Microbial Mats

Scientists of diverse interests, and even laymen, are beginning to understand what is meant by a “microbial mat”. But this knowledge is still superficial even for the specialist, both figuratively and literally. Most of what we know of microbial mats up to this point is still relegated to the superficial upper layer or layers which are usually composed mainly of photosynthetic prokaryotes. It is these that are responsible for the accretion and for the conversion of atmospheric and aqueous CO₂ into reduced energy-rich compounds on which the rest of the community is dependent. It is also these superficial layers that define most of the Precambrian remnants of microbial mats, the lithified mats or stromatolites which also contain the microfossils that are often recognizable as morphotypes of cyanobacteria. The laminae or underzones of even living mats are still “black boxes” for the most part, with little known of the actual microbial makeup, the matrix of exopolymers, and of the species interactions and metabolic activities. Although the microbial mat community may be regarded as an ecosystem characterized by primarily a molecular flow of carbon and chemical energy (since grazing invertebrates and even smaller phagotrophs are generally rare or absent), most mat communities are now known to be of considerable taxonomic and biochemical complexity, and a great diversity of communities exists.

Microbial mats develop in a wide variety of environments such as hot springs, hypersaline ponds, dry and hot deserts, alkaline lakes and coastal intertidal sediments (Cohen et al. 1984; Cohen and Rosenberg 1989). A noticeable similarity of these environments is their relative low species diversity. Particularly, multicellular organisms are excluded from such environments and it has been conceived that the absence or limited activity of grazing organisms is an important prerequisite for the development of microbial mats (Walter 1976). At first sight such conditions seem not to apply to coastal environments. However, intertidal sediments are alternatingly inundated and exposed to the atmosphere and this will cause fluctuations of water content, salinity and temperature. Fluctuations of environmental conditions are more extreme in the higher parts of the intertidal flats where microbial mats develop particularly well (Stal et al. 1985; Pierson et al. 1987).

We have conducted long-term studies of microbial mats of hot springs as model systems for investigating composition and structure of natural microbial communities and as modern analogs of stromatolites (Ward et al. 1984, 1987, 1989b). As recently as the last symposium on microbial mats our knowledge of the biodiversity within these communities was based solely on microbial species cultivated from such mat systems. The view has changed dramatically in the intervening years, because of the application of culture-independent techniques to recover and probe ribosomal RNAs (rRNAs) as biomarkers representing individual mat community members (Ward et al. 1992; Weller et al. 1992 and references cited therein), and more recently because of the renewed attempt to cultivate more relevant species. Here, we summarize what is currently known. The evidence suggests that the molecular methods we are using (Ward et al. 1992) may still not permit a complete understanding of the true complexity of the community. However, the approach does provide insight into understanding the basis behind this large biodiversity and into how more relevant species can be cultivated.

The thermomineral sulfurous waters at Mangalia in southeastern Dobrogea, Romania, have been known and used as spa facilities for well over 2,000 years (Feru and Capotà 1991). Hydrogeologieal studies performed during the last 60 years (Macovei 1912; Ciocîrdel and Protopopescu-Pache 1955; Moissiu 1968; Feru and Capotà 1991) identified a deep captive sulfurous aquifer located in Barremian-Jurassic limestones, extending 15 km to the North and 50 km to the South of Mangalia. In the Mangalia region, a system of geological faults allows the deep water to ascend toward the surface and mix with the Sarmatian oxygenated waters (Lascu et al. 1993). The biological investigation of the subsurface ecosystems associated with the sulfurous waters at Mangalia commenced in the late eighties, after the discovery of Movile Cave and its unique subterranean chemoautotrophically based ecosystem (Sarbu, 1990).

Laminated microbial mat ecosystems, dominated by phototrophic microorganisms are widely distributed in geothermal, hypersaline and intertidal marine environments (Castenholz 1984; Cohen 1984; Jørgensen and Cohen 1977; Stal et al. 1985; Van Gemerden et al. 1989a, b). The laminated structure of these vertically stratified benthic ecosystems is the result of the development of the constituent microflora along physical and chemical gradients. Thus, in a typical intertidal marine microbial mat dense populations of diatoms, cyanobacteria, purple and green sulfur bacteria, sulfate reducing bacteria and colourless sulphur oxidising bacteria coexist and compete for light, CO₂ and sulfide in the top few millimetres of the beach surface. In these systems a green layer of oxygenic cyanobacteria and unicellular green algae overlies a pink/red layer of anoxygenic purple sulfur bacteria which in turn is underlain by a black layer composed of precipitated ferrous sulfide (FeS). Cyanobacteria are the key component of most microbial mats since they thrive in nutrient poor, unstable and hostile environments. Filamentous species belonging to the genus Oscillatoria, able to fix dinitrogen, are commonly found as the initial colonisers in such environments (Stal and Krumbein 1981). They not only enrich intertidal marine sediments with low molecular weight organic compounds as a consequence of photosynthate excretion and cell lysis but also play an important role in stabilising sediments and reducing evaporative water loss (Stal et al. 1985). The released organic carbon is rapidly mineralised by heterotrophic bacteria resulting in the development of anoxic conditions. In the absence of oxygen populations of sulfate reducing bacteria become established.

