pathways microbiology thiosulfate transformations sulfate ...pathways of the oxidative part of the...

APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Mar. 1991, p. 847-856 Vol. 57, No. 30099-2240/91/030847-10$02.00/0Copyright © 1991, American Society for Microbiology

Pathways and Microbiology of Thiosulfate Transformations andSulfate Reduction in a Marine Sediment (Kattegat, Denmark)t

BO BARKER J0RGENSEN* AND FRIEDHELM BAKDepartment of Ecology and Genetics, University of Aarhus, Ny Munkegade, DK-8000 Aarhus C, Denmark

Received 25 October 1990/Accepted 27 December 1990

Reductive and oxidative pathways of the sulfur cycle were studied in a marine sediment by parallelradiotracer experiments with 35so42-, H235S, and 35S2032- injected into undisturbed sediment cores. Thedistributions of viable populations of sulfate- and thiosulfate-reducing bacteria and of thiosulfate-dispropor-tionating bacteria were concurrently determined. Sulfate reduction occurred both in the reducing sedimentlayers and in oxidized and even oxic surface layers. The population density of sulfate-reducing bacteria was>106 cm-3 in the oxic layer, high enough that it could possibly account for the measured rates of sulfatereduction. The bacterial numbers counted in the reducing sediment layers were 100-fold lower. The dominantsulfate reducers growing on acetate or H2 were gas-vacuolated motile rods which were previously undescribed.The products of sulfide oxidation, which took place in both oxidized and reduced sediment layers, were 65 to85% S2032 and 35 to 15% S042-. Thiosulfate was concurrently oxidized to sulfate, reduced to sulfide, anddisproportionated to sulfate and sulfide. There was a gradual shift from predominance of oxidation towardpredominance of reduction with depth in the sediment. Disproportionation was the most important pathwayoverall. Thiosulfate disproportionation occurred only as cometabolism in the marine acetate-utilizing sulfate-reducing bacteria, which could not conserve energy for growth from this process alone. Oxidative andreductive cycling of sulfur thus occurred in all sediment layers with an intermediate "thiosulfate shunt" as animportant mechanism regulating the electron flow.

The microbiology of dissimilatory sulfate reduction and itsecological significance for sulfur cycling in marine sedimentshave been intensively studied over the past 15 years. Newdiscoveries which show that the pathways of the cycle andthe metabolic capacities of the bacteria are much moreflexible than has been expected continue to appear. Forexample, the sulfate-reducing bacteria (SRB) are not boundto reducing environments but may live under oxidizingconditions, in which they may respire with nitrate or evenwith oxygen (13, 33). There is also recent evidence thatsulfate reduction may proceed under highly oxic conditions(10). SRB may also use sulfite, thiosulfate, or elementalsulfur, either as alternative electron acceptors or for dispro-portionation of sulfur compounds (a type of inorganic fer-mentation) in their anaerobic energy metabolism (4, 5, 43).During the disproportionation of thiosulfate, the two sulfuratoms are transformed into sulfate and sulfide. Finally,experiments with washed cell suspensions have establishedthat some SRB are even able to reverse their sulfur metab-olism and oxidize reduced sulfur (Sred) species such assulfide or thiosulfate to sulfate in the presence of oxygen(13).Compared to these rapid advances in our understanding of

the ecology and physiology of SRB, the microbiology andpathways of the oxidative part of the sulfur cycle are stillpoorly known. On the basis of studies of the depth distribu-tions of inorganic sulfur transformations in sediments, it hasbeen concluded that sulfide oxidation generally takes placein the anoxic zone (19, 24). Sulfide oxidation even occursunder strongly reducing conditions in deeper parts of sedi-ments (16, 26). The electron acceptors in the reducing

* Corresponding author.t Dedicated to Norbert Pfennig on the occasion of his 65th

birthday.

sediment layers have not been quantitatively identified butprobably include oxidized manganese and iron minerals (1,2).

Thiosulfate may be formed by biological sulfide oxidationor by chemical oxidation of free and iron-bound sulfides (7,18, 34). Thiosulfate concentrations range from low values of<1 to 10 ,uM in sediments to high values of 100 ,uM in saltmarshes (32). Recently, thiosulfate has been found to be themost important product of sulfide oxidation in sediments, soits further transformations are therefore key reactions in thesulfur cycle (25, 26). A "thiosulfate shunt" seems to exist inthe sulfur cycle of both freshwater and marine sediments.Because of the diverse capabilities of SRB both to reducesulfate and thiosulfate and to disproportionate sulfur com-pounds, these bacteria may play a decisive role in theregulation of the electron flow in the sulfur cycle.We traced the main pathways of S042-, H2S, and S2032-

transformations in a marine sediment and studied the viablepopulations of SRB. The goal was to reach an integratedunderstanding of the physiological capabilities of the bacte-ria relative to their activities in the sulfur cycle. The concur-rent transformations in whole sediment cores were moni-tored for four radiotracers: s042-, H2S, and S2032- witheither the inner (sulfonate) or the outer (sulfane) sulfur atomlabeled. The distributions of SRB, thiosulfate-reducing bac-teria, and disproportionating bacteria were studied by im-proved viable counting techniques, and the bacteria werepartly characterized.

MATERIALS AND METHODS

Samples. Sediment samples were obtained during a cruisewith RV Gunnar Thorson (Danish National Agency forEnvironmental Protection) in April 1989. Sediment coreswith intact vertical zonations were collected with a Hapscorer (28) at station 1243 (56°25'N, 11°55'E) in southern

847

on August 15, 2020 by guest

http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/

848 J0RGENSEN AND BAK

Kattegat, Denmark. The station is situated at the transitionbetween the Baltic Sea and the North Sea at a water depth of33 m. Subcores were sampled on deck with 26- and 36-mm-inner diameter Plexiglas tubes, some of which were suppliedwith 1-mm-wide silicone-sealed holes for injection of theradiotracer. To avoid oxygen gradients in the overlyingwater and thereby impede oxygen flux into the sedimentduring incubation, we siphoned off the water from the coresto leave a film only a few millimeters thick over the sedi-ment. The cores were kept at the in situ temperature of 5.5°Cfor immediate radiotracer experiments on board the ship orfor microbiological studies in the laboratory 4 days later.

Radiotracer experiments. Studies of sulfur transformationswere done on different sediment cores with four different35S-labeled tracers: S042-, H2S, and S2032- with either theinner or the outer sulfur atom labeled. The tracers wereinjected horizontally into intact sediment cores at 1-cmdepth intervals. After incubation at the in situ temperaturefor 3, 7, or 18 h, the cores were cut into 1- or 2-cm segmentsand fixed with ice-cold Zn2+ (see below) for later determi-nation of 35S distributions in sulfate, thiosulfate, or Sred (Sred= HS- + FeS + SO + FeS2). Duplicate cores were analyzedfor each tracer and incubation period. Details of the ra-diotracer techniques are described elsewhere (16, 26).

