, Qld, 4101, Australia; 1 June 1998. EVIDENCE FOR SYMBIOTIC ALGAE IN SPONGES FROM TEMPERATE COASTAL REEFS IN NEW SOUTH WALES, AUSTRALIA D.E. ROBERTS, S.P. CUMMINS, A.R. DAVIS AND C. PANGWAY Roberts, D.E., Cummins, S.P., Davis, A.R. & Pangway, C. 1999 06 30: Evidence for sym- biotic algae in sponges from temperate coastal reefs in New South Wales, Australia. Memoirs of the Queensland Museum 44: 493-497. Brisbane. ISSN 0079-8835. The symbiotic relationships between tropical reef sponges and cyanobacteria (microalgae) has been well documented. Preliminary evidence suggests that these relationships may be justas common in temperate reef species. Screening of sponges from temperate reefs in New South Wales, Australia, found 5 out of 8 species tested were *chlorophyll positive'. Of those tested, Cymbastela concentrica (Lendenfeld) had the greatest concentration of chlorophyll-a pigment within its tissue (139.4+9.4ug/g). An estimate of the percentage of temperate reef sponges that potentially contained symbiotic algae was made based on their in situ colour pigmentation. It was predicted that over 65% of temperate reef sponges potentially contain symbiotic algae, although it is unknown how many may be phototrophic. O Porifera, symbi- otic algae, cyanobacteria, temperate sponges, ecology. D.E. Roberts* (email: Dannio(a)bigpond.com.au), Department of Biological Sciences, University of Wollongong, Wollongong 2522, NSW; S.P. Cummins**, Water Science Section, Environment Protection Authority, Locked Bag 1502, Bankstown 2200, NSW; A.R. Davis, Department of Biological Sciences, University of Wollongong, Wollongong 2522, NSW; C. Pangway, Analytical Chemistry Laboratory, Environment Protection Authority, PO Box 29, Lidcombe 2141, NSW, Australia; Present address: *Wyong Shire Council, PO Box 20, Wyong 2259; **Wyong Shire Council, PO Box 20, Wyong 2259; 27 January 1999. Many sponges from tropical reefs contain and benefit from symbiotic relationships with cyano- bacteria or microalgae (Sarà, 1971; Vacelet & Donadey, 1977; Wilkinson, 1978, 1981, 1983; Larkum et al., 1988). Wilkinson (1983) suggest- ed that in some species these algae can provide a substantial source of nutrition for the host sponge (see also Wilkinson, 1981; Wilkinson et al., 1988). In tropical regions, light levels may limit the maximum depth to which phototrophic sponges are distributed. Photosynthetic symbionts are unlikely to provide nutrition beyond the net photosynthetic compensation point (Cheshire & Wilkinson, 1991). In striking contrast, symbiotic algae may even provide photo-protection to some species of sponges from excessive light in shallow water (Sarà, 1971; Wilkinson, 1983). Until recently, phototrophic sponges were thought to be restricted to sub-tropical and shallow tropical reef habitats (Wilkinson, 1983), however a temperate reef phototrophic sponge (Cymbastela sp.) is now known to be an ex- ception (Cheshire et al., 1995). At least one tropical sponge with symbiotic algae, Cymbastela concentrica (Wilkinson, 1983; Larkum et al., 1988; Hooper & Bergquist, 1992; Hooper & Lévi, 1994), also has a temperate range where it is found between 10-60m depth or more (Roberts & Davis, 1996). We predicted that symbiotic relationships be- tween temperate reef sponges and algae may be widespread because of the number of species exhibiting a colour, indicative of microalgae within their tissue (Hooper & Bergquist, 1992). Here we report the results of a screening study where eight common temperate reef sponges were tested for the presence of symbiotic algae. Furthermore, we extrapolated from our records on pigmentation for over 100 temperate species to gain some estimate of the potential for temp- erate reef sponges to contain symbiotic algae. MATERIALS AND METHODS To test for general evidence of symbiosis in temperate reef sponges we sampled a range of common species, some of which were replicated to estimate variation and consistency in both colour and chlorophyll concentration. Eight species of Demospongiae (Table 1) were hap- hazardly collected from the subtidal reefs at Henry Head and Inscription Point at the entrance to Botany Bay, New South Wales (NSW) Australia (Fig. 1), and the concentration of chlorophyll-a within their tissue was determined. 494 Fä 5, PortJackson ) | BotanyBaj , | y ; E* FIG. 1. Location of reefs at Inscription Point and Henry Head, NSW, where sponges were collected to determine the presence of symbiotic algae (1=Henry Head; 2= Inscription Point). The collections were made on sponge-dominated reefs between 10-20m depth. Other common sessile macrobenthic organisms found on these reefs included ascidians, bryozoans, cnidarians, foliose macroalgae and a nondescript matrix of microorganisms and silt (Davis et al., 1997). Spatial and temporal patterns of abundance for these assemblages have been described else- where (Davis et al., 1997). The sponges were photographed, their in situ colour recorded, brought to the surface, field weighed (to the nearest gram), and immediately transported back to the laboratory under dark conditions in fresh seawater. Variation in the concentration of chlorophyll-a within a species was also examined for two of the eight sponges, by collecting five replicate specimens of each of C. concentrica and Clathria striata. All sponges were identified and were included in a voucher collection lodged with the Queensland Museum. To determine whether symbiotic algae were present in each of the sponges, 1g of tissue was removed and processed with 90% aqueous acetone solution. A sub-sample of 1g was found to give a suitable chlorophyll-a absorbance read- ing for a variety of sponge samples. To ensure the complete extraction of the chlorophyll-a pigment, the sponge tissue was mechanically broken down MEMOIRS OF THE QUEENSLAND MUSEUM using a mortar and pestle. The resultant extracted slurry was transferred to a screw-cap bottle and the total volume was adjusted to a constant vol- ume (10ml) and stored at -20°C until analysis. Immediately before spectrophotometric deter- mination samples were removed from the freezer and returned to room temperature in the dark. The sample extracts were filtered through a solvent- resistant disposable filter (Whatman GF/C 47mm diameter), and 3ml of the solution was transferred to a lcm cuvette. Absorbance was measured on a Cary 1E UV-visible spectrophotometer at 664nm for chlorophyll-a, and 750nm, 647nm and 630nm for turbidity, chlorophyll-b and chlorophyll-c corrections, respectively. Before sample analysis, the spectrophotometer absorbance was adjusted to zero by inserting a 90% acetone blank. The concentration of chlorophyll-a was calcu- lated using the methods described in Clesceri (1985). It should be noted that the samples were not subsequently acidified and analysed for pheophytin-a (a common degradation product of chlorophyll-a). All work with chlorophyll extracts was undertaken in cool, dark conditions to minimise potential degradation of chlorophyll. The in situ colour of a sponge may be a good preliminary test or indicator for the presence of symbiotic algae (Hooper & Bergquist, 1992). To estimate the percentage of temperate reef sponges that potentially contained symbiotic algae we examined our earlier collections of over 100 temperate reef species from Sydney to Port Stephens, NSW (Roberts & Davis, 1997; Roberts et al., 1998). Subjective estimates were made on the presence or absence of symbiotic algae based on the in situ colour of each species from field records and photographs. RESULTS Five of the eight species of temperate reef sponges we tested (in four orders and seven families) were found to be ‘chlorophyll positive’. These were Clathria striata, Phoriospongia cf. kirki, Cymbastela concentrica, Callyspongia sp. and Spirastrella areolata (Table 1). Of the five species, C. concentrica had the greatest con- centration of chlorophyll-a within its tissue, followed by Callyspongia sp. (Table 1). Little within-species variation was found in the mean concentration of chlorophyll-a (+S.E.) in the tissue of C. concentrica (139.4+9.4) and C. striata (19.3+1.4) (Table 1). The other three common species tested were also chlorophyll positive (Table 1), although SYMBIOTIC ALGAE IN SPONGES TABLE 1. Summary of sponges from temperate reefs (10-20m depth) screened for the presence of symbiotic algae (* n= 5 sponges). 495 algae based on their in situ colour. It was estimated that of these Chlorophyll-a species, 65% potentially Order Fail Species (ugg) contained symbiotic algae. Hadromerida Spirastrellidae Spirastrella areolata Dendy 20.7 DISCUSSION Halichondrida Axinellidae mban eia conceniriga 139.4 + 9.4 * Mau É ; 7 , Our findings show that over Haplosclerida Callyspongiidae | Callyspongia sp. 85.3 60% of sampled temperate teat Poecilosclerida — | Tedaniidae Tedania digitata (Schmidt) Nil sponges have chlorophyll levels Microcionidae Clathria striata Whitelegge 193+1.4* consistent with the presence of cts arborea Nil microalgae. Of the 5 species i Ae which we identified as being Phoriospongiidae | (Bowerbank) 10,2 ‘chlorophyll positive’, we ex- R iliid Ceratopsion aurantiaca ? pected C. concentrica and espa’ | (Lendenfeld) n Callyspongia sp. to contain chlorophyll-a concentrations were considerably lower than in C. concentrica and Callyspongia sp. Spirastrella areolata is a massive ridge shaped sponge, which has an olive-yellow colour, whilst C. striatais a fan-shaped, orange-tan coloured sponge. Phoriospongia cf. kirki grows as a massive- ridged shaped form and is character- istically a cream-tan colour throughout. It had the lowest measurable concentration of chlor- ophyll-a and contained small dark-brown nodules along the side of each ridge within the ectosomal outer layer. We are not certain whether the chlorophyll positive result we obtained for this species was due to these dark-brown nodes and further work would be required to identify the distribution of symbiotic algae within this species. Other species T. digitata, C. aurantiaca and H. arborea all returned a ‘chlorophyll negative’ response for the presence of symbiotic algae (Table 1). Both 7. digitata and C. aurantiaca were bright orange in colour, which probably reflects metabolically produced carotenoid pigments within the sponge (Hooper, 1996). These pigments may be photo-protective and have been predominately observed in the Poecilosclerida and Axinellida (Hooper, 1996). Holopsamma arborea, a common white, honey- combed reticulated sponge (Hooper, 1996) on shallow and deep reefs in NSW (Roberts & Davis, 1996) was also chlorophyll negative. Four orders of sponges examined in this pre- liminary study had chlorophyll positive species (Table 1). Wilkinson (1993) also found chloro- phyll positive species in each of 5 orders he examined from tropical reefs (see Table 2). Over 100 species of temperate reef sponges were ex- amined for their potential to contain symbiotic chlorophyll-a, based entirely on the colour of their external tissue. Since Cymbastela spp. were reported as phototrophic in temperate and tropical waters (Cheshire et al., 1995; Wilkinson, 1983; Larkum et al., 1988), it was highly likely that this would be the case for related species in NSW. Wilkinson (1983) found that nine out of ten of the most common sponges on a tropical reef (representatives from five orders and six families) contained symbiotic algae. It has been estimated that up to 50% of tropical sponges may rely on this symbiosis because of the relatively low levels of available nutrients in these waters (Cheshire & Wilkinson, 1991). It is our opinion that a large proportion of shallow water temp- erate reef sponges may also have these associations, given that five of the eight species we screened were 'chlorophyll positive’. Furthermore, in our subjective examination of over 100 species from these temperate waters, at least 65% had similar colour to those that proved to be chlorophyll positive in this preliminary analysis. On shallow water temperate reefs, C. con- centrica was found to have on average 139.4+9.4ug/g of chlorophyll-a within its tissue. Wilkinson (1983) reported a concentration of 93.41g/g of chlorophyll-a in C. concentrica (de- scribed as Pseudaxinyssa sp. in his work) from a tropical reef. Seddon et al. (1993) investigated the ability of C. concentrica to photoacclimate in shallow water, and suggested that factors other than visible light were important in restricting its distribution. A preliminary study by Cheshire et al. (1995) on a Cymbastela sp. from southern Australian waters (possibly C. notiana Hooper & Bergquist), demonstrated that these sponges were capable of maintaining themselves 496 TABLE 2. Sponge orders found to be chlorophyll positive on temperate versus tropical reefs (nt = not tested). Sponge Order (lis ody) (Wilken a 983) Astrophorida nt 1/1 Dictyoceratida nt 5/6 Hadromerida 1/1 nt Halichondrida 1/1 1/1 Haplosclerida 1/1 nt Poecilosclerida 2/5 1/1 Clathrinida nt 1/1 photosynthetically. Given the results we have obtained in this study, we anticipate that C. concentrica and many other temperate reef sponges may have this ability. Whether or not temperate reef sponges rely on symbionts to enhance nutrition is unknown, however the work by Cheshire et al. (1995) in South Australia would suggest that this is the case. Cymbastela concentrica is typically an olive- brown colour (Hooper & Bergquist, 1992) at its surface, but variations in the shade of colour have been observed between shallow and deep waters. This may indicate changes in the concentration of symbiotic algae within the sponge associated with light gradients. Initial experiments with C. concentrica suggest that its colour can ‘lighten’ within days of manipulating its position with respect to incident light. Cymbastela concentrica and a number of other temperate reef sponges have been shown to be adversely affected by the discharge of sewage effluent into shallow and deep-water habitats (Roberts, 1996; Roberts et al., 1998). We speculate that any symbiotic relationship between sponges and algae may be altered through reductions in available light and/or increased nutrients as a result of sewage effluent (Roberts et al., 1998). We believe that many temperate reef sponges contain symbiotic algae however the significance of any symbiosis has to be quantified. We need to identify the types of symbionts within various species and examine these relationships with light gradients. Although temperate reef sponges may contain symbiotic algae, their role in the nutrition of the sponge needs to be quantified. ACKNOWLEDGEMENTS We thank John Hooper (Queensland Museum) for his continued support, assistance and taxonomic input with the sponge fauna from MEMOIRS OF THE QUEENSLAND MUSEUM temperate NSW. Clive Wilkinson (Australian Institute of Marine Science) is thanked for his advice on colour morphs associated with sym- biotic algae. Klaus Koop and Tony Church (Environment Protection Authority, NSW) provided significant management support for the study. We are indebted to Alan Butler (CSIRO Marine Research - Marmion Laboratories), Brian Bayne (Centre for Research on Ecological Impacts of Coastal Cities, University of Sydney) and David Ayre (University of Wollongong) for their critical reviews of the manuscript. This paper represents contribution no. 185 from the Ecology and Genetics Group, University of Wollongong. LITERATURE CITED CHESHIRE, A.C., BUTLER, A.J., WESTPHALEN, G., ROWLAND, B., STEVENSON, J. & WIL- KINSON, C.R. 1995. Preliminary study of the distribution and photophysiology of the temperate phototrophic sponge Cymbastela sp. from South Australia. Marine and Freshwater Research 46: 1211-1216. CHESHIRE, A.C. & WILKINSON, C.R. 1991. Modelling the photosynthetic production by sponges on Davies Reef, Great Barrier Reef. Marine Biology 109: 13-18. CLESCERI, L.S., GREENBERG, A.E. & TRUSSEL, R.R. 1985. Standard methods for the examination of water and wastewater. 16th Edition. (American Public Health Association: Washington DC), DAVIS, A.R., ROBERTS, D.E. & CUMMINS, S.P. 1997. Rapid invasion of a sponge-dominated deep-reef by Caulerpa scalpelliformis (Chlorophyta) in Botany Bay, New South Wales. Australian Journal of Ecology 22: 146-150. HOOPER, J.N..A 1996, Revision of Microcionidae (Porifera: Poecilosclerida: Demospongiae), with description of Australian species. Memoirs of the Queensland Museum 40; 1-626. HOOPER, J.N.A. & BERGQUIST, P.R. 1992, Cymbastela, a new genus of lamellate coral reef sponges. Memoirs of the Queensland Museum 32: 99-137. . HOOPER, J.N.A. & LEVI, C. 1994, Biogeography of Indo-west Pacific sponges: Microcionidae, Raspailiidae, Axinellidae. Pp. 191-212. In Soest, R.W.M. Van, Kempen, T.M.G. van, Braekman, J.C. (eds) Sponges in time and space. (Balkema: Rotterdam). LARKUM, A.W.D., COX, G.C. & DIBBAYAWAN, T.P. 1988. Prokaryotic algal symbionts of coral reef sponges. Proceedings of the 6th International Coral Reef Symposium 3: 163-169. ROBERTS, D.E. 1996. Patterns in subtidal marine assemblages associated with a deep-water sewage outfall. Marine and Freshwater Research 47: 1-9. SYMBIOTIC ALGAE IN SPONGES ROBERTS, D.E. & DAVIS, A.R. 1996. Patterns in sponge (Porifera) assemblages on temperate coastal reefs off Sydney, Australia. Marine and Freshwater Research 47: 897-906. ROBERTS, D.E., SMITH, A., AJANI, P. & DAVIS, A.R. 1998. Rapid changes in encrusting marine assemblages exposed to anthropogenic point- source pollution: a ‘Beyond BACT approach. Marine Ecology Progress Series 163: 213-224. SARA, M. 1971. Ultrastructural aspects of the sym- biosis between two species of the genus Aphanocapsa (Cyanophyceae) and Ircinia variabilis (Demospongiae). Marine Biology 11: 214-221. SEDDON, S., CHESHIRE, A.C. & WILKINSON, C.R. 1993. Photophysiology and acclimation of the coral reef sponge Cymbastela concentrica across a depth gradient. Proceedings of the 7th International Coral Reef Symposium 2: 847-852. 497 VACELET, J. & DONADEY, C. 1977. Electron microscope study of the association between some sponges and bacteria. Journal of Experimental Marine Biology and Ecology 30: 301-314. WILKINSON, C.R. 1978. Microbial associations in sponges. 1. Ecology, physiology and microbial populations on coral reef sponges. Marine Biology 49: 161-167. 1981. Significance of sponges with cyanobacterial symbionts on Davies Reef, Great Barrier Reef. Proceedings of the 4th International Coral Reef Symposium 2: 705-712. 1983. Net primary productivity in coral reef sponges. Science 219: 410-412. WILKINSON, C.R., CHESHIRE, A.C., KLUMPP, D.W. & McKINNON, A.D. 1988. Nutritional spectrum of animals with photosynthetic symbionts — corals and sponges. Proceedings of the 6th International Coral Reef Symposium 3: 27-30. 498 MEMOIRS OF THE QUEENSLAND MUSEUM NEW COLONIAL VACELETIA-TYPE SPHINCTOZOAN FROM THE PACIFIC. Memoirs of the Queensland Museum 44: 498, 1999:- Three new morphotypes of a Recent colonial sphinctozoan coralline sponge are presented, All types show close relationships to the taxon Vaceletia crypta, a non-colonial form from Indo-Pacific reef caves. The first two types were discovered in shallow water reef caves of Osprey Reef, N Queensland Plateau in the Coral Sea. These sponges are common in these caves. The third type of colonial sphinctozoan was found only at two localities at North Astrolabe Reef and Great Astrolabe Reef in Fiji. This variety shows similarities with a previously described deep water variation of Vaceletia from New Caledonia. The first two morphotypes of colonial Vaceletia from Osprey Reef show more similarities to the cryptic, non-colonial form V. crypta from reef caves of the Great Barrier Reef and reefs of the Indo-Pacific, than to the deep-water colonial species described by Vacelet (1988) and Vacelet et al. (1992) from New Caledonia. The third variation from Astrolabe Reef is more similar to this deep water variation from New Caledonia. All three variations will be described elsewhere in detail as multidisciplinary taxonomic and geochemical investigations of these taxa are still in progress (Reitner & Worheide, 1995; Wórheide & Reitner, 1996). The discovery of these three new colonial variations from shallow water reef caves of the SW Pacific clearly demonstrates that colonial forms of Recent Vaceletia are not restricted to deep waters, as previously thought. Sphinctozoan sponges were primary reef building organisms during the Permo-Triassic. They are chambered calcified sponges with morphological similarities to Cambrian Archaeocyaths. The Vaceletia-type of coralline sponges occured first in the middle/late Triassic (Reitner. 1992). Sphinctozoans were considered to be rare since the end of the Triassic, and were thought to be extinct since the end of the Cretaceous; that is until the ‘living fossil’ Vaceletia was discovered by Vacelet (1979) in the Indian Ocean. The solitary, non-colonial form Vaceletia crypta has no reef building potential and is found only sparsely dispersed in the darker areas of Indo-Pacific reef caves. These recently discovered colonial variations of Vaceletia from shallow water reef caves retain a colonial growth mode and a reef building capability. They provide, therefore, clues to understand the modalities of skeletal construction and biocalcification, as well as the ecology of Permo-Triassic sphinctozoan sponges. O Porifera, coralline sponges, mud-mounds, Vaceletia, colonial reef-building sphinctozoans, Osprey Reef, Coral Sea. Literature cited. REITNER, J 1992. ‘Coralline Spongien’ - Der Versuch einer phylogenetisch-taxonomischen Analyse. Berliner geowissenschaftlichen Abhandlungen (E) 1: 1-352. | REITNER, J. & WORHEIDE, G. 1995, New Recent sphinctozoan coralline sponge from the Osprey Reef (N’ Queensland Plateau, Australia). Fossil Cnidaria & Porifera 24(2, B): 70-72. VACELET, J. 1979. Description et affinités d'un éponge Sphinctozoaire actuelle. Colloques internationale du CNRS 291: 483-493, 1988. Colonial growth in a Recent Sphinctozoa (Porifera). Berliner geowissenschaftlichen Abhandlungen (A) 100: 47. VACELET, J., CUIF, J.-P., GAUTRET, P., MASSOT, M., RICHER DE FORGES, B. & ZIBROWIUS, H. 1992. Un Spongiaire Sphinctozoaire colonial apparenté aux constructeurs de récifs triasiques survivant dans le bathyal de Nouvelle-Calédonie. Comptes Rendus de l'Academie des Sciences, Paris (Biologie Marine, Paléontologie) 314(3): . 379-385. WORHEIDE, G. & REITNER, J. 1996. ‘Living fossil’ spinctozoan coralline sponge colonies in the shallow water caves of the Osprey Reef (Coral See) and the Astrolabe Reefs (Fiji Islands). Pp. 145-148. In Reitner, J., Neuweiler, F., & Gunkel, F. (eds) Globale und regionale Steuerungsfaktoren biogener Sedimentation. . Göttinger Arbeiten zur Geologie und Paläontologie SB2 (University of Göttingen: Germany). Joachim Reitner (email: jreitne@gwdg.de) & Gert Würheide*, Institut und Museum für Geologie und Paldontologie, Universitat Göttingen, Goldschmidt-Strasse 3, D-37077 Göttingen, Germany; John N.A. Hooper, *Queensland Museum, P.O. Box 3300, South Brisbane, Qld. 4101, Australia; 1 June 1998. NEW HEXACTINELLID SPONGES FROM THE MENDOCINO RIDGE, NORTHERN CALIFORNIA, USA HENRY M. REISWIG Reiswig, H.M. 1999 06 30: New hexactinellid sponges from the Mendocino Ridge, Northern California, USA, Memoirs of the Queensland Museum 44: 499-508. Brisbane. ISSN 0079-8835. Two hexactinellid sponges collected by ROV from the Mendocino Ridge by Dr. Andrew G. Carey Jr, are representatives of new taxa of Hexactinellida. The first, Poliopogon mendocino sp. nov. (Amphidiscophora, Pheronematidae) is an orange, sheet-like fragment from a 60cm wide, flaring, funnel-shaped sponge sampled at 2332m depth. The lower part of the specimen, and thus the basal spicules, were not collected and are unavailable. A new generic diagnosis for Poliopogon is provided. The second sponge, Nubicaulus careyi gen. nov., sp. noy. (Hexasterophora, Euplectellidae, Corbitellinae) is a nearly complete specimen in the form ofa soft white cup on long hollow stalk encountered at 2102m depth. It bears distinctive drepanocomes, spirodiscohexasters and aspidoplumicomes, a combination previously unknown among hexactinellids. A reformed diagnosis of Trachycaulus is provided from re-inspection of the type specimen. O Porifera, Hexactinellida, new species, new genus, Mendocino Ridge, Poliopogon, Nubicaulus, Trachycaulus, California. Henry M. Reiswig (email: cxhr(a)ymusica.mcgill.ca), Redpath Museum & Biology Dept., McGill University, 859 Sherbrooke St. West, Montreal, Quebec, H3A 2K6, Canada; 6 January 1999. Hexactinellid sponges from northern California (San Francisco to the Oregon border) are known from only three publications. Schulze (1899) recorded Rhabdocalyptus dawsoni trom Albatross stn 3349 (1890) W of Point Arena, 437m depth. Talmadge (1973) reported three species, Aphrocallistes vastus, Chonelasma tenerum (now Heterochone tenera) and Bathyxiphus subtilis, along with numerous other partially identified specimens obtained by the dragboat fishing fleet out of Humboldt Bay, N. California; specific collection sites were not included. The specimens were reportedly housed at Stanford University, but repeated attempts by the author to locate them have failed. Since the specimens’ identities cannot be authenticated, and the original identifier is unknown, Talmadge’s report cannot be considered adequate for acceptance as documented species records. Carey et al. (1990) listed hexactinellids collected on Gorda Ridge, at the California/Oregon border. Among the 10 forms reported, only 3 of them — Staurocalyptus fasciculata, Farrea aculeata and Aphrocallistes vastus — were identified to species by a recognised authority. The seven other incompletely resolved forms are still under review; their number is indicative of considerable diversity remaining to be documented in the poorly known hexactinellid fauna of this region. All of the above deal with new occurrences of taxa originally known from other locations. The present report is the first description of previously unknown taxa of hexactinellid sponges from the N. California region. MATERIALS AND METHODS Large sponges were encountered during operations of the Remotely Operated Vehicle (ROV) Advanced Tethered Vehicle (ATV) of the US Navy Deep Submergence Group on a NOAA Undersea Research Program on the Mendocino Fault Ridge off N. California (Fig. 1). This fauna was recorded on video tape and specimens collected by A.G. Carey Jr. in the course ofa local faunal survey. After manipulator collection and recovery to shipboard, the specimens were air dried. The two specimens that form the basis of this present report were deposited to the invertebrate zoology collections of the California Academy of Sciences, San Francisco (CASIZ). Comparative material from The Natural History Museum, London (BMNH) was also reviewed. Sections of the sponge body wall as well as fragments of dermal and gastral surfaces were either whole-mounted in balsam for light microscopy or digested in hot nitric acid. Large spicules in the resulting spicule suspensions were rinsed, spread on microscope slides and mounted 500 129° Juan de Fuca FIG. 1. Collection sites of Poliopogon mendocino sp. nov. (P) and Nubicaulus caryi sp. nov. (N) and major physiographic features of the northern California / Oregon region. in balsam. Smaller spicules were dispersed on 25mm diam., 0.2mm pore-size, nitrocellulose filters by filtration; the filters were rinsed, dried and mounted in balsam. Spicules were measured by computer via a microscope-coupled digitiser. Data are reported as: mean + st. dev. (range, number of measurements). Spicule drawings were prepared from video-captured microscope images imported into a computer drawing program and traced on-screen. Samples for scanning electron microscopy (SEM) were nitric-acid-cleaned and either filtered onto 13mm diam., 0.2mm pore-size, membrane filters or deposited directly onto cover-glasses after rinsing in distilled water. Following gold-palladium coating, specimens were viewed and photographed with a JEOL JSM-840 SEM. SYSTEMATICS Subphylum Symplasma Reiswig & Mackie, 1983 Class Hexactinellida Schmidt, 1870 Subclass Amphidiscophora Schulze, 1886 Order Amphidiscosa Schrammen, 1924 Family Pheronematidae Gray, 1872 Poliopogon Thomson, 1873 TYPE SPECIES. Poliopogon amadou Thomson, 1873:29, Fig. 1. DIAGNOSIS (summarised from Schulze, 1893:166 and Ijima, 1927:9; emended here). Body lamelliform, either ear-shaped involute MEMOIRS OF THE QUEENSLAND MUSEUM plate, disc attached on edge, or widely open funnel attached centrally. Attached to hard substrate by short, broad, brush-like pad of thin bidentate basalia with shafts, extending into the body, entirely smooth; columella lacking. Conspicuous marginal fringe composed of sceptres and uncinates as marginal prostalia. Lateral surfaces smooth; lateral prostalia absent. REMARKS. Poliopogon, established by Thomson (1873) for the type species, P. amadou, was augmented by Schulze's (1886) addition of P. gigas. It was briefly synonymised with Pheronema from 1894 through 1902 due to Schulze’s (1894) discovery of two Pheronema lacking lateral prostalia, thus removing the only sustaining difference between the genera. After discovery of the missing lateral prostalia in other specimens of those species, Schulze (1902) re-established Poliopogon with its earlier complement of two species. Ijima (1927) suggested removal of P. gigas from the genus because of its different body shape (a barrel), an action taken up by Tabachnick (1990). The genus presently includes only the type species, P amadou, P. maitai (Tabachnick 1988), and the new species, P. mendocino, described below. The form erected as P. amadou pacifica by Tabachnick (1988) cannot be accepted as a member of the genus due to poor condition of the specimen and/or incomplete description, e.g., lack of sceptres. On the basis of available information, even its family placement cannot be confirmed. It is relegated to Amphidiscosa incertae sedis until more details are revealed. Proper rhabdodiactin megascleres (excluding uncinates) are completely absent in this genus as in all Pheronematidae, a point needing reinforcement. Poliopogon mendocino sp. nov. (Figs 2A, 3A-B, 4A-J, 5) MATERIAL. HOLOTYPE: CASIZ 113631: Mendocino Ridge, 300km W. of Cape Mendocino, N. California, 40?21.6"N, 129°23.7'W, 18.ix.1995, 2,332m depth, coll. A.G. Carey Jr., US Navy Deep Submergence Advanced Tethered Vehicle from R/V ‘Laney Chouest’, stn, MRF-1, dive no. 95-52-153 (Fig. 1). ETYMOLOGY. Named after the type locality, Mendocino Ridge. NEW HEXACTINELLIDS FROM MENDOCINO RIDGE 501 FIG. 2. Gross morphology of the holotypes of A, Poliopogon mendocino sp. nov. and B, Nubicaulus caryi sp. nov. in situ before collection, traced from video clips; the recovered samples are indicated by “R” and the dashed lines; arrow indicates point of attachment. DESCRIPTION. Shape. Single specimen recorded live in situ on video tape before sampling (Fig. 2A) with broad flaring funnel, ca. 60cm diam., 20cm high. Robust marginal fringe to 2cm wide extends continuously around strongly undulated distal edge. Large openings of exhalant channels, about 4mm wide, spaced 2cm apart, clearly evident through transparent gastral cover layer within funnel (upper surface). Attachment point at base of a central cone which carried an inferior lateral opening. Approximately one third of the specimen was torn from one side and recovered for study. Recovered dried sample (Fig. 3A-B) a sheet, 23x29x1.8cm maximum dimensions, with marginal fringe intact on one-half of edge. Gastral surface smooth with exhalant channels clearly evident through gastral cover layer. Irregular, vein-like network of white lines formed by strands of overlapping tangential rays of hypodermal pentactins. Dermal surface more irregular and opaque, with openings of inhalant canals visible through thin transparent dermal cover only in marginal areas. Hypodermal strand system not evident to eye. Both dermal and gastral surfaces lack lateral prostalia. Colour. Gold-colored in situ (video recording). Dry sample pumpkin orange, intensified when wetted. Skeleton. Skeleton composed of completely separate spicules; synapticular fusion does not occur. Both surfaces lined by pinnular pentactins i 7 forming a delicate and fragile quadrate lattice by FIG. 3. Poliopogon mendocino sp. nov. Holotype. A, overlapping of basal rays; square meshes have Dermal surface. B, Gastral surface. 502 A. | B. «lo 5mm MEMOIRS OF THE QUEENSLAND MUSEUM FIG. 4. Poliopogon mendocino sp. nov. Holotype spicules. A, Sceptre with three distal tips magnified. B, Uncinate with magnified central segment. C, Macropentactin. D, Three macrohexactins. E, Pinnular pentactin. F, Spiny mesomonactin. G, Mesohexactins. H, Mesamphidisc. I, Two micramphidiscs. J, Microhexadisc. sides of ca. 100mm. These are supported on tangential rays of large, irregular hypodermal pentactins, with rays overlaping to form subdermal strands on both surfaces, but only macroscopically evident on gastral side. Choanosome supported by principalia which are pentactins and hexactins, both irregular in form and shape of rays. Marginal fringe composed mainly (ca. 95%) of large sceptres, with distal tips mostly broken off, and small component (ca. 5%) of uncinates. Proper rhabdodiactin megascleres (excluding uncinates) absent. Attachment point and associated basal spiculation not included in sample and unavailable for characterisation. Megascleres. Principalia large pentactins and hexactins with thin irregular, curved rays; the same pentactins as hypodermalia and hypogastralia; pentactin (Fig. 4C): tangential ray length 3.1+0.1.3mm (range 1.5-7.3mm; n=50), ray width 37.7+8.4um (range 19.7-55.5um; n=50), proximal ray length 2.5+1.2mm (range 0.5-5.9mm; n=50); hexactin (Fig. 4D): ray length 3.4+1.4mm (range 1.2-7.2mm; n=50), ray width 39.3+7.9um (range 22.8-63.2um; n=50). Sceptres (Fig. 4A), restricted to marginal fringe, have mainly smooth shafts when mature, with varying degrees of proclined spination on distal extremity just proximal to the tip which bears the axial cross (centrum); the proximal tenth usually roughened; smaller (younger) sceptres entirely spined; length 12.2+2.4mm (range 7.6-19.4mm; n=50), width 41.349.9um (range 19.2-58.6um; n=50). Uncinates (Fig. 4B) with very low barbs, NEW HEXACTINELLIDS FROM MENDOCINO RIDGE 3 3.4 3.8 4.2 4.6 5 Ln Amphidisc Length Cum) FIG. 5. Frequency distribution of amphidisc length (natural logarithm) from Poliopogon mendocino sp. nov. Holotype. not projecting from spicule profile, also restricted to marginal fringe; length 6.1+1.3mm (range 3.6-9.0mm; n=50), width 20.4+3.8um (range 8.3-29.] um; n=50). Mesoscleres. Dermal and gastral pinnular pentactins (Fig. 4E) similar but differ significantly in all dimensions (P«0.05); pinnulus has cylindrical profile and basals are long, moderately spined throughout, straight, taper uniformly to a sharp tip, and cross at right angles; no evidence of ray curvature or *figure-8" form. Dermalia: pinnulus length 226+25um (range 180-292um; n=50), pinnulus total width 4745.1um (range 36.3-56.3um; n=50), basal ray length 148-11um (range 124-171um; n=50), basal ray width 9.6+1.3um (range 6,8-12.0um; n=50). Gastralia: pinnulus length 204+24um (range 145-238um; n=50), pinnulus total width 42.9+5.7um (range 26.2-53.8u4m; n=50), basal ray length 128+-13 um (range 104-155um; n=50), basal ray width 8.7+1.lpm (range 6.6-12.2um; n=50). Mesohexactins (Fig. 4G) exceedingly abundant throughout entire wall thickness; rays perfectly straight, regularly arrayed, strongly spined and highly variable in size and robustness; ray length 84+20um (range 49-144um; n=50), ray width (excluding spines) 6.3+2.5um (range 3.1-12.2um; n=50). Spiny mesomonactins (Fig. 4F) occur throughout body wall in low numbers, probably representing extreme reduction of mesohexactins; length 152+29um (range 105-243um; n=50), width at head (excluding spines) 12.9+4.2um (range 7.3-25.9um; n=50). 503 Microscleres. Amphidiscs of a single shape occur in large numbers in surface layers and throughout wall; frequency distribution (Fig. 5) shows two distinct size classes. Mesamphidiscs (Fig. 4H) have uniformly spined shafts and umbels with 8 round-tipped tines, slightly in-turned at tips, length 85+15um (range 61-131um; n=166), width 22+3um (range 14-35um; n=100). Micramphidiscs (Fig. 41) similar but umbels carry 11-14 tines, length 39+6um (range 26-60um; n=850), width 12+1.5um (range 9-18um; n=100). Hexadiscs (Fig. 4J) rare; umbels differ from those of micramphidiscs and invariably differ in size on the two ends of each axis; diameter 33+5um (range 26-44um; n=35). REMARKS. This species differs qualitatively from the other two members of the genus, P amadou and P. maitai, in body form, lack of microdiactins and perpendicular junction of pinnule basal rays. It also differs from them in size of largest amphidiscs, and pinnulus ray length. The closer relative of P. mendocino appears to be P. maitai, the form also occurring in the northern Pacific basin, but detailed similarity cannot be assessed on the basis ofthe sparse data so far available on P. maitai. Subclass Hexasterophora Schulze, 1886 Order Lyssacinosa Zittel, 1877 (sensu ljima 1927) Family Euplectellidae Gray, 1867 Subfamily Corbitellinae Ijima, 1902 Nubicaulus gen. nov. TYPE SPECIES. Nubicaulus careyi sp. nov. ETYMOLOGY. Descriptive combination from Greek: nubis = cloud and caulus = stalk; the cloud-stalk or stalked cloud sponge. DIAGNOSIS. Body form a cup on a long, thin, hollow stalk. Principalia are diactins and hexactins. Dermalia and gastralia are pinnulate hexactins. Microscleres are drepanocomes, spirodiscohexasters, and aspidoplumicomes. DISTRIBUTION. Known only from the type locality ofthe type species: Mendocino Ridge off Cape Mendocino, N. California, U.S.A. (Fig. 1). Nubicaulus careyi sp. nov. (Figs 2B, 6A-G, 7A-F) MATERIAL. HOLOTYPE: CASIZ 113632: Mendocino Ridge, 300 km w of Cape Mendocino, northern California, 40722,5"N, 128°08.4’ W, 23.1x.1995, 2,074m depth, coll. A.G. Carey, Jr, US Navy Deep Submergence Advanced 504 MEMOIRS OF THE QUEENSLAND MUSEUM FIG. 6. Nubicaulus caryi sp. nov. Holotype, specimen and spicules (SEM). A, Recovered sample in lateral view. B, View of the distal surface. C, Drepanocome. D, Distal pinnulus tip of a dermal pinnule. E, Spirodiscohexaster. F, Magnified view of central secondary ray bundles showing counterclockwise spiralling. G, Aspidoplumicome showing secondary rays originating in a single marginal whorl. Tethered Vehicle from R/V Laney Chouest, stn. MRF 5-10, dive no, 95-54-155 (Fig. 1). ETYMOLOGY, Named in acknowledgment of the extensive effort and accomplishment made by the collector, Andrew G, Carey, Jr., in his numerous surveys of the NE Pacifie deep-water benthos. DESCRIPTION. Shape. White in situ, nearly spherical goblet, 15cm diam., on a very long, thin stalk, estimated from video records as 60cm long, attached to hard bottom. During collection most of stalk left in place and terminal 2-3cm of main body lost during manipulation (Fig. 2B). Dried E NEW HEXACTINELLIDS FROM MENDOCINO RIDGE 505 recovered specimen has basic form of a calyx (body) with convoluted surface on a long, thin, hollow stalk (Fig. 6A-B). Squat, slightly laterally flattened calyx is 14.7x10.5cm diam., 9.4cm tall. Large superior upper opening, 4.2cm diam., is the upper margin of a cylindrical atrial cavity, the osculum proper and its marginal structures lost during collection. Atrial cavity extends axially 5cm down into body dividing into 4 large, radial, exhalant canals, 1.7cm diam., separated by broad tissue septa, and into stalk lumen aperture located on a central conical prominence on floor of atrium (Fig. 6B). Smaller exhalant canals, 0.4-1.0cm diam, extend radially from lateral atrial walls and four main exhalant canals deep into diverticula of lateral and inferior body wall. Diverticula are manifest on external body surface as softly rounded protrusions, 0.8-2.5cm diam., often joined as ridges which circumscribe deep embayments of outer surface; protrusions and ridges exhibit no regular arrangement. On lower third of body, protrusions lengthen to 4.2cm, and terminate in parietal oscula 0.5-0.8cm diam. Proper body wall only 3.3-4.1mm thick at any point, but due to protrusions and their anastomosis, the convoluted and cavernous wall is effectively 3.5-4.3cm thick. A thin, delicate hydrozoan colony permeates entire wall of body, with unprotected terminal polyps located on all sponge body surfaces: external, atrial and larger exhalant canals. Two recovered stalk pieces total 23.3cm long, ca. 40% of entire stalk in place; tapers from base of body, 1.5cm diam., to lower broken end, 1.1cm diam. Stalk wall thickness is uniform, 1.0mm. All recovered stalk living (sponge tissues present throughout). Colour. White alive (video tape) and dried. Skeleton. Dermal surface of body smooth and covered by a tight quadratic lattice of pinnular hexactins of 200um mesh. Immediately below are openings of inhalant canals, mean diameter 0.6mm (range 0.5-0.9mm). Spirodiscohexasters abundant in, and just below, dermal layer. Gastral surface not covered by a spicule lattice, but consists of open ends of small calibre exhalant canals, mean diam. 1.2mm (range 0.6-2.0mm). Ridges between adjacent canal apertures carry 2-4 ranks of overlapping pinnular hexactins, pinnular rays spaced 150um apart; ridges densely tufted. Ridges supported by conspicuous bundles of diactins coursing sinuously between exhalant canals in plane parallel to, but just below, aperture margins. Parenchyme supported by loose network of macrohexactins and diactins, the latter as single spicules or in small bundles of 2-6 spicules. No discernable layering or orientation of parenchymal megascleres. Stalk skeleton mainly composed of tightly synapticula-joined diactins, oriented randomly with respect to stalk axis; diacts oriented longitudinally only in outermost layer. A few loose pinnules remain on outer surface, but most lost by abrasion during collection and subsequent handling. Some hexactine pinnules fused into outer layer by synapticulae. Internal surface without free gastralia; their absence not attributable to abrasion. Megascleres. All megascleres sparsely and inconspicuously spined throughout, not apparent below 100x magnifications. Principalia are parenchymal diactins and hexactins. Diactins (Fig 7A) smooth (at low magnification), thin, with 4 conspicuous central tubercles; tips rounded or parabolic, not densely microspined; length 2.78+0.66mm (range 1.18-5.06mm; n=100), width 11.5+2.lum (range 7.1-18.2um; n=100). Parenchymal hexactins (Fig. 7C) have thin, often curved rays, ray length 738+239um (range 258-1,34lum; n=100), ray width 11.8+2.0um (range 5.3-16.7um; n=100). Dermalia and gastralia are pinnulate hexactins; most have long, sharp-tipped, spindle-form pinnulus (Figs 6D, 7B, left) but 10% of dermalia have shorter, blunt-tipped, club-shaped pinnulus (Fig. 7B right); dermal and gastral spicules significantly different in dimensions (t-test; P<0.05). Dermalia pinnulus length 496+32um (range 392-595 um; n=100), tangential ray length 203433 um (range 116-297um; n=100), proximal ray length 373+106um (range 106-604um; n=100). Gastralia pinnulus length 478+67um (range 338-730um; n=100), tangential ray length 319+75um (range 193-655um:; n=100), proximal ray length 502+158um (range 122-841um; n=100). Microscleres. Spirodiscohexasters (Figs 6E, 7E) usually spherical in aspect but secondary bundles often restricted in angular splay and central-most secondaries are longer than peripherals, producing noticeable cruciate profile. Secondaries number 15-24 and spiral is sinistral (counterclockwise) (Fig 6F). Terminal discs hemispherical with 15-20 marginal teeth, occurring most commonly in and near dermal and gastral surfaces, but found throughout wall and in stalk; diameter: 147+13um (range 118-176um; n=100). Drepanocomes (Figs 6C, 7D) are large oxyhexasters with recurved secondary rays (hooks); secondaries 4-8, occasionally branched 100um MEMOIRS OF THE QUEENSLAND MUSEUM VIG. 7, Nubicaulus caryi sp. nov., spicules ofthe Holotype. A, Diactins. B, Dermal and gastral pinnular hexactins. C, Macrohexactin. D, Drepanocome, two rays perpendicular to the page omitted. E, Spirodiscohexaster. F., Aspidoplumicome, two rays perpendicular to the page omitted. at terminal bend; uncommon, occuring in stalk and throughout body wall, but not closely associated with surface layers. Body drepanocomes slightly smaller than those of stalk (t-test; P«0.01). Body drepanocome: diameter 265+40um (range 142-329um; n=68); stalk drepanocome: diameter 297+19um (range 234-326um; n=38). Aspidoplumicomes (Figs 6G, 7C) delicate hexasters with shield-like primary terminations nearly contacting adjacent shields. About 60 secondaries emanate from each primary shield in a single marginal whorl but differ in length and angle of curvature, thereby forming a series of 4-5 apparent ‘layers’ when seen in profile (Fig. 7F); distributed throughout body wall and stalk. Diameter 65+6um (range 50-76um; n=72), primary ray length (to distal surface of shield) 11.2+0.9um (range 8.9-14.2um; n=87). REMARKS. Drepanocomes have been reported from three monospecific genera of Corbitellinae, Dictyaulus Schulze, Hertwigia Schmidt, and Trachycaulus Schulze, and from a single Euplectellinae, Holascus belyaevi Koltun, 1970. The new species differs from Dictyaulus in at least 10 significant characters, including (in Dictyaulus): gastral pentactins, floricomes, codonhexasters, non-spiral discohexasters. The NEW HEXACTINELLIDS FROM MENDOCINO RIDGE genus Hertwigia differs in at least four characters including its possession of gastral pentactins, floricomes, oxyhexasters, and codonhexasters, all of which are absent in Nubicaulus careyi. Koltun’s H. belyaevi differs from N. careyi in at least seven characters, including its tubular (Euplectella-like) body form, possession of tetractins as principalia, oxyhexasters, oxyhexactins, and graphiocomes as well as absence of spirodiscohexasters and aspidoplumicomes. Thus the new species, N. careyi, shows little affinity with any of these taxa. The poorly known genus and species, Trachycaulus gurlitti, consists entirely of only two short sections of hollow stem totalling 9cm length, collected by the ‘Challenger’ in the mid-southern Pacific Ocean (Schulze, 1887). Schulze reported its spiculation as including pinnular hexactins, parenchymal hexactins, and drepanocomes, all compatible with the new species. He also reported simple oxyhexasters which were never figured and are here considered to have been a mistaken earlier statement from Schulze’s original description (Schulze, 1886). The holotype of 7. gurlitti (BMNH 1187.10.20.42) was re-examined using filtration methods to recover small pieces of spicules lodged in the stalk framework. It was found to contain, in addition to the spicules reported by Schulze, two classes of codonhexasters. There was no indication of the presence of any other microsclere such as plumicomes, graphiocomes, floricomes, etc. If any of these were present in the remaining stem tissues, their secondary rays or parts of them, would certainly have been detected by this method. The presence of codonhexasters and absence of spirodiscohexasters and aspidoplumicomes in Trachycaulus gurlitti are considered sufficient differences to prevent inclusion of the new species in that genus. This new information on Trachycaulus, which is added to its revised diagnosis (below), pertains to a suggestion by Lévi (1964) that 7. gurlitti be synonymised with Hertwigia falcifera. The combination of dermal pinnular hexactins, drepanocomes and codonhexasters is shared by both of these species, but the same are also shared by Dictyaulus elegans. The absence of any evidence of floricomes in T. gurlitti, as well as its geographic location, argue against Lévi’s suggestion. For the present, it is recommended that both the genus and species represented by 7. gurlitti be maintained as distinct, but still poorly known, entities. 507 Trachycaulus Schulze, 1886 TYPE SPECIES. Trachycaulus gurlittii Schulze, 1886:46. DIAGNOSIS (based on re-inspection of holotype BMNH 1887.10.20.42, and modification of the description summary by Schulze, 1887:373). Corbitellinae with principalia as long diactins and thin oxyhexactins. Known only as a hollow stalk; upper body remains unknown. Diactins arranged in parallel longitudinal series in stalk and united by profuse synapticula. Dermalia are thin pinnular hexactins; gastralia unknown. Microscleres include large drepanocomes with 4 secondary rays per primary, and two classes of small codonhexasters. DISTRIBUTION. Mid south Pacific, 4665m depth. ACKNOWLEDGEMENTS Iam very grateful to Dr. Andrew G. Carey, Jr., for entrusting me with his valuable specimens and for providing access to video recordings and collection data. I also thank Dr. William C. Austin for his collaboration in determination of these new species. I acknowledge once again the critical access to comparative material from the BMNH London, provided by Clare Valentine, without which this work could not have been completed. The field collections were supported by contract with NOAA Undersea Research Program (NURP). Support for specimen analysis was provided by the Natural Sciences and Engineering Research Council of Canada. LITERATURE CITED CAREY, A.G., STEIN, D.L. & RONA, P.L. 1990. Benthos of the Gorda Ridge axial valley (NE Pacific Ocean): taxonomic composition and trends in distribution. Progress in Oceanography 24: 47-57, GRAY, J.E. 1867. Notes on the arrangement of sponges, with the description of some new genera. Proceedings of the Zoological Society of London 1867: 492-558, pls XXVII-XXVIII. 1872. Notes on the classification of the sponges. Annals and Magazine of Natural History (4) 9: 442-461. IJIMA, 1. 1902. Studies on the Hexactinellida. Contribution II. (The genera Corbitella and Heterotella). Journal of the College of Science, Imperial University, Tokyo 17: 1-34. 1927. The Hexactinellida of the Siboga Expedition. Siboga Expedition Reports 6: vii, 1-383, pls 1-26. 308 KOLTUN, V.M. 1970. Sponge fauna of the northwestern Pacific from the shallows to the hadal depths. Communication 1. Pp, 177-233. In Bogorov, V.G. (ed.) “Fauna of the Kurile - Kamehatka Trench and its Environment’. (Akademiya Nauk SSSR. Trudy Instituta Okeanologij: Israel Program for Scientific . Translation, Jerusalem No. 600496). LEVI, C. 1964. Spongiaires des zones bathyale, abyssale et hadale. Galathea Report 7; 63-112, REISWIG, H.M. & MACKIE, G.O. 1983. Studies on hexactinellid spanges IH. The taxonomic status of Hexactinellida within the Porifera, Philosophical Transactions of the Royal Society of London B Biological Sciences 301: 419-428. SCHMIDT, O. 1870. Grundzüge einer Spongien-fauna des Atlantischen Gebietes. (Engelmann: Leipzig). SCHRAMMEN, A. 1924. Die Kieselspongien der oberen Kreide von Nordwestdeutschland, LL, und letzter Teil. Mit Beiträgen zur Stammeeshichte, Monographie zur Geologie und Paläontologie, Berlin (1) 2: 1-159. SCHULZE, F.E. 1886, Uber den Bau und das System der Hexactinelliden, Abhandlungen der Königlichen Akademie der Wissenschafien zu Berlin (Physikalisch-Mathematisch Classe) 1886: 1-97, 1887. Report on the Hexactinellida collected. by H.M.S. ‘Challenger’ during the years 1873-1876. Pp. : 1-513, pls 1-104; In ' Repart on the Scientific Results of the voyage of H,M,S. Challenger during the years 1873-76, Zoology' Volume 21, MEMOIRS OF THE QUEENSLAND MUSEUM 1893. Revision des Systemes der Hyalomenatideri. Sitzungsberichte der Kóniglich Preussischen Akademie der Wissenschaften zu Berlin 1893: 541-589. 1894, Hexactinelliden des Indische Oceans. I. Theil. Die Hyalonematiden. Abhandlungen der Königlichen Akademie der Wissenschaften zu Berlin (Physikalisch-Mathematisch Classe) 1894: 1-60. 1899. Amerikanische Hexactinelliden nach dem materiale der Albarross-Expedition. (G. Fischer: Jena), 1902. An Account of the Indian Triaxonta collected by RA.M.S. ‘Investigator’, 1-113, pls 1-23. (Translated by R.V. Lendenteld: Calcutta). TABACHNICK, KR. 1988, Hexactinellid sponges from the mountains of West Pacific. Pp. 49-64. In Shirshoy, P.P. (ed.) ‘Structural and Functional Researches of the Marine Benthos'. (Academy of Sciences of the USSR: Moscow) [in Russian]. 1990. Hexactinellid sponges from the Nasca and Sala- Y-Gomes. Transactions of the P.P, Shirshov Institute of Oceanology 124: 161-173 [in Russian]. TALMADGE, R.R. 1973. Glass sponges, phylum Poritera, class Hexactinellida, Of Sea & Shore 4(2): 57-58. THOMSON, C.W. 1873, Notes from the ‘Challenger’. Nature (London) 8: 28-30. ZITTEL, K.A. 1877. Studien über fossile Spongien. l Hexactinellidac, Abhandlungen der Königlich Bayerischen Akademie der Wissenschaften (Mathematisch-Physikalisch Classe) 13: 1-63, THE SPONGES OF PARAISO NEARSHORE FRINGE REEF, COZUMEL, MEXICO. Memoirs of the Queensland Museum 44: 508. 1999:- Although sponges as a group are an easily recognisable life form. in ecological studies the identification of individual species of this phylum can be problematic. The objective of this study was to identify and describe the sponge species of the Paraiso nearshore fringing reef off the island of Cozumel, Mexico, A survey of sponges living within an 80m x 40m permanent study site was conducted using underwater video. Sponge tissue samples were also collected, A field guide based on morphological characteristics was compiled describing 42 different sponges, representing 9 orders, 18 families and 21 genera of the class Demospongiae, Comparing the results of this study with earlier descriptions of the diversity of this sponge community indicate the importance of correct sponge identifications for accurate evaluation of changes in reef community structure. The results of this study suggest that regional identification guides are necessary for life forms such as sponges that have a plastic morphology that can be dramatically affected by environmentally induced variables, O Porifera, taxonomy, species list, reefs, biodiversity, Cazumel, Mexico, field guide. J. Ritter (email: jritter(@jbbsredu), Bermuda Biological Station for Research, Ferry Reach GEO1, Bermuda; June 1998, EXPRESSION OF HOMEOBOX-CONTAINING GENES IN FRESHWATER SPONGES EVELYN RICHELLE-MAURER & GISELE VAN DE VYVER Richelle-Maurer, E. & Van de Vyver, G. 1999 06 30: Expression of homeobox-containing genes in freshwater sponges. Memoirs of the Queensland Museum 44: 509-514. Brisbane. ISSN 0079-8835, Homeoboxes have been particularly valuable to identify genes involved in development. This prompted us to look for homeobox-containing genes in sponges, the most primitive metazoans, and to explore the potential role of these genes in sponge development. Using RT-PCR, we have shown that two homeobox-containing genes, EmH-3 and prox! are present in five freshwater sponge species: Ephydatia muelleri, E. fluviatilis, Spongilla lacustris, Eunapius fragilis and Trochospongilla horrida. EmH-3 is expressed differentially during gemmule germination and hatching in E. muelleri as well as in E. fluviatilis. The expression pattern of EmH-3 suggests a role during cell differentiation. Hydroxyurea, which specifically blocks the differentiation of choanocytes and the aquiferous system, seems not to affect the expression pattern of EmH-3. Contrary to EmH-3, prox is expressed almost at the same level throughout development. O Porifera,homeobox-containing genes, development, expression. Evelyn Richelle-Maurer (email: emaurer(a)ulb.ac.be) & Gisele Van de Vyver, Laboratoire de Physiologie Cellulaire et Génétique des levures, CP 244, Université Libre de Bruxelles, Bd du Triomphe, 1050 Brussels, Belgium; 18 January 1999, Homeobox-containing genes are important developmental genes that play a central role in the early development of a variety of organisms. It was thought for a time that they were only involved in spatial and temporal organisation in segmented animals, whereas it is now known that they are also active in non-segmental organisms and systems, and are implicated in axial patterning and cell-fate decisions during differentiation (Davidson, 1995; Dolecki et al., 1986; Garcia et al., 1993; Lawrence & Morata, 1994; Salser & Kenyon, 1996). Homeobox-containing genes have been identified throughout the animal kingdom, from primitive phyla such as cnidarians, nematodes, flatworms and more recently sponges, to chordates (Ruddle et al., 1994). They have been isolated both from freshwater sponges (Coutinho et al., 1994; Richelle et al., 1998; Seimiya et al., 1994; Seimiya et al., 1997), and from marine sponges (Degnan et al., 1995; Kruse et al., 1994). The presence of homeobox-containing genes in Porifera is of particular interest, and of evol- utionary significance, as sponges are considered to be the most primitive metazoans: they do not display any type of symmetry nor polarity, nor do they contain distinct organs or a nervous system. Therefore, elucidating the structure, function and role of homeobox-containing genes in sponges is essential to comprehend the evolution of these genes in metazoans. As previously reported, we have isolated and sequenced three homeobox-containing genes: EfH-I and EfH-2 from Ephydatia fluviatilis using the PCR reaction and degenerated Antennapedia primers (Coutinho et al., 1994), and EmH-3 from Ephydatia muelleri by screen- ing an E. muelleri genomic library with EfH-/ (Richelle et al., 1998). The nucleotide and predicted amino acid seq- uences of EfH-/ and EfH-2 are very different whereas EfH-/ is very similar to EmH-3 (85%- 86%). The comparison of EfH-/ and EmH-3 homeo- domains with all known sponge homeodomains proxl, prox2, prox3 from E. fluviatilis (Seimiya et al., 1994), SHOX from Geodia cydonium (Kruse et al., 1994), SpoxTA1 from Tethya aurantia, and SpoxH1 and SpoxH2 from Haliclona sp. (Degnan et al., 1995) has revealed the highest similarity with prox2 and SpoxTA1 (Tablel). EfH-/ and EmH-3 share a lesser degree of similarity with prox3 and prox/, and are not more closely related to them than to Cnox3, Cnox2 and Cnox! from Hydra (Schummer et al., 1992; Shenk et al., 1993). They exhibit only a low level of similarity (19%) with SHOX homeodomain which seems at present not to belong to homeo- box genes as it does not contain the critical sequence of standard homeodomains (Seimiya et al., 1998). 510 TABLE 1. Levels of similarity between sponge and hydra homeodomains in percent of identical amino acids including conservative substitutions. Gene Efh-1 (37 aa) EmH-3 prox2 92 98 SpoxTA1 (23aa) 96 96 prox3 73 70 SpoxH2 (23aa) 74 70 | prox! 67 70 SpoxH1 (23aa) 52 52 SHOX 19 17 Cnox3 70 68 Cnox2 70 66 Cnox1 65 60 Phylogenetic studies have shown that EmH-3 is closerto metazoan homeodomains than to those of yeast/ fungi and plants (Richelle et al., 1998). EfH-1, EmH3, prox2 and probably SpoxTA/ are representatives of the Hox/1/ Om (1D) class; proxl, prox3 and SpoxH2 representatives of the NK-3, msh and Chox7 class respectively and SpoxH1 may be a representative of the Antp-class (Degnan et al., 1995). Nevertheless, no clustered homeobox genes have yet been reported in sponges. A realignment of msh related genes by Master et al. (1996) has indicated that all four VK-class homeoboxes from D. melanogaster clustered with the sponge homeoboxes prox1, prox2 and prox3 to the exclusion ofall other homeodomain family. The NK family is a large widespread family of non-clustered genes that appears to have been conserved throughout the evolution of animals and may be involved in specifying cell fate rather than specifying regional patterns (Shenk & Steele, 1993). In the present study, we investigate the occurr- ence of EmH-3 and prox! genes in three fresh- water sponge species, common in Belgium: Spongilla lacustris, Eunapius fragilis and Trochospongilla horrida in addition to E. fluviatilis and E muelleri from which they where initially isolated. The expression of these genes was followed during gemmule germination and hatching. The effect of hydroxyurea on the expression of EmH-3 was analyzed. MATERIALS AND METHODS SPONGE CULTURE. sponges were raised in the laboratory from gemmules in Petri dishes filled with sterile mineral medium (Rasmont, 1961) and incubated at 20°C. For some experiments, MEMOIRS OF THE QUEENSLAND MUSEUM they were grown in mineral medium containing hydroxyurea at a final concentration of 100ug/ml (HU-medium). Finally some sponges were cult- ivated in the field. For this purpose, six-day-old sponges, hatched from gemmules on glass plates, were transferred to the outflow of a pond and were allowed to grow for several weeks. RT-PCR EXPERIMENTS. This sensitive method for the detection and estimation of the levels of RNA transcripts was applied to analyse EmH-3 gene expression during development. The expression of two other genes was followed in the same conditions: prox] homeobox- containing gene isolated from E. fluviatilis, known to be expressed at all stages of development for comparison (Seimiya et al., 1994); EmA 1 actin gene isolated from E. muelleri as a control (Ducy, 1993). Total RNA was extracted at different stages of development, from gemmules to functional sponges, using TRIzol reagent as described in the instructions for use (Life Technologies). Before hatching, gemmules were collected and ground in a Potter homogeniser, on ice, in the presence of TRIzol reagent. After hatching, sponges were scraped and mechanically dissociated by pipetting. The gemmules were discarded, the dissociated cells were pelleted by low speed centrifugation (500g, 10mins, 4°C) and resus- pended in TRIzol reagent. For sponges grown in the field, a small piece of the sponge was squeezed in cold mineral medium and the cells dissociated as described for laboratory sponge cultures. The quantity and purity of RNA was estimated by optical absorbance at 260nm and 280nm according to standard procedures (Sambrook et al., 1989). Its quality was checked on an agarose gel. RT-PCR reactions were carried out using the Promega single tube, two-enzyme Access RT-PCR System which provides quick and reproducible analysis of even rare RNAs. All components necessary for RT-PCR were mixed in one tube with 10ng of total RNA and reverse transcription was automaticaly followed by PCR cycling without additional steps according to the manufacturer”s protocol. The conditions were: 1 cycle of 45mins at 48°C; 1 cycle of 2mins at 94°C; 40 amplification cycles: 30sec at 94°C, Imin at 55°C (EmH-3, EmA 1) or 57°C (prox!) and 2mins at 68°C; followed by a final extension cycle of 7mins at 68°C. The amplification products were analysed by agarose gel (1%) electrophoresis of 10% of the total reaction. FRESHWATER SPONGE HOMEOBOX-CONTAINING GENES The primers were supplied by Eurogenetec (Belgium). They were all gene specific and designed to flanked introns in order to discrim- inate between products that had been amplified from RNA and those that had been amplified from DNA. 1) For the study of EmH-3, the upstream primer, 5’-ATGGACAACTGCAGGG GTGA-3’, was complementary to nt 1-20 of the first exon of the genomic sequence and the downstream primer, 5”-CATTCTCCTATTTTG GAACC-3’, was complementary to nt 716-736 of the third exon containing the homeobox. 2) For the study of prox/, primers were those chosen by the authors Seimiya et al. (1994): the upstream primer, 5”-GGACAGATACGCTTCCGATCT- 3”, was complementary to nt 19-39 of the genomic sequence of the first exon and the down- stream primer, 5”-ATATCGTCTGTTCTGAAA CCA -3”, was complementary to nt 347-367 of the second exon. 3) For the study of EmA I: the upstream primer, 5”-AACTGGGACGACATGG AGAA-3”, was complementary to nt 15-35 ofthe published EmA 1 Actin cDNA sequence (Ducy, 1993) and the downstream primer, 5”-GATCCA GACACTGTACTTGC-3”, was complementary to nt 787-807. According to the author, there must be at least one intron between the sequences chosen for the two primers. A | d 440 Efl Em Sl Efr Th M - M 511 The nature of the amplified products was checked by digestion with specific restriction enzymes. The length of transcripts were: about 440bp for EmH-3, 240bp for prox], 390bp for prox2 and 792bp for EmA 1. RESULTS Gemmules hatched after 3-4 days incubation according to the species. Subsequently, choano- cytes and aquiferous system became differentiated and the osculum appeared around 4-5 days of incubation. Seven-day-old sponges were consid- ered to be fully functional. In HU-medium, hatching was postponed by about 2 days and sponges had a typical hollow- dome structure (Rozenfeld & Rasmont, 1976). Neither choanocytes nor an aquiferous system were differentiated. The investigation of EmH-3 and prox/ in S. lacustris, E. fragilis, T. horrida indicated that these genes were expressed in fully functional sponges in the three species as in E. muelleri and E. fluviatilis (Fig. 1). However, as far as EmH-3 expression was concerned, there was a noticeable difference between the length of E. fluviatilis transcripts and those of S. lacustris, E. fragilis and T. horrida (Fig. 1A). The latter were «4 240 M " Efl Em SI Efr Th M FIG.1. Expression of homeobox-containing genes in five freshwater species. Amplified products of RT-PCR of total RNA isolated from 7-day-old sponges. A, expression of EmH-3 gene. B, expression of prox] gene. Abbreviations: Efl=Ephydatia fluviatilis; Em=Ephydatia muelleri, SF-Spongilla lacustris; Efr=Eunapius fragilis; Th=Trochospongilla horrida; - = negative control without RNA template; M=molecular-size marker. Arrows indicate the size in bp of the amplified products for each gene. FIG.2. Expression of homeobox-containing genes in the course of development. Amplified products of RT-PCR of total RNA isolated from gemmules to the formation of fully functional sponges. A,Expression of EmH-3. B, Expression of proxi. C, Control, expression of EmA / (Actin gene from £. muelleri), 1=E. muelleri; 25E. fluviatilis, Developmental stages are expressed as days after incubation at 20°C in mineral medium; Abbreviations: - 7negative control without RNA template; M=molecular-size marker. Arrows indicate the size in bp of the amplified products for each gene. approximately 440bp long, the same size as £. muelleri transcripts and about 50bp longer than £. Jluviatilis transcripts. E. fluviatilis transcripts were 390bp Jong, the expected prox2 transcript size according to Seimiya et al. (1994). We noticed also that prox/ was expressed at a slightly lower level in E. fragilis (Fig, 1B). The study of the temporal expression of EmH-3,prox1,and EmA 1, summarised in Figure ¿ PE i 5 2, reveals a clear-cut. difference in the level and pattern of expression of these genes, though the 2 MEMOIRS OF THE QUEENSLAND MUSEUM absolute amounts of expression cannot been directly compared from one gene to another because of possible differences in amplif- ication efficiency between the different sets of primers. EmH-3 gene was expressed differentially in the course of development in both species (Fig. 2, Al and A2). In gemmules, transcripts were present in very small amounts as they were almost undetectable by RT-PCR. The level of expression increased very slightly until hatching, 3 days and 4 days of incubation, respectively, At that time, a high level of expression was observed. This level was maintained during several days, even in sponges transferred to the field for three weeks (27-day-old sponges). On the other hand, prox/ gene appeared to be expressed at nearly the same level throughout development: transcripts were already discernible in the gemmules and their level varied little although a slight enhance- ment could be detected at the moment of hatching (Fig. 2, Bl and B2), In the control set of experiments, EmA I Actin gene from £. muelleri (Ducy, 1993), was strongly expressed at all stages of development (Fig. 2, C1 and C2). In HU-treated sponges, the evolution of the expression of EmH-3 was roughly the same as in non-treated sponges (Fig. 3). The level of transcripts, very low during the first days of incubation reached already high values one day before hatching (6th day of incubation). Actin expression was high at all stages. DISCUSSION The results of the RT-PCR survey of EmH-3 and prox] in 5 freshwater species corroborate previous Southern hybridisations realised with EfH-1 as a probe (Richelle et al, 1995). They indicate that $. lacustris, E. fragilis and T. FRESHWATER SPONGE HOMEOBOX-CONTAINING GENES 513 FIG, 3. Expression of EmH-3 in hydroxyurea-treated sponges in the course of development. Amplified products of RT-PCR of total RNA isolated from gemmules to the formation of fully functional sponges in Æ. fluviatilis, Developmental stages are expressed as days after incubation at 20°C in mineral medium. Abbreviations: A=Expression of EmA 1 (Actin gene from Æ. muelleri); - template; M=molecular-size marker. Arrows indicate the size in bp =negative control without RNA of the amplified products for each gene. horrida possess an EmH-3-like gene but that this gene differs in structure from the Æ. fluviatilis EfH-1/ prox2 gene. This is clearly evidenced by the difference in length of their transcripts. This difference could be explained by a differential splicing as is the case for E. muelleri the first exon of which is 54bp longer than that of E. fluviatilis (Richelle et al., 1998). The presence of an EmH-3-like gene in 5 species of freshwater sponges together with the high identity of sequence with SpoxTA/ from Tethya aurantia, a marine sponge, may indicate that this type of gene could be widespread among Porifera and could represent one of their ancestral homeobox-containing gene. This hypothesis is supported by the data of Larroux & Degnan (1999), showing that a prox2-like gene is present in two other marine sponge species, /otrocota baculifera and Tedania digitata. On the contrary, prox | although present in the 5 species of freshwater sponges, does not show a high degree of similarity with other sponge homeobox-containing genes isolated to date. The temporal pattern of expression of EmH-3 clearly demonstrates a differential expression of EmH3 gene during gemmule germination and hatching. The enhancement of the expression at the moment of hatching suggests that this gene is particularly involved at that stage of develop- ment and provides evidence for a role in cell-fate decisions during differentiation. Actually, at hatching, all cells began to differentiate from the undifferentiated gemmular archaeocytes in a definite sequence: first the pinacocytes and the sclerocytes, then the choanocytes which arise by repeated divisions undergone 4 7» by the archaeocytes (Rasmont & Rozenfeld, ^ 1981). The persistence of the expression of 439) £"H-3 in the adult sponge is probably related to the continuous replacement and/ or differentiation of cells occurring in the organism, in particular the turnover of the choanocytes (Rozenfeld & Rasmont, 1976). In their work, Seimiya et al. (1994) concluded that prox2 transcripts were identi- fied at all stages of differentiation in Æ. fluviatilis. This discrepancy with our results arises from the fact that the authors studied only one undefined stage before hatching and that obviously, as demonstrated by our results, the main events occur during gemmule hatching. On the other hand, the kinetics of expression of prox] show that this gene is expressed almost at the same level at all stages of development in E. muelleri as in E. fluviatilis. In HU-treated sponges, the overall pattern of expression of EmH-3 is similar to that in untreated sponges. The time-dependent increase in expression of EmH-3 is not delayed in the presence of hydroxyurea, even though hatching is delayed to day 6 rather than day 4. Thus, in contrast to control sponges, the increased expression of EmH-3 in HU-treated sponges appears to precede hatching, since it occurs before the migration of the cells through the micropyle. Consequently, it would be interesting to determine if differentiation processes observed in control sponges at hatching have already started in unhatched HU-treated gemmules. These experiments are of special interest because hydroxyurea inhibit the differentiation of only one type of cells, i.e. the choanocytes, the other cell types being insensitive to its action. In addition, HU-blocked sponges provide a suitable source for the isolation of pure populations of embryonic archaeocytes that can be brought to differentiate and achieve normal development by removal of hydroxyurea from the medium (Rozenfeld & Rasmont, 1976). Indeed, to gain more understanding of the role played by EmH-3 and prox/ in sponge develop- ment, it would be essential to determine what happens when archacocytes differentiate into other cells, in particular into choanoeytes but also to determine in which cells these genes are expressed. LITERATURE CITED Coutinho, CC., Vissers, S. & Van de Vyver., G. 1994. Evidence af homeobox genes in the freshwater sponge Ephydatia fluviarilis. Pp. 385-388. In Soest, R.W.M. van, Kempen, Th. M. G. van & Braekman, J.C, (eds) Sponges in Time and Space. (Balkema; Rotterdam). DAVIDSON, D. 1995. The function and evolution of Ms genes: pointers and paradoxes. Trends in Genetics 11: 405-411, DEGNAN, B.M, DEGNAN, S.M., GLUSTI, A. & MORSE, BLE. 1995. A hox-hom homeobox gene in sponges. Gene 155(2): 175-177, i DOLECKL G.J, WANNAKRAIROJ, S. LUM, R. WANG, G., RILEY, H.D., CARLOS, R., WANG, A, & HUMPHREYS, T. 1986. Stage-specific expression ofa homeo box-containing gene in the non-segmented sea urchin embryo. EMBO Journal 5; 925-930, DUCY, P. 1993, Etude d'isomorphes d'actine chez le sponuiaire Ephvelutia mülleri Lieb. (These de Doctorat, Université Claude Bernard: Lyon - | France). GARCIA-FERNANDEZ, J, BAGUNA, J, & SALO, E. 1993. Genomic organization and expression of Ihe planarian homeobox genes 01/71 and Dih-2, Development 118: 241-253. KRUSE, M., MIKOC, A., CETKOVIC, H., GAMULIN, V, RINKEVITCH, B... MULLER, LM, & MULLER, W.E.G. 1994, Molecular evidence for the presence ofa developmental gene in the lowest animal: identification of a homeobox-like gene in the marine sponge Geodia evdonium. Mechaniims of Ageing and Development 77(1 ): 43-34. LARROUX, C. & DEGNAN, B.M. 1999. Abstract. Homeobox genes expressed in the adult and reaggregating sponge. Memoirs of the Queensland Museum (this volume). LAWRENCE, PA. & MORATA, Ci. 1994. Llomeobox penes: their function in Drosophila segmentation and pattern formation, Cell 78: 181-189. MASTER, V.A,, KOURAKIS, M.J. & MARTIN- DALE, M.Q. 1996, Isolation, characterization, und expression of Le-msx, a maternally expressed member of the msx gene family from the Glossiphoniid Leech, Helohdella, Developmental Dynamics 207: 404-419. RASMONT, R. 1961. Une technique de culture des éponges d'eau douce en milieu controle, Annales aeia Société Royale Zoologique de Belgique 91: 147-156. RASMONT, R. & ROZENFELD, E, 1981. Etude micro-cinématographique de la formation des chambres choanocytaires chez une éponge d'eau MEMOIRS OF THE QUEENSLAND MUSEUM douce. Annales de la Société Royale Zoologique de Belgique 111: 33-44. RICHELLE-MAURER, E., KUCHARCZAK, J, VAN DE VYVER, G. & VISSERS, S. 1996, Southern- blot hybridization, a useful technique in fresh- water sponge taxonomy. Bulletin de l'Institut Royal des Sciences Naturelles de Belgique 66: 227-229. RICHELLE-MAURER, E, VAN DE VYVER, G. VISSERS, S. & COUTINHO, C.C. 1998, Homeo- box-containing genes in freshwater sponges; characterization, expression and. phylogeny. Pp. 157-175. In. Müller, W.E.G. (ed) Progress in Molecular and Subcellular Biology. Vol. 19. (Springer-Verlag: Berlin, Heidelberg. ROZENFELD, F. & RASMONT, R. 1976, Hydroxy- urea: an inhibitor of the difTerentiation of choano- cytes in fresh-water sponges and a possible agent for the isolation of embryonic cells. Differentiation 7: 53-60. RUDPLE, F.H., BARTELS, Li. BENTLEY, K.L., KAPPEN, C., MURTHA, M.T. & PENDLETON, LW. 1994. Evolution of HOX genes. Annual Review of Genetics 28: 423-442, SALSER, S.J. & KENYON, C. 1996, A C, elegans Hox gene switches on, off, on and off again to regulate proliteration, differentiation and morphogenesis. Development 122: 1651-1661. SAMBROOK, J., FRITSCH, E.F., & MANIATIS, T. 1989. Molecular Cloning. A laboratory manual. 