02 and I-I S Coexistence in Stromatolites A Model for the Origin of Mineralogical Lamination in Stromatolites and Banded Iron Formations W.E. ICrumbein, H.Buchholz, P. Franke, D. Giani, C. Giele and K. Wonneberger Fachbereich IV der UniversitS~t, D-2900 Oldenburg

The coexistence of 02 and H 2 S in recent microbial mats is demonstrated. Oxygen and HzS are not only coexisting in considerable concentrations within recent stromatolites but high concentrations of the respective gases occur in alternating laminae. This "sandwich structure" is produced by alternating populations of oxygenic and anoxygenic photosynthetic bacteria and cyanobaeteria. The finding of such conditions may explain the alternation of oxidized and reduced iron minerals in Precambrian stromatofites and possibly also in the banded iron formations (BIFs).

"Stromatolites are organosedimentary structures produced by sediment trapping, binding and/or precipitation of carbonate or silicate minerals as a result of growth and metabolic activities o f microorganisms, principally cyanobacteria." This definition of stromatolites, which can be applied to more than 3. 109-yearold fossil stromatolites as well as to recent examples of laminated microbial ecosystems, has been presented by Awramik and Margulis [1]. The microbiology of recent stromatolites has been studied extensively [2-9]. Castenholz [10] has described examples from thermal environments. Most authors agree about the following findings which have been summarized in the above cited publications and other contributions to the subject [11]: 1. Recent microbial stromatolitic environments are finely laminated. 2. The lamination is expressing daily, monthly or annual cycles and also different layers of microorganisms one below the other. 3. The types and physiological activities of the microorganisms producing the mat structures vary largely and depend on several different parameters such as sedimentation rates, water and gas exchange between the mat and the overlying water, salinity of the water, light regime, chemistry of the interstitial and overlying water, grazing organisms and competition among the organisms. Naturwissenschaften 66, 381-389 (1979)

9 by Springer-Verlag 1979

4. Stromatolitic systems in recent environments are widespread and range from the subtidal of recent seas over the intertidal zone, to peak in hypersaline lagoons and embayments and in so-called" sabkhas ", where the influence of macroorganisms is largely excluded. Another widespread occurrence of stromatolitic microbial mats is in thermal springs (e.g., the Lake View District in Oregon or Yellowstone National Park). Furthermore, such mats occur in dry desert environments at high evaporation and low precipitation rates. 5. All stromatolitic microbial mats of the marine environment have been described as systems in which within a few millimeters oxygen concentration drops to zero, while sulfate reduction underneath causes constant upward migration of hydrogen sulfide and high concentration of sulfide within the mat [9]. 6. Light penetration is very limited and rarely exceeds 1 to 3 mm depth within the mat [3, 12]. 7. The lamination within a single system is usually very fine and the active photosynthetic system rarely exceeds more than 3 to 4 mm depth within the mat. 8. Under certain conditions a considerable portion of the organic material produced within the mats is preserved for many years under continuous sulfate reduction and transformation of organic matter into calcium carbonates [7, 9, 13]. 9. The processes mentioned above often lead to the preservation of laminated structures within the sediments throughout diagenesis to solid rocks (e.g. [14]). 10. Until 1975 it was generally accepted that the main primary producers within such mats are oxygenic cyanobacteria, which have some potential of thriving at very low oxygen pressures in nature. It has been shown, however, that high productivity of cyanobacteria is also possible under anoxic conditions and by anoxygenic photosynthesis of cyanobacteria. This modifies former assumptions [15, 16]. 11. Stromatolitic microbial mats of recent environments serve as study cases for the development of stromatolites throughout the past, namely in the Precambrian, as they were the major living ecosystems on earth for more than two thirds of life history [11, 17]. 12. Within stromatolitic microbial ecosystems most of the primitive organisms are represented which have ruled life on earth for almost 2,109 years. Within these mats anaerobic and facultative anaerobic heterotrophic bacteria, hyphomicrobes, photosynthetic bacteria, cyanobacteria and some chemosynthetic bacteA= dominate [7, 19, 20]. 13. Measurements of photosynthesis and degradation of organic material within such environments are extremely complicated because the actively living system usually consists of several distinct layers of only 1 to 3 mm thickness [2]. 381

14. The evolutionary step from anoxygenic photosynthesis, using only photosystem I, to oxygenic photosynthesis using both, photosystem I and II, may have occurred within stromatolitic environments and, thus, may have initiated the drastic change from anoxygenic environments to the present-day era of a mixture of oxygenated with anoxic systems which seem to be characterized mainly by the presence or absence of eukaryonts [13, 15, 1@ 15. Stromatolitic microbial ecosystems are considered to be major places of production of hydrocarbons and possible sources of fossil fuels, because they are extremely productive shallow-water environments with a high preservation potential caused by anaerobiosis [18].