[35S]sulfate reduction. Carrier-free 35 o42- (2 RI with <0.1nmol of S042-; 80 kBq; Isotope Laboratory, Ris0, Den-mark) was injected (22). After incubation, samples werefixed in 5 ml of cold 20% ZnCl2. The formation of reduced35S was analyzed by a single-step chromium reductionmethod in which all Sred is converted into H2S and trappedas ZnS (14, 41a, 47). In some cores, reduced 35S wasanalyzed separately in pools of acid volatile sulfide (AVS;mostly FeS), S0, and FeS2 (39). Elemental sulfur was firstextracted with CS2 from fixed samples. AVS was subse-quently solubilized as H2S in an acid solution at roomtemperature. By a final boiling with Cr2+ in acid, FeS2 wasextracted. Sulfate reduction rates were calculated as de-scribed by J0rgensen (22).

[35S]sulfide oxidation. To trap radiolabeled thiosulfate pro-duced during incubation by the oxidation of [35S]sulfide, weinjected a nonradioactive solution of thiosulfate (three injec-tions of 50 [lI of 15 mM Na2S203 at 1-cm depth intervals) intothe cores. The sediment was preincubated for 1 h while thethiosulfate gradually dissipated and approached a meanconcentration of 0.5 mM. The sulfide tracer was preparedfrom 35S0 as follows. 35S0 (Amersham), dissolved in acetone,was reduced to H235S by Cr2+ in an acid solution, and thetracer was trapped under N2 in 1 ml of 0.01 N NaOH. Thelabeled sulfide was immediately injected (20 ,ul; 5 ,uM; 100kBq per injection).

After incubation, samples were fixed in 5 ml of cold 0.2%ZnCl2 which contained 25 mM S2032- as a carrier. The fixedsamples were centrifuged, and a 100-,ul subsample was takenfrom the supernatant for the direct determination of 35S in alldissolved oxidation products. Another 100-,l subsamplewas injected into an ion chromatograph for separation of the35S-labeled oxyanions. Only peaks of S2032- and S042-were detectable at an instrument sensitivity of 1 puM for50-fold-diluted samples. These peak fractions were col-lected, and their radioactivities were determined. As acontrol of radioisotope recovery among the sulfur oxyan-ions, the sum of 35S2032- and 35 o42- accounted for 90 to105% of the radioactivity of 35S in all dissolved oxidationproducts (obtained from the direct determination), indicatingthat other, undetected oxyanions could not carry significantamounts of radioactivity. The amount of 35S remaining in

Sred pools was determined by single-step chromium distilla-tion of Sred (14) and determination of the radioactivity in thetrapped Zn35S. In some samples, separate determinations of35S in AVS, S, and FeS2 were made (20, 39).

[35S]thiosulfate transformations. Injections were done intwo sets of cores with 35S2032- (5 RI of 10 mM Na2S203; 80kBq; Amersham) in which either the inner or the outer sulfuratom was labeled. After incubation, radioisotope separationswere done as described for the H235S experiments.

Cultivation of bacteria. (i) Medium and growth conditions.Anaerobic sulfur compound-utilizing bacteria were enumer-ated and cultivated in a bicarbonate-buffered, sulfide-re-duced mineral medium containing (per liter) the following:NaCl, 15.0 g; MgCl2 .6H20, 2.0 g; CaCl2- 2H20, 0.2 g;KCl, 0.5 g; NH4Cl, 0.1 g; KH2PO4, 0.2 g; NaHCO3, 1.75 g;Na2S 8H20, 0.18 g; trace element solution, 2 ml (42); andvitamin solution, 1 ml (44). The pH was adjusted to 7.2. Themedium was prepared under an 02-free N2 atmosphere asdescribed by Widdel and Pfennig (44). Substrates (electrondonors and acceptors) were added from sterile 0.5 to 2 Mstock solutions before inoculation. H2 was used in a mixturewith CO2 (H2-CO2, 5:1 [vol/vol]) in the gas phases of tubes orbottles sealed with black rubber stoppers (liquid-gas, 2:3[vol/vol]). The Hungate technique was used for gassing ofculture vessels. Cultures with H2 were incubated horizon-tally to obtain a large surface to facilitate gas exchange. Allcultures received about 100 puM Na2S204 as an additionalreducing agent prior to inoculation. All incubations weredone at 19°C in the dark. Growth tests were conducted with20-ml screw-cap- or stopper-sealed tubes. Growth was mon-itored by measuring the turbidity at 500 nm in a spectropho-tometer.

(ii) Isolation and maintenance of strains. Pure cultures wereobtained by repeated application of deep-agar dilutions asdescribed by Widdel and Pfennig (45). Isolated colonies werechecked for purity with microscopic controls and withgrowth tests in complex medium containing yeast extract(0.25%), pyruvate (5 mM), glucose (5 mM), and fumarate (5mM). Stock cultures were kept at 4°C in the dark andtransferred to fresh medium at monthly intervals.

(iii) Enumeration procedures. Numbers of viable sulfate-or thiosulfate-utilizing bacteria were estimated with most-probable-number (MPN) dilutions in liquid medium (3). Thedilution series were prepared as described by Bak andPfennig (Sb). Samples containing sulfate but no electrondonors served as controls. The counting tubes were incu-bated at 19°C in the dark for a maximum of 4 months andchecked every week for growth. Grown cultures were testedfor H2S production, and cultures expected to contain thio-sulfate-disproportionating bacteria were also checked forS042- formation. Three different sets of enumerations weremade: (i) enumeration of H2-oxidizing SRB or thiosulfate-reducing bacteria with the substrates H2 (1.0 mM acetate asthe carbon source), 5.0 mM S042-, and 5.0 mM S2032-7 (ii)enumeration of acetate-oxidizing SRB or thiosulfate-reduc-ing bacteria with the substrates 10.0 mM acetate, 5.0 mMS042-, and 5.0 mM S2032-; and (iii) enumeration of thiosul-fate-disproportionating bacteria with the substrate 10.0 mMS2032- (1.0 mM acetate as the carbon source). Enumera-tions were carried out for four depth intervals: 0 to 1, 1 to 2,2 to 4, and 4 to 6 cm.Chemical determinations. (i) Pore water. Pore water from

the sediment was obtained by pressure filtration through0.45-,um-pore membrane filters under N2. Sulfate was deter-mined by nonsuppressed ion chromatography (Waters).Samples for sulfide determinations were collected without

