SEARCH

SEARCH BY CITATION

Keywords:

  • Carbonate precipitation;
  • freshwater microbial biofilms;
  • mesocosm experiments;
  • microbialites;
  • nanospheres;
  • thrombolitic micro-fabrics

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Experimental results
  6. Multiple layering within the biofilm
  7. Discussion
  8. Conclusions
  9. Acknowledgements
  10. References

Microbialites with laminar (stromatolite) and thrombolitic (thrombolite) fabrics are ubiquitous within the Cenozoic freshwater sedimentary record. However, the biology and physiology of the living prokaryote–microphyte biofilms which produced them is only now becoming understood. The present contribution describes a flowing water experimental mesocosm study spanning over 2·5 years and run under near-natural conditions. This work focussed on microbial biofilm precipitation mechanisms which produce thrombolitic carbonate micro-fabrics capable of preservation in the geological record. In particular, the roles of microbial guilds and carbonate precipitation processes were examined and recorded at all stages of thrombolite development. The mesocosm experiments convincingly demonstrated that the biofilm community actively encouraged calcium ion precipitation derived from flowing waters. This precipitation took the form of amorphous calcium carbonate nanosphere clusters. These clusters were not randomly distributed within the biofilm extracellular polymeric substances but were focussed in the close vicinity of living filament and coccoid bacterial clusters within individual living biofilm layers. Significantly, the precipitates never replaced microbial cell walls and never buried the living microbes. During nanosphere precipitation extracellular polymeric substances were progressively occluded from between the developing nanosphere clusters. However, extracellular polymeric substances were never totally removed from within the amorphous calcium carbonate clusters until they had neomorphosed into microspar crystals. The orientation of precipitating microspar crystals within the biofilm appeared to be controlled by the host extracellular polymeric substance fabric (cf. typical crystal growth from solid substrates). Precipitates were organized around the margins of a cancellate microfabric developed by a range of microbial guilds within each biofilm layer. This produced a distinct thrombolitic fabric within the biofilm which was quite distinct from laminar stromatolite fabrics. It is concluded that the mesocosm grown freshwater biofilms and their associated microbialite calcite micro-fabrics present a universally applicable model. Importantly, they provide a mechanism for thrombolite micro-fabric developments throughout the geological record.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Experimental results
  6. Multiple layering within the biofilm
  7. Discussion
  8. Conclusions
  9. Acknowledgements
  10. References

Freshwater carbonates (tufas) develop ubiquitously throughout the World wherever calcium rich waters are present at the Earth's surface. Consequently they are often associated with riverine deposits but can also develop adjacent to resurgences on karstic hillsides, as spring mounds within artesian basins and also within static water lacustrine settings (Ford & Pedley, 1996). Some examples even develop locally around resurgences within hypersaline lakes. The carbonates associated with all these situations form a lithofacies continuum ranging from lime mud dominated fabrics in static waters, detrital phytoclastic and coated grains (oncoids) in areas of turbulent flow, to catchment-scale constructional phytohermal morphologies (tufa barrages) in areas of perennial active flow (Pedley, 1990). The fundamental building block of all these bioconstructions is the microbial biofilm (prokaryote–microphyte biofilm of Pedley, 1992). This is composed of a massive to layered fabric, containing an association of phototroph and heterotrophic bacteria and cyanobacteria, though perhaps of simpler structure than its marine counterparts. Calcium carbonate precipitates are deposited within the extracellular polymeric substances (EPS) of these microbial biofilms. Consequently, the lasting evidence of freshwater biofilm activity within the geological record is the microbialite, a non-specific term proposed by Burne & Moore (1987) to encompass both laminar stromatolitic and micro-cavernous thrombolitic fabrics.

Considerable progress has been achieved over the past two decades in understanding the construction and function of living microbial biofilms in natural carbonate rich aqueous environments. Much of the earlier work was carried out on marine stromatolites (e.g. Arp et al., 2003; Reid et al., 2003; Decho et al., 2005; Dupraz & Visscher, 2005). However, the role of freshwater biofilms in carbonate precipitation has also been recognized, especially in the past decade (e.g. Pedley, 1994; Turner & Jones, 2005; Bissett et al., 2008a,b; Rogerson et al., 2008; Pedley et al., 2009; Arenas et al., 2010; Arp et al., 2010; Gradziński, 2010; Jones & Peng, 2012). These studies have obtained information from microbial biofilms by a range of approaches from the study of biofilms on artificial substrates introduced in rivers (for example, marble and glass tiles) to the use of Eh, pH, conductivity and water temperature microprobes and flow velocity meters. The results from these biofilm surveys reveal high rates of calcite precipitation, much of which appears to be microbially biomediated. Arenas et al. (2010) have convincingly shown how multiple-visit measuring of in situ living biofilm growth can monitor tufa precipitation rates within natural sites. Gradziński (2010) has demonstrated the advantages of re-sampling natural sites by means of carefully prepared artificial substrates and Bissett et al. (2008a,b) have shown the value of in situ microprobe analysis at natural river sites. Turner & Jones, 2005, for example, showed how some forms of calcium carbonate precipitates are characteristically organized into skeletal crystals (triads) which are frequently arranged around cyanobacterial trichome sheaths. Eventually however, living samples must be removed to the laboratory to investigate the roles of individual coccoid and filamentous cyanobacteria (e.g. Arp et al., 2010). This investigation often involves scanning electron microscopy (SEM) work which typically demands desiccation and death of the materials.

Such valuable snapshots of natural microbial communities and their products significantly add to current knowledge of the freshwater carbonate forming process. However, colonized field sites are often distant from the laboratory and much essential information may be lost during sample preparation because it invariably involves dehydration of the associated communal EPS. Equally significantly, prevailing conditions may change rapidly between sampling visits at the field site. These variations are not only beyond the control of the researcher but may stimulate unpredictable responses from the microbial community at indeterminable times.

An alternative approach involving the use of laboratory based biofilms cultured in a mesocosm has been pioneered by Rogerson et al. (2008) and Pedley et al. (2009). This non-invasive mesocosm approach (Fig. 1A and B) has also been employed in the present study in order to examine the calcium carbonate precipitating processes associated with freshwater microbial biofilms.

image

Figure 1. (A) Mesocosm apparatus early in an experiment. (B) Details of the construction of the mesocosm apparatus (modified from Pedley et al., 2009). (C) Photomicrograph of a flume grown biofilm showing a typical thrombolitic fabric. Outer surface of the biofilm is at the top. Thin section of an air dried sample collected after 2·5 years. (D) Photomicrograph of micrite in a typical clotted fabric developed on cyanobacterial filaments. Same sample as (C). (E) Details of individual opaque calcium carbonate crystals from the clotted fabric seen in (C). Note how the cloudy centres grade out into clear sparite. NB: The crystal on the lower left shows a typical EPS-associated pock marked (‘Swiss cheese’) appearance which contributes significantly to the opacity of the biomediated crystals.