The scenario of oxygenic and anoxygenated ecosystems emerging from anoxygenic and anaerobic Precambrian stromatolites seems to be simple inasmuch as the evolution of oxygenic photosynthesis was an advantage to the photosynthetic microorganisms because water is the most abundant electron donor on earth. Therefore the stromatolitic ecosystem must have shifted from complete anaerobiosis to create oxygenated subsystems. Cyanobacteria seem to be the "inventors" of oxygenic photosynthesis. Thus, they started to change the chemical gradients in the water column and later in the atmosphere. Subsequently many ecosystems have evolved which are studied extensively today and are characterized by a steep oxygen decrease to zero values at a given point and increasing concentrations of H2S or methane downwards from this chemocline. Well-described marine examples are the Black Sea, many sediment systems with an aerobic layer on top and an anaerobic layer underneath, and stagnant meromictic lakes. From the classical chemical and physicochemical redox diagrams of such environments it is usually derived that oxygen and H2S exclude each other in these transitory ecosystems and that the chemical and biological processes within them are mainly ruled by the strong redox pairs sulfate-sulfide, oxygen-hydrogen, and by methane-CO2. Additional features arise from the oxidation-reduction potentials of ferric and ferrous iron, nitrate, ammonia and N2 and in some special cases Mn(VII) and Mn(IV). The extremely fast chemical and biological oxidation rates of Hz S to S and SO ] seem to emphasize the mutual exclusion of O2 and HzS. Therefore the coexistence of oxygen and H2S has been recorded so far only for very small amounts

[211. The microbial mats of stromatolitic environments have always been described as ecosystems with steep oxygen gradients leading to total depletion of oxygen within the first 1 to 3 mm of the mat and increasing sulfide concentrations further downwards. The distribution of microorganisms in such mats generally fits with these findings inasmuch as it is usually described as one or two layers of photosynthetic oxygenic mi382

croorganisms underlayn by one or several layers of different photosynthetic anoxygenic microorganisms (e.g., green and purple photosynthetic bacteria, anoxygenic cyanobacteria and flexibacteria [3, 4, 8, 22, 23]). It is generally accepted, that the system is ruled mainly by oxygen at the surface, by light penetration into depth, and accessible energy sources and electron acceptors for respiration under anaerobic conditions in a sequence of availability and efficiency. Sulfide produced upon respiration would serve in turn as electron donor for photosynthetic populations in the anaerobic layers and chemolithotrophic or heterotrophic H z S oxidizing bacteria at the H z S - O 2 interface. Recently we have found, however, stromatolitic microbial mats in a sabkha environment which exhibited some astonishing and initially confusing deviations from the overall picture presented here. Based on morphological and ultramorphological analyses of the mat types in this environment and on measurements of physicochemical and productivity parameters within the mat, we shall reconsider the classical model of stromatolitic ecosystems and their possible evolution from the Precambrian to the present-day state. Material and Methods

The "Sabkha Gavish", located a few km south of the Oasis of Nabq on the western shore of the Gulf of Elat, is a small saline lake fed by sea-water springs seeping through a bar (Fig. 1). The salinities of the open water vary from 49 to 330%o. Extended strips of different types of microbial laminated mats have been encountered and are described elsewhere [24]. The mats have been studied extensively during working periods in fall (September 1977), winter (January-February 1978), spring (March 1978) and summer (July 1978). Transmittant light microscopy and scanning electron microscopy (SEM) as well as transmittant electron microscopy (TEM) were combined with element analysis by energy-dispersive X-ray equipment. Primary productivity was estimated by using the oxygen method with BOD bottles, into which sections of the mat were inserted. Samples were taken by means of syringes with 14 mm diameter. Several assays were also done with plastic jars of 1 1 volume inserted into the mat and with plastic jars with a support for the mat below. Algal mats were extracted, cleaned from below, washed in water of the environment, sewed with a needle and thread to a sheet of mm paper to ensure exact surface area and to avoid upfloating, placed on a support, covered by the persphex jar and replaced at the same depth into the environment. Parallel dark measurements were always conducted. This enabled us to minimize the effect of H2S, which migrates from below into the mat and increases in migration rate with any disturbance of the mat. Productivity was also measured by inserting plastic jars into places where the microbial mats were taken away prior to measurements. Besides the oxygen measurements 14C-incorporation experiments were run simultaneously in BOD bottles into which mat sections were placed. The comparison of both methods enabled us to cope with the various problems Naturwissenschaften 66, 381-389 (1979)

9 by Springer-Verlag 1979

Gavish" were most suitable for our purposes since they were the thickest and most expanded mats we have found so far. The Sabkha is called locally "Sabkha Garish" after the first description [32].