2™ edition. (Cold Spring Harbor Laboratory Press: Cold Spring Harbor, NY). SCHUMMER, M., SCHEURLEN, 1., SCHALLER, C. & GALLIOT, B. 1992, HOM/ HOX homeobox genes are present in hydra (Chlorohydra viridissima) and are differentially expressed during regeneration. EMBO Journal 11: 1815-1823. SEIMIYA, M., ISHIGURO, H., MIURA, K., WATANABE, Y. & KUROSAWA, Y. 1994, Homeobox-containing genes in the most primitive metazoa the sponges. European Journal of Biochemistry 221; 219-225, SEIMIYA, M. , NAITO, M. , WATANABE, Y. & KUROSAWA, Y. 1998. Homeobox genes in the freshwater sponge Ephydctia fluviatilis. Pp. 133- 155. In Müller. W.E.G. (ed) Progress in Molecular and Subcellular Biology. Vol. 19 (Springer-Verlag: Berlin, Heidelberg). SEIMIYA, M., WATANABE, Y. & KUROSAWA, Y, 1997. Identification of POU-class homeobox genes in a freshwater sponge and the specific expression of these genes during differentiation. Furopean Journal of Biochemistry 243: 27-31. SHENK, M.A., BODE, H.R, & STEELE, R,E. 1993. Expression of Ciox-2, a HOM/HOX homeobox gene in hydra, is correlated with axial pattern formation. Development 117: 657-667. SHENK, MA. & STEELE, RE. 1993. A molecular snapshot of the metazoan ‘Eve’. Trends in Biochemical Sciences 18: 459-463, ABSTRACTS ORIGIN AND EARLY FOSSIL RECORD OF SPONGES - A GEOBIOLOGICAL APPROACH. Memoirs of the Queensland Museum 44: 515, 1999:- The Porifera are Precambrian active filter feeding metazoans which exhibit a reproductive strategy as known from the Eumetazoa. However, most morphological characters of the sponges differ from those of the Eumetazoa. The defining unique character of Porifera is the possession of aggregates of choanocytes, which demonstrate a phylogenetic relationship with the protozoan taxon Choanoflagellata (Reitner & Mehl, 1996). Sponges have various amounts of symbiotic bacteria (e.g. Reitner, 1993; Schumann-Kindel et al., 1996, 1997)which control metabolic processes. As an hypothesis, sponges originated from biofilms which were associated with choanoflagellates. The first remains of sponges are known from Middle Proterozoic (1.8bya) blackshales (biomarker C30-sterane, 24-isopropylcholestane) (Moldowan et al., 1994; Thiel et al 1999). First spicules and entirely preserved sponge bodies are known from the Late Proterozoic (Ediacaran, various microbialite reefs) (Steiner et al., 1993; Gehling & Rigby, 1996 ). In the early Cambrian all main groups of sponges were known to exist, including the Calcarea (Reitner & Mehl, 1995). During the Phanerozoic six major sponge events are noticed. The first one is represented by the development of the Archaeocyaths in the Lower Cambrian. The occurrence of typical stromatoporoids started in the Ordovician. Rigid hexactinellids are known since the Late Devonian. First modern demosponge taxa occurred after the Late Devonian extinction event. Modern types of coralline sponges, e.g. Ceratoporellida, occured in the Permian, and have an optimum diversity in the Late Triassic. The last significant development is seen in the Jurassic - starting point of the fresh water sponges - when some marine taxa (Haploscerlida: Poecilosclerida: ?Hadromerida) moved into fresh water environments. Sponges were important in reef buildung, and many are still specialised reef dwelling organisms. Their importance as main reef buildung organisms decreased in the Late Jurassic - Lower Cretaceous, when fast growing modern zooxanthellate corals became more important. O Porifera, symbionts, reef-building sponges, fossil sponges. Literature cited. MOLDOWAN, J.M., DAHL, J., JACOBSON, S.R., HUIZINGA, B.J., MCCAFFREY, M.A. & SUMMONS, R.E. 1994. Molecular fossil evidence for late Proterozoic-early Paleozoic Environments, Terra Nova 6: 4, GEHLING, J.G. & RIGBY, J.K. 1996. Long expected sponges from the Neoproterozoic Ediacara fauna of South Australia. Journal of Paleontology 70: 185-195. REITNER, J. 1993. Modern Cryptic Microbialite/Metazoan Facies from Lizard Island (Great Barrier Reef, Australia). Formation and Concepts. Facies 29: 3-40. REITNER, J. & MEHL, D. 1995. Early Palaeozoic Diversification of sponges: New Data and Evidences. Geologie Paläontologie Mitteilungen der Innsbruck 20: 335-347. 1996. Monophyly of the Porifera. Verhandlungen der naturwissenschaftlichen Vereins in Hamburg (NF) 36: 5-32. SCHUMANN-KINDEL, G., BERGBAUER, M. & REITNER. J. 1996. Bacteria associated with Mediterranean Sponges. Pp. 125-128. In Reitner, J., Neuweiler, F. & Gunkel, F. (eds) Globale und regionale Steuerungsfaktoren biogener Sedimentation. Góttinger Arbeiten zur Geologie und Paláontologie SB2 (University of Góttingen: Germany). SCHUMANN-KINDEL, G., BERGBAUER, M., MANZ, W., SZEWZYK, U. & REITNER, J. 1997. Aerobic and anaerobic microorganisms in modern sponges: a possible relationship to fossilisation processes. Pp. 268-272. In Neuweiler, F., Reitner, J. & Monty, Cl. (eds) Biosedimentology of Microbial Buildups IGCP Project No.380, Proceedings of 2™ Meeting, Göttingen, Germany 1996. Facies 36: 268-272. STEINER, M., MEHL, D., REITNER, J. & ERDTMANN, B.-D. 1993. Oldest entirely preserved sponges and other fossils from the Lowermost Cambrian and a new facies reconstruction of the Yangtze platform (China). Berliner geowissenschaftlichen Abhandlungen (E) 9: 293-329. » THIEL, V., JENISCH, A., WORHEIDE, G., LOWENBERG, A. REITNER, J, MICHAELIS, W. 1999, Mid-chain branched alkanoic acids from ‘living fossil’ demosponges: a link to ancient sedimentary lipids. Organic Geochemistry 30: 1-14. Joachim Reitner (email: jreitne@gwdg.de), Institut und Museum für Geologie und Paläontologie, Univ.Góttingen, Goldschmidtstr.3, D-37077 Göttingen, Germany; Gabriela Schumann-Kindel, Okologie der Mikroorganismen, Technische Universität Berlin, Franklinstr.29, D-10587 Berlin, Germany; Volker Thiel, Institute of Biogeochemistry and Marine Chemistry, Univ Hamburg, Bundesstr.55, 20146 Hamburg, Germany; 1 June 1998. 516 MEMOIRS OF THE QUEENSLAND MUSEUM TAPHONOMY AND PRESERVATION POTENTIAL OF SPONGE TISSUE. Memoirs of the Queensland Museum 44: 516. 1999:- The preservation potential of sponge tissues is mainly controlled by sponge related bacteria (Reitner, 1993; Reitner Neuweiler, 1995). [n situ hybridization of the associated microbial populations in few modern demosponges (Petrosia, Chondrosia) showed that the majority of bacteria are members of the gamma-subclass of Proteobacteria (Schumann-Kindel et al., 1996, 1997). Using highly specific oligonucleotide probes for detecting sulfate-reducing bacteria, distinct signals were found scattered in native sponge tissue of both investigated sponges. Also, other fermentative bacteria are involved in the degradation of sponge tissue. Sulfate reducing bacteria may control the calcification of the sponge tissue during degradation, increasing the carbonate alkalinity (Schumann-Kindel et al., 1997; Reitner & Schumann- Kindel, 1997). Therefore, isolated pyrite crystals are common in mineralized (automicritic) sponges tissues. In the surrounding sediment, pyrite is absent or rare. The sponge tissue automicrites are often dark-coloured due to statistically distributed very fine pyrite crystals (ca. lum diameter). Besides the small pyrite, larger erystals often exhibit patchy concentrations or they are arranged in rows. Pyrite formation is probably linked with sulfate reducing symbiotic bacteria in the sponge mesohyle. During early decaying processes of the sponge tissue the internal sponge space becomes entirely anaerobic which favours the growth of the sulfate reducing bacteria. This process may explain the rapid calcification of sponge tissue in modern marine microbialites and ancient sponge mud mounds. In mud mounds siliceous sponges contribute to buildup development with considerable amounts of sponge body-related micrite produced in place. These sponge container automicrites form during the biodegradation of soft tissues, resulting in various 'classical' microbial fabrics. The initial formation of carbonate crystals is controlled by reactive organic compounds (macromolecules) during conditions of elevated carbonate alkalinity (ammonification) (Reitner, 1993; Reitner et al., 1995). The resulting carbonate microfabrics correlate with different soft tissue precursors (mesohyle). The mesohyle structure varies from: bacteria-containing (minipeloidal); bacteria- bearing, rich in choanocyte chambers (peloidal); to bacteria-poor or syncytial structures (aphanites). Intermediate reactive states of organic matter also lead to the in situ preservation of non-rigid demosponges, which are recognized by spicular architecture, spatially restricted occurrences of unsorted spicules, or even by spicule bearing minipeloids, peloids or aphanites. Principally non-spicule bearing sponges should be recognized by the outer (e.g. nodular) shape of microbial fabrics. Organically induced automicrites (organomicrites) are high-Mg calcites with an inorganic signature of 8"C (+3 to +4). The enhanced identification of an autochthonous sponge fauna within mud mounds provides new insight into the nature and origin of these structures. Semi-quantitive data of Cretaceous mounds reveal that 50-8095 of mound micrites were produced in place from which up to 60% of automicrites can be related to metazoans. Therefore, the origin of reactive organic matter is the crucial point to evaluate the pure microbial vs metazoan character of Paleozoic and Mesozoic mud mounds, as well as Precambrian micrites within biostromal and biohermal deposits (Reitner & Arp, 1999). O Porifera, mud-mounds, micrites. Literature cited. REITNER, J. 1993. Modern Cryptic Microbialite/ Metazoan Facies from Lizard Island (Great Barrier Reef, Australia) - Formation and Concepts. Facies 29: 3-40. REITNER, J. & ARP, G. 1999, Evolutionary trends in Processes of Calcareous Organo- and Biomin- eralization. In Fliigel, E. & Freiwald, A. (eds) Evolution of Reefsystems. (Springer) (in press). REITNER, J. & NEUWEILER F. (eds) 1995, Mud Mounds: A Polygenetic Spectrum of Fine-grained Carbonate Buildups. Facies 32: 1-70. REITNER, J. & SCHUMANN-KINDEL, G. 1997. Pyrite in mineralized sponge tissue- Product of sulfate reducing sponge related bacteria ? Pp. 272-276. In Neuweiler, F., Reitner, J. & Monty, Cl. (eds) Biosedimentology of Microbial Build- ups IGCP Project No.380, Proceedings of 2™ Meeting, Gottingen, Germany 1996. Facies 36: 272-276. REITNER, J., GAUTRET, P., MARIN, F. & NEUWEILER, F. 1995, Automicrites in a modern microbialite - Formation model via organic matrices (Lizard Island, Great Barrier Reef, Australia). Bulletin de l'Institut d'oceanographique Monaco 14(2): 237-263. SCHUMANN-KINDEL, G., BERGBAUER, M. & REITNER. J. 1996. Bacteria associated with Mediterranean Sponges. Pp. 125-128. In Reitner, J., Neuweiler, F. & Gunkel, F. (eds) Globale und regionale Steuerungsfaktoren biogener Sedimentation. Góttinger Arbeiten zur Geologie und Paläontologie SB2 (University of Göttingen: Germany). SCHUMANN-KINDEL, G., BERGBAUER, M., MANZ, W., SZEWZYK, U. & REITNER, J. 1997. Aerobic and anaerobic microorganisms in modern sponges: a possible relationship to fossilisation processes. Pp. 268-272. In Neuweiler, F., Reitner, J. & Monty, Cl. (eds) Biosedimentology of Microbial Buildups IGCP Project No.380, Proceedings of 2" Meeting, Góttingen, Germany 1996. Facies 36: 268-272. Joachim Reitner (email: jreitne@gwdg.de) & Fritz Neuweiler, Institut und Museum fiir Geologie und Paläontologie, Univ.Góttingen, Goldschmidtstr.3, D-37077 Göttingen, Germany; Gabriela Schumann-Kindel, Okologie der Mikroorganismen, Technische Universitcit Berlin, Franklinstr29, D-10587 Berlin, Germany; Volker Thiel, Institute of Biogeochemistry and Marine Chemistry, Univ. Hamburg, Bundesstr.55, 20146 Hamburg, Germany; 1 June 1998. MORPHOLOGICAL PHYLOGENETIC CONSIDERATIONS ON THE RELATIONSHIPS OF ISODICTYA BOWERBANK, 1864 TOUFIEK SAMAAI, MARK J. GIBBONS AND MICHELLE KELLY Samaai, T., Gibbons, M.J. & Kelly, M. 1999 06 30: Morphological phylogenetic consider- ations on the relationships of /sodictya Bowerbank, 1865. Memoirs of the Queensland Mu- seum 44: 517-523. Brisbane. ISSN 0079-8835. Isodictya Bowerbank was recently transferred from the Order Poecilosclerida to the Order Haplosclerida, based upon the hypothesis that skeletal architecture and spiculation are homologous between /sodictya and some genera in the haplosclerid family Niphatidae. We examined this hypothesis by determining the diagnostic morphological characters of Isodictya, comparing these with a selection of poecilosclerid and haplosclerid genera, which also have reticulate spongin skeletons. A phylogenetic analysis of diagnostic morphological characters such as spicule morphology, choanosomal skeletal architecture, and surface fibre ornamentation was carried out to determine the ordinal affinities of /sodictya. The analysis of this data set produced 21 equally parsimonious trees of 29 steps with a consistency index (C.I) of 1.000 and retention index (RI) of 1.000. Major characters that separate /sodictya from the haplosclerid genera include the nature of the surface skeletal outgrowths, the amount of spongin associated with these outgrowths, the absence of a paratangential skeleton, the presence of chelae, and the presence of small cigar-shaped oxeas. Results strongly suggest the retention of /sodictya and allied taxa Cercidochela and Esperiopsis within the Order Poecilosclerida. CJ Porifera, Isodictya, Poecilosclerida, Haplosclerida, phylogeny. Toufiek Samaai & Mark J. Gibbons, Zoology Department, University of The Western Cape, Bellville 7535, South Africa; Michelle Kelly* (email: m.kelly@ niwa.cri.nz), Zoology Department, Natural History Museum, Cromwell Road, London SW7 5BD, UK; *Present address: National Institute of Water and Atmospheric Research, Private Bag 109-695, Newmarket, Auckland; 14 April 1999. The order Poecilosclerida is the largest and most taxonomically difficult and unstable of all Demospongiae (Van Soest, 1984; Bergquist $ Fromont, 1988; Hajdu et al., 1994a); at present there is little consensus on family composition and classification. One major debate in poecilosclerid classification has questioned the integrity of Desmacididae Schmidt, 1870 (also incorrectly known as Desmacidonidae Gray, 1872). It is now considered to be polyphyletic (Van Soest, 1984; Hajdu et al., 1994a,b) because the family contains species with monactinal and/ or diactinal megascleres, with myxillid, mycalid, and microcionid skeletal architecture, and in some genera sand replaces the megascleres. Isodictya Bowerbank, 1864 has been tradition- ally placed in this family because it contains diactinal megascleres in a reticulate skeleton and possesses chelae. Recent rearrangement of the Desmacididae by Van Soest (1984), Bergquist & Fromont (1988) and Hajdu et al. (1994a), resulted in the transfer of all desmacidid genera to other poecilosclerid families, leaving /sodictya unassigned. Recently Van Soest (1987) and de Weerdt (1989) postulated that Desmacididae was a sister group of Haplosclerida, with the primary synapomorphy being the presence of small oxeas (100-250um) (de Weerdt, 1989). Hajdu et al. (1994a ,b) subsequently transferred /sodictya to the order Haplosclerida arguing that the skeletal architecture is homologous to that of genera in the haplosclerid family Niphatidae. The presence of chelae in /sodictya, which are absent in all haplosclerids as presently defined, was considered by Hajdu et al. (1994b) to have been secondarily lost in other haplosclerids, /sodictya alone retaining this plesiomorphic character. A comprehensive historical overview of this problem is given in Hadju et al. (1994a) and Bergquist and Fromont (1988). The validity of this transfer rests on the question of whether oxeas and the pattern of their reticulation in /sodictya and the Haplosclerida are truly homologous. Our objectives were firstly to determine the diagnostic morphological characters for /sodictya, and then to consider this genus in the light of other poecilosclerid genera 518 TABLE 1. Material examined and locality data for Isodictya spp., Esperiopsis informis Kirkpatrick, Cercidochela lankesteri Kirkpatrick, Niphates spp., Amphimedon spp., Cribrochalina spp. MEMOIRS OF THE QUEENSLAND MUSEUM were analysed using PAUP Version 3.1.1 (Swofford, 1993). Characters were coded as unordered and multi- state and were unweighted. Wagner Species Registration Number J Locality ^ : Isodictya palmata BMNH 1830.7.3.381 NW. Atlantic parsimony (Kluge & Farris, 1969) Isodictya multiformis BMNH 1997.5.12.18 Ouderkraal, South Africa was used as it «Bv mati evolu- Isodictya sp. BMNH 1895.6.8.140 Indo-Pacific tionary steps by making no Esperiopsis informis BMNH 1997.5.12.30 Ouderkraal, South Africa assumptions about the direction of Cercidochela lankesteri BMNH 1826.10.26.179 Winter Quarters, character changes. Analyses were performed using an exhaustive Antarctica TAL Niphates digitalis BMNH 1928.5.12.202 Bahamas search to find the pannum length S Es : trees. Halichondria moorei ipnateeap. CN 513008 ioronesm (Halichondrida, Halichondriidae) Amphimedon compressa BMNH 1928,5.12.921 St Thomas was defined as the outgroup data El Amphimedon sp. OCDN 4249-C Micronesia obtained from Bergquist (1970). | Cribrochalina vasculum BMNH 1997,3.20.1 Bahamas One hundred bootstrap replicates Cribrochalina sp. OCDN 4159-G L Micronesia (Felsenstein, 1985) were carried out such as Esperiopsis and Cercidochela. Finally, a selection of reticulate spongin skeletons in haplosclerid genera were examined and compared to those in species of Isodictya. MATERIALS AND METHODS Collections were made by the authors using SCUBA unless otherwise stated. Methods of collection, preservation, histological preparation for light microscopy examination, and scanning electron microscopy, were carried out according to Bergquist & Kelly-Borges (1995). Material examined and considered in this study is listed in Table 1. Taxa for comparison with /sodictya and allied genera were selected on the basis of their superficial similarity to haplosclerid genera. Genera in the haplosclerid families Chalinidae, Callyspongiidae, and Oceanapiidae were not considered here as their reticulate skeletons are quite separate from the Niphatidae whose genera have been strongly compared to /sodictya. Abbreviations:BMNH, Natural History Museum, London; 0CDN, Specimen sample numbers for the United States National Cancer Institute shallow-water collection programme contracted to the Coral Reef Research Found- ation, Micronesia. Twelve morphological characters (Table 2) were identified from direct examination of the specimens listed in Table 1. Parallelism may be quite a common feature in sponges (de Weerdt, 1989), at least when simple characters such as consistency, colour and habitats are concerned. We have avoided this by excluding them from the phylogenetic analysis. Each character was scored for each species in a taxon/character data matrix (Table 3). The data to provide confidence estimates on groups contained in the most parsimonious trees. RESULTS Phylogenetic analysis produced 21 equally parsimonious trees of 29 steps with a consistency index (CI) of 1.000, a retention index (RI) of 1.000, and homoplasy index (HI) of 0.000 (Fig. 1A,B). Even under the hypothesis of poly- morphisms for multiple character states, the CI was 1.000. The advantage of using multiple states with the hypothesis of polymorphism is that this procedure permits the detection of hidden homoplasies and reversions in a study at the generic level, which otherwise would be omitted by exclusively affecting isolated species in each genus (Maldonado, 1993). Isodictya spp. form a monophyletic clade in all 2] equally parsimonious trees, with essentially the same arrangement of species in each tree (Fig. 1). Cercidochela lankesteri is grouped with Esperiopsis and /sodictya in 15 of the 21 reconstructions, but in the remaining 6 trees the position of /sodictya is unresolved with respect to Cercidochela and Esperiopsis. The overall position of /sodictya is stable within the poecilo- sclerid clade, and this is clearly separated from haplosclerid genera which also form a distinct, yet internally unresolved clade. Removing Cercidochela from the analysis resolves Isodictya spp. as a single clade that it is more closely related to Esperiopsis informis than to haplosclerid genera. A strict consensus tree yielded a polytomy between the different orders but /sodictya remains clearly separated from Niphatidae (Fig. 2). PHYLOGENY OF /SODICTYA TABLE 2. Characters and character states of /sodictya spp., Esperiopsis informis Kirkpatrick, Cercidochela lankesteri Kirkpatrick, Niphates spp., Amphimedon spp., Cribrochalina spp. CHARACTERS 1-2. GENERAL SKELETAL STRUCTURE: 1. Body compression: a. three-dimensional b. planar; 2| General skeletal organisation: a. plumoreticulate with interstitial isodictyal reticulation, b. square-meshed reticulation; c. confused halichondroid. CHARACTERS 3-7. FIBRE DEVELOPMENT: 3. Primary fibres: a. small square-meshed reticulation, b. large polygona reticulation, c. fine plumose fibres, d. robust plumose fibres, e. Absent; 4. Secondary fibres: a. regular spongin-bound ladder-like fibres, b. irregular fascicular spongin-bound fibres, c. primary fibres bridged by single spicules and tracts, d. primary, fibres bridged by semi-isodictyal reticulation of spicules, e. absent; 5. Mesh shape: a. small square, b. large square, c. large elongate, d. small irregular, e. absent; 6. Ornamentation associated with termination of primary fibre: a. low blunt conule, b. arge spiky conule, c. plumose tufi, d. absent; 7. Spongin development in primary fibres: a. spongin joining spicules; b. spongin| entirely enclosing fibres; c. absent. CHARACTER 8. ECTOSOME: 8, Ectosome (between primary fibres): a. tangential fibres, b. ectosomal brushes, c. tangentia detachable ectosome. CHARACTER 9. MEGASCLERES: 9. Morphology: a. small hastate oxeas, uniformly thick; b. small centrally angulate and thickened fusiform oxeas; c. styles, d. large fusiform oxeas. CHARACTER 10. MICROSCLERES: 10. Chelae: a. palmate isochelae, normal form; b. palmate isochelae, modified; c. canonochelae; d. absent, CHARACTER 11. BIOCHEMISTRY: 11. Manzamine alkaloids: a. present; b. absent. (Magnier & Langlois, 1998) PHARA CTER 12. REPRODUCTION: 12. Viviparity: a. present; b. Absent. Isodictya palmata (Fig. 3A-D) and Isodictya sp. are characterised primarily by a plumo- reticulate skeleton (character 2a) (Fig. 3B) with tufted surface outgrowths (character 6a), the possession of small fusiform oxeas (character 12c) which are often angulate and thickened centrally (Fig. 3C), and palmate isochelae (character 13a) (Fig. 3D). Isodictya multiformis is separated from other species of /sodictya by the nature of the secondary fibres (character 4b->c), the differences in mesh shape and size (character 5a->b) and morphology of the palmate isochelae (character 13a->b). Esperiopsis informis (Fig. 3L-N) and Cercidochela lankesteri (Fig. 3H-K) are joined in a common clade linked by the possession of an irregular anastomosing interstitial network (character 3d) (Fig. 31,M), absent in all other taxa except Isodictya. These two genera form a common clade with /sodictya spp. sharing plumose surface outgrowths (character 6a), absence of special dermal skeleton found only in Niphatidae, and isochelae. Canonochelae (Fig. 3K) are unique to Cercidochela lankesteri TABLE 3. Character state matrix. Characters and states are described in Table 2. For certain characters some taxa may not logically possess a given state, or the authors are unsure of the character state assignment, in which case these character (character 13a->c) while Esperiopsis has palmate isochelae (character 13a) (Fig. 3H). Esperiopsis informis is unique in this analysis as it has styles (character 12a->d). states are coded as ‘unknown’ which is indicated by ‘?’; * = the outgroup. Niphatidae are united by several Charact 1 Species ranee: - synapomorphies. They all have hastate 112 /3/4/5/6/7/8 9/10 11 12/ oxeas(Fig.3G)(character 12b->a), a para- Isodictya palmata blajejejeje a bjb]b ba | tangential ectosome and surface conules Isodictya multiformis blajele[die e bib a ba | surrounded by spongin (character 7a->b). Isodictya sp. blaleclelelela plo alb la The interstitial skeleton of Niphatidae is Esperiopsis infovihis blalelalalclelblelalbla| limited to individual spicules rather than Cercidochela lankestri bPlal|d|ald|e|e|b|a|c|p|a4| an anastomosing network as in the 7 "- Isodictya group. Major characters that Niphates digitalis a|b|b|bj|bibjbjajajd ala ~ = 3 A 4 [S ki eik e 4 separate /sodictya from Niphatidae iphales . > — : LA include the nature of the surface skeletal SRP Ge BB aj blajalala}biajaldia al outgrowths (plumose tufts in /sodictya Amphimedonsp a|bsj|2/2/2 bie 2)d|2|2| and conules in Niphates), the amount of Amphimedon compressa a|bjaj|ajaja|bjaj[ajda|a spongin associated with these outgrowths Halichondria moorei* ?2|ejejejejdjejce[djd|b|[a in Niphatidae, spongin being absent in 520 Isodictyapalmata Isodictyasp. A Isodictyamultiformis Esperiopsisinformis Niphatesdigitalis Niphatesspl Cribrochalinasp2 Amphimedonspl Amphimedoncompressa t— Cereidochelaelankestri Halichondriamoorei Strict Y Isodictyapalmata Isodictyasp. Isodictyamultiformis Esperiopsisinformis Cercidochelaelankestri L—— Niphatesdigitalis L— Niphatesspl Cribrochalinasp? L— — Amphimedonspl — ——— Amplimedoncompressa Halichondriamoorei FIG. 2. Strict consenses tree of the 21 MP trees generated from the morphological data for the 12 sponge species. MEMOIRS OF THE QUEENSLAND MUSEUM Isodictyapalmata Isodictyasp. B Isodictyamultiformis Esperiopsisinformis — Cercidochelaelankestri Niphatesdigitalis 7 Niphatesspl Cribrochalinasp2 p Amphimedonspl Amphimedoncompressa Halichondriamoorei FIG. 1. A-B, Alternative, equally likely hypothetical phylogenies for /sodictya spp. with respect to allied genera Cercidochela and Esperiopsis, and selected haplosclerid genera. Isodictya outgrowths, the presence of a paratangential ectosomal skeleton in Niphatidae, the presence of chelae in /sodictya, and the presence of small curved hastate oxeas of uniform thickness, in Niphatidae. A bootstrap 50% majority rule consensus cladogram (Fig. 4) provides good support for separation of the Isodictya group from niphatid genera, strongly suggesting that they belong to separate orders. DISCUSSION The diagnostic character reversal suggested by Hajdu et al. (1994b) as being synapomorphic (viz. loss of chelae) is inconsistent with the present analysis. It is more parsimonious to regard the appearance of chelae in the Poecilo- sclerida, rather than their loss, as a subsequent achievement in the evolution of this order. Thus, the presence of chelae is a synapomorphic character for the poecilosclerid genera under study. Moreover, although suggested as symplesiomorphic at the species level, palmate PHYLOGENY OF /SODICTYA 52] FIG. 3. Morphological characters of /sodictya, Amphimedon, Cercidochela and Esperiopsis. A-D, Isodictya allia Bowerbank, 1864 (BMNH 1895.6.8.140). A, Holotype. B, Skeletal architecture («31). C, Oxea morphology (500). D, Profile view of palmate isochela (*3,000). E-G, Amphimedon compressa Duch. & Mich., 1864 (BMNH 1928.5.12.921). E, Specimen. F, Skeletal architecture (x31). G, Oxea morphology (* 700). H-K, Cercidochela lankesteri Kirkpatrick, 1906 (BMNH 1826.10.26.179). H, Holotype. I, Skeletal architecture (x31). J, Oxea morphology of (*300). K, Profile view of canonochela (*3,000). L-N, Esperi iopsis informis Stephens, 1915 (BMNH 1997.5,12.30). L, Specimen. M, Skeletal architecture (*31). N, Profile view of palmate isochela (*3,000) Isodictyapalmata Isodictyasp. Isodictyamultiformis Esperiopsisinformis Cercidochelaelankesin Niphatesdigitalis Niphatesspl Cribrochalinasp2 Amphimedonspl Amphimedoncompressa Halichondriamoorei FIG. 4. Bootstrap 50% majority rule tree ofthe 21 MP trees generated from the morphological data for the 12 sponge species. The percentage of these trees that contain each component is shown along each branch. isochelae can be treated as a synapomorphy when discussing generic or familial affinities (Hajdu et al., 1994b). The similarity between the principal morpho- logical characters of /sodictya, Esperiopsis and Cercidochela are striking; the independent development of the skeletal architecture, ocolourea morphology, and chelae morphology is considered to be unlikely. Hajdu et al, (1994b) based their transfer of Isodictya to Niphatidae MEMOIRS OF THE QUEENSLAND MUSEUM primarily on the presence of what they regarded to be an ‘isodictyal’ reticulation, and secondarily on the presence of palmate isochelae in /sodictya. These microscleres are absent in all other Haplosclerida, and are considered to be simply an underlying synapomorphy to both orders by Hadju et al. (1994a). Close examination of the reticulate skeletons of pertinent genera within these two groups shows here that niphatid skeletons are not isodictyal in the strict sense, a term that should be reserved for genera within the haplosclerid family Chalinidae, and that they are very different from the reticulate skeletons of Isodictya, Esperiopsis and Cercidochela. Moreover, additional characters examined here, such as the nature of surface ornamentation, megasclere morphology, interstitial spiculation, nature of the actual fibres and the exclusive presence of manzamine alkaloids (Magnier & Langlois, 1998) in the haplosclerid family Niphatidae and Chalinidae add confidence to the separation of these to groups. Even though palmate isochelae are present in Esperiopsis, it is presently placed in the poecilosclerid family Mycalidae whose genera all contain anisochelae. If other skeletal characters are considered, such as the nature of the primary fibres and surface ornamentation, Isodictya may also fit within this family (see Hooper, 1997). Many of these decisions on affinities may be further corroborated and illuminated with additional taxa and additional tools such as molecular systematics. Until further studies are carried out on the detailed nature of the skeletal morphology, spiculation, and chemistry of Isodictya, this genus should be retained within Poecilosclerida. ACKNOWLEDGMENTS We thank The Foundation for Research and Development, South Africa, and the Royal Society London, for financial support for TS. We are grateful to Clare Valentine, Klaus Borges, and the Photographic Unit at the Natural History Museum, London, for access to specimens, histological assistance, and photography, respectively. Thanks to the Coral Reef Research Foundation, Micronesia, for collection support in Micronesia, and thanks to Zaahir Toeffie, Goosein Isaacs, and Martin Hendricks, University of The Western Cape, for assistance with collections in South Africa.This is a Coral Reef Research Foundation contribution. PHYLOGENY OF /SODICTYA LITERATURE CITED BERGQUIST, P.R. 1970. The marine fauna of New Zealand: Porifera, Demospongiae, Part 2 (Axinellida and Halichondrida). New Zealand Oceanographic Institute Memoir (51): 5-86. BERGQUIST, P.R. & FROMONT, J.P. 1988. The Marine Fauna of New Zealand: Porifera, Demospongiae. Part 4 (Poecilosclerida). New Zealand Oceanographic Institute Memoir (96): 1-197. BERGQUIST, P.R. & KELLY-BORGES, M. 1995. Systematics and Biogeography of the genus lanthella (Demospongiae: Verongida: lanthellidae) in the South-west Pacific. The Beagle, Records of the Northern Territory Museum of Arts and Science 12: 151-176. FELSENSTEIN, J. 1985. Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39: 783-791. HAJDU, E., SOEST, R.W.M. VAN & HOOPER, J.N.A. 1994a. Proposal for a phylogenetic subordinal classification of poecilosclerid sponges (Demospongiae, Porifera). Pp. 123-140. In Soest, R.M.W. van, Kempen, Th.M.G. van & Braekman, J.C. (eds) Sponges in Time and Space. (Balkema: Rotterdam). HAJDU, E., DE WEERDT, W.H. & SOEST, R.W. M. VAN 1994b. Affinities of the * Mermaid's Glove’ sponge (Isodictya palmata) with a discussion of the synapomorphic value of chelae microscleres. Pp. 141-150. In Soest, R.W.M. van, Kempen, 523 Th.M.G. van & Braekman, J.C. (eds) Sponges in Time and Space (Balkema: Rotterdam). HOOPER, J.N.A. 1997. ‘Sponguide’. Guide to sponge collection and identification. (Queensland museum: Brisbane) (http://www.Qmuseum. qld.gov.au). KLUGE, A.G. & FARRIS, J.S. 1969. Quantitative phyletics and the evolution of anurans. Systematic Zoology 18: 1-32. MAGNIER, E. & LANGLOIS, Y. 1998. Manzamine alkaloids: synthesis and synthetic approaches. Tetrahedron 54:6210-6258. MALDONADO, M. 1993. The taxonomic significance of the short-shafted mesotriaene reviewed by parsimony analysis: validation of Pachastrella ovisternata Von Lendenfeld (Demospongiae: Astrophorida). Bijdragen tot de Dierkunde 63(3): 129-148. SOEST, R.W.M. VAN. 1984. Marine sponges from Curacao and other Caribbean localities. Part 3. Poecilosclerida. Studies on the Fauna of Curagao and other Caribbean Islands 199: 1-167. 1987. Phylogenetic exercises with monophyletic groups of sponges. Pp. 227-241. In Vacelet, J. & Boury-Esnault, N. (eds) Taxonomy of the Porifera. (NATO ASI Series, Springer-Verlag: Berlin). SWOFFORD, D.L. 1993. PAUP: Phylogenetic analysis using parsimony Version 3.1. (Illinois Natural History Survey: Champaign, Illinois). WEERDT, W.H. DE 1989. Phylogeny and vicariance biogeography of North Atlantic Chalinidae (Haplosclerida, Demospongiae). Beaufortia 39(3): 55-95. MEMOIRS OF THE QUEENSLAND MUSEUM EFFECTS OF PHOTOSYNTHETIC ACTIVITY IN ENDOSYMBIOTIC ZOOCHLORELLAE ON GEMMULE GERMINATION OF A FRESHWATER SPONGE, RADIOSPONGILLA CEREBELLATA. Memoirs of the Queensland Museum 44: 524. 1999:- Many of freshwater sponges thrive in oligotrophic clear waters in cooperation with photosynthetic endosymbionts, such as zoochlorellae in their mesenchymal cells. They also withstand unfavourable winter season in dormant forms as gemmules. Annandale sponge Radiospongilla cerebellata is a green freshwater demospongiae with zoochlorellae in their archaeocytes and flourishes only in warmer season in southern district of Japan. Gemmules of the Annandale sponge also contain zoochlorellae in thesocytes and germinate only under illumination even if all other conditions are properly provided. Although the light sufficient to induce the germination was very low in intensity and extremely short in illuminating period, photosynthesis seems to be essential for the germination, because a photosynthetic inhibitor, atrazine, strongly inhibited the germination under optimal condition. Since gemmules of the Annandale sponge contain a rich storage of nutrients in the thesocytes, photosynthetic nutrients, produced during the incubating period under very low intensity and short length of illumination, seem to have little effect on the induction of the gemmule germination. We undertook to observe the effects of other factors on gemmule germination, that is, gaseous components such as oxygen evolved and carbon dioxide consumed by photosynthesis. To accomplish gas experiments, we devised a glass slide with a hollow chamber of 3cm? sealed with a glass plate. In gas experiments, gemmules were placed in the chamber with M-medium (previously boiled to eliminate dissolved gaseous components). The chamber was tightly sealed with a glass plate and the desired amount of gas was introduced to the medium as a bubble under the glass plate. Gemmules were illuminated at 3000 Ix through the glass plate by ordinal fluorescent tubes for 10hrs daily and kept at 24°C for 8 days. When gemmules were incubated in degassed medium that was tightly sealed and isolated from the atmospheric gases, no gemmules were germinated. When an air bubble was introduced to the incubating chamber by one tenth or more the volume of the incubation medium 100% germination was achieved. Of the major elements of air, only oxygen induced germination efficiently, and brought full germination at much less quantity than the air (about one hundredth of the incubating medium in volume). On the other hand, nitrogen and carbon dioxide, showed no effects on the gemmule germination under the optimal condition. On the contrary, carbon dioxide showed strong inhibition of gemmule germination in the oxygenated or aerated media. These results show that gemmules of the Annandale sponge are induced to germinate by cytoplasmic oxygen concentration under the favourable condition, but can be substantially suppressed by carbon dioxide. Thus, it is regarded that light applied to the gemmules would initially promote photosynthesis in the symbiotic chlorellae in the thesocytes, which would absorb cytoplasmic carbon dioxide and block the initiation and/or progression of gemmule germination, and evolve oxygen that promotes gemmules to germinate. To confirm this assumption, we have tried to induce germination in darkness with fully oxygenated and carbon dioxide free media. However, no gemmules have germinated in total darkness, in spite of other optimal conditions. The failure of germination in darkness can be understood as follows: when optimal temperature and oxygen were provided to the gemmule cells, (which had been dormant under the cold temperature), they seemed to rouse and begin to respire, and generate carbon dioxide. The carbon dioxide could not be eliminated from the cytoplasm readily, due to the lack of photosynthesis under darkness, so its accumulation in the thesocytes appears to inhibit germination of the gemmules. O Porifera, freshwater sponges, gemmule, germination, symbionts, photosynthesis, O» CO». Y Satoh, Y. Fujimoto, A. Yamamoto & Y. Kamishima (email: ykam@cc.okayama-u.ac.jp), Department of Biology, Faculty of Education, Okayama University, Okayama, 700-8530, Japan; I June 1998. TAXONOMIC EVALUATION OF JASPLAKINOLIDE-CONTAINING SPONGES OF THE FAMILY COPPATIIDAE MIRANDA SANDERS, M. CRISTINA DIAZ AND PHILLIP CREWS Sanders, M., Diaz, M.C, € Crews, P. 1999 06 30: Taxonomic evaluation of jasplakinolide-containing sponges of the family Coppatiidae, Memoirs of the Queensland Museum 44; 525-532. Brisbane. ISSN 0079-8835. Much interest has been generated by the isolation of jasplakinolide (or jaspamide), a novel cyclodepsipeptide originating from a variety of marine sponges. but there have been taxonomic problems in assigning specimens trom the South Pacific and Indo-Pacific, possessing a similar spicule composition of oxeas and cuastersand sharing parallel chemical profiles. From a comparative study of museum types and recent collections of jasplakinolide-containing sponges in the family Coppatiidae, skeletal and external morphology indicate that only one genus (Jaspís Gray) and five species are valid Y splendens (de Laubenlels), J digonoxea (de Laubenfels), J. johnstoni (Schmidt), J. serpeniina Wilson, and a Jaspis sp. probably new to science), with the genera Dorypleres Sollas and Zaplethea de Laubenfels relegated into synonymy. O Porifera; jasplakinolide, cvelodepxipepride, Coppatiidae, Jaspis, Dorypleres, Zaplethea. Miranda Sanders (email; sandersiuchemisiry ucsc.edul, M. Cristina Diaz& Phillip Crews, Department of Chemistry & Biochemistry, University af California, Santa Cruz, CA 95064, USA; 21 December 1998, Jasplakinolide (jaspamide) represents a structurally novel compound which has been the focus of several natural products and synthetic studies (Zabriskie et al., 1986; Crews et al., 1986; Braekman et al., 1987; Grieco et al, 1988; Rao et al, 1993; [maeda et al., 1994). The anticancer activity identified m the NCI 60 cell assay pro- moted this marine secondary metabolite as à potential therapeutic lead. Of particular value is jasplakinolide’s biological profile against PC-3 (human prostate cancer) cells. Although several synthetic derivatives have been examined, jasplakinolide proves to be the most potent agent against cancer models. Mechanism based studies have shown this compound to be a specific actin inhibitor which stabilises microfilaments (Senderowicz et al., 1995). Its potential for drug development may therefore be hindered by its inherent toxicity. However, as a tool for study of the cytoskeleton and the numerous processes mediated by actin in eukaryotes, jasplakinolide provides unique qualities. Three genera have been used to refer to jasplakinolide-containing sponges, all very closely related morphologically and chemically: Jaspis Gray, 1867, Dorypleres Sollas, 1888 and Zaplethea de Laubenfels. 1950. Jaspis, with type species ioa johnstoni Schmidt, 1862, was originally described as having bwo spicule types: fusiform and stellate (Dendy, 1916), but now contains several disparate species justifying tts division into at least two separate genera (Boury Esnault, 1973; Hajdu & Van Soest, 1992). Currently speeies with oxeas, a single category of euasters, and a confused choanosomal arrangement are included in Jaspis (Bergquist, 1968; Wiedenmeyer, 1989). Dorypleres contains species like Jaspis that have more than one category of asters, but the genus has been used by few authors. Topsent (1904) synonomised Dorvpleres with Jaspis, an action subsequently reversed hy de Laubenfels (1954), and reversed again by Bergquist (1968) on the basis that the type species, D. dendyi Sollas, lacked two categories of asters. Zaplethea includes species with oxeas and euasters, in which the oxeas were characteristically bent (twice-bent), but only one species, Z. digonoxea de Luubenfels, 1950, was assigned to it. A report of a jasplakinolide producing sponge from Laing Island (Madang, PNG) refers to Z digonoxea as a synonym of Jaspis Johnstoni Schmidt, 1862 (Braekman, 1987), implicitly merging the two genera. At least four tàxonomie names have been used for thick encrusting jasplakinolide-containing sponges in the South Pacific and Indo-Pacilic with a skeleton of oxeas and euasters: Jaspis sp. (Zabriske, 1986; Crews, 1986), Jaspis johnstoni, Zaplerhea digonoxea (Braekman, 1987), and Dorvpleres splendens (Schmitz and Kelly- Borges, pers. comm.). The present study aims to determine the appropriate generic assignment of 526 these species and their relative conspecificity. Two other species MEMOIRS OF THE QUEENSLAND MUSEUM TABLE 1. List of voucher and type specimens examined (*=specimen known to contain jasplakinolide). (J. serpentina Wilson, 1925, and AR . ; : : : Reference/institution Material Locality an unidentified species referred to TOUR here as Jaspis sp. 2), have similar sp iculation and external USNM 23037 Dorypleres splendens de Caroline Islands, Ponape oórbholo to the Laubenfels 1954 + p : 8y . re USNM 21270 Jaspis serpentina Wilson 1925 Philippines jasplakinolide-containing a STATENS species, and are included here as a USNM 22746 re rer ORG Oahu, Hawaii comparison to these species. MHNG MATERIALS AND METHODS EE SUR x ae cx em E x" A 25 Viva johnstoni Schmidt 2 Sebenico, Adriatic Specimens were collected from LMJG 15258/0 | Vioa johnstoni Schmidt 1862 Sebenico, Adriatic various localities in the South | Pacific and Indo-Pacific. AII POR ARE EMI UR P . ME. ^d ronoxed diastra A samples were collected using SME E104 Vacelet & Vasseur 1976 Tuléar, Madagascar SCUBA at approximate depths of SME Tu 120 Jaspis ef. johnstoni Tuléar, Madagascar 10-25m. Sponges were preserved naai . E ù É aspis johnstoni (Z. Laing Island, Papua New in 3.7% formalin for one week, digonoxea) Braekman 1987 Guinea and then transferred and stored in | Uo á : 70% ethanol. Morphological 35-T-93 *Dorypleres splendens Palau characterisation was made from [ese thick sections (Permount embed- . 3 y * : : 91601 * Jaspis sp.1 Walindi, Papua New Guinea ded) and spicule preparations of Bs SH DIE. each specimen. Spicule sizes 91622 datis sp.2 alin i Papua New Guinea represent mean and range values 92102 *Jaspis sp. 1 Pacific Harbour, Fiji (minimum and maximum) of the 92204 *Jaspis sp. 1 Tioman Island, Malaysia spicule length and width for the 97238 *Jaspis sp. 1 Madang, Papua New Guinea oxeas, ray length and width for 92402 * Jaspis sp. 1 Bali, Indonesia oxyasters I and II and total 92405 Jaspis sp. 2 Bali, Indonesia diameter for oxyasters HI. 15 94541 * Jaspis sp. 1 N Sulawesi, Indonesia iW De 8 Ver per 95077 *Jaspis sp. 1 Milne Bay, Papua New Guinea spicu e type. pecia attention 96555 *Jaspis sp. 1 N Sulawesi, Indonesia was given to surface details in Sat Fic Rp UO PPge nS : . is sp. 2 ulawesi, S order to assess whether this was a Spe Sp eee valid taxonomic character for Dorypleres splendens in particular. Specimens examined are listed in Table 1. These were acquired from the United States National Museum (USNM), the Landes-Museum Joanneum of Graz (on loan to the Museum d’Histoire Naturelle Genéve (MHNG)), the Station Marine d’Endoume, Centre d’Ocean- ologie de Marseille (SME), the University of Oklahoma (UO) and the University of California, Santa Cruz (UCSC). RESULTS AND DISCUSSION Of the material examined (Table 1) five species were differentiated on the basis of their external morphology and skeletal structure (Table 2). In all this material there is great similarity in skeletal composition and arrangement (Figs 1-6). All species possess oxeas and euasters, a confused choanosomal skeletal arrangement and paratangential arrangement of small spicules at the surface. The external morphology of Jaspis sp. 1, (UCSC collections) J. johnstoni/Z. digonoxea (Braekman, 1987) and D. splendens (both USNM 23037 and UO 35-T-93) is very similar, consisting of an encrusting growth form (2-4cm thick) with oscules on lobate protrusions (1-2cm high). The texture is dry and crumbly and the sponges easily torn. The dry voucher of J. serpentina (USNM 21270) is similar except that it appears to have grown away from the substrate, anchored by a stalk (Fig. 3A). External morphology of voucher specimens Z. digonoxea (USNM 22746) and Vioa johnstoni (LMJG 15258/0, 15256/0) are quite different from the other species described above: they are much more delicate and thinly encrusting (1-3mm thick). JASPLAKINOLIDE-CONTAINING SPONGES We propose to synonymise all these species under the senior generic name Jaspis Gray. The twice-bent oxeas of Z. digonoxea de Laubenfels, 1950, are interpreted here as a diagnostic character at the species level only, similar to the serpentine rhabds of J. serpentina (Fig. 3B). The proposed assignment ofthe studied species are: J. splendens (de Laubenfels, 1954; Figs 1, 4); J. digonoxea (de Laubenfels, 1950; Fig. 5); J. johnstoni (Schmidt, 1862; Fig. 6); J. serpentina Wilson, 1925, (Fig. 3) and Jaspis sp. 2 (Fig. 2), which is probably new to science. 527 FIG. |. Jaspis sp. 1 (94541) collected in the Sangihe Islands, N Sulawesi, Indonesia. A, underwater photo oflive specimen showing encrusting form with raised oscules; B, external detail ofthe dry voucher showing smooth surface with occasional grooves; C, surface detail of the ethanol voucher; D, cross section showing paratangential arrangement of asters at the ectosome; E, cross section showing confused organisation of oxeas and euasters in the choano- some. All of the jasplakinolide-containing sponges studied here (Table 1) were found to be conspecific. Jaspis splendens (de Laubenfels, 1954) is the senior-most available name for these specimens. The species description requires emendation to de-emphasise the significance of certain surface characteristics given importance by de Laubenfels (1954): conules or protruberances occur as a rather uncommon feature both in the type (Fig. 4B) as well as in more recent collections (Fig. 1B,C). Figure 2 shows an additional Jaspis sp., which although very similar to the others, does not 528 MEMOIRS OF THE QUEENSLAND MUSEUM TABLE 2. Skeletal analysis of specimens of Jaspis mentioned in this paper. Measurements given in micrometres as mean (range) of lengths and widths or diameter. 1 = in some specimens this category represents chiasters; 2 = this sample also has serpentine rhabds, 1625 (1550-1700) x 30 (20-40)um. Specimen Reference # Large oxeas Small oxeas Oxyasters 1 Oxyasters II Oxyasters m! Jaspis sp. 1 Crews 92102 (Gi ERE rapes ite d Ld 1 AA H 16 (13-20) 11(8-15) Jaspis.sp. 1 Crews 94541 |636 E H^ 139 rid Ho 22 e H 21 (15-28) 8 (5-10) Jaspis sp. 1 Crews95077 | 975 ano H [1222 yb HD| 2 i^g H 18 (13-23) 8 (5-10) Jaspis sp. 1 Crews96555 | 37 copie H, 11428 vo haj aCA 12 H 20 (15-30) 10 (8-13) Jaspis sp. 1 Crews 97238 | 9407 x B 113 E E [A in H 17 (13-23) 10 (8-13) dagpisisi 2 Crews 96591 | 777 ERE. ke [1115 pá ial | ae on H 19 (15-28) 11 (8-13) Jaspis sp. 2 Crews 91629 ctt) H 158 ETN HS) 23 ny H 17 (13-20) 10 (8-13) Dres Er M0 ld ay | SOS 13 (8-20) 11 (8-15) pc pud pa A A | expat 18 (13-25) 10 (8-13) ie. | Brockman, 1987 [SUCIA] HOGER TEE. | TI 20 (15-25) 10 (8-13) apis ef, VacslecTta 120 [CANAS ae none 22 (18-30) 11 (8-15) naa dias | Vesti | SEENON) MOOR none 24 (18-30) 1 (8-15) Eso USNM22746 A rN ye none 21 (15-28) 10 (8-15) Vioa jolinstont LMIG15257/0 144 ey He ¿26 pose H none 11 (8-20) noñe Jaspis serpentina USNM212702 1225 Sr Hif ST voy H none 13 n u5 8 (6-10) contain jasplakinolide and differs in external ACKNOWLEDGEMENTS morphology and growth form: it is generally found growing away from the substrate in an encrusting to fan-like form. The topside is very similar to J. splendens shown in Fig. 1, but the reverse side is dominated by regularly spaced holes (Fig. 2B). Based on these differences, we suggest that this species is also probably new to Science. CONCLUSIONS Comparison of type and recently collected material previously assigned to three genera (Dorypleres splendens, Jaspis johnstoni, Zaplethea digonoxea, Jaspis serpentina, Jaspis sp.) are all referred to Jaspis. All Jaspis-like species reported in the literature to contain jasplakinolide were examined and referred to J. splendens (de Laubenfels), and it is probable that jasplakinolide-containing sponges belong to this species. Funding for this project was provided by NIH grants CA 47135 and CA 52955. We would like to thank Rob van Soest for his advice and sug- gestions. We are also grateful to Clare Valentine, Kate Smith, Ruth Desqueroux-Faundez, Jean Vacelet and Fritz Schmitz for providing voucher specimens, skeletal preparations, photographs and literature. LITERATURE CITED BERGQUIST, P.R. 1968. The marine fauna of New Zealand, Porifera, Demospongiae, Part 1: Tetractinomorpha and Lithistida. New Zealand Oceanographic Institute Memoir 37: 1-105. BOURY-ESNAULT, N. 1973. Spongiaires. Resultats Scientifiques des Campagnes de la “Calypso” au large des cótes atlantiques de l' Amérique du Sud (1961-1962) 10(29): 263-295. BRAEKMAN, J. C., DALOZE, D. & MOUSIAUX, B. 1987. Jaspamide from the marine sponge Jaspis johnstoni. Journal of Natural Products 50: 994. JASPLAKINOLIDE-CONTAINING SPONGES FIG. 2. Jaspis sp. 2 (96591) collected at Tifore Island, N Sulawesi, Indonesia. A, underwater photo of live specimen showing encrusting to fan-like morphology; B, external morphology of the dry voucher, topside (left) and underside (right) (scale in cm); C, surface detail of the ethanol voucher; D, cross section showing the concentration of asters at the ectosome and confused organisation of the choanosome. J 2 CENTIMETERS FIG. 3. Jaspis serpentina Wilson 1925 (USNM21270). A, dry voucher showing raised oscules and smooth surface; B, cross section of choanosome, dominated by serpentine rhabds, 529 ae ee CREWS, P., MANES, L.V. & BOEHLER, M. 1986. Jasplakinolide, a cyclodepsipeptide from the marine sponge Jaspis sp. Tetrahedron Letters 27: 2797. DENDY, A. 1916. Report on the Homosclerophora and Astrotetraxonida collected by H.M.S. “Sealark” in the Indian Ocean. In Reports of the Percy Sladen Trust Expedition to the Indian Ocean in 1905. Vol, 6. Transactions of the Linnean Society of London (2, Zoology) 17: 225-271. GRAY, J.E. 1867. Notes on the arrangements of sponges, with the description of some new genera. Proceedings of the Zoological Society, London (1867): 492-588. GRIECO, P.A., HON, Y.S. & PEREZ-MEDRANO, A. 1988. Convergent, enantiospecific total synthesis of the cyclodepsipeptide (+)-jasplakinolide (jaspamide) Jaspis sp. Journal of the American Chemical Society 110: 1630. HAJDU, E. & SOEST, R.W.M. VAN 1992. A revision of Atlantic Asteropus Sollas, 1888 (Demonspongiae), including a description of three new species, and with a review of the family Coppatiidae Topsent, 1898. Bijdragen tot de Dierkunde 62(1): 3-19. IMAEDA, T., HAMADA, Y. & SHIORI, T. 1994, Efficient sytheses of geodiamolide A and jaspamide, cytotoxic and antifungal cyclic depsipeptides of marine sponge origin. Tetrahedron Letters 35: 5918 MEMOIRS OF THE QUEENSLAND MUSEUM FIG. 4. Dorypleres splendens de Laubenfels 1954 (USNM23037). A, dry voucher showing raised oscules; B, surface detail of the ethanol voucher showing shallow grooves; C, cross section of choanosome with confused arrangement of oxeas and euasters. LAUBENFELS, M.W, De 1950. Sponges of Kaneohe Bay, Oahu. Pacific Science 4(1): 3-36. 1954. The sponges of the west-central Pacific. 1954 Oregon State Monographs. Studies in Zoology 7: 1-306. RAO, A.V.R., GURJAR, M.K., NALLAGANCHU, B.R. & BHANDARI, A. 1993. Studies on cyclo- depsipeptides. 2. The total synthesis of jaspamide and geodiamolide D. Tetrahedron Letters 34: 7085. SENDEROWICZ, A.M.J., KAUE, G., SAINZ, E., LAING, C., CREWS, P. & SAUSVILLE, E. 1995. Jasplakinolide’s inhibition of the growth of prostate carncinoma cells in vitro with disruption of the actin cytoskeleton. Journal of the National Cancer Institute 87: 46-51. SCHMIDT, O. 1862. Die Spongien des Adriatischen Meeres. (Engelmann: Leipzig). SOLLAS, W.J. 1888. Report on the Tetractinellida collected by H.M.S. ‘Challenger’. In Report on the Scientific Results of H.M.S. ‘Challenger’ during the years 1873-1876. Zoology. Vol. 25. (Her Majesty’s Stationery Office: London). TOPSENT, E. 1904. Spongiaires des Acores. Résultats des campagnes scientifiques Prince Albert I 25: 1-280. 1898. Introduction à l'étude monographique des spongiaires de France, III. Monoaxonida (Hadromerina). Archives de Zoologie Expérimentale et Générale. (3) 6: 91-113. un Ly —. JASPLAKINOLIDE-CONTAINING SPONGES FIG. 5. Zaplethea digonoxea de Laubenfels 1950 (USNM22746). A, smooth surface of the ethanol voucher; B, cross section of the ectosome; C, cross section of the choanosome. CAMA TIR \ 7 8 9 i | FIG. 6. Vioa johnstoni Schmidt 1862. A, dry voucher encrusting on coralline rock (LMJG15258/0); B, dry voucher encrusting on coralline rock (LMJG15256/0). Scales in cm. VACELET, J., VASSEUR, P. & LEVI, C. 1976. Spongiaires de la pente externe des récifs coralliens de Tulear (sud-ouest de Madagascar). Mémoirs du Museum National d'Histoire Naturelle, Paris (A, Zoologie) 99: 1-116. WIEDENMAYER, F. 1989. Demospongiae (Porifera) from northern Bass Strait, southern Australia. Memoirs of the Museum of Victoria 50(1): 1-242. MEMOIRS OF THE QUEENSLAND MUSEUM WILSON, H.V. 1925. Silicious and horny sponges from Philippine waters. Bulletin of the United States National Museum 100(2, 4): i-vii, 273-506. ZABRISKIE, T.M., KLOCKE, J.A., IRELAND, C.M., MARCUS, A.H., MOLINSKI, T.F., FAULKNER, D.J., XU, C. & CLARDY, J.C. 1986. Jaspamide, a modified peptide from a Jaspis sponge with insecticidal and antifungal activity. Journal of the American Chemical Society 108: 3123. RECOVERY AND GROWTH OF THE GIANT BARREL SPONGE (XESTOSPONGIA MUTA) FOLLOWING PHYSICAL INJURY FROM A VESSEL GROUNDING IN THE FLORIDA KEYS. Memoirs of the Queensland Museum 44: 532. 1999:- On February 2, 1997, the 187m (614 feet) container ship *Contship Houston' ran aground on the Florida reef tract near Maryland Shoal within the Florida Keys National Marine Sanctuary. This incident resulted in significant injury to coral reef resources over an area 650m (2,132 feet) in length. Hundreds of the Giant Barrel Sponge (Xestospongia muta) were damaged or destroyed as the ship approached the final grounding site, along with thousands of scleractinian corals and other reef organisms. A major coral reef restoration project is currently underway to address the physical and biological injury caused by the grounding. Over 3,000 broken and dislodged corals were reattached to the substrate within the inbound tract of the vessel, and large areas of rubble created by the ship's hull have been stabilised through a variety of techniques. The purpose ofthis study was to assess the response of injured Xestospongia to the physical injury caused by the vessel grounding. As the vessel approached the grounding site, sponges which were in the path of the ship were subjected to various degrees of injury. This injury ranged from the minor breaking off of the tops of the sponges to the complete destruction of the sponge except for the basal tissue attached to the substrate. I located and marked 37 injured specimens with individual tags attached to plastic cable ties positioned tightly on the upper injured surface at two locations of each sponge. | monitored the sponges at two to three month intervals and measured upward linear growth from the cable ties. I also observed the condition and vitality of each sponge and the method by which the sponges responded to their injuries. All sponges were photographed at regular intervals. During the course of the study, seven of the tagged sponges disappeared from the study site. Four of these were observed to have died from a wasting disease that was reported from numerous locations in the Florida Keys and the Caribbean. The causes of the disappearance of the other three were not directly observed. The 30 remaining sponges have survived and recovered from the direct physical injury at a minimum by healthy tissue regeneration of the damaged areas. The rate of upward linear growth ranged from zero to 4.48cm over 13 months, with an average upward linear growth of 1.42cm for all sponges. Eight specimens (27%) showed no upward growth over the observation period. The average growth rate for the sponges that did exhibit upward growth was 1.94cm. The most significant period of growth was in the late summer and throughout the fall, which corresponds to the period of warmest seawater temperatures. Upward linear growth was correlated with the degree of injury, with the moderate or slightly injured specimens growing at a faster rate than the badly injured ones. Of the four sponges that died from the wasting disease, three had been categorised as badly injured, which may suggest that injured sponges may be more susceptible to disease than non-injured sponges. O Porifera, Giant Barrel Sponge, Xestospongia muta, sponge growth rates, recovery from physical injury, coral reef injury and restoration, Florida Keys, Florida Keys National Marine Sanctuary. George P. Schmahl (email: george.schmahl(a)noaa.gov), Flower Garden Banks National Marine Sanctuary, 216 W. 26" St., Suite 104, Bryan, Texas 77803, USA; 1 June 1998. AN IMPROVED METHOD OF TISSUE DIGESTION FOR SPICULE MOUNTS IN SPONGE TAXONOMY CHRISTINE H, L. SCHONBERG Schonberg. CILL: 1999 06 30: An improved method of tissue digestion for spicule mounts in sponge taxonomy. Memoirs of the Queensland Museum 44: 333-539, Brisbane, ISSN 0079-8835. Digestion of bioeroding Sponges is difficult as the tissue of many samples cannot be dissolved easily using traditional techniques involving 70% nitric acid. A numberof factors reduce the effectiveness of the acid. such as dilution by water from fresh and alcohol preserved specimens, and buffering effects, particularly contamination with calcium carbonate. In these preparations spicules are often obscured by a while precipitate and tissue/spongin residues. This traditional method was emeñded to produce clean preparations, with the added benefit that digestion time was greatly reduced. Ln contrast to repeated treatments with 70% nitric acid in an 80°C sandbath under the traditional method, I propose the use of alternate applications of 70% aqua regia and 70% nitric acid in à 140°C sandbath with whirl shaking after each application. Drying the tissue samples prior to digestion was important to speed up the process. Nevertheless, under the modified methad some impurities remained resistant to acid digestion, including diatoms, clay and quartz particles. O Porifera. Demospongiae, bloeroding sponges, tissue digestion, taxonomy, spicule preparations. Christine HL. Schánberg* (email: sehoenbereta biologie,uni-oldenbure. de), Australian Institute of Marine Science, PMB 3, Townsville MC, OLD 4810, Australia; *Present address: Carl von Ossietzky Université Oldenburg, FB 7 Biologie, AG Zoasystematik und Morphologie, PF 2503, 26111 Oldenburg, Germany; 18 January 1999. Spicule morphology temains a fundamental criterion in sponge identification, yet their preparation tor examination under light microscopy has changed very little since last century. Traditionally, demosponge spicule preparations have been obtained by digesting sponge tissue and calcareous particles in either sodium hypochlorite (‘bleach’) or heating in nitric acid, leaving the siliceous spicules remaining. Nitric acid digestions have been applied in several ways: 1) a small piece of tissue placed directly onto a microscope slide and digested by dropping small amounts of acid and boiling off the supernatant, with the spicules fixed in place with a mounting medium; 2) tissue digested in a test tube, then spread over slides by burning a drop of the resuspended mixture, and subsequentially fixed onto the slide with a standard mountant. The first method, henceforth referred to as the traditional method, has been widely used over time and is adequate for sponges containing few or no foreign particles. It also has the advantage over the second method in minimising the potential loss of rare spicules or microscleres from slides, given thal preparation occurs directly on the slide medium, whereas using separate platforms for digestion and viewing introduces the possibility that some spicules may be lost during their transfer to glass slides (usually via pipette). However, this traditional method is clearly inadequate for arenaceous species and bioeroding species (i.e. those that bore into calcitic substrata), Under the traditional method calcitic debris is retained on slides, obscures spicules, and often make spicule preparations too thick to be useful. The second method, henceforth referred to as the modified method, was described most recently by Schönberg & Barthel (1997, 1998), and enables clean spicules to be pipetted from the test tube onto the slide, leaving behind the contaminating material. However, recent tests on bioeroding sponges boiled in 70% nitric acid (Schénberg, unpublished data), found both methods were unsatisfactory for this group of sponges. The present study aimed to identify the reasons fot the reduced effectiveness of acid during digestion, and to develop improvements in the modified method. 534 MEMOIRS OF THE QUEENSLAND MUSEUM 1A 100 100 2 90 90 = 89 80 E_ 70 70 2s 60 60 3 2 50 50 2 340 40 93230 L4 30 5 £ 20 old method yx 22 10 ©~ new method Ff 10 = 0 T T Den 0 123 4 5 6 7 8 910 11-15 16-20 21-25 26-30 number of fresh acid treatments B cm n un 2 80 d —%— MYR 80 E 707 *— PAN 70 ER —*— ACH 60 s —9— JBR 7 5.2 °° —*— RIB Oe 40 —8— FAN 40 € g 1] —— PEL 30 253207 —*— MAG 20 ES 10 4 *— ORPH 10 E 0 4 oT a e a DES ER: pe eT 0 123 4 5 6 7 8 910 11-15 16-20 21-25 26-30 number of fresh acid treatments 1C un Q a E AS r= 5.2 OS Er ale! 5 E old method E 9— new method im] ra 123 4 5 6 7 8 910 11-15 16-20 21-25 26-30 number of fresh acid treatments FIG. 1 A-C. Bioeroding sponge tissue digestion speed expressed in number of acid washes necessary to obtain clean spicule samples. A, Digestion speed for samples from all sample sites combined. Black circles - traditional method 2a (N=255); white circles - new method 2b (N=217); B, Digestion speed ordered by sample site (digestion under method 2b only). Black circles - Myrmidon Reef (MYR; N=14), white circles - John Brewer Reef (JBR, N=22), white rhombus - Rib Reef (RIB, N=15), white squares - Pelorus Island (PEL; N=40), black squares - Orpheus Island (ORPH; N=107), subdivided white squares - Fantome Island (FAN; N=19), white triangles - Acheron Island (ACH; N=7), black triangles - Magnetic Island (MAG; N=27), black rhombus - Pandora Reef (PAN; N=8). C, Digestion speed for samples from a single sample site, Little Pioneer Bay, Orpheus Island. Black circles - method 2a (N=200), white circles - method 2b (N=99). Values represent counts, hence no error bars are included. SPICULE PREPARATION METHODOLOGY un tod Un TABLE 1. Relationship of bioeroding sponge tissue digestion speed and were mixed on a whirl shaker distance to the shore through possible uptake of fine terrestrial sediments. with each fresh application of acid (Fig. 1A). % of samples clean afier a giv ; . Sample site Distance to shore | Visibility during ny Eb cede y apra E e 1 Us ing both . methods > (km) dive (m) digestion was considered to be 50% 80% | 100% LE . ae - LA T 3 1 finished when the solution » - - - = = = looked clear (i.e. without any 2 ET 3 : * Ipfa Brewer ieot Je 20 2 3 5 yellow colouration or white Rib Reef 55 - 20 2 2 5 debris remaining in the spicule Fantome Island 20 5-7 2 5 7 sediment). In a few cases, Acheron Island 18 ~10 3 4 4 digestion was stopped after Orpheus Island 16 <5 3 8 >30 more than 20 acid applications Pandora Reef 15 o > 2 4 even though the sample still Pelorus Island 14 5-7 2 5 10 | appeared whitish, and there Magnetic Island 5 25 4 10 | 10.15 | Was no evidence of nitrous gases indicating that the acid was still reacting with organic material. The supernatant acid was removed, and MATERIALS AND METHODS residual acid in the test tubes was allowed to react Bioeroding sponges were collected from the central region of the Great Barrier Reef, Queensland. Material was acid digested using three methods. 1) Traditional method of placing a small fragment of sponge in 70% nitric acid directly on a microscope slide, and boiling under low heat (e.g. alcohol flame or a sand bath). This simple traditional method is described by Hooper (1996). 2a) Modified method of Schónberg & Barthel (1997) whereby pieces of sponge tissue 1-3mm? were digested in test tubes using repeated washes with 70% nitric acid in an 80°C sand bath. Each application of fresh acid was allowed to react over night. This method is similar to that described by Hooper (1996) for digesting sponge tissue for examination under scanning electron microscopy. 2b) Pieces of sponge tissue 1-3mm! were dried at room temperature for 2 days or overnight at 80°C, then pre-digested in test tubes in a 140°C sand bath using aqua regia (1 volumetric unit of 70% nitric acid and 3 units of 70% hydrochloric acid). The high temperature was chosen to maximise acid reactivity by keeping it at boiling level (hydrochloric acid: 110°C; nitric acid: 122°C; Falbe & Regitz, 1992), allowing for lower temperatures inside the test tubes than at the bottom of the sandbath. Spent acid was removed after 24hrs and replaced by 70% nitric acid. The samples were then again left overnight in the 140°C sandbath. This combined acid treatment was repeated as required. Samples off with 70% ethanol added dropwise. The re-settled spicules were washed twice with 100% ethanol to remove remaining acid and to dehydrate the sample. Spicules were stored in 100% ethanol until mounting. The proportion of sample concentrations of spicules to alcohol was adjusted to a ratio of about 1:10 in volume, or less, to ensure optimal spicule concentrations on the microscope slides when applying the same amount of spicule suspension (i.e. 300u1). The suspension was haphazardly spread on a level slide and then burnt off (Schónberg & Barthel, 1998). Before fixing the cover slip with DPX mountant, a flame was held underneath the slide to remove vapour still adhering to the spicules and the microscope slide. The efficiency of both methods was compared by noting the number of acid washes necessary to obtain clear suspensions. In addition, the quality of the preparations was checked by noting whether microscleres or spines on megascleres were still obscured, and if so by what. In 69 sponges, ectosome and choanosome regions were sampled separately to assess whether it was possible that different proportions of spongin and cell material were responsible for producing different digestion results. RESULTS The traditional method of tissue digestion, as well as the modified method of Schónberg & Barthel (1997, 1998), were both found to be inadequate in producing clear, useful slide preparations for bioeroding sponge spicules, irrespective of whether digestion was attempted 536 MEMOIRS OF THE QUEENSLAND MUSEUM directly on a microscope slide (method 1) or in a test tube (method 2a). The traditional method (1) usually produced a thick, white precipitate on the microscope slide, which obscured the spicules, especially microscleres. It was almost impossible to obtain clean preparations, even when carefully washing with alcohol. Moreover, washing increased the risk of losing spicules. The modified digestion method (2a) was extremely slow, requiring many fresh acid washes, and not always dissolving all the tissue. As a consequence, microscleres and spines on megascleres were often obscured, minimising their usefulness of these slides for sponge identification. With up to 20-30 acid washes there was also the risk of losing small and rare spicules during repeated pipetting off the supernatant spent acid. Using method 2b proposed here, the efficiency of sponge tissue digestion was markedly improved. A comparison of methods 2a and 2b is as follows. 1) After two acid applications, one each of aqua regia and 70% nitric acid, spicules were clean in 50% of samples using method 2b. By comparison, using 70% nitric acid in an 80°C sand bath, an average of four acid applications were FIG. 2 A-G. Comparison in quality of bioeroding : Ü sponge spicule preparations and possible required to produce 50% of clean contaminants (scale bars A, B, E-G, 100um; C-D, samples under method 2a. 10um). A, Undissolved tissue obscuring skeletal 2) After four acid applications elements ofan unidentified bioeroding sponge after 21 75% of samples were clean under acid washes under digestion method 2a. B, Clean method 2b; whereas nine preparation of Cliothosa cf. hancocki spicules after 4 applications were required to acid washes under digestion method 2b. C-E, produce the same result under Contaminating particles in SEM spicule preparations method 2n. ARE hime held of Cliona viridis sensu Bergman (1983). C, Diatom. ^ Heatins 95% nPadirinled wele D, Clay particle. E, Quartz particle. F, Permanent Pp o p preparation of C. viridis skeletal elements with clean under method 2b. residual water adhering to the spicules, obscuring j aai : finer details and microscleres (arrow). G, Clear 3) After 10 acid applications using spicule preparation of C, viridis which has been dried method 2b all samples were by heating with a flame prior to mounting (arrow = generally clean (Fig. 2B), whereas clearly visible spirasters). using method 2a samples still SPICULE PREPARATION METHODOLOGY contained debris, obscuring microscleres in particular, which adhered to debris or remained embedded in tissue (Fig. 2A). 