<<

South Afllcan Journal of Botany 2001 67 399-410 CopYright © NfSC Ply Lid PrmtM m South Africa - AlIlIgllts roserved soutH AFRICAN jOURNAL OF BOTANY ISSN 0254-6299

Minireview

Coastal microbial mats: the physiology of a small-scale

LJ Stal

Netherlands Institute of , Centre for Estuarine and Marine Ecology, Department of Marine , PO Box 140, NL -4400 AC Yerseke, Netherlands e-mail: stal @cemo.nioo.knaw.nl

Received 10 May 2001, accepted in rev ised form 25 May 2001

Coastal inter-tidal sandy sediments! salt marshes and and the sulfide that they produce is oxidised by mangrove forests often support the development of anoxygenic phototrophic bacteria and by colorless SUl­ microbial mats. Microbial mats are complex associa­ fur bacteria. Growth and of these microor­ tions of one or several functional groups of microor­ ganisms result in markedly fluctuating vertical gradi­ ganisms and their formation usually starts with the ents of oxygen and sulfide that shift during a day-night growth of a cyanobacterial population on a solid sub­ cycle. This review discusses the metabolic contribu­ strate. They are considered as analogues of tions of the different functional groups of microorgan­ Precambrian . by the isms and how their joint effort results in the formation of fuels the metabolism of sulfate reducing the mat.

Introduction

In their , may occur matolites have been built th rough the growth and metabolic essentially in one of the following 3 ways: free-living. asso­ activity of microbial mals that lithilied through calcification ciated with an d often inside other or fo rm ing mul­ and subsequently by silicification and other diagenetic ticellular, mono- or muftispecies aggregates. In the latter processes (Walter 1976). The lamination in the rock rep re­ form they may occur as colonies, biofilms or microbial mats. sents the seasonal or erratic growth pattern of the microbial I will not attempt to give an all-embracing definition of any of mats, comparable to the growth rin gs in the trunk of a tree . the 3 forms of microbial aggregates, because this seems to In stromatolites fossi l remnants of microorganisms have be a nearly impossible tas k and also the borderlines been found Ihat morphologically resemble modern between the diffe rent forms are in fact continuous. cyanobacteria, which also today are common builders of Nevertheless, it is important to indicate in a more general microbial mats (Schopf 2000) . Since there is little doubt that way of what is understood in Ihe framework of th is review by many of these stromatol ites were lormed through au totroph­ a . Microorganisms th at grow on a solid sur­ ic metabolism, it is tempting to believe that they were built by face may eventually form microbial mats. Microbial mats cyanobacteria. However, the recent discovery that cyanobac­ have been the subject of 3 international meetings of which teria are an evolutionary relatively young group within the th e results have been published (Cohen et al. 1984, Cohen , argues against this hypothesis (Gupta 2000). and Rosenberg 1989, Stal and Caumette 1994). A mat dif­ Modern microbial mats have been termed recent stroma­ fe rs fro m a biofilm mainly because of size and the coherent to li tes in order to distinguish them from th e fossil ones. structure typical for the former. In its ultimate form, microbial However, the majority of the modern microbial mats do not mats resemble something like a doormat, which can be lithify and therefore th e term '' was considered peeled from the surface as a whole. This explains the origin not appropriate, except in the few examples that are known of the concept of a microbial mat. Before this te rm became to calcify. The same is true for 'potential stromatolites' common property, microbial mats were known as 'laminated because it is uncertain whether non-calcifying microbial microbial ' or recent or potential stromatolites mats in fact possess the potential of lithification. Similar as (Krumbein 1983) . Stromatolites are fossil laminated rock for­ stromatolites, modern microbial mats retain a lamination, mations of biogenic origin . The oldest stromatolites date representing older, partly degrad ed mats. However, the term back to more than 3.5 Gyr B.P. and represent the earliest 'laminated microbial ecosystems' usually does not refer to indications of life on earth. It is generally accepted that stro- this historical lamination but rather to a vertical zonation of 400 Sial

different functional groups of microorganisms , which is often colonise low- environments. In inter-tidal coastal visible to the naked eye because of the different colors. sediments they prefer fine sandy sediments , which combine Thus, the actual active microbial mat is laminated through moderately strong hydrodynamics and low sedimentation differen t groups of organisms. rales (Yallop et al. 1994). Moreover, the quartz sand grains However, neither the historical nor the instantaneous lam­ allow excellent Iransmission of lighl into Ihe sediments (Stal ination is the law of the Medes and Persians . Microbial mats , et al. 1985, Kuhl et al. 1994). Initial colonisation of the sedi­ after growlh has ceased, may be decomposed completely, ment by Ihe usually fi lamentous cyanobacte ria is Ihrough not leaving a visible trace and the substrate is co lonised adhesion to sand grains probably by sticky extracellular every season anew. Likewise, a microbial mat may be com ­ polymers. This property allow cyanobacteria to colon ise posed of one species or different species may nol be sepa­ environments with relatively high energy, Without being rated into differenl strata or they may not be dislinguished as washed away. Some species are particularly well equipped such . Such systems are obviously not 'laminated' but Ihey to settle in high-energy environments. For instance, in the do nol principally differ from Ihe 'laminated microbial ecosys­ Bahama's Schizothrix spp. is known to colonise envi ron­ tems' and can all be em braced by the term 'microbial mat'. ments with strong wave currents where other organisms are Theorelically, microbial mats may be composed of one unable to sellie (Reid and Browne 1991 ). These organi sms particu lar species. However, as a ru le a complex microbial give rise to the formation of modern stromatolites. ecosystem forms, in which a variety of different functional groups of microorganisms rep resent a structu ral and physi­ Photorespiration ological unit (Van Gemerden 1993). II has been proposed that microbial mats are structural and physiological equiva­ Once a successful colonisation has occurred, the cyanobac­ lents of tissues (Wachendorfer 1991). teria th rough growlh and their pholosynlhetic aclivity enrich In this review I wi ll describe the processes and organisms the sediment with organic matter, which becomes available that are involved in Ihe formation of one Iype of microbial to other microorganisms. There are a number of different mal Ihat is buill by cyanobacteria Ihal is frequently found in mechanisms by which the pholosynthetically fixed is coastal inteHida1 sediments, mangrove forests and salt liberated into the environment. An important mechanism marshes all around Ihe globe. Microbial mats developing in could be Ihrough pholorespiralion . The cyanobacterial mal extreme environments such as hypersaline ponds, thermal is characterised by a high concentration of . The sp rings or hot or cold deserts, nor non-phototrophic systems oxygen thai is produced through pholosynthesis accumu­ such as mals of the gliding bacleri um or lates in the mal and can only leave it through diffusion. Thioploca (Larkin and Strohl 1983) are nol discussed here. Although Ihe diffusion coefficienls of the polysaccharide matrix of the cyanobacterial mat is not much different from The de velopment of a cyanobacterial mat: primary pro­ that of water, the medium is stagnant and no turbulence can duction is the motor of the ecosystem aid the exchange of gas with the overlying water or air. Hence, the cyanobacterial mat may become supersaturated Colonisation with oxygen (Revsbech et al. 1983). Two to three-fold oxy­ gen supersaturation in cyanobacterial mats is not exception~ Inter-tidal sandy sediments are high energy environments , al. At the same time the inverse is true for C02. Carbon diox­ exposed to strong hydrodynamic conditions. Sediment parti­ ide is fixed during and depleted Irom Ihe cies of small grain size such as silt and cl ays will not deposit mat. It can only be repl enished by diffusion from the overly­ here , or they will be eroded away. On the most exposed ing medium . Rlbulose-l ,5-bisphosphale ca rboxylase/oxyge­ areas only the heavier quartz sand grains are deposited. nase (Rubisco) , the key enzyme of the reductive pen tose Inter-tidal sand flats can be consi dered as extreme environ­ cycl e (Calvin) an d Ihe enzyme responsible for ments because besides Ihe physical forces they are low in the fixation of CO.!, possesses also oxygenase activity, Le. and the periodic inundation causes desiccation and uses 0 2 for the oxidative cleavage of ribu l ose~ 1 ,5-bisphos· strong variations in salinity and temperature. Few organisms phate (Tabila 1994) . In facl, Rubisco has a much beller affin­ are capable of colon ising these environments. ily for 0 , than fo r CO, and in order 10 be able to effectively Cyanobacteria have remarkably few nutritional require­ fix CO" its concenlration musl be much higher Ihan 0 ,. In an ments. Their main way 01 life is photoautotrophic, which aerobic environment, many cyanobacteria (and other means Ihat they use as the source of energy, water as ) achieve th is requ irement by concentraling CO, the electron donor and C02 as the sole source 01 carbon (Carbon Concenlraling Mechanism, CCM) and Rubisco in (Stal 1995). Moreover, many species are capable of fixing carboxysomes (Price and Badger 1991 , Reinhold et a/. atmospheric nit rogen, which makes them independent on 1991). Howeve r, when both the ambient CO, concentration sources of combi ned nitrogen such as nit rate or is low and the 0 , level is high, the GGM may nol be able 10 or organic nitrogen , which are generally in low supply in the prevenl the oxygenalion of ribulose-l ,5-bisphosphate, whi ch marine environmenl (Paerl et al. t 996, Bergman et al. evenlually will lead to the formalion of glycolale Ihat is 1997). Hence, the on ly cri lical nulri ent fo r these cyanobacte­ excreted inlo Ihe medium (Renslrom and Bergman 1989). ria is phosphate. Cyanobacteria are well-known for their high Hence, the oxygenation reaction leads to a loss of fixed car· affinity towards this important nutrient , which they can store bon. This photorespiration might be a major source of organ­ inlracellularly as polyphosphale (Ri egman and Mur 1985). ic carbon lor the microbial associated with the Cyanobacteria are therefore excellent candidates to cyanobacterial mat. Soulh Afri can Journa l of BOlany 200 1. 67 399-410 401

