SiliceousSpiculesandSkeletonFrameworksinSponges:Origin,Diversity,Ultra - PDF document

SiliceousSpiculesandSkeletonFrameworksinSponges:Origin,Diversity,UltrastructuralPatterns,andBiologicalFunctionsA-J.URIZ,XAVIERTURON,MIKELA.BECERRO,GEMMAAGELLCenterforAdvancedStudies(CSIC),Girona,SpainDepartmentofAnimalBiology(Invertebrates),UniversityofBarcelona,Barcelona,SpainKEYWORDSsilication;spicules;ultrastructure;sponges Contractgrantsponsor:CICYT;Contractgrantnumber:MAR98-1004-C02;Contractgrantsponsor:GeneralitatofCatalonia;Contractgrantnumbers:1999SGR00184andREN2001-2312-CO3/MARandINTERREG-III-2002(toMJUandXT). ofdesmaarms(whichinterlocktoformrigidskeletons)areonlyimplicitlyaccepted(e.g.,Garroneetal.,1981).SPICULEDIVERSITYSiliceousspongespiculeshavetraditionallybeenseparatedintotwocategoriestermed,accordingtotheirsize,megascleresandmicroscleres(e.g.,Le1973).However,sizealonedoesnotsufcetoseparatethetwocategoriesinallcases.Somemicroscleres(e.g.,toxasof,sigmasoforonychaetesof)canbelargerthansomemegascleres(e.g.,oxeasin).Moreover,therearespecieswithspiculesofintermediatesize(e.g.,familyPlakinidae),whicharecalledmesoscleresonlyinhexactinellidsponges(Reiswig,2002).Ontheotherhand,severalspiculeshapesareexclusivelypresentamongeithermegascleresormicroscleres,buttherearealsoanum-berofexceptions(e.g.,microxeasandmicrostrogylesaresimilarinshapeto,butsmallerthan,oxeasandAmorediagnosticcharacterusedtosep-aratemegascleresfrommicroscleresistheirroleinskeletonorganization.Megascleresusuallyformthemainskeletalframework.Incontrast,microscleresarewidespreadinthespongebodyandonlyrarelyaretheyembeddedincollagenousmaterial(e.g.,whentheyareconcentratedinaperipherallayerthecortexasinGeodiidae,Ancorinidae,orTethyidae).Anaccessoryroleofmicroscleresinthespongeskeletoncanbeas-sumedafterobservingspongespeciesthatgrowandreproducenormallybutlackoneorseveraltypesofmicrosclere.Inadverseenvironmentalconditions,suchaslimitedsilicicacidavailability(YourassowskyandRasmont,1983;Maldonadoetal.,1999),spongesmaynotproducemicroscleres,whichhasnodetectableef-fectonthemainskeletonframework,speciesshape,orthicknessorecologicalsuccess.Forinstance,isoneofthefewspongespeciesthatcompetessuccessfullyforthesubstratewithseaweedsinthewesternandcentralMediterraneansublittoral(e.g.,Urizetal.,1992),althoughsilicaconcentrationinthesewatersdoesnotallowthespeciestoproduceitschar-acteristicmicroscleres(UrizandMaldonado,1995).Siliceousspiculesarehighlydiverseinspongesandtheselectionpressuresresponsiblearedifculttoen-visage.Thereareover12basictypesofmegasclereand25typesofmicrosclerereportedinDemospongiae,20basictypesofmegasclere,and24typesofmicro-sclereinHexactinellida,besidesalonglistofvaria-tionsofthebasictypes(Fig.3)(Boury-Esnaultandtzler,1997;TabachnickandReiswig,2002).Inseveralordersofdemosponges,somemegascleresormicroscleres(desmas)becomehypersiliedandin-terlocktoformacompact()skeleton(Fig.4) Fig.1.Typicalmegascleresofdemo-spongesandhexactinellids.Lightmi-croscopeimageoftetraxons(triaenes)ofseveralfamiliesofastrophoridDemo-SEMphotographofatriaxonSEMacanthotriaeneof,modiedfromUriz,1988;edfromUriz,2002).M.-J.URIZETAL. thatmayconferastonyconsistencytothesponge.Inspeciesthathavethegeneticpotentialtoproducedes-mas,theconcentrationofsilicicacidintheenviron-mentmaydeterminewhetherthesespiculesareex-pressed(e.g.,C.crambe,Maldonadoetal.,1999).Con-versely,inhighSiconcentrations,nodesmoidspiculesaccumulateadditionalsilica,givingriseto(Fig.5).InupwellingregionssuchastheNamibia Fig.2.SEMimagesofmicroscleresfromdifferentordersofDemospongiaeandHexactinellida.spirastersofasterosemicroacanthostylediscorhabdsofsterrastersofmicroscleresofricome(microhexaster)of,modiedfromUriz,1988;edfromBoury-Esnaultetal.,SILICEOUSSPICULESANDSKELETONFRAMEWORKS shelf(Uriz,1988),thetylostylesofahavegreatlyenlargedshaftsandroundedpoints,asaresultofwhichtheyresembleabnormaltylostrongyles(Fig.6).Inotherspecies,suchasGuitarraßamencasilicaspheresmayaccompanynormalspicules(Car-balloandUriz,1998;Fig.7).Furtherexamplesofhy-cationincludetheproximaltyleofsometylo-stylesofCrambe,Discorhabdella(UrizandMaldonado,1995),ortzlerandSmith,1993).Allthesetylostyleshaveanaxiallament,whichispolyaxialatthetyloid(proximal)end.Consequently,youngspic-uleshaveapolyactinateheadwithshortactines,whichareseenaspseudospinesorpronouncedlobes.Thisproximalzoneisfurtherenlargedbysilicaappositionandbecomeasubsphericaltylesimilartothoseofnormaltylostyles.Spiculesuptoseveralcmlong(e.g.,Levi,1989)andstronglyhypersiliedskeletonsaretypicalofmanyhexactinellidsponges.SKELETALFRAMEWORKSSpiculesandmostparticularlymegasclerescanbedistributedthroughoutthespongemesohylbuttheygenerallyframetwo-orthree-dimensionalstructuresjoinedbyspongin(mostdemosponges),interlockwitheachother(lithistids),orarecementedbyadditionalsilica(Hexactinellids).Incontrasttothehighdiversityofspicules,therearerelativelyfewbasictypesofskeletalframeworkinDemosponges.Sixelementaltypesofskeletonswithintermediateformscanbedifferentiated:hymedes-moid,plumose,axial,radiate,reticulate,andarrangedinstrengthconfusion(Boury-EsnaultandRu1997).Similararrangementsmaybefoundinunre-latedtaxonomicgroupsasaresultofconvergentevo-lution(e.g.,radialarrangementinthehadromeridTethydaeandthespirophoridTetillidae).Thespongegrowthhabitusuallyreectstheunder-lyingskeletalarrangementindemosponges.Withfewexceptions,thinlyencrustingspongeshaveahymedes-moidskeleton(megascleresarrangedsinglywiththeirpointsdirectedupwards(e.g.,Eurypon,Hymedesmiaearlystagesof).Thicklyencrustingformsmayhideaplumoseskeleton(e.g.,Dictyonella,Phor-),whilesolid,cylindrical,andbranchingformsgen-erallyhaveaxialskeletons(e.g.,Raspailia,Axos).Sub-sphericalorglobularformsusuallyhavearadiatespiculearrangement(Tethya,Geodia,Polymastia,).Massive,irregularformsmayhaveeitherdis-arranged(Topsentia,Epipolasis,Halichondria)orre-ticulatedskeletons(e.g.,Petrosia,Haliclona).Finally,tubularspongesareoftenmadeofreticulatedskeletonwalls(Fig.8)(e.g.,).Othermoresophisti-catedskeletonsoccurasaresultofcombinationsofthebasictypes(e.g.,inpapillaorstule-bearingspecies,Fig.9). Fig.3.SEMimagesofplumules(modiedtriactines)ofhexacti-Pentactines(ModiedfromUriz,1988). Fig.4.SEMimageofinterlockeddesmasofCrambeacuataedfromUriz,1988).M.-J.URIZETAL. Conversely,therelationshipbetweenbodyshapeandskeletalframeworksislessevidentforhexactinellids.Theyareusuallymassivespongeswithaclearlyde-nedshape:erect,pedunculated,orsessile,eitherun-branchedorbranched,tubularsaccular,mushroom,fan,blade,andfunnelshapes.Manyoftheseforms(Fig.10)shareaprimarilydictyonalframeworkandthespecicbodyshapereliesonthesecondarystruc-tureofthereticulateframework.Thetwostructuralthatcanbedistinguishedinademosponge,theectosomeandthechoanosome(e.g.,Galeraetal.,2000),usuallyhavedifferentspiculetypesandarrangements.Thechoanosomalskeletonplaysthemainsupportiveroleinasponge(Boury-EsnaultandRutzler,1997).Buttheperipheral(ecto-somal)skeletons,whichareoriginalinmanycases,maycontainmegascleres,microscleres,orboth.Someofthemostfrequentarrangementsofthespongeecto-someareasfollows: Fig.5.SEMimagesofabnormal(hypersilicied)astersofdrillanuculaedfromBavestrelloetal.,1993). Fig.6.SEMimageofanhypersiliciedtylostyleofSuberitestylob- Fig.7.SEMimageofanabnormalhypersiliciedspiculeinsp.