Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

1 Introduction

The term microbial mat is used for multilayered microbial communities growing on sediments in diverse habitats such as tidal sand flats, hypersaline ponds, hot springs and other. Microbial mats are generally formed by filamentous, entangled organisms that produce a macroscopic ‘mat-like’ structure. In some cases such mats can indeed be peeled off from the sediment as a large coherent piece. However, benthic microbial communities of unicellular organisms, that usually do not form such coherent structures, are also called mats. Microbial mats exhibit great variety in morphology and composition, and they may include mats of diatoms and other biofilms of immobilized microorganisms (Bauld 1984). Nevertheless, eukarya are few or excluded altogether from environments in which extreme conditions prevail but in some cases meiofauna and other grazers are active in the habitats in which microbial mats are formed. One reason for the exclusion of eukarya is the wide spectrum of metabolic capabilities of bacteria and archaea and the great capacity these ‘prokaryotes’ display to adapt to changes and fluctuations in environmental conditions. Purple and sometimes green sulphur bacteria are normal components of most cyanobacterial mats (Nicholson et al. 1987; Pierson et al. 1987). This review focuses on mats formed by cyanobacteria.

The reason why cyanobacteria are typically the most successful mat-building organisms may be found in the combination of a number of the characteristic properties of this unique group of microorganisms. Cyanobacteria are the only oxygenic phototrophic bacteria and this metabolism is absent in archaea. As their predominant metabolism is oxygenic photosynthesis, cyanobacteria use light as an energy source, water as an electron donor and CO2 as a carbon source. These are the major requirements for growth and are abundant in the environments where most microbial mats are found. Another important property of many cyanobacteria, which is not shared by eukarya (and hence not by algae), is their ability to fix atmospheric N2, allowing them to grow independent of a source of combined nitrogen. Photosynthesis in cyano­bacteria saturates at low light intensity, cyanobacteria have a high affinity for light, and maintenance requirements are extremely low (Van Liere and Mur 1979). These properties allow photosynthesis even under extremely low light conditions. Moreover, several species are capable of sulphide-dependent anoxygenic photosynthesis (Garlick et al. 1977). Mat-forming cyanobacteria are well-adapted for life under anoxic conditions. In addition to the normal aerobic dark respiration, virtually all species of cyanobacteria in microbial mats are capable of fermentation (Stal and Moezelaar 1997). These properties of cyanobacteria are essential for life in microbial mats in which environmental conditions strongly fluctuate.

A typical property of microbial mats is their laminated structure in which different functional groups of microorganisms occur in vertically stratified layers (Stal et al. 1985). In addition to biological stratification biomineralogical stratification can be distinguished (Monty 1976). This type of lamination can be attributed to different growth periods, seasonal events, periodical events (e.g. tides) or episodic or erratic events (e.g. storms). Often, this laminated pattern is restricted, since most of the organic matter of the mat is degraded. When conditions allow, mats precipitate minerals, mainly calcite (Golubić 1973; Monty 1976; Krumbein 1979). This calcification is strongly associated with microbial metabolism and it may therefore give rise to the formation of distinct laminae and eventually to consolidated rock. Laminated rocks dating from the Precambrian and later eras are known as stromatolites (Krumbein 1983). Modern microbial mats built by cyanobacteria show remarkable similarities to fossil stromatolites. Stromatolites date back to 3.5 billion years (Mason and Von Brunn 1977; Lowe 1980; Walter et al. 1980; Orpen and Wilson 1981; Chap. 2). In some of these stromatolites well-preserved microfossils have been found that in some cases showed a remarkable resemblance to modern cyanobacteria (Schopf and Walter 1982; Awramik 1984; Chap. 2). It is therefore attractive to consider modern microbial mats as analogues of Precambrian stromatolites; however, structural differences do not always seem to justify this comparison. A major problem is the fact that the great majority of present day microbial mats does not form consolidated rock.

This review will discuss the metabolic activities of cyanobacteria that allow them to form microbial mats and stromatolites. This is a revised, updated and extended version of the chapter with the same title that appeared in the first edition of The Ecology of Cyanobacteria (Stal 2000).

2 Microbial Mats, Stromatolites and Their Environments

2.1 What Are Microbial Mats and Stromatolites? Some Definitions

Krumbein (1983), referring to the work of Kalkowsky (1908), proposed the following definition: “Stromatolites are laminated rocks, the origin of which can clearly be related to the activity of microbial communities, which by their morphology, physiology, and arrangement in space and time interact with the physical and chemical environment to produce a laminated pattern which is retained in the final rock structure”. This definition includes fossil as well as recent formations. Modern stromatolites that fit this definition are rare. Awramik and Margulis (in Walter 1976) defined stromatolites as: “Organosedimentary structures produced by sediment trapping, binding and/or precipitation as a result of the growth and metabolic activity of microorganisms, principally cyanophytes”. This definition includes fossil and recent consolidated stromatolites as well as unconsolidated microbial mats. Both definitions, however, emphasize the role of microbial mats and their microflora in the formation of stromatolites. Walter (1976) formulated the following conditions necessary to form a microbial mat:

  1. (i)

    environmental conditions must allow growth of the mat-building microorganisms; growth rate of the mat-building organisms must be faster than consumption by grazers;

  2. (ii)

    sedimentation rates should not be exceedingly high to allow stabilized colonization of the surface by the mat-building organisms;

  3. (iii)

    destructive forces from burrowing organisms and mecha­nical and chemical erosion must be absent or at least not prevent accretion of organisms.

In order to produce a stromatolite, preservation of the structure must occur. In modern day environments unconsolidated microbial mats are formed, i.e. systems that do not have the potential to preserve its structure, defined by Krumbein (1983) as: “Unconsolidated laminated systems, clearly related to the activity of microbial communities, often called recent stromatolites or living stromatolites are defined as potential stromatolites”. Indeed, stromatolites sensu Krumbein are still being formed today. Excellent examples of consolidated, well-laminated stromatolites formed by the growth and metabolic activity of a microbial mat are found in the Exuma Cays, Bahamas (Reid and Browne 1991; Pinckney et al. 1995). Stromatolites are just one form of calcified microbial mats that are jointly termed microbialites, a term that includes thrombolites, characterized by a cohesive macrofabric, and leiolites, which are without defined structure (Dupraz et al. 2009).

Also, non-consolidated, non-lithified microbial mats may leave traces of microbially induced sedimentary structures especially in siliciclastic deposits in shallow coastal environments (abbreviated as MISS) (Noffke et al. 2006). These structures have in some cases been preserved in the fossil record going back to the early Archean, emphasizing that not all Archean microbial mats were microbialites.

Since their discovery in 1961 in Shark Bay, Western Australia, poorly-laminated consolidated calcareous stromatolites (thrombolites) have been presented as strikingly similar to Precambrian stromatolites (Logan 1961). Contemporary calcareous stromatolites are also formed in Polynesian atolls (Kopara) (Défarge et al. 1994a, b) and other lacustrine and perimarine settings (Kempe et al. 1991; Kempe and Kazmierczak 1993). Such stromatolites can be called ‘modern stromatolites’ to distinguish them from fossil formations. It is not necessary to name unconsolidated microbial mats ‘potential stromatolites’, since they are not stromatolites (sensu Krumbein) and most of these microbial mats will never become such. Nevertheless, unconsolidated mats may keep in themselves the capacity for consolidation. This was shown by a transplantation experiment in which a non-lithifying microbial mat was placed in an environment with lithifying microbial mats. This mat calcified, demonstrating its potential to become a microbialite. Calcification clearly depends on the prevailing local environmental conditions (Dupraz et al. 2009).

Calcification is required for consolidation and preservation of microbial mats. In many microbial mats, calcification does not occur and the reasons for this are discussed later on in this chapter. Many consolidated rocks without a clear lamination are formed by microbial communities. It may be that laminations were lost during the process of diagenesis or that neither vertical stratified communities of microorganisms nor clear seasonal variations were involved in growth and metabolic activity. Another proposed mechanism is that carbonate sand accretes through trapping and binding in the microbial mat without in situ calcification. Such cohesive but poorly laminated microbialites are known as thrombolites (Kennard and James 1986). The stromatolites of the Exuma Cays, Bahamas, in contrast with other recent formations, possess a fine laminated structure. There, in addition to trapping and binding of carbonate sand, in situ precipitation of calcium carbonate produces distinct layers of cement (Reid and Browne 1991).

2.2 Microbial Mats and Stromatolites: The Geological Evidence

The Hadean era from the origin of the Earth 4.5  ×  109 years before present to 3.9  ×  109 is the period of which no rock record exists. The oldest rock known from the early Archean may not be of biogenic origin. The oldest stromatolites date back in the Archean about 3.5  ×  109 years ago. Only a few examples are known from this era, but they were undoubtedly biologically produced. Microfossils have been found in these stromatolites, but it is premature to identify them as cyano­bacteria. From measurements of carbon isotope compositions in these rocks it was deduced that autotrophic microorganisms must have been active at that time. However, it could have been chemoautotrophs that fixed the CO2, rather than photoautotrophs. The morphology of the microfossils from these oldest stromatolites also does not give an unequivocal clue about the identity of the organisms. Cyanobacteria are a group of oxygenic phototrophic organisms and it is well established that the Archean atmosphere did not contain oxygen.

During the Proterozoic, which started about 2.5  ×  109 years ago, stromatolites became abundant (Fig. 4.1) and occur in a wide variety of facies. They occupied every major ecological niche, marine and lacustrine, shallow and deep water. Most of limestones, dolomites and magnesites as well as many phosphorites and iron formations of the Proterozoic contain stromatolites. Like modern microbial mats, it seems certain that the Proterozoic stromatolites were produced by growth and metabolic activity of cyanobacteria. The Proterozoic stromatolites contain a wealth of very well preserved microfossils that show striking similarity to present day cyano­bacteria. Over 1,100 microfossils have been described from 190 stromatolite formations (Walter et al. 1992). Many of these fossils are preserved in the cherts of stromatolites. The best preservation occurred following early silicification of the stromatolites. Silica precipitation occurred possibly spontaneously because of its supposed high concentration in the seawater. Diatoms with their silicate frustules had not evolved yet and no other sink for silica is known.

Fig. 4.1
figure 1

Relative abundance of stromatolites plotted against time (After Awramik 1984)

Full size image

Oxygen was present in the atmosphere at 2.3  ×  109 years before present. It is now well accepted that the oxygenation of the atmosphere was the result of oxygenic photosynthesis. It might have taken considerable time after the origin of oxygenic photosynthesis until the atmosphere became oxygenated, since a large amount of reduced compounds had to be oxidized. Banded iron formations (BIFs) are known from 2.5  ×  109 years before present. These are huge formations consisting of oxidized iron and they have been taken as evidence for the presence of oxygenic photosynthesis. However, iron oxidation could also have taken place by the activity of anoxygenic phototrophic bacteria under anaerobic conditions (Widdel et al. 1993; Ehrenreich and Widdel 1994) or perhaps even by iron-dependent anoxygenic photosynthesis by cyanobacteria (Cohen 1989). Oxygenic photosynthesis most likely evolved at the beginning of the Proterozoic. Evidence for this is the presence of molecular markers such as the methylhopanes that are supposed to be specific for cyano­bacteria 2.7  ×  109 years before present (Brocks et al. 1999). Also, phylogenetic analysis date the origin of cyanobacteria at 2.6  ×  109 (Hedges et al. 2001). However, many cyanobacteria in present day microbial mats are capable of anoxygenic photosynthesis and it seems likely that cyanobacteria were anoxygenic phototrophs before they evolved oxygenic photosynthesis (Olson 2006).

The morphology of the stromatolites of the 3.1  ×  109 old Insuzi group of South Africa hints at the involvement in their formation of tactic filamentous organisms (Mason and Von Brunn 1977). Although it is tempting to suspect phototaxis and hence potentially photosynthetic organisms, a chemotactic response would also explain the structure of this formation (Schopf and Walter 1982).

Of the many different morphological forms of microfossils, some can be traced back to present day cyanobacteria such as Oscillatoria and Lyngbya (Schopf and Walter 1982). These organisms are common in modern microbial mats where they may be involved in N2 fixation. It is difficult to determine whether these ancient mats were diazotrophic, although the signature of the stable isotope 15N might give some hints in the direction of diazotrophic (N2-fixing) cyanobacteria (Bauersachs et al. 2009). Microfossils resembling cyanobacteria of the heterocystous genera Nostoc and Scytonema were abundant in Archean stromatolites (Schopf and Walter 1982), and this is taken as evidence that N2 fixation might have been important. No remnants of heterocysts are known, probably because these cells did not fossilize well. Fossil remnants of akinetes which are survival stages of cells that are known only from heterocystous cyanobacteria are known dating back 1.5  ×  109 years (Srivastava 2005). However, they may not have originated from microbial mats and it is unknown whether they were already associated with heterocystous cyanobacteria. To date, heterocystous cyanobacteria are uncommon in most microbial mats.

Proterozoic stromatolites reached maxima in numbers and diversity towards the end of this era, after which it showed a rapid decline (Walter and Heys 1985) (Fig. 4.1). It has been postulated that metazoa, which then appeared on Earth, were responsible for this decline (Walter and Heys 1985). The grazing activity of these animals would prevent the accumulation of the benthic mat-forming organisms and destroyed the fabric of microbial mats by bioturbation. After the appearance of metazoa, microbial mats would be much more limited in their distribution and developed in environments in which these grazers are largely excluded (so-called extreme environments). Nevertheless, based on evidence from a modern hypersaline lagoon in Venezuela, Gingras et al. (2011) have suggested that the early evolution of mobile complex animals may have been in cyanobacterial dominated mats during the Ediacaran period (635–542 million years ago).

The appearance of algae that competed successfully for light and nutrients in many environments could help explain the eventual pushing back of cyanobacterial mats to extreme environments. Also, sea level changes, caused by changes of climate and by tectonic processes, could explain the sudden decrease in stromatolite abundance (Gebelein 1976). And finally, the seawater in the Proterozoic might have been greatly oversaturated with respect to calcium carbonate (alkaline soda ocean) facilitating the calcification and preservation of stromatolites, which is less the case in the modern moderately alkaline ocean (Kempe and Kazmierczak 1990a).

Proterozoic stromatolites probably formed through one or more of the following (Walter et al. 1992):

  1.  (i)

    in situ precipitation as cement;

  2. (ii)

    in situ precipitation as micrite either accreted passively from suspension or through trapping and binding of the grains by the mat microorganisms;

  3. (iii)

    precipitation of micrite imported from adjacent environments.

The fine and distinct lamination of Proterozoic stromatolites hints at in situ precipitation. Most Phanerozoic stromatolites are probably produced by trapping and binding of carbonate and sand grains and therefore show poor or no lamination (Cloud and Semikhatov 1969).

2.3 Stratification and Structure of Microbial Mats and Stromatolites

Microbial mats are characterized by the vertical stratification of different functional groups of microorganisms. This structure is the result of the physicochemical gradients that are present in mats and in fact produced by the metabolic activity of the mat organisms themselves (Jørgensen et al. 1983). The typical structure of a microbial mat build by cyanobacteria is schematically depicted in Fig. 4.2.

Fig. 4.2
figure 2

Schematic drawing of a building of a typical microbial mat formed by cyanobacteria. The layer of green sulphur bacteria has been observed in only a few occasions

Full size image

Cyanobacteria evidently form the top layer of microbial mats although they are sometimes overlain by a film of diatoms. These organisms need to harvest light for photosynthesis and are essentially aerobic organisms. The cyanobacteria may further be covered by a layer of sand or sediment of varying thickness or be covered by an organic-rich mucilaginous layer which may contain photoprotective pigments such as scyto­nemin, which is produced by cyanobacteria. It occurs predominantly in the extracellular polysaccharide sheaths. Scytonemin is highly recalcitrant remaining in the empty sheaths that are left behind by the cyanobacteria. Scytonemin protects the underlying community from damage caused by UV irradiation (sunglass effect) (Garcia-Pichel and Castenholz 1991; Chap. 19). The organic matter introduced in the sediment through the photosynthetic activity of the cyanobacteria is decomposed by a variety of chemotrophic microorganisms. The degradation of organic matter and the accompanying demand of oxygen result in permanent anoxic conditions below the layer of cyanobacteria (Fig. 4.3). Obligate anaerobic sulphate-reducing bacteria play a major role in the decomposition of organic material in marine cyano­bacterial mats and other sulphide dominated environments. These bacteria produce sulphide, which is used by anoxygenic phototrophic bacteria.

Fig. 4.3
figure 3

Vertical profiles of oxygen, sulphide and pH at night (left panel) and during the day (right panel) in a mat of Microcoleus chthonoplastes from Solar Lake, Sinai (Redrawn from Revsbech et al. 1983)

Full size image

Purple sulphur bacteria are very common in microbial mats and are often seen as a pink layer below the cyanobacteria. Purple sulphur bacteria are essentially anaerobic bacteria, but species that occur in microbial mats are usually metabolically versatile (van Gemerden 1993). Anoxygenic photosynthesis in purple sulphur bacteria saturates at even much lower light intensities than photosynthesis in cyanobacteria. In addition, these organisms use a different part of the electromagnetic spectrum, not used by cyanobacteria (Pierson et al. 1987). This far red light is also least attenuated by the sediment (Fig. 4.4) (Stal et al. 1985; Jørgensen and Des Marais 1988). The biological stratification is thus the result of gradients of light, oxygen and sulphide and is found in virtually all cyanobacterial mats (Fig. 4.3). In some rare cases a layer of green anoxygenic phototrophic bacteria can be found underneath the purple bacteria (Nicholson et al. 1987).

Fig. 4.4
figure 4

Spectral light penetration through a 1.5-mm mat of Microcoleus chthonoplastes of the North Sea island of Mellum (Germany). Far-red light is least attenuated by the mat. The light absorption by chlorophyll a and phycobiliproteins at respectively 680 and 600 nm is clearly visible

Full size image

A distinct layer of oxidized iron may be present between the cyanobacteria and the purple sulphur bacteria (Stal 1994). It is not clear how this layer is formed. It may be formed by chemical oxidation by the oxygen produced during photosynthesis. An alternative explanation is the anaerobic oxidation of iron by anoxygenic photosynthesis in a specific group of purple bacteria (Widdel et al. 1993; Ehrenreich and Widdel 1994). Aerobic oxidation of iron by chemotrophic bacteria seems unlikely at the alkaline pH that are usually encountered in microbial mats (Fig. 4.3) although it should also not be excluded as a possibility as shown by Emerson and Revsbech (1994a, b). Other forms of anoxygenic photosynthesis that could potentially be important are those using nitrite (Griffin et al. 2007) and arsenate (Budinoff and Hollibaugh 2008; Kulp et al. 2008), although expected only in special cases.

The deeper layer of the mat is often black or gray, indicating the presence of iron sulphide (FeS) or pyrite (FeS2). This layer has often been referred to as the layer of the sulphate-reducing bacteria but it has become clear that these bacteria in fact do not form a distinct layer and occur throughout the sediment (Visscher et al. 1992; Stal 1993). They are both abundant and highly active in the top layers of microbial mats. At first sight this distribution of sulphate-reducing bacteria is unexpected and odd. However, their substrates are mainly produced by the cyanobacteria and it is certainly beneficial to the organisms to be close to the site of production. At night when photosynthesis ceases, the mat turns anoxic (Fig. 4.3), providing the appropriate environment for sulphate-reducing bacteria. Sulphate reducing bacteria appear to be quite tolerant to oxygen and some are even capable of low rates of aerobic respiration although they are unable to grow aerobically (Dilling and Cypionka 1990; Marschall et al. 1993). It has also been shown that sulphate reduction in a microbial mat can occur in the presence of oxygen (Canfield and Des Marais 1991; Fründ and Cohen 1992).

Chemotropic bacteria can also oxidize sulphide and represent an important group of organisms in microbial mats. As many as 2  ×  109 cm−3 sediment of colourless sulphur bacteria has been detected in the top layer of microbial mats (Visscher et al. 1992). These bacteria are quantitatively important in microbial mats. Colourless sulphur bacteria are essentially aerobic and gain energy from the aerobic oxidation of sulphide. They are autotrophic organisms and fix CO2 through the reductive pentose phosphate pathway (Calvin-Benson-Bassham cycle). Colourless sulphur bacteria have high affinities for their substrates and their presence cause highly dynamic oxygen and sulphide gradients, thereby overruling the chemical oxidation of sulphide. Since the sulphide-oxygen interface is highly dynamic and not fixed at a certain depth in the sediment (Fig. 4.3), these bacteria also do not form a distinct layer, although they are clearly most abundant in the top layer (Visscher et al. 1992). The joint metabolic activity of microorganisms in microbial mats results in steep physicochemical gradients of e.g. light, oxygen, sulphide, carbon dioxide and pH; these gradients shift markedly during a 24-h cycle (Fig. 4.3) (Jørgensen et al. 1979; Revsbech et al. 1983) and also respond to fluctuations of incident light. All microorganisms in microbial mats must therefore be highly versatile and flexible in order to respond to the continuous changes in environmental conditions.

The biological stratification in microbial mats may however be far more complex than described above. Cyanobacteria may be sandwiched between layers of anoxygenic photo­tro­phic bacteria and even perform oxygenic photosynthesis there. Microbial mats may be also just ‘inverted’, with cyanobacteria occurring underneath the layer of anoxygenic phototrophic bacteria (Van Gemerden et al. 1989). This type of microbial mats can develop when much organic matter is deposited on the sediment and its degradation results in very high rates of sulphide production.

The construction of extensive clone libraries has shed new light on the structure and composition of the microbial mat community. While microscopic examination of the mat suggests that the cyanobacteria are the dominant component, clone libraries of the 16S rRNA gene tell a different story. Ley et al. (2006) found cyanobacteria only important in the top most layer of a microbial mat but most of the 16S rRNA sequences belonged to other bacteria. These authors found 752 species in 42 bacterial phyla and Chloroflexi were identified as the dominant organism both in terms of biomass and in numbers of 16S rRNA genes and were present throughout the mat. Most cyanobacteria are large microorganisms. They have only few copies of the 16S rRNA gene and their DNA is often difficult to extract from natural environments and to amplify by PCR. Even so, the importance of the diversity and biomass of other bacteria in microbial mats is probably severely underestimated. For instance, the clone libraries of the microbial mats of Hamelin Pool, Shark Bay, Western Australia, contained only less than 5% cyanobacteria. Ninety percent were bacteria and ten percent belonged to archaea. These mats did not reveal sequences belonging to eukarya (Papineau et al. 2005). The average sequence identity was only 92% emphasizing the high diversity of these microbial mats.

The organic matter produced by photosynthesis is actively cycled in mats. In mats net growth is often close to nil (Nold and Ward 1996). When growth occurs seasonally, a new mat may grow on top of the old one, resulting in a ‘historical’ lamination (Monty 1976). In many coastal environments the organic matter is fully degraded and a historical lamination is then absent. However, in other mats, particularly those growing in hypersaline environments, degradation may be incomplete. Examples of such mats are the well-investigated hypersaline mats of ‘pond 5’ in Guerrero Negro, Baja California, Mexico and those of Solar Lake, Sinai, Egypt (D’Antoni D’Amelio et al. 1989). The Guerrero Negro ‘pond 5’ mats are about 5–6 cm thick and are well-laminated (Fig. 4.5). This mat grows at a rate of about 1 cm year−1 but there is no net accretion, so that the thickness remains about the same. This means that the microbial decomposition of the mat also must proceed at a rate of approximately 1 cm year−1. This mat therefore seems to be in steady state (Canfield and Des Marais 1994). The mineralization of the mat of Solar Lake is incomplete, although up to 99% of the primary production is immediately recycled in the mat, leaving only 1% for net accretion (Jørgensen and Cohen 1977; Krumbein et al. 1977). The Solar Lake microbial mat is about 1 m thick and the lamination goes back almost 2,000 years.

Fig. 4.5
figure 5

The mat of Microcoleus chthonoplastes of hypersaline pond 5 of the saltern of Exportadora de Sal, S.A., Guerrero Negro, Baja California Sur, Mexico, showing the typical multilaminated structure

Full size image

2.4 Environments Supporting Cyanobacterial Mats

2.4.1 Coastal Microbial Mats

Coastal tidal sand flats often are excellent environments for microbial mats to develop, particularly when the flats extend over a large area and when the slope of the flat is low (Fig. 4.6). Large areas will be covered by water for only a short period during the tide and often the sediment is not inundated for several days during neap tides. Such sediments often experience large fluctuations in water content, salinity and temperature, resulting in extreme conditions that limit the range of organisms able to inhabit this environment. The near absence of grazing organisms allows mat-building cyanobacteria and diatoms to accumulate. Coastal sand flats are usually nutrient-poor, but the phototrophic cyano­bacteria have low nutrient demands and they can fix N2. Moreover, most cyanobacteria resist long periods of drought, tolerating large fluctuations of salinity and temperature. Often these coastal microbial mats are composed of filamentous cyanobacteria that form a dense entangled mass which traps and binds sediment particles. Such mats are clearly visible to the naked eye as massive structures that to a large extent resist erosion (Fig. 4.6b). Their sediment stabilizing effect is of great importance for coastal morphogenesis. Typical examples are found in tidal sand flats of islands of the southern North Sea (e.g. Mellum, Germany), along the east coast of North America (e.g. Great Sippewisset Marsh, Cape Cod, Massachusetts; Bird Shoal, North Carolina), Pacific Coast (e.g. Guerrero Negro, Mexico) (Fig. 4.7), Shark Bay and Spencer Gulf in, respectively, Western and South Australia. A more complete list is given by Pierson (1992). Most such coastal microbial mats are not stromatolites, but examples of stromatolites in coastal sediments can be found in El Hamira Bay, Sinai, where stromatolitic beachrock is formed (Krumbein 1979) and along the Exuma Cays of the Bahamas intertidal and subtidal stromatolites are formed (Reid and Browne 1991) (Fig. 4.8). Schizothrix sp. can settle there in spite of the high wave energy to which the Atlantic coast of the Bahamas is exposed. Calcification of the Exuma Cays microbial mats renders stability to the system, which is necessary to cope with the strong wave energy.