Results

-29 ~

28 ~

3/, ~

35 ~

Fig. 1. Map of the study area in the southern Sinai

involved in productivity measurements in such environments (e.g., 02 consumption by H2S migrating into the upper parts, 14C absorption and incorporation to sediments and into the heterotrophic community etc.). Salinity, nutrients, alkalinity, pH, redoxpotential, temperature and light were measured for several 24-h periods in the water column and the stromatolitic mats, partially by using multiple electrodes. Chlorophyll and pheophytin were determined using the methanol-extraction method with a correction for bacteriochlorophyll [25]. Nitrogen fixation was determined by the acetylenereduction method. These data, however, were not corrected by 15N measurements as has been suggested [26]. Light penetration was measured in the field and in the laboratory using a Umea Inst. Quantaspectrometer and a light meter measuring only white light. Different layers of the mat were dissected with scalpels under the dissecting microscope to determine the light-absorption factors of different communities of the mats. Mat samples were also measured after depigmentation using succesively methanol and acetone. Mineral associations within the mats were analyzed with a Philips X-ray diffractometer. Initially a rough survey was done of oxygen and H~S within the mats by collecting gas bubbles from the mat with a funnel and dissolving the gas in Winkler reagent flasks. Later the mat was dissected in horizontally distinct layers and the layers were analyzed for H2S and oxygen by extraction and fractionate centrifugation and filtration in parallel sets for oxygen and H2S according to a combination of the Pachmeier and Winkler methods [28]. Samples of the mat were dissected after each round of measurements. The situation and placement of the different electrodes were related to the microbial layers by measuring the distinguishable layers with a micromanipulator in the laboratory. Thus, it was possible to attribute each of the multiple electrodes to a defined microbial layer and also to relate HaS and oxygen analyses of sections of the mat to the microorganisms building it. The mats of the "Sabkha Naturwissenschaften 66, 381-389 (1979)

9 by Springer-Verlag 1979

Figure 2 shows blown-up sketches of the microbial mats of the "Sabkha Gavish", disregarding the fact that horizontally in all these environments several different types of mats also occur. The mat type presented here, however, is the most abundant one. Every distinguishable layer of microorganisms (light and scanning electron microscopy) was assigned a number. The abundance of microorganisms in relative figures has been summarized in Fig. 2. We also incorporated the positioning of the multiple-electrode platinum strips in the same scale as the thickness of the layers (enlarged 1 : 10). Figure 3 shows the redox-potential readings over a 24-h period, relative to a calomel electrode situated in the supernatant water column. The data were corrected for hydrogen reference electrode. During all measurements an adaptation time of about 6 to 8 h was monitored but disregarded for the results. Oxygen and H z S distribution in the different sections of the mat during day and night are summarized in Fig. 4 for the day maximum and in Fig. 5 for the night minimum situation. Oxidation-reduction potential readings and the oxygen and hydrogen sulfide determinations on sections of the mat show clearly that from day (photosynthetic activity and respiration) to night (only respiration activity) three major peaks of positive oxidationreduction potential change into relatively negative ones. In general terms the production of oxygen during daytime photosynthesis occurs in the water column itself, and in layers 2, 4, and 6, mainly, while layers 1, 3, 5, 7 and 8 do not show considerable active oxygenic photosynthesis. Oxygen bubbles were visible macroscopically in layers 2, 4 and 6. Very few oxygen bubbles formed directly at the surface of the mat. Some oxygen bubbles, however, were visible at the top end of small channels leading downwards into the mat through the leathery surface layer. The oxygen and sulfide analyses of 4 distinguishable layers which we were able to cut in horizontal sections exhibit the same pattern. Oxygen dropped considerably in all layers but kept supersaturation throughout the whole night period in layers 2 and 4. In laboratory experiments with displaced mats of sizes of 25 cm 2 surface area we have tested the time period it takes to reduce the total oxygen stored within the microbial mats in the main oxygenproducing layers. The results are summarized in Fig. 6 as expressed by redox-potential readings. It took a 383

ASSOCIATED MICROORGANISMS COLOUR ZONES AND REDOX - PROBES

DOMINANT F O ~

x

;m

s

MM o

ORANGE -

~ ;m

SYNECHOCOCCUS ( TYPE 1 )

Z - -

u~

YELLOW Z

_

_

E PALE GREEN [

_

_

2

r.

