APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Dec. 1981, p. 985-9920099-2240/81/120985-08$02.00/0

Vol. 42, No. 6

Evidence for Coexistence of Two Distinct Functional Groupsof Sulfate-Reducing Bacteria in Salt Marsh Sediment

IBRAHIM M. BANAT,' E. BORJE LINDSTROM,2 DAVID B. NEDWELL,'* AND

M. TALAAT BALBA'Department ofBiology, University ofEssex, Colchester, United Kingdom,' a,d Department ofMicrobiology,

University of Umea, S-901 87 Umed, Sweden2

Received 27 October 1980/Accepted 18 August 1981

Oxidation of acetate in salt marsh sediment was inhibited by the addition offluoroacetate, and also by the addition of molybdate, anrinhibitor of sulfate-reducing bacteria. Molybdate had no effect upon the metabolism of acetate in a

freshwater sediment in the absence of sulfate. The inhibitory effect of molybdateon acetate turnover in the marine sediment seemed to be because of its inhibitingsulfate-reducing bacteria which oxidized acetate to carbon dioxide. Sulfide was

not recovered from sediment in the presence of molybdate added as an inhibitorof sulfate-reducing bacteria, but sulfide was recovered quantitatively even in thepresence of molybdate by the addition of the strong reducing agent titaniumchloride before acidification of the sediment. Reduction of sulfate to sulfide bythe sulfate-reducing bacteria in the sediment was only partially inhibited byfluoroacetate, but completely inhibited by molybdate addition. This was inter-preted as showing the presence of two functional groups of sulfate-reducingbacteria-one group oxidizing acetate, and another group probably oxidizinghydrogen.

Sulfate-reducing bacteria (SRB) have untilrecently been regarded as metabolizing only anarrow range of organic electron donors, princi-pally lactate (17), although some at least areknown to utilize hydrogen (2, 4, 20). Abram andNedwell (1, 2) have shown that these hydrogen-utilizing SRB are capable of scavenging hydro-gen, causing the "hydrogen transfer" reactioninitially described (6) when methanogenic bac-teria were the principal hydrogen scavengers.The SRB outcompeted methanogenic bacteriafor the available hydrogen. In addition, it hasbeen demonstrated (5) that the SRB are thepredominant hydrogen scavengers in the sedi-ment of the Colne Point salt marsh (the site hasbeen described by Nedwell and Abram [12]). Asimilar observation has been reported by Mount-fort et al. (11) in New Zealand intertidal sedi-ments. The reported inhibition of methanogen-esis in the presence of active sulfate reduction infield situations (7, 15, 23) has therefore been atleast partially attributed to the SRB outcom-peting methanogens for available hydrogen (2,15).

It has previously been suggested (9, 11) thatthe SRB might be the major acetate oxidizers inat least some anaerobic sediments, and Widdeland Pfennig (22) have already described an ace-tate.oxidizing SRB named Desulfotomaculumacetoxidans. To investigate whether the SRB

were significant acetate oxidizers in the ColnePoint salt marsh sediment, the following exper-iments were carried out.

MATERIALS AND MEETHODSProducts of acetate metabolism in salt marsh

sediments. A large vertical core was taken with aplastic coring tube (7-cm diameter) from intertidalsediment in Ray Creek, Colne Point salt marsh. Smallcores were removed from the large core at depths of 0to 1, 4 to 5, and 9 to 10 cm by using 5-ml plastichypodermic syringes with the needle end removed.The syringes were capped with rubber teats to excludeair and returned to the laboratory. Each core wasinjected via the rubber seal with 50A of [U-_4C]acetatesolution (10 uCiml', 57.8 mCiemmol- ), withdrawingthe syringe while injecting to distribute the tracersolution as evenly as possible. The samples were in-cubated for 2 h at the temperature of the sedimentwhich had been measured in the field.