Download figure to PowerPoint

In particular, freshwater microbialites typically show two contrasting fabrics which frequently occur in alternation with each other. One consists of laminar and often dense isopachous spar fringe cements (for example, fig. 14 in Pedley, 1994). The other consists of less well-understood open textured, thrombolitic fabrics (tufa thrombolite of Riding, 2000) which may occur in thin, isopachous bands or may totally dominate the tufa fabric (Fig. 1C to E and fig. 16 in Pedley, 1994). Despite their uncertain origins, thrombolitic fabrics have a distinguished history throughout the late Proterozoic and Phanerozoic in marine (and probably freshwater) environments. Examples from both environments are typified by clotted micro-textures within larger, fenestal fabrics. However, freshwater (tufa) thrombolites frequently also contain microspar and spar (cf. Hoffmann, 1976; Turner et al., 2000). The alternating sparry isopachous fringe cements and particularly the thrombolitic cements demand further laboratory-based study. Consequently, the laboratory development and function of freshwater thrombolitic micro-fabrics forms the subject of this study. The aim of this work was to culture freshwater microbial biofilms over a period of years within laboratory-based mesocosm and microcosm experiments. The experiments were designed to encourage the precipitation of calcium carbonate by the biofilms; these were examined at intervals to explore the processes and products which contributed to the development of freshwater microbialite micro-fabrics.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Experimental results
  6. Multiple layering within the biofilm
  7. Discussion
  8. Conclusions
  9. Acknowledgements
  10. References

Laboratory-based mesocosm studies permit a close matching of the laboratory environment to that of the natural environment. The equipment used for this study was designed to monitor the physical development of a living freshwater microbial biofilm under flowing water conditions over a period of ca 2·5 years. Two identical mesocosms supplied by separate but identical water sources were used in order to demonstrate reproducibility of results. Details of the construction and operation of the mesocosm apparatus are outlined in Pedley et al. (2009). For the present study, the mesocosm was set up with a 60 l sump filled with 40 l of spring water and submersible circulating pump, diaphragm air supplier, and a calcium reactor maintaining the circulating conductivity regulated at ca 150 micro seimens (μS) concentration. Water temperature was maintained at 10·0 to 11·0°C and flow rate maintained at ca 350 l h–1 throughout the experiment. A sodium lamp provided the illumination and was set to 8 h daylight/24 h via a timer (Fig. 1B). Water for the experiment was obtained from a chalk spring supplying Welton Beck, East Yorkshire. This water naturally produces oncoids and has been used successfully in previous experiments (Pedley et al., 2009). It has a very low phosphate and nitrate content and a conductivity (principally of calcium ions) of 150 to 200 μS.

Microbial biofilm cultures were obtained from a well-described, active barrage tufa site (Pedley, 1993; Pedley et al., 2000) on the River Lathkill, Derbyshire. These cultures were seeded into two 1 m lengths of polycarbonate gutter, held within two identical mesocosm flumes (Fig. 2A). Each gutter contained two transverse bars which were provided to create pools and turbulence (see Pedley et al., 2009, for details). Full colonization of the flumes was achieved in ca three weeks during routine recirculation of the water over the gutter surfaces. Microbial biofilm samples were removed for study either on 10 mm square, frosted glass tiles previously placed into regions of active flow or from random spot sampling of biofilm development at the margins and developing spillover bars.

image

Figure 2. (A) Mesocosm flume showing microbial biofilm developments building to produce transverse bars. (B) Mesocosm after desiccation at the end of an experiment. The tufa development visible after the biofilm outer surface has been removed is entirely composed of thrombolitic calcium carbonate.

Download figure to PowerPoint

The experiment was run for 30 months during which time the biofilm deposited an irregular covering of tufa varying from a few millimetres in thickness beneath the biofilm in the ‘pool’ areas to >2.0 cm thick at the transverse bar spillover points (see Fig. 1B and Pedley & Rogerson, 2010). Throughout the experiment pH, conductivity and water temperature were monitored at 10 min intervals on a 24/7 basis via a PC data logging system. In addition, colony viability was maintained by the monthly addition of 250 ml of macerated leaf extract obtained from the Welton catchment.

A second series of 24 samples taken randomly from the evolving mesocosm biofilms were cultured in static water, Petri dish (microcosm) experiments run under natural light conditions. Twelve were propagated by means of deionized water and twelve in Welton spring water. Collectively, these samples provided additional data on the microbial colonization process. Significantly, it was noted that colonization was more rapid on these polycarbonate Petri dish surfaces than on the flume glass slides (cf. Characklis, 1990; Donlan, 2002).

Experimental results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Experimental results
  6. Multiple layering within the biofilm
  7. Discussion
  8. Conclusions
  9. Acknowledgements
  10. References

Initial biofilm colonization without calcium ions

Initial colonization of the mesocosm gutter surfaces comprised filamentous blue green cyanobacteria, diatoms, coccoid heterotrophic, other indeterminate photic bacteria and eukaryotic algae, all associated with copious amounts of EPS. Although operating parameters remained constant throughout this phase of the experiment, progressive changes did occur to the initial biofilm fabric which appeared to be related to changes in the evolving biofilm community. When living samples from glass tile inserts were examined, by light microscopy during the initial three weeks, it was clear that prokaryotes and microphytes were not distributed randomly on the colonized surface but progressively segregated into sub-communities (guilds) within the surficial biofilm. For example, low sinuosity filamentous cyanobacteria (cf. Phormidium/Lyngbya sp.) with very little associated EPS became well- established in vertical-growing tufts and curtains. In contrast, open associations of coiled chains of other cyanobacteria (predominantly Nostoc sp.) enclosed in copious, viscous EPS colonized intervening areas which were generally avoided by the filamentous tufts and curtains. Each EPS dominated Nostoc colony varied from about 0.2 to 0.5 mm in diameter. In effect, the initial randomly settled biofilm community progressively became segregated into guilds at an early stage of biofilm development.

In detail, the microbial colonization process was initiated by random settlement of cyanobacteria, eukaryotic algae and heterotrophic bacteria. Within minutes, however, many of the filamentous forms became attached by one end and the segregation into guilds commenced. Over several days, each Phormidium trichome progressively lengthened to produce a randomly flexed prostrate growth (Fig. 3A) which adhered to the substrate (also replicated in the microcosm experiments, see Fig. 3B). Once several filaments had clumped together, their growing ends effected a polarity change in direction and began to extend vertically from the substrate (Fig. 4A). Progressively, these tufts also extended laterally, linked by additional filaments, until felted tufts and vertical curtains became established on a polygonal pattern (Pedley & Rogerson, 2010), best seen in plan view (Fig. 4B), leaving intervening vacuolar areas occupied by low numbers of Nostoc trichomes and coccoid Pleurocapsacea. The polygonal curtains between adjacent vacuoles did not always extend vertically. Consequently, many vacuolar areas developed a three-dimensional vermiform shape during subsequent wall growth to produce a labyrinthine, polygonal (cancellate) porous fabric.

image

Figure 3. (A) Initial biofilm attachment to the substrate by means of adhering, loosely coiled cyanobacterial filaments associated with photic coccoid bacteria. Note the EPS clouds (bottom centre and top left) containing concentrations of heterotrophic bacteria. Biological transmitted light microscope of a colonized glass slide. (B) Pure stands of filamentous cyanobacteria (Nostoc) within a small EPS-filled polygonal shaped vacuolar area. Note the Phormidium which surrounds the vacuole. Biological transmitted light microscope (experiment using deionized water).

Download figure to PowerPoint

image

Figure 4. (A) Profile view of a cluster of cyanobacterial filaments forming a porous curtain which defines the margins of a vacuole. The top of the biofilm is indicated by the growth direction arrow (microcosm experiment using deionized water). (B) View looking vertically down on to a group of vacuoles showing a typical cancellate fabric (microcosm experiment using deionized water).