Fig. 4.6
figure 6

(a) Extended tidal sand flat of the island of Mellum (Southern North Sea, Germany) at low tide covered with microbial mats. (b) Mature mats of Microcoleus chthonoplastes of the island of Mellum have accreted and fixed much sediment so that they are often not submersed at high tide. This decreases the grazing pressure

Full size image
Fig. 4.7
figure 7

Microbial mats on intertidal flats of the Pacific coast of Guerrero Negro, Baja California Mexico. (a) Two types of mat develop close to each other, with smooth mat (shown as dark areas) in the lower intertidal. (b) Smooth mat, showing cracks caused by desiccation at low tide. (c) Pustular mat. (d) Differences between the mats shown in cross-section: smooth mat has laminated structure typical of microbial mats – with a thin and dense layer of cyanobacteria on top, next a layer of anoxygenic purple sulphur bacteria, then a black layer of FeS, indicating that the mat is permanently anoxic below the layer of cyanobacteria. (e) Pustular mat showing a much looser mat of cyanobacteria on top, while layers of purple sulphur bacteria and FeS are absent, indicating that the sediment below the cyanobacteria is predominantly aerobic. (f) Smooth mat is composed of the non-heterocystous (but N2-fixing) Lyngbya aestuarii, the trichomes of which are surrounded by a thick polysaccharide sheath and the organisms are embedded in a dense matrix of mucilage. (g) Pustular mat composed of Calothrix

Full size image

Other coastal mats are present in protected lagoons and are semi-permanently inundated. Examples of such coastal lagoons in which benthic microbial mats develop are found in many countries. Mats develop in the shallow parts of coastal lagoons, where large fluctuations of temperature and salinity may occur (Stal et al. 1996; Villbrandt and Stal 1996).

2.4.2 Hypersaline Microbial Mats

Hypersaline environments can be found in shallow and sheltered coastal lagoons and tidal channels with high rates of evaporation and low precipitation. In the Mediterranean, hypersaline lagoons may form when they have narrow connections to the open sea and exchange of water is limited because a tide is virtually absent. Alternatively they can be totally disconnected from the sea and are fed by sea water through a sand bar as is the case in Solar Lake on the Red Sea coast of Sinai, Egypt. When virtually all water in such coastal lagoon environments evaporates, a natural salt pond develops, forming structures known in the Sinai desert as Sabkhas. In those geographical regions were the combination of sun and wind result in sufficient evaporation, artificial salt ponds have been constructed. Other hypersaline environments are inland seas which can be found at many different locations on the globe (Oren 1988) and in shallow lagoons of many of these lakes cyanobacterial mats develop (e.g. Zavarzin et al. 1993).

Cyanobacterial mats develop under these hypersaline conditions thanks to the potential of certain cyanobacteria to accumulate compatible solutes such as betaine that allow osmoregulation up to high salinities (up to 25%). Depending on the salinity cyanobacteria with different osmoprotectants prevail. Cyanobacteria are not found under saturating salinities or when the salt composition differs strongly from that of seawater. In general, hypersaline environments are also strongly alkaline. Cyanobacteria tolerate high pH thanks to their capacity of taking up bicarbonate and accumulating inorganic carbon. As a result of the fixation of CO2 cyanobacteria generate alkaline conditions themselves. The best studied hypersaline microbial mats are from the salt ponds in Guerrero Negro, Baja California Sur, Mexico and Solar Lake, Sinai, Egypt (D’Antoni D’Amelio et al. 1989; Des Marais et al. 1992). Salts and brines are discussed more fully by Oren in Chap. 15.

2.4.3 Hot Spring Mats of Cyanobacteria

Thermal hot springs are environments in which the high temperature in combination with H2S or acidic conditions decreases biodiversity enormously. Cyanobacterial mats are most common in hot springs at near neutral or alkaline conditions and are described more fully in Chap. 3. Cyanobacteria are generally alkaliphilic organisms. Acidic hot springs are more likely to inhabit eukaryotic microalgae. Thermal springs that contain sulphide may limit the formation of mats since thermophilic cyanobacteria do not tolerate the combination of high temperature and high levels of sulphide (Castenholz 1976, 1977). At moderate concentrations of sulphide, mats of Oscillatoria spp. have been shown to lower the sulphide concentration by anoxygenic photosynthesis by the cyanobacteria (Cohen et al. 1986; Ward et al. 1989). Another strategy is found in the so-called inverted mats (Castenholz 1976). Here, mats of the anoxygenic phototroph Chloroflexus overlay the cyanobacterial mat. Anoxygenic photosynthesis scavenges the sulphide and protects the underlying mat of the oxygenic heterocystous cyanobacterium Chlorogloeopsis sp. (Jørgensen and Nelson 1988). The maximum temperature at which photosynthesis can take place is slightly above 70°C.

Fig. 4.8
figure 8

(a) Stromatolites formed by cyanobacteria in the intertidal of Exuma Cays, Bahamas: these structures are considered as modern examples of known fossil stromatolites. (b) A closer look at these lithified sedimentary structures, which consist of trapped carbonate sediment cemented by micritic (microcrystalline) carbonate. (c) Surface of stromatolites is covered by a cyanobacterial-algal mat, which is thought to be involved in formation of the micritic horizons. (d) Cross-section of the top part of the stromatolite shows a distinct green layer of cyanobacteria. (e) Schizothrix is the dominant cyanobacterium in these modern stromatolites

Full size image

2.4.4 Terrestrial Cyanobacterial Mats

Terrestrial cyanobacterial mats can be found in a variety of different environments. De Winder et al. (1989a, b) described a cyanobacterial-algal crust in coastal dunes. Sand dunes have a poor capacity of retaining water and are therefore extremely dry environments that are characterized by a low biodiversity. Under certain conditions there develops a mat of Crinalium epipsammum, a unique band-shaped filamentous cyanobacterium (Fig. 4.9); its unusual cell envelope is exceptionally well-adapted to desiccation (De Winder et al. 1990). This species is important in the Netherlands in stabilizing and protecting dune sand from wind erosion. Once this organism has established a matrix the community is taken over by the green alga Klebsormidium flaccidum.

Fig. 4.9
figure 9

Scanning electron micrograph of Crinalium epipsammum, a cyanobacterium forming phototrophic microbial crusts on coastal dunes. This organism is unusual because of the flat trichomes and the fact that the cell wall contains cellulose. Scale bar  =  3 μm

Full size image

Desiccation and life under low water potential are also the controlling factors for the development of cyanobacterial mats and stromatolites in the hot desert. Microcoleus chthonoplastes, which occurs in some desert crusts (Brock 1975), has a polysaccharide sheath which plays an important role in protection from desiccation. After re-wetting the sheath absorbs water and the cyanobacterium resumes activity immediately (Campbell 1979). The sheaths of the unicellular desert Chroococcus sp. and Chroococcidiopsis sp. also have this function and these species are found hypolithically on rocks in the Negev desert (Potts and Friedman 1981; Potts et al. 1983; Caiola et al. 1993, 1996). Cyanobacterial mats are particularly well investigated in the Negev desert in Israel (Friedman et al. 1967; Berner and Jensen 1982). Krumbein and Giele (1979) found a calcifying mat of a unicellular cyanobacterium producing stromatolitic structures in the desert. Cyanobacterial mats also seem to be involved in the formation of rock varnish in the desert. Desert rock varnish is composed of iron and manganese oxides that are preci­pitated by the metabolic activity of mat microorganisms. The cyanobacterial mat is usually present underneath this hard brownish layer where they are protected from direct sunlight and are capable of retention of some water (Krumbein and Jens 1981).

Carbonate caves are other terrestrial environments that support the formation of microbial mats of the unicellular N2-fixing Gloeothece (Gloeocapsa) and of the heterocystous Nostoc on walls that receive some daylight (or when artificial illumination is present) (Cox et al. 1981; Griffiths et al. 1987). Another example of such a low-light terrestrial environment is the mats of Leptolyngbya sp. described by Albertano and Kovacik (1996) on the walls of Casa Aureum in Rome. Terrestrial mats of Nostoc have been reported from a variety of desert environments, including the cold desert in Antarctica (Davey 1983; Davey and Marchant 1983). Cyanobacterial mats from cold deserts have been described by Davey and Clarke (1992) and Vincent et al. (1993a, b).

3 The Organisms: Cyanobacteria That Build Microbial Mats

Cyanobacteria that build microbial mats include a variety of filamentous and unicellular species. The filamentous non-heterocystous Microcoleus chthonoplastes dominates marine intertidal microbial mats all over the world (Stal et al. 1985), hypersaline environments (Garcia-Pichel et al. 1996) and in the hot desert (Campbell 1979).

A notable characteristic of M. chthonoplastes is its occurrence in bundles containing many trichomes, often twisted like a rope. The bundles are enclosed in a common polysaccharide sheath (Fig. 4.10) which may be partitioned in different compartments (Fig. 4.10b). The rope morphology has been suggested to be an adaptation evolved to colonize unstable substrates (Garcia-Pichel and Wojciechowski 2009). Garcia-Pichel et al. (1996) investigated and compared cultures isolated from a variety of mats from geographically distant locations, both marine and hypersaline. Based on morphological and genetic characteristics, the authors concluded that all these isolates were closely related and belonged at least to the same genus and probably the same species.

Fig. 4.10
figure 10

Scanning electron micrographs of a mat of Microcoleus chthonoplastes of the island of Mellum, Germany. (a) Overview of the mat showing the large polysaccharide ensheathed bundles of trichomes. Scale bar  =  30 μm. (b) Detail showing one end of a bundle of M. chthonoplastes. The inner room of the bundle is composed of different compartments separated by polysaccharide walls. This bundle contains a large number of trichomes. Scale bar  =  10 μm. (c) A side view of a bundle of M. chthonoplastes. The outside is colonized by other microorganisms, including diatoms. Scale bar  =  10 μm. (d) Detail of the end of the bundle with the individual trichomes sticking out. The trichomes can move freely in and out of the bundle. Scale bar  =  10 μm

Full size image

The analysis of the 16S rRNA gene and morphological characteristics of a large number of strains that were assigned to Microcoleus led Siegesmund et al. (2008) to conclude that this genus belongs in fact to two families: the Oscillatoriaceae and the Phormidiaceae. The marine and hypersaline mat-forming M. chthonoplastes belongs to the latter and formed a coherent group that was proposed a new genus name, Coleo­fasciculus (only species so far C. chthonoplastes), in order to separate them from the freshwater and terrestrial Microcoleus (type strain M. vaginatus). However, in this Chap. 1 will refer to the better known name M. chthonoplastes.

More than 20 morphotypes of cyanobacteria were isolated from the mats of the intertidal sediments of the German North Sea island Mellum (Stal and Krumbein 1985) and a similar number with the same range of morphotypes were later isolated from a similar mat of the Dutch North Sea barrier island Schiermonnikoog. These also included several heterocystous species that were rarely observed in the field except in supratidal mats that received more freshwater than those in the intertidal regions. The cyanobacterial community composition of these mats varied considerably during the course of a year but also between different years. Sometimes the mats consisted virtually exclusively of Spirulina or Merismopedia (Palinska et al. 1996). Often these mats were composed of mixtures of different species (Fig. 4.11).

Fig. 4.11
figure 11

(a) Scanning electron micrograph (SEM) of a N2-fixing mat of Oscillatoria spp. and other cyanobacteria. (b) SEM photo of a detail of a mat of Lyngbya sp. with the typical coiled filament of Spirulina subsalsa, which is a typical component of these mats and a single trichome of Microcoleus chthonoplastes

Full size image

An important species that was always present and was frequently dominant was originally assigned to Oscillatoria limosa strain 23 (Stal and Krumbein 1985), which was shown to be the diazotrophic component of these mats. This strain was capable of aerobic N2 fixation in culture (Stal and Krumbein 1981). Based on the 16S rRNA gene sequence analysis O. limosa was found to be related to Lyngbya aestuarii. This morphotype is observed frequently in microbial mats all over the world and as far as known all of these mats are diazotrophic. The trichomes have a thick polysaccharide sheath which is often pigmented (Fig. 4.7). Mats of Lyngbya/Oscillatoria can be found in geographically distant locations and are characterized by high rates of N2 fixation.

Although many if not all microbial mats are capable of diazotrophy, the specialized heterocystous forms are only rarely the dominant component. That in some cases they can be isolated proves their presence, but obviously the prevailing conditions in the mat prevent their proliferation. Nevertheless, a few exceptions are known. In parts of the tidal flat in Guerrero Negro (Baja California Sur, Mexico) extensive mats of the heterocystous Calothrix sp. are present (Stal 1995) (Fig. 4.7). In a coastal lagoon near Bordeaux, France, mats of Anabaena sometimes develop (Villbrandt and Stal 1996). Mangroves often support extensive mats of the heterocystous Scytonema sp. (Potts 1979). Calothrix sp. is also known to form mats on rocks in the spray zone at the seashore (Whitton and Potts 1982). The development of these heterocystous diazotrophic mats is discussed further in Sect. 4.7.

Hot spring microbial mats such as Octopus Spring in the Lower Geyser Basin of Yellowstone National Park in the USA and similar microbial mats are dominated by the rod-shaped unicellular cyanobacterium Synechococcus lividus (Brock 1978). For a long time, strain Synechococcus sp. Y-7c-s was the only cultivated species. This strain was isolated from the 50–55°C Octopus Spring mat and considered to be representative for all thermophilic Synechococcus species since they possessed DNA with almost identical G+C ratios (Waterbury and Rippka 1989). In fact, at least seven different strains, all with identical morphology, are present (Ward et al. 1994). Y-7c-s was only detected in Clearwater Spring, which is slightly acidic (Ruff-Roberts et al. 1994; Ferris et al. 1996) and from which this strain may have been originally isolated. Hybridization probes have shown that Synechoccoccus sp. strain Y-7c-s was present in the Octopus Spring mats in extremely low frequency. Ferris et al. (1996) demonstrated that enrichment cultures selected for this strain. Although morphological indistinguishable, the populations of Synechococcus sp. in Octopus hot spring microbial mats belong to a phylogenetic diverse group. Analysis of the 16S rRNA gene sequences of Octopus Spring revealed a high diversity of Synechococcus distantly related to S. lividus. These thermophilic Synechococcus are an ecologically diverse group of cyanobacteria that are distributed hori­zontally along a temperature gradient and vertically along light and oxygen gradients (Allewalt et al. 2006). However, by diluting the inocula prior to enrichment new strains of Synechococcus were obtained in axenic cultures. The different strains of hot-spring Synechococcus sp. have growth optima at different temperatures (Ward et al. 1994) and it was shown that their temperature ranges and optima were consistent with their distributions in the mats. Other adaptations may include those to pH and light.

In the hypersaline mats of Pond 5 of the Guerrero Negro saltern and the shallow flat mat of Solar Lake M. chthonoplastes is the dominant species and forms gelatinous organic mats (Fig. 4.5). Other cyanobacteria that may occur in these hypersaline mats are Oscillatoria sp. and Spirulina sp. Unicellular cyanobacteria may also be present. The Pond 5 mat of Guerrero Negro grows at salinities from 60‰ to 95‰. The salt content of the shallow flat mat of Solar Lake ranges from 45‰ to 180‰. At salinities that are permanently above 100‰ M. chthonoplastes does not proliferate well but Spirulina subsalsa may be found up to 150‰. At higher salinities up to 200‰, the unicellular Aphanothece halophytica usually dominates the mats (Dor and Paz 1989). The salinity tolerances for cyanobacteria seem to be higher in mats of the Sabkha, where A. halophytica occurs at 250‰, which is close to saturation, while S. subsalsa is present up to ∼200‰ (Dor and Paz 1989). The reason for these differences is unclear. Salinity tolerance may be influenced by temperature. Although the sediment surface of the Sabkha may become hot from the solar radiation, the submersed mats in solar ponds may also be exposed to high temperatures. Therefore, halophilic cyanobacteria may also be moderately thermophilic such as was shown for a newly discovered organism with very thin trichomes of 1 μm Halomicronema excentricum that grows in the range of 3.2–12% salt and 28–50°C (Abed et al. 2002).

Lithified microbial mats found in the Exuma Cays, Bahamas, are built by the filamentous cyanobacterium Schizothrix (Pinckney et al. 1995). This forms thin trichomes about 1 μm wide, with cells 2–5 times as long as wide. The trichomes may be enclosed by a thick sheath. Communities of Schizothrix may form dense and tough mats that are often associated with calcium carbonate precipitation. The lithified microbial mats of Exuma Cays, Bahamas, result in the formation of stromatolites, a process which still goes on (Reid and Browne 1991; Pinckney et al. 1995; Reid et al. 2000).

4 Motility, Chemo- and Phototaxis of Cyanobacteria in Microbial Mats

Microbial mats are characterized by steep and fluctuating physicochemical gradients. In order to experience optimum conditions at all times, cyanobacteria must position themselves continuously in the mat. Microbial mats also often occur in environments with high sedimentation rates. This demands a light-oriented motility, in order to prevent permanent burial.

Cyanobacteria possess essentially three different ways in which they respond to light: phototaxis, photokinesis and photophobic response (Häder 1987a, b). Phototaxis is a movement, which is oriented along the direction of light. Phototaxis can be either positive or negative. Positive phototaxis is towards the direction of light whereas negative phototaxis is the movement away from the light. Both positive and negative phototaxis are important for cyanobacteria in microbial mats. Most cyanobacteria are adapted to growth at low light intensities. Excessive light may result in photooxidative stress and can cause damage. The combination of positive and negative phototaxis will allow the organism to obtain an optimum position in the mat. Most of the research on light responses of cyanobacteria has been carried out on Phormidium and Anabaena and more recently also on the unicellular Synechocystis. Little work has been carried out on light responses in cultures of mat-forming cyanobacteria.

Photokinesis is the term used for the phenomenon where speed of movement increases with light intensity. This is because of the greater supply of energy. Only positive photokinesis is known (negative photokinesis would be the decrease of speed at higher light intensities). The photophobic response is the reversal of the direction of movement as a result of a sudden change in light intensity. This response is very important for cyanobacteria. Both step-down and step-up responses are known (Häder 1987a). A step-down response causes the accumulation of the organisms in the light. At very high light intensities a step-up response may result in the accumulation of the organisms in a shaded area. Photophobic responses are clearly related to photosynthesis as could be concluded from action spectra (Häder 1988).

The light required for phototaxis might be extremely low (0.001 μmol photon m−2 s−1) and also saturates at very low photon irradiance (1 μmol m−2 s−1) (Ng et al. 2003). In most cases the action spectrum of phototaxis follows the photopigments of the cyanobacteria, the phycobiliproteins and chlorophyll a (Bhaya 2004). The low threshold for phototaxis is important for cyanobacteria in microbial mats to direct them to the surface after a large deposition event. The low light intensity and the complex action spectrum suggest that it is not required for providing the energy for locomotion, which is confirmed by the ineffectivity of inhibitors of photosynthesis (Choi et al. 1999).

Motility is an extremely important property for most mat-forming cyanobacteria and occurs by gliding, which can be defined as a self-propulsion along a surface. This surface can also be the interior of the polysaccharide sheath. Trichomes may move forwards and backwards in their sheaths and may move out of it, leaving an empty sheath behind. Trichomes may also glide along each other. The hypotheses to explain gliding motility that have received most attention are:

  1. (i)

    secretion of mucilage

  2. (ii)

    contractile structures that cause surface undulations.

Some motile cyanobacteria possess junctional pore complexes, organelles that penetrate the cell wall and through which it is assumed that mucilage is secreted (Hoiczyk 2000). According to this hypothesis the mucilage adheres to the substrate and flows in tight contact with the trichome, thereby producing the propulsive force. The reversal of the movement would be obtained by using junctional pore complexes at the other end of the cell that are directed opposite. The rotation along the long axis in some Oscillatoriaceae could be produced through the presence of helically arranged fibrils. The arrangement of these fibrils determines indeed the left or right rotation, which is species specific. However, the highly motile Phormidium uncinatum does not possess junctional pore complexes, although it secretes polysaccharide (Häder 1987b).

Halfen and Castenholz (1971) and Castenholz (1973) suggested that the helically arranged microfibrils which they found in the external layers of the cyanobacterium Oscillatoria princeps can contract producing a surface wave that contacts the surface producing the force needed for the gliding movement.

Gliding movement is not restricted to filamentous cyano­bacteria. The unicellular Synechocystis exhibits a motility that has been described as twitching and depends on type IV pili which moves the organism by pilus extension, adhesion and retraction (Bhaya 2004). Hence, gliding may well be a collection of different types of locomotion, each with their own specific mechanisms. One pelagic marine unicellular Synechococcus is capable of swimming, even if it lacks flagella (Waterbury et al. 1985). Swimming depends on swmA, a cell surface glycoprotein of Synechococcus and which is needed for the generation of thrust (McCarren et al. 2005). This mode of locomotion does not seem practical in the dense matrix of the microbial mat.

It is not certain how important the three different responses to light are in microbial mats. Ramsing and Prufert-Bebout (1994) concluded from light measurements in mats made by fiber-optic micro light sensors that light fields in microbial mats are uniform. This is caused by scattering of light, and it means that there is in fact no direction of light. Moreover, light intensity will not be subject to sudden changes in microbial mats. These authors therefore conceived that phototactic and photophobic responses would not be especially beneficial for mat-forming cyanobacteria. Studies with M. chthonoplastes indicated that the strategy of this organism is to minimize movement when conditions are favourable. Instead of varying the speed of movement (photokinesis) it moves less frequently. M. chthonoplastes also reverses its movement frequently. This is not a photophobic response because this would imply a step-down or step-up change in light intensity which is not the case. Ramsing and Prufert-Bebout (1994) further observed that M. chthonoplastes bends more frequently at optimum light intensity. In the long-term this could lead to curling of trichomes into bundles. Motility in such bundles is likely to be restricted. Such cyanobacteria are likely to be confined to a fixed position in the mat. In the hypersaline Guerrero Negro mats M. chthonoplastes was present throughout the mat and did apparently not migrate, whereas other bacteria, including other cyanobacteria, migrated through the mat and occupied different positions during the daily cycle (Dillon et al. 2009).

Garcia-Pichel et al. (1994) demonstrated that mat-forming cyanobacteria Oscillatoria sp. and Spirulina subsalsa mig­rated up and down in the mat in a daily manner (Fig. 4.12). At sunset these cyanobacteria moved towards the mat surface and stayed there throughout the night. At sunrise they migrated downwards. The depth to which they migrated appeared to be related to the light intensity, reaching the maximum depth in the mat at mid-day when the light intensity was highest. Interesting was also that Oscillatoria sp. and S. subsalsa contained unusually high amounts of chlorophyll a (3.9% d. wt). A unicellular cyanobacterium in this mat was non-motile and contained only 0.3% chlorophyll a (Garcia-Pichel et al. 1994). It was calculated that if Oscillatoria and S. subsalsa did not migrate they would be photoinhibited for most of the time, whereas the daily movement guaranteed optimum photosynthesis throughout the light period. Many cyanobacteria move deeper into the sediment at high light intensities (Pentecost 1984; Whale and Walsby 1984; Richardson and Castenholz 1987a) (Fig. 4.12).

Fig. 4.12
figure 12

Movements of cyanobacterial layer in mats during a 24-h period. Upper panel shows an example of the daily sinusoidal light curve. Example Type I is a mat which does not displace itself during a day-night cycle. This is either the case with unicellular cyanobacteria that are not motile and grow at optimal light intensity or by species that minimize movement when conditions are favorable, which may be the case in some populations of Microcoleus chthonoplastes. Example Type II is a mat that moves toward the surface at sunset and moves into the sediment during the day. Upwards movement may be controlled by chemical factors such as oxygen or sulphide. Downwards movement is in most cases attributed to negative phototaxis. Mats of Oscillatoria often show this type of displacement during a 24-h cycle. Example Type III is exhibited by the hot-spring O. terebriformis. In the dark the organism moves randomly, but motility is inhibited by sulphide, which eventually results in the accumulation of the population in the sulphide-rich layer deep in the sediment. Positive phototaxis occurs at low light and negative phototaxis at high light. This forces the organism to move deeper into the sediment during the middle of the day

Full size image

Other researchers have also noticed that cyanobacteria migrate to the surface during the night or when the mat is shaded. Migration occurs also during the dark and Whale and Walsby (1984) therefore concluded that this upward movement was not controlled by light. Since these authors could not find any evidence for geotactic or magnetotactic responses, they assumed that migration of cyanobacteria was directed through chemotaxis in a chemical gradient. On the other hand, not all cyanobacteria are capable of moving in the dark. Malin and Walsby (1985) observed that motility of Oscillatoria sp. was strictly dependent on light and gliding stopped in the dark after a short period, presumably because energy reserves were exhausted. These authors demonstrated responses of Oscillatoria sp. to oxygen (aerotaxis) and CO2 and bicarbonate. A light-dependent response to CO2 would be advantageous. Photosynthetic activity in microbial mats causes depletion of CO2 and the high pH usually encountered in these environments as a result of photosynthetic activity and CO2 fixation will shift the carbonate equilibrium resulting in even lower concentrations of CO2. A light-dependent positive response to oxygen seems to be less advantageous. High concentrations of oxygen in the light may cause photo-oxidative reactions (Eloff et al. 1976) and photorespiration with therefore less efficient CO2 fixation (Lorimer 1981; Reinhold et al. 1991).

Aerotaxis would be a useful strategy for dark migration. This would allow aerobic degradation of endogenous storage carbohydrate. The migration of M. chthonoplastes (Whale and Walsby 1984), Oscillatoria sp. and S. subsalsa (Garcia-Pichel et al. 1994) to the mat surface during the dark can be explained by a positive aerotaxis (Fig. 4.12). Alternatively, migration to the surface during the dark can be explained by a negative response to sulphide which is very toxic. In the dark the concentration of sulphide will increase because anoxygenic photosynthesis is absent and no oxygen is available for biological or chemical oxidation (Fig. 4.3). Castenholz (1982) therefore conceived a chemophobic response towards sulphide in cyanobacteria that migrate to the mat surface during the dark.

Chemotaxis in chemotrophic bacteria has received much attention but cyanobacteria have hardly been investigated for such migratory behaviour. Several cyanobacteria are capable of assimilating organic compounds such as glucose and fructose in the light (photoheterotrophy) and some even display a fully chemoorganotrophic metabolism (Smith 1982). Oscillatoria terebriformis is capable of fermenting extracellular compounds such as fructose and glucose (Richardson and Castenholz 1987b). Chemotactic responses of cyano­bacteria to organic compounds are largely unknown. Fechner (1915) reported a negative chemotactic response to organic acids and Richardson and Castenholz (1989) observed inhibition of gliding of O. terebriformis by fructose. This effect was similar to that observed for sulphide. Glucose, the other substrate for this organism, did not have an effect, nor did lactate which is one of the fermentation products produced by O. terebriformis.