Fermentation Extracellular polymeric substances

Another major source of organic matter is the fermentative Mat-bUilding cyanobacteria also excrete large amounts of metabolism of cyanobacte ri a. Whereas in th e lig ht the polymeric su bstances (extracellular polym eric su bstances, cyanobacte ri al mat may be su persaturated wilh oxygen, in EPS), wh ich are largely composed of polysaccharides and the dark not on ty no oxygen is produced but also there is a with minor componenls of and lipidS (Stal 1994, high demand for oxygen. Initially, the cyanobacte ria will Decho 1994). One may distinguish roughly two types of mobilise their storage (glycogen) and resptre it EPS: one is more or less intimately associ ated with the aerobically in order to generate bIOch emical energy. (cells or trichomes) and is usually design ated as However, this will deplete the oxygen in the cyanobacterial the sheath (De Philippis and Vincenzini 1998) . This is a mat quickly and diffu sion from the overlying water or air is more or less structural cell component, albeit that it is out­ slow. This will result in th e factual anoxic conditions in the side the . Depending on th e organism this sheath mat. Measurements of oxygen dynamics in microbia l mats may be thick and may in fact be wider than the tri ch om e have repeatedly show n th at an oxic conditions are often itself, or it may be vanishing th in or even absent. Some uni­ establ ished withi n minutes aft er darkening, meaning that the cell ular cyanobacteria such as Gloeolhece spp . or cyanobacteria in the mat are confronted with anoxia during Gloeocapsa spp. form colonies of which the cells are virtually the whole night (Revsbech et al. 1983). In order to embedded in a polysaccharide sheath (Tease et al. 1991). cover their energy demands during dark anoxic cond itions The trichomes of the cosmopolitan mat-forming cyanobac­ th es e cyanobacteria switch to a fermentative metabol ism terium Microcoleus chthonoplastes form bundles that are (Stal and Moezel aar 1997) . The reserve compound glyco­ enclosed by a common sheath (Garcia-Pichel et al. 1996) . gen and sometimes also energy-rich com patible so lutes Cyanobacteria such as aestuarii that produce very (osmotica) are mobilised and fermented to a variety of low­ thick sheaths are usually not motile (R ippka et al. 1979). The molecular compounds that are excreted into the med ium . sheath may serve d i ~erent functions. The highly hydrated Depending on the species, cyanobacteria have been shown polysaccharide s are an effective protectIon from desiccation to possess homo- and heterolactic ferm entation, mixed acid that may occur when the mat is exposed to the air. The Fermentation and homoacetic fermentation. A considerable sheath is also effective as an adhesive that attaches the part of the glycogen is fe rm ented and has to be replenished organism to the su bstrate. It may further scave nge rare by photosynthetic CO, Itxation during the subsequent day mi cro nutri ents and protect the organism from a variety of period. Hence, an important part of this fixed CO?e nd s up as exte rnal th reats, including grazing. The sheath is always low-molecular ferm entation products in the medium and produced in the cyanobacteri a that have one, bu t its prope r­ becomes available to other mic roorganisms. ties may vary wit h the environmental condition s in which the organism thrives. Th e other type of EPS produced by Compatible so/utes cyanobacteria is not inti mately associated with th e organism and is excreted into the medium as mucilage. In many cases Although fenmentation and photo respiration are probably the this type of EPS seems to be produced as th e result of major mechanisms by wh ich organic matter becomes avai l­ unbalanced growth . This occurs when grow th of the organ­ able in the cyanobacteri al mat, thei r exist also other mecha­ ism is timited by a nu tri ent (o~en N) while its photosynthetic ni sms, that may be important under certain conditions. and COr fixing capacity are not impaired. Under such condi­ Marine cyanobacteria accumulate compatible solutes that tions. cyanobacteria in itially produce more of the intracell u­ serve as osmoprotectants (Reed et a/. 1986). Th e most lar storage com pound glycogen, but the space for this inside common osmoprotectant in marine cyanobacteria is glycosyl the cells is limited (Lehmann and W6ber 1976) . glycerol, but also the disaccharides trehalose and to a tess­ Cyanobacteria rarely produce more than 50% of dry weight er extent sucrose can be found. Glycine-betaine is mo re as glycogen. Excess of fixed CO, is su bsequently excreted common as compatible solute in cyanobacteria in hyper­ as mucilage into the medium. This seems a waste but in this saline environments. Particul arly the disaccharides and glu­ way, cyan ob acteria dissipate the light energy that they har­ cosyl glycerol are energy rich compounds and it has been vest through their pigments. This EPS forms a gel matrix in shown that they may be used in addition to the storag e com­ which the cyanobacteria are embedded and it may have pounds as energy reserve in the dark, particu larly under similar advantages for the community as a whole as the anoxic conditions. Cyanobacterial mats in inter-tidal sedi­ sheath has for individuats. In addition , EPS excretion may ments a.n d on rocky sh ores may be exposed to dramatic occur as part of the mechanism of gliding motil ity of some changes in salinity. This is the case when during exposure cyanobacte ria (Castenholz 1982). th e mats experien ce a salinity down-shock when it ra ins. Although EPS represents a high quality and energy-rich The only way for cyanobacteria to protect them from such a substrate for microorganisms, it must be hydrolysed extra­ sudden osmotic down shock is to quickly excrete the com ­ cel lularly before it can be taken up and utilised. Sh eath EPS, patible sol ute into th e environment where it subsequently but also the mucilage, can be considered as rather recalci­ becomes available to other microorgani sms. Th is mecha­ trant compoun ds. In many coastal microbial mats the nism of transfe r of organic matter to the community is rather amount of mucilage is low as compared to microbial mats in erratic. hypersaline environments. In the latter, the degradation of complex organ ic molecules appears to be very slow proba­ bly because of high salinity. In the well-investigated hyper- 402 Sial

saline microbial mats 01 Solar Lake (S inai, Egypt) and the b l oom~forming cyanobacte ri a, but benthic cyanobacteria have salterns of Guerrero Negro (Baja California, Mexico), the top not been investigated in th is respect. Recently, it has been layer of 1 and 0.1 m, res pectivel y, is purely organic in nature suggested that cyanobacteria, including mat-forming species, and formed through the accu mulation of the successive may be produce compounds with antibiotic activities or sub­ active cyanobacterial mats. The mat (Pond 5) of Guerrero stances that are involved in cell-to-cell Signaling (Kreitlow et