APPL. ENVIRON. MICROBIOL.


http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/

THIOSULFATE TRANSFORMATIONS AND SULFATE REDUCTION

So4 red. (nmoL cm-3 d-1)0 5 10 15

2_

6_

8_

10

mM so0240 5 10 15 20 25

pmol S cm-3

20

AVS 0So 0 FeS2 E AVS S So 0 FeS2

FIG. 1. (A) Sulfate reduction (red.) rates in the upper 0 to 10 cm of the sediment as determined from radiotracer experiments. The barsshow the total reduction rates and the distribution of the reduced label in the pools of AVS, elemental sulfur (S0), and pyrite (FeS2). The dataare mean values for two cores incubated for 3 h and two cores incubated for 18 h. d, Day. (B) Distributions of sulfate in the pore water andof the particulate sulfur compounds in the sediment (AVS, elemental sulfur, and pyrite). The broken line indicates the mean gradient of totalparticulate sulfur accumulation with depth.

contact to the atmosphere in 0.2% ZnCl2, and the amount oftrapped ZnS was determined by the spectrophotometricmethylene blue method (9).

(ii) Sediment. Elemental sulfur was extracted with CS2from ZnCl2-fixed sediment samples and analyzed spectro-photometrically after cyanolysis and complexation with iron(40). The amounts of AVS and chromium-reducible sulfidewere determined from the ZnS traps after distillation of theradiolabeled pools described above. The amount of ZnS was

determined by the methylene blue method (9) after appro-

priate dilutions. The FeS2 concentration was calculated bydifference from chromium reducible sulfur (CRS) minus S5and AVS.

(iii) Cultures. The amount of sulfate in bacterial cultureswas determined by the turbidimetric BaSO4 precipitationmethod (37). Sulfide was analyzed photometrically afterreaction with a CuSO4-HCl reagent by rapidly measuring thelight extinction at 480 nm caused by the CuS precipitate (11).

RESULTSSediment. The sediment consisted of sandy and silty mud

with a mean organic content of 4.4% (dry weight). Oxygenpenetrated to a depth of 9 mm below the sediment surface(35a). The sediment was brownish and oxidized in the upper0 to 2 cm, below which the color shifted to black or darkgrey. Free sulfide was not detectable (<1 ,uM) in the porewater of the upper 0 to 10 cm. The sediment and theoverlying seawater had a temperature of 5.5°C, and thesalinity was 32.4%o. The oxygen concentration of the seawa-

ter was 80% of air saturation.Sulfate reduction. The depth distribution of sulfate reduc-

tion rates is shown in Fig. 1A. The highest rates, 16 to 19nmol of s042- cm-3 day-', were found in the reducedsediment layers below 2 cm. Significant activity of the SRB

was, however, also observed in the suboxic zone at a 1- to2-cm depth and even in the oxic zone at a 0- to 1-cm depth.AVS was not detectable in the oxic zone (Fig. 1B). Sulfatereduction could therefore be detected from the 35S incuba-tions only because the reduced label was trapped in elemen-tal sulfur and pyrite. Only below 2 cm was a part of thelabeled sulfide recovered from the AVS fraction. On aver-

age, for the 0- to 10-cm depth interval studied, 31% of thereduced label was found in AVS, 33% was found in S, and36% was found in FeS2.AVS accumulated in the sediment below the oxidized

zone. Elemental sulfur constituted a small fraction of theparticulate sulfur compounds at all depths, while pyrite was

quantitatively the dominant sulfur pool. The sulfate concen-

tration had a slight peak at a 2-cm depth, below which theconcentration gradually decreased (Fig. 1B). The peak coin-cided with the redox transition at a 2-cm depth, which was,as expected, the zone in which the most intensive oxidationof Sred compounds to sulfate took place.

Sulfide oxidation. The maximal degree of [35S]sulfide oxi-dation, 46% of the injected tracer, was found near the redoxtransition at a 2-cm depth (Fig. 2A). [35S]sulfide oxidationwas less intense in the oxic zone, but even in the reducedsediment it proceeded down to a 10-cm dep.h. Thiosulfatewas the main product, constituting up to 85% of the oxidized35S products after 3 h of incubation (Fig. 2B). The mostefficient oxidation to sulfate was found in the oxic surfacelayer. During prolonged incubation of up to 18 h, some of thethiosulfate was oxidized further to sulfate.

Thiosulfate transformations. The time course of [35S]thio-sulfate transformation to 35S042- and 35Sred in the sedimentcores is shown in Fig. 3 for the uppermost 0 to 1 cm. The

rates of thiosulfate depletion were constant and similar in

experiments with the inner or the outer sulfur atom labeled.

E

-C-I-

a,D)

VOL. 57, 1991 849


http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/


% of H35S oxidized20 10 20 30 40 50

2 - /07//M32

4 -

6 21s203

8 0mso;2ic4

S203 formation (% of SOX)

FIG. 2. Sulfide oxidation to thiosulfate and sulfate in the sediment. (A) The bars show the percentage of injected H235S which was oxidizedto the two products. The data are mean values for 3-, 7.5-, and 18-h incubations. (B) Relative contribution of thiosulfate to the products ofH235S oxidation (S0x) after 3 and 18 h.

The "S-labeled products of thiosulfate differed strongly,however, between the two sets of experiments. When theinner sulfur atom was labeled (Fig. 3A), the "S-labeledproducts were 16% Sred and 84% S042-. Since a conversionof the oxidized, inner sulfur atom to sulfide represents a netreduction, 16% of the thiosulfate was reduced. When theouter sulfur atom was labeled (Fig. 3B), the "S-labeledproducts were 76% Sred and 24% S042. Since a conversionof the reduced, outer sulfur atom to sulfate represents a netoxidation, 24% of the thiosulfate was oxidized. The remain-ing 60% [100 - (16 + 24)] of the thiosulfate was metabolized,not by net reduction to sulfide or net oxidation to sulfate, butby disproportionation concurrently to sulfide and sulfate (24,26).A larger fraction of thiosulfate was transformed in the oxic

zone than in the reduced sediment (Fig. 4). The formation of35sred and 35042- from thiosulfate with the inner and outersulfur atoms labeled also changed with depth in the sedi-ment. The formation of 31SO42- decreased with depth withboth tracers, while the formation of 35Sred increased.

100

80 2-\S203-

60 o\ Zo-

mZ 40~0 _\-_A2-

Like the calculations from the time course experiments,the data in Fig. 4A and B were used to calculate the depthdistributions of thiosulfate reduction, oxidation, and dispro-portionation. The total rates of thiosulfate transformationdecreased threefold between the surface and the reducingzone (Fig. 5A). The relative contribution of thiosulfatereduction increased strongly with depth (Fig. 5B). Thedistribution of thiosulfate oxidation was similar to that ofsulfide oxidation, with a maximum near the redox transitionat a 2-cm depth. Disproportionation decreased slightly withdepth and was overall the most important pathway ofthiosulfate transformation.