Cyanobacteria that form symbioses with plants were attracted by plant extracts, by certain sugars and particularly by mucilage (Nilsson et al. 2006). Higher temperature and darkness decreased chemotaxis, although this may be explained that light provides the energy for motility, rather than that it controls chemotaxis itself. Chemotaxis may be much more widespread in cyanobacteria than known until now, since operons for this process have been found in their genomes (Wuichet and Zhulin 2003) and the advantages for life in microbial mats with their steep and fluctuating chemical gradients are obvious.

An interesting behaviour has been encountered in O. terebriformis, which occurs in hot spring microbial mats and has a light-oriented motility. In the dark, this organism continues to move, albeit randomly. It may thus happen that it moves down into the sediment reaching the sulphide layer. Sulphide inhi­bits motility in O. terebriformis and therefore the organism is trapped in this layer (Richardson and Castenholz 1987a) (Fig. 4.12). Under laboratory conditions, 0.7 mM sulphide completely inhibited its gliding motility. Sulphide inhibited motility only in the dark or in the light when photosystem II was blocked by 3,-(3,4-dichlorophenyl)-1,1-dimethylurea (DCMU). This inhibition was reversible and was abolished in the light. Since every individual organism has a high probability to become trapped in the sulphide layer, virtually the whole population will end up there. During the day the sulphide horizon will move down into the sediment relieving the inhibition of motility and at the same time the organism will move towards the light at the mat surface. At mid-day, when light intensity is high, O. terebriformis shows negative phototaxis and moves deeper into the sediment in order to prevent photo-oxidative damage (Fig. 4.12). The majority of motile mat-forming cyanobacteria will prefer low light intensities and move deeper into the sediment during the day. The trapping of O. terebriformis in the sulphide layer during the dark is unusual, but essential for this organism to survive the dark period. In the presence of oxygen dark respiration will cause a rapid depletion of the endogenous storage carbohydrate which will result in the death of the organism in a matter of hours. The sulphide layer is of course anoxic. O. terebriformis is capable of fermentation and this process is slow, allowing for an extended period of energy generation. Indeed, the amount of energy that can be generated is small, but sufficient to cover its maintenance requirements (Stal and Moezelaar 1997). Most cyanobacteria, including mat-forming species, have low rates of dark respiration, allowing them to overcome long periods in the dark in the presence of oxygen. In the dark many microbial mats are virtually anoxic up to the surface, and fermentation is probably the only metabolism possible for the majority of cyanobacteria in the mat, with the exception of those that are exposed to the air. When the mat is submersed, oxygen decreases to zero within the diffusive boundary layer and no oxygen will be available to the mat (Fig. 4.3).

An unusual and new form of taxis is directed to gradients of water or water potential and has been termed hydrotaxis (Garcia-Pichel and Pringault 2001; Pringault and Garcia-Pichel 2004). Hydrotaxis has been found in a desert crust cyanobacterium related to a marine Oscillatoria. It was shown that migration depended on energy and probably occurred by gliding movement which is the mode of motility in this genus. When the surface of the crust dried out, the cyanobacteria migrated deeper into the crust where the water potential is higher. After a rain shower, the cyanobacteria moved quickly back to the surface. It is not known what exactly the cyano­bacteria senses but it was speculated that it might be water potential or hydrophobicity (Pringault and Garcia-Pichel 2004). Migration towards water makes sense in microbial mats or crusts in desert environments but is probably unlikely in microbial mats in aquatic environments, even if coastal intertidal mats may become desiccated for prolonged periods as well. Hydrotaxis has hitherto only been described for this one occasion.

Mats of diatoms on intertidal mudflats in estuaries and bays have been reported to migrate into the sediments on a high tide (Serôdio et al. 1997). This migration might be under the control of an endogenous rhythm and was maintained for a certain period of time even when the trigger of the tidal cycle was taken away experimentally. For these diatoms it is important to migrate into the sediments when the tide comes in, in order to avoid grazing, even when this greatly limits their window for photosynthesis. In cyanobacterial mats such migration triggered by the tidal cycle has not been reported. Cyanobacterial mats are usually found in the higher reaches of the tidal flats and are not or for shorter periods inundated with water and therefore less subject to the tidal cycle. It is probably the lack of the tide-triggered migration that cyano­bacteria can not escape grazing and therefore microbial mats do not occur there where diatom mats are found. In contrast to what had been assumed, it was demonstrated by Garcia-Pichel and Bebout (1996) that ultraviolet radiation penetrates well in microbial mats. The amount of penetration varies with the type of sediment on which microbial mats developed. Silty mud absorbed UV light most and quartz sand the least. Mats that are mainly organic take an intermediate position. UV light was absorbed in these mats more or less exponentially, in a similar way to visible light. There were two important aspects of the penetration of UV light in microbial mats, regardless of their sedimentological characteristics. In some mats the intensity of UV-B at the surface is considerably higher than the incident intensity; this is caused by scattering. Secondly, the total amount of UV-B in the euphotic zone of the mat ranged from 15% to 33% of incident irradiance which is high, particularly when compared with aquatic systems, where this number varies from 3% to 9%. These measurements carried out by Garcia-Pichel and Bebout (1996) were the first to demonstrate unequivocally that cyanobacterial mats develop under UV stress. Garcia-Pichel and Castenholz (1994) and Bebout and Garcia-Pichel (1995) provided also strong evidence that vertical migrations are partly under control of UV light. Garcia-Pichel and Castenholz (1994) reported that only 1.3 W  m−2 of UV-A (315–400 nm) was sufficient to keep the cyanobacteria Oscillatoria sp. and Spirulina subsalsa deep in the sediment. This intensity is only 3–4% of the level that these organisms would experience at mid-day. These cyanobacteria responded by negative phototaxis. In another study of microbial mats in Solar Lake (Sinai) it was shown that M. chthonoplastes responds clearly to UV-B light (310 nm). Exposure of the mat to 0.35–0.79 W  m−2 was sufficient to cause a downwards migration of the cyanobacteria. The effect of UV-B was about two orders of magnitude stronger than normal visible light. Also UV-A had this effect but was about five times less efficient than UV-B (Bebout and Garcia-Pichel 1995). It was concluded from these experiments that M. chthonoplastes is capable of sensing UV light, particularly UV-B.

There is no doubt that UV light causes serious damage to oxygenic phototrophic organisms (Cullen and Neale 1994) and has therefore negative effects on primary productivity (Smith et al. 1992). A mat-forming cyanobacterium will therefore benefit from the capability of sensing low levels of UV radiation and combining it with negative taxis. This will nevertheless result in a negative effect on total gross photosynthesis and productivity during exposure to UV light, but it is largely reversible (Bebout and Garcia-Pichel 1995). Due to the downwards migration of the cyanobacterium, the biomass at the surface, and thus gross photosynthesis, decreases. In addition, surface photosynthesis may be partly inhibited by UV irradiation. Because in the deeper layers more biomass accumulates gross photosynthesis is even higher but due to the low level of photosynthetic active radiation (PAR), biomass specific photosynthesis is low. Not all cyanobacteria exhibit negative phototaxis with respect to UV light. Donkor and Häder (1991) and Donkor et al. (1993) showed that moti­lity in the cyanobacteria they investigated was rather impaired by UV-B (280–315 nm). This may also have been the case in a mat of M. chthonoplastes of the temperate southern North Sea, where photosynthesis was strongly inhibited by UV-B radiation and did not recover during the subsequent 3 h when UV was excluded (Garcia-Pichel and Castenholz 1994).

Instead of migrating up- and down in a microbial mat, cyanobacteria have other possibilities to control the amount of light that they must absorb. Pierson and Parenteau (2000) observed for instance that cyanobacteria in the top layer oriented themselves vertically in the mat. This orientation may also have important consequences for the morphology of the microbial mat and may explain some of the morphologies of fossil stromatolites. Merismopedia is a unicellular cyanobacterium that occurs frequently in coastal microbial mats. It is characterized by its occurrence in plates in which the cells are well ordered. I have observed frequently that these plates may change its orientation from the large surface towards the light so that it receives maximum light to the side (one cell layer thick) to receive a minimum of light. The same was observed for the flat band-shaped filamentous cyanobacterium Crinalium epipsammum (de Winder et al. 1990).

5 Carbon Metabolism

5.1 Introduction

Cyanobacteria are the principal primary producers in the majority of microbial mats, although in some cases diatoms contribute as well. Oxygenic photosynthesis and sometimes anoxygenic photosynthesis and even chemosynthesis drives CO2 fixation. Cyanobacteria enrich the microbial mat with organic matter. CO2 fixation results in the formation of structural biomass of the cyanobacteria. This organic matter may become available to other organisms in the mat by the death and subsequent lysis of the cyanobacteria. However, it appears that, in spite of the high rates of photosynthesis usually observed, net growth of the cyanobacteria is often low in mature mats (Nold and Ward 1996). Hence, other processes must be involved in order to divert photosynthate to the mat community.

The benthic microbial mat community of the hypersaline lake, ‘La Salada de Chiprana’, northeastern Spain, produced dissolved organic carbon during the day and the night (Jonkers et al. 2003). These authors estimated that 14% and 49% of the mat gross and net photosynthetic production, res­pectively, diffused out of the mat in the form of low molecular weight fatty acids, although these compounds made up only 2% of the total dissolved organic carbon pool. The high flux of the dissolved organic carbon was generated by nutrient deficiency of the cyanobacteria. Photoheterotrophic Chloroflexus-like bacteria grew on top of the cyanobacterial mat at the expense of these phototrophic exudates. Also, large numbers of sulphate-reducing bacteria were found in the fully oxygenated surface layers. Another process that degrades dissolved organic carbon in microbial mats is by exposure of UV-B radiation (e.g. Häder et al. 1998). As will be discussed below, the flow of organic carbon from the cyanobacteria to the heterotrophic mat community may include the excretion of glycolate during photorespiration, the excretion of compatible solutes after an osmotic down shock, the excretion of fermentation products during dark anoxic conditions and the secretion of extracellular polymeric substances (EPS).

5.2 Oxygenic Photosynthesis

Oxygenic photosynthesis requires the presence of two photosystems (PS I and II). Cyanobacteria contain chlorophyll a in the reaction centers of both PS I and II, but the former contains about 2–3 times as many molecules of chlorophyll a. This chlorophyll may also contribute to the light harvesting, but the phycobiliproteins are far more important pigments as light-harvesting antennae. Jørgensen et al. (1987) demonstrated by recording photosynthetic action spectra in cyano­bacterial mats that chlorophyll a contributed hardly to these action spectra even when additional 600 nm light was given to excite PS II.

Light is strongly attenuated in microbial mats, both by the sediment and by absorption by the dense phototrophic community. Sediments are transparent to light of long wavelengths (Stal et al. 1985). Dry sediments consisting of fine sandy quartz attenuate light much stronger than the same sediment saturated with water (Stal et al. 1985). Through the upper 1 mm of the latter more than 10% of surface irradiance penetrated, while this was only 2.5% of the dry sediment. The photic depth of the bare wet fine sandy sediment was about 4 mm. Through a 1.5 mm mat of cyanobacteria (0.5 g chlorophyll a m−2), 0.45% of photosynthetically active radiation (PAR) penetrated. However, due to the specific absorption of the mat, wavelengths that would support oxygenic photosynthesis are specifically attenuated and oxygenic photosynthesis would not be possible below the cyanobacterial mat (Stal et al. 1985; Jørgensen et al. 1987; Pierson et al. 1987, 1990; Jørgensen and Des Marais 1988) (Fig. 4.4). Sediment and cyanobacterial mats are relatively transparent to light of wavelengths above 700 nm, which explains the occurrence of communities of bacteriochlorophyll a-containing anoxygenic phototrophic purple sulphur bacteria (Pierson et al. 1987, 1990). These findings were confirmed by using fiber-optic microprobes connected to diode array detector (Kühl and Jørgensen 1992) and by a scalar irradiance microsensor which allowed spectral light measurements in sediments at the scale of the phototrophic microorganisms (Lassen et al. 1992a). More than 50% of the incident irradiance of 800 nm light penetrated a 1-mm thick microbial mat (Ploug et al. 1993). Lassen et al. (1992b) used this technique to study photosynthesis and photosynthetic efficiency in a microbial mat in Limfjorden, Denmark. This mat consisted of a top layer of diatoms and a cyanobacterial layer (Oscillatoria spp.) underneath. Using an oxygen micro-sensor, two peaks of oxygenic photosynthesis were found, corresponding to the diatom biofilm and the second deeper maximum corresponded with the layer of cyanobacteria. This latter maximum at 1 mm depth occurred at a light intensity of only 12 μmol photon m−2 s−1 i.e. 1.5% of incident light intensity. However, photosynthetic efficiency (rate of photosynthesis at a specific depth divided by the scalar irradiance at that depth) appeared to be tenfold higher in the cyanobacterial mat compared to the diatom film. This increased photosynthetic efficiency at low light intensity was the result of both the content of cyanobacteria at the depth of the second maximum of photosynthesis as well as a likely increased efficiency with which the available light was absorbed by the organisms (Lassen et al. 1992b). The report (Chen et al. 2010) that a cyanobacterium from a Shark Bay stromatolite can form a newly discovered chlorophyll, chl f ,with the ability to absorb light in the infrared as well as red part of the spectrum, suggests the possibility that this pigment may also contribute to efficient use of light deeper in the stromatolite.

The dense biomass of cyanobacteria in the upper photic zone of microbial mats results in high rates of photosynthesis, and on a surface basis it compares to the productivity of rain forests, which are usually considered as the most productive ecosystems on Earth (Guerrero and Mas 1989) (Table 4.1). Revsbech et al. (1983) measured a total daily photosynthesis in a cyanobacterial mat in Solar Lake (Sinai) of 156 mmol O2 m−2 and similar rates were found by Villbrandt et al. (1990) for a cyanobacterial mat in a temperate region (North Sea). The daily rates of photosynthesis measured by Revsbech et al. (1983) in Solar Lake microbial mats followed the light intensity during the day. The efficiency of photosynthesis was highest at low light intensity (up to 120 μmol photon m−2 s−1). Photosynthesis was not inhibited at high light intensity. The same was found by Villbrandt (1992) when diurnal photosynthesis data on several days were plotted against light intensity (Fig. 4.13a). Photosynthesis increased in a linear way with light intensity. This photosynthesis versus light intensity curve of a microbial mat is completely different from isolated cultures of cyanobacteria (Fig. 4.13b). At low light intensity the photosynthesis curve is steep, depending on photosynthetic efficiency of the organism. At certain, usually moderate, light intensity photosynthesis saturates (P max ) and decreases again at higher intensity as a result of photoinhibition. The different P versus I curve of a mat is explained by the fact that all light is absorbed and used for photosynthesis. At higher light intensity cyanobacteria in the deeper parts of the mat will exhibit higher rates of photosynthesis. The diurnal variation of photosynthesis on a bright, cloudless day in July in a mature mat of M. chthonoplastes on the island of Mellum, North Sea (chlorophyll a content typically 0.3 g  m−2: Stal et al. 1985), showed a different pattern as the one measured by Revsbech et al. (1983). The rates of photosynthesis were highest during the morning hours and showed a sharp drop after mid-day (Villbrandt et al. 1990) (Fig. 4.14b). This mid-day drop in photosynthesis has been observed by other workers both in microphytobenthos as well as in phytoplankton (Paerl et al. 1989; Storch et al. 1990). An explanation for this observation may be that in the morning hours the concentration of dissolved inorganic carbon in the pore water of the mat is high as a result of decomposition of organic matter during the preceding night. After some hours of photosynthetic CO2 fixation the pore water becomes depleted of dissolved inorganic carbon, which could explain the much lower rates of photosynthesis measured in the afternoon. A similar pattern as the one measured by Revsbech et al. (1983) was measured at another site of tidal flat on the island of Mellum. This site was characterized by freshly colonized sediment with Oscillatoria limosa (Lyngbya aestuarii) as the dominant species and with only 1/10 of the biomass compared to the mature mat (chlorophyll content typically 0.03 g  m−2, Stal et al. 1985) (Villbrandt 1992) (Fig. 4.14). Photosynthesis on a surface basis was much lower which is of course due to the much lower biomass and therefore the supply of dissolved inorganic carbon may have been sufficient throughout the day. Photosynthesis on a biomass basis was about twice as high as in the mature mat with comparable surface incident light intensity. This is attributed to the much smaller attenuation of light when biomass is low, i.e. the individual cyanobacteria receive much more light in the freshly colonized sediment compared to the mature mat.

Table 4.1 Comparison of primary productivity in microbial mats with other ecosystems (After Stal 1993)
Full size table
Fig. 4.13
figure 13

(a) Photosynthesis versus light curve in a microbial mat of the North Sea island of Mellum (Germany) (Data from Villbrandt 1992). A large number of depth integrated measurements of photosynthesis recorded at different days and during different times of the day at the same location in the mat were used to plot in this curve. The curve was fitted by linear regression: P (mmol O2 l−1 min−1)  =  0.01062  +  (7.58 10−4)I (R  =  0.99 ± 0.08; N  =  27 P  <  0.0001). (b) Typical photosynthesis versus light intensity curve of a cyanobacterial culture

Full size image
Fig. 4.14
figure 14

Daily light curve (shaded area) and depth integrated photosynthesis of: (a) freshly colonized sediment with Lyngbya sp. as the dominant cyanobacterium; (b) mature mat of Microcoleus chthonoplastes. Both mats were located on tidal sand flats on the North Sea island of Mellum, Germany (Data from Villbrandt et al. 1990)

Full size image

The high rates of photosynthesis often observed in microbial mats may cause supersaturated oxygen conditions. The dense organic matrices represent a diffusion barrier that limits gas exchange. Oxygen bubbles that eventually develop may also be trapped in this matrix, causing the mats to become buoyant and lift off the sediments. Erosion of microbial mats as a result of this phenomenon can be regularly observed. Pieces of mat may be carried to vegetated areas and become desiccated when the tide has gone. Such desiccated pieces may be transported by wind over long distances. This phenomenon was first reported in 1686, becoming known as “Meteorpapier” because of the belief that it came from space. In his publication “Über das im Jahre 1686 in Curland vom Himmel gefallene Meteorpapier und über dessen Zusammensetzung aus Conferven und Infusorien” Ehrenberg (1838) identified this meteor paper as desiccated microbial mats.

5.3 Anoxygenic Photosynthesis

Anoxygenic photosynthesis in microbial mats is not the exclusive trait of purple- or green sulphur bacteria. Some species of cyanobacteria are capable of anoxygenic photosynthesis in which only photosystem I is involved. As with phototrophic sulphur bacteria, anoxygenic photosynthesis in cyanobacteria depends on sulphide as the electron donor. Roughly two categories of cyanobacteria can be distinguished with respect to the capacity of anoxygenic photosynthesis. In one group oxygenic photosynthesis is inhibited at low concentrations of sulphide and anoxygenic photosynthesis is induced. Inhibition of oxygenic photosynthesis by sulphide is probably at the level of the manganese-containing, water-splitting enzyme (Oren et al. 1977). Both types of photosynthesis are mutually exclusive in these organisms. In the other group anoxygenic and oxygenic photosynthesis occur concurrently. At low sulphide concentrations oxygenic photosynthesis is more important and with increasing sulphide concentrations anoxygenic photosynthesis gradually takes over.

In both types of cyanobacteria, anoxygenic photosynthesis must be induced, a process which depends on a certain threshold of sulphide concentration and on light. Induction of anoxygenic photosynthesis in some organisms may take several hours. Therefore, cyanobacteria that possess the capability of carrying out oxygenic and anoxygenic photosynthesis concurrently have an ecological advantage in environments in which the sulphide concentration fluctuates, as is the case for instance in many marine and hypersaline microbial mats. Cyanobacteria that can carry out only one type of photosynthesis at a time are typical of environments with a constant supply of sulphide, as in certain hot spring microbial mats with an indigenous supply of sulphide. Moreover, these cyanobacteria tolerate higher concentrations of sulphide.

In cyanobacteria, anoxygenic photosynthesis is defined as energy generation through cyclic electron flow driven by photons absorbed by the reaction center of photosystem I and the fixation of CO2 using sulphide as the electron donor. This means that this process also takes place when photosystem II is inhibited by the herbicide 3-(3,4-dichlorophenyl)-1,1-dimethylurea (DCMU) or when illuminated by far red light (>700 nm) which cannot be used by photosystem II. Cyanobacteria that perform anoxygenic photosynthesis oxidize sulphide to elemental sulphur according to the following stoichiometric equation:

$$ 12{\rm{H}}_{2}\rm{S }+{\rm{ 6 CO}}_{2}\to {\rm{C}}_{6}{\rm{H}}_{12}{\rm{O}}_{6}+12{\rm{S}}^{\rm{o}}+6{\rm{H}}_{2}\rm{O}$$

The elemental sulphur that is produced is deposited outside the cell but is often found attached to the outer sheath of the cyanobacterium as finely dispersed particles. The elemental sulphur is not further oxidized, as is the case in purple sulphur bacteria. In the mat-forming Microcoleus chthonoplastes (this particular strain is now assigned to Geitlerinema and is therefore closely related to the anoxygenic Solar Lake cyanobacterium ‘Oscillatoria limnetica’, now also reassigned to Geitlerinema) thiosulphate was found to be the product of sulphide oxidation (De Wit and Van Gemerden 1987):

$$ 6{\rm{H}}_{2}\rm{S}+\\3{\rm{H}}_{2}\rm{O}+{\rm{6 CO}}_{2}\to {\rm{C}}_{6}{\rm{H}}_{12}{\rm{O}}_{6}+3{\rm{H}}_{2}{\rm{S}}_{2}{\rm{O}}_{3}$$

While the oxidation of sulphide to sulphur yields only two electrons, the oxidation to thiosulphate yields four electrons per sulphide oxidized. Thus, the oxidation of sulphide to thiosulphate in M. chthonoplastes (Geitlerinema) seems to be twice as efficient as in other cyanobacteria in which zero-valent sulphur is the product. Rabenstein et al. (1995) reported that sulphite was the intermediate in those cyanobacteria that oxidized sulphide to thiosulphate. It is possible that thiosulphate is formed in a chemical reaction of sulphite with sulphide.

It is not clear why most anoxygenic cyanobacteria oxidize sulphide only to elemental sulphur. Phototrophic sulphur bacteria usually oxidize sulphide to sulphate, which yields eight electrons. However, in these bacteria the oxidation to elemental sulphur is usually much faster as the further oxidation to sulphate. Depending on the species this elemental sulphur is accumulated intra- or extracellularly. It can be hypo­thesized that the rapid oxidation of sulphide to elemental sulphur would be advantageous since it assures a rapid removal of the toxic sulphide. Moreover, in anoxygenic phototrophic bacteria, no matter whether they store the elemental sulphur intra- or extracellularly, it is not available for other organisms (Van Gemerden 1987). In that way, when sulphide is depleted these organisms continue anoxygenic photosynthesis at the expense of stored sulphur without competition with other organisms. In cyanobacteria the rapid consumption of sulphide most likely serves for detoxification since they can immediately switch to oxygenic photosynthesis when it has been completely oxidized. In addition, cyanobacteria can also use elemental sulphur as electron acceptor in anaerobic dark metabolism. O. amphigranulata is capable of using elemental sulphur for assimilatory purposes (Castenholz and Utkilen 1984). In the unicellular Anacystis nidulans low rates of CO2 fixation are supported by the oxidation of thiosulphate (Utkilen 1976; Peschek 1978).

M. chthonoplastes (Geitlerinema) belongs to the group that performs oxygenic and anoxygenic photosynthesis concurrently. In this organism the growth rate decreased exponentially with increasing concentrations of sulphide. At a concentration of 1 mM sulphide (pH 8.0) growth was completely inhibited. Oxygenic photosynthesis was gradually inhibited with increasing concentrations of sulphide. The relative contribution of anoxygenic photosynthesis to total photo­synthesis was >95% at concentrations of sulphide exceeding 0.35 mM (pH 8.0). The inhibition of growth at 1 mM of sulphide is probably caused by the complete inhibition of oxygenic photosynthesis, rather than anoxygenic photosynthesis. It was shown that M. chthonoplastes (Geitlerinema) requires oxygen for growth. When oxygenic photosynthesis was inhibited by DCMU, M. chthonoplastes (Geitlerinema) was not capable of sulphide-dependent anoxygenic photo­trophic growth unless some oxygen was present. Therefore, a small contribution of oxygenic photosynthesis may be necessary in order to provide the essential oxygen. Oxygen may be required for the oxidation of fatty acids. Anacystis halophytica, for instance, possesses an oxygen-dependent desaturation mechanism (Padan and Cohen 1982). Several cyanobacteria contain polyunsaturated fatty acids (Kenyon et al. 1972) and Padan and Cohen (1982) suggested that such cyanobacteria may be incapable of anaerobic growth. O. limnetica (Geitlerinema) does not contain polyunsaturated fatty acids which may explain its capacity for anaerobic growth (Padan and Cohen 1982). M. chthonoplastes (Geitlerinema) SAG 3192 contains considerable amounts of linoleate (18:2), linolenate (18:3) and tetradecadienate (14:2), regardless whether the culture was grown in the presence or absence of sulphide (De Wit et al. 1988).

The affinity of M. chthonoplastes (Geitlerinema) for sulphide is extremely low. The K m for sulphide has been calculated as 974 μM, approximately the concentration at which growth of M. chthonoplastes (Geitlerinema) ceased (De Wit and Van Gemerden 1987). These authors analyzed the data of Jørgensen et al. (1986), who investigated the transition of anoxygenic photosynthesis to oxygenic photosynthesis in a mat of M. chthonoplastes. A K m of 710 μM was calculated for sulphide oxidation in this mat. This affinity is close the one calculated in culture by De Wit and Van Gemerden (1987). These affinities for sulphide are extremely low when compared to the value of 5 μM for an anoxygenic phototrophic purple sulphur bacterium as Thiocapsa roseo­persicina ,which is frequently present in microbial mats (De Wit and Van Gemerden 1988). The similarity of the K m of sulphide oxidation estimated in a culture to that estimated in a natural microbial mat of M. chthonoplastes indicates that this organism was probably responsible for the sulphide oxidation in the microbial mat. This was also suggested by Jørgensen et al. (1986) although they concluded this from the fact that purple sulphur bacteria constituted only a minor fraction in that mat.