Morphogenesis in stromatolites refers to the pattern of development that creates a megascopic laminated form through repeated cycles of sediment accretion. In general, all sedimentary laminations reflect periodic interruptions in deposition. For example, in stromatolites formed in detrital sedimentary systems, laminations may arise through variations in current velocity and rates of sedimentation (Farmer and Des Marais 1993). In mineralizing sedimentary systems, laminae may form by fluctuations of temperature, pH, or saturation (Guo and Riding 1992), or by seasonal growth cycles of organisms (Chafetz et al. 1991). Laminations can also originate by post-depositional (diagenetic) processes, such as dissolution (Ekdale and Bromley 1988).

In microbial mats, the growth of constituent populations and sediment accretion are accompanied by burial of phototrophic organisms in sediment layers that are anoxic and not reached by light. Nevertheless, cyanobacterial remains are often preserved in the sedimentary column. In the virtual absence of bioturbation, the original stratigraphy remains conserved in the sedimentary record. Hence, sediments with microbial mats often comprise an actively growing mat on top of a laminated system corresponding to buried mats of previous years, as described for Solar Lake (Egypt) by Jørgensen et al (1988). We have studied similar laminated systems in solar saltern ponds that have been exploited for more than one century.

The shortest wavelengths of the solar spectrum reaching the ground, known as the solar ultraviolet (UV, 280–400 nm), amount to only a small percentage of the total incident irradiance (Table 1). However, they may produce disproportionately large biological effects because they activate many photochemical reactions involving molecules of central biological importance. The consequence of these photoreactions is usually detrimental to the cellular metabolism. Mutation and lethal DNA and RNA damage, lipid peroxidation, photoinhibition of photosynthesis and respiration, and specific enzyme inactivation can all be a result of exposure to UV (Jagger 1985). These physiological effects are not always a laboratory curiostity but have ecological implications under natural conditions: the solar UV has been recognized as an important environmental stress factor in a variety of microbially dominated ecosystems. UV has been shown to cause sustained inhibition of both primary productivity and bacterial production in planktonic communities and to drive marked changes in species composition in periphyton communities (see Worrest 1982), but not much is known about the significance of ultraviolet radiation in benthic ecosystems. We have gathered some direct experimental evidence, as well as strong indirect inference from ecophysiological studies in diverse mat communities, suggesting that UV may be of importance, at least for the top phototrophic layers.

The presence of inland saline and alkaline lakes is one of the most original limnological features of the Iberian Peninsula. These waterbodies are associated with endorheic phenomena that occur in arid and semi-arid climates. On a global basis, inland saline lakes are common (Williams 1986), but in Western Europe, these lakes are confined to the Iberian Peninsula. The Spanish continental salt and soda lakes lakes occur in groups known as endorheic centers (see Fig. 1). These lakes encompass a broad variety of different physico-chemical conditions, seasonal fluctuations and sediment features. Lake Chiprana (Aragon) is the only deep permanent hypersaline lake in Spain (Guerrero et al. 1991). In contrast, all other inland saline and alkaline lakes in Spain are temporal water basins which contain water during 3 to 10 months of the year. During the inundation periods, the water level and concomitantly the salinity is subjected to extreme temporal fluctuations. The ionic composition of these lakes has been studied previously (Montes and Martino 1987; Comin and Alonso 1988). In summary, the Spanish inland lakes encompass a wide spectrum of ionic types: in the Andalusian lakes, sodium predominates with either chloride or sulphate as the dominant anion; the lakes in La Mancha mainly contain magnesium sulphate; and the lakes in Aragon encompass a whole range from sodium chloride to magnesium sulphate.

Hypersaline evaporitic systems are one of the places where microbial mats usually develop. In such environments the evaporation of seawater leads to a mixture of salt deposits, in which NaCl predominates and MgCl₂, MgSO₄, CaSO₄, KCl, NaBr, CaCO₃ are present as minor components. These salts precipitate at different stages of seawater concentration due to their different solubilities. Calcite is the first salt that precipitates, followed by gypsum and finally halite. In solar salterns the process of salt precipitation is spatially distributed in different ponds with increasing salinities. Generally, the salinity in the different ponds is kept constant and the environmental conditions are well defined (Barbé et al. 1990).