After incubation each sediment core was extrudedinto a flask containing 35 ml of potassium hydroxidesolution (1 M). The flask was immediately stoppered,and the sediment was dispersed in the liquid by shak-ing and then gassed with a stream of air from whichcarbon dioxide had been removed by a soda lime trap.The alkali present in the flask prevented carbon diox-ide being carried over during this first stage of distil-lation. This was checked by adding H'4CO3 solution toa control sediment sample containing no labeled ace-tate. No radioactivity was detected when the airstream was passed through a carbon dioxide-absorbingscintillant (14C Absorber P; Fisons, United Kingdom),

985

on January 16, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

986 BANAT ET AL.

and then scintillation counted. Methane in sedimentsamples injected with [U-'4C]acetate was carried bythe air stream through a copper oxide column heatedto 850°C, which oxidized methane to carbon dioxide.This was then absorbed by bubbling through 5 ml ofcarbon dioxide-absorbing scintillant. The efficiency ofmethane oxidation was checked by passing knownvolumes of methane through the copper oxide column,collecting the effluent gas over water, and analyzingfor residual methane by gas-liquid chromatography(Perkin Elmer F33 gas chromatograph with flameionization detector; Poropak N column at 70°C; nitro-gen carrier gas at 20 ml-min-'). The oxidation effi-ciency of methane was higher than 99%.

In the second stage of the distillation the residualsuspension was acidified to lower than pH 1.0 by theaddition of 50% (vol/vol) hydrochloric acid and againgassed with air. The effluent gas was passed throughduplicate scintillation vials in series, each containing5 ml of carbon dioxide-absorbing scintillant, to trap14CO2 formed from labeled acetate.To measure radioactivity in the residual acetate the

sediment suspension in the flask was finally centri-fuged at 2,000 x g for 15 min, and 0.5 ml of the clearsupernatant was added to 4.5 ml of Triton X scintillant[Triton X-100, 0.33 liters; toluene, 0.66 liters; 2,5-di-phenyloxazole, 2.64 g*liter-1; 1,4-bis-(5-phenyloxa-zolyl)benzene, 0.033 g.liter-1]. This scintillant was rou-

tinely used because it was miscible with up to 10%(vol/vol) water, whereas the 14C Absorber P was mis-cible with less than 1% aqueous volume.The counting efficiency for [U-_4C]acetate differed

in the two scintillants, so the radioactivity measuredin the Triton X scintillant was corrected to that in 14CAbsorber P by using a correction factor previouslydetermined by counting a 10-pl sample of [U-_4C]ace-tate solution in each scintillant.The radioactivity in all samples was counted in a

scintillation counter (Packard Tri-Carb 460C), usingan external standard to correct for quenching. Theradioactivity of residual acetate was corrected for sub-sample volume. It was assumed in these experimentsthat residual dissolved radioactivity was due to labeledacetate. This had been previously checked by extract-ing a sediment slurry sample with ether and analyzingthe extract for radioactivity in organic acids by gas-liquid chromatography linked to a gas proportionalcounter. The only radioactivity detectable was in theacetate peak.The radioactivity of residual dissolved acetate could

be affected by low extraction if some labeled acetateadsorbed to, or was entrained by, the centrifugedsediment particles. This was corrected for by using a

reference sediment core which was injected with thesame volume of [U-_4C]acetate solution and then im-mediately extracted by the method described above.As the radioactivity injected into the sample was

known, the soluble acetate radioactivity measured inwater from this control sample could be used to correctother extracted samples for low extraction efficiency.Moreover, although the gas stream was bubbledthrough 14C Absorber P, no radioactivity was detect-able. This confirmed that there was no significantvolatilization and carry-over of labeled acetate fromthe acidified sediment suspension during the gassingperiod used here to strip carbon dioxide.

Experiments with sediment slurries. Materialwas collected from the top 5 cm of sediment in RayCreek, Colne Point salt marsh in completely filled andcapped glass jars. Sediment slurries were prepared inthe laboratory by mixing equal volumes of anaerobicsediment and aged seawater deoxygenated by flushingwith oxygen-free nitrogen. Hungate anaerobic tech-nique was used to maintain anaerobiosis during prep-aration. The sediment contained abundant sulfide,which is an effective redox buffer, and the slurryremained black and reduced.