4) Under method 2a 50/255 (20%) of samples contained incompletely digested material after more than 10 acid washes, whereas only 5/217 (2%) of samples contained ‘impurities’ under method 2b. Clearly, using the modified method 2b there are far fewer ‘impurities’ in samples than under method 2a. Moreover, any ‘impurities’ that did remain in spicule suspensions were inorganic debris rather than organic tissue remains. This remaining debris was checked under the electron microscope and identified as diatoms, clay and quartz particles (Ross Freeman pers. comm.; Rothwell, 1989; Fig. 2C-E). Therefore, it is possible that part of the cloudiness remaining in samples may be a product of the habitats from which samples were collected, such as differing levels of turbidity and sedimentation rates in different localities. Empirical support for this hypothesis comes from a comparison between samples collected from offshore and mid-shelf sites, which were digested much faster than samples collected from inshore sites. These latter samples were the only ones which could not be cleaned entirely of debris (Fig. 1B). This finding largely correlates with visibility recordings made during field collections, although samples from Pandora Reef and Acheron Island digested surprisingly better than expected (Table 1). The distribution of data comparing the different methods was not random in terms of sample sites. To test for bias in site-effect, data were re-evaluated for one site (Little Pioneer Bay, Orpheus Island), which had both the lowest visibility and highest number of samples. These data provided weaker support for method 2b over 2a, than did analysis of the entire data set, although both data sets followed the same trend (Fig. 1C). For the site-effect data, methods 2a and 2b showed 50% of the samples were clear of debris after 4 and 3 acid washes, respectively, 75% after 10-15 and 6 washes, 90% after 15-20 and 9 washes, and 100% of samples were clear after more than 30 washes of acid. Whereas under method 2a 51/200 samples (26%) were still cloudy after 10 washes, in method 2a only 4/99 samples (4%) were cloudy. Too few data were available to statistically evaluate the effect of collection localities (zones) on coral reefs, but visually the data suggest that acid digestion was slightly better from samples collected from the fore-reef zones than from lagoons or back reefs, possibly related to differential sedimentation rates between the various zones. Potential bias due to seasonal effects were not tested given the patchy collection schedule. Differences in the type of sponge tissue did not appear to effect the speed of digestion except in Aka cf. mucosa, in which tissue samples from the soft choanosome digested on average twice as fast as those taken from of the long, brittle papillae. In general, it was found that tissue samples with minimum calcium carbonate digested better than more calcitic samples, although this is often difficult to achieve for most bioeroding sponges which often incorporate calcitic debris into the choanosome. Nevertheless, some samples that contained more than 50% of their volume with incorporated coral skeleton were sometimes clean within the first two acid washes. No differences were observed in acid digestion rates between the several different (but still unidentified) species of Cliona and Aka sampled, apart from the example of A. cf. mucosa mentioned above. In general, however, samples which contained more soft tissue (such as those eroding large chambers), generally digested slightly better than those in which tissue penetrated carbonate substrates in honeycomb patterns. Finally, in the modified method 2b, repeated alcohol washes were necessary after digestion to avoid residual acid that masked spicule features (such as spines). These washes also reduced the amount of water in samples, as did heating microscope slides before adding the mountant, which can obscure finer details of spicules (Fig. 2F-G). DISCUSSION A general problem in sponge tissue digestion is the high water content of fresh tissue samples, which dilutes the acid. Similarly, the readily available ‘concentrated nitric acid’ is actually 70% concentrated, whereas the more effective ‘fuming nitric acid’ is both extremely expensive and unavailable in certain countries. However, dehydration is easily achieved by fixing samples in ethanol, or drying fresh samples, prior to acid digestion, with the latter method being the most effective. Another explanation for the reduced effect of acid in tissue digestion is the buffering effect of calcium carbonate debris, which is a particular problem in bioeroding sponges. Most samples examined in this study contained a relatively high proportion of calcium carbonates, particularly those that produce small excavations in which the tissue cannot be separated entirely from the host coral skeleton. In species which excavate large chambers, where tissue can be more easily be separated from carbonate, eroded coral fragments (termed “sponge chips” by Cobb (1969) are always present in the tissue. Inorganic contaminants also occur in samples, mostly skeletal remains of diatoms, clay and quartz particles (Fig. 2C-E). Clay minerals may also produce a decreased effectiveness in acid digestion, because they are phyllosilicates, which are negatively charged and thus attract cations resulting in a specific ion-exchange capacity depending on the nature of the mineral (Brownlow, 1979). In acidic solution, these cations can be substituted against hydrogen ions until an equilibrium is reached, which could produce an additional buffering effect (Schréder, 1978). Silicate contaminants produce a cloudy supernatant in acid digestions, superficially appearing to be a problem involving undissolved organic matter and suggesting that digestions should be repeated. This is incorrect, however, and repeated digestions with nitric acid or aqua regia do nothing further to siliceous particles given that these are chemically similar to the demosponge spicules. One solution to this problem is foresight in the design of the sampling program, armed with the knowledge that specimens living close to the shore are likely to contain higher amounts of such debris than those sampled further away (Fig. 1B, Table 1). Tissue of bioeroding sponges was surprisingly resistant to acid digestion. There is evidence that bioeroding sponges are able to shift the calcium carbonate solubility equilibrium with the aid of carbonic anhydrase in favour of substrate dissol- ution (CO; + H;O <= HCO; e 2H' + COy?; Pomponi, 1980). The production of hydronium ions in close vicinity to the etching cells of the sponge would require a special resistance to their corrosiveness. However, the etching process is not yet entirely explained. The improved method for sponge tissue digestion described here is primarily based on increasing the efficiency of acid by reducing its dilution and buffering effects as far as possible. MEMOIRS OF THE QUEENSLAND MUSEUM 1) The greatest improvements were gained by drying tissue samples and increasing the sand bath temperature from 80°C to 140°C. 2) Alternating rinses using aqua regia and nitric acid provided some improvement, although the reason for this is not clear. Due to the development of activated chlorine and nitrose chloride (HNO; +3 HCl <>NOCI+Cl, +2 H20), aqua regia is more corrosive than both nitric acid or hydrochloric acid alone (Falbe & Regitz, 1990). Curiously, aqua regia alone seemed to be less efficient than when used in alternation with nitric acid. The reason for this is also not clear. 3) Repeated stirring during acid digestion, using a whirl shaker, makes some small difference to the efficiency of the process by providing a better saturation of organic shreds with acid, and by physically breaking down larger particles. ACKNOWLEDGEMENTS All field and laboratory work was conducted at the Australian Institute of Marine Science, and I thank the staff for their frequent assistance, particularly scientists in Module 4 for their patience and provision of equipment and space. D. Barthel and O. Tendal first showed me how to digest sponge tissue in test tubes. J. Hooper, S. Cook and J. Kennedy at the Queensland Museum, Brisbane, provided demonstrations of digestion directly on a microscope slide and discussed possible improvements. H. Windsor at the James Cook University, Townsville, assisted with electron microscopy. C. Wilkinson, H. K. Schminke and two anonymous reviewers provided useful comments on the manuscript. Studies of bioeroding sponges were supported by a scholarship from the German Academic Exchange Service. This is AIMS publication no. 931. LITERATURE CITED BERGMAN, K.M. 1983. The distribution and ecol- ogical significance of the boring sponge Cliona viridis on the Great Barrier Reef, Australia. (unpublished MSc Thesis, Mc Master University: Hamilton), BROWNLOW, A.H. 1979. Geochemistry. (Prentice- Hall Inc.: Englewood Cliffs, N.J.). COBB, W.R. 1969. Penetration of calcium carbonate substrates by the boring sponge, Cliona. American Zoologist 9: 783-790, FALBE, J. & REGITZ, M. (eds) 1990. Rómpp Chemie Lexikon 3, H-L. Pp. 1679-2580. 9th improved edition. (Georg Thieme Verlag: Stuttgart & New York). SPICULE PREPARATION METHODOLOGY 1992. Rómpp Chemie Lexikon 5, Pl-S. Pp. 53467- 4428. 9th improved edition. (Georg Thieme Verlag: Stuttgart & New York). HOOPER, J.N.A. 1996. Revision of Microcionidae (Porifera: Poecilosclerida: Demospongiae), with description of Australian species. Memoirs of the Queensland Museum 40: 1-626. POMPONI, S.A. 1980. Cytological mechanisms of calcium carbonate excavation by boring sponges. International Review of Cytology 65: 301-319. ROTHWELL, R.G. 1989. Minerals and mineraloids in marine sediments. An optical identification guide. (Elsevier Applied Science: London & New York). 539 SCHONBERG, C.H.L. & BARTHEL, D. 1997. Inorganic skeleton of the demosponge Halichondria panicea. Seasonality in spicule production in the Baltic Sea. Marine Biology 130: 133-140. 1998. Unreliability of demosponge skeletal charac- ters: the example of Halichondria panicea. Pp. 41-53. In Watanabe, Y. & Fusetani, N. (eds) Sponge Sciences. Multidisciplinary Perspectives. (Springer Verlag: Tokyo, Berlin, Heidelberg, New York). SCHRODER, D. 1978. Bodenkunde in Stichworten, 3rd edition. (Verlag Ferdinand Hirt: Kiel). 540 MEMOIRS OF THE QUEENSLAND MUSEUM THE INFECTIVENESS OF A BIOERODING CLIONID SPONGE. Memoirs of the Queensland Museum 44; 540. 1999:- Sponges play a major role in reef bioerosion. Early impressions suggested that only dead coral skeleton was infected. Cores of a very abundant Great Barrier Reef clionid sponge,Cliona sp., probably new, were removed with an underwater drill, allowed to heal and fixed onto living surfaces of nine coral spp. at Orpheus Island. Sponge survival varied greatly. lt was best on control surfaces of dead massive Porites, on live massive Porites and on Astreopora myriophtalma. It was least on Lobophyllia hemprichii and two branching Porites spp. Several individuals in seven of the nine coral species were infected within eight weeks. The areas of infection varied widely, However, after removal of grafts, the sponges died, regardless of their size. The risk of epidemics by fragmentation of this sponge species is considered to be low. O Porifera, Cliona, bioerosion, Great Barrier Reef, Coral Sea. C.H.L. Schónberg* (email: schoenberg@b iologie.uni-oldenburg.de) & C.R. Wilkinson, Australian Institute of Marine Science, PMB 3, Townsville MC, OLD 4810, Australia.*Present address: Carl von Ossietzkv Universität Oldenburg, Department of Zoology, PF 2503, 26111 Oldenburg, Germany; 1 June 1998. WANTED: THE NAMES OF COMMON BIOERODING SPONGES OF THE CENTRAL GREAT BARRIER REEF. Memoirs of the Queensland Museum 44: 540. 1999:- Bioeroding sponges have been well studied in the Mediterranean and the Caribbean Seas. Only few bioeroding sponge species are properly described from the Australian Great Barrier Reef. All other descriptions of Australian species are based on von Lendenfeld (1884-85) and de Laubenfels (1954). Detailed surveys in the central section of the Great Barrier Reef resulted in a large amount of new reference material suitable for revisions and new descriptions. Field descriptions and preliminary studies of spicule mounts made it possible to clearly distinguish 15 species from the rest ofthe samples, which are harder to categorise. Some of the sampled species appear to be sponges described previously and occurring in other oceans, some new species are likely to be endemic to the Coral Sea. Descriptions of the most common species are presented with preliminary names, distinguishing morphological features, and spicule assemblages with a request that participants assist by comparing with species they are familiar with or have observed elsewhere. O Porifera,Cliona, Cliothosa, Aka, bioerosion, Great Barrier Reef, Coral Sea, taxonomy. C.H.L. Schónberg* (email: schoenberg@ biologie.uni-oldenburg.de), Australian Institute of Marine Science, PMB 3, Townsville MC, OLD 4810, Australia. *Present address: Carl von Ossietzky Universität Oldenburg, Department of Zoology, PF 2503, 26111 Oldenburg, ER, Germany; 1 June 1998. FRESHWATER SPONGES (PORIFERA: SPONGILLIDAE) OF THE GUANACASTE CONSERVATION AREA, COSTA RICA: A PRELIMINARY SURVEY. Memoirs of the Queensland Museum 44: 540. 1999:- A survey of freshwater sponges in the Guanacaste Conservation Area (GCA), NW Costa Rica, was conducted in August and December of 1996 and March of 1997. The GCA occupies 110,000 hectares and includes Pacific dry forest, cloud forest and Atlantic rain forest. Objectives were to find out what sponges occur in the GCA, their distribution and preferred habitats. Sites were chosen to represent aquatic habitats from each biome in the GCA. At each site water temperature, pH, specific conductance and current velocity were measured. A unique aspect of this project entailed the measurement of particulate organic carbon (POC). POC can be useful for estimating food availability and has not been used to describe habitat preferences of freshwater sponges. To estimate POC, methods outlined by Wetzel & Likens (1979) were used. Listed in order of decreasing frequency, the following taxa were observed: Radiospongilla sp., Dosilia sp., Corvomeyenia sp., Spongilla cenota, Trochospongilla sp., and unidentifiable colonies without gemmules. The Radiospongilla species is currently being described in a separate paper (Poirrier, in prep.), and prior to this survey, Spongilla cenota was known only as far south as Florida and Mexico and therefore represents a significant range extension and a new record for Central America. The other genera are of uncertain taxonomic position and will be considered in a future paper. Sponges in the GCA were restricted to temporary, slow moving streams and ponds in dry tropical forest. These habitats have a drought season of up to six months and have POC content greater than 560g/l. No sponges were found in the clear, fast, permanent streams ofthe cloud and rain forests. Due to the high volume of fast flowing water, POC values are extremely low in these streams. These data suggest that natural aquatic habitats within evergreen tropical forests do not provide adequate food for freshwater sponges and that more favourable habitats are found in the dry tropical forest biome. This may be an important point for the conservation of Central American freshwater sponges, because dry tropical forest is considered the most endangered of all tropical ecosystems. O Porifera, Costa Rica, ecology, habitat, POC, Spongillidae, taxonomy, conservation. Scott A. Roush (email: s005sar(a discover wright.edu), Department of Biological Sciences, Wright State University, Dayton, Ohio 45435, USA. CHEMISTRY, ECOLOGY AND BIOLOGICAL ACTIVITY OF THE HAPLOSCLERID SPONGE OCEANAPIA SP; CAN ECOLOGICAL OBSERVATIONS AND EXPERI- MENTS GIVE A FIRST CLUE ABOUT PHARMACOLOGICAL ACTIVITY? P. SCHUPP, C. EDER. V. PAUL AND P. PROKSCH Schupp, P., Eder, C., Paul, V. & Proksch, P. 1999 06 30: Chemistry, ecology and biological activity of the haplosclerid sponge Oceanapia sp.; Can ecological observations and experi- ments give a first clue about pharmacological activity? Memoirs ofthe Queensland Museum 44: 541-549, Brisbane. ISSN 0079-8835. A new species ol Oceanapiu was discovered around Moen Island, Chuuk Lagoon, Micronesia, during collections made in 1994-1995. The appearance and growth form closely resembled that af Oceanapia sagiltaria. which is widely distributed throughout the Western Pacific. Unlike the Chuuk species, however, O, sayillaria has microscleres (siginas and roxas) and thicker oxeas, suggesting fo us it may be new. We examined whether this conspicuous, red sponge was chemically defended against generalist and more specialized fish predators, Methanol extracts of the sponge were highly deterrent in field feeding assays against generalist reef fish at physiological concentrations. This extract also deterred teeding by the spongivorous angelfish, Pomacamhus imperator, in laboratory feeding experiments at the same concentration. Based an our field observations and fish feeding experiments, different pharmacological screens were undertaken to demonstrate possible targets of these compounds. Kuanoniamine C and D showed insecticidal activity against the polyphagous larvae of Spodoptera littoralix, and toxic effects against the brine shrimp Artemia salina; N-deacy! derivative did not show a pronounced activity; whereas cytotoxicity screens. against HELA and MONO-MAC-6 tumor cells found that all three compounds revealed considerable cytotoxicity. O Porifera, Oceanapia, secondary metabolites, predation, Seeding deterrency, Micronesia, kuanoniamines. P. Schupp (email: schupp_marine_ica\hatmail.cam) & P. Praksch, Julius- von-Sachs-Institur Jür Biowissenschaften, Lehrstuhl Tür. Pharmazeutische Biologie, Julius-von-Sachs-Platz 2, D-97082 Würzburg, Germany; €. Eder, Hoechst Marion Roussel, Lead Generation, Natural Products Research, H 780, D-63926 Frankfurt, Germany; Y Paul, University of Guam, LOG Marine Laboratory, Mangilao, Guam 96923. USA; 20 May 1999. Despite considerable progress achieved in modern synthetic chemistry, natural products are still an indispensable source for the development of new pharmaceuticals and plant protectants. The frequent occurrence of bioactive compounds in nature is due to selective forces (e.g. predation or attack by pathogens), which have shaped and optimized a highly efficient chemical defense armory in many plants, animals and micro- organisms (Harborne, 1993), For decades scientists have relied on ethno-pharmacological knowledge to obtain initial indication on the biological and pharmacological properties of terrestrial planis. In contrast, there is almost no equivalent information available for the marine environment, requiring the implementation of Other strategies to achieve fast and environ- mentally sensible bioprospecting. Marine organisms are a rich source of novel, often unusual, secondary metabolites (Faulkner, 1996). Researchers often focus only on the ecology of secondary metabolites from marine organisms (Pawlik, 1993; Hay, 1996; Wink, 1998), or on the chemistry and pharmacological activity of the new compounds (Eder et al., 1998; Faulkner, 1996; Eder et al., 1999). Ecological studies focus mostly on assumed defensive functions of secondary metabolites, such as predator deterrence (Paul, 1992; Pawlik et al., 1995), prevention of fouling (De Nys et al., 1991), or inhibition of overgrowth (Thacker et al., 1998). Chemical studies report mainly on the structure elucidation of new compounds and their pharmacological activities, such as cytotoxicity (Steube et al.. 1998) and anti-microbial activity (Edrada et al.. 1996). Paul (1988) suggested that ecological experiments could be useful as a first indication for the presence of pharmacological active compounds, although at that time there was no clear empirical support for this idea. In this study, we show that ecological observations and HN HN Oen Y [6] [6] A. B. H NOS y SS | N Z N ` A D Sa =z S N S N H NH, ^N. H^ ~CH, c. D. FIG. 1. Major metabolites isolated from Oceanapia sp. A, kuanoniamine D; C, N-acetyl- kuanoniamine C; B, kuanoniamine D; D, dercitin. experiments can indeed be used as initial indicators of possible, potential pharmacological properties ofthe compounds involved (Schupp et al., 19992, 1999b). MATERIALS AND METHODS COLLECTION AND ISOLATION. Oceanapia sp. was collected at 1-3m depth on reef flats near the southern tip of Moen Island, Chuuk Lagoon, Federated States of Micronesia. Immediately after digging the basal ‘root’ of the sponge from the substrate, samples were separated in situ into the translucent capitum, fistule and base. After separation the different sample components were frozen and stored at - 20°C, until subsequent freeze-drying and extraction in 100% methanol. CHEMISTRY. The crude methanol extract was evaporated under reduced pressure and chromatographed on a silica gel column (elution with CH»Cl,/MeOH/NH,OH, 70:30:3). The pyridoacridine alkaloids kuanoniamine C (Fig. 1A), kuanoniamine D (Fig. 1B), and N-deacetyl- kuanoniamine D (Fig. 1C) were obtained after a final purification on a RP-18 silica gel column (MeOH/H,0/TFA, 70:30:0.5)(Eder etal., 1998). MEMOIRS OF THE QUEENSLAND MUSEUM The N-deacyl derivative could only be isolated from the sponge in trace amounts. Therefore, we synthesized the compound through acid hydrolysis of kuanoniamine C, which provided a sufficient quantity ofthe compound for subsequent bioassays. Kuanoniamine C (65mg) was dissolved in 120mL MeOH and the same volume of 2N HCl was added. The reaction mixture was stirred for 48hrs at 85?C under reflux. The sample was dried and the resulting red solid was purified by RP-18 column chromatography, using the solvent system described above, to yield 49mg of the new compound. Structure elucidation was accomplished by NMR and mass spectrometry (Eder et al., 1998). BIOLOGICAL ACTIVITY. Experiments with insects. Neonate larvae of the vigorous pest insect, Spodoptera littoralis, were used to test if the isolated compounds had any insecticidal properties. This assay allowed determination whether compounds were toxic towards insect larvae, or if they reduced or inhibited larval growth. It was not possible to determine whether reduced growth was due to inhibition of physiological processes and metabolism, or reduced food consumption caused by feeding deterrence. Larvae of S. littoralis were obtained from a laboratory colony, reared on an artificial diet under controlled conditions, as described by Srivastava & Proksch (1991). Feeding studies were conducted with neonate larvae (n=20) maintained on an artificial diet which had been treated with various concen- trations (13-373ppm) of the compounds under study. Spodoptera littoralis was offered the spiked diet over a range of 5-6 concentrations in a chronic feeding experiment. After 6 days, surviving larvae were counted, weighed, and compared to controls. LCsos were calculated from the dose-response curves by probit analysis (Eder et al., 1998). Brine Shrimp Assay. This assay was used to determine if the isolated compounds were toxic towards small marine organisms, although no information was forthcoming on the mode of action, or on how specific the toxicity was directed towards the test organisms (Artemia salina). Eggs of this species were kept for 48hrs in artificial CHEMICAL ECOLOGY OF OCEANAPIA SP. 54 seawater, as described previously (Meyer et al., 1982), and the nauplii (n=20) were introduced into vessels with brine, containing various concentrations (5-100ug/mL) of the compounds under test (each concentration in triplicate). DMSO (10uL/mL brine) was added to improve solubility. After 24hrs the surviving larvae were counted and compared to controls. LCs 9s were calculated from the dose-response curves by probit analysis (Eder et al., 1998). Cytotoxicity Studies. Information on the selectivity and possible intracellular targets were tested by analysing the effects ofkuanoniamine C and D, and the new compound N-deacetyl- kuanoniamine D, on cell growth and differentiation of two different human cell lines. These cell lines, MONO-MAC 6 and HELA, were deposited in the German Collection of Microorganisms and Cell Cultures (DSMZ, Braunschweig, Germany). Cultures were mycoplasma-free, and were cultivated under standardized conditions (Drexler et al., 1995). For all experiments, exponentially-growing cells were used, with a viability exceeding 90%, as determined by trypan blue staining. The final concentration used in experiments was 10°cells/mL. Experiments were conducted using 1001U/mL penicillin G and 100mg/mL streptomycin. Concentrated stock solutions of the test compounds were prepared in ethylene glycol monomethyl ether, and stored at -20°C. For the cytotoxicity analysis, cells were harvested, washed, and resuspended in a final concentration of 10°cells/mL. These solutions were seeded in triplicate, in 90uL volumes, in 96-well flat bottom culture plates (Nunc). Test compounds in 10uL, obtained by diluting the stock solution with a suitable quantity of growth medium, were added to each well (Eder et al., 1998). Cultures were incubated for 48hrs at 37°C in a humidified incubator with 5% CO;. Cytotoxicity was determined by incorporation of ['H]- thymidine (Steube et al., 1992), Radio- active-incorporation was carried out for the last 3hrs of the 48hr incubation period. One pCi of [methyl-^H]-thymidine (Amersham-Buchler, Braunschweig, Germany; specific activity 0.25mCi/umol), was added in 20uL volumes to each well. Cells were harvested on glass fiber filters with a multiple automatic sample harvester, and radioactivity was determined in a liquid scintillation counter (1209 Rackbeta, LKB, Freiburg, Germany). Media with 0.2% ethylene glycol monomethyl ether were included in the experiments as controls (Eder et al., 1998). od ECOLOGY. Distribution of secondary metabolites. Extract and secondary metabolite concentrations were determined for the basal, fistular and capitum subsamples of Oceanapia sp., to see whether there were any differences in compound concentrations between these subsamples of the sponge, correlated with differences in exposure of these sponge structures (e.g. burrowing versus exposed). This strategy was essential to test the extract and sponge compounds at ecological relevant concentrations. These freeze-dried sponge subsamples were weighed before extraction, and the extract weights were also determined after removing all solvents with a rotary evaporator. Concentrations of kuanoniamine C and D in the different subsamples of the sponge were determined by HPLC quantif- ication, following methods described by Schupp et al. (1999b). Field feeding assays. Field feeding experiments using natural fish assemblages were conducted as an initial test for biological activity of the crude extract. We used reef fishes to determine whether the crude extract had antifeedant properties. The methanol fraction was incorporated in an artificial diet at a concentration of 7.4 % of dry mass, which represents the methanol extract concentration in the basal part of the sponge. We also tested the three major metabolites (Fig. 1 A-C) at their respective fistule concentrations, against several fish species in the field at Western Shoals, Guam. Kuanoniamine C and D and the N-deacyl derivative were tested at fistule concentrations since this was considered to be the most easily accessible structure of the sponge susceptible to fish feeding (Schupp et al., 1999b). The food was prepared according to Schupp et al. (1999b). Four food cubes, with or without one of the metabolites, were attached to a poly- propylene rope by a safety pin. Ropes were placed on the reef in pairs of one treated (with one metabolite), and one control (without the meta- bolite) rope, and attached to coral heads at 3-5m depth. Several pairs of ropes (replicates) were set at the same time. Ropes were removed when approximately half the cubes were completely eaten. These assays were scored as the number of cubes completely eaten, and the results analyzed with a Wilcoxon signed-rank test for paired comparison (two-tailed; Schupp et al., 1999b). Laboratory feeding assays. The angelfish, Pomacanthus imperator, was used as an experi- mental subject to determine whether or not there were differences in susceptibility between generalist and specialist predators (Thacker et al., 544 MEMOIRS OF THE QUEENSLAND MUSEUM TABLE 1. /n vitro cytotoxicity of the kuanoniamine derivatives (ICso values, ug/mL) to HELA and MONO-MAC-6 tumor cells in the [3H]-thymidine incorporation assay (MTT), (mean + SD, n=3; Eder et al., 1998). i — | Compound Cell line = 1 2 3 HELA 5.113 14207 — 1230.2 MONO-MAC-6 1.20.4 0.8 + 0.1 2.0+ 0.5 1998). Experimental food was prepared following methods described by Schupp et al. (1999b), resulting in strips of window screens that had one rectangle of treated food and one rectangle of control food (2.5x2.0cm) embedded in the screens. To determine the amount of control and treated food eaten by angelfish, the squares in the window screen served as a grid, and the number of squares where the food had been completely removed were counted (Hay et al., 1998). These results were analyzed with a paired t-test (Schupp et al., 1999b). RESULTS AND DISCUSSION TAXONOMY. While diving on reef flats, in sandy areas and coral rubble, a bright-red, short stocked capitum was observed protruding from the sand. The capitum was found sitting on top of a long fistule, buried in the sand, and attached to the turnip-shaped base of the sponge burrowed in the substrate (Fig. 2). Sometimes this basal portion can be buried up to 20cm deep into the substrate. The capitum is easily broken off by stronger water movements, found ‘rolling’ on the substrate. The capita are between 0.5-1.5cm diameter. The hollow fistules vary in length from 6-12cm and up to 1.3cm diameter. Fistules are attached to an irregular turnip-shaped main body (base), up to 10cm high, 6cm diameter, and sometimes several fistules are attached to the Sediment FIG. 2. Schematic drawing of Oceanapia sp. (Redrawn from Hooper et al., 1993). base of the sponge. In some cases one fistule branches into several smaller ones, protruding from the sand. The burrowing basal portion often grows around large pieces of coral rubble, or completely incorporates smaller pieces of rubble, shells, or sand into the base. The color of the sponge ranges from bright- to dull-red. The surface of the sponge is smooth, with a distinct skin-like ectosomal layer, and a very crumbly, fragile choanosome, with an abundance of foreign material. Spicules are exclusively curved oxeas, with pointed or blunt tips, ranging in size from 260-320x3-5um. This material probably belongs to an undescribed species of Oceanapia (M. Kelly, pers. comm.), subsequently verified by Dr. Rob W.M. van Soest (Eder et al., 1998), and a voucher of the sponge is deposited in the Zoólogisch Museum, Amsterdam (registration number ZMA POR 11007). CHEMISTRY. Three pyridoacridine alkaloids were isolated from Oceanapia sp. (Eder et al., 1998): kuanoniamine C (Fig. 1A), kuanoniamine D (Fig. 1B), and N-deacetylkuanoniamine D (Fig. 1C), of which kuanoniamine C and D have been previously described from an unidentified tunicate and its prosobranch mollusc predator, Chelynotus semperi (Carroll & Scheuer, 1990). Additionally, kuanoniamine C has also been isolated from a deep water sponge of the genus Stelletta (Gunawardana et al., 1989, 1992), and from Oceanapia sagittaria (Salomon & Faulkner, 1996). The fact that these alkaloids occur in different species, comprising more than one major phylum, gives support to the specul- ation that marine microorganisms could possibly be the true source of these compounds (Molinski, 1993). Conversely, there is some evidence that sponge cells are the actual producers of the alkaloids (Faulkner et al., 1999, this volume). BIOLOGICAL ACTIVITY. The insecticidal activity of the three metabolites (Fig. 1A-C) was determined by assessing survival and growth rates after six days of exposure. From the dose- response curves obtained, LCs) values of 156ppm (+0.46 S.E.) for kuanoniamine C, and CHEMICAL ECOLOGY OF OCEANAPIA SP. 545 TABLE 2. Results of the fish feeding experiments testing the deterrent effect of the extract and the kuanoniamine derivatives against an array of reef fish at Western Shoals, Guam. Sponge part (compound/ extract concentration, % Mean number of cubes eaten STD N P of dry mass) Control food Treated food Methanolic extract from base , 7.4% 3.1 202 0.7 € 0.7 17 «0.001 Kuanoniamine C from E fistule, 1.2% 3.5 £05 0.5 + 0.6 19 «0.001 Kuanoniamine D from fistule, 0.4 % 3.6 + 0.5 LI£L2 19 «0.001 N-deacetylkuanoniamine e D from fistule, 0.04 % 2701 2.5411 20 0.27 59ppm (+0.30) for kuanoniamine D, were calculated by probit analysis. The new compound N-deacetylkuanoniamine D (Fig. 1C), was tested up to a concentration of 934ppm, although, in contrast to the other two compounds, no significant insecticidal activity was observed. Only the growth of larvae was reduced. This can be expressed by an EDs value of 141 ppm (+0.13) for the new compound (Fig. 1C). General cytotoxicity was assessed by testing the three alkaloids (Fig. 1A-C), in the brine shrimp assay. After 24hrs of exposure, the number of surviving nauplii was counted. LCso values calculated from the dose-response curves by probit analysis were 37ug/mL (+17) for one compound (Fig. 1A), and 19ug/mL (+4) for the second compound (Fig. 1B). The third compound (Fig. 1C), did not show any toxic effect up to 100ug/mL. No LCsy of this alkaloid was obtained due to its limited solubility at concentrations greater than 100ug/mL. Several pyridoacridine alkaloids have been reported to exhibit significant cytotoxicity towards murine and human tumor cell lines (Molinski, 1993). In this study, we analyzed the effects of the three compounds (Figs 1 A-C), on cell growth towards two different human cell lines using [' ‘HI- -thymidine incorporation. This cytotoxicity assay is used to determine the capability of cells to synthesize DNA during the cell cycle. The validity of the method applied in this study has been documented previously (Arnould et al., 1990). Table 1 summarizes the results obtained with the two cell lines. The small amount of ethylene glycol monomethyl ether (0.1%), used to solubilize the compounds, did not affect growth of the tumor cells. Each alkaloid was tested for its cytotoxic activity at a range of concentrations (0.1-201g/mL). Suppression of ['H]-thymidine incorporation into cells treated with the three compounds (Fig. 1A-C), was observed for each of the two cell lines. Intercalation with DNA has been demonstrated previously for dercitin (Fig. 1D) (Gunawardana et al., 1988; Burres et al., 1989), a marine natural product closely related to the kuanoniamines. Based on obvious structural similarities of these three alkaloids (Fig. 1A-C) with dercitin, it may be hypothesized that the kuanoniamines also interact with DNA by intercalation. In addition to our own testing, kuanoniamine C was tested with the im vitro disease-oriented primary antitumor screen on a panel of 60 cell lines at the National Cancer Institute (NCI), in Bethesda, Maryland, U.S.A. The in vitro assays showed that most cell lines were fairly sensitive against kuanoniamine C at concentrations, ranging from 4.93 x 10 molar to 3.11 x 10% molar for the Gls, and a concen- tration of 1,86 x 10% molar as the mean graph midpoint (mean concentration required over all cell lines; Monks et al., 1991). The Glo is an interpolated value representing the concentration at which the percentage growth of exposed cells used in the assay is 50% compared to that of non-exposed cells (Monks et al., 1991). One of the breast cancer cell lines (MCF7), was extremely sensitive against kuanoniamine C, with a concentration of 3.11 x 10% molar for the Glso. Based on the successful performance of the in vitro screen, NCI subsequently requested further material for in vivo studies. The cytotoxicity data are not corroborated by insecticidal activity towards neonate larvae of S. littoralis, or the toxic activity against A. salina. In the latter tests, the new derivative was nearly inactive, but it showed similar activity in the 546 cytotoxicity assay to kuanoniamine C and D. Thus, it is possible that the different activities observed are not caused by a general cytotoxicity, but may be due to a different mode of action. ECOLOGY. Distribution of secondary metabolites in the different sponge parts. Extrac- tion of the base, fistule and capitum of the sponge yielded significantly different crude extract concentrations. Concentrations increased from the base, with an extract yield of 26.4% of dry mass, 34.2% in the fistule, up to 52.7% crude extract concentration in the capitum (Schupp et al., 1999b), Concentrations of kuanoniamine C and D increased sharply from the base to the capitum of the sponge. The lowest concentrations of these alkaloids were found in the base, with 0.4% of dry mass for kuanoniamine C and 0.1% of dry mass for kuanoniamine D. The fistule showed a four fold increase of these metabolites, with 1.2% of kuanoniamine C and 0.4% of kuanoniamine D. Highest concentrations were found in the translucent capitum, with 3.5% of dry mass for kuanoniamine C and 1.2% of dry mass for kuanoniamine D, representing a 9 to 12-fold increase in secondary metabolite concentration compared to the base (Schupp et al., 1999b). This huge increase of the two major metabolites kuanoniamine C and D in Oceanapia sp. stresses the importance not only to determine an overall yield of secondary metabolites, but also to look at compound concentration at an intraspecimen level. This is important for both ecological investigations and pharmacological screening for new bioactive compounds. Secondary metabolites are often concentrated in parts where they are first encountered by pred- ators, such as the skin of mollusks or in biologically valuable parts like reproductive regions (Pawlik et al., 1988; Avila & Paul, 1997). When whole organisms are extracted, compounds used by animals to deter possible predators might show lower yields than they actually have in the organs or regions where they are deposited. Therefore, it is possible that bioactive compounds are over- looked in pharmacological screens, because their concentrations are too low. Fish feeding assays. The methanol fraction was highly deterrent towards reef fishes at base concentration (Table 2). It also deterred feeding by the angelfish Pomacanthus imperator in laboratory feeding experiments at the same concentration, although the deterrent effect was MEMOIRS OF THE QUEENSLAND MUSEUM not as pronounced as with other reef fishes (p = 0.047; N = 7; Schupp et al., 1999b). The field and laboratory fish feeding experiments showed that even the base is chemically well protected. This is surprising, since the base is normally not accessible to fish because it is buried in the sand. A possible explanation could be that the high extract concentrations might be a defense against crabs, flatworms and other sediment-dwelling invertebrates (Schupp et al., 1999b). The laboratory feeding experiments with P. imperator demonstrated the potency and broad effectiveness of the compounds, since they deterred the more specialized angelfish in addition to being a deterrent towards generalist reef fishes. Angelfishes are known to feed primarily on ascidians and sponges, and are thought not to be as susceptible to sponge secondary metabolites as are other fishes (Randall & Hartman, 1968; Wulff, 1994). The field feeding assays with the pure compounds demonstrated that the pyridoacridine alkaloids were at least partly responsible for the deterrent effect of the crude extract. Kuanoniamine C and D clearly deterred fish feeding at fistule concentration (Table 2). Each compound was by itself effective in deterring reef fishes. Schupp et al. (1999b) suggested that it might be important for a sponge to have more then one deterrent chemical, or to increase the overall concentration of defensive compounds, to prevent feeding by different predators. The N-deacyl derivative did not reduce feeding (Table 2). One possible explanation is the low concentration of the compound in the sponge. It was tested at natural concentrations, which is approximately 1/10 of the concentration used for kuanoniamine D. There are numerous studies showing that benthic invertebrates can reduce predation through the production of secondary metabolites (Bakus et al., 1986; Wylie & Paul, 1989; Paul, 1992; Pawlik, 1993; Ebel et al., 1997; Thacker et al., 1998), but there is little information on the pharmacological activity of these compounds - this study being one of them. One problem might be that researchers looking at the ecological importance of new compounds, seldom perform pharmaceutical screenings, and conversely, those searching for new pharmacologically active substances seldom conduct ecological experi- ments with the substances or organisms. Fish feeding assays could be used as a first indication for the presence or absence of pharmacologically CHEMICAL ECOLOGY OF OCEANAPIA SP. active substances, and also providing inform- ation on the ecological relevance of these compounds as feeding deterrents. Schupp et al. (1999b) demonstrated that the intraspecimen distribution of secondary metab- olites in Oceanapia sp. was in accordance with the optimal defense theory (McKey, 1974, 1979; Baldwin & Ohnmeiss 1994). This theory suggests that organisms optimize the production of secondary metabolites and deter possible predators with the least amount of energy, assuming that the production of secondary meta- bolites is costly. Therefore, biologically important parts (with a high contribution to fitness), like reproductive regions (e.g. the capitum in Oceanapia sp.; Hooper et al., 1993), and parts that are first encountered by predators (e.g. the fistule and capitum in Oceanapia sp.), should show higher secondary metabolite concentrations. This is certainly the case in Oceanapia sp. (Schupp et al., 1999b). It might be advantageous to keep the optimal defense theory in mind, while collecting and screening for pharmacologically active compounds. By collecting mostly exposed and vulnerable parts of marine invertebrates, it is highly probable that biologically active compounds are present and that they are tested at ecologically relevant concentrations. This could be also interesting from a conservation standpoint, since it should be possible to collect the exposed parts of certain organisms (e.g. sponges), and leave the remainder to regrow. Oceanapia sp. can be used again as an example, since it is possible to collect the capitum without severely damaging the sponge. The capitum grows back and can be harvested repeatedly. The harvested amounts are small, but show on the other hand the highest secondary metabolite concentrations. CONCLUSION This study demonstrates, that ecological observations and experiments can be useful as initial indicators for possible pharmacological properties of secondary metabolites (Schupp et al., 1999a, 1999b). Field observations and fish feeding experiments demonstrated the biological activity of the pyridoacridine alkaloids found in Oceanapia sp. Through different pharmacological screens we were able to show possible physio- logical targets of the compounds, although these results do not explain the mode of action as a fish deterrent agent. ACKNOWLEDGEMENTS Financial support by the DFG (SFB 251) and by the ‘Fonds der Chemischen Industrie’ (both to P.P.) is gratefully acknowledged. P. Schupp wants to thank the DAAD for a Kurzzeitstipendium during summer 1994. V. Paul acknowledges support from the National Institutes of Health (GM 38624) for work conducted on Guam. We would like to thank Dr Michelle Kelly-Borges and Dr Rob W. M. van Soest for their assistance with the identification of the sponge. We acknowledge the National Cancer Institute, in Bethesda, Maryland, U.S.A., for performing the in vitro disease-oriented primary antitumor screens. We also thank David Ginsburg and the staff of the UOG Marine Laboratory for their help, especially during the fish feeding assays. Dr Chuck Birkeland provided advice on statistical analyses. We also thank the Coral Reef Research Foundation for assistance during collection ofthe sponge. We also thank John Hooper for his editorial comments and suggestions. LITERATURE CITED ARNOULD, R., DUBOIS, J., ABIKHALIL, F., LIBERT, A., GHANEM, G., ATASSI, G., HANOCQ, G. & LEJEUNE, F.J. 1990. Comparison of two cytotoxicity assays: tetra- zolium derivative reduction MTT and tritiated thymidine uptake on three malignant mouse cell lines using chemotherapeutic agents and investigational drugs. Anticancer Research 10: 145-154, AVILA, C. & PAUL, V.J, 1997. Chemical ecology of the nudibranch Glossodoris pallida: is the location of diet-derived metabolites important for defense? Marine Ecology Progress Series 150: 161-170 BAKUS, G.J., TARGETT, N. & SCHULTE, B. 1986. Chemical ecology of marine organisms: an overview. Journal of Chemical Ecology 12: 951-987 BALDWIN, I.T. & OHNMEISS, T.E. 1994. Coordination of photosynthesis and alkaloidal responses to damage in uninducible and inducible Nicotiana sylvestris. Ecology 75: 1003-1014. BURRES, N.S., SAZESH, S., GUNAWARDANA, G.P. & CLEMENT, J.J. 1989. Antitumor activity and nucleic acid binding properties of dercitin, a new acridine alkaloid isolated from a marine Dercitus species sponge. Cancer Research 49: 5267-5274. CARROLL, A.R. & SCHEUER, P.J. 1990. Kuanoniamines A, B, C, and D: Pentacyclic alkaloids from a tunicate and its prosobranch mollusk predator Chelynotus semperi. Journal of Organic Chemistry 55: 4426-4431. DE NYS, R., COLL, J.C. & PRICE, LR. 1991. Chemically mediated interactions between the red alga Plocamium hamatum (Rodophyta) and the octocoral Sinularia cruciata (Alcyonacea). Marine Biology 108: 315-320 DREXLER, H.G., DIRKS, W., McLEOD, R.A.F., QUENTMEIER, H. & STEUBE, K.G. 1995. DSMZ Catalogue of Human and Animal Cell Lines, 5th Edition (DSMZ: Braunschweig, Germany). EBEL, R., BRENZIGER, M., KUNZE, A., GROSS, H.J. & PROKSCH, P. 1997. Wound activation of protoxins in the marine sponge Aplysina aerophoba. Journal of Chemical Ecology 23: 1451-1462 EDER, C., SCHUPP, P, PROKSCH, P., WRAY, V., STEUBE, K., MULLER, C.E., FROBENIUS, W., HERDERICH, M. & SOEST, R.W.M. VAN 1998. Bioactive pyridoacridine alkaloids from the Micronesian sponge Oceanapia sp. Journal of Natural Products 61: 301-305 EDER, C., PROKSCH, P., WRAY, V., STEUBE, K., BRINGMANN, G., SOEST, R.W.M. VAN, SUDARSON, O., FERDINANDUS, E., PATTISINA, L.A., SUMALI, W. & MOKA, W. 1999. New alkaloids from the indopacific sponge Stylissa carteri. Journal of Natural Products 62: 184-187 EDRADA, R.A., PROKSCH, P., WRAY, V. & CHRIST, R. 1996. Bioactive isoquinoline quinones from an undescribed Philippine marine sponge of the genus Xestospongia. Journal of Natural Products 59: 973-976 FAULKNER, D.J. 1996. Marine natural Products. Natural Product Reports 13: 259-302. GUNAWARDANA, G.P, KOHMOTO, S., GUNASEKERA, S.P, McCONNELL, OJ. & KOEHN, F.E. 1988. Dercitin, a new biologically active acridine alkaloid from a deep water marine sponge, Dercitus sp. Journal of the American Chemical Society 110: 4856-4858. GUNAWARDANA, G.P., KOHMOTO, S. & BURRES, N.S, 1989. New cytotoxic acridine alkaloids from two deep water marine sponges of the family Pachastrellidae. Tetrahedron Letters 30: 4359-4362 GUNAWARDANA, G.P., KOEHN, F.E., LEE, A.Y., CLARDY, J., HE, H. & FAULKNER, D.J. 1992. Pyridoacridine alkaloids from deep-water marine sponges of the family Pachastrellidae: Structure revision of dercitin and related compounds and correlation with the kuanoniamines. Journal of Organic Chemistry 57: 1523-1526, HARBORNE, J.B. 1993. Ecological Biochemistry. (Academic Press: London). HAY, M.E. 1996. Marine chemical ecology: what's known and what's next, Journal of Experimental Marine Biology and Ecology 200: 103-134. HAY, M.E., STACHOWICZ, J.J., CRUZ-RIVERA, E., BULLARD, S., DEAL, M.S. & LINQUIST, N. 1998. Bioassays with marine and freshwater MEMOIRS OF THE QUEENSLAND MUSEUM macroorganisms. Pp. 39-141. In Haynes, K.F. & Millar, J.G. (eds) Methods in chemical Ecology. Vol. 2. Bioassays Methods. (Chapman & Hall: New York). HOOPER, J.N.A., KELLY-BORGES, M. & RIDDLE, M. 1993, Oceanapia sagittaria from the Gulf of Thailand. Memoirs of the Queensland Museum 33: 61-72, McKEY, D. 1974. Adaptive patterns in alkaloid physiology. American Naturalist 108: 305-320. 1979, The distribution of secondary compounds within plants. Pp. 55-133. In: Rosenthal, G.A. & Janzen, D.H. (eds) Herbivores: their interactions with secondary plant metabolites. (Academic Press: New York). MEYER, B.N., FERRIGNI, N.R., PUTNAM, J.E., JACOBSEN, L.B., NICHOLS, D.E. & McLAUGHLIN, J.L. 1982. Brine shrimp: A convenient general bioassay for active plant constituents. Planta Medica 45: 31-34. MOLINSKI, T.F. 1993. Marine pyridoacridine alkaloids: Structure, synthesis, and biological chemistry. Chemical Reviews 93: 1825-1838. MONKS, A., SCUDIERO, D., SKEHAN, P., SHOEMAKER, R., PAUL, K., VISTICA, D., HOSE, C., LANGLEY, J., CRONISE, P., VAIGRO-WOLFF, A., GRAY-GOODRICH, M., CAMPBELL, H., MAYO, J. & BOYD, M. 1991. Feasibility of a high-flux anticancer drug screen using a diverse panel of cultured human tumor cell lines. Journal of the National Cancer Institute 83: 757-766. PAUL, V.J. 1988. Marine chemical ecology and natural products research. In Fautin, D.G. (ed.) Biomedical importance of marine organisms. Memoirs of the California Academy of Science 13: 23-27. 1992. Ecological roles of marine secondary metabolites. (Comstock Publishing Associates: Ithaca, New York). PAWLIK, J.R. 1993. Marine invertebrate chemical defenses. Chemical Reviews 93: 1911-1922. PAWLIK, J.R., KERNAN, M.R., MOLINSKI, T.F., HARPER, M.K. & FAULKNER, D.J. 1988. Defensive chemicals of the Spanish dancer nudibranch Hexabranchus sanguineus and its egg ribbons: macrolides derived from a sponge diet. Journal of Experimental Marine Biology and Ecology 119: 99-108. PAWLIK, J.R., CHANNAS, B., TOONEN, RJ. & FENICAL, W. 1995, Defenses of Caribbean sponges against predatory reef fish. 1. Chemical deterrency. Marine Ecology Progress Series 127: 183-194, RANDALL, J.E. & HARTMAN, W.D. 1968. Sponge feeding fishes of the West-Indies. Marine Biology 1: 216-225. SALOMON, C.E. & FAULKNER, D.J. 1996. Sagitol, a pyridoacridine alkaloid from the sponge Oceanapia sagittaria. Tetrahedron letters 37: 9147-9148. CHEMICAL ECOLOGY OF OCEANAPIA SP. SCHUPP, P., EDER, C., PROKSCH, P., WRAY, V., SCHNEIDER, B., HERDERICH, M. & PAUL, V. 1999a. Staurosporine derivatives from the ascidian Eudistoma toealensis and its predatory flatworm Pseudoceros sp. Journal of Natural Products (in press). SCHUPP, P., EDER, C., PAUL, V. & PROKSCH, P. 1999b. Distribution of secondary metabolites in the sponge Oceanapia sp. and its ecological implications. Marine Biology (in press). SRIVASTAVA, R.P. & PROKSCH, P. 1991. Contact toxicity and feeding inhibitory activity of chromenes from asteraceae against Spodoptera littoralis (Lepidoptera; Noctuidae). Entomologia Generalis 15: 265-274. STEUBE, K.G., GRUNICKE, D., PIETSCH, T., GIGNAC, S.M., PETTIT, G.R. & DREXLER, H.G. 1992. Dolastatin 10 and 15: Effects of two natural peptides on growth and differentiation of leukemia cells. Leukemia 6: 1048-1053. STEUBE, K.G., MEYER, C., PROKSCH, P., SUPRIYONO, A., SUMARYONO, W. & 549 DREXLER, H.G. 1998. A new calyculin deriv- ative from the sponge Theonella swinhoei is a novel and potent inhibitor of tumor cell proliferation. Anticancer Research 18: 129-138. THACKER, R.W., BECERRO, M.A., LUMBANG, W.A. & PAUL, V.J. 1998. Allelopathic inter- actions between sponges on a tropical reef, Ecology 79: 1740-1750, WINK, M. 1998. Chemical ecology of alkaloids. Pp, 265-300. In Rogers, M.F. & Wink, M. (eds) Alkaloids: biochemistry, ecology, and medicinal applications. (Plenum Press: New York). WULFF, J.L. 1994. Sponge feeding by Caribbean angelfishes, trunkfishes, and filefishes. Pp. 265-271. In Soest, R.W.M. van, Kempen, T.M.G. van & Braekman, J.-C. (eds) Sponges in time and space. (Balkema: Rotterdam). WYLIE, C.R. & PAUL, V.J. 1989. Chemical defenses in three species of Sinularia (Coelenterata, Alcyonacea): effects against generalist predators and the butterflyfish Chaetodon unimaculatus Bloch. Journal of Experimental Marine Biology and Ecology 129: 141-160. 550 MEMOIRS OF THE QUEENSLAND MUSEUM MICROBIAL INFLUENCED PYRITISATION OF MARINE SPONGES. Memoirs of the Queensland Museum 44: 550. 1999:- Although biomineralisation plays an important role in the history of earth, our knowledge about the involved processes is rather limited. In this context microorganisms take a significant part within these global processes. Pyrite crystals are often found in taphonomically mineralised sponge tissues but are absent or rare in the surrounding sediment. In terms of microbiology, the role of sponge-associated, anaerobic species such as sulfate reducing and fermentative bacteria is fairly unknown. During early decaying processes of the sponge tissue the internal sponge becomes entirely anaerobic which is a necessary prerequisite for the growth and metabolical activity of the sulfate reducers. Therefore, pyrite formation is probably linked with sulfide producing bacteria. In the marine environment, sulfate reducing bacteria are most likely to play this role besides heterotrophic sulfidogenic bacteria. In this study enrichment cultures of native sponge tissue of the mediterranean sponges Chondrosia reniformis and Petrosia ficiformis under sulfate reducing conditions were investigated using a combination of rRNA-targeted in situ probing and classical cultivation techniques. Recently developed specific oligonucleotide probes for sulfate reducing bacteria elucidated the occurrence and abundance of various sulfate reducing-species affiliated to different phylogenetic taxa within the Desulfovibrionaceae and Desulfobacteriaceae. Furthermore, 16S-rRNA based phylogeny revealed to hitherto unknown anaerobic bacteria inhabiting the sponge mesohyl. Additionally, the sulfate reducing activity was confirmed by established physiological methods. In living sponges, sulfate reducing bacteria were evenly distributed within the tissue. This indicates the presence of anoxic microniches, which allow not only the survival, but even the subsequent growth of these anaerobic bacteria. Further investigations of culturable sponge-associated sulfate reducing bacteria are currently carried out to investigate their ecology and ability to produce pyrit in culture. CJ Porifera, biomineralisation, pyritisation, anaerobic microbes, sulphate reducing bacteria, phylogeny, molecular biology, microniches. G. Schumann-Kindel (email: schu0654@ mailszrz.zrz.tu-berlin.de), M. Bergbauer, W. Manz & U. Szewzyk, Technische Universität Berlin, FG Ökologie der Mikroorganismen, FranklinstraBe 29, 10587 Berlin, Germany; J. Reitner, Institut und Museum für Paläontologie und Geologie, GoldschmidtstraBe 3, 37077 Gottingen, Germany; 1 June 1998. RELATIONSHIP OF SAND AND FIBRE IN THE HORNY SPONGE, PSAMMOCINIA CHUNG JA SIM AND KYUNG JIN LEE Sim, C.J. & Lee, K.J. 1999 06 30: Relationship of sand and fibre in the horny sponge, Psammocinia. Memoirs of the Queensland Museum 44: 551-557. Brisbane. ISSN 0079-8835. Five species of the genus Psammocinia (Irciniidae) are described from Chejudo Island, Namhaedo Island and Wando Island, Korea (Psammocinia jejuensis, P. mosulpia, P. mammiformis, P. samyangensis and P. wandoensis). Psammocinia is characterised by large quantities of sand in spongin fibres, mesohyl matrix and as a thick superficial cortex. In addition to the primary and secondary branching fibres, fine filaments emerge from individual pores in the fibres. Occasionally short secondary fibres are connected to large sand grains, forming bridges between adjacent sand grains. The skeleton formed from sand grains associated with fibres provides additional support for the body of the sponge. O Porifera, horny sponge, Dictyoceratida, Psammocinia, Korea. Chung Ja Sim (email: cjsim@eve.Hannam.ac.kr) & Kyung Jin Lee, Department of Biology, Hannam University, Daejeon 300-791, Korea; 6 January 1999. Psammocinia (Irciniidae) is characterised in having many sand grains within spongin fibres and the mesohyl matrix, and a surface crust of sand (Bergquist, 1980; Cook & Bergquist, 1996). Lendenfeld (1888, 1889) reported eight species in Psammocinia, at that time included as a subgenus of Hircinia (Ircinia). Of these, only four are currently included in Psammocinia: H. rugosa Lendenfeld, 1889, H. arenosa Lendenfeld, 1888, H. tenella Lendenfeld, 1889 and H. halmiformis Lendenfeld, 1888, and of these two (H. rugosa and H. tenella) are synonyms of P. vesiculifera (Poléjaeff, 1884) (Hooper and Wiedenmayer, 1994). More recently, Bergquist (1995) reported one new species from New Caledonia, and Cook & Bergquist (1998) described five new species from New Zealand. In Korea, three species of Psammocinia were reported from Chejudo Island and Namhaedo Island (Sim, 1998), and two new species, P samyangensis Sim & Lee, 1998 and P. wandoensis Sim 8 Lee, 1998, were described from the South Sea of Korea (Sim & Lee, 1998). In the present study we examine five species from Korean waters with the aim to determine the extent of skeletal support provided by these foreign skeletal elements, showing that sand is closely associated with the fibres of the sponge, sometimes providing support to the sponge as bridges of primary and secondary fibres connect sand grains in the body. The principal diagnostic characteristic of Irciniidae is the possession of a third element of the skeleton consisting of fine collagenous filam- ents beyond the fasciculate primary fibres and uncored secondary fibres (Bergquist, 1980). Filaments of Psammocinia emerge from pores in the fibres. MATERIALS AND METHODS Specimens of Psammocinia were collected from Chejudo Island, Namhaedo Island and Wando Island in Korea; P. mammiformis (Manjaedo, Namhaedo Island), P. mosulpia (Mosulp'o, Chejudo Island), P jejuensis (Kimnyung, Chejudo Island), P. samyangensis (Samyang | dong, Chejudo Island) and P. wandoensis (Wando Island). Specimens were collected by SCUBA, 10-25m depth, and by fishing-net. For identi- fication of horny sponge, light microscopy and SEM (AKASHI ISI-SS40) were used to determine fibre arrangement. RESULTS DISTRIBUTION OF SAND. In all species examined the primary fibres are completely filled with sand grains that form a loosely packed sand core. Secondary fibres may be either partially cored with sand, or lack any sand grains within their cores (Fig. 1D, E). Larger size sand grains are attached to the outside ofthe fibres (Fig. 2G). Beneath the surface, a sand crust is mixed with foreign spicules (Fig. 1A, B). In one species (P Jejuensis) the external surface of the sponge is armoured with pieces of shell debris. SAND ON THE ECTOSOMAL CRUST. Surfaces of P. wandoensis, P. mosulpia and P. mammiformis have a sand crust mixed with foreign spicules, whereas P. samyangensis has a un nN to MEMOIRS OF THE QUEENSLAND MUSEUM FIG. 1. A-B, Psammocinia mosulpia; A, sand crust; B, under-crust mixed with sponge spicules. C, P. wandoensis, sand attached inside of fibres. D-E, P. Samyangensis; D, primary fibre with sand; E, large oxea supporting fibre. F, P. mosulpia, primary fibre with sand. G-H, P. jejuensis; G, large oxea in fibre; H, sand in fibre. I, P. mammiformis, primary fibre with sand. (Scale bars; A-B, 300um; C-D, 150um; E, 300um; F- I, 150um) SAND & FIBRE IN PSAMMOCINIA FIG. 2. A, Psammocinia mammiformis, choanosome with sand grains (SG). B, P. samyangensis, choanosome. C, P. jejuensis, choanosome. D, P. mosulpia, choanosome. E, P. wandoensis, choanosome. F, P. samyangensis, choanosome secondary fibre with sand. G, P. mosulpia, choanosome fibre with sand (Scale bars: A-D, 4001m; E, 100um; F, 80um; G, 300pm). thin filamentous membrane mixed with large sand grains and pieces of shell, each 1-2.5mm diameter, not strictly a sand crust. The texture of this species is soft and easily torn because the fibre and filament arrangement is very loose. Psammocinia jejuensis also has filamentous membrane instead of true sand crust, with large grains of sand and pieces of shell distributed within the surface armour. In P. mosulpia there is a black sand crust making this species appear darkly pigmented (Fig. 1A, B). SAND IN THE MATRIX. The choanosomal matrix also contains many sand grains, with sizes of sand grains varying between each species. In P. Jejuensis and P. samyangensis large sand grains, 0.7-4.0mm diameter, are combined with the filament network (Fig. 2B). In P. wandoensis the 554 mesohyl matrix contains smaller size sand grains, 10-70umdiameter (Fig. 2E), whereas P. mosulpia and P. mammiformis have sand grains of intermediate size, 350-600um diameter. SAND IN SPONGIN FIBRES. Sand grains in P wandoensis are attached to the inside of fibres, with grains approximately 50um diameter and uniformly distributed in a unidirectional plane. In this species the fibres are difficult to differentiate from the closely packed, amalgamated sand grains. In P. samyangensis the accumulation of large sand grains, 10-180um diameter, completely obscure the axis of primary fibres, and often a single foreign oxea connects adjacent primary fibres like a bridge, further supporting fibres. Rarely, smaller sand is distributed among the primary fibres (Fig. 1D, E). Psammocinia mosulpia has large sand grains, 100-400um diameter, contained within trans- parent fibres which are simple, not fasciculated (Fig. IF). Sand grains are attached in and outside of fibres. Psammocinia jejuensis has sand grains within its primary fibres that make up stout fasciculate columns. Sand grains measure 20-220um diameter. Fibres are easily torn. Secondary fibres may have a large single foreign oxea included, up to approximately 1,440um long and 801m wide, appears to support the sponge (Fig. 1G, H). These oxeas are unbroken within fibres. In P mammiformis there are thick, strong fibres with chain-like, small sand grains included in the centre of fibres (Fig. 11). Secondary fibres connected to large sand grains, form bridges between adjacent sand grains (Fig. 2F). Sands and fibres are tightly bound together (Fig. 2G). FILAMENTS AND FIBRES. In all five species of Psammocinia we observed that filaments emerge from pores on fibres (Figs 3A-F, 4A-H). MEMOIRS OF THE QUEENSLAND MUSEUM These filaments are visible under light microscopy, but are more clearly observed using SEM. In P. samyangensis fibre pore sizes vary greatly, apparently correlated to the thickness of filaments emerging. At their base several adjacent filaments fuse to form a central filament (Fig. 3B). At their terminal ends filaments are usually composed ofa single terminal knob (Figs 3C, 4H). The fibres seem to be in the form of a branch but we are able to confirm that there is an opening on the end of the branch from which the filament emerges (Figs 3E, F, 4D). DISCUSSION In Psammocina spongin fibres are thin and either simple or weakly fasciculated. As such, fibres probably do not provide sufficient support for the sponge body. Through the incorporation of sand grains into fibres and at the fibre core, sufficient structural integrity is achieved by these species. In addition to the rigidity received from association with sand grains, fibres of Psammocinia are also supported by a large, single oxea in many places within the mesohyl (Fig. 1E). Filaments also serve an important role in the sponge. If the sponge surface lacks a true sand crust, filaments produce a filamentous, cotton-like membrane at the surface. Together these structures provide some measure of skeletal support for the sponge, perhaps compensating for some inadequacy in their own organic skeletal elements. Bergquist (1995) stated that organic filaments were separated from spongin fibres, whereas we have shown that filaments emerge from the numer- ous pores throughout the fibre, and thus integral to the sponge fibre system. We also noted that several filaments emerge from a pore on the fibres (Fig. 3A), merge, and continue as a single filament ending in a terminal knob (Fig. 3D, E, TABLE 1. Main characteristics of the five Psammocinia species. 350-600um diameter chain-like Species Consistency Surface Mesohyl Fibre and Sand Filament . LATET ; ope "rs ae Small sand grains, Much small sand inside | Filament pores difficult Pwandaesris Very resilient; |; Thick: sand trust 10-120um diameter fibres to detect v m ; Medium sand grains, Sand in and outside of fi- | Filament pores difficult Fostaiuipia Resilient 7 Thin saríd crust 400-600um diameter bre to detect P. mammiformis Resiliént Sand'omst Medium sand grains, Small sand in fibre, Filament pores slightly visible P. jejuensis Hard but not re- silient, easily torn No sand crust, filamentous membrane Large sand grains, 700-4,0004m diameter Much sand in fibre Filament pores easily visible P. samyangensis Very soft, eas- ily torn No sand crust, filamentous membrane Large sand grains, 700-4,000um diameter Much sand in fibre Filament pores easily visible SAND & FIBRE IN PSAMMOCINIA un Un Un FIG. 3. A-F, Psammocinia samyangensis; A, many pores on a fibre (F, fibre; Fi, filament; P, pore); B, base of filaments and pores; C, filament emerging from pore on the fibre (T, terminal knob); D-E, filament and pore on a fibre; F, base of filament and pore (Scale bars: A, 100m; B-C, 20um; D-F, 10um). H). Very rarely, we noted filaments emerging Cook & Bergquist (1998) stated that the fibre skeleton in Psammocinia is supplemented at fine collagenous filaments, each enlarged terminally at both ends, whereas in the five species examined appear to emerge only from pores. here these terminal knobs appear at one end only. from both fibre pores and longitudinal slits along the fibres (Fig. 3C), but most commonly filaments MEMOIRS OF THE QUEENSLAND MUSEUM FIG 4, A-B, Psammocinia jejuensis, base of filament (F, fibre; Fi, filament; P, pore), C-E, Psammocinia mammiformis; filament and pore on a fibre. F, Psammocinia wandoensis, base of filament and pore. G-H, Psammocinia mosulpia; G, base of filament and pore; H, filament emerging from pore on a fibre (T, terminal knob) (Scale bars: A-B, 20um; C-E, 10m; F-H, 20m). Several questions still remain regarding the nature of the filaments of Psammocinia. One such question concerns the origin and development of the filaments along the fibres, which is a topic for further study. Another question concerns the quantity of filaments in relation to the quantity of fibre pores. In all sponges we examined we observed a large number of filaments, whereas there were far fewer pores from which the filaments emerged, and we assume, with empirical support from SEM studies, that a single pore can produce several filaments over time. Due to the complex morphology of the fibres and filament arrangement in Psammocinia, we were fortunate to observe the multi-based fila- ments (Fig. 3A) not previously described for this genus. Further studies are required, however, to SAND & FIBRE IN PSAMMOCINIA determine whether this type filament is exclusively a characteristic of Psammocinia, or is also found within other sponges of Irciniidae. As noted in Table 1, all species with a true sand crust are tough and it is difficult to observed filament pores given that so many sand grains are attached to fibre. The two species without a true sand crust are not tough, easily torn, have many filaments and many filament pores were observed. ACKNOWLEDGEMENTS This work was supported partly by the Basic Science Research Institute Programme, Korean Ministry of Education through Reasearch Fund (BSRI-97-4428). LITERATURE CITED BERGQUIST, P.R. 1980. A revision of the supraspecific classification of the orders Dictyo- ceratida, Dendroceratida and Verongida (Class Demospongiae). New Zealand Journal of Zoology 7: 443-503. 1995. Dictyoceratida, Dendroceratida and Verongida from the New Caledonia Lagoon (Porifera: 557 Demospongiae). Memoirs of the Queensland Museum 38(1): 1-51. COOK, S, DE C. & BERGQUIST, P.R. 1998, New species of dictyoceratid sponges Porifera: Demo- spongiae: Dictyoceratida) from New Zealand. New Zealand Journal of Marine and Freshwater Research 30:19-34. HOOPER, J.N.A. & WIEDENMAYER, F. 1994. Porifera. Pp. 1-626. In Wells, A. (ed.) *Zoological Catalogue of Australia’, Vol. 12. (CSIRO Australia: Melbourne). LENDEFELD, R. VON 1888. Descriptive catalogue of the sponges in the Australian Museum, Sydney. (Taylor & Francis: London). 1889, A monograph of the horny sponges. ( Trübner & Co.: London). POLEJAEFF, N, 1884. Report on the keratosa collected by H.M.S. “Challenger? during the years 1873-1876. Pp. 1-88. In Report on the Scientific Results of the voyage of H.M.S. ‘Challenger’ during the years 1873-76, Zoology. Vol. 11. (Her Majesty’s Stationery Office: London, Edinburgh, Dublin) SIM, C. J. 1998. Three new horny sponges ofthe genus Psammocinia (Dictyoceratida: Irciniidae) from Korea. Korean Journal of Systematic Zoology 14(1): 35-42. SIM, C. J. & LEE, K. J. 1998. New species of two Psammocinia Horny Sponges (Dictyoceratida: Irciniidae) from Korea. Korean Journal of Systematic Zoology 14(4): in press. 558 MEMOIRS OF THE QUEENSLAND MUSEUM CHONDROSIA RENIFORMIS: HITHERTO UNKNOWN BACTERIA. Memoirs of the Queensland Museum 44: 558. 1999:- Marine sponges as evolutionarily ancient Metazoa are nowadays in the focus of great interest. However, the implication of closely associated bacterial populations within the sponge tissue is completely unknown. One of the sponges investigated to clarify this relationship is the Mediterranean species Chondrosia reniformis. Bacteria associated with this sponge were examined by aerobic and anaerobic cultivation, culture-independent methods as whole cell in situ hybridization, PCR assisted rDNA sequence retrieval and comparative sequence analysis. /n situ hybridization of bacteria within the sponge tissue with fluorescently labeled rRNA-targeted oligonucleotides for the major subclasses of Proteobacteria revealed a great part of the population to be affiliated to the alpha-, gamma- and delta-subclasses. Interestingly, no organism was found to be a member of the beta-Proteobacteria. On the other hand, there are also many microorganisms that only gave signals with the universal probe for all bacteria, whereas group- or species-specific probes did not hybridize with these bacteria. Some of them are culturable and could successfully be characterised using the polyphasic approach. 16S rDNA-sequencing and subsequent analysis of the cultivated bacteria obtained from sponge-tissue showed that these metazoa are closely associated with a great diversity of hitherto unknown bacteria. Further studies of culturable sponge-associated bacteria are currently carried out to investigate the ecology of sponge-associated bacteria. This survey will elucidate the physiological properties and, by molecular analysis, the phylogenetical affiliation of these organisms. Occurrence and spatial distribution ofeven unculturable bacteria in sponge tissue will be analyzed by in situ hybridization with specifically designed rRNA-targeted oligonucleotides. O Porifera, Bacteria, in situ sequencing, phylogeny, PCR, microecology. G. Schumann-Kindel (email: schu0654@ mailszrz.zrz.tu-berlin.de), M. Bergbauer, W. Manz & U. zewzyk, Technical University Berlin, Dept. of Microbial Ecology, FranklinstraBe 29, 10587 Berlin, Germany; J. Reitner, IMPG, University of Göttingen, Goldschmidtstrafe 3, 37077 Göttingen, Germany; 1 June 1998. CYANIDE AND THIOCYANATE-BASED BIOSYNTHESIS IN TROPICAL MARINE SPONGES JAMIE S. SIMPSON AND MARY J. GARSON Simpson, J.S. & Garson, M.J. 1999 06 30: Cyanide and thiocyanate-based biosynthesis in tropical marine sponges. Memoirs of the Queensland Museum 44: 559-567. Brisbane. ISSN 0079-8835. The sponge Axinyssa n.sp. incorporates both sodium [!4C] cyanide and sodium [!4C] thiocyanate into 2-thiocyanatoneopupukeanane as well as into 9-isothiocyanatopupukeanane, however these 2 precursors were not incorporated into 9-isocyanopupukeanane. The specificity of incorporation into the thiocyanate carbon was confirmed by chemical degradation. Stylotella aurantium incorporates sodium [!4C] cyanide and sodium [!4C] thiocyanate into the dichloroimine functionality of the stylotellanes A and B, as well as into farnesyl isothiocyanate. The specificity of incorporation into the dichloroimine carbon atom was confirmed by chemical degradation. These experiments represent the first detailed study of the biosynthetic origin of organic thiocyanates and dichloroimines, and extend the range of functionality known to be biosynthesised from cyanide and thiocyanate. Our results raise the interesting question of the interconversion of inorganic cyanide and thiocyanate and/or the interconversion of the resulting organic metabolites in marine sponges. An isothiocyanate-isocyanide conversion was demonstrated in Amphimedon terpenensis by incorporation of a !4C-labelled sample of diisothiocyanatoadociane into diisocyanoadociane. O Porifera, Amphimedon terpenensis, Axinyssa, Stylotella aurantium, biosynthesis, cyanide, dichloroimines, isocyanides, isothiocyanates, secondary metabolites, terpenes, thiocyanates. Jamie S. Simpson & Mary J. Garson (email: garson(üchemistry.uq.edu.au), Department of Chemistry, University of Queensland, St Lucia 4072, Australia; 30 November 1998. Marine sponges of the order Axinellida, Halichondrida and Haplosclerida often contain bioactive terpenes with isocyanide, isothio- cyanate and formamide functionality; the rarer isocyanate and thiocyanate substituents are also known (Scheuer, 1992; Chang & Scheuer, 1993; Garson et al., 1998), These unique metabolites have been novel targets for study with '*C- and C-labelled precursors to determine the bio- synthetic origin of the non-terpenoid carbon atom (Garson, 1989; Chang & Scheuer, 1990; Garson, 1993; Garson et al., 1999). Work by our research group on the sponge Amphimedon terpenensis has shown that marine isocyanides such as diisocyanoadociane (Fig. 1 A) are derived by functionalisation of a terpene precursor with inorganic cyanide (Garson, 1986; Fookes et al., 1988). Karuso & Scheuer (1989) subsequently showed that both diterpene (eg. Fig. 1B) and sesquiterpene (eg. Fig. 1C) isocyanides are cyanide-derived, and further demonstrated the intact incorporation of the N,-C, unit. Our recent work with Acanthella cavernosa has shown the utilisation of cyanide for the biosynthesis of both a sesquiterpene isocyanide (Fig. 1D) and an isothiocyanate (Fig. 1 E) in this axinellid sponge. Furthermore inorganic thiocyanate was shown also to be a precursor to both the isocyanide and the isothiocyanate metabolites in this sponge. From these experimental results a biosynthetic link was inferred between the two inorganic precursors or between the two metabolite types (Dumdei et al., 1997). The origin of the thiocyanato group has been the subject of much biosynthetic speculation (Garson, 1993). Pham et al. (1991) suggested the cyanation of a thiol, which appears to be a reasonable pathway to the amino acid-derived psammaplin thiocyanate (Jimenez & Crews, 1991). In contrast, in those sponges in which thiocyanates co-occur with isocyanides or with isothiocyanates, the involvement of the ambident thiocyanate ion has been invoked (He et al., 1989; 1992; Walker & Butler, 1996). The dichloroimine (= carbonimidic dichloride) moiety represents a rare example of a functional group containing both nitrogen and carbon which has previously been found in terpene metabolites of the Indo- Pacific sponge Pseudaxinyssa pitys (Wratten & Faulkner, 1977; 1978a; 1978b). The coocurrence of an isothiocyanate together with dichloroimines in P pitys suggested to us the involvement of 560 D &-5Nc E R=Ncs FIG. 1. Structures of isocyanide and isothiocyanate metabolites investigated in biosynthetic experi- ments. A, diisocyanoadociane. B, kalihinol F. C, 2-isocyanopupukeanane. D, axisonitrile-3. E, axisothiocyanate-3. cyanide/thiocyanate in the biosynthesis of the dichloroimine group. In this paper we present the results of bio- synthetic experiments with the sponge Axinyssa n. sp. which provide evidence for a cyanide/ thiocyanate origin of the thiocyanato function- ality. We test the possibility using Stylotella aurantium that dichloroimine metabolites are biosynthesised from farnesyl pyrophosphate using cyanide or thiocyanate to supply the N¡-C; moiety. The role of inorganic thiocyanate and of an organic isothiocyanate in diisocyanoadociane biosynthesis are also explored. MATERIALS AND METHODS Abbreviations. GC-MS, gas chromatography- mass spectrometry; TLC, thin layer chromatography; NMR, nuclear magnetic resonance; HPLC, high performance liquid chromatography. Chemicals and biochemicals. Solvents used in the extraction of compounds from sponge samples were glass distilled. All radioactive precursors were purchased from Sigma Chemical Co. (St Louis, MO). Biological materials. Samples of Axinyssa n. sp. (Halichondrida: Halichondriidae), Stylotella aurantium (Halichondrida: Halichondriidae), Kelly-Borges & Bergquist, 1988, and Amphi- medon terpenensis (Haplosclerida: Niphatidae) Fromont, 1995, were collected using SCUBA at Coral Gardens, Experimental Gardens or Coral Spawning dive sites (12-16m depth), Heron MEMOIRS OF THE QUEENSLAND MUSEUM Island (23°27’S, 151?55'E) or at North Point (12-16m depth), Lizard Island (14?39'S, 145°27°E) on the Great Barrier Reef, Australia under permit numbers G96/050, G97/097, G98/037 and G98/227 issued jointly by the Great Barrier Reef Marine Park Authority and the Queensland National Parks and Wildlife Service. Sponge samples used in biosynthetic exper- iments were maintained in running seawater at ambient temperature and light conditions prior to use. Voucher specimens of the sponges Axinyssa n. sp., (accession number QMG312575), Stylo- tella aurantium (QMG307133) and Amphimedon terpenensis, (AMZ4978; QMG3 14228), are held at the Queensland Museum (QM), Brisbane or the Australian Museum (AM), Sydney. Isolation of metabolites. 