Bacteria. All bacteria observed during the enumerationswere SRB. Other dissimilatory sulfur compound-utilizingbacteria did not occur. In some cases, it took up to 3 monthsfor the count to fully develop. The total numbers of SRBrecorded (on H2, acetate, and thiosulfate) are shown in Fig.6, together with the sulfate reduction rates. Maximum pop-ulation densities of SRB, 2 x 106 cells cm-3, were surpris-ingly found in the uppermost 0 to 2 cm, where the sediment

80

601

401

201

Time (h)

0 5 10 15 20

Time (h)FIG. 3. Time course of thiosulfate transformation to sulfate and Sred in the uppermost 0 to 1 cm of the sediment. (A) Inner atom of

thiosulfate labeled with 35S. (B) Outer atom of thiosulfate labeled with 35S.

E

cLacCD

0 0S

-o \S-Sred

0 2-

I



http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/


35s 02 (%)

E

-C.4-C.ciuCD

60 70 80 90 100I

2 -

Sred

4

6 0 o

8 0 S-

if' II

35s 02- (%)

10 20 30 40 10 20 30 40

H35 35so2 (%) H35S, 35so2- M)FIG. 4. Depth distribution of thiosulfate transformation to sulfate and Sred after 18 h of incubation. Upper scales show the percent label

remaining in thiosulfate after incubation. Lower scales show the percent label present in S042- or Sred after incubation. (A) Inner atom ofthiosulfate labeled with 35S. (B) Outer atom of thiosulfate labeled with 35S.

was oxidized and the measured sulfate reduction rates wererelatively low.

H2-utilizing SRB. The numbers of H2-oxidizing SRB de-creased from 1.1 x 106 cells cm-3 in the uppermost layer toabout 100-fold-lower numbers at a 4- to 6-cm depth (Fig. 7).Two morphological types dominated: small, mostly motileDesulfovibrio species (Fig. 8A, arrow) and relatively large,motile rods with gas vacuoles (Fig. 8B). The latter specieswere numerically dominant but occurred only in MPNcounts from the upper 0 to 2 cm, whereas Desulfovibriospecies were present at 104 to 105 cells cm-3 at all depths.Four representative Desulfovibrio strains and one gas-vacu-olated SRB (strain WK1) were isolated in pure cultures andtested for sulfate and thiosulfate utilization. All five strains

s2o02 transformations (%)

reduced sulfate or thiosulfate to sulfide with H2 as theelectron donor. None of the strains was able to dispropor-tionate thiosulfate or to grow on H2 in the absence of sulfateor thiosulfate. The gas-vacuolated strain, WK1, isolatedfrom the highest growth-positive dilution of the H2 enumer-ation series, was moderately psychrophilic. The strain grewwell at 19°C but not at 25°C. Strain WK1 efficiently utilizedH2, lactate, malate, and pyruvate as electron donors forsulfate reduction. Older cultures of the Desulfovibrio speciesoften formed coccoid bodies (Fig. 8A), especially uponprolonged storage.

Acetate-utilizing SRB. The numbers of SRB growing onacetate were highest in the 0- to 2-cm interval and decreasedsignificantly in deeper layers (Fig. 7). Oval or slightly curved

2-S203 transformation (%)

E O4X ; RediOisp.ca.

o 6M Red. 6

oxi.8 8-

ElO isp.

FIG. 5. Depth distributions of thiosulfate transformation by reduction (Red.), oxidation (Oxi.), and disproportionation (Disp.) after 18 hof incubation. (A) Thiosulfate transformation calculated as a percentage of the total injected 35S2032- pool. (B) Relative contributions of thethree pathways calculated as percentages of the 35S2032- transformed (reprinted [redrawn] from Science [25] with permission of thepublisher).

VOL. 57, 1991 851


http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/


2- 3So4 reduction (nmol cmr3 d-1 )

0 5 10 15 20 1o2I I I

MPN: bacteria cm-3

Eu

-Cau

,7

E

4-

ca

1

2

4

FIG. 6. Total numbers of SRB, as determined by the MPNtechnique, on acetate or H2 substrate at different depths in Kattegatsediment. The concurrently measured sulfate reduction rates areshown for comparison. The depth of oxygen penetration into thesediment (0.9 cm) and the transition depth between the brown,oxidized surface sediment and the black, reduced sediment areshown. d, Day.

cells, morphologically resembling SRB of the genus Desulfo-bacter or Desulfobacterium, developed. In the highestgrowth-positive dilutions of counts from the 0- to 1- and 1- to2-cm zones, bacteria containing gas vacuoles were againobserved (Fig. 8C). Most SRB, which developed in enumer-ations with acetate, had the tendency to grow in chains ordense clumps. Higher dilutions sometimes showed weak butsignificant H2S production (1 to 2 mM) without an increase inturbidity. Upon careful observation, however, a thin bacte-rial film or small colonies on the glass wall were detectable,mainly in the bottom portions of the culture tubes. Thebacteria within the film morphologically resembled smallerSRB of the genus Desulfobacter or Desulfobacterium. Inone of the tubes from the 2- to 4-cm interval, an undescribed,thick, spirilloid SRB developed (Fig. 8D).Two Desulfobacter-like species from higher dilutions (0-

to 1- and 2- to 4-cm depths) were subcultured and tested forsulfur compound utilization. With 15 mM acetate as theelectron donor, both cultures reduced sulfate and thiosul-fate. In the presence of only small amounts of acetate (1mM), both cultures also disproportionated thiosulfate. Thio-sulfate disproportionation was, however, not coupled togrowth.

Thiosulfate-disproportionating SRB. The estimated num-bers of thiosulfate-disproportionating bacteria were about10-fold lower than were the numbers of H2- or acetate-utilizing SRB at all depths (Fig. 7). With one exception (onetube of a 10-1 dilution from the 0- to 1-cm depth), only asmall fraction (10 to 20%) of the added thiosulfate wasdisproportionated and significant growth could not be ob-served. Only small flocs or clumps consisting of Desulfobac-

MPN: bacteria cm 3

103 104 105 106

a H2 [1 Acetate 13 s 02-2 ~~~23FIG. 7. Depth distributions of H2-utilizing, acetate-utilizing, and

thiosulfate-disproportionating SRB, as determined by the MPNtechnique.

ter- or Desulfobacterium-like SRB were observed in thebottom portions of sulfate-positive tubes. When the bacteriawere transferred into fresh medium, growth remained weakunless the amount of acetate was increased.