(ModiedfromCarballoandUriz,1998). Fig.8.SEMimageofamonactines-madereticulatedskeletonofanhaploscleriddemosponge.SILICEOUSSPICULESANDSKELETONFRAMEWORKS Densepalisadeofupward-pointingmegascleresofthesametypebutshorterthanthoseinthechoano-some(e.g.,inSuberitidae,Polymastidae).Bundlesofshortmegascleresthathispidatethespongesurface(e.g.,Raphidophlus,Willardia)orsurround(echinate)alargerchoanosomalmegas-clere(e.g.,styloidsinTangential,parchment,andparatangentialdisar-rangedlayer(e.g.,acanthoxeascrustin,oxeas,parchmentinParticularreticulationofspiculesofsimilarshapeasthoseinthechoanosome(e.g.,insomeMicroscleresconcentratedinacrust(e.g.,disco-rhabdsin,sterrastersin,spi-rastersinFurthermore,whenanectosomalskeletonisabsent(e.g.,inAxinellida),theectosomeisoftenhelduportraversedbychoanosomalspicules,whichhispidatethespongesurfaceandallowsubectosomalcanalstoSiliceousskeletonsareparticularlywelldevelopedindesma-bearingdemosponges(Fig.11)andhexactinel-lids.Somehexactinellidskeletons(i.e.,subclassAm-phidiscophoraSchulze,1886)areformedbylooseme-gascleresofdifferenttypes(hexactins,pentactins,stauractins(tetractines),tauactins(triactines),anddi-actins),usuallyatspecicsitesofthespongebody.Microscleresarealwaysamphidiscsandnevermicro-hexactinsorderivedforms.Incontrast,megascleresfuseinmostskeletonsofHexasterophoraSchulze,1886,formingarigiddictyonalnetworkinwhichtheoriginalspiculesareobscuredbysilica(Fig.12)orareonlyjoinedbysilicacementatcontactpoints.Theframeworksofdictyonalhexactinellidsshowverydif-ferentpatterns,withmeshesbeingrectangular,rounded,irregular,smooth,orspiny,andformedbyfusionofdifferenttypesofmegasclere:pentactines,diactines,andhexactines.Microscleresaremicrohex-astersthatcanbelooseinthespongebodyorpartiallyfusedtothemainskeleton.Fusionofhexactineorhexactine-derivedspiculesmaybesimultaneoustospic-uleformation(Reiswig,2002).However,thebasallongdiactinesthatanchorthespongetothesubstratefuselongafterthespiculesaresecreted(Reiswig,2002).Hexactinellidspongeshavemoresilicaperbio-massunitthandemosponges.Asaconsequenceoftheirsilicademands,theyareconnedtosilica-richenvironmentssuchasdeepbottomsandupwellingSPICULEFORMATION:ULTRASTRUCTURALANDMOLECULARPATTERNSUltrastructuralAspectsWhethersecretionofspiculesinspongesisintra-orextracellulariscontroversial.Microscleresaremostlysecretedintracellularly(e.g.,Custodioetal.,2002)withonlyoneknownexception(Simpson,1968).Similarly,severalstudiessupportintracellularsecretionofmegascleres(Garrone,1969;SimpsonandVaccaro,1974;Garroneetal.,1981;Hartman,1981),andonlyonesupportsextracellularsecretion(megascleresofthedemospongeCrambecrambe,Urizetal.,2000).However,allthestudiesthatsupportedtheintracellu-larsecretionofmegascleresindemospongesdealtex-clusivelywithsmallspicules.Asinglespongecellcan-notcompletelyincludealargemegasclere,likesomeofAstrophorid,Spirophorid,Hadromerid,orcertainAx-inellidsponges,andthereforeseveralsclerocytesap-peartobenecessaryforsecretingasinglespicule.Incontrast,intracellularformationwithingiantmultinu-cleatesclerocyteshasbeensuggestedforthemega-scleresinHexactinellida(MackieandSingla,1983),adhocformationoftransientjunctionsamongsclerocytestosecretesomecentimeterslonghexactscannotberuledout(Boury-EsnaultandVace-let,1994).Consequently,bothmodes,intra-andextra-cellularsecretion,appeartooccurdependingoneitherthespiculetypeorthetaxonomicpositionofthespeciesconsidered.Thefactorsdeterminingoneortheothertypeofspiculesecretionremainelusive.Thecellsthatsecretethespicules(sclerocytes)arewelltypiedindemospongesbutlesssoinhexactinel-lids.Theyarecharacterizedbyabundantsmallclearvesiclesandmitochondria(Fig.13),whichareanindi-cationoftheirhighmetabolicactivity,andaGolgiapparatus,andmayoccurbothisolatedinthespongemesohylandformingclusters(Urizetal.,2000;Wilkin-sonandGarrone,1980).Megasclerocytesarelargerthanmicrosclerocytesandhaveanucleolatednucleus(e.g.,Custodioetal.,2002;Garroneetal.,1981;Simp-son,1968;WilkinsonandGarrone,1980).Moreover,inexperimentalculturesofSpongillalacustris,micro-andmegasclerocytesalsoresponddifferentlytochangesinsilicicacidconcentration(Jrgensen,1944).Althoughnodifferenceshavebeenfoundattheul-trastructurallevel,sclerocytesseemtobespecicin Fig.9.Longitudinalsectionofanastrophoridspongeshowingacomplexskeletalarrangement:radialinthemainspongebodyandaxialinthepapilla.M.-J.URIZETAL. thetypeofspiculetheysecrete.Thedifferencesrelyonthegeneticcontroloftheaxiallamentthatchiedeterminesspiculeshape.Eachtypeofsclerocyteap-pearstorequireaparticularsiliconconcentration,be-lowwhichitdoesnotsecretethecorrespondingspicule.InthedemospongeC.crambe,thisthresholdappearstobehigherformicrosclerocytesthanformegasclero-cytes,formicrosclerocytesthatsecretedesmathanforthoseproducingisochelae,andformegasclerocytesse-cretinglargestylesthanthosesecretingsmallones.Consequently,smallstylesaretheonlyspiculetypeproducedinC.crambelivinginthesilicon-poorwest-ernMediterraneanwaters(Maldonadoetal.,1999).Aunit-typemembranesurroundsmostgrowingspic-ules.Thismembrane,calledsilicalemma(Fig.14),wasthoughttobesomethingdifferentfromtheplasma-lemmaorcellmembrane(GarroneandLethias,1990).However,aconnectionbetweenbothmembraneswasrstspeculated(Simpson,1984)andlatershowninTEMimages(Urizetal.,2000).Thesecretionofspiculeswithoutasurroundingmembranehasbeenreportedonlyonce:themicroxeasofbulariairata,whicharese-cretedalltogetherinavacuole(WilkinsonandGarrone,1980).Theexistenceofspiculeswithoutanindividualmembranequestionstheroleofthesilicalemmainshap-ingthenalspiculebycontributingtotheformationofswellingsandspines(Simpson,1984).However,desmasusuallyformarticulateframeworks,andthepresenceofintermediatemembranesmaybetheonlystructurepre-ventingfusionamongneighboringdesmas(Fig.15).AxialFilament:RoleinSilicaPolymerizationandSpiculeShapingAllthesiliceousspiculeshaveacentralcorelledbyaproteinaceousmaterialwithsilicaboundtotheor-ganicmatrix(Fig.16),calledtheaxiallament.SilicaformspartoftheaxiallamentinC.crambe(Urizetal.,2000)andstronglyimpregnatestheaxialofspiculesinCorticiumcandelabrum(thisarticle).Intheformer,silicananospheres(Shimizuetal.,1998)areseenattheexternalzoneoflament(Fig.17),whilein,transversalsectionsofthelamentshowsimilarfracturestothoseofthespiculewalls(Fig.18A),whichisanindicationoflamenthardnessduetosilica.Desmoid(hypersilicied)spiculesalsohaveapro-teinaceousaxialcorebutdifferfrominthatthelamentisshorterthanthespiculearms(Fig.19)andtheirwidthmayvary.Thequestionre-mainsastohowsilicadepositiontakesplaceinthe Fig.10.SEMimagesofdictyo-nalframeworksinHexactinellida.edfromUriz,1988).SILICEOUSSPICULESANDSKELETONFRAMEWORKS terminalarmsofdesmas,whicharedeprivedofaxiallaments.Pisera(pages312326inthisissue)sug-geststhepresenceoforganicmoleculesthatwouldinducesilicapolymerizationwithoutformingdiscreteThecross-sectionoftheaxiallamentintheseveralmegascleresandmicroscleresexaminedhasgeneratedagreatdealofdiscussion,asitwasthoughtthatthelamentshapemightexplainspiculemorphology(Reis-wig,1971).Thereisagreementinthatthecross-sectionoftheaxiallamentinhexactinellidspeciesisqua-drangular(e.g.,Sandford,pages336355inthisissue),butwhetheritistriangularorhexagonalindemo-spongesisstillcontroversial.However,accordingtoSEMimagesbothshapesarepresentinDemosponges.AstrophoridandSpirophorid(Butschli,1901;Reiswig,1971;Simpsonetal.,1985),Hadromerid(Bu1901),andPoecilosclerid(Urizetal.,2000)aretriangularincross-section.Moreover,onalshapeshavebeenfoundintheaxiallamentsofHaploscleridspicules,whicharealsothinnerandclearlyshowaparacrystallinestructure(Garrone,1969;Donadeyetal.,1990;WilkinsonandGarrone,1980).AdistinctparacrystallinestructureisalsoseenintriangularlamentsofthePoeciloscleridP.tenacior.Threecrystallizationplanes,paralleltothethreesidesofatriangle,overlaptoproducehexagonalcells,3nmindiameter(Fig.18B).Thishexagonalnanostructurematchesrecentndingsonnestructureoftheaxiallamentsofoxeas(dif-fractionstudies),whichshowa2Dhexagonallatticewitharepeatingdistanceof6nm(Croceetal.,pages381inthisissue).Onthebasisofthesinglespeciesstudiedtodate,theorderHomosclerophoridaappearstoshowcleardiffer-enceswiththeotherDemospongiaeintheshapeofthelament.ThespiculesofC.candelabrum Fig.11.SEMimageofanasteroiddesmaofCrambeacuataedfromUriz,1988). Fig.12.SEMphotographofhexactinellidspiculesembeddedinsil-Partiallyembeddedspicule(arrow).transversalsectionshow-ingspicules(arrows)surroundedbylargeamountsofadditionalsilica.M.