1) Axinyssa n. sp. An or- ganic extract was prepared from frozen sponge (49.6g wet wt) and further purified by normal phase flash chromatography (gradient elution with hexanes/EtOAc) and normal phase HPLC using 0.25% EtOAc in hexanes to give (-)-9-isocyanopupukeanane (Fig. 2A; 107.6mg), (-)-9-isothiocyanatopupukeanane (Fig. 2B; 3.5mg), and (-)-2-thiocyanatoneopupukeanane (Fig. 2C; 31.4mg) together with smaller amounts of other isocyanides and isothiocyanates as described by Simpson et al. (1997b). 2) Stylotella aurantium. An organic extract was prepared from frozen sponge (204g wet wt) and further purified by normal phase flash chromatography (gradient elution with hexanes/EtOAc) and by normal phase HPLC using 0.2% EtOAc in hexanes to give stylotellane A (Fig. 2E; 9mg), stylotellane B (Fig. 2F; 75.6mg), and farnesyl isothiocyanate (Fig. 2G; 2mg) as described by Simpson et al. (1997a). 3) Amphimedon terpenensis. Diisocyanoadociane (Fig. 1A; l6mg) was isolated from frozen sponge (25g wet wt) as described by Fookes et al. (1988). Biosynthetic experiments. 1) Pieces of Axinyssa n. sp. (approx. 80g wet wt) were placed in an aquarium containing 200ml aerated seawater at ambient temperature (20-23*C). Sodium ['*C] cyanide (100uCi) or sodium ['*C] thiocyanate (Q5uCi) was added and the sponge allowed to assimilate radioactivity for 12hr. The sponge was kept in running seawater in a 10 litre aquarium at ambient temperature for 16 days, then frozen for subsequent radiochemical analysis. Metabolites were purified according to the above protocol. The radioactivity content was monitored at each stage of the purification sequence, and terpenes were subjected to repeated HPLC until the specific activity was constant. 2) Stylotella aurantium (24g BIOSYNTHESIS IN TROPICAL MARINE SPONGES wet wt) was placed in an R aquarium containing 200ml aerated seawater at ambient Lee (20-23°C), Sodium C] cyanide (50uCi) was Lael and the sponge allowed io E assimilate radioactivity for 12hr overnight, The sponge was kept Her e e R in running seawater in a 10L n aquarium at ambient temp- : Fes 9 erature for 9 days, then lrozen ; H R-NHCO CH JR NI for subsequent radiochemical n = analysis. Metabolites were pür- B. ified according lo the above — * C R-SCN protocol. The radioactivity con- DK -5H x x »: a tent was monitored at each stage A G R-NCS of the purification sequence. sodium da C] thiocyanate (I3uCi; 9 days incorporatior) experiment, used a 12g piece of sponge. 3) Amphimedon ter- penensis (26g wet wt) was placed in an aquarium containing 400mL acrated seawater at ambient temperature (20- -23*C) ["C]- Diisothiocyanatoadociane (11101) was added and the sponge allowed to assimilate radioactivity for I2hrs overnight. The sponge was kept in running seawater in a 20L aquarium at ambient temperature for 19 days, then frozen for subsequent radiochemical analysis. Metabolites were purified according io the above protocol. The radioactivity content was monitored at cach Stage of the purification sequence, A sodium [^C] thiocyanate (50nCi; 19 days incorporation) experiment used à 45g piece of sponge. Procedures used in the synthesis of ["C]- diisothiocyanatoadociane will be described elsewhere (Simpson & Garson, in preparation), FIG, RESULTS 1) Axinyssa n.sp. collected at Heron 1. contained sesquiterpene metabolites by GC-MS, TLC and NMR; the hexane-solubles were processed as described in Simpson et al. (1997b) to give the 9-pupukeanane isocyanide/isothiocyanale pair (Fig. 2A,H) and 2-thiocyanatoneopupukeanane (Fig. 2C). The GC-MS profile of the sesqui- terpene fraction showed a number of other peaks including isocyanides and isothiocyanates. Light and electron microscopic inspection of Axinyssa n.sp. revealed ihe presence of microbial sym- bionts. The outer layers of sponge tissue were rich in cyanobacteria of a type morphologically similar lo Aphanocupsa feldmanni while the . Structures of terpenes isolated from Jximyssa.sp., dutan iail and degradation products, A, 9-isocvanopupukeanane. P 0-isothioeyanatopupukeanane: C, thiol from 2-thiocyanatoneopupukeanane. E, stylotellane A, F, stylotellane B. G, farnesyl isothiocyanate. G. methyl carbamate From stylotellane B. 1, amine from stylotellane B Styluiella 3-thiocyanatoneopupukeanatie, Et, inner tissue contained high populations of diverse bacterial cell types in addition to sponge cells. An Archaea-like symbiont with a membrane-bound nucleoid was found (see Fuerst et al., this volume; Fuerst et al., 1998). 25yCi Sodium [C] thiocyanate was supplied to a specimen of Axinyssa n.sp. maintained in a small aquarium (Dumdei et al., 1997; Simpson & Garson, 1998). After 16 days aquarium incu- bation, the sponge sample was frozen and 2-thiocyanatoneopupukeanane (Fig. 2C) was isolated and rigorously purified by HPLC tu constant specific radioactivity. The thioeyanate (Fig. 2C) was significantly radioactive, as shown in Table |, consistent with the use of thiocyanate for the biosynthesis of the thiocyanato group as shown in Fig, 3. To test the specificity of in- corporation, 2-thiocyanatoncopupukcanane (Fig, 2C) was degraded to the thiol (Fig. 2D) using LiAIH4. The thiol product Was not radio- active (Table 2) therefore ihe [C] label was exclusively associated with the thiocyanato carbon, Incorporation of sodium [PC] cyanide into a second piece of sponge also gave radio- active 2-ihiocyanaloneopupukeanane (Table 1), Degradation resulted in unlabelled thiol product (Table 2) indicating (he label was again ex- clusively associated with the thnocyanato motety. Our experiments also allowed us to monitor isocyanide/isothiocyanate biosynthesis in this sponge. When the isocyanide/isothiocyanale pair 562 TABLE 1. Molar specific activities of Axinyssa n. sp. metabolites. a, published incorporation values were not percentage values (Simpson & Garson, 1998); b, incorporation of 25uCi; c, <10-2 %; d, incorporation of 100uCi. Compound Precursor et Incorporation" (Fig. no.) (mCi/mMole) (9) 2A Na[ "C]SCN? 0.004 l 2B Na[ C]SCN" 2.630 0.08 2C Na[ ^C]SCN^ 0.150 0.02 2A Na[ "C]CN* 0.014 g 2B Na[ C]CN? 13.900 0.3 2C Na[ "C]CN? 1.230 0.2 (Fig. 2A,B) were isolated, the isothiocyanate samples were radioactive (>150,000dpm/mg), whereas the isocyanide samples from both thiocyanate and cyanide feedings were not significantly labelled («100dpm/mg). The specificity of labelling of 9-isothiocyanato- pupukeanane is currently under investigation. 2) Extracts of the sponge Stylotella aurantium weakly inhibited the growth of a P388 mouse leukaemia cell line and contained terpenes by TLC and NMR. The DCM-soluble components of the extract were processed as described in Simpson et al. (19972) to give the stylotellanes A and B (Fig. 2E,F), together with farnesyl isothiocyanate (Fig. 2G). Light microscopic in- spection of sponge tissue revealed the absence of microbial symbionts other than bacteria. 50uCi Sodium ["C] cyanide was supplied to a specimen of S. aurantium according to our established protocols (Dumdei et al, 1997; Simpson et al., 19972). After 9 days aquarium incubation, the sponge sample was frozen and stylotellanes A and B were isolated and rigorously purified by HPLC to constant specific radioactivity. The samples of stylotellanes A and B (Fig. 2E,F) were significantly radioactive, as . SCN, FIG. 3. Incorporation into isothiocyanate and thiocyanate metabolites of Axinyssa n. sp. MEMOIRS OF THE QUEENSLAND MUSEUM TABLE 2. Molar specific activities of Axinyssa n. sp. degradation products. Compound | precursor | ivi | Radioastivity (mCi/mMole) 2C Na[^C]SCN 0.150 100.0 2D Na[ ^C]SCN 0.001 0.3 2C Na["C]CN 1.230 100.0 2D Na[ C]CN 0.001 0.1 shown in Table 3, consistent with the use of cyanide for the biosynthesis ofthe dichloroimine group (Fig. 4, route notation ‘a’). The percentage incorporation levels measured were low as a result of loss of volatile metabolites during the purification process combined with the chemical instability ofthe dichloroimine group. To test the specificity of incorporation, stylotellane B was degraded to the methyl carbamate (Fig. 2H) and the amine (Fig. 21) using 0.1N phosphoric acid in 95% methanol. The carbamate product was radioactive, whereas the amine was devoid of radioactivity (Table 4), therefore the ['*C] label was exclusively associated with the imine carbon. Incorporation of sodium ['*C] thio- cyanate into a second piece of sponge also gave radioactive metabolites (Table 3), however there was insufficient material for chemical degrad- ation. In each experiment, the isolated farnesyl isothiocyanate (Fig. 2G) was also radioactive. 3) 50uCi Sodium ['*C] thiocyanate was then supplied to a specimen of 4. ferpenensis according to our established protocols (Fookes et al., 1988; Dumdei et al., 1997). After 19 days aquarium incubation, the sponge sample was frozen and diisocyanoadociane isolated and rigorously purified by HPLC, then recrystallised to constant specific radioactivity. The sample was significantly radioactive consistent with the use of thiocyanate for the biosynthesis of the isocyanide group (Fig. 5). Degradative exper- iments are in progress to confirm the specific labelling. When a sample of diisothiocyanatoadociane, "C-labelled in both isothio- cyanate groups, was provided to A. terpenensis, the diisocyanoadociane isolated was found to be radioactive. The specificity of labelling is under investigation. BIOSYNTHESIS IN TROPICAL MARINE SPONGES 563 TABLE 3. Molar specific activities of S. aurantium metabolites. a, incorporation of 50uCi; b, incorporation of 13uCi. TABLE 4. Molar specific activities of S. aurantium degradation products. a, after dilution with unlabelled metabolite. Compound | precursor | Cae | corporation (mCi/mMole) 2E Na[ ^C]CN* 1.36 | 0.004 2F Na[ C]CN* 1472 0.033 2E Naf "C]SCN^ 0.354 0.00034 2F Nal ^C]SCN" 0.224 0.00056 DISCUSSION Our biosynthetic experiments with Axinyssa n. sp. and with S. aurantium, together with the earlier work on A. ferpenensis and A. cavernosa (Garson, 1986; Fookes etal., 1988; Dumdei etal., 1997), reveal that cyanide and thiocyanate are precursors involved in the biosynthesis of four N;-C; functional groups found in marine ter- penes, namely isocyanides, isothiocyanates, thiocyanates and dichloroimines. A number of different biosynthetic pathways can be invoked to explain the origin of the thio- cyanate group. Pham et al. (1991) suggested the cyanation of a terpene thiol, however this proposal does not adequately explain the co- occurrence of thiocyanates and isothiocyanates in the same sponge. The insertion of sulphur into an organic cyanide or isocyanide to give a thiocyanate is mechanistically unprecedented. Sulphur insertion into an isocyanide to give an isothiocyanate (Hagadone et al., 1984), perhaps using an enzyme functionally equivalent to rhodanese (Westley, 1973), followed by iso- merisation of the isothiocyanate to the thiocyanate is a plausible biosynthetic pathway. In the laboratory however, the isothiocyanate- thiocyanate equilibrium usually favours an isothiocyanate over a thiocyanate (Hughes, 1975). A final biosynthetic possibility is the use of an ambident thiocyanate anion to attack a terpene carbenium ion or its functional equiv- alent (He et al., 1989; 1992; Walker & Butler, 1996). Thiocyanate either reacts through the nitrogen centre generating an isothiocyanate derivative or through the sulphur generating a thiocyanate. Results on the biosynthesis of the thiocyanate moiety are particularly informative. The incorp- oration of inorganic thiocyanate into the thiocyanate and isothiocyanate metabolites (Fig. 2B,C) of Axinyssa n. sp. is consistent with direct Compound Molar specific Radioactivity (Fig, no.) Precursor activity (%) oo - (mCi/mMole) 2F Na[ ^C]CN 0.332" 100.0 | 2H Na[C]CN 0.326 982 | 21 Na[^C]CN 0.004 1.2 utilisation of this ambident precursor. Likewise cyanide is utilised for both thiocyanate and iso- thiocyanate biosynthesis. Our proposal is that cyanide is converted in Axinyssa n. sp. to thio- cyanate by the action of an enzyme similar to rhodanese (Scheivelbein et al., 1969; Westley, 1973) and then incorporated into either 9-isothiocyanatopupukeanane or 2-thiocyanato- neopupukeanane. The alternative possibility that cyanide is converted first to the isocyanide then by sulphur insertion to an isothiocyanate is less likely since cyanide was not utilised for the bio- synthesis of 9-isocyanopupukeanane in the same specimen of Axinyssa n. sp. Stylotella aurantium uses both cyanide and thiocyanate as precursors for the biosynthesis of the dichloroimine and isothiocyanate groups. Fig. 4 shows two plausible biosynthetic routes to the stylotellanes A and B, one route (a) using an isonitrile intermediate and the other (b) invoking an isothiocyanate intermediate. The isolation of farnesyl isothiocyanate, but not of farnesyl isocyanide (Fig. 4A), from this sponge is con- sistent with the operation of path (b). The dichloroimine metabolites are among the most unusual of the cyanide and thiocyanate-derived terpenes. In the laboratory, isocyanide dihalides can be synthesised by addition of chlorine to isocyanides or by chlorination of isothiocyanates (Kühle et al., 1967). The biosynthetic mechanisms proposed invoke the use of a chloroperoxidase enzyme to chlorinate intermediates (Butler & Walker, 1993; Walker & Butler, 1996). In A. terpenensis, our results are consistent with the use of both thiocyanate and of cyanide for isocyanide biosynthesis. Thiocyanate may perhaps be converted to cyanide by use of a peroxidase enzyme, as has been demonstrated in some bacteria (Ohkawa et al., 1971; Pollock & Goff, 1992; Westley, 1981), which is then utilised for isocyanide biosynthesis (Fig. 5). We have previously suggested that A. cavernosa is able to interconvert inorganic cyanide and thiocyanate (Dumdei et al., 1997). Our current pA SN m X ES ES ES A pe | A ENS us us * N=c—0| Su SS [e] = = A FIG. 4. Biosynthesis of dichloroimines in Stylotella aurantium. A, farnesyl B, farnesyl isothiocyanate. C, stylotellane A. D, stylotellane B. isocyanide. results suggest Axinyssa n. sp., S. aurantium and A. terpenensis are able to interconvert these 2 inorganic precursors. We have also speculated that enzymic transformations which parallel the cyanide-thiocyanate interconversion may transform organic isocyanides into isothio- cyanates, or the reverse, in marine sponges (Dumdei et al., 1997). Figure 6 illustrates these suggested biosynthetic relationships for Axinyssa n.sp. Thiocyanate is used to make 9-isothio- cyanatopupukeanane which then undergoes desulphurisation to give 9-isocyanopupukeanane; alternatively, cyanide is used for isocyanide bio- synthesis, then the isocyanide is converted into the isothiocyanate by an enzyme functionally equivalent to rhodanese. In pioneering biosynthetic experiments, Hagadone et al. (1984) inferred the precursor status of an isocyanide terpene metabolite in isothiocyanate formation in Ciocalypta sp. They explored the in vivo conversion of 2-iso- cyanpupukeanane into the corresponding formamide and isothiocyanate metabolites. The natural product status of formamide metabolites has however been questioned by Tada et al. (1988). A second concern with the work of Hagadone et al. (1984) is their use of the relatively insensitive "C label in conjunction with mass spectrometric detection. Our preliminary results with A. ferpenensis suggest that an isothiocyanate to isocyanide transformation may occur in this sponge. The MEMOIRS OF THE QUEENSLAND MUSEUM sponge contains isothio- cyanates as minor met- abolites (unpublished results). When radiolabel- led diisothiocyanatoadociane was supplied to samples of A. terpenensis, the diisocyanoadociane isolated was shown to be radioactive. Chemical degradation is currently in progress to confirm the specificity of labelling in this advanced precursor experiment. In view of the previous successful experiments with both diterpene isocyanides (Garson, 1986; Fookes et al., 1988; Karuso & Scheuer, 1989) and sesquiterpene isonitriles (Karuso & Scheuer, 1989; Dumdei et al., 1997), it 1s quite extraordinary that we have not demonstrated the incorporation of cyanide into the major iso- cyanide component (Fig. 6) of Axinyssa n. sp. Likewise thiocyanate appears not to be used for isocyanide biosynthesis in this sponge, in contrast to A. cavernosa in which thiocyanate is used for isocyanide biosynthesis (Dumdei et al., 1997) and also in contrast to A. terpenensis (this paper). The lack of incorporation of thiocyanate into 9-isocyanopupukeanane (Fig. 6) suggests that either the thiocyanate to cyanide conversion is inefficient in this sponge or that the conversion of isothiocyanate into isocyanide does not occur. We are currently isolating some of the other minor isocyanide metabolites from Axinyssa samples labelled by thiocyanate or cyanide in order to investigate the role of cyanide and thiocyanate in isocyanide biosynthesis in this sponge. A clearer picture of the complex meta- bolic interrelationships in Axinyssa n. sp. will emerge when we test the utilisation of CN- SCN- FIG. 5. Biosynthesis of diisocyanoadociane in A. terpenensis. BIOSYNTHESIS IN TROPICAL MARINE SPONGES 56 9-isothiocyanatopupukeanane 9-isocyanopupukeanane FIG. 6. Possible biosynthetic interconversions in Axinyssa n. sp. Solid lines indicate incorporation results or possible conversions, dotted lines indicate non-incorporation. C-labelled isocyanide and isothiocyanate precursors by this sponge. The origin of the cyanide or the thiocyanate used by marine sponges remains a tantalising mystery. Plants generate hydrogen cyanide by hydrolysis of cyanogenic glycosides (Seigler, 1975). Some bacteria are known to produce hydrogen cyanide (Knowles, 1976) or to convert the amino acid cysteine to thiocyanate (Voet & Voet, 1995), while methionine has been implicat- ed in the formation of cyanide as a byproduct of ethylene biosynthesis (Pirrung, 1985). To date experiments to determine an amino acid origin for the isocyano group in diisocyanoadociane have been unsuccesful (Fookes et al., 1988). Two sponges used in our biosynthetic ex- periments have interesting symbiotic profiles. Amphimedon terpenensis has previously been shown to contain high bacterial populations of eubacteria together with a cyanobacterial symbiont which morphologically resembles Aphanocapsa feldmanni (Garson et al., 1992). Axinyssa n.sp. contains a cyanobacterial sym- biont together with numerous bacteria, in particular an archaeal-like bacteria which con- tains a highly unusual membrane-bound nucleoid (Fuerst, this volume; Fuerst et al., 1998). We have previously demonstrated that 4. terpenensis isocyanides are localised in sponge cells, primarily archaeocytes and choanocytes, and infer that this is the site of synthesis of the metabolites (Garson et al., 1992). Terpene metabolites in 2 other sponges have been shown to be localised in sponge cells rather than sym- biont cells (Uriz et al., 1996; Flowers et al., 1998). The range of sponges with which we are un now exploring N;-C, biosynthesis provide us with additional candidates to study the cellular localisation ofterpene metabolites and to explore the role of symbionts in biosynthesis. TAXONOMIC NOTE. The sponge which we have identified as ‘Amphimedon’ terpenensis in this paper has a chequered taxonomic history. It was first named in the literature as an Adocia sp. by the Roche group (Baker et al., 1976). Fromont (1993) placed the sponge within Amphimedon in her taxonomic studies on haplosclerid sponges of the Great Barrier Reef and proposed the species name for the large proportion of terpene metabolites. Van Soest et al. (1996) considered the skeletal characteristics were too irregular to be compatible with Amphimedon. Based on structural characteristics and spicule analysis, they proposed the combination Cymbastela terpenensis, but acknowledged however that the skeletal morphology, growth form and texture for the sponge were not typical of Cymbastela, as described by Hooper & Bergquist (1992). The documented secondary metabolite chemistry of Cymbastela spp. consists of pyrrole metabolites from a New Caledonian species (Ahond et al., 1988). Samples of Cymbastela sp. collected from Heron I. and Lizard I. do not have a secondary metabolite profile by NMR and GC-MS, but have been shown to contain 24-isopropyl-& sterols (Stoilov et al, 1986), whereas A. terpenensis contains 5°’-sterols (Garson et al., 1988). A more thorough taxonomic assessment of “4.” ter- penensis (and the related C. hooperi), including consideration of live specimen characteristics, growth form, texture and spongin content and skeletal structure is required. It is possible that a new genus is required for these species but this requires substantially more corroborative evidence than is presently available (e.g. genetic analyses). For the present we retain the taxon ' A." terpenensis, but acknowledge it does not belong with typical members of Amphimedon (Haplo- sclerida; Niphatidae). ACKNOWLEDGEMENTS We thank John Hooper, Queensland Museum, for taxonomy, the Australian Research Council for funding and the Australian Government for an APRA scholarship. The assistance of the staff of Heron Island and Lizard Island Research Stations in performing field work is gratefully acknow- ledged. This research was performed under permits 96/050, G97/097, G98/037 and G98/227 issued jointly by GBRMPA and QNPWS. LITERATURE CITED AHOND, A., ZURITA, M.B., COLIN, M., FIZAMES, C., LABOUTE, P., LAVELLE, F., LAURENT, D., POUPAT, C., PUSSET, M. & THOISON, D. 1988. A new antitumoural compound isolated from the sponge Pseudaxinyssa cantharella sp. nov. (Axinellidae). Comptes Rendus de l'Académie des Sciences Série II. 307: 145-148. BAKER, J.T., WELLS, R.J., OBERHANSLI, W.E. & HAWES, G.B. 1976. A new diisocyanide of novel ring structure from a sponge. Journal of the American Chemical Society 98(13): 4010-4012. BUTLER, A. & WALKER, J.V. 1993. Marine halo- peroxidases. Chemical Reviews 93: 1937-1944. CHANG, C.W.J. & SCHEUER, P.J. 1990. Biosynthesis of marine isocyanoterpenoids in sponges. Comparative Biochemistry and Physiology 97B: 227-233. 1993, Marine isocyano compounds. Pp. 34-75. In Scheuer, P.J. (ed.) Marine Natural Products. Diversity and Biosynthesis: Topics in Current Chemistry. Volume 167. (Springer-Verlag: Berlin and New York). DUMDEI, E.J., FLOWERS, A.E., GARSON, M.J. & MOORE, C.J. 1997. The biosynthesis of sesqui- terpene isocyanides and isothiocyanates in the marine sponge Acanthella cavernosa (Dendy); evidence for dietary transfer to the dorid nudi- branch Phyllidiella pustulosa. Comparative Biochemistry and Physiology 118A: 1385-1392. FLOWERS, A.E., GARSON, M.J., WEBB, R.L., DUMDEL E.J. & CHARAN, R.D. 1998. Cellular origin of chlorinated diketopiperazines in the dictyoceratid sponge Dysidea herbacea (Keller). Cell and Tissue Research 292: 597-607. FOOKES, C.J.R., GARSON, M.J., MACLEOD, J.K., SKELTON, B.W. & WHITE, A. H. 1988. Biosynthesis of diisocyanoadociane, a novel diterpene from the marine sponge Amphimedon sp.: crystal structure of a monoamide derivative. Journal of the Chemical Society Perkin Trans- actions I: 1003-1011. FROMONT, J. 1993. Descriptions of the Haplosclerida (Porifera: Demospongiae) occurring in tropical waters of the Great Barrier Reef. The Beagle, Records of the Northern Territory Museum of Arts and Science 10(1): 7-40, FUERST, J.A., WEBB, R.L., GARSON, M.J., HARDY, L. & REISWIG, H.M. 1998. Membrane-bounded nucleoids in microbial symbionts of marine sponges. FEMS Microbiology Letters 166: 29-34. GARSON, M.J. 1986. Biosynthesis of the novel diterpene isonitrile diisocyanoadociane by a marine sponge of the Amphimedon genus: Incorporation studies with sodium ['*C] cyanide and sodium [2-'*C] acetate. Journal of the Chemical Society Chemical Communications 35-36. 1989. Biosynthetic studies on marine natural products. Natural Product Reports 6: 143-170. MEMOIRS OF THE QUEENSLAND MUSEUM 1993, The biosynthesis of marine natural products, Chemical Reviews 93: 1699-1733. GARSON, M.J., PARTALI, V., LIAAEN-JENSEN, S. & STOILOV, I.L. 1988. Isoprenoid biosynthesis in a marine sponge of the Amphimedon genus. Incorporation studies with [1-' ^C] acetate, [4- C] cholesterol and [2- C] mevalonate. Comparative Biochemistry and Physiology 91B(2): 293-300. GARSON, M.J., THOMPSON, J.E., LARSEN, R.M., BATTERSHILL, C.M., MURPHY, P.T. & BERGQUIST, P.R. 1992. Terpenes in sponge cell membranes: cell separation and membrane fractionation studies with the tropical marine sponge Amphimedon sp. Lipids 27: 378-388. GARSON, M.J., SIMPSON, J.S., FLOWERS, A.E. & DUMDEI E.J. 1999. Cyanide and thiocyanate- derived functionality in marine organisms — structure, biosynthesis and ecology. Studies in Natural Products, in press. HAGADONE, M.R., SCHEUER, PJ., & HOLM, A. 1984. On the origin of the isocyano function in marine sponges. Journal of the American Chemical Society 106: 2447-2448. H.-Y., FAULKNER, D.J., SHUMSKY, J.S., HONG, K. & CLARDY, J. 1989. A sesquiterpene thiocyanate and three sesquiterpene isothio- cyanates from the sponge Trachyopsis aplysinoides. Journal of Organic Chemistry 54: 2511-2514. HE, H.-Y., SALVA, J., CATALOS, R.F. & FAULK- NER, D.J. 1992. Sesquiterpene thiocyanates and isothiocyanates from Axinyssa aplysinoides. Journal of Organic Chemistry 57: 3191-3194. HOOPER, J.N.A. & BERGQUIST, P.R. 1992. Cymbastela, a new genus of lamellate coral reef sponges. Memoirs of the Queensland Museum 32(1): 99-137. HUGHES, M.N. 1975, General chemistry. Pp. 3-67. In Newman, A.A. (ed.) Chemistry and biochemistry of thiocyanic acid and its derivatives. (Academic Press: London). JIMENEZ, C. & CREWS, P. 1991. Novel marine sponge derived amino acids. 13. Additional psammaplysin derivatives from Psammaplysilla purpurea. Tetrahedron 47; 2097-3102. KARUSO, P. & SCHEUER, P. J. 1989. Biosynthesis of isocyanoterpenes in sponges. Journal of Organic Chemistry 54: 2092-2095. KELLY-BORGES, M., & BERGQUIST, P.R. 1988. Sponges from Motupore Island, Papua New Guinea. Indo-Malayan Zoology 5: 121-159. KNOWLES, C.J. 1976. Microorganisms and cyanide. Bacteriological Reviews 40(3): 652-680. KUHLE, E., ANDERS, B. & ZUMACH, G. 1967. New methods of preparative organic chemistry IV. Syntheses of isocyanide dihalides. Angewandte Chemie International Edition 6: 649-653. OHKAWA, H & CASIDA, J.E. 1971. Glutathione- S-transferases liberate hydrogen cyanide from organic thiocyanates. Biochemical Pharmacology 20: 1708-1711. HE, BIOSYNTHESIS IN TROPICAL MARINE SPONGES PHAM, A.T., ICHIBA, T., YOSHIDA, W., SCHEUER, P.J., UCHIDA, T., TANAKA, J., & HIGA, T. 1991. Two marine sesquiterpene thiocyanates. Tetrahedron Letters 32: 4843-4846, PIRRUNG, M.C. 1985. Ethylene biosynthesis. 3. Evidence concerning the fate of C¡-N, of l-aminocyclopropanecarboxylic acid. Bioorganic Chemistry 13: 219-226. POLLOCK, J.R. & GOFF, H.M. 1992. Lactoper- oxidase-catalysed oxidation of thiocyanate ion. A carbon-13 nuclear magnetic resonance study of the oxidation products. Biochimica Biophysica Acta 1159: 279-285. SCHEUER, P.J. 1992. Isocyanides and cyanides as natural products. Accounts of Chemical Research 25: 433-439, SCHIEVELBEIN, H., BAUMEISTER, H. & VOGEL, R. 1969. Comparative investigations on the activity of thiosulphate-sulphur transferase. Naturwissenschaften 56: 416-417. SEIGLER, D.S. 1975. Isolation and characterisation of naturally occurring cyanogenic compounds. Phytochemistry 14: 9-29, SIMPSON, J.S., RANIGA, P. & GARSON, M.J. 1997a. Biosynthesis of dichloroimines in the tropical marine sponge Stylotella aurantium. Tetrahedron Letters 38(45): 7947-7950. SIMPSON, J.S. GARSON, M.J., HOOPER, J.N.A., CLINE, E.I. & ANGERHOFER, C.K. 1997b. Terpene metabolites from the tropical marine sponge Axinyssa sp. nov. Australian Journal of Chemistry 50:1123-1127. SIMPSON, J.S. & GARSON, M.J. 1998. Thiocyanate biosynthesis in the tropical marine sponge Axinyssa n. sp. Tetrahedron Letters 39: 5819- 5822. SOEST, R.W.M. VAN, DESQUEYROUX- FAUNDEZ, R., WRIGHT, A.D. & KONIG, G.M. 1996. Cymbastela hooperi sp. nov. (Halichondrida: Axinellidae) from the Great Barrier Reef, 567 Australia. Bulletin de l'Institut Royal des Sciences Naturelles de Belgique 66: 103-108. STOILOV LL., THOMPSON, J.E. & DJERASSI, C. 1986. Biosynthetic studies of marine lipids 7. Experimental demonstration of a double alkylation at C-28 in the biosynthesis of 24-iso- propylcholesterols in a sponge. Tetrahedron 42(15): 4147-4160. TADA, H., TOZYO, T. & SHIRO, M. 1988. A new isocyanide from a sponge. Is the formamide a natural product? Journal of Organic Chemistry 53: 3366-3368. URIZ, M.J., TURON, X., GALERA, J. & TUR, J.M. 1996. New light on the cell location of avarol within the sponge Dysidea avara (Dendro- ceratida). Cell and Tissue Research 285: 519-527. VOET D. & VOET J.G. 1995. Biochemistry. 2nd Edition (Wiley: New York). WALKER, J.V. & BUTLER, A. 1996. Vanadium bromoperoxidase-catalysed oxidation of thiocyanate by hydrogen peroxide. Inorganica Chimica Acta 243: 201-206. WESTLEY, J. 1973. Rhodanese. Advances in Enzym- ology 39: 327-368. 1981. Cyanide and sulfane sulfur. Pp. 61-75. In Vennesland, B., Conn, E.E., Knowles, C.J., Westley, J. & Wissing, J. (eds) Cyanide in Biology. (Academic Press: London). WRATTEN, S.J. & FAULKNER, D.J. 1977. Carbo- nimidic dichlorides from the marine sponge Pseudaxinyssa pitys. Journal of the American Chemical Society 99: 7367-7368. WRATTEN, S.J., FAULKNER, D.J., VAN ENGEN, D. & CLARDY, J. 1978a. A vinyl carbonimidic dichloride from the marine sponge Pseudaxinyssa pitys. Tetrahedron Letters 16: 1391-1394. WRATTEN, S.J. & FAULKNER, D.J. 1978b. Minor carbonimidic dichlorides from the marine sponge Pseudaxinyssa pitys. Tetrahedron Letters 16: 1395-1398. 568 MEMOIRS OF THE QUEENSLAND MUSEUM REGENERATION ABILITIES OF SPONGILLID LARVAE. Memoirs of the Queensland Museum 44: 568. 1999:- Free-swimming larvae of Spongilla lacustris were cut into two halves with a razor blade under a binocular microscope. Two samples of the larvae were used. In the first sample, larvae were cut in the tangential plane (two equal halves with a similar set of cells and structures). In the second sample, larvae were cut in the transverse plane (two unequal parts). The ‘anterior’ fragment contained a large cavity lined with pinacocytes, the halved amount of the surface flagellated cells, and underlying collencytes. The ‘posterior’ part of the larva had a halved amount of the surface flagellated cells and all of the internal structures typical for the fully developed spongillid larvae. Each half of the larva was maintained in a separate Petri dish with the celloidine-covered bottom in well-aerated river water. Visual observations and transmission electron microscopy yielded the following preliminary results. The halves of the larvae closed the edges of the wound immediately after dissection while continuing to move. However, trajectory, velocity and direction of the movements differed in different types of experimentally cut larvae. This was directly related to the presence or absence and the development of the larval cavity. Thirty minutes following the dissection, the tangentially split halves of the larvae looked normal (movement, attachment and metamorphosis generally similar), but half the size of the control larvae. In 2-3 days after the settlement these half-larvae metamorphosed into normally functioning small sponges. Developmental capabilities of the transverse halved larvae were different. The anterior half-larvae soon closed the cut edges, acquired a shape of a hollow sphere and swam easily and rapidly in the water. They maintained activity for two or more days, attached, formed pinacoderm and few flagellated chambers, and the sponges died. The posterior halves recovered the integrity of the flagellar cover in an hour following the dissection. They acquired the shape of spheres tightly packed with cells, covered with slightly elongated flagellated cells, and swam heavily and slowly near the bottom, with a maximum free life of 18 hours. After the settlement and attachment they formed pupae covered with pinacoderm and within 2 days developed into normal sponges. Transmission electron microscopy showed the surface flagellated cells played an important role. These cells provided restoration of the surface cell cover, however, their role greatly differed in development of the ‘anterior’ and ‘posterior’ half-larvae. In the ‘anterior’ halves, the flagellated cells migrated inside; some were ingested by underlying collencytes (phagocytosis); some transformed into choanocytes giving rise to few flagellated chambers. During development of the ‘posterior’ half- larvae, some surface flagellated cells transformed into the pinacocyte-like cells in situ and still retained flagella for a long time. The leading role in the transformation of the flagellated cells belonged to the centrioles (both flagellated and flagellum-less) and to the root structures of flagellum connected with the centrioles. Collencytes played the important role in the attachment and development of the settled half-larvae. These cells actively migrated to the surface of the settled larva, phagocyted the cells damaged during dissection, secreted a large amount of collagen and contributed to the flattening of the half-larvae and their attachment to the substrate. The post- settlement fate of flagellated surface cells of the half-larvae was partially dependent on the amount and the activity of collencytes. The next major morphogenetic role belonged to archaeocytes, the main source for the formation of choanoblasts, spiculocytes, collencytes, pinacocytes and other cells. The archaeocytes mitotically divided several times, losing their storage inclusions, and thus gave rise to several differentiated cell lineages. Probably, the lack of the necessary amount of the cells is responsible for the developmental retardation of the settled ‘anterior’ half-larvae. O Porifera, Spongillidae, larva, metamorphosis, development, ontogeny, transmission electron microscopy. VV. Semenov, Biological Institute of St. Petersburg State University, Oranienbaumskoje shosse, 2, Stary Petergof, St. Peterburg, 198904, Russia; L.V. Ivanova (email: inna(a)sokolzoo.spb.su), Pedagogical State University named after A.I.Herzen, River Moika Emb., 48, St.Petersburg, 191186, Russia; 1 June 1998. CHEMOSYSTEMATICS OF PORIFERA: A REVIEW R.W.M. VAN SOEST AND J.C. BRAEKMAN Soest, R.W.M. van & Braekman, J.C. 1999 06 30: Chemosystematics of Porifera: a re- view. Memoirs of the Queensland Museum 44: 569-589. Brisbane. ISSN 0079-8835. All compounds isolated from Porifera were reviewed in an attempt to discover what level of reliability may be attached to chemistry data when applied to sponge systematics. To date (May 1998) more than 3500 different compounds have been described from 475 species of marine sponges, belonging to two of the three classes (Calcarea and Demospongiae), all major orders of Demospongiae, 55 families and 165 genera. Previous studies suggested that several ordinal, family and genus patterns may exist, with unique types of compounds apparently restricted to discrete sponge taxa. Based on this premise, the impressive chemical dataset is potentially valuable in solving persistent problems and disagreements over the systematics of various taxa. However, compounds may be produced by sponge cells (and thus regarded as sponge characters), or by microsymbionts (which may not be necessarily species- or group-specific). Large numbers of proven or suspected microsymbiont compounds appear to be present from the lack of correspondence between sponge identity and compound structure, e.g. macrolides and cyclic peptides dispersed amongst most demosponge groups are suspected products from various microbes. Reported chemistry is distributed heterogeneously over the various sponge taxa, with highest diversity of compounds reported from Dictyoceratida and Dendroceratida (1,250 compounds from 145 species), Haplosclerida s.l. (665 from 85 species) and Halichondrida s.l. (approximately 650 from 100 species); other groups have an intermediate (Astrophorida-Lithistida, Hadromerida and Poecilosclerida) or very low (Calcarea) diversity of compounds. Despite previous claims that particular compounds occur exclusively in particular sponge taxa, we found that in most, if not all, cases compound distribution does not exactly match sponge classification. Some classes of compounds are predominant in particular taxa (e.g. bromotyrosines in Verongida, furanoterpenes in Dictyo- and Dendroceratida, straight-chain acetylenes and 3-alkylpiperidine derivatives in Haplosclerida s.l.), but almost invariably there are also reports of these classes of compounds from unrelated sponges. Furthermore, in rare cases where a compound type is restricted to a certain sponge group (e.g. pyrrole-2-carboxylic derivatives in Halichondrida s.l), their distribution amongst the families within the group appears to be inconsistent. Possible reasons for this fuzzy distribution include: 1) parallel biosynthetic pathways leading to the same structure; 2) involvement of microsymbionts; 3) careless specimen handling (contamination by epibionts, confused labels, etc.); 4) incorrect identification/classification. Currently, the degree of inconsistency is such that direct use of chemical data to solve classification problems, or to erect new higher taxa, is inadvisable. Inconsistent occurrence of compounds cannot be dismissed without further study. Large scale re-examination of voucher specimens, or recollection and chemical analysis, as well as cooperative studies between systematists, microbiologists and bio-organic chemists, are necessary to demonstrate whether or not chemical characters are true indicators of sponge systematics. O Porifera, chemistry, chemotaxonomy, bioactive compounds, review. Rob W.M. van Soest (email: soest@bio.uva.nl), Institute for Systematics and Ecology, University of Amsterdam, The Netherlands; Jean-Claude Braekman, Laboratoire de Chimie Bio-Organique, Université Libre de Bruxelles, Belgium; 4 April 1999. Natural products chemistry described from sponges is reminiscent of terrestrial plant chem- istry in its diversity and distribution throughout the phylum. Secondary metabolites, such as terpenoids, alkaloids and peptides, as well as bioactive fatty acid-, polyketide- and sterol derivatives, are common amongst most sponge groups. Biological activity of sponge compounds is very diverse (MarinLit lists more than 20 activity categories for various sponge compounds; Blunt & Munro, 1998), but cytotoxic (see also Schmitz, 1994), antibiotic, antifungal, antitumour, antiviral, antifouling and enzyme- inhibitory activities are the most common. Among marine organisms, sponges are the most productive sources of bioactive compounds: they 570 have so far yielded more than twice the number of structures reported from Cnidaria and from Algae, five times the number from Mollusca and Echinodermata, and seven times the number from Ascidiacea (Garson, 1994; Baker, 1996; Blunt & Munro, 1998). Geographic areas and habitats with the highest reported numbers of bioactive compounds from sponges are the Indo-West Pacific (ca. 800 structures), Australian - South Pacific (600) and Caribbean coral reefs (600). The Mediterranean (550) and Japanese waters (750) are also prolific source areas. East Pacific (250), East Atlantic (150), Indian Ocean (150), Red Sea (150) and New Zealand waters (100) are intermediate in diversity. This pattern is similar to the pattern of sponge species diversity over the seas and oceans of the world (Van Soest, 1994), and thus cannot be directly linked to ecological phenomena such as increased predation and competition (e.g. Green, 1977). Natural products continue to be described from sponges at an increasing rate (Table 1), such that the extent of sponge bioactivity is not yet apparent. So far, chemical structures have been elucidated from about 475 species of sponges, but many more have been shown to be bioactive in various bioassays. Previous reviews (e.g. Bergquist, 1979; Bergquist & Wells, 1983; Sarma et al., 1993), demonstrated that many types of compound are restricted to discrete groups of sponges, the prime example being bromotyrosine derivatives which appeared to be restricted to Verongida. A large body of literature has appeared since these last reviews of sponge chemotaxonomy. This lit- erature is now easily accessible through the MarinLit database (Blunt & Munro, 1998), which provides bibliographic references, structures and key words to virtually all marine natural products publications since the early sixties. The origin of bioactive compounds isolated from sponges is still a controversial issue. Many bio-organic chemists believe that microsymbionts are likely to be the source of most compounds rather than sponge cells themselves. In some recent studies (e.g. Faulkner et al., 1994), it has been demonstrated that sponge microsymbionts may indeed be the source of bioactive com- pounds, but that sponge cells themselves also appear to produce them. It is possible, although yet to be demonstrated, that endosymbionts living inside the sponge cells are the true source of such compounds. In that case, such symbionts MEMOIRS OF THE QUEENSLAND MUSEUM TABLE 1. Numbers of published articles in which sponge chemistry is described and numbers of chemical structures reported in the past decades since 1960 (source MarinLit data base; Blunt & Munro, 1998). publication No. articles No, structures H 1960-1969 4 2 1970-1979 | 314 361 1980-1989 1016 1275 1990-1997 1691 2484 total (1998) 3025 4122 may be obligatory symbionts which have co-evolved with the sponge hosts and sponge- symbiont chemical interactions may be indicative of chemo-taxonomic affinities. An example of such a scenario is explored in a recent study by Van Soest et al. (1998). It is the purpose of the present paper to review conclusions of previous chemotaxonomic studies and to determine whether new chemo- taxonomic evidence has come forward to support these conclusions. For this purpose we reviewed all sponge compounds and examined their distribution over the classes, orders, families and genera of sponges and, if relevant, over other marine phyla. METHODS AND DATA SOURCES The MarinLit database (Blunt & Munro, 1998) was consulted using taxonomic keywords for the various taxonomic groups yielding lists of references, species, and trivial names of com- pounds, as well as drawings of structures of compounds extracted from the species. The card system built up by one of us (JCB) was used as a supplementary source. These data provided a compilation of sponge chemistry arranged taxonomically by order, family and genus. Sub- sequent searches were made using trivial names of compounds or compound types as key words, to establish the distribution of classes of compound over the various sponge groups and other marine phyla. Since there is still no firmly established classification for sponges nor for secondary metabolite chemistry, chemosystematic sig- nificance of the various compounds and classes of compound was also determined by ad hoc discussions between the two authors. Compounds considered to be related and oc- curring in two or more clearly different taxa (species, genera, families, orders) are listed CHEMOSYSTEMATICS OF PORIFERA below arranged according to the sponge group in the order given in Table 2. The usually recognised sponge orders and TABLE 2. Numbers of structures reported from various sponge taxa, arranged by ordinal group and the numbers of species, genera and families from which the compounds were isolated. * including Chondrosiidae; **including Halichondriidae, Axinellidae, Bubaridae, Agelasidae, Ceratoporellidae; *** including ‘Nepheliospongida’/ Petrosida. families (Bergquist, 1978) are employed, with the exception Taxon No. Structures | No. Species | No. Genera | No. Families of orders Halichondrida s.l. |Homosclerophorida 120 14 5 1 (following Van Soest et al., | astrophorida 200 38 14 4 1990), and Haplosclerida |spirophorida 14 5 1 1 (following Van Soest, 1980). |i itnistida 200 20 T 5 Orders Dictyoceratida and 7 : Hadromerida* 185 40 13 7 Dendroceratida are treated cate F together for reasons explained Ammamma s, en 100 = : below Poecilosclerida 350 63 33 14 ~ «leri +e* » Relatedness of compounds, a s.l 665 85 17 5 in a phylogenetic sense, is not | Lubomirskiidae 3 2 3 f unequivocal as most com- |Dictyo/Dendroceratida 1,250 145 36 6 pounds consist of building |Verongida 240 22 8 3 blocks and side-chains with | Calcarea 40 9 3 2 often diverse biosynthetic origin. Unless biosynthetic experiments have been performed, homology of the seemingly related compounds remains tentative in most cases. In accordance with the chemical literature and to acknowledge discrepancies between chemical and morphological characters, we use the term ‘markers’ for shared compounds rather than ‘synapomorphies’. Examples of structures of ‘markers’ for the various taxa are included (Figs 1-43). Unique compounds reported from single species, though possibly significant as finger- prints, are ignored here, because they cannot be used for classification. RESULTS NUMBERS OF COMPOUNDS ISOLATED FROM SPONGES. To date (May 1998) more than 3,500 different chemical compounds have been extracted from 475 species of marine sponges belonging to two of the three classes (Demospongiae and Calcarea), all major orders of Demospongiae, one major order of Calcarea, ca. 55 families and 165 genera (Blunt & Munro, 1998). The various orders demonstrate large dif- ferences in numbers of compounds (Table 2). The total diversity of species compounds listed in Table 2 (3,917) maybe misleading because it has not been possible to check whether the same compounds were sometimes isolated from dif- ferent sponge taxa. However, since that is of relatively rare occurrence, a conservative esti- mate is approximately 3,500 different structures. Similarly, the total number of sponge species from which these compounds have been isolated (544) is inaccurate because of the large numbers of indeterminate identifications; quite a few of these may concern the same species. A conserv- ative estimate, based on arguments of geographic nearness of localities of indeterminate identif- ications, is approximately 475 different species of sponges. GENERALLY DISTRIBUTED COMPOUNDS. The first category delineated includes compounds which are apparently found over several or many different sponge groups, without any clearly restricted distribution amongst any particular sponge group. Some of these are suspected to be products of microsymbionts because sponges are known to have a rich bacterial flora which they use as food. The chemosystematic significance of these types of compounds is usually limited to be, at most, a fingerprint for individual species. Fre- quently, however, even that cannot be confirmed because symbionts may not be species specific. The following classes of compounds do not generally have much value for sponge classifi- cation because of their distribution amongst unrelated groups of sponges: Fatty acids and derived lipids. (Fig. 1A). These are ubiquitous and are often primary metabolites, although some specialised branched or unsat- urated fatty acids appear to be restricted in their distribution (see below). The chemosystematic significance of the presence and concentration of fatty acids with particular carbon-chain lengths has been explored by Bergquist et al. (1984), but from their patterns of distribution there is no hard 572 FIG. 1. Sponge chemistry. eve branched unsaturated fatty acid. B, sterol. C, carotenoid. D, cyclic peptide. evidence for their applicability to sponge systematics. Sterols. (Fig. 1B).These are ubiquitous and are often primary metabolites. Some sterols with specific side chains or functionalities (e.g. cyclopropene-, polyhydroxylated- or sulfated sterols) have a more restricted distribution and may have chemotaxonomic significance. Cyclopropene sterols have been used as a chemotaxonomic character to support the erection of a new order (Nepheliospongida or Petrosida; Bergquist, 1980), although subsequent research (Fromont et al., 1994) failed to demonstrate the consistent presence of these sterols amongst members of this ‘order’, whereas several similar cyclopropene sterols were isolated from the disparate taxa Spheciospongia (Hadromerida) (Catalan et al., 1982), Halichondria sp. (Halichondrida) (Ravi et al., 1978), and Lissodendoryx topsenti (Poecilosclerida) (Silva & Djerassi, 1991). The chemosystematic sig- nificance of the presence and concentration of sterols with particular side-chains and function- alities has been explored by Bergquist et al. (1986) and Fromont et al. (1994), but again there is no hard evidence for their consistency and applicability for sponge classification. They may be useful for fingerprinting at the species level, but even then care must be exercised (see e.g. Kerr & Kelly-Borges, 1994). Carotenoids. (Fig. 1C). These are basically derived from ingested autotrophic organisms and MEMOIRS OF THE QUEENSLAND MUSEUM modified in various ways by the sponges. A review is found in Liaaen-Jensen et al. (1982). The distribution of sponge carotenoids coincides with orange colour. They are reported from Astrophorida, Hadromerida, Halichondrida, Agelasida, Poecilosclerida, Haplosclerida, Dictyoceratida and Verongida. Cyclic and linear peptides. (Fig. 1D). These elaborate molecules have been reported from Astrophorida, Spirophorida, Lithistida, Halichondrida, Poecilosclerida, Haplosclerida, Dictyoceratida and Dendroceratida. In a review published by Fusetani & Matsunaga (1993) it is concluded that microsymbiont involvement is the most likely explanation for the widespread occurrence of these compounds. Compounds similar to those isolated from sponges are also reported from Cyanobacteria and several other marine invertebrates, notably ascidians. Macrolides. (Fig. 2). Similarly, these elaborate molecules have a wide distribution among Porifera: Calcarea, Astrophorida, Spirophorida, Lithistida, Hadromerida, Halichondrida, Poecilosclerida, and Dictyoceratida. Their apparent absence from Haplosclerida, Dendroceratida, Dysideidae and Verongida is noteworthy. Some of the molecules reported from sponges are almost identical to those of terrestrial Cyanobacteria or marine bacteria (Kobayashi & Kitagawa, 1998). Acridine derivatives. (Fig. 3A). These compounds are not common, but recorded from Homosclerophorida, Astrophorida, Haplosclerida and ascidians. Some of these sponges and ascidians are brightly coloured due to the possession of acridine derivatives. Nucleosides. (Fig. 3B). These have been recorded from Astrophorida, Hadromerida and Poecilosclerida. They are very likely to be microbial. Sesquiterpene quinones. (Fig. 3C). Related structures have been recorded from Chondrosia (Hadromerida or Chondrosida), Halichondria (Halichondrida), Strongylophora (Haplosclerida) and many Dictyoceratida and Dendroceratida, and even from Verongida. No taxonomic significance or pattern can be attributed to this distribution. Tetracyclic triterpenes. (Fig. 3D). Similar structures have been reported from Siphono- chalina siphonella (Haplosclerida), Axinella weltneri (Halichondrida) and Raspaciona aculeata (Poecilosclerida). No taxonomic signif- icance can be attributed to this distribution. CHEMOSYSTEMATICS OF PORIFERA 57 OCH; FIG. 2. Sponge macrolide. TAXONOMICALLY DISTRIBUTED COM- POUNDS Homosclerophorida compounds. About 120 different secondary metabolites have been reported from about 14 species belonging to 5 genera: Peroxy-polyketides (recorded from at least 9 species belonging to at least 2 genera) and acridine derivatives (recorded from 4 species belonging to 2 genera), are common constituents of Homo- sclerophorida. However, both these types of compounds are also found in other sponge groups. The Plakortis peroxy-polyketides (e.g. Higgs de Faulkner, 1978) (Fig. 3E) are particularly similar to those isolated from Callyspongia sp. (Toth & Schmitz, 1994), Cladocroce incurvata (D' Auria et al., 1993) (both Haplosclerida) and from Chon- drosia and Chondrilla (order Hadromerida or Chondrosida) (Wells, 1976; Stierle & Faulkner, 1979). So, despite the apparent concentrated occurrence of these structures in Homosclero- phorida, it is not entirely justified to consider peroxy-polyketides as valid markers for the group. Further investigations are required as to why similar structures can be found in the unrelated Haplosclerida and Hadromerida/ Chondrosida. Aminosteroids isolated from Plakina sp. (Rosser £ Faulkner, 1984: Fig. 3F) and Corticium sp. (Jurek et al., 1994), have an unusual nitrogen-bearing side chain, and these may be Ly considered to be a valid marker for these two taxa despite their being sterols. Astrophorida compounds. More than 200 secondary metabolites have been reported from at least 38 species belonging to 14 genera and all major families. Chemosystematic markers are listed below, Saponines (steroid-saccharides, e.g. Carmely etal., 1989: Fig. 4A) are reported across families: from 6 species of Ery/us (Family Geodiidae), | species of Melophlus (as Asteropus) (Family Ancorinidae) and 1 species of Pachastrella (Family Pachastrellidae). Related compounds are very common in Echinodermata, notably Asteroidea and Holothuroidea. Possibly the number of saccharides attached to the sterol part is species specific. The apparent absence of saponines in the other genera of the Astrophorida make them of dubious value for classification. Triterpenes (malabaricane and derivatives, e.g. McCabe et al., 1982) (Fig. 4B) were reported from 2 species of Stelletta, | species of Rhab- dastrella and from Jaspis stellifera. The latter is probably a Stelletta lacking triaenes and nota true Jaspis. Vhus, these triterpenes are a good marker for Srelletta s.l. (including closely related FIG, 3. Sponge chemistry. A, acridine alkaloid. B. nucleoside. C, sesquiterpene quinone. D, triterpene. E, peroxy-polyketide from — Plakortis halichondrioides. E, aminosteroid from Plakina sp. 574 Rhabdastrella and ‘Jaspis’ stellifera ) (family Ancorinidae). Penaresidins, peculiar straight-chained azetidine alkaloids (e.g. Kobayashi et al., 1991) (Fig. 4C) were independently isolated from two species of Penares (Ancorinidae) and thus may be a good marker for that genus. There are also compounds suspected or proven to be of microsymbiont origin. Sulfated sterols were reported from Pachastrella and Poecil- lastra, and thus could be a potential marker for the family Pachastrellidae. However, sulfated sterols are also found in Polymastia (Hadromerida), Hymedesmia (Poecilosclerida) and several Halichondrida, so their value as marker is dubious. Also, like saponines, these compounds are very common in Echinodermata. Cyclic peptides, macrolides and polyketides are commonly reported from species of Geodia (Geodiidae) and Jaspis (Coppatiidae), but these are often similar to Lithistid compounds and very probably of microsymbiont origin (discussed elsewhere in this review). Geodia barretti and Pachymatisma johnstonia (Geodiidae) share a bromoindole compound of very similar structure, and thus these may be considered to be a valid marker for these two genera. Dercitus (Pachastrellidae) and Stelletia (Ancorinidae) share similar acridine derivatives; however, related compounds are also reported from Homosclerophorida and Haplosclerida, as well as from ascidians as noted above. Common-place sterols and (un)saturated fatty acids were reported from many species, but their chemosystematic value is low and they will not be discussed further here. Spirophorida compounds. Fourteen secondary metabolites have been reported from about 5 species belonging to a single genus, Cinachyrella (Tetillidae). The fatty acids, sterols and macro- lides shared between species do not seem to have chemosystematic value. The macrolides are similar to those of *Lithistida' (e.g. Theonella), and to those isolated from the marine bacterium Vibrio sp. (Kobayashi & Kitagawa, 1998). No compounds are shared with the ‘lithistid’ family Scleritodermidae which is — on morphological grounds — assumed to be closely related to Spirophorida. ‘Lithistida’ compounds. Approximately 200 structures have been reported from at least 20 species belonging to 11 genera and 5 families. ‘Lithistida’ are certainly polyphyletic, with some MEMOIRS OF THE QUEENSLAND MUSEUM FIG. 4. A, Saponine from Erylus lendenfeldi. B, malabaricane-type triterpene from Stelletta sp. C, penaresidine from Penares sp. D, aaptamine from Aaptos aaptos. E, clionamide from Cliona celata. families showing distinct synapomorphies with Spirophorida (e.g. Scleritodermidae) and Astrophorida (e.g. Corallistidae). Commonplace sterols were reported from several species, but their chemosystematic value is low and they will not be discussed further here. Dominant compounds are cyclic peptides and macrolides, shared between families and genera. Related cyclic peptides are shared between several species of Theonella, 3 species of Discodermia and 1 species of Neosiphonia (Theonellidae), 1 species of Callipelta (Corallistidae), 1 species of Aciculites and 1 species of Microscleroderma (=Amphibleptula) (Scleritodermidae). Thus, it would seem that these cyclic peptides are straightforward markers for the ‘Lithistida’. However, similar compounds are found in unrelated Astrophorida (Geodia, Jaspis), Hadromerida (Hemiasterella) and several Halichondrida (Halichondria, Stylissa). Discodermin E, the cyclic peptide isolated from Discodermia kiiensis (Ryu et al., 1994) is almost identical to the halicylindramides of Hali- chondria cylindrata (Li et al., 1995). These facts, coupled to the recorded presence of a rich microsymbiont flora in ‘lithistids’, support Fusetani & Matsunaga’s (1993) conclusion of probable microsymbiont origin of these peptides. CHEMOSYSTEMATICS OF PORIFERA Related macrolides are shared between several species of Theonella, 2 species of Discodermia, | species of Neosiphonia, | species of Reidispongia (Theonellidae) and 1 species of Callipelta (Coral- listidae). Again, however, similar macrolides have been isolated from many different sponges belonging to widely divergent orders. As with cyclic peptides the chemosystematic value of ‘lithistid’ macrolides is thus compromised. Hadromerida compounds. Approximately 185 secondary metabolites have been reported from at least 40 species belonging to 13 genera and 7 families. Fatty acids (except those mentioned below), sterols and carotenoids occur across several families and genera, but will not be discussed further. Cyclic peptides, macrolides and polyketides have been reported from a few Hadromerida and will also be left out of consideration. No distinct hadromerid compounds can be identified. However, several compounds appear to be useful markers for families (or at least genus groups) and genera. Aaptamine-type alkaloids (e.g. Nakamura et al., 198) (Fig. 4D) have been isolated from sev- eral species of Aaptos and Suberites and thus may be considered tentative markers for Suberitidae. They have been reported previously as markers for the order Hadromerida (Bergquist et al., 1991), but this is unwarranted in view of their limited occurrence. Carballeira et al. (1989) maintained that 4,8,12-trimethyltridecanoic acid was a useful marker for the families Spirastrellidae and Clionidae, as they isolated this fatty acid from both Anthosigmella varians and Cliona aprica. The occurrence of this admittedly unusual fatty acid needs further investigation before this can be accepted. Two presumably different species of Cliona apparently share the possession of clionamides (e.g. Stonard & Andersen, 1980) (Fig. 4E), which could serve as a useful marker for the genus. However, the amide shows some structural relationships with cyclic peptides and is thus a suspect microsymbiont-produced compound. Peroxy-sesterterpenoids and derivatives (e.g. Capon et al., 1987) (Fig. 5A) have been isolated from 1 species of Sigmosceptrella and 3 species of Latrunculia. These genera were previously considered synonymous. But the value of these compounds as markers for Latrunculiidae is diminished by the isolation of closely related sesterterpenoids from 3 species of Mycale (Mycalidae) and from Prianos spec. (a name of uncertain affinity) (Manes et al., 1984). Pyrroloquinoline alkaloids (e.g. Perry et al., 1986) (Fig. 18) have been isolated from 5 species of Latrunculia, but again the value of these com- pounds as markers for this genus is compromised by the reports of very similar and undoubtedly related compounds from another group of Poecilosclerida, viz. Zyzzya (lophonidae) (see Van Soest et al., 1996; Dumdei et al., 1998). If both compound types were genuine sponge compounds then their shared occurrence in Latrunculiidae - Mycalidae and Latrunculiidae - Iophonidae could indicate that 1) Latrunculia s.. has poecilosclerid affinities; and 2) Latrunculia is polyphyletic supporting the data of Kelly-Borges. However, Perry et al. (1988) suggested that microsymbionts may be involved in the product- ion of the pyrroloquinoline alkaloids. Chondrosida compounds. This group is con- sidered a family of the order Hadromerida in most previous classifications, but the apparent absence of morphological synapomorphies may justify their separation as an order of their own. Approximately 18 secondary metabolites have been reported from at least 4 species belonging to 2 genera. Apart from straight-chained and branched unsaturated fatty acids and sterols, two other compound types have been reported from this group: Peroxy-polyketides similar to, or some identical to, those of Homosclerophorida (e.g. Fig. 3E) were isolated from Australian Chondrilla (Wells, 1976) and Caribbean Chondrosia (Stierle & Faulkner, 1979), Mistaken identification is unlikely, though not impossible, and corroboration of the occurrence of peroxy-polyketides in Hadromerida/ Chondrosida would be welcome. Halichondrida s.l. compounds. For morpholog- ical and chemosystematic reasons Halichondrida are here treated in a very wide sense, including the families Halichondriidae, Desmoxyidae, Dictyonellidae, Axinellidae, Bubaridae, Agelas- idae and Ceratoporellidae. The nominal orders Axinellida (Hemiasterellidae and Raspailiidae excluded), Agelasida and Halichondrida s.s. are here included but not separately treated because the groups are in a taxonomic flux, with several recent revisions and proposed rearrangements. Morphologically, the recognised families are perceived by us to intergrade from Agelasidae at one end to Halichondriidae at the other end. 576 FIG. 5. A, trunculin-type sesterterpene from Latrunculia brevis. B, discorhabdin from Latrunculia sp. C, oroidin from Agelas oroides. D, isocyanosesquiterpene from Halichondria sp. E, sulfated steroid from Halichondria cf.moorei. F, cyclic terpene from Myrmekioderma styx. Chemically there appear to be shared classes of compounds amongst family clusters with an overlap between Halichondriidae and Axinellidae (cf. Braekman et al. 1992, and see below). Many identifications of sponges with secondary met- abolites are suspected or proven to be slightly or widely off the mark, which makes detailed re- examination of vouchers an essential prerequisite. From Halichondrida, as employed here, approximately 650 structures have been reported from at least 100 species belonging to 23 genera and 7 families. Compounds with chemo- systematic significance include the following. Pyrrole-2-carboxylic derivatives (e.g. Braek- man et al, 1992) (Fig. 5C) have been isolated from 12 species of Agelas (Agelasidae), 1 species of Astrosclera, 1 species of Goreauiella (Ceratoporellidae), 3 species of Axinella, 3 species of Stylissa, 1 species of Phakellia, | species of Cymbastela, | species of Ptilocaulis (as Teichaxinella) (Axinellidae) and 3 species of Hymeniacidon (Halichondriidae). At least one of the Hymeniacidon species (H. aldis) is a suspect Hymeniacidon as H. aldis is a junior synonym of Stylissa massa. lt is possible that true Hymen- iacidon (i.e., those with a detachable tangential skeleton) do not synthesise this class of com- pounds, and all reported Hymeniacidon with that compound type are in reality Stylissa. Thus, it MEMOIRS OF THE QUEENSLAND MUSEUM appears that pyrrole-2-carboxylic derivatives are at least a marker for Agelasidae-Ceratoporellidae- Axinellidae, illustrated for example by the shared possession in Agelas oroides (Agelasidae), Goreauiella sp. (Ceratoporellidae) and Stylissa carteri (Axinellidae) of the same compound oroidin (Fig. 5C) (e.g. Braekman et al., 1992; Rinehart, 1989; Supriyono et al., 1995). A possibly related pyrrole compound is recorded from Pseudoceratina purpurea | (Verongida) (Tsukamoto et al., 1996), but it may also be a case of convergent synthetic pathways. Isocyanoterpenes (Burreson et al., 1975) (Fig. 5D) have been isolated from 2 species of Axinella, 1 species of ‘Stylotella’, 2 species of Cymbastela, 5 species of Acanthella (all Axinellidae), 1 species of Bubaris (Bubaridae), 5 species of Halichondria, 3 species of Hymeniacidon, 2 species of Ciocalypta, | species of Topsentia, | species of ‘Leucophloeus’, 3 species of Axinyssa (partly as Trachyopsis ), and 1 of Epipolasis (all Halichondriidae). Even though it is suspected that identifications may not be entirely accurate, this is overwhelming evidence, that isocyanoterpenes are shared be- tween families Axinellidae and Halichondriidae. Sulfated sterols (Fusetani etal., 1981) (Fig. 5E) were isolated from 2 species of Halichondria, 4 species of Topsentia, | species of Axinyssa, 1 species of Epipolasis and | Halichondriidae not further identfied. Thus it would seem that they are a marker for the family Halichondriidae. However, sulfated sterols are also common in Pachastrellidae (Astrophorida) and have been isolated from a species of Polymastia (Kong & Andersen, 1996) and a species of Hymedesmia (as Stylopus) (Prinsep et al., 1989); they are also common in Echinodermata. Cyclic diterpenes (Sennett et al., 1992) (Fig. 5F) occur in Myrmekioderma and Higginsia and thus may be a potential marker for the family Desmoxyidae. Linear diterpenes (Albrizio et al., 1992: Fig. 6A) described from Myrmekioderma and Didiscus appear to be unrelated or only distantly related to the cyclic diterpenes. Moreover, the record from Didiscus is a suspect identification because it concerns an E. Pacific species, and so far the genus Didiscus is not known from that area. It could be a case of a mistaken Myrmekio- derma, because Myrmekioderma and Didiscus share similar habit characters. Consequently, the linear diterpenes may be a marker for Myrmekio- derma only. CHEMOSYSTEMATICS OF PORIFERA $77 o [ Qu d 27 À. + ko a H hil r CN D Ky | pe NH. Hj NH FIG. 6. A, linear diterpene from Myrmekiaderma styx. B, curcuphenol from Didiseus oxeata. C, topsentin from Spongosorites genirrix. D, agelasine from Agelas nakamurai. E. agelasine G from Agelas sp- F, crambescine A from Crambe crambe. Curcuphenol (Wright et al., 1987) (Fig. 6B) and related sesquiterpenes have been recorded from Didiscus flavus and Epipolasis sp. Some Didiscus specimens may have few of the char- acteristic didiscorhabs and then are easily mistaken for related genera such as Topsentia or Epipolusis. If that has been the case, it would mean curcuphenol and related sesquiterpenes are markers for Didiscus, Topsentins (Bartik et al, 1987) (Fig. 6C) were considered a marker for Spongosorites since 4 species of that genus have yielded these cam- pounds. However, this neat marker is threatened by the record of topsentins ftom 2 species of the Axinellidae genus Dragmacidon, which shows no close relationship with Spongosorites. The reported occurrence of both bromotyrosine der- ivatives (a Verongida marker compound) and topsentins in Hexadella (Dendroceratida) (Morris & Andersen, 1989; Morris & Andersen, 1990) is one of the more intriguing inconsis- tencies, It is also possible that bisindole compounds isolated from Hamacantha (Poecilo- sclerida) (Komoto & McConnell, 1988) are related to topsentins. Terpene compounds (diterpenes (Wu et al., 1984) (Fig. 6D) and sesquiterpenes) are found in several Agelas species and thus may be markers of species groups within that genus (Braekman et al., 1992). A remarkable and significant com- pound was isolated from Agelas sp. (Ishida et al., 1992) (Fig. 6E): an apparent combination of a terpene and a pyrrole-2-carboxylic substructure. This indicates that in Agelasidae, in contrasi to Axinellidae, some species have the ability to synthesise both terpenes and pyrrole-2-carboxylic acid moieties and even to combine these. Compounds with low chemosystematic significance are the following: Macrolides and polyethers are commonly reported from Heli- chondria, cyclic peptides from Halichondria, Axinella, Phakellia, Cymbastela and Stylissa, Carotenoids are found in Acanthella and Agelas. Sterols and fatty acids are ubiquitous in this group. A host of unrelated smaller and larger compounds have so far been isolated from single species. Further exploration is needed 10 assess their potential as taxonomic markers. Poecilosclerida compounds. Approximately 350 secondary metabolites have been reported trom ai least 63 species belonging to 33 genera and 14 families. Despite this large number of compounds, very few appear to have chemosystematic significance. Apart from sterols, fatty acids, macrolides and cyclic peptides, many indoles, pyrroles and carotenoids have been reported from Poecilosclerida, but in most cases there is no consistent taxonomic pattem, Polycyclic guanidine alkaloids (Berlinck et ál., 1990) (Fig. GF) have been isolated from 2 species of Monanchora and | species of Crambe and are thus a potential marker for the family Cram- beidae (Van Soest et al, 1996). However, these are also reported from Arenochalina mirabilis (Barrow et aL, 1996), which is supposedly u Mycalidae, The voucher musi be verified because Arenochalina has subtylostyles rather similar io those of Monanchora or Cranibe, and reduced spiculation is very common in those genera. Peroxy-sesterterpenoids and related derivatives (Capon & Macleod. 1987) (Fig. 7A) have been isolated from about 5 species of Mycale, but very similar compounds are known from Latrunculia (Hadromerida, see aboye and Fig. 5A), The chemosystematic value of these compounds is thus dubious. Trikentrin (Capon et al., 1986) (Fig. 78) und related compounds were isolated from two species of Trikentrion (Raspailiidac) and these might be a marker for that genus. But similar indoles are also reported for a species identified as Axinella 578 (Halichondrida) (Herb et al., 1990). Re-examination of the voucher might reveal that the characteristic triactines have been overlooked (they are often rare in various Trikentrion species and the further skeletal characters are similar to those of Axinella). Haplosclerida compounds. Haplosclerida are here considered in a wide sense, including the order Nepheliospongida or Petrosida. The issue of one or two orders has been debated at several occasions using morphological, life cycle and chemistry arguments. Since both groups appear to share unique chemistry it is practical to unite the two groups for our purpose. From this group of 5 (marine) families, approximately 665 secondary metabolites have been reported from at least 85 species belonging to 17 genera and 5 families. Chemosystematic markers appear as follows: straight-chain acetylenic compounds occur across 4 of the 5 families, they appear to be lacking in Niphatidae (see an extended review in Van Soest et al., 1998). The compounds are a clear marker for Haplosclerida s.l. However, related compounds have been described from Phakellia carduus (Halichondrida) and Raspailia ramosa (Poecilosclerida), which make it likely that the compounds are produced by microsymbionts. There are distinct types of acetylenic compounds based on the number of carbon atoms, the number and position of acetylenic bonds and the nature and position of the side chains. For example, Petrosia characteristically has hydroxyl-groups as side chains (e.g. Fusetani et al., 1983) (Fig. 7C), whereas the massive Xestospongia species (X. muta, X. testudinaria) characteristically have terminal bromine atoms (e.g. Patil et al., 1992) (Fig. 7D). 3-Alkylpiperidine derivatives have been reported from all 5 families of Haplosclerida and thus are a good marker for the order (see review in Andersen et al., 1996). Their occurrence in Phloeodictyidae is based on Pellina and Pachypellina, the family assignment of which is considered dubious. The type of Pe//ina is con- sidered to be a Halichondria, but most species assigned to Pellina are either Haliclona (Chalinidae) or Oceanapia (Phloeodictyidae). The type of Pachypellina is considered to be a Xestospongia (Petrosiidae). It appears as if'straight alkylpiperidines such as niphatesines (e.g. Kobayashi et al., 1992) (Fig. 7E) and halitoxins occur in families Niphatidae and Callyspongiidae, whereas cyclic alkylpiperidines (e.g. Sakai et al., 1986) (Fig. 7F) occur in Chalinidae, Petrosiidae MEMOIRS OF THE QUEENSLAND MUSEUM FIG. 7. A, sigmosceptrelline from Mycale ancorina. B, trikentrine from Trikentrion flabelliforme. C, straight-chain acetylene from Petrosia sp. D, straight-chain acetylene from Xestospongia muta. E, niphatesine D from Niphates sp. F, manzanine A from Haliclona sp. G, calysterol from Calyx nicaensis. and (perhaps) in Phloeodictyidae. Much remains uncertain, because identification and assignment of species in this order are a specialist job. Voucher re- examination and repetition of collec- tion and extraction is necessary to properly assess the chemotaxonomic significance of this group of compounds at the genus and family level. 