FACIES 1291 3-40 PI. 1-8 110 FigS·1 1 ERLANGEN 1993 1 Modern Cryptic Microbialite/Metazoan Facies from Lizard Island (Great Barrier Reef, Australia) Formation and Concepts . Joachim Reitner, Berlin KEYWORDS: ORIGIN OF MICROBIAL CARBONATES - REEF CA VES - PELOID FORMA- DFG-Schwerpunkt TION - BIOFILMS - SPONGES - GREAT BARRIER REEF - RECENT BIOGENE SEDIMENTATION CONTENTS Sununary 1 Introduction 2 Location of studied reef areas and sampie sites 3 Material and methods 3.1 Material 3.2 Methods 3.2.1 In vivo fluorochroming of hardtissue of sponges and microbialites Non fluorochroming stainings Fixation of sponges and microbialites Sampie preparations Hardpart microtome 3.22 3.2.3 3.3 3.3.1 3.3.2 3.4 Electron microsopy (TEM. SEM) Experiments within reef caves and aquaria Definition of "Microbialite" 4 4.1 Previous studies on Microbialites - a short overview 5 Cryptie microbialites of the Lizard Island Seetion 5.1 Spatial facies distribution within shallow water reef caves of Lizard Island and North Direction Island 5.2 5.2.1 5.2.2 5.3 Strueture of cryptie microbialites Outer surfaces Vertieal facies sueeession Interpretation 6 6.1 Microbialite formation 6.2 Microbialite mineralization model based on organie maeromoleeules Ca1cifying soluble and insoluble matriees within micro- eavities and poeket-like struetures 6.3 In situ peloid formation 6.4 Sponge tissue diagenesis - micropeloid formation 6.5 Biofilms and mierobes 6.5 .1 Caleified microbes 6.5 .2 Fe/Mn biofilms 6.6 Geochemistry and stable isotopes 6.6.1 Geochemical parameters 6.6.2 Cements 6.6.3 Stable isotopes 7 Proposedrelatioship between "bacterio-sponges" and biofilms of cryptie microbialites 7.1 Baeteria in sponges 7.2 Heterotrophie baeteria within studied eave sponges 7.2.1 Proposed funetion of symbiontie heterotrophie baeteria 7.2.2 Hypothetie relation- ship between "bae- terio-sponges " and biofilms 8 Problems of mierobialite formation - the alkalinity question 9 Are modem mierobialites from Lizard Island a model for aneient sponge reefs? - a perspeetive 10 Conc1usions Referenees SUMMARY ..... Rlff-Eyolutlon und Krelde-Sedlment.tlon From shallow water caves of fringing reefs related to continental islands of the Lizard Island Section thrombolitic micritic microbialites were observed. The microbialites ex- hibit always a light decreasing facies succession. The succes- sion starts with a coralgal community and ends with light independent microbial biofilms and benthos (coralline spon- ges). The sessile mineralized benthos community is construct- ed of crustose foraminifera, serpuli$, thecidean brachiopods, bryozoans, and coralline sponges. The observed benthic com- munity is very similar 10 thoseone bbserved in cryptic habitates of Aptian and Albian reefs of northem Spain. For longtime studies of the m~crobialite formation and growth rates of coralline sponges the specimens were stained in vivo, within their natural habitat with histochemical fluorochromes and nonfluorescent agents. Main results are a very slow growth of the microbialite and associated sponges (50-100 Ilm/y). Only few calcifying microbes are participa- tors during microbialite formation. Calcifying acidic organic macromolecules are mainly responsible for microbialite for- mation by cementing detritical material. FelMn-bacterial biofilms are responsible for strong corrosion of the microbialite. Beside the corrosive activity of the Fe/Mn-bacterial biofilms boring sponges (Aka, Cliona) are the main destructors. Address: Doz. Dr. J. Reitner, Freie Universität Berlin Institut für Paläontologie, Malteserstr. 74-100, Haus D, D-12249 Berlin; Fax: Germany-030-7762070 4 Locations of studied Reefs Eagle Islet (sand cay~~ [) Byrie Ree( ~-'~ , "!:ll M"""';~, "" ~izard Island (Le: 1450 2S'E; La: 140 41'S) ~ North DirectiOD (SloDd t N The investigations began in 1990 with a pilot study in fringing reefs of Lizard Island (Great Barrier Reef) and reefs of the outer barrier for comparison. Main goals are to study this environ- ment under normal conditions and for compari- son under controlled artificial conditions in seawater running aquaria. Two reef caves were selected for longtime research and all experi- ments were carried out within these caves. H.ZANKL(Marburg)hasrecentlyfoundnearly similar microbialites in cryptic habitates of reefs of St.Croix (US Virgin Islands, central Carib- bean realm) (ZANKL 1993). Comparative fossil examples were studied by NEUWEILER (1993) on middle Albian reefs of northem Spain and KEupp et al. (this volume) for Jurassic ones. Both studies refer to the presented data. Purpose of this study is to demonstrate the growing procedure of a modem type of cryptic, light independent microbialite, the interaction with associated benthos, and its significance as a key facies to understandfossil metazoan-micriticl microbialitereefs. To understand these very com- plex processes it is neecessary to review and compile modem concepts of biomineralisation, biofilms, and e.g. the alkalinity question. There- fore mixing of own results and reviews is planned and intended! Fig. 1. Location of studied reefs of the Lizard Island Section. Lizard Island is situated in the Oreat Barrier Reef (Australia) at Le: 1450 28' E; La: 140 41 ' S. Main goal of this study is creation of a work- ing hypothesis to understand the processes of formation of micritic/microbial build ups ("mud mounds"). Geochemically the observed microbiali tes are composed of mainly high-Mg calcites and exhibit high positive ö13C (+3 to +4) values. 1 INTRODUCTION The presented study of modem micritic crusts in reef caves is part of an investigation of surviving Cretaceous benthic communities in light degreased environments of reef habitates. Within the simultanously studied Lower Cretaceous reefs, in northem Spain (REnNER 1982,1986, 1987) we have observed in all cases a elose relationship of throm bolitic or stromatolitic micrite crusts with a benthic community char- acterized by coralline sponges. This relationship of micro- bialite crusts with sponges and further sessile benthic organ- isms was observed in shallow water cryptic habitates and within larger reef mound struc tures in deeper parts of the reef slopes ("sponge reefs"). The investigations of modem occurrences in reef caves which may represent a facies telescoping of deeper biofacies in shallower ones, could be a key to understand the formation of fossil sponge reefs. 2 LOCATION OF STUDIED REEF AREAS AND SAMPLE SITES The work was carried out at l:izard Island Research Station on Lizard Island (Le 145"28E; La: 1400 41'S) (Fig. 1). The reefs studied are located jn the northem part of the Great Barrier Reef (GBR) around Lizard Island within the Lizard Island Seetion of the GBR Marine Park. The fringing reefs ofLizard Island are the main reef areas studied (Fig. 2). The investigations are focused on two reef caves within the fringing reefs of the Lizard Island Group. One of the seleeted reef caves is located at the southern corner of South Island in a water depth of 6-9m, and the second selected reef cave is located elose to Bommie Bay on the northeastem side ofLizard Island in a water depth of ca. 10m (Fig. 2). Within these two caves all life habitat controlled longtime experiments are carried out. Additional investigations were made in reef caves of the "Washing Machine" , Pidgin Point, and Crystal Beach (Fig. 2). In the surroundings of the Lizard Island Group, reef caves were studied at North Direction Island, Eyrie Reef, and MacGillivray Reef ("Macs-Reef') (Fig. 1). Additional 5 the cave walls were la~led also to study the re- generation of the disturbed parts. The inital growth of microbialites and settle- ment of biofilms and benthos were studied beside the regeneration of disturbed parts using artificial substrates. Crystal Beach From 160 selected specimens more than 500 rock thin-sections and histological sections were made. All specimens including the sections are deposited in the Institut für Paläontologie der Freien Universität Berlin. Pidgin Point 3.2 Methods 3.2.1 In vivo fluorochroming of hardtissue of sponges andmicrobialites In order to study growth and calcification rates of living specimens of two species of coralline sponges, the chaetetid Spirastrella (Acantho- chaetetes) wellsi and stromatoporoid Astrosclera willeyana, and associated calcified microbialites SE Trade Winds were stained with different types of fluorochromes in vivo. Research Caye Fig. 2. Location of investigated reef caves around Lizard Island. studies were carried out in reef caves and overhangs of the Outer Barrier. The studies were focused on Hicks Reef, Hilder Reef, Jewell Reef, and Ribbon ReefNo.10 (fig.l). 3 MATERIAL AND METHODS 3.1 Material All specimens (organisms and rock sampies) used for this study were collected at the sampie sites and prepared for further investigations. In the caves the sampies were laken from recognizable different facies types. All selected speci- mens were photographed and documented by using under- water video films. The investigations were focused on coralline sponges (Astrosclera willeyana, Spirastrella (Acanthochaetetes) wellsi, Vaceletia crypta, Murrayona phanolepis, and Plectroninia neocaledoniense) and micro- bialites. Sp. (Acanthochaetetes) wellsi and Astrosclera wi- lleyana are the most common taxa and therefore selected for longtime experiments. Within the caves 5-10 specimens of each species were selected and labeled for in situ experi- ments. From different types of microbialites sampIes were laken using a big (1.5 kg) hammer and the damaged parts on The specimens were stained principally with two types of fluorescent dyes, with the antibiotics tetracycline base (Achromycin; SERV A 35865), tetracycline HCL (Achromycin-HCL; SERVA 35866), 7-Chlorotetracycline-HCL (CTC, Aureo- mycin; SERVA 14200),and with theindiCatorsfor metal titration (fluormetric determination of Cal- cium) calcein (MERK2315, C30H26N2013),and calcein-Na2 (FLUKA 21030 C3oH24N2Na2013)' In all cases high concentrations of 250 mgll sea- water were used. The fluorescence dyes were in- jected into the living portions ofthe specimens, and they were covered with a plastic bag for an incubation time of 10-15 min. only. The used fluorochromes are Ca2+ binding chelat com- plexes which are bound on the surfaces of the newly formed Ca-salt crystals. Calcein has for example 5 free carboxyl- groups (COO-) which bind Ca2t . They allow a polychrome fluorescence labeling of the calcareous skeletons and microbialites. The polychrome labeling allows to calculate , the growth rates and to recognize the loci of mineralization! Calcein was first used as an intravital fluorescence tag to study the growing mammal bones (RAHN 1977). WILLENZ & HARTMAN (1985) has used it with success to study the growth rates of the aragonitic coralline sponge Ceratoporella nicholsoni from Jamaica (Caribbean). They used a concen- tration of l00mg/1 and an incubation time between 12 and 24 hours and treated the specimen again after one week and six months. The sponge exhibited a growth rate of200 ~/a. We have done it more or less in the same manner, however, we used higher concentrations and less incubation time, but the results are similar. Especially the labeled specimens of the high-Mg calcite S. (Acanthochaetetes)wellsi exhibitastrong green fluorescence. Same results were observed within the microbialites, but a distinctive growing was not observed, 6 however, the loci of calcifications were localized. The same procedure was performed with still living specimens in aquaria. The aquaria experiments were not directly comparable because the sponges were too stressed due to rapid temperature change. The calcification rates of the aquaria specimens were therefore much higher (1-3 j..lIll/ week) as observed in the natural habitat (1j..lIll) probably caused by increased stressed Ca-detoxification. To localize the loci of mineralization, specimens fixed with buffered glutaraldehye and stored in 70% ethanol can also be used. The Ca-binding proteins with seed crystals exhibit a similar fluorescence behavior which is different from any autofluorescence of different tissues types. Best results in shorttime