Download figure to PowerPoint

Concurrently, small, low viscosity EPS bodies (‘cloud-like’ EPS of Arp et al., 2010), initially present on the colonized substrate, developed within the entwined vertical filament tufts and curtains (especially at hammock sites where filament bunching occurred). In all cases, these cloud-like EPS areas were associated with small (<2 μm long) bean shaped and sometimes highly motile coccoid bacteria (Fig. 5A and B). Extracellular polymeric substance cloud density increased with time but was always associated with marginal and basal areas of the developing cancellate fabric. Significantly, this low viscosity, coccoid associated EPS also developed within the non-illuminated sump during the course of the mesocosm experiments.

image

Figure 5. (A) Scanning electron microscope photomicrograph of an air dried specimen showing small heterotrophic coccoid isolated from an EPS ‘cloud’ associated with filamentous trichomes from a vacuole margin area. (B) An EPS ‘cloud’ suspended in a hammock of filamentous cyanobacteria. Note the fine white dots within the cloud which are the motile heterotrophic coccoid bacteria seen in (A). Biological light microscope image.

Download figure to PowerPoint

Concurrently, the loosely associated, coiled Nostoc sp. colonies (Fig. 3B) were restricted in development to the intervening vacuole areas surrounded by the filamentous cancellete curtain networks. Sometimes, however, they also formed isolated, botryoidal shaped viscous EPS cushions around the periphery of this developing cancellate fabric. The copious volumes of viscous EPS associated with Nostoc imparted considerable rigidity to each of the vacuoles which they occupied, whereas the general coherence to the entire biofilm was provided by the surrounding felted filamentous trichome framework of the cancellate curtains.

These observed results were replicated in twelve microcosm culture (Petri dish) experiments run under de-ionized, static water conditions. It was found that the microcosm cultures consistently replicated the biofilm developments described for the flowing mesocosm experiments supplied with calcium ions.

Inevitably, changes did occur within the microbial mat (biofilm) over time. The most obvious one was the development of bubbles from intra-biofilm biogenic gas accumulation and irregular conduits which provided water ingress to the biofilm. These bubbles were randomly scattered throughout the biofilm as fenestral cavities from one to tens of millimetres in diameter; they developed most readily within the microcosm experiments but were also encountered within the shallowest depth areas of the mesocosm biofilms (see fig. 13B and D in Pedley & Rogerson, 2010). These bubbles frequently caused the upper surface of the biofilm to bulge. Extreme examples within the mesocosm flumes caused biofilms to float as partly anchored rafts around their downstream margins. Additionally, changes occurred in the biota composition of the mesocosm biofilm colonies. In particular, after ca four months diatom populations reduced and filamentous green algae increased (cf. Burns & Walker, 2000).

Another variation in growth pattern was the infrequent development of lamination. This resulted from a process of over-arching of the vacuoles by cyanobacterial filaments associated with the fenestrate curtains of the cancellate fabric. Trichomes subsequently grouped to form a felted flat ‘roof’ to each Nostoc-filled vacuole (cf. ‘fibrillar meshwork’ of Gerdes et al., 2000). Continued colonization of these vacuole roof areas was achieved by the migration and attachment of further loosely coiled cyanobacteria filaments and coccoid associated EPS clouds. These additions served to reinforce the links between adjacent vacuole margins and effectively formed a tabular base to the next incipient cancellate layer. Life conditions were not changed during the experiments. Consequently, no obvious cause could be identified for the laminar developments. However, Gerdes et al. (2000) suggested that such changes might be induced by light intensity caused by changes in water depth.

Thrombolite development

Once the biofilms were established, a controlled supply of calcium ions was introduced into the mesocosms in order to study the biofilm controls upon microbialite fabric development. In the mesocosm experiments the calcium ion supply rate within the flumes was closely matched to natural flow conditions monitored at the spring water sampling site. Calcium carbonate commenced to precipitate within the microbial biofilms during the initial weeks of the colonization period (Figs 6A and 7). The precipitates always developed in two locations: (i) within the biofilm, especially where loosely adherent cyanobacterial filaments lying close together formed hammocks and were associated with ‘cloud-like’ low viscosity EPS; and (ii) within discrete, isolated low viscosity EPS clouds on the vacuolar floor areas.

image

Figure 6. (A) View looking vertically down onto the base of a biofilm to show the development of a vacuolar cancellate fabric. Calcium carbonate precipitates occur throughout the biofilm but primarily within the filamentous cyanobacterial curtains which define the cancellate vacuole margins. Central vacuole areas show some precipitates in basal areas but the central area of each vacuole is precipitate free – SEM photomicrograph. (B) Scanning electron microscope photomicrograph of calcium carbonate nanospheres. These are the initial ACC precipitates developed within biofilm EPS.

Download figure to PowerPoint

image

Figure 7. (A) Living mesocosm biofilm showing vacuolar cancellate fabric (green areas) and calcium carbonate precipitates (white areas). The white arrow indicates the direction of growth and its apex marks the outer growing surface of the biofilm. Biological light microscope image. (B) Detail of vacuolar areas defined by calcium carbonate precipitates and infilled by green coloured EPS. Darker patches at the top are leaf fragments added as part of regular nutrition additions to the experiment. Biological light microscope image.

Download figure to PowerPoint

These small EPS clouds were invariably associated with the coccoid bacteria (Fig. 5B) and never occurred in association with Nostoc sp. Consequently, the cancellate vacuolar fabric became further emphasized by the calcium carbonate precipitation sites (Fig. 7A and B). Circumstantial supporting evidence that the cloud-associated coccoids were dominantly heterotrophic bacteria and paramount in the precipitation process was derived from the activities of identical coccoid associated EPS clouds within the unlit mesocosm sump (‘infection’ of Pedley et al., 2009). When drained, the floor and sides of the sump were found to be coated in EPS associated with identical coccoid bacteria and calcite precipitates. Anchored, living phototrophs were absent in this permanently unlit environment. The coccoid, heterotrophic bacteria appeared to survive here on account of the continuous supply of nutrients. This consisted of dislodged and dead cell material derived from the established ‘sun lit’ mesocosm biofilm, via the recirculated water.

Calcium carbonate fabrics

Nanospheres

The initial mesocosm flume precipitates took the form of spherical, amorphous calcium carbonate (ACC) nanospheres (= nanospherulites of Pedley et al., 2009; nanoparticles of Jones & Peng, 2012), each about 20 nanometres (nm) in diameter (X-ray diffraction analysis verified a total absence of vaterite, aragonite and crystalline calcite and no evidence for opal-A). Nanospheres generally developed at localized sites within the low viscosity EPS clouds associated with the heterotrophic bacteria (Fig. 5B) and never in close vicinity of the more viscous EPS associated Nostoc colonies. Consequently, the first precipitates developed either close to the substrate base or within suspended EPS clouds in the fibrillar meshwork curtains around the cancellate vacuole margins (Fig. 6A).