3-alkylpiperidine derivatives have been reported from Stelletta (Astrophorida), Theonella (‘Lithistida’) and /rcinia (Dictyoceratida), but these are very likely cases of overgrowth by epibiont Haplosclerida, because identical com- pounds were also isolated from Haplosclerida. Unpublished information (M.K. Harper, in litteris) indicates that a particular 3-alkylpiper- idine derivative, manzamine A, occurs also in a few Poecilosclerida (Clathria and Mycale). Thus, some doubts exist as to the true origin of these compounds, but no microsymbiont sources have so far been identified (see Kobayashi & Kitagawa, 1998). A rather striking observation is that straight- chain acetylenes are recorded from several massive volcanoe-shaped Xestospongia, whereas 3-alkylpiperidines are recorded from compact, fine-grained and less elaborate Xestospongia. Previous authors have employed different names (Xestospongia s.l. and Neopetrosia) for these CHEMOSYSTEMATICS OF PORIFERA sponges and chemistry appears to support this subdivision. Cyclopropene sterols (e.g. Itoh et al., 1983) (Fig. 7G) are a marker for the ‘order’ Nephelio- spongida/Petrosida (Bergquist, 1980), as they have been recorded from 2 species of Xesto- spongia, 2 species of Petrosia, 1 species of Cribrochalina (all Petrosiidae), 1 species of Oceanapia and from 2 species of Calyx (Phloeodictyidae). However, a recent study (Fromont et al., 1994) has shown that they are absent in most investigated members of these families. Moreover, similar sterols have been reported from species in the Hadromerida, Halichondrida and Poecilosclerida. Their chemo- systematic significance is probably low and certainly debatable. Tetrahydropyrans (e.g. Ciminiello et al., 1992) (Fig. SA) are independently recorded from two species of Haliclona. They may be a marker for that genus, although such functionalities are widely distributed in natural compounds. Low value markers for Haplosclerida are as follows: acridine alkaloids of closely related structure were isolated from Amphimedon sp. (Niphatidae) (Schmitz et al., 1983), Petrosia sp. (Petrosiidae) (Molinski et al., 1988), Oceanapia sagittaria (Salomon & Faulkner, 1996) and Oceanapia sp. (Eder et al., 1998) (Phloeodic- tyidae), and would seem to be a potential marker for those three genera/families. However, they contain an acridine moiety and are closely related to similar acridine compounds reported from Homosclerophorida and Astrophorida (as report- ed above). Sesquiterpene and diterpene quinones are rather commonly found in Chalinidae (3 species), Petrosiidae (6 species) and Phloeodictyidae (2 species). However, similar compounds are also common in Dictyoceratida and Dendroceratida and are reported occasionally from Hali- chondrida and Chondrosida. Isoquinolinoquinones are recorded from blue sponges assigned to Reniera, Haliclona (Chalinidae), Petrosia, Xestospongia and Cribrochalina (Petrosiidae). In view ofthe rather unusual blue colour, it is possible that these records all concern only a single species. In any case, identical or closely similar compounds are produced by a terrestrial Streptomyces (bacteria), and thus the chemistry is probably symbiont- derived. The usual complement of fatty acids, sterols, cyclic peptides, polyketides and carotenoids have 579 been reported across the families of Haplo- sclerida, but no taxonomic significance can be attributed to them. Many unrelated compounds were isolated from single species. Freshwater sponge compounds. Only fatty acids and sterols have been reported from several species of the family Lubomirskiidae. These seem to have no taxonomic value. Dictyoceratida and Dendroceratida compounds. We choose here to treat the orders Dictyoceratida and Dendroceratida in tandem because there is shared chemistry between the two and there is also a ‘border conflict’ over the assignment of the family Dysideidae. Bergquist (e.g. Bergquist, 1996) retains Dysideidae in Dictyoceratida, whereas Boury-Esnault et al. (1990) assign it to Dendroceratida. From this assemblage approximately 1,250 structures have been recorded from at least 145 species belonging to about 36 genera and 6 families. Chemosystematic markers are as follows: Furan- or lactone terpenes (sesterterpenes: e.g. De Giulio et al., 1989 (Fig. 8B); sesquiterpenes, e.g. Guella et al., 1985 (Fig. 8C); and diterpenes: e.g. Bobzin & Faulkner, 1989 (Fig. 8D)) are shared by many species and genera of the group. Although there is a predominance of sester- terpenes in families Spongiidae, Irciniidae and Thorectidae (undisputed Dictyoceratida), a predominance of sesquiterpenes in Dysideidae, and a predominance of diterpenes in Darwin- ellidae and Dictyodendrillidae (both undisputed Dendroceratida), the occurrence is never absolute and many inconsistent records exist. Bergquist (e.g. Bergquist, 1996) chose to dismiss these inconsistencies announcing that they can be resolved by reassigning species to different families and genera. However, in view of the number of inconsistencies, this seems rather too optimistic. The evidence for this opinion is as follows. Sesterterpenic furans or lactones are found in: 11 species of Spongia, 3 species of Hippospongia, 3 species of Carteriospongia, 3 species of Phyllospongia, | species of Strepsi- chordaia, 1 species of Collospongia, 1 species of Leiosella, | species of Dactylospongia, | species of Rhopaloeides, | species of Hyattella (all Spongiidae), about 9 species of Ircinia, 3 species of Sarcotragus, | species of Psammocinia (all Irciniidae), 3 species of Cacospongia, 3 species of Luffariella, 3 species of Fasciospongia, 2 species of Hyrtios, 2 species of Lendenfeldia, 2 species of Thorecta, | species of Petrosaspongia, 1 species of Fascaplysinopsis (all Thorectidae), 2 species of Dysidea, | species of Spongionella 580 MEMOIRS OF THE QUEENSLAND MUSEUM TABLE 3. Distribution of furan- and lactone terpenes over Dictyoceratida and Dendroceratida. Compound Spongiidae Thorectidae Irciniidae Dysideidae Darwinellidae | Dictyodendrillidae Sesterterpene 26 17 13 3 0 2 Sesquiterpene 1 0 0 14 1 2 Diterpene 8 1 0 3 11 2 (Dysideidae), and 2 species of /gernella (Dictyo- dendrillidae). Sesquiterpenic furans or lactones are found in: | species of Spongia (Spongiidae), 12 species of Dysidea, 2 species of Euryspongia (Dysideidae), l species of Pleraplysilla (Darwinellidae) and 2 species of Dictyodendrilla (Dictyodendrillidae). Diterpenic furans or lactones are found in: 5 species of Spongia, 1 species of Hippospongia, 1 species of Dactylospongia, 1 species of Hyattella (all Spongiidae), 1 species of Luffariella (Thorectidae), 2 species of Dysidea, 1 species of Spongionella (Dysideidae), 3 species of Aplysilla, 3 species of Chelonaplysilla, 3 species of Darwinella, 2 species of Dendrilla (Darwiellidae), 1 species of /gernella and 1 species of Dictyo- dendrilla (Dictyodendrillidae). From the overview presented in Table 3 it is evident that the number of cases that do not match the simple scheme presented by Bergquist (1996), viz. Dictyoceratida: sesterterpenes, Dendroceratida: diterpenes, Dysideidae: sesqui- terpenes, is substantial, involving about 20% of all investigated species. It will take more than just reassigning a few possible mistakes. Moreover, the sesterterpenes, diterpenes and sesquiterpenes are biogenetically related. Several species (e.g. Spongia agaricina) apparently are able to synthesise both furanosesterterpenes and furano- sesquiterpenes, or (e.g. Spongia officinalis) both furanosesterterpenes and lactone diterpenes. Because all three terpene types basically originate from a common biosynthetic pathway, which only at the end part of the synthesis of the terpenes will have divergent pathways, it is con- ceivable that the inconsistent occurrence of the terpenes is the product of independent (con- vergent) development. It seems best at present to emphasise the shared presence of furan- and lactone terpenes as a marker for both Dictyo- ceratida and Dendroceratida. A possible use of this compound type as marker for family or genus levels will have to await further studies com- bining voucher re-examination, morphological and molecular taxonomy, microsymbiont research and biosynthetic experiments. Pending this, it would be unwise to rearrange Dictyo- and Dendroceratida species and genera based only on terpene chemistry. Of course, emphasis of shared compounds which appear to confirm morph- ological synapomorphies remains a justified course of action. A single inconsistent occurrence of sester- terpenic lactones is reported from a Japanese Amphimedon spec. (Ishibashi et al., 1993). This is possibly a case of mistaken labelling as the same group of chemists reported the occurrence of a typical Haplosclerid compound, manza- mines, from an /rcinia spec. (Kondo et al., 1992). Low value markers are as follows: Sesqui- terpene quinones are reported from Spongia (4 species), Hippospongia, Coscinoderma, Dactylospongia, Hyattella, Ircinia, Sarcotragus, Fasciospongia, Smenospongia, Hyrtios, Thorectandra, Fenestraspongia, Dysidea (6 species) and Euryspongia. Thus, they seem to be good markers for Dictyoceratida including Dysideidae. However, these compounds are closely similar to the sesquiterpene quinones reported from Haplosclerida, Halichondrida and Chondrosida. Their apparent absence from Dar- winellidae and Dictyodendrillidae is noteworthy. Diketopiperazine derivatives, resulting from the condensation of two aminoacids, and di- phenylether derivatives have been isolated from several species of Dysidea. However, sophist- icated research by Faulkner et al. (1994) proved beyond doubt that bacteria are responsible for the production of these compounds. Diketopiper- azines isolated from 7edania (although with different amino acid building blocks than those of Dysidea) also appeared to be produced by a bacterium (Stierle et al., 1991). Polyhydroxylated sterols have been isolated from Spongia (2 species), Hippospongia, Ircinia (2 species), Dysidea (4 species), Euryspongia and Spongionella. Thus, they seem to be a marker for the Dictyoceratida including the family Dysideidae. However, these compounds are reported from isolated species belonging to almost all orders of the Demospongiae. Moreover they are commonly reported from Echinodermata and soft corals. CHEMOSYSTEMATICS OF PORIFERA FIG. 8. A, haliclonol from Haliclona hogarthi.B, scalarine from Spongia officinalis. C, furanosesquiterpene derivative from Dysidea avara. D, polyrhaphin from Aplysilla polyrhaphis. V. bromotyrosine derivative from Aplysina aerophoha. Indole derivatives occur scattered over all fam- ilies. These compounds oceur in many different sponge groups (and indeed other animal phyla), and they are assumed to be of microsymbiont origin. In any case, they appear quite diverse in the various species and genera. Sterols and fatty acids have been isolated from many Dictyoceratida and Dendroceratida. Macrolides occur scattered over à few species in the families Spongiidae and Thorectidae; their absence in Dendroceratida is perhaps note- worthy. Cyclic peptides occur sparsely (here and there) over all families. The family Halisarcidae is usually attributed to Dendroceratida, but was recently raised to ordinal level: Order Halisarcida Bergquist. 1996. No compounds have been isolated from members of this group so far: Verongida compounds. Approximately 240 secondary metabolites are reported from at least 22 species belonging to $ genera and 3 families. Bromotyrosine derivalives (e.g. Ciminiello et al.. 1997) (Fig. SE) are uniformly present in all families and genera of Verongida (11 species of Aplysina, 2 species of Ferongula, 2 species of Janthella, | species of Anomoianthella, 2 species of Pseudoceratina, 2 species of Suberea, ] species of Aiolochroia and | species of Aplysinella). Within this large group of derivatives, macrocyclic bromotyrosines (e.g. Pordesimo & Schmitz, 1990) (Fig. 9A).are shared between two species of Janthella, so they could be a marker for that genus; however, a macrocyclic bromo- tyrosine is also recorded trom Pseudoceratina purpurea (Carney et al, 1993). The value of bromotyrosine derivatives as a marker for the Verongida is diminished by the isolated occurrence of similar bromotyrosine compounds in Jotrochota hirotulata (Poecilo- sclerida) (Constantino et al., 1994) and Agelas (Agelasidae) (Kónig & Wright, 1993), Related bromotyrosines have been found also in an ascidian, Botrvilus (McDonald et al., 1995) and a green alga, Avrainvillea (Colon etal., 1987). The reported occurrence in Hexadella (Dendro- ceratida) of both bromotyrosine derivatives (Verongida) (Morris & Andersen, 1989) and topsentins (Spongosorites compounds) (Morris & Andersen, 1990) is one of the more intriguing inconsistencies. Sterols. fatty acids, carotenoids, nucleosides and sesquiterpene quinones have been reported across families and genera. Their value for tax- onomy is low. The apparent absence of cyclic peptides and macrolides in this order is note- worthy. Calcarea compounds, Approximately 40 secondary metabolites were reported from at least 9 species. So far the subclass Calcaronea did not yield any compounds (the record of phospholipid fatty acids and sterols from Caribbean OH H B N I 3 p^ EN o Br Ha B OH | € | Br o Z g OH c H OH A — NAL ARS 5 B i CH 0 E Í 1 N M 7 Sy o wf wo FIG. 9. A, bastadine from Janthella hasta. B, clathridine from Clathrina clathrus. C, rhapsamine from Leucetta leptorhaphis. 582 MEMOIRS OF THE QUEENSLAND MUSEUM TABLE 4. Chemical markers (bold print) and non-exclusive markers (plain print) for sponge taxa, with numbers of species from which the compound was isolated. Figure numbers refer to figures presented in this review. For further comments and explanations see text. Sponge group (No. spp. studied) Compound type (example of structure) Homosclerophorida (9) peroxy-polyketides (Fig. 3E) Plakina-Corticium (2) steroid-amines (Fig. 3F) Astrophorida (8) saponines (Fig. 4A) Stelletta s.l. (4) triterpenes (Fig. 4B) Penares (2) penaresidins (Fig. 4C) Pachastrellidae (2) sulfated sterols (Fig. 5E) Suberitidae (3) aaptamines (Fig. 4D) Spirastrellidae/Clionidae (2) 4,8,12-trimethyltridecanoic acid Cliona (2) clionamides (Fig. 4E) Latrunculiidae (4) peroxy-sesterterpenoids (Fig. 5A) Latrunculiidae (5) pyrroloquinoline alkaloids (Fig. 5B) Axinellidae- Agelasidae-Ceratoporellidae (26) pyrrole-2-carboxylic derivatives (Fig. 5C) Axinellidae-Bubaridae-Halichondriidae (32) isocyanoterpenes (Fig. 5D) Halichondriidae (9) sulfated sterols (Fig. 5E) Desmoxyidae (3) cyclic diterpenes (Fig. 5F) Myrmekioderma (2) linear diterpenes (Fig. 6A) Didiscus (2) sesquiterpene phenols (Fig. 6B) Spongosorites (4) topsentins (Fig. 6C) Agelas (6) di- and sesquiterpenes (Fig. 6D) Crambeidae (3) polycyclic guanidine alkaloids (Fig. 6F) Mycale (5) peroxy-sesterterpenoids (Fig. 7A) Trikentrion (2) trikentrin indoles (Fig. 7B) Haplosclerida s.l. (ca. 17) straight-chain acetylenes (Figs 7C-D) Haplosclerida s.l. (ca. 22) 3-alkylpiperidine derivatives (Figs 7E-F) Petrosia (ca. 7) polyhydroxylated acetylenes (Fig. 7C) Xestospongia s.s. (ca. 3) brominated acetylenes (Fig. 7D) Niphatidae + Callyspongiidae (ca. 6) linear 3-alkylpiperidines (Fig. 7E) Chalinidae + Petrosiidae (ca. 8) cyclic 3-alkylpiperidines (Fig. 7F) Petrosiidae + Phloeodictyidae (8) cyclopropene sterols (Fig. 7G) Haliclona (2) tetrahydropyrans (Fig. 8A) Dictoceratida + Dendroceratida (102) furano-or lactone terpenes (Figs 8B-D) Spongiidae + Thorectidae + Irciniidae (56) furano-or lactone sesterterpenes (Fig. 8B) Dysideidae (14) furano-or lactone sesquiterpenes (Fig. 8C) Darwinellidae + Dictyodendrillidae (13) furano - or lactone diterpenes (Fig. 8D) Verongida (22) bromotyrosine derivatives (Fig. 8E) lanthella (2) macrocyclic bromotyrosines (Fig. 9A) Clathrinida (4) guanidine-imidazoles (Fig. 9B) Clathrinida (3) long-chained aminoalcohols (Fig. 9C) Leucosolenia canariensis (Carballeira & Shalabi, 1995) almost certainly concerns Clathrina, which is a member of the Calcinea). Within the subclass Calcinea compounds were isolated only from members of the order Clathrinida. Guanidine-imidazoles (e.g. Ciminiello et al., 1989) (Fig. 9B) are recorded across families: from 1 species of Clathrina (Clathrinidae) and 3 species of Leucetta (Leucettidae), and are thus markers for the order Clathrinida. Long-chained aminoalcohols (Jayatilake et al., 1997) (Fig. 9C) are recorded across families: from | species of Clathrina (Clathrinidae) and 2 species of Leucetta (Leucettidae), and are thus also markers for the order Clathrinida. Sterols, fatty acids, a macrolide (similar to those of various Demospongiae) and a pteridine (similar to compounds from terrestrial organisms) make up the remaining compounds reported from CHEMOSYSTEMATICS OF PORIFERA 583 Calcarea. These do not appear to have chemo- taxonomic significance. SUMMARY Table 4 summarises the conclusions, showing that a total of 38 chemical markers have been identified for 35 sponge groups of different taxonomic levels (7 orders, 8 family groups, 7 families and 13 genera). However, only 22 of these markers show some consistency, the re- maining 16 presenting substantial inconstencies preventing their use as reliable markers (‘non-exclusive markers’). The overview pre- sented above clearly demonstrates that, despite huge numbers of compounds isolated from sponges, only a fraction shows potential as chem- ical markers for larger or smaller groups. When the distribution of the thousands of compounds over the various sponge groups is viewed as a whole, most demonstrate an erratic, scattered distribution, occurring either in single species only, or shared by unrelated species. Thus, the 22 markers and 16 non-exclusive markers comprise only a small part of the chemical database. Moreover, markers often are not well-founded because only a handful of species have so far been recorded to contain them. Further exploration may or may not establish their consistent occurrence in the group. Nevertheless, about 260 sponge species (out of a total of about 475) appear to contain secondary metabolites belonging to the 22 markers with utility for sponge chemosystematics. DISCUSSION CHEMISTRY AS A CHARACTER FOR SYS- TEMATICS. The major problem preventing the use of chemotaxonomic markers as characters for sponge classification is their lack of consistency. Not one marker is problem-free, either because it is reported to occur outside the group for which it is supposed to be a marker, or because the marked groups overlap partially. To be useful for classification, markers should either include each others groups or exclude them completely, Seemingly solid markers known from dozens of closely related species in a particular taxo- nomic group have also been reported from a few sponges phylogenetically unrelated from that particular sponge group. Examples include: bromotyrosine compounds (Verongida markers) in the poecilosclerid Jotrochota; furanoterpenes (Dictyo- and Dendroceratida markers) in the haplosclerid Amphimedon; straight-chain acetylenes (Haplosclerida markers) in halichondrid Phakellia and poecilosclerid Raspailia, iso- cyanoterpenes (Halichondrida marker) in the haplosclerid Amphimedon. The same lack of consistency is apparent when the distribution of marker derivatives is viewed within the group of which they are assumed to be characteristic. Examples include: furanosester- terpenes concentrated in Dictyoceratida overlap in distribution with furanoditerpenes concen- trated in Dendroceratida (as discussed in detail above); linear 3-alkylpiperidine derivatives con- centrated in Callyspongiidae and Niphatidae overlap in distribution with macrocycle 3-alkyl- piperidines concentrated in Chalinidae and Petrosiidae; isocyano-diterpenes, concentrated in most Halichondrida s.l. (including Axinellidae) are lacking in Agelasidae. These inconsistencies may be the result of one or more of the following four explanations. Parallel biosynthetic pathways. These may lead to compounds with structural similarity. Sec- ondary metabolites are built from very generally distributed precursors of the primary metabolism. Different enzymes may have the property for allowing the biosynthesis of structurally related compounds. Such an explanation may be valid for the occurrence of bromotyrosine derivatives in Jotrochota, a genus which cannot conceivably be mistaken fora verongid. Tyrosine and bromine are very general molecules in marine organisms and their combination may be achieved by different enzymes. This explanation, however, requires empirical support through biosynthetic experiments. Microsymbiont involvement. Although evidence for the involvement of microsymbionts in prod- uction of sponge secondary metabolites is largely circumstantial, several studies have established that compounds suspected to be of microsymbiont origin were indeed produced by bacteria and fungi isolated from sponges. Examples of sponges known to harbour microsymbionts which are the source of bioactive compounds are: Theonella swinhoei, Halichondria okadai, Mycale sp., Tedania ignis, Calyx podatypa, Dysidea herbacea and Darwinella rosacea. Schmitz (1994) provides a review of such proven or sus- pected cases. On the basis that compounds are universally inconsistently distributed, it is conceivable that all natural products isolated from sponges are of microsymbiont origin. Studies which have allegedly identified sponge cells as the source of 584 a given compound (Faulkner et al., 1994; Garson, 1994; Uriz et al., 1996) did not address the real possibility that isolated sponge cell fractions, free from bacterial cells, contained endosymbionts ultimately involved in the production of com- pounds. Microsymbionts involved in natural products biosynthesis may be species- or group-specific, in which case a close correspondence between sponge phylogeny and compound type is most likely. Such cases would not be easily dis- tinguished from true sponge cell origin of compounds. It is possible that symbionts may be occasionally transferred to other organisms, in- cluding other sponges, which would explain ‘pockets’ of concentrated occurrence of com- pound types in unrelated organisms. Careless specimen handling. During earlier days of natural products exploration, in particular, collected specimens were not always treated in a way required to get unequivocal results. Spec- imens that were not ‘cleaned’ from overgrowing algae and epizootic invertebrates were quoted as sources for compounds not actually produced by them. There are quite a few of these suspected cases, which can only be solved if a voucher including the epibionts has been retained. Labels have also been confused, resulting in reciprocal mismatch of compounds and sponge identities. Such cases are unlikely to be easily solved and will continue to ‘pollute’ the database. Incorrect identification/classification. Some sponge groups are extremely difficult to identify to family or genus level and considerable tax- onomic experience is needed. Moreover, classification of such groups is often a source of disagreement among reigning classification sys- tems. Specimens have often been identified by dozens of taxonomists with very diverse ex- periences and views on the classification, resulting in an almost Babylonic confusion of sponge sources of interesting chemistry, especially in certain genera (e.g. Batzella, Hy- meniacidon, Halichondria, Amphimedon, Reniera, Xestospongia and Cribrochalina). It is obvious that re-examination of vouchers — if at all retained — is needed to correct the more obvious mistakes. It is essential that voucher specimens of important sources of compounds be lodged in museums or collections maintained in perpetuity, and not thrown away after a certain period. IMPACT OF CHEMISTRY ON CURRENT CLASSIFICATION. Several taxonomists have MEMOIRS OF THE QUEENSLAND MUSEUM used chemical data to underbuild existing clas- sifications or to support proposals for changes in the classification. Bergquist (1978) erected the Verongida on the basis of the universal occur- rence of bromotyrosines in the group, which was earlier recognised only at the family level. This proposal remains unchallenged. Bergquist (1980) erected Nepheliospongida (later to be re- named Petrosida for nomenclatorial reasons), based on the occurrence of cyclopropene sterols. This proposal has been challenged on morph- ological (e.g. Van Soest, 1990) as well as chemical grounds (Fromont et al., 1994). Recent chemosystematic analyses (Andersen et al., 1996; Van Soest et al., 1998) yielded strong chemical arguments for the integrity of the Haplosclerida s.l. Van Soest (1991) also used chemical data to unite Axinellidae and Halichondriidae into a Hali- chondrida s.l. Braekman et al. (1992) on the basis of chemical arguments, suggested including Agelasidae into that group, although without making a formal proposal to do so. Chemical data are additional proof that soft bodied Age/as and the sclerosponges Goreauiella and Astrosclera are closely related, supporting morphological indications (Rinehart, 1989; Braekman et al., 1992; Williams & Faulkner, 1996). It appears as if the chemosystematic evidence for order-level relationships are more-or-less exhausted. A final proposal could be made to formally unite Dictyoceratida and Dendro- ceratida into a taxon of the ordinal level, because both groups contain furanoterpenes. Such a proposal has the added advantage of avoiding border disputes at the ordinal level over the position of Dysidea (Boury-Esnault et al., 1990; Bergquist, 1996). Further rearrangement of suborders, families and genera may well be necessary, but will remain within unchallenged ordinal boundaries. Promising chemosystematic conclusions may be expected in the near future especially at the genus level. Examples are Crambe - Monanchora (Poecilosclerida: Crambeidae) sharing the same compounds; some morphological subgroups of Xestospongia (Haplosclerida: Petrosiidae) appear to share similar chemistry. Intriguing prob- lems to be solved are Latrunculia (Hadromerida or Poecilosclerida) - Mycale (Poecilosclerida) relationships, which share similar chemistry but lack morphological correspondence. CHEMOSYSTEMATICS AS A FUTURE DISCIPLINE. It seems imperative that micro- symbiont involvement in the production of CHEMOSYSTEMATICS OF PORIFERA compounds used to underpin classifications is investigated exhaustively. It goes without saying that classifications of sponges should be based on hypotheses of evolutionary developments in the group and not — unwittingly — on those of mierosymbionts. Techniques to investigate whether or not microsymbionts are involved in this process are available, but these are sophist- icated, and lie beyond the reach of the average sponge taxonomist. Thus, certainty of this outcome will be slow in arriving and may depend heavily on non-relevant factors, such as pharm- acological interest in the compounds. In view of the observed widespread inconsistencies it is judged to be unwise at the present time to propose new classification schemes that depend heavily on chemical mformation, Conversely, however, m the face of overwhelming morphological evidence, support from chemistry should also be underlined. In cases where microsymbionts are dem- onstrated to be the source of the compounds, chemosystematic conclusions may still be possible on the basis that many microsymbionts may be obligatory and co-evolved with their sponge hosts. However, different analytical techniques are necessary to arrive at such con- clusions, which involve ‘mapping’ phylogenetic data of microsymbionts on those of the sponge hosts (see for example Van Soest et al., 1998). Chemosystematic studies are hampered by certain aspects of current practice of natural products chemical research. Examples are: bias caused by limited biosassays and extraction pro- cedures, and a widespread reluctance to report on the re-discovery of already known structures. Future studies would benefit greatly from hroad-spectrum bioassays including organisms vr cell-lines from all five kingdoms, to maximise efforts of discovery of useful compounds. Nat- ural products chemists should report their results, irrespective of the news value for chemists, through networks and databases. Confirmation of repeated occurrence of particular compounds in particular sponges has much greater strength of conviction. Chemosystematic conclusions should take into account inconsistent results of previous studies. Rather than dismissing inconvenient data, ii should be attempted to find out why inconsisten- cies are there on a case-by-case basis (parallel biosynthetic pathways, microsymbiont involve- ment, careless specimen handling, or mistaken identification). Such attempts can be made only fruicful in cooperative efforts of taxonomists, chemists, cell biologists and microbiologists. LITERATURE CITED ABRELL, L.M., BORGESON, B. & CREWS, P. 1996. A new polyketide, secocurvularin, Irom the saltwater culture of a sponge-derived fungus, ‘Tetrahedron Letters 37. 8983-8984. ALBRIZIO, S. FATTORUSSO, E., MAGNO, S., MANGONT, A, € PANSINI, M, 1992, Linear diterpenes from the Caribbean sponge Myrmekioderma styx. Journal of Natural Products 55: 1287-1293. ANDERSEN, RJ. SOEST, R.W.M, VAN & KONG, F 1996. 3-Alkylpiperidine alkaloids isolated from marine sponges in the order Haplosclerida. Pp. 301-356, In Pelletier, S.W. (ed.) Alkaloids: Chemical and biological perspectives. Vol, 10, (Pergamon; Oxford). BAKER, BJ, 1996, B-Carboline and isoquinoline alkaloids from marine organisms. Pp. 357-407. In Pelletier, SW. (ed.) Alkaloids: Chemical and tuolozieal perspectives. Vol, 10. (Pergamon: Oxford). BARROW, R.A. MURRAY, L.M., LIM, LK. de CAPON, RJ. 1996. Mirabilios (AT: new alkaloids from a southern Australian marine sponge, Arenochalina mirabilis. Australian Journal of Chemistry 49: 767-773. BARTIK, K., BRAEKMAN, J.C.. DALOZE, D., STOLLER, €., HUYSECOM, 1, VAN DE VYVER, G. & OTTINGER, R. 1987. Topsentins, new toxic bisindgle alkaloids from sponge Topsentia genitrix. Canadian Journal of Chemistry 65: 2118-2121, BERGQUIST, P.R. 1978. Sponges. (Huteliinson: London). 1979, Sponge chemistry - A review. Pp, 383-392. In Lavi, C. & Boury-Esnault, N. (eds) Biologie des spongiaires, Editions du Centre National de la Recherche Scientifique (291) (CNRS: Paris). 1980, The ordinal and subelass classification of the Demospongiae (Porifera): appraisal of the present arrangement, and proposal of a new arder. New Zealand Journal of Zoology 7: 1-6. 1996, The marine fauna of New Zealand: Porifera: Demospongiae Part 5. Dendroceratida and Halisarcida. New Zealand Oceanographic Memoirs 107: 7-53. BERGQUIST, PR. de WELLS, R.J. 1983. Chemo taxonomy of the Porifera, The development and current status of the field. Pp. 1-50. In Scheuer P.J. (el) Marine natural products. Vol. 5. (Academic Press: New York). BERGOQUISI, PR, LAWSON, M.P.. LAVIS, A. de CAMBIE, R,C. 1984, Fatty acid composition and the classification of the Porifera. Biochemical Systematics and Ecology 12; 63-84- BERGOUTST, PR, LAVIS.A, & CAMBIO RA, 1986, Sterol composition and the classification of tlic 586 Porifera. Biochemical Systematics and Ecology 14: 105-112. BERGQUIST, P.R., CAMBIE, R.C. & KERNAN, M.R. 1991. Aaptamine, a taxonomic marker for sponges of the order Hadromerida. Biochemical Systematics and Ecology 19: 289-290. BERLINCK, R.G.S., BRAEKMAN, J.C., DALOZE, D., HALLENGA, K., OTTINGER, R., BRUNO, I. & RICCIO, R. 1990. Two new guanidine alkaloids from the Mediterranean sponge Crambe crambe. Tetrahedron Letters 31: 6531-6534. BLUNT, J.W. & MUNRO, M.H.G. 1998. MarinLit. A database of the literature on marine natural products for use on a Macintosh computer prepared and maintained by the Marine Chemistry Group (Department of Chemistry, University of Canterbury: Canterbury, New Zealand). BOBZIN, S.C. & FAULKNER, D.J. 1989. Diterpenes from the marine sponge Aplysilla polyrhaphis and the dorid nudibranch Chromodoris norrisi. Journal of Organic Chemistry 54: 3902-3907. BOURY-ESNAULT, N., VOS, L. DE, DONADEY, C. & VACELET, J. 1990. Ultrastructure of choano- some and sponge classification. Pp. 237-244. In Rützler, K. (ed.) New perspectives in sponge biology. (Smithsonian Institution Press: Wash- ington DC). BRAEKMAN, J.C., DALOZE, D., STOLLER, C. & SOEST, R.W.M. VAN 1992. Chemotaxonomy of Agelas (Porifera: Demospongiae). Biochemical Systematics and Ecology 20: 417-431. BURRESON, B.J., CHRISTOFERSON, C. & SCHEUER, P.J. 1975. Co-occurrence of two terpenoid isocyanide-formamide pairs in a marine sponge (Halichondria sp.). Tetrahedron 31: 2015-2018. CAPON, R.J., MACLEOD, J.K. & SCAMMELLS, PJ. 1986. The trikentrins, novel indoles from the sponge Trikentrion flabelliforme. Tetrahedron 42: 6545-6550. CAPON, R.J. & MACLEOD, J.K. 1987. Structural and stereochemical studies on marine norterpene peroxides; part 2. Journal of Natural Products 50: 225-229. CAPON, R.J., MACLEOD, J.K. & WILLIS, A.C. 1987. Trunculins A and B; norsesterterpene cyclic peroxides from a marine sponge; Latrunculia brevis. Journal of Organic Chemistry 52: 339-342. CARBALLEIRA, N.M., MALDONADO, M.E., RIVERA, E. & PORRAS, B. 1989. The fatty acid 4,8,12-trimethyltridecanoic as a common constituent of the phospholipids of the sponge families Spirastrellidae and Clionidae. Bio- chemical Systematics and Ecology 17: 311-314. CARBALLEIRA, N.M. & SHALABI, F. 1995. The rare Caribbean sponge Leucosolenia canariensis: phospholipid fatty acids and sterols. Lipids 30: 467-470. CARMELY, S., ROLL, M., LOYA, Y. & KASHMAN, Y. 1989. The structure of eryloside A, a new antitumour and antifungal 4-methylated steroidal MEMOIRS OF THE QUEENSLAND MUSEUM glycoside from the sponge Erylus lendenfeldi. Journal of Natural Products 52: 167-170. CARNEY, J.R., SCHEUER, P.J. & KELLY-BORGES, M. 1993. A new bastadin from the sponge Psammaplysilla purpurea. Journal of Natural Products 56: 153-157. CATALAN, C.A.M., LAKSHMI, V., SCHMITZ, F.J. & DJERASSI, C. 1982. Minor and trace sterols from marine invertebrates, 39. Cyclocholest- 5-en-3B-ol; a novel cyclopropyl sterol from Spheciospongia. Steroids 40: 455-463. CIMINIELLO, P., FATTORUSSO, E., MAGNO, S. & MANGONI, A. 1989. Clathridine and its zinc complex, novel metabolites from the marine sponge Clathrina clathrus. Tetrahedron 45: 3873-3878. CIMINIELLO, P., CONSTANTINO, V., FATTO- RUSSO, E., MAGNO, S. & MANGONI, A. 1992. Haliclonol, a new tetrahydropyranone from the Caribbean sponge Haliclona hogarthi. Heterocycles 34: 765-770. CIMINIELLO, P., FATTORUSSO, E., FORINO, M., MAGNO, S. & PANSINI, M. 1997. Chemistry of Verongida sponges. 8. Bromocompounds from the Mediterranean sponges Aplysina aerophoba and Aplysina cavernicola. Tetrahedron 53: 6565-6572. COLON, M., GUEVARA, P., GERWICK, W.H. & BALLANTINE, D. 1987. 5’-Hydroxyiso- avrainvilleol, a new diphenylmethane derivative from the tropical green alga Avrainvillea nigricans. Journal of Natural Products 50: 368-374. CONSTANTINO, V., FATTORUSSO, E., MANGONI, A. & PANSINI, M. 1994, Three new brominated and iodinated tyrosine derivatives from lotrochota birotulata, a non-Verongida sponge. Journal of Natural Products 57: 1552-1556. D'AURIA, M.V., PALOMA, L.G., MINALE, L., RICCIO, R., ZAMPELLA, A. & DEBITUS, C. 1993. Metabolites ofthe New Caledonian sponge Cladocroce incurvata. Journal of Natural Products 56: 418-423. DUMDEI, E.J.. BLUNT, J.W., MUNRO, H.M.G., BATTERSHILL, C.N. & PAGE, M.J. 1998. The whys and whats of sponge chemistry: why chemists extract sponges and what problems this cause. Pp. 353-364. In Watanabe, Y. & Fusetani, N (eds) Sponge sciences. Multidisciplinary perspectives. (Springer: Tokyo). EDER, C., SCHUPP, P., PROKSCH, P., WRAY, V., STEUBE, K., MULLER, C.E., FROBENIUS, W., HERDERICH, M. & SOEST, R.W.M. VAN 1998. Bioactive pyridoacridine alkaloids from the Micronesian sponge Oceanapia sp. Journal of Natural Products 61: 301-305. FAULKNER, D.J., UNSON, M.D. & BEWLEY, C.A. 1994, The chemistry of some sponges and their symbionts. Pure and Applied Chemistry 66(10/11): 1983-1990. CHEMOSYSTEMATICS OF PORIFERA FROMONT, J., KERR, S., KERR, R., RIDDLE, M. & MURPHY, P. 1994. Chemotaxonomic relation- ships within, and comparisons between, the orders Haplosclerida and Petrosida (Porifera: Demospongiae) using sterol complements. Bio- chemical Systematics and Ecology 22: 735-752. FUSETANI, N., MATSUNAGA, S. & KONOSU, S. 1981. Halistanol sulfate; an antimicrobial steroid sulfate from Halichondria cf. moorei Bergquist. Tetrahedron Letters 22: 1985-1988. FUSETANI, N., KATO, Y., MATSUNAGA, S. & HASHIMOTO, K. 1983. Polyacetylene alcohol; inhibitor of cell division in fertilized sea urchin eggs; from Petrosia sp. Tethrahedron Letters 24: 2771-2774. FUSETANI, N, & MATSUNAGA, S. 1993. Bioactive sponge peptides. Chemical Review 93: 1793-1806. GARSON, M.J. 1994. The biosynthesis of sponge secondary metabolites: Why it is important. Pp. 427-440. In Soest, R.W.M. van, Kempen, T.M.G. van & Braekman, J.C. (eds) Sponges in time and space. (Balkema: Rotterdam). GIULIO, A. DE, ROSA, S. DE, VICENZO, G. DI & ZAVODNIK, N. 1989. Terpenoids from the North Adriatic sponge Spongia officinalis. Journal of Natural Products 52: 1258-1262. GREEN, G. 1977. Ecology of toxicity in marine sponges. Marine Biology 40: 207-215. GUELLA, G., MANCINI, L, GUERRIERO, A. & PIETRA, F. 1985. New furano-sesquiterpenoids from Mediterranean sponges. Helvetica Chimica Acta 68: 1276-1282. HERB, R., CARROLL, A.R., YOSHIDA, W.Y., SCHEUER, P.J. & PAUL, V.J. 1990. Poly- alkylated cyclopentindoles: cytotoxic fish antifeedants from a sponge, Axinella sp. Tetrahedron 46: 3089-3092. HIGGS, M.D. & FAULKNER, D.J. 1978. Plakortin; an antibiotic from Plakortis halichondrioides. Journal of Organic Chemistry 43: 3454-3457. ISHIBASHI, M., KUROSAKI, M., MIKAMI, Y. & KOBYASHI, J. 1993. The stereochemistry of variabilin, an antimocrobial sesterterpene isolated from marine sponges. Natural Products Letters 3: 189-192, ISHIDA, K., ISHIBASHI, M., SHIGEMORI, H., SASAKI, T. & KOBAYASHI, J. 1992. Agelasine G, a new antileukemic alkaloid from the Okinawan marine sponge Age/as sp. Chemical Pharmacological Bulletin. 40: 766-767. ITOH, T., SICA, D. & DJERASSI, C. 1983. Minor and trace sterols. 36. (24S)-24H-Isocalysterol; a new steroidal cyclopropene from sponge Calyx nicaensis. Journal of Organic Chemistry 48: 890-892. JAYATILAKE, G.S., BAKER, B.J. & McCLINTOCK, J.B. 1997. Rhapsamine, a cytotoxin from the Antarctic sponge Leucetta leptorhaphis. Tetrahedron Letters 38: 7507-7510. JUREK, J., SCHEUER, P.J. & KELLY-BORGES, M. 1994. Two steroidal alkaloids from a sponge, 587 Corticium sp. Journal of Natural Products 57: 1004-1007 KERR, R.G. & KELLY-BORGES, M. 1994. Bio- chemical and morphological heterogeneity in the Caribbean sponge Xestospongia muta (Petrosida: Petrosiidae). Pp. 65-74. In Soest, R.W.M. van, Kempen, T.M.G. van & Braekman, J.C. (eds) Sponges in time and space. (Balkema: Rotterdam). KOBAYASHI, J., MURAYAMA, T., KOSUGE, F., ISHIBASHI, M., KOBAYASHI, H., OHIZUMI, Y., OHTA, T., NOZOE, S. & SASAKI, T. 1990. Niphatesines A-D, new antineoplastic pyridine alkaloids from the Okinawan marine sponge Niphates sp. Journal of the Chemical Society Perkin Transactions 1: 3301-3303. KOBAYASHI, J., CHENG, J., ISHIBASHI, M., WALCHLI, M.R., YAMAMURA, S. & OHIZUMI, Y. 1991. Penaresidin A and B, two novel azetidine alkaloids with potent actomyosin ATPase-activating activity from the Okinawan marine sponge Penares sp. Journal of the Chemical Society Perkin Transactions 1: 1135-1137 KOBAYASHI, M. & KITAGAWA, I. 1998. Likely microbial participation in the production of bioactive marine sponge chemical constituents. Pp. 379-389, In Watanabe, Y. & Fusetani, N (eds) Sponge sciences. Multidisciplinary perspectives. (Springer: Tokyo). KOMOTO, S. & MCCONNELL, O.J, 1988, New anti- cancer indole alkaloids from marine sponges. European Patent Applications 111: 1-10. KONDO, K., SHIGEMORI, H., KIKUCHI, Y., ISHIBASHI, M., SASAKI, T. & KOBAYASHI, J. 1992. Ircinals A and B from the Okinawan marine sponge /rcinia sp.: plausible biogenetic precursors of manzamine alkaloids. Journal of Organic Chemistry 57: 2480-2483. KONG, F. & ANDERSEN, R.J. 1996, Polymastiamides B-F, novel steroid/aminoacid conjugates isolated from the Norwegian marine sponge Polymastia boletiformis. Journal of Natural Products 59: 379-385. KONIG, G.M. & WRIGHT, A.D. 1993. Agelorin-A and Agelorin-B, and ll-epi-fistularin-3, three new antibacterial fistularin-3 derivatives from the tropical marine sponge Agelas oroides. Heterocycles 36: 1351-1358. LI, H., MATSUNAGA, S. & FUSETANI, N. 1995. Halicylindramides A-C, antifungal and cytotoxic depsipeptides from the marine sponge Halichondria cylindrata, Journal of Medical Chemistry 38: 338-343. LIAAEN-JENSEN, S, RENSTROEM, B., RAMDAHL, T., HALLENSTVET, M. & BERGQUIST, P.R. 1982. Carotenoids of marine sponges - Accumulation and transformation highest in Poecilosclerida and Axinellida. Biochemical Systematics and Ecology 10: 167-174, 588 MANES, L.V., BAKUS, G.J. & CREWS, P. 1984. Bioactive marine sponge norditerpene and sesterpene peroxides. Tetrahedron Letters 25: 931-934. McCABE, T., CLARDY, J., MINALE, L., PIZZA, C., ZOLLO, F. & RICCIO, R. 1982. A triterpenoid pigment with the isomalabaricane skeleton from the marine sponge Stelletta sp. (Choristida). Tetrahedron Letters 23: 3307-3310. McDONALD, L.A., SWERSEY, J.C., IRELAND, C.M., CARROLL, A.R., COLL, J.C., BOWDEN, B.F., FAIRCHILD, C.R. & CORNELL, L. 1995. Botryllamides A-D, new brominated tyrosine derivatives from styelid ascidians of the genus Botryllus. Tetrahedron 51: 5237-5244. MOLINSKI, T.F., FAHY, E., FAULKNER, D.J., VAN DUYNE, G.D. & CLARDY, J. 1988. Petro- samine, a pigment from marine sponge Petrosia sp. Journal of Organic Chemistry 53: 1340-1341. MORRIS, S.A. & ANDERSEN, R.J. 1989. Nitrogen- ous metabolites from the deep water sponge Hexadella sp. Canadian Journal of Chemistry 67: 677-681. 1990. Brominated bis(indole) alkaloids from the marine sponge Hexadella sp. Tetrahedron 46: 715-720. NAKAMURA, H., WU, H., ABE, R., KOBAYASHI, J., OHIZUMI, Y. & HIRATA, Y. 1983. Structures of physiologically active substances from the Okinawa Sea sponges (neural blocking activity). Tennen Yuki Kagobotsu Toronkai 26: 118-125. PATIL, A.D., KOKKE, W.C., COCHRAN, S., FRANCIS, T.A., TOMSZEK, T. & WESTLEY, J.W. 1992. Brominated polyacetylenic acids from the marine sponge Xestospongia muta: inhitors of HIV protease. Journal of Natural Products 55: 1170-1177. PERRY, N.B., BLUNT, J.W., MCCOMBS, J.D. & MUNRO, H.M.G. 1986. Discorhabdin C; a highly cytotoxic pigment from a sponge of the genus Latrunculia. Journal of Organic Chemistry 51: 5476-5478. PERRY, N.B., BLUNT, J.W., MUNRO, H.M.G., HIGA, T. £ SAKAI, R. 1988. Discorhabdin D, an antitumour alkaloid from the sponges Latrunculia and Prianos sp. Journal of Organic Chemistry 53: 4127-4128. PORDESIMO, E.O. & SCHMITZ, F.J. 1990. New bastadins from the sponge /anthella basta. Journal of Organic Chemistry 55: 4704-4709, PRINSEP, M.R., BLUNT, J.W. & MUNRO, H.M.G. 1989. A new sterol sulfate from the marine sponge Stylopus australis. Journal of Natural Products 52: 657-659. RAVI, B.N., KOKKE, W.C.M.C., DELSETH, C. & DJERASSI, C. 1978. Isolation and structure of 26;27-cycloapsterol (petrosterol), a cyclo- propane-containing sterol from sponge Halichondria sp. Tetrahedron Letters 1978: 4379-4380. MEMOIRS OF THE QUEENSLAND MUSEUM RINEHART, K.L. 1989. Biologically active marine natural products. Pure and Applied Chemistry 61: 525-528. ROSSER, R. M. & FAULKNER, D.J. 1984. Two steroidal alkaloids from marine sponge; Plakina sp. Journal of Organic Chemistry 49: 5157-5160. RYU, G., MATSUNAGA, S., FUSETANI, N. 1994. Bioactive marine metabolites. 64, Discodermin E, a cytotoxic and antimicrobial tetradecapeptide, from the marine sponge Discodermia kiiensis. Tetrahedron Letters 35: 8251-8254, SAKAI, R., HIGA, T., JEFFORD, C.W. & BER- NARDINELLI, G. 1986. Manzamine A; an antitumor alkaloid from a sponge. Journal of the American Chemical Society 108: 6404-6405. SALOMON, C.E. & FAULKNER, D.J. 1996. Sagitol, a pyridoacridine alkaloid from the sponge Oceanapia sagittaria. Tetrahedron Letters 37: 9147-9148. à SARMA, A.S., DAUM, T. & MULLER, W.E.G. 1993. Secondary metabolites from marine sponges. (Ullstein Mosby: Berlin). SCHMITZ, F.J. 1994, Cytotoxic compounds from sponges and associated microfauna. Pp. 485-496. In Soest, R.W.M. van, Kempen, T.M.G. van & Braekman, J.C. (eds) Sponges in time and space. (Balkema: Rotterdam). SCHMITZ, F.J., AGARWAL, S.K., GUNASEKERA, S.P., SCHMIDT, P.G. & SHOOLERY, J.N. 1983. Amphimedine; new aromatic alkaloid from a Pacific sponge, Amphimedon sp. Journal of the American Chemical Society 105: 4835-4836. SENNETT, S.H., POMPONI, S.A. & WRIGHT, A.E. 1992. Diterpene metabolites from two chemotypes of the marine sponge Myrmekioderma styx. Journal of Natural Products 55: 1421-1429. SILVA, C.J. & DJERASSI, C. 1991. Sterols in marine invertebrates. Part 64. Isolation, stereochemistry and biosynthesis of sormosterol, a novel cyclopropane-containing sponge sterol. Collected Czech Chemical Communications 56: 1093-1105. SOEST, R.W.M. VAN 1980, Marine sponges from Curacao and other Caribbean localities. Part II. Haplosclerida. Studies on the Fauna of Curaçao and other Caribbean Islands 62: 1-173. 1990a. Toward a phylogenetic classification of sponges. Pp. 344-348. In Rützler, K. (ed.) New perspectives in sponge biology. (Smithsonian Institution Press: Washington). 1990b. Demosponge higher taxa classification re-examined. Pp. 54-71. In Reitner, J. & Keupp, H. (eds) Fossil and recent sponges. (Springer Verlag: Heidelberg). 1994, Demosponge distribution patterns. Pp. 213-224. In Soest, R. W.M. van, Kempen, T.M.G. van & Braekman, J.C. (eds) Sponges in time and space. (Balkema: Rotterdam). SOEST, R. W.M. VAN, DIAZ, M.C. & POMPONI, S.A. 1990. Phylogenetic classification of the CHEMOSYSTEMATICS OF PORIFERA Halichondrids (Porifera, Demospongiae). Beaufortia 40: 15-62. SOEST, R.W.M. VAN, BRAEKMAN, J.C., FAULK- NER, D.J., HAJDU, E., HARPER, M.K. & VACELET, J. 1996. The genus Batzella; a chemosystematic problem. Bulletin de l'Institut Royal des Sciences Naturelle de Belgique 66(suppl.): 89-101. SOEST, R.W.M. VAN, FUSETANI, N. & ANDER- SEN, R.J. 1998, Straight-chain acetylenes as chemostaxonomic markers of the marine Haplosclerida. Pp. 3-30. In Watanabe, Y. & Fusetani, N. (eds) Sponge sciences. Multi- disciplinary perspectives. (Springer: Tokyo). STIERLE, D.B. & FAULKNER, D.J. 1979. Meta- bolites of the marine sponge Chondrosia collectrix. Journal of Organic Chemistry 44: 964-968. STIERLE, A.A, CARDELLINA, J.H. & SINGLE- TON, F.L. 1991, Benzothiazoles from a putative bacterial symbiont of the marine sponge Tedania ignis. Tetrahedron Letters 32: 4847-4848. STONARD, R.J. & ANDERSEN, R.J. 1980. Linear peptide alkaloids from the sponge Cliona celata (Grant). Celenamides C and D. Canadian Journal of Chemistry 58: 2121-2126. SUPRIYONO, A., SCHWARZ, B., WRAY, V., WITTE, L., MULLER, W.E.G., SOEST, R.W.M. VAN, SUMARYONO, W. & PROKSCH, P. 1995. Bioactive alkaloids from the tropical marine sponge Axinella carteri. Zeitschrift für Naturforschung 50c: 669-674. 589 TOTH, S.I. & SCHMITZ, F.J. 1994, Two new cytotoxic peroxide-containing acids from a New Guinea sponge Callyspongia sp. Journal of Natural Products 57: 123-127. TSUKAMOTO, S., KATO, H., HIROTA, H. & FUSETANI, N. 1996. Pseudoceratine: a new antifouling spermidine derivative from the marine sponge Pseudoceratina purpurea, Tetrahedron Letters 37: 1439-1440. URIZ , M.J., BECERRO, M.-A., TUT, J.M. € TURON, X. 1996. Location of toxicity within the Mediter- ranean sponge Crambe crambe (Demospongiae: Poecilosclerida). Marine Biology 124: 583-590, WELLS, R.J. 1976. A novel peroxyketal from a sponge; genus Chondrilla. Tetrahedron Letters 1976: 2637-2638. WILLIAMS, D.H. & FAULKNER, D.J. 1996. N-methylated ageliferins from the sponge Astrosclera willeyana from Pohnpei. Tetrahedron 52 (15): 5381-5390. WRIGHT, A.E., POMPONI, S.A., McCONNELL, O.J., KOHMOTO, S. & McCARTHY, P.J. 1987. Curcuphenol & curcudiol, sesquiterpene phenols from Didiscus flavus. Journal of Natural Products 50: 976-978. WU, H., NAKAMURA, H., KOBAYASHI, J., OHIZUMI, Y. & HIRATA, Y. 1984. Agelasine -e & -f, novel monocyclic diterpenoids with 9-methyladeninium unit possessing inhibitory effects on Na,K-ATPase from Okinawan sponge Agelas nakamurai. Tetrahedron Letters 25: 3719-3722. 590 MEMOIRS OF THE QUEENSLAND MUSEUM BIOLOGY OF SPONGE NATURAL PRODUCTS. Memoirs of the Queensland Museum 44: 590. 1999:- This is to announce the EC-MAS 3 project, which started April 1, 1998. The biological and chemical aspects of selected sponge natural products (secondary metabolites) of interest to human use will be studied to obtain understanding of: 1) the cellular origin and possible microsymbiont involvement, and 2) the ecological significance of sponge secondary metabolites, and 3) the patterns in these processes enabling rationalisation of exploration for and exploitation of sponge secondary metabolites. The results will have a direct bearing on policy decisions concerning industrial production of sponge secondary metabolites, which are too difficult or too costly to synthesise. A major deliverable product of the proposed research will be the formulation ofa standard protocol of research steps needed as a basis for such policy decisions. The research will be structured in three phases: 1) exploration and pattern recognition, 2) testing of hypotheses using experiments with selected sponges, 3) protocol construction. Initially, investigations will be directed towards two sponge groups (Haplosclerida and Halichondrida) and towards a limited number of molecule types. Known secondary metabolite occurrence will direct exploration for related sponges and related secondary metabolites. For the experimental phase a choice for 3-4 target sponges will be made based on suspected production of secondary metabolites by own sponge cells (1-2 target sponges) or microsymbionts (1-2 target sponges). The biological aspects include: determination of the identities and phylogenetic relationships of bioactive sponges; within-sponge spatial distribution of sponge cells and microsymbiont cells; experimental observation of variability of biological activity of selected sponges in various environmental (biotic and abiotic) situations; identification of target microsymbiont cells; fractionation, isolation and culture of target sponge cells. The chemical aspects include: extraction, isolation and structure determination of selected bioactive compounds; development of qualitative and quantitative analytical methods for their spatial distribution in the sponge and for their distribution between and within different species. Summary of methodologies: Sponges will be collected using SCUBA (shallow water) and/or dredges (deep water) and photographed upon collection. Various types of fixations of material will be made immediately after removal from the water. Voucher specimens will be studied for identity and phylogeny using routine morphological as well as molecular (18S / 28S rDNA) characters. Collected sponges preserved in methanol or as freeze-dried material will be extracted with methanol and dichloromethane. The primary extracts will then be tested for their biotoxicity using an invertebrate bioassay organism, the Artemia toxicity test, and several prokaryote and eukaryote bioassay organisms (bacteria, fungi, yeast). Cytological analyses will consist of two different approaches, one using glutaraldehyde-fixed material, the other using live sponges: 1) sponges will be fixed in glutaraldehyde, (a) for microsymbiont detection; thick sections will be stained with suitable fluorochromes and viewed by fluorescence microscopy and confocal scanning light microscopy. If microsymbionts are present, populations will be characterised by different parameters. Microsymbionts will be further identified by fluorescence in situ hybridisation using rRNA-targeted oligonucleotides as probes; (b) for sponge cell spatial distribution, samples will be postfixed in 1% osmium tetroxide and thin sections will be studied by Transmission Electron Microscopy (TEM). 2) Live sponges will be dissociated into single-cell suspensions. Recognition of secondary metabolite production will be realised using two advanced techniques: (a) cell fractionation into pure cell populations using continuous or discontinuous Percoll gradients; (b) symbiont-free sponge cultures, initiated either from pure cell populations or from dissociated sponge cell suspensions. Experimental observations will be made in situ using various types of manipulations (caging, artificial standard lesions, confrontation with substrate competitors, crude extract assays with substrate competitors and potential predators). O Porifera, secondary metabolites, microsymbionts, chemical ecology, exploration and exploitation. R.W.M. van Soest (email: soest@bio.uva.nl), Department of Coelenterates and Porifera, Institute for Systematics and Population Biology (Zoologisch Museum), University of Amsterdam, PO.Box 94766, 1090 AT Amsterdam, The Netherlands; G. Van De Ver & E. Richelle-Maurer, Physiologie Cellulaire et Genetique des Levures, Université Libre de Bruxelles, Bd du Triomphe - CP 244, B-1050 Bruxelles, Belgium; C. Woldringh, J.C. Braekman & R. Tavares, Université Libre de Bruxelles, Faculté des Sciences, Laboratoire de Chimie Bio-Organique CP 160/70, Avenue ED. Roosevelt 50, B-1050 Bruxelles, Belgium; 1 June 1998, PATTERNS OF INTRA AND INTERSPECIFIC GENETIC DIVERGENCE IN MARINE SPONGES ANTONIO M. SOLE-CAVA AND NICOLE BOURY-ESNAULT Solé-Cava, A. M. & Boury-Esnault, N. 1999 06 30: Patterns of intra and interspecific genetic divergence in marine sponges. Memoirs of the Queensland Museum 44: 591-601. Brisbane. ISSN 0079-8835, Since the first molecular systematic studies on marine sponges in the 1980’s, many papers have been published about levels of allozyme divergence between conspecific and congeneric sponge populations. Those genetic studies have indicated that sponges are more divergent than other marine invertebrates, a fact that was attributed to the high levels of genetic variation and morphological conservativeness found in Porifera. However, an analysis of 55 interspecific and 87 intraspecific pairwise genetic identity (/) values indicates a more complex picture. This study found that the average of / over all interspecific comparisons (/=0.42) was not much smaller than that found among other marine invertebrates (150,54), and the frequency distribution of /, for intraspecific comparisons, appears to be bimodal. Some genera were consistently highly divergent (/<0.30; Cinachyrella, Oscarella, Cliona, Spirastrella and Tethya), whereas others were within the normal range of gene divergence (0.40 < / < 0.80; Chondrosia, Suberites, Petrosia, Plakina and Phyllospongia). Furthermore, in the genera Axinella, Chondrilla and Clathrina, both low and high levels of intrageneric genetic differentiation were found (0.13 < 7 « 0.82). This pattern may reflect a large variance in the evolutionary age of genera in sponges, with very large levels of intrageneric gene divergence for some. We conclude with two non-mutually exclusive scenarios: a) genetic identity levels are too variable among sponge species to be of any use to evaluate taxonomic rank above species, or b) the range of evolutionary divergence in some genera of sponges is so broad that they may need revision. O Porifera, gene divergence, allozymes, heterozygosity, molecular systematics, larval dispersal. Antonio Solé-Cava (email: sole@centroin.com. br), Departamento de Genética, Instituto de Biologia, Universidade Federal do Rio de Janeiro, Bloco A, CCS, Ilha do Fundáo, 21941-490 - Rio de Janeiro, Brazil; and Department of Environmental and Evolutionary Biology, University of Liverpool, Port Erin, Isle of Man, United Kingdom; Nicole Boury-Esnault,Centre d'Océanologie de Marseille, Station Marine d'Endoume, UMR-CNRS DIMAR 6540, 41 rue de la Batterie-des-Lions, F-13007, Marseille, France; 16 March 1999, For marine organisms genetic markers have been extremely useful both for estimating levels of gene flow in structured populations (Burton, 1996), and for the detection of sibling species (Knowlton, 1993; Thorpe & Solé-Cava, 1994), Allozyme electrophoresis has become the method of choice for alpha (i.e. at the species level) molecular systematics of marine organisms (Thorpe & Solé-Cava, 1994; Knowlton & Weigt, 1997). The main advantage of allozyme electrophoresis for taxonomic studies is that it represents an independent set of characters for the detection of sibling species (Solé-Cava & Thorpe, 1987). Genetic markers such as allozymes are particularly powerful for alpha-taxonomy (Hillis et al., 1996) because they can be used to detect reproductive isolation in sympatry (i.e. the biological species concept of Mayr, 1981), and describe unambiguous diagnostic characters (ie. the phylogenetic species concept of Cracraft, 1987). In addition, as they are ubiquitous, allozymes offer a yardstick to compare levels of evolutionary divergence in relation to taxonomic rank in widely different taxonomic groups. Through molecular methods, it has become easier to verify whether ichthyologists, entomologists and spongologists infer the same thing when they talk about generic taxa in their respective groups. Since 1978, over 3000 intraspecific and interspecific allozyme comparisons have been performed between marine populations (literature data based on search on the Aquatic Sciences and Fisheries Abstracts database, between 1978 and 1998). The most commonly used measure of genetic similarity is the index of gene identity (1; Nei, 1972), which varies from 1.0 (=complete identity) to zero. An analysis of the large database of genetic studies, mostly for terrestrial vertebrates and Drosophila, demonstrated that mean levels of gene identity were, as expected, very different when conspecific populations, congeneric species or confamilial genera were compared (Thorpe, 1982; Thorpe, 1983). It was shown that less than 5% of all conspecific comparisons fell below an identity level of 0.8 (Thorpe, 1982). Consequently, the / value of 0.8 has been used as a threshold for deciding about specific differentiation using allozyme data to define species, especially for comparing allopatric popul- ations, where the more straightforward use of diagnostic loci (sensu Ayala, 1983) is not possible, and the biological species concept (Mayr, 1981) is not practical (Aron & Solé-Cava, 1991; Claridge et al., 1997). However, that value may be still too high for making decisions about the taxonomic rank of some marine invertebrates from geographically distant populations. This is because the number of allozyme loci detectable in marine invertebrates is usually smaller than in other organisms, with a consequent increase in the variance of estimates of gene identity (Nei, 1978), and also because gene flow is expected to be limited by geographical distance, with a consequent lowering of gene identities (Palumbi, 1992). Considering that decisions about species’ borders in complex groups, using genetic attributes, are best taken using what has become known as ‘fuzzy logic’ (Van Regenmortel, 1997), the use of a threshold value becomes very important for the comparison of allopatric sponge populations. Allozyme electrophoresis was first employed for molecular systematics of sponge populations by Solé-Cava & Thorpe (1986) and recently for sponge population genetics (Benzie etal., 1994), Molecular data are also very useful for inferring patterns of genetic flow linked to larval dispersal (Burton, 1996). Sponge larvae are usually short lived (e.g. Borojevic, 1970; Fry, 1971; Sarà & Vacelet, 1973), which suggests that geographical distance could determine levels of gene differentiation in sponge populations. On the other hand, the pattern of gene flow observed in many marine invertebrates is often chaotic, depending mostly on rare but long-ranging broadcasting events (Johnson & Black, 1984). It would be interesting, therefore, to verify whether gene flow among sponge populations is also chaotic or supports the ‘isolation by distance’ model of genetic differentiation (Wright, 1978). It has been suggested that Porifera might MEMOIRS OF THE QUEENSLAND MUSEUM display much higher levels of interspecific gene divergence than other invertebrates, possibly due to the presence, in the former, of high levels of gene variation (Solé-Cava et al., 1991a; Klautau et al., 1994; Boury-Esnault et al.,1999). If this is true, then a re-calibration of the threshold value of conspecific gene identity should be performed, in order to reduce possible type I errors (i.e. deciding that putative species are different when they are not), due to a shift in gene identities between sponge populations in relation to other organisms. This calibration would be fundamental both for the analysis of evolutionary rates in the Porifera and for the continuing study on putative cosmopolitanism in the group. The aims of this paper are to: 1) correlate levels of intraspecific gene identity with geographical distance, in order to estimate the importance of larval dispersal to the composition of sponge populations; 2) verify whether patterns of interspecific gene similarity in sponges are indeed different from those of other marine invertebrates; and 3) re-evaluate the threshold gene identity value for making taxonomic decisions for sponges. MATERIALS AND METHODS Data were gathered from the literature and from unpublished studies made by our laboratory (see references listed in the table legends). Whenever necessary, values of mean heterozygosity and genetic identity (Nei, 1978) were calculated from tables of gene frequency. Geographical distances were measured as the shortest distances by sea, using a large scale map (1 cm=60km; Christie et al., 1995). The possible relationship between pairwise geographical and genetic distances for intraspecific populations was tested using a Mantel test, with 1,000 replicates (Sokal & Rohlf, 1995). Pooled data of pairwise gene identity measures of intraspecific, interspecific and intergeneric comparisons were used to construct frequency histograms, in a similar way as those built by Thorpe (1982, 1983). The significance of differences between mean identity levels in interspecific (intra- generic) and intergeneric comparisons was tested using a Mann-Whitney U test (Sokal & Rohlf, 1995). RESULTS From all available data, 87 intraspecific, 55 interspecific and 8 intergeneric comparisons were compiled (Tables 1-3 respectively). No significant correlation (Mantel test; P>0.40) was found GENETIC DIVERGENCE IN SPONGES 593 TABLE 1. Levels of gene identity between conspecific populations. Key: Km, distance in kilometers; NL, number of loci; /, unbiased mean genetic identity (Nei, 1978); H, mean Hardy-Weinberg expected heterozygosity (Nei, 1972). References: 1, Benzie et al. (1994); 2, Klautau et al. (in press); 3, Cristiano Lazoski (unpublished results); 4, Solé-Cava et al. (1992); 5, Boury-Esnault et al. (1992); 6, Bavestrello & Sarà (1992); 7, Boury-Esnault et al. (1999); 8, Sarà et al. (1992). Species Locality 1 Locality 2 Km NL I h Ref fascia Willis Island (Aust) | Middle Island 8.7 6 1.00 0.19 1 Willis Island Magdelaine 44 6 0.90 0.26 1 Middle Island (Aust) Magdelaine 52 6 0.89 0.27 1 Lihou NE (Aust) Lihou SW 80 6 0.93 0.22 1 Magdelaine (Aust) Lihou SW 175 6 0.69 0.28 1 Magdelaine Lihou NE 200 6 0.77 0.22 1 Willis Island Lihou SW 210 6 0.59 0.25 1 Middle Island Lihou SW 210 6 0.62 0,21 1 Willis Island Lihou NE 245 6 0.64 0.20 1 Middle Island Lihou NE 245 6 0.67 0.16 1 2. Chondrilla nucula Marseille (Fr) Ligurian (It) 350 9 0.91 0.11 2 3. Chondrilla sp.3 Anjos (Braz) Praia do Forno 2 9 0.99 0.27 2 Búzios (Braz) Anjos 30 9 0.87 0.56 2 Buzios Praia do Forno 30 9 0.90 0.30 2 Itacuruca (Braz) Picinguaba 60 9 0.95 0.24 2 Buzios Itacuruca 240 9 0.95 0.30 2 Anjos Itacuruca 240 9 0.92 0.27 2 Praia do Forno (Braz) Itacuruca 240 9 0.94 0.30 2 Picinguaba (Braz) MOR 280 9 0.91 0.22 2 Anjos Picinguaba 300 9 0.95 0.22 2 Praia do Forno Picinguaba 300 9 0.89 0.25 2 Búzios Picinguaba 310 9 0.89 0.25 2 Ttacuruca Ilha do Mel 340 9 0.91 0.27 2 Anjos Ilha do Mel 560 9 0.89 0.25 2 Praia do Forno Ilha do Mel 560 9 0.98 0.27 2 Buzios Ilha do Mel 700 9 0.98 0.28 2 Noronha (Braz) Buzios 2400 9 0.84 0.34 2 Noronha Anjos 2400 9 0.88 0.31 2 Noronha Praia do Forno 2400 9 0.91 0.34 2 Noronha Itacuruca 2600 9 0.88 0.33 2 Noronha Picinguaba 2700 9 0.90 0.28 2 Noronha Ilha do Mel 3200 9 0.84 0.59 2 bi id La Ciota (Fr) Callelongue 17 13 0.96 0.14 3 La Ciota Endoume 25 13 1.00 0.16 3 Callelongue (Fr) Endoume 8 13 0.99 0.12 3 La vesse (Fr) Endoume 10 13 0.99 0.11 3 La vesse La Ciota 15 13 0.97 0.12 3 La vesse Callelongue 2 13 0.99 0.08 3 5. Chondrosia sp. Bermudas Recife 6640 13 0.89 0.27 3 Bermudas Büzios 8300 13 0.95 0.33 3 Bermudas Forno 8330 13 0.95 0.30 3 Bermudas Angra 8600 13 0.92 0.28 3 Recife (Braz) Buzios 1860 13 0.94 0.27 3 594 MEMOIRS OF THE QUEENSLAND MUSEUM TABLE 1. Continued. 5. Chondrosia sp. (cont.) Recife Forno 1890 13 0.94 0.25 3 Recife Angra 2160 13 0.94 0.22 3 Büzios Forno 30 13 0.93 0.30 3 Büzios Angra 300 13 0.93 0.28 3 Forno (Braz) Angra 270 13 0.93 0.25 3 6. Collospongia auris Willis Island Middle Island 8.7 6 1.00 0.30 1 Willis Island Lihou SW 210 6 0.95 0.31 1 Middle Island Lihou SW 210 6 0.91 0.28 1 M cian La Vesse Riou (Fr) 25 16 0.97 0.24 4 8. Oscarella lobularis La Vesse Riou 25 16 1.00 0.11 4 La Vesse Riou 25 12 0.98 0.12 5 9. Petrosia clavata Paraggi (It) Zoagli (It) 2 9 0.96 0.12 6 10. Petrosia ficiformis Paraggi Zoagli 9 0.90 0.09 6 o ospongia Willis Island Middle Island 8.7 6 0.96 0.38 1 Willis Island Holmes 210 6 0.89 0.35 1 Willis Island Lihou SW 210 6 0.85 0.26 1 Middle Island Holmes 210 6 0.86 0,34 1 Middle Island Lihou SW 210 6 0.87 0.32 1 Holmes (Aust) Osprey 300 6 0.74 0.34 1 Holmes Lihou SW 370 6 0.77 0.27 1 Willis Island Osprey 430 6 0.74 0.31 1 Middle Island Osprey 430 6 0.76 0.38 1 Osprey (Aust) Lihou SW 630 6 0.49 0.24 1 a bosponete Willis Island Middle Island 8.7 6 0.91 0.30 1 Diamond (Aust) Lihou SW 50 6 0.89 0.25 1 Lihou NE (Aust) Lihou SW 80 6 0.93 0.26 1 Diamond Lihou NE 120 6 0.86 0.31 l Willis Island Diamond 175 6 0.96 0.35 1 Lihou NE Marion 175 6 0.91 0.26 1 Lihou SW (Aust) Marion 175 6 0.99 0.19 1 Willis Island Holmes 210 6 0.96 0.34 1 Willis Island Lihou SW 210 6 0.85 0.27 1 Middle Island Holmes 210 6 0.90 0.24 1 Middle Island Lihou SW 210 6 0.76 0.19 1 Diamond Marion 220 6 0.93 0.24 1 Willis Island Lihou NE 245 6 0.85 0.36 1 Middle Island Lihou NE 245 6 0.80 0.27 1 Holmes Diamond 300 6 0.93 0.29 1 Holmes Lihou SW 370 6 0.85 0.27 1 Willis Island Marion 380 6 0.87 0.30 1 Middle Island Diamond 380 6 0.91 0.24 1 Middle Island Marion 380 6 0.80 0.19 1 Holmes Lihou NE 420 6 0.85 0.31 1 Holmes Marion 500 6 0.85 0.24 1 13. Spirastrella hartmani San Blas 1 (Pan) San Blas 2 1 8 0.95 0.30 7 San Blas 1 Galeta (Pan) 100 8 0.87 0.28 7 San Blas 2 Galeta 100 8 0.95 0.29 7 14. T. citrina Marsala (It) Torbay (GB) 3600 11 0.74 0.15 8 Average - - 0.89 0.26 - GENETIC DIVERGENCE IN SPONGES 10 n * a , . ? * + + $ $ * * 09* e * l y» E. > 0,8 oe MY » I9 " * * 06 T 05 $ 04 + ——— c + —a -1 10 1 2 3 4 logio Km 45 4 40 = —O- Interspecific 35 LE Intraspecific £x 30 71 H Interuenerio 3 25 = 20 E 15 10 54 04 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 I 74 —O- Group 1 $5 — Group 2 54 —&- Group 3 F a 47 v 3 d 2 3 [e 2 B ] B 0 E 01 02 03 04 05,06 0. 08 09 1 I FIG. 1. A, Relationship between geographical distance (in logio km) and pairwise gene identities (Mantel test; 1,000 replicates; P>0.40). B, Frequency histogram of gene identity (/) and taxonomic rank for species of sponges. C, Frequency histogram of gene identity (7) and taxonomic group. Group 1 - Chondrosia, Suberites, Petrosia, Plakina and Phyllospongia; Group 2 - Axinella, Chondrilla and Clathrina; Group 3 - Oscarella, Cinachyrella, Tethya, Cliona and Spirastrella. between geographic distance and genetic identity (Fig. 1A). The empirical frequency distribution of intraspecific (Table 1), interspecific (Table 2) and intergeneric (Table 3) gene identities studied on the different genera of Demospongiae and Calcarea (Fig. 1B) was similar to that found for other organisms (Thorpe & Solé-Cava, 1994). The average of / over all interspecific sponge comparisons was 0.42, which is similar to that found among other marine invertebrates (770.54). However, the distribution of interspecific pairwise gene identities in sponges was bimodal (Fig. 1C). Species of some genera were consistently highly divergent (/<0.30; ‘Group 3°: Cinachyrella, Oscarella, Cliona, Spirastrella and Tethya), whereas others were within the normal range of gene divergence (0.400.10). The bimodal distribution of interspecific gene identities is more puzzling, and seems to result from different patterns of gene divergence in different sponge genera. The genera analysed can be roughly broken into three groups in relation to levels of interspecific gene identities: 1) genera whose species have similar levels of gene identity as other invertebrates (Chondrosia, Petrosia, Phyllospongia, Plakina and Suberites); 2) genera where some pairwise species comparisons give very low identity values (I«0.3), whereas others have levels of gene identity comparable to those of other organisms (0.4 350-370 /22-25 380/30 540-560/30 (rare) - - Amphiasters 15-35 15-30 15-40 10-30 20-47/7-10 25/83 Discotriaenes 110-120 90-120 z - È Styles 280-610/2-3 370-650/2-3 - + a MEMOIRS OF THE QUEENSLAND MUSEUM 1. Alectona wallichii, A-B, Pits on the surface of excavated scleractinian skeleton from Tuléar, showing a double system of concentric and radiating lines (Scale bar: A=152um; B=51 um). C-H, Smooth and tu diactines from Hawaiian specimens (Scale bar: C=374um; D=126um; E=1264m; F-19.6um; G=136um; H=142um). Scale bar on bottom right. PLANKTONIC PROPAGULES OF ALECTONA 6 Amphiasters, or more correctly sanidasteroid discorhabds, have short, blunt actines. Actines are generally equal and disposed in two whorls near the middle of the axis, but spicules with additional or unequal actines are frequent. Spicules are microspined, except in the central part of the axis between the two whorls of actines. A few small spicules, with an axis as short as 7.5m, resemble the nodulose amphiasters found in the genus Thoosa. Size 15-35/2-2.5um in Hawaii, 10-40um in Tuléar. Discotriaenes of the embryos have an irregularly circular outline, sometimes roughly triangular. Edge of the disc is slightly inwardly curved, with outer central part depressed. The inner surface bears a few small tubercles. The rhabdome is rarely acerate, most often with a blunt or inflated tip which may divide. An axial canal is clearly visible in the rhabdome, but cannot be traced in the disc. Size 110-120um diameter with a rhabdome 20-30/8-10um in Hawaii, 90-120um in Tuléar (the latter observed in only a single specimen). Styles of the embryos, are straight, slightly enlarged at some distance from the rounded end, with a large axial canal up to 1.2um diameter. Two size categories are present in specimens containing advanced embryos: 280-360/1.5-3um and 610-650/2-3um, the former dominating in Hawaii and the latter in Tuléar. The diactines are dispersed without order in the choanosome inside cavities. Walls of the papillary canals contain a greater concentration of longitudinally arranged tuberculate diactines (Fig. 2D). Amphiasters are dispersed throughout the whole tissue, although more numerous in papillary canals. The styles and discotriaenes are present only in embryos, the discotriaenes as an external layer, the styles as three fascicles made up of two spicules crossing in the centre of the embryo. Living tissue. Living tissue of specimens from Hawaii is rather dense, with few canals (Fig. 3A-B). Choanocyte chambers are spherical, 14-20um diameter. The mesohyl contains a large number of bacteria of the morphotypes usually found in bacteriosponges. The walls of the papillary canals are reinforced by dense, intertwined collagen fascicles containing elongated collencytes (Fig. 2D). Specimens from Tuléar have the same general features as those from Hawaii, although their tissues are more poorly preserved. Choanocyte chambers are not visible. One specimen that has uy) probably suffered from delayed fixation contains many rod-like bacteria, different in shape from the usual symbiotic bacteria. Reproductive stages (Fig. 3A-F). Several stages of sexual reproduction are simultaneously present in the choanosome of most specimens. They are described here mostly from the Hawaiian specimens, which are better preserved and display more numerous stages. Spermatogenesis occurs in spermatic follicles, which are spherical cavities, 25-40um diameter, surrounded by a thin pinacocyte envelope. They contain densely stained spermatids or spermatocytes in various stages grouped as morulae at the centre of the follicle. In the most advanced stage observed (Fig. 3B), the spermatozoa are dispersed within the spermatic follicle. They have an elongated head measuring 2-3/0.5-0.6um, and a long flagellum. Oocytes are rare, with only two stages of their development observed. The youngest is a rounded cell lying in the mesohyl without a special collagen envelope, 20um diameter, with the cytoplasm containing a few large inclusions, and a prominent, 64m diameter nucleus containing a 2-3um nucleolus. Another stage, which was observed only once in a thick section with poor definition, is 90um diameter with an 18um nucleus and a 4um nucleolus. The cytoplasm contains large vitelline inclusions, and is surrounded by a follicular envelope of flattened cells and by a dense collagen envelope. Embryos are spherical and uniform in size, approximately 200-320um, regardless of their developmental stage (Fig. 3A). However, the most mature stages are slightly larger and elongated. Segmentation is total and equal. A four-cell stage has been observed, with apparently equal blastomeres 70-110um diameter, filled up with heterogeneous vitelline inclusions 3-8um diameter (Fig. 3C). Blastomeres divide without formation ofa blastocoele. The early stages, up to the formation of the larval skeleton, are surrounded by a thin outer envelope made up of very thin, elongated cells, and by a dense inner layer, up to l10um thick, made of collagen fascicles. Similar fascicles also individually wrap the blastomeres (Fig. 3C, D). This collagen development, which is highly unusual in sponge embryos, is maintained during the following stages. The most advanced embryos observed are made up of cells, 10-15um size with inclusions 5pm diameter, which are widely dispersed in a loose collagen matrix also containing dispersed 632 MEMOIRS OF THE QUEENSLAND MUSEUM FIG. 2. Alectona wallichii. A-C, Spicules from Hawaiian specimens. A, Amphiasters. B, C, Discotriaenes (Scale bar: A=1 1 um, 6.2um, 12.5um from left to right; B=40um; C=44um). D, Semi-thin section through the papilla after desilicification, showing collagen strands, collencytes and axial filament (arrows) in the spicule ghosts (Scale bar=52um). E-F, Spicules from Tuléar specimens. E, Amphiasters. F, Smooth, spinose and tuberculate diactines (Scale bar: E=15.8um; F=86um, 724m, 98m from left to right). Scale bar on bottom right. PLANKTONIC PROPAGULES OF ALECTONA 63 symbiotic bacteria. A group of elongated cells perpendicular to the embryo surface is observed near one pole of the mature embryo on a slightly prominent button (Fig. 3A). Most cells still appear to be undifferentiated. Contrary to the free-living larvae of Alectona millari (Garrone, 1974), these embryos possess neither choanocyte chambers nor spherulous cells. The larval skeleton appears when the blastomeres have grown to approximately 201m diameter and are still apparently undifferentiated. The three spicule categories appear simultaneously, first as very thin (1-1.5um) styles and incomplete discotriaenes, in which the smaller disc observed attained only 20um diameter (Fig. 3E). Amphiasters are recognisably the same as in later stages, but are less numerous than those found in the most advanced embryos. In advanced stages (Fig. 3F), the embryo is surrounded by a single layer of discotriaenes in which the shaft is inwardly directed. The cover is first discontinuous, then becomes continuous with a total number of approximately 12 discotriaenes. Styles, consistently six in number in each embryo, are disposed in three fascicles made up of two parallel spicules, which cross approximately at right angles to each other near the centre of the embryo. Amphiasters are mostly located near the discotriaene cover, although a few are dispersed within the internal tissue. The most advanced embryos observed had relatively short styles, which will probably eventually elongate and finally rearrange at one pole, as observed in Thoosa (Topsent, 1904), before being shed through the canals at later develop- mental stages. Specimens from Tuléar also have several reproductive stages. Rare spermatic follicles were observed, with the same features as those seen in Hawaii. Embryos were mostly in early stages of development, with blastomeres 30-50um diameter, and without spicules. They differ from those of Hawaiian samples by a greater development of the collagen envelope, which is up to 25um thick and made up of intertwined collagen fascicles (Fig. 3D). The fascicles in between the blastomeres are also larger (approximately 10pm) than in Hawaiian material. Mature stages retaining a larval skeleton were not observed in sections, although they were present in one specimen, as indicated by the presence of larval spicules on dissociated spicule preparations (Table 1). uy REMARKS. As summarised in Pang (1973, 1977), Alectona wallichii was originally described in the genus Gummina by Carter (1874), from isolated acanthodiactines collected from the Cape of Good Hope at 146-182m depth. These remarkable spicules were previously described by Bowerbank (1864), also as isolated spicules washed off corals. Carter (1879a) later succeeded in finding the whole sponge and its complete spiculation from the same skeletons of Stylaster studied by Bowerbank, and subsequently transferred the species to Corticium. He discussed its possible excavating habitus and gave three probable localities inferred from the known occurrences of the highly diagnostic spicules: the ‘South Sea’, Cape of Good Hope and the Seychelles. Carter (1879b) later transferred the species to Alectona, where it has been considered either as the type-species of the genus A/ectona or as a synonym of A. millari (Laubenfels, 1936), although no additional material was known until the species was rediscovered in the NW Pacific by Bavestrello et al (1998). It is likely that A. wallichii was already present in the Early Miocene, as suggested by tuberculate spicules found in sediments from the W Atlantic figured by Wiedenmayer (1994). Although some individual variations were observed between the various specimens from the Indian and Pacific Oceans (Table 1), the species appears to be well-characterised by its tuberculate diactines. These diactines are generally smaller in material from Tuléar (with the exception of specimen 1371/1, which has the largest megascleres). In all specimens, however, megascleres were smaller than in those recorded in the type-specimens by Carter (1874, 1879a) and in material from the NW Pacific (Bavestrello et al., 1998). The most important difference between material from Hawaii and Tuléar, however, is the presence of spinose or bumped diactines, which are intermediary stages between tuberculate and smooth diactines, in Tuléar specimens. The presence of these intermediary spicules in Tuléar specimens, which have few or no advanced embryos, could indicate that diactine production, first appearing as smooth spicules, is reduced in actively reproducing specimens. The discotriaenes and styles are present only in specimens containing advanced embryos. 