staining (15min) of living specimens within their natural habitat we got with Calcein- Na2 fluorochrome «(FLUKA 21030 C30H24N2 Na2013» (100mg/l). The yellow/green fluorescence is very prominent and marks few newly formed calciumcarbonate seed crys- tals. The second group of fluorochromes which work suc- cessfully are different types of the antibiotic tetracyclines. Tetracyclines are used as antibiotic agents and were the first fluorochromes which allow a staining of Ca-bound hardtissues in vivo (MILcH et.al 1958). The fluorescence behavior was discovered because the growing teeth of young children which had got tetracyclines as drugs exhibit under UV -light a yellow fluorescence. Tetracyclines are used widely for calcification and Ca- binding protein studies in vivo (FINERMAN & MILCH 1963 binding of tetracyclines to Ca; KOBA Y ASHI & T AKi 1969 sea urchins; BARNES 1981 corals). Aureomycin (CTC) was used to study intra- and extracellular Ca-ions especially to show the role of Ca during e.g. the moving of amoebae (GA WUIT A & STOCKEM 1980) and the growth of algae (HAussER & HERTH 1983). For the here presented study the tetracycline-HCL and Aureomycin (CTC) exhibit the best fluorescence results. Especially the calcification areas of the microbialites could be detected with tetracycline-HCL. Aureomycin shows a strong fluorescence within newly formed skeletons of S. (Acanthochaetetes). The main problem with these dyes is their antibiotic character. Sponge tissues are, as a rule, penetrated with symbiontic bacteria which play an impar- tant role in the metabolic pathways of a sponge. The tetracyclines are disturbing or inhibiting this procedure. Wll.LENZ & HARTMAN (1985) had no results using tetracyclines and they argue that the staining for more than 12 hours would be lethal for the symbiontic bacteria. Therefore we have used tetracyclines only for shorttime (10-15 min). Normally the staining results are visible within some minutes after incuba- tion. The water flow through the spange canal systems guarantees a rapid transport to the calcification loci. The microbialites do not show this effect because mi- crobes do not playa very importantroleduring calcification. The yellow/golden fluorescence is easy to distinguish from the autofluorescence of different tissue types. The fluorochromed specimens were studied with an epifluorescence microscopeZEISS Axiophot. The filter sets used for calcein and tetracyclines are: high-performance narrow-band pass filter BP 365/12 nm, LP 397 nm (UV- light, no. 487701); high performance wide-band pass filter BP 395-440, LP470 (blue violett, no. 487705). To study acidic mucus substances (mucoproteins, glyco- proteins, acidic mucopolysaccharids (= proteoglycanes = glycosaminoglycans & proteins; glycosaminoglycane= Uron acid + aminosugars) which play an important role during biomineralisation, the fluorochrome acridine orange and non-fluorescent thiacine stainings (toluidine blue, methyl- ene blue) were used. Acridin~ orange is an important fIuo- rescence dye to recognize bacteria. Some specimens were stained with Acridineorangeand ca-chelating fIuorochromes ( calcein, tetracyclines) to localize ca2+ concentrations within the acidic mucopolysaccharids. Acridine stainings were done with living tissues and microbialites as weIl as with glutaraldehyde/formol fixed ones. Acridine organe exhibits an orange fIuorescence using a high performance wide-band pass filter (blue,BP450-490,LP 520; no.487709) (STRUGGER 1940, H!CKS & MAITHAEI 1958). Living sponge tissues (proteins) including collagenous supporting skeletons are easy to determine with the fluorochrome thiacine red. Using this dye, it was easy to recognize the very thin sponge crusts (50j..lIll) which grow in close contact with microbial biofilms, and which exhibit a strong staining behavior using acridine orange andlor toluidin blue, and methylene blue (acidic mucopolysaccharids!). Thiacine red exhibits a strong red fluorescence using a high performance wide-band pass long wave filter set (green, BP 515-565/LP 590; no. 487714) or in combination with acri- dine orange (blue, BP 450-490/LP 520) or in combination with tetracyclines (blue violet, BP 395-440/LP470). In all cases autofluorescence behavior of different tis- sues and minerals gives additional information beside the artificial fIuorescence. Many substances exhibit a strong autofluorescence as chlorophyll, collagen, and organic mac- romolecules enclosed in newly formed crystals (DRAVIS & YUREWICZ 1985). The best way to study autofIuorescence is to use the longwave high-performance narrow-band pass filter BP546/12, LP590 (green, no. 487715). The fluores- cence is red. The main problem is to get good micrographs. It is useful to reduce the shutter timbialite was stained with red fuchsine and differentiated with methylene blue. This staining indicates a reacti ve protein dominance (red) and a reactive sugar dominance (blue). The intensity of the color is value of the maturity of the crust. The non- r stained portions are completely cemented. Scale 2 mm Thin section of a top area of a growing microbialite under dark field conditions. The uppermost portion is incompletely cemented and exhibits a cloudy structure. The interparticle space is filled by organic mucus. The particles are bind allochems. The bright area is an entirely cemented crust. Scale 2 mm Growing mode of internal parts of a fixed microbialite. The Mg -calcitic crust is bored by sponges and the pore space is filled with methylene blue stained basophilic organic mucus. The inner surfaces are covered with Fe/ Mn-microbial biofilms (arrows). All observed microbialites are structured by these Fe/Mn-crusts! The loss of the intensity of the blue color is a result of calcification events (1-4). Within the pore with stage 2 are early cemented peloids (PE) visible, a result of a direct microbial cementation. Scale 2 mm Growth mode of an upper portion of a fixed methylen blue microbialite below a thin poecilosclerid sponge crust (SPC). The small sediment pocket is internally covered with Fe/Mn-biofilms which have a corrosive potention, and a corrosion pattern is always recognizable (CP). The Mg-calicte crystal seeds are growing within the blue stained basophilic sugar-rich organic mucus (star). Scale 1 mm Same specimen under x Nicols. Plate 2 15 16 ous particles, molecules, ions, and foreign atoms are present within the solution. Tbe extraneous particles have a catalysator effect which decreases the activating energy of crystal forming ions. This process is most successful when the catal ysator (=nucelation) surfaces of the extraneous parti- cles are very similar to the forming crystal seed (discussion in SIGG & STUMM 1989). An example for abiogenic calcites formed during heterogenic nucleation is the chalky fine grained calcite in freshwater lakes ("Seekreide"). Aragonite and vaterite need organic marker molecules as extraneous particles (AnOADI & WEINER 1985; WHEELER & SIKES 1989). The heterogenous crystal nucleation is a paradigm for seed The binding ability of divalent cations by acidic macro- molecules is an assumption for crystal nucleation. Important is the formation of a flat molecular monolayer of the acidic macromolecules in polypeptid ligatures chains. The chains are linked with H-bridges and must lay on a neutral non reactive surface (AnOADI etaI. 1990). Tbe monolayer and its basement are linked chemically weakly only. The CH- chains of the monolayer are arranged in a faul ted secondary structure which are called "ß-sheet structure" (AnOADI & WEINER 1985, WORMS & WEI!'ffiR 1986) (Fig. 5). The monolayer of large acidic molecules represents the EDT A "soluble matrix", and the neutral basement is called "insoluble matrix". , crystal nucleation under control of acidic macromolecules (Fig 5, PI. 5/4,6; PI. 6/4, 5). The negatively charged free COO- groups are trapping divalent cations such as Ca2+. Tbis ability is of great significance for an organism because the process allows the elimination the cell toxic surplus of Ca2+ and a directed transport via enzyms like calmodulin and membrane bound Ca2+ -A TPase to loci where calcium is used as a physiological controlling factor. The negatively charged COO- groups of the ß-sheets may bind Ca2+. In certain cases the carboxyl groups of asp and glu in the ß-sheets are arranged distally (far from the IM) in one plane. The trapped Ca2+ ions have therefore the same direction. If the COO- groups have definite distances, the Ca2+ ions form an initial crystal plane, e.g. the (001) plane which is perpendicular to C-axis of a CaC03 crystal. Plate 3 Fig.1. Fig.2. Fig.3. Fig.4. Fig.5. Fig.6. Fig.7. Fig.8a. Fig.8b. Modern cryplic microbialile/metazoan facies of Lizard Island (Great Barrier Reef, Australia). Epifluorescence of in vivo labeled microbialites by different fluorochromes (Lizard Island, Bommie Bay reef cave, 15 m water depth). All specimens are fixed with glutaraldehyd and stored in 70% ethanol. The epifluorescense was carried out with a ZEISS Axiophot light microscopc. Calcein staincd microbialite. The calcification fronts exhibit a strong yellow fluorescence which indicates a high amount ofCa2+-ions. The color gradient represents the decrease of free Ca2+-ions (arrows). Filter 05 High-performance wide-band pass filter blue violct, BP 395-440 LP 470. Scalc 1 ()() ~m Tetracycline-HCL stained uppermost portion of a microbialite crust. The Fe/Mn-biofilms demonstrate dark brownish crusts (FEC). Below this crust a yellow and a very bright fluorescence is present which proves reactive cementation fronts and the presence of free Ca2+ -ions (arrows). The foraminiferal test (F) does not exhibit any fluorescence and is initially overgrown by a calcifying biofilm. Filter 05 High-performance wide- band pass filter blue violet, BP 395-440 LP 470. Scale 50 ~m Tetracycline-HCL stained uppermost portion of a microbialite crust. The Fe/Mn-biofilms demonstrate dark brownish crusts (FEC). The in statu nascendi calcifying biofilms overgrow a serpulid tube. The corrosive behavior of the Fe/Mn-biofilms is responsible for the irregular surfaces of the different growing stages. Tbe calcifying areas exhibit a yellow fluorescence. Filter 05 High-performance wide-band pass filter blue violet, BP 395-440 LP 470. Scale 50 ~m Aureomycine (Chlorotetracycline) stained microbialite. The yellow fluorescence ön the surface of the crust indicates Ca-rich cells and tissues of sponges (SP). Below one sponge the sedimentfilled boring cavity shows a bright fluorescence of calcifying organic mucus (CO) between allochems. The mucus may be a product of decaying sponge tissue. The nearly completely cemented areas exhibit a strong increasing bright fluorescence to the inner margins (arrows). An intense microbial activity in located there. Filter 05 High-performance wide- band pass filter blue violet, BP 395-440 LP 470. Scale 50 ~m Tetracycline-Base stained small sediment trap with a high amount of calcifying proteins. The cementation fronts are clearly visible! (arrows). Filter 05 High-performance wide-band pass filter blue violet, BP 395-440 LP 470. Scale 100~m Calcein stained microbialite with many Fe/Mn-biofilms. The fluorescence is restricted to small areas and thin vertical filamentous structures (arrow) oftraces of endolithic fungi borings. Filter 05 High-performance wide- band pass filter blue violet, BP 395-440 LP 470. Scale 50 ~m Tetracycline-HCL stained calcifying bacteria (arrows) on a Fe/Mn-biofilm. High-performance wide-band pass filter 05, blue violet, BP 395-440 LP 470. Scale 25 ~m Active Fe/Mn-biofilm. The siderocapsid bacteria form grape-like structures. The basement exhibits typical corrosion features for this type of biofilm. Normal light Same specimen as 8a stained with calcein. The living biofilm (arrow) exhibits a strong bright fluorescence. The Fe/Mn grape-like aggregates are the product of a special respiration of these microbes under aerobic to disaerobic conditons. UV High-performance narrow-band pass filter 01; BP365/12 LP 397; Scale 10 ~m Plate 3 17 18 Carboxyl groups with bind Ca2+ ions fonn an interface between the acidic organic macromolecule and the inor- ganic crystal (Fig.5). The distances of the Ca2+ ions within the (001) plane detennine the calcium carbonate mineral (4.99Ä = calcite, 4.96Ä = aragonite, 4.13Ä = vaterite) (AoDADI & WEINER 1989, WHEELER & SIKES 1989). ß-sheets of acidic macromolecules are soluble matrices which allow initial crystal plane fonnation (Fig. 5). However, crystal growth is lirilited or controlled by these very large and acidic macromolecules too (AoDADI & WEINER 1985, REITNER et aI. in press). Most of the matrix fonned crystals have strange fonns like the aragonitic nacre of the mollusks. Extremly acidic macromolecules, predominantly glycoproteins may trapp Ca2+ ions on certain crystal planes oron inhibit the formation ofOOl planes. These very acidic types are inhibitors of any mineralisation, or they allow the crystal to grow only in selected directions (furtherdiscussion in AoDADI etal.