(i) Experiment with ["4C]acetate in salt marshsediment. Samples (160 ml) of homogenized slurrywere added to each of 10 250-ml conical flasks. Dupli-cate flasks were prepared with the following treat-ments: (i) no addition; (ii) sodium monofluoroacetate(5mM final concentration); (iii) sodium molybdate (20mM final concentration); (iv) no addition; and (v)monofluoroacetate as for (ii).

All chemicals used were of analytical reagent grade.The flasks were closed with Suba-seals and gassed for15 min. Flasks i through iii were gassed with 80% (vol/vol) N2-20% (vol/vol) C02, and flasks iv and v weregassed with 80% (vol/vol) H2-20% (vol/vol) C02.The flasks were incubated at 25°C overnight on an

orbital shaker at a speed just sufficient to preventsettlement of sediment particles. This preincubationwas necessary as experience had shown that some-times a lag occurred before molybdate or fluoroacetatebecame totally inhibitory. As acetate turnover was sorapid, it was necessary to ensure that inhibition hadoccurred before labeled acetate was added to theslurry. Each flask was then injected with 500 Al ofsodium [ U-'4C]acetate solution (lO,uCi * ml-'; 57.8 mCi.mmol-1, Radiochemical Centre, Amersham).

Samples (3 ml) were withdrawn with a 5-ml plastichypodermic syringe and introduced in a universal con-tainer closed with a Suba-seal and containing 0.5 mlof concentrated hydrochloric acid. The universals werethen gassed with oxygen-free nitrogen for 15 min, andthe effluent gas was passed through 4 ml of carbondioxide-absorbing scintillant in a plastic scintillationtube to collect labeled carbon dioxide.The acidified slurry was then centrifuged at 2,000

x g for 10 min, and 0.5 ml of the clear supernatantwas added to 4 ml of Triton X scintillant to count theremaining radioactive acetate.

During the experiment a sample was taken imme-diately after the addition of reagents to the slurry toprovide a zero time measurement of radioactivity, andsamples were thereafter taken at regular intervals upto 48 h. The flasks were regassed only once after 2 hduring the course of the experiment.

(ii) Experiment with [U-'4C]acetate in fresh-water sediment. To elucidate the effect of the inhib-itors molybdate and fluoroacetate upon acetate turn-over an identical experiment was carried out with [U-14C]acetate in slurries of an anoxic freshwater sedi-ment devoid of sulfate. This sediment was obtainedfrom the lake in the grounds of Wivenhoe Park, Uni-versity of Essex, Colchester, United Kingdom. A sam-ple of sediment was removed from the top 5 cm of thelake sediment and used, with the precautions to ex-clude oxygen described for the salt marsh sediment, tomake a 50% (vol/vol) slurry of sediment with distilledwater. Distilled water was used rather than lake water

APPL. ENVIRON. MICROBIOL.

on January 16, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

VOL. 42, 1981

as the latter contained low levels of sulfate. The sulfatecontent of the slurry was below the limits of detection(<0.1 mM) when measured by a barium sulfate turbi-dimetric technique (3). The slurry was dispensed intoflasks and gassed out with N2-CO2. The details of theexperiment were identical to those described for thatwith the salt marsh sediment, except that the lakeslurry was preincubated in the presence of the addedinhibitors for only 3 h, rather than overnight, beforelabeled acetate was added, and the experiments wereonly carried out under N2-CO2. Residual dissolvedradioactivity was measured as described for the exper-inment with salt marsh sediment and subsamples ofslurry were periodically withdrawn over a 96-h period.