Amorphous calcium carbonate precipitates commenced growth loosely dispersed within the EPS (Fig. 6B). These sites progressively became more closely packed by the precipitation of additional ACC nanospheres (Fig. 8A) until SEM scale, regular sheet-like nanosphere stacking patterns developed. In areas of dense stacking and little interstitial EPS (5 to 10 μm diameter aggregates) the entire nanosphere association ripened and neomorphed into a coherent rhombic microspar crystal face (Fig. 8B). However, while this neomorphism typically produced well-formed rhombic calcite faces, their interfacial angles and corners were often incomplete. These imperfections were completed by additional nanosphere precipitation with corners being the last to form (Fig. 9A). Frequently, however, imperfections were seen within the crystal faces due to the remaining presence of films and strands of EPS within the developing calcium carbonate fabric. Extracellular polymeric substance filled blind cavities were the most obvious manifestation of this in incipient crystals (Fig. 9B). These developed where nanosphere precipitation failed to occlude all the primary EPS fabric and were common even in relatively perfectly formed crystals. It is these imperfectly occluded EPS sites which give crystals the characteristic ‘Swiss cheese’ appearance. Such fabrics, in natural and experimental mesocosm situations, have only been recognized by the author to occur in crystals grown within the EPS of microbial biofilms.

image

Figure 8. (A) Vertical stacking of nanospheres towards the outer surface of a proto-crystal (direction of the white arrow). Note the alignment of nanospheres into a tabular outer surface. Extracellular polymeric substance occupies all the inter-nanosphere areas of the tabular surface facing the observer – SEM photomicrograph of an air dried sample. (B) Close stacking of nanospheres into layers within the proto-crystal [cf. panel (A)]. Note the ripening into well-formed rhombic crystal faces by further nanosphere nucleation and neomorphism. From this point, crystal growth is by the addition of calcite layers to the faces of the proto-calcite rhomb – SEM photomicrograph of an air dried sample.

Download figure to PowerPoint

image

Figure 9. (A) Construction of interfacial angles in a rhombic crystal; this is achieved by the addition of nanospheres to the edges of three converging sheets which effectively define the incipient crystal faces. Note how ripening of the crystal faces has created perfect planar rhomb faces by progressive occlusion of the remaining EPS-filled cavities – SEM photomicrograph of an air dried sample. (B) A typical nanosphere aggregate (micro-peloid) at an early stage in spar development (first crystal face developing top left). This aggregate shows the characteristic cavernous ‘Swiss cheese’ fabric. Remaining areas of EPS within the growing peloid fabric (now desiccated but formerly visible in the holes) provided a route for further calcium ion streaming into the growing fabric (these depressions correspond to pock marks within the lower left crystal in Fig. 1E) – SEM photomicrograph of an air dried sample.

Download figure to PowerPoint

Continued growth of these isolated microspar crystals commonly resulted in the development of imperfectly interlocking spar and microspar crystal fabrics. These provided support to the cancellate pattern of fibrillar filament curtains surrounding the vacuoles occupied by the Nostoc colonies (Figs 6A and 7). Significantly, the calcite crystals developed suspended within the EPS and many were never in direct contact with the underlying solid substrate during their development. The majority of crystals appeared to be randomly orientated, although Fig. 10A shows that many (crystals A and B) had grown down towards the substrate. In addition, an imperfect cover of slower growing bio-precipitated microspar crystals, growing normal to the substrate (Fig. 10A, area C) developed within the vacuole basal areas.

image

Figure 10. (A) View of an air dried sample removed from a glass slide formerly within the mesocosm. The calcite crystals, suspended during growth within an EPS fabric (‘A’ and ‘B’), have extended their c-axes in the direction of the white arrows and have made contact with smaller crystals growing normal to the basal layer substrate (‘C’). Scanning electron microscope photomicrograph of an air dried sample (modified from Pedley & Rogerson, 2010). (B) Profile of skeletal calcite overgrowths coating a cyanobacterial trichome (filament occupied circular cavity at top left). Note the absence of EPS within this framework and the absence of ‘Swiss cheese’ texture to the skeletal crystal framework. The crystallographic orientation of the long ray triads conform to the rhombic crystal face angles within the final calcite crystal. NB: In this association the crystal c-axis typically lies parallel to the long axis of the cyanobacterial filament – SEM photomicrograph of an air dried sample.

Download figure to PowerPoint

Calcite triads

It is important to note that these intra-biofilm precipitates developed in a quite different way to the calcite crystals which seed onto individual cyanobacterial filaments or filament clusters. In the case of filament hosted calcite, the incipient crystal growth developed into a skeletal scaffold of long-ray triad calcite (Fig. 10B). Importantly, this was always on the external surface of the EPS sheath (cf. short triads of Turner & Jones, 2005). Crystal c-axes were typically orientated parallel to the cyanobacterial filament long axes and, commonly, single crystals grew to entomb several parallel oriented erect, living filaments. In natural sites, the end product of such precipitation on exclusively filamentous cyanobacterial colonized surfaces was a spar lamina.

Unfortunately, the mesocosm experiments have been less successful in propagating this form of lamina, possibly because of the general reduction in filamentous cyanobacteria early in the experiments. Natural occurrences of laminae, where filament density is high, typically consist of laterally continuous palisade crystal layers. These may alternate with thrombolitic carbonate microfabrics or may make up repeated palisade laminae. In all cases studied the resulting crystals differed from intra-EPS thrombolitic precipitates by the absence of true ‘Swiss cheese’ texture to the microspar rhombs. In contrast, cylindrical tubes within crystals with circular exit holes were common where individual cyanobacterial filaments or clusters had been exposed to sheath surface precipitation. A tentative suggestion that such isopachous spar bands may also develop as faster flow physico-chemical precipitates is supported by past mesocosm experiment studies (Pedley & Rogerson, 2010) and by the present work.

Grazers and their effects

Significantly, it was noted that some biofilms within the microcosm experiments generated no tufa precipitates, even when propagated under suitably alkaline conditions. Typically, these failures appeared to be due to the inhibiting effects of specialist grazers, such as ostracods (Lawrence et al., 2002) and micro-gastropods. This effect was investigated further using static water microcosm experiments and showed that in the presence of Candona ostracods the biofilm structure was effectively dismantled over several days by grazing out all filamentous phototrophs. The surviving parts of the biofilm comprised sub-millimetre peloid shaped flocculated EPS clouds containing coccoid bacteria; these continued to thrive but, in several cases, failed to precipitate observable calcium carbonate. In contrast, protists and metazoans (especially nematode worms) burrowed extensively through living biofilm EPS without detriment to the developing fabrics. There was no observable evidence that they were directly responsible for encouraging either the fenestal conduit development or thrombolitic architecture (cf. Walter & Hayes, 1985; Konishi et al., 2001).