(1990), BORBA,S etal. (1991), and REImER et al. (in press). Most successful crystal formation is realized in weak to medium acidic glycoproteins which have asp and glu concentrations of 25-30%. BERMAN et aI. (1988) have compared the SM of irregular echinids with those one of Mytilus calijornianus. The Ca-binding proteins of the nacre of Mytilus calijornianus exhibit 46mol% asp+glu. This protein inhibits the crystal growth in direction of the c-axis. The SM proteins of the echinids exhibit only 29mol% of asp+glu which allow a much more regular growing of the stereom single crystals. The same phenomenon was ob- served during the high-Mg calcite formation in the spange S, (Acanthochaetetes) wellsi (REImER et al. in press). Crystal growth Very impartant for crystal growth are vertical side chains ofß-sheets which are constructed of glycosidic linked acidic saccharid-groups mostly of oligosaccharids. These sugars have negatively charged sulfate groups which are not ar- ranged in one plane. The sulfate groups concentrate Ca2+ ions, but they do not neutralize them! This phenomenon is visible in organic mucilages using Ca-bin ding fluorochromes (REITNER 1992). A spatial concentration of Ca2+ ions on the carboxyl groups of the ß-sheet occurs which fixes the cati- ons. The seed crystals are formed under control by the ß- sheet amI/or through the organisms. Rapid epitactically crystal growth is a result of Ca supersaturated seawater with HC03 - and C032- getting in contact with the crystal seeds. The inhibition ability of very acidic macromolecules, Plate 4 Fig. la. Fig.lb. Fig. 2. Fig. 3. Fig.4. Fig.5. Fig.6. Modern cryptic microbialite/metazoan facies of Lizard Island (Great Barrier Reef, Australia) Active biofilm of a microbialite surface stained with calcein. The calcifying microbes exhibit a strong yellow fluorescence (cf. PI. 5/5). Large calcifying microbes within bore hole of a sponge exhibiting a strong green/yellow fluorescence (compare PI. 5/7) la, b High-perfonnance wide-band pass filter 17 green, BP 485/20 BP 515-565. Scale 10 11m Same microbes as in fFig.l b stained with acridine orange, a microbe tracer. High-performance wide-band pass filter OS, blue violet, BP 395-440 LP 470. Scale 10 11m Peloid formation 3a. Large borehole of a sponge within a fixed microbialite. The internal surface is covered with a Fe/Mn- biofilm. The cavity is irregularly filled with Mg calcite spherulites (peloids) of various sizes. Newly formed peloids are loose and exhibit fuchsine and methylene blue stained cores, tracing proteins and complex sugars. In some cases bacteria cells are forming the peloid cores. The peloid organic cores (soluble organic matrices) are altered into a crypto-crystalline low Mg calcite. Marginally on the cores Mg 'Calcite crystals grow. The peloids grow within organic slime pockets. If the muco-substances are resorbtrd~ the peloids condense and grow together (arrow).Scale 100 jlffi • I 3 b. Detail ofFig .3a. The peloids grow on a biofilm . Peloids with dark cores are deri ved from microbes (arrow). Scale 50 jlffi 1 Spange tissue decaying features (Aka, boring spange). The living part exhibits a clear red color (red fuchsine staining). The dead parts of the sponge are separated by a prominent organic phragma (p). The boundary area of the living tissue is characterized by collagenous fibres, indicating the moving process of the tissue. Some of the spicules (S) are penetrating the phragma. The decaying tissue below the phragma exhibits a crumply structure which calcifies, may be via ammonification. Scale 100 11m Double fluorochrome stained upper partion of a microbialite with calcein and red thiacine. Yellow calcein fluorescence indicates calcifying parts below the Fe/Mn-crust. The prominent red color fluorescence is in all cases typical of sponge tissues (here a poecilosclerid thin spange crust). High-performance wide-band pass filter OS, blue violet, BP 395-440 LP 470. Scale 100 11m Small spange reefs (Zone 4, LIZ92/45) 6a. LargeS. (Acanthochaetetes) colony (Ac) is overgrown by Spongia (1), Clathrina (2), yellow poecilosclerid spange (3), Mycale (4), ascidian (5), and hydrozoans (6). Scale 1 cm 6b. Fuchsine stained thinsection of a small coralline spange reef (Zone 4). Astrosclera willeyana (AS), S, (Acanthochaetetes) wellsi (AC), and Stromatospongia micronesica (S). Scale 5 mm Plate 4 19 20 however, is limited when most of the negatively charged valences are neutralized. When this takes place, the now wealdy acidic mucus mineralizes rapidly. The process de- scribed may explain the very rapid calcifications e.g. within the polysaccharid layers ofbacteria (e.g. stromatolite forma- tion, algae, foraminifera) LOWENSTAM (1981) and WEINER et aI. (1983) dis tin- guished two different evolutionary grades of mineralisation processes. They distinguish a "biologically matrix medi- ated" and a "biologically matrix controlled" mineralization process. The matrix controlled process is the most complex and sophisticated one which forms functional skeletal ele- ments. For microbialite formation the matrix mediated miner- alisation process is the most important one which explains calcium carbonat crystal formation without direct control of organisms but under presence of reactive medium to weak acidic macromolecules. 6.2 Calcifying soluble and insoluble matrices within microcavities and Itocket-like structures All studied microbialites do not show clear horizontal laminae as observed in true stromatolitic structures, except the Fe/Mn crusts which exhibit sometimes pseudo- stromatolitic features. They are constructed of irregular pocket structures (IPS) or microcavities of less than one millimeter to centimeter size which are limited and sur- rounded by brownish ferro-manganese crusts, short fibreous Mg calcite cements, or by truncation surfaces linked with changes of the textural character of the pocket filling caused by the boring activity of sponges (pI. 2/4, 5; PI. 3/1, 5). The IPS and microcavities exhibit different generations of growth and maturity grades depending on the amount of organic substances (pI. 2/2,4; PI. 3/5). They are filled with organic mucus and with trapped and bind fine grained detritical material (PI. 2/3). It is possible to distinguish three different types of microcavity structures: I. interpillar space of the thrombolitic microbialites (pI. 2/1c), 2. fillings of the bore holes of excavating sponges (pI. 215; PI. 3/4),3. detritus bind on the uppermost growing parts of thrombolitic pillars (pI. 2(2,3). Pieces of formol or glutaraldehyde fixed microbialites were block stained with the non-fluorescent dyes basic red fuchsine, methylene blue, and toluidine blue O. Using these basophilic dyes, it was possible to stain the different organic mucilages. The staining behavior depends on the amount of organic substances and on the biochemical type. Most of the organic fillings of the pockets exhibit a red to pink staining which indicatesahighamountofacidicproteins (pI. 2/1c,2). Acidic proteoglycane (sugar-) rich mucilages stain blue or lilac, sometimes pink, if using methylene blue and toluidine blue 0 (pI. 2/2,4,5). This metachromatic effect (discussion in BOCK 1989) indicates a den se storage of negati ve valences of carboxyl and sulphat groups! This observations are very significant because acidic proteins and proteglycanes playa central role during mineralisation. All observed in mature IPS exhibit a certain staining behavior. The intensity of staining at least depends on the grade of mineralisation. Plate Fig.1. 5 Modem cryptic microbialite/metazoan facies of Lizard Island (Great Barrier Reef, Australia) Microbialite crust surface with Spirillina sp. which is initially overcrown by a microbial biofilm. Lizard Island Reef cave 15m water depth; Zone 4; (LIZ92/43). SEM; Critical Point dried (CP). Scale 50 ~m Fig.2. Fig. 3. Fig.4. Fig.5. Fig.6. Fig.7. Fig.8. Detail of the biofilm of Fig.l exhibiting various collapsed non-mineralized microbes (1), fungi myceles (2), and bound allochthonous coccoliths (3). Scale 10 ~m Biofilm surface with different unknown bacteria (1) and cyanobacterium (Oscillatoriales) (2). The biofilm is formed by a ca. 100 ~m thick mucus. Lizard Island reef cave 10m water depth; zone 4; (LIZ92/50/1). SEM, CP. Scale 10 ~ Burst rod-shape bacterium (type 11) with wine grape-shaped seed aggregates of calcite crystals located on a biofilm . The calcite seeds have a diameter of ca. 20-50 nm). This type of calcite forming bacteria are relatively I common. The chalky calcareous muds within borings are maybe linked with this type ofbacteria. Lizard Island reef cave; zone 4; (LIZ92/50/1). SEM, CP Scale 1 ~m. Spindle-shaped calcified bacteria (type I) cell colonies on a biofilm. These bacteria are common in biofilms with a high amount of fungi filaments. The sparry calcified cells are often trapped in small cavities and open bore holes. (Compare same species in PI. 4/1a. under epifluorescence). Lizard Island reef cave; zone4; (LIZ921 32). SEM,CP Detail of Fig. 5. The calcified cells are hollow and exhibit in some cases remains of the murein layer. These specimens exhibit the calcified outer polysaccharid membrane layer. The calcite sheats are formed by very small (20 nm) grains (arrow). Scale 2 ~m Calcified coccoid bacteria in a sponge bore hole (type III). This type is formed under disaerobic conditions. The holes are closed by Fe/Mn-bacteria biofilms. (Arrow: dividing cells). (Compare same species in PI. 4/1 b under epifluorescence). Lizard Island reef cave; zone 4; (LIZ92/43). SEM, CP. Scale 20 ~m Fe/Mn-bacteria biofilm. The coccoid Fe/Mn-bacteria (Siderocapsa?) are forming rows. A single coccoid cell (arrow) has a size of ca. 1-2 ~m. (Compare same species in plate 3, fig. 8a, b; under epiflurescence). Lizard Island reef cave; zone 4; (LIZ92/45). SEM, CP. Scale 20 ~m Plate 5 21 22 Young IPS with acidic organic mucilages and a packstone fabric are semidurable and exhibit a strong blue color (pI. 2/ 4, 5). Under X Nicols they exhibit only few mineralisation events (pI. 2/5a, b). If metachromatic effects show a red/pink color, the consistence of the IPS is more or less durable (pI. 2/2). Completely mineralized portions do not exhibit any staining. The newly formed micrite is honey brown. lEM micrographs of these portions exhibit a lot of remains of collagenous fibres, fibre networks, sheets and further not clearly distinctable EDTA insoluble remains which represent the insoluble matrices (PI. 6/1). Rod shaped bacteria are rarely observed (pI. 6/1). Fluorescence dyes were also used to localize the calcifi- cation fronts. The specimens were injected with Ca-chelating (=binding in chelate complexes) fluorochromes (tetra- cyclines, calcein) in vivo. Using these dyes it was possible to recognize the loci of calcification fronts in statu nascendi (PI. 3/1-7). The acidic proteins, glycoproteins, and proteoglycanes have a strong affinity to bind and trapp divalent cations especially Ca2+. They enrich these iones and within the basophilic mucilages (soluble matrices) grow seed nuclei of different Ca-salts. The used fluorochromes mask on the surfaces of the very small seed crystals Ca2+ or trapp free Ca2+ ions. The use of tetracyclines tumed out as most successfuI. The calcifying mucilages exhibit a strong yellow fluorescence, using a high performance wide-band pass filter BP 395-440nm LP 470nm (blue-violet) or a high performance narrow-band pass filter set BP 365nm LP 397nm (UV) (pI. 3/2-5, 7). Calcein exhibits with the same procedure a more or less green-yeUow fluorescence (pI. 3/ 1). The studied slimes in IPS have only a weak autofluores- cence behavior. The observations within the acidic mucilages of the IPS are very peculiar. The intense yellow fluorescence is re- stricted to spots and thin fronts (PI. 3/7, 5).These spots are areas where Ca2+ ions are extremely enriched in very acidic mucu- or glycoproteins which inhibit calcification, still the concentration is too high and the acidity decreases. The thin fluorescing fronts are areas where seed crystals are formed in statu nascendi when acidity decreases (pI. 3/5). This phenomenon was also observed within very acidic gl ycoproteins of the calcium waste chambers of the coralline sponge Vaceletia crypta (REITNER 1992, GAlITRET & MARIN 1991). The mineralizing process takes place via Ca-binding macromolecules (EDT A soluble matrices) which are rich in acidic aminoacids (asp 19.2 mol%, glu 12.8 mol%) and common insoluble matrices are mostly dislocated collagen- ous fibres (hypro4-6 mol %) (first chromatographic analyses were made from microbialites ofNorth Direction Islands by Dr.Pascale Gautret, Univ. Orsay, Paris Sud). This mineralization is a typical irregular matrix mediatied calcifaction process without any control by organisms. 