DISCUSSIONSulfate reduction. The depth distribution of sulfate reduc-

tion showed a typical rate maximum in the uppermost part ofthe black, reduced sediment (Fig. 1A). There was, however,also rather high sulfate reduction in the oxidized (Eh >0 mV)sediment above a 2-cm depth and even in the oxic (02-containing) sediment above a 0.9-cm depth. Similar obser-vations of sulfate reduction in oxidized surface layers havebeen made in other marine and freshwater sediments (Sb,21). In the absence of free sulfide or AVS in the oxidizedsurface sediment, the radioactivity of the H2S formed was allretained in the less reactive sulfur compounds, elementalsulfur and pyrite (8, 39). Because of the oxidizing conditions,a significant but unknown fraction of the sulfide formed fromsulfate reduction may have been rapidly reoxidized withinthe uppermost 1 to 2 cm; thus, the measured sulfate reduc-tion rates may have been underestimated.The potential for high sulfate reduction under oxic condi-

tions was emphasized by the observation of about 50-fold-higher population densities (MPN counts) of SRB at the 0- to1-cm depth than at the 3- to 5-cm depth (Fig. 6). There areseveral possible explanations for this distribution. The SRBmay be able not only to tolerate oxygen for prolongedperiods (12) but also perhaps to carry out sulfate reduction inthe presence of oxygen. The latter has been observed in02-supersaturated surface layers of photosynthetically ac-tive cyanobacterial mats (10). It is also possible that the SRBin the surface sediment are clustered within anoxic andreducing microenvironments (21). Because of diffusive andmetabolic constraints, however, such microenvironmentswould be expected to have a characteristic size of severaltens or hundreds of micrometers. Although they should thus

...:

3ZZYZZA

r



http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/


be easy to recognize during measurements with oxygenmicroelectrodes, we have not encountered such microenvi-ronments during our studies. Finally, a versatile metabolismhas been demonstrated among the SRB in recent years,including respiration with nitrate or even oxygen (13, 33).Although such respirations could help to explain the ob-served bacterial distribution, the quantitative importance ofthese metabolic capabilities in sediments is still unknown.

Sulfide oxidation. Oxidation of sulfide was directly dem-onstrated by the use of H235S in both the oxidized surfacesediment and the black, reduced sediment at a 2- to 10-cmdepth (Fig. 2A). A comparison of the high sulfate reductionrates and the slow accumulation of Sred in both zonessimilarly showed that intensive reoxidation of sulfide musthave taken place. Sulfate reduction in the upper 0 to 10 cmwas 2.15 mmol m-2 day-1. The mean accumulation of Sredwith depth was 4 ,umol of S cm-3 per cm or 40 ,umol of Scm-3 over a 0- to 10-cm depth (Fig. 1B, broken line). With amean sediment accumulation rate of 1 mm year-1 for thearea (31a), the rate of sulfide burial over the top 10 cm was0.11 mmol m-2 day-'. Thus, only 5% [(0.11 x 100)/2.15] ofthe sulfide produced from sulfate reduction was retained asFeS2, FeS, and So. The remaining 95% was reoxidized.Since there was no detectable sulfide in the pore water,reoxidation cannot be explained by diffusion of sulfide up tothe oxic and oxidized surface layers. It is therefore inaccordance with these mass balance considerations thatreoxidation of sulfide was found to take place throughout thesediment column.Anoxic reoxidation of [35S]sulfide also has been directly

demonstrated in other sediments from both marine andfreshwater localities (16, 26). The degree of reoxidationvaries from between 60 and 97% of all sulfide produced fromsulfate reduction in the area between the Baltic Sea and theNorth Sea (27). In Lake Constance sediment, 98% of thesulfate used by the SRB was derived from reoxidized sulfideand only 2% was derived from "new" sulfate diffusing downfrom the overlying lake water (Sb, 5c). The oxidants foranoxic sulfide oxidation have not been directly identified.However, oxidized iron and manganese minerals may playan important role (1, 2, 24, 29).The addition of nonlabeled thiosulfate, which was neces-

sary to transiently trap 35S2033- formed from [35S]sulfide,was not expected to have changed the relative pathways ofsulfide oxidation. Thiosulfate addition increased the rate ofsulfide formation, especially from disproportionation in theoxidized zone. The increase was, however, less than twofoldthroughout most of the sediment and could therefore nothave changed the chemistry of the sediment significantlyduring the short incubations.

It was an experimental problem in our earlier radiotracerstudies of sulfide oxidation that only a small fraction of theradiolabel was found in more oxidized sulfur compounds,even in the oxidized surface layers of the sediment. Thereason for this was that the radiolabel was rapidly trans-ferred from H235S into other reduced, but unreactive, sulfur

FIG. 8. Photomicrographs of undescribed SRB observed inMPN enumerations. Bars: 10,um in A and C and 5,um in B and D.(A) Coccoid bodies observed in a 2-week-old culture of an H2-oxidizing Desulfovibrio sp. The arrow shows an intact Desulfovibriocell. (B) Strain WK1, an H2-oxidizing SRB with gas vacuoles. Thearrow shows a spherical cell which had just exploded. (C) SRB withgas vacuoles observed in enumerations with acetate. (D) Acetate-utilizing, sulfate-reducing spirillum.

VOL. 57, 1991 853


http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/


pools, such as S or FeS, by isotopic exchange (15, 16a). Thespecific radioactivity of the reacting sulfide thereby becamevery low and the radiotracer transformations could not beused for calculations of the amount of sulfide oxidized. Anoxidation of up to 46%, as found in the present study (Fig.2A), is the highest that we have recorded so far. The factorscontrolling isotopic exchange are not quantitatively under-stood, but the presence of polysulfides seems to be highlycatalytic for the exchange (15).

Isotopic exchange may also be an important mechanismfor the formation of 35S-labeled S° and perhaps FeS2 in theoxidized surface sediment during sulfate reduction measure-ments (Fig. 1A). Isotopic exchange into more stable sulfurpools is thus a mechanism which transiently preserves 35S inreduced form and helps to reveal in situ rates of sulfatereduction despite rapid reoxidation of the H2S formed.

Thiosulfate transformations. The rates of thiosulfate deple-tion were similar in the experiments with the inner or theouter sulfur atom labeled (Fig. 3), indicating, as should beexpected, that the two radioactive thiosulfate pools were

similarly metabolized by the bacteria and that isotopicexchange reactions, which may occur between the outersulfur atom and sulfide at higher temperatures (41), were notof significance. Thiosulfate was the main product of sulfideoxidation throughout the sediment (Fig. 2), in accordancewith radiotracer results from other marine and freshwatersediments (16, 26) and with observations that thiosulfate isone of the main products from the chemical oxidation ofH2S, FeS, and FeS2 (7, 18, 34).