The results on anoxygenic photosynthesis in a culture of M. chthonoplastes (Geitlerinema) obtained by De Wit and van Gemerden confirmed those for a natural mat by Jørgensen et al. (1986). These authors found that oxygenic and anoxygenic photosynthesis occurred concurrently. However, at higher concentrations of sulphide oxygenic photosynthesis was insignificant. Oxygenic photosynthesis in this mat started when the sulphide concentration decreased to about 0.3 mM, which is in agreement with the results obtained by De Wit and van Gemerden. The experiments of Jørgensen and co-workers showed that oxygenic photosynthesis could occur even when the microbial mat was exposed to 5–6 mM of sulphide in the overlying water. An oxygen peak was sandwiched between layers of sulphide. This sandwiching of cyanobacteria in microbial mats is often observed. Apparently, M. chthonoplastes is capable of resisting high concentrations of sulphide and recovers oxygenic photosynthesis. It is also capable of oxidi­zing this high concentration of sulphide in the light but this does not mean that the organism is indeed growing or even fixing CO2.

Taking together the extremely low affinity of M. chthonoplastes for sulphide, the low growth rate with anoxygenic photosynthesis and the fact that this organism cannot grow in the absence of oxygen, the major function seems to be the detoxification of sulphide.

5.4 CO2 Fixation

Carbon dioxide is the most important source of carbon for cyanobacteria and it is therefore crucial for the functio­ning of microbial mats. Cyanobacteria use the energy and low potential reductant (NADPH) produced during photosynthesis to fix CO2 through the reductive pentose phosphate pathway (Calvin-Benson-Bassham cycle). The same pathway in the opposite direction, the oxidative pentose pathway, is used for the oxidation of storage carbohydrate during the dark in combination with aerobic respiration (Smith 1982; Schmetterer 1994).

High rates of photosynthesis will deplete the sediment of CO2 and raise the pH. The pH may even reach values of over 9.5 (Revsbech et al. 1983) (Fig. 4.3) and any dissolved inorganic carbon will be present as bicarbonate or carbonate. Cyanobacteria are capable of adapting to growth at extremely low concentrations of dissolved inorganic carbon. Both CO2 and bicarbonate are taken up by cyanobacteria. However, CO2 is the substrate for RubisCO, the key enzyme of the Calvin-Benson-Bassham cycle and responsible for the fixation of CO2. As is the case in other autotrophic organisms, this enzyme has also a low affinity for CO2 in cyanobacteria, which means that both a high concentration of CO2 and of RubisCO are prerequisites for the efficient fixation of carbon dioxide. Cyanobacteria possess an inorganic carbon-concentrating mechanism (CCM) which may result in up to 1,000-fold accumulation of inorganic carbon in the cell. A tentative model of this CCM in cyanobacteria proposes that either bicarbonate or CO2 is taken up but that the former is the predominant species of inorganic carbon in the cytoplasm (Kaplan et al. 1994). Bicarbonate enters the carboxysome, a cell inclusion in autotrophic bacteria also known as polyhedral bodies. Carboxysomes contain virtually all RubisCO in organisms that possess these inclusions. The importance of carboxysomes for the CCM is also shown by the observations of Turpin et al. (1984) and McKay et al. (1992) that the number of these inclusions increases during adaptation of cyanobacteria to low CO2 concentrations.

The fixation of CO2 in microbial mats can be investigated by measuring the 12C/13C carbon isotope ratio in the organic matter. RubisCO discriminates between carbon isotopes with a slight preference for the lighter isotope 12C. This fractionation factor α equals 1.029 (Roeske and O’Leary 1984), which means that organic matter may become 29‰ depleted in the heavy isotope 13C when its origin is from RubisCO mediated CO2 fixation. This isotopic discrimination is only achieved when the CO2 concentration is sufficiently high. This is generally not the case in microbial mats. Moreover, the measured value may differ from expected one because other organisms that were responsible for RubisCO independent CO2 fixation may have been present in the system. Also cyanobacteria may fix significant amounts of CO2 via alternative pathways such as PEP carboxylase or carbamylphosphate. Furthermore, the dissolved inorganic carbon produced from the decomposition of organic matter may be recycled and give rise to different net isotope discrimination. At the low concentrations of CO2 that usually occur in cyanobacterial mats active transport of HCO −3 becomes important (Badger and Andrews 1982) which results in a much smaller isotope discrimination than in the case of CO2 uptake (Des Marais and Canfield 1994). Microbial mats are usually not much depleted in 13C (δ13Cmat is not very negative) because the pool of dissolved inorganic carbon is small compared to the rate of CO2 fixation. This minimizes the isotope discrimination. The most negative values of δ13Cp (photosynthate) are expected when CO2 does not become depleted from the medium and when exchange between the medium and the site of fixation is rapid.

Des Marais and Canfield (1994) investigated the carbon isotope discrimination in two microbial mats in Guerrero Negro, Baja California, Mexico. The δ13Cmat in these mats was slightly negative (−70‰). In the mat of Lyngbya aestuarii this value corresponded with the fractionation factor 1.007. This low value was evidently attributed to the closed reservoir behaviour of the system. The dissolved inorganic carbon that was produced by the mat during the night possessed the same negative value and therefore no changes in the δ13Cmat were expected in the L. aestuari mat. In the mat of M. chthonoplastes photosynthesis did not discriminate between the lighter and heavier isotopes. At present a conclusive explanation for the negative value of δ13Cmat in the M. chthonoplastes mat is not available (Des Marais and Canfield 1994). Processes such as excretion, fermentation and respiration do not change isotopic discrimination. Diagenesis of organic matter does not alter its isotopic composition (Des Marais et al. 1992) and the δ13CDIC is similar as δ13Cmat (Bauer et al. 1991). It is known that the ‘pond 5’ mats of M. chthonoplastes are more or less in ‘steady state’, i.e. most of the organic matter that is produced by photosynthesis is mineralized in the mat. It is likely that photosynthesis scavenges very efficiently the dissolved inorganic carbon that is produced in the mat, thereby limiting net isotope fractionation. In fossil Proterozoic stromatolites δ13C is much more negative than in present day microbial mats and stromatolites. This may reflect the higher levels of dissolved inorganic carbon in the Precambrian compared to today’s concentrations (Kempe and Kazmierczak 1990a) or a more negative δ13CDIC.

5.5 Photorespiration and Glycolate Excretion

During daylight, the dense phototrophic biomass in the cyano­bacterial mat depletes CO2 and accumulates oxygen, which sometimes may reach high supersaturation. It is assumed that such conditions will support photorespiration. Besides carboxylation, RubisCO also possesses oxygenase activity and can oxidize ribulose-1,5-bisphosphate to one molecule of each 2-phosphoglycolate (2PG) and 3-phosphoglycerate (3PGA) instead of two molecules of the latter during the carboxylation reaction (Lorimer et al. 1973; Lorimer 1981; Miziorko and Lorimer 1983). In fact, RubisCO has a much better affinity for oxygen as substrate than for carbon dioxide (Pierce 1988) and it has been suggested that the original function of the enzyme was an oxygenase rather than a carboxylase (Tabita 1988). Warburg (1920) discovered that O2 inhibited CO2 fixation in algae. Schau et al. (1950) showed that glycolate was a product of CO2 fixation and Warburg and Krippahl (1960) demonstrated that its synthesis could be stimulated by oxygen. Glycolate is produced from 2PG by phosphoglycolate phosphatase and is metabolized via the glycine-serine (C2) pathway (Renstrom-Kellner and Bergman 1990), resulting in the production of 3PGA, CO2 and NH3. This light-dependent oxygen uptake and CO2 evolution is called photorespiration. Photorespiration may represent a loss of fixed carbon which may be as high as 15–50% of net photosynthesis (Artus et al. 1986; Gerbaud and Andre 1987). What function of photorespiration is so important to justify this loss of fixed C?

In plants, the C2 pathway that metabolizes the toxic intermediate of photorespiration 2PG is essential for photosynthesis. Mutations in the C2 pathway result in plants with a high CO2-requiring phenotype (high CO2 would minimize photorespiration). In cyanobacteria the oxygenase reaction has been considered as irrelevant because these organisms possess a carbon concentration mechanism (CCM) which would allow sufficient high CO2 concentration at the site of RubisCO. Cyanobacteria were known to convert 2PG to glycolate. However, it has now become clear that 2PG meta­bolism is probably present in all cyanobacteria, even in those with the smallest genomes, such as the marine picocyanobacteria Prochlorococcus and Synechococcus. The model strain Synechocystis PCC 6803 possesses three routes to meta­bolize 2PG: the plant-type C2 route, the bacterial glycerate pathway and the conversion of glyoxylate via oxalate to formate and subsequently to CO2 (Eisenhut et al. 2008). It was also shown that 2PG metabolism was obligatory and that it could not be compensated by the CCM of this organism. Eisenhut et al. (2008) postulated that 2PG metabolism evolved simultaneously with oxygenic photosynthesis in cyanobac­teria to allow them to cope with the toxic products generated by photorespiration that could occur because of the oxygen that accumulated in the cell and in the microbial mats, while the CCM evolved only recently (Raven et al. 2008). Hence, the C2 pathway in plants was probably inherited from the cyanobacteria as well (Eisenhut et al. 2008).

Although it seems reasonable to assume that the metabolism of 2PG serves the elimination of toxic intermediates produced by the oxygenase reaction of RubisCO, there may be other benefits from photorespiration. 3PGA is regenerated for the Calvin-Benson-Bassham cyclus and other metabolic intermediates may be produced (Husic et al. 1987). The CO2 produced may be re-fixed and the NH3 may be re-assimilated even if this at the expense of ATP in both cases. In microbial mats photorespiration may help to prevent photooxidative damage by the lowering of O2.

It is questionable whether in microbial mats the CCM would be sufficient to prevent photorespiration, because of the depletion in CO2 during the daytime. If all cyanobacteria are capable of 2PG metabolism as is the case in Synechocystis PCC 6803, then it seems unlikely that glycolate is excreted, and this compound would therefore not be important as a substrate in microbial mats.

Many microbial mats are characterized by oxygen supersaturation in the light and CO2 depletion and by very high pH (sometimes above 10). They are also subject to high light intensities and are chronically nitrogen depleted. All these factors will force the cyanobacteria to maximum rates of photorespiration. Glycolate metabolism, and therefore photo­respiration, is closely associated with nitrogen metabolism. Renstrom-Kellner and Bergman (1989) demonstrated that the excretion of glycolate by Anabaena cylindrica decreased drastically in the presence of a source of nitrogen such as NH4Cl or glutamate. As was shown by these authors, N2-fixing cyanobacteria could lose up to 60% of photosynthetic fixed CO2 as glycolate. Heterocystous cyanobacteria can access nitrogen through N2 fixation. However, most mat-forming cyanobacteria are non-heterocystous and probably grow under severe nitrogen limitation. This suggests that these organisms may even lose the greater part of net photosynthesis. Bateson and Ward (1988) showed the importance of glycolate as a substrate for the microbial community in microbial mats. Glycolate may be utilized by sulphate-reducing bacteria, even in the presence of oxygen (Fründ and Cohen 1992). Glycolate-oxidizing sulphate-reducing bacteria have indeed been isolated from marine sediments (Friedrich and Schink 1993, 1995), but these organisms were strictly anaerobic.

5.6 Organic Compatible Solutes

Cyanobacteria exposed to high salinity or drought, accu­mulate osmoprotectors and extrude sodium ions through the activation or adaptation processes. These include (1) the uptake or biosynthesis of compatible solutes, (2) the active extrusion of sodium ions through the enhancement of ATPase activity, (3) modifications of the membrane lipid composition and (4) increase the energetic capacity through cyclic electron transport through photosystem I and through respiration.

In marine and hypersaline environments, micro-organisms accumulate solutes in order to obtain a sufficient turgor pressure necessary to allow cell division and growth (Taiz 1984). The cytoplasmic membrane is permeable to water and an organism that is exposed to an elevated salt concentration in the surrounding medium would tend to lose water. In order to retain water inside itself the cell can either take up ions until an osmotic equilibrium with the environment is obtained or accumulate low molecular weight organic solutes. High concentrations of inorganic ions are not compatible with the metabolism of cyanobacteria and cause inhibition of enzyme activity (Warr et al. 1984).

Cyanobacteria can be subdivided into three groups with respect to the type of organic solute they accumulate in response to osmotic stress (Reed et al. 1986a). Halotolerant freshwater cyanobacteria accumulate disaccharides (either sucrose or trehalose). Marine cyanobacteria accumulate the heteroside glucosylglycerol (2-O-α-d-glucopyranosyl-glycerol) and the very halotolerant hypersaline cyanobacteria accumulate quaternary ammonium compounds (glycine betaine and in one case glutamate betaine) (Mackay et al. 1984). There is no clear link between the type of solute and the taxonomic group of cyanobacteria, although all strains of Anabaena that were screened accumulated sucrose in response to osmotic stress. A habitat relation is suggested among species of the unicellular Synechococcus. Of the 33 strains investigated, all originating from freshwater environments accumulated sucrose, those isolated from marine systems accumulated glucosylglycerol and those from hypersaline habitats without exception betaine (Reed et al. 1986a). Stal and Reed (1987) screened 25 strains of cyanobacteria isolated from a microbial mat in the North Sea and found glycosylglycerol as well as trehalose and sucrose as osmolytes, suggesting no habitat relationship with the type of solute. Glucosylglycerol was nevertheless typically the dominant osmolyte in this marine ecosystem. The two dominant cyanobacteria in this mat, M. chthonoplastes (Geitlerinema) and O. limosa (Lyngbya) accumulated glucosylglycerol and trehalose, respectively. This property has been used to estimate the respective biomass of both species in these microbial mats (Stal and Reed 1987). Karsten (1996) measured the compatible solutes of a variety of strains of M. chthonoplastes isolated from various geographic locations and found that they contained glucosylglycerol as well as trehalose. However, the latter prevailed under sub-optimal salinities, while it appeared that the glucosylglycerol served as the only osmolyte. Betaine seems not to be limited to cyanobacteria from hypersaline environments since it has been identified in marine picocyanobacteria Synechococcus (Lu et al. 2006).

Although cyanobacteria normally accumulate a single low-molecular weight organic compound in response to osmotic stress, many species may produce a secondary compound. The synthesis of disaccharide is much faster than glucosylglycerol. Within 8 h of an osmotic upshock the disaccharide pool has reached 90% of its maximum, while with glucosylglycerol this is only the case after 24–48 h (Reed and Stewart 1988). Therefore the synthesis of disaccharide as secondary osmolyte may help for a quicker response to salt stress. Thus cyanobacteria that thrive under relative constant salinities may prefer glucosylglycerol while those that are exposed to fluctuating salinities may be better off with trehalose for example. This difference could explain why the pioneer in microbial mats, Oscillatoria (Lyngbya) sp., contains trehalose while the typical organism in established microbial mats, M. chthonoplastes contains glucosylglycerol. For the same reason hypersaline species contain sucrose in addition to betaine. Since betaine is a nitrogen-containing compound, nitrogen deficiency may also lead to the accumulation of sucrose as secondary osmolyte (Trüper and Galinski 1989). It has been shown that only glycine betaine provided a significant protection of enzyme activity against Na+ ions, suggesting that sugars and polyols protect by a different mechanism (Warr et al. 1988).

Osmotic down shock exerted on betaine-containing Aphanothece halophytica resulted in the release of this osmolyte into the environment (Reed and Stewart 1988). This may have important consequences for an ecosystem such as a microbial mat because it may allow chemotrophic bacteria that cannot synthesize betaine to take it up from the environment (Reed and Stewart 1988). Moreover, betaine may serve as substrate for sulphate-reducing bacteria and the product of its metabolism, trimethylamine (TMA) is known as a so-called non-competitive substrate for methanogenic bacteria (Heijthuijsen and Hansen 1989).

Osmotic downshock in Rivularia atra resulted in a corresponding decrease of the osmoticum trehalose but only 10% was recovered from the medium and the rest was apparently metabolized or converted to glycogen (Reed and Stewart 1983). The glucosylglycerol accumulating strain Synechocystis PCC6714 and the sucrose-containing Synechococcus PCC6311 released 50% of their carbohydrates and over 70% of their amino acids after experiencing hypo-osmotic shock (Reed et al. 1986b). However, in some other cyanobacteria there is no evidence for the release of low molecular weight compounds upon hypo-osmotic shock (Reed and Stewart 1988). The release by cyanobacteria of low molecular weight compounds into a microbial mat would have a great impact on the ecosystem. The cellular concentration of these osmolytes is considerable and at full seawater salinity it may amount to as much 270 mM. These carbohydrates are easy accessible substrates for chemotrophic bacteria in the mat. Except in the case of a hypo-osmotic shock, which may occur in exposed microbial mats after a rain shower for instance, osmotica will also be liberated after death and lysis of the organism.

Microbial mats have often been found to evolve dimethylsulphide (DMS), a sulphur-containing organic volatile compound. It is known that DMS can be produced from dimethylsulphoniopropionate (DMSP) by microbial activity or by chemical decomposition at high pH (Kiene and Visscher 1987). DMSP occurs in a number of algae where its most likely function is that it serves as osmoprotectant (Turner et al. 1988). Vogt et al. (1998) suggested that some cyano­bacteria might perhaps contain minor amounts of DMSP and Visscher and Van Gemerden (1991) suggested that M. chthonoplastes (Geitlerinema) may produce DMSP as a secondary osmolyte and could be the source of DMS in microbial mats. However, Van Bergeijk and Stal (1996) found a correlation between the number of diatoms in these mats and the amount of DMS that evolved from it. Some benthic diatoms accumulate large amounts of DMSP as osmoticum. It is possible that cyanobacteria take up DMSP from the environment (Vila-Costa et al. 2006).

5.7 Fermentation

When in microbial mats photosynthesis ceases, they may rapidly, sometimes even within minutes, turn anoxic. Cyano­bacteria are essentially aerobic organisms that during the dark normally have a respiratory metabolism in which the endogenous storage carbohydrate glycogen is degraded (Smith 1982). When oxygen is absent, aerobic respiration is evidently not an option. Many cyanobacteria die and lysis occurs within 2–3 h after transfer to dark anoxic conditions. However, mat-forming cyanobacteria survive dark anoxic conditions for much longer time, often for several days. A number of these cyanobacteria were investigated in more detail and it was discovered that they were capable of fermenting glycogen. Fermentation as a constitutive property would have a number of advantages. Primarily, it would greatly increase the reactivity of the organism. Microbial mats are generally environments in which steep gradients of light and oxygen occur and these factors fluctuate strongly. If oxygen disappears rapidly, fermentation can immediately provide energy for maintenance, allowing the organism to survive. The excretion of fermentation products by cyanobacteria is an important process in microbial mats because it supplies other microorganisms, notably the sulphate-reducing bacteria, with substrate.

There is a great diversity of fermentation pathways in cyano­bacteria. In some cases the pathways have been elucidated by the demonstration of the enzymes involved. In O. limosa (Lyngbya) the presence of the key enzymes of the homoacetate pathway has been demonstrated as well (Heyer et al. 1989). In the majority of pathways the Embden-Meyerhof-Parnas pathway (glycolysis) was involved in fermentation. Only the heterolactate fermentation makes use of parts of the oxidative pentose phosphate pathway. The oxidative pentose phosphate pathway is used by cyanobacteria during aerobic dark respiration and it is essentially the reverse of the reductive pentose pathway, which serves CO2 fixation in the light (Smith 1982). In all cyanobacteria capable of fermentation the capacity for fermentation appears to be constitutive (Stal and Moezelaar 1997).

Stal and Moezelaar (1997) reviewed fermentation in cyano­bacteria. Table 4.2 lists cyanobacteria capable of fermentation. The phenomenon was first discovered in O. limnetica (Geitlerinema) (Oren and Shilo 1979), which occurs in the sulphide-rich hypolimnion of Solar Lake, Sinai, and is typically adapted to anaerobic growth. In the dark this organism ferments glycogen to lactate. Since no other fermentation product was found, it was assumed that the homolactic acid pathway was used in this organism. In the non-heterocystous diazotrophic mat-building O. limosa (Lyngbya), heterolactic acid fermentation was found (Heyer et al. 1989). This organism produced equimolar amounts of lactate and ethanol from glycogen. In addition, it is capable of homoacetic fermentation, for which its osmoprotectant trehalose was used as substrate. Trehalose was degraded to 5–6 acetate and some hydrogen and CO2. The occurrence of homoacetate fermentation in cyanobacteria is remarkable, since it further only occurs in a group of specialized anaerobic bacteria, the acetogenic bacteria. Homoacetate fermentation is energetically efficient. The occurrence of homoacetate fermentation has been proposed in a number of other cyanobacteria (De Philippis et al. 1996). The degradation of the osmoprotectant in O. limosa (Lyngbya) was another unexpected phenomenon. Trehalose represents a large amount of energy, which may be important for the organism to use under a situation of severe starvation. The question of how the organism compensates for the loss of compatible solute has not been answered, but it has been suggested that this may be through a temporary accumulation of inorganic ions such as K+ (Stal and Moezelaar 1997). Also, the mat building cyanobacterium M. chthonoplastes (Geitlerinema) has been shown to ferment part of its osmoprotectant (Moezelaar et al. 1996). M. chthonoplastes (Geitlerinema) accumulates the heteroside glucosyl glycerol which is only degraded in cultures that contain low amounts of glycogen. Unlike in O. limosa (Lyngbya), M. chthonoplastes (Geitlerinema) possesses just one fermentation pathway. Glycogen and the glucose part of glucosyl glycerol are fermented via a mixed acid fermentation, resulting in the formation of formate, acetate, ethanol, lactate, H2 and some CO2. The presence of the homoacetogenic pathway allows O. limosa (Lyngbya) acetogenesis from CO2 and H2 (Stal, unpublished observations). Acetogenesis from CO2 has been observed in anoxic sediment (Hoehler et al. 1999).

Table 4.2 Cyanobacteria capable of fermentation (After Stal and Moezelaar 1997)
Full size table

Hydrogen is often a product of fermentation in cyanobacteria. Hydrogenases in cyanobacteria have been extensively reviewed by Tamagnini et al. (2002, 2007). Cyanobacteria possess different hydrogenases. N2-fixing cyanobacteria produce hydrogen as a by-product of nitrogenase. Because nitrogenase obligatory produces hydrogen during N2 fixation, aerobic N2-fixing cyanobacteria usually possess an uptake hydrogenase. This enzyme carries out an oxy-hydrogen reaction. The third type of hydrogenase in cyanobacteria is reversible hydrogenase. This enzyme is frequently found in obligate anaerobic bacteria. Depending on the conditions it catalyses either the uptake or the production of hydrogen at approximately equal rates. Although its function in cyanobacteria has been debated for some time, reversible hydrogenase plays an important role in fermentation (Stal and Moezelaar 1997). Hydrogen concentrations are kept low in microbial mats because sulphate reducing bacteria, acetogenic bacteria, methanogenic bacteria and anoxygenic phototrophic bacteria use it as energy source and/or as electron donor (Hoehler et al. 2002). The escape of reduced gases from microbial mats into the atmosphere may have contributed to the oxygenation of the oceans on early earth (Hoehler et al. 2001).

Elemental sulphur may serve as electron acceptor in cyanobacteria. Many cyanobacteria have been shown to be able to reduce elemental sulphur to sulphide. It has been suggested that this process in O. limnetica (Geitlerinema) might represent a form of anaerobic respiration (Oren and Shilo 1979). However, in other cyanobacteria the advantage of the reduction of sulphide is probably that it serves as electron sink, allowing the formation of more oxidized pro­duct (acetate) which results in a higher amount of substrate phosphorylation.

Stal and Moezelaar (1997) have discussed the bioenergetics of fermentation in a number of different cyanobacteria for which sufficient information is available. Although evidently the amount of energy that is generated during fermentation is low, calculations showed that it usually exceeded the minimum amount required for maintenance. This remaining energy could potentially drive metabolic processes. For instance, O. limosa (Lyngbya) is even capable of maintaining a considerable rate of N2 fixation under anaerobic conditions in the dark (Stal and Heyer 1987).

Hot spring microbial mats consisting of Synechococcus switched on a variety of genes involved in fermentation when in the dark and anaerobic conditions although the fermentation products were not identified (Steunou et al. 2006). This fermentation supported the fixation of N2 that occurred in the dark in these mats. Anderson et al. (1987) investigated the fate of representative fermentation products (acetate, propionate, butyrate, lactate, and ethanol) in hot spring cyanobacterial mats. Fermentation occurred mainly in the top 4 mm of the mat. In the light, filamentous bacteria resembling Chloroflexus aurantiacus photoassimilated the fermentation products. In the dark under anaerobic conditions, only lactate was oxidized and also the extended incubation under these conditions did not enhance the metabolism of acetate, propionate, or ethanol. Acetogenic bacteria converted butyrate into acetate. In mats occurring at temperatures ranging from 50°C to 70°C acetate and propionate accumulated under dark anaerobic conditions.

5.8 Extracellular Polymeric Substances (EPS)

Extracellular polymeric substances are important components of microbial mats. They are involved in the attachment of cyanobacteria to the substrate and are essential structuring molecules by producing a matrix in which the organisms are embedded (Decho 2000). This polymeric matrix fulfils a number of other important functions in microbial mats, which will be discussed below.

Cyanobacterial exopolysaccharides are complex molecules composed of 6 or more different monosaccharides out of a suite of at least 12 sugars (De Philippis and Vincenzini 1998). The variety of linkages that is possible gives a broad range of possible structures. Cyanobacteria produce polysaccharides which can be roughly categorized in three groups: (i) endo­genous polysaccharides that serve as storage compounds; (ii) cell envelope polysaccharides; (iii) extracellular polysaccharides. The endogenous polysaccharides in cyanobacteria can be found in the so-called α-granules, which are composed of a branched glycogen-like polymer. This polymer consists of α(1–4) and α(1–6) linked glucose molecules. The cell envelope consists of the cell wall polysaccharides and the external layers (glycocalyx). The glycocalyx can be subdivided in (i) the well-structured polysaccharide sheath, (ii) a polysaccharide capsule which extends outside the sheath but is clearly associated with the organism and is less structured and (iii) mucilage polysaccharide. The latter is not or very loosely associated with the organism. In fact, the polysaccharides that form the glycocalyx should all be considered as extracellular polysaccharides or exopolysaccharides (Fig. 4.16). Exopolysaccharides that are released into the surrounding environment may be the colloidal suspended molecules originating from any of the glycocalyx components. The different fractions are often poorly defined and mostly based on the different extraction procedures. Relatively little is known about cyanobacterial exopolysaccharides and their biosynthetic pathways are complex and not well known (Pereira et al. 2009).