(iii) Recovery of sulfide from sediment in thepresence of molybdate. We used [3S]sulfate as atracer to measure the rates of sulfate reduction inslurries of salt marsh sediment in the presence ofdifferent inhibitors. The method used was that de-scribed by Sorokin (19) and depends upon the acidifi-cation of a sample of sediment to volatilize any [3S]-sulfide which is then gassed over with oxygen-freenitrogen into traps containing zinc acetate solution.We frequently observed that when 20 mM molybdatewas added to a sediment sample to inhibit sulfatereduction little or no sulfide was subsequently evolvedor trapped after acidification of the sample, eventhough some sulfide was initially present. This hasalso been reported by Oremland and Silverman (14).At the same time, in the presence of molybdate a darkblue color developed in the sediment interstitial waterafter acidification of the sample. A centrifuged sampleof this interstitial water showed an absorption peak at870 nm. The addition of further sulfide to the sampleintensified the blue color, indicating that its opticaldensity was dependent upon the amount of sulfidepresent. Thus, the apparent lack of measured [3S]-sulfide activity recovered in the zinc acetate traps maybe a reflection of sulfide chemically reacting in thesediment rather than inhibition of sulfate reduction tosulfide.

Strickland and Parsons (21) described a method foranalysis of phosphate in seawater where a phospho-molybdate complex is formed under acidic conditionsin the presence of added molybdate solution, and thisis then reduced by ascorbic acid to a blue-coloredcomplex. Trivalent antimony is also required to formthis blue complex, which has a peak absorption at 885nm.To check whether this blue complex could be

formed by sulfide we added excess phosphate to sea-water followed by the standard reagents described byStrickland and Parsons (21), but omitted the ascorbicacid. When the solution was acidified and then hydro-gen sulfide was passed through, the characteristic bluecolor developed. Thus, it appears that the lack of anyevolved sulfide from acidified sediment in the presenceof molybdate may be the result of the sulfide acting asa reducing agent in forming the blue phospho-molyb-date complex. Many anoxic sediments contain abun-dant phosphate, and the availability of sulfide as areducing agent appears to be the factor limiting thecomplex formation. The slight difference in wave-length of the absorbtion peak of the complex formedin the sediment (870 nm) compared with that pro-

SRB IN SALT MARSH SEDIMENT 987

duced in the standard phosphate analysis (885 nm)may be because trivalent ions other than antimonywere involved in the formation of the colored complexin the sediment.

However, the reoxidation of sulfide upon acidifica-tion in the presence of molybdate interfered withmeasurements of the [3S]sulfide evolved. To over-come this we added an excess of strong reducing agentwhich would be preferentially used to form the bluephospho-molybdate complex, leaving the sulfide un-affected. If excess titanium chloride solution wasadded immediately before acidification, sulfide wasevolved when the sample was acidified and gassedover through zinc acetate traps. In the absence oftitanium chloride, no sulfide was evolved. In the pres-ent work 65 mM titanium chloride proved to be suffi-cient for this purpose, but the amount required maydiffer in other sediments with higher concentrationsof phosphate. Quantitative recovery was checked byusing a slurry of 50% (vol/vol) sediment in deoxygen-ated seawater. Samples of sediment slurry (50 ml)were placed in 150-ml conical flasks sealed with Subaseals and gassed out with 80% (vol/vol) H2-20% (vol/vol) CO2. To each flask 0.1 ml sodium [3S]sulfatesolution (10 1uCiml-1) was added, and three treat-ments were used (four replicates for each): (i) noaddition as controls, (ii) 20 mM molybdate added atthe start of the incubation, and (iii) 20mM molybdateadded at the end of the incubation. The flasks wereincubated at 23°C for 1 week, after which titaniumchloride solution was added to the treatments ii andiii. The flasks were then acidified and gassed out withoxygen-free nitrogen, and evolved sulfide was trappedin zinc acetate solution (1%, wt/vol). Half of the zincacetate trap was titrated iodometrically to measuretotal evolved sulfide; the other half was filtered, andradioactivity was measured by scintillation counting(see below).