Negro grows at a rate of approximately 1 em y 1, close to the al. 1999). All these subjects are still in their infancy. rate of of the older layers, keeping the sys­ tem more or less at a thickness of 10cm (Des Marais 1995). Oxygen dynamics in microbial mats However, some net accumulation must have occu rred in order to produce the layer of 10em of organic matter. The Cyanobacteria are oxygenic phototrophic organisms. Solar Lake microbial mat system has been estimated to be Therefore. highly dynamic vertical oxygen profi les are an 2000 years old (Krumbein ef al. 1977). An average yearly important phenomenon in microbial mats (J0rgensen et a/. net accumulation rate of O.5mm wou ld have yielded the 1979). These are in the lirst place produced as a result from present day thickness. This IS on ly a fraction of the yearly the daily light curve. During the night oxygen is consumed by gross accum ulation, which is in the sa me order of magn itud e respiration. The oxygen demand is usually high so that the as in the Guerrero Negro mat. This means that 99.5% of the entire mat may become anoxic. Only the mat surface may organic matter produced is degraded in these hypersaline receive oxygen, particularly when it is exposed directly to air. mats and that only th e most recalcitrant molecules remain During the day oxygenic photosynthesis is only possibl e in as refractory matter in the system. In coasta l inter-tidal the upper part of the mat in which sufficient light penetrates. microbial mats this net accumulation is usually not observed The euphotic zone of the mat is defined as the layer in which (Stal ef al. 1985, Stal 1994). This may be attributed to the gross oxygenic photosynthesis occu rs . This is not the same fact that the degradation of recalcitrant organic matter may as the depth at which oxygen may penetrate. In the light a be easier at lower salinity, bu t it is more likely that these sys­ typical oxygen profile shows a concentration maximum at a tems are exposed to dynamic conditions, causing erosion or depth of 0.1-0.3mm. This peak usually coincides with th e oxygenation. The occasionally in troduction of oxygen would maximum photosynthetic activity or a concentration of pho­ facilitate the degradation of recalcit rant organic molecules. tosynthetic biomass or both. Cl oser to the surface the high­ er irradiation may be sub-optimal, resu lting in lower photo­ Grazing synthetic rates or even in negative values when photooxida­ lion prevails. Below the oxygen optim um, irradi ance levels After having discussed photo respiration , fermentation, are not saturating, tikewise resulting in lower oxygen pro­ osmotic down-shock and EPS exudation as major mecha­ duction rates. Obviously, the actual lotal oxygen production nisms by which photosynthetic fi xed CO, is lib erated as rate at any place in the mat is a function of the active ph oto­ organic matter in the environment, a few other mechanisms synth etic biomass and the level of irradiance. The latter should be mentioned here as well. Obviously, organi sms varies during a diurnal cycle and the oxygen maximum is and that includes cyanobacteria, have a limited life. Cells therefo re pushed from the surface down to reach its maxI­ may stop dividing and eventually die and disintegrate. In mum depth at noon, when the level of irradiance has coastal micro bial mats grazing may be more important as a become maximal. On the other hand, in some microbial cause 01 cell death (Fenchel 1998, Fenchel and KuhI 2000). mats the standing stock of photosynthetic biomass is also T he possibility of viral or bacterial attack has a cause of cell not constant at a particular place during a diurnal cycle. lysis in microbial mats has received only little attention in lit­ Many mat-bUilding cyanobacteria are motile by gliding and erature (Margulis ef al. 1990), but may prove to be an impor­ may constantly move in order to position themselves at opti­ tant process. However, conSidering the fact that the total mum light conditions. Hence, the cyanobacterial mat opti­ amount of fixed CO? that is liberated as non-structural com­ mizes its photosynthetic pertorm ance continuously and this pou nds may exceed 90%, the con tribution of structural is reflected in the oxygen profile in the mat. In addition, light cyanobacterial cell materi al to the chemotrophic community impinges rarely as a pure sinus curve and may fluctuate is comparatively small. tremendously during a diurnal cycl e du e to clo uds or to water covering the mat. Photosynthesis responds immedi­ Secondary metabolites ately to any fluctuation in light intensity, consequently chang­ ing th e oxygen profile. Other organic compounds that may be produced and excret­ The actual oxygen concentration at any location in the mat ed by mat-forming cyanobacteria are the result of secondary is the result of its production by photosynthesis, its con­ metabolism (Jultner 1987, Carmichael 1992). The amounts sumption by respiratory and chemical oxidation processes of these compounds are usually small and do not contribute and diffusion. This is the basis of a poputar method to meas­ to the organic substrate that is available for the microbial ure photosynthesis in microbial mats using micro-elec­ community, but they may be of great importance for the trodes: the decrease of oxygen concentration in the first cou­ ecosystem functioning. Examples of products of secondary ple of seconds after the darkening of the mat equals the pho­ metabolism include geosmines and other volatile organic tosynthetic rate (changing the minus sig n to positive) odorous compou nds. Not much is known of th ese com­ (Revsbech ef al. 1983). Downward diffusion of oxygen may pounds from microbial mats. The same is true for cyanobac­ result in aerobic or micro-aerobic conditio ns below the terial toxins. They are well-known from certain planktonic, euphotic zone. On the other hand respiration and chemical South Alrlcan Journal of Botany 2001. 67 ' 399-410 403

oxidations may exceed the photosynthetic oxygen produc­ (Risatti ef al. 1994). Sullate reduction has been demonstrat­ tion, resulting in anoxic conditions within part of the euphot­ ed even in the light in the fully oxygenated cyanobacterial ic zone. In this latter case, the method for measuring photo­ mat. Since a search for anoxic micro-niches in this ma t was synthesis as described above, does not work. unsuccessful, it was conceived that sulfate-reduction would take place under fully aerobic conditions (Canfietd and Des Microbial mats: a joint venture of several functional Marais 1991) . Attempts to isolate sulfate-reducing bacteria groups of micro-organisms that carry out sulfate reduction under aerobic conditions have failed so far. The substrate for aerobic sulfate reduction As we have seen, much of the photosynthetic lixed CO" co uld be glycolate, excreted by th e cyanobacteria in the light becomes readily available to the microbial community. A as the result 01 photorespiration . Glycolate has been shown large portion of this organic matter is easily degradable low­ to be an important substrate for chemotrophic bacteria in molecular compounds. All 01 this material is produced in the microbi al mats and su lfate-reducing bacteria using glycolate photosynth etically active cyanobacteria l mat and it can be have been isolated from cyanobacte rial mats (FrD nd and anticipated that is also decomposed there. In the presence Cohen 1992, Nold and Ward 1996). These observations of oxygen it can be assumed that aerobic processes pre­ question the importance 01 aerobic degradation 01 the bulk dominate in the decomposition of this organic matter. During of the organic matter in these microbial mats. Ae robic the night, oxygen is depleted often within minutes, which is metabolism may be limited to the oxidative attack of com­ mostly attributed to the respiratory activity of the cyanobac­ plex, recalcitrant compounds. teria themselves, subsequently switching to fermentation. Th e major fermentation products are acetate, formic acid, Sulfide lactate, ethanol, H, and CO,. These are excellent substrates for the obligate anaerobic sulfate-reducing bacteria (Hansen Sulfate reduction is a form of anaerobic respiration, using 1994). Methan ogenic or acetogenic bacteria could also be sulfate as the term inal electron acceptor an d producing SUl­ involved in the degradation 01 these compounds but since fide. Next to oxygen, sulfide is a major compound determin­ sulfate-reducing bacteria are superior in their affinity for ing microb ial activities and the vertical stratification of differ­ these substrates and sui late reduction yields more energy. ent functional groups of microorganisms tha t is typical for these bacteria do not playa role of importance in coastal marine microbial mats. It should be noted that sulfate reduc­ microbial mats as long as sulfate is present (Raskin et al. tion is not the on ly process that results in the fo rmation of 1996). Seawater contains 28mM 01 sullate and its depletion sulfide. For ins tance, cyanobacteria under anaerobic condi­ in marine microbial mats is only envisaged at extraordinary tions may use zero-valence 'elemental' sulfur as electron high or when th e supply of seawater lags behind. sink fo r fermentation, which also results in the formation of In hypersaline microbial mats methanogenic bacteria may be sulfide (Stal 1991). A variety of other bacteria use zero­ important because of the presence of non-competitive sub­ valence sulfur as electron acceptor for anaerobic respiration, strates (Orem land and Polci n 1982, Cytryn ef al. 2000). but it is not known how important these organisms are in marine mi crobial mats. Sulfate reduction Sulfide is chemically and biologicalty very reactive and in addition it is toxic to almost all organisms, including those We may assume that a substantial pa rt of the fermentation who produce it or depend on it as substrate. In the lirst products excreted by the cyanobacteria is metabolised by place, sulfide reacts instantan eously with iron. It is oxidised sulfate- reducing bacteria. It should be noted that these bac­ to zero-valence su lfur by ferric iron, wh ich itself is reduced to teria should occur in the Vicin ity of where their su bstrate is ferrous iron . Sulfide preci pitates with ferrous iron forming the produced, I.e. in the cyanobacterial mat. This is remarkable, virtually insoluble FeS, which produces the characteristic since sulfate-reducing bacteri a are considered as ob ligate intense black sediment. FeS may subsequently react with anaerobic organisms and the cyanobacterial mat may be zero-valence sulfur to form the very stable pyrite, which is a even supersaturated with oxygen during daytime. This relatively slow process (Howarth and Jergensen 1984, seems contradictory. Although it could be conceived that Thode-Andersen and J0rgensen 1989) . either the sulfate-reducing bacteri a move up and down in the Sulfide also readily reacts chemically with oxygen and mat with the sh ifting oxygen profiles or that the fe rme ntation therefore both species can not co -exist at high concentra­ products diffuse into the lower permanent anoxic part of the tions. Co-existence of low concentrations of oxygen and sul­ mat, it has been clearly demonstrated that large numbers of fide has been shown to occur in microbial mats, but more sulfate-reducing bacteria are perman ently present in th e Irequently th is is hardly measurable because the ra te 01 bio­ cyanobacterial mat and that the highest rates 01 su lfate logical oxidation of sulfide is much faster than the chemical reduction indeed are found in th is layer (Visscher ef a/. 1992, oxidation (Krumbein ef al. 1979, Revsbech et al. 1983). Two Tes ke ef a/. 1998). More recent work has shown that many important functional groups of microorganisms are involved sulfate-reducing bacteria tolerate oxygen and that some are in sulfide oxidation in microbial mats. The colorless sulfur able even to perform limited aerobic respiration (Dillin g and bacteria are chemosynthetic organisms that oxidise sulfide Cypionka 1990). The vertical distribution 01 sulfate-reducing to sulfate using oxygen as electron acceptor. Many of these bacteria reveals that typica l oxygen-tOlerant species domi­ species can live ic, Le. are capable of fixing C02. nate the top layer of the mat, wh ereas those that can not tol­ Very high numbers (up to 10'cm' ) of these bacteria have erate oxygen are found in the permanently anoxic layers been lound in microbial mats (Visscher et al. 1992). Some 404 SIal