5 +

3

++

++

+

t

if-

++

--+4- ++ 4-

-t-

.+_4- +

+

++

+

4-

q _

_

5

[__

_

_

SYNECHOCOCCUS 6

GREEN

[

_

_

7

i

10" SYNECHOCOCCUS ( TYPE 2 )

~HITE -

4-4- 4-4-]-t++

4-4- 4-

I. . . . . . . + ++ +

4 - + -I-44-

4-+

++

4 - + 4-

4- + 4-+

-~-

++

++

++ + ++ 4--I- -t-

+-I-

+

4-

8 E

-

-

PINKISH -

llqOCAPSA

15YELLOW

9

PLEUROCAPSA

4

GREEN (BLUE)

LPp - B

+

4-

5

PINK

lIH I OCAPSA

+ 4 - 4-

+

6

GREEN (OLIVE)

LPP

4-+

+

-

B

+

20CHROMATIUM

7

F VIOLET

8

BLACK

-

-

++

4-4- Jr

lO CHLOROBIACEAE

DESULPOV IBR I0

4-+ + +

dark period of more than 30 h to consume all oxygen in the oxygen-rich layers under the experimental conditions. Under natural undisturbed conditions the process may develop anaerobic conditions faster than in the experiment with dissected mats. ,During these measurements we took all precautions to avoid oxygen introduction from external sources. A summary of the primary production of the total of the mat according to the different situations is given in Table 1. Productivity was measured during the main light period in the pond and therefore does not give a total daily production-consumption picture as has been worked out in the waters of the Solar Lake [30]. The productivity measurements clearly show that a combined production occurs by oxygenic and anoxygenic photosynthesis. The measurements, however, cannot be regarded as absolute data since non-photosynthetic 1~C incorporation to the mat material and the interference of HzS with the oxygen method were not under total control and thus deviations from the correct values may occur. Astonishingly oxygen production was observed in a depth of more than 16 mm within the mat (Figs. 3 and 4), while light measurements clearly indicate extremely low light penetration into this depth. Light 384

4-44-444-

4-

++

+4-

+44+ + 4 - 4-4- 4 4 - 4-4+-b~ 4--I-

i

Fig. 2. Multilayered microbial ecosystem. The left arabic numbers are the numbers of the various microbial communities which form the stromatolitic unit. Arabic numbers and arrow designate size and placement of electrodes. + present, + + few, + + + many, + + + + most abundant microorganisms of a defined genus

penetration for sections of the visible and the total visible spectrum are summarized in Fig. 7. They show that different light qualities are penetrating to the different depths, thus indicating selective advantages of the microorganism associations in question. Light penetration altogether is largely regulated by the various photosynthetic pigments present, as can be seen from depigmented samples of the same mat (Fig. 7). From redox measurements, oxygen and H2S determinations and from the occurrence of 8 different layers of microbial populations within the "Sabkha Gavish" microbial mats it is concluded: 1. The microbial mat system of the '~ Gavish" produce s very high amounts of oxygen during daytime photosynthesis. 2. Large amounts of H2S are produced within the mat and namely in the black zone of sulfate-reducing bacteria (layer 8). 3. Oxygen produced within the mat is partially diffusing into the overlying water column producing up to four times oversaturation regarding oxygen. 4. Constant migration of H2S through the mat produces distinguishable concentrations of H2S in the overlying water. 5. Coexistence of as much as 4 mg oxygen/1 and 25 ppm H2S within one section of 2 mm thickness (0.5 ml mat material) was found. 6. The main places of oxygenic photosynthesis and 02 production are an extensive layer of coccoid oxygenic cyanobacteria Naturwissenschaften 66, 381--389 (1979)

9 by Springer-Verlag 1979

mm

mV

o 2

4o8,o1

,2 Fig. 3. Oxidation-reduction potentials (corrected for hydrogen electrode) in the mats during a 24-h cycle. Two maxima of oxygen below minima and related minima and maxima of H2S are indicated and followed during the day cycle. The map depicts continuous recordings of the Eh electrodes in Fig. 2

181

2ol

x

x H2S

o

o

"-

"

02

mg/I

mm

O-

21

10 20 30 40 50 60 70 80 go 100 i

I

i

[

i

i

i

i

t

i

0

1

2

3

4

5

6

7

8

9

10

~

t

i

I

t

I

I

~

I

i

t

~

mg/I

mV

0

-200

-100

I

I

water mats

~0 1

x..

+100 I

0

o

02

+200 +300 I

4- yellow

mm 0 2 4-

0 1 2 3 4 5 6 7 8 9 1 0

mg/I mV

I

J"

10 20 30 40 50 60 70 80 90100

~ H 2 S m g / l l l l l l i i l l l l ~

t

i

-200

-100

I

I

water mats ...... yellow

~ "~. ~X~

I

i

•

I

i

+100 I

I

~

~

~

i

+200 +300 I

I

,.~

6 ~ orange

6-" orange

8-

8

10-" green

I0-

12; white-

12-" white-

green

14-" yellow-

14-_ yellow-

16-" pink 18- green pink 20- 9ree n 22-" violet

16- pink 18- green pink 20-- green 22-" violet 24-

24-" black 25

black

/

t ~'/

26

Fig. 4. Distribution of oxygen, H2S and redox-potentials during day (maximum photosynthesis period) which clearly indicate two oxygen maxima simultaneously with constant migration of I-IzS