6.3 In situ peloid formation Some of the microcavities are filled with peloids only, and no further allochthonous material was observed. The peloids have a mean size of 20-60 11m and they are irregu- larly distributed within the microcavities (pI. 4/3a, b). The surfaces of the small cavities are covered by active FeIMn- biofilms (pI. 4/3a). The holes are partly filled with mucus substances which exhibit a strong basophilic acidic charac- ter. The peloids grow within the mucus which exhibit a clumpy structure. The clumpy structures are the starting points of peloid formation. All observed peloids possess a core of these clumps which have a diameter of 10-20 Jlffi. The peloid cores exhibit a strong mainly blue staining, using a dye mixture of basic fuchsine and methylene blue (pI. 4/ 3b). On the surfaces of the cores an euhedral dentate Mg calcite rim is always observed which does not exhibit any staining behavior. Within medium mature peloids the or- ganic core exhibits a microcrystalline framework of anhedral Mg calcite crystals. The calcite crystals grow inside of the mucus. Using Ca-binding fluorochromes, the same fluores- cence behavior is seen as observed in the larger slime pockets (IPS). The more or less irregular distribution of Plate 6 Fig.l. I f Modem cryptic microbialite/metazoan facies of Lizard Island (Gre!lt Barrier Reef, Australia) Fig.2. Fig.3. Fig.4. Fig. 5. Structural EDT A-insoluble matrices ofthe surfaces ofLizard Island microbialites. The insoluble matrices are commonly formed by collagenous fibres and combind sheats (1). Sometimes bacleria are present (2). LIZ92/ 45; Transmisson Electron Microscope (lEM) No.37909. Scale 10 11m Biofilm with Vibrionacae-type bacteria some with loosely bound membranes intercalated in very thin fibres (collagen). LIZ92/32, TEM No.; 37383. Scale 5 11m Electronic dense FeIMn-bacteria from a FeIMn-crust (not stained with Os04!). Cell and outer membranes exhibit the same dense structures (arrows). Probably the cells are not vital. The grey portions are probably remains of the soluble matrices of FeIMn seeds produced by microbes. They are connected with a thin collagenous fibre network. (Compare PI. 5/8, PI. 3/8). LIZ92/45; lEM No.: 36599. Scale 2 11m Filaments ofunknown origin (maybe fungia) with a dense inner layer of rhombic crystal seeds (maybe calcite) and a diffuse outer layer with scattered rhombic crystals. The crystal seeds are surrounded by an insoluble organic sheet. The grey inner portions of the seed crystals are probably remains of the soluble organic matrix. LIZ92/32; lEM No.: 36596. Scale 1 11m Single rod-shape bacteria with a faulted outer membrane containing small vacuoles (arrow) which exhibit the same size and shape as seen in the calcite seed crystals in Fig. 4. Probably these vacuoles have contained metabolic Ca waste. LIZ92/45; lEM No.: 37869. Scale 500 nm Plate 6 • ~< ~ d ~ 9-. " . ~, 23 24 peloid concentrations is a result of the elumpy structures of the organic mucus. The elumpy structure may be a result of decaying of organic slimes, but no traces of bacteria were observed by using SEM and TEM analyses. The cores of the observed peloids are definitely no bacteria remains! The insoluble matrices are very small (> 11lJl1) electron dense particles of unknown origin. The peloid problem is widley discussed and variable hypothesis exist. MACINTYRE (1985) has summarized the different ideas on the origin of peloidal formation. He favors an abiogentic chemical precipitation of Mg calcite peloids. MARSHALL (1983) discussed the possibility that Mg calcite crystal nucleation takes place when seawater is passing through cavities and the crystals rapidly grow on these nuclei to form a microcrystalline fabric (peloid cores!). The often observed geopetal structures of peloids were inter- preted as a mechanical sorting caused by flow regimes. MARSHALL (1983) and MACINTYRE & MARSHALL (1988) have discussed peloid nuclei settling down to the floor of the microcavity which may explain the geopetal filling. How- ever, the so-called geopetal filling is in most cases an artefact which results from the clumpy structure of the organic mucus and the insoluble matrix. The newly formed peloids were bound by remaining mucilages and often concentrated on the roof or on the side walls of a cavity wh ich have the function of insoluble matrices (pI. 4/3a,b). The binding mucus becomes sometimes calcified, and a dense crusts of cryptocrystalline Mg calcite is precipitated. More or less geopetal fillings with many unconformities and crypto- crystalline crusts are observed within larger microcavities. However, these cavities are filled with calcifying mucilages and no water movement is indicated. CHAFETZ (1986) discussed a different model and inter- preted the marine peloids as a product of bacterically in- duced calcification. He has used etched surfaces for study and came to the result that bacteria are responsible for peloid formation. Main arguments are the presence of bacterial fatty acids (LAND & GOREAU 1970, LAND 1971 have found saturated, strait-chain 16-carbon acid which is characteristic for bacteria) and brownish color of the peloid cores an indicative of organic matter. His conelusions are very elose to my recent observa- tions. But the peloid cores do not exhibit bacteria remains according to TEM histology, SEM analyses, and staining. The observed presence of bacteria fatty acids is not contra- dictory. Most of the observed slimes are a product of bacterial decaying processes. The cores of marine peloids are formed within clumpy basophilic (acidic) organic mucilages (mucoproteins). The later euhedral Mg calcite rim grows under the presence of remaining Ca-binding mucus epitactically on the peloid core. The peloid formation is also a product of matrix medi- ated crystallisation. 6.4 Sponge tissue diagenesis - micropeloid formation The hardground type microbialites are intensivly bored by sponges, mainly (ca. 90%) by Aka cf. coralliphaga which forms larger cavities of some millimeter to lcm centimeter size. Different types of Cliona (e.g. cf. viridis) produce smaller cavities often arranged like pearl chains. The bore holes of the sponges are settled by the boring sponges itself or occupied by a secondary settlement of sponges, mainly of poecilosclerids, hymedesmiids, and axinellids, but never by coralline sponges or Calcarea. In many cases the sponge tissues exhibit a different staining behavior if using basic fuchsine. The bore holes are often separated by intensively stained organic phragmas which separate the living part from the dead part of the sponge (pI. 4/4). The phragmas are normally penetrated by the spicules of the sponge which indicate that the dead portion is related to the sponge. The living cellsofthe sponge tissue exhibit a normal fuchsine red staining, whereas the dead portions of the sponge are stained in a dark brownish/ P I a te 7 Modem cryptic microbialite/metazoan facies of Lizard Island (Great ~arrier Reef, Australia) I Fig. l. Spirastrella (Acanthochaetetes) wellsi (HARTMAN & GOREAU 1975) la. Large collagenous fibres forming cells with many huge vacuols filled with reserve material (Icg-cells) (1) ofthe lower dermallayer. Boundary between the decalcified Mg-calcite basal skeleton (2) is marked by a baso- pinacoderm (3). The pinacoderm is penetrated by collagenous anchor fibres (4) . Within the basal skeleton a boring facultative heterotrophic cyanobacterium of the Anabaena-group is common (5). Within the mesohyl collagenous fibrenetworka very small rod-shape bacterium is common (6). LIZ92/8; TEM No.: 36637. Scale 10 Ilm and 2 llJl1 Ib. Very sm all bacteria (6 in la) within the intercellular space (mesohyl). SEM, ep. Scale 500 nm Fig.2. Astrosclera willeyana LI STER 1900 2a. Choanosom with small choanocyte chambers (C) and bacteria within the intercellular space (B). LIZ92/ 45; TEM No.: 35948. Scale 5 Ilm 2b. SEM micrograph of 2a. (C: choanocyte chambers; bacteria (B». SEM, CP. Scale 21lm Fig. 3. Vaceletia crypta (V ACELET 1977) 3a. Choanosom with medium sized choanocyte chambers (C) and many intercellular different bacteria (e.g. Vibrionacae) which are probably facultative anaerobe. (p: baso-pinacoderm); LIZ9O/320; TEM No.: 35932. Scale 5 llJl1 3b. SEM micrograph of 3a. SEM, CP. Scale 5 Ilm Plate 7 25 ' . Q o \ , " , .-." B • , , • • ~ ~ , @ • • A 26 red manner. Under x Nicols these portions exhibit a strong birefringence of newly formed calcite crystals. The sponge tissue is degraded into a irregular clumpy structure during decaying (pI. 4/4). These clumps become mineralized in the same manner as observed within the peloid formation. Using thiacine staining dyes as methylene blue and toluidine blue 0, the decayed organic matter is stained into a dirty blue and exhibits therefore a strong basophilic behavior. The calcify- ing soluble matrix is extremely enriched with sulphated organic matter (proteoglycanes, glycoproteins) which con- centrate Ca2+ ions. An important source for Ca2+ ions in this particular case are lysing cells of the sponges and bacteria. The mineralisation takes place under aerobic and disaerobic to anaerobic conditions. Fe/Mn-bacteria films at the side walls of the cavities are common. The soluble matrices of acidic proteins and sugars initiate the seed crystal formation within the small glumpy structures. Insoluble matrices are still preserved collagen fibres of the sponge mesohyle and may be murein sheets ofbacteria. The decaying proteins via bacteria allow a supporting further calcification process via ammonification (BERNER 1 %8). Ammonification is the proc- ess by which ammonia is released from the organic matter by microbes. Ammonification occurs under aerobic and anaerobic conditions. Decaying proteins increase the car- bonate alkalinity as a result of ammonia and C02 due to bacterial decomposition (simplified reactions: 2NH3 + C02 + H20 > 2NH4+ + C03~-; NH3 + C02 + H20 > NH4+ + HC03 -). The sulfate-groups of the glycoproteins are partly reduced to sulfide via sulfate reducing-bacteria 2CH20 + S042- > H2S + 2 HC03-). Pyrite is sometimes observed within axial canals of sponge spicules and very rarely within the decaying tissue. This indicates occasional anaerobic conditions. The increased alkalinity supports the calcifica- tion. Calcification via ammonification may result by follow- ing simplified stoichiometric reaction: Ca2+ + C02 + 2(N) + 6 (H) > CaC03 + 2NH4 + Matrix mediated seed crystal formation and calcifica- tion via ammonification is one model to explain the calcified decaying sponge tissues. The endproduct is an irregular clumpy micropeloidal structure which is very common within the studied microbialites and in ancient sponge reefs! The spicules of the recently decaying sponges are partly altered into calcite! It is now evident that the clumpy peloids, which are missing the characteristic euhedral Mg calcite rim of the "normal" peloids, are in situ <;alcified remains of decaying sponge tissue! (but only partly comparable with the "Verwesungsfällungskalk" ofFRITZ 1958). 