Microbial exopolymers, including those produced by cyanobacteria, are high molecular weight mucous secretions that often have a complex structure. The molecular weight is often more than 100,000 Da. Although polysaccharides are quantitatively the most important part of these exopolymers, other components are present as well. Proteins make up a significant part of the exopolymers (Decho 1990). These polymers are also known as extracellular polymeric substances (EPS). The composition and structure of EPS vary widely among different microorganisms (Tago and Aida 1977; Bertocchi et al. 1990; Decho 1994; Stal 1994) and even one single strain may produce more than one type EPS simultaneously or at different stages of growth (Christensen et al. 1985). Most of the polysaccharides in EPS are hetero­polysaccharides that are composed of a variety of different monosaccharides, arranged in repeating units. EPS often contain uronic acids such as D-glucuronic acid, D-galacturonic acid and D-mannuronic acid. These are important functional groups because they contain carboxyl groups that are responsible for interactions with other EPS molecules or the binding of metals. However, other types of EPS are composed of neutral sugars. Extracellular polymeric substances may be hydrophilic or hydrophobic. Many are hydrophilic and may contain over 95% water by weight (Decho 1994). Depending on the chemical composition and the functional groups present, the tertiary structure of EPS is determined. The tertiary structure of EPS determines whether it is a cohesive gel or in a colloidal form. An intermediate form could be described as nonconsolidated mucilage (Decho 1994). The tertiary structure of EPS not only depends on the chemical composition but also strongly on temperature. Microbial mats and intertidal mudflats during emersion are subject to large variations in temperature and this will thus affect the cohesiveness and rheological properties of the sediment.

A large number of functions have been ascribed to EPS (Decho 1990). These include adhesion and immobilisation of the organism, protection against desiccation, protection from grazing, protection from toxic substances, scavenging of trace metals, and (anti-) calcification. Some of these functions will be discussed below as far as they are relevant to microbial mats.

Organisms in microbial mats are often subject to desiccation. EPS may retain large amounts of water and form a gel that stabilizes the macromolecular components and the cell structure of the cyanobacteria and organisms that produce it may overcome long periods of drought by forming hydrogen bonds with proteins, membrane lipids and DNA, thereby replacing the water shell surrounding these cell constituents (Caiola et al. 1996; Potts 1994). Some cyanobacterial EPS may be hydrophobic due to the presence acetyl-groups, peptide moieties or desoxysugars which determine the emulsifying properties and the rheological properties (Neu 1992). Particularly, EPS containing uronic acids or hydrophobic proteins may be important for micro-organisms, including cyanobacteria and diatoms, enabling these to attach to surfaces (Robins et al. 1986). For benthic organisms, it is important to stay on surfaces when conditions are optimal for growth. Some cyanobacteria are capable of modifying EPS from hydrophobic to hydrophilic and they may thus detach from a surface when conditions become inappropriate (Bar-Or et al. 1985). Benthic communities of diatoms may attach to the surface of intertidal mudflats by the production of hydrated and hydrophilic exopolymers during periods of emersion. During immersion, these polymers go into solution releasing the diatoms into the water column (Talbot et al. 1990). Benthic cyanobacteria may secrete flocculants, exopolymers that produce flocs with detritus and other material in the overlying water. These flocs eventually sediment, thereby clearing the overlying water and hence improving the conditions for these benthic phototrophs (Bar-Or and Shilo 1987, 1988).

Mat-forming cyanobacteria that excrete EPS produce a matrix that stabilizes the sediment (Fig. 4.15). This is also the case with benthic films of diatoms that grow on intertidal mudflats (Paterson 1989; Stal 1994; Yallop et al. 1994). In the desert, hydrophobic EPS of microbial crusts cause the run-off of water preventing erosion (Mazor et al. 1996; Kidron et al. 1999).

Fig. 4.15
figure 15

Capsular and slime extrapolymeric substances (EPS) and the formation of a microbial mat

Full size image

Uronic acids are important components of EPS because these charged groups interact with sediment particles. Thus, EPS with a large content of uronic acids are more efficient in the stabilization of sediments (Martin 1971; Stal 1994). Sulphated groups and uronic acids contribute to the anionic nature of exopolysaccharides which determines the sticky pro­perties of these molecules. EPS may also contain sulphated sugars. As the uronic acids, sulphate groups are also important for the tertiary structure of the polysaccharide and influence the stability of the microbial mat matrix (Decho 1990). Uronic acids as well as sulphate groups interact with a variety of metals. This may either result in the immobilization of toxic metals or scavenge trace metals that are important nutrients. The uronic acid groups of polysaccharides may be involved in the regulation of calcification. Sulphated polysaccharides are often encountered in algae but rarely in archaea and bacteria, including cyanobacteria (Bertocchi et al. 1990). Nevertheless, sulphated polysaccharides have been found in cyanobacteria (Tease et al. 1991; Ortega-Calvo and Stal 1994) and a more thorough investigation of mat-forming cyanobacteria may reveal that such polysaccharides are more common in this group of organisms as previously thought (Pereira et al. 2009).

The polysaccharide produced by the mat-forming cyano­bacteria fulfils an important function as a matrix for exoenzymes, plasmids and DNA (Decho 1990). Extracellular DNA is protected from DNases in the sediments (Romanowski et al. 1991), and may give rise to natural transformation in these ecosystems (Lorenz and Wackernagel 1990, 1994). Hence, in microbial mats, gene exchange may take place.

A considerable number of functions can be attributed to EPS but it is not clear what controls the formation of this polysaccharide and how this relates to one or more of the possible functions. It may very well be that the production of mucilage by cyanobacteria is the result of unbalanced growth caused by nutrient deficits (Lange 1976). Particularly a shortage or deficiency of nitrogen and sulphur results in the stagnation of protein synthesis while the full photosynthetic capacity remains. Under such conditions cyanobacteria accumulate large amounts of glycogen (Allen and Smith 1969; Lehmann and Wöber 1976). The capacity of the cell to store glycogen is limited and any additional polysaccharide may be excreted as mucilage. Old starved cultures often become viscous as a result of excess mucilage production. In the modern marine stromatolites of Highborne Cay, Bahamas, the maximum EPS production represented 7% of the total CO2 fixation while most the fixed carbon was released as low-molecular weight dissolved organic carbon.

Little is known about the fate of EPS in mats. Some polysaccharides appear to be recalcitrant to microbiological degradation, whereas others are not. EPS that is newly formed in a microbial mat may be transformed rapidly (within 12 h) through the degradation by heterotrophic organisms, particularly by sulphate reducing bacteria (Decho et al. 2005). This is interesting because sulphate reducing bacteria are thought to degrade preferentially low-molecular weight organic compounds such as acetate, lactate and ethanol. The degradation of EPS was incomplete causing the accumulation of a more-refractory remnant polymer that was enriched in nitrogen. Net production of EPS in the Highborne Cay stromatolites was less than 2% of the total inorganic carbon uptake (Decho et al. 2005). A similar model of the origin of different fractions of EPS has been proposed for diatom biofilms (Stal 2010). In this model diatoms were supposed to produce one type of EPS which was enriched in glucose. Degradation and the preferential utilization of the glucose component left an EPS that was relatively depleted in glucose and enriched in uronic acids which was refractory and accumulated in the biofilm.

6 Calcification in Mats and Stromatolites

The biological control over calcium carbonate precipitation in the ocean leads to overproduction. It is estimated that 5 Gt calcium carbonate is annually produced in the ocean of which 3 Gt is removed from the system by incorporation and accumulation in sediments, while the other 2 Gt is dissolved (Milliman 1993). The weathering of rock on the continents causes a continuous runoff of calcium and carbonate into the sea. Therefore the oceans tend to be supersaturated with calcium carbonate. In order to maintain a steady state, the amount of calcium carbonate removed from the oceans must be the same as that entering. However, it is estimated that twice as much calcium is removed from the ocean by calcium carbonate precipitation than is brought in (Milliman 1993). This means that the ocean is not in equilibrium or that, sources and sinks are respectively under- or overestimated. The equilibrium of calcium carbonate in the oceans could be maintained by the dissolution of the excess calcium carbonate. Part of this dissolution is biologically controlled because it acts as a pH buffer for respiratory and fermentative processes. Another part dissolves in the deep sea, which is under saturated with calcium carbonate. Some calcium carbonate leaves the system by sinking as fecal pellets to the ocean floor or by fast burial. Although the surface waters of the ocean are supersaturated with calcium carbonate spontaneous precipitation does not normally occur.

Calcification is responsible for the lithification of microbial mats and is the basis of the formation of stromatolites. In most cases calcification seems under a stringent biological control, but the mechanisms by which living organisms influence the precipitation of calcium carbonate are poorly understood. Whereas the function of calcium carbonate precipitation in many organisms is obvious (e.g. shell or skeleton formation) this is not the case in microorganisms, including algae and cyanobacteria. Calcification in bloom-forming algae such as the coccolithophore Emiliania huxleyi would predominantly serve the production of CO2 for subsequent photosynthetic fixation:

$$ {\rm{2HCO}}_{3}{}^{-}+{\rm{ Ca}}^{2+}\leftrightarrow {\rm{CaCO}}_{3}+{\rm{ CO}}_{2}+{\rm{H}}_{2}\rm{O}$$

The so-called ‘whitings’, clouds of aragonite needles, that often occur in tropical lagoons and that have been considered as inorganic precipitates may be produced by the photo­synthetic activity and CO2 fixation of dense communities of picoplankton that increase carbonate ion (Robbins and Blackwelder 1992). Calcification in microbial mats may serve as a mechanism of producing CO2 or it results from an increase in the concentration of the carbonate ion. Due to the dense phototrophic biomass and high rates of photosynthesis and the alkaline conditions it is likely that mats become depleted in CO2. Another function that has been proposed is the detoxification of intracellular calcium. But whatever the function, calcification can be generally inferred from the changes in the concentration of inorganic carbon and from the low solubility product of calcium carbonate.

Calcium carbonate is rather insoluble. Aragonite, which is often thought to be a product of biological calcium carbonate precipitation, has a solubility product (K sp) of 10−6.19, and the more stable form calcite 10−6.37 (at 25°C and salinity of 35) (Zeebe and Wolf-Gladrow 2001). This means that when the ion activity product (IAP) (molar concentrations multiplied by their activity coefficients) of {Ca2+} and {CO 2−3 } in a solution exceeds 10−6.19 calcium carbonate is saturated, although several-fold supersaturation is normally required for spontaneous precipitation (Arp et al. 2001).

$$\left\{{{Ca}}^{2+}\right\} + \left\{{{CO}}_{3}\right\} \to {{CaCO}}_{3}$$

The concentration of calcium-ion in seawater is about 10−2 M (Ehrlich 1996). When the concentration of calcium ion is assumed to be constant, then the carbonate concentration determines calcification. CO2 reacts with water to form bicarbonate and this dissociates according to the following reversible reactions:

$$ {\rm{CO}}_{2}+{\rm{H}}_{2}\rm{O}\leftrightarrow {\rm{H}}^{+}+{\rm{ HCO}}_{3}{}^{-}$$
$$ {\rm{HCO}}_{3}{}^{-}\leftrightarrow {\rm{H}}^{+}+{\rm{ CO}}_{3}{}^{2-}$$

Typical physicochemical processes that cause calcification are degassing of CO2 and evaporation of water due to solar radiation. Evaporation leads to the formation of brines in which minerals precipitate, of which calcium carbonate is only a minor component (Yechieli and Wood 2002). In hot spring microbial mats degassing of CO2 shifts the equilibrium towards the carbonate ion and to the formation of travertine deposits (Fouke et al. 2000).

In microbial mats, a number of biological processes influence the equilibriums of carboxy species and hence may control calcification (Dupraz et al. 2009). These include specific metabolic processes in which CO2 is consumed such as photosynthesis, chemosynthesis and, less importantly, heterotrophic CO2 fixation. Metabolisms in which CO2 is produced such as respiration and fermentation may cause an acidification of the medium and eventually result in dissolution of calcium carbonate rather than precipitation. Nevertheless, Krumbein (1974) demonstrated the formation of aragonite on the surface of marine bacteria as the result of their metabolism of substrates such as glucose, sodium acetate and sodium lactate. However, this also strongly depends on the environmental conditions that apply (Canfield and Raiswell 1991). A variety of metabolic processes influence the equilibrium of inorganic carbon by the production of acids and bases.

Photosynthesis is an important process in the vast majority of microbial mats, at least in those occurring in illuminated environments. Because photosynthesis normally involves CO2 fixation it has often been considered important for calcification (Golubić 1973; Krumbein and Giele 1979; Pentecost 1988). The fixation of CO2 from a bicarbonate solution will result in an increase in carbonate ion:

$$ {\rm{2HCO}}_{3}{}^{-}\leftrightarrow {\rm{CO}}_{2}+{\rm{H}}_{2}\rm{O }+{\rm{ CO}}_{3}{}^{2-}$$

Other important metabolic processes in microbial mats which remove CO2 are chemosynthesis and heterotrophic CO2 fixation. Organisms that carry out chemosynthesis include colourless sulphide oxidizing bacteria, nitrifying bacteria, autotrophic sulphate-reducing bacteria as well as methanogenic- and acetogenic bacteria. Heterotrophic CO2 fixation occurs in virtually all organisms but is limited and negligible compared to the CO2 produced during the oxidation of organic compounds. This process is therefore not important for calcification. Many reports mention calcification as a result of photosynthesis by cyanobacteria (e.g. Golubić 1973; Krumbein and Giele 1979; Pentecost 1988; Pentecost and Bauld 1988). Calcification associated with anoxygenic photosynthesis or anoxygenic phototrophic bacteria has only been reported in one case in which the purple non-sulphur bacterium Rhodopseudomonas palustris was shown to stimulate calcification in a solution that was oversaturated with calcium carbonate (Bosak et al. 2007). Phototrophic sulphur bacteria produce sulphuric acid and they will therefore cause calcium carbonate dissolution rather than its precipitation although anaerobic sulphide oxidation has been reported to be involved in calcium carbonate precipitation in a marine stromatolite (Visscher et al. 1998). The same holds true for most chemosynthetic metabolisms and for heterotrophic CO2 fixation. Precipitation of calcium carbonate may be indirectly associated with oxygenic photosynthesis and due to an increase of pH and/or a shift in the equilibrium of inorganic carbon (Golubić 1973; Krumbein and Cohen 1977). This was elegantly demonstrated in a hypersaline microbial mat (Ludwig et al. 2005). These authors showed that calcification was solely due to the increase of the carbonate ion as the result of photosynthesis and that the changes in the activity of calcium were not important. The contribution of heterotrophic bacteria was indirect as these organisms kept the concentration of dissolved inorganic carbon high in the pore water. Sulphate reducing bacteria did not change the pH and their effect was solely maintaining high concentrations of dissolved inorganic carbon. On the other hand, Chafetz and Buczynski (1992) found that calcification in stromatolithic microbial mats was associated with heterotrophic bacteria rather than with the cyanobacteria and these seemingly contradictory observations emphasize the complexity of the process and the fact that the actual environmental conditions may cause quit different outcomes.

Aerobic or anaerobic oxidation of organic compounds results in the production of CO2 and/or HCO −3 and affects pH and consequently causes a shift in the equilibrium of inorganic carbon. Organic matter possessing the “Redfield” stoichiometry of C:N:P of 106:16:1 is oxidized by O2, NO −3 and SO 2−4 according to the following reactions (Boudreau and Canfield 1993):

$$[\begin{array}{lll}{\left({{CH}}_{2}{O}\right)}_{106}{\left({{NH}}_{3}\right)}_{16}{{H}}_{3}{{PO}}_{4}+138{{O}}_{2}\\ \to 122{{H}}_{2}{O}+10{{6 CO}}_{2}+{{16 NO}}_{3}+16{{H}}^{+}+{{H}}_{3}{{PO}}_{4}\\ \end{array} $$
$$[\begin{array}{lll}{\left({{CH}}_{2}{O}\right)}_{106}{\left({{NH}}_{3}\right)}_{16}{{H}}_{3}{{PO}}_{4}+94.{{4NO}}_{3}\\ \to 84.8{{H}}_{2}{O}+13.{{6 CO}}_{2}+92.{{4 HCO}}_{3}+55.2{{N}}_{2}+{{H}}_{3}{{PO}}_{4}\\ {}\end{array}$$
$$\begin{array}{lll}{\left({{CH}}_{2}{O}\right)}_{106}{\left({{NH}}_{3}\right)}_{16}{{H}}_{3}{{PO}}_{4}+{{53SO}}_{4}\\ \to 10{{6 HCO}}_{3}+53{{H}}_{2}{S}+{{16 NH}}_{3}+{{H}}_{3}{{PO}}_{4}\\ {}\end{array}$$

The formation of CO2 and the acidification of the medium could result in the dissolution of calcium carbonate rather than cause its precipitation. Anaerobic respiration results in the formation of bicarbonate and could give rise to supersaturation of calcium carbonate. The effects of the sequential oxidation of organic matter by the three electron acceptors oxygen, nitrate and sulphate on pore water pH and calcium carbonate saturation are complex and depend on the prevailing conditions (Boudreau and Canfield 1993).

In the microbial mats of Solar Lake (Sinai), it has been shown that sulphate reduction and CaCO3 formation were stoichiometrically related and organic carbon was transformed into a number of different carbonate minerals (Jørgensen and Cohen 1977; Krumbein and Cohen 1977; Krumbein et al. 1977). However, whether sulphate reduction in microbial mats in reality results in calcium carbonate precipitation depends largely on a variety of conditions that prevail in these microbial mats. Most important is the development of alkaline conditions, the removal of excess CO2 or the presence of a suitable buffer (Ehrlich 1996). The precipitation of sulphide as iron sulphide acts as a pH buffer. In the absence of iron, sulphate reduction produces equal amounts of H+ and HCO −3 , which will thus cause a decrease of carbonate saturation. In many microbial mats high rates of sulphate reduction occur but despite this, calcification is absent. In the modern stromatolites of the Exuma Cays the tightly associated sulphate reduction and anaerobic sulphide oxidation promoted calcification, while the couple oxygenic photosynthesis and aerobic respiration cause calcium carbonate dissolution (Visscher et al. 1998).

Modern microbial mats are often considered as the structural analogues of Precambrian stromatolites. By definition stromatolites are lithified laminated formations. Precambrian stromatolites were formed in shallow marine areas. Lithification of present day coastal microbial mats is extremely rare and it is still an enigma why this should be so. Kempe and Kazmierczak (1990a, b) and Kazmierczak et al. (1996) investigated stromatolites in the sea-linked Satonda Crater Lake in Indonesia and alkaline Lake Van in Turkey, both formed under extreme alkaline conditions. They hypo­thesized that the greater abundance of stromatolites during the Precambrian should be attributed to the much greater alkalinity of the marine environment during that era (Kempe and Kazmierczak 1990a). The hypothesis of a Precambrian soda ocean may certainly offer an explanation for the greater abundance of stromatolites and the discovery of modern calcifying stromatolites in alkaline seas supports this. Nevertheless, calcification in these stromatolites is still under biological control rather than being a spontaneous occurrence. Moreover, other recent stromatolites are formed under less alkaline or normal marine conditions such as those found in the French Polynesian atolls (Défarge et al. 1994a, b) or in the Bahamas (Reid and Browne 1991). Even if the early oceans were more alkaline, the marine environment today is still supersaturated with calcium carbonate. Furthermore, in microbial mats several biological processes predominate which presumably increase the concentration of carbonate ion, which theoretically should lead to calcium carbonate precipitation. As a result of active photosynthesis and CO2 fixation in the top layer of cyanobacterial mats the pH in these mats may reach values as high as 9.5 (Fig. 4.3) (Revsbech et al. 1983). Although these conditions would normally promote calcification, this does not happen in most marine microbial mats.

There is also abundant evidence of non-lithifying Proterozoic microbial mats that are recognized as microbially induced sediment structures (MISS), suggesting that the past may not have been so different from the present (Noffke et al. 2006; Noffke 2009). An experiment in which a non-lithifying mat was transplanted into an environment with lithifying microbial mats showed that it was now capable of calcification (Dupraz et al. 2009). Kremer et al. (2008) showed abundant calcification in an otherwise typically non-lithifying coastal microbial mat. This indicates that calcification is probably not so unusual and that the absence of lithification of these mats may be due to a subsequent dissolution of the calcium carbonate.

Hence, spontaneous calcification does not seem to be important in stromatolites. Biological control of calcification may not only exist in the change of the carbonate ion concentration and equilibrium but also in a mechanism that inhibits calcification (anti-calcification) (Westbroek et al. 1994). Biologically controlled calcification must distinguish between supersaturation of calcium carbonate in a solution (which is the thermodynamic force) and those factors that influence the kinetics of the process. The latter may be either inhibitory or stimulatory factors. Supersaturation of calcium carbonate in the ocean is the primary driving force of calcification that can be dramatically increased in the immediate vicinity of phototrophic organisms. Because uncontrolled calcification in or around organisms is unwanted it may be clear that some mechanism must exist that can inhibit the process. Crystal poisons such as Mg2+, SO 2−4 , PO 3−4 that complex with CO 2−3 and Ca2+, respectively, are not sufficient and additional mechanisms must be postulated. It is known that some small acidic molecules may inhibit crystallization. An example is the binding of Ca2+ to oxalate (Verrecchia et al. 1990). Acidic polysaccharides are also very effective in binding Ca2+ or interact with it (Dupraz and Visscher 2005; Braissant et al. 2009). These interactions will doubtless influence calcification. Figure 4.16 depicts the way in which such polysaccharides could influence crystallization or crystal growth (Westbroek et al. 1994). The association of a polyanion with Ca2+ ions will lower the activity of the latter below the saturation of calcium carbonate and prevent subsequent crystallization. Likewise, polyanions may associate with a growing calcium carbonate crystal and prevent its further growth. A layer of charged polymers may bind calcium carbonate crystal and arrest its growth. The structure of such polymers may determine crystal shape. Evidence has been obtained that this mechanism is involved in the formation and morphology of coccoliths in the coccolithophore Emiliania huxleyi (Borman et al. 1982, 1987).

Fig. 4.16
figure 16

Simplified model of the possible interactions of charged extracellular polymeric substances with calcium carbonate: (a) nucleation of calcium (Ca2+) and carbonate (CO 2−3 ) ions; (b) Inhibition of nucleation by a polyanion; (c) Inhibition of crystal growth by association of a crystallization nucleus with a polyanion; (d) Calcium carbonate crystal bound to a layer of charged polymers. The growth of the crystal may be arrested and the charged polymer may determine the eventual shape of the calcium carbonate crystal

Full size image

In microbial mats it is hypothesized that extracellular polymeric substances (EPS) which are mainly composed of polysaccharides serve as agents that inhibit calcification. EPS produced by cyanobacteria are often rich in uronic acids and contain other acidic groups such as pyruvate, succinate, sulphate and phosphate groups and hence are negatively charged polyanions (Bertocchi et al. 1990; De Philippis et al. 2001; Sutherland 2001). Many microbial mats are composed of vast amounts of EPS in which the cyanobacteria and other organisms are embedded. It is possible that this EPS acts as an anti-calcification agent. Cyanobacterial EPS may bind 55–183 mg Ca g−1 EPS (Li et al. 2001; Ortega-Morales et al. 2006). When heterotrophic bacteria decompose this EPS, high concentrations of calcium carbonate may exist locally, leading to precipitation (Dupraz and Visscher 2005). In some non-lithifying microbial mats such as in Solar Lake (Sinai), aragonite needles are formed in the deeper layers of the mat, where the organic matter is subject to degradation. Several others have observed the association of calcification with bacterial activity (Chafetz and Buczynski 1992; Krumbein 1979; Krumbein and Giele 1979).

Mucilage EPS is often produced by cyanobacteria as an overflow metabolism when they experience nutrient limitation. This is particularly the case under nitrogen depleted conditions, a situation common in the marine environment. In microbial mats, where extremely dense communities of cyanobacteria are present there is a high demand for nitrogen, while there is often a shortage of this important nutrient. Therefore, many microbial mats are diazotrophic, i.e. the cyanobacteria that build these mats fix atmospheric dinitrogen. However, most diazotrophic mats consist of non-heterocystous cyanobacteria. As argued in the section on N2 fixation these cyanobacteria are not efficient N2 fixers because the process is seriously hindered by oxygen. It is likely that such cyanobacteria in fact still are nitrogen limited. Cyanobacteria that grow under nitrogen limited conditions tend to produce a lot of mucilage (Ortega-Calvo and Stal 1994). It is possible that Precambrian calcifying microbial mats were not nitrogen-limited and that the growth of the organisms therefore might have been balanced with less overflow metabolism and mucilage production. This might also hold true for modern calcifying microbial mats. While some of these mats may rely on a sufficient external supply of combined nitrogen, others comprise heterocystous N2-fixing cyanobacteria that satisfy their nitrogen demand. One example from freshwater environments is the Rivulariaceae. This group of heterocystous cyanobacteria produces extent calcium carbonate formations (Whitton 1987). The marine cyanobacterium Calothrix spp. belongs also to this taxonomic group and is known in some cases to form microbial mats but usually does not calcify.

A model for calcification and the development of stromatolites in the Exuma Cays, Bahamas is presented in Fig. 4.16. Subtidal and intertidal stromatolites that can be found in the Exuma Cays (Bahamas) are characterized by mats of the cyanobacterium Schizothrix sp. The model for calcification in these mats is based on a number of observations and assumptions. Two types of mats of Schizothrix can be distinguished. Lithifying microbial mats of Schizothrix are usually characterized by low ratios of photosynthesis over respiration while the opposite is true non-lithifying mats (Pinckney et al. 1995). Moreover, calcium carbonate was not associated with the cyanobacteria but with heterotrophic bacteria (Chafetz and Buczynski 1992; Chafetz 1994). It was further assumed that extracellular polymeric substances (EPS) may interfere by binding calcium and magnesium ions, thus locally inhibiting carbonate precipitation (Borman et al. 1982, 1987; Westbroek et al. 1994).