-J.URIZETAL. irregularlyshapedlaments(Fig.20),closelyinter-mingledwithsilicamicro-ornanospheres(Fig.17).lamentsvaryinwidthacrossspiculetypes.Spiculeswiththickerwallscouldbeexpectedtohavethickeraxiallaments,sincethesewouldcontroltheamountofsilicapolymerized.However,lamentandspiculediametersdonotappeartobepositivelycorre-lated.Theratioaxiallament/totalspiculewidthis0.3forC.crambespicules(Urizetal.,2000),0.20.4forthethin(0.1mindiameter)raphidesof(Donadeyetal.,1990),0.4in,andbetween0.20.7forthetetraxonidspiculesofC.candelabrum.Conversely,thisratioonlyreaches0.03fortheastersandthicktriaenesofS.grubei(Simpsonetal.,1985).Reiswig(1971)attributedaphylogeneticvaluetothecross-sectionshapeoftheaxiallament.Hefoundtri-lamentsinAstrophorids,Spirophorids,Hadromerids,Axinellids,andsomePoecilosclerids,andhypothesizedthatboththetriangularmorphologyandapolyactinalspiculeshapewouldbeplesiomorphicwithindemospongesandwouldbepotentiallypresentineitherofthetwosubclassesestablishedbyLe(1973)(i.e.,TetractinomorphaandCeractinomorpha)independentlyofwhetherthepolyactinalshapeisvis-ible.ReiswigshypothesisiscurrentlygainingstrengthafterTEMandSEMexaminationofmorespeciesbe-longingtoseveralordersofDemospongiae.Further- Fig.13.TEMphotographofasclerocytecontaininganaxialment,abundantmitochondria,andclearvesicles. Fig.14.TEMdetailofasclerocytecontaininganaxialSeethemembrane(arrow)surroundingthelament(silicalemma). Fig.15.SEMimageofinterlockeddesmas(arrows)(ModiedfromUrizandMaldonado,1995). Fig.16.TEMtransversalsectionofaspiculeofshowingatriangularaxiallament(a),impregnatedwithsilica.SILICEOUSSPICULESANDSKELETONFRAMEWORKS more,thetriangularaxiallamentsreportedinsomemonaxonidspiculesofpoecilosclerids(UrizandMaldo-nado,1995)andhadromerids(RutzlerandSmith,1993)haveapolyaxialorigin,sincetheiraxialbranchesatoneend.Incontrast,allthehaploscleridsponges(eitherwithmonactinalordiactinalspicules)examinedsofarshowedaxiallamentshexagonalincross-section(Garrone,1969;SimpsonandVaccaro,1974;WilkinsonandGarrone,1980;WeissenfelsandLangenbruch,1985).Althoughlamentswithhexago-nalcross-sectionhavealsobeenattributedtothespic-ulesofthePoeciloscleridbulariairataandGarrone,1980)andtheraphidesofAxinellapol-(Donadeyetal.,1990),otherprolescanbefound:intheTEMguresshownbytheseauthors:triangular,triangularwithcutangles(i.e.,fauxhex-),rounded,andirregularshapes.ThespiculesoftheorderHomosclerophoridamaybeexceptionalsinceintheonlyspeciesexaminedsofar(C.candela-),thelamentsareirregularincross-section.Thetriangularorhexagonallamentsectionscorre-latewiththetwomodelsofsecretionpostulatedbySimpson(1990).Atriangularsectionisassociatedwithmonactinesandhexagonalwithhaploscleriddiactines.Oxeasoffresh-waterspongeswithhexagonalments,whenculturedeitherinasilica-richmedium(Simpson,1981)orinthepresenceofgermanium(Simpson,1990),producedcentralswellings,whichhavebeeninterpretedascentersfromwhichsilicaextendsinoppositedirectionstowardsbothspiculepointedends.Incontrast,thegrowthofmonactinespiculeswithtriangularlamentswouldbeunidirec-tionalfromaproximalsilicationcenter(Simpson,1990).Thesealternativehypothesesofspiculegrowtharesupportedbytheobservationofyoungdiactinesormonactinesthatshowaswellingatacentralorproxi-malposition,respectively(Corrieroetal.,1996;Uriz,ChemicalandMolecularAspectsThechemicalcompositionofdemospongeandhexactinellidspiculesvariesslightlydependingonthespeciesandthewatercompositioninthehabitat,butitismostlysilica(SiO)andwater(e.g.,J1944),withsometraceelements(seeSandford,pages355inthisissue).Othermineralions,whichdonottakepartofthespiculecomposition,suchas),appeartobedecisiveinactivatingtheen-zymesthatcatalyzesilicapolymerization(LePennecetal.,2002;Mulleretal.,pages368377inthisissue).Spongestakeupsiliconintheformofsolublesilicicacid.Siuptakebyspongeshasbeenmeasuredinlab-oratoryexperiments(FrohlichandBarthel,1997;Re-inckeandBarthel,1997;Maldonadoetal.,1999)andmayvaryaccordingtoSiconcentrationinthewater,temperature,andotherenvironmentalfactorsthataf-fectspongephysiologyandmetabolism.Positivecorre- Fig.17.TEMtransversalsectionofaC.candelabrumNotethesilicananospheresinthezonebetweentheaxialandthespiculewall(arrows). Fig.18.TEMtransversalsectionsofTheaxiallamentshowssimilarfracturetothatofthespiculewall.Theaxiallamentshowsaparacrystallinestructurewiththreedifferentcrystallizationplanes.M.-J.URIZETAL. lationsbetweentheconcentrationofsilica,thenutri-tionalconditionofthesponge,anditssilicauptakehavebeenreportedforHalichondriapaniceaandBarthel,1997),whiletemperaturedidnotaffectuptakerates.Incontrast,Siconcentrationaboveathresholdmayevenreducesilicauptake(Maldonadoetal.,1999).Ithasbeenspeculatedthatsomespongesmaycompeteforsilicawithdiatomsinsummer(FrohlichandBarthel,1997).However,thiscompeti-tionseemsunlikelyaccordingtoReinkeandBarthel(1997),giventhefasteruptakeandlowersaturationpointofdiatomsanddifferencesintheirspatialdistri-bution(i.e.,watercolumnandseabottom,respectively).Therehasbeenspeculationastowhethersilicicacidistransferredfromthewatertotheinsideoftheunit-typemembranethatwrapsthegrowingspiculebymeansofvacuoles,ordirectlythroughthecytoplasmofthesclerocyte.X-rayanalysiscoupledtoTEMimageshasshownthatthesilicicacidmayenterthespongemesohyldirectlythroughtransientspacesbetweentheepithelialcells,sinceitwaspresentinthecytosolbutnotcontainedinparticularvacuoles(Urizetal.,2000).X-rayanalysisprovidesaccuratecomparisonsbetweencellcompartmentsinthesamesample,providedaninternalcontrolisused.Byusingthismethod,Urizetal.(2000)foundrelativelylargeramountsofSiinthelamentofgrowingspicules(5070%ofSirela-tivetothespiculewall)thaninmaturespicules(3040%),whichindicatestheintimaterelationshipbe-tweenproteinandsilicaduringtherststepsofsilicadeposition.NoSiwasdetectedintheperispicularcol-lagen,inarcheocytes,choanocytes,orinthemesohylfarfromthesclerocytespiculecomplex.Whensclero-cytessecretingspiculeswereanalyzed,mitochondriaandvesicles,whetherelectron-denseorelectron-clear,didnotcontainsignicantamountsofsilica.Incon-trast,silicawasconcentratedinthecytoplasmofthepseudopodiathatsurroundedthegrowingspicule(50%ofSirelativetothatcontainedinthespicule-wall)andintheextracellularspacebetweenthespiculeandthesclerocyte(50Themechanismbywhichsilicaisconcentratedandpolymerizedaroundtheaxiallamenttoformspiculeswaspoorlyknownuntilrecently(Shimizuetal.,1998; Fig.19.SEMphotographsofdifferentphasesintheformationofanasterosedesma.Firststage:silicaisdepositedalongtheaxiallamentgivenrisetoanapparentlyLaterstagesofcationshowadditionalsilicadepositedaroundtheinitialaster-osespicule(notetheendsoftheinitialasterprotrudingfromthedesmaarms(arrows)). Fig.20.TEMtransversalsectionsofC.candelabrumspicules.Thelamenthasanirregularlyroundedshape.SILICEOUSSPICULESANDSKELETONFRAMEWORKS Chaetal.,1999;Kraskoetal.,2000).Shimizuetal.(1998)identiedthemainproteinsthatconstitutethelament(i.e.,silicateins,and)andprovedthattheyareenzymeshighlysimilartomembersofthecathepsin-Landpapainefamilyofproteases.Theseenzymescatalyzethehydrolysisandpolycondensationofsiliconalkoxides(Chaetal.,1999)butthisactivitywasabolishedafterthermaldenaturation,suggestingthattheenzymaticactivityreliesonthetertiarystruc-tureoftheprotein.Thephysicalstructureofthepro-teinhasbeenelucidated(Croceetal.,pages378381inthisissue)and,accordingtotheseauthors,silicateinunitswouldrstactasatemplatefortheformationofahighlyorderedstructuresimilartothatinmesoporematerials,uponwhichsecondarydepositionofamor-phoussilicawouldproceed.Theproposedarrangementofthisstructureissupportedbycross-sectionsoflamentsobservedathighmagnication(Fig.18B).Sipolymerization,eitherinabioticconditionsorbybiomineralization,givesrisetoanetworkofmicro-ornanospheres.Nanospherenetworkshavebeenob-tainedexperimentallyinthelaboratorybyusingpro-teinsfromdiatoms(Krogeletal.,1999)andsponges(Shimizuetal.,1998;Chaetal.,1999).TheycanalsobeobservedingrowingspiculesofmembersofthefamilyClionidae(Schonberg,2000;RosellandUriz,2002)andPoeciloscleridalivinginSi-poorenviron-ments(UrizandMaldonado,1995).Duringspiculefor-mation,thesenanospheresfusetoeachothertoformlargerspheres.Amorphoussilicaisaddedatlaterstages,givingrisetoanevenspiculesurface(Fig.21).Variousnanospherestructuresarevisibleingrowingdes-masofseveralLithistida(seePisera,pages312326inthisissue).Silicananospherescanberecognizedinthe120nmlong,dense(silica-rich)zonesthatsurroundtheaxiallamentofC.crambe(Urizetal.,2000).GENOTYPICVS.ENVIRONMENTALCOMPONENTSINTHEPHENOTYPICEXPRESSIONOFSPICULESIthasbeenexperimentallydemonstratedthatspongesclerocytesproducetheaxiallament(protein)atverylowconcentrationsofsilicon(YourassowskyandRastmont,1983).However,severalspiculetypes,whichareabsentfromnaturalpopulationslivinginlowconcentrationsofsilicon,canbeproducedinthelaboratorywhentheconcentrationofsilicicacidisin-creasedarticially(Maldonadoetal.,1999).Hence,thepotentialnumberofspiculetypesinaspongespeciesappearstobegeneticallyxed,buttheenvironmentalconditions,specically,theavailabilityofsilicon,maydeterminewhetherageneticallydeterminedspiculetypeisnallyexpressed.SeveralinstancesillustratetheabsenceofcertainspiculetypesinMediterraneanspecimensduetochronicsiliconlimitation.TheMedi-AxinellapolypoidesdiffersfromthesiblingAtlanticspeciesA.dissimilisintheabsenceofra-phides.However,anultrastructuralstudy(Donadeyetal.,1990)demonstratedthatMediterraneanspecimensA.polypoideshavemicrosclerocytescontainingmlong,0.1mwideraphides,whicharehardtodetectinspiculepreparationsunderlightmicroscopyduetotheirextremelysmalldiameter.Furthermore,Merlialipoclavidisca,whichlivesinsilica-poorwateroftheBalearicIslands,differsfromM.normaniintheMediterraneansilica-richwatersoftheMedesislandsbytheabsenceofclavisdiscs(Fig.22)(VaceletandUriz,1991).SPICULESARRANGEMENT:HOWTHEYGETTHERIGHTPLACEANDORIENTATIONTheabove-mentionedskeletalarrangementsaremostlybuiltbyseveralspiculetypes,whichareplacedinprecisesiteswithintheskeletalframeworkasthespongedevelopsandgrows.Howeachspiculetypereachesthepreciselocationwithinthosecomplicatedskeletonsisdifculttounderstandwithoutconsidering Fig.21.SEMphotographsoftwoisochelaeindifferentphasesofcation,showinganunevenandevensurface,respectively. Fig.22.SEMimageofaclavidiscfromaspecies.Thisspiculemaybemissingfromspecimenslivinginsilica-poorenviron-M.-J.URIZETAL. theroleofthedevelopmentalgenes,whichinvolveexpressionofdistinctproteinsatseveralspongezonesduringdevelopment.Thereisaclearpolarityintheskeletalorganizationfromthechoanosometotheecto-someinmostspongespecies,andaclearradialsym-metrycanbeseenintheskeletonofsubsphericalsponges(Uriz,2002).HomeoboxeshavebeenidentiinDemosponges(Coutinhoetal.,1994;Kruseetal.,1994;Seimyiaetal.,1994;Degnanetal.,1995;Ri-chelle-Maureretal.,1998,1999)andcalcareoussponges(ManuelandLeParco,2000)andtheirtempo-ralandspatialexpressionhasbeenshownbothinadults(Seimyiaetal.,1997;Richelle-Maureretal.,1999;LarrouxandDegnan,1999)andinembryos(Leysetal.,1999).Thesehomeoboxgenesmayberesponsiblefortheestablishmentofpolarityduringthespongeslifecycle.Calcareousspongesshowaparticularlycom-plexskeletalorganizationinwhicheachdistinctspi-culetypeoccupiesapreciselocation.Jones(1998)high-lightedtheimportanceofmechanicalforcescausedbytheinitialcontiguityofthefoundercellsonmorpho-genesisof(Johnston).AlthoughSimpson(1984)statedthattheexactlocalizationofthedifferentspiculetypesindemospongeskeletonsisrarerthaninCalcarea,therearealsomanyinstancesofcomplexframeworksindemospongeswithspecisitesforanyoftheseveralspiculetypes.Theaxialskeletonofthegenusillustratesthispoint.Itconsistsofacentralcoreformedbylongstylesjoinedbyspongininanirregularlyreticulatedpattern.Long,perpendicularstylesarisefromthisaxialcore,protrud-ingthrough(hispidating)thespongesurface.Smallacantho-tylostylesareplacedperpendiculartothelongperipheralstyles,withtheirbasesonthestyleandtheirpointsupwards.Furthermore,bundlesofthinstyloidsobliquelysurroundthehispidatingstylesattheectosomelevel.Simpson(1984)posedthequestionastowhetherclustersofsclerocytesaredevelopedsimultaneouslyinaparticularzoneofthespongeandthusspiculesareformedintheirdeniteplaceor,alternatively,spiculesaresecretedanywhereinthespongemesohylandthentransportedtotheappropriatedsitetoformtheskele-talframework.Crawlingmesohylcellsapparentlyex-ertforcesthatmayenablethemtomovespicules.Spic-ulestransportbycellsacrossthespongemesohyltodistancesuptoseveralhundredmicrometershasbeenrecordedinbothmarine(BondandHarris,1988;Bond,1992)andfreshwater(Elvin,1971;TokyoCinema,1996;Custodioetal.,2002)sponges.In,sclerocytescontainingorsurroundinggrow-ingspiculeswereobservedwanderingquicklythroughthespongemesohyl(Elvin,1971).TEMimagesofsclerocytessurroundingyoungspiculesbypseudopodiahavebeenreportedinC.crambe(Urizetal.,2000),(Fig.23),and,andtheymaybeanindicationofspiculetransportwhilecompletingsil-icadeposition.Thelongdirectionalpseudopodiaofspic-ule-secretingsclerocytesinScopalinalophyropodavaeshownthroughTEM(Fig.24)alsosuggestactivemovement(Urizetal.,2002).Transportbysclerocyteshasalsobeenreportedformicroscleres(Custodioetal.,2002).Sclerocytesmoveacrossthespongemesohylfol-lowingthegeneralcellow(Elvin,1971;Bond,1992;TokyoCinema,1996)andtheyappeartomovewhilesecretingboththespiculesandtheaxiallament(i.e.,beforesilicasecretionstarts).Oncesilicationhasnished,spongocytessecretebrilsaroundthespicules.Maturespiculesmightbeasforeignmaterialbythespon-gocytes,inthesamewayasbasopinacocytessecretespongintoisolatethespongecellsfromthesubstrate(Fig.25).Afterbeingentrappedandxedtothebasallayerortootherspiculesbyspongin,spiculesmayconstituteatoothertransportingspiculecellsthatmoveacrossthemesohylfollowingthegeneralcellow,therebyintegratingthetransportedspiculesin Fig.23.TEMimageofasclerocytesurrounding(transporting?)aspiculeof Fig.24.TEMimageofalarvalsclerocyteofScopalinalophy-,whichcontainsanaxiallament.Seethelongthinpseudop-odiaatoneendofthecell(arrows),whichindicatesdirectionalactiveSILICEOUSSPICULESANDSKELETONFRAMEWORKS thegrowingspiculebundle.Theiterationoftheprocesswouldleadtotheformationofanelementalskeletonstructure:theprimaryspiculetracks(Bond,1992).Someoftheseentrappedspiculesareabandoneddur-ingthecontinuousmorphogeneticprocessesofthesponge(Borojevic,1971;BondandHarris,1988;Mal-donadoandUriz,1998)sincetheydonotfollowthegeneralcellIncontrast,spiculefusionappearstooccurinparal-leltospiculesecretionindictyonalhexactinellids(Reis-wig,2002).Inthesesponges,spiculesshouldbepro-ducedattheirdenitesiteatxeddistancesfromeachothertoconformtotheresultingregularlyreticulatedTheroleoftensionforcesinspongemorphogeneticprocesseshasbeendiscussedpreviously(e.g.,Borojevic,1971).Teragawa(1990)concludedthatthemechanicalforcesoftensioninthedermalmembraneDysideaavaracouldinuenceskeletogenesisinthiskeratosespeciesbyaffectingbothprimarybergrowthandthetransportofsandparticlestotheprimarybers.Membranetensioncouldalsoinuencespongemorphogenesis,sincethegrowthofprimarybersattheconulesdeterminestheexternalshapeofthesponge.Manysiliceoussponges,particularlythosewithplumoseorascendingreticulateskeletons,showaconulosesurfaceandthemechanicalforcesactinginthecongurationofplumoseprincipalspiculetractsareexpectedtobesimilartothosedeterminingtheorganizationofhornyskeletons.AccordingtoTera-gawa(1990),tensileregulationofskeletogenesiscouldbeachievedthroughadynamicbalancebetweenten-silestressesinthedermalmembrane(ectosome)andcompressivestressesintheprimaryspiculetracts.Inmanycasesbundlesofmicroscleresaresecretedinasinglecellandarethereforeassociatedinorigin.Multiple-spiculesecretionbyasinglecelloccursformicroxeasorraphidesinbulariairataandGarrone,1980),raphidesinAxinellapolypoides(Donadeyetal.,1990)andMycaleangulosaetal.,2002),ortoxasinM.angulosadioetal.,2002).Conversely,inothermicrosclerebundles,suchastheanisochelaerosettesinthegenus,thespiculesaresecretedindividuallyandappeartobelateraggregatedbymicrosclerocytemigrationandag-gregationwhilesecretingthespicules,inasimilarwaytothatdescribedformegascleres(seeabove).Duringanisochelaesecretion,thespiculefootseemstoberst(Custodioetal.,2002)andisthenextrudedtothemesohyl.Onceextruded,theisochelaefeetwouldxedbyspongintoeachother,sinceadensecolla-gen-likematrixhasbeenobservedsurroundingtherosettes(Custodioetal.,2002).Consequently,bothinsitusecretionandtransportofspiculesmaycombineduringtheframeworksforma-SKELETALFUNCTIONSIntuitively,mostspiculeshapesappeartobeusefulinmicro-construction.Forinstance,monaxons(Fig.26)canbecombinedinseveralwaysbyfollowingthesameprinciplesofphysicsusedinhumanbuilding.Proximalswellings(tyles)ofmonaxonsmayprovideanchoringpointsforspiculesinhymedesmoid,plumose,andre-ticulateskeletons.Thesponginusuallyenrobesthespiculebaseandaterminalswelling(Fig.27)increasesthespiculesurfacetobeimmersedinthespongin.Thecladomeoftetraxons(Fig.28)anchoredinanarmoredlayerofmicroscleres(thecortex)givescohesiontothewholestructurebyinterlacingcortexandchoanosomeandallowssubglobularshapesthroughtensionforces.Thequestionarisesastowhyevolutionhasallowedthemaintenanceofmultiplespiculeshapesandframe-worksthatmayplayanalogousskeletalfunctionsinsponges.Forinstance,whyarethereseveralmonax-onidtypeswitheitheroneortwoendsbluntorbothendssharplypointed(e.g.,styles,tornotesoxeas, Fig.25.TEMimageofthesponginthatsurroundsmaturespic-Spongocytesproducingsponginbrils(arrows)inclosecon-tacttothesponginmass.Twospicules(sp)surroundedbyspongin,whichisinvadedbyarcheobacteria(arrowheads).M.