6.5 Biofilms and microbes It is taken for gran ted that microbes play an important role in forming a microbialite structure. In all publication dealing with this topic it is obvious to the authors that microbes are directly responsible for carbonate crust forma- tion, e.g. JONES & HUNTER (1991) discuss that filamentous cyanobacteria are responsible for microbialite formation, CHAEFE1Z & BUCZYNSKI (1992) have shown that.non thylakoid bearing bacteria play central role by the lithification of intertidal microbial mats. KlmMBEIN (1976, 1979), GERDES & KRUMBEIN (1987), and BURNE & MOORE (1987) listed differ- ent ways and types of microbial rock formation. In most cases phototrophic procaryotes are made responsible as a main resource of calcium carbonate formation. The role of bacteria which induce the precipitation of calcium carbonte is discussed since the first decades of our century (e.g. DREW 1911, KELLERMAN 1915, KELLER MAN & SMlTH 1914). More recentlY the investigations were focused more on artificial bacterial cultures to study the pure CaC03 products (e.g. KRUMBEIN 1974 , MORITA 1980, GREENFlELD 1963, CARROLL et a1. 1966, NOVITSKY 1981, BUCZYNSKI & CHAFETZ 1991, P 1 a t e 8 Modem cryptic microbialite/metazoan facies of Lizard Island (Great ~arrier Reef, Australia) Fig. 1. Murrayona phanolepis KlRKPATRICK 1910 I I" la. View on the broken surface of the calcinean sponge with dermal plate spicules (1), inner dermallayer with triradiate calcitic spicules (2), and choanosomal high-Mg calcite basal skeleton (3) . LIZ90/284; SEM. Scale ~~ I 1 b. TEM seetion of the boundary between the choanosom (1) and the decalcified basal skeleton (2) with fibres. The boundary layer is formed by many empty membrane bounded vesicles (3) (partl y fixation artefacts). Some of the vesicles are dark (4) and exhibit bacteria affinities. LIZ92/8; TEM No.: 36741. Scale 5 ~ lc. Detail of a dark vesicle from Fig. 1 b which is probably a special bacterium. The outermembrane is faulted, and exhibits sometimes empty small vesicles. TEM No.: 36756. Scale 500 nm Id. SEM micrograph of the outer membrane of the special bacterium exhibiting very small Mg-calcite granules (20-50 nm) (arrow) which have the same as the empty membrane vesicles of fig. lc. SEM. Scale 500 nm Fig.2. TEM seetion of a very thin (100~) poecilosclerid sponge crust which shows similarities to biofilms. These morphological types of "bacterio-sponges" are very common in zone 3 and 4 within cryptic reef caves. (p: endopinacocyte with a large nucleus); LIZ92/45 zone 3; TEM No.: 37903. Scale 3 11m Fig.3. Discodermia sp.; lithistid demosponge 3a. Dermallayer with phyllotriaene dermal spicules. LIZ92/32, zone 4; SEM; CP. Scale 100 ~ 3b. Common rod-shape bacteria within the intercellular space (mesohyle). Scale 211m Plate 8 27 ;'~ , , - '" . ~ . ( t , -. , ,..[ 2 · .~ .,. I' - --~ () ",. 'I ~ ~ r ,$P \. (f1 , , -' ~ "> .., 6 4· ,..... t;-~ 28 CIIAFETZ 1986, and many others). Many of the observed calcified products of bacteria exhibit a dumbbell structure with brnsh crystals with a size between 10-50 ~m! Investigations on well fixed material of cryptic micro- bialites has shown that microbes playa secondary role only as mineralizing agents, and their direct influence on CaC03 fonnation is minor. Rowever, they play an important role on the acti~e (=living) surface of the microbialites and some- times in small cavities. The entire surfaces of an observed types of microbialites are initially covered by variably diverse microbial single- and multilayered sheets. These microbial sheets exhibit a vertical structure known from so-called "bio films " described by WILDERER & ÜlARACKUS (1989). Biofilms are complexe structures of microorganisms which nonnally occupy the entire surfaces sometimes in a patchy distribution also. They can have a thickness of 100 ~. A biofilm-reactor is a self- organized system compartment in which the microorgan- isms have a high ability to survive and grow. Biofilms have always a certain vertical structure and consists of microbial cells which are embedded in an organic polymere matrix (mucilages, slimes) fonned by microbes. Biofilms are prin- cipall y two-Iayered and they need in most cases an imperme- able substratum of solid material to settIe upon (Fig. 6). On the substratum, e.g. a lithified microbialite crnst, a basis layer is located which is called base film (Fig. 6). This layer is more or less firm and contains microbial cells, exo- enzymes, different types of bind allochthonous particles, and extracellular polymeric substances (EPS) which hold the cells together on the substrate. The EPS have a central function, because this is an open porous system filled with moving water through which liquids and particles of a certain size can move and be transported. Larger particles are entrapped this way in the base film. Within the base film the metabolic processes happend. The second layer is called surface film (Fig. 6) and normally exhibits a very rough topography. The surface film is an interface between the base film and the surrounding water or bulk liquid in the sence of ROSENBERG (1989) and WILDERER & ÜlARACKLIS (1989). The surface film exhibits a more or less firm portion similar to the base film which shows the rough topography. Between the the more or less firm base film extrusions a liquid phase is present which contains biofilm metabolites, microbial cells, inert substances, and at least allochthonous particles. The bulk liquid or nonnall y surrounding watennass compartment supplies the biofilm reactor via eddy diffu- sion. In special cases the substratum is semipermeable and porewaters which are moving within the substratum may supply the biofilm reactor or decayed and/or metabolic products are deli vered to the substratum space!. This process plays an important role during microbialite fonnation be- cause this may explain the transport of Ca-binding organic macromolecules to the loci of later mineralisation. Various biofilms are observed on the microbialite sur- faces and the bioactivity is visible using fluorochromes in vivo and in fixed material (pI. 3/1-3; PI. 4/1, 5). Calcifying bacteria were detected with tetracyclines and non-calcifing ones with acridine orange and thiacine red (pI. 3n; PI. 4/1 .). The very common Fe/Mn-bacteria exhibit a strong fluores- cence using calcein (UV) (pI. 3/8a, b). The great variability of the biofilms is seen when studying Critical Point or PELDRI 11 dried specimens (pI. 5). All biofilms exhibit strong fluorescence behavior, and it is possible to recognize the interaction with the substrates (pI. 3(2; PI. 4/5). It is the rnle that all observed biofilms have an interaction with the porewater systems and microcavities of the microbialite which are filled with calcify.ing mucilages. The observed biofilms are always a system of bacteria, fungi, and some- times protozoans like foraminifera. Algae are never ob- served within the dark environments. The base films are thin with a thickness of only 10-12 ~m, and they are stabilized by entrapped allochthonous collagenous fibres. The bacteria, mainly round and rod-shaped ones, are commonly linked with the fibres (PI. 6/2). Similar features are seen within the mesohyle of sponges and it is very difficult to distinguish biofilms from very thin sponge crnsts with only few choanocyte chambers per area unit (PI. 8(2). The interfaces of the base films to the liquid film are very different and not yet extensively studied. Some of the observed surfaces are knoby and very irregular to extend the surface planes. The knoby structure is caused by microbes near the internal surface (PI. 5(2). On the surface films many different often stalked microbes and fungi hyphen are present which are linked with the interior surface layer (PI. 5/2,3). Some of the microbes which are lying on the surface layer were primarly enriched in the overlaying liquid phase. The surface films have sometimes small openings (1~m) likeostias in sponges. Within the base films pores were observed too, wh ich may be identical with the "extracellular polymeric substances" postulated by WILDERER & CHARACKLIS (1989) (pI. 5/5). They are part of a pore system through which natural products as liquids and non-liquids are transported. 6.5.1 Calcified microbes Within the base film of few biofilms special types of calcified bacteria and filamentous' structures were observed. One calcified bacteria type (I)' is relatively common (pI. 5/5,6). Single specimens exhi\)ifa spindie like shape of l- I 2 ~m size. Nonnally they occur in rosettes like aggregates. Broken individuals are hollow and the cavity is covered by an organic phragma which may be a remain of the murein sheet. The complex membrane layer of the bacterium is completely mineralized by a low-Mg calcite. The calcite exhibits an extremely fine grained structure due to aggre- gates of seed crystals of 50-80 nm. This bacteria type is sometimes enriched in small cavities. The calcifaction on the membranes of a gram negative bacterium was described by MORITA (1980). Divalent cati- ons (Ca2+, Mg2+) are linked with protruding acidic polysaccharid chains of the outer membrane. The cations are reacting either with metabolic CO2 wh ich reacts in sea water e.g. with ammonia and amines to RC03- and C032- (s. ammonification) increasing the alkalinity or with RC03- andC032-from sea water. MORITA (1980) hasevidenced this process with 14C-marked aspartate. Liquid phase cells metabolites M allocht. particles exo-enzymes E I. EPS J", ooe acidic macromolecules 29 seawater basement (microbialite) Fig.6. Biofilm model after WILDERER & CHARACKLlS (1989) adapted to microbialites. The biofilm is a reactor which takes up from the ambient seawater DOC and rigid paricles via EPS channel systems. The reactor filters and produces DOM ~d DOC, e.g. waste products, and releases it back to the seawater or into the open system of the upper microbialite (microcavities, irregularpocket structures (IPS). Some products are acidic macromolecules which enhancemineralization. Organic matterreleased into the seawater is anurientsupply forclosely related filter feeding benthos as "bacterio-sponges". Another type (11) of calcifing bacterium produces fine grained calcareous particles (100-200 nm) when decaying (pI. 5/4). Thenon-calcified bacterial cellenvelope iscracked and the fine grained calcified particles surround the cell envelope. A third type (III) of calcifying rod-shape microbe was observed in microcavities (pI. 5/7). This type exhibits a strong fluorescence when using acridine orange, and this is often surrounded by brownish non fluorescent small ferroan- rich particles (PI. 4/1b, 2). This form is very large (10-15 Ilm), and the calcified body exhibits the same aggregates of seed crystals as seen in bacteria type I. The taxonomic affinities are unknown. Calcified filamentous structures are common, and they play in mierobialites of zone 3 a significantrole. The filaments are not related to eyanobacteria. They have diameters be- tween 4-15 )lffi. No traces of thylakoid membranes were observed studying TEM seetions (pI. 6). Two types of mineralized filaments were observed. Both types exhibit an internallayer with tangentially arranged erystals. One type exhibits elongated irregularly arranged aragonitic crystals in the second outer layer, and the other type exhibits small rhomboedric calcite seed crystals whieh grow in a mucilage (PI. 6/4). In TEMs the seed crystals are surrounded by insoluble matrix layers and they exhibit inside grey parts which may be remains of soluble matriees which are pro- tected during decalcification with EDT A. However, the observed ealcifing microbes play only an insignificantrole during the mierobialite formation. But the activity of the biofilms may control the transport of calcify- ing organic substances to the microcavities underneath the biofilms via the EPS systems. The particle trapping ability of the basal films within the biofilms explains in an excellent way the formation of the thrombolitic pillars of the microbialites of zone 3. The base films can be mineralized. because the main physiological processes occured there. Base films are firm mucus layers with a lot of insoluble substances, and decay- ing base films have a strong ability to mineralize as detected by Ca-binding chelating complexes. MORITA (1980) has observed under artifieial condition that decreasing amount of living bacteria eells inerease the amount of calcium carbonate precipitation. He suggested that this process is not microbial but is supported by microbial activities. 