634 MEMOIRS OF THE QUEENSLAND MUSEUM PLANKTONIC PROPAGULES OF ALECTONA 635 Alectona mesatlantica sp. nov. (Fig. 4) MATERIAL. HOLOTYPE: MNHN D JV 63: Saint Peter & Saint Paul Rocks, mid-Atlantic Ridge, 00?94"N, 25?29'W, 2030m depth, 6.1.1998, coll. ‘Nautile’ submersible, R.V. ‘Saint Paul" (sample SP13-16). ETYMOLOGY. From mes, Greek, middle, and Atlantic, pertaining to the type locality on the Mid-Atlantic Ridge. HABITAT. In deep water (2030m depth), near to Saint Peter & Saint Paul Rocks in the Equatorial Atlantic; excavating large cavities in calcareous rock of probable organic origin, encrusted by ferromanganese oxyhydroxides. DESCRIPTION. Morphology and living tissue (Fig. 4A). The sponge is a fleshy mass growing in tunnels or in subspherical cavities, up to 5cm maximum dimension, single but irregular in shape. Communication with the outside is by a few papillary canals, 2-3mm diameter, ending in papillae of the same size with a single aperture, Imm diameter. Colour is white in alcohol. The tissue is rubbery, compact with few aquiferous canals 2-3mm diameter. The mesohyl is typical of bacteriosponges, containing numerous, densely packed symbiotic bacteria. There are no visible cells with inclusions. Choanocyte chambers were poorly preserved, 22-30um diameter. Aquiferous canals are lined by elongated cells aligned parallel to the canals, collagen fascicles and diactines. The skeleton, composed of acanthodiactines and amphiasters, is developed only along the canals and near the border of the cavities. It is reduced and nearly absent in the choanosome. Styles and discotriaenes are found only in spiculate embryos. Cavities display the characteristic pitted surface of excavating sponges, with subcircular pits 25-42um diameter (Fig. 4B). Pits have an irregular surface, with concentric lines and secondary small holes. This structure is visible on all the components of the limestone construction, indicating that it is not related to the microstructure of the bored substrate. Spicules (Fig. 4C-F). Acanthodiactines are of very irregular shape, usually with a swelling and bent in the middle, sometimes with a more-or-less developed third ray, or one of the two rays absent. Axial canal are large, frequently vesiculated or making a loop near the middle of the spicule or near the occasional branching of a third ray. Spines are acerate, short, irregularly distributed. Size: 410-530/21-28um without spines. Amphiasters, with a relatively thick axis, are microspined except near the centre of the axis, with rounded actines predominantly central, sometimes disposed in two irregular whorls especially in the shorter spicules. Size: 20-60/3-7m, Amphiasters of the spiculate embryos have a thinner axis than those of the body. Styles of embryos are very thin, flexuous, slightly enlarged at some distance from the rounded end, mostly broken on the slides. Size: up to 1125/4-Sum. Discotriaenes of embryos have a short rhabd (20-40um), frequently dichotomous or with lateral expansions. Disc is 130-150um diameter, circular or slightly triangular, with the three branches of the axial canal clearly visible. Incomplete spicules, including a ‘triaene’ with free clads whose surface is irregular, were observed in embryos in the early stage of spiculogenesis. A few large, irregular siliceous plates, probably of foreign origin, were observed in dissociated spicule preparations. Their position in the sponge tissue is unknown. Acanthodiactines and asters are very rare or absent in the choanosome. They are more abundant near the walls of the excavation, and in the lining of the choanosomal and papillary FIG. 3. Alectona wallichii. Stages in reproduction in specimens from Hawaii (except D, from Tuléar). A, Semi-thin section of desilicified tissue, general view of choanosome with young embryos at diverse stages in segmentation, and an embryo with ghost of discotriaene cover (arrow) enlarged in the inset (Scale bar=218um, inset=142um). B, Semi-thin section of desilicified tissue, nearly mature spermatic follicle and advanced embryo with ghost of discotriaene cover (arrows) (Scale bar=56um). C, Semi-thin section of desilicified tissue, early stage in segmentation with collagen strands between the blastomeres (Scale bar=56um). D, Semi-thin section of desilicified tissue, specimen from Tuléar, embryo with thick collagen strands (Scale bar=38um). E, Non-desilicified thick section, early stage in segmentation (arrow) and three embryos with developing discotriaenes and styles; a fully developed discotriaene of a mature embryo on bottom left (Scale bar=200pm). F, Non-desilicified thick section, advanced embryo with discotriaene cover and three fascicles of two styles (Scale bar=132um). Scale bar on bottom right. MEMOIRS OF THE QUEENSLAND MUSEUM FIG. 4. Alectona mesatlantica sp. nov. A, Holotype (Scale bar=15.6mm). B, Pits on the excavated surface (Scale bar=64um). C-D, Acanthodiactine (Scale bar: C=84um; D=78um). E., Amphiasters (Scale bar20um, 17.21m, 17.2um from left to right). F, Discotriaene (Scale bar=72um). Scale bar on bottom right. PLANKTONIC PROPAGULES OF ALECTONA canals. Styles and discotriaenes are found only in advanced embryos, with the same arrangement as in 4, wallichii. Reproductive stages. The tissue contains numerous spermatic follicles, approximately 25um diameter, at various stages of development ranging from a central mass of dense, 4um diameter cells, to nearly mature spermatozoa with an elongated head 3/1 um. The single stage of oogenesis observed was an oocyte with a poorly preserved cytoplasm, 40/20um, with a nucleolate nucleus. Numerous ovoid embryos are present in the tissue. They display several developmental stages with a uniform size, 200-230/150m. The earliest stage observed is a four-cell blastula with equal blastomeres, 100um diameter, with numerous vitelline inclusions, 2-8um diameter. These blastulae are surrounded by a thin envelope. Collagen strands, up to 35m thick, are located between this envelope and the blastomeres. Thinner strands also wrap the blastomeres individually. Spicules first appear in embryos with an undetermined number of blastomeres, whose size is reduced to approximately 201m, with inclusions smaller than Sum. In these embryos the amphiasters are thinner than in the adult tissue, and the disc of the discotriaenes is incomplete, with the four axial filaments clearly visible. In the most advanced stages containing a complete larval skeleton, the blastomeres are smaller (approximately lOum diameter), but apparently still undifferentiated and contain heterogeneous inclusions up to 4um diameter. Intercellular bacteria are numerous in the embryo tissue. Discotriaenes are disposed as an outer cover over the embryo, and are surrounded by a thin tissue strand. Long styles are grouped either in a fascicle protruding from one pole into the maternal tissue, or in three perpendicular fascicles of two spicules. REMARKS. The new species is characterised by the large size of its acanthodiactines, amphiasters with rounded spines, the unusually large cavities it excavates, the small size and peculiar aspect of the pits, and the distribution of megascleres which are virtually absent from the choanosome. The most closely related species appears to be the common North Atlantic-Mediterranean A. millari Carter, 1879, which bores smaller cavities with pits up to 100um diameter, thus considerably larger than those of A. mesatlantica, which are smaller than 50um, and whose diactines are smaller (Bavestrello et al., 1998) - although a Mediterranean specimen of A. millari has been recorded with tuberculate diactines 240-460/27-54um (Pulitzer-Finali, 1983). This is the deepest record for the genus. Alectona millari, previously considered to be a deep-sea species (Topsent, 1900), has now been found at 0.5m depth and is not recorded deeper than 1190m. The present species is also the deepest record for an excavating sponge after Cliona levispira, recorded at a depth of 2165m (Topsent, 1928b). DISCUSSION SEXUAL ORIGIN OF THE EMBRYOS. Observations made here have some similarities to those reported for Thoosa armata by Topsent (1904). Both these data demonstrate, in my opinion, that the ‘armoured gemmules” are clearly larvae of sexual origin. However, this point is equivocal and requires further discussion because these are peculiar larvae, with characteristics generally found not in larvae, but rather in asexual gemmules or buds, and also because an asexual origin for ‘larvae’ of incubating demosponges has been alleged by several authors (Wilson, 1894; Sivaramakrishnan, 1951; Bergquist et al., 1970) — although their interpretation has been challenged and has never really gained general acceptance (Bergquist, 1978; Simpson, 1984). According to these authors, “asexual larvae’ may be found together with stages of sexual reproduction. The occurrence of sexual products (oocytes, spermatic follicles and indisputable embryos) simultaneously with the armoured bodies in Alectona is thus not full evidence that they are of sexual origin. However, considering the observed sequence it appears very unlikely that the armoured bodies would not be derived from the early segmentation stages. It may also be argued that oocytes are fewer in number than embryos, but this is a frequent phenomenon in incubating sponges. The presence of collagen fascicles between the blastomeres is highly unusual in sponge early embryos, as discussed below, but this is not a convincing argument for an asexual origin because collagen fascicles are also unknown between gemmular archaeocytes. Finally, the present data, particularly the observed sequence in development, leave no doubt as to the sexual origin ofthese reproductive bodies. These embryos, however, display a number of unique peculiarities among poriferan larvae. 1) They possess a larval spicule skeleton which is absent from the adult, a feature known only so far in the hexactinellid trichimella larva. 2) They are devoid of flagellae. 3) They have a strong development of collagen structures appearing at the end of the oocyte development and during the first segmentation stages. 4) They presumably have a long planktonic life with special flotation devices. Such larvae clearly cannot be included in any other types of embryonic development in Porifera (Borojevic, 1970; Simpson, 1984; Fell, 1989), and the new term ‘hoplitomella’ (from hoplites, Greek, armoured, with heavy armour, and the suffix mella, used for some other types of sponge larvae), is proposed here. The hoplitomella is presently known only from the two genera Alectona and Thoosa. MODE OF DEVELOPMENT. The early embryonic development of the hoplitomella larva is normal in that the first cleavages are equal and give rise to a solid stereoblastula, as in most incubating demosponges. However, a strong development of collagen occurs, both around the mature oocyte and in the embryo. The origin of collagen fascicles is not clear, and an ultrastructural study is needed to resolve this point. However, it is speculated that collagen fascicles surrounding the oocyte could be made either by the maternal sponge tissue, or by the oocyte itself. A synthesis of collagen fibrils by oocytes, which has been very rarely reported in animals (Wischnitzer, 1966), has been described in some oviparous demosponges (Gallissian & Vacelet, 1976; Watanabe & Masuda, 1990). The collagen envelope in early stages of the embryo may be derived from the oocyte envelope, as in the direct development of Tetilla, in which radiating fiber bundles of the egg surface are enclosed in the perivitelline space after fertilisation (Watanabe, 1978). This hypothesis, however, would neither explain the greater thickness of the embryo envelope, nor the presence of thick collagen fascicles between the blastomeres. It appears that in hoplitomella development the blastomeres are able to secrete thick collagen strands, even during the first cleavages. Although collagen fibrils are present in mature sponge larvae, they appear at a relatively late stage when most cell categories are already differentiated and they never form such thick, intertwined fascicles. Another peculiarity is the early synthesis of the larval spicule skeleton, which also occurs at a time when most of the blastomeres are apparently MEMOIRS OF THE QUEENSLAND MUSEUM still undifferentiated. The precocious appearance of siliceous spicules has been described in some demosponge embryos (Brien & Meewis, 1938; Fell, 1969; Simpson, 1984), but this occurs after differentiation of most cell types, It appears that the cells at the surface of the embryo, which in other sponges differentiate into a flagellated layer, differentiate here into discrete sclerocytes. As for precocious collagen synthesis, an ultrastructural study of the phenomenon would be of the greatest interest. Differentiation of macromeres and micromeres, which is the subsequent step in embryonic development in other poriferans having a stereoblastula, resulting in differentiation of flagellated cells, appears to be skipped in this scheme. There is no early separation of a flagellated cell lineage. This separation probably occurs in a late developmental stage when the first choanocyte chambers differentiate in the free larvae, as described in A. millari (Garrone, 1974). Embryos observed here, as well as the free larval stages found in the plankton (Trégouboff, 1942: Garrone, 1974), are devoid of surface flagellae. This is again unique amongst the Porifera, where larvae are always more-or-less mobile due to the presence of flagellated cells, even in the creeping larvae of Polymastia (Borojevic, 1967). Flagellated cells are absent only in the development of Tetilla, whose zygotes develop directly without any larval stage (Watanabe, 1978). It remains to be confirmed, however, whether the mature larva is actually devoid of flagellae. The elongated cells grouped onasmall protuberance in a pole of the embryo of A, wallichii (Fig. 3A) vaguely suggest a future small tuft of flagellae, although it has not yet been reported in free larva of A. millari (Trégouboff, 1942: Garrone, 1974). After long debate, it has finally been convincingly demonstrated, both in Calcarea and in Demospongiae, that at metamorphosis the choanocytes of the young sponge may derive from larval flagellated cells (Amano & Hori, 1993, 1996, 1998). There is, however, evidence from other species for an absence of such a ‘reversal of layers’ (Misevic et al., 1990). The present observations, reporting the absence of flagellated cells in alectonid larvae, suggest that a direct lineage from larval flagellated cells to choanocytes cannot be retained in this development. The absence of surface flagella, and consequently of swimming behaviour, is PLANKTONIC PROPAGULES OF ALECTONA surprising in these larvae which are the only sponge elements regularly found in full planktonic conditions. It might have been supposed that such larvae with a presumed relatively long pelagic life would have a well developed swimming apparatus. Although the swimming devices of sponge larvae do not allow significant movement, they could play an important role by changing the vertical position in the water column or for final microhabitat selection before settlement. Instead, the hoplitomella appears to rely on flotation devices produced by its long protruding styles and, at least at the stages which have been observed, is completely at the mercy of the currents. The possibility that a swimming apparatus will develop at the end of the pelagic life, assuring refinement in the choice of substratum as in other poriferan larvae, is unknown but cannot be excluded at present. The absence of surface flagella would exclude photonegative behaviour, which is nevertheless likely as most species of Alectona and Thoosa are sciaphilous. Whether or not this peculiar mode of larval life has some effect on their dispersal ability and geographic distribution remains to be investigated. GALLERY SIZE AND PIT ORNAMENTATION. The shape, size, and organisation of the camerate borings of excavating sponges have been tentatively used in systematics of clionids and for identification of species of the ichnogenus Entobia Bronn, 1837, with an evident interest as indicators of palaeoenvironments (Bromley, 1970, 1978; Bromley & D'Alessandro, 1984; Pleydell & Jones, 1988; Bromley et al., 1990; Edinger & Risk, 1996). The chambers of the two species of Alectona appears, from present data, to be larger than in most clionids, especially those of A. mesatlantica which range up to 5cm diameter. Another difference is that these large chambers are simple, apparently without any growth stage in which they would be camerate or catenate as in clionids. The possible use of the fine features of the pits for a correlation between extant species and ichnospecies of excavating sponges has yet to be explored. The genus A/ectona appears to have a special ornamentation of the pits, with a double system of concentric and radiating lines whatever the nature of the substratum. This has been observed in the skeleton of the calcified calcareous sponge Petrobiona massiliana bored by A. millari (Omnes, 1991), and is confirmed here in A. wallichii. The small size and the peculiar ornamentation of pits in A. mesatlantica, whose spiculation is not very different from that of A. millari, appears as an interesting additional taxonomic feature. This demonstrates that characterisation of both Recent species and the ichnospecies of genus Entobia using microsculpture of the boring walls is worth considering. SYSTEMATIC POSITION OF ALECTONA. The systematic position of A/ectona and related genera ( Thoosa and Delectona Laubenfels, 1936, possibly with Thooce Laubenfels, 1936 and Annandalia Topsent 1928, which are probable synonyms of Thoosa) is puzzling. These genera are generally classified within Clionidae (Hadromerida) based on their excavating habit — a character of doubtful value given that it occurs within other orders of Demospongiae — and the presence of amphiasters that resemble those of Cliothosa Topsent, 1905, which undoubtedly belongs to the Clionidae (Topsent, 1928a; Rosell & Uriz, 1997). This classification has been questioned by several authors. Topsent (1891, 1900, 1928a) considered that the spinose diactines of A. millari were giant microscleres derived from oxyasters rather than from mega- scleres, and suggested that clionids, and in particular Alectona, were intermediary between Hadromerida and Tetractinellida, whereas Alander (1942) firmly classified Alectona and Thoosa in Tetractinellida. De Laubenfels (1936) suggested that Clionidae could be divided into two groups, possibly of the subfamily rank, i.e. Cliona and related genera, and Thoosa and related genera. More recently, recognition of Thoosidae as a distinct family of Hadromerida has been proposed, first as Alectonidae (Rosell, 1996), then as Thoosidae (Rosell & Uriz, 1997). This distinction appears fully justified from my results. The presence of discotriaenes in larvae, to which may be added now the unique features of the sexual development, precludes their classification in Clionidae. These spicules are undoubtedly of tetraxonial origin, and thus also preclude the classification of Alectona in Hadromerida. Alternatively, they may be considered as an ‘ancient adult character’ (Jágersten, 1972), still present during embryonic development but disappearing in the adult during ontogeny. I will leave for the moment the family Thoosidae incertae sedis. Based on presently accepted criteria, its classification in Tetractinellida 640 would rest only on the presence of discotriaenes in larvae of Alectona, whereas the tetractinellid character is unclear in Thoosa and Delectona. Furthermore, the larval development is markedly different from both Tetractinellida and Hadromerida, which are oviparous. Possibly a new order will be needed for these sponges when additional data becomes available on their full life cycle, cytology and molecular sequences. ACKNOWLEDGEMENTS I am grateful to Roger Hékinian (IFREMER/GM, Brest) chief scientist of the ‘SAINT PAUL’ cruise, organised and sponsored by IFREMER; Bertrand Sichler (scientific diver), and Michel Ségonzac (IFREMER/EP-CENTOB, Brest) for providing specimens of the new species. 1 thank Chantal Bézac and Christian Marchal for technical assistance. LITERATURE CITED ALANDER, H. 1942. Sponges from the Swedish west- coast and adjacent waters. Ph.D. Thesis (Univer- sity of Lund: Lund). AMANO, S. & HORI, I. 1993. Metamorphosis of calcareous sponges. II. Cell rearrangement and differentiation in metamorphosis. Invertebrate Reproduction and Development 24(1): 13-26. 1996. Transdifferentiation of larval flagellated cells to choanocytes in the metamorphosis of the demosponge Haliclona permollis. Biological Bulletin 190: 161-172. 1998. Larval flagellated cells transform to choan- ocytes in Demosponge metamorphis. Pp. 167-177. In Watanabe, Y. & Fusetani, N. (eds) Sponge Sciences. Multidisciplinary perspect- ives. (Springer-Verlag: Tokyo). BAVESTRELLO, G., CALCINAI, B., CERRANO, C. & SARA, M. 1998. Alectona species from Western Pacific (Demospongiae: Clionidae). Journal of the Marine Biological Association of the United Kingdom 78: 59-73. BERGQUIST, P.R. 1978. Sponges. (Hutchinson: London). BERGQUIST, P.R. & SINCLAIR, M.E. 1973. Seasonal variation in settlement and spiculation of sponge larvae. Marine Biology 20: 35-44. BERGQUIST, P.R., SINCLAIR, M.E. & HOGG, J.J. 1970, Adaptation to intertidal existence: reproductive cycles and larval behaviour in Demospongiae. Pp. 247-271. In Fry, W. (ed.) Biology of the Porifera. Zoological Society of London Symposia. Vol. 25. (Academic Press: London). BOROJEVIC, R. 1967. La ponte et le développement de Polymastia robusta (Démosponges). Cahiers de Biologie Marine 8: 1-6. MEMOIRS OF THE QUEENSLAND MUSEUM 1970. Différentiation cellulaire dans l'embryo- génese et la morphogénése chez les Spongiaires. Pp. 467-490. In Fry, W.G. (ed.) The biology of the Porifera. Zoological Society of London Symposia. Vol. 25. (Academic Press: London). BOURY-ESNAULT, N. & VACELET, J. 1994, Preliminary studies on the organization and development of a hexactinellid sponge from a Mediterranean cave, Oopsacas minuta. Pp. 407-415. In Soest, R.W.M.van, Kempen, Th.M.G.van & Braekman, J.C. (eds) Sponges in time and space; Biology, Chemistry, Paleontology. (Balkema: Rotterdam). BOWERBANK, J.S. 1864. A monograph of the British Spongiadae. (Robert Hardwicke: London). BRIEN, P. 1967. L'embryogénése d'une éponge d'eau douce africaine: Potamolepis stendelli (Jafté). Larves des potamolepides et des spongillides. Polyphylétisme des éponges d'eau douce. Bulletin de la Classe des Sciences 53: 752-777. BRIEN, P. & MEEWIS, H. 1938. Contribution à l'étude de l'embryogénése des Spongillidae. Archives de Biologie de Liege 49: 178-250. BROMLEY, R.G. 1970. Borings as trace fossils and Entobia cretacea Portlock, as an example. Pp. 49-90. In Crimes, T.P. & Harper, J.C. (eds) Trace fossils. (Geological Journal, Special Issue 3: Liverpool). 1978. Bioerosion of Bermuda reefs. Palaeogeog- raphy, Palaeoclimatology, Palaeoecology 23: 169-197. BROMLEY, R.G. & D'ALESSANDRO, A. 1984, The ichnogenus Entobia from the Miocene, Pliocene and Pleistocene of southern Italy. Rivista Italiana di Paleontologia e Stratigrafia. 90: 227-296, BROMLEY, R.G., HANKEN, N.M. & ASGAARD, U. 1990. Shallow marine bioerosion: preliminary results of an experimental study. Bulletin of the Geological Society of Denmark 38: 85-99, CARTER, H.J. 1874. Description and figures of deep-sea sponges and their spicules, from the Atlantic Ocean, dredged up on board H.M.S. ‘Porcupine’, chiefly in 1869; with figures and descriptions of some remarkable spicules from the Agulhas Shoal and Colon, Panama. Annals and Magazine of Natural History 14: 207-221, 245-257. 1879a. Contributions to our knowledge of the Spongida. Annals and Magazine of Natural History 5: 284-304, 343-360. 1879b. On a new species of excavating sponge (Alectona millari); and on a new species of Raphidotheca (R. affinis). Journal of the Royal Microscopical Society 2: 493-499. DE VOS, L., RUTZLER, K., BOURY-ESNAULT, N., DONADEY, C. & VACELET, J. 1991. Atlas de morphologie des Eponges-Atlas of sponge morphology. (Smithsonian Institution Press: Washington D.C.). PLANKTONIC PROPAGULES OF ALECTONA EDINGER, E.N. & RISK, M.J. 1996. Sponge borehole size as a relative measure of bioerosion and paleoproductivity. Lethaia 29: 275-286. FELL, P.E. 1969, The involvement of nurse cells in oogenesis embryonic development in the marine sponge, Haliclona ecbasis. Journal of Morphology 127: 133-149. 1989. Porifera. Pp. 1-41, In Adiyodi, K.G. & Adiyodi, R.G. (eds) Reproductive biology of Invertebrates. Vol. 4. (Oxford & IBH Publishing C*: New Delhi) GALLISSIAN, M.F. & VACELET, J. 1976. Ultrastructure de quelques stades de l'ovogenese de spongiaires du genre Verongia (Dictyoceratida). Annales des Sciences Naturelles, Biologie Animale 18: 381-404. GARRONE, R. 1974. Ultrastructure d'une “gemmule armée" planctonique d'éponge Clionidae. Inclusions fibrillaires et genése du collagéne. Archives d'Anatomie Microscopique et de Morphologie Expérimentale 63(2): 163-182. JAGERSTEN, G. 1972. Evolution of the metazoan life cycle. (Academic Press: London). KARAWAIEW, W. 1896. Uber ein neues Radiolar aus Villafranca. Zoologischer Anzeiger 19: 501. 1897, Nachtrag zu meinem Artikel: Uber ein neues Radiolar aus Villafranca. Zoologischer Anzeiger 20: 193, LAUBENFELS, M.W. DE 1936. A discussion of the sponge fauna of the Dry Tortugas in particular and the West Indies in general, with material for a revision of the families and orders of the Porifera. Publications of the Carnegie Institution of Washington, Papers from Tortugas Laboratory 30: 1-225. MISEVIC, G.N., SCHLUP, V. & BURGER, M.M. 1990. Larval metamorphosis of Microciona prolifera: evidence against the reversal of layers. Pp. 182-187. In Rútzler, K. (ed,) New Perspectives in Sponge Biology. (Smithsonian Institution Press: Washington D.C.). OKADA, Y. 1928. On the development of a hexactinellid sponge, Farrea sollasii. Journal of the Faculty of Science Imperial University of Tokyo 2(1): 1-27. OMNES, P. 1991. Micromorphologie des galeries de démosponges perforantes. Unpublished degree in Oceanography (Université d'Aix Marseille Il: Marseille). PANG, R.K. 1973. The systematics of some Jamaican excavating sponges (Porifera). Postilla 161: 1-75. PLEYDELL, S.M. & JONES, B. 1988. Boring of various faunal elements in the Oligocene-Miocene bluff formation of Grand Cayman, British West Indies. Journal of Paleontology 62: 348-367. PULITZER-FINALI, G. 1983, A collection of Mediterranean Demospongiae (Porifera) with, in appendix, a list of the Demospongiae hitherto recorded from the Mediterranean Sea. Annali del 641 Museo Civico di Storia Naturale di Genova 84: 445-621. ROSELL, D. 1996. Systematics, biology and ecology of the Mediterranean excavating sponges. Ph.D.Thesis (Universitat de Barcelona: Barcelona). ROSELL, D. & URIZ, M.J. 1997. Phylogenetic relationships within the excavating Hadromerida (Porifera), with a systematic revision. Cladistics 13: 349-366. RUTZLER, K. & RIEGER, G. 1973. Sponge burrowing: fine structure of Cliona lampa penetrating calcareous substrata. Marine Biology 21: 144-162. SIMPSON, T.L. 1984. The cell biology of sponges. (Springer Verlag: New York). SIVARAMAKRISHNAN, V.R, 1951. Studies on early development and regeneration in some Indian marine sponges. Proceedings of the Indian Academy of Sciences 34: 273-310. TOPSENT, E. 1891. Deuxiéme contribution à l'étude des Clionides. Archives de Zoologie Expérimentale et Générale (2) 9: 555-592, 1900. Etude monographique des Spongiaires de France. III.-Monaxonida (Hadromerina). Archives de Zoologie Expérimentale et Générale (3) 8: 1-331. 1903. Sur les larves cuirassées de Thoosa armata. Archives de Zoologie Expérimentale et Générale (4) 1, Notes et Revue 1: 1-3. 1904. Spongiaires des Açores. Résultats des Campagnes Scientifiques du Prince Albert I" de Monaco 25: 1-280. 1920. Caracteres et affinités des Thoosa Hanc. et des Alectona Cart. Considérations sur leurs germes à armure. Bulletin de la Société Zoologique de France 45: 89-97, 1928a. Nouvelle étude sur les Cliothosa, Bulletin de l'Institut Océanographique de Monaco 525: 1-7. 1928b. Spongiaires de l'Atlantique et de la Méditerranée, provenant des croisiéres du prince Albert 1% de Monaco. Résultats des Campagnes Scientifiques du Prince Albert I de Monaco 74: 1-376. 1941. Sur les germes des Alectona et des Thoosa. Bulletin Scientifique de Bourgogne 9: 37-40. 1948. Considérations sur les “plasmodes planctoniques d'Eponges". Bulletin de l'Institut Océanographique de Monaco 923: 1-4. TREGOUBOFF, G. 1939, Sur les larves planctoniques d'éponges. Comptes-Rendus de l'Académie des Sciences de Paris 208: 1245-1246. 1942. Contribution à la connaissance des larves planctoniques d'éponges. Archives de Zoologie Expérimentale et Générale 82: 357-399, 1957. Porifera (Spongiaires). Pp. 259-262. In Trégouboff, G. & Rose, M. (eds) Manuel de planctonologie méditerranéenne. Vol. 1. (CNRS: Paris). 642 TREGOUBOFF, G. & ROSE, M. (eds) 1957. Manuel de planctonologie méditerranéenne. Vol. 1 (CNRS.: Paris). VACELET, J. 1981, Etude qualitative et quantitative des salissures biologiques de plaques expérimentales immergées en pleine eau. 6. Les Eponges. Téthys 10(2): 165-172. WATANABE, Y. 1978. The development oftwo species of Tetilla (Demospongie). Natural Science Report of the Ochanomizu University 29: 71-106. WATANABE, Y. & MASUDA, Y. 1990. Structure of fiber bundles in the eggs of Tetilla japonica and their possible function in development. Pp. 193-199, In Rützler, K. (ed.) New Perspectives in Sponge Biology. (Smithsonian Institution Press: Washington, D.C.). WIEDENMAYER, F. 1994, Contributions to the know- ledge of post-palaeozoic neretic and archibenthal MEMOIRS OF THE QUEENSLAND MUSEUM sponges (Porifera): The stratigraphic record, ecology, and global distribution of intermediate and higher taxa. Schweizerische Paláontologische Abhandlungen 116: 1-147. WILSON, H.V. 1894. Observations on the gemmule and egg development of marine sponges. Journal of Morphology 9: 277-408. WISCHNITZER, S. 1966. The ultrastructure of the cytoplasm of the developing amphibian egg. Advances in Morphogenesis 5; 131-179. WULFF, J.L. 1985. Dispersal and survival of fragments of coral reef sponges. Proceedings of the Fifth International Coral Reef Congress, Tahiti 2: 119-124, 1991. Asexual fragmentation, genotype success, and population dynamics of erect branching sponges. Journal of Experimental Marine Biology and Ecology 149: 227-247. FRESITWATER SPONGES FROM A NEOTROPICAL SAND DUNE AREA CECILIA VOLKMER-RIBEIRO, MARIA M, F. CORRELA, SERGIO L. A, BRENHA AND MAURICIO A. MENDON(GA Volkmer-Ribeiro, C., Correia, M.M.F., Brenha, S.L.A. & Mendonca, M.A. 1999 06 30: Freshwaler sponges from a Neotropical sand dune area. Memoirs of the Queensland Museum 44; 643-649, Brisbane, ISSN 0079-8835. A survey ofa freshwater sponge community from a sand dune belt area, NE coast of Brazil. is reported for the first time. Corvolieteromeyenia heterosclera was the only sponge living in crystal clear seasonal ponds nestled among the sand dunes. The sponge forms fan-shaped growths around the leaves of Eleocharis sp.(Cyperaceae) in the shallow border of ponds, or massive crusts on sporophytie plants of Equiserum sp.(Equisetaceae) in deeper parts of ponds, Towards the boundary, between sand dunes and savanna (cerrado), ponds are less subject to wind action, thus more stable and allowing the establishment of shrubby vegetation and palm trees. Accumulation of decaying vegetation produces brownish acid waters in which some of the five sponge Species live, all characteristic of the savanna fauna: Metania spinata, Corvomeyenia thumi. Dosilia pydanieli, Radiospongilla amazanensis and Trachaspongilla variabilis. An association between two of the five savanna species with C. heteroselera was observed in some ponds at the verge of the mobile sand dunes. This large ecotone seems to represent the "patch concept’ in the dynamics of its ponds and their sponge communities. Porifera, freshwater ephemeral habitats, sand dunes, savanna, ecotone, survival, dispersal, adaptations. Cecilia Volkmer-Ribeira (email: ebibipampa.iche.br), Museu de Ciencias Naturais, Fundação Zoobatanica do Rio Grande do sul, Cx, Postal 1188, 90.001-970, Porto Alegre - R;, and Graduate Program of Zoology, Pontificia Universidade Católica da Rio Grande da Sul, Cx. Postal 1429, 90.619-900, Porto Alegre — R;, and Fellow Researcher of CNPq , Brazil; Maria Murlúcia Ferreira Correia, Sérgio Luis A. Brenha & Mauricio Arcuijo Mendonça, Departamento de Biologia, Universidade Federal da Maranháo, Pr. Gongalves Dias, 21, 65.020-240, São Luiz - MA, Brazil; 10 February 1999, The eastem part of Maranhào State, Brazil, is a coastal sand dune field containing an amazing array of seasonal freshwater ponds nestled among dunes, stretching along the municipalities of Tutóia, Barreirinhas and Humberto de Campos and penetrating 10-50km into ihe countryside where il borders with the ‘cerrado’ (Brazilian savanna). The National Park of ‘Lengois Maranhenses” was created in 1981 to protect the western part of this area. Under Koeppen's classification this region is classed as AW climate, with the rainy season between December-May, and the annual mean air temperature. between 26,8-27,.2°C.(Padua, 1983). Prior to the present work we knew nothing about the freshwater sponge fauna of this large seasonal lentic system, which appears to be unique in the world and certainly within the Neotropical Region. The study is part of ongoing investigations of the Neotropical freshwater sponge habitats and communities undertaken by the senior author and collaborators (Volkmer-Ribeiro, 1992: Volkmer-Ribeiro € Tavares,1993; Volkmer- Ribeiro et al., 1983, 1998), MATERIALS AND METHODS Our survey of freshwater sponges in the sand dune lentic system was undertaken at the E and W borders of the Lençóis Maranhenses, Maranhão State, Brazil (in proximity to the villages of Tutóia and Santo Amaro), with some sites inside the National Park, where sand dunes border with the savanna (Fig 1). In response to valuable information provided by local residents on current water levels (since dry and wet seasons do not occur exactly at the same period each year), our survey was conducted between 20-30 October 1995. At this time of year the largest ponds (approximately 1500m^) contained only about 30% of their potential water capacity, which was considered an opportune time to sample because it was most likely that sponges would have gemmules. Several variables were measured in some of the ponds in the western area (pH, temperature and salinity). The pH was taken in situ with Merck pH paper. Salinity was also recorded in situ with a salinometer. Samples of sediments were also 644 CERRADO-RESTINGA ==] wameut E ourros E TERRENOS MUNDAVEIS 77 LIMITE DO PARQUE DOS |: LENGOIS MEMOIRS OF THE QUEENSLAND MUSEUM BARRA DAS N PR EQUIGAS ATLÁNTICO BARBA DE PTA, LAZOU, TUTOIA BARRA DO CAJU q AQ FIG. 1. The Lençóis Maranhenses. The villages of Tutóia and of Santo Amaro (arrows) are shown, at the E and W area ofthe Lencóis Maranhenses, respectively, as well as the National Park limits and the contact areas with the savanna, collected from the top layer of pond beds, stored in small glass jars, for ponds in which sponges were not conspicuous. This was undertaken in order to determine, through the presence of spicule sediments, whether sponges had ever been present in these ponds (Volkmer-Ribeiro & Turcq, 1996). These sediment samples are also currently being used by the senior author to study recent and paleo sediments. Entire or cross-sectioned dry specimens of C. heterosclera were glued to a stub and gold-coated for scanning electron microscope (SEM) observations and photographed on a MCN JEOL-5200 microscope equipped with a Pentax SF7 35mm camera. Part of the dry specimens was deposited in the collection of Departamento de Biologia of Universidade Federal do Maranháo. The other part was deposited in the Porifera Collection of Museu de Ciências Naturais of Fundação Zoobotanica do Rio Grande do Sul (MCN/POR) with Catalog numbers listed in Tables | and 2. RESULTS A succession of ponds with different spatial areas (Fig. 2A-C) and water color was observed as one progressed from the mobile dunes towards the savanna boundary. Ponds situated among the TABLE 1. Identification of the sponges sampled at the ponds from the E area of Lencóis Maranhenses, together with their environmental characteristics and substrates. Ponds sampled around o " ; = "m. A : Catalog number . Tutóia Village Type of sample Environment Substrate Water color Sponge species MCN/POR Lagoa do Vidro specimens sand dunes Equisetum sp clear abr anal 3829-3840 B | | heterosclera | Lagoa da Ponta do M" y ; Corvoheteromeyenia 4 Arpoador specimens sent dunes Eleocharis a clear heterosclerd 3830-3839 Radiospongilla amazonensis, Lagoa da Pedra Hume sediment sand dunes brownish clear piney ti 3854-3856 Corvoheteromeyenia u | heterosclera Radiospongilla Lagoa da Ponta do : m. : amazonensis, Mangue bottom sediment oasis brownish Trochospongilla 3859 variabilis Lagoa da Coceira sediment grassy field black Corvomeyenia thumi, 3858 oe _Metania spinata NEOTROPICAL SAND DUNE SPONGES 645 FIG. 2. Study area. A, Lagoa do Vidro, at the E boundary of the Lengóis Maranhenses. B, Lagoa da Ponta do Arpoador, showing the stiff Cyperacean vegetation exposed to the sun as the pond becomes fully dry. C, Lagoa do Pico, at the western border of the Lengóis, with its surrounding dunes covered by bushy vegetation and palm trees (photos C. Volkmer Ribeiro). mobile dunes had crystal clear waters whereas those approaching the savanna boundary had increasingly brownish color and acidic pH (Tables 1-2). Lagoa do Vidro (Fig. 2A), Lagoa da Ponta do Arpoador (Fig. 2B) and Lagoa do Cajueiro are clear water ponds amongst the mobile dunes. The dominant plants in these ponds, as potential sponge substrates, were Eleocharis sp. and Equisetum sp. The former plant is a small but very abundant Cyperaceae which inhabits the shallow marginal area of ponds, and is the first to emerge from the water (Fig. 2B) as ponds begin to dry up. The second species is also relatively abundant in deeper parts of ponds, close to the leeward face of the barchan dunes (Fig. 2A). Lagoa da Pedra Hume, Lagoa da Ponta do Mangue, Laguinho, Lagoa do Pico (Fig. 2C) and Lagoa da Coceira represent a succession of ponds having brownish clear to black water, and distributed from the mobile sand dunes into the almost flat, fixed dunes, covered with grassy or bushy vegetation. Lagoa da Ponta do Mangue is an oasis pond with surrounding patches or islands of palm trees and containing shrubby vegetation. Lagoa da Coceira was reduced to a fetid, tiny muddy pond with black water surrounded by grass. Laguinho was a relatively deep pond, and together with Lagoa da Ponta do Mangue, is a popular fishing site. Taxonomic identification of the sponges sampled in the E and W dune areas are presented in Tables 1 and 2, respectively, together with their 646 MEMOIRS OF THE QUEENSLAND MUSEUM TABLE 2. Identification of the sponges sampled at the ponds from the W area of Lengóis Maranhenses, together with their environmental characteristics and subsirates. Ponds sampled around Catalog number Icuco) amazonensis Santa Amaro Village Type of sample | Environment Substrate Water color Sponge species MCN/POR E IDA suem (pH 6; Specimens & Sand duris Eleocharis s Eler Corvoheteromeyenia 3853 perature 28.5°C at 10am) sediments » SR heterosclera Laguinho Specimens sand dunes stones brownish clear | Dosilia pydanieli 3857 sobinersed Metania spinata, Lagoa do Pico (pH 5.5; sand dunes pranchas ofri Corvomeyenia thumi, Jagoa c d 25.5; Specimens & i a d : Trochospongilla " 0 " em- é ra salinity pt dí “a nr Sediments with bushy veg-| parian bushes brownish variabilis. 3089 perature 32°C at 0.10pm) etation (Crysobalanus Radiospongilla environmental characteristics and substrates. Ponds in which only sediments were reported indicate that living samples of sponges were not found. These tables indicate that Corvoheteromeyenia heterosclera and M. spinata were abundant, as were gemmules, particularly in the two extremes of the ‘pond succession’ (i.e. very clear ones and the brownish colored ones), confirming that the sampling period was well chosen and our prediction that this huge seasonal lentic system contained such a fauna. The pond sponge communities also revealed interesting distribution patterns. The only species found in clear-water sand dune ponds was C. heterosclera (Ezcurra de Drago, 1974), with fan-shaped growths on the leaves of the Cyperaceae Eleocharis sp. (Fig. 3A) at the pond margins (Fig. 2B), or thick irregular growths on the sporophitic plants of Equisetum sp.(Fig. 3B) in deeper parts of the pond close to the leeward dune face (Fig. 2A). Sponges on both substrates were fully developed and full of gemmules. Many specimens were seen dried out, not far from the pond margin, their gemmules continually swept away by the wind. On the other hand specimens still submerged at the pond margins were already half buried by sand blown in from the dune, the same way as were specimens encrusting Equisetum located close to the dune inner wall. Corvoheteromeyenia heterosclera appears to be a species fully adapted to such an environment. SEM study of cross sections of Eleocharis leaves encrusted with this species (Fig. 3C) disclosed a skeletal structure of very slim fibers producing a very open and irregular network enclosing sand grains of variable sizes. The presence of oscular sieves (Fig. 3D) is another device used to prevent oscula being clogged with sand. Gemmoscleres were also seen to take part in skeletal fibers, together with megascleres and two categories of microscleres (Fig. 3E-F), which might be due to the accelerated production of gemmules during this time of the year. Living specimens of C.heterosclera had a light green color, probably due to its association with a microscopic green algae, and this green color could be seen around the pond margins where sponges had been buried and were decaying in the sand. The sponge community towards the savanna boundary included one or more of the following species, but never all five of them: Metania spinata (Carter, 1881), Corvomeyenia thumi (Traxler, 1895), Dosilia pydanielli Volkmer- Ribeiro, 1992, Trochospongilla variabilis Bonetto & Ezcurra de Drago, 1973, and Radiospongilla amazonensis Volkmer-Ribeiro & Maciel, 1983. However, leaving this area and heading towards the clear-water ponds we observed associations of C. heterosclera with R, amazonensis and T. variabilis. TAXONOMIC REMARKS. Ezcurra de Drago (1979) erected Corvoheteromeyenia for Corvomeyenia australis Bonetto & Ezcurra, 1966, and C. heterosclera Ezcurra de Drago,1974. The holotype of the first species comes from Laguna Setübal, next to the town of Santa Fé, Argentina, whereas the holotype of the second species comes from NE Brazil, with no precise locality data. The distinction between the two species was based particularly on their respective gemmoscleres, which are composed of two categories differing in size and shape in the first species, and one category in the second species. Specimens from the Lençóis Maranhenses agree with the second species in spiculation and geographical origin. Corvoheteromeyenia heterosclera has also been collected from other sand dune or paleo sand dune areas in S Brazil (Volkmer-Ribeiro, unpublished data), and its spicular remains may be useful indicators of such environments. NEOTROPICAL SAND DUNE SPONGES 647 385301 500mm 3853093 X: ieddm 3033h4 FIG. 3. Freshwater sponges. A, Fan-shaped specimens of Corvoheteromeyenia heterosclera full of gemmules growing on Eleocharis sp. B, Specimens of Corvoheteromeyenia heterosclera growing on Equisetum sp. C, Skeleton of Corvoheteromeyenia heterosclera with an open network where sand grains are enclosed side-by-side with the gemmules. D, Cross section ofa leaf of Eleocharis sp. (arrow, upper left) encrusted by the skeleton of Corvoheteromeyenia heterosclera, showing an irregular distribution of the thin skeletal fibers and (arrow, upper right) the presence of an oscular sieve. E, Skeleton of Corvoheteromeyenia heterosclera with two gemmules in the process of completion at the bottom as well as abundant free gemmoscleres (arrow). F, Skeletal fiber of Corvoheteromeyenia heterosclera (see arrow in Fig. 3E) showing the reduced amount of spongin and the presence of the two categories of microscleres (O, microspined oxea; M, microanfidiscs) and several large anfidisc gemmoscleres (A) around the megascleres (MG). 648 MEMOIRS OF THE QUEENSLAND MUSEUM FIG. 4. Schematic drawing of the wind-driven dynamics that seasonal ponds around the moving dunes of Lencóis Maranhenses are subjected to, together with the vegetation and sponges they contain. E= Eleocharis, Eq= Equisetum. DISCUSSION Recent literature dealing with the study of Neotropical freshwater sponge communities has shown that the five species recorded in the present study, found in the ponds close to the savanna boundary, generally thrive in seasonal ponds N to S throughout the Brazilian savanna (Volkmer-Ribeiro, 1992; Volkmer-Ribeiro et al., 1998). Thus, two characteristic faunal assemblages are present: one in the savanna ponds with five species, and one which is mono- specific and lives in the sand dune ponds. Observations made here on the characteristics of C. heterosclera match well with our present understanding of the processes associated with constant disturbances imposed by dune movement on populations. The living barchan dunes, which move in the direction ofthe wind in coastal areas, is a well known geomorphological phenomenon (Termier & Termier, 1963). The Lengois Maranhenses is, however, unique in its situation being in a tropical area with a marked rainy season, resulting in temporary accumulation of water in ponds on the leeward sides of moving dunes, with the ponds subsequently displaced next season. The wind, which was seen to blow the sand into the ponds, was also seen to blow ahead the gemmules from dry, exposed sponges growing on Eleocharis at the perimeter of ponds. Moreover, the erect position of plant leaves together with the fan- shaped morphology of the sponges appears to facilitate gemmule dispersal via wind action. These are the pioneers of new sponge populations, opportunistically waiting in place where the new border of the pond with Eleocharis will be situated next season. The same process applies to sponges encrusting on Equisetum, which will be passed over by the moving dune when again they will be set against the inner side of a barchan dune, protected against drought and ready to again form the Equisetum/Corvoheteromeyenia association (Fig. 4). The region extending between the areas occupied by these two faunas thus biologically and physically defines a savanna/sand dune ecotone. This particular ecotone best fits the dynamic ‘patch concept’ (White & Picket, 1985), driven by constant disturbance through dune movement. The Lençóis Maranhenses is an outstanding example ofthe patch concept with its pond system, spatial and temporal relationships to dune movement, and dynamic sponge fauna. It also illustrates the remarkable ability of freshwater sponges to manage constant disturbance via asexual reproduction when gemmules produced by these sponges are especially selected mechanisms to withstand drought and to passively disperse via the wind, enabling all six species to opportunistically colonise new pond systems as they develop. ACKNOWLEDGEMENTS The authors acknowledge the kind assistance of Dr. Marcio F.Costa Vaz dos Santos of the Biology Dept., Universidade Federal do Maranháo, Sao Luiz, for identification of plant specimens. The authors are indebted to two anonymous reviewers as well as to the editor, for valuable comments on the manuscript. Thanks to Rejane Rosa and M.Sc. Laura Maria G. Tavares (MCN) for, respectively, preparation of Figs | and 11, and operation of the SEM. A grant from the National Council for the Scientific and Technological Development of Brazil (CNPq.)- RHAE Program to the senior author funded her travelling and field expenses.. NEOTROPICAL SAND DUNE SPONGES LITERATURE CITED EZCURRA DE DRAGO, I. 1979. Un nuevo genero sudamericano de esponjas: Corvoheteromeyenia gen. nov. (Porifera: Spongillidae). Neotropica. 25(74): 109- 118. PADUA, M.T.J. 1983. Os parques nacionais e reservas biológicas do Brasil. (Instituto Brasileiro de Desenvolvimento Florestal: Brasília). TERMIER, H. & TERMIER, G. 1963. Erosion and Sedimentation. (D. Van Nostrand: London). VOLKMER-RIBEIRO, C. 1992. The freshwater sponges in some Peat-bog ponds in Brazil. Amazoniana. 121(2): 317-335. VOLKMER-RIBEIRO, C. & De ROSA-BARBOSA, R. 1972. On Acalle recurvata (Bowerbank, 1863) and an associated fauna of other freshwater sponges. Revista Brasileira de Biologia 32(3): 303-317. VOLKMER-RIBEIRO, C., De ROSA-BARBOSA, R. & MELLO, J.F. 1983. The unexpected occurrence of Drulia browni (Bowerbank, 1863) (Porifera:Spongillidae) in an oxbow lake at the 649 extreme South of Brazil. Iheringia, Zoologia 63: 3-10. VOLKMER-RIBEIRO, C., MANSUR, M.C.D., MERA, P.A.S. & ROSS, S.M. 1998. Biological indicators of the quality of water on the island of Maracá, Roraima, Brazil. Pp. 403-414. In Miliken, W. & Ratter, J. (eds) MARACA: The biodiversity and Environment of an Amazonian Rainforest. (John Wiley & Sons Ltd: Chichester). VOLKMER-RIBEIRO, C. & TURCQ, B. 1996. SEM analysis of siliceous spicules of a freshwater sponge indicate paleoenvironmental changes. Acta Microscopica 5(B): 186-187. VOLKMER-RIBEIRO, C. & TAVARES, M.C.M. 1993. Sponges from the flooded sand beaches of two amazonian clear water rivers (Porifera). Iheringia, Zoologia 75: 187-188. WHITE, P.S. & PICKET, S.T.A. 1985. Natural disturbance and Patch Dynamics: An Introduction. Pp. 3-13. In Picket, S.T.A. & White, P.S. (eds) The Ecology of Natural Disturbances and Patch Dynamics. (Academic Press: Orlando, Florida). 650 MEMOIRS OF THE QUEENSLAND MUSEUM BIOGEOGRAPHY AND TAXONOMY OF THE REEF CAVE DWELLING CORALLINE DEMOSPONGE ASTROSCLERA WILLEYANA THROUGHOUT THE INDO-PACIFIC. Memoirs of the Queensland Museum 44: 650. 1999:- Astrosclera willeyana Lister, 1900, is a pyriform-half spherical, predominantly bright orange coralline demosponge. The habitat of Astrosclera is generally restricted to cryptic and light reduced environments of the Indo-Pacific, found mainly in reef caves, but sometimes also in the dim-light areas of cave entrances and overhangs. Its spicule skeleton consists of megascleres only, whereas microscleres are absent. The basic spicule type is a sub-verticillate to verticillate acanthostyles, of the Age/as type, with a mean length of 80um. The spicule morphology and size is highly variable, depending on the geographic origins of specimens. Variability in spicule morphology of Astrosclera from different geographic localities was previously reported by several authors (Vacelet, 1967, 1977, 1981; Ayling ,1982; Wórheide etal., 1997), and Vacelet (1981) discussed the idea that there might be more than one species of Astrosclera. Empirical testing of the question - whether variation in spicule morphology represents geographic variation or separate species - was undertaken in this study, examining the spicule morphology of specimens from 26 geographically distinct populations. Corroborative evidence from a restriction fragment length analysis of the ribosomal DNA was also undertaken for twenty specimens from five geographically distinct populations of Astrosclera. Analysis of spicule morphology showed that variation was not random but specifically linked to geographical origin of the specimen. Six groups were recognized with similar spicule morphology (group with similar spicule morphology: GSSM’s), based on spicule length, spination, proximal thickening, and abundance. These GSSM's comprise populations from adjacent localities. These spicule data, therefore, support the concept that there may be more than one species of Astrosclera, whereas analysis of ribosomal DNA did not lend support to this hypothesis. The RFLP-method of rDNA-analysis was sensitive enough to detect species-level differences in sponges, as shown by comparative studies on other demosponges, but until further divergent characters are found, the different GSSM’s are still regarded as one species. The GSSM's seem to represent geographic subspecies, whose genetic differences, expressed by different (non-random) spicule morphology, were not detected by rDNA-analysis. It is supposed that Astrosclera is currently in the process of species separation. Each GSSM (or subspecies) is likely to have its own history with respect to radiation, isolation and evolution, and a model of the biogeographic and phylogenetic relationships of the GSSM's are presented. See Wórheide (1998) for details. O Porifera, biogeography, taxonomy, Astrosclera, ribosomal DNA, spicule morphology, phylogeny. Literature cited. AYLING, A. 1982. Redescription of Astrosclera willeyana Lister 1900 (Ceratoporellida, Demospongiae), a new record from the Great Barrier Reef. Memoirs ofthe Natural Museum of Victoria 43: 99-103. VACELET, J. 1967. Quelques éponges pharétronides et "silico-calcaires" des grottes sous-marines obscures. Recueil de Travaux de Station Marine d'Endoume 58(42): 121-132. VACELET. J, 1977. Eponges pharétronides actuelles et sclérosponges de Polynésie Francaise, de Madagascar et de la Réunion. Bulletin du Muséum National d'Histoire Naturelle, Paris 3(444) (Zoologie, 307): 345-368. VACELET, J. 1981. Eponges hypercalcifeé (“Pharétronides”, *Sclérosponges") des cavités des récifs cralliens de Nouvelle-Calédonie. Bulletin du Muséum National d'Histoire . Naturelle, Paris (4,A) 2(3): 313-351. WORHEIDE, G., GAUTRET, P., REITNER, J., BOHM, F., JOACHIMSKI, M.M., THIEL, V., MICHAELIS, W. & MASSAULT, M. 1997. Basal skeletal formation, role and preservation of intracrystalline organic matrices, and isotopic record in the coralline sponge Astrosclera willeyana Lister 1900. Boletin Real Sociedad de Espanola de Historia Natural. Seccion Geologica ... 91(1-4): 355-374. WORHEIDE, G. 1998. The reef cave dwelling ultraconservative coralline demosponge Astrosclera willeyana Lister from the Indo-Pacific - Micromorphology, Ultrastructure, Biocalcification, Taxonomy, Biogeography, Phylogeny, Isotope Record. Facies 38: 1-88. Gert Worheide* (email: gertw@ibm.net) & Joachim Reitner, Institut und Museum für Geologie und Paldontologie (IMGP), University of Goettingen, Goldschmidtstr. 3, D-37077 Goettingen, Germany. *Present address: Marine Biology Laboratory, Queensland Museum, PO. Box 3300, South Brisbane, Old, 4101, Australia; 1 June 1998 EXTANT AND FOSSIL SPONGIOFAUNA FROM THE UNDERWATER ACADEMICIAN RIDGE OF LAKE BAIKAL (SE SIBTRIA) E. WEINBERG, C. ECKERT, D. MEHL, J. MUELLER, Y. MASUDA AND S. EFREMOVA Weinberg, E., Eckert, C.. Mehl, D., Mueller, J., Masuda, Y. € Efremova, S. 1999 06 30: Ex- tant and fossil spongiofauna [rom the underwater Akademician Ridge of Lake Baikal (SE Sibiria). Memoirs of the Queensland Museum 44: 651-657, Brisbane. ISSN 0079-8835. Sediments of Lake Baikal contain unique and highly diverse siliceous assemblages. Spange spicules and diatom frustules are both well preserved and abundant, A bottom sediment core STX3GC (48m long) was taken from the Academician Ridge using a yravily corer and studied at intervals of Oem, At discrete intervals sediments consisted Of 10-30% by weight of biogenic silica. Major contributors were diatams, which are a good tool for stratigraphic assignments. Sponges were also widely distributed in space and lime, and may also be valuable stratigraphic markers in Lake Baikal, if addition to their important ecological role, Analysis of core STX3GC revealed 4 genera of Lubomirskidae consisting of 9 species, together with scattered mega- and microseleres of Spongillu sp, and Eplivdalía sp. (Spongillidae). Spicule abundance and diversity were highest during periods of warm climate, whereas during cold intervals Spongilla sp. and Swarrsehewskia papyracea spicules were missing. and ubundanee and diversity of other spicules also significantly decreased These observed morphometric changes may be applied in Tracing palacoccological and climatic changes in Lake Baikal during the Holocene and Pleistocene, or even earlier, and spicules of new species may hold information on the evolution of Lubomirskiidac and their probable sponyillid roots. J Porifera, Lubomirskiidae. Spongillidae. freshwaier sponges, spicules in sediments, palaeolimnology, Luke Baikal, Pleistocene, Holucene, Elena Weinberg (email: info(alin irk ru), Limnolagical Institute Irkutsk. Sihivian Branch af RAS, Ulan-Batorkava str. 3, 664033 Irkutsk. Russia: Johannes Mueller & Corsten Fckerl, Alfred Wegener Institute Jor Pular and Marine Research, RD Potsdam, Telegrafenberg 143, 14473 Potsdam, Germany; Dorte Mehl-Janussen, Freie Universita: Berlin, Institute of Paleontology, Malteserste 74-100, Hans D 12249 Berlin, Germany: Yoshiki Masuda. Kawusaki Medical School, Department af Biology. Kurashiki City, Okavama 701-01, Japan; Solia Efremova, Biological Institute, St Petersburg State Universin, Oranienbuumskoyc seh. 2, 108904 St Petersburg, Russia; 21 January 1999, Compared to most existing fresh water lakes Lake Baikal has many unique leatures, one of which is its sedimentary fossil fauna. Spicules of siliceous sponges and diatom frustules are well preserved and abundant in sediments, In the Academician Ridge region the biogenic silica content of sediments reaches 10-30% by weight, the main part consisting of diatoms which are traditionally used in palaeostratigraphic analyses. According to Martinson (1936b) sponge spicules in deeper paris of Lake Baikal are of allochthon- eous origin because, in his opinion, freshwater sponges live on hard substrata in shallow waters, However, the sedimentary bed of the Academician Ridge represents a silt layer 1,000-1,500m thick, and the water column height above it is 300m. The present work was undertaken to tind how sponge spicules have come into the Academician Ridge sediments — if they are autochtonous elements of sediments or have come in trom ather parts of the lake. MATERIAL AND METHODS SAMPLE COLLECTION. Living sponges were sampled between 1996-1997 from the Academician Ridge by dragging the lake bed using a deep water trawler, in a transect extending from the north part of Olchon Island to the south part of Great Ushkany Island, 38 specimens of sponges were collected using this method. with an additional 50 specimens collected by divers trom the litoral zone of Great Ushkany 1. Sediment samples were collected by box carers and a gravity corer on board the RV Vereshagin, a scientific research vessel from the Limnological Institute, Irkutsk. Altogether 9 surface samples and a gravity core with a length of 480cm were sampled (see Fig. 3). Sampling was undertaken at 652 side view top view FIG. 1. Schematic illustrations of Petri dish with cover glases for sample sedimentation. every 10cm interval of the gravity core, i.e. 48 samples for the entire core. SAMPLE PREPARATION. 5g of each sediment sample was freeze dried for 12hrs, oxidised and disaggregated on a sample shaker using 100m1 3% hydrogen peroxide with a drop of conc. NH¿OH. The solution was then wet sieved through a 32mm mesh. The fraction <32mm was taken for further sedimentological analysis, whereas the fraction >32mm was carefully washed out from the sieve into PE bottles using 50ml distilled HO, and prepared as follows. Depending on numbers of samples several Petri dishes (4.9cm diameter) were put on a horizontal base (Fig. 1). Two cover glasses were placed into each dish and the dishes filled to approx. 2/3 with a gelatine solution (0.06g of 700ml distilled water). After shaking the PE bottles containing the >32mm fraction a 1ml aliquot was pipetted onto the gelatine solution, slowly stirring with a pipette in order that the suspension would ideally be distributed evenly within the Petri dish. Petri dishes were left undisturbed for at least 2hrs to ensure even settlement of the suspended material. Subsequently, prepared filter strips were inserted into the Petri dishes with one end touching the bottom and the other end touching the base outside. Thus, due to cohesive attraction, water from the basin flowed to the base and was picked up with a pipette or absorbing paper. For this purpose, a sedimentation ‘stairway’ was constructed at the Alfred Wegener Institute (Zielinsky, 1993), making this preparation step much easier. In the Petri dishes only the dry cover glasses remained, with an equally dispersed >32mm fraction. These were put onto glass slides using canada balsam as mounting medium. Nopicules/g 2m (0.3925 (n¡+n>) Vio NV P M,) (where N= amount of spicules per gram of freeze dried sediment; n;+n.=amount of spicules counted in the preparation; V};29=volume of distilled water MEMOIRS OF THE QUEENSLAND MUSEUM added to the sample; V,=volume of the sample aliquote; l-length of the cover glass; M;-weight of the freeze dried sample; d=diameter of the Petri dish). The coefficient of quantity, in this case 0.3925, has to be newly determined depending on the diameter of each Petri dish using the following formula: N=(0.5(n/+n2) Vio (d/2Y p d) (V4 M,) RESULTS LIVING SAMPLES. Over the Academician Ridge transect we sampled 38 living sponges from two species, of which 31 specimens belonged to a new species Baikalospongia sp. nov.] (identified by S. Efremova, in prep.). This species is pillow shaped with a strongly porous surface and very stable consistency. Lateral surfaces are covered by a cornea-like layer protecting the sponge body from silt penetration. Size varies from 1-5cm. Colour of the upper surface is green- blue or brownish. Baikalospongia sp. nov.l has spicules of a characteristic shape: both ends of the spicule have unique crown of thorns. The sponges grow on large, compressed, iron-manganese concretions which are embedded abundantly in the silts of the Academician Ridge. The spicules of this species were found in all sediment samples here. Two sponges externally similar to Baikalospongia sp. nov.l, but with different spicule forms — short strongyles 170-215um long, 15-22um wide, with tiny spines diverging on all sides — may be a subspecies of Baikalospongia bacilifera, based on similarities in their skeletal forms. Also interesting is the find of 5 specimens of soft, friable, brown-blue sponges, 170-215um long, 15-22um wide, with round oscules and unattached to any substratum. In skeletal structure and spicule morphology this sponge is similar to a variety of B. intermedia (Dyb.) described by Rezvoy (1936) as morpha profundalis. Rezvoy noted that this variety differs fromB. intermedia s.s. in its spicule size (330-470um long, 20-25um wide), spicule morphology (consisting of slightly curved strongyles covered with small spines, with a dense accumulation of small pointed spines at the rounded (basal) end), and possession of weak skeletal structure (with an irregular net-like reticulation of spicules bonded at their ends by collagen). This species is previously known only SPONGE FAUNA OF LAKE BAIKAL Lubomirskia Baikalospongia Lubomirskia Baiklospongia Swarischew- abietina intermedia n.sp.1 2000 0 4000 0 9000 100 200 300 Depth of sediment core (cm) 400 0 Spongilla sp. Ephydatia sp. n.sp.l skia sp. 700 0 200 0 2000 0 600 Spicules per gm of sediment FIG. 2. Distribution of sponge spicules in sediments collected from the gravity core STX3GC. from a single specimen collected in 1932 from 890m depth. The taxonomy used here is based on the spicule classification system of E. Weinberg which is in some details controversal among the authors (e.g. the differentiation of isolated spicules between Ephydatia and Spongilla). However, we still consider the spicular signals sufficiently distinct to establish the specific ecological response among the fresh water sponges, Lubomirskiidae and Spongillidae. SUBRECENT AND FOSSIL SAMPLE MATERIAL. Analysis of slides preparations of sediment sample revealed 7 different species of Lubomirskiidae. Analysis and description of the latter are in progress. The most prevalent and widely distributed spicules found in surface sediment samples came from Baikalospongia intermedia, followed by Lubomirskia abietina, Baikalospongia sp. nov.l, Baikalospongia sp. nov.2, Baikalospongia sp. nov.3 and Lubomirskia gen. et sp. nov., whereas spicules of Swartschewskia sp. were relatively rare. The concentration of the spicules in surface sediment samples was inversely correlated with water depth, showing a general decrease in abundance of spicules in sediments as water depth increased (Fig. 5). TABLE 1. Distribution of sponge species in the central part of Lake Baikal. Akadem -ician Ridge Surfacial Ushkany Is ground Sponges species + 1 1 i . Baikalispongia bacillifera . B. intermedia . B.interm. m. profund. 2 3 4. B.bacillifera, ssp. 1 - 5. 6 7 8 + + |+ i Baikalispongia nsp. 1 . Baikalispongia nsp. 2 . Baikalispongia nsp 3 . Baikalispongia nsp. 4 9. 10. Lubomirskia nsp. 1 Lubomirskia abietina ++ le le + BÀ Í [+ + [+ + + [+ + [+ [+ [+ i i 11. Swartschewskia sp. 12. Lubomirskiidae n. G. n. sp. ds 13. Spongilla sp. 14. Ephydatia * È - 15. New spicule 1 - r 4 + + + ]4 16. New spicule 2 A " n 654 Watercontent Sponge spicules Diatoms in% total total 30 40 50 60 70 80 0 100 200 300 400 0 100 200 0 ü I 4 MET 20-22 4 ES E A” n Li 100 pa = 27 £ 200 — = — a a 300 400 MEMOIRS OF THE QUEENSLAND MUSEUM Sulfur in% Carbon in % Nitrogen in 9o Magnetic susceptibility l 2 30 01 02 030 05 10 1060 FIG. 3, Sediment parameters of the gravity core STX3GC. Analysis of the gravity core STX3GC revealed four genera of Lubomirskiidae in nine species, in addition to mega- and microscleres of Spongillidae Spongilla sp. and only megascleres of Ephydatia sp. (Fig.3, Table 1). We also found spicules that could not be attributed to any known living species. Some of these were extraordinarily long (360-700um), relatively thin (12-16um) and with small spines spread over the whole rhabd. Others had a smooth surface and a relatively thick rhabd with a length of 225-280um and width of 24-40um, and small spines found only on the ends. The abundance of sponge spicules throughout the core indicates that cyclic environmental changes had taken place between these sedimentary strata as indicated by maximum and minimum spicule concentrations. The first maximum concentrat- ion showed up in Holocene sediments, down to a depth of about 40cm. It is dominated by Baikalospongia intermedia, as for the subrecent surface sediment samples, and Lubomirskia abietina. The second maximum concentration commenced at a sediment depth of 200cm and ended at 300cm. This has to be classified as the Karginsk interstadial, and is indicated by the presence of spicules of Baikalospongia sp. nov. 1. The intervals, which according to diatom stratigraphy (Grachev et al., 1997) can be classified as glacials and stadials, are charact- erised by distinctly low spicule diversity and total spicule concentration, perhaps reflecting low species diversity and abundance during these periods. Spicules of Swartschewskia sp. and Spongilla sp. disappear completely during these intervals. DISCUSSION The presence of both living sponges and their spicules in subrecent sediments on the Academician Ridge suggest that sponge spicules are an autochtonous element of these sediments. This opinion is confirmed by the presence of large numbers of spicules from Baikalospongia sp. nov.1 in the surface samples of ST 14, located near living populations. Sponges have adapted to life on a loose silty substrata in different ways. Firstly, they may live without a fixed anchorage to the bottom, as seen in Baikalospongia intermedia morpha profundalis living on the Academician Ridge. Rezvoi (1936) described similar examples of sponges living on loose substrata without any fixed anchorage. Secondly, sponges may live on SPONGE FAUNA OF LAKE BAIKAL Lake Baikal 10730 107° 0 10730 1085 01 108° ( 108°30' 7^ NORTH BASIN ( OF LAKE BAIK 9/7 109" 0 syao Legend: e | sample stations of surfacial ground station of gravity core * X3GC Scale: 1: 1 400 000 y 108730" 109? 