-J.URIZETAL. Asforthedemospongemicroscleres,manyofthemalsohaveapparentlyusefulformsforframingskele-tons.Sigmas,isochelae,anisochelae,diancisters,for-ceps,plachochelae(Fig.29),mayhelpholdtogethersomemegascleres,providedtheyhavetheappropriatesize.Asters(poly-actinatemicroscleres),ifinsufdensityandaccompaniedbyspongin,mayproducecon-sistentbutexibleexternallayersandmayeasetheorganizationofcomplicatedinhalantstructures,asinmanyAstrophorids(Uriz,2002).However,theprecisestructuralfunction(ifany)ofsomeclustersofmicro-scleres,suchastherosettesofanisochelaeinthetrichodragmas(raphidesclusters)inremainsenigmatic.SupportingCellsandImprovingSpongeStrengthThemostobviousfunctionofskeletonsinsponges,asinanyotherlivingorganism,istoallowprogrammedcellarrangementinafunctionalbodyplan(supportingfunction).Spiculesxedtothebasalsubstratebyspon-gincanplayadecisiveroleinspongemorphogenesisandposteriorrearrangementprocesses(e.g.,Bond,1992),byactingasanchoringpointsagainstmovingcellstheycanexertforces(Borojevic,1971).Furthermore,skeletonsconferthenecessarystrengthtomarineanimalstoendurehydrodynamicforces:eddies,currents,andwaves(e.g.,McDonaldetal.,2002;Palumbi,1896).Aexiblemainskeletonisnecessaryforerectbranchingformstocounteractstrongcurrents.ErectbranchingformssuchastheMediterraneandemospongeAxinellapolypoidesrangetheirbranchesinaplaneperpendiculartothepredominantcurrents,thustakingadvantageofthefoodtransportedinthewater.Onlythankstoaexibleskeletoncanthespongesurvivehigh-speedcurrents.Athickectosomalcrustcombinedwithara-dialmainskeleton(e.g.,Geodia,Tethya,Aaptos),ahighdensityofspicules(e.g.,)ordesma-madeskeletons(Lithistida)alsoconferresistancetomassivesponges.Encrustingsponges,whichareparticularlyabundantinenvironmentsexposedtostrongwater Fig.26.SEMimageofmonaxons(stylesandstrongyles)ofdemo- Fig.27.SEMimageofaterminaltyleinamonaxon(tylostyle) Fig.28.Lightmicroscopeimageoftriaeneswiththeircladomearrangedtoformtheskeletoncortex(ModiedfromUriz,2002).SILICEOUSSPICULESANDSKELETONFRAMEWORKS motion,counteractthehydrodynamicforcesbyad-rmlytothesubstratumthroughtheirwholeSpiculesmayalsoplayasupportingroleforcellsduringtheprocessofformationofexternalpropagulainspeciessuchasTethyaaurantium,Haliclona,orMycalecontarenii,tocitesomeexamples.ReachingtheWaterColumnAnotherroleoftheskeletalframeworksistotrans-portthespongecellstowardthewatercolumn(i.e.,afewcmabovethesubstrate)byallowingthespongetogrowupwards.Thebenetsof3Dgrowthforfeedingorganismshaveoftenbeenreported(Jackson, Fig.29.SEMimagesofseveralhookedmicroscleresofDemospongiae,whichmayjoinmegascleresinskeletalDiancistersofBipocilloofSigmaofAnchoratechelaofPlacochelaeofedfromUriz,1988).M.-J.URIZETAL. 1979).Thisabilitymayhelpthespongetoeludelateralovergrowthorcompetitionforspacewithcloseneigh-borsaswellasthecloggingofinhalantoricesbysediment(Jackson,1979).Furthermore,thankstotheirparadigmaticplasticity(e.g.,Becerroetal.,1994;Sara`andVacelet,1973),spongesmayundergomorphologicalandskeletalchangesalongtheirlife,dependingontheenvironmen-talconditions(McDonaldetal.,2002),adaptingtheirshapetoperformthephysiologicalfunctionswithaminimalmetaboliccost(Riisgardetal.,1993;RiisgardandLarsen,1995).Spongesthathaveageneticallyxedcapacityforbuilding3Dskeletonscanmaintaintheirthinencrustingshapeduringtheirwholelifeorcanbecomethickencrustingormassive,dependingonfactorssuchassedimentrates,foodavailability,andhydrodynamicforces.Forinstance,thinlyencrustingindividualsofthegenusthickencrustingormassive,andtheirskeletonschangefromhymedesmoidtoplumose,whenwatermotionisloworthetrophicconditionsarefavorable.Spiculesseemtorepresentharmfulelementsinthefood,particularlyifweconsiderthattheymayrepre-sentupto75%ofthespongebiomass(RutzlerandMacintyre,1978;Desqueyroux-Faundez,1990)andthattheyareoftenarrangedintheskeletonwiththeirsharpendtowardsorprotrudingthespongesurface.Asaconsequence,thesiliceousskeletonofspongeshasoftenbeeninterpretedasaneffectivemechanismfordeterringpredation(RandallandHartman,1968;SaraandVacelet,1973).SpiculeconcentrationappearstobeaplastictraitthatcanbeinducedbydamageinsomemorphotypesofAntosigmellavariansandMichelotti(HillandHill,2002).Theseauthorsfoundthatspongesunprotectedfrompredatorsin-creasedspiculeyields,andsuggestthatthelargespic-ule-richcortexofthesemorphotypesisaninduciblestructuraldefense.However,studiesspecicallydesignedtotestthedeterrentcapabilitiesofspongestructuraldefenses(spicules)haveproducedcontrastingresults,depend-ingontheenergycontentofthearticialfoodassayed(ChanasandPawlick,1995,1996;Urizetal.,1996).ChanasandPawlick(1995)foundthatspiculesde-terredpredatorswhenincorporatedintopreparedfoodofanutritionalqualitylowerthanthatofthespongetissue.Acombinationofpoorlydigestiblesponginandindigestiblesilicamayresultintissueoflownutri-tionalqualitythatmaybemoreeffectiveindeterringpredatorsthanaharmfuleffectofspiculesperse(e.g.,WaddellandPawlik,2000).Thespiculearrangementandconcentrationatthespongesurfacemightalsoexertsomeinuenceinavoidingpredation.Thefoodofferedtoshorechino-dermsintheabove-mentionedexperiments(ChanasandPawlick,1995,1996;Urizetal.,1996)didnotmimicthespiculeorientationorconcentrationatthespongeperiphery(e.g.,protrudingthespongesurfaceordenselypacked).However,largepredatorsofspongessuchsh(Wulff,1994)andturtles(Meylan,1988,1990)mayeludethetroublesomespiculesthankstoadifferentsizescaleofthepredatorsmouth(cm)andthespicules(m),respectively.Fishandturtlesusuallypredateonseveralspongespecies,whichmaybetheresultofcoevolutionsincepartialpredationbyshisnotfatalforsponges(Wulff,1997).Moreover,somespecializedinvertebratescanalsocircumventthedefensesbycrushingthespiculesthankstospecialtoolsforgrazinghardmaterials(e.g.,mollusksseaurchinslantern,polychaetesmandibles)(e.g.,CiminoandSodano,1994;MartinandBritayev,1998;Urizetal.,1996),orobtaincellularmaterialbymeansofthinsuctiontrunks(e.g.,Siphonostomatoidcope-pods,MarianiandUriz,2001)orexternaldigestionsuchassea-starsthatfeedexclusivelyonspongesintheAntarctic(Dayton,1974,1979),despitethefactthatsiliceousskeletonsareparticularlywelldevelopedinthisocean.However,fewvertebratesormacro-invertebratesfeedontemperate(e.g.,SalaandBallesteros,1997)ortropicalsponges(e.g.,RandallandHartman,1968).Productionofdeterrentandnoxiouschemicalsappearstobemoreeffectivethanspiculesindeterringpreda-tors(e.g.,Pawliketal.,1995;Urizetal.,1996)andspongesseemtohavearecurrentselectionforchemicaldefensesasapartoftheirlifestrategy(Becerroetal.,Spiculesmaybepresentinspongelarvae.Hexacti-nosespiculesareabundantinthetrychymellalarvaofthehexactinellidFarreasollasiiSchulze(Okada,1928)andtheirarrangement,coveringthewholelarvalbodypointstoasupportingfunction.Someauthorshaveproposedthatlarvalspiculeswouldprotectspongelarvafrompredators(e.g.,YoungandChia,1987).However,thisdoesnotseemveryreliablebecauselar-valspiculesareratherscarce,andifhighamountsofspiculesdonotprotectadultspongesinmostcases(seeabove),itisunlikelythatafewspiculeswouldprotectlarvae.Fewexperimentalstudiesaddressedlarvalde-fenseofsponges,but,forinstance,spicule-bearinglar-vaeofC.crambewerepalatable,whilespicule-freelarvaeofDysideaavarawereunpalatable(Urizetal.,OtherFunctionsSeveralotherfunctions,sometimescomplementarytothemainsupportingrole,canbeenvisagedforspongespicules,althoughmostofthemappearexclu-siveofparticularspecies.Inextremeenvironmentswhereanactivelterfeederhasalowyield,thefamilyCladorhizidaeDendyhasdevelopedparticulartoolsforcapturinglivingpreypassively,whichconsistoflongthintentaclesprovidedwithadenselayerofprotrud-ingupraisedhook-shapedmicroscleres(Fig.30A,B).Thesecarnivorousspongescapturesmallcrustaceans(lessthan1mminsize),whichareentrappedthankstothespongemicroscleres(VaceletandBoury-Esnault,Spiculesalsoappeartoplayaroleingameteandlarvaedispersalofsomespecies.Certainlarvalspic-ulessuchasthediscotriaenesthatcoverthehoplito-mellalarvaofAlectonidae(Vacelet,1998)arenotpresentintheadults.Someofthelarvalspiculesno-tablyfavorlarvalbuoyancyandthusmayincreaselarvaldispersal,asreportedforthestylesthatlargelyprotrudethroughthearmoredhoplitomellaofAlec-tonidae,whichcanbefoundamongtheoceanicplank-ton(Tregouboff,1942).Awell-knowncaseofparticularSILICEOUSSPICULESANDSKELETONFRAMEWORKS