Almost all natural sediment deposits support the growth of microbial assemblages. On occasions these assemblages become sufficiently developed to be recognisable as microbial mats or biofilms but even before the development of such extensive structure, microbial colonisation can influence the microstructure of the bed and the critical erosion threshold for the resuspension of particles. The entrainment of particles is a critical event in transport of sediments and as yet there is still insufficient knowledge to accurately predict the response of natural sediments to hydrodynamic stress. Early attempts to predict sediment behaviour were for the most part based on the relationship between hydrodynamic stress and particle size (see Miller et al. 1977). This type of relationships, such as the Shield’s curve, have proved to be reasonably reliable for abiotic well-sorted sandy sediments. However, the erosional behaviour of sediment mixtures, small cohesive sediments and, to a large extent, natural sediments

in situ

defies adequate description. Many workers now recognise that predictions based purely on physical parameters are insufficient to explain even the erosional behaviour of non-cohesive sediments

in situ

(Paterson and Daborn 1991; Daborn et al. 1993). It has long been suggested that the biological “status” of the sediment may influence transport processes (Carter 1933), but there are still relatively few studies that deal directly with this problem. There is increasing evidence in the literature that as Montague (1986) stated:

“The physical and chemical properties of sediments that contain a significant biotic community are radically different from those of abiotic sediments ”.

The cyanobacterial species “Microcoleus chthonoplastes” is defined by the presence of many typically oscillatorian filaments with tapered, conically shaped terminal cells within a common sheath both in the botanical (Geitler 1932) and, more recently in the bacteriological literature (Castenholz 1988). The significant and often dominant ecological role of Microcoleus species in the formation and stabilization of intertidal and hypersaline microbial mat systems around the world is well recognized. However, it is not known if geographically disjunct populations referred to as “Microcoleus chthonoplastes” on the basis of gross morphological traits alone constitute a close genetic unit, or are in fact examples of evolutionary convergence driven by the adaptive advantages of a “microcoleus-like” morphology in benthic marine environments. What adaptive advantages this morphology may confer has been difficult to ascertain since these features, most notably the defining features of bundle formation, and presence of conically-shaped terminal cells, are not observed in currently available cultures.

In the coastal dunes of the Netherlands phototrophic organisms are colonising the sand surface (De Winder et al. 1989a). In previous studies the response of the three representative organisms, the cyanobacteria Crinalium epipsammum and Tychonema sp. and the green alga Klebsormidium flaccidum, to re- and dehydration (De Winder et al. 1989b, 1990b) was investigated. Equilibration with water-saturated air did only enable photosynthesis when this was granted by the water-retention characteristics of the physical environment of the samples. Rehydration enabled the recovery of photosynthesis of desiccated samples on substrata with a good water retention only. Photosynthetic activity is thus ascertained at environmental conditions of non-liquid water when this is enabled by a “compatible matrix”. The cyanobacterial strains showed a recovery of photosynthesis instantaneously. In contrast, rewetted cells of the green alga showed a recovery of photosynthesis only after a time-lag. During controlled dehydration K. flaccidum and to a lesser extent C. epipsammum proved to be able to maintain full carbon-fixing activity until the water content of its surroundings approached zero. Tychonema sp. appeared to reduce its activity in response to becoming dehydrated. The experiments described substantiated the view on the ecology of these drought tolerant organisms. The cyanobacteria as the initial colonizers inhabiting the barren sand are able to react quickly to changing water availability, by an on and off switching of their metabolism. As a result of the growth of the cyanobacteria the water retention of the top soil is improved to conditions were the green alga K. flaccidum is adapted.

Microorganisms that live in hypersaline environments need to maintain a high osmotic pressure in their cytoplasm to balance the osmotic pressure caused by the high salt concentrations. Most halophilic and halotolerant eubacteria and eukaryotic protists exclude salt, and synthesize or accumulate high concentrations of organic solutes that do not interfere with intracellular enzymatic activities (“osmotic solutes” or “compatible solutes”).

During the recent few years there has been a considerable expansion of our capabilities for microsensor analysis of chemical microenvironment and rates of microbial metabolism within microbial mats. Since 1987 we have thus seen published accounts of new electrochemical microsensors for N₂O, NO₃-, and NH₄+; the O₂ microsensor has been dramatically improved, and computerized equipment for remote operation of microsensors at great water depths has been developed. It now also seems that the electrochemical microsensors for analysis of chemical species in microbial mats may get competition from microsensors with optical signal transduction. The technological advances have enabled us to perform detailed studies of the nitrogen cycle within intact sediments and microbial mats. It has also been possible to perform very detailed studies including total oxygen budgets of photosynthetic microbial mats, and we have thereby verified the microsensor method for determining photosynthetic rates. This chapter will describe both the recent technological advances and the new insights we have gained by the new technologies.