Multiple layering within the biofilm

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Experimental results
  6. Multiple layering within the biofilm
  7. Discussion
  8. Conclusions
  9. Acknowledgements
  10. References

Thrombolitic carbonate precipitates formed continuously during the experiment in areas sheltered from direct turbulence. However, an intriguing aspect of the microbial biofilm was its ability to produce successive (ca 0.5 mm) thick biofilm layers. These developed best in areas of active turbulent flow even under constant experimental conditions. Layering was made more apparent by the growth of nanospheres and subsequent microspar crystals within fibrillar filament meshworks bridging the upper surfaces of vacuoles (Fig. 7A). These filaments rapidly became linked laterally into thin, continuous sheets upon which the next biofilm layer was developed. Regular observations of glass tiles seeded with living biofilm (both within the flowing water mesocosms and also in static water microcosm experiments) revealed the following process leading to the development of a new layer within the precipitating biofilm (see Fig. 11B): Establishment of the cancellate vacuolar fabric with development of fibrillar filament curtain margins by segregation of microbial guilds as previously described. Progressive over-arching, by growth and migration of the filamentous cyanobacteria around the polygonal vacuole margins leading to roofing of the vacuoles to produce a cancellate fabric. Coccoid associated EPS clouds in the roof and curtain areas were preferential sites where the fabric became reinforced with calcium carbonate precipitation, as described previously for the polygon floor, tufts and curtains. Generation of the next layer appeared as a continuous process achieved by a combination of further vertical growth from the underlying tabular layer colonists possibly reinforced by new colonization from within the surrounding waters. The positioning of new vacuoles and the new cancellate fabric within each succeeding biofilm layer did not appear to be directly controlled by vacuole positioning in the previous layer. Progressively the earlier biofilm layers receded from the light as they became buried by younger layers but maintained vigour, even at depths of several tens of millimetres. Distortion of some vacuoles in the underlying cancellate fabrics (see the irregular vacuoles in Fig. 7B) may be present. However, the basic cancellate vacuolar pattern was typically maintained within the progressively deeper buried layers of the biofilm, aided by reinforcement from continuing microspar development.

image

Figure 11. (A) View of the underside of a biofilm basal layer and associated calcite precipitates (seen after the glass tile has been removed). Note the dark grey lines (‘B’) within the pale calcium carbonate precipitates which are cyanobacterial filaments. These were attached to the glass slide substrate and alive at the time of sampling. The outer face of the biofilm (mainly EPS) lies on the lower left of the sample (partly hidden by the scale bar). Also note the abundant cavities within the basal layer, all of which are filled by EPS and directly connected the substrate to the living colony surface. Backscatter SEM photomicrograph of a wet stage preparation. (B) Conceptual two-dimensional model to demonstrate the precipitation processes within a tufa biofilm: Surrounding water is white; EPS is yellow; filamentous cyanobacteria are black; ion diffusion gradients are indicated by blue arrows. Contour density is related to relative concentrations of calcium ions; small rhombs represent calcium carbonate precipitates (darker ones are in older basal lamina and paler ones in younger). Note how calcium ion gradients within the EPS increase towards the basal lamina and cancellate wall fabrics of the vacuole (caused by streaming of chelated calcium ions within the EPS). Here the ions precipitate as ACC nanospheres and proto-crystals in the vicinity of coccoid heterotrophic bacteria. The interconnected EPS also permits further streaming of calcium ions down into underlying layers where earlier crystals continue to grow.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Experimental results
  6. Multiple layering within the biofilm
  7. Discussion
  8. Conclusions
  9. Acknowledgements
  10. References

Previous freshwater mesocosm experiments (Rogerson et al., 2008, 2010; Pedley et al., 2009; Pedley & Rogerson, 2010) have highlighted the excessive supply of calcium ions in the surrounding waters and the ways in which this problem is dealt with by the biofilm community. The present work has built on this by examining and describing the intra-EPS precipitation sites from the inception of nanosphere precipitation to the development of thrombolitic microspar fabrics which survive beyond the lifetime of the microbial biofilm. This work revealed that microbial biofilms operate a sophisticated precipitation process whereby ions are carried away from the EPS/water interface and individual cells, and are stored in pre-determined sites which contribute to colony structure, stability and security. This contrasts with the process associated with exposed, filamentous cyanobacterial trichomes where calcium ions precipitate as skeletal calcite crystals directly onto the EPS coat of the sheath.

Calcium, the most abundant cation in karstic riverine systems, appears to be a particular problem to biofilms in all aqueous environments, possibly because it interferes with intracellular conductivity gradients generally and possibly also interferes with the calcium signalling processes (Norris et al., 1996; Clapham, 2007). Hartley et al. (1996) demonstrated that living ‘algal’ biofilms are exposed to a build-up of pH associated with calcium ionic gradients within a 500 μm deep zone directly overlying the living microbial biofilm surface. These calcium ions are actively captured by chelation into the EPS and are redistributed within it. However, there appears to be a limit to the amount of chelated calcium which can be accommodated here. Consequently, much is precipitated as ACC (nanospheres), not at the outer surface of the EPS but at precise sites within the communal biofilm. Although the experimental data are few in comparison to studies of marine biofilms, the processes leading to calcium carbonate precipitation appear to be similar to those described for marine systems and possibly involve ion build-up within intra-biofilm micro-domains.

The precipitation sites lie within the EPS and never involve entombment of the associated heterotrophic bacteria (also confirmed from natural hydrothermal sites by Jones & Peng, 2012). Significantly, this precipitation process can involve other Group 2 elements (for example, magnesium, Bontognali et al., 2008; barium, Rogerson et al., 2008) and biofilm selectivity varies according to how great a threat the cations pose to the well-being of the community.

This chelation and stabilization of cations is clearly essential but its precise function is still not fully understood. Obst et al. (2009) suggested that its value to the biofilm community may be that it serves as a protection mechanism against uncontrolled precipitation of a thermodynamically stable phase calcite directly onto cyanobacterial cell surfaces.

Bissett et al. (2008a,b) and Shiraishi et al. (2008) showed that the precipitation process was not uniform throughout the diurnal cycle but was focussed within the daytime. These authors found a strong pH increase, coupled with Ca2+ consumption at the biofilm surface, during daylight conditions. Rogerson et al. (2010) refined this observation by employing pH microprobes seeded with biofilm in order to determine the precise timing of calcium carbonate precipitation within the biofilm. These authors demonstrated significant diurnal pH changes internally within the EPS but particularly showed that the onset of calcium carbonate precipitation was associated with a significant and instantaneous rise in pH within the biofilm immediately after ‘sunrise’. In addition, it is apparent from the current flowing water mesocosm experiments that the highest precipitation rates coincided with areas affected by the fastest flow rates. Although the precipitation trigger has generally been ascribed to CO2 loss by evaporation in natural sites, additional unpublished mesocosm data indicates that this may only represent part of the explanation. A more likely hypothesis is that the chelation of calcium ions is a supply limiting process. Consequently, any acceleration in cation supply (i.e. by increasing the flow rate) speeds up the chelation and ion streaming processes, resulting in faster growth of stable precipitates within the EPS. Significantly, the current mesocosm experiments reveal that the process does not lead to random precipitation. It is suggested that some form of selective intra-EPS diffusion process may be operating which is capable of conveying cations to specific extracellular precipitation sites within the EPS. In addition, it is also clear from earlier mesocosm studies (see Rogerson et al., 2008) that chelated cations within the EPS can very rapidly be repatriated to the surrounding water body should the need arise.

The driving force causing internal laminar development within the mesocosm biofilms remains unresolved. Clearly it is not controlled by the precipitation process because laminations developed with equal frequency in the static water microcosm experiments grown in deionized water. In other situations (for example, Fig. 7A) successive calcite laminae developed at regular intervals throughout the experiments. Current flume data supports a view that turbulent flow causes damage and vertical biofilm growth slows. Concurrently, the accelerated calcium ion supply to these areas encourages a local increase in the intra-EPS precipitation rate (see fig. 13 in Pedley & Rogerson, 2010).