spiculesinspongepropagulaisthatofgemmoscleresthatformapneumaticlayerintheresistancegem-mulesofthefreshwaterspongesofthefamilySpongil-lidae,whichfavorsgemmabuoyancyanddispersal(Hartman,1981).ThecarnivorousspongeVaceletandBoury-Esnaulthasmicroscleres(forceps)thataccompanythespermaticcystsatre-lease.Theshapeofthesespicules(Fig.30C)andtheirprotrudingpositionincreasespermatocystbuoyancy.Theycanremainsuspendedinthewatercolumnforalongtime,enhancingthepossibilityofcontactingthelamentsofanotherindividual.Theshapeofthespermatocystforcepsandthehookedanisochelaeofthelamentsalsofavorscapturing(entrappingandphago-cytosis)ofthespermatocysts(Boury-EsnaultandVace-let,pers.commun.).Incontrasttotheroleofspiculesinthebuoyancyofgametes,ithasbeenspeculatedthatin-ternalspiculesfavorlarvalsinkingbydiminishinglar-valbuoyancy,sincespiculesincreaseinnumberasthelarvaegetolder(Meewis,1939;Woollacott,1993;Mal-donadoetal.,1997).Accordingtosomeoftheseau-thors,abuoyancydecreasewouldenhancethechancesoflarvaecontactingthesubstratum,thuspromotingsettlementsuccess.However,larvaemaysettlebeforespiculesecretionstarts(e.g.,Urizetal.,2001)andmanylarvaedonotproducespicules(i.e.,keratosesponges)andsettlenormally.SPICULESANDSKELETALFRAMEWORKS:INCONSISTENCIESINTHEIRTRADITIONALUSEINTAXONOMYSpiculesize,shape,andarrangementaretradition-allyusedasthemaindiagnosticcharactersinspongetaxonomy(e.g.,Levi,1973;HooperandvanSoest,2002).However,thereisnoconsensusastothetaxo-nomicvalueofagivencharacter,andthesamechar-actertypeisusedatdifferenttaxonomiclevels(e.g.,HooperandvanSoest,2002).Forinstance,thetypeofskeletalframeworkissometimesusedtodifferentiategenera(e.g.,),families(e.g.,RaspailidaefromEury-ponidae,Levi,1973)ororders(e.g.,theorderHexacti-nosidaSchrammenfromAulocalycoidaTabachnickandReiswigintheHexactinellida(Reiswig,2002)).Thepresence/absenceofatypeofmicroscleresepa-ratesgenerainmanycases(e.g.,withandwithoutisochelae,respectively)alsoallowsustodistinguishbetweenfamilies(e.g.,GeodiidaewithsterrastersfromStellettidaewithoutsterrasters,andfromPachastrellidaewithestreptasters;orHamacan-thidae,withdiancisters,fromtheremainingfamiliesofMycalina).Inaddition,thepresenceofatypeofmicro-sclereisconsideredthesynapomorphiccharacterforanorder(e.g.,spinispiresforSpirophoridainDemo-spongiaeoramphidiscsAmphidiscosidawithintheHexactinellida(Reiswig,2002)).Spiculesizeisoftenusedtodistinguishbetweenspeciesofthesamegenus(e.g.,spp.),al-thoughenvironmentalconditionsmayaffectthistrait.Individualsofseveralspecieslivingininsular,silicon-poorMediterraneanwatersproduceshorterandnota-blythinnerspiculesthantheirAtlanticcounterparts(Bibiloni,1990),particularlythoseinupwellingzones(Uriz,1888).Moreover,spiculesundergoseasonalvari-ationinsizeinHalichondriapanicea(Hannaetal.,1998)andChondrillanucula(Bavestrelloetal.,1993).Spiculeshapemayalsochangewithenvironmentalconditions.Forinstance,thepresenceofterminalorsubterminalswellings,whichtransformstyloidspic-ulestotylostylesorsubtylostyles(seeBoury-EsnaultandRutzler,1997),maydependonthesiliconconcen- Fig.30.SEMimagesofthecarnivorousspongeSEMphotographofthelaments()providedwithprotrudinghook-shapedmicroscleres(arrows).ofahook-shapedmicrosclere.spermatogoniaspicule(forceps),whichfavorsspermatogoniabuoyancyanddispersal(ModiedfromVaceletandBoury-Esnault,1996).M.-J.URIZETAL. trationinthewater.Moreover,ithasbeenshownex-perimentallyforbothfreshwater(YourassowskyandRasmont,1983)andmarine(Maldonadoetal.,1999)spongesthattheconcentrationofsilicicacidmayaffectthephenotypicexpressionofseveralspiculetypes,whicharenotproducedinsilicon-poorenvironments.Alltheseconsiderations,besidesconvergentspiculeshapes(FromontandBergquist,1990)andskeletalarrangements,maycomplicatetheuseofskeletonele-mentsinspongetaxonomy.Althoughthespiculetypesxedgenetically,andmaythusbeusefulinthereconstructionofthespongesphylogeny,variationinspiculesize,shape,andtypesduetoenvironmentalconditionsshouldbetakenintoaccount.Severalrelevantstudiesonspongeskeletogenesishaverecentlybecomeavailable.Thediversityofsuchstudiesgoesbeyondtaxonomicboundariestoincludeareassuchasspiculechemistryandfunctionalroles,orcellularandmolecularcontroloftheprocessofsilicadepositionandframeworksbuilding.Thisongoingre-searchisimprovingourunderstandingofroleofspic-ulesandskeletalframeworksinthebiology,ecology,andevolutionofspongesaswellasthemechanismsdeterminingspiculessecretion.Wearenowuncoveringthecontributionoftheaxiallamenttospiculediver-sity.Asweapplymoleculartechniquesandecologicalexperimentstospongeresearch,weareunravelingtheroleofgeneticandenvironmentalfactorsinspiculeformation.Althoughwehaveclearlyadvancedinourunderstandingof,forexample,thegeneticcontrolofsilicadeposition,othersubjectssuchastherole(ifany)ofthemembranesremainunresolved,astheyhaveforthelast20years,andthesewarrantfurtherinvestiga-tion.Wenowknowthatbasicspiculetypesaregenet-icallydeterminedand,accordingly,theymaybeusefulinthereconstructionofspongephylogeny.Yetwehaveincreasingevidenceoftheroleofenvironmentalfactorsinmodulatingnotonlyspiculesizeandshape,butalsothepresenceandabsenceofspiculetypes.Theimpli-cationsofsuchvariationsinspongetaxonomyloomlargeandcannotbetakenforgranted.Contrarytospongespicules,thereisalowdiversityofspongeskel-etalframeworkswithsimilararrangementspresentinunrelatedtaxonomicgroupsasaresultofconvergentevolution.Skeletalframeworksareclearlyassociatedwithspongegrowthhabitsindemosponges,butthisassociationislessevidentinhexactinellids.Thesevari-ationsinspiculeformationandskeletalframeworksasafunctionofenvironmentalandgrowthhabitsmaypartiallyrespondtotherolethatthesestructureshaveinthebiologyandecologyofsponges.Thereiscontrast-inginformationontheroleofspongespiculesandskel-etonsindefenseagainstpredators,althoughitseemswidelyacceptedthatspongesdependmoreonchemicaldefensesthanonstructuraldefenses.Besidessupport-ingspongecellsandtransportingthemtowardthewatercolumn,spiculesmayhelplarvaestaybuoyantwhileintheplankton,reachthebottomatsettlement,enhancereproductionsuccess,orcatchprey.However,ourunderstandingofnumerousprocessesisstillbasedonalimitednumberofspecies.Wethereforehavetobecautiousaboutthewidespreadoccurrenceofmanyas-sumptions,trends,andhypothesescurrentlyavailableintheliterature.Still,spongescienceremainsanopeneldandintheyearstocomewewillincreaseourknowledgeofthebiologicalfunctionsofsiliceousspic-ulesandskeletonframeworksaswellasourunder-standingoftheultrastructuralandmolecularpatternsofsilicadepositioninsponges.BavestrelloG,BonitoM,Sara`M.1993.Inuenceofdepthonthesizeofspongespicules.In:UrizMJ,RutzlerK,editors.Recentadvancesinecologyandsystematicsofsponges.Barcelona:ScientiaMarina57(4).p415BecerroMA,UrizMJ,TuronX.1994.TrendsinspaceoccupationbytheencrustingspongeCrambecrambe:variationinshapeasafunctionofsizeandenvironment.MarBiol121:301BecerroMA,ThackerRW,TuronX,UrizMJ,PaulVJ.2003.Bioge-ographyofspongechemicalecology:comparisonsoftropicalandtemperatedefenses.Oecologia135:91BibiloniMA.1990.FaunadeEsponjasdelasIslasBaleares.Variaciocualitativaycuantitativadelapoblaciondeesponjasenungradi-entebatimetrico.comparacionBaleares-CostaCatalana.Ph.D.the-sis,UniversitatdeBarcelona.BondC.1992.Continuouscellmovementsrearrangeanatomicalstructuresinintactsponges.JExpZool263:284BondC,HarrisAK.1988.Locomotionofspongesanditsphysicalmechanism.JExpZool246:271BorojevicR.1971.Lecomportamentdescellulesdpongelorsdeprocessusmorphogetiques.AnnBiol10:533Boury-EsnaultN,RutzlerK.1997.Thesaurusofspongemorphology.SmithContribZool596:1Boury-EsnaultN,VaceletJ.1994.Preliminarystudiesontheorga-nizationanddevelopmentofahexactinellidspongefromaMedi-terraneancave,Oopsacasminuta.In:vanSoestRWM,vanKempenThMG,BraekmanJC,editors.Spongesintimeandspace.Proc4thIntPoriferaCongr1993,Amsterdam.Rotterdam:AABalkema.ptschliO.1901.EinigeBeobachtungenuberKiesel-undKalknadelmvonSpongien.ZWissZool69:235CarballoJL,UrizMJ.1998.sp.nov.