Fig. 5. Distribution of oxygen, H2S and redox-potentials during night (4.00 h)just before onset of photosynthesis. Oxygen produced during daytime is preserved within the mats throughout the whole night and recovers fast by new addition of photosynthetic oxygen production and H2S consumption

(layer 2) and a layer of filamentous oxygenic cyanobacteria (layer 4).

thesis occur within the microbial mats of the "Sabkha Gavish ". 9. The thickness of one annual layer of the stromatolite system is higher than in any other known recent microbial mat system. It consists of a maximum of 24 mm sectioned into 7 discernible layers above the black sulfate reduction zone. This special situation enabled selective measurements, almost impossible in mats with only 2 to 5 mm of active populations.

7. Photosynthetic potential was preserved in all layers down to layer 7 (Chromatium and Chlorobiaceae). 8. As a result of the hypersaline conditions, desert light regime and shallow clear water overlaying the mat with less than 10 cm water cover, extremely high oxygenic and anoxygenic photosynNaturwissenschaften 66, 381-389 (1979)

9 by Springer-Verlag 1979

385

light off=

.200

light on

+150 J

"++o; iiiiiiiiiiiii o -~,

-100 -

%%, I

17oo

l

1800

25.07.'197B

/"

9 --"

o~176

"::':'.=:-':=2 i

'

600

,,

........... i .++,.~

- - .

i

~ . "'""'"":"~', "~"~,:".:~ ':-=:-:.......... : ~ - ~ ,i i * Ill1*l

.+oo

26.07.1978

i

ooo

; oo

9

ii

IpO t+,,+

27.07.1978

Fig. 6. Oxygen distribution in layers 2, 3, 5 and 8 during a long-term test period of darkness in the laboratory. The oxygen produced and stored in the slime material of the cyanobacteria in layers 2 and 4 mainly is only exhausted after almost 48 h of permanent darkness. The oxygen reservoir recovers very fast after switching on the light Table 1. Primary productivity of the mats measured by the oxygen method and ~+C method in the field and laboratory using different methods and sections of the mats. Field data measured from 11-13.00. Laboratory data using 10 000 lux m- a Sample

Oxygen method [ g C m - 2 h -1]

14C method [gCm-Zh-t I

Dome inserted into mat light dark

0.15 -0.10

0.75 0.08

Active mat separated from sediment and washed light dark

0.47 - 0.12

0.62 0.05

Layer 1 4 light (lab.) dark (lab.)

0.52 -0,08

0.64 0.07

Layer 1-4 light (lab.) dark (lab.)

0.46 -0.1

0.55 0.05

t0. The diurnal variation of oxygen, HzS and redox-potential values shows clearly that oxygen and HzS maxima are produced within the mats in a vertical separation of less than 2 mm, one below the other in alternating layers. Below a zone of low oxygen HzS increases and beneath this surface layer two more oxygen-rich layers are separated by layers of higher H2S concentrations. This clearly indicates that instead of a steady increase of H2S towards the deeper parts of the mat, oxygen peaks twice beneath layers, which are rich in sulfide. 386

11. According to the microscopic analyses of the mat in relation to the measurements of productivity, redox-potential and oxygen, tile whole mat consists of eight vertically separated layers of defined adaptation potential and ecological meaning. 12. No major differences were observed between summer and winter situation at this place. 13. The productivity of the total mat, according to different methodological approaches and by using measurements of photosynthetic potentials of different sections of the mats, clearly indicates that oxygenic photosynthesis occurs in layers i, 2, 3, 4 and 6 in different intensities and that anoxygenic photosynthesis is mainly restricted to layers 3, 5, 6 and 7. 14. Within the "Sabkha Gavish" microbial mat no switch from anoxygenic to oxygenic photosynthesis of cyanobacteria could be ovserved, though some of the cyanobacteria may have the potential (especially those of layer 6). 15. Active photosynthesis within the mat was established down to a depth of 20 ram, which is almost tenfold that of other microbial mats described in literature and occurs at light intensities of less than 10 ~tEinstein cnl-2 s-~+

For the first time coexistence of oxygen supersaturation and up to 25 p p m HaS in microbial mats is described. The main results were also presented in a poster at the IInd International Congress of Ecology [27].