By contrast, the literature on marine organomineralization (Dupraz et al., 2009) within biofilms is much more extensive but lacks experimental simulations. Nevertheless, the freshwater precipitation process appears to be closely related to that of the marine realm generally, and particularly to the work of Dupraz et al. (2004) on hypersaline lake biofilms on Eluthra Island (Bahamas). Here, precipitation progressively occluded the organic framework but, unlike the mesocosm biofilms, is described as being initiated inside the vacuolar structures. Other points of difference are the limited development of precipitates in the filamentous photosynthetic lower layer of the Bahaman mats compared with the higher rates in the dead upper layer. This is in contrast to the freshwater mesocosm biofilms which contain alternations of precipitate rich horizontal layers (associated with both filamentous and coccoid forms) and layers where there is marked lateral partitioning of microbial guilds, some areas of which remain precipitate free. However, these differences between marine and freshwater biofilms may be more apparent than real. For example, fig. 11B in Dupraz et al. (2004) showed that the same polygonal (cancellate) vacuolar fabric also existed within their hypersaline ‘mats’ but no annotation or comment was made. In addition, Decho (2009) observed that amorphous ‘organominerals’ precipitation occurred not at the biofilm surface but at intra-EPS sites where local alkalinity increased, sometimes to saturation point. This produced amorphous calcite gels which developed into ACC nanospheres.

Perhaps the closest morphological parallel to the mesocosm microbialite biofilms is to be found in the microbial mats recorded from natural Tunisian peritidal environments (Gerdes et al., 2000). However, the Tunisian examples were from a biofilm in a siliciclastic setting. The Tunisian microbial mats typically grew as thin sheets composed of horizontal or prostrate growing fibrillar meshworks of unsheathed cyanobacterial filaments. Tufts and pinnacles of vertical orientated trichomes rose millimetres from this meshwork surface (see Gerdes et al., 2000, fig. 3C to F) to form millimetre-scale reticulate fabrics very similar to the cancellate mesocosm fabrics. Similarly, coccoid bacteria colonized the vertical filaments generating EPS which served to reinforce the vertical fabrics. The Tunisian microbial mats also produced alternations between laminar sheets, in which the filamentous cyanobacteria were woven horizontally into fibrillar network layers and macroscopic clotted fabric (thrombolitic) layers. The latter were associated with ‘hydroplastic’ cushion shaped and pink coloured viscous EPS developments ‘several centimetres thick’ (Gerdes et al., 2000), here associated with coccoid cyanobacteria.

Thrombolitic fabric versus laminar calcite sheets

One unresolved aspect of the study was the cause of alternating spar and micrite fringe cements in natural sites. Within the mesocosm, this was found not simply to be temperature controlled because the experiments simulating summer and winter temperatures failed to generate significant spar fringe cements. In many natural situations the spar fringes were closely associated with filamentous cyanobacterial colonies, each entombed in single spar crystals (for example, fig. 9D in Pedley, 1994; fig. 12A and B in Gradziński, 2010). Frequently these fringe cements have been equated with ‘winter’ precipitates (e.g. Pedley, 1987), although Arp et al. (2010) suggests that dense microcrystalline layers are summer/autumn precipitates. The climatic control becomes more tenuous when fringes, seen to be present in the same tufa lamina, are shown to be spar dominant on the stoss side of the plant substrate while thrombolitic micrites dominate on the leeward side of the same object (for example, figs 21 and 22 in Pedley, 1994). This partitioning of fabrics according to environmental energy (see Discussions in Gradziński, 2010 and Arenas et al., 2010) is further reinforced by the total absence of spar fringe cements in tranquil water deposits (e.g. Pedley, 1994; Pedley et al., 2003). At present it is considered that the two types of fringe cement developments might be controlled by whether calcite is precipitated at intra-EPS or extra-EPS sites. In both situations the highest flow sites will encourage the fastest calcium carbonate precipitation. In the case of intra-EPS sites, the precipitation rate is controlled by the external calcium ion supply rate and the biofilm requirement to chelate and relocate it as quickly as possible. In contrast, at extra-EPS sites faster flow (and calcium ion supply) would lead directly to rapid precipitation of calcium carbonate. On filamentous microbial EPS surfaces this would encourage further rapid growth of skeletal triad crystallites. In contrast, well-formed calcite crystals might grow directly from purely physico-chemical precipitation on crystalline carbonate substrates.

The reason why crystal sizes range from micrite to coarse spar on extra-EPS derived filamentous microbial precipitates remains unclear. In fact, many coarse thrombolitic fabrics are solely the product of micrite and microspar precipitation around randomly orientated filamentous networks (see Gradziński, 2010). Because it is considered that these macro-thrombolitic fabrics are extra-EPS in origin, they are outside the scope of the current article.

Early diagenesis

Another intriguing aspect of geologically older freshwater carbonates is the preservation of the open fenestrae and vacuoles within the thrombolitic tufa microfabric. This effect is unexpected because of the observed tendency in living biofilms for continuing precipitate development within the progressively deeper buried parts (for example, Fig. 12A and B). The implication here is that, as long as EPS is present within the buried layers, further calcium ion transfer and precipitation will continue. In this scenario the expected final tufa fabric should consist of a solid mass of interlocking calcite crystals. However, this is rarely the case because the EPS link to the surface is eventually broken, thereby isolating the remaining cavities from further calcium ion supply. The present mesocosm experiments illustrate this process well. Here, the living biofilm EPS, together with associated bacterial filaments, ramify through the full thickness of the tufa precipitates (ca 20 mm) although their presence was diminished where extensive crystal growth was present in the oldest layers (Fig. 11A). This additional cement was seen to progressively displace the remaining EPS (Fig. 13A and B) and, in rare cases, the EPS became totally occluded by calcite (Fig. 12B). In one flume spillover area, unmineralized former vacuoles in the deeper buried layers were preserved but only where no EPS was detected within these layers. Clearly, once the EPS link to the living biofilm surface is lost further precipitation by intra-EPS ion streaming is impossible. In addition, the living biofilm cover effectively seals off the underlying ‘dead’ tufa, thus restricting porewater circulation and additional precipitation. Consequently, the open thrombolitic fabric frequently is preserved into the geological record.

image

Figure 12. (A) Details of laminations (indicated by black lines) within a living biofilm fabric. Note the increase of calcite volume in some laminae (‘a’) which may have resulted from a slowing down of biofilm growth during exposure to turbulent flow and associated greater calcium ion supply. Transverse section viewed under normal light microscope. Arrow points in the direction of biofilm growth. Biological transmitted light microscope of a colonized glass slide. (B) Calcite precipitation at an advanced stage where most of the EPS has been occluded with virtual loss of the primary vacuolar fabric. Arrow points in the direction of biofilm growth and apex of arrow lies at the living biofilm surface. Biological transmitted light microscope of a colonized glass tile. Transverse section viewed under normal light microscope.

Download figure to PowerPoint

image

Figure 13. (A) Sketch to show the typical stacking pattern of nanospheres within an embryonic calcite rhomb. Note the ripening of the non-crystalline nanosphere fabric (darker outer zone) into planar rhombic faces. The inset (top right) shows the orientation of the embryonic crystal but with interfacial angles complete. (B) An embryonic crystal showing typical imperfections before growth of the interfacial angles is completed. The cavernous and incomplete nature of the crystal fabric is typical of EPS hosted bioprecipitates. NB: Many areas of EPS remain to be occluded during development of this fabric – SEM photomicrograph of an air dried sample.