species are facultative anaerobic and capable of denitrifica· The intracellular stored sulfur serves another important lion , using nitrate as the te rminal electron acceptor, but function in purple sulfur bacteri a. As is the case with since marine microbial mats are usually nitrogen·depleted, cyanobacteria, the purple sulfur bacteria switch to fermenta­ this mode of metabolism is considered unimportant. The col ­ tion in the dark, using elemental sulfur as electron sink. In this orless sulfur bacteria are more or less homogeneo usly dis­ way, purple sulfur bacteria also contribute to the production tributed in the mat and are probably only active at the sui· of su lfide in the microbial mat (Van Gemerden 1968). fide-oxygen interlace , thereby generating an almost perlect Whereas the anaerobic purple sulfur bacteria display aero separation between these two. Becau se the oxygen gradient obic metabolism when exposed to oxygen , most ma t-fo rm­ is moving up and down during a diurna l cycle, the ing cyanobacteria are capable of ano xygenic photosynthesis sulfide-oxygen interlace follows it. It seems that the color· when exposed to sulfide (Cohen et al. 1986). Sulfide is a less sulfur bacteria would have only a very limited time to be potent inhibitor of oxygenic photosynthesis, but it may metabolically active, unless they move with the su lfide-oxy· donate electrons to photosystem 1 and in this manner allow gen interlace. It is not known wh ich of these options apply in C02 fixation . In some species th is property must be induced microbial mats. and requires de novo protein synthesis. Th is is for example the case with the cyanobacterium Oscillatoria limnetica from Anoxygenic, purple sulfur, phototrophic bacteria the sulfide· rich hypolimnion of the hypersaline Solar Lake (Sinai) (Arieli et al. 1989). However, oxygen and sulfide gra· The other important functional group of s in dients in microbial mats are strongly fluctuating and th ere­ microbial mats thai are involved in the oxidation of sulfide fore mat-forming cyanobacte ria such as Microcoleus are the anoxygenic phototrophic bacteria (De Wit and Van chthonoplastes usually possess the capacity of anoxygenic Gemerden 1988). These are essentially obligate anaerobic photosynlhesis constitutively and perlorm oxygenic and organisms that perlorm photosyn thesis and fix CO,. In stead anoxygenic photosynthesis simultaneously at sulfide con­ of water they us e sulfide as the electron donor. Most com· centrations lower than 1mM, with the relative importan ce of monly found in marine microbial mats are the purple sulfur oxygenic photosynthesis decreasing with increasing sulfide bacteria. Because they need light, they are found immedi· concentration (De Wit et a/. 1988). Above 1mM sulfide oxy· ately below the cyanobacterial mat. Although the cyanobac· genic photosynlhesis is complelely inhibited. It is likely th at teria filter out most of the visible light. purple sulfur bacteria in this case growth of the cyanobacterium becomes impos­ use a different part 01 the spectrum, particularly in the sible, since M. chthonoplastes has an indispensable require ­ infrared, which penetrates the mat extreme ly well. Moreover. ment lor oxygen. purple sutfur bacteria are low·light adapted organisms that perlorm photosynthesis at light levels as low as 0 .1 % of full Vertical stratification of function al groups of microor­ (Overmann et al. 1992). Another important reason ganisms in microbial mats why purple sulfur bacteria are found in microbial mats is that these organisms are also very oxygen-tolerant or are even Four major functional groups of microorganisms have been capable of aerobic metabolism (De Wit and Van Gemerden distinguished so far. These are: the oxygenic cyanobacteria, 1990). This is important, because of their occurrence imme­ the anoxygenic phototrophic bacteria, the sulfate·reducing diately below the cyanobacterial mat, where oxygen may be bacteria and the colorl ess sullur bacteria. Only the first 2 present. In the presence of oxygen. purple sutfur bacteria . groups lorm clearty stratified layers. visible with naked eye. continue photosynthesis but they are unable to synthesise The latter 2 groups are distributed throughout the microbial the major lighl·harvesting pigment: bacteriochlorophyll a. mat, allhough Ihe sullate reducing bacteria may be parti· After prolonged exposure to oxygen. they wilt eventually end tioned into a more oxygen·tolerant population in the top lay· up as colorless organisms. not capab le anymore of photo­ ers and the truly Obligate an ae robic species in the pe rma­ synthesis. Although these cells are capable of a chemosyn· nent anoxic layers. thetic mode of metabolism, oXidising sulfide wit h oxygen . This vertical stratification of microorganisms may be more identical to the colorless sulfur bacteria, they can not com­ comp lex. In some microbial mats a layer of green sulfur bac­ pete WIth the latter organisms because of poor substrate teria is found beneath the purple sullur bacteria (Nicholson affinity. et al. 1987). The green sutfur bacteria are another group of Purple sulfur bacteria oxidise sulfide in two major steps. anoxygenlc phototro phic bacteria. Although they share part They first oxid ise it to ze ro-valence sulfur, which is stored of the light spectrum with the cyanobacleria, their light require· Intra cellularly (in fact outside the cytoplasm membrane). ments are extremely low and when sufficient light impinges on This reaction is relatively quick. by which the su lfide avail· the mat it can be anticipated that these organisms may live able is rapidly depleted. The oxidation of sulfide to sulfur phototrophically. are extremely oxygen produces only 2 electrons with which on ly y, CO, can be sensitive but resist high sulfide concentrations. fIxed. The subsequent oxidation of sulfur to sulfate occurs The occurrence of a green layer be neath the purple sulfur after the su lfide has been depleted , is slower, and because bacteria does not necessanly indicate the presence of green it yields 6 electrons it allows for the fixat ion of 3 times as sulfur bacteria. Sometimes a second layer of cyanObacteria much CO,. This strategy is ecologically advantageous when can be found below the layer of purple sullur bacteria. different species compete for sulfide. Once the sulfur is Measurements of oxygen and photosynthesis have shown stored inside the cells . it is not available to oth er organisms that these cyanobacteria periorm predominantly an oxy· (Van Gemerden 1983). genic mode 01 photosynthesis (Krumbein et al. t 9771. The South African Journal 01 Botany 2001. 67 399- 410 405 activity of the pu rp le sulfur bacteria takes care of the elimi­ nation of sulfide that otherwise would prevent the occur­ rence of oxygenic photosynthesis. The oxygen profile as measured by micro-electrodes shows then 2 peaks, sepa­ rated by an anoxic layer. Although the species composition of the deep layer of cyanobacteria differs from th e surface mat, it seems likely that it represents the original surface mat that has been overgrown by a new one and that succes­ sively shifts in species composition have occurred. In other occasions 'inverted mats' are encountered where the purple u 5" 5" 5" 0 , 0, X sulfur bacteria form the top layer and the cyanobacteria 0 2 3 occur beneath them (Van Gemerden et al. 1989, Van u Fe 3• .. Fe2~ • F 5 .. Fe3~ "Fe + --Fe " X Gemerden et al. 1989). Such mats form on sediments that 0 u Z 2 2 rec eive a high load of exogenous produced organic malter, 5. - 5' 5 -.. S2- X <: , 5' 0 >,. ,'.. •• -i ...... / ~ Z for instance or grasses deposited on the beach. <: lis decomposition result in the production of large amounts FeS of sulfide, preventing growth of cyanobacteria. Mats of pur­ ple sulfur bacteria may develop that scavenge the sulfide, allowing cyanobacteria to grow below them. Figure 1: A conceptual model of a multi-layered microbial mat wilh a layer of iron as the buffer between the aerobIc and anaerobic Layer of ferric iron communities 01 phototrophic microorganisms. At night, sulfate­ reducing bacteria (SRB) produce sulfide throughout the mat and the top layer of cyanobacteria and the layer of purple sulfur bacteria The separation of the oxygenic and anoxygenic ­ below, reduce zero valence ('elemental') sulfur to sulfide. This ic bacteria l communities is clearly the result of the opposite reduces ferric iron (Fe)') whi le the sulfide is ox[dised back to sutfur. gradients of oxygen and sulfide and of course of light. Sulfide subsequently reacts with ferrous iron (Fe2. ) to produce the Although the separation appears perfect to the naked eye, black insoluble FeS. During the day, when the cyanobacteria evolve there is overlap between the two when observed at the O~ in the cou rse of photosynthesis, FeS is oxidised. This prevents micro-meter scale, giving rise to competitive interactions and OJ from diffusing into the layer of the anaerobic, anoxygenic bacte­ the exposure to sulfide and oxygen in the aerobic and anaer­ ria. The sulfide produced by the SRB reacts with the Fe" and pre­ obic communities, respectively. In some microbial mats an vents it from reaching the cyanobacteria. additional layer can be distinguished between the cyanobac­ teria and the purple sulfur bacteria. This layer has a rusty color and presumably is composed of iron hydroxides (Stal Mat-forming cyanobacteria are protected by an iron coat 1994). This would represent an ideal barrier between the aerobic cyanobacteria and the anaerobic purple sulfur bac­ The rusty layer that sometimes separates the green and pur­ teria (Figure 1). Any sulfide diffusing upwards will react with ple communities may prevent sulfide diffusing from below ferric or fe rrous iron before it reaches the cyanobacterial into the cyanobacterial mat but it does not help against su l­ mat. Vice versa, any oxygen diffusing downwards will react fide which is produced inside the cyanobacterial mat by sul­ with ferrous iron or FeS and be unable to interfere with the fate~reducing bacteria or by the cyanobacteria themselves. purple sulfur bacteria. II is presumed that this layer will tend The mat-forming cyanobacterium M. chthonoplastes has to reduce at night and oxidise during daytime. This hypothe­ been shown to accumulate iron in its polysaccharide sheath sis so far has not been proven experimentally. Apart from (Stal 1994). Th is layer of iron may serve a similar function as being a pure chemical barrier, this rusty layer may also rep­ the rusty layer of the mat (Figure 2). When during fe rmenta­ resent a community of an oxygenic phototrophic (purple) tion in the dark M. chthonoplasles reduces sulfur to sulfide, bacteria that uses ferrous iron as electron donor, oxidising it the latter reacts with fe rric iron, oxidising the su lfide back to to ferric iron (Widdel et al. 1993). Such bacteria have been zero valence 'elemen tal' sulf.ur, and producing ferrous iron. isolated from a variety of environments but it is not known The net result of this fermentation is the reduction of iron. whether th ey are important in microbial mats. It has also Another process coupled to fermentation by which iron is been proposed that cyanobacteria may be capable of iron­ reduced is the oxidation of formic acid to COl. During the dependent anoxygenic photosynthesis but experimental subsequent light period the ferrous iron is oxidised back to fer­ proof for this hypothesis is lacking (Cohen 1984). The ric iron by the oxygen produced during photosynthesis. This involvement of chemosynthetic bacteria in the oxidation of wrll keep the oxygen partial pressure low in the cell and in its iron in microbial mats is less likely. Some colorless sulfur immediate vicinity which is beneficial for the organism because bacteria are capable of oxidising ferrous iron aerobically, but it reduces losses of photosynthate by photorespiration. this process occurs only at extremely low pH (-2), which does not occur in coastal microbial mats. Species such as Nitrogen fixation: without it coastal microbial mats Gallionella ferruginea or Sphaerotilus natans which oxidise would not develop ferrous iron at neutral pH are unlikely to be able to compete with the chemical oxidation of iron in microbial mats Nitrogen comprises 7-1 0% of cell dry weight matter and rep­ (Emerson and Revsbech 1994). resents therefore the second most important element. In the 406 SIal