Photosynthetie microbial mats occur as dense stratified communities in top layers of sediments, where microorganisms and sediment particles are embedded in an extracellular polymer matrix (e.g., Farbstreifensandwatt, Stal et al. 1985) or growing as a thin photosynthetic biofilm on solid substrata, e.g., stone or plant surfaces (Kühl 1993). In extreme environments, like hypersaline salt marshes and hot springs, regular mats several mm’ to cm’ thick and composed of almost pure biomass and exopolymers develop (Cohen and Rosenberg 1989). Photosynthetic microorganisms are the predominant component of these microbial mats, which often exhibit a vertical stratification of different colored layers due to the presence of photosynthetie microalgae and bacteria containing different photopigments with depth (Nicholson et al. 1987; Pierson et al. 1990). High metabolic rates due to the high density of microorganisms in mats combined with molecular diffusion acting as the major transport mechanism result in steep chemical gradients with depth as has been demonstrated by microelectrode measurements at < 50–100 µm spatial resolution (Revsbech and Jørgensen 1986; Revsbech this volume). Production and consumption of the major electron acceptors can be calculated from measured microprofiles and different functional layers can thus be identified in microbial mats from microelectrode measurements. The typical sequence found is an upper oxygenic photosynthetie layer with concurrent oxygen respiration and a lower anoxic layer with denitrification and sulfate reduction as the predominant respiratory processes and with anoxygenic photosynthesis, provided sufficient light is penetrating from above (Jørgensen et al. 1983; Jørgensen and Des Marais 1986b; Revsbech et al. 1989; Kühl 1993).

The usual procedure to measure CH₄ concentration profiles in sediments or submerged soils is by extraction of the porewater and analysis of the extracted CH₄ by gas chromatography (e.g., Martens and Berner 1977). However, this method has a very low spatial resolution of usually > 1 cm, is destructive for the environment, and is also laborious so that per day only a few vertical profiles can be analyzed. An alternative to extraction is the application of peepers (Hesslein 1976) which, however, have to be left for equilibration in the sediment for at least one week and furthermore, have a limited spatial resolution and require special precautions if used in anoxic sediments (Adams and Fendinger 1986; Brandi et al. 1990).

A variety of traditional techniques for light microscopy (e.g., differential interference contrast, phase contrast, fluorescence) and electron microscopy (transmission and scanning) have been used to obtain valuable information on the structure and species composition of microbial mat communities. Recent advances in techniques and instrumentation for microscopy are providing new tools for studying these communities in situ. Innovations in immunohistochemistry and molecular biology allow in situ labeling of specific microorganisms with fluorophores or gold particles which have been conjugated to antibodies or oligonucleotides. Digital image analysis software and confocal laser microscopy have made serial sections and 3-D reconstruction possible, providing structural data on biofilms, microbial mats, and microfossils in petrographic sections. This review will highlight a few examples of these new technologies with emphasis on those which have potential use in microbial mat studies.

Light dependent movement of gliding cyanobacteria has been studied for more than a hundred years and several types of response to light have been observed (Häder 1987). The following concepts have been recommended by an ad hoc committee on behavioral terminology, (Diehn et al. 1979): Phototaxis: Orientation of movement with respect to a directional light field; photokinesis: by which the steady-state speed of movement is related to the total light intensity and a photophobic response: where the movement is altered by a spatial or temporal change in light intensity. Most photosynthetic organisms display one or more of these behaviours in their search of suitable light conditions (Castenholz 1982).

The character of studies of microbial mats is being changed by the use of molecular genetic techniques. Since these techniques are not dependent upon the vagaries of enrichment and pure culture isolation, they promise a complete accounting of commumty structure and direct access to the study of microorganisms at the levels of populations and single cells. Thus, the integration of molecular techniques with more standard techniques (e.g. microscopy, microelectrodes, stable isotopes, radiotracers, and analytical chemistry) should prove a powerful synthesis. Our goal is not to provide a review of all the recent literature (and we apologize for any omissions in this regard) but rather to provide perspective on available and developing techniques.

Microbial ecosystems contain a large diversity of bacterial species. They are dominated by complex interactions between the different microorganisms, whereby each of the individual species has a specific role in the maintenance of the system. The active communities can efficiently scavenge nutrients from the environment and eliminate toxic compounds. However, not all of the species are active; most of them are dormant until environmental conditions change to favour their growth. Because of the great metabolic diversity and flexible organisation microbial ecosystems can be found nearly everywhere. Examples are the bacterial biofilms in waste water treatment reactors, on ship walls, or the microbial mats, found in hypersaline environments, tidal sediments and hot springs.