Download figure to PowerPoint

An additional very important aspect to note is that in the event of the biofilm dying before the thrombolitic fabrics are fully formed the micrite and microspar fabrics quickly become dismantled. Consequently, thrombolitic microbial biofilms are also potentially important producers of detrital lime mud.

Wider perspectives

Thrombolitic fabrics developed into an essential element of marine (and possibly freshwater) microbialites in the Late Neoproterozoic and became abundant in the lower Palaeozoic (Aitken & Narbonne, 1989; Harwood & Sumner, 2011). The appearance of these fabrics in the record signalled the focussing of solitary microbes into inter-collaborating guilds and the first development of a truly communal EPS matrix.

This intra-EPS commensalism imparted important new advantages over solitary microbial growth. The additional cell anchorage enabled biofilms to colonize areas of greater turbulence than was possible by individual cells. Importantly, the communal EPS functioned as an improved filter which protected colonists from adverse chemistries and direct burial by cements derived from the surrounding environment. Finally, the development of an intra-EPS thrombolitic calcite framework provided the necessary rigidity to the biofilm, thus permitting multi-storey laminated thrombolite construction within considerable thicknesses of living biofilm.

Critically, the benefits imparted by a rigid internal structure could only become reality after EPS secretion became a shared contribution by the diverse prokaryote−microphyte biofilm population that had already overcome its mutual incompatibilities over millions of years of association. Microbial biofilms thus evolved to become the first life forms to show ‘intercellularity’ comparable to multicellular organisms (Shapiro, 1998).

Conclusions

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Experimental results
  6. Multiple layering within the biofilm
  7. Discussion
  8. Conclusions
  9. Acknowledgements
  10. References

The following points, distilled from the mesocosm study, help to explain many observed carbonate precipitation processes in the natural freshwater environment:

  1. The biofilm community encourages streaming of chelated calcium ions within the extracellular polymeric substances (EPS) to pre-selected precipitation sites within the micro-fabric (Fig. 11B).
  2. These ions are precipitated in the form of amorphous calcium carbonate (ACC) nanospheres, probably because these unwanted cations interfere with other ionic transfer processes during biofilm metabolism (Fig. 6B).
  3. Nanosphere precipitates form specifically around the margins of filament/coccoid associated vacuoles. Importantly, they never replace microbial cell wall material and never bury living microbes.
  4. Extracellular polymeric substances are progressively occluded as the calcium carbonate (nanosphere) precipitates evolve into microspar. Increasingly closer stacking of nanospheres and progressive occlusion of EPS results in the development of multi-sheet nanosphere aggregates assembled into a neo-crystalline framework. Importantly, living EPS-filled cavities remain within the nanosphere aggregates and between growing crystals and provide essential routes for the further transfer of calcium ions, thereby reinforcing the evolving crystal framework (Fig. 8).
  5. The closely stacked, multiple sheets of nanospheres neomorphically ripen into well-ordered microspar crystal fabrics. Crystal face edges and interfacial angles are the last parts of these crystals to precipitate (Fig. 13A and B).
  6. The orientation of individual intra-EPS crystals is controlled by the EPS in which they grow. Unlike normal precipitation sites, many biofilm grown crystals have no direct connection with a solid substrate but develop suspended within EPS. The effects of this on crystal c-axis orientations and morphologies have yet to be investigated (Fig. 10A).
  7. The morphological dissimilarities between the mesocosm grown biofilms and those from natural sites may be accounted for by the planar substrates of the experimental cultures which simplify the colony morphology and internal structures. The experimental fabrics were generated under relatively slow flow conditions compared with natural sites and this may have contributed to the dissimilarities.

Significantly, biofilms from varied natural environments all show similarities in structure, function and precipitates to those generated in the laboratory experiments. These similarities suggest that there may be a universal pattern to the structure and function of aqueous microbial biofilms. In all situations these biofilms are achieving the same goal of internally stabilizing externally derived ions within the colony for the benefit and well-being of the entire microbial community.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Experimental results
  6. Multiple layering within the biofilm
  7. Discussion
  8. Conclusions
  9. Acknowledgements
  10. References