(Porifera:Poecilosclerida)withaSEMrevisionofthespinyisochelaeandplacochelaeinthegenus.JMarBiolAssocUK78:807ChaJN,ShimizuK,ZhouY,ChristiansenSC,ChmelkaBF,StuckyGD,MorseDE.1999.Silicateinlamentsandsubunitsfromamarinespongedirectthepolymerizationofsilicaandsiliconesinvitro.ProcNatlAcadSciUSA96:361ChaJN,StuckyGD,MorseDE,DemingTJ.2000.Biomimeticsyn-thesisoforderedsilicastructuresmediatedbyblockpolypeptides.Nature40:289ChanasB,PawlikJR.1995.DefensesofCaribbeanspongesagainstpredatoryreefsh.II.Spicules,tissuetoughnessandnutritionalquality.MarEcolProgSer127:195ChanasB,PawlikJR.1996.Doestheskeletonofaspongeprovideadefenseagainstpredatoryreefsh?Oecologia107:225ChombardC,Boury-EsnaultN,TillierS.1998.Reassessmentofho-mologyofmorphologicalcharactersintetractinellidspongesbasedonmoleculardata.SysBiol47:351CiminoG,SodanoG.1994.Transferofspongesecondarymetabolitestopredators.In:vanSoestRWM,vanKempenTMG,BraekmanJC,editors.Spongesintimeandspace.Proc4thIntPoriferaCongr1993,Amsterdam.Rotterdam:AABalkema.p459CorrieroG,LiaciL,PronzatoR.1996.Twonewspeciesof(Porifera:Demospongiae)fromtheMediterraneanSea.BullInstRSciNatBelg66(Suppl):197CoutinhoCC,VissersS,vandenVyverG.1994.Evidenceofhomeoboxgenesinthefresh-watersponge.In:vanSoestRWM,vanKemperTMG,BraekmanJ-C,editors.Spongesintimeandspace.Proc4thIntPoriferaCongr1993,Amsterdam.Rotterdam:AABalkema.p385CroceG,FracheA,MilanesioM,ViterboD,BavestrelloG,BenattiV,GiovineM,AmenitschH.2003.Fiberdiffractionstudyofspiculesfrommarinesponges.MicroscResTech62:378dioMR,HadjuE,MuricyG.2002.Invivostudyofmicrosclereformationinspongesofthegenus(Demospongiae,Poecilo-sclerida).Zoomorphology121:203DaytonP.1979.Observationsofgrowth,dispersalandpopulationdynamicsofsomespongesinMcMurdoSound,Antarctica.Col-loquesInternationauxduC.N.R.S.BiologydesSpongiaires291:SILICEOUSSPICULESANDSKELETONFRAMEWORKS DaytonP,RobilliardGA,PaineRT,DaytonLB.1974.BiologicalaccommodationinthebenthiccommunityatMcMurdoSound,Ant-arctica.EcolMonogr44:105DegnanBM,DegnanSM,GiustiA,MorseDE.1995.Ahox/homhomeoboxinsponges.Gene155:175ndezR.1990.SilicacontentoftheNewCaledonianfaunaofHaploscleridaandPetrosiidaanditspossibletaxonomiccance.In:RutzlerK,editor.Newperspectivesinspongebi-ology.3rdInternationalConferenceontheBiologyofSponges,1985.Washington,DC:SmithsonianInstitutionPress.p279DonadeyC,ParisJ,VaceletJ.1990.OcurrenceandultrastructureofmicroraphidesinAxinellapolypoides.In:RutzlerK,editor.Newperspectivesinspongebiology.3rdInternationalConferenceontheBiologyofSponges,1985.Washington,DC:SmithsonianInstitutionPress,p259ElvinD.1971.Growthratesofthesiliceousspiculesofthefresh-waterEphydatiamuellerihn).TransAmMicroscSocFrohlichH,BarthelD.1997.SilicauptakeofthemarinespongeHalichondriapaniceainKielBight.MarBiol128:115FromontJP,BergquistPR.1990.Structuralcharactersandtheiruseinspongetaxonomy;whenisasigmanotasigma?In:RutzlerK,editor.Newperspectivesinspongebiology.3rdInternationalCon-ferenceontheBiologyofSponges,1985.Washington,DC:Smith-sonianInstitutionPress.p273GaleraJ,TuronX,UrizMJ,BecerroMA.2000.Microstructurevari-ationinspongessharinggrowthform:theencrustingspongesideaavara(Dictyoceratida)andCrambecrambeActaZool81:93GarroneR.1969.Collage`ne,sponginetsquelettemineralchezlHaliclonarosea(O.S.)(Demosponge,Haploscleride).JMicroscGarroneR,LethiasC.1990.Freezefracturestudyofspongecells.In:tzlerK,editor.Newperspectivesinspongebiology.3rdInterna-tionalConferenceontheBiologyofSponges,1985.Washington,DC:SmithsonianInstitutionPress.p121GarroneR,SimpsonTL,PottuJ.1981.Ultrastructureanddepositionofsilicainsponges.In:SimpsonTL,VolcaniBE,editors.Siliconandsiliceousstructuresinbiologicalsystems.NewYork:Springer.pHannaC,SchonbergL,BarthelD.1998.UnreliabilityofDemospongeskeletalcharacters:theexampleof.In:WatanabeY,FusetaniN,editors.Spongescience.Multidisciplinaryperspectives.Tokyo:Springer.p41HartmanWD.1981.Formanddistributionofsilicainsponges.In:SimpsonTL,VolcaniBE,editors.Siliconandsiliceousstructuresinbiologicalsystems.NewYork:Springer.p453HildebrandM,VolcaniBE,GassmannW,SchroederJI.1997.Agenefamilysilicontransporters.Nature385:688HillMS,HillAL.2002.MorphologicalplasticityinthetropicalspongeAnthosigmellavarian:responsestopredatorsandwaveenergy.BiolBull202:86HindeGJ.18871893.AmonographoftheBritishfossilsponges.PalaeontogrSocMonogr1HooperJNA.1990.Characterstability,systematicsandafbetweenMicrocionidae(Poecilosclerida)andAxinellida.In:RuK,editor.Newperspectivesinspongebiology.3rdInternationalConferenceontheBiologyofSponges,1985.Washington,DC:SmithsonianInstitutionPress.p284JacksonJBC.1979.Morphologicalstrategiesofsessileanimals.In:LarwoodG,RosenBR,editors.Biologyandsystematicsofcolonialorganisms2.London:AcademicPress.p499JonesWC.1998.Spiculeformandmorphogenesisinthecalcareous(Johnston).In:WatanabeY,FusetaniN,editors.Spongesciences.Multidisciplinaryperspectives.Tokyo:Springer.p41rgensenCB.1944.Onthespicule-formationofSpongillalacustris(L.)1.Thedependenceofthespicule-formationonthecontentofdissolvedandsolidsilicicacidofthemilieu.DetKglDanskeVi-densksSelskBiolMedd19:2KraskoA,BatelR,SchroderHC,MullerIM,MullerWEG.2000.ExpressionofsilicateinandcollagengenesinthemarinespongeSuberitesdomunculaiscontrolledbysilicateandmyotrophin.EurJBiochem267:4878gerN,DeutzmanR,SumperM.1999.Polycationicpeptidesfromdiatombiosilicathatdirectsilicananosphereformation.ScienceKruseM,MikocA,CetkovicH,GamulinV,Rinkevich,MullerIM,llerWEG.1994.Molecularevidenceforthepresenceofadevel-opmentalgeneinthelowestanimals:identicationofahomeobox-likegeneinthemarinespongeGeodiacydonium.MechAgeingDevLarrouxC,DegnanB.1999.Homeoboxesgenesexpressedintheadultandreaggregatingsponge.MemQueenslandMuseum44:306.LarsenPS,RiisgardHU.1994.Thespongepump.JTheorBiol168:LePennecG,PerovicS,ShokryA,AmmarA,GrebenjukVA,SteffenR,BrummerF,MullerWEG.2003.CultivationofprimmorphsfromthemarinespongeSuberitesdomuncula,morphogeneticpotentialofsiliconandiron.Areview.JBiotechnol100:93viC.1973.SystematiquedelaclassedesDemospongiaria(Desponges).In:GrasseP,editor.Spongiaires.TraitedeZoologie3(1).Paris:Masson.p577viC.1989.Aremarkablystrongnaturalglassyrod:Theanchoringspiculeofthesponge.JMatSciLett8:337LeysSP,HinmanVF,DegnanBM.1999.Whichwayup?Phototaxisandpolarityinasponge.AmZool39:38A.MackiGO,SinglaCL.1983.Studiesonhexactinellidsponges.I.His-tologyofRhabdocalyptusdawsoni(Lambe,1973).PhilosTransRSocLondB301:365MaldonadoM,UrizMJ.1998.Anexperimentalapproachtotheeco-logicalsignicanceofmicrohabitat-scalemovementinanencrust-ingsponge.MarEcolProgrSer185:239MaldonadoM,GeorgeSB,YoungCM,VaquerizoI.1997.Depthreg-ulationinparenchymellalarvaeofademosponge:relativerolesofskeletogenesis,biochemicalchangesandbehavior.MarEcolProgrSer148:115MaldonadoM,CarmonaMC,UrizMJ,CruzadoA.1999.DeclineinMesozoicreef-buildingspongesexplainedbysiliconlimitation.Na-ture401:785ManuelM,LeParcoY.2000.HomeoboxgenediversicationinthecalcareousspongeSyconraphanus.MolPhylEvol17:97MarianiS,UrizMJ.2001.Copepodsofthegenus(Copepoda:Siphonostomatoida)feedingonsponges:behavioralandecologicaltraits.InvertBiol120:269MartinD,BritayevTA.1998.Symbioticpolychaetes:reviewofknownspecies.OceanogrMarBiolAnnuRev36:217MeewisH.1939.Contributionatudedel`sedesHalichonalimbata(Mont.).AnnSocRZoolBelg70:MeylanA.1988.Spongivoryinhawksbillturtles:adietofglass.Science239:393MeylanA.1990.Nutritionalcharacteristicsofspongesinthedietofthehawksbillturtle,Eretmochelysimbricata.In:RutzlerK,editor.Newperspectivesinspongebiology.3rdInternationalConferenceontheBiologyofSponges,1985.Washington,DC:SmithsonianInstitutionPress.p472llerEG,KraskoA,LePennecG,MullerCI,SchroderHC.2003.Biochemistryandcellbiologyofsilicaformationinsponges.MicroscResTech62:368OkadaY.1928.Onthedevelopmentofahexactinellidsponge,.JFacSciImpUnivTokyo2:1PalumbiSR.1986.Howbodyplanslimitacclimation:responsesofademospongetowaveforce.Ecology67:208PawlikJR,ChanasB,ToonenRJ,FenicalW.1995.DefensesofCaribbeanspongesagainstpredatoryreefsh.I.Chemicaldeter-rence.MarEcolProgSer127:183PiseraA.2003.Someaspectsofsilicadepositioninlithistiddemo-spongedesmas.MicroscResTech62:312RandallJE,HartmanWD.1968.Sponge-feedingshesoftheWestIndies.MarBiol1:216ReinckeT,BarthelD.1997.SilicauptakekineticsofinKielBight.MarBiol129:591ReiswigHM.1971.Theaxialsymmetryofspongespiculesanditsphylogeneticsignicance.CahBiolMar12:505ReiswigHM.2002.ClassHexactinellidaSchmidt,1870.In:HooperJNA,vanSoestRWM,editors.SystemaPorifera:aguidetothecationofsponges.NewYork:KluwerAcademic/Plenum.pRichelle-MaurerE,vandeVyverG.1999.TemporalandspatialexpressionofEmH-3,ahomeobox-containinggeneisolatedfromthefreshwaterspongeEphydatiamuelleri.MechAgeingDev109:203Richelle-MaurerE,vandeVyverG,VissersS,CouthinhoCC.1998.Homeobox-containinggenesinfreshwatersponges:characteriza-tion,expressionandphylogeny.ProgMolSubcellBiol19:1RiisgardHU,LarsenPS.1995.Filter-feedinginmarinemacro-invertebrates:pumpcharacteristics,modellingandenergycost.BiolRev70:67RiisgardHU,LarsenPS,ThomassenS,JacobsenH,WeeksJM.1993.SuspensionfeedinginmarinespongesHalichondriapaniceaM.-J.URIZETAL. Haliclonaurceolus:effectsoftemperatureonltrationrateandenergycostofpumping.MarEcolProgSer96:177RosellD,UrizMJ.1997.Phylogeneticrelationshipswithintheexca-vatingHadromerida(Porifera),withasystematicrevision.Cladis-tics13:349RosellD,UrizMJ.2002.Excavatingandendolithicspongespecies(Porifera)fromtheMediterranean:speciesdescriptionsandidenti-cationkey.OrgDiversEvol2:55tzlerK,MacintyreIG.1978.Siliceousspongespiculesincoralreefsediments.MarBiol49:147tzlerK,SmithKP.1993.Thegenus(Suberitidae)andnewspeciesinthecomplex.In:UrizMJ,RutzlerK,editors.Recentadvancesinecologyandsystematicsofsponges.Barcelona:ScientiaMarina.p381SalaE,BallesterosE.1997.Partitioningofspaceandfoodresourcesbythreeshofthegenus(Sparidae)inaMediterraneanrockyinfralitoralecosystem.MarEcolProgSer152:273SandfordF.2003.Physicalandchemicalanalysisofthesiliceousskeletonsinsixspongesoftwospongegroups(DemospongiaeandHexactinellida.MicroscResTech62:336SangsterAG,ParryDW.1981.Ultrastructureofsilicadepositsinhigherplants.In:SimpsonTL,VolcaniBE,editors.Siliconandsiliceousstructuresinbiologicalsystems.NewYork:Springer.p`M,VaceletJ.1973.EcologiedesDemosponges.In:Grasseeditor.Spongiaires.TraitedeZoologie3(1)Paris:Masson.p462nbergCHL.2000.BioerodingspongescommontotheCentralAustralianGreatBarrier:descriptionofthreenewspecies,twonewrecords,andadditionstotwopreviouslydescribedspecies.Senck-enbergMarit30:161SeimyiaM,IshiguroH,MiuraK,WatanabeY,KurosawaY.1994.Homeobox-containinggenesinthemostprimitivemetazoa,thesponges.EurJBiochem221:219SeimyiaM,WatanabeY,KurosawaY.1997.IdenticationofPOU-classhomeoboxgenesinafreshwaterspongeandthespecicex-pressionofthesegenesduringdifferentiation.EurJBiochem243:ShimizuK,ChaJH,StuckyGD,MorseDE.1998.Silicateinalpha:cathepsinL-likeproteininspongebiosilica.ProcNatlAcadSciUSASimpsonTL.1968.Thestructureandfunctionofspongecells:newcriteriaforthetaxonomyofPoeciloscleridsponges(Demospongiae).PeabodyMusNatHist(YaleUniv)25:1SimpsonTL.1981.Effectsofgermaniumonsilicadepositioninsponges.In:SimpsonTL,VolcaniVE,editors.Siliconandsiliceousstructuresinbiologicalsystems.NewYork:Springer.p527SimpsonTL.1984.Thecellbiologyofsponges.NewYork:Springer.pSimpsonTL.1989.Silicationprocessesinsponges:Geodiaasterstheproblemofmorphogenesisofspiculeshape.In:CrickRE,editor.Origin,evolutionandmodernaspectsofbiomineralizationinplantsandanimals.NewYork:KluwerAcademic/Plenum.p125SimpsonTL.1990.Recentdataonpatternofsilicicationandtheoriginofmonaxonsfromtetraxons.In:RutzlerK,editor.Newperspectivesinspongebiology.3rdInternationalConferenceontheBiologyofSponges,1985.Washington,DC:SmithsonianInstitutionPress.p264SimpsonTL,VaccaroCA.1974.AnultrastructuralstudyofsilicadepositioninthefreshwaterspongeSpongillalacustris.JUltrastrRes47:296SimpsonTL,LangenbruchPF,Scalera-LiaciL.1985.SilicaspiculesandaxiallamentsofthemarinespongeStellettagrubiiDemospongiae).Zoomorphology105:375TabachnickKR,ReiswigHM.2002.DictionaryofHexactinellida.In:HooperJNA,vanSoestRWM,editors.SystemaPorifera:aguidetotheclassicationofsponges.NewYork:KluwerAcademic/Plenum.p1224TeragawaCK.1990.Mechanicalfunctionandregulationoftheskel-etalnetworkin.In:RutzlerK,editor.Newperspectivesinspongebiology.3rdInternationalConferenceontheBiologyofSponges,1985.Washington,DC:SmithsonianInstitutionPress.pTokyoCinema.1996.Lifeofthefreshwatersponge.Tokyo:TokyoCin-gouboffG.1942.Contributiona`laconnaissancedeslarvesplanc-toniquesdponges.ArchZoolExpGen82:357UrizMJ.1983.Contribucionalafaunadeesponjas(Demospongiae)deCataluna.AnSecCiencColUnivGerona(Spain)7:1UrizMJ.1988.Deep-waterspongesfromthecontinentalshelfandslopeofNamibia(South-WestAfrica):ClassesHexactinellidaandDemospongia.In:ContributionsonthemarinefaunaofSouthernAfrica.MonografasdeZoologaMarina,vol.3.Barcelona:InstitutodeCienciasdelMar(C.S.I.C).p9UrizMJ.2002.FamilyGeodiidae.In:HooperJNA,vanSoestRWM,editors.SystemaPorifera:aguidetotheclassicationofsponges.NewYork:KluwerAcademic/Plenum.p134UrizMJ,CarballoJL.2001.Phylogeneticrelationshipofspongeswithplacochelaeorrelatedspicules(Poecilosclerida,Guitarridae)withasystematicrevision.ZoolJLinnSoc132:41UrizMJ,MaldonadoM.1995.AreconsiderationoftherelationshipbetweenpolyaxonidandmonaxonidspiculesinDemospongiae:newdatafromthegenera(Porifera).BiolJLinneanSoc55:1UrizMJ,RosellD,MartinD.1992.ThespongepopulationoftheCabreraArchipelago(BalearicIslands):characteristics,distribu-tion,andabundanceofthemostrepresentativespecies.PSZNMarEcol13:101UrizMJ,TuronX,BecerroMA,GaleraJ.1996.Feedingdeterrenceinsponges.Theroleoftoxicity,physicaldefenses,energeticcontents,andlife-historystage.JExpMarBiolEcol205:187UrizMJ,TuronX,BecerroM.2000.SilicadepositioninDemosponges:spiculogenesisinCrambecrambe.CellTissueRes301:299UrizMJ,TuronX,BecerroMA.2001.MorphologyandultrastructureoftheswimminglarvaeofCrambecrambe(Demospongiae,Poecilo-sclerida).InvertBiol120:295UrizMJ,TuronX,BecerroM.2002.Silicadepositionindemosponges.In:MullerWEG,editor.Siliconbiomineralization.Progressinmo-lecularandsubcellularbiology.Berlin:Springer.VaceletJ.1998.Planktonicarmouredpropagulesoftheexcavating(Porifera):Demospongiae)arelarvae:evidenceAlectonawallichiiA.mesatlanticasp.nov.MemQueenslMus627VaceletJ,Boury-EsnaultN.1995.Carnivoroussponges.Nature(Lond)373:333VaceletJ,UrizMJ.1991.Decientspiculationinanewspeciesof(Merliida,Demospongiae)fromtheBalearicislands.In:ReitnerJ,KeuppH,editors.Fossilandrecentsponges.Berlin:Springer.p170VogelS.1974.Current-inducedowthroughthesponge,.BiolBullMarBiolLabWoodsHole147:443VolcaniBE.1981.Cellwallformationindiatoms:morphogenesisandbiochemistry.In:SimpsonTL,VolcaniBE,editors.Siliconandsiliceousstructuresinbiologicalsystems.NewYork:Springer.pWaddellB,PawlikJR.2000.DefensesofCaribbeanspongesagainstinvertebratepredators.I.Assayswithhermitcrabs.MarEcolProgSer195:125WeissenfelsN,LangenbruchPF.1985.LangzeitkulturenvonSummen(Porifera,Spongillidae)unterLaborbedingun-gen.Zoomorphology105:12WilkinsonCR,GarroneR.1980.Ultrastructureofsiliceousspiculesandmicrosclerocytesinthemarinespongebulariairatan.sp.JMorphol166:51WulffJL.1994.SpongefeedingbyCaribbeanangelshes,trunkshes.In:vanSoestRWM,vanKempenThMG,BraekmanJC,editors.Spongesintimeandspace.Biology,chemistry,paleon-tology.Rotterdam:AABalkema.p265WulffJL.1997.ParrotshpredationoncrypticspongesofCaribbeancoralreefs.MarBiol129:41YoungCM,ChiaF-S.1987.Abundanceanddistributionofpelagiclarvaeasinuencedbypredation,behavior,andhydrodynamicfactors.In:GieseAG,PearseJS,PearseVN,editors.Reproductionofmarineinvertebrates.PaloAlto,CA:BlackwellScienticandBoxwoodPress.p385YourassowskyC,RasmontR.1983.Thedifferentiationofsclerocytesinfresh-waterspongesgrowninasilica-poormedium.Differentia-tion25:5SILICEOUSSPICULESANDSKELETONFRAMEWORKS