6.5.2 Fe/Mn biofilms All observed mierobialite types exhibitnumer9us brown- ish to black/brownish thin crnsts. They are more or less horizontally orientated and show a pseudo-stromatolitic strueture. Nearly all inner surfaces of microcavities and some ares of the upper surfaces of the microbialites show these crusts (PI. 3; PI. 4/3a, 4,5). EDAX and X-ray diffrac- tion analyses have shown that light brownish crnsts are rieh in Fe (goethide) and the darker ones have a significant amount of Mn (Mn oxids). They are also enriched in clay minerals. The surfaces of the Fe/Mn crusts are extremely irregular and sometimes form ing grape-like structures. High magnifications show clumpy aggregates of spherical ele- ments with a size of2-5)lffi (p~. 3/8; PI. 5/8». Using calcein stained specimens, the surf~e 'of the clumpy aggregates exhibits a strong fluorescence under UV -light (pI. 3/8). This fluorescence indicates a microbihl activity, and the microbes are seen in TEM sections as well as under SEM (pI. 6/3; PI. 5/8» . Some of the bacteria have affinities with the taxon Siderocapsa sp .. Within zone 4 of the reef caves, this type of biofilm is very common on the surface of the reef cave walls and it is responsible for the dark brownish color (pI. 11 3bAb). In this area the film is growing under oxygene rich conditions. In some eases living Fe-biofilm surfaces were detected in microcavities which are may be reduced in oxygene. Thick ferroan hydroxyid crnsts were observed in microcavities of zone 3 linked with thrombolitic facies. Mn and Mo rich portions were only observed in zone 4. All Fel Mn-erusts exhibit at their base corrosive structures (pI. 2/4; PI. 3/3,6). Parts of the calcified microbialites and organisms are dissoluted and exhibit strong eorrosive patterns ("pseudo- 30 stylolitic "). The microbialites in zone 4 have lost by this way more than 40% of the lithified volume! They exhibit an extreme dense pattern of Fe/Mn crnsts. In one 4 cm thick crnst over 100 larger isochronic Fe/Mn crnsts were counted. Corrosion is an electrochemical process characteristi- cally of all metals except noble ones caused by the flow of electrons from one metal to another if an electrolyte is present Dissolution (= corrosion) occurs at the anode and at the cathode the electrones are consumed. An important cathodic reaction in aerated solution at nearly ph 7 is an oxygen reduction. VIDELA (1989) describes the principal processes of microbiological corrosion which allows to explain the observed dissolution phenomena. The observed metal biofilms, the underlying calcium carbonate crnsts, and the sea water is a natural electrochemical cel!. The calcium carbonate crnsts have an anode and the biofilms have a cathode function. VIDELA (1989: 309, fig.4) has observed colloidal hydrated iron oxid spheres as a corrosion product after settlement of marine Vibrio bacteria on stee!. These goethide sphernles are very similar to those ones in the brownish crnsts and may be interpreted also as a corrosion product at the cathode. The calcium carbonate corrosion increases the alkalinity and may support the calcification in closely related microcavities and thin underlaying areas despite of their dissolution ability!. In all cases below active Fe/Mn-biofilms Ca2+ is enriched detected by Ca-chelating fluorochromes (pI. 3/2,6). This observation may explain the extreme slow growth of the microbialite. At present, an environmental interpretation of Fe/Mn- biofilms is still difficult. It is obvious that these crnsts occur at certain times and cover the complete surfaces of the microbialite and bury all sessile organisms (pI. 3(2). Envi- ronmental crises as eutrophic events might cause a rapid growth of this biofilm type. This idea is supported by the linked formation of crypt cells in S. (Acanthochaetetes) wellsi as a special surviving strategy. 6.6 Geochemistry and stahle isotopes 6.6.1 Geochemical paramenters The first geochemical results of lithified microbialites show a dominance of high-Mg calcites with 12-16 mol% MgC03 and 2000-2500 ppm Sr. The analyses were carried out with EDX, x-ray diffraction, and electron microprobe. Low-Mg calcites are present in bulk analyses and probably related to allochems (bivalves, brachiopods). However, some calcified bacteria sheets exhibit low-Mg calcites 100 (EDAX). Aragonite israre andrelated to allochems (ascidian sclerites, coral remains) except some calcified filaments which ex- hibit an autochthonous aragonite. Calcifying organic rich areas in microcavities or decay- ing sponges exhibit high amounts ofS andP and sometimes Ba. The sulfur is related to Ca-binding sulphate groups of glyco- proteins and P is related to calcium phosphates and free P of lysing cells. Silicon is detected in high arnounts at certain places related to spange spicules and clay minerals. Mn and Fe are enriched significantly in the Fe/Mn-microbialites sometimes associated with certain arnount Mo. 6.6.2 Cements Cements are rarely observed within the microbialites. An acicular equant aragonitic marine cement was observed in coral skeletons of the coralgal cores of the microbialite only. Within this facies a granular calcite cement is rarely present and is interpreted as an early diagenetic fresh water cement. Some microcavities exhibit a rim of short dentate Mg calcite cement which exhibit a strong fluorescence using tetracyclines (pt 3/4). Margins of some boring sponge cavities are covered with a short fibrous Mg calcite cement which may be related to the sponge metabolic activity (Ca- detoxification). However, this phenomenon has not been studied up to now. Other typical marine cements play no significant role during microbialite early diagenesis. 6.6.3 Stable isotopes For further characterisation of the calcium carbonates, the Ö13C and Öl SO ratios of some microbialites were studied. All measured data from Mg calcites of a microbialite crnst from the Bommie-Bay research cave (Lizard Island) exhibit high positiveöl3C (+3 - +3.84) andölSO (-0.37 to -1.1). The öl3C and Ö180 values expected for equilibrium inorganic precipitation of calcite in surface waters of the northern Great Barrier Reef are approximatel y +2,2 % and -1 % re- spectively. (WEBER & WOODHEAD 1970) (Fig. 7). The equi- librium values for Mg calcites must be corrected of +0,06%/ mol MgC03 (T ARUfANI et a!. 1969). Therefore, the observed microbialites are precipitated elose to the expected equilib- rium values. This is evident also for most of the asseciated coralline sponges. The high-Mg calcite of S. (Acantho- chaetetes) and minchinellid sponges (Plectroninia) as weIl as aragonitic ones (Astrosclera. Vaceletia) fall within this range (REImER 1992). The aragonitic ones exhibit higher positive values due to the fractionation during aragonite precipitation. Calculated aragonite equilibrium values are +4 to +5% Ö13C and -1.6 to -1.8%ölSO. These data must be corrected -1.8 for ö l3C and -0.6 for ölSO values (TARUfANI et a!. 1969). I f I The observed data within one in detail measured profile of a microbialite exhibit some significant chances of the isotope values (Fig. 8). The coralgal cores exhibit a mean of light Ö13C values. The coral core has a corrected value of a mean of -0.8 Öl3C and -0.43 ölSO which is characteristic of hermatypic corals. The overlaying coralline red alge have moderately positive values of Ö13C (+1). The lowermost microbialite layer has intermediate heavy carbon values. All following data are in a range between + 3 and +3.5 öl3C. The heaviest value was measured in portion where Fe/Mn-crnsts are very common (point 9 in Fig. 8). This may be a hint for eutrophic conditions. Non-lithified portions of the microbialite e.g. in microcavities exhibit relatively light öl3C values (+ 1 to + 1.5) (Fig. 7). Semi-lithified areas have moderate high Öl3C values (+2.5) (Fig. 7). The measured initially formed peloids fall also in the field of lithified microbialites (Fig.7). Lizard 'sllgd Ccpde Mjefobialjtcs Ud CO[Jlljne SponlCS Stable Carbog agd DXleen iSQtope distributiQg CO(Jlljge 'pogees · AS Astrose/era willeytuIQ Ae AcafUhochaetetes wdLsi M MiebineUidae Mur Mu"ayollQ phanolepis Vae Vacdetia crypta SWS Seawater standard (Heran !sland) MC Mierobialile erust CAM Coral,aI + mierobialite aust CA Coral,al frame . ~MC Red Sea mierobialite crust RULM Red Sea uDiithified mud (Braebert ci: Dullo 1991) Ce Microbialite crust of Cebu (eolI . HiJlmer/Scbolz) Cx Mierobialite erusta of St.Craix (toll. ZantJ) MUR ~ ~ 1} 8UOCI\Sed. s~s The shift from light o13C values (+1) to heavy ones (+3.5) within mature microbialites is a result of a slow calcification (McCoNNAUGHEY 1989). Sponges and micro- bialites exhibit more or less the same stable isotope fractioning behavior. The observed values are very close to an expected equilibrium. A problem is that normally biologically pre- cipitated carbonates are in disequilibrium with the ambient sea water. In coralline sponges, a moderate biological con- tral of calcification is observed, and disequilibration should occur (REITNER 1992, REITNER et al. in press). Only one species of the reef caves is in disequilibrium, the high-Mg calcite pharetronid Murrayona phaMlepis, exhibits ex- tremely light 013C (-3.5 to -4) and light 0180 (-3.5) values (Fig. 7). Thecalcification in this particularcaseoccurs under the control of symbiontic bacteria which explains the strong vital effect (REITNER 1992). The microbialite formation occurs under the control of different organic substances (matrix mediated calcifica- tion). But in all cases the calcification is extremely slow (ca. 100 f..lJll/y) and CO2 resource is from dissolved inorganic carbon (DIC) of the surrounding sea water. In all studied cases no photosynthetic fractionated car- bonates were detected except those from coralline algae (point 3 in Fig. 8). This observation fits with the histological data. RMC - .-..) ...- + +1 31 Fig. 7. Distribution pattern ö 13C and 1)180 (v. PDB) of modern cryptic microbialites from Lizard Island, St.Croix and Cebu (Philippines), and coralline sponges. - 1 -3 All published isotope data of modem microbialites have more or less the same values (KEupp et al. 1993). Isotope analyses from fossil occurrences exhibit the same distribu- tion patterns as in modem ones (discussion by NEUWEILER (1993) and KEupp et al. (1993». 7 PROPOSED RELATIqNSHIP BETWEEN "BACTERIO-SPONGES" AND BIOFILMS OF CRYPTIC MICROBIALITES 7.1 Bacteria illtsponges Very important eonstitutent of the reef eave inhabitants are thin sponge crnsts whieh eovering the living bio fIlms of the microbialites (pi. 1/4b; Pi. 5/5). The sponges have a thickness of 50-100 f..lJll only and it is very diffieult to make any taxonomie determinations. Probably most of them are new species or only poorly known taxa. More than 80% are related to the poeeilosclerida a major taxon of the Demospongiae. Calcareous sponges are rare and observed non-rigid genera are Sycon, Leucetta, and Clathrina (pi. 4/ 4a). Rigid pharetronid type species are Plectroninia neocaledoniense, P. hindei, and the calcinean Murrayona phanolepis. Coralline sponges with demospongüd affmities are the stromatoporoid Astrosclera wi//eyana, new species ofAstrosclera, the ehaetetid Spirastrella (Acanthochaetetes) 32 Mjerobjalite emst Bommje 81X Caye (juni (sland Cmst surfaee 12 Coralg.1 Frame 1 hm:!!lIinc Spongcs Lil.ard Island Sceli!!n 3,2 -1-2,9 -0,7 -I 6180 1.' 