Thiosulfate reduction occurred throughout the sedimentbut became a relatively more important pathway in thedeeper layers (Fig. 5). All of the SRB strains from thesediment could also reduce thiosulfate. Bacteria utilizingthiosulfate but not sulfate, e.g., Campylobacter spp., were

not observed. Consequently, SRB appeared to be the prin-ciple thiosulfate-reducing bacteria in the sediment. Thiosul-fate is generally a preferred electron donor in these bacteriabecause, unlike sulfate, it does not require ATP-consumingactivation. The addition of thiosulfate to freshwater sedi-ment slurries has accordingly been found to significantlyinhibit sulfate reduction (26).The oxidation of thiosulfate to sulfate in all sediment

layers, with a maximum at the redox transition, is consistentwith the observation of a similar distribution of sulfideoxidation (Fig. 5). About 50% of the thiosulfate in the 0- to10-cm layer was, however, disproportionated to H2S andS042-. The oxidized, inner sulfur atom was thereby trans-formed to S042- while the reduced, outer sulfur atom was

transformed to H2S. An H2S-S2032- cycle in which half ofthe sulfur atoms remained at the oxidation state of -2 was

thus established. Since thiosulfate was the main product ofsulfide oxidation, this internal recycling of sulfur is quanti-tatively an important pathway of the sulfur cycle. Such a

thiosulfate shunt has now been found in a range of marineand freshwater sediments (16, 26).

Viable counts of SRB. The population sizes of SRB in thesediment were studied by an MPN technique. Incubationswere done at an ecologically realistic temperature, 19°C, andthe growth of bacteria was checked weekly for up to 4months at the highest dilutions. During recent studies withfreshwater sediment, the MPN enumeration technique usedhere proved to be more efficient (i.e., yielded higher recov-

eries of SRB) than previously applied techniques, such as

deep'agar dilution with complex media and an Fe2+ indica-tor for sulfide production (Sb). With the MPN technique it ispossible to overcome the problem inherent in the deep-agar

method, in which fast-growing colonies tend to overgrowmore slowly growing organisms, which are, as expected,more abundant, and to blacken the agar. In the MPNtechnique, even bacteria with doubling times of several daysmay eventually be recognized. Densities of up to 2 x 106SRB cm-3 were found in the oxidized sediment zone (Fip.6). These densities are 1 to 3 orders of magnitude higher thahthose previously reported for SRB from marine sedimentts(16, 23).A comparison of the measured population sizes of SRB

and the measured in situ rates of sulfate reduction indicatedthat the bacteria were respiring at specific rates of 3 x 10-15to 5 x 10-l' mol cell-1 day-' in the upper 0 to 2 cm. Theserates are in the same range as the calculated specific respi-ration rates for SRB growing in pure cultures on H2, lactate,or pyruvate: 0.2 x 10-15 to 50 x 10-15 mol cell-1 day-' (23).The bacterial densities are thus high enough to potentiallyaccount for their metabolism in the present study. Thespecific respiration rates below 2 cm were 100-fold lower,while those calculated from several earlier studies wereabout 1,000-fold lower. We therefore conclude that theapplied MPN technique has improved the viable countingtechniques for SRB in sediments by 10 to 1,000 fold. The cellnumbers determined are still highly underestimated in thedeeper sediment layers, but in the surface sediment the cellniumbers are approaching a realistic range as compared withthe measured activity of the bacteria, mnaking the viablecounts much more interesting with respect to characteriza-tion of the different species of SRB. Furthermore, the typesof bacteria that develop in the MPN tubes are different fromthose that develop in the deep-agar tubes. We thus concludethat these viable counting techniques can be very useful forthe quantification of natural populations of SRB and for theirphysiological characterization. It is necessary that countingtechniques and media that allow the bacteria to grow underecologically realistic conditions are established. Bak andPfennig (Sb) found similarly high recoveries of SRB infreshwater sediments.SRB. It was an important new result that gas-vacuolated

types were the most abundant SRB in the upper 0 to 2 cm ofthe sediment. These organisms dominated in the highestpositive dilutions of the MPN counts, both with acetate andwith H2 as electron donors. In marine enrichment cultures ofSRB, which are incubated at low temperatures (4 to 5°C),such gas-vacuolated types often develop (20a). Qnly at lowerdilutions on H2 did the classical, fast-growing Desulfovibriospecies dominate in abundances which were similar to thosereported in earlier studies (17, 23). Species with gas vacuolesmust thus be among the dominant SRB in the oxic andoxidized layers of the sediment. So far, only one straincontaining gas vacuoles has been tentatively named Desulfo-bacterium vacuolatum (43), but this strain has not yet beendescribed.Although morphologically clearly different from all SRB

so far described, strain WK1 resembled physiologically theclassical H2-oxidizing Desulfovibrio species. A further un-usual feature was the formation of spherical structures incultures of strain WK1 (Fig. 8B). These structures wereeither free swimming or attached to cells of strain WK1. Theformation of coccoid bodies in older cultures of the Desul-fovibrio species (Fig. 8A) was a similar phenomenon and iscurrently being studied.

Several other undescribed SRB morphologies, e.g., thespirilla shown in Fig. 8D, developed in the acetate MPNtubes. Moreover, acetate-oxidizing SRB were at least asabundant as H2-oxidizing SRB. This result is in contrast to



http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/


those for freshwater sediments, in which SRB physiologi-cally resembling the H2-oxidizing Desulfovibrio species weremore abundant, or to those for Wadden Sea sediments, inwhich lactate oxidizers (and probably the classical Desul-fovibrio species) were found in the highest numbers (Sb, 31).Because of the lower recovery in earlier studies, however,these relative distributions may be influenced by the selec-tivity of the applied counting procedures as much as by theactual abundances of the physiological types of SRB.The present study thus demonstrates directly for the first

time that acetate-oxidizing SRB dominate in marine sedi-ments. A predominance of acetate oxidizers is consistentwith observations of the compositions of fatty acid biomar-kers of SRB in marine sediments (38). It is also in accor-dance with repeated demonstrations of acetate as the mainelectron donor for sulfate reduction in marine sediments(e.g., 6, 35, 36, 46).