(iv) Experiments with [36S]sulfate in saltmarsh sediment. Samples (160 ml) of salt marshsediment slurry were added to 250-ml conical flasksclosed with Suba-seals. Three flasks (i through iii)were gassed with 80% (vol/vol) N2-20% (vol/vol) C02,and two flasks (iv and v) were gassed with 80% (vol/vol) H2-20% (vol/vol) CO2. To each flask was added400 pl of sodium [3S]sulfate solution (20 ,iCiml-',2,500 Ci*mmoli'; Radiochemical Centre), and theflasks were incubated at 25°C on an orbital shaker.After 2.5 h the following additions were made: (i) noaddition (as control); (ii) sodium monofluoroacetatesolution (5.0 mM final concentration); (iii) sodiummolybdate solution (20 mM final concentration); (iv)no addition; and (v) sodium monofluoroacetate as inii. The incubation was continued at 25°C on an orbitalshaker. Samples of slurry (3 ml) were withdrawn withsterile plastic hypodermic syringes and introduced intoa universal tube closed with a Suba-seal and contain-ing 0.3 ml of concentrated hydrochloric acid. Theuniversals were previously gassed out with oxygen-freenitrogen to avoid oxidation of sulfide by air. Gassingwith oxygen-free nitrogen was then continued, andevolved sulfide was trapped in 10 ml of 4% (wt/vol)zinc acetate solution. In sediment samples where mo-lybdate was present 0.2 ml of titanium chloride solu-tion (25%, wt/vol) was added to the subsamples of

on January 16, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

APPL. ENVIRON. MICROBIOL.

sediment to liberate sulfide (65 mM final concentra-tion). The zinc sulfide precipitated in the trap wasfiltered through a Sartorius membrane filter (0.45-,umpore size, 2.4-cm diameter) and dissolved in 2 ml ofBray scintillant. Sufficient fumed silicate (Fisons,United Kingdom) was then added to form a gel inwhich the sulfide particles were evenly distributed.Radioactivity was measured in a scintillation counterby using the external standard method to correct forquenching. Initial samples were taken from all flasksimmediately after the addition of radiotracers to pro-vide a zero time reading; samples were then taken atregular intervals to determine the reduction of sulfateto sulfide with time.

RESULTSProducts of acetate metabolism in salt

marsh sediment. The results ofthe experimentusing injected [U-'4C]acetate in sediment sam-ples are shown in Table 1. Approximately 80%of the added label was recovered as methane,carbon dioxide, and residual soluble radioactiv-ity. The proportion of acetate metabolized tomethane was negligible, carbon dioxide beingthe major end product of acetate turnover.Experiments with sediment slurries. (i)

Experiment with acetate in salt marsh sed-iment. For the first experiment with salt marshsediment slurry using labeled acetate the rela-tive rates ofremoval of acetate are shown in Fig.1 as the radioactivity due to [U-14C]acetate in a0.5-ml subsample of the water from the slurryafter centrifugation. Radioactivity appearing indissolved carbon dioxide is shown in Fig. 2 asactivity in a 3-ml subsample of slurry. The dis-solved carbon dioxide is assumed to be in equi-librium with that in the overlying gas atmos-phere, and the graphs therefore show relativerates of conversion of acetate to carbon dioxide.The apparent decrease in radioactivity after 2 h(Fig. 2) is due to the flasks being regassed afterthis time to maintain an abundant supply ofhydrogen in these slurries held under a hydrogenatmosphere.The experiment with labeled acetate revealed

that its turnover was rapid under both a hydro-gen and a nitrogen atmosphere, the majority(90%) of the label having disappeared from theslurry within the first 2 h of incubation. The

TABLE 1. Percentage conversion after 2 h ofincubation of[U-'4CJacetate to 14CO2 and I4CH4 incores of salt marsh sediment from Ray Creek, Colne

Point salt marshDepth of % Acetate % 14C02 % 14CH4 % Re-cores (cm) remaining formed formed covery

Oto 1 18.0 62.6 0 80.64 to 5 16.0 74.7 0 90.79 to 10 44.7. 39.9 0 84.6

FIG. 1. Radioactivity remaining in [U-"4CJacetatein 0.5-ml subsamples of slurries of salt marsh sedi-ment. Symbols: A, slurry under N2-CO2; A, slurryunder H2-CO2; 0, slurryplus 20mM molybdate underN2-CO2; E, slurryplus 5 mMfluoroacetate under N2-C02; , slurry plus 5 mM fluoroacetate under H2-CO2.