Exuma Cays stromatolites are formed at high energy sites. Intertidal stromatolites can be found on the Atlantic Ocean coast and are exposed to high wave energy. Subtidal stromatolites are almost exclusively encountered in channels with high currents. The cyanobacterium Schizothrix sp. is a filamentous organism composed of thin (often less than 1 μm wide) trichomes which are enveloped by a thin rigid polysaccharide sheath. This organism is capable of colonizing a solid substrate. Because grazing pressure will be low in these high-energy areas, a community of Schizothrix sp. may develop. Under conditions of low sedimentation rate a mat of Schizothrix sp. and associated microorganisms will develop (Fig. 4.17a). These mats are rigid and tightly associated with the underlying substrate. The cyanobacteria will grow and produce sheath material and possibly some mucus. It is assumed that this EPS will bind Ca2+ or that uronic acids prevent further growth of crystallization nuclei (Borman et al. 1982). During a period of sedimentation, Schizothrix sp. will move rapidly upwards by phototaxis and continue growth in the top layers where optimum light conditions prevail. The trichomes form a dense network in which carbonate sand grains (ooids) are trapped and agglutinated by EPS, while Ca2+ is further bound. Empty sheaths and other organic matter that has been abandoned deeper in the sediment will be decomposed (Fig. 4.17b). During a subsequent period of low rates of sedimentation a dense mat of Schizothrix sp. will develop in the top layer of the sediment. This layer is characterized by active growth of the cyanobacteria and may be associated with abundant production of EPS. The matrix of EPS in which the mat is embedded may bind Ca2+ efficiently and condenses EPS to the gel state (Rees 1969; Decho and Moriarty 1990; Decho 1994). It is conceived that this will lower the activity of this ion so that calcium carbonate will not precipitate. Depending on the physicochemical gradients that typically develop in microbial mats due to phototrophic and heterotrophic activities, some dissolution and re-precipitation of CaCO3 and re-crystallization of the carbonate sediment grains may occur (Fig. 4.17c). During the next stage of development, Schizothrix sp. moves upward after another period of high rate of sedimentation. While growth of the cyanobacterium and the production of EPS in the new top layer trap and agglutinate the carbonate sand, the large amount organic matter that was left behind is decomposed by heterotrophic bacteria. Because EPS is also decomposed, Ca2+ that was bound will be released (Decho et al. 2005). This will locally result in supersaturation of calcium carbonate resulting in the formation of a microcrystalline crust of precipitated carbonate (Fig. 4.17d). A similar type of bacterial calcification occurs during the degradation of calcium oxalate. Oxalate is a product of metabolism of fungi and other organisms and is capable of immobilizing calcium (Verrecchia et al. 1990). In the Exuma Cay stromatolites another type of biologically influenced calcification occurs. The unicellular cyanobacterium Solentia bores in the calcium carbonate grains, dissolving them partly and the subsequent precipitation fuses the ooids, forming a solid lithified structure (Reid et al. 2000). This model explains that in situ calcium carbonate precipitation and lithification of the mat is controlled by biology. It is indirectly associated with the cyanobacteria but degradation of organic matter, notably EPS, by heterotrophic bacteria is required for this process. Alternating periods with high and low rates of sedimentation are responsible for the formation of the laminated structure of the lithified microbial mats, which could therefore be termed stromatolites.

Fig. 4.17
figure 17

Simplified scheme of the development of lithified micritic layers in Bahamas stromatolites. For explanation, see text

Full size image

7 Nitrogen Metabolism and Nitrogen Fixation

7.1 Introduction

In cyanobacteria nitrogen content may amount up to about 10% of dry weight and is quantitatively the third most abundant element. Any shortage of it will immediately affect the amount of phycobiliproteins and, consequently, the efficiency of light harvesting for photosynthesis (Allen and Smith 1969). Cyanobacteria may produce a unique nitrogenous compound known as cyanophycin or multi-L-arginyl-poly(L-aspartic acid). Its high nitrogen content means that it can serve as a nitrogen reservoir (Mackerras et al. 1990a, b).

Cyanobacteria use a variety of nitrogen sources (Flores and Herrero 1994). Ammonia can be taken up by passive diffusion or the protonated form ammonium (NH +4 ) by a specific uptake system (Fig. 4.18). The amino acids arginine, asparagine and glutamine have been reported to serve as nitrogen sources in cyanobacteria (Flores and Herrero 1994). Nitrate and nitrite are important sources of nitrogen for cyanobacteria. This involves the uptake of nitrate or nitrite and its subsequent reduction to ammonia. This process involves ferredoxin as an electron donor and is therefore intimately associated with photosynthesis. Urea appears also to be a common nitrogen source for cyanobacteria (Moore et al. 2002; Valladares et al. 2002). Many cyanobacteria are capable of using dinitrogen (N2) as the source of nitrogen.

Fig. 4.18
figure 18

The nitrogen cycle in a cyanobacterial mat. Cyanobacteria take up and assimilate ammonium into amino acids (AA) which are used for protein synthesis or can be stored as cyanophycin. Amino acids and ammonia may leak out the cells and oxidized to nitrate (nitrite) by nitrifying bacteria. In the anoxic part of the mat ammonium and nitrate (nitrite) is converted to dinitrogen by anammox and dinitrifying bacteria. Nitrate (nitrite) can be taken up by the cyanobacteria and assimilated. N2-fixing cyanobacteria produce ammonium which is subsequently assimilated into cell material

Full size image

7.2 The Nitrogen Cycle

Nitrogen occurs in different chemical oxidation states, varying from its most reduced form ammonia (NH3) (−3), to hydroxylamine (NH2OH) (−1), dinitrogen (N2) (0), nitrous oxide (N2O) (+1), nitric oxide (NO) (+2), nitrite (NO −2 ) (+3) to its most oxidized form nitrate (NO −3 ) (+5). All of these oxidation states are biologically significant and microorganisms may carry out reduction and oxidation reactions transforming one form into another. The element nitrogen therefore is subject to microbiological cycling in nature. In microbial mats all steps of the nitrogen cycle may be present and cyanobacteria play a particular important role (Fig. 4.17).

Ammonia is assimilated into amino acids that are used for the synthesis of proteins. Luxury uptake of nitrogen may occur and be stored as cyanophycin. Ammonia and amino acids may leak out of the cell. When oxygen is present, ammonium may be oxidized via nitrite to nitrate by nitrifying bacteria. Nitrate may be taken up by the cyanobacteria and assimilated or under anoxic conditions converted to gaseous nitrogen by denitrifying bacteria. Anaerobic ammonium oxidation (anammox) is another process that leads to the conversion of fixed nitrogen to N2 (Jaeschke et al. 2009; Porubsky et al. 2009). Hence, these processes cause a loss of combined nitrogen, which is counteracted by the capacity of some cyano­bacteria to fix dinitrogen (Joye and Paerl 1994). Nitrogen cycling in microbial mats contributes to the nutrient limitation patterns of mangrove trees. In dwarf habitats, microbial mats serve as a source of nitrogen via the fixation of dinitrogen, while in fringe and transition habitats, mats compete with the trees for nitrogen via denitrification (Lee and Joye 2006).

In microbial mats the decomposition of organic matter may be incomplete which could mean that part of the nitrogen is not recycled and so enters the geological record (Fig. 4.17). Hence this nitrogen is withdrawn from the microbial nitrogen cycle. It is not clear how important this process is since even in mats in which a net accretion of organic matter occurs, up to 99% of the produced organic matter may be recycled within the mat (Krumbein et al. 1977). In exceptional cases organic nitrogen produced by N2-fixing cyanobacteria can be transformed to nitrate deposits known as guano or caliche nitrates (Ehrlich 1996).

Most of the nitrogen in the biosphere is present in the atmosphere in the form of dinitrogen (N2), which amounts to 3.9  ×  1018 kg N. In the oceans and on land the amount of combined nitrogen (organic and inorganic) amounts each about 1015 kg N. The amount of nitrogen in living biomass on earth amounts only 1.3  ×  1013 kg (Ehrlich 1996). It is generally assumed that primary production in the marine environment is limited by nitrogen. In particular marine microbial mats with their dense and compressed biomass often experience a shortage of nitrogen. The majority of organisms cannot use the most abundant form of nitrogen, N2. Only N2-fixing organisms (diazotrophs) are capable of using dinitrogen. All these organisms possess nitrogenase. Cyanobacteria are among the most important diazotrophs and in all marine microbial mats that have been investigated to date, N2 fixation has been observed.

7.3 Nitrogenase

In all N2-fixing organisms the enzyme complex nitrogenase is present. This enzyme is similar in all organisms that contain it. The complex is composed of two enzymes: dinitrogenase reductase which is a dimer of identical subunits and also termed the iron-protein, which is encoded by nifH, and dinitrogenase, a tetramer composed of two different subunits (α2 β2), encoded by nifDK. Dinitrogenase is also known as the molybdenum-iron protein (Howard and Rees 1994).

Nitrogenase catalyzes the following reaction:

$$ \begin{array}{l}{{N}}_{2}+8{{e}}^{-}+8{{H}}^{+}+{16 MgATP}\\ \to {{2 NH}}_{3}+{{H}}_{2}+{16 MgADP}+16{{P}}_{{i}}\\ {}\end{array}$$

Reduced ferredoxin is the electron donor of nitrogenase. The equation shown above makes clear that the fixation of N2 is at the expense of considerable amount of energy and low potential electrons. This high energy demand of nitrogenase presents often a problem for diazotrophic organisms except for cyanobacteria which produce reduced ferredoxin and convert light energy into chemical energy during photosynthesis. However, nitrogenase is extremely sensitive to oxygen and therefore diazotrophic organisms must provide an anaerobic environment in order to be able to fix N2. Cyanobacteria as oxygenic phototrophic and principally aerobic organisms need special adaptations.

7.4 Dinitrogen-Fixing Cyanobacteria

Diazotrophic cyanobacteria are capable of using dinitrogen (N2) as the sole source of nitrogen for growth. These organisms can be subdivided in three main groups (Table 4.3).

Table 4.3 Types and characteristics of N2-fixing cyanobacteria
Full size table

Group I consists of heterocystous cyanobacteria. These filamentous organisms differentiate special cells, heterocysts, which have lost the capacity of oxygenic photosynthesis and have evolved a modified thick cell envelope. Heterocysts are the site of N2 fixation in these cyanobacteria. The thick cell wall contains special lipopolysaccharides and forms a diffusion barrier for gases, limiting the entry of oxygen. Respiration scavenges the little oxygen that enters the heterocyst. Since photosystem II is absent from the heterocyst, no photosynthetic oxygen is evolved by these cells. Therefore the heterocyst is virtually anoxic and provides an excellent environment for the oxygen-sensitive nitrogenase. Photosystem I-mediated conversion of light energy in the heterocyst provides nitrogenase with ATP. However, for reducing equivalents nitrogenase depends on the import of carbohydrates from the neighbouring vegetative cells. The strategy that heterocystous cyanobacteria have developed in order to be able to grow diazotrophically can be best described as the spatial separation of the two incompatible processes of N2 fixation and oxygenic photosynthesis. Among the cyanobacteria heterocystous species are the ultimate adapted organisms for N2 fixation. The vast majority of heterocystous cyanobacteria can be found in freshwater or terrestrial systems, both free-living and as symbionts.

Heterocystous cyanobacteria occur in some brackish basins but are rare in the marine environment, including microbial mats. Fischerella and Mastigocladus form mats in thermal springs. The heterocystous Calothrix sp. has been found as the dominant organism in microbial mats in the tidal area of the Pacific coast in Baja California Sur, Mexico. Calothrix is also known from a variety of other marine and brackish habitats such as the spray zone of rocky shores (Jones and Stewart 1969; Whitton and Potts 1982). Mats of the heterocystous cyanobacterium Anabaena have been found in a coastal lagoon in southwest France (Villbrandt and Stal 1996) and Nodularia occurs in coastal microbial mats of the Dutch barrier islands (Severin and Stal 2008). However, these are exceptions rather than a rule. The vast majority of microbial mats are built by non-heterocystous cyanobacteria, notwithstanding the facts that in many cases N2 fixation is a crucial process in these systems.

Group II consists of filamentous and unicellular cyano­bacteria that do not show cell differentiation and are capable of N2 fixation only under virtually anoxic conditions with no oxygenic photosynthesis occurring. These organisms, although possessing the genetic capacity of synthesizing nitrogenase, have obviously not evolved a mechanism to protect effectively nitrogenase from oxygen inactivation. Consequently, their strategy can be characterized as avoidance of oxygenated environments. Such environments usually also prevent oxygenic photosynthesis. Among the non-heterocystous cyanobacteria up to about 50% may belong to this group of organisms but for virtually all of them it is uncertain whether they live diazotrophically in their natural environment. Non-heterocystous cyanobacteria that are capable of inducing nitrogenase activity under anaerobic conditions can be found in many environments, including microbial mats. However, most environments in which these cyanobacteria occur are permanently oxygenated and therefore diazotrophic growth is unlikely. In contrast, microbial mats are often characterized by steep and fluctuating gradients of oxygen and sulphide. Anoxia frequently occurs in microbial mats; this as a rule coincides with high levels of sulphide, a very potent inhibitor of oxygenic photosynthesis. Thus, it is not surprising that evidence hinted to anaerobic N2-fixing cyanobacteria growing diazotrophically in microbial mats in which H2S was present (Villbrandt and Stal 1996).

Group III cyanobacteria also comprise non-heterocystous filamentous and unicellular cyanobacteria but they are remarkable as they possess the capacity of inducing nitrogenase and growing diazotrophically under fully aerobic conditions. To date our knowledge of how these organisms are protecting their undoubtedly oxygen-sensitive nitrogenase is incomplete. Although the number of species that we know that possess this capability is still relatively rare, their numbers are increasing at steady pace. Examples can be found in terrestrial environments such as cave-walls, paddy fields and thermal springs, and in the marine environment. Freshwater lakes apparently do not harbour aerobic N2-fixing non-heterocystous cyanobacteria. In the ocean the planktonic colony-forming Trichodesmium spp. is known as an efficient diazotrophic growing, non-heterocystous cyanobacterium. In addition, several unicellular diazotrophic cyanobacteria occur in the oceans. These include Crocosphaera and Cyanothece (Needoba et al. 2007). The pico-sized and hitherto uncultivated ‘Group A’ cyanobacteria are abundant and metagenomic analyses suggest that these organisms may lack the oxygenic photosystem II (Zehr et al. 2008). Their mode of life is unclear and could either be photoheterotrophic or symbiotic. Strikingly, diazotrophic cyanobacteria in the oceans are confined to the tropical and subtropical regions with surface water temperature well above 20°C. In microbial mats aerobic N2-fixing non-heterocystous cyanobacteria are reported to belong predominantly to the genera Oscillatoria and Lyngbya, which are morphologically and phylogenetically closely related to Trichodesmium. However, in a variety of environments unicellular N2-fixing cyanobacteria such as Cyanothece, Gloeothece and Synechococcus are known to build microbial mats.

The first report of a culture of a filamentous non-heterocystous aerobic N2-fixing cyanobacterium was by Pearson et al. (1979). This organism was originally identified as Microcoleus chthonoplastes but later re-named as Symploca sp. (Janson et al. 1998). Symploca sp. is also morphologically related to Oscillatoria, and was isolated from a tidal microbial mat (Pearson et al. 1979; Malin and Pearson 1988). It has been proposed that the strategy of aerobic non-heterocystous cyanobacteria, in analogy with the heterocystous species, is temporal separation of the incompatible processes of photosynthesis and N2 fixation. The latter would than occur during the dark (Mullineaux et al. 1981; Stal and Krumbein 1987). However, not all species in this group follow this strategy. Trichodesmium spp. fix N2 during the day (Capone et al. 1990). Moreover, all species that have been cultured are capable of growing diazotro­phically under continuous light and, in the unicellular Gloeothece sp., culture conditions can be chosen under which N2 fixation occurs during the light period of a light dark cycle (Ortega-Calvo and Stal 1991).

7.5 Daily Variation of N2 Fixation

N2 fixation has a high demand of energy and low-potential reducing equivalents. For the oxygenic phototrophic cyano­bacteria light is the source of ATP generation and electrons are derived from water and transferred to ferredoxin mediated by photosynthetic electron transport. Thus, in cyanobacterial mats N2 fixation ought to be directly linked to light. However, since oxygen exerts a negative effect on nitrogenase, daily variations of N2 fixation in microbial mats can be expected. The patterns of these daily variations will depend on the type of diazotrophic cyanobacterium and on the dynamics of light and oxygen in the mat. In Fig. 4.19 five daily patterns of nitrogenase activity, measured in different microbial mats, are depicted.

Fig. 4.19
figure 19

Five typical patterns of daily variation of N 2 fixation (ARA, relative units) in microbial mats: (a) Mat of the heterocystous Calothrix sp. in Baja California, Mexico (Data from Stal et al. 1994); (b) Mat of the unicellular Gloeothece sp. on the wall of a carbonate cave (Data from Griffiths et al. 1987); (c) Mixed mat of the non-heterocystous Microcoleus chthonoplastes and Lyngbya sp. of a tidal flat of the North Sea island of Mellum (Data from Villbrandt et al. 1990); (d) Mat dominated by Lyngbya sp. (location as c) (Data from Villbrandt et al. 1990); (e) Mat of non-heterocystous Lyngbya aestuarii (location as a) (Author, unpublished)

Full size image

N2 fixation in heterocystous cyanobacteria is intimately linked to light. The heterocyst is not capable of CO2 fixation and therefore does not accumulate storage carbohydrate, as is the case in vegetative cells. Dark energy generation in heterocysts will be limited because at one hand the reducing equivalents must be imported from the vegetative cells while at the other hand the oxygen entry in the heterocyst is limited as a result of the diffusion barrier provided by the cell wall (Walsby 1985). Therefore it is not surprising that daily variations of N2 fixation in communities of heterocystous cyanobacteria are strongly light dependent (Griffiths et al. 1987; Storch et al. 1990; Stal 1995; Falcón et al. 2007) (Fig. 4.18a). However, considerable dark nitrogenase activity may occur in such communities. The ratio of light over dark nitrogenase activity in different populations of heterocystous cyanobacteria varies considerably and is possibly dependent on the species, the light history or other conditions.

In microbial mat communities composed of non-heterocystous cyanobacteria the daily pattern of N2 fixation is less predictable (Paerl et al. 1989, 1996) (Fig. 4.18b–e). This depends largely on the type of organism and prevailing conditions in the mat. Moreover, due to the fact that these conditions may also vary from day to day (tidal movement, light and overcast, temperature and other factors), the daily pattern of N2 fixation may change considerably.

The daily pattern of N2 fixation in non-heterocystous cyanobacteria is the result of the combined effects of oxygen, light, and in some cases sulphide. As in heterocystous cyano­bacteria, non-heterocystous species must supply nitrogenase with sufficient energy and low-potential reducing equivalents. This condition is satisfied in the light but the serious drawback is the evolution of oxygen. In such mats, photosynthesis obviously must occur at daytime and N2 fixation is confined to the night (Fig. 4.18b). For instance this is the case in mats of Gloeothece and Oscillatoria. Whereas during the day microbial mats often become supersaturated with oxygen because the diffusion of this gas is limited, at night they may turn anoxic within minutes (Stal 1995). The microbial community, including the cyanobacteria, consumes oxygen in the dark by respiration. Obviously, anoxic conditions are ideal for N2 fixation, but pose a problem with respect to the supply of energy and reducing equivalents. However, all cyanobacteria isolated from marine microbial mats and tested appeared to be capable of fermentation of endogenous storage carbohydrate (Stal and Moezelaar 1997). Although the energy generation by fermentation is undoubtedly small, it has been shown that it exceeds many times the extremely low maintenance requirements of cyanobacteria (Stal and Moezelaar 1997). It has also been shown that dark anoxic conditions supported considerable nitrogenase activity in the filamentous, non-heterocystous cyanobacterium Oscillatoria limosa (Lyngbya aestuarii) (Stal and Heyer 1987). In microbial mats in which this cyanobacterium occurred, daily patterns of N2 fixation were found in which this activity was low but totally confined to the dark (Villbrandt et al. 1990). However, other patterns were also observed at different times in the same mats with the same organism. For instance, it could often be seen that nitrogenase activity peaked around sunset and sunrise (Fig. 4.18c). This was confirmed by experiments with O. limosa grown in the laboratory under an alternating light dark cycle and with anoxic conditions established 1 h after the onset of the dark period and aerobic conditions 1 h after the onset of the light period (Stal and Heyer 1987). Highest nitrogenase activities in these cultures were obtained in the light in the absence of oxygen. Also, in natural samples it has been observed that highest nitrogenase activities occurred at sunrise (Villbrandt et al. 1990) (Fig. 4.18d). Exactly the same observation was done for the hot spring unicellular cyanobacterium Synechococcus (Steunou et al. 2008). NifH was transcribed at the end of the day but the nitrogenase was present during the whole night and dis­appeared once the mat became enriched by oxygen the next day. Nevertheless, during most of the night nitrogenase activity was low, and revealed a small peak at sunset and a large peak at sunrise. This is because light is available while oxygen is still absent. After sunset, oxygen may have been present for some time, allowing energy generation through aerobic respiration. Once anoxic conditions are established only low rates of nitrogenase activity can be supported by the lower rate of fermentative energy generation. Vertical profiles of oxygen measured during a 24 h period have shown that the mat which possessed this type of fluctuating nitrogenase activity, oxygen was indeed present during the first hours after sunset and that it appeared again in the morning only hours after sunrise.

In freshly colonized sediments of North Sea tidal sand flats, O. limosa (L. aestuarii) is often the pioneer cyano­bacterium (Stal et al. 1985). This is most likely because of its capacity to grow diazotrophically. In this pioneer state biomass is low and therefore so is the oxygen demand in the dark. Such sediments normally do not turn anoxic. However, during the light they may accumulate oxygen up to several fold saturation (Villbrandt et al. 1990). N2 fixation in such systems is typically confined to the night, peaking at sunrise when light becomes available but at oxygen levels well below air saturation (Fig. 4.18d). In other systems such as in mats of the unicellular N2-fixing Gloeothece ,which grows on carbonate cave walls, a peak of nitrogenase activity is observed immediately after sunset and then decreasing gradually until hardly any activity is detectable at the end of the night (Fig. 4.18b). This organism depends on oxygen for respiratory energy generation and it is possible that in the course of the dark period this organism depletes its endogenous storage carbohydrate (Maryan et al. 1986).

Another type of daily pattern of N2 fixation in microbial mats of non-heterocystous cyanobacteria is more or less constant activity or fluctuations scattered throughout the day and night (Fig. 4.18e). This is often the case when Group 2 diazotrophic cyanobacteria are involved. These cyanobacteria are only capable of fixing N2 under anoxic conditions or at least when oxygen concentrations are low and oxygenic photosynthesis is inhibited. Such a situation can be expected in microbial mats in which high concentrations of sulphide inhibit oxygenic photosynthesis. In the light, sulphide at the same time may serve as electron donor for nitrogenase in these situations. In most cases oxygenic photosynthesis is continuing in the surface layers of the mat. Sulphide in inhibitory concentrations for photosynthesis is usually present in the deeper layers where light intensity will also be low. In the dark, N2 fixation may be supported by fermentation of endogenous storage carbohydrate. Thus both in the light and in the dark relatively low nitrogenase activities can be expected.

Severin and Stal (2008) recorded light-response curves of nitrogenase activity in coastal microbial mats. They observed changes in the fitted parameters of nitrogenase activity during a 24 h cycle and used these parameters and the monitored natural light intensities to calculate the daily amount of N2 fixation. The daily variations of nitrogenase activity in the different types of microbial mats agreed with those that have been found previously and were typical for the cyanobacterial communities present in these mats (Fig. 4.20). Severin and Stal (2008) also integrated the daily amount of N2 fixation during different days with different total daily irradiances and found that it was independent on the amount of light received by the microbial mat, even if nitrogenase activity had a strong light response. Also, the two types of microbial mats which were investigated and which were characterized by totally different daily patterns of nitrogenase activity, did not differ in their total daily integrated amount of N2 fixation (Table 4.4). These authors concluded that N2 fixation in these microbial mats was tuned to a maximum by the concerted action of a diverse diazotrophic community in which different components become active at different times as the result of the changing conditions.

Fig. 4.20
figure 20

Daily patterns of N 2 fixation in two different types of microbial mat on the green beach of the North Sea island Schiermonnikoog: (a) Mat containing a variety of filamentous cyanobacteria, including heterocystous species; (b) Mat containing predominantly Lyngbya sp. Nitrogenase activity (circles) was calculated from the actual ambient irradiance (grey area) and the light response curves (ARA) recorded for both mats at hourly intervals (From Severin and Stal 2008)

Full size image
Table 4.4 Daily integrated nitrogenase activity (μmol C2H4 mg chl a−1) and photon flux (μmol m−2) for two microbial mats (After Severin and Stal 2008)
Full size table

7.6 Vertical Distribution of N2 Fixation in Microbial Mats

Little is known about the vertical distribution of N2 fixation in microbial mats, but the vertical distribution and dynamics of factors that control it such as light, oxygen and sulphide have been investigated in considerable detail. Light is attenuated strongly in microbial mats. The wavelengths that are absorbed by the cyanobacteria in the top layers are obviously attenuated most strongly. Far red light (<700 nm), however, is absorbed by the cyanobacteria to only a small extent; also the attenuation of this light in (wet) sediment is small compared to shorter wavelengths. Far red light does not support oxygenic photosynthesis but anoxygenic photosynthesis depends on it. It can thus be assumed that the cyanobacteria in the lower part of the mat are not capable of oxygenic photosynthesis. This has been shown by microelectrode measurements of oxygen concentration and photosynthesis. Such measurements have also shown that in some microbial mats sulphide is present in these layers. In an attempt to measure potential nitrogenase activity in microbial mats it was shown that in a mat of 3-mm maximum surface related nitrogenase activity occurred in the depth horizon of 1–2 mm (Stal et al. 1984). However, when nitrogenase activity was expressed on the basis of chlorophyll a, highest specific nitrogenase activity was present in the lowest layer of the cyanobacterial mat (2–3 mm) (Fig. 4.21). Cyanobacterial biomass was highest in the top layer, decreasing gradually until about 3 mm depth. Thus it is likely that a spatial separation of N2 fixation and oxygenic photosynthesis had occurred in this mat. The top layer carries out oxygenic photosynthesis and CO2 fixation, while N2 is fixed in the lower layers.