Discussion

The " S a b k h a G a v i s h " is an extremely fascinating example of a microbial stromatolitic environment for Naturwissenschaften 66, 381-389 (1979)

9 by Springer+Verlag 1979

0.3E +06 2

0.9E +04

0 . 3 E + 06

0.8 E + 04

0.2E +06

O,TE +04 (N 69

~1 ~

E 0.2E + 06

E 0.5E +04

s 0.1E +06

0.4E +04

0.gE +05

0.3E +04

O , 4 E + 05

0.1E+04

0.1E+03 without )610 )695 )640 Fitter [nm]

0 400

450

500

550

600

650

700

[nm]

Fig. 7. Light penetration in the microbial mat system. Left: light penetration under the application of different cut-off filters. Right: light penetration over the spectral range by using a system of 13 interference filters. Values are indicated in quanta m - z s - 1 Light penetration through Z 4cm of hypersaline sea water, [] distilled water, zx layer 1 after depigmentation, o layer 4 after depigmentation, ~ layer 4-7 after depigmentation, o layers 1 3 under normal light conditions, + layer 4~5 under normal conditions, + layers 1-6, o layers 1-7 under normal light conditions

it differs from the so far described examples of recent stromatolites in several important points. It has an almost tenfold thicker vertical extension regarding its active photosynthetic communities. It has 8 discernible different layers, while microbial mats usually consist of only up to four different subsystems. Specific layers of the mat are oxygen-saturated during day and night, though sulfide is constantly produced within the mat and constantly migrates into the water column. In contrast to the Solar Lake mats [8, 9, 13], oxygen is preserved in the deeper layers during the whole dark period, while in other mats either no oxygen is produced beneath the surface layer or a switch is observed from oxygenic to anoxygenic photosynthesis under given conditions [13]. It may also occur, that oxygen produced during day and stored in minute gas bubbles is consumed completely during night, as this seems to be the case in the Solar Lake mats [28]. From the data presented here we draw the following conclusions regarding recent gradient ecosystems and their evolution since the Precambrian: The coexistence of large amounts of oxygen together with high concentrations of sulfide has never been observed and measured clearly in recent sediments. Some indications have been found of an overlap of oxygen and sulfide in lakes [21] but this was only a matter of small amounts and short periods of the day. Naturwissenschaften 66, 381-389 (1979)

9 by Springer-Verlag 1979

From our measurements and observations it emerges that eight different microbial communities have adapted themselves with some overlap of the related organisms within the microbial mat system of the "Sabkha Gavish". These stromatolitic communities consist of: Layer 1 A protective layer of light-resistant coccoid cyanobacteria of the Synechococcus type (formerly grouped into the Aphanothece and Aphanocapsa genera [24, 29]). It do~s not photosynthesize much and oxygen produced by it diffuses immediately into the water above. The distribution of cyanobacteria within the uppermost layer is scattered and few cells are embedded in large amounts of capsule material. A few diatoms indicate high light conditions, while some colorless sulfur bacteria at the surface demonstrate the presence of H2S migrating through from below into the water of the pond.

Layer 2 In the second layer at 7 to 10 mm depth a light optimum occurs for a population of coccoid cyanobacteria dominating the layer. Synechococcus sp. (Aphanothece, Aphanocapsa) is dominating with healthy, densely packed cells. Chlorophyll is peaking in this layer and the slime is loaded with oxygen bubbles. Spirulina sp. and several types of the LPP group (Lyngbya, Phormidium, Plectonema) are admixed. At the lower end some colonies of green photosynthetic bacteria and Gloeothece occur as well.

Layer 3 Below this layer a population with low oxygen concentration during day and night and high sulfide values occurs. Photosynthetic bacteria of an extremely large Thiocapsatype (4 to 7 gm), densely packed with thylakoids, are mixed with some cyanobacteria of the Synechococcus and Gloeothece type. Pleurocapsa sp. does occur as well. 387

winter and summer, consisting mainly of Chromatium and Chlorobiaceae with few other species involved. In all samples no Chloroflexus-type bacteria were identified. Light conditions at this depth indicate that some of the activity of this layer must be based on dark heterotrophic growth of the community; and possibly only in summer major photosynthetic activity occurs.

Layer 8 Black layer with mainly sulfate-reducing bacteria.

Fig. 8. SEM micrograph of filamentous and coccoid photosynthetic bacteria and cyanobacteria of zone 3 and 5. They were tentatively determined as Thiocapsa sp. and a member of the LPP group (Schizothrix). Magnification 8 000 x This layer has relatively high production during the day but does not keep oxygenated during the night (Fig. 8).

Layer 4 The fourth layer has increasing amounts of Pleurocapsa with large pseudofilamentous colonies containing many baeocyte-carrying cells. Calcification is frequent within this layer. Photosynthesis is high and so is the chlorophyll a concentration. Oxygen bubbles are distributed within the layer and occur also in the slime above it. In this layer the second oxygen peak occurs during day and night. In several places large bundles of members of the LPP group occur (Microcoleus).