E

0

0 8 16 24Time (hours)

36 48

FIG. 2. Radioactivity from [U-'4CJacetate appear-ing as dissolved carbon dioxide in 3-ml subsamplesof slurries of salt marsh sediment. Symbols as forFig. 1.

turnover constant (k) for acetate in the controlswas 2.07 h-1. The presence of fluoroacetate, aninhibitory analog of acetate, virtually stoppedacetate turnover (k = 2.84 x 10-3 h-1), with 85%of the label still remaining in the slurry undernitrogen after 48 h. The presence of 20 mMmolybdate, which is an inhibitor of sulfate re-duction (15, 16) also inhibited acetate turnover(k = 1.82 x 10-3 h-'), although a very slow butcontinued rate of disappearance of label seemedto proceed throughout the experiment, with 65%of the initial acetate label remaining after 48 h.When the accumulation of 14C label in the

dissolved carbon dioxide was examined therewas rapid accumulation in the controls underboth hydrogen and nitrogen during the first 2 hof the experiment when label rapidly disap-peared from the acetate. Thereafter, there waslittle further accumulation of label in the carbon

EcL

x

5 24Time (hours)

luu

75 A

50

25 O-

0 .

988 BANAT ET AL.

on January 16, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

SRB IN SALT MARSH SEDIMENT 989

dioxide. The presence of fluoroacetate almostentirely eliminated the accumulation of label incarbon dioxide throughout the experiment; thepresence of molybdate resulted in a small initialamount of accumulation during the first 16 h,but thereafter no further accumulation of labeloccurred.

(ii) Experiment with [U-'4C]acetate infreshwater sediments. In the slurries of lakesediment acetate turnover (k = 1.4 x 10-2 h-1)was slower than in the salt marsh sediment, theresidual dissolved radioactivity disappearingafter approximately 96 h (Fig. 3). Fluoroacetateagain almost totally inhibited acetate metabo-lism (k = 1.09 x 1O-3 h-1), but the addition ofmolybdate had no effect upon acetate turnover.The rate of disappearance of dissolved residualradioactivity due to acetate in the presence ofmolybdate was identical to that in the controls.

(iii) Recovery of sulfide from sediment inthe presence ofmolybdate. Table 2 shows theeffect of the addition of titanium chloride uponrecovery ofsulfide from sediment in the presenceof molybdate. Comparison of treatments i andiii shows that recovery of sulfide was >99% whentitanium chloride solution was added to the sed-iment slurry before acidification. In contrast, intreatment ii, where molybdate (20 mM) was

present throughout the incubation period, a

25 I.

20 0aE

01.In0

0 24 48 72 96Time (hours)

FIG. 3. Radioactivity remaining in [U-14C]acetatein 0.5-ml subsamples of slurries of freshwater sedi-ment from Wivenhoe Park Lake. Symbols as forFig. 1.

smaller amount of sulfide was present at the endof the experiment and in addition there was no

reduction of [3S]sulfate to [3S]sulfide. This con-

firned that molybdate was an effective inhibitorof sulfate reduction, and that the lack of radio-activity in sulfide was not an artifact resultingfrom a low recovery of sulfide in the presence ofmolybdate.

(iv) Experiments with [36S]sulfate in saltmarsh sediment. The relative rates of sulfatereduction are shown in Fig. 4 as the radioactivitydue to [3S]sulfide in a 3-ml subsample of slurry.Under a nitrogen atmosphere the presence offluoroacetate substantially decreased the rate ofsulfate reduction, as measured by the reductionof 35SO42- to 'S- Over the first 48 h the rate ofsulfide formation was reduced by approximately60% in the presence of fluoroacetate. Molybdateentirely eliminated labeled sulfide formation. Nolabel was present in the sulfide even when tita-nium chloride was added to ensure successfuldistillation of sulfide, and inhibition of sulfatereduction was complete.