1 Jlffi) andrare ovoids (pI. 7/1a, b). The choanocyte chambers are very large, 40-50 Jlffi in diameter. Within the calcareous basal skeleton endolithic heterotrophie eyanobacterians oftheAnabena group areeommon (pI. 7/a). Completely different arethe bacteria in the calcinean pharetronid Murrayona phanolepis (pI. 8/1). This still unde- tected baeteria type plays a significant role during caleifica- tion of the Mg calcite basal skeleton. The ovoid to round shaped bacteria (1-2 Jlffi) exhibit outer membrane bound vesicles of 100 nm size in wbich caleium carbonate is seereted (pI. 8/1 c). It eannot exefuded that these vesicles are may be overprinted ftxation aI1~facts. However, in SEM micrographs small seed CaC03 erystals (pI. 8/1d) were observed, which exhibit the same size as the vesieles. The baeteria are eoncentrated at thb boundary between living tissue and basal skeleton (pI. 8/1b), and probably they are controlling the mineralization of the basal skeleton. One possiblity is that the caleified bacteria are aggregated 10- gether and build initiali, the basal skeleton. This process is evidenced by the light 0 3C( -3.5) and 0180 (-3.5) values of the skeleton whieh exhibit a strong bacterial metabolie fractioning Fig. 7). 7.2.1 Proposed funetion of symbiontie heterotrophie baeteria Only little is known about the function of heterotrophie bacteria in contrast to the phototrophic ones (WILKINSON 1978a,b, 1983; WILKINSON & FAY 1979). Most of the ob- 34 served and cultivated bacteria from sponges are facultative anaerobes able to metabolize a wide range: of components. The described ones in Ceraloporella nicholsoni, which are very similar to those studied, are able to ferment sucrose and fucose but exhibit an inability to ferment glucose which is typical for the most aeromonads. W ithin all observed Lizard Island coralline sponges high amounts of fucose and galac- tose were detected which indicates the fermenting processes (REITNER 1992). Therefore the bacteria may playa role during polysaccharid mucus formation. Important is the observation that bacteria take up dissohled amino acids (Wll.KINSON & GARRONE 1980) and bacteria are able to degrade sponge collagen (Wll.KINSON 1979). REISWIG (1981) has observed that bacteria rich sponges have a gross deficit in the energy budgets and partial carbon uptake compared with sponges poor in bacteria. This means that dissolved organic carbon produced by bacteria must be an important nu trient source for the sponge. The fermentative metabolism of the symbionts and delivered DOC allows the sponge to survive ecological crisis when the pumping rates are de- creased. SANr A VY etal. (1990) have postulated that anaerobic zones may enhance the calcification of the basal skeleton of Ceraloporella maintaining an acidic environment. Some- times 1 have observed an uptake of symbiontic bacteria by archaeocytes withinAstrosclera willeyana but in most cases the phagocytized bacteria are not identical with symbiontic ones. Comparable obersavtions for example were made by REISWIG (1990) within further sponges. He has observed in situ feeding in two shallow-water hexactinellid sponges. The one species Rhabdocalyplus dawsoni uses only dis- solved organic carbon (DOC) and the other species Aphro- callisles vaslus feeds bacteria. Both types take up organic matter in colloidal form. In conclusion the natural products of symbiontic bacte- ria in sponges play a primary role as a nutrient source, control metabolic processes, refuses metabolic waste, and may enhance calcification. The facultative anaerobic char- acter of most of the symbiontic bacteria allows the sponge to survive environmental disturbances. 7.2.2 Hypothetic relationship between "bacterio-sponges" and biofIlms The pumping activity of sponges supplies in first order thesymbiontic bacteria with dissolved organic carbon (DOC) because they are the dominant part of the sponge biomass. Sponge accumulations need therefore a Fesource of DOC and dissolved organic matter (DOM) to supply their sym- biontic bacteria. Bacteria themselves produce high amounts of DOM, and therefore the closely related external biofIlms may playa role as bacteria food producing "farms". It is obvious that the observed biofIlms on the microbialites are tightly linked with sponges (pI. 4/5). The coralline sponges are restricted to areas within the caves where Fe/Mn micro- bial biofilms are present and pharetronid calcareous sponges are linked with non-Fe/Mn-bacteria bearütg biofilms. This may explain why many investigated reef eaves are free of coralline sponges because the sponge reIated biofilms are missing. The sponge larvae settl~ment may depend on the type of biofIlms. Different types of biofilms may release biochemical signals to attract the larvae to settle down and metamorphize. This hypothesis is supported by the observa- tion that young small coralline sponges are restricted to certain Fe/Mn-biofIlms. Similar larval behaviour was de- scribed from some bryozoan and serpulid larvae by STEBBING (1972). He has observed that both organisms prefer to settle on the younger partof the large brown algaLaminaria which is covered by special types ~f biofIlms. Same experiments and observations were made by RYLAND (1959) and CRISP & R YLAND (1960). SCHOU ( this volume) describes the relation- ship between bryozans and microbial mats. The stolon settlement in the scyphopolyp of Aurelia aurita is induced bya single species of the Micrococcaceae bacterium (SCHMAL 1985). A further important aspectofbiofIlm/metazoan relation- ship is the interaction of the settled organisms with the substrate. Within some thin sponge crusts it was observed that the basopinacoderm is linked with the base film of the underlying biofilm. The interface exhibits a strong fluores- cence detecting free Ca2+ ions and the presence of acidic polysaccharid slimes. SOULE (1973) has observed that a simple acid proteoglycane is responsible for the biological adhesion of bryozoa to the substrate. Comparable observations of the relations hip between metazoans and microbial films were made within arctic deep water spiculite mats and sponge mounds of the Vesteris- banken Seamount (HENRICH et al. 1992). Observed sponge larvae, young bryozans, and serpulids settled onlyon spicules which are covered by microbial films. The internal spiculite mats are extremly enriched in biofilms and therefore an important food resource (DOC and DOM). An adapted strategy was recently observed by the author in many deep water sponges (e.g. the hexactinellidR habdocalyplus dawsoni - see above!, the demosponges Polymaslia sol. and Thenea abyssorum ) which have many extern al microbial fIlms surrounding protruding spicules. These external bacterial gardensare a very sophistigated nutrient supply! . The proposed relationship between sponges and biofIlms in cryptic reef caves is summerizect in Fig. 9. This model tries to explain the circuits betweeJ;t biofilm activity, sponge development, and microbialite formation. Form a functional viewpoint sponges and biofIlms are very closely related. This may' explain why microbes and sponges always playa central role e.g. in carbonate buildup formation since the beginning of the Cambrian (REITNER in prep.). 8 PROBLEMS OF MICROBIALITE FORMATION· TUE ALKALINITY QUESTION The investigation of reef caves in the Lizard Island Section has shown that microbialite formation does not occur in all caves studied. The microbialites occur only in fringing reefs of continental islands (e.g. Lizard Island, North Direction Island, South Island) which have an island core formed by an Late Permian granite. Microbialites were Lizard Island 5 35 Fig. 10. Simplified N-S proflle Lizard !sland Hieks-Reef o ~~~:;-==-=:--=:----==---==---===-=::;r::::;:::::=:--:=--=- Tide Mierobialites oeeur only in reefs whieh are linked with continental island (e.g.Lizard IsIand). Weathering silicates as feldspars and freshwater influx increase the carbonate alkalinity whieh favors ealeitication of mierobialites. 20m 100m 2 sea milcs Submerged Reefs also observed in mid-Holocene cores of reef islands close to the mainland of Australia (e.g. Nymph Island). The reef caves of the outer barrier are free of microbialites except in same cases when very thin crusts of different origin are presentonly. The link of microbialites with continental crust islands supports the idea that an increased carbonate alkalin- ity enhances calcification. However firstresults of alkalinity measurements exhibit no significant increase of carbonate alkalinity [HC03 -]; [C032-] in the ambientreef cave water. A slightly increased alkalinity was measured only on microbialite surfaces of zone 3 within the caves. This inter- nal alkalinity increases partly under anaerobic conditions via ammonification. We have done the measurements during the dry season only (mesurements during the wet season are in progress). It is postulated that during the wet seasan carbonate alkalinity increases by an intense weathering of silicates of the Permian granite or through input of weathered. product of silicates. The microbialites exhibit periodically a concentration of kaolinite, a clay mineral wh ich is a product of feldspar weathering. Weathering of e.g. plagioclase increases car- bonate alkalinity (KEMPE et al. 1989, KEMPE & DEGENS 1985). Freshwater with a high carbonate alkalitiy is during the wet season running into the reef cave systems and increases slightly the carbonate alkalinity which favors the calcifica- tion of the microbialites. In many supratidal places ofLizard Island where granite boulders weather the reef debris, re- mains of continental rocks, glas bottles, and beer cans are strongly cemented by thick microbialitic/micritic crusts. The reef systems close to the mainland exhibit the same type microbialitic/micritic crusts. These reefs may be influenced by higher concentrations of [HC03 -] and [C032-] by river transport from rain forest areas which exhibit a strong silicate weathering. The reefs of the outer barrier are not influenced by silicate weathering and therefore no microbialites occur in reef caves although the same biofilms and linked benthos community are present (Fig. 10). Comparable and supporting observations were made by ZANKL (this volume) from reef caves of SlCroix (Carib- bean). StCroix is a continental island and silicate weather- ing occurs. In reef caves of the Bahamas and linked deeper slope Queensland Trough 9O-120m Terrace Rhodoliths Microbialites are nearly absend in reef ~ystems without continental influx (e.g. reefs of the outer barrier). areas no microbialites of the Lizard types were observed ,(REnNER unpublished data) because no continental crust is exposed. BRACHERT & DULLO (1991) have described Lizard Island-type microbialites from deeper parts (120m) of Red Sea reefs close Port Sudan in neighborhood of crystalline basements. If this observation is confirmed that the influence of silicate weathering may be correlated with freshwater influx supporting the formation of microbialitic crusts - this phe- nomenon would be an important controlling factor for microbialite formation. NEUWEll.ER and KEupp et al. (1993) have argumented parly in the same manner. NEUWEll.ER for example has discussed that an increase of carbonate alkalin- ity via silicate weathering from the Spanish mainland by river systems, cold seeps and brines of salt solutions of diapirs, and weathering of carbonates (paleokarst) are re- sponsible for the wide distribution of microbialite reefs in the Middle Albian of northem Spain. 9 ARE MODERN MICROBIALITES FROM LIZARD ISLAND A MODEL FOR ANCIENT SPONGE REEFS? - A PERSPECTIVE Investigated microbialites and combind benthos ofLiz- ard Island are found in cryptic 1md dark reef cave systems. Similarobservations were made by JACKSON etal. (1971) in Caribbean reef caves, VACEI.ET1& VASSEUR (1965,1971) in reefsoftheislandMadagascar,andBAsll.Eetal.