Microbial mat environments are characterized by thin-layered stratifications of microorganisms (Cohen 1989; Krumbein and Stal 1991) and geochemical gradients over small (μm) spatial scales (Jørgensen et al. 1983; Revsbech 1983). Sharp gradients in the concentrations of a variety of ions and molecules, such as O₂, CO₂, SO₄2-, and PO₄3-, have been observed within the layers of the mat. These gradients largely originate from the concentrated activities of specific groups of microorganisms within the mat layers (Jørgensen and Des Marais 1990). The relative stability of geochemical gradients (i.e. preventing rapid fluctuations) is important to maintain the efficient interactions between the different physiological groups of microorganisms contained within the mat layers.

Extracellular polymeric substances (EPS) represent a major fraction of microbial biofilms. The EPS within biofilms form a 3-dimensional network which is called biofilm matrix. The EPS immobilize the microbial cells within the biofilm and at the interface (Neu and Marshall 1990). The matrix is highly hydrated as it consists of more than 95% water. For this reason the biofilm matrix can be regarded as a layer of immobilized water (Cooksey 1992). So far the presence of the biofilm matrix has been demonstrated by various electron microscopic techniques which have been adapted from studies on microbial cell surfaces (Hancock and Poxton 1988; Mozes et al. 1991). The significance of EPS in biofilms has been elaborated in several review articles (Characklis 1973a; Characklis 1973b; Christensen 1989; Christensen and Characklis 1990; Geesey 1982; Neu 1992a). Nevertheless, the important compositional aspects of this matrix, mainly polysaccharides but also proteins, have been neglected in many studies on biofilms.

Algal mats are one example of a large range of microbial ecosystems which are spatially heterogeneous and dominated by physico-chemical gradients. Such gradients can be expressed over a huge range of physical dimensions. pH gradients at the nm level occur around clay crystal domains Bacterial colonies show oxygen gradients from fully saturated to anoxic over 20 to 30 µm whilst biofilms and algal mat communities range from tens of µm to mm in depth. Gradients over soil profiles are expressed in the cm range whilst stratified water bodies incorporate gradients that range from meters to hundreds of metres. As these examples indicate, probably the majority of microbial systems are spatially heterogeneous.

On a bacterial scale, the interface between a microbial mat and the overflowing water is a remarkable environment governed by low Reynolds number hydrodynamics, by diffusional solute transport, and by exposure to extreme chemical fluctuations and gradients. The thin surface layer of benthic phototrophic mats, in which all photosynthesis and most of the respiration of the mat community takes place, is generally about a mm thick. Yet, these mats have a productivity and organic turn-over of a similar magnitude as planktonic ecosystems. Since the euphoric zone of mats is typically 103–105-fold thinner than that of the water column, the microbial activity per unit volume is correspondingly 103–105-fold higher. As a consequence, there is a dynamic balance between the rapid production and consumption of oxygen, the concentration of which may fluctuate between > 1 atm. partial pressure and total depletion within minutes during shifting light conditions. The microorganisms living at the mat surface are physiologically adapted to these chemical extremes. Many also have a highly developed motility and tactic response to the environmental factors. The responses are often simple for the individual cells, but together they may lead to complex behavioral patterns of the whole populations. This paper will review some of the microbiologically important properties of the interface and how some bacteria are adapted to this environment.

Microbial mats are among the most productive aquatic ecosystems on Earth, yet, in many cases, the waters from which they grow are depleted in the basic nutrient elements. How, then, are nutrients cycled to allow for such high rates, and what ultimately controls these rates? To begin to address these issues, the cycling of carbon, oxygen, sulfur and nutrients has been explored over several years in Microcoleus chtholoplastes-dominated cyanobacterial mats from the hypersaline salt ponds in Guerrero Negro, Baja California Sur, Mexico (D’Amelio et al. 1989; Canfield and Des Marais 1993).

Although N-fixation has been assumed to contribute substantially to the overall N requirements of primary production in benthic microbial mat communities, few quantitative studies have been undertaken. Several factors have contributed to the lack of quantitative data including: 1) pronounced temporal variability in rates of N-fixation (which necessitate rate measurements over entire diel cycles), 2) difficulties in relating rates of nitrogenase activity (acetylene reduction) to actual N fixed (due to possible natural variations in this ratio, documented in other systems; Bothe, 1982), and 3) differences in methodologies used by various authors to assay N-fixation. These restraints on the acquisition of quantitative N-fixation data, as well as comparable restraints on obtaining reliable estimates of carbon fixation in these communities (Revsbech et al. 1981), result in some uncertainty as to the importance of N-fixation in supporting primary productivity in microbial mats.