DISCUSSION

The data from salt marsh sediment cores in-jected with labeled acetate showed that virtuallyno methane was derived from acetate metabo-lism, acetate oxidation to carbon dioxide beingthe predominant process. This is similar to otherworkers' data (9, 11, 23) for sediments rich insulfate, although in low-sulfate sediments meth-anogenesis predominates (23).The experiments with slurries of salt marsh

sediment showed that labeled acetate was rap-idly metabolized under both an N2-CO2 and an

H2-C02 atmosphere. Label accumulated in car-

bon dioxide during the initial 2 h when acetatemetabolism was rapid (Fig. 2). Acetate metabo-lism was almost entirely inhibited by the pres-ence of fluoroacetate, whereas the presence ofmolybdate, an inhibitor of SRB (15, 16), alsogreatly diminished acetate metabolism to an

extent almost equal to that in the presence offluoroacetate.An overnight preincubation of slurry was used

TABLE 2. Sulfide concentrations and radioactivity present as [3S]sulfide in sediment slurry after 1 week ofincubation at 23°CFinal sulfide concn (mmol-l' of sedi- Radioactivity as [3S]sulfide

Treatment ment) (cpm.ml-' of sediment)

i Control 10.2 x 10-3 (± 1.77 x 10-4)a 80,797 (± 671)ii 20 mM molybdate added at start of 0.38 x 10-3 (+ 7.22 x 10-5) 18 (+ 4)

incubationbiii 20 mM molybdate added at end of 9.94 x 10-3 (± 1.88 x 10-4) 81,063 (± 2159)

incubationba Numbers within parentheses indicate standard errors.b Titanium chloride (65 mM) added before acidification and distillation of sulfide.

VOL. 42, 1981

on January 16, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

APPL. ENVIRON. MICROBIOL.

E

U 80

40

16

12

EcLu 8

0

4 ci

0 20 40 60 80 100 120

Time (hours)

FIG. 4. Appearance ofradioactivity from 35SO4- as35'2- in 3-ml subsamples of slurries of salt marshsediment. Symbols as for Fig. 1.

to ensure complete effectiveness of the addedinhibitors before labeled acetate was injectedinto the slurry, as its turnover was so rapid. Thispreincubation period did not result in depletionof acetate metabolism which was active in theunsupplemented controls when labeled acetatewas later injected. The lack of activity in thepresence of both molybdate and fluoroacetatemust have been the result of effective inhibitionof acetate-metabolizing bacteria during thepreincubation period.Our data strongly suggested that SRB were

involved in acetate turnover in the salt marshsediment, as suggested by other workers (9, 23).Peck (16) and Oremland and Taylor (15) haveshown that SRB are inhibited by molybdate,and our data indicated that acetate oxidationwas largely inhibited by molybdate addition.However, caution is required when interpretingdata using supposedly specific inhibitors in nat-ural mixed microbial communities where theremay possibly be nonspecific side effects by theinhibitor on other groups ofmicroorganisms. Forexample, Wolin and Miller (24) have shown thata molybdate-sulfide complex affects synthesis ofhydrogenase, but not activity of preformed hy-drogenase, in Ruminococcus albus, although thecomplex had no effect at all in five other bacteria

examined. The other major group of acetate-metabolizing bacteria, the methanogens, do notappear to be adversely affected by molybdate,as addition of 20 mM molybdate almost invari-ably stimulates methanogenic rates in slurries ofsalt marsh sediment (15; Nedwell and Banat,Microb. Ecol., in press). Therefore, inhibitoryside effects of molybdate upon acetate-dissimi-lating methanogens seem to be discounted. In-deed, the slow residual rate of dilution of labeledacetate in the presence of molybdate (Fig. 1)indicates the presence of acetate-dissimilatingbacteria which are not molybdate sensitive.However, the experiment with labeled acetate

in slurries of lake sediment suggested that mo-lybdate had no significant side effects upon othergroups of bacteria involved in acetate turnover.In this experiment fluoroacetate again totallyinhibited acetate turnover, but molybdate hadno inhibitory influence. In the absence of sulfatein this freshwater sediment the SRB would notbe actively respiring sulfate and hence wouldnot be susceptible to inhibition by molybdate.The identical results for acetate turnover in boththe controls and the molybdate-treated slurry(Fig. 3) argue that molybdate had no effect uponother groups of bacteria, other than the SRB,which influenced acetate turnover. While notentirely conclusive, this evidence from a fresh-water sediment suggests that molybdate has lit-tle significant influence other than upon theSRB, at least as far as acetate turnover is con-cerned, and that the effect of molybdate uponacetate turnover in the salt marsh sedimentmust be because of its effect upon acetate-oxi-dizing SRB.This conclusion was in turn supported by the