(1984)ofthe Enewetak Atoll. The observed1benthos community is com- pletely different from the normal photic shallow water reef habitat and represents a deeper water community which occurs normally in water depth from ca. lOO-25Om. The cave systems have environmental conditions which allow the telescoped deeper water faunas to live within these protected areas. The cave and channel systems within the cemented reef bodies are very widespread and the surfaces are vaster than the outer active growing zOlles of the coralgal reefs. The cave and pore systems are successively filled up by the growing of cryptic sessile benthos and, if present, by microbialites. Therefore also ancient reefbodies must dem- onstrate commonly this facies type. Very close affinities to the modem ones were recognized within Albian coral reef pla~orms of northem Spain (RErrNER 36 1987,1989, REllNER & ENGESER 1987, NEUWEILER this vol- urne). The best studied example is the Late Albian Albeniz- Eguino carbonate platfonn. This platfonn exhibits an outer barrier of coral reefs, an open lagoonal reef system with rudists, and deeper water micritic reef mounds with sponges. The platfonn is paleogeographically situated on a tilted crust segment in the front of a larger deltatic system. The most important observation in this respect is that the cryptic surfaces are occupied by a microbialte fades and the benthic community which has a great taxonomically corre- spondence with the modem ones (Acanthochaetetes-com- munity) (REllNER 1987, 1989, REl1NER & ENGESER 1987). The microbialites in the caves exhibit the same vertical facies successions as observed in modem ones (cf. KEupp et al.1993). The deeper water micritic/microbialitic sponge mounds ("mud mounds ") were located in a reconstructed waterdepths of 80-15Om at the platfonn margins. They demonstrate the same microbialite facies and benthic successions as seen within the shallow water reef caves today. This sponge mound fades is of widespread distribution at the platfonn margins and it migrates towards the coral fades as a result of rapid subsidence of the reef platfonn. The maximum distribution of sponge reef mounds in Albian platfonns is explained by trangressive events (start- ing phase of the "Cenomanian 11 -transgression) (further dis- cussion see NEUWEILER this volume). Comparable microbialite facies successions were stud- ied in J urassic sponge reefs (KEupp et al. this volume), in late Cretaceous and Eocene ones of the southern Pyrenees, and within early Camian sponge reefs of the Cassian Fonnation of northem ltaly. As a perspective and concept for further studies is postulated that within mostPhanerozoic micritic reef mounds (" mud mounds") sponges and associated microbes (symbionts and external ones linked to biofilms) playa central role and control the mound fonnation. All micritic reefs should demonstrate this community type. 10 CONCLUSIONS 1. The investigated modem microbialites from Lizard Island are mainly a product of matrix mediated calcification via acidic organic macromolecules and baffled detritus. The calcifying organic substances were detected by using histochemical methods. 2. Microbial biofilms cover the surfaces of the microbialites and control the input of dissolved organic substances. They also control the settlement and distribution ofbenthic organ- isms. The slimey surfaces ofbiofllms are able to fix detritus. Biofilm based layers can mineralize and in few cases calci- fied microbes were observed. 3. The microbialites show two structural types: a thrombolitic one growing under relativly high sedimenta- tion rates, and a more laminar 10 structurless one under reduced sedimentary influx (hardground-type). 4. All types are structured by Fe/Mn-microbial crusts. These crusts have a strong corrosive potential caused by electro- chemical anode dissolution. In some microbialites more than 40% of the volumen is corroded. The brownish layers exhibit a pseudostromatolitic feature. 5. The studied caves have four main facies zones: Zone 1: entrance area with nonnal reef corals (Acropora) Zone 2: dim light area with deeper water zooxantellate corals (Agaricia, Leptoseris) and thick crusts of coralline red algae. Zone 3: dimlight to dark area with slightly increased sedi- mentation rates and thrombolitic microbialite fonnation in eomparison with zone 4. Coralline sponges are present but rapidly growing benthos is dominant (r-strategists). Zone 4: dark conditions with redu~ed sediment particles in the water eolumn, hardground-type microbialite and the Acanthochaetetes-community (k-strategists). 6. The vertical succession ofthe microbialites demonstrates a transgressive character. 7. Geochemically the microbialites are high-Mg calcites and the stable isotopes show a micrite fonnation close 10 the isotopic equilibrium of sea water. 8. The growth of microbialites within the investigated reefs is most probalbly eontrolled by silicate weathering which increases the carbonate alkalinity. Therefore microbialitic erusts were only observed in reefs linked 10 continental islands (e.g.Lizard Island). In reefs of the outer barrier no microbialites were found. 9. The modem microbialites and associated organisms from the Lizard Island Section are very similar 10 those from Tertiary and Cretaceous occurrences. ACKNOWLEDGEMENTS I thank the co-directors Dr. L: Vail and Dr. A. Hoggart at the Lizard Island Research S~tion as well as Terry Ford and Lois Wilson, the crew of t\le RV "Sunbird", for exten- sive help and support. The investigations were pennitted by the Great Barrier Reef Marine Park Authority (G92/041; G93/046;). The Deutsche Forschungsgemeinschaft is ac- knowledged for financial support (Re 665/1-2, 4-1,2). I wish to thank Dr.Erlenkeuser (Univ.Kiel) and DrJoachimski (Univ.Erlangen) to earry out the stable isotope analyses and I acknowledge Prof.Franke (FU-Berlin) for the X-ray dif- fraction and electron microprobe analyses. The ESR-analy- sis were carried out by Dr.Eisenhauer (Heidelberg), and Dr.P .Röpstorf (Berlin) has done the TEM preparations, both are gratefully acknowledged. Many thanks 10 the Berlin Lizard Island working group, Dipl.Geol. Neuweiler, Prof. Keupp (all Berlin), Prof. Dullo (Kiel), Dr. Scholz (Ham- burg), Prof. Hillmer (Hamburg), and Prof. Zankl (Marburg) for eritical diseussions, eomments and leaving comparative material. Dr. Mehl and P. Böttcher (both Berlin) kindly improve the language and make helpful suggestions. Very constructive reviews by Prof. Schlichter. and an anonymous reviewer are gratefully acknowledged. REFERENCES ADDADI. L. & WEINER, S. (1985): Interactions between acidie pro teins and Crystals: Stereochemical requirements in biomineralization. - Proc. Natl. Acad. Sei. USA, 82, 4110- 4114 -- & -- (1989): Stereochemical and Structural Relations between Macromolecules and Crystals inBiomineralization. -In: MANN, S.; WEBB, J . & WIlLIAMS, R. J. P. (eds.): Biomineralization. - 133-156, Weinheim (VCH) & -- (1992): Kontroll- und Designprinzipien bei der Bio- mineralisation. - Angew. Chemie, 104, 159-176, Weinheim ADDADI, L.; BERMAN, A.;MORADIAN-OLDAK, J. & WEINE.R, S. (1990): Tuning of Crystal Nucleation and Growth by Proteins: Mo- lecular Interactions at Solid-Liquid Interfaces in Biomineral- ization. - Croatica Chem. Acta, 63, 539-544, Zagreb BARNES, DJ. (1981): Improvedmeasurement of coral calcification by sequential incubations with tritiated tetracycline and 45- calcium. - Proc. 4th internat. Coral Reef Symp., Manila, 2, 249-255, Manila BASILE, L.L., CUFFEY, R.J. & KOSICH, D.F. (1984): Sclerosponges, Pharetronids, and Sphinctozoans (relict cryptic hard-bodied porifera) in the modern reefs of Enewetak Atoll. - J. Paleont., 54,636-650;,Tulsa BE.RMAN, A.; ADDADI, L. & WEINE.R, S. (1988): Interactions of sea- urchin skeleton macromolecules with growing calcite crystals - a study of intracrystalline proteins. - Nature, 331, 546-548, London BERNER, R.A. (1968): Calcium carbonate concretions formed by the decomposi tion of organie matter. - Science, 159, 195 -197, Washington BORBAS, J. E.; WHEELER, A. P. & SIlms, C. S. (1991): Molluscan Shell Matrix Phosphoproteins: Cirrelation of Degree of Phosphorylation to Shell Mineral Microstructure and to In Vitro Regulation ofMineralization. - J. Exp. Zool., 258, 1-13 BRACHERT, T. & Duuo, C. (1991): Laminar micritic crusts and associated foreslope processes, Red Sea. - J. Sed. Petrol., 61, 354-363, Tulsa BUCZYNSKI, C. & CHAfE17., H. S. (1991): Habit of Bacterially induced Precipitates of Calcium Carbonate and the InfIuence of medium Viscosity on Minerology. - J.Sed.Petrol., 61, 226- 233, Tulsa BURNE, R. V. & MOORE, L. S. (1987): Microbialites: Organo- sedimentary deposits of benthic microbial communities . - Palaios, 2, 241-254, Tulsa CARROLL, J. J., GREENFIELD, L. J & JOHNSON, R. F. (1966): The Mechanism of Calcium and Magnesium Uptake from Sea Water by a Marine Bacterium. - J. Cello and Cop. Physiol., 66, 109-118 CHAfE17., H. S. (1986): Marine Peloids: A Product of Bacterillay induced Precipitation of Calcite. - J. Sed. Petrol., 56, 812-817, Tulsa CHAfE17., H. S. & BUCZYNSKI, C. (1992): Bacterially induced Lithification of Microbial Mats. - Palaios, 7, 277-293, Tulsa CRlSP, D. J. & RYLAND, J. S. (1960): Influence of filming and of surface texture on the settlement of "marine organisIns. - Nature, 185, p. 119, London CUlF, J.-P., GAUTRET, P., LAGHI, G. F, MASTANDREA, A., PRADIER, B. & Russo, F. (1990): Recherche sur la fIuorescence UV du squelette aspiculaire chez les demosponges calcitiques Triasiques. - Geobios, 23, 21-31, Lyon DEGENS, E. T. (1976): Molecular mechanisms on carbonate, phos- phate and silica deposition in the living cello - Topics on CUIT. Chem. 64,l-112 1979): Why do organisms calcify? - Chern. Geol., 25, 257- 37 269, AInsterdam (1989): Perspectives on Biogeochernistry. - 423 p., Berlin (Springer) DEGENS, E. T ., CAREY, F. G., SPENCER. D. W. (1967): Amino-acids and amino-sugars in caleified tissues of portunid crabs. - Nature, 216, 601-603, London DEGENS, E. T . & ITTEKKoT, V. (1986): Ca2+ -Stress, biological response and particle aggregation in the aquatic habitat. - Netherlands. J. Sea Res., 20, 109-116 DRAVIS, J. J. & YUREWIcz, D. A. (1985): Enhanced carbonate petrography using fluorescence microscopy. - J. Sed. Petrol., 55, 795-804, Tulsa . DREw, G. H. (1911): The action of some denitrifying bacteria in tropical and temperate seas, and the bacterial precipitation of calcium carbonate. - J. Mar. Biol. Assoc. U.K., 9, 142-155 DRUFFEL, E. R. M. & BENA VIDES, L. M. (1986): Input of excess C02 to the surface ocean based on 13C/12C ratios in a banded Jamaican sclerosponge. - Nature, 321,58-61, London FINE.RMAN, G. A. M. & Mn..CH, R. A. (1963): In vitro Binding of Tetracyclines to Calcium. - Natuie, 196,486-487, London FRICKE, H. & MmSCHNE.R, D. (1985): Depth limits of Bermudan scleractinian corals: a submersible survey. - Mar. Biol., 88, 175-187, Berlin FRICKE, H. & SCHUMACHER, H. (1983): The depth limits of Red Sea stony cora1s: an ecophysiological problem (a deep diving survey by submersible). - Pubbl. della Staz. Zool. de Napoli I Mar. Ecol., 4, 163-194, Naple FRI17., G. K. (1958): SChwamInstotzen, Tuberoide und Schutt- breceien im Weißen Jura der Schwäbischen Alb. - Arb. Geol. Paläont. Inst. TH Stuttgart, N.F., 13, 118 p., Stuttgart GAUTRET, P. (1989): Premieres donnees sur lesmassesmoleculaires des composes organiques associes aux squelettes aspiculaires de spongiaires calcifies (Demosponges et Calcarea). - C. R. Acad. Sei. Paris, 309 (ser. 11), 1083-1088, Paris GAUTRET, P. &MARlN, F. (1990): Compositionenacides amines des phases proteiques solubles et insolubles du squelene calcaire de trois Demosponges actuelles: Ceratoporella nicholsoni (HICKSON), Astrosclera willeyana LISTER et Vaceletia crypta (V ACELET). - C. R. Acad. Sei. Paris, 310 (ser. 11), 1369-1374, Paris. GAWLITTA, W, STOCKEM, W., WEHLAND, J. & WEBER, K. (1980): Pinocytosis and Locomotion of Amoebae. - Cell and Tissue Res., 213, 9-20, Berlin GERDES, G. & KRUMBEIN. W. E. (1987): Biolaminated Deposits. - Lecture Notes in Earth Sciences, 9, 283 p., Berlin GRANT, R. E. (1980): Brachiopods ofEnewetak Atoll. - In: Dev ANEY, D. M, REESE, E. S., BURCH, B. L & HELFRICH, P. (eds.), The Natural History of Enewetak Atoll; Vol.lI,- 77-84, Honolulu GREENFIELD, L. J. (1963): Metabolism and concentration of Cal- cium and Magnesium and preGipitation of Calcium carbonate by a marine bacterium. - Ann. N. Y. Acad. Sei., 130,23-45, New York • GUNTHORPE, M. E., SIl