A B ',,-<:,-,- \+ .... ,..\ o O2 ;o~ ~~~ .!-=- . EPS .. __ Fe 2~ . ! ~- ~ Cyanobacterium •. Fe2 ~ t NH3 - .. CyanobactsJium NH3' j I r j Glycogen I : Glycogen , I C0rl:l~lible sol,ute ,-- .C~l1lprf solul~ - c:'] Acetate (formate, lactate, ethanol S2·

CO2, H2 ~sg.B 1

Figure 2: A conceptual model showing the possible role of iron accumulating in the sheath of a mal-forming cyanobacterium. The inserted photo shows a transmission electron microscope image of the cyanobacte rium Microcoleus chthonoplastes with iron accumulated al the sheath visi ble as electron-dense particles. A. During the day, the cyanobacterium fixes CO, as result of photosynthesis. The photosynthate is used for the largest part to fill up the glycogen sto rage pool. It is further used for the synthesis of structural cell material and compatible solute (osmotica) and is excreted as extracellular polysaccharides (E PS). The organism evolves 0 2. This may result in photorespiration and a loss of photosynthate as glycolate. In order to prevent 0 2 from accumulation, ferrous iron in the sheath is oxidised. B. During the night, gl ycogen and the compatible solute is fermented and concomitantly sulfur is reduced to sulfide. The fermentation products are end-oxidised by sulfate-reducing bacteria (SRB) and sulfide is formed. Th e sulfide reacts with ferric iron, preventing th e accumulation of sulfide to toxic lev­ ers while regenerating ferrous iron.