Layer 5 This layer consists almost exclusively of the above described Thiocapsa-like photosynthetic anoxygenic bacterium, which for its size (4 to 7 gm) is only tentatively identified as Thiocapsa. Large amounts of bacteriochlorophyll a occur. Many sulfur drops indicate anoxygenic photosynthesis. Some members of the LPP group and namely of the Microcoleus type are frequent within this layer. Some of them occur in bundles within a common sheath, others without a sheath. At least three different group members can be identified according to cell size, form and diameter, a firm definition only being possible on isolated strains. The cleaning procedure is presently underway (Fig. 8).

Layer 6 Below the pink to orange layer, consisting mainly of Thiocapsa-like photosynthetic bacteria, a thick layer of filamentous cyanobacteria of three different types follows. It ranges from appr. 18 to 20 mm depth. According to the cyanobacterial nomenclature they all belong to the LPP group. According to the classical botanical code Mierocoleus, Schizothrix and Phormidium were identified. This layer did not exhibit major oxygen production or concentrations. It is possible that the eyanobacteria within this layer were thriving on anoxygenic photosynthesis but light-penetration data (Fig. 6) indicate rather low activity altogether.

Layer 7 It is striking that beneath this layer of cyanobacteria another layer of photosynthetic bacteria always occurred in the mats throughout 388

As can be seen from the depigmented samples and light penetration through depigmented samples (Fig. 7) it is expected that some photosynthetic activity is still going on at these depths, limited to a narrow strip of the visible and near-infrared light which still allows light harvesting under these conditions. Oxygen production within the mats is considerable and the primary productivity data calculated from incubations in the light and in the dark by either including or excluding H2S migration from underneath indicate clearly that oxygenic photosynthesis occurs in layers 1, 2, 4 and 6, and less in layers 3, 5 and 7. In the latter layers anoxygenic bacterial photosynthesis is presumably taking place, as can be deduced from differences between CO2 fixation and Oz production as well as from the sulfur globules in the cells of Thiocapsa and Chromatiurn. Altogether this specific stromatolitic microbial mat is indicative of conditions and developments which may well have taken place in the early times of life history on earth and then may have developed diverging lines of adaptational pattern. In our case oxygenic cyanobacteria have evolved and established themselves within a generally reducing environment, creating oxidized microenvironments and survival capacity by storing oxygen throughout the night for respiration activity and possibly also for depoisoning the action of H 2 S . In the immediate vicinity high production of sulfide by sulfate-reducing bacteria in and below the mat creates the favorable conditions for anoxygenic photosynthesis of cyanobacteria and other photosynthetic bacteria. All this takes place in a few millimeters of a column consisting mainly of organic laminated sediments. Conclusion The coexistence of oxygen with sulfide and the occurrence of photosynthetic potential at extremely low irradiation as well as the very distinct lamination of different populations of photosynthetic bacteria within a few millimeters of a microbial mat demonstrates the high degree of specialization and adaptation within these environments. The presence of H2S-resistant oxygenic cyanobacteria as well as of anoxygenic photobacteria and the documented protective adaptations to the environment are witnesses of the long Naturwissenschaften 66, 381-389 (1979)

9 by Springer-Verlag 1979

history and the different lines of adaptation which have been evolving in such systems. The very high productivity under extreme conditions is also indicative of the ecology of the Precambrian era and of possible ways of hydrocarbon accumulation in former periods. Aerobic decay within the mats is limited to the upper 7 layers. Beneath this active zone, several layers exist, in which the activity of sulfate-reducing bacteria and fermentation processes are the only ways of oxidation of organic material produced above. For the first time it has been demonstrated that oxygenic cyanobacteria may thrive below oxygen-free layers of other microorganisms without switching from oxygenic to anoxygenic photosynthesis but by storing large amounts of oxygen within the slime material during night [27]. Light sensitivity, sensibility to hydrogen sulfide, selective absorption within the visible spectrum, quanta efficiency, stability towards desiccation and possibly heterotrophic growth potential regulate the spatial separation o f this microbial mat into eight distinct layers. It is not surprising that the main place of oxygen production is below the protective layer of slime in which only few cyanobacteria are embedded. One interesting point is that H2S is capable of migrating through three layers of oxygen formation with concentrations close to fourfold supersaturation and bubble formation, without being oxidized completely during the diffusion process. This can be explained only by extremely fast turnover rates of organic material and by the presence of high amounts of sulfates available for reduction in these hypersaline waters. The brine reaches salinities almost tenfold those of normal sea water. The high production potential of sulfide is explained as well by the extremely high primary production within the mats furnishing large amounts of organic matter. Since this production is located in layers which also expose anaerobic microenvironments, it is concluded that sulfate reduction is taking place immediately upon the production of organic matter and almost no diffusion of the produced organic energy sources is needed. Fig. 7 clearly indicates that oxygenated conditions rule the interior of cyanobacterial populations, while outside of them in some horizontal and/or vertical distance from these, sulfate is reduced and H z S used immediately for anoxygenic photosynthesis. The occurrence of oxygenic and anoxygenic photosynthesis as well as aerobic and anaerobic respiration and fermentation within a few mm 3 of a microbial mat is an excellent demonstration of the conditions which presumably must have existed during the first periods of evolution of oxygenic photosynthesis in the shallow-water sediments of the, until then, anoxic and anoxygenic Precambrian. It is justified to assume that not only the lamination of communities but also Naturwissenschaften 66, 381-389 (t979)