Professor Brian Jones and another anonymous reviewer are both thanked for their valuable comments which helped simplify this article by focussing my ideas. Dr Mike Rogerson is also thanked for the many informal discussions we have had on biofilm processes during our mesocosm experimentation at Hull.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Experimental results
  6. Multiple layering within the biofilm
  7. Discussion
  8. Conclusions
  9. Acknowledgements
  10. References
  • Aitken, J.D. and Narbonne, G.M. (1989) Two occurrences of Precambrian thrombolites from the Mackenzie Mountains, northwestern Canada. Palaios, 4, 384388.
  • Arenas, C., Osacar, C., Sancho, C., Vazquez–Urbez, M., Auque, L and Pardo, G. (2010) Seasonal record from recent fluvial tufa deposits (Monasterio de Piedra, N.E. Spain): sedimentological and stable isotope data. In: Tufas and Speleothems: Unravelling the Microbial and Physical Controls (Eds H.M. Pedley and M. Rogerson), Geol. Soc. London Spec. Publ., 336, 119143.
  • Arp, G., Reimer, A. and Reitner, J. (2003) Microbialite formation in seawater of increased alkalinity, Satonda Crater Lake Indonesia. J. Sed. Res., 73, 105127.
  • Arp, G., Bissett, A., Brinkmann, N., Cousin, S., De Beer, D., Friedl, T., Mohr, K. I., Neu, T. R., Reimer, A., Shiraishi, F., Stackebrandt, E. and Zippel, B. (2010) Tufa–forming biofilms of German karstwater streams: microorganisms, exopolymers, hydrochemistry and calcification. In: Tufas and Speleothems: Unravelling the Microbial and Physical Controls (Eds H.M. Pedley and M. Rogerson), Geol. Soc. London Spec. Publ., 336, 83118.
  • Bissett, A., de Beer, D., Schoon, R., Shiraishi, F., Reimer, A. and Arp, G. (2008a) Microbial mediation of stromatolite formation in karst–water creeks. Limnol. Oceanogr., 53, 11591168.
  • Bissett, A., Reimer, A., de Beer, D., Shiraishi, F. and Arp, G. (2008b) Metabolic microenvironmental control by photosynthetic biofilms under changing macroenvironmental temperature and pH conditions. Appl. Environ. Microbiol., 74, 63066312.
  • Bontognali, T.R.R., Vasconcelos, C., Warthmann, R.J., Dupraz, C., Bernasconi, S.M. and McKenzie, J.A. (2008) Microbes produce nanobacteria-like structures, avoiding cell entombment. Geology, 36, 663666.
  • Burne, R.V. and Moore, L.S. (1987) Microbialites; organosedimentary deposits of benthic microbial communities. Palaios, 2, 241254.
  • Burns, A. and Walker, K.F. (2000) Effects of water level regulation on algal biofilms in the River Murray, South Australia. Regulated Rivers: Res. Manage., 16, 433444.
  • Characklis, W. G. (1990) WG Microbial fouling. In: Biofilms (Eds W. G. Characklis and K.C. Marshall), pp. 52384. John Wiley & Sons, New York.
  • Clapham, D.E. (2007) Calcium signaling. Cell, 131, 10471058.
  • Decho, A.W. (2009) Overview of biopolymer-induced mineralization: What goes on in biofilms? Ecol. Eng., 36, 137144.
  • Decho, A.W., Visser, P.T. and Reid, P. (2005) Production and cycling of natural microbial exopolymers (EPS) within a marine stromatolite. Palaeogeogr. Palaeoclimatol. Palaeoecol., 219, 7186.
  • Donlan, R.M. (2002) Biofilms: microbial life on surfaces. Emerg. Infect. Diseases, 8, 881890.
  • Dupraz, C. and Visscher, P.T. (2005) Microbial lithification in marine stromatolites and hypersaline mats. Trends Microbiol., 13, 429438.
  • Dupraz, C., Visscher, C. P. T., Baumgartner, L. K. R. and Reid, P. (2004) Microbe–mineral interactions: early carbonate precipitation in a hypersaline lake (Eleuthera Island, Bahamas). Sedimentology, 51, 745765.
  • Dupraz, C., Reid, R.P., Braissant, O., Decho, A.W., Norman, R.S. and Visscher, P.T. (2009) Processes of carbonate precipitation in modern microbial mats. Earth Sci. Rev., 96, 141162.
  • Ford, T.D. and Pedley, H.M. (1996) A review of tufa and travertine deposits of the World. Earth Sci. Rev., 41, 117175.
  • Gerdes, G., Klenke, T. and Noffke, N. (2000) Microbial signatures in peritidal sediments: a catalogue. Sedimentology, 47, 279308.
  • Gradziński, M. (2010) Factors controlling growth of a modern tufa: results of a field experiment. In: Tufas and Speleothems: Unravelling the Microbial and Physical Controls (Eds H.M. Pedley and M. Rogerson), Geol. Soc. London Spec. Publ., 336, 143191.
  • Hartley, A.M., House, W.A., Leadbeater, B.S.C. and Callow, M.E. (1996) The use of microelectrodes to study the precipitation of calcite upon algal biofilms. J. Colloid Interf. Sci., 183, 498505.
  • Harwood, C.L. and Sumner, D.Y. (2011) Microbialites of the Neoproterozoic Beck Spring Dolomite, Southern California. Sedimentology, 58, 16481673.
  • Hoffmann, H.J. (1976) Environmental diversity of Precambrian stromatolites. In: Stromatolites (Ed. M.R. Walter), pp. 599612. Elsevier, New York.
  • Jones, B. and Peng, X. (2012) Amorphous calcium carbonate associated with biofilms in hot spring deposits. Sed. Geol., 269–270, 5868.
  • Konishi, Y., Prince, J. and Knott, B. (2001) The fauna of thrombolitic microbialites, Lake Clifton, Western Australia. Hydrobiologia, 457, 3947.
  • Lawrence, J.R., Scharf, B., Packroff, G. and Neu, T.R. (2002) Microscale evaluation of the effects of grazing by invertebrates with contrasting feeding modes on river biofilm architecture and composition. Microbial Ecol., 44, 199207.
  • Norris, V., Grant, S., Freestone, P., Canvin, J., Sheikh, F.N., Toth, I., Trinei, M., Modha, K. and Norman, R. (1996) Calcium signalling in bacteria. J. Bacteriol., 178, 36773682.
  • Obst, M., Dynes, J.J., Lawrence, J.R., Swerhone, G.D.W., Benzerara, K., Karunakaran, C., Kaznatcheev, K., Tyliszczak, T. and Hitchcock, A.P. (2009) Precipitation of amorphous CaCO3 (aragonite-like) by cyanobacteria: a STXM study of the influence of EPS on the nucleation process. Geochem. Cosmochim. Acta, 73, 41804198.
  • Pedley, H.M. (1987) The Flandrian (Quaternary) Caerwys tufa, North Wales: an ancient barrage tufa deposit. Proc. Yorks. Geol. Soc., 46, 141152.
  • Pedley, H.M. (1990) Classification and environmental models of cool freshwater tufas. Sed. Geol., 68, 143154.
  • Pedley, H.M. (1992) Freshwater Phytoherm reefs: the role of biofilms and their bearing on marine cementation. In: Ramps and Reefs (Eds H.M. Pedley, G. Harwood and B. Sellwood), Sed. Geol., 79, 255274.
  • Pedley, H.M. (1993) Sedimentology of the Late Quaternary tufas in the Wye and Lathkill valleys, North Derbyshire. Proc. Yorks. Geol. Soc., 49, 197206.
  • Pedley, H.M. (1994) Prokaryote–microphyte biofilms: a sedimentological perspective. Kaupia. Darmstader Betr. Naturgesh., 4, 4560.
  • Pedley, H.M. and Rogerson, M. (2010) In vitro investigations of the impact of different temperature and flow velocity conditions on tufa microfabric. In: Tufas and Speleothems: Unravelling the Microbial and Physical Controls (Eds H.M. Pedley and M. Rogerson), Geol. Soc. London Spec. Publ., 336, 193210.
  • Pedley, H.M., Hill, I., Denton, P. and Brasington, J. (2000) Three-dimensional modelling of a Holocene tufa system in the Lathkill valley, N. Derbyshire, using ground penetrating radar. Sedimentology, 47, 721735.
  • Pedley, H.M., Gonzales–Martin, J.A., Ordonez, S. and Garcia del Cura, M.A. (2003) The sedimentology of Quaternaryperched springline and paludal tufas: criteria for recognition, with examples from Guadalajara Province, Spain. Sedimentology, 50, 2344.
  • Pedley, M., Rogerson, M. and Middleton, R. (2009) Freshwater calcite precipitates from in vitro mesocosm flume experiments: a case for biomediation of tufas. Sedimentology, 56, 511527.
  • Reid, R.P., Visscher, P.T., Decho, A.W. and Stolz, J.F. (2000) The role of microbes in accretion, lamination and early lithification of modern marine stromatolites. Nature, 406, 989992.
  • Riding, R. (2000) Microbial carbonates: the geological record of calcified bacterial–algal mats and biofilms. Sedimentology, 47, 179214.
  • Rogerson, M., Pedley, H.M., Wadhawan, J.D. and Middleton, R. (2008) New insights into biological influence on the geochemistry of freshwater carbonate deposits. Geochim. Cosmochim. Acta, 72, 49764987.
  • Rogerson, M., Pedley, H.M. and Middleton, R. (2010) Microbial influence on macroenvironment chemical conditions in alkaline (tufa) streams: perspectives from in vitro experiments. In: Tufas and Speleothems: Unravelling the Microbial and Physical Controls (Eds H.M. Pedley and M. Rogerson), Geol. Soc. London Spec. Publ., 336, 6582.
  • Shapiro, J.A. (1998) Thinking about bacterial populations as multicellular organisms. Annu. Rev. Microbiol., 52, 81104.
  • Shiraishi, F., Bissett, A., de Beer, D., Reimer, A. and Arp, G. (2008) Photosynthesis, respiration and exopolymer calcium-binding in biofilm calcification (Westerhöfer and Deinschwanger Creek, Germany). Geomicrobiol J., 25, 8394.
  • Turner, E.C. and Jones, B. (2005) Microscopic calcite dendrites in cold-water tufa: implications for nucleation of micrite and cement. Sedimentology, 52, 10431066.
  • Turner, E.C., James, N.P. and Narbonne, G.M. (2000) Taphonomic control on microstructure in Early Neoproterozoic reefal stromatolites and thrombolites. Palaios, 15, 87111.
  • Walter, M.R. and Hayes, G.R. (1985) Links between the rise of the metazoan and the decline of stromatolites. Precambrian Res., 29, 149174.