cell it is mainly present in its redu ced form . Ironically, in its isms can be distinguished with resp ect to their strategy to fix most ubiquitous form, atmospheric dinitrogen (N,) it is not nitrogen in two major groups (Stal 1995, Bergman ef al. accessible to most organisms and the bound forms such as 1997). One grou p is capable of fixing nitrogen on ly under nitrate, ammonia or organiC nitrogen are usually in low sup­ anaerobic conditions and its strategy can be described as ply, particularly in the marine environment. A limited group 'avoidance of oxygen'. There are a number of ways by which of, excl usively, prokaryotic organisms possesses the ability this can be achieved. First of all, nitrogen fixation ca n be to reduce the extre me stable triple bond between the two confined to the dark period, when the mat has become anox­ molecules of N in N2. These organisms all co ntain the ic. The disadvantage of this strategy is that on ly a limited en zyme co mplex nitrogenase which catalyses the redu ction nitrogenase activity can be sustained under such conditions of N, to NH, at the expense of a large amount of energy and because 01 the low energy yield of fermentalion. low-potential electrons (ferredoxin) (Peters ef al. 1995). Experiments and calculations have unequivocally demon­ Among cyanobacteria many species are known to be capa­ strated that some nitrogen fixation can be supported by fer­ ble of lixing nitrogen. As oxygenic photo-autotrophic organ­ mentation by mat-forming cyanobacteria such as isms they are particularly we ll equipped to meet the energy Oscil/aforia Iimosa (Stal and Moezelaar 1997). Another way and electron demands of nitrogen fixation. On the other is to realise a spatial separation of oxygenic photosynthesis hand, nitrogenase is extremely sensitive to oxygen and dia­ and nitrogen fixation in Ihe mat. At the bottom of the zotrophic cyanobacteria the refore developed strategies to cyanobacterial mat oxygenic photosynthesis may not be circumvent this problem (Gallon 1992). Clearly, the best possible because only lar red light (>700nm) predominates ad aptation has been evolved by the heterocystous here, which is specilically harvested by the anoxygenic pho­ cyanobacte ria . These filamentous cyanobacteria differenti­ tosystem I (Stal ef al. 1985). Moreover, sulfide may al so ate special cells, the heterocysts, which have lost the ability reach higher concentrations in this part of the mat and this is of oxygenic photosynthesis and CO, fixation and wh ich have a potent inhibitor of oxygenic photosynthesis. It has indeed become the sites of nitrogen fixation. Nitrogen fixation in been shown that the specific (chlorophyll-normalised) nitro­ these organisms is strong ly light dependent although some genase activity increased with depth in the mat. Sulfide has activity can be sustained in the dark, driven by aerobic res­ been shown to induce nitrogenase in mat-forming cyanobac­ piration. Heterocystous cyanobacteria are remarkably rare teria in the light (Villbrandt and Stal 1996). The problem with in coastal microbial mats (Stal ef al. 1994). More frequently, this strategy is that the nitrogen-fixing cyanobacteria do not non-heterocystous filamentous cyanobacteria have been fix CO, (except when sulfide-dependent anoxygenic photo­ shown to fix nitrogen in these environments. These organ- synthesis occurs) and the cyanobacteria in the suriace lay- South Alrlcan Journal of Botany 2001. 67 399- 410 407

ers are still devoid of nitrogen. Tra nsport or fixed nitrogen tion in most microbial mats is accomplished by non-hetero­ outside the cel ls seems ineHicient and would rather favor cystous cyanobacteria and that th is occu rs predominantly non-diazotroph ic organisms in the mat. It seems more likely under anaerobic conditions implies that th is process is far that in this case th e cyanobacteria move up and down from eHldent. It can be anticipated that growth of the between the two sites . Mo tility of mat-rorming cyanobacteria cyanobacteria is strongly nit rogen-limited in these mats. has been shown in several occasions but these were all con­ trolled by light (Garcia-Pichel et al. 1994, Kruschel and Is nitrogen-limited growth of the cyanobacterial mat pre­ Castenholz 1998). Motility controlled by the nitrogen status of venting the formation of a stromatolite? cyanobacteria still awaits experimental proof. The second group of non-heterocystous diazotroph ic Wh en nitrogen is limiting growth of the mat-building cyanobacte ri a is capable of fi xing nitrogen under fu lly aero­ cyanobacteri a these organisms will divert th e fixed CO! to bic conditions (Bergman et al. 1997). It is not precisely non-nitrogenous compou nds, mainly . This know n by wh ich mechanism these organisms protect nitro­ mode of growth is termed 'un balanced'. Du ring balanced genase from oxygen inactivation. Although they are capable growth all cell components are synthesised in the right pro­ of di azot rophic grow th in culture under fu lly aero bic co ndi ­ po rtion s, but when nitrogen is unava ilabl e, , nucleic tions and under continuous illumination , they fix nitrogen acids and cell walls can not be produced. However, photo­ excluSively during the dark when grown under an alternating synthetic COl fixation is not impaired and carbohydrates are light-dark cycle. The sam e day-night pattern of nitrogen fix­ synlhesised. Intracellular, glycogen is stored, but since limit­ ation can be found in most microbial mats. Presu ma bly, oxy­ ed space is available polysaccharides are also excreted into gen concentrations reach too high levels in the mat during the medium as mucilage and sheath materia l. Many marine daytim e. Som etimes , two peaks of nitrogenase activity can microbial mats are the refore embedded in a th ick gelatinous be obse rved during a day-night cycle in a mat of aerobic polysaccharide matrix. nit rogen-fixing cyanobacteri a, on e each at sunris e and ­ Photosynthetic C02 fixation res ults in the formation of car­ set an d low or zero activity at night and during th e day bonate ion (CO..'). It has therefore been hypothesised th at (Villbrandt et at. 1990) . cya nobacterial photosynthesis would promo te calcification In the few exampl es of microbial mats formed by hetero­ (Krum bein and Giele 1979). However, in most marine micra­ cys tous cyanobacteri a, nitrogen fixation occurs during day­ bi al mats , calcification was not spatially associated with the tirne (Stal 1995). It is not precisely known why these organ­ cyanobacteria (Lyons et al. 1984, Chafetz and Bu czynski isms are excluded in many microbial mats. It has been sug ­ 1992) . Th is is rem arkable because with the concentration of gested that heterocystous cyanobacteria are more sensitive ca lcium ion in seawater it would easily exceed the solubi lity to su lf ide tha n non-heterocystous species (Howsley and product of calcium carbon ate such as aragonite Pearson 1979) . However, it is H,S which freely diffuses into and calcite. Thu s. it was obvious to su ppose a mechanism the cell and exerts is tox icity and because of Ihe alkaline that would prevent calcification. Since it is known that poly­ con di tions in coastal microbial mats the concentrations of saccharides can bind Ca ~· and/or Mg2. one possib le mecha­ this gas remains extrem ely low. Another suggestion is that nism that has be en proposed is that the extracellular poly­ heterocystou s cyanobacteri a do not tole rate anoxic condi­ saccharides produced by the cya nobacteria in the mat tions in the dark. Indeed, mats of heterocystous and non­ would lower the activity of these io ns to such extend that heterocystous cyanobacteria occurring close to each other calcium carbonate precipitation does not occur (Borman et on an inter-tidal fl at in Baja California, Mexico , differed al. 1982, Westbroek et al. 1994). Alternatively, polysaccha­ markedly in their oxygen dynamics (Stal 1995). The mat of rides may interact with crystallisatio n nuclei , preven ting their the heterocystous cyano bacterium Calothrix sp. did not turn growth, or a com binatio n of both mechanisms. anoxic during the night and lacked the black layer of FeS Hence, a chronic nitrogen depletion occurring in many and purple sulfur bacteria. Th is also hinted to the absence of marine mic rob ial mats, lead ing to excessive extracellul ar su lfide in th is mat. During daytime oxygen concentrations in polysacchari de prod uction , may offer an explanation for the the Calothrix mat did not reach excessive high levels. This, fact that th e majority of these systems do not lith ify and pro­ and th e availability of oxygen during the dark could have been du ce stromatolites. critical for the heterocystous cyanobacte rium to proliferate. The rarity of heterocystous cyanobacteria is not only the Concluding remarks case in marin e microbial mats but extends to the whole marine environment. The most important nitrogen-fixing The classic example of a microbial mat is built by cyanobac­ cyanobacterium in the marine is the non-hetero­ te ria and these oxygenic phototrophiC microo rganisms lulfill cystous Trichodesmium , whereas in fres hwater and brackish a key role in the systems function and metabolism. Most, if environments only heterocystous species occur (Paerl 1990, not all . of the primary production of these cyanobacleria is Moisander and Paerl 2000). Hence, it could also be that high more or less directly transferred to the micro bial mat system sal inity se lects again st heterocystous species , although the and not primarily used for the synthesis of structural cell precise mechanism remains to be elucidated. The mats of material. Probably the most important mechanism is through Calothrix in Baja California occur high in the inter-tidal sedi­ the excretion of fermentatio n products. But also other mech­ men ts and are rarely su bmersed and may be are therefore anisms are responsible for the transfer of fixed carbon to the not strongly influenced by the seawater. microbial comm unity that is fueled in this way. Th e solar Whatever may be the reason, the fact th at nitrogen fixa- en ergy fixed by the cyanobacteria is utilised by th e sulfate 408 Stal