0° FIG. 4. Map of the study locality in Lake Baikal indicating station sites for surface sediments and gravity core samples. ferro-manganese crusts in which they build up a horny layer for protection against sedimentation in the muddy environment. This strategy is seen in Bakalospongia n. sp | and Baikalospongia bacillifera. Thirdly, sponges may grow on other sponges, using their horny layer as a substrate. An example of this commensalism was seen in a specimen of Baikalospongia sp. nov. 1 covered by two smaller individuals of the same species living on its horny external layer, collected by our expedition to the Bolshye Koty area in 1998, at a water depth of 100m. The expedition of the Irkutsk State University found a similar specimen at Academician Ridge, kindly provided to us by Dr. V. Takhteev. In this case it is also possible that the two smaller individuals may be buds of the larger ‘parent’ sponge. Spicules were found in both subrecent and fossil sediment samples, including those of species living today in near shore areas and shallow waters of Lake Baikal. These mainly concern Spongillidae but also some represent- atives of Lubomirskia. The Recent spectra of species in the area of the S part of Bolshye Ushkany I. is generally comparable with the spectra found in the gravity core STX3GC (see Table 1). Assuming that prevailing wind directions have changed from E to NE, these sponge spicules are probably allochthoneous material brought in from the Barguzin Bay and Ushkany Is. Even during periods of more prevalent SW winds the cyclonic centre remains above the Ushkany Is. Thus, sediment material is also transported from near shore areas into central parts of Lake Baikal. Based on these facts, we consider that a part of the spicules found by us from the Academician Ridge are allochtonous whereas the others are autochtonous in origin. In general sponge spicules are not regularly distributed within sediments, showing distinctly different patterns during colder and warmer periods of the Lake. Maximum concentrations of spicules occur in the Holocene and late Pleisto- cene (Figs 3-4), During this peak there is also a maximum concentration of diatoms present in sediments, demonstrating that it was a period of longer lasting warm weather and high bio- productivity. In contrast, sediments laid down during the long lasting Sartan stadial (latest Pleistocene), a relatively cool period, show 16 000: 1000 content of spicula per gram Station 9 Station 8 656 MEMOIRS OF THE QUEENSLAND MUSEUM Legend: E i S Baikal i dis] L 8000 aikalospongia intermedia = Baikalospongia recta zal | 9 Baikalospongia n.sp. 1 2000 w Baikalospongia bacillifera 1000 Lubomirskia abietina Rezinkovia sp. Swartschewskia sp. 0 100 200 300 IR Susion 3 Station 6 400 Station 3 Station 4 FIG. 5. Relative composition of sponge spicules in subrecent sediments in relation to water depth. variable concentrations of diatoms ranging from very low to virtually absent (Fig. 4), whereas sponge spicules are present, and spicule abundance of Baikalospongia sp. nov. | is ata maximum, during this period. It is possible that this deep water species can better survive or adapt to changing environmental conditions due to particular nutrient regimes and possession of symbionts, because its growth is not strongly dependent on the availability of organic nutrient supplies. A change in water chemistry, resulting in dissolution of diatom frustules during colder climates, as hypothesised by some scientists (e.g. Grachev et al., 1997), seem unlikely given that the biogenic silica of sponge spicules and diatoms is identical. If this were so then it would be expected that during a change in pH conditions there would be dissolution of both sponge spicules and diatom frustules. A better explanation proposed by Back & Strecker (1998) is that during colder climates, and largely during glacial activity, high amounts of suspended material were transported into the Lake largely reducing light transmission in surface waters. This could have led to a distinct decrease in the diatom population, whereas its effect on the sponge fauna, if any, would at worst have led to a change in their symbiont relationships with little or no impact on their general living conditions. Not all sponge species were present consis- tently over time. Spicules of Swartschewskia sp. l and Spongilla sp. occur only in the concentration maxima of the Holocene and the Karginsk interstadial. These species are typical representatives of the littoral environment, and are obviously more prone to environmental changes than deeper water species. These species are potentially useful palaeomarkers as indicators of relatively warm periods. The occurrence of unidentified spicules in sediments, so far are unknown to any species, suggest that the Lake Baikal fauna may contain undescribed species of sponges, particularly in the endemic family Lubomirskiidae. Of special interest in this regard is our further investigations of deep drilling cores BDP-96 from the Academician Ridge, which contains nearly 5m.y. of sedimentary records. Thus, even single spicules of new species can provide information about the evolution of Lubomirskidae as well as their probable spongillid ancestry. SPONGE FAUNA OF LAKE BAIKAL ACKNOWLEDGEMENTS We are grateful to Mikhail A. Grachev, Limnological Institute of Irkutsk, whose devoted work on Lake Baikal made these international studies possible. We wish him all the best for a fast recovery. We also thank I.B. Mizandrontsev for help in the mathematical analysis of data. Thanks to John Hooper of Queensland Museum for organising the Sponge Symposium in Brisbane, for editing the proceedings and for his, and two other anonymous referees, critical review of the manuscript. LITERATURE CITED BACK, S. & STRECKER, M.R. 1998. Asymmetric late Pleistocene glaciations in the North Basin of the Baikal Rifi, Russia. Journal of the Geological Society, London 155: 61-69, GRACHEV, M.A., LIKHOSHWAY, E.V., VOROBY- OVA, 8.8, KHLYSTOV, O.M., BEZRUKOVA, E.V., WEINBERG, E.V., GOLDBERG, E.L., GRANINA, L.Z., VOLOGINA, E.G., LAZO, F.L, LEVINA, O.V., LETUNOVA, P.P., OTINOV, P.V, PIROG, V.V., FEDOTOV, A.P., YASKEVICH, S.A., BOBROV, V.A, SUKHORUKOV, F.V., REZCHIKOV, V.I., FEDORIN, M.A. ZOLOTARYOV, K.V. & KRAVCHINSKY, V.A. 1997. Signals of the paleoclimates of upper 657 Pleistocene in the sediments of Lake Baikal. Russian Geology and Geophysics 38(5): 927-956. MARTINSON, G.G. 1936a. Izkopaemaya spongio- fauna tretichnych otloshenii Pribaikalya [The fossil spongiofauna in tertiarnary sediments of Pribaikalian], Doklady Akademii nauk SSSR, Paleontologiya 21 (4): 212-214. (in Russ.) 1936b. Razpredeleniye spikul gubok v skvashine glubokovo bureniya u s. Posolska na Baikale [The distribution of sponge spikula from a deepth drilling core near by the village Posolska in Lake Baikal aera]. Doklady Akademii nauk SSSR, Geologiya 21(4): 261-264, (in Russ.) 1955. Iskopaemaya presnovodnaya fauna i eye znacheniye dlya stratigrafii. Vestnik Akademii nauk SSSR (12): 32-35 (in Russ.). REZVOI, P.D. 1936. Fauna SSSR. Presnovodnye gubki, sem. Spongillidae i Lubomirskiidae [The fauna of USSR. Freshwater sponges, families of Spongillidae and Lubomirskiidae]. P. 104. In Zoologicheskii institut Akademii nauk SSSR, novaya sertya No.3, izd. (Akademii nauk SSSR: Moskva-Leningrad). (in Russ.) ZIELINSKI, U. 1993. Quantitative Bestimmung von Paláoumweltparametern des Antarktischen Oberflachenwassers im Spátquartiár anhand von Transferfunktionen mit Diatomeen [Quantitative estimation of paleoenvironmental parameters of the Antarctic Surface Water in the Late Quarternary using transfer functions with diatoms]. Berichte zur Polarforschung (126): 1-148 (Bremerhaven). (in Germ.). 658 MEMOIRS OF THE QUEENSLAND MUSEUM CLIMATIC CHANGES OF THE LAST 450 YEARS RECORDED IN THE SKELETON OF THE CORALLINE DEMOSPONGE ASTROSCLERA WILLEYANA. Memoirs of the Queensland Museum 44: 658. 1999:- Stable isotope time series of 8'0 and 5'*C were measured in successive growth layers of the largest and oldest Astrosclera ever found (diameter of 25cm, max. age 550yrs) from Ribbon Reef #10 (GBR) (Wórheide et al, 1997; Wórheide, 1998). Astrosclera forms its skeletal aragonite in equilibrium with the ambient seawater, and represents, therefore, a high precision recorder of the isotopic history of the ambient seawater. "°C of surface water dissolved inorganic carbon in the northern Great Barrier Reef has apparently decreased continuously since the mid-16" century. The total decrease is 0.7?6o. The major decline of 0.5%o occurred during the industrial period of the 19th and 20th century, likely to be due to the increased release of CO; by deforestation and burning of fossil fuel during the period of industrialization after 1850 (increased input of lighter carbon isotopes). The oxygen isotope history shows a slightly colder (and/or dryer) phase before 1850, which correlates with the *Little Ice Age'. A considerable shift to lighter values occurred during the 20th century (warming of SST). This may be due to an anthropogenic greenhouse effect. Most of the major climatic changes caused by ENSO/EI Niño events, as reported by Quinn et al. (1987), as well as by large volcano eruptions (see LaMarche & Hirschbroek, 1984) in the last four and a half centuries seem to be recorded in the oxygen isotope record of Astrosclera. Further, more detailed isotope analyses on replicate samples are needed to corroborate present preliminary data. (J Porifera, Astrosclera, growth layers, isotopes 80 and 6"C, seawater. Literature cited. LAMARCHE, V.C. & HIRSCHBOECK, K.K. 1984. Frost rings in trees as records of major volcanic eruptions. Nature 307: 121-126. QUINN, W.H., NEAL, V.T. & ANTUNEZ DE MAYOLO, S.E. 1987. El Niño occurrences over the past four and a half centuries. Journal of .. Geophysical Research 92(C13): 14,449-14,461. WORHEIDE, G. 1998. The reef cave dwelling ultraconservative coralline demosponge Astrosclera willeyana Lister from the Indo-Pacific - Micromorphology, Ultrastructure, Biocalcification, Taxonomy, Biogeography, . Phylogeny, Isotope Record. Facies 38: 1-88. WORHEIDE, G., GAUTRET, P., REITNER, J., BOHM, F., JOACHIMSKI, M.M., THIEL, V., MICHAELIS, W. & MASSAULT, M. 1997. Basal skeletal formation, role and preservation of intracrystalline organic matrices, and isotopic record in the coralline sponge Astrosclera willeyana Lister, 1900. Boletin de la Real Sociedad Espanola de Historia Natural, Seccion Geologica 91: 355-374. Gert Worheide* (email: gertw@ibm.net) & Joachim Reitner, Institut und Museum für Geologie und Paläontologie (IMGP), University of Goettingen, Goldschmidtstr. 3, D-37077 Goettingen, Germany. *Present address: Marine Biology Laboratory, Queensland Museum, P.O.Box 3300, South Brisbane, Qld, 4101, Australia; 1 June 1998. SEROTONIN IN PORIFERA? EVIDENCE FROM DEVELOPING TEDANIA IGNIS, THE CARIBBEAN FIRE SPONGE (DEMOSPONGIAE) SIMON WEYRER, KLAUS RUTZLER AND REINHARD RIEGER Weyrer, S., Rützler, K. & Rieger, R. 1999 06 30; Serotonin in Porifera ? Evidence from developing Tedania ignis, the Caribbean fire sponge (Demospongiae). Memoirs of the Queensland Museum 44: 659-665. Brisbane. ISSN 0079-8835. Histochemical study of larvae and freshly settled juveniles of the Caribbean fire sponge Tedania ignis (Tedaniidae, Poecilosclerida) reveals evidence of serotonin-like immuno-reactivity, a possible indication for the presence of precursors of nerve cells in this species. Already in the earllest stages of its life, 7. ignis is made up of two discernable cell types: monociliated cells arranged in quasi-epithelial fashion and covering the larva and the developing settled organism. and mesohylal cells (archeacytes). In the adult sponge, several mesohylal cell types can be distinguished which form a complex connective tissue. Serolonin-like immuno-reactivity demonstrated by us occurs only in two cell types: in some archeocytes of the parenchvmella larvae, and in similar archeocytes and ina second, bipolar cell type of the settled, juvenile sponge. The discovery of a neuroactive substance in cells of developing sponges before and alter metamorphosis provides new insights into the origin and evolution of nerve and muscle cells in the Eumetazoa. O Porfferu, hisrachemistry, seratonin, Tedania ignis, larval development, evolution. Simon Weyrer & Reinhard Rieger, Institute af Zaolagy and Limnology, University of Innsbruck. Technikerstrasse 25, A-6020. Innshruck, Austria; Klaus Rützler (email: ruetzler(anmnh.si.edu), Department of Invertebrate Zoology, National Museum of Natural History, Smithsonian Institution, Washington D.C. 20560, USA; 18 January 1999. Although sponges, one of the oldest metazoan groups, possess the greatest diversity of biologically active compounds of any marine phylum, the neurotransmitter serotonin (5-hydroxytryptamine, 5-HT) has been reported only once, in myocyte-like cells of Sycon ciliatum (Sycettidae, Calcarea) (Lenz, 1966). Serotonm appeared early in the evolution of eucaryotes. For example, it is used in chemical signal chains in Protista where, in a species of the ciliate Blepharisma. a serotonin-like substance is known to function as a mating pheromone (Haldane, 1954; Miyake, 1984). It has also been shown that a number of lower organisms use serotonin as an internal messenger in their neuro- transmitter-receptor systems (Carr et al., 1989) and that some of these characteristics of molecular structure that arose in unicellular organisms may have been inherited and modified by metazoans (Mackie, 1990; Van Houten, 1990), It seems obvious that nerve cells developed gradually over a long period of time but the sequences of changes that must have occurred are difficult to establish. Being a primitive outgroup of the Eumetazoa, Porifera do not have neurons or myocytes that are present in organisms of higher levels of organisation. A common phylogenetic hypothesis such as the Planula or Phagocytella hypothesis (see Hyman, 1951: Ivanov, 1988; Rieger et al, 1991; Ax, 1995) encouraged the authors to search for precursors of nerve and muscle cells in sponge larvae in early developmental stages rather than in adults, a neglected area of research so far (Harrison & De Vos, 1991; Woollacott & Pinto, 1995). Such precursors of nerve cells and myocytes in sponges could represent the first stage in the evolution of integrative systems (e.g. Pavans de Ceccatty. 1974a, 1989; Mackie, 1990), MATERIALS AND METHODS Larvae of Tedania ignis (Durchassaing & Michelotti) ( Tedaniidae, Poecilosclerida, Demo- spongiae) were sampled in the laboratory seawater system of the Smithsonian Coral Reef Field Station at Carrie Bow Cay, Belize, in March 1994 and November 1995. Larval release was induced in adult specimens collected in the nearby mangrove of Twin Cays (Rützler & Feller, 1996) and maintained in aerated seawater by exposing them to natural sunlight following a 12—24hr period of dark adaptation ( Woollacott, 1993). The larvae were kept in seawater-rinsed glass dishes (l0cm diameter) and fixed immediately after release and 80-100hrs after attachment to the substrate. To provide a 660 substrate suitable for fixation and removal for subsequent processing, the bottom of these dishes was coated with polymerised epoxy resin (Spurr). Specimens were fixed in 4% paraform- aldehyde (PFA) in phosphate-buffered saline (PBS; 0.1M, pH 7.4) for 6-8hrs, rinsed in PBS, and treated with Triton X-100 (0.2%, 1hr) to permeabilise membranes. After labelling with the primary antibody (rabbit anti-serotonin, IMMUNOTECH 0601; 2.5%) overnight at 4°C, fluorescence-labelling was done for Ihr with a tetrarhodamine-isothiocyanate-(TRITC )- conjugated secondary antibody (swine anti-rabbit; DAKO, 1%) for Lhr. Specimens were then rinsed in PBS, whole-mounted (Gelmount) on slides, and examined under a REICHERT Polyvar epifluorescence microscope. Incubation in bovine-serum albumin-Triton (BSA-T) without primary antibody was used as the control for nonspecific binding of the secondary antibody. Three larvae and three settled juvenile sponges were sectioned (epoxy-embedded, lum thick, stained with toluidin blue) and investigated in detail. The immuno-staining of both larvae and freshly settled sponges was carried out by the labelled streptavidin-biotin method (LSAB kit; DAKO). Histochemical staining of peroxidase with amino-ethyl-carbazole (AEC, substrate buffer) was used to enhance visibility of the labelled cells. RESULTS Tedania ignis has a parenchymella larva composed of two types of cells (Bergquist et al., 1979). Peripherally, flagellated epithelium-like cells cover the organism. This “epithelium” is monociliated and 10-25um high. The free- swimming larva exhibits coordinated ciliary action. A distinct basal lamina and typical eumetazoan apical junctional complexes are apparently lacking (but see below). Interior, appar- ently motile mesenchymal cells (mesohyl cells) are arranged beneath the epithelium-like sheath (Woollacott, 1990, 1993; Amano & Hori, 1994) (Fig. 1A). The live larvae are ovoid and have a size of 700-900um long, 500-600um wide, but the necessary Triton X-100 treatment weakens the cell membranes and larvae usually shrink and collapse (Fig. 1A, B). In the juvenile, settled sponge too— as in the adult-the exopinacoderm which covers the ectosome acts as the protective layer, Inside the sponge, choanocyte chambers connected by canals lined with endopinacocytes MEMOIRS OF THE QUEENSLAND MUSEUM lie embedded among mesohyl cells (Fig. 2A). Using a whole-mount fluorescence technique, we found serotonin-like immuno-reactivity in special mesohyl cells of both larval and juvenile T. ignis. Spherical serotonergic cells appear to be randomly distributed and occur alone or in clusters (Figs 1B, 2B). In one larva, for example, 6 clusters of such serotonin positive cells were found, each composed of 2-4 single spherical cells with a diameter of 4-6m. In some clusters as well as in several single cells the nuclei are clearly visible and appear as non-fluorescent regions (Fig. 1B). In the juvenile, settled sponge, a few bipolar cells were found in addition to the spherical cells that superficially resemble bipolar neurons or “myocyte-like” cells (actinocytes) reported by Bagby (1966) (Fig. 2B, C). These bipolar cells have a maximum length of 20-50um. Both types of serotonin-positive cells (spherical and bipolar) appear to be located in the mesohyl as spicules can be seen on top ofthe labelled cells (Fig. 2C). No information is yet available on whether interactions between these morphological types of serotonin-positive cells occur, nor do we know whether the bipolar cells differentiate from the spherical type. If these serotonergic cells are part of an integrative system, both cellular commun- ication at a distance (spherical cells) or cell-cell contacts (bipolar cells) could be expected. DISCUSSION Our study is the first to report serotonergic cells in Demospongiae, a spherical type in both larva and post-metamorphosed sponge, and a second bipolar cell type that is exclusive to the post- larval developing organism . Up to now, serotonin was only demonstrated histo- chemically in myocyte-like cells of Calcarea (see below). Among the most primitive Eumetazoa, serotonin is well known to act as a neuro- transmitter (e.g. in Anthozoa, Umbriaco et al., 1990). Actually, 5-HT has a wide range of functions in invertebrates, such as control of regeneration processes in Planaria (Kimmel & Carlyon, 1990) and of beat of cilia in echinoderm embryo (Mogami et al., 1992), and as inhibitors and activators of muscle of molluscs (Welsh, 1953; Twarog, 1988). As in Porifera, members of the phylum Placozoa do not differentiate nerve or muscle cells and are therefore counted among the most primitive eumetazoans (Grell, 1974; Ax, 1989; Grell & Ruthmann, 1991). Schuchert (1993) demonstrated in Trichoplax adherens SEROTONIN IN PORIFERA ? 661 ae FIG. 1. Tedania ignis, histology of larva. A, Longitudinal section of an entire larva stained by toluidin blue showing epithelial-like cell layer (e) and dark-staining cells (archaeocytes, arrow) in clusters near the posterior pole (p) (scale bar=100um). B, Serotonin-positive cells (s) are randomly distributed in the larval body; at least 6 clusters of 2-4 labelled cells are evident (one marked, asterisk). The nucleus in some of the cells is visible as a non-fluorescent region (arrow). As control for non specific binding of the secondary antibody specimen were incubated in BSA-T without primary antibodies. (Scale bar=100um.) C, Nomarsky contrast view of the same specimen as in Fig. 1B. The collapsed and shrunk appearance is due to a necessary Triton X-100 treatment that weakens the cell membranes. (Scale bar=100mup..) (Placozoa) bottle-shaped cells (2.74um) that stain specifically with the neuropeptide RF- amide. The author speculated about a possible sensory function of the bottle-shaped cells because the RF-amide positive cells were localised at the margin of the disc-like body of T. adhaerens and the neuropeptide RF-amide is regarded as functionally conservative in lower invertebrates. Much effort has been made toward identifying attachment complexes between adjacent cells of the pinacoderm in adult sponges, as this layer controls the sponge’s internal milieu which differs from the surrounding environment (Ledger, 1975; for review see Harrison & De Vos, 1991). This altered chemical composition in the tissue fluid of the sponge is regarded as a basic precondition for the evolution of nervous systems. Bandshaped, (epithelial-type) apical cell junctional complexes seem to be present in adult Porifera (e.g. apposed membrane junctions in Bagby, 1970; simple junctions in Ledger, 1975; parallel membrane junctions in Green & Bergquist, 1982; fig. 6 in Garonne & Lethias, 1990) in ultrastructural investigations of Tedania ignis, comparable structures seem to be evident (authors, unpublished). However, unequivocal clarification of the organisation of apical junctional complexes is still lacking in the Porifera. It has been stated repeatedly that perm- anent junctional complexes in epithelially organised cells, if present, are different in 662 MEMOIRS OF THE QUEENSLAND MUSEUM FIG. 2. Tedania ignis, histology of freshly metamorphosed juvenile sponge. A, Overview of a cross section stained by toluidin blue. The cells have begun to differentiate into pinacocytes (p) and multiple subtypes of mesohyl cells (m). A water canal in process of forming is indicated (arrow). (Scale bar=150um.) B, Detail of the ectosome. Several spherical (arrow) and one bipolar (asterisk) serotonin-positive cells are visible. To enhance visibility of the labelled cells a histochemical peroxidase staining was used. (Scale bar=1 51m.) C, Bipolar serotonergic cell (same specimen as in Fig 2A). Spicules (arrow) can be seen above the labelled cell. (Scale bar=4um.) construction from those of Metazoa (e.g. Green ,1978; Green & Bergquist, 1979). One reason for this difference is seen in the high motility and frequent rearrangement of sponge cells (e.g. Miiller, 1982). However, it seems likely that development of permanent communicating structures goes hand in hand with less mobility and a highly differentiated state of cells. For example, Lethias et al. (1983) are of the opinion that additional freeze-fracture- and TEM-investigations of sponges might reveal membrane specialisations and connections to cytoskeletal components. On the other hand, several investigators of freshwater and marine sponges could not identify such junctional complexes (e.g. Lethias et al., 1983; Weissenfels, 1990; Woollacott, 1990). While there is yet no evidence for chemical synapses in sponges, several reports discuss gap junction-like structures in the Porifera that may function as electrical synapses. For example, Green & Bergquist (1979) interpreted structures observed by SEM as temporary intercellular communication canals. Also, Revel (1988) and Garrone & Lethias (1990) describe various particle fields in freeze-fracture replicas, some of which superficially resemble gap junctions or rhomboid panicle fields of Eumetazoa but cannot unequivocally be claimed as typical gap junctions. Contractile cells in the mesohyl of sponges-actinocytes (see Boury- Esnault & Riitzler, 1997), but often called myocytes or myocyte-like cells -are arranged in networks (Bagby, 1966; Prosser, 1967; Bergquist, 1978; Burlando et al., 1984; Wachtmann et al., 1990; Harrison & DeVos, 1991) and have been seen as an early stage in the evolution of nerve and muscle cells (see literature in Lentz, 1966; Pavans de Ceccatty, 1974a; Mackie, 1990). Two types of microfilaments, that is, thin (5-7nm) and thick (15-25nm) filaments can be observed in the cytoplasm. In Tedania ignis and Hippospongia communis, only thick filaments have been reported (Harrison SEROTONIN IN PORIFERA + & De Vos, 1991). The ability to contract or condense in response to endogenous events or external stimuli is a common feature in Calcarea and Demospongiae. It results in a decrease in body size and concurrent increase in number of cell contacts (review in Simpson, 1984; Weissenfels, 1989). One can speculate whether the increased number of cell contacts is only the result of a reduced body volume or possibly serves the intensification of ‘signal transduction’ in the network of actinocytes. Frequent cell-cell contacts between myocyte-like cells were also observed in H. communis (Pavans de Ceceatty et aL, 1970). According to Pavans de Ceccally (19744), the microfilament-containing piranócyios play an important role in this process, oth for cell contraction and cell communication. Owing to the dynamic, ‘loose’ organisation of cellular features (Pavans de Ceccatty. 1974b; Bond, 1992) there are no nervous cells evident in sponges. but one can expect temporary, fixed pathways through connected mesohyl cells at the points of stable intercellular connections. Lentz (1966) reported acetylcholinesterase, monoamine oxidase, epinephrine, norepinephrin, and SHT (serotonin) in *myocyte-like" cells of Sycon ciliatum. These observations along with the association of cholinesterase and myofilaments in myocyle-like cells and the report of actin filaments in pinacocytes (Pavans de Ceccatty, 1989) may signify myoid and neuroid elements from a common integrative system that is coordinating “tissue” contractions in sponges ulthough electrophysiological evidence of a conducting mechanism is still lacking (Lawn, 1982). In conclusion, we believe our findings of serotonergic cells in the Parazoa suggest an evolutionary specialisation of serotonin, separate from its function in Protists, We interpret our observations as supporting the recently emphasised sister-group relationship with the Eumetazoa (Morris, 1993; Miiller, 1995; Ax, 1995) by exhibiting the very first steps in the evolutionary development of the integrative system of the Metazoa. Further research involving additional species and immuno- cytochemical, electrophysiological, and other approaches is clearly needed. ACKNOWLEDGMENTS We are indebted to G. Rieger, D. Reiter, P. Ladurner (University of Innsbruck), and W.E.G. Milller (University of Mainz, Germany) for valuable discussions and for critical reading, of the manuscript. We thank K.P. Smith (Dept. of Invertebrate Zoology, National Museum of Natural History, Washington, D.C.) and M.C. Diaz (University of California, Santa Cruz) for help with fieldwork in Belize. The staff uf the Department of Ultrastructure and Evolutionary Biology of the University of Innsbruck are thanked for the use of facilities and technical advice (W, Salvenmoser, in particular), This paper was presented at the 31h International Sponge Symposium in Brisbane, 1998. and we wish ta acknowledge British Airways tor the donation of guest tickets to the Smithsonian Institution which allowed K. Rützler and K.P. Smith to attend, Research in Belize, Innsbruck, and Washington was supported by grants from Fond Wissenschafilicher Forschung, (Austria) and funds from the Caribbean Coral Reel Ecosystems Program (CCRE), National Museum of Natural History. Smithsonian Institution (contribution no. 484). LITERATURE CITED AMANU, S. & HORI, I. 1994, Metamorphosis of a demosponge. 1. Cells and structure of swimming larval. Invertebrate Reproduction and Development 25: 193-204, AX, P, 1989. Basic phylogenic systematization of the Metazoa, Pp. 229-245, In Fernholm, B., Bremer, K. & Joernvall, H. (eds) The Hierarchy of Life. (Elsevier: Amsterdam.) 1995, Das System der Metazoa 1. (Gustav Fischer Verlag: Stuttgart, Jena, New York.) BAGBY, R.M: 1966. The fine structure of myocytes in the sponges Microciona prolifera (Ellis and Solander) and Tedania ignis (Durchassaing and Michelotti). Journal of Morphology 118:167- 182, 1970, The fine structure of pinacoevtes in the marine sponge Mierociona prolifera (Ellis and Solander). Zeitschrift für Zellforsehung und mikroskopische Anatomie 105: 579-594, BERGQUIST, PR. 1975. Sponges. (Hutchinson: London.) BERGQUIST, APR, SINCLAIR, M.E., GREEN, C.R. & SILYN-ROBERTS, H. 1979, Comparative morphology and behaviour of larvae of Demospongiae. Pp. 103-111. In Levi, C, & Boury-Esnault, N. (eds) Biologie des Spongiaires, Colloques Internationaux de CNRS Number 291. (Centre National de la Recherche Scientifique: Paris.) — . BOURY-ESNAULT, N. & RUTZLER, K. (eds) 1997. Thesaurus of Sponge Morphology. Smithsonian Contributions to Zoology 596: 1-55. BOND, C. 1992. Continuous cell movements rearrange anatomical structures in intact sponges, Journal of Experimental Zoology 263: 284-302. 664 BURLANDO, B.E., GAINO. E. & MARCHISIO, P.C. 1984, Actin and tubulin in dissociated spurge cells. Evidence for peculiar actin containing microextensions. European Journal of Cell Biology 35: 317-321, CARR, W.E.S,. GLEESON, R.A, & TRAPIDO-ROSENTHAL, H.G, 1989, Chemosensory systems in lower organisms: Correlations with internal receptor systems. for neurolransmitters and hormones. Advances in Comparative Environmental Physiology 5: 25-52. GARRONE, R. & LETHTAS, C. 1990. Freeze-fracture study of sponge cells. Pp. 121-128. In Rützler, K. (ed.) New Perspectives in Sponge Biology. (Smithsonian Institution Press: Washington. D.C.) GREEN, C.R. 1978. Variations in septate junction structure in the invertebrates, Pp. 338-339. In Sturgess, J.M, (ed.) Electron Microscopy II, (Microscopical Society of Canada: Toronto.) GREEN, C.R. & BERGQUIST, PR. 1979. Cell membrane specializations in the Porifera. Pp. 153-158, In Lévi, C, & Boury-Esnaull, N. (eds) Biologie des Spongiaires, Colloques Internationaux du CNRS Number 291. (Centre Nalional de la Recherche Scientifique: Paris.) 1982, Phylogenetic relationships within the Invertebrata in relation to the structure of septate junctions and the development of “occluding” junctional types. Journal nf Cell Science 53: 279-305. GRELL, K.G, 1974. Vam Einzeller zum Vielzeller. Hundert Jahre Gastrea-Theorie. Biologie in unseter Zeit 4: 65-71. GRELL, K.G. & RUTHMANN, A. 1991. A new culture method for Trichoplax adherens F. E. Schulze. Cytobiology 4: 216-240, HALDANE, J,B. 1954, La signalisalion animale. Année Biologique 58: 89-98. HARRISON, FW. & DE VOS, L. 1991, Porifera. Pp. 29-89. In Harrison, F.W. & Wesrtall, J.A. (eds) Microscopic Anatomy of Invertebrates, 2: Placozoa, Porifera, Cnidaria, and Ctenophora. ( Wiley-Liss: New York.) HYMAN, L.H. 1951. The Invertebrates. LT. Plathyhelminthes and Rhynchocoela. The acoelomate Bilateria. (McGraw Hill: New York.) IVANOV, A.V. 1988, On the early evolution of the Bilateria. Fortschritte der Zoologie 36: 349-352, KIMMEL, H.D. & CARLYON, W.D. 1990. Persistent effects of a serotonin depletor (p-chloro- phenylalanine) in regenerated planaria (Dugesia dororocephala). Behavioral Neuroscience 104: 127-34. LAWN, LD. 1982. Porifera. Pp. 49-72. In Shelton, G.A.B, ted.) Electrical Conduction and Behaviour in *Simple” Invertebrates. (Clarendon Press: Oxford.) LEDGER, P.W. 1975. Septate junctions in the calcareous sponge Syeon ciliatum. Tissue and Cell 7; 13-18, MEMOIRS OF THE QUEENSLAND MUSEUM LENTZ, TL. 1966. Histochemical localisation af neuroliumors in a sponge. Journal of Experimental Zoology 162: 171-180. LETHIAS, C., GARRONE, R. & MAZZORANA, M. 1983. Fine structure of sponge cell membranes: Comparative study with treeze-fracture and conventional thin section methods. Tissue and Cell 15: 523-535, MACKIE, G.O. 1990, The elementary nervous system revisited. American Zoologist 30: 907-920. MIYAKE, A. 1984. Blepharismane: A conjugation inducting trypotophan derivate in the ciliate Blepharisma. Pp. 807-813, In Schlossberger, H. G., Hocken, W., Linzen, B. & Steinhart, H. (eds) Progress in Tryptophan and Serotonin Research. (De Gruyter: Berlin.) MOGAMI, Ya WATANABE, K.. QOSHIMA, C. KAWANDO, A. & BABA, S.A. 1992. Regulation of ciliary movement in sea urchin embryos: Dopamine and 5-HT change the swimming behavior. Comparative Biochemistry and Physiology 1010: 251-254. MORRIS, PJ. 1993. The developmental role of the extracellular matrix suggests a monophyletic origin of the Kingdom Animalia, Evolution 47: 152-165. MÜLLER, W.E.G. 1982. Cell membranes in sponges. International Revue of Cytology 77: 129-181. 1995, Molecular phylogeny of Metazoa (Animals): Monophyletic origin. Nalurwissenschaften 82: 321-329, PAVANS DE CECCATTY, M. 19743, The origin of the integrative system: A change in view derived from research on Coelenterates and Sponges. Perspectives in Biology and Medicine 17: 379-390, 1974b, Coordination in sponges: The foundation of integration. American Zoologist 14: 895-903. 1989. Les éponges, a l'aube des communications cellulaires. Pour la Science 142: 64-72. PAVANS DE CECCATTY. M.. THINEY, Y. & GARRONE, R. 1970, Les bases ultrastructurales des communications intercelluluires dans les oscules des quelques éponges. Pp. 449-466, In Fry. W.G, (ed.) The Biology of the Porifera. (Academic Press; London.) PROSSER, C.L. 1967. Problems in the comparative physiology of non-striated muscles. Pp.133-144. In Wiersma, C,A.G, (ed,) Invertebrate Nervous Systems: Their Significance for Mammalian Neurophysiology. (University of Chicago Press: Chicago.) REVEL, J.P. 1988. The oldest multicellular animal and its junctions. Pp, 135-149, In Hertzberg, E.L.(ed.) Gap Junctions, (A.R. Liss: New York.) RIEGER, R.M.. HASZPRUNAR, G. & SCHUCHERT, P. 1991. On the origin of the Bilateria: Traditional views and new concepts. Pp. 107-113. In Simonetta, A.M, & Conway Morris, S. (eds.) The Early Evolution of Metazoa and the Significance of Problematic Taxa. (Claredon Press: Oxford.) SEROTONIN IN PORIFERA ? RUTZLER, K. & FELLER, C. 1996. Caribbean mangrove swamps. Scientific American 274(3): 94-99, SCHUCHERT, P. 1993. Trichoplax adherens (Phylum Placozoa) has cells that react with antibodies against the neuropeptide RF-amide. Acta Zoologica 74: 115-117. SIMPSON, T.L. 1984. The cell biology of sponges. (Springer-Verlag: New York.) TWAROG, B.M. 1988. Serotonin: History of a discovery. Comparative Biochemistry and Physiology 91: 21-24. UMBRIACO, D., ANCTIL, M. & DESCARRIES, L. 1990. Serotonin-immunoreactive neurons in the cnidarian Renilla koellikeri. Journal of Comparative Neurology 291: 167-78. VAN HOUTEN, J. 1990. Chemoreception in unicellular eucaryotes. Pp 343-356. In Anderson, P.V.A. (ed.) Evolution of the First Nervous Systems. (Plenum: New York.) WACHTMANN, D., STOCKEM, W. & WEISSENFELS, N. 1990. Cytoskeletal organization and cell organelle transport in basal 665 epithelial cells of the freshwater sponge Spongilla lacustris. Cell and Tissue Research 216: 145-154. WEISSENFELS, N. 1989. Biologie und mikroskopische Anatomie der Siisswasserschwámme (Spongillidae). (Gustav Fischer Verlag: Stuttgart & New York.) 1990. Condensation rhythm of fresh-water sponges (Spongillidae, Porifera). European Journal of Cell Biology 53: 373-383. WELSH, J.H. 1953. Excitation of the heart of Venus mercenaria. Archives of Experimental Pathology and Pharmacology 219: 23-29. WOOLLACOTT, R.M. 1990. Structure and swimming behaviour of the larva of Halichondria melanadocia (Porifera: Demospongiae). Journal of Morphology 205: 135-145. 1993. Structure and swimming behaviour of the larva of Haliclona tubifera (Porifera: Demospongiae). Journal of Morphology 218: 301-321. WOOLLACOTT, R.M. & PINTO, R.L. 1995. Flagellar basal apparatus and its utility in phylogenetic analyses of the Porifera. Journal of Morphology 226: 247-265. 666 MEMOIRS OF THE QUEENSLAND MUSEUM BIOCALCIFICATION IN THE INDO-PACIFIC CORALLINE DEMOSPONGE ASTROSCLERA WILLEYANA LISTER - THE ROLE OF BASOPINACODERM. Memoirs of the Queensland Museum 44; 666. 1999:- The aragonitic calcareous basal skeleton of Astrosclera is composed of 20-60um-sized aragonitic spherulites, produced by a combination of three processes. First, the spherulites are formed in large vesicle cells (LVC's) inside large vesicles in the ectosome. In a second process, after release from LVC's, basopinacocytes transport the spherulites to the tips of the skeletal pillars, where they fuse together by epitaxial growth; and in a third process, during upward growth, the soft tissue is slowly rejected from the lowermost-parts of the skeletal cavities and the remaining spaces are subsequently filled by epitaxially-growing aragonite fibers. In the second and third process, basopinacocytes produce either the insoluble intracrystalline organic matrix, which does not consist of collagen, as well as the soluble intracrystalline matrix, which consists of highly acidic Ca?'-binding mucus substances. Basopinacocytes control speed and direction of epitaxial growth in both of the latter two biocalcification processes. It is hypothesized that Astrosclera is able to control the rate of calcification by the regulation of its bacterial population. The mean growth rate of Astrosclera was measured at 230m per year. A detailed description of soft tissue ultrastructure and its cellular composition has recently been published by Wórheide (1998). O Porifera, Astrosclera, skeletal development, calcification regulation, ultrastructure. Literature cited. WORHEIDE, G. 1998. The reef cave dwelling ultraconservative coralline demosponge Astrosclera willeyana Lister from the Indo-Pacific - Micromorphology, Ultrastructure, Biocalcification, Taxonomy, Biogeography, Phylogeny, Isotope Record. Facies 38: 1-88. Gert Worheide* (email: gertw@ibm.net) & Joachim Reitner, Institut und Museum fiir Geologie und Paldontalogie (IMGP), University of Goettingen. Goldschmidtstr. 3, D-37077 Goettingen, Germany. *Present address: Marine Biology Laboratory, Queensland Museum, P.O.Box 3300, South Brisbane, Old, 4101, Australia; 1 June 1998. NITROGEN FIXATION IN SYMBIOTIC MARINE SPONGES: ECOLOGICAL SIGNIFICANCE AND DIFFICULTIES IN DETECTION CLIVE R. WILKINSON, ROGER E. SUMMONS AND ELIZABETH EVANS Wilkinson, C.R., Summons, R. & Evans, E. 1999 06 30. Nitrogen fixation in symbiotic ma- rine sponges: ecological significance and difficulties in detection. Memoirs of the Queensland Museum 44: 667-673. Brisbane. ISSN 0079-8835. There has been considerable speculation, and some evidence, that coral reef sponges can fix atmospheric nitrogen through some of their microbial symbionts, particularly symbiotic cyanobacteria. Many Indo-Pacific coral reef sponges can satisfy much of their requirement for carbon energy compounds via translocation from photosynthetic symbionts, and asimilar mechanism has been invoked to explain how some sponges could supplement the low amount of available nitrogen in clear tropical waters. Attempts to measure nitrogen fixation using the acetylene reduction test have proven technically difficult and given ambiguous results. However, fixation was demonstrated unambiguously with incorporation of the stable isotope !5N> into the amino acids glutamine, glutamate and aspartate of Callyspongia muricina, although at relatively low rates. The variability in measuring acetylene reduction in 23 sponge species is attributed to several factors: the number of cellular and matrix barriers that must be passed by acetylene and ethylene; the difficulty of maintaining sponges alive under experimental conditions; possible metabolism of ethylene by symbiotic bacteria; and possible toxicity of the reagents. O Porifera, nitrogen fixation, acetylene reduction activity, cyanobacteria, coral reef sponges, Callyspongia muricina. Clive R. Wilkinson (email: c.wilkinson@aims.gov.au) & Elizabeth Evans, Australian Institute of Marine Science, PMB No. 3, Townsville 4810, Australia; Roger E. Summons, Australian Geological Survey Organisation, GPO Box 378, Canberra 2601, Australia; 4 April 1998. Photosynthetic symbionts convey a distinct advantage to marine sponges living in low nutrient, tropical waters. These sponges have been shown to receive fixed nutrient carbon translocated from cyanobacterial symbionts (Wilkinson, 1979), such that some are virtually ‘phototrophic’, i.e. the symbionts can provide the bulk of carbon energy requirements (Wilkinson, 1983; Cheshire & Wilkinson, 1991). Such sponges are distinctly flattened to enhance light capturing, possibly at the expense of the ability to act as filter feeders (Wilkinson et al., 1988). These abilities are similar to other major photo- synthetic symbioses on coral reefs: zooxanthellate corals (Muscatine et al., 1981); tridacnid clams (Trench, 1987); and didemnid ascidians (Griffiths & Thinh, 1983). The potential to fix nitrogen was demonstrated in Red Sea sponges using the acetylene reduction test (Wilkinson & Fay, 1979). However, subsequent attempts to apply similar methods to a range of sponges from other coral reef areas proved to be highly variable. After many experiments showing equivocal results in 23 species, more direct and unambiguous nitrogen fixation methods were applied with the incubation of a sponge in the stable isotope N;. We chose a sponge from the Great Barrier Reef that had previously shown some promise at fixing nitrogen and from which cyanobacterial cell preparations could be easily obtained. Marine sponges also have many other photosynthetic and non-photosynthetic symbionts (Wilkinson, 1992). Many sponges have large populations of symbiotic bacteria, which may occupy up to 40% of the animal volume, comparable to the volume of the matrix and exceeding the volume of animal cells (Vacelet & Donadey, 1977; Wilkinson, 1978). These symbionts are particularly variable with the possibility of six or more ‘species’ occurring within the mesohyl matrix. However, few, if any, roles have been demonstrated for these sym- bionts, until recent experiments demonstrated a possible role in nitrogen metabolism within the sponge that also may be significant within coral reef ecosystems (Corredor et al., 1988; Diaz & Ward, 1997). Recently, many different bacterial forms have been observed, including Archaea-like cells (Preston et al., 1994; Fuerst et al., 1998) and methane-oxidising bacteria in sponges (Vacelet et al., 1996). 668 MATERIALS AND METHODS All field experiments were performed at Davies Reef (18°15’S; 147°38’E) on the central part of the Great Barrier Reef aboard the RV Harry Messell (location in Wilkinson & Evans, 1989). NITROGEN FIXATION BY ACETYLENE REDUCTION. The technique of Stewart et al. (1967) was used with incubations in 10-20% acetylene over acetylene saturated seawater (Flett etal., 1976). In some instances additional 5-20% oxygen gas was added to ensure that the sponges were not stressed through anaerobic conditions. Pieces (frequently transplanted several months prior to experimentation; Wilkinson & Thomp- son, 1997) or whole sponges of 23 species (116 replicates) were incubated in air tight containers, temperature buffered in running seawater and illuminated by filtered sunlight. Controls were either sponge tissue incubated in the dark, or boiled sponge tissue, or live coral rubble incubat- ed in the light. Cyanobacterial cell preparations (see below) were incubated similarly. Regular gas samples were collected over 4-6 hour incubation periods in evacuated tubes and assayed in a Tracor 222 gas chromatograph, with concentrations determined against an internal methane standard with corrections for gas solubility (Wilkinson & Sammarco, 1983). NITROGEN FIXATION WITH "N;. The sponge Callyspongia muricina was chosen because it had frequently shown some acetylene reduction activity and because large quantities of cyanobacteria can be extracted from sponge tissues relatively easily. Seawater was degassed in a stream of argon gas and then saturated overnight under a headspace of 4 parts ^N; (Amersham) to 1 part O, with stirring. Cyanobacterial preparations were obtained by gently crushing pieces of sponge, and blending the suspension in a glass blender with a teflon piston. After repeated centrifugation, washing and resuspension, a pellet was obtained that was predominantly cyanobacteria (examined microscopically). Replicate whole pieces of sponge and cell suspensions were incubated in closed chambers with no gas spaces in "N3 seawater (as above) for 20, 40 and 60mins. Approximately 3cm? of sponge tissue was fixed in ethanol: water: acetic acid (50:45:5) and then pulverised using a 2em probe diameter polytron probe blender in a 50ml polyethylene centrifuge tube in 25ml of distilled water. The MEMOIRS OF THE QUEENSLAND MUSEUM mixture was filtered through ‘Miracloth’ until the spicule mass was colourless with all cyano- bacterial cells removed. The suspension was extracted through Dowex 50 and 1 ion exchange columns to yield amino acid fractions. Centri- fuged pellets of the cyanobacterial suspensions were similarly fixed and then blended. Each fraction was reduced to 2ml and Kjeldahl digested at 150°C for 1.5hrs, followed by 330°C for 3hrs to remove all water (Bergensen, 1980). Digests were immediately converted to ammonium sulphate salts to prevent ambient contamination (Volk & Jackson, 1979) by digesting in 0.1 M H5SO, followed by 10M NaOH to convert to an alkaline solution. After 4 days incubation at 50°C, digests were evaporated to dryness under argon and analysed. The fixed nitrogen in samples was driven off by mixing with NaOBr under an argon atmosphere, and the gas injected directly into VG 602D (Cheshire, UK) computerised mass spectrometer through an ethanol-dry ice moisture trap. Calibration was with CIG ultra high purity nitrogen calibrated against N-1 and N-2 international standards (IAEA, Vienna) and results expressed as delta notation on air nitrogen scale with a reprod- uctibility of 20 replicate samples being less than 0.1ppm. Control samples of unlabelled sponge, cell suspensions and coral rubble were similarly assayed to detect natural levels of ^N;. SPONGE PARAMETERS. Wet weight, dry weight, surface area and chlorophyll a content (Jeffrey & Humphrey, 1975) were as outlined in Wilkinson (1983). RESULTS NITROGEN FIXATION BY ACETYLENE REDUCTION. Assays of 23 sponge species revealed either negative, ambivalent or slightly positive results for ethylene production (Table 1), Three species, Callyspongia muricina, Ircinia ramosa and Collospongia auris frequently showed slight, or on occasions significant positive acetylene reduction. However, there was little consistency or reproducibility of positive results in these species, with or without symbiotic cyanobacteria. Similar ambiguous results were obtained with incubations of 6 sponge species from coral reefs off the coast of Western Australia (Wilkinson, unpublished data). Variations in oxygen or acetylene concentrat- ions on incubations of whole tissue or cell preparations had no apparent effect on acetylene reduction rates. Control incubations of coral NITROGEN FIXATION IN SPONGES 669 TABLE 1. Results of acetylene reduction assays on coral reef sponges, compared to pieces of rubble as positive controls. Negative controls of incubations in the dark or with dead sponge are not reported as all were negative. Results from experiments using different concentrations of oxygen and or acetylene are combined, as there were no discernible patterns. Symbiont = lists of the type of cyanobacterial symbiont. The Results are reported as the number of sponge incubations in each category: no evolution of ethylene, and less than contamination; +/—= inconclusive result with traces of ethylene; + = slight release of ethylene but less than 0.5 nM cm2 hr-!; ++ = significant increase in ethylene >2.0 nM cm hr-!. Nutrition = whether or not sponge can obtain the bulk of their carbon energy via symbiont photosynthesis (phototrophic), versus none (heterotrophic), or both (mixed). Order Species Symbiont Nutrition = +/- + ++ Dictyoceratida | /rcinia ramosa Unicellular Phototrophic 1 9 9 0 Phyllospongia lamellosa Unicellular Phototrophic 0 1 0 0 Phyllospongia papyracea Unicellular Phototrophic 0 2 1 0 Carteriospongia foliascens Unicellular Phototrophic 1 9 1 0 Carteriospongia flabellifera Unicellular Phototrophic 2 2 0 0 Strepsichordaia lendenfeldi Unicellular Mixed 0 3 0 0 Collospongia auris Unicellular Phototrophic 1 1 2 0 Ircinia sp. 1 Multicellular Mixed 0 2 0 0 Rhopaloeides odorabile None Heterotrophic 0 2 0 0 Dendroceratida | Dysidea herbacea Multicellular | Phototrophic 0 3 0 0 Dysidea sp. 1 Multicellular Phototrophic 0 1 0 0 Haplosclerida | Callyspongia muricina Unicellular Phototrophic 7 23 4 0 Callyspongia sp. 1 None Heterotrophic 0 1 0 0 Amphimedon sp. 1 Unicellular Phototrophic 0 2 0 0 Petrosida Xestospongia exigua Unicellular Mixed 0 2 0 0 Aida, da Cymbastela sp. | Uni & Multi Phototrophic 7 20 1 0 Cymbastela sp. 2 Uni & Multi Phototrophic 0 2 0 0 Phakellia aruensis None Heterotrophic 0 1 0 0 Acanthella sp.1 None Heterotrophic 0 2 0 0 Poecilosclerida | Neofibularia irata Unicellular Phototrophic 0 3 0 0 Astrophorida Jaspis stellifera Unicellular Mixed 0 2 0 0 Hadromerida Cliona sp. BP Zooxanthellae | Phototrophic 0 1 0 0 Class Calcarea | | Pericharax heteroraphis Unicellular Mixed 0 1 0 0 None Coral rubble controls Turf algae 0 0 0 8 rubble with natural turfalgal populations showed consistent, relatively high rates of acetylene reduction, comparable to those shown by Wilkinson et al. (1984). NITROGEN FIXATION WITH "N, Definite nitrogen fixation of "N; was observed in whole sponge and cyanobacterial cell preparations of C. muricina. The highest rates of enrichment were observed in the amino acids: glutamine, glutamate and aspartate (Fig. 1). Similar rates of enrichment in the amino acid fraction were found in pieces of rubble incubated in ^N;. CELLULAR NATURE OF THE SYMBIONTS. The nature and location of cyanobacterial sym- bionts varies between sponges (Table 1). These were observed during transmission electron microscopic study of these sponges for other studies (Wilkinson, unpublished data). Three dis- tinct categories can be observed: a) cyanobacteria free living in the mesohyl; b) cyanobacteria predominantly within large vacuoles within special mesohyl cells, cyanocytes; or b) and c) cyanobacteria both within vacuoles and in the mesohyl (Fig. 2; Wilkinson, 1978). In addition, some sponge species have other symbionts including multicullular cyanobacteria and zooxanthellae (Table 1; Wilkinson, 1992). DISCUSSION Two conclusions are evident from these studies: 1) at least one sponge with cyanobacterial 670 100 KlAmino acids - E LlAqueous soln - E 80 | Camino acids - C H Aqueous soin - C Intact sponge Sponge cells SAMPLE Coral rubble FIG. 1. Total enrichment of ^N; in whole sponge and cell suspensions of Callyspongia muricina compared to coral rubble controls. Data are delta enrichment values of ^N; compared to non-labelled N within control (C) aqueous and amino acid solutions and experimental (E) aqueous and amino acid fractions. Differences between E and C are significant at P<0.001. symbionts fixes atmospheric nitrogen, but hot at rapid rates, as indicated by direct fixation of | "Ns; and 2) the relatively easier technique of acetylene reduction is unreliable and inapplicable. A presumptive conclusion is that nitrogen fixation may occur in many other sponges with symbiotic cyanobacteria, but this can only be verified with individual testing of direct incorporation of N;. The first conclusion confirms the earlier ob- servations of acetylene reduction in two Red Sea sponge species by Wilkinson & Fay (1979). We subsequently questioned those earlier results when repeated acetylene reduction tests on a larger number of species showed inconsistent results (Table 1). However, the direct incorporation of *N; as gas has demonstrated that sponges with cyanobacterial (or possibly other prokaryotic) symbionts do contain active nitrogenase. The progressive enrichment of glutamine, glutamate and aspartate demonstrate that this fixed nitrogen is of potential benefit to the host sponge as these compounds can be incorporated into sponge and symbiont protein, and metabolised for energy. Translocation of these amino acids, however, was not demonstrated, but may be assumed because the population size of microbial symbionts is usually stable with little need for protein for cell growth and there is a parallel release of fixed carbon as glycerol (Wilkinson, 1979). Any nitrogen fixation would be valuable to those coral reef sponges that live in clear tropical waters, as it can supplement the particularly low MEMOIRS OF THE QUEENSLAND MUSEUM levels of particulate nutrients and dissolved fixed nitrogen (Wiebe etal., 1975). Moreover, many of these sponges obtain the bulk of their energy from the photosynthetic symbionts as trans- located glycerol, which is rich in carbon but devoid of nitrogen (Wilkinson, 1979, 1983). Without this added source of nitrogen, sponge growth rates could be reduced through a lack of nitrogen to produce proteins, particularly for the production of the fibrous protein skeleton. The majority of sponges in clean water on the Great Barrier Reef are distinctly flattened to enhance light capture (Wilkinson, 1988). These are sponges that exhibit particularly phototrophic nutrition and have the potential to obtain virtually all of their nutrition from the symbionts down to a depth of 30m (Cheshire & Wilkinson, 1991). The following possible explanations are advanced to explain the inconsistency in acetylene reduction assays compared to coral rubble controls: 1) poor diffusion between multiple layers of cell and matrix; 2) disturbance to host sponges; 3) possible removal of ethylene by symbiotic bacteria; and 4) possible acetylene toxicity to sponges and cells. 1) The turf algae on the rubble are totally exposed to the seawater containing dissolved acetylene, with only the algal cell barriers remaining for ethylene to diffuse back into the water. Therefore, there is efficient and rapid transfer of both the acetylene into turf cyano- bacteria and similar transfer of the ethylene back into seawater, evident as high and consistent rates of ethylene production (Table 1). The situation in sponges is different as the symbionts in many sponges are contained within specialised vacuoles (cyanocytes; Wilkinson, 1978) embedded in the sponge mesohyl matrix (Fig. 2). For these symbionts, there are multiple cell and matrix layers that must be passed for both acetylene to diffuse into the cyanobacteria (or bacterial symbionts) and then for the ethylene to diffuse back out to the water, where it can be detected in water samples. This double diffusion process may be slow and inefficient, as it would rely on diffusion gradients across cell and matrix barriers. Against this argument is the fact that low molecular weight gaseous molecules like acetylene (M.W. 26) and ethylene (M.W. 28) diffuse rapidly through membranes, comparable to other gases such as nitrogen (N; and BN, ; M.W. 28 and 30) and oxygen (M.W. 32) (Cheung & Marshall, 1997), NITROGEN FIXATION IN SPONGES 671 FIG, 2. Electron micrograph of the Great Barrier Reef sponge Jaspis stellifera showing unicellular cyanobacterial symbionts both free in the mesohyl matrix and contained within vacuoles of specialised cells, the cyanocytes. Scale bar Sum. 2) Sponges have the ability to contract and cease pumping when disturbed, which would reduce water and gaseous exchanges. This was demonstrated by Reiswig (1971) and has been a consistent problem with physiological experi- ments with sponges on the Great Barrier Reef, because they frequently contract and close their oscules when placed in experimental chambers. This is most evident in massive species with large oscules, like Rhopaloiedes odorabile and Jaspis stellifera, but may also occur in small oscule species like the Phyllospongia and Carterio- spongia spp., but observing any contraction in the field is particularly difficult. Cessation of pumping activity would restrict water movement through canals and prevent a free exchange of acetylene and ethylene, thereby reducing the potential for acetylene reduction. 3) There is the possibility that symbiotic bacteria exist which can oxidise either or both acetylene and ethylene and interfere with concentration measurements. Sponges contain a large variety of bacterial symbionts (Vacelet & Donadey, 1977; Wilkinson, 1978) and recently it has been shown that there are methane-oxidising bacteria in marine sponges (Vacelet et al., 1996), as well as a wide range of Archaea-like bacteria (Preston et al., 1994; Fuerst et al., 1998). The majority of bacterial symbionts cannot be isolated in culture and have only been observed using electron microscopy. 4) Acetylene toxicity has not been shown in these sponges, but has been demonstrated in other nitrogen fixing systems (David & Fay, 1977). Most acetylene reduction assays have been applied to plants, rather than animals. Thus acetylene toxicity may have a greater impact on animal respiration and reduce or block the transfer of water through the canal system. This would result in the effects in 2) above. Irrespective of the reason for inconsistencies with the acetylene reduction method compared to the use of the stable nitrogen isotope technique, it is concluded that acetylene reduction should not be used with these animals as a method to detect nitrogen fixation. One problem is that stable isotope analysis is more expensive and difficult to apply under field conditions. However, the ready availability of new continuous-flow isotope analyses methods for carbon, nitrogen and hydrogen isotopes means that enrichment experiments are very easily evaluated at the molecular level using compound-specific isotope analyses (Merrit & Hayes, 1994; Macko et al., 1997), Verified nitrogen fixation in one sponge species has demonstrated the potential for fix- ation to be found in other sponges with microbial symbionts. Although it is more probable that the cyanobacterial symbionts are responsible for the activity, the possibility exists that bacterial sym- bionts may also fix atmospheric nitrogen in other sponges. More research is needed to confirm the origin of the nitrogen fixing enzyme, nitrogenase. Irrespective of the source, any fixed nitrogen would supplement nutrition in coral reef sponges that must make a living in low nutrient tropical waters. ACKNOWLEDGEMENTS Professor J.M. Hayes and Z. Roksandic are thanked for advice and assistance with nitrogen isotopic analyses. LITERATURE CITED BERGENSEN, F.J. 1980. Measurement of nitrogen fixation by direct means. Pp. 65-110. In Bergensen, F.J. (ed.) Methods for evaluating biological nitrogen fixation (Wiley: Chichester). CHESHIRE, A.C. & WILKINSON, C.R. 1991. Modelling the photosynthetic production by sponges on Davies Reef, Great Barrier Reef. Marine Biology 109: 13-18. CHEUNG, A.T. & MARSHALL, B.E. 1997. The inhaled anesthetics. Pp. 75-87. In Longnecker, D.E. & Murphy, F.L. (eds) Introduction to Anesthesia. 9th Edition. (Saunders: Philadelphia). CORREDOR, J.E., WILKINSON, C.R, VICENTE, V.P., MORELL, J.M. & OTERO, E. 1988. Nitrate release by Caribbean reef sponges. Limnology and Oceanography 33: 114-120. DAVID, K.A.V. & FAY, P. 1977. Effects of long-term treatment with acetylene on nitrogen-fixing MEMOIRS OF THE QUEENSLAND MUSEUM microorganisms. Applied Environmental Microbiology 34: 640-646. DIAZ, M. C. & WARD, B. B. 1997, Sponge-mediated nitrification in tropical benthic communities. Marine Ecology Progress Series 156: 97-107. FLETT, R.J., HAMILTON, R.D. & CAMPBELL, N.E.R. 1976. Aquatic acetylene-reduction techniques: solutions to several problems. Canadian Journal of Microbiology 22: 43-51. FUERST, J.A., WEBB, R.L, GARSON, M.J., HARDY, L. & REISWIG, H.M. 1998. Membrane-bound nucleoids in microbial symbionts of marine sponges. FEMS Microbiology Letters 166: 29-34. GRIFFITHS, D.L. & THINH, L.V. 1983. Transfer of photosynthetically fixed carbon between the prokaryotic green alga Prochloron and its ascidian host. Australian Journal of Marine and Freshwater Research 34: 431-440. JEFFREY, S.E. & HUMPHREY, G.F. 1975. New spectrophotometric equations for determining chlorophyls a, b, el and c2 in higher plants, algal and natural phytoplankton. Biochemische Physiologie Pflanzen 167: 191-194. MACKO, S.A., UHLE, M.E., ENGEL, M.H. & ANDRUSEVICH, V. 1997. Stable nitrogen isotope analysis ofamino acid enantiomers by gas chromatography/combustion/isotope ratio mass spectrometry. Analytical Chemistry 69: 926-929. MERRIT, D.A. & HAYES, J.M. 1994. Nitrogen isotopic analysis by isotope-ratio-monitoring gas chromatography/mass spectrometry. Journal of the American Society for Mass Spectrometry 5: 387-397. MUSCATINE, L., McCLOSKEY, L.R. & MARIAN, R.E. 1981. Estimating the daily contribution of carbon from zooxanthellae to coral animal respiration. Limnology and Oceanography 26: 601-611. PRESTON, C.M., WU, K.Y., MOLINSKI, T.F. & DELONG, E.F. 1996. A psychrophilic crenarchaeon inhabits a marine sponge: Cenarchaeum symbiosum gen. nov., sp. nov. Proceedings ofthe National Academy of Science USA 93: 6241-6246. REISWIG, H.M. 1971. In situ pumping activities of tropical Demospongiae. Marine Biology 9: 38-50. STEWART, W.D.P., FITZGERALD, G.P. & BURRIS, R.H. 1967. Jn situ studies on N; fixation using the acetylene reduction technique. Proceedings of the National Academy of Science USA 58: 2071-2078. TRENCH, R.K. 1987. Dinoflagellates in non-parasitic symbioses. Chapter 12. Pp. 530-570. In Taylor, F.J.R. (ed.) Biology of Dinoflagellates (Blackwell: Oxford). VACELET, J. & DONADEY, C. 1977. Electron microscope study of the association between some sponges and bacteria. Journal of Experimental Marine Biology and Ecology 30: 301-314, NITROGEN FIXATION IN SPONGES 673 VACELET, J. FIALA-MEDIONI, A., FISHER, C.R. & BOURY-ESNAULT, N. 1996. Symbiosis between methane-oxidizing bacteria and a deep-sea carnivorous cladorhizid sponge. Marine Ecology Progress Series 145: 77-85. VOLK, R.J. & JACKSON, W.A. 1979. Preparing nitrogen gas for nitrogen-15 analysis. Analytical Chememistry 51: 463. WIEBE, W.J., JOHANNES, R.E. & WEBB, K.L. 1975, Nitrogen fixation in a coral reef community. Science 188: 257-259. WILKINSON, C.R. 1978. Microbial associations in sponges. III, Ultrastructure of the in situ associations in coral reef sponges. Marine Biology 49: 177-185. 1979. Nutrient translocation from symbiotic cyanobacteria to coral reefsponges. Pp. 373-380. In Levi, C. & Boury-Esnault, N. (eds) ‘Biologie des Spongiaires’ Colloques Internationaux du Centre National de la Recherche Scientifique No. 291 (CNRS: Paris). 1983. Net primary productivity in coral reef sponges. Science 219: 410-412. 1988. Foliose Dictyoceratida of the Australian Great Barrier Reef. I1. Distribution and ecology. P.S.Z.N.I. Marine Ecology 9: 321-327. 1992, Symbiotic interactions between sponges and algae. Pp. 112-151. In Reisser, W. (ed.) Algae and Symbioses: Plants, Animals, Fungi and Viruses, Interactions Explored. (Biopress: Bristol). WILKINSON, C.R., CHESHIRE, A.C., KLUMPP, D.W. & McKINNON, A.D. 1988. Nutritional spectrum of animals with photosynthetic symbionts — corals and sponges. Pp. 27-30. In Choat, J.H. et al. (eds) Proceedings of the 6th International Coral Reef Symposium, Townsville 1988. Vol.3 (6th International Coral Reef Symposium Executive Committee: Townsville). WILKINSON, C.R. & EVANS, E. 1989. Sponge distribution across Davies Reef, Great Barrier Reef, relative to location, depth, and water movement. Coral Reefs 8: 1-7. WILKINSON, C.R. & FAY, P. 1979. Nitrogen fixation in coral reef sponges with symbiotic cyanobacteria. Nature 279: 527-529. WILKINSON, C.R, & SAMMARCO, P.W. 1983. Effects of fish grazing and damselfish territoriality on coral reef algae. II. Nitrogen fixation. Marine Ecology Progress Series 13: 15-19. WILKINSON, C.R. & THOMPSON, J.E. 1997. Experimental sponge transplantation provides information on reproduction by fragmentation. Pp. 1417-1420. In Proceedings of the Sth International Coral Reef Symposium, Panama, June 1996. Vol. 2 (8th International Coral Reef Symposium Executive Committee: Panama). WILKINSON, C.R., WILLIAMS, D.McB., SAMMARCO, P.W., HOGG, R.W. & TROTT, L.A. 1984. Rates of nitrogen fixation on coral reefs across the continental shelf of the central Great Barrier Reef. Marine Biology 80:255-262, 674 MEMOIRS OF THE QUEENSLAND MUSEUM RAPID CHANGE AND STASIS IN A CORAL REEF SPONGE COMMUNITY. Memoirs of the Queensland Museum 44: 674. 1999:- Four censuses of a sponge community on a shallow coral reef in San Blas, Panama, have revealed a combination of both extreme change and also apparent stasis over the 14 years between 1984-1998. Total biomass of the sponge assemblage varied little over the first 11 years, with the exception of the first few years after a hurricane decreased sponge populations in 1988. However, relative contributions to total biomass by the different species have changed to the extent that over half of the original species are now altogether absent from the censused area. Species lost were not necessarily those that had been rare initially, and the hurricane does not appear to have been responsible for the loss of species. The most striking pattern of loss is that keratose species account for a disproportionately large number of the species and also of the volume of biomass lost. Growth form also seems to influence vulnerability to loss, as massive forms were lost disproportionately and no erect branching forms were lost. Pathogens appear to be the agents of at least some of the mortality, with high rates of infection by what seem to be species-specific pathogens in the most common species. CJ Porifera, coral reef sponges, population dynamics, disease, environmental ecology. Janie L. Wulff (email: wulff@jaguar middlebury.edu), Biology Department, Middlebury College, Middlebury, VT 05753, USA; 1 June 1998, GROWTH AND REGENERATION RATES OF THE CALCAREOUS SKELETON OF THE CARIBBEAN CORALLINE SPONGE CERATOPORELLA NICHOLSONI: A LONG TERM SURVEY PHILIPPE WILLENZ AND W.D. HARTMAN Willenz, Ph. & Hartman, W.D. 1999 06 30: Growth and regeneration rates of the calcareous skeleton of the Caribbean coralline sponge Ceratoporella nicholsoni: a long term survey. Memoirs of the Queensland Museum 44: 675-685. Brisbane. ISSN 0079-8835. The growth rate of the aragonitic skeleton of the Caribbean ‘sclerosponge’ Ceratoporella nicholsoni has been studied by in situ staining of specimens with calcein in a reef tunnel, 28m depth, near Discovery Bay, Jamaica. Experiments were performed up to five times from 1984 to 1997 ona population of 10 specimens ranging from 10-20cm maximum diameter. In each experiment small skeletal samples were removed from the periphery of sponges, and specimens were left in place for further studies on growth and regeneration. Perpendicular sections, ground to a thickness of about 10um, were photographed by fluorescence microscopy. Annual skeletal growth rates were calculated from measurements of the linear extension between calcein stained lines along growth axes. Data indicate that although average annual growth rates remained in the same range for different periods (214.6+54,5-233.3+33.0um yr!), significant differences occurred from one individual to another within the same period. The annual growth rate of a given individual also varied significantly in time (191.1+30,0-269.9+37.0um yr!). A second population of smaller individuals, measured after a single period of one year, revealed a strikingly lower average annual growth rate (124,4+35.0um yr-!), Regeneration of the skeleton of injured specimens was also characterised by an initial slower growth rate. Nevertheless, after the first year, it was comparable to normal growth, and exceeded it slightly thereafter. This first long term study of Ceratoporella provides important information on the variability in growth rates, with implications on the use of sclerosponges as paleoenvironmental proxies. O Porifera, sclerosponges, coralline sponges, growth rate, aragonite, skeleton, regeneration, calcein, Ceratoporella nicholsoni. Philippe Willenz (email: pwillenz@ulb.ac.be), Department of Invertebrates, Royal Belgian Institute of Natural Sciences, 29 Rue Vautier, B-1000 Brussels, Belgium; W.D. Hartman, Yale University, Peabody Museum of Natural History, P.O. Box 208118, New Haven, CT, USA 06520-8118; 18 January 1999. Although the first specimen of the coralline sponge Ceratoporella nicholsoni was dredged off Cuba in 1878, and described as a new alcyon- arian coelenterate more than thirty years later (Hickson, 1911), it was not until the mid-1960s that this species was rediscovered (Hartman & Goreau, 1966), and subsequently shown to be a sponge (Hartman & Goreau, 1970). At the same time, thanks to the increased use of SCUBA diving, the extent of the diversity of ‘sclero- sponges’ became evident (Hartman, 1969; Hartman & Goreau, 1970, 1975). Among the nine known species of Caribbean coralline sponges, Ceratoporella nicholsoni secretes the most massive basal skeleton of calcium carbonate. Despite the fact that the ecology and ultrastructure of Ceratoporella have now been extensively investigated (Lang et al., 1975; Willenz & Hartman, 1989), the growth rate of its aragonitic skeleton is still unknown, more than a century after its discovery. Both direct staining and indirect techniques have been used to evaluate the growth rate of Ceratoporella: the former using alizarin red (Dustan & Sacco, 1983) or calcein stains (Willenz & Hartman, 1985); the latter based on ^C and 7'°Pb chronologies (Benavides & Druffel, 1986; Druffel & Benavides, 1986) or focusing on carbon and oxygen isotope studies (Joachimski et al., 1995; Böhm et al., 1996). Considering the estimated slow calcification rate of this species, the latter techniques are the most convenient for elucidating long term information (time scales of tens of years to centuries). Direct methods, however, have the potential advantage of revealing data on growth rates for shorter time scales (years to decades). Calcein was first used in invertebrates to mark 676 TABLE 1. Ceratoporella nicholsoni. Successive in situ labeling with calcein (*). Abbreviations: Tl, 9.V11.1984; T3, 15.11.1985; T4, 29.1V.1986; T5, 1.V.1987; T6, 1.V.1997. Specimen Experimental period number Ti T3 T4 T5 T6 4 * * * * * 7 d * * * * 8 E * * * * 9 * * * * 10 * * * * 1 ] * * * | 14 * * * * 16 * * * + * 17 | * * * 27 * * 29 * * DURATION E T1-T3 221 days T3-T4 438 days T4-T5 363 days T5-T6 10 years the newly deposited calcium carbonate of the basal skeleton of Ceratoporella (Willenz & Hartman, 1985). Subsequently, this chemical has been used to record calcification amongst a wide variety of taxa such as brachiopods, bryozoans, molluscs and echinoderms (reviewed in Rowley & Mackinnon, 1995). More recently it has also been employed in studies ofthe growth dynamics of calcareous sponge spicules (Ilan et al., 1996), or to estimate the growth rate of the Indo-Pacific coralline sponges Acanthochaetetes wellsi (Reitner & Gautret, 1996) and Astrosclera willeyana (Wórheide, 1998). From these studies, calcein appears to be permanently bound to calcium carbonate that forms in the presence of the dye, although the chemistry of the process has yet to be studied. Calcein has the advantages of fluorescing brightly under UV light and having only weak toxicity. Several specimens of Ceratoporella nicholsoni, including the four individuals used in the first experiment by Willenz & Hartman (1985), were repeatedly stained and sampled at different intervals over 13 years, in order to evaluate potential growth rate variations among the sponges during extended periods of time. MEMOIRS OF THE QUEENSLAND MUSEUM MATERIALS AND METHODS EXPERIMENTAL DESIGN. Two size categories of Ceratoporella nicholsoni were studied in a reef tunnel at depths ranging from 25-29m at Pear Tree Bottom, 5km E of Discovery Bay, Jamaica. The largest individuals, 10-15cm diameter, were labelled with calcein (Fluka 21030) in situ without being removed from their substrate. Labelling was performed from 1984 to 1997 at intervals given in Table 1. After the initial labelling in July 1984 (T1) a second incubation was performed six days later (T2), to test the FIG. 1. Ceratoporella nicholsoni. Natural growth pattern. Ground section from specimen no. 16 sampled in May 1997, viewed by epifluorescence microscopy. Successive labeling events with calcein are indicated at apex of walls separating pseudocalicles, along aragonitic skeleton (A) extension axis. Living tissue (T) is brightly fluorescent. N=natural growth axis, S=surface of living tissue. (Scale bar=500um). CALCAREOUS SKELETON OF CERATOPORELLA TABLE 2. Ceratoporella nicholsoni. Annual growth rates during the 10 year experimented period. A, Kruskal-Wallis ANOVA on ranks (H=379.5 with 9 degrees of freedom; P<0.0001). B, All pairwise multiple comparison procedures (Dunn’s Method). NS indicates no significant difference and * indicates significant (P<0.05) difference. Both statistics indicate a significant variability between specimens (P<0.005). 677 concentration of 100mg/l. Bags were removed from the sponges after 12 or 24hrs. For large specimens, samples of the skeleton, with attached living tissue, about 1-3cm? in volume, were removed with hammer and cold chisel A. Specimen | Median 25% 75% Mean N from the periphery ofthe sponge, each i 236.0 228,0 24210 234.6 42 specimen was left in place for further 7 224.0 2215 | 2280 2243 29 growth and regeneration. Small 9 280.0 275.5 284.0 2794 33 specimens were sacrificed after one 10 219.0 210.0 224.0 216.9 36 year. Following dehydration in a T 248.0 242.0 254.5 249.5 37 graded series of alcohols, samples were 14 290.0 281.5 296.0 2873 61 embedded in Spurr's medium (Spurr, ié 2596 2. |. cased 426 s $ 1969). Sections, cut with a low speed 17 215.0 212.0 222.0 215.8 18 diamond saw (Bennet Labcut 1010) o E 2145 bas Mtis "t were mechanically ground on a series > ina a T TT 2 of diamond grinding disks (Buehler - : 2 £ : ultra-prepTM) using a semiautomatic grinder (Buehler Minimet 1000) to a B. Specimen) 4 | 7 | 9 | 10 | 11 | 14 | 16 | 17 | 27 | 29 | thickness of 5-10um and observed 4 F under epifluorescence microscopy A wr 3 A (Nikon Optiphot-2 microscope, excit- E Tr ation filter 340-380nm, barrier filter " Fareed eis: 420nm). fi zm Fue osi gen, Growth increments of the aragonitic skeleton were established 1 ra pear i Eisen by measuring the linear extensions Li NS NS} * | * | NS | * | - (in micrometers) between stained 17 EAN ANS AGUA AS BI lines along growth axes at the apical 27 * |NS|*|NS *|* | * [NS | - edges of the wall separating two 29 * * [+ [| + | + | + | * | NS | NS] - pseudocalices, or, for the most recent sensitivity of the method. Although distinct bands could be detected at a distance of about 441m (Willenz & Hartman, 1985), interval T1-2 was omitted in this analysis because of the shortness of the time period involved. Additional smaller specimens (11-25mm diameter) were removed from the substrate and cemented in situ to Plexiglas plates (5 specimens /12x12cm plate) with epoxy underwater patching compound (Pettit Paint Co no. 7050 & 7055). Plates were stored in Plexiglas racks placed on a ledge of the tunnel at the depth of collection. To label the sponges with dye, the large speci- mens were individually enclosed within a plastic bag (of 4L volume) that was secured around the base of the sponge with nylon cords or rubber bands. In the case of plates bearing small speci- mens, the Plexiglas racks securing the plates were enclosed in a plastic bag. Calcein, dissolved in sea water, was injected in each bag to reach a period, between stained lines and the surface of the skeleton. DATA ANALYSIS. Statistical analyses were performed using SIGMASTAT and SIGMAPLOT (Jandel Scientific) data analysis and graphics software. All linear extension measurements were normalised as annual growth rate prior to their analysis. Non parametric Kruskal-Wallis analysis of variance (ANOVA) on ranks was performed to test two null hypotheses. 