Thiosulfate-disproportionating bacteria. Thiosulfate-dis-proportionating bacteria were counted by an anaerobic MPNtechnique with thiosulfate as the energy substrate and asmall amount of acetate as the carbon source. The additionof acetate at a low concentration was necessary to promotethe growth of these chemolithoheterotrophic bacteria (5). Todiscriminate them from acetate-oxidizing SRB with S2032-as an electron acceptor, we found it necessary to check forthe simultaneous production of S042- and H2S.The anaerobic, thiosulfate-disproportionating bacteria iso-

lated from the marine sediment were distinctly different fromthe types that have been found in freshwater sediments (Sb).In contrast to the freshwater strains, the marine strainsappeared to be unable to grow by S2032- disproportionationalone. Obviously, S2032- was only cometabolized by thebacteria which grew marginally by acetate oxidation coupledto thiosulfate reduction. No specialized disproportionatingbacteria, such as Desulfovibrio sulfodismutans, which iscommonly present in enrichments from freshwater sources,were found (5). It seems that marine bacteria cannot couplethe disproportionation of sulfur compounds to energy con-servation. This idea is consistent with pure-culture studieswhich demonstrated that marine SRB are unable to grow bydisproportionation reactions (30). In enrichments from fresh-water sediments, only H2-utilizing SRB could disproportion-ate thiosulfate (Sa). The reason for these differences betweenmarine and freshwater strains is yet unexplained.

It was found in the present study that all of the SRB wereable to use S2032- as an electron acceptor and that all of theS2032-_reducing bacteria were SRB. The depth distributionsof SRB and thiosulfate-reducing bacteria were thus identical.It was furthermore demonstrated that SRB, which candisproportionate thiosulfate, may carry out this reactionconcurrently with the reduction of thiosulfate in the pres-ence of suitable electron donors (30). It has also been shownrecently that some SRB are able to oxidize sulfide andthiosulfate during aerobic respiration in the presence ofoxygen (13). It is therefore likely that it is largely the samegroups of bacteria which carry out the different oxidativeand reductive transformations of sulfur compounds in thesediment. In conclusion, it appears that SRB have a verybroad metabolic capacity and that they may play a muchmore diverse role in sediments than we have previouslythought.

ACKNOWLEDGMENTSWe thank Dorthe T. Thomsen for assistance and Birger Kruse for

an invitation to join a cruise on RV Gunnar Thorson in Kattegat,Denmark.

Financial support was provided by the Danish Natural ScienceResearch Council and by the National Agency of EnvironmentalProtection. A research fellowship (to F.B.) was provided by theGerman Science Foundation.

REFERENCES

1. Aller, R. C., J. E. Mackin, and R. T. Cox, Jr. 1986. Diagenesisof Fe and S in Amazon inner shelf muds: apparent dominance ofFe reduction and implications for the genesis of ironstones.Cont. Shelf Res. 6:263-289.

2. Aller, R. C., and P. D. Rude. 1988. Complete oxidation of solidphase sulfides by manganese and bacteria in anoxic marinesediments. Geochim. Cosmochim. Acta 52:751-765.

3. American Public Health Association. 1969. Standard methods forthe examination of water and wastewater, including bottomsediments and sludge, p. 604-609. American Public HealthAssociation, Washington, D.C.

4. Bak, F., and H. Cypionka. 1987. A novel type of energymetabolism involving fermentation of inorganic sulphur com-pounds. Nature (London) 326:891-892.

5. Bak, F., and N. Pfennig. 1987. Chemolithotrophic growth ofDesulfovibrio sulfodismutans sp. nov. by disproportionation ofinorganic sulfur compounds. Arch. Microbiol. 147:184-189.

5a.Bak, F. 1988. Ph.D. thesis. University of Konstanz, Konstanz,Federal Republic of Germany.

5b.Bak, F., and N. Pfennig. FEMS Microbiol. Ecol., in press.5c.Bak, F., G. Scheff, and K. H. Jansen. FEMS Microbiol. Ecol., in

press.6. Banat, I. M., E. B. Lindstrom, D. B. Nedwell, and M. T. Balba.

1981. Evidence for coexistence of two distinct functional groupsof sulfate-reducing bacteria in salt marsh sediment. Appl. En-viron. Microbiol. 42:985-992.

7. Chen, K. Y., and J. C. Morris. 1972. Kinetics of oxidation ofaqueous sulfide by 02. Environ. Sci. Technol. 6:529-537.

8. Christensen, J. P. 1989. Sulphate reduction and carbon oxida-tion rates in continental shelf sediments, an examination ofoffshelf carbon transport. Cont. Shelf Res. 9:223-246.

9. Cline, J. D. 1969. Spectrophotometric determination of hydro-gen sulfide in natural waters. Limnol. Oceanogr. 14:454-458.

10. Cohen, Y. 1989. Photosynthesis in cyanobacterial mats and itsrelation to the sulfur cycle: a model for microbial sulfur in-teractions, p. 22-36. In Y. Cohen and E. Rosenberg (ed.),Microbial mats. Physiological ecology of benthic microbialcommunities. American Society for Microbiology, Washington,D.C.

11. Cord-Ruwisch, R. 1985. A quick method for the determinationof dissolved and precipitated sulfides in cultures of sulfate-reducing bacteria. J. Microbiol. Methods 4:33-36.

12. Cypionka, H., F. Widdel, and N. Pfennig. 1985. Survival ofsulfate-reducing bacteria after oxygen stress, and growth insulfate-free oxygen-sulfide gradients. FEMS Microbiol. Ecol.31:39-45.

13. Dilling, W., and H. Cypionka. 1990. Aerobic respiration insulfate-reducing bacteria. FEMS Microbiol. Lett. 71:123-128.

14. Fossing, H., and B. B. J0rgensen. 1989. Measurement of bacte-rial sulfate reduction in sediments: evaluation of a single-stepchromium reduction method. Biogeochemistry 8:205-222.

15. Fossing, H., and B. B. J0rgensen. 1990. Isotope exchangereactions with radiolabeled sulfur compounds in anoxic seawa-ter. Biogeochemistry 9:223-245.

16. Fossing, H., and B. B. J0rgensen. 1990. Oxidation and reductionof radiolabeled inorganic sulfur compounds in an estuarinesediment, Kysing Fjord, Denmark. Geochim. Cosmochim. Acta54:2731-2742.

16a.Fossing, H., and B. B. J0rgensen. Unpublished data.17. Gibson, G. R., R. J. Parkes, and R. A. Herbert. 1987. Evaluation

of viable counting procedures for the enumeration of sulfate-reducing bacteria in estuarine sediments. J. Microbiol. Methods7:201-210.

18. Goldhaber, M. B. 1983. Experimental study of metastable sulfuroxyanion formation during pyrite oxidation at pH 6-9 and 30°C.Am. J. Sci. 283:193-217.

19. Howarth, R. W. 1984. The ecological significance of sulfur in the

VOL. 57, 1991 855


http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/


energy dynamics of salt marsh and coastal marine sediments.Biogeochemistry 1:5-27.