reducing bacteria and the sulfide they produce, is utilised by macroscate cohesiveness of exopolymers. In: Krumbein WE, anoxygenic phototrophic bacteria and colorless sulfur bacte­ Paterson OM , Stal LJ (eds) Bioslabilization of sediments. BIS ria. In this way, the microbial mat functions as an internally Verlag, Oldenburg, pp 135-148 closed system , driven by solar energy. The mal behaves like Des Marais DJ (1995) The biogeochemistry of hypersaline microbial a lissue, displaying its own physiology. The nel exchange mats. Advances in 14: 251-274 with the atmosphere and geosphere is limited because the Dilling W, Cypionka H (1990) Aerobic respiration in su lfate-reducing bacteria. FEMS Microbiology Letters 71 : 123- 127 cycles of the elements are essentially closed. In many Emerson 0, Revsbech NP (1994) Investigation of an iron-oxidizing microbial mats, decomposition is complete but in a few other microbial mat community located near Aarhus, Denmark: field systems a slow accumulation of organic matter may occur. studies. Applied and Environmental Microbiology 60: 4022-4031 Rarely, microbial mats calcify and lithification results in the Fenchel T (1998) Formation of laminated cyanobacterial mats in the formation of stromatolites. absence of benthic fauna. Aquatic Microbial Ecology 14: 235-240 Fenchel T, KOhl M (2000) Artificial cyanobacterial mats: growth, Acknowledgements - This is publication 2803 of NIOO-CEMO. structure, and vertical zonation patterns. Microbial Ecology 40: 85-93 References FrOnd C. Cohen Y (1992) Diu rnal cycles of sulfate reduction under oxic conditions in cyanobacterial mats. Applied and Arieli B, Binder B, Shahak Y, Padan E (1989) Sulfide induction of Environmental Microbiology 58: 70-77 synthesis of a peripfasmic protein in the cyanobacterium Gallon JA (1992) Reconciling the incompatible: N2 fixation and 0 2. Oscillatoria fimnetica. Journal of Bacteriology 171: 699-702 New Phylologisl 122: 571 - 609 Bergman B, Gallon JR, Rai AN , Stal LJ (1997) N? fixation by non­ Garcia-Pichel F, Mechling M, Castenholz RW (1994) Oiel migrations heterocystous cyanobacteria. FEMS Microbiotogy Reviews 19: of microorganisms within a benthic, hypersaline mat community. 139-185 Applied and Environmental Microbiology 60: 1500-1511 Borman AH , De Jong EW, Huizinga M, Kok OJ, Westbroek p, Bosch Garcia-Pichel F, Prufert-Bebout L, Muyzer G (1996) Phenotypic and phylogenetic analyses show Microcoleus chlhonoplastes to be a L (1982) The role in CaC03 crystallization of an acid Ca~ ' -b in din g polysaccharide associated with of . cosmopolitan cyanobacterium. Applied and Environmental European Journal of Biochemistry 129: 179-183 Microbiology 62: 3284-3291 Canfield DE, Des Marais OJ (1991) Aerobic sulfate reduction in Gupta RS (2000) The phylogeny of proteobacteria: relationships to microbial mats. Science 251 : 1471- 1473 other eUbacterial phyla and . FEMS Microbiology Carmichael WW (1992) A review: Cyanobacteria secondary Reviews 24: 367-402 metabolites - The cyanotoxins. Journal of Applied Bacteriology Hansen TA (1994) Metabolism of sulfate-reducing prokaryoles. 72: 445-459 International Journal of General and Castenholz AW (1982) Motility and taxes. In: Carr NG, Whinon BA Molecular Microbiology 66: 165- 185 (eds) The of cyanobacteria. Blackwell Scientific Howarth RW, J0rgensen BB (1984) Formation of :\$S-labelled ele­ Publishers. Oxford. pp 413-439 mental sulfur and pyrite in coastal marine sediments (Limfjorden Chafetz HS. Buczynski C (1992) Bacterially induced lithification of and Kys ing Fjord, Denmark) during short·term 3 $ SO~~ reduction microbial mats. Palaios 7: 277- 293 measurements. Geochimica et Cosmochimica Acta 48; Cohen Y (1984) The Solar Lake cyanobacterial mats: strategi es of 1807-1818 microbial life under sulfide. In: Cohen Y, Cas tenholz AW. Howsley R, Pearson HW (1979) pH dependent sulphide toxicity to Halvorson HO (eds) Microbial Mats: Stromatolites. Alan A liss. oxygenic photosynthesis in cyanobacteria. FEMS Microbiology New York, pp 133-148 Lellers 6: 287-292 Cohen Y, Castenholz AW, Halvorson HO (1984) Microbial Mats: J0rgensen BB, Revsbech NP, Blackburn TH . Cohen Y (1979) Stromatolites. Alan A liss, New York Diurnal cycle of oxygen and sulfide microgradients and microbial Cohen Y. J0rgensen BB . Revsbech NP, Poplawski R (1986) photosynthesis in a cyanobacterial mat sediment. Applied and Adaptation to sulfide of oxygenic and anoxygenic pho­ Environmental Microbiology 38: 46-58 tosynthesis among cyanobacteria. Applied and Environmental Jullner F (1987) Volatile organic substances. In: Fay P, Van Baalen Microbiology 51: 398-407 C (eds) The cyanobacteria. Elsevier Science Publishers BV Cohen Y, Aosenberg R (1989) Microbial Mats. ASM, Washington, Amsterdam. pp 453-469 ' pp 494 Krei\low S, Mundt S, Undequist U (1999) Cyanobacteria - a poten· Cytryn E, Minz D. Oremland AS . Cohen Y (2000) Distribution and tial source of new biologically active substances. Journa l of diversity of corresponding 10 the limnological cycle of a Biotechnology 70: 61 - 63 hypersaline stratified lake (Solar Lake. Sinai, Egypt). Applied and Krumbein WE (1983) Stromatolites. The chaUenge of a term in Environmental Microbiology 66: 3269-3276 space and time. Precambrian Research 20: 493-531 De Philippis R. Vincenzini M (1998) Exoceffular polysaccharides Krumbein WE, Buchholz H, Franke P, Giani D. Giele C, from cyanobacteria and their possible applications. FEMS Wonneberger K (1979) 02 and H ~S coexistence in stromatolites. Microbiology Aeviews 22: 151-175 Naturwissenschafien 66: 381-389 Oe Wit R. Van Boekel WHM. Van Gemerden H (1988) Growth of the Krumbein WE, Cohen y, Shilo M (1977) Solar Lake (Sinai). 4. cyanobacterium Microcoleus chthonoplastes on sulfide. FEMS Stromatolitic cyanobacterial mats. and Oceanography Microbiology Ecology 53: 203-209 22:635-656 De Wit A, Van Gemerden H (1988) Interactions between pho­ Krumbein WE , Giele C (1979) Calcification in a coccoid cyanobac­ totrophic bacteria in sediment ecosystems. Hydrobiological terium associated with the formation of desert stromatolites. Bulletin 22: 135- 145 Sedimentology 26: 593-604 De Wit A, Van Gemerden H (1990) Growth of the phototrophic pur­ Kruschel C, Castenholz RW (1998) The effect of solar UV and visi­ ple sulfur bacterium Thiocapsa roseopersicina under oxic/anoxic ble irradiance on the vertical movements of cyanobacteria in regimens in the light. FEMS Microbiology Ecology 73: 69- 76 microbial mats of hypersaline waters. FEMS Microbiology Decho AW (1994) Molecular-scale events influencing the Ecology 27: 53-72 South African Journal of Botany 2001, 67 : 399-410 409