Fig. 4.21
figure 21

Vertical distribution of chlorophyll a in: (a) potential nitrogenase activity (acetylene reduction, ARA); (b) specific, chlorophyll a-based ARA; (c) a microbial mat (Data from Stal et al. 1984)

Full size image

7.7 Effects of Anoxia and Sulphide on N2 Fixation in Microbial Mats

Among the non-heterocystous diazotrophic bacteria those that are capable of N2 fixation under fully aerobic conditions are rare (Bergman et al. 1997). Since they grow by oxygenic photo­synthesis, these organisms not normally perform N2 fixation. It has been questioned whether this capacity of N2 fixation is of any importance in the natural environment (Rippka and Waterbury 1977). Padan and Cohen (1982) mention that the facultative anoxygenic photosynthetic cyano­bacterium Oscil­la­toria limnetica (Geitlerinema) is capable of N2 fixation when carrying out sulphide-dependent anoxygenic photosynthesis. Villbrandt and Stal (1996) investigated the effect of sulphide on N2 fixation in cyanobacterial mats and on cultures of cyanobacteria isolated from these mats. They compared a mat dominated by a heterocystous cyanobacterium (Anabaena) with another one dominated by non-heterocystous filamentous organisms (Oscillatoria and Phormidium). Sulphide inhibited nitrogenase activity in the mat of Anabaena, but greatly stimulated it in the mat of non-heterocystous cyanobacteria (Fig. 4.22). Both light and dark nitrogenase activity was inhibited by sulphide in the mat of Anabaena but when DCMU was added in order to inhibit oxygenic photosynthesis virtually no effect of sulphide on N2 fixation was seen in this mat. Therefore the effect of sulphide was mainly through the inhibition of oxygenic photosynthesis and respiration. Only the addition of 10 mM sulphide resulted in the almost complete inhibition of dark nitrogenase activity which depends on respiratory energy generation. In the light, this sulphide concentration resulted in a decrease of nitro­genase activity to the level obtained in the presence of DCMU, which in this case was about 40% of the control. This demonstrated that this amount of sulphide caused the complete inhibition of oxygenic photosynthesis. Due to the large amount of iron in this mat the actual concentration of free sulphide was probably much lower. Moreover, as was shown in a laboratory culture of Anabaena isolated from this mat, the effect of sulphide strongly depended on the pH. At pH 9.5 a total sulphide concentration of 5 mM had no effect on N2 fixation but at pH 6.5 this concentration almost completely inhibited nitrogenase activity. This shows that the effect of sulphide is through the gaseous species H2S. This gas will passively diffuse into the cell. Because the pH in these mats is usually high, very little H2S will be present, even when the total concentration of sulphide is high.

Fig. 4.22
figure 22

Effect of sulphide on nitrogenase activity (acetylene reduction, ARA) in: (a) mat of the heterocystous Anabaena sp. (lagoon, Atlantic coast of France); (b) mat of non-heterocystous cyanobacteria (Oscillatoria sp. and Phormidium sp.) (lagoon, Mediterranean coast of France) (Data from Villbrandt and Stal 1996)

Full size image

In non-heterocystous mats the situation is totally different. In the control, without sulphide, nitrogenase activity is low in the light. In the dark or when oxygenic photosynthesis was inhibited by DCMU, nitrogenase activity was greatly stimulated. This can obviously be explained by the sensitivity of N2 fixation in these organisms for photosynthetic and atmospheric oxygen. Sulphide stimulated nitrogenase activity in the light, in the dark and with DCMU. Stimulation was most marked in the light and reached a maximum at 5 mM (Fig. 4.21b). However, even at 10 mM sulphide nitrogenase activity was about tenfold the control. The stimulation in the dark was small (it doubled) and reached already maximum at 0.5 mM. Also, with DCMU, stimulation was maximal at 0.5 mM, the same order of magnitude as the effect in the dark (Villbrandt and Stal 1996). The effect of sulphide on the light activity of nitrogenase is best explained by its inhibition of oxygenic photosynthesis in concert with a lowering of environmental oxygen concentration.

None of the cyanobacteria isolated from this mat possessed the capacity for aerobic N2 fixation, but all of the strains were capable of inducing nitrogenase under anaerobic conditions (Villbrandt and Stal 1996). In experiments with Phormidium, it was shown that sulphide (total concentration up to 8 mM) had no effect on nitrogenase activity when this was induced anaerobically with DCMU. Therefore it was concluded that sulphide did not act as an electron donor to nitrogenase and that the stimulatory effect observed in the mat was most probably due to the scavenging of environmental oxygen. Sulphide very efficiently induced nitrogenase in Phormidium and other non-heterocystous cyanobacteria with anaerobic nitrogenase. About 4 mM total sulphide was sufficient for full induction of nitrogenase. However, in contrast with what was seen with the heterocystous cyanobacterium and to what was expected, it appeared that induction of nitrogenase with sulphide was optimal at high pH, which means that the ions HS− and/or S2− were more efficient than the gas H2S. This suggested that Phormidium could actively take up sulphide ion. Thus the uptake of sulphide ion may be essential to allow diazotrophic growth in these organisms.

7.8 Oxygen Protection of Nitrogenase in Microbial Mats

Because cyanobacterial mats often contain a high density of biomass and have low rates of molecular diffusion, they may become markedly supersaturated with oxygen. This poses the question of oxygen protection of nitrogenase in microbial mats. Although in heterocystous cyanobacteria nitrogenase is confined to the heterocysts, and protected from oxygen under normal atmospheric conditions, it has been shown that N2 fixation may be seriously impaired at oxygen pressure well above atmospheric levels. In the majority of cases, non-heterocystous cyanobacteria are the dominant organisms in mats. Many of these species possess only the capacity of anaerobic nitrogenase activity because the lack of an adequate oxygen protection mechanism. They may be able to grow diazotrophically under anaerobic conditions and when sulphide inhibits oxygenic photosynthesis. This may under circumstances lead to a vertical spatial separation of oxygenic photosynthesis in the top layer of the mat and N2 fixation in the deeper parts. Paerl and Prufert (1987) and Paerl et al. (1995) emphasized the importance for N2 fixation of anoxic microzones in microbial mats and other systems. In a few cases, microbial mats have been shown to be built by non-heterocystous cyanobacteria that are capable of N2 fixation under fully aerobic conditions (Pearson et al. 1979; Stal et al. 1984; Villbrandt et al. 1990; Gallon et al. 1991; Paerl et al. 1991). Since nitrogenase in these organisms is as sensitive to oxygen as in any other organism, these cyano­bacteria obviously must possess a protection mechanism. Despite a large amount of research on this problem the precise mechanism by which these species protect nitrogenase from oxygen inactivation is still not known (Bergman et al. 1997). In fact, in all cases in which aerobic N2-fixing cyano­bacteria form microbial mats, N2 fixation is confined to the night. Thus, a temporal separation of N2 fixation and photosynthesis (respectively during the night and during the day) is maintained (Stal 1995). Because these mats turn anoxic during the night, there is no need for oxygen protection. The problem of oxygen protection of nitrogenase in microbial mats is therefore hardly relevant.

Aerobic N2-fixing non-heterocystous cyanobacteria isolated from microbial mats include the filamentous Oscillatoria, Lyngbya, and Microcoleus, and the unicellular Gloeothece, Cyanothece and Synechococcus (Bergman et al. 1997). Among the different mechanisms that have been proposed for oxygen protection of nitrogenase, the uptake and reduction of oxygen, seems to be the most promising. Such systems may act in concert with enzymes that remove oxygen radicals.

In Table 4.5 the effects of different treatments of a diazotrophic microbial mat composed of Oscillatoria on nitrogenase activity are shown. When these mats were incubated in the laboratory and exposed to elevated salinity, phosphate fertilization or to a tidal movement of the water, all these treatments resulted in a dramatic increase of nitrogenase activity. The application of a tidal movement (alternating immersion and emersion of the mat) resulted in a two orders of magnitude increase in nitrogenase activity. The vertical profiles of oxygen in these mats, measured at the same time, showed that the increase of N2 fixation was probably the result of markedly decreased concentrations of oxygen. The reference showed oxygen supersaturation, peaking at about 250 μm depth, typical for these mats. When the mats were subject to increased salinity, phosphate fertilization or to a tidal movement, oxygen profiles decreased dramatically. The decrease of oxygen concentration was most pronounced in the mats subject to both phosphate fertilization and tidal movement. These treatments also resulted in the strongest stimulation of nitrogenase activity. It is obvious that the decreased oxygen concentration is associated with the increased potential to fix dinitrogen.

Table 4.5 Effect of different treatments of a nitrogen-fixing microbial mat of Oscillatoria sp. from the island of Texel, The Netherlands
Full size table

One possibility that must be investigated is the capacity of nitrogenase in these cyanobacteria to reduce oxygen (autoprotection) (Bergman et al. 1997). This causes the reduction of O2 to H2O2, which can be further reduced by peroxidases.

7.9 Heterocystous Versus Non-heterocystous Cyanobacteria in Microbial Mats

There is no doubt that heterocystous cyanobacteria are particularly well adapted for diazotrophic growth. They can fix N2 in the light while carrying out oxygenic photosynthesis. In this way they make optimal use of light energy to satisfy the large demands of nitrogenase. Oxygen protection of nitrogenase in these organisms is assured by the heterocyst. Anoxic conditions usually result in scarcely higher nitrogenase activities and the inhibition of oxygenic photosynthesis by DCMU invariably results in lower activities, apparently because it inhibits the flow of reduction equivalents from the vegetative cells. Non-heterocystous cyanobacteria either cannot fix N2 at all in the presence of oxygen or those that can invariably can much better in the dark or when transferred to anoxic conditions (Stal 1995). Often the inhibition of oxygenic photosynthesis by DCMU also stimulates nitrogenase activity considerably. Notwithstanding these facts, the vast majority of marine microbial mats are composed of non-heterocystous cyanobacteria. Thus the question is raised as to why heterocystous cyanobacteria are not more common in these mats.

On the tidal flats in Guerrero Negro, Baja California Sur, Mexico, two types of microbial mats can be found in close vicinity of each other (Stal et al. 1994; Stal 1995) (Fig. 4.7). The smooth mat is composed of the non-heterocystous cyanobacterium Lyngbya aestuarii and covers the lower areas of the tidal flat. On the upper tidal flat and on slightly elevated spots a pustular mat develops which is composed of the heterocystous cyanobacterium Calothrix. Both mats fix N2 but show distinct differences in their daily nitrogenase patterns. Calothrix fixes predominantly during the day while nitrogenase activity in the mats of L. aestuarii is confined to the night. Due to their locations on the tidal flat the mats of L. aestuarii are covered more often and during longer periods of time at high tide than the mats of Calothrix. During inundation of the mats of L. aestuarri diffusion is limited. This causes oxygen supersaturation during the period of photosynthesis and anoxic conditions at night. These anoxic conditions also allow the development of a community of sulphate-reducing bacteria. This mat has a very dense biomass and is characterized by steep gradients of oxygen and sulphide, typical for microbial mats. Due to this dense mat structure the gradients of oxygen and sulphide exist, regardless whether the mat is inundated or not.

The situation in the mat of Calothrix is totally different. This pustular mat has a porous structure. Due to this structure in this mat there is a free exchange of oxygen with the atmosphere and oxygen supersaturation or anoxic conditions are not usually the case. In exceptional cases when the mat is inundated for a prolonged period of time anoxic conditions or oxygen supersaturation may occur but normally the mat will be inundated for short periods or not at all. Stal et al. (1994) hypothesized that heterocystous cyanobacteria would not be able to maintain themselves in an environment in which either dark anoxic conditions or high concentrations of sulphide occur. Cyanobacteria incapable of fermentation will die within 2–3 h of dark anoxic conditions (Stal and Moezelaar 1997). However, there is no reason why heterocystous cyanobacteria should be incapable of fermentation and this has been demonstrated in a number of symbiotic Nostoc spp. (Margheri and Allotta 1993; De Philippis et al. 1996).

In order to investigate the possibility of sulphide as a selecting factor, Villbrandt and Stal (1996) compared a heterocystous and a non-heterocystous N2-fixing mat in two coastal lagoons in France. The mat of heterocystous cyanobacteria was found in a lagoon with exceptional high amounts of iron, while this was not the case in the other system (Stal et al. 1996). As a result of the high amount of iron the sediment on which this microbial mat was found it did not contain any free sulphide because it precipitated as iron sulphide (Schaub and Van Gemerden 1996). Villbrandt and Stal (1996) hypothesized that the absence of sulphide would allow the proliferation of heterocystous species. On the one hand it was indeed demonstrated that N2 fixation in heterocystous cyanobacteria was sensitive to sulphide. But on the other hand, unrealistic high concentrations of sulphide (10 mM) were required to obtain full inhibition. Sulphide inhibition of nitrogenase in heterocystous cyanobacteria depended on H2S, which in microbial mats is present in very low concentrations as a result of the alkaline conditions. However, in heterocystous cyanobacteria it is uncertain whether other metabolic processes than N2 fixation are more severely influenced by sulphide. In microbial mats and living stromatolites from Cuatro Cienegas, Mexico, evidence for the presence of heterocystous N2-fixing cyanobacteria was obtained from the pattern of nitrogenase activity with highest activities during the day and its inhibition of oxygenic photosynthesis (Falcón et al. 2007). The negative effect of molybdate addition on nitrogenase activity observed by these authors was interpreted as a contribution of sulphate reducing bacteria to N2 fixation. Molybdate inhibits sulphate reduction and this would therefore decrease the sulphide production. The ecological effects of such experiments are complex and difficult to interpret.

Another reason why heterocystous cyanobacteria are absent from the majority of microbial mats could lie in the fact that such organisms generally are not motile and that the link between the heterocyst and the vegetative cell is weak (Stal et al. 1994). Non-heterocystous cyanobacteria that form microbial mats are mostly motile by gliding movement. This is an important property since it facilitates optimal vertical positioning. It allows the cyanobacteria to compensate for the rapidly shifting physicochemical gradients in microbial mats. Gliding motility is also important because microbial mats often develop in environments that are characterized by high rates of sedimentation. In the rare cases that heterocystous cyanobacteria dominate microbial mats their filaments are orientated in a uniform manner at the mat surface. The tapered trichomes of Calothrix are orientated vertically with the terminal heterocysts situated away from the surface. These filaments do not glide freely. The same is the case for another mat-building heterocystous cyanobacterium, Scytonema, which likewise reveals a vertical orientation. In this organism intercalary heterocysts are formed located in the centre of the aggregates. It is likely that shear forces produced during gliding in these highly compressed microbial mats would result in the breakage of the weak link between heterocysts and the neighbouring vegetative cells. Hence, although gliding motility is an essential property for cyanobacteria in microbial mats, it has at the same time a serious disadvantage for heterocystous species. It is therefore expected that mats of such cyanobacteria can only develop in environments with low rates of sedimentation and relatively constant physicochemical gradients.

7.10 Other Diazotrophic Organisms in Microbial Mats and the Case of Microcoleus chthonoplastes

The capacity of N2 fixation is widespread among bacteria and archaea and such organisms are among those that form the microbial community of microbial mats. The question is therefore relevant whether or not microorganisms other than cyanobacteria contribute to the fixation of N2 in microbial mats and to what extent. Chemotrophic bacteria and archaea will be confronted with a limited supply of substrate to satisfy the energy demand of nitrogenase. Although they do not evolve O2 they will still have to cope with an aerobic environment and oxygen supersaturation as the result of the oxygenic photosynthesis of the cyanobacteria. Or they avoid the aerobic environment but this would limit their energy generation capabilities.

Anoxygenic phototrophs will have ample energy (sunlight) but may be limited in their sources of electron donor (i.e. sulphide or sulphur, ferrous iron, organic compounds). The problem of non-heterocystous cyanobacteria is to provide an anoxic environment for nitrogenase. Steppe et al. (1996) proposed a joint venture between diazotrophic bacteria and cyanobacteria. The latter would provide the former with organic matter and oxygen and the bacteria provide CO2 and fixed nitrogen to the cyanobacteria. This model evolved from the observation that cultures and natural samples of the common and cosmopolitan mat-building cyanobacterium Microcoleus chthonoplastes possessed nitrogenase genes belonging to the γ- or δ-Proteobacteria, while cyanobacterial nif genes were lacking.

Although in the literature M. chthonoplastes has repeatedly been presented as a diazotroph, this may have been due to wrong identification (Garcia-Pichel et al. 1996; Siegesmund et al. 2008). For instance, the aerobically N2-fixing M. chthonoplastes isolated by Pearson (Pearson et al. 1979; Malin and Pearson 1988) was later identified as Symploca sp. (Janson et al. 1998) and the anaerobic N2-fixing M. chthonoplastes ‘strain 11’ is related to Geitlerinema (Siegesmund et al. 2008) to which genus also the Solar Lake strain ‘Oscillatoria limnetica’ belongs. Dubinin et al. (1992) and Sroga (1997) reported also on N2-fixing Microcoleus but their correct assignments await confirmation. Rippka et al. (1979) were unable to detect nitrogenase activity in the type strain of M. chthonoplastes PCC7420 and Villbrandt and Stal (unpublished results) were unable to induce nitrogenase activity under strictly anaerobic conditions in the collection of the ‘true’ M. chthonoplastes (Garcia-Pichel et al. 1996).

Bolhuis et al. (2010) discovered that a collection of M. chthonoplastes from distant geographic locations possess the structural genes for nitrogenase (nifHDK), but that they were not typical cyanobacterial but rather belong to the δ-Proteobacteria. The type strain of M. chthonoplastes PCC7420 possesses a full nitrogenase operon. These authors were unable to express the nitrogenase genes in any of these strains, though they showed expression in a microbial mat, indicating that the laboratoy conditions used were inappropriate for expressing nitrogenase in this strain. It was conceived that M. chthonoplastes obtained the nitrogenase operon from a sulphate reducing bacterium through lateral gene transfer. Hence, the attribution of N2 fixation to organisms other than cyanobacteria may have been erroneous in a number of reports.

Severin et al. (2010) showed that filamentous and unicellular cyanobacteria dominated the clone libraries of nifH and their transcripts in two coastal microbial mats, but that δ- and γ-Proteobacteria also contributed importantly. The nifH of the γ-Proteobacteria belonged predominantly to anoxygenic phototrophic purple sulphur bacteria. The nifH of the δ-Proteobacteria belonged partly to M. chthonoplastes and for the other part might have belonged to sulphate reducing bacteria. Other reports also mention the predominance of nifH belonging to δ- and γ-Proteobacteria (Zehr et al. 1995; Olson et al. 1999; Omoregie et al. 2004) in microbial mats and Steppe and Paerl (2002) showed the transcription of δ-proteobacterial nifH.

8 Cyanobacteria and the Sulphur Cycle in Microbial Mats

The sulphur cycle has a large impact on microbial mats either when sulphate is present and the end-oxidation of organic matter is carried out by sulphate reducing bacteria, or when the ecosystem receives primary sulphide as is the case in sulphur springs. Seawater contains abundant sulphate (28 mM) and therefore sulphate reduction is usually a dominant process in coastal and hypersaline microbial mats. Sulphate-reducing bacteria are essentially anaerobic micro-organisms that oxidize simple organic compounds using sulphate as electron acceptor, which results in the formation of sulphide. A variety of different chemotrophic microorganisms as well as cyano­bacteria and purple sulphur bacteria are capable of reducing elemental sulphur to sulphide. Sulphide is eventually oxidized back to sulphate. This can be done anaerobically by anoxygenic phototrophic bacteria such as purple and green sulphur bacteria or also by some cyanobacteria. Elemental sulphur is produced as an intermediate in this process. The cycling between elemental sulphur (S0) and sulphide (S2−) is also called the ‘mini sulphur cycle’ and is probably a dominant process in microbial mats (Van Gemerden 1993).

Sulphide-dependent anoxygenic photosynthesis by purple sulphur bacteria may account for more than 25% of the total photosynthetic carbon fixation as was found in microbial mats in the Ebro Delta in Spain (Martínez-Alonso et al. 2004). Some cyanobacteria are capable of sulphide-dependent anoxygenic photosynthesis, but oxidize sulphide only to elemental sulphur or to thiosulphate (Fig. 4.23). Colourless sulphur bacteria oxidize sulphide aerobically to sulphate but some species can carry out this oxidation anaerobically using nitrate as electron acceptor (denitrification). This anaerobic chemotrophic sulphide oxidation is probably not important because of the limited availability of nitrate in microbial mats. Hence, colourless sulphur bacteria and anoxygenic phototrophic bacteria compete for sulphide.

Fig. 4.23
figure 23

A simplified scheme showing the role of cyanobacteria in the cycle of sulphur in a microbial mat. SRB sulphate-reducing bacteria, CSB colourless sulphur bacteria, PSB photosynthetic sulphur bacteria. For further explanation, see text

Full size image

De Wit et al. (1995), using a mathematical model of a microbial mat, discovered a strikingly clear interaction between purple and colourless sulphur bacteria. Depending on the environmental parameters the model predicted that either the colourless sulphur bacteria dominate or that they coexist with purple sulphur bacteria. In the latter case purple sulphur bacteria can outweigh the colourless sulphur bacteria by more than an order of magnitude. This may explain why coastal or hypersaline microbial mats sometimes exhibit a purple layer and sometimes not, even when the black layer of FeS is present in both cases, pointing at anaerobic conditions and the presence and production of sulphide.

Another process that consumes considerable amounts of sulphide is when it reacts with ferric iron producing elemental sulphur. A peak of elemental sulphur has been observed underneath the aerobic zone where high rates of sulphate reduction occurred while sulphide remained very low due to oxidation by anoxygenic photosynthesis or by ferric iron (Wieland et al. 2005). Other processes in the sulphur cycle in microbial mats include the disproportionation of thiosulphate, sulphite and elemental sulphur. In these reactions one part of the molecule is oxidized while the other is reduced (Bak and Pfennig 1987; Canfield and Thamdrup 1996). The biomass of the major functional groups of microorganisms involved in the sulphur cycle, the purple sulphur bacteria, colourless sulphur bacteria and sulphate reducing bacteria may account for 40% of the total bacterial community (Visscher and Van Gemerden 1993).

In microbial mats most of the organic matter produced by photosynthetic CO2 fixation is recycled. The organic matter (dissolved organic matter, DOM) is liberated into the mat environment by a variety of different mechanisms. Fermentation by the cyanobacteria results in the excretion of low-molecular organic carbon compounds (acetate, ethanol, and lactate) that serve directly as substrate for sulphate reducing bacteria. Photorespiration by cyanobacteria results in the formation and excretion of glycolate, which has also been shown to be used by sulphate-reducing bacteria (Fründ and Cohen 1992; Friedrich and Schink 1995) (Fig. 4.22). Degradation of more complex DOM, which is produced as a result of cell lysis or the exudation of extracellular polymeric substances (EPS), requires the combined action of several different microorganisms until it is eventually end-oxidized by sulphate reducing bacteria. Hence, the metabolic activity of the mat cyanobacteria can directly influence sulphate reduction.

It has been assumed that sulphate-reducing bacteria are present only below the euphotic depth in the microbial mat because only there, conditions are permanently anoxic. This layer is recognized by its black colour that indicates the presence of FeS. There are several lines of evidence that this may be incorrect. Sulphate reduction itself may take place under oxygenated conditions although the majority of sulphate-reducing bacteria are obligate anaerobes (Canfield and Des Marais 1991). However, sulphate-reducing bacteria are found throughout the microbial mat as is the case with sulphate reduction (Visscher et al. 1992; Stal 1993). In fact, 16S rRNA gene sequence analysis of microbial mats have demonstrated that the oxygen tolerant sulphate-reducing bacteria are predominantly found in the top layers of the mat while the obligate anaerobic species are found in the deeper layers of the mat (Risatti et al. 1994). Thus a vertical stratification of different groups of sulphate-reducing bacteria is likely and those in the top layer co-exist with cyanobacteria. Several reports have demonstrated the intimate association of cyanobacteria and sulphate reducing bacteria (Baumgartner et al. 2006). Throughout the hypersaline cyanobacterial mat of Solar Lake (Sinai, Egypt) sulphate-reducing bacteria were present and the rates of sulphate reduction were sometimes higher in the oxygenated layer than in the deeper permanent anoxic parts of the mat (Teske et al. 1998). The dominant filamentous sulphate-reducing Desulfonema migrated during a day night cycle moving from the cyanobacterial layer and probably following the oxygen chemocline (Minz et al. 1999). This showed that some sulphate reducing bacteria tolerate substantial levels of oxygen. Facultative aerobic respiration and motility were considered as essential adaptations for these sulphate reducing bacteria to thrive in a microbial mat.

Many sulphate-reducing bacteria are much less oxygen-sensitive than had previously been assumed and sulphate-reducing bacteria are known that are even capable of aerobic respiration (Cypionka et al. 1985; Dilling and Cypionka 1990; Marschall et al. 1993). However, cultures that carry out dissimilatory sulphate reduction in the presence of oxygen have not been isolated thus far. It is possible that sulphate reduction in the oxygenated part of the microbial mat occurs in anoxic microniches, e.g. in aggregates.

Sulphate reduction also plays a crucial role in calcium carbonate precipitation and thereby controls the lithification process in stromatolites. This is supposed to be the result of two intertwined processes. First, the degradation of EPS liberates calcium that was bound to it and the sulphate reducing bacteria produce carbonate and raise the pH (Visscher et al. 2000; Dupraz et al. 2004).

Cohen et al. (1986) distinguished four groups of cyano­bacteria with respect to the degree of sulphide inhibition and the possibility to carry out sulphide-dependent anoxygenic photosynthesis (Table 4.6). Cyanobacteria belonging to Group 1 are extremely sulphide sensitive. Oxygenic photosynthesis is inhibited at low levels of sulphide (<0.1 mM) and these species are not capable of anoxygenic photosynthesis. Cyanobacteria belonging to this group are evidently not important in marine microbial mats but are likely to be found in freshwater lakes, in the oceans or in terrestrial systems in which sulphide is absent or present at insignificant concentrations. Examples of such cyanobacteria are Anacystis nidulans (Synechococcus elongatus) and Plecto­nema boryanum (Leptolyngbya boryana) in which CO2 fixation was inhibited at 60 and 75 μM (Cohen et al. 1986). In Group 2 cyano­bacteria are represented that are incapable of anoxygenic photosynthesis but that resist considerable levels of sulphide. Oxygenic photosynthesis in these organisms is often stimulated at mode­rate (<1 mM) sulphide concentration. This type of adaptation is typical for marine microbial mats with fluctuating sulphide concentrations. The mat-forming and diazotrophic cyanobacterium Oscillatoria limosa (Lyngbya aestuarii) is a typical example of this group (Stal 1995). Also Group 3 cyanobacteria are typically found in marine microbial mats. These cyanobacteria are characterized by sulphide-insensitive oxygenic photosynthesis concurrent with sulphide-dependent anoxygenic photosynthesis. The cosmopolitan mat-forming cyanobacterium ‘Microcoleus chthonoplastes’ (Geitlerinema) belongs to this group (De Wit and Van Gemerden 1988). Oxygenic photosynthesis in Group 4 cyanobacteria is as sensitive to sulphide as in those belonging to Group 1. The difference is in their capacity of carrying out sulphide-dependent anoxygenic photosynthesis. The sulphide tolerance of this group of cyanobacteria varies considerably from less than 1–10 mM. Oscillatoria limnetica (Geitlerinema) is the best-studied cyanobacterium belonging to that group. Photosystem II in this cyanobacterium is switched off when exposed to <0.1 mM of sulphide. Anoxygenic photosynthesis is induced in a process requiring protein synthesis. O. limnetica tolerates up to 9.5 mM sulphide but anoxygenic photosynthesis is gradually inhibited at concentrations exceeding 4 mM. A good example of how a mat community is structured with respect to sulphide was presented for microbial mats in Fuente Podrida, a cold sulphur spring located in East Spain (Camacho et al. 2005). Three filamentous cyanobacteria were found that fitted Group 1 (sensitive) UVFP3, in areas where sulphide was absent and Group 2 (tolerant) UVFP2 and Group 3 (anoxygenic photosynthesis) UVFP1. The latter isolate was related to Planktothrix, while the other cyanobacteria did not have close relatives. Oscillatoria boryana is also a typical Group 4 organism. This organism employs sulphide-dependent ano­xygenic photosynthesis in the early morning which depleted sulphide locally (Castenholz et al. 1991). Sulphide concentrations of over 1 mM inhibit oxygenic photosynthesis completely. With increasing light intensity oxygenic photosynthesis became dominant. At low light sulphide-dependent anoxygenic photosynthesis remained the dominant mode. O. boryana is also able to photosynthesize at substantial rates over a wide range of sulphide concentrations by shifting between oxygenic and anoxygenic modes and possibly by combining both.