Microbial mats are characterized by strong diel fluctuations in oxygen concentration which to a large extent can be attributed to the physiology of the cyanobacteria (Revsbech et al. 1983). In the light these organisms carry out oxygenic photosynthesis resulting in oxygen supersaturation of the mat. In the dark the primary mode of energy generation is respiration of endogenous glycogen (Smith 1982). However, in well-established microbial mats diffusion of oxygen in the mat will normally be insufficient to cover the demands and, as a result, the mat will turn anoxic. Evidently, the cyanobacteria have to switch to another way of energy generation in order to survive. Organisms isolated from such habitats were investigated for their mechanisms of anaerobic dark energy generation and turned out to be capable of fermentation. Forinstance, Oscillatoria limnetica, which thrives in the sulfide-rich hypolimnion of Solar Lake (Sinai) displays a homolactic fermentation using endogenous storage glucan as the substrate (Oren and Shilo 1979). Alternatively, this organism is capable of anaerobic respiration using elemental sulfur as the electron acceptor. Another mat-building cyanobacterium, Oscillatoria limosa, carries out a heterolactic and homoacetic fermentation simultaneously (Heyer et al. 1989). In the presence of elemental sulfur this organism stops producing hydrogen and sulfide is produced instead. Oscillatoria terebriformis was isolated from a hot-spring microbial mat and produces lactate form exogenous glucose or fructose (Richardson and Castenholz 1987).

Marine microbial mats typically support high rates of sulfate reduction. As a result, sulfur cycling is one of the dominant biogeochemical processes in these ecosystems. Detailed studies on inorganc sulfur cycling are available, but little is known about organic sulfur transformations. Sulfur-containing amino acids and dimethylsulfoniopropionate (DMSP) are the major biogenic precursors of volatile organosulfur compounds. DMSP, which functions as an osmolyte, is cleaved to acrylate and dimethylsulfide (DMS; White 1982). Production of DMS and methane thiol (CH₃SH) from decaying microbial mats in Yellowstone was reported by Zinder et al. (1977). Upon hydrolysis with NaOH, 200 μmol DMS 1-1 sediment was retrieved from intact marsh sediments (Kiene 1988), and DMSP, DMS, CH₃SH and dimethyldisulfide (DMDS) were found in slurried samples of a marine microbial mat (Visscher et al. 1991; Visscher and Van Gemerden 1993). Kiene and Visscher (1987) found production of CH₃SH from methionine and DMS from DMSP in anoxic salt marsh sediments. Visscher and Van Gemerden (1991b) reported rapid production of DMS, CH₄ and traces of CH₃SH in oxic and anoxic mat slurries when DMSP was added. In addition to cleavage, DMSP also undergoes demethylation, during which 3-methiolpropionate (MMPA) and 3-mercaptopropionate (MPA) are formed successively (Mopper and Taylor 1986; Kiene and Taylor 1989). MPA was a dominant thiol in anoxic sediments of Biscayne Bay, and DMSP addition resulted in increasing MPA production in slurries (Kiene and Taylor 1989). Sulfur-containing amino acids added to coastal sediments generated a variety of thiols, including CH₃SH and MPA (Kiene et al. 1990).

Studies of the distribution of stable carbon isotopes within ecosystems frequently offer useful insights into the structure and function of those ecosystems. The isotopic composition of the organic matter in a community such as a microbial mat is controlled by those enzymes which create and subsequently alter organic compounds. The extent to which these enzymes can affect the isotopic composition of organics is influenced by the relative fluxes of carbon, both within a community and between the community and its environment. Because the isotopic composition of organic matter in sedimentary rocks resists thermal alteration better than organic molecular structures, isotopic studies of ancient rocks have been important for Precambrian paleobiology (Hayes et al. 1992; Des Marais et al. 1992a).

In this manuscript we briefly discuss some of the open questions in relation to the response of photosynthetic microorganisms to changes in the ambient concentration of CO₂. We focus on the regulation and activity of the inorganic carbon-concentrating mechanism (CCM) in cyanobacteria and its contribution to the fractionation of stable carbon isotopes (δ13C).

Marine intertidal sediments are often colonized by dense populations of prototrophic microorganisms forming stratified communities with diatoms at the very surface and an underlying population of cyanobacteria (Stal et al. 1985). Underneath the layers of oxygenic phototrophs, purple and green sulfur bacteria frequently form additional colored bands (Nicholson et al. 1987). Microalgae in the top layers shade the underlying sediment of those regions of the light spectrum which they preferentially absorb (Jøgensen et al. 1987; Pierson et al 1987; Lassen et al. 1992b; Ploug et al. 1993). The distinct stratification, which is often observed for different types of phototrophic organisms, may thus be strongly influenced by their complementary utilization of the light spectrum. Below the layers of the oxygenic phototrophs, scalar irradiance in the visible spectrum (400–700 nm, PAR) is depleted 10–100 times more than light in the near infrared spectrum (NIR) (Pierson et al. 1990; Lassen et al. 1992b; Kühland Jørgensen 1992). The NIR absorption bands of the bacteriochlorophyll-protein complexes of the anoxygenic phototrophs are essential for the presence of thedense populations of these organisms underneath the oxygenic phototrophs.