9 by Springer-Verlag 1979

those of physicochemical biogenic parameters may have contributed to the development of distinct laminations in Precambrian stromatolites. Furthermore at least those banded iron formations (BIF), that have been formed at periods younger than 2.6.109 years before present, may be derived from the alternation of O2- and H2S-~rich laminae described here. BIFs older than 2.6 - 109 years, e.g. the 3.8 9 109 years old systems in Greenland, may have been influenced by similar microbial systems, but we have no proof for such an assumption. This work was supported financially by D F G grant Kr 333/13. The MBL Elat and especially Dr. Y. Cohen have helped us in field work. I gratefully acknowledge discussions with B.B. Jorgensen and G. Kuenen during our joint sabbatical at Elat. Technical aid by K. Liebig, I. Raether and H.G. Schwenzer as well as patient typing by G. Koch is appreciated. 1. 2. 3. 4. 5. 6. 7.

Walter, M.R., in: [11], p. 1 Brock, T.D., in: ibid., p 21 Brock, T.D., in: ibid., p. 141 Golubi6, S., in: ibid., p 113 Monty, C.L.V.: Ann. Soc. Geol. Belg. 90, 55 (1967) Monty, C.L.V., in: [11], p. 193 Krumbein, W.E. (ed.): Envii:onmental Biogeochemistry and Geomicrobiology, Vol. 1 (Proc. III. Int. Syrup. Environment. Biogeochem. 1977, Wolfenb~ttel). Ann Arbor, Michigan: Ann Arbor Science 1978 8. Krumbein, W.E., Cohen, Y., Shilo, M.: Limnol. Oceanogr. 22, 635 (1977) 9. Jorgensen, B.B., Cohen, Y. : ibid. 22, 657 (1977) 10. Castenholz, R.W. : Bacteriol. Rev. 33, 476 (1969) 11. Walter, M.R. (ed.): Stromatolites. Developments in Sedimentology 20. Amsterdam: Elsevier 1976 12. Fenchel, T., Straarup, B.J.: Oikos 22, 172 (1971) 13. Krumbein, W.E., Cohen, Y., in: Fossil Algae, p. 37 (Fl%el, E., ed.). Berlin-Heidelberg-New York: Springer 1977 14. Cohen, Y., et al. : Limnol. Oceanogr. 22, 597 (1977) 15. Krumbein, W.E., Cohen, Y.: Geol. Rdsch. 63, 1035 (1974) 16. Cohen, Y., Padan, E., Shilo, M.: J. Bacteriol. 123, 855 (1975) 17. Schopf, J.W.: Sci. Amer. 239 (3), 110 (1978) 18. Friedman, G.M., in: [7], p. 227 19. Hirsch, P., in: ibid., p. 189 20. Padan, E. : Adv. Microb. Ecol. (in press) 21: Sorokin, Y.I. : Arch. Hydrobiol. 66, 391 (1970) 22. Golubi6, S., in: The Biology of Blue-green Algae, p. 434 (Cart, N.G., Whitton, B.A., eds.). Oxford: Blackwell 1973 23. Castenholz, R.W.: Microb. Ecol. 3, 79 (1977) 24. Krumbein, W.E.: Cyanobakterien Bakterien oder Algen? Ergebn. I. Oldenburger Syrup. Cyanobakt., p. 130 (1979) 25. Potts, M.: P h . D . Thesis, Univ. of Durham, England, 1977 26. Potts, M., Krumbein, W.E., Metzger, J., in: [7], p. 753 27. Krumbein, W.E.: Oxygen and H 2 S - d o they coexist in microbial stromatolitic mats? Abstr. and poster, IInd Int. Congr. Ecol., Jerusalem 1978 28. Jorgensen, B.B., et al.: Appl. Environ. Microbiol. (submitted) 29. Rippka, R., et al.: J. Gen. Microbiol. 111, 1 (1979) 30. Cohen, Y., Krumbein, W.E., Shilo, M.: Limnol. Oceanogr. 22, 609 (1977) 31. Krumbein, W.E. : Salinity related zonation of microbial ecosystems (Sinai). II. Int. Congr. Fossil Algae, Paris 1979 (submitted) 32. Gavish, E , et al., in: X. Int. Congr. Sedimentol. Field excursion geuide book, p. 347. Jerusalem 1978 Received February 12, 1979 389