20. Howarth, R. W., and B. B. J0rgensen. 1984. Formation of35S-labelled elemental sulfur and pyrite in coastal marine sedi-ments (Limfjorden and Kysing Fjord, Denmark) during short-term 35s042- reduction measurements. Geochim. Cosmochim.Acta 48:1807-1818.

20a.Isaksen, M., and F. Bak. Unpublished observations.21. J0rgensen, B. B. 1977. Bacterial sulfate reduction within re-

duced microniches of oxidized marine sediments. Mar. Biol.41:7-17.

22. J0rgensen, B. B. 1978. A comparison of methods for thequantification of bacterial sulfate reduction in coastal marinesediments. I. Measurement with radiotracer techniques. Geomi-crobiol. J. 1:11-28.

23. J0rgensen, B. B. 1978. A comparison of methods for thequantification of bacterial sulfate reduction in coastal marinesediments. III. Estimation from chemical and bacteriologicalfield data. Geomicrobiol. J. 1:49-64.

24. J0rgensen, B. B. 1988. Ecology of the sulphur cycle. Symp. Soc.Gen. Microbiol. 42:31-63.

25. J0rgensen, B. B. 1990. A thiosulfate shunt in the sulfur cycle ofmarine sediments. Science 249:152-154.

26. J0rgensen, B. B. 1990. The sulfur cycle of fresh water sedi-ments: role of thiosulfate. Limnol. Oceanogr. 35:1329-1343.

27. J0rgensen, B. B., M. Bang, and T. H. Blackburn. 1990. Anaer-obic mineralization in marine sediments from the Baltic Sea-North Sea transition. Mar. Ecol. Prog. Ser. 59:39-54.

28. Kanneworff, E., and W. Nicolaisen. 1973. The "Haps", a framesupported bottom corer. Ophelia 10:119-128.

29. King, G. M. 1990. Effects of added manganic and ferric oxideson sulfate reduction and sulfide oxidation in intertidal sedi-ments. FEMS Microbiol. Ecol. 73:131-138.

30. Kramer, M., and H. Cypionka. 1989. Sulfate formation via ATPsulfurylase in thiosulfate- and sulfite-disproportionating bacte-ria. Arch. Microbiol. 151:232-237.

31. Laanbroek, H. J., and N. Pfennig. 1981. Oxidation of short-chain fatty acids by sulfate-reducing bacteria in freshwater andin marine sediments. Arch. Microbiol. 128:330-335.

31a.Larsen, B., and P. P. Madsen. 1986. Report. Marine PollutionLaboratory, Charlottenlund, Denmark.

32. Luther, G. W., III, T. M. Church, J. R. Scudlark, and M.Cosman. 1986. Inorganic and organic sulfur cycling in salt-marsh pore waters. Science 232:746-749.

33. McCready, R.- G. L., W. D. Gould, and F. D. Cook. 1983.Respiratory nitrate reduction by Desulfovibrio sp. Arch. Micro-biol. 135:182-185.

34. Moses, C. O., D. K. Nordstrom, J. S. Herman, and A. L. Mills.1987. Aqueous pyrite oxidation by dissolved oxygen and byferric iron. Geochim. Cosmochim. Acta 51:1561-1571.

35. Parkes, R. J., G. R. Gibson, I. Mueller-Harvey, W. J. Bucking-

ham, and R. A. Herbert. 1989. Determination of the substratesfor sulphate-reducing bacteria within marine and estuarine sed-iments with different rates of sulphate reduction. J. Gen. Micro-biol. 135:175-187.

35a.Rasmussen, Hanne. Personal communication.36. S0rensen, J., D. Christensen, and B. B. J0rgensen. 1981. Volatile

fatty acids and hydrogen as substrates for sulfate-reducingbacteria in anaerobic marine sediment. Appl. Environ. Micro-biol. 42:5-11.

37. Tabatabai, M. A. 1974. Determination of sulfate in watersamples. Sulphur Inst. J. 10:11-13.

38. Taylor, J., and R. J. Parkes. 1985. Identifying different popula-tions of sulphate-reducing bacteria within marine sedimentsystems, using fatty acid biomarkers. J. Gen. Microbiol. 131:631-642.

39. Thode-Andersen, S., and B. B. J0rgensen. 1989. Sulfate reduc-tion and the formation of 35S-labeled FeS, FeS2, and S incoastal marine sediments. Limnol. Oceanogr. 34:793-806.

40. Troelsen, H., and B. B. J0rgensen. 1982. Seasonal dynamics ofelemental sulfur in two coastal sediments. Estuarine CoastalShelf Sci. 15:255-266.

41. Uyama, F., H. Chiba, M. Kusakabe, and H. Sakai. 1985. Sulfurisotope exchange reactions in the aqueous system: thiosulfate-sulfide-sulfate at hydrothermal temperature. Geochem. J. 19:301-315.

41a.Westrich, J. T. 1983. Ph.D. thesis. Yale University, NewHaven, Conn.

42. Widdel, F. 1986. Growth of methanogenic bacteria in pureculture with 2-propanol and other alcohols as hydrogen donors.Appl. Environ. Microbiol. 51:1056-1062.

43. Widdel, F. 1988. Microbiology and ecology of sulfate- andsulfur-reducing bacteria, p. 469-585. In A. J. B. Zehnder (ed.),Biology of anaerobic microorganisms. John Wiley & Sons, Inc.,New York.

44. Widdel, F., and N. Pfennig. 1981. Studies on dissimilatorysulfate-reducing bacteria that decompose fatty acids. I. Isola-tion of new sulfate-reducing bacteria enriched with acetate fromsaline environments. Description of Desulfobacter postgateigen. nov., sp. nov. Arch. Microbiol. 129:395-400.

45. Widdel, F., and N. Pfennig. 1984. Dissimilatory sulfate- orsulfur-reducing bacteria, p. 663-679. In N. R. Krieg and J. G.Holt (ed.), Bergey's manual of systematic bacteriology, vol. 1.The Williams & Wilkins Co., Baltimore.

46. Winfrey, M. R., and D. M. Ward. 1983. Substrates for sulfatereduction and methane production in intertidal sediments. Appl.Environ. Microbiol. 45:193-199.

47. Zhabina, N. N., and I. I. Volkov. 1978. A method of determina-tion of various sulfur compounds in sea sediments and rocks, p.735-746. In W. E. Krumbein (ed.), Environmental biogeochem-istry and geomicrobiology, vol. 3. Ann Arbor Science Publish-ers, Ann Arbor, Mich.



http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/

pathways microbiology thiosulfate transformations sulfate ...pathways of the oxidative part of the...

Documents