During surface waterbloom formation a population of buoyant cyanobacteria becomes telescoped to the lake surface in the absence of mixing (Reynolds and Walsby 1975). Bloom formation involves an abrupt change in environmental conditions, most notably the exposure to full sunlight of cells in the top-layer of the bloom. In contrast, cells in deeper layers have to cope with anoxic, dark conditions (Ibelings and Mur 1992). Abeliovich and Shilo (1972) already observed photooxidative death of cyanobacteria in surface blooms. In addition to high irradiance stress, cells in the surface bloom are exposed to elevated temperature, and formation of a dry, desiccated crust has also been observed (Zohary and Pais Madeira 1990). The presence of additional stresses enhances the likelihood of photoinhibition (Demmig-Adams and Adams 1992). Photoinhibition is the decrease in quantum yield of photosynthesis induced by exposure to an irradiance, higher than that which can be used with maximum quantum efficiency (Powles, 1984). In this study we further investigated the impact of such conditions in a bloom on photosynthesis of the cyanobacterium Microcystis. Our investigations concentrated on the performance of Photosystem II, which is not only sensitive to photoinhibition, but to other stress factors as well (Havaux 1992).

The mat building photosynthetic microbes need nitrogen to produce biomass. The actual availability of combined nitrogen depends on the balance between sources and sinks in the mat. Sources are dissolved nitrogen in the overlying water, nitrogen from mineralization processes, and nitrogen fixation, and sinks are burial of nitrogen, efflux of dissolved nitrogen, and denitrification.

(N) limitation is a widespread feature of estuatine and oceanic waters (Dugdale 1967; Carpenter and Capone 1983). Despite this, benthic microbial mats frequently flourish in geographically-diverse intertidal and subtidal sand/mud flat, lagoon, reef, marsh and mangrove habitats characterizing such waters (Cohen et al. 1984; Cohen and Rosenberg 1989). Mats are particularly prolific in oligotrophic tropical and subtropical waters, where an appreciable portion of the primary production can be attributed to benthic microalgae (Whitton and Potts 1982; Bauld 1984). The remarkable success of mats in these N-limited environments has been attributed to the ability of specific groups of mat microorganisms (cyanobacteria, photosynthetic-, heterotrophic- and lithotrophic bacteria) to “fix” (reduce) atmospheric N (N₂), thereby providing biologically-available N (NH₃) to mat flora and fauna (Wiebe et al. 1975; Potts and Whitton 1977; Carpenter et al. 1978; Paerl et al. 1981; Stal et al. 1984). Nitrogen fixation is often the sole biologically-derived source of “new” N supporting primary production in these waters. In the face of N loss (as N₂) in mats (and other ecosystem components) via denitrification (Joye and Paerl 1993) and to a lesser extent ammonification, N₂ fixation is also of importance in regulating flux and mass balance of this important nutrient.

Microbial mats develop under a wide range of environmental conditions, and can be found in hypersaline coastal lagoons, hot springs, alkalinelakes, and marine intertidal flats (Cohen 1984, 1989; Jørgensen and Cohen 1977; Javor and Castenholz 1981, 1984; Jørgensen et al. 1983; Bauld 1984; Stal et al. 1985; Nicholson et al. 1987; Pierson et al. 1987). These laminated ecosystems characteristically are dominated by only a few functional groups of microbes. The driving force of most microbial mats is photosynthesis by cyanobacteria (CyaB) and algae. Subsequently, dissimilatory sulfate-reducing bacteria (SRB), using excretion-, lysis-, and decomposition products of CyaB, produce sulfide. The sulfide can be reoxidized to sulfate by colorless sulfur bacteria (CSB) and purple sulfur bacteria (PSB). Aerobic heterotrophic organisms are functionally important as their activity leads to oxygen depletion, and fermentative organisms provide growth substrates for SRB. In microbial mats these metabolically different groups of microbes live together in a layer of 5–10 mm thickness. Their combined metabolic activities result in steep environmental microgradients, particularly of oxygen and sulfide. Sulfide is inhibitory for most oxygenic phototrophs. Sulfide production immediately underneath the layer of CyaB might inhibit their growth, and, consequently, that of the entire ecosystem. On the other hand, anaerobic PSB and SRB are hampered by oxygen.

Marine microbial mats often develop under conditions, where diel extremes in oxygen and sulfide are found (Fig 1), which are a result of the fluctuation of incident light. During the dark period, when photosynthesis ceases, free sulfide can reach the surface of the sediment, while during the light period super-saturating oxygen concentrations prevail at the same depth horizon. As a result of changing gradients of O₂ and sulfide concentration, the flux of these compounds displays a diel pattern as well (Fig 2).