Table 4.6 Groups of cyanobacteria with different types of adaptation to sulphide (After Cohen et al. 1986; Stal 1995)
Full size table

In microbial mats most of the sulphide is present as ‘acid-volatile sulphide’ (AVS) which is mostly in the form of ferrous sulphide (FeS) (Fig. 4.22). In this form sulphide is virtually insoluble. Only free sulphide may be toxic. Free dissolved sulphide occurs as hydrogen sulphide or sulphide ions in a pH-dependent equilibrium.

$$ {\rm{H}}_{2}\rm{S}\leftrightarrow {\rm{HS}}^{-}\leftrightarrow {\rm{S}}^{2-}$$

Below pH 7, H2S becomes gradually more important while above pH 9 it is S2−. Between pH 7 and 9 virtually all sulphide is present as HS−. H2S is a gas that can enter the cell by passive diffusion. However, cyanobacteria capable of anoxygenic photosynthesis are apparently capable of uptake of the sulphide ion. Sulphide may also react with elemental sulphur to form polysulphides. It was thought that this process could occur only in microbial mats in which the amount of iron is not sufficient to keep the level of free sulphide low (Jørgensen and Cohen 1977). However, Visscher (1992) measured very high concentrations of polysulphides in a cyanobacterial mat, indicating that this compound may be more common than previously assumed. While on the one hand polysulphides are an order of magnitude more toxic for most organisms than sulphide, on the other hand it may serve as the form of elemental sulphur that is transported in cells (Steudel et al. 1990). The microbial mat purple sulphur bacterium Thiocapsa roseopersicina is capable of anoxygenic photosynthesis at the expense of polysulphide (Visscher et al. 1990).

9 Interactions of Cyanobacteria with Iron

Iron is one of the most abundant elements on Earth and it has several important functions in microbial mats. Iron occurs in three oxidation states: elemental iron Feo, ferrous (reduced iron), Fe2+ and ferric (oxidized iron), Fe3+. Ferric iron is virtually insoluble and in the presence of oxygen ferrous iron is readily oxidized, except under acidic conditions (pH  <  2). Elemental iron is not stable in nature because it will also be oxidized. Thus, in the presence of oxygen at physiological pH iron is hardly available for organisms and aerobic microorganisms often produce compounds that have a high affinity for ferric iron. These siderophores bind iron and transport it into the cells.

Iron is an essential micronutrient for all organisms and also its redox behaviour gives rise to its biological importance. Iron occurs in a number of enzymes that act as electron carriers, such as cytochromes, in respiratory electron transport chains and ferredoxins which serve as electron donors to a variety of processes (including N2 fixation, nitrate- and sulphate reduction) in the cell, and indirectly, CO2 fixation. Moreover, iron is an important co-factor in enzymes such as nitrogenase and nitrate reductase. In addition to this assimilatory metabolism of iron, the dissimilatory iron metabolism is of importance in microbial mats and other environments. Ferric iron may serve as electron acceptor in anaerobic respiration (Lovley 1991). Under acid conditions, ferrous iron can be oxidized aerobically by the chemolithotrophic, autotrophic bacterium Thiobacillus ferrooxidans (Leduc and Ferroni 1994). Under neutral conditions, ferrous iron is rapidly oxidized by oxygen (Druschel et al. 2008). However, Gallionella ferruginea and Leptothrix ochracea oxidize iron to support an autotrophic mode of metabolism (Hallbeck and Pedersen 1991; Carlile and Dudeney 2000). Instead of competing with the chemical reaction they rather seem to compete with the autocatalysis of iron oxidation as the result of their own activity (Rentz et al. 2007). Ferrous iron may also serve as electron donor in anoxygenic photosynthesis by specialized purple bacteria (Widdel et al. 1993; Ehrenreich and Widdel 1994). The fourth biologically controlled iron transformation is the formation of magnetite in magnetotactic bacteria and in a variety of other organisms (Stolz 1993).

In microbial mats iron is often present in high amounts (Wieland et al. 2005). In coastal and hypersaline microbial mats suspended iron oxides present in seawater precipitate in the sediment. Iron may precipitate either as oxides and hydroxides, siderite (FeCO3) or as iron sulphide (FeS) and pyrite (FeS2). Iron readily reacts with sulphide:

$$ {\rm{2Fe}}^{3+}+{\rm{H}}_{2}\rm{S}\to {\rm{2Fe}}^{2+}+{\rm{S}}^{\rm{o}}+2{\rm{H}}^{+}$$
$$ {\rm{2Fe}}^{2+}+2{\rm{H}}_{2}\rm{S}\to \rm{2FeS }+4{\rm{H}}^{+}$$
$$ {\rm{2Fe}}^{3+}+3{\rm{H}}_{2}\rm{S}\to \rm{2FeS }+{\rm{S}}^{\rm{o}}+6{\rm{H}}^{+}$$

FeS is virtually insoluble. Thus both ferric and ferrous iron is important in immobilizing the toxic sulphide. Ferrous iron including FeS will react with oxygen both chemically as well as biologically. In microbial mats oxygen supersaturation may present a problem for cyanobacterial growth and the presence of ferrous iron may aid in keeping the partial pressure of oxygen low (Wieland et al. 2005). Moreover, the oxidation of iron in siderite will result in the libera­tion of CO2.

In coastal and hypersaline microbial mats a layer of oxidized iron is often observed between the layer of cyano­bacteria and the anoxic layers below (Fig. 4.2). When purple sulphur bacteria are present, this layer of oxidized iron usually separates them from the cyanobacteria. The origin of this layer of oxidized iron is not precisely known. Since this layer is generally found at the transition of the oxic-anoxic layers, ferric iron may be produced by chemical oxidation of ferrous iron or biologically by iron-oxidizing bacteria. When oxygen is unavailable due to the continuously migrating oxic-anoxic transition zone, denitrifying bacteria could be responsible for the anaerobic iron oxidation (Straub et al. 1996) although in many microbial mats the amount of nitrate is probably too low to allow for this process. Cohen (1989) supposed that some mat-forming cyanobacteria were capable of iron-dependent anoxygenic photosynthesis. However, it is more likely that the iron is in fact oxidized by oxygen evolved by photosynthesis. Alternatively, anoxygenic phototrophic purple bacteria that use iron as electron donor may have produced the layer of ferric iron. Such organisms have been isolated from freshwater and marine sediments, including intertidal mud (Widdel et al. 1993; Ehrenreich and Widdel 1994). Also green sulphur bacteria have been shown to use Fe2+ as an electron donor in anoxygenic photosynthesis (Crowe et al. 2008). Although such anoxygenic phototrophic bacteria may be responsible for the layer of ferric iron found in microbial mats this has not been demonstrated (Pierson and Parenteau 2000). Moreover, other evidence showed that the oxidation of iron in microbial mats was entirely the result of the oxygen production by the cyanobacteria (Trouwborst et al. 2007). In fact, reduced iron stimulated photosynthesis in cyanobacterial mats which led to higher oxygen levels and higher pH resulting in the precipitation of iron oxides (Pierson et al. 1999).

The formation of distinct layers of oxidized iron in microbial mats may well have resulted in the formation of so-called Banded Iron Formations (BIFs) formed during the Archean and Proterozoic ages. Banded Iron Formations are finely layered sedimentary rocks composed mainly of silica and iron oxides (James and Trendall 1982). BIFs were deposited over large areas and several thousands are known. Although the majority is only few meters thick and covers a limited area, others are several hundreds of meters thick and extend over many thousands of square kilometres (James and Trendall 1982). The iron content of BIFs is typically in the range of 24–35%, which is 5–7 times more than normally found in the crust. These iron formations are therefore of great economic importance. The silica content of BIF’s is about 45%. Together, iron oxides and silica may make up to 90% of the weight of BIF. Iron oxides and silica (chert) occur in alternating layers. The cherty banded iron formation of Hamersley Basin, Australia, is one of the largest in the world and is characterized by stratification at different scales. At the millimetre scale microbands of iron minerals are recognized, separated at the centimetre scale by mesobands of chert. Regular banding is seen at the meter scale (macrobands) (James and Trendall 1982). Over 90% of the deposits are from the early Proterozoic age (2,500–1,900 Million years). Although it is tempting to assume a biological basis for the genesis of these cherty iron stromatolites, so far evidence of biogenesis has not emerged. Because of the fact that BIFs were overwhelmingly present during the early Proterozoic this has also been taken as evidence for the oxygenation of the earth’s atmosphere which started 2,300 Million years ago. One mechanism for BIF formation may be the chemical oxidation of ferrous iron with oxygen evolved by oxygenic photosynthesis, most likely by cyanobacteria (Trouwborst et al. 2007). The huge amounts of ferrous iron in the earth’s crust would act like a buffer and prevent the oxygenation of the atmosphere until most of the iron was oxidized. Although less abundant, the fact that BIF’s are also known from the mid Archean (3.4–2.9  ×  109 years) might indicate other mechanisms. Geological evidence shows that until 2.0  ×  109 years ago the oxygen level of the earth’s atmosphere was still quite low. High energy solar UV irradiation (200–300 nm range) could freely reach the surface of the earth where it could be absorbed by ferrous iron, resulting in the formation of ferric iron and H2 which escaped into the atmosphere (Cairns-Smith 1978). This reaction has been experimentally proven to be a possible explanation for the precipitation of ferric iron.

Ferrous and ferric iron strongly absorbs in the region 220–270 nm, UV light that is deleterious for organisms. It has been suggested that both ferrous and ferric iron play an important role in protection from UV irradiation because they provide an effective UV screen (Pierson and Olson 1989). It is assumed that the flux of UV irradiation that reached the earth surface during the early Precambrian was very high since the oxygen-free atmosphere would scarcely attenuate it. It is also known that microbial life developed on earth during this period, particularly in stromatolites. This life was apparently not arrested by the high UV flux. Although UV-C light does not reach the earth surface because it is completely absorbed by the earth’s atmosphere, it is interesting that some mat-forming cyanobacteria such as Microcoleus chthonoplastes accumulate large amounts of iron at the outer polysaccharide sheath (Stal 1994). Iron is bound to negatively charged polysaccharides, particularly through the presence of uronic acids and precipitates at the sheath (Bender et al. 1994). Acidic extracellular polymeric substances containing carboxyl groups have been shown to mediate iron oxide mineralization (Chan et al. 2009). This has also been seen in a new cyanobacterium (‘Chroogloeocystis siderophila’) isolated from an iron-depositing hot spring microbial mat (Brown et al. 2005). This strain possesses elevated requirements for iron and also tolerates high levels of iron, making this organism well adapted to thrive in high iron environments. ‘C. siderophila’ failed to grow at low (8 μM) and at very high (1,000 μM) concentrations of Fe3+. Although this iron is present in an insoluble form, its toxicity at high concentrations may be through the binding of iron precipitates to the outer sheath. Ferric iron is reduced through this EPS or through other cellular processes associated with it, or even under the influence of light. The ferrous iron produced in this way can be taken up by the cell but when in excess it may cause a problem inside the cell or it may draw excessively on the pool of reducing equivalents. Brown et al. (2005) suggested that the accumulation of iron served as a pool of iron for times of low iron availability but this seems unnecessary when thriving in a high iron environment. However, the iron precipitates bound to the outer sheath may present a way for the uptake of iron in organisms that lack siderophores. The sheath EPS would then serve as a siderophore.

There may be other advantages of accumulating iron precipitates by cyanobacteria. One of the possibilities is that it evolved from an ancient UV screen to serve new functions for the cyanobacteria or for the ecosystem as a whole. For instance, the accumulation of iron by mat-forming cyanobacteria has been shown to protect the organism from sulphide produced either by sulphate-reducing bacteria living in the immediate vicinity of the cyanobacteria or produced by themselves through the reduction of elemental sulphur. Hence, the bound ferric hydroxides may represent a buffer against toxic sulphide, which reacts to produce ferrous iron and iron sulphide. Another possibility is that ferrous iron will react with oxygen, keeping its concentration low and minimizing photorespiration which would lead to a loss of fixed carbon. A low partial pressure of oxygen is essential for the cell in order to minimize the oxygenase reaction of the CO2-fixing enzyme ribulose-1,5-bisphosphate carboxylase/oxygenase (RUBISCO) The layer of ferric iron may therefore present an efficient barrier between the aerobic and anaerobic parts of the system (Stal 2001). Finally, M. chthonoplastes is also capable of reducing ferric iron probably using it as an electron acceptor during anaerobic dark metabolism (Stal 1994).

Iron oxides form complexes with phosphate which is then immobilized and unavailable as source of phosphate. It is liberated when the iron is reduced. Hence, the cycle of oxidation and reduction of iron may also be important for the temporal binding and storage of phosphate in a microbial mat.

10 Phosphorus in Microbial Mats

Few studies have addressed the role of phosphorus in microbial mats. This is remarkable because phosphate is involved in a variety of geochemical reactions that are important in mats and it is indispensible for growth and metabolic activity for all forms of life, including cyanobacteria. The almost complete ignorance of phosphorus in the study of microbial mats is also in strong contrast with the attention it receives in the study of phytoplankton. Generally, nitrogen or phosphorus limits growth of phytoplankton and most likely this applies also to cyanobacteria that form microbial mats.

Typically about 3% dry mass of cells consists of phosphorus, but some cells can store phosphate as polyphosphate, which increases their phosphorus content. Cyanobacteria take up orthophosphate (H3PO4) which is the most common form of inorganic phosphorus.

The solubility of orthophosphate is controlled by elements such as Ca2+, Mg2+, Fe2+, Fe3+ and Al3+. In seawater, the solubility of orthophosphate is predominantly controlled by Ca2+, which at a suitable pH (7.4–8.1) produces the virtually insoluble hydroxyapatite (Ca10(PO4)6(OH)2; solubility product 1.53  ×  10−112) (Ehrlich 1996). In addition, phosphate may also form an insoluble precipitate with ferric iron (FePO .4 2H2O, strengite, solubility product 1.35  ×  10−18). Phosphate may be liberated from these insoluble minerals by microbial activity. The mechanisms include:

  1. (i)

    production of organic acids

  2. (ii)

    production of chelators

  3. (iii)

    dissimilatory reduction of ferric iron

  4. (iv)

    production of sulphide (Ehrlich 1996).

The latter can react with ferric iron phosphate according to:

$$ {\rm{2FePO}}_{4}+3{\rm{H}}_{2}\rm{S}\to \rm{2FeS }+2{\rm{H}}_{3}{\rm{PO}}_{4}+{\rm{S}}^{\rm{o}}$$

All these processes are likely to occur in microbial mats, but the phosphate liberated will be taken up immediately by the microbial community. Hence, the occurrence of free orthophosphate ion in microbial mats is expected to be negligible. Any organic phosphates must be cleaved hydrolytically by phosphatases.

As argued in Sect. 4.7, the growth of most cyanobacterial mats in coastal environments seems to be limited by nitrogen. However, microbial mats formed by heterocystous cyanobacteria are more likely to become phosphate-limited because N2 fixation provides all the nitrogen needed for their growth. Mats built by non-heterocystous cyanobacteria are probably still nitrogen-limited as a result of impairment of N2 fixation by oxygen. However, as is shown in Table 4.5, phosphate fertilization of such a mat resulted in a dramatic increase of N2 fixation. Similar observations were made by Camacho and De Wit (2003) for hypersaline microbial mats and by Pinckney et al. (1995) in stromatolitic microbial mats in the Bahamas. While in hypersaline microbial mats the addition of nitrogen resulted in a shift in the community from cyanobacteria to diatoms without increasing the photosynthetic capacity of the mat, phosphate additions greatly stimulated the cyanobacterial community and their capacity of N2 fixation. Obviously, these mats were co-limited by phosphate and nitrogen. It is known that N2 fixation requires a certain amount of phosphate for optimal performance (de Nobel et al. 1997). The effect of phosphate fertilization on N2 fixation may also have been indirect since it also caused a strong decrease in dissolved oxygen in the mat. The latter explanation is supported by the fact that N2 fixation was also stimulated by other treatments that resulted in a decrease of oxygen (Table 4.5). Phosphate fertilization may also have stimulated heterotrophic bacterial activity and consequently oxygen uptake. The addition of phosphate stimulated gross photosynthesis and oxygen consumption equally well in the hypersaline microbial mat so that net photosynthesis remained unaltered (Ludwig et al. 2006). Hence, the microbial community as a whole was phosphate limited. However, high phosphate (1 mM) additions inhibited photosynthesis but not oxygen consumption. This was possibly due to chemical interactions of phosphate with iron or calcium ions, influencing their availability (Elser et al. 2005). Phosphate limitation caused very high C:P ratios which constrains the quality of the food for herbivores and therefore minimizes grazing activity (Elser et al. 2005). Phosphate ferti­lization of coastal mats resulted in a considerable increase of chlorophyll a and a shift in cyanobacterial species composition from an Oscillatoria-dominated community to one with mainly Phormidium-type forms (Stal unpublished).

Although phosphorus may occur in other oxidation states (from +5 to −3), it is not important in redox reactions, as is the case with nitrogen and sulphur. Bacteria readily oxidize any reduced phosphorus, both aerobically and anaerobically. The reduction of orthophosphate is thermodynamically not favorable and is therefore not important for dissimilatory purposes. Hence, the microbial phosphorus cycle consists predominantly of the uptake of inorganic phosphate and the liberation by excretion or autolysis of organic phosphate, which is subsequently mineralized by phosphatases.

Although almost all phosphate on earth is present in the oxidized (+5) form, it has become clear that the more reduced form phosphonate (+3) plays an important role in many organisms (White and Metcalf 2007). Phosphonates are characterized by a very stable C-P bond. The potential of the use of phosphonate as a source of phosphorus has been proposed for the marine planktonic filamentous cyanobacterium Trichodesmium (Dyhrman et al. 2006) and the transporter gene phnD has been found in marine picocyanobacteria (IIikchyan et al. 2009). The genes that code for the enzymes that are capable of hydrolyzing this bond may have been spread by lateral gene transfer (Huang et al. 2005). In the ocean the source of phosphonates may in fact be the cyanobacterium Trichodesmium, of which 10% of the phosphorus is present as phosphonate (Dyhrman et al. 2009). Hot spring microbial mats may constitute up to 5% as phosphonate and may therefore represent an important source of phosphorus, when other sources become unavailable (Adam et al. 2008). The unicellular mat-forming cyanobacterium Synechococcus OS-B’ possesses genes for the transport of metabolism of phosphonates which were transcribed upon phosphate starvation. This organism could become adapted to growth at the expense of phosphonate even when inorganic phosphate is the dominant source of phosphorus in these mats and phosphonate appeared to be inhibitory in the short term (Adam et al. 2008). The source of phosphonate in microbial mats remains unknown as well as whether it is common in microbial mats.

Phosphate may be stored in mineral deposits such as phosphorite, apatite, strengite, and other forms. Phosphorite deposits are usually found in coastal waters or shallow seas. They can be formed authigenically when soluble phosphate reacts with calcium to form calcium phosphate or by dia­genesis when phosphate replaces carbonate in calcareous concretions (Ehrlich 1996). Both processes are probably biologically controlled. The model of Piper and Codespoti (1975) explains phosphorite formation in the marine environment from the mineralization of organic matter below the oxygen minimum layer, where it is coupled to denitrification. This results in excess inorganic phosphate compared to combined nitrogen. Upwelling transports phosphate to the sea surface, where it precipitates with calcium. This model could also apply to microbial mats where the same processes take place. Such phosphorite accumulation has been observed in cyanobacterial mats found on the bottom of small brackish ponds of atolls in French Polynesia called kopara (Rougerie et al. 1997). Dahanayake and Krumbein (1985) also reported phosphorite formed by a microbial mat, but concluded that fungi rather than cyanobacteria produced this particular fossil mat.

11 Conclusions

Laminated microbial mats are often considered to be recent analogues of fossil Precambrian stromatolites. Stromatolites are laminated lithified structures that have been formed by growth and metabolism of microorganisms. Studies of carbon isotope ratios provide evidence that photosynthesis was involved in the formation of stromatolites and the discovery of microfossils supports the idea that cyanobacteria have built these formations. However, modern microbial mats rarely lithify and doubts have been raised as to whether these systems really can be considered as analogues. Moreover, the sedimentary record may be biased because lithified mats have a greater potential of preservation. Nevertheless, non-lithifying mats have also left their traces in the fossil record and therefore we know that they have existed throughout the geological history. There are a limited number of examples of microbial mats that calcify and form more or less laminated lithified structures which have morphologies very similar to the Precambrian examples. The comparison of lithifying and non-lithifying microbial mats has provided deeper understanding of the factors that determine the processes leading to lithification.

In the majority of examples of microbial mats, cyanobacteria play a key role in their formation. Cyanobacteria are oxygenic phototrophic bacteria and many species are capable of using dinitrogen (N2) as their only source of nitrogen. Hence, these organisms have a minimum requirement to proliferate, which is important considering the harsh conditions in which microbial mats often develop. Only extreme conditions will limit the biodiversity and exclude higher grazing organisms so that cyanobacteria accumulate to the dense community that produces a mat. Cyanobacteria have a number of additional properties that make them excellent model organisms for forming microbial mats. Many species are motile through gliding movement, which allows them to position themselves under optimal conditions. Light and possibly chemical factors serve as signals to direct the movement of the organisms. Many cyanobacteria are further characterized by a high affinity for light and reach maximum rate of photosynthesis at very low light intensities and have low compensation points. The requirement for energy for maintenance purposes is low. Cyanobacteria often have high affinities for nutrients and perhaps even more important, they possess storage possibilities for a variety of growth factors. In addition, their metabolic versatility and reactivity are important properties of cyanobacteria. For instance, cyanobacteria are not only photoautotrophs that perform oxygenic photosynthesis but many are also capable of anoxygenic photosynthesis. The majority of mat-forming cyanobacteria is even capable of performing oxygenic and anoxygenic photosynthesis in concert, allowing maximum flexibility and reactivity to quickly changing environmental conditions. Whereas aerobic respiration of endogenous glycogen seems to be the normal metabolism in the dark, this does not usually occur in microbial mats, which often are devoid of oxygen during the night. However, most, if not all, mat forming cyanobacteria are capable of fermentation.

Growth and metabolic activity of the cyanobacteria introduce organic matter in the microbial mat system and its degradation will drive the growth of other micro-organisms in microbial mats. Although some organic matter may become liberated into the environment by death and lysis of the cyanobacteria, this seems not to be most important. In mature microbial mats there is hardly any net growth of cyanobacteria despite the high rates of photosynthesis. Organic matter may become liberated as a result of photorespiration, fermentation, excretion of organic solutes and the secretion of extracellular polymeric substances (EPS), notably polysaccharides. Cyanobacteria may produce a well-defined polysaccharide sheath. This is often a structural component of the cell envelope of cyanobacteria. However, cyanobacteria may also produce vast amounts of mucilage which is not or only partly associated with the organism. Mucilage is often composed of recalcitrant polysaccharides. It produces a matrix in which the microbial mat is embedded. This material is sticky and it will glue sediment particles and organisms together, giving stability to the sediment surface. Since this matrix cannot be mixed it presents a diffusive barrier. The polysaccharide matrix is therefore responsible for the accumulation and supersaturation with oxygen in the light and likewise for the anoxic conditions that occur during the night. Of particular importance is the role that EPS probably plays in calcification. It is likely that EPS inhibits this process and that it serves as an anti-calcification agent. This may be either by preventing growth of small calcite crystallization nuclei or by binding Ca2+. Therefore cyanobacterial mats in which a high amount of mucus is produced will not calcify. The production of mucus in cyanobacteria may be related to unbalanced growth. Unbalanced growth occurs when one growth factor is in shortage. Nitrogen limitation is well known as a factor that stimulates the secretion of mucus. Marine microbial mats are often developing under conditions of nitrogen shortage. This is evidenced by the fact that these mats are built by N2-fixing cyanobacteria. These mats usually consist of non-heterocystous diazotrophic cyanobacteria that are inefficient in fixing dinitrogen because they lack an effective mechanism to protect nitrogenase from oxygen. This is particularly the case when during the day very high levels of oxygen occur in the mat. At night oxygen is absent and therefore N2 fixation in these mats occurs predominantly then. However, the limited amount of energy and low-potential electrons that can be generated under such conditions will not allow the fixation of ample dinitrogen. Heterocystous cyanobacteria are optimally equipped for N2 fixation in the light and will not usually face nitrogen-limited growth. However, as stated above, such cyanobacteria are largely excluded from microbial mats. Possibly heterocystous cyanobacteria cannot tolerate high levels of sulphide or dark anoxic conditions. Also the absence of gliding motility in heterocystous cyanobacteria and the weak connection between the heterocyst and the vegetative cell may exclude these organisms from environments with high rates of sedimentation. It seems reasonable to assume that Precambrian microbial mats did not face nitrogen limitation and thus might have produced much less mucus, calcification therefore not being inhibited. Future research should investigate the nitrogen state of modern calcifying microbial mats and the role of EPS as anti-calcification agents in non-lithifying mats.