KOhl M, lassen C, Jorgensen 86 (1994) light penetration and light Microbiology 142: 72- 76 intensity in sandy marine sediments measured with Irradiance Rippka R, Oeruelles J, Wa terbury JB, Herdman M, Stanler AY and scalar irradiance fiber-optic microprobes. Marine Ecology (1979) Generic assignments strain histories and properties of Progress Series 105: 139- 148 pure cul tures of cyanobacteria. Journal of General Microbiology Larkin JM , Strohl WR (1983) Beggiatoa, Thiothrix and Thioploca. 111 : 1-61 Annual Review of Microbiology 37: 34 1-367 Rlsa tti JB, Capman we, Stahl OA (1994) Community struclure of a Lehmann M, Weber G (1976) Accumulation, mobilization and lurn­ microbial mat: the phylogenetic dimension. Proceedings of the over of glycogen in the blue-green bacterium Anacystis nidulans. National Academy of Sciences of the United Stat es of America Archives of Microbiology 111 . 93--97 91 : 10173-10177 Lyons WB, Long OT, Hines ME, Gaudette HE, Armstrong PB (1984) Schopf JW (2000) The fossil record: tracing the roots of th e Calcificati on of cyanobacterial mats in Solar Lake, Sinai. Geology cyanobaclerialltneage. In: Whitton SA, Potts M (eds) The ecolo+ 12:623-626 gy of cyanobacteria. Their diversity in time and space. Kluwer Margulis L, Hinkle G, Stolz J, Craft F, Esteve I, Guerrero A (1990) Academic Publishers, Oordrecht, pp 13-35 MobilifIJum chase/ - Morphology and ecology of a spirochete SIal LJ (1991) The su lfur metabolism of mal-building cyanobacteria from an intertidal stratified microbial mat communit y. Archives of in anoxic marine sediments. Kieler Meereslorschungen 8: Microbiology 153: 422-427 152- 157 Moisander PH , Paerl HW (2000) Growth , primary productivity, and Stal LJ (1994) Microbial mats in coastal environments. In: Stal LJ, nitrogen fixation potential of Nodularia spp. (Cyanophyceae) in Caumette P (eds) Microbia] Mats. Structure, Development and water from a subtropical es tu ary in the United States. Journal of Environmental Signi ficance. Springer Verlag, Heidelberg, NATO Phycology 36 : 645-658 AS I Senes, Vol G 35 : 21-32 Nicholson JAM , Stolz JF, Pierson BK (1987) Slructure of a microbia! SIal LJ (1 995) Physiological ecology of cyanobacteria in microbial mat at Great Sippewisselt Marsh, Cape Cod, Massachusells. mats an d other communities. New Phytologist 131 : 1-32 FEMS Microbio logy Ecology 45: 343-364 Sial LJ , Caumette P (1994) Microbial mats. Structure, development Nold SC, Ward OM (1996) Photosynthate partitioning and fermen ­ and environmental significance. Springer Verlag, Heidelberg, tation in hot spring microbial mat communities. Applied and NATO ASI Seri es G35. pp 463 Environmental Microbiology 62: 4598-4607 SIal LJ , Moezelaar R (1997) Fermentation in cyanobactena. FEMS Oremland AS, Polcin S (1982) Methanog enesis and sulfate reduc­ Microbiology Reviews 21 : 179-211 tion : competitive and noncompetitive substrates in estuarine sed­ Sta l LJ , Paerl HW, Bebout B, Villbrandt M (1994) Heterocystous ver­ iments. Applied and Environmental Microbiology 44: 1270-1276 sus non-h eterocystous cyanobacteria in microbial mats. In: Stal Overmann J, Cypionka H, Pfennig N (1992) An extremely low-light­ LJ , Caumette P (eds) Microbial mats. Structure, development an d adapted phototrophic sulfur bacterium from the . environmental significance. Spri nger Verlag, Heidelberg, NATO limnology and Oceanography 37: 150-155 ASI Series vol. G 35: 403-414 Paerl HW (1990) Physiological ecology an d regulation of N! fixation Stal LJ , Van Gemerden H, Krumbein WE (1985) Structure and in natural waters. Advances in Microbial Ecology 11 : 305-344 development of a benthic marine microbial mat. FEMS Paerl HW, Fitzpatrick M, Bebout BM (1996) Seasonal ni trogen fixa­ Microbiology Ecology 31: 111-125 tion dynamics in a marine microbial mat: Potential roles of Tabita FR (1994) The biochemistry and molecular regulation 01 car­ cyanobacteria and microhe terotro phs. Limnology an d bon dioxide metabolism in cyanobactena. In: Bryant OA (ed.) The Oceanography 41 : 419-427 molecular biology of cyanobacteria. Kluwer Academic Publishers, Peters JW, Fisher K, Dean OA (1995) Nitrogenase structure and Dordrech l. pp 437-467 function: A biochemical-genetic perspective. Annual Reviews of Tease B, Jurgens UJ, Golecki JA, Heinrich UR. Rippka R, Microbiology 49: 335- 366 Weckesser J (1991) Fine-structu ral and chemical analyses on Price GO , Badger MR (1991) Evidence for the role of carboxysomes inner and outer sheath of the cyanobacterium Gloeorhece sp . in the cyanobacterial CO ~ -conce ntra t ing mechanism. Canadian PCC-6909. Antonie Van Leeuwenhoek International Journal of Journal of Botany 69: 963--973 General and Molecular Microbiology 59: 27-34 Raskin L, Rittma n BE, Stahl OA (1996) an d coexistence Teske A, Aamsing NB, Habicht K, Fukui M, KOver J, Jorgensen BB, of sulfate-reducing and metha nogenic populations in anaerobic Cohen Y (1 998) Sulfate-reduci ng bacteria and their activities in blofilms. Applied and Environmental Microbiology 62: 3847-3857 cyanobacterial mats of Sola r Lake (Sinai, Egypt). Applied and Aeed RH , Borowitzka LJ , Mackay MA, Chudek JA, Foster R, WarT Environmental Microbiology 64 : 2943--2951 SCA, Moore OJ, Stewart WOP (1986) Organic solu te accumu la­ Thode-Andersen S, Jorgensen SS (1989) Sulfate reduction and the ti on in osmo tically stressed cyanobacteria. FEMS Microbiology formation of l~ S-labeted FeS, FeS2, and sai n coastal marine sed­ Reviews 39: 51-56 iments. Li mnology and Oceanography 34: 793-806 Aeid RP, Browne KM (1991) Intertidal stromatolites in a fringing Van Gemerden H (1968) On the ATP generation by Chromatium in Holocene reef complex, Bahamas. Geology 19: 15-18 darkness. Archiv fOr Mikrobiologie 64: 118-124 Reinhold L, Kosloff R, Kaplan A (1991) A model for inorganic carbon Van Gemerden H (1983) Physiological ecology of purple and green fluxes and photosynthesis in cyanobacterial carboxysomes. bacteria. Annales de Microbiologie (Inst. Pasteur) 134 8 : 73-92 Canadian Journal of Botany 69: 984-988 Van Gemerden H (1993) Microbial mats: A joint venture. Marine Renstrom E, Bergman B (1989) Glycolate melabolism In cyanobac­ Geology 113: 3-25 teri a. I. Glycolate excretion and phosphoglycolate phosphatase Van Gemerden H, De Wi t A, Tughan CS. Herbert AA (1 989) activity. Physiologia Plantarum 75' 137-143 Development of mass blooms of Thiocapsa roseopersicina on Revsbech NP, Jorgensen BB, Blackburn TH, Cohen Y (1983) sheltered beaches on the Orkney Islands. FEMS Microbiology Microelectrode studies of the photosynthesis and 0 " H ~ S and pH Letters 62 : 11 1-118 pro files of a microbial mat. Limnology and Oceanography 28: Van Gemerden H. Tug han CS, De Wit A. Herbert RA (1989) 1062-1074 Laminated microbial ecosystems on sheltered beaches in Scapa Riegman R, Mur LR (1985) Effects of photoperiodicity and light irra­ Flow, Orkney Islands. FEMS Microbiology Ecology 62 : 87- 102 diance on phosphate-limited Oscilfatoria agardhii In chemosta! ViJibrandt M, Stal LJ (1996) The effect of sullide on nilrogen fixation cultures. [I. Phosphate uptake and growlh. Arch ives of in he terocys tous and non-heterocystous cyanobacterial mat com- 410 Stal

munities. Algological Studies 83: 549-563 20: 1-790 Villbrandt M, Star LJ, Krumbein WE (1990) Interactions between Westbroek P, Buddemeier B, Coleman M. Kok DJ, Fautin D, Sial LJ nitrogen fixa tion and oxygenic photosynthesis in a marine (1994) Strategies for the study of climate forcing by ca lcification. cyanobacterial mat. FEMS Microbiology Ecology 74 : 59-72 Past and Presence Processes: 37-60 Visscher PT, Prins AA. Van Gemerden H (1992) Rates of sulfate Widdel F, Schnell S, Heising S, Ehrenreich A, Assmus B, Schink B reduction and thiosulfate consumption in a marine microbial mat. (1 993) Ferrous iron oxidalion by anoxygenic phototrophic bacte­ FEMS Microbiology Ecology 86 : 283-293 ria. Nature 362: 834-836 Wachendorfer V (1991) Parahislologische und sedimentmikrobiolo­ Yallop ML. De Winder B, Paterson OM, Sial LJ (1994) Comparative gische Untersuchungen an einem potentiellen silikoklaslischen structure, primary production and biogenic stabilization of cohe­ Stromatolithen. Thesis University of Oldenburg, Oldenburg, sive and non-cohesive marine sediments inhabited by microphy­ Germany tobenthos. Estuarine, Coastal and Shelf Science 39: 565-582 Walter MR (1976) Stromatolites. Developments in Sedimentology

Edited by M Bjork and RN Pienaar