A typical inhabitant of a microbial mat is able to control its metabolism or to migrate vertically in response to the diel changes of light, redox conditions, and the concentrations of O₂, sulfide, organic substrates etc. The dissimilatory sulfate-reducing bacteria, however, were considered to be strict anaerobes, active in the reduced zones at the bottom of microbial mats, only. This view has changed now. New observations concerning the metabolism of sulfate reducers are coming from activity measurements in mats and from laboratory studies. However, to this point field and pure culture results give different pictures. My aim in this review is to discuss the metabolic versatility of sulfate-reducing bacteria. Some general principles derived from pure culture studies might be worthwile as guidelines, but will not close the gap. First, I shall describe the ambivalent relations of sulfate-reducing bacteria to oxygen. Then, the consequences of varying environmental conditions for sulfur metabolism are discussed.

Microbial mats are found along the outflow of continental thermal springs, on marine littoral sediments, in thalassic and athalassic (inland) hypersaline ponds and lakes, and in the deep-sea along hydrothermal vents. Hydrothermal vent microbial mats consist of chemotrophic sulphur bacteria thriving in the dark on sulphide which is mainly supplied by geochemical processes. However, it has been shown that sulphate reduction in these systems does occur up to 110 °C (Jørgensen et al. 1992), thus indicating that complete sulphur cycling also takes place. Microbial mats that are exposed to sunlight often comprise dense populations of oxygenic and anoxygenic phototrophic microorganisms together with chemoorganotrophs and chemolithotrophs. In habitats where the sulphur cycling is not a dominant process, microbial mats often originate from an association of cyanobacteria with anoxygenic filamentous phototrophic bacteria. For example, in thermal mats, it was shown that the Chloroflexus-like filamentous bacteria incorporated glycolate which was a major excretion product of the cyanobacteria found in the same mats (Bateson and Ward 1988). This commensalistic relationship is a nice example of a positive interaction of two organisms coexisting in the same environment. In other microbial mat environments, sulphide oxidation plays a predominant role. In the absence of geochemically formed sulphide, sulphide formation results mainly from the activities of sulphur and sulphate-reducing bacteria. In this paper, we discuss the biodiversity of sulphur bacteria and their ecological interactions in microbial mats.

Among the metals involved in biological processes, iron stands forth by a multiplicity of physiological and biochemical functions. It appears that organisms make use of iron in so many ways because this element not only exhibits a versatile chemistry, but is also wide-spread and in principle available in almost every type of environment.

Relative to biological demands, marine waters are often depleted in combined forms of nitrogen (Dugdale 1967). As such, the development of microbial mats in coastal marine environments is dependent on the fixation of atmospheric dinitrogen as the dominant source of biologically available N.

Today, microbial mats are confined to a restricted range of habitats including hypersaline and coastal marine environments, hot springs and alkaline lakes. In the last few years, these systems have been considered a major subject of study and many articles (Cohen et al. 1984; Cohen and Rosenberg 1989) have been published, which extensively revise the structure, physiology, genetics and evolution of microbial mats.

Sulfate-reducing bacteria and moreover methanogenic bacteria have long been considered to be strictly anaerobic organisms (Morris 1979; Skyring 1987; Widdel 1988). No light harvesting system has ever been described for these bacteria, on the contrary, several methanogens were found to be inhibited by exposure to light. It is therefore surprising to find accumulating data indicating that elevated levels of sulfate reduction and methane emission are found to be associated with photosynthetically active cyanobacterial mats when exposed to high intensity of solar irradiation in close proximity to supersaturation concentrations of oxygen (Cohen 1984; Giani et al. 1984; Canfield and Des Marais 1991; Fründ and Cohen 1992). Oxygen, which was found to be toxic at much lower concentrations to many isolates of both sulfate reducing and methanogenic bacteria does not seem to inhibit both processes (Hastings and Emerson 1988). Moreover, there are some indications that these processes are enhanced in cyanobacterial mats during noontime when supersaturation of oxygen is present in the photic zone (Fründ and Cohen 1992).

The round table discussion on “bioremediability and biological value of microbial mats” was unique within the context of the workshop for two reasons. Firstly, it was the only round table session that was not preceeded by a series of lectures. Secondly, this session was clearly focused on the interface between society and science. It thus draw attention to the role and the specific responsabilities of the scientist, and it posed the question how to develop applied research proposals. The session was split up in two parts focusing on the following subjects: 1) the use of microbial mats for bioremediation and in environmental technology; 2), biodiversity and conservation of microbial mats.

The discussion on new concepts in biogeochemical cycling and geophysical implications was chaired by D. Des Marais with H. Pearl, B. B. Jørgensen, and B. Bebout as panelists. Five major topics were addressed, controls on productivity, physical architecture of mats, nutrient recycling and conservation, the basis of depth zonation, and diversity and versatility in mat communities.

What have we learned from studying cyanobacterial mats? What is the influence of this study on microbiology and general biology?