1) Ho: there are no differences in the average growth rates among specimens of Ceratoporella within a given TABLE 3. Ceratoporella nicholsoni. Comparison between linear annual growth rate and regeneration growth rate. Unavailable data due to bioerosion in a specimen are indicated (*). Period Lear eit piede n Specimens T3-4 230.5 61.2 194.2 + 42.2 7-9-10-17* T4-5 232,0 + 59.6 238.4 + 38.8 7-9-10-11-17 T5-6 244.4 + 25.10 272.2 + 36.9 9-10-11-16 678 MEMOIRS OF THE QUEENSLAND MUSEUM FIG. 2. Ceratoporella nicholsoni. A, ground section of fragment sampled at T5, view at base of pseudocalyx. Two possible orientations of section plane indicate that measurement can easily be biased when section is not parallel to growth axis (T1’-T3’ < T1-T3; T3’-T4’ < T3-T4). T= living tissue. (Scale bar=250um); B, ground section of fragment sampled at T6, view at apex of wall separating two pseudocalices. Narrow structure of walls (arrow) prevents reading errors. Here, space between two walls has been filled as sponge grew upward. (Scale bar=250um). TABLE 4. Ceratoporella nicholsoni. Measurement reproducibility test. A Mann-Whitney test indicates that differences obtained from two different slides of the same sample are not significantly different (P < 0.005), except for specimen 16 in period T3-4. Specimen Period T3-4 _ Period T4-5 | slide no. | Mean | n P Mean | n P 7a | 2734 | 24 |0.1120| 287.6 | 24 | 0.0395 |7b — | 2609 | 20 | 266.6 | 21 E 9a 2681 | 35 |0.0272| 2543 | 35 | 0.6680 9b | 2585 | 35 | — | 2548 | 35 | [10a 160.7 | 26 [0.7190| 204.8 | 26 | 0.7490 | ob | 1591 | 22 | — | 2024 | 24 | téa — | 177.0 | 35 |0.0054| 2116 | 35 | 0.7960 | 16b — | 1897 | 35. lama | as | period; 2) Ho: there are no differences in the average growth rates among various periods for a given specimen (Sokal & Rohlf, 1981; Fox etal., 1994). Where the ANOVA on ranks rejected the null hypothesis the Dunn’s all pairwise multiple comparison procedure was used to determine the groups that differed from each other (P<0.005 level). A Mann-Whitney rank sum test was used where only two groups were to be compared. Significant differences were concluded at P<0.005 level. Data reproducibility was tested for four of the largest specimens by comparing measurements from a second ground section prepared from the same fragment. CALCAREOUS SKELETON OF CERATOPORELLA Average: 214.6 + 54.5 um 4 7 8 16 Specimens FIG. 3. Ceratoporella nicholsoni. Mean linear annual growth rates (um yr ^') of 4 specimens during period T1-3 (221 days). Average of all specimens is indicated. Numbers of measurements are indicated within bars. A Kruskal-Wallis ANOVA on ranks and all pairwise multiple comparison test (Dunn's method) indicate a significant variability between specimens (P«0.005), except between specimen numbers 4 vs 8. RESULTS LINEAR ANNUAL GROWTH RATES. Observation of ground sections of fragments of Ceratoporella nicholsoni in epifluorescent micro- scopy revealed successive labelling with calcein (Fig. 1). Figure 2A-B shows details of the bright fluorescent bands at the base of a pseudocalicle and at the apex of walls separating two units, respectively. It is shown that variations in the orientation of sections can induce larger measurements errors at the base than at the apex. Consequently, only measurement at the apexes were considered. Figure 3 presents the means of measurements of the four specimens of Ceratoporella successively marked during the first period T1-3 (221 days), as well as the average growth rate of the population. The results of Kruskal-Wallis ANOVA on ranks test indicate significant variability between the means (P<0.0001). However, a Dunn’s all pairwise multiple comparison procedure determined that specimens no. 4 and 8 did not differ from each other. 679 Average: 217.4 + 54.5 um Specimens FIG. 4. Ceratoporella nicholsoni. Linear annual growth rates (um yr ~) of 8 specimens during period T3-4 (438 days). Conventions indicated as in Figure 3. A Kruskal-Wallis ANOVA on ranks and all pairwise multiple comparison test (Dunn’s method) indicate a significant variability between specimens (P<0.005), except between specimens 11 vs 14, 14 vs 16, 8 vs 16, 16 vs 10, and 10 vs 4. Figures 4-6 present the same procedure for periods T3-4 (438 days), T4-5 (363 days) and T5-6 (10yrs), respectively. Identical statistical tests also indicate a significant variability between specimens, except for pairs indicated in the captions. In period T4-5, a batch of smaller samples gave rise to an ¡Average annual value (124,4+35.0um yr!) reaching only half the average annual growth rate of the larger specimens (233.3+45.0um yr'). However, individual measurements did not reveal a direct correlation between the size of sponges and their growth rate. Table 2 presents the results of a Kruskal-Wallis ANOVA on ranks on data available from the longest interval (T5-6), showing a significant variability between specimens (P<0.005). An all pairwise multiple comparison procedure indicates in detail which pairs are significantly different. Comparison of the mean annual growth rates of specimens from one period to the other (Fig. 7) also indicates a significant variability (191.1+30-269.9+37, Oum yr'), revealing that individual growth rate amongst Ceratoporella was not steady either. 680 MEMOIRS OF THE QUEENSLAND MUSEUM 350 Average: 233.2 + 45.0 um 300 A | 250 200 um 150 100 33] 45] 36] 69] 50} 23) 441701117 4 7 8 9 10 11 14 16 17 109] 41 Average: 124,4 + 35.0 um B 23] 40] 45165] 48/64] 47 | 36) 28) 45] 60] 10 42 43 44 45 47 62 63 64 65 66 74 107108 109 Specimens FIG. 5. Ceratoporella nicholsoni. Linear annual growth rates (um yr ') of A, 9 large specimens and B, 14 smaller ones during period T4-5 (363 days). Conventions indicated as in Figure 3. A Kruskal-Wallis ANOVA on ranks and all pairwise multiple comparison test (Dunn’s method) indicate a significant variability between specimens (P<0.005), except between specimens 4 vs 10, 8 vs 11,8 vs 16, for the large specimens and specimens 42 vs 62, 45 ys 64, 108 ys 109, 109 ys 107, 107 vs 44 for the small ones. Both populations have distinct average annual growth rates. REGENERATIVE ANNUAL GROWTH RATES, Each sampling caused an injury to the sponge, leaving a bare fracture of the skeleton (Fig. 8A), The living tissue rapidly extended over this fracture within a few weeks and covered it completely after a month (observations reported by a local diver). No detailed measurement was done, but at most, after 200 days the naked fracture was healed (shortest interval between personal observations). Subsequent labelling followed by re-sampling in the healed zone (Fig. 8B) provided direct observation on ihe regeneration pattern of the skeleton. Ground sections show that new walls are progressively erected to form new pseudo- calices perpendicularly oriented toward the fracture (Fig. 9). Measurements of the extension of the skeleton show that in the first period following an injury, the average regeneration rate is lower than the average normal linear growth rate measured on the same specimens (Fig. 10, Table 3). In the subsequent periods, the regeneration rate increased, exceeding progressively the normal average growth rate. Average: 233.3 + 33.0 um 4 7 9 10 11 14 16 17 27 29 Specimens FIG. 6. Ceratoporella nicholsoni. Linear annual growth rates (um yr") of 10 specimens during period T5-6 (10 yr). Conventions indicated as in Figure 3. A Kruskal-Wallis ANOVA on ranks and all pairwise multiple comparison test (Dunn's method) indicate a significant variability between specimens (P<0.005), except as indicated in Table 2. CALCAREOUS SKELETON OF CERATOPORELLA 681 350 350 350 350 300 300 300 300 250 250 250 250 um 200 200 200 200 150 150 150 150 100 100 100 100 50 50 50 50 0 0 0 0 T1-3 T34 T45 AAGR: 211.5 + 43.4 T3-4 T4-5 T56 AAGR: 263.0 * 24.2 T1-3 T3-4 T4-5 T5-6 AAGR: 193.3 * 45.7 T1-3 T3-4 T4-5 T5-6 Periods AAGR: 269.9 + 37.0 350 350 350 350 300 300 300 300 250 250 250 250 um 200 200 200 200 150 150 150 150 100 100 100 100 50 50 50 50 0 0 0 0 T1-3 T3-4 T4-5 T5-6 AAGR: 269.9 * 37.0 T34 T4-5 T56 AAGR: 266.3 + 25.5 T3-4 14-5 T5-6 AAGR: 268.1 * 30.0 T3-4 T4-5 T5-6 AAGR: 191.1 * 30.0 Periods Fig. 7. Ceratoporella nicholsoni. Growth rate variability within samples, from one experimental period to another. A Kruskal-Wallis ANOVA on ranks and all pairwise multiple comparison test (Dunn's method) indicate a significant variability between means (P«0.005), except for the following specimens and periods: specimen 4 (T4-5 vs T5-6), specimen 7 (T1-3 vs T4-5; T 3-4 vs T4-5), specimen 10 (T4-5 vs T5-6), specimen 11 (T3-4 vs T 4-5), specimen 16 (T1-3 vs T4-5). Specimen numbers are indicated within each graph. AAGR- Average annual growth rate. Obvious but unexpected marks from sampling were noticed on several specimens when revis- iting them in May 1997 (Figs 8A-8B). Measure- ment of the extension of those particular regeneration areas indicates that someone had shown interest in our samples exactly one year before. Luckily, only one specimen had disapp- eared (specimen no. 8). REPRODUCIBILITY OF THE MEASURE- MENTS. In order to test the reliability of the method, measurements of pairs of sections prepared from the same fragment for four of the largest specimens were compared. A Mann- Whitney test indicates that differences obtained from two different slides were not significant, except for one specimen (Table 4). DISCUSSION Both tetracycline and calcein were used to initiate this study (Willenz & Hartman, 1985). First experiments found that tetracycline failed to label the skeleton of Ceratoporella. Although no harm was apparent to the organism, further experimentation in an attempt to adjust the concent- ration of the dye, to improve its recovery in the skeleton, was abandoned. This was decided in consideration of the potential effects the antibiotic might have on the abundant intercellular symbiotic bacteria present in sponge tissues (Willenz & Hartman, 1989; Hartman & Willenz, 1990). Calcein, first used then on invertebrates, was considered as a more appropriate benchmark to measure ‘growth since marking’. At that time, there was no indication of the permanence of its strong fluorescence for long term measurements, whereas this study clearly showed that calcein is stable enough to mark aragonite for at least 13 years. Based on calcein-labelling experiments, the average annual growth rates of Ceratoporella nicholsoni were shown to remain in the same range throughout the different experimental periods (214.6454.5-233.3433.0um yr ). Measurements made over the largest time interval (10yrs) are Obviously most indicative of average annual growth rates (233.3+33.0um yr). These values are comparable to other studies based cither on indirect radiata methods ¿270um yr! using "C, and 220um yr! using "Pb (Benavides & Druffel, Druffel & Benavides, 1986), 180- -260um ' (Joachimski et al., 1995), and 220um yr MEMOIRS OF THE QUEENSLAND MUSEUM FIG. 8. Ceratoporella nicholsoni. A, Specimen no. 14 after a particularly severe sampling in May 1987. Arrow indicates fractured aragonitic skeleton. Encircled zone was found missing in 1997. (Scale bar-5cm); B, same specimen as in A, seen in May 1997. Zone fractured in 1987 has healed and shows round edges created by skeleton regeneration. Arrows indicate sharp edge of a more recent unexpected injury. (Scale bar=5cm). CALCAREOUS SKELETON OF CERATOPORELLA 683 FIG. 9. Ceratoporella nicholsoni. Regeneration growth pattern. Ground section from specimen no. 10 sampled from its regeneration area in May 1997, Fracture, indicated by arrows occurred in May 1986, when a fragment was sampled. T6 indicates surface of skeleton marked with calcein 363 days later. R= regeneration growth axis. Epifluorescence microscopy. Captions as in Figure 1. (Scale bar= 500um). (Böhm et al., 1996)], or on direct methods (Dustan & Sacco, 1983; Willenz & Hartman, 1985). No details of analytical methods were reported by Dustan & Sacco (1983), who used alizarin red staining; their data remain approximate (0.1-0.2mm yr). Initial values using calcein (Willenz & Hartman, 1985) were slightly lower (184.2+19.4um yr”) than in the present study because measurements at the infilling zone at the base of pseudocalices were included in the latter study. Such artifacts were avoided here by rejecting measurements made in this zone, because they are too sensitive to minor deviations in the orient- ation of skeletal sections. This study also provides clear evidence that statistically significant differences occur in the growth rate from one individual to another, within the same period of time. Moreover, the annual growth rate of a given individual also varied significantly with time. Variations in growth rate were shown to occur in two other circumstances, suggesting that young tissues produce aragonite at a slower rate than tissues of older individuals. Firstly, measure- ments on a second population of smaller sized individuals, revealed a striking lower average annual growth rate (124.4+35.0um yr”). Secondly, injuries made to large specimens induced horizontal regeneration zones which also appeared to reduce growth rate. In this latter case, however, after one year, growth was recorded but it was never higher than 18% of the normal growth figure. The impressive lateral regeneration growth rate (102-154 times faster) reported by Lehnert & Reitner (1997) corresponds to an ordinary extension of the living tissue produced to repair a damaged zone of the sponge prior to calcification. However, these authors did not record growth rates of the skeleton itself. The present study examined populations of Ceratoporella, whereas most previous reports on growth rates were based on measurements of single specimens. These studies assumed a constant growth rate during the lifetime of the sponge. In this work we found that statistically significant variations are consequential, espec- ially in researches relating growth rates and marine paleoenvironmental conditions such as water temperature or salinity, using skeletal chem- istry of sclerosponges. ACKNOWLEDGEMENTS We are indebted to Roland Baron, Department of Internal Medicine and Cell Biology, Yale University School of Medicine, for his cordial reception and for suggesting the use of calcein to us. During the many years that this work has proceeded, underwater work was made possible with the assistance of the following buddy divers: Jeremy Woodley, Peter Gayle, Robin Hall, Debbie Santavy and Pierre Van de Steen. Field work assistance by Sylvain and Donat Willenz has been much appreciated. We are indebted to Mr Eric Regout for making special arrangements with his company to lend us a Nikon Optiphot-2 epifluorescent microscope on several occasions, prior to its purchase. This work was supported by National Science Foundation Grant BSR-8317690 to Yale University, and by the following to one of us (Ph.W): two NATO science fellowships, two Fondation Agathon de Potter subventions, a Lerner-Gray Fund for 684 BOBO 0294949494949. 9.94 SII VL QS vw ow .. SKC AOUE LX A M N wo -à o MEMOIRS OF THE QUEENSLAND MUSEUM Averages: T3-4 194.2 + 42.2 ym EJ 14-5 238.4 + 38.8 um EX T5-6 272.2 + 36.9 um X ¿2 9.9.4 LIX) XX XXX Q e > SX xX Y xX 9, 2 EHO = SOX x KX Sx > <> X 99 KRS Y, e LX UT eT em SS OS or: WOE we = KX 16 17 Specimens FIG. 10, Ceratoporella nicholsoni. Regenerative annual growth rate (um yr!) of 6 specimens after injury caused by sampling. Conventions indicated as in Figure 3. A Kruskal-Wallis ANOVA on ranks and all pairwise multiple comparison test (Dunn’s method) indicate a significant growth rate increment within a given specimen, from one period to the next one. Marine Research of the American Museum of Natural History (1987), a subvention from Yale University (1987), and from the Léopold III Funds for Nature Exploration and Conservation (1997). Presentation of these data at the Sth International Sponge Symposium, Origin & Outlook, was supported by the Queensland Museum and the Symposium sponsors, and a travel grant of the Fonds National de la Recherche Scientifique. Mr, Yves Barette patiently helped with slide preparations. We extend our gratitude to Dr Jackie Van Goethem for his encourage- ments and significant support in the completion of this work. This is Contribution No. 607 from the Discovery Bay Marine Laboratory, University of the West Indies, Discovery Bay, Jamaica, W.I. LITERATURE CITED BENAVIDES, L.M. & DRUFFEL, E.R.M. 1986. Sclerosponge growth rate as determined by ? "Pb . and 5'*C chronologies. Coral Reefs 4: 221-224. BOHM, F., JOACHIMSKI, M.M., LEHNERT, H., MORGENROTH, G., KRETSCHMER, W., VACELET, J. & DULLO, W.-CHR. 1996. Carbon isotope records from extant Caribbean and South Pacific sponges: Evolution of 8"C in surface water DIC. Earth and Planetary Science Letters 139: 291-303. DRUFFEL, E.R.M. & BENAVIDES, L.M. 1986. Input of excess CO; to the surface ocean based on 15C/PC ratios in a banded Jamaican sclerosponge. Nature 321: 58-61. DUSTAN, P. & SACCO, W.K. 1983. Hidden reef builders: the sclerosponges of Chalet Caribe Reef. Discovery 16(2): 12-17. FOX, E., KUO, J., TILLING, L. & ULRICH, C. 1994. Sigmastat User's guide. (Jandel Scientific Erkrath: Germany). HARTMAN, W. D. 1969. New genera and species of coraline sponges (Porifera) from Jamaica. Postilla, Peabody Museum of Natural History, Yale University 137: 1-39. HARTMAN, W.D. & GOREAU, T.F. 1966. Ceratoporella, a living sponge with stromato- poroid affinities. American Zoologist 6(4): 262. 1970. Jamaican coralline sponges: Their morphology, ecology and fossil relatives. Pp. 205-243. In Fry, W.G. (ed.) The Biology of the Porifera. Symposia of the Zoological Society of London, No. 25 (Academic Press: London). 1975. A Pacific tabulate sponge, living repres- entative of a new order of sclerosponges. Postilla, Peabody Museum of Natural History, Yale University 167: 1-21. HARTMAN, W.D. & WILLENZ, PH. 1990. Organization of the choanosome of three Caribbean sclerosponges. Pp. 228-236. In Rützler, K. (ed.) New Perspectives in Sponge Biology. (Smithsonian Institution Press: Washington D.C.). CALCAREOUS SKELETON OF CERATOPORELLA HICKSON, S.J. 1911. On Ceratopora, the type ofa new family of Alcyonaria. Proceedings of the Royal Society of London (B) 84: 195-200. ILAN, M., AIZENBERG, J. & GILOR, O. 1996. Dynamics and growth patterns of calcareous sponge spicules. Proceedings of the Royal Society of London (B) 263: 133-139. JOACHIMSKI, M.M., BOHM, F. & LEHNERT, H. 1995. Longterm isotopic trends from Caribbean Demosponges: Evidence for isotopic disequilibrium between surface waters and atmosphere. In Lathuliére, B. & Geister, J. (eds) Proceedings of the 2nd European Regional Meeting of the International Society for Reef Studies, Luxembourg. Publications du service géologique du Luxembourg 29: 141-147. LANG, J.C., HARTMAN, W.D. & LAND, L.S. 1975. Sclerosponges: Primary framework constructors of the Jamaican deep fore-reef. Journal of Marine Research 33(2): 223-231. LEHNERT, H. & REITNER, J. 1997. Lebensdauer und Regeneration bei Ceratoporella nicholsoni (Hickson, 1911) und Spirastrella (Acantho- chaetetes) wellsi (Hartman & Goreau, 1975). Geologische Blaetter fiir Nordost-Bayern und angrenzende Gebiete 47(1-4): 265-272. REITNER, J. & GAUTRET, P. 1996 Skeletal Formation in the Modern but Ultraconservative Chaetetid Sponge Spirastrella (Acanthochaetetes) wellsi (Demospongiae, Porifera). Facies 34: 193-208. 685 ROWLEY, R.J. & MACKINNON, D.I. 1995. Use of the fluorescent marker calcein in biomineralisation studies of brachiopods and other marine organisms. Bulletin de l'Institut océanographique 14(2): 111-120. SOKAL, R.R. & ROHLF, F.J. 1981. Biometry. The Principles and Practice of Statistics in Biological Research. 2nd edition (W.H. Freeman and Company: New York). SPURR, A.R. 1969. A low-viscosity epoxy resin embed- ding medium for electron microscopy. Journal of Ultrastructure Research 26: 31-43. WILLENZ, PH. & HARTMAN, W.D. 1985. Calcification rate of Ceratoporella nicholsoni (Porifera: Sclerospongiae): an in situ study with calcein. Pp. 113-118. In Delesalle, B., Galzin, R. & Salvat, B. (eds) Proceedings of the Fifth International Coral Reef Congress, Tahiti, 1985. Volume 5. (Antenne Museum-EPHE: Moorea, French Polynesia). 1989. Micromorphology and ultrastructure of Caribbean sclerosponges. I. Ceratoporella nicholsoni and Stromatospongia norae (Cerato- porellidae: Porifera). Marine Biology 103: .. 387-401. WORHEIDE, G. 1998 The Reef Cave Dwelling Ultraconservative Coralline Demosponge Astrosclera willeyana Lister 1900 from the Indo- Pacific. Micromorphology, Ultrastructure, Biocalcification, Isotope Record, Taxonomy, Biogeography, Phylogeny. Facies 38: 1-88. 686 MEMOIRS OF THE QUEENSLAND MUSEUM A SPONGE THAT CHEATS ON DIFFUSE MUTUALISM AMONG OTHER SPONGE SPECIES. Memoirs of the Queensland Museum 44: 686. 1999:- The demosponge Desmapsamma anchorata is frequently found growing on other organisms, especially gorgonians and other sponges. For paired D. anchorata individuals of the same genotype and initial size, growth rates were lower and mortality rates were higher on carbonate substrata than they were on sponges of other branching species. The three sponge species that served as hosts in the experiments, Jotrochota birotulata, Amphimedon compressa, and Aplysina fulva, grow and survive better when they are intimately intertwined with each other, and do not therefore discourage other sponges from adhering to them. However, D. anchorata does not improve the quality of life for these species when it participates in associations with them. Desmapsamma anchorata grows many times more rapidly than the other species, and appears to accomplish this by skimping on skeletal quality such that it requires skeletal support produced by other organisms in order to withstand physical disturbances. In the early stages of its growth on sponges of other species, D. anchorata does not decrease growth rates of its hosts, but as it continues to grow, it can entirely overwhelm the other sponges, smothering and killing enveloped tissue. The extreme fragility of Desmapsamma anchorata makes it vulnerable to being swept away by physical disturbance, and this prevents it from becoming a chronic hazard for the other sponges. Intimate association with D. anchorata may provide one benefit to other sponge species, which is to facilitate reattachment of loose fragments. Because D. anchorata is able to reattach to carbonate substrata within one day, fragments of other species to which it is attached are anchored for the few additional days that they require in order to establish their own stable attachments to solid substrata. O Porifera, mutualism, parasite, growth, mortality, asexual fragmentation. Janie L. Wulff (email: wulff@jaguar.middlebury.edu), Biology Department, Middlebury College, Middlebury, VT 05753, USA; 1 June 1998. PRODUCTION OF BIOACTIVE FURANOSESTERTERPENE TETRONIC ACIDS AS POSSIBLE INTERNAL CHEMICAL DEFENSE MECHANISM IN THE SPONGE IRCINIA FELIX (PORIFERA: DEMOSPONGIAE) SVEN ZEA, FERNANDO J. PARRA, ALEJANDRO MARTINEZ AND CARMENZA DUQUE Zea, S. Parra, F.J. Martínez, A, & Duque, C, 1999 06 30: Production of bioactive furanosesterrerpene tetronic acids as possible internal chemical defense mechanism in the sponge /reinia felix (Porifera; Demospongiae). Memoirs of the Queensland Museum 44: 687-696, Brisbane, ISSN D079-8835. Marine sponges of the genus /rcinia (Poritera, Demospongiae, [rciniidae) are known to produce several linear furanosesterpene tetronic acids (FTAs) with antimicrobial, cytotoxic and antitumoral properties. freinia felix is a common and abundant sponge from Santa Marta, Colombian Caribbean Sea. containing FTAs in quantities up ta4.5% of its ash-tree dry tissue weight. FTA concentration was quantified by HPLC after organic extraction in individuals of L felix. The following results were obtained: 1) peripheral tissues had greater concentration than internal tissues; 2) total body PTA concentration was inversely and significantly related lo the ambient illumination where individuals lived (in relation to depth. and comparing locations shaded vs. open to light, and localities differing in water turbidity); 3) there was no significant variation in PTA concentration throughout the time of study (June-December 1995); 4) over a 2 month period it was found that experimental shading induced a significant increase in total body F'TA concentration; 5) there was a strong FTA increase (in a scale of 1 week-2 months) when sponges were manipulated in depth transference experiments and when they were purposely injured and; 6) intact or injured individuals did not exude measurable quantities of FTAs into the surrounding medium in laboratory conditions. ‘Together, these results indicate that FTAs have some adaptive value, but probably nat in mediating external ecological interactions, but instead acting as allomonal internal suppressors and/or antibiotics, The shade-dependent production of FTAs suggest that these substances may prevent parasitisation by photosynthetic Aphanocapsa feldmani-type endosymbionts, when the ambient illumination is below their compensation point. Additionally, as the sponge becomes more heterotrophic under lower light levels it may have an increased need for antibiotics in the choanosome to prevent bacterial food from becoming infectious. Finally, during wound healing, increased FTA levels may also act as internal antibiotic protection, O Porifera, furanosesterterpene tetronic acids, Ircinia, chemical internal defense, Caribbean. Sven Zea (email szealalinvemarorg.co) & Fernando J. Parra, Universidad Nacional de Colombia (Departamento de Biologia), Instituto de Investigaciones Marinas y Costeras José Benito Vives de Andreis', INVEMAR, AA 10-16, Santa Marta, Colombia; Alejandro Martinez, Facultad de Quimica Farmecéytica, Universidad de Antioquia, AA 1226, Medellin, Colombia: Carmenza Duque, Departamento de Quimica. Universidad Nacional de Colombia. AA 100690, Bogotá, Colombia; 27 March 1999. The search for new drugs has led to the discovery of a variety of bioactive secondary metabolites in many terrestrial and aquatic organisms (Garson, 1994). In the marine realm, however, little is known about the use that source organisms make of these substances. |t is commonly thought that sessile organisms use bioactive secondary metabolites as signals to communicate with conspecifics, to deter predators of adults (e.g. Bakus et al., 1986; Pawlik et al, 1995; Wulff, 1994; 1995), or propagules (Thompson et al., 1983), to actively or defens- ively compete for space (c.g. Sullivan et al., 1983; Aerts & Van Soest, 1997), or to prevent epibiosis or external damage by roaming browsers (e.g. Thompson, 1985; Walker et al., 1985; Thompson etal., 1987; Davis, 1991). Intra- specific variation in toxicity or in secondary metabolite composition has been documented in several groups of benthic sessile organisms (sec review in Becerro et al., 1995), In sponges there are a few cases in which intraspecific variation of bioactive secondary metabolites, both in type and concentration, has been documented (Thompson et al., 1983, 1985, 1987; Kreuter et al., 1992; Becerro et al., 1995), This variation has heen attributed to individual physiological defensive responses to differential environmental 688 pressures within its habitat range, especially the degree of spatial competitive interactions with neighboring macrobiota, and epibiosis prevention (Thompson et al., 1987; Becerro et al., 1995). Intra-individual (intercellular) variation in the location of bioactive secondary metabolites has been documented in a few sponges. In two cases, bioactive metabolites were found in inclusions within spherulous cells, located mostly near the surface of the sponge or around exhalant canals. These cells appear to disintegrate and release their inclusions, resulting in the exudation of metabolites through the pinacoderm, into the excurrent canals, and then into the boundary layer around the sponge. These metabolites are thought to be used for external defense-offense, but they may also be released into the mesohyl matrix for internal use (Thompson et al., 1983; Uriz et al., 1996b). Species of /rcinia (Demospongiae, Dictyo- ceratida, Irciniidae) are known to produce linear furanosesterterpenes (Cimino et al., 1972a, b; Lumsdon et al., 1992; Urban & Capon, 1992; Capon etal., 1994; Davis & Capon, 1994; Murray et al., 1995). Recently we reported on three Caribbean species of /rcinia (1. felix, I. strobilina and /. campana), containing the novel (7E, 12E, 18R, 20Z)-variabilin as the major (58%-59.8%) furanosesterterpene tetronic acid, followed by a mixture of (8E, 13Z, 18R, 20Z)- strobilinin plus (TE, 13Z, 18R, 20Z)-felixinin (27.1%-28.6%) and a mixture of the new compounds (8Z,13Z,18R,20Z)-strobilinin and (7Z,13Z,18R, 20Z)-felixinin (13.1%-13.9%) (Martínez et al., 1995b, 1997). The greatest concentration of FTAs occur in 7. felix, followed by 7. campana and J. strobilina (Martínez, 1996). These FTAs were also found to occur as branched chain fatty acid esters, a unique combination never reported before in nature (Martínez et al., 1995a). FTAs have been demonstrated to have a variety of pharmacological properties: e.g. antibiotic (Faulkner, 1973; Martínez, 1996), cytotoxic (Martínez, 1996), antimicrobial and antitumoral (Gamboa & Pinzón, 1997), analgesic and anti- inflammatory (Del Valle & Vargas, 1997), calcium transport inhibition (Beveridge et al., 1995). Pawlik et al. (1995), argued that structurally complex secondary metabolites, which are usually present at high concentrations in sponges, can be physiologically expensive to produce, and thus must have an adaptive purpose. To test whether FTAs play an ecological role in /rcinia we studied the intraspecific and intra-individual variation in FTA concentration in Z. felix, under various MEMOIRS OF THE QUEENSLAND MUSEUM natural and experimental conditions. We found an inverse relationship between FTA concentration and ambient illumination, a greater concentration of FTAs in internal tissues, and an induced production of FTAs in wounded sponges. Here, we report on, and interpret these results in terms of internal defense mechanisms. MATERIALS AND METHODS STUDY AREA AND SOURCE ORGANISM. Ircinia felix (Duchassaing & Michelotti, 1864) was collected from rocky shores and mid-depth fringing reefs of Punta de Betín, and adjacent port dock-pilings in the bay of Santa Marta city (11°15°N, 74°13’W), and in rocky shores and fringing reefs of Isla Aguja, further to the NE (11°19°N, 74°12’ W), Colombian Caribbean Sea. Compared to Isla Aguja, Punta de Betin generally has more turbid waters, and is subjected to greater sedimentation loads from an adjacent river, the city sewage outflows and commercial port activities. Reef corals in this locality have also suffered greater mortality, and reefs are amply colonised by sponges (Zea, 1994), especially by species of /rcinia (Parra, 1997). Ircinia felix has been described from this locality in detail by Zea (1987), and reference material is deposited in the collections of INVEMAR, Santa Marta, and the Instituto de Ciencias Naturales, Universidad Nacional de Colombia, Bogota. In the investigated localities this species lives in densities from 4-50 individuals/40m”; it is usually thickly encrusting to cushion- -shaped, occupying areas from about 30-100cm?, and having maximal thickness from about 1- 6cm: surface is conulose, usually clean and free of epibionts, with several interspersed oscules, 3-5mmdiameter, slightly raised by a membranous collar; external color in life varies from shades of maroon and amber in specimens in well illuminated locations, to dirty cream in shaded or deep locations; internal color is cream (Parra, 1997). Below a sand-filled ectosomal reticulation (cortex), there is a layer with dense aggregations of cyanobacteria of the Aphanocapsa feldmani-type, whose pigments are responsible for the color of the sponge (Riitzler, 1990; Vicente, 1990). This species is typical for the genus in being very tough and difficult to cut or tear largely due to a dense reinforcement of spongin fibrils throughout the mesohyl. Species of Ircinia also yield a characteristic sulfur-garlic stench when handled (Bergquist, 1978), releasing several sulfur and cyanide volatiles (Bonilla, 1997). Two morphotypes are readily distinguishable in Santa Marta populations of /. felix: 1) encrusting, CHEMICAL DEFENSE IN /RCINIA FELIX dark amber surface, oscular skin collar dark brown, and,; 2) encrusting to cushion-shaped, maroon surface, oscular skin collar white (Zea, 1987). Only the latter morphotype, which is the most abundant in the study area (Parra, 1997) was used for the chemical ecology studies presented here. EXTRACTION AND QUANTIFICATION OF FTAs. Whole sponges or fragments were immediately frozen upon return to the field laboratory base (INVEMAR, Santa Marta). Frozen material was air-shipped to the Natural Products Laboratory of the Universidad Nacional de Colombia at Bogotá, for chemical analyses. Extraction and quantification of bioactive FTAs was developed by Martínez (1996) and Martínez et al. (1997), and standardised for this study as follows: 5-10g of wet sponge were cut and macerated first in methanol (MeOH) and then in ethyl acetate (EtOAc), each for 15mins, removing the supernatant by filtration after each solvent addition. Each supernatant was separately vacuum-dried at 35?C in a rotatory evaporator, then diluted in EtOAc, mixed, and partitioned repeatedly with H5O to eliminate sea- water salts. The EtOAc fraction was collected, dried with anhydrous sodium sulfate, filtered, vacuum-dried and weighed. Ash-free dry weight of the solid sponge residual was obtained by subtracting from the oven-dry weight (at 115?C) the ash weight obtained after combustion in a muffle furnace at 400°C. To prolong its stability during storage, the extract was then acetylated in a 20ml mixture of acetic anhydride-pyridine (1:1). The acetylated FTAs were then purified by silica-gel column chromatography, dried and stored under nitrogen atmosphere at 0°C until use. Acetylated FTAs were subjected to HPLC for final purification using MeOH-H;O (85: 15) as mobile phase at a flow rate of 1ml min’ and a Capcell Pak C¡g (250x46mm i.d.) column as stationary phase, monitoring at 270nm. Chromatograms typically gave three peaks between 9 and 12mins, whose subfractions were known to contain five different acetylated FTAs (Martinez et al., 1997). Since the largest and latest subfraction contained pure (7E, 12E, 18R, 20Z)-variabilin acetate (henceforth refered to as variabilin), a 10ug ul” solution of this compound previously obtained was used to construct a calibration curve to calculate sample concen- trations. Initial quantifications were done only on variabilin, but as the structure of all FTAs were being elucidated and found to be bioactive (Martinez, 1996), data for the three peaks were 689 pooled for further analyses and calculated as mg FTA g' ash-free dry weight of sponge. FTA CONCENTRATION IN /RCINIA FELIX. Depth and ambient illumination factors. To initially explore if there was variation in natural variabilin concentration in tissues between individuals across various environmental conditions, two specimens of /. felix were collected in June 1995 at each of four depths (5, 10, 15 and 20m), in conditions of open exposure to ambient illumination at the rocky shore and fringing reef of Punta de Betín, and two more at 4-9m depth in the adjacent well-shaded pilings of the Santa Marta port dock. Statistical differences in variabilin concentration between depths and in pilings were tested by one-way ANOVA; variabilin concentration in relation to depth was investigated by regression analysis. For all statistical tests, including those mentioned below, data was tested for homogeneity of variances between treatment combinations (Bartlett test), and for normality of residuals (Kolmogorov- Smirnov test); when suitable, transformations were applied and means and standard errors back- transformed for presentation (Sokal & Rohlf, 1981). Variabilin concentrations in peripheral vs. choanosomal tissues. To compare variabilin and total FTA concentration in peripheral tissues (including pinacoderm and peripheral choanosomal tissues a few mm below the ecto- some) vs. internal tissues (deeper within the choanosome), in open vs. shaded locations, two specimens were collected at 6-7m in the Punta de Betín rocky shore in June 1995 (variabilin only), and two more at 5-6m in the adjacent dock pilings in September 1995 (variabilin and total FTA). Peripheral tissues were dissected upon return to the laboratory and stored and processed separately from choanosomal tissues. Statistical differences in variabilin and FTA concentration between tissues were compared by one-way ANOVA for each habitat separately. Exudation. Two assays were carried out to test whether FTAs are released by undisturbed and wounded /. felix. 1) Nine darkened and aerated aquaria filled with 0.5L of filtered sea-water were set up in the laboratory. Six specimens of /. felix were carefully collected with the substratum at 10m depth in Punta de Betin in October 1995 using hammer and chisel, and each placed in an aquarium. Three specimens were deeply wounded with a razor blade whereas the other three were left undisturbed; the three aquaria 690 without sponges were used as controls. Unwounded specimens were checked for vitality by observing the pumping of water through oscules. After 4 hours, sponges were removed and the water stored in cold (4°C), while it was being vacuum-filtered through RP-8 cartridges to retain organics. Cartridges were then kept frozen in the dark. 2) This assay was carried out in a similar manner to the first, in November 1995, with a single specimen for each treatment. The water was immediately partitioned in EtOAc, the organic fraction dried, put under nitrogen atmosphere and stored frozen in the dark. After shipment to Bogota, cartridges were flushed with MeOH and EtOAc to release organics and the combined extracts dried. Extracts from both experiments were acetylated and quantified as mentioned above. Ambient- and time-related differences, and experimentally induced production. To compare total FTA concentration in localities having different environmental conditions, three specimens of /. felix were collected at each of two depths (10 and 20m) in Punta de Betin and Isla Aguja, in September 1995. Three more specimens were collected at each of the same depths in Punta de Betin in December 1995. Simultaneously, shading and depth transference experiments were carried out at Punta de Betin from September to December 1995 to test for additional production of FTAs. The above- mentioned specimens served as initial and final controls for natural FTA levels. At each of the two depths, three individuals were shaded by a canopy of wire nailed to the coralline bottom and covered by a black soft polyethylene plastic. As controls, two individuals were also covered by a canopy but with transparent plastic. Also, at each depth, a PVC tube frame holding four, 5cm-wide Plexiglas beds was nailed to the bottom. Four specimens at each depth were carefully collected together with the substratum using hammer and chisel, and fixed tightly on each bed with plastic cable ties; in each frame, two specimens came from the same depth as controls for manipulative experiments, and two specimens were transferred from the reciprocal depth. All sponges were collected simultaneously in December 1995, frozen and processed. Natural, total FTA levels for two localities, two times (initial and final), and under shading and transference treatments and controls were statist- ically compared against depth in a two-way ANOVA (type III sums of squares), and a Tukey multiple comparisons procedure (separately for MEMOIRS OF THE QUEENSLAND MUSEUM each depth), both appropriate for unbalanced designs (SAS Institute Inc., 1988). Time changes were tested, comparing in a separated one-way ANOVA, samples from Punta de Betin collected in June (see above), September and December. Wound-induced production. To test for FTA production in tissues after wounding, six speci- mens were located and tagged at 10m depth from Punta de Betín in November 1995. Initially, a wound was produced by cutting free from the edge a 2-3cm wide fragment from each specimen, which was immediately frozen to measure initial, natural FTA levels. A similarly-sized fragment was again taken, cutting parallel to the initial wound, from each of three specimens seven days later, and from each of the other three specimens 14 days later; these were frozen to measure FTA concentration changes in the wound area in the intervening period as aresult of the initial wound. Initial and final FTA concentrations were compared for each time interval set by a t-student test. FTA changes for individual sponges were compared (null hypothesis of no change) by a paired t-student test, separately for each time interval. RESULTS NATURAL FTA LEVELS. Total body content of FTAs in /. felix ranged from about 1-46mg g' of ash-free dry tissue weight (Table 1; 0.1%-4.6% by weight). Significant variation in total FTA and/or variabilin (58-6096 of total) content in samples was found as follows. Ambient illumination factor. There was a significant increase in variabilin concentration in tissues of /. felix with increased depth at the fringing reef of Punta de Betín in June 1995 (Fig. 1). This trend followed a potential regression model [variabilin] = 0.0453 Z'*, R°=0.95, 0.01«P«0.025 (see Martínez, 1996). Since light intensity decreases exponentially with depth in sea-water, ambient illumination (not measured) was assumed to at least relate partly to variabilin concentration. Deviations from regression, however, accounted for a significant part of the model (R?=0.04, 0.025 g E Ta 40 E 40 2 2 30 c i 30 BC É 20 E 20 a 3 E 10 B E 10 0 0 Transp. Dark Transp. Dark Shading treatment Shading treatment 50 D G = 50 c c PES CD To 40 m 2 E 30 E so <, , o- o... . 50,360 II SOLS. zoe Renee MEAS 45 DEEZOYÍO: Lese rete e Ee de thee 163 Bermuda .. oo... o... ..-«...- 360 Biemna tubulata .. 4... osos 180 Biemna ,... s. n^ ad abt oc dide a d cde 259 bioactive meraboliles , 222.022. 167, 569 biodiversity survey... o... .. 175,249, 263, 361, 508 biocroding sponges... o... 2.202... 533 binerosion, o . . 540,627 biogeography, > -o coo. ee (54,175,249, 263,650 biological activity (natural products). >.. 161,306, 541 biological testing. o oe ee 282 D MD 422 biomimeralisation, 5... 2.2222... 76, 550 bimstrames 4d breue d Rea Ren 99 biosynthetic pathways... -3 ....... e- 560, 583 bio-toxicity oo 5M bishomoscalaranes. o... ooo... 57,61 Blastochgetetes pealiiki 16 body plan .-... RA AAA A 27 body shape 22, 2220 ooo... 456,607 boring sponges. o 77 Biachiolitas-- o. iio oe e omie o ea ae 467, 469 Brazilian coast... 2.2... ee - , - 299, 369 British Columbia... .....o.o.o.ooo.o.o. 44 brominated indoles. |... 0.0 0.0000... 162 bromopyrroles ,.... 22.2 rs 286 bromotyrosines. «52s ne 581 Cacospongia mollior... .. 0... 2. C. - 486 Cacospongia scalaris ooo... EIA 456 Calcarea, .12t100itori data Pd SIO CRIGET Meme rotor ecran miae AUR EH calcification regulan... 666 calcification ... 222.22 ooo o... u 297 cgieimicrobe 5... 222 1... . rns 298 Calcispongiae (Calcarea) -. oig eo. .... 44 DIIRE : ¿ario be eee oe Pol ii AA 1 76 calcium-binding proteins... o... ooo... 76 A AR E 499 Callyspongia diffusa... 2s. vss s 248, 258 - Callyspongia mollis ocios oo. oo 8 Callyspongia mauricio ooo 667 Callyspongia ramosa. cc 180 Cailyspongla ridleyi <, 22202 ee, 258 Callyspongia vaginis. «ne 457 Callyspongia. o.. E e 132,179, 258,445, 669 Callysponglidae ... 22 ee ee ee MM A etn 603 Calycosóma -Se css ee ee 603 Candidaspongia jlabelleta . 00 ee SY MEMOIRS OF THE QUEENSLAND MUSEUM Candidaspongia. 222.2. s... 58 carbon isotope history o... oo... oo... 91 carbon metabolism. -.....-.- CUm Rie 06 Caribbean reef ecosystems... llus ,92 Carm iii be dy hig ed EA AA 225 camivoroussponges .- sss ee 1,27, 289 carotenoids. ooo a eee 572 carrier telt, A 44 Carteriospongia flabellifera ..--...... , 593,669 Carteriospongiafoliascens |... sss 61,259. 669 Carteriospóhgid . a.s 0.42222 ee. A Lasearg. succ ci e de o 418, 464 Caulophacella |... 2s oe cres 603 Caylophacidae |, 5.0. 0.2... ee, . 603 Caulophaeus ooo. $e s. 603 Caulospongia amplexa... oe. ee 177 Caulospongia plicata... 2 ee 180 CARVESyo Ve uua e Ta oe eee xp > 360) ODA yg der ad ad re 184 cell adhesion ...... EA E 184 cell odltür& e mi ame o epee ber aee we Se ZAR io IE eee. ee 4 A A A a 91, 160 Ceratódietyónspongiosum ..--... 124,204, 602, 606 Ceratoporellanicholsont s., s.s. 84,91, 492,675 Ceratapsion aurdhlided. , 2. ee ek 495 GeratopsiD: (ai l2-..: doas .4-BA 258 Cercidochela lankesteri ooo.» 518 ¡E ii e A ee ts shee a ae 16 chaelitids -.. 1.1.0.1... .mo.saiod E AA, ett hin os ae 132 ChürülüBemüs NA eee 613 chemical defense. .-.......... 92, 426,438, 687 chemicalecology ~.. -a-o -a . 161,214,411,590 chemia e. ede be hea Sb batten 569 chemotaxonomy 2... 260s irl o^ ay tr ote one Chinese Guizhouspongedepasity - ,.-..-..2,18,20 AS AAA mens 342 choanocyte chambers. <. oaceae 289 Chondrilla australiensis - peor. 177,185,257 Chondrillanucula o., 0 ee ee 257, 309, 593 Chondrilla, fe +e fe yg tt ba bee: 257, 593 Chondrocladia gigantea. oe 289 Chondracladià. oa 2 2 o el 29 Chondrosiareniformis |... oe es 85,550,558, 501 Chondhusig. ub bs e i bdo e isle S RA e A e L, a 485 Cinachyra australiensis ooo. 257 CARAC -10 E. eve lia uam t 257 Cinachvrellaulloclada- ooo... 299,317, 597 Cinachyrella apion 26 co. A die ea 299, 597 Cinaehyrella, ita 2.222222 ee 257, 299 Ciocalypta fenestratus o o oaoa ee 259 Ciocalypia .-... a a Jae E, fn 259 Cladorhisa; 221.102.201 2.. Eeg Cladorhizidae. 2. 2 o peole eid ihoa eden 289 SUBJECT INDEX Clathria abietina ©... oa naag 2 eee 178. 258 Clathria abrolhosensis. 2... 224. ss. R0 Clathria aphylla |... 2 ss e sn 180 Clathria australiensis a o.oo aoea een 180 Clathria cactifarmis ; ooo. xv eo. ==. 177 Clathria cancellari «22e sn 180 Clathria cervicornis io... eli 258 Clathria coralliophila |... 22.222... 258 Clathria eccentried. ooo. 259 Clathriagrisea.. oe ee 180 Clathria lendenfeldi ooo ee ee es 258 Clathria patula... ptt att. ott? 4 , 180 Clathria pyramida- oo oaa we 125, 180 Clathria reinwardti |... aaau 36 Clathria selaċhit. o... j.- B: 177 Clathria striata, iio eese e ad e s ne 495 Clathria sivloprothesis. os 0 0222... 0... MBO Clathria tingens ooo ee ee oo. 258 Clathria vulpina >.. ooog s 258 Glammria= 3 PPE E og Qe Ee, 258 Clathrina asvandroides o... RAP PR 506 Clathrina aspina . 22s 96 Clathrinaaured ooo cocos S96 Clalhrina brasiliensis -ooa sell 506 Clathrina éerebrum. ooo. 36, 596 Clathrina clathrus «2... o... co. ta ols elke 596 Claihrina cylindractina o. coso 596 Clathrina primordialis |... ooo. +. - 309,596 CURR D- o E A A eM 31 Clathrocoilona. .. osse e ea 99 Claviscopulia furcillata «c os een 626 cleavage(larvà) o 22. osos ess 160 climate change records- -3 o... 222... 658 CUBRA lampa a. 2o ig de MEC 360 Cliona nigricans «s ce mh 77, 597 Cliona viridis |... 2 oso... . 536,597 Cuang. e eit. chipped sg segets 540, 669 clionamides. | 22 ooo. ee DTS Cliothasa cf. hancocki |... 2 22s les 536 CIRWRHOSU aps capado ote ede oE e RR 540, 639 Cnemidiasiruim ooo. s.m... ll 467 CO» (history). 14 bei sez 524 Cuelocarteria singaporensis, ooo... sss 259 Caelasphaera. |... ee e 185 Coenostróma. , . c ee s 99 Collospongia auris... .. oo... : 594. 668 colonial reel-building sphinetozoans. +... 498 eolonisalion. ooo ee, 360 commensalism . . . 2.2.2 sens 427 community distribution. --.. o... ee 148 commubity ecology... ooo... ns 45 community structure +... 288 competition... ooo. roo. ra medirase 161,457 Composataly oo ee 613 A IA AAA 455 conduction o... o... ... --. M2 699 congruence. 6... 5: . ttt 445 4:5 eae uS d 274 contractile filaments ..., 102003000... 398 Coppatiidae. |... 525 coral reeF communities... o... 0... ...- 360 coral reefinjury andrestoralion, .... 0... 532 coralreefsponges 20... 2 ee ee n 667, 674 ceoralreefs. |... ee uie - 45. 307,411,457 CoralSea ........ ooo... 263, 462,498, 540 coral toxicity, -e oo... o... .- . 307 coralline spoñges- - . - , 84,91, 462,492, 498,675 Corallistes typus... ooo... a afte: a 345 Gorallisigs.i i2. t+ be pu iile: 343, 469 Corallistidae |.. oaoa ee en 274 ERIE ae Ie A ce eei te M ly, 411 Corticiumcandelabrum o... 178, 400, 594 Corticium cf. simplex. ©. o sss TPS AY 178 Mi 22. ee bard r egn dae 333 Corvoheteromeyenia heteroselera |... ss. 644 Corvomevenia thui ooo... - 644 CoryRahemizv o. 756653 st. eae: : i4 BID Coscinoderma malhewsl. 0. 2 len 259 Coscinoderma pesleonis. ooo... - 180 Caseimonema. occ ee oe 612 Cozumel, Mexito. n. o . e... eaa 508 Crambe crambe coco... 616 Crateromorpha s es s 604 Cratiomlaria , 2.222 222 zio 222.2: . 4687 Crawling (larva), noaa 2222s 9] Crella incrustans tuelet o. o ee 177 Crélla spimnulata «.. ee et 177 Crella. ccs ARN NULLI CR S 259 Cribrochalina vasculum . ooo. coo SIS Cribrachalina... .... ee ee ee. 136,518 cross-shel distributio... o... o... 151 cryptic species ooo ee ee 239 gulis: hamm els ed pe eles toto ee shee di 617 cyanide; ; tty es beet tae . 561 cyanobacteria, |... sss. 154, 167, 238, 493, 667 cyclic and linear peptides es Me Oe. n 572.574 cyclic depsipeptides. ... o... 17), $25 cycliediterpenes -. o... ee eee sen a 576 cyclopropene sterols a. o,a aano sss 579 AV A A M Svlindrophyma coo ee 466 CymbastelaconcentPiea® . o.an aoaaa 359, 483, 495 Cymbastela coralliophila ......... iss. 259 Cymbastela hooperi .....3 Phys 01449 492 Cymbastela marshae |... ,- $33 1.23 see Cymbastela cf. vespertina -nonpa naana 178 Cymbaslel. , u adinin ee ey ees 669 Cypellospongia, «sea 463 cytology ¿ocio cocido AA 399 cytotoxic alkaloids . .....,-- . 205 cytotoxicity. . . -a 22s 282,342, 438, 525,541 D3domain288rDNA ... 02.0... ooo... 43 Ductylaspongia elegans... ee o. 4 259 700 DUABIRD a eiusd by dos ada ert ares 132 Dactylocalys |... 222222 sae s 464 Darwinellidac, |... eee 353 Dasychalimg see ee i ea 136 Weed" AS 289 defining characters, <.. 222 nn 27 Ehlen: sq. sete kb be batting. 627 Demospongiae. , 33,63, 100, 131, 175,185,297, 299, 361, 369, 410, 533,617, 627 Dendrilla cirsioide$ io. yo... -- 354 A dace kt AA 179 Dendrnceratida: 55 222.2222 biii 353 densitie&. ooo ee eR ren 45 depth zonation. . . ooo... ooo... o... 132, 307 demande, i. 5.45, A ferc itl. 17] AI AA 333 Desmapsamma anchorata .. ci, oa 686 Desmapsamrma oc 258 development... o... ..-. 306, 509, 568 dichloroimines <, -oo ee en 561 Dierunaclonella oi ee 469 Dieryaults elegans |... oe ns 507 Dictyoceratida, o... 2.2225 57,63, 238,551 Dictyodendrillidag |... 2... ooo... . 353 Üigestion A A a pee ce ee eor 262 digestive phagasomes , .. 0... 1... .. V. 262 diisucyanoudociune. ...... o... . mm... 282 diketopiperazine . . -o 2.2 22252-0222. . 580 dinoflagellates +. -n Jtt eea ee n 205 Discodermia ef. laevidiseus «2 2e 432 Discodermia polydiseus . 2 2. ee s. - ADO Discodermia .. 2 el 329, 343, 345 iege 201 e trios se nr Repone iba d e G74 dispersal . 2.2... mOS nts ::: sed O49 distribution pattems... sss. 288,307,368, 602 disturbance. +07... 0.2.2.2 -0- 1360 diterpeneisonitriles. ..-.. rer 282 diterpene quinones, ooo 1.0... pom... STD DNA .:1*»i:85.:.502:2.:: bg 5: 5155 BR Dorypl&res. see oi e Je oema i efe 167, 257, 525 Dosilia pydanieli . . ooo. oes 646 Dragmaiella i g ia padma ogg as a nog - 178 Druinella purpurea, o: o o ii siose es 260 dynamies i6 j irat i i vi Dysidea dakini. 2. aa e 180 Dysideaherbaceà , |... 0. oe. 167,238, 260, 669 idea: E spt t io egi iua 179, 260, 669 FEBRONI usi Louer ce betae do nem bd 214 Echinachalina tubulosa >a 2222222 cols 259 Echinodermata... 0 ee ooo... os 125 Echinodictyum clathridides ooo. ss. 177, 185 Echinadictyum mesenterinum . . - eh: , 39 Echinodictyum nitulus . . ossaa aiaga eaa 180. ecology 4. . » 93,101, 160,426,438,493 o osu i Dank. PA a oic 643 Eemadokyt . s uf seme eia ete 17% MEMOIRS OF THE QUEENSLAND MUSEUM Ectyoplasiaferox ,.. 2e els 309, 317 electron microscopy ++... o... o ae 193 electrophysiology, ©... 2... ees 342 embryology... ee ee 160, 627 AS A XX * e 616 Endectyon éelyakoyi. -a 02 ese ses 259 endemism , 2.2, a ey se 263, 361 endocytosiS. «ee 262 endolithic sponges. ©. oo... see 77 envirommentalecology . .. +... 174,368,674 environmental heal... o... o... ... 50 Ephydatia fluviatilis ooo ee ee os 215,398, 509 Ephydatia muelleri . . 0.022 cL 215,275,509 3 ce «eee a s orm m mn 653 epiobiotie fauna... ee 8! Epipolasis i 0 ee 432 Erylus allen... es 372 Erylus amarphus 0.2 ze 429 Erylus-Burtotl «eeu ide esten bep e nm 429 Eryluscartetl i pap cia tese te iet 429 Eryluscorneus oc 375 Erylus diminati... scs 371 ErylusfOrmosus ios coo 309, 375 Erylus.geodioides. e 0 s ee 429 Erylus lendenfeldi |... 2022 2.2 os... MIT Erylus nigra. ee 429, 432 ETVE ONISE A oe i o r ee ee ee 369 Erylus proximus. 2 0 zs eas 180, 429 Erylustapsenli ociosa cla s 369 EFBARS A AA i BO ee 369, 432 Esperiopsis desaphiórd. 2. oe conos. 30 Esperiopsis informis «sse o. ss 518 ESPErapsis. Lo do (dieta eio t4 BS ghi. ner dor iss me iii 77 Ethiopian region . o 22222 361 Euchelipluma |... ee ee es 29 Euhyalonema. e a ~ irb aew si wia b ee td 613 Bu-Mefazoa..s. 042. - 86d E 381 Eunupius carteri . ee 215 Eundpius fragilis ooo. vill 215, 509 Euplacella ooo 132 Buplectallà A cR e e ot bà 612 Eurétldad T.t. ote ge e Ra Re o LCS . 626 evolution... 2.22 ss ss 27.33, 154.343, 381,659 excavating sponges. 1... ee es 627 exploration and exploitation 2. 0. ke, 590 expression ... ooo 509 external surface. <- . o -ie a p ......-. 617 farming method, o -.-.-.-5,-+-3.---;3 155 Porrgtibted' dato d bee 3d oe s 626 Fascaplysinapsis reticulata oo coo. . 260 Fascaplysinópsis ooo 260 fatty acidsand derived lipids -. -. -2 ........ 571 faunasurvey ....,..-.-, 93,249, 263, 249, 307,439 fauna—Atelia. .. ek, y - 160,262 SUBJECT INDEX 701 fauna—Catibbean , . . 84,91,92, 160,239, 307,329, 345, 360, 426, 438. 457,473, 659, 686, 687 fauna—European (limnotic), -. 20... o... 215 faunà—European Late Jurassic... ........ 297 fauna— Indian region, -. -o ... ooo... . 439 fauna Indonesia... ... 0... 0... 147, 477 fauna—Indo-west Pacific... ...,., 0... 174,650 fauna—tranian Middle Cambrian , -.--......298 fauna—Mediterranean . . . 77,85, 101,353, 399,485,617 fauna—N American (limnotic) .............% fauna—N African (limnotic). . .. -..,.., +. 361 fauna—NE Allanic. ...... ee ee 101,627 fauna—NE Australia 57, 63, 124, 161, 193, 204, 205, 238, 249, 263, 281, 479, 602, 606,667 fauna—NE Pacilic ... ..- . 93,288,342, 499,508, 627 fauna- New Zealand .,,.....-,.-. 76,155,342 fauna NW Pacific... 1.11... o... 345, 541 tauna—Queensland Middle Devonian -........ 99 fauna—Red Sea ...... ee ee 248 fauna SAmerican(limnotic) .......--- 643,651 launa—SE Asia s.a 2... eso sss SS fauna—SE Australia |... ss 125,288,493 fauna—Siberian (limmotic), .., ... - - . 275,368,651 fauna—SW Atlantic... 2. 214,299, 306, 317, 369 fauna—SW Australia . . o... ee 175, 185 fauna—SW Pacific. . 5... 1... «00... .. 498 feedingbiology .......<0 ooo... 51,262,457 feedingdeterrency 2. ee, 205, 360, 541 leedingecology. oo ooo 214 Eli. lass LE. se Idee des 688 Ferestromáloperü . ssa cl cs rns 99 A AAA aces cbe e Ee E A o 44 Beld'guide . hace pe bee bit: pi DE filamentous eubacteria |... 2. 0... 167 Bnestructur& i. aaa atte A 6l7 fishfeeding ,-..-..---- -. 4. 92,207,543 flooding impact... eee 215 Florida Keys National Marine Sanctuary +... a.. 532 Fullicdtdn de >- : : secto 1 acetato ira O food-poor environment, |... ee ee ee 289 Poreepia biceps... oo ee 177 forsipgn matter. 2. ee ee p eodera 85 fossil SPONBES +... ze te 418,515 fossilisation. coo 2222222 coss ss. 297 fragmentation. n o u 2222s 602 freshwater ephemeral habitats , ........0..2 643 freshwater sponges . . . . 93,215,275, 368, 398, 524,651 furanosesterterpene tetronic acids . oo opaa ana 687 furano-terpenes. 52.222222 cl oss 579 Gastrophanelia, o aooaa Y ENS cc a e ee nce RR ye he MA Gelliodes cf abusu... 2c 178 Gelliodes pumilus .. csse es 258 MSAD . et E bi epe ot ila, 136, 258 Guille Bi: 3-34 7s ye fan rm. ML 178, 258 gommule. sss eoe 5 324 gene divergence 5... ee 59] gene sequencing... 2.) . ee ee ee eee 329 generic revision. -onago sess 131 genetic diversity 2 ee 317 o AO EA epst: 76 @eMusMOV, ... 2.2 ee 57,361,499, 503 Geadiacydonium , coo oe ee +. 31,381,509 Geodia gibberosa o., poo ganran naan: 160 Geodia cf. parasitica... ee 432 Geodidae. |... ee ee ee ee 360 géographie distribution. , , .. 2... +. .+ 361 germination, |... o... o... oo... 2... .. 524 Girvanella |... 22222222 iii sss s 298 Ehem s. irem. tuu eben Pn in 617 glycosphingolipids. ©... ee 285 graff rejection... coo esee 154 NAAA AAA 225 Gravestockia pharetronensiy -- 020 222222 17 AAA A ce eee a 248 Bravingtcfugia. o... 125 Great BarrierReef. 2 , .-- , . - 84,249, 263,282,540 Piwi OTAS 322 t) ur pe HERE eem s 77 growth layers. .,.....2 222220223 ek 658 growth monitoring ~- -a o-oo oea ons 419 prowthrate. o ee a 156,422,675 BOW wg t o pee Die t eng: 50,479, 686 guanidine-IMidazoles. |... 2... lee 582 Guetiardiscyphia «os s sess 469 CTHilUrP eei iaa i-o os iil 31, 178 habitat specialisation»... 2, 0. o... 288. 307 hadromerid subordinal classification. ......... 100 Hadromerida. .. o... olio o.c... 100 Haliehondria japonica. coo corsa 275 Halichondrlapanicea . oo coco... 100, 262. 288 Halichondria phakelliodes a.nn agaaa 177 Halichondria stalagmites 2.2.0. ee 259 Halichondria 00 0... ee ee, . 289,323 Halichondrida ,..,....:.. 2... eee 100 Haliclona amboinensis: «22.2... 2.2. 77 Halielona camerata ooo es 257 Haliclona clathrata, oo 2 e 257 Halielonacoérulà |... ee 309 Haliclona cymaeformis [eymiformis] . 124, 177, 204, 258, 602, 606 Haliclona miner oe e 258 Haliclona pigmentifera. ... 22.2 ll . 257 Haliclona tenuispiculata oso 2o... 258 Haliclona laxius isses 258 Haliclona cf 1esotes ooo c. s. M8 Halielana- oe 178,185,205. 214.257 haliclonacyclamines -.... oo... ....... 205 Haliclonissa - occ. eros e 136 Halisarca dujardini ois ss. 160 Maplosclerida . ooo... ee et 131,517 hard substrata, o ee ee 288 Hemigellius. ..... cc eee 136 702 Flermatostroma |... 04er wede 99 Hertwigiafaleifera. ooo sse 507 heterozygosity 2... ee rs e SO Hexactinellida 33, 44, 297, 342, 410,418, 499, 603, 607, 626 Hexactinoderma |... oe ee ee 463 Hgxactinoósu. ... ee ee 463 Hexasterophora, +... ee ee 603 high alkalinity --.-..ooo0. 0... 20... «477 Hippospongia communis. oc 486 Hippospongia. .. oe s 259 histochemistty +... ooo... o... ens 659 histocompatibilily. |... o... . 2... 184, 602 historical review .-.. ooo... ooo coros. l history of classification... o a 9 Holascus bélvaevi o onou ee ee 506 Holééehé o ms cee sse ruta rh epim 651 Holopsamma arborea... 2s ss 180, 495 Halopsumma crassa... ess 180 Holopsamma faYus . e 177 Holoxea furtiva. ooo ee ee 432 homeobox gents. | 2. o... os... - 306, 509 hemoplasy .... eee n n 626 homoscalaranes . coo 222 22 ee a 61 Homosclerophorida |... ..... ro... ee, 399 horny sponge... 6 oa o 551 Houtman Abrolhos. |... saag ee eee eg 175 Hwlaktüs pti bib tat tig l 603 Hyalonema cretacead coso cc 418 Hyalonema ni3 A uii batane ndi] -- , 607 Hyalonematidae ..........o ee 607 Hyalosinicd. ooo solle 418 Hyatella intestinalis |... sese 180 Hymeniacidon heliophila |... . oo 100, 317 Hymeniacidon . 2... ee s e ne 259 hypercalcified basal skeletod.......-.. ee 9 lanthellabasta ooo... e... 260, 483 lanthella flabelliformis. «ees 260 Iberian Peninsula. oo... ooo mo... 101 (A AS Bete ad 174 Igernella mirabilis |... 2 cese 354 Igernella noiabilis |... 2222s 354 immunoeytes. 2... 2e s e aen 248 immuno-gold technique... es 248 immuno-histuchemical technique -.,..-¿.-.. 248 immunology ooo 248 impact... erai eee ee o. 255 insite sequencing . 6... ee n 558 in vito culture. a oce 329 incorporated foreignmatenal .. 2... 2. - 85,533,551 incorporation... 4. sse 77 indi- c LAM. ovile Y ronm: To de 439 indicators... os trn 50 indoles Coa. Sicha’ see A ARA 161, 581 ÁS ¡ir caro bbe arts By 147, 477 inteltidal ui. oei e iHe CA 262, 288 intra-reefal distribution. 1.0... 0... 153 MEMOIRS OF THE QUEENSLAND MUSEUM invertebrate immunity... 2... 2. ee ee 184 lotrochota acerdta ooo cor 177 lotrochota baculifera ooo... 177, 306 lotrachota birotulata .. , co 309,686 Totrochota foveolarid, .. 2s es 258 fotrachola .. 22. 2... . 160, 178, 258 IAS Lupe O A 298 dreinia campana s; jatomi e e 687 NAAA eds e catat s 687 drvinia rantosa, oo ee 259, 668 Ircinia spinosula ooo 456 Ircinia strobilina o... ee 309, 687 Ircinia variabilis o. 2.22 22.22. ii, . ABÓ 2 "4 E A E CN ve 259, 669, 687 isocyanidés. o o orro S6l isocyanolerpenes «4 re 576 lsodictya multiformis. . o.o paaa 518 Isodictya palmata .. 2.222 ill. 518 Iddi o os te te eh tbe ep a A ST? isoquinolinoquinones. +... +... .<., aooo ay eroa 575 Petrobiona: i. 222 ned s pi Bargi di. et dd PRONE ie ot e Jee A DO 306 Pétrasaspongia mycafijiensis ooo ooo... 61 MEMOIRS OF THE QUEENSLAND MUSEUM Petrosia cf. cancellata |... coca 179 Petrosiaclavatg , 222 antaa zl 594 Petrostaficiformis o... nna nauan 82, 486,550, 594 RAROS. T. remorum eror a offs 258 Phakellia arwensis «see nne 669 Phakellia cavernosa s.s sse ee 259 AS A 607 phenotypic plasticity. ooo... o... ... 239 Phéronema.. a ee e a a 607 Pheronematidae +... «so... 2... 607 Phialonema, coosccosniona hn 612 Philippines ,... 222 enn 45, 84. 411 Phorbas fictitioidés, |... ccoo... 177 Phoriospongia cf. Kirk... ee 495 Phor ROB E. sot lnBb-v gag e Lrepe 178 phospholipids, |... cess iren 581 photoadaptation i- a soca asao an 606 photosynthate ©. -o 222 ee 1... 204, 238 photosynthesis... o... a 524, 606 phototrophy. .. oo... ooo... +... 147 Phycopsis. ooo ee eee 178 Phyllospongia alcitormis. «ees 594 Phyllospangia lamellosa ... .... 22i. 594, 669 Phyllospongia papyracea .. 2. een 669 Phyllospongit ooo... mm 60 Phyllospongiinae ..2 3.9.0... 57 phylogenetic reconstruction... 2... 22s sss 225, 607 phylogenetics, |... a sso eas 626 phylogeny 43, 225, 274, 275, 352, 381. 399, 410, 418, 517, 550, 558, 650 physicáldéfen$es ..... 2. e.c s 9E pinacocyéles;4 od rra edi id 398 Plakina endoumenyis.. 0 2222222. 400 Plafinajaom........ cese as 400 Plakina monolopha .. 2s sls 400, 596 Plakina trilopha. . ee ee s . 400,596 Plakinalopha mirabilis. |. oe ro... 274 Plakinastrella mammillaris o. os. 222i. 177 Plakinastrella minor |... o.a, ooo... Q7 Plakortis simplex... . seen 309 Plakortis ooo rms momo. cr. 178 317,401 planktoniclava e H: i gid o ioh ee .. 027 PBN bate ele Bener br 608, 613 Plamlisitut 22 ts pe ba bade ea tid 610 O A rmm CY de $51 Pleraplysilla reticulata. ooo... ee ee 355 Fleraplysiiiaes). tt AL Esaa bet Data 353 IA A A pee neat tA 469 Poecilosglerida 20... 0.0.0 2.24 101, 517 Poliopogon amadou . . 2l 500 Poliopógon müilal ooo 500 Poliopogon mendocina. , ... o 5 ee 500 Poliopogon .. . 2.2 ee 499, 610 polybrominated biphenyl ethers |... sss 169 polychlorinated amino aids... 2 eee - 169 polycyclic guanidine alkaloids, ..-......... 577 SUBJECT INDEX 705 pelyetherg. |... us esee ee tee eE a 577 polyhydroxylated sterols... 2-2... 2-0... 580 pülyketid$. tio. 9. q2* e hor yas 574 Polymastia croceus. «s s 51, 156 Polymastig janeirensis ooo 306 Polvmastia megasclera, ooo... 257 DA se. Sethe O et SE 257 population dynatnics, +... e... .. ee 674 population structure 2. ees 239 Porifera genes o.oo aoua s is 184 PostPaleozoic history- ooo o... 463 Precambrian’... o... momo... o... 17,27 predation ...oo.o coco. o. 92, 160, 207, 288, 426, 541 predator/prey interaction sss . 92, 288 Prionema 0... pete By arit 2612 Prachlorovoccus bacteria... 2 es §3, 457 Proeuplectella |... ee 453 projected body areà. coco 419 Propachastrella |... ee 463 proteoglycans, , 2... 205% fos tf. ess IBA Psammocinia jejuensis ooo. i 58 Psammocinia mammiformis sn 551 Psammocinia mosulpia, oo. com 581 Psammocinia samyangensis . o. 2.5 0-2. cs. 55 Psammocinia wandoensis . 0... 2s 55] Psammocinia . 32411 - 179, 260, 551 Pyeudaxinella mnes I X US i elke 8 ten ye 259 Pseudaxinella reticulata 2. ls. 306, 317 Pseudoglteromonas ...,...---,--------72 Pseudoceratina crassa... ss Cid trios: 193 Pseudoceralina «<. cocos... 179, 260 Pseudocorticium jarret. ooo... coo. 400 Pseiudosuberites andrewsi... o oe iss 257, 419 Pseudotrupetostroma ooo O pteridine rd ips ido o va. 582 PIEPÓREMO es sts aga e ere ea ae 612 Ptilocaulis spiculifera o ooo... ... 438 pumping ~ eee o SL DRAKMO trees a pamu aue Ae 286 pyridoacridine alkaloids... .. ... aoa ++. 171,541 pyritisation, 6. ooo’ eroaa e 150 pyrrole-2-carboxylic derivatives... 2212.21 - 576 pyrroloquinoline alkaloids. .......o....... 575 Quadrolaminiella. |... aoaaa aa lll 418 Racodiscula sceptrellifera |... .. llis 429 radiocarbon dating. |. a -o ee o... 34 Radiospongilla amazonensis, ooo... 644 Radiospongilla cerebellata. -anana oo. 324 Rankenella |... 22222 222r. 298 RAPD, Pad A PME RERO Bil Raphidotethya errieination AENA XN 257 Raphidotethva - 2. 222222 oll 22i li. 257 Raspailia reticulata «oos 259 Web (Strid? AA fe uu lcm s e Ire 306 ps oL olv et TT siitE- 381 recovery from physical injury. o... ooo... 332 teapveh. c.a oa olor reme ne jer rs p e 455 recruitment- >. 22. 2 ee 288, 479, 616 Poets ive o SIE Be 2... es et ta T3 298, 508 reef-building sphinctozoans, . .... dsl tact 4 462, 492 reef-building sponges -o 0200 222g coo. 515. Regudrella o ee 418,463,613 regénerütlon . oos o oa 306, 568, 675 Wad S Fa, 5a Na © Bag, 178, 257 Reniochalinu stalagmitis, o. niem AAA 259 reproduction... 2.2 185,248,602,617 reproductive output, 2. -c-a o | 616 reproductive liming. ©... ee eee 190 fesllictice. e ee e RR 455 respiration. a eaeoe enaena onea es. 238,606 resting Stage o... os... cao wet g UBI DESEE i a etn eed eA ios i alaa isi 2009 VEN CIERRE MCN ET Den on Rhabdastrellarowi. |... oe en 180 Rhabdocalyptusdawsoni ..... 0... . 44,342,391 Rhabdocalypius ooo... 342 Rhaphidoteca, |. 0222 sso co... o.. 230 Rhizomorine lithistids |... 5... 22s <<. 473 Rhogostomium ooo . 464 Rhopaloeides odorabile oo... - - 63,669 RiverRhine. ....... dot DON a v 215 ROY a age oo lara Pi scm d o A ATA running water. |... Jeb n i oe e n in i i 215 sanddünés . .. e 643 SOJA ucl o op ma IIT rm t eet o oe 573 Savanna ifi te ER 643 seqlarenes: 2 o o pe ha eiia sti... vb s ssl scanning electron microscopy. . 2.2... 02. 361,617 Sehaudinnia . 0 . -eose ee 604 Sahiha, e aa pae a ee 610 Scleritoderma, oo... «n As ho E 473 sclerosponges. 2. s e sees 07S Scopalina laphyropoda. . 22 lel. 616 RBawaler o e e ee 058 secondary metabolites ...-.-. 205, 438,541, 561,590 secondary structure +2... 22222. -- - 43 sedimentstudies -.. o... ooo ooo... 368 selectivity |... a cc oc TI Semperella... o 0 ee ooo s ls 607 Sgricdlóphus*.. «lieu al ie e a a a 610 Serolónil E OE 2635 659 sesquiterpene quinones, o oo... ses 572 sesquiterpenes ooo 222. oss 169,577, 580 sessile fauna, o.. ee ee 455,479 SestertérpeneS -me eee 57, 6l Setidiumoblectum |... oaan lel. 473 Selisliint. 53.3.14 4 aee apes ce dab tm e alm 466, 473 settlement. 22202222 cz22 zoo ns 606 sexual reproduction. . oo... +... 627 shallow- subtidal «......... ses 44, 288 Shape changes. a os ea ee ee ee ee 5 Sigmoscepirella fibrosa |... ss. AA Y 706 Silesiaspongia rimosa ooo . 464 silica production . .... oo... o... 48, 85,87, 92 Siliquariaspongia japonica. ooo. oe sas 429 siliquariid molluscs. s «22er 427 Siphonochalina, sapaan aagana nbbepii 132 size-frequeficy ooo... no... V. 0602 skeletalarchitecture ....-..-.<.... 410,666,675 Skeletonema o p e iage ee ee 72 Solactiniella |... 2c a ee 418 southern Brazil. 2. ee 306 spatial and temporal variation .. > ,..... . 360,616 Species distribution patterns . . o -o-a ee 263 Species diversity. . -o o oao 147, 508 Specjesnoy. 3... ..- 57,353,361, 369, 499, 503,627 Spermatozoon , ....... xo Meet 15. 44 spermiocyst . Umatropt zh shan feb c inea a eT Spermonde Shelf. * A5.nm.x5-...-83x- 147 Sphenaüldx. occ ee 464 sphinctozoans. rr Y spicule morphology ..-. «o... mo... 650 spicule preparations , s- o cso oc... po... 533 spicules in sediments, o. o... +... 651 JS er ar or 92. 160, 626 SpiBofElia. wey oh) et me ode lr v ue Spt oe Snirastrellaareofata oo. eee ee e... 495 Spirasirella aurivillil, |... 4 2.2.0... 287 Spirastrella hartmant, 5... ee ee ois. 504 Spirastrelld inconstans oc 257 Spirastrellasabogae ooo 597 Spirastrella vagabunda. . o o.o osagaia ee 177 Spirasirella, o csse es do dd 257 Spirophorida |... 2.4... o. 2.4 299 sponge biology... ooo. ee 1 sponge distribution. 2. ee ee 147 sponge growth rates or o or or ee 932 Spongia agaricina. 1 22222 naaa nG 456, 486 Sponpia officinalis ooo... ee 259, 456, 486 Spongia tubulifera . . os 457 Spongiazimoved o 456 SpbHSide. bg 8 EE te eb ayt ss mi dut QUA Spongilla lacustris o.a 6 ciis 93,215, 275,568 NA 653 Spongillidas .-.......... 215,275,368, 568.651 Spongionella |... 22e ee 179 spongivory .. s n n n 160, 214, 239 Spongosorites ruetzleri, «s e 429 Spongusorites cj. salomonensis ovio. 432 Spongosarites siliquaria ooo 429 Spongosorites suberitaides, ooo... oo... - 352 Spongosorites fopsenfi, o... 02.2... 29 Spongosarités, . h cieli ees ib ern ea 0832 Sparadopyle 2,2 -aiaa naacaas s.s ss. 465 Sporadoscinid. occ 467. 469 SuCa. Eo e Ru AAA Q1 174 starfish feeding... 2. ee 0... -.2 207 stalus of research «oe 23 MEMOIRS OF THE QUEENSLAND MUSEUM Stauractigella. a eo iriste a ere tham ale aoe 463 Stelletta ef brevis oao o oaae ee 178 Stéelletadebilis, .. .... 2.2222. ee 180 Stelletta siamairiaena ... o. os... ss 180 sterols . 534 AER APT 286, 572 stevensine, enti O tia aa? 438 strengths (of research) . ERES E ene m. 23 Strepsichordaialendenfeldi oo ooo... 61, 669 Strepsichordald 3 «o. 9 <=. ee 60 strobilinimn .,........ vsat iar faita 688 Strobilospongia. >. 4... ano (jen dct ahead 418 Stromatopord n nadira s Ras 99 sttomatopordids, o 9,99 Stromatospongia micronésica 5... 6... ... . 193 Sirongylacidon . 2 ee 259 Strongylophora cf. strongylata. 2. o csaa oaaae 179 Stromeplophora. coco 258 structural habitat complexity... 22s ss 125 Stylotella aurantium |... 220... ee. . 561 Suberites domuncula: «2.2222 les 382 Suberites pagurearum «s 596 Suberitespeleia. s 2.22 ssl. 257 Suberites rubrus ooo 596 SLAVE TSENG os O StL 477 subfamily DOY... -33 +... +. Y $55 sulfatereducingbacteria. ............. 516,550 sulfated sterols o. ro oc 576 CHRIS BP. 156, 643 Swartschewskia papyracéa. «se 275 Swarischewskia, |... 22.02.2222... . 683 A pee dn a 170 Sycon calearavis, «2.2 2222s oss. 36 Sveon gelatinosum |... ee ee 257 Sycon raphanus. . . ee 394 Symbiodinium microadriaticum , .. . id . 205 symbionts... . . « 63,124, 193, 342, 493, 515, 524 symbiosis... 2.2 2 si. 167, 204, 238,427, 602, 606 Svmpagella nux, o . 36, 345 Swnpagella... ss Lee 603 Synechococcus-type cvanobacteria. . o.. 53,457 A e a eb re 418, 473, 607 [cuum or 479 taxonomic Checklist... 0.0.0... +... + 439 taxonomic overview... 2. ee 9 tàxonomy , 101, 131, 249, 282, 361, 368, 508, 533, 540, 650 Tedania cf, anielaBs , «see 178 Tedania digitata... ek 495 Tedania ignis o., oog ooo 214, 306, 659 temperate SPONBES . 2 2. o. 493 ac se pe E Yn EA o II Dion oao 205,561 Tethya aurartid oo coria ees 509 Tethya aurantium. ooo 897 Tethya cirina, ocio e 594 Tetliya coceinea ooo ee 257 Teihva cf. multistelld 6 socorro NT SUBJECT INDEX Tethya norvegica., occ 597 Telltya-ürplielis. e pto aud © eaa os 597 Tethya robusta o 180,257, 597 Tethya seychellensis |... 2... ee 897 EA AA m nasa dee s , 185, 257 Tetilla japonica +... o... 36. tetrabrominated metabolites ...-.........- 168 tetracyclic triterpenes. . , 2... «o... 1... ....- 572 tetrahydropyrans +... 0.01... moco... es $79 Thamnonema ooo 612 Theonella swinhoei. ooo co... 167 Theonella tubulata |... coo .oo.os TERE. 274 Theonella. .. ...... ev ee rs 178, 343, 345 theonellapeptolides. . 222... 22-2. 342 Theonellidae |... 22er 274 theopalauamide, ... o... +... +... 0 2-2 170 thiocyamates 1.0... llo e... . . 961 ARA IAE 639 Thoosa oc 027 Vhorectidae .... o... o... ......... 57 threats (to research)... 0... ole 23 Thrinacophora ... oo o 259 TIMES zo cA. e nd de 257 fime-lapse Agr. . $4 Side age iron dor 5 9| tissue digestion. .. 2... isses tot 7333 Topsentia ophiraphidites. o.a 0 ee s. 317 Tonseniü.i me... idim? sr ir ews re 432 iopsentins:- s'ese beer ee de trp e hon n OTT toxins... : 300. Loa O asa 205 Toxochdlib 42:24 ae tees dese gee 132 Trachycaulus gurlitti, a. 22... ee we 807 Trachycaulus 5.0 0 paaa 499, 507 Trachycladus laevispirulifer . ooo...» 177 iránsects cc tte ce Ee a eG tor LAS translocation .... 0... om... ee 204 transmission electron microscopy . -----.,- 568.617 transplantation... 2.2222 e 51, 242 wawling. © 4! 622.0 OS: sr Saber caus 455 Tremabolites s u hyi.“ sse ee 465 tridecapeptides o a m por es m d toe ole pb nn 342 trikentrins. - m iè: njarita: ,.. 577 707 > Lus qd to dem ate A 573 Trochospongilla horrida ....... 2.4 -- 215, 509 Trochospongilla variabilis. , ooo... vos, 644 tropical-temperate boundaries . ..... 2.1.1 181 Trupetostoma . ss ess tee 99 ultraplankton diet. |.. 0. les 51,457 ultrastructure .. o... 0... 4 160, 193, 398, 666 underwater weight |... 419 üptike 4n i eje aa € pet e eR aa o ed a 0033 85 Uralanema r op aa ee eee 607 uranium-thorium dating «2... 5... sss 84,9] urchin grazing |.. ooo oaa 125 Vaceletiacrypta ooo. eo. -- 12, 498 Vaceletia ooo... os. 76,462,492, 498 vatiabilin; ¿¿.:.,3to ee ia bn 688 variability, 0.2.2... 0... - 456 Verulina stalactites |... 2.22 36, 467 Vetulina. . ... 466 Vibrid T. a Y at Mey sl el eoa uou s 63, 68 vidéo, Aa m7 ee fps Fb 455,479 Witla AAA en ao ttt Rs 604 Waldoschmittia schmidti o o... o 22 css. 180 water flow rates |.. ooa a ee s. M water turbulence...) osea cana 2224 oe 85 weaknesses (in research) ..... 1... ens 23 AAR ORSINI- « cece y onde ds aa o eaa lhe 93 Xestospongia carbonaria. . . . pa., 22-222. 309 Xestospongia exigia . . ooo. 257, 669 Xestospongia mula «a llo e 160, 532 Xestospongia nigricans, . «2 ss 258 Xestospongia pacifica .. . sse 258 Xestospongia similis ooo... ee 180 Xestospongia testudinaria. ooo ooo... 258, 483 Xestospongia cc 411 x-ray crystallography...) ....--2..-+,5+-., 342 Zaplethea digonoxea aa ee ee ee J25 zonation... 2, ... prs: kai: 0460 307 zooxanthellae 2... o... ooo... .. 77 Zygamycalé ov Re 22 DEBO og Llc pel g orc pg c9 s 44 Zyzzya massalis. «e re 177 Zyzzyü oo. Prid vc. o IH F554 259 ASTRA 3: IAN POTTER * ANN Astra Australia BEEN. QU shine m Commonwealth Bark