slurry experiment investigating reduction of[35S]sulfate. Fluoroacetate, which has beenshown to totally inhibit acetate metabolism, de-pressed the measured rate of sulfate reductionunder an N2-CO2 atmosphere (Fig. 4) during thefirst 2 days by about 60% of the rate in theabsence offluoroacetate. However, fluoroacetatedid not entirely inhibit sulfate reduction. Aslower but continued rate of sulfate reductionproceeded throughout the experiment. We inter-pret this as showing the presence of at least twodistinct functional groups of SRB both of whichare inhibited by the addition of molybdate,which entirely inhibits sulfate reduction (Fig. 4).Molybdate apparently acts by inhibiting adeno-sine triphosphate sulfurylase (18), which is pre-sumably common to both groups. One of thegroups ofSRB comprises acetate oxidizers whichare inhibited by the addition of fluoroacetate,thus depressing the rate of sulfate reduction. Asecond group of SRB comprises organisms thatmetabolize electron donors other than acetate

I

990 BANAT ET AL.

on January 16, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

SRB IN SALT MARSH SEDIMENT 991

and are not inhibited by the presence of fluo-roacetate. Thus, this second group continues toreduce sulfate in the presence of fluoroacetate.As the SRB seem to be the predominant hydro-gen scavengers in this sediment the implicationis that hydrogen is the major electron donor forthis second group.This interpretation is supported by the obser-

vation (Fig. 4) that under a hydrogen atmos-phere the rate of sulfate reduction was greatlystimulated, and although fluoroacetate dimin-ished the amount of sulfate reduction the inhib-itory effect was proportionately less than undera nitrogen atmosphere. It has been shown else-where (16) that sulfate reduction is limited byelectron donor availability in these sediments,and the presence of abundant hydrogen wouldstimulate the activity of the hydrogen-metabo-lizing SRB, but not that of the acetate-oxidizingSRB. We propose, therefore, that the SRB arenot only the predominant hydrogen-scavengingbacteria in this sediment (5), but that they arealso the predominant acetate oxidizing bacteria.Most of the SRB known, predominantly of

the genus Desulfovibrio (17), are able to utilizea restricted range of reduced organic acids suchas pyruvate and lactate or ethanol. Many ofthese bacteria are also known to utilize hydro-gen, and indeed some are able to use hydrogenas the sole electron donor (4). Acetate is thepredominant end product when organic electrondonors are being metabolized by these SRB, andit is not further oxidized to carbon dioxide. Afew acetate-oxidizing sulfate reducers have beenreported, including Desulfotomaculum acetoxi-dans (22) and more recently Desulfobacter (9a).Both ofthese reported genera have been isolatedfrom marine anoxic sediments, and both oxidizeacetate to carbon dioxide. Neither is able tometabolize hydrogen. It seems, therefore, thatthese acetate-oxidizing sulfate reducers repre-sent a distinct functional group occupying a met-abolic niche separate from that of the majorityof sulfate reducers so far known. Their distri-bution may be wider than so far appreciated.The observation that active sulfate reduction

inhibits methanogenesis has already at least par-tially been attributed to competition for availa-ble hydrogen (2, 15), but our present data sug-gest that competition for acetate between theacetate-oxidizing SRB and the acetate-dissimi-lating methanogenic bacteria may also be signif-icant. Winfrey and Zeikus (23) have proposed asimilar competition in Lake Mendota sediment,where sulfate inhibition of methanogenesis wasreversed by the addition of either hydrogen oracetate.

Figure 5 illustrates the suggested relationshipbetween SRB and the organotrophic bacteria

preceding them in the pathway of carbon flowin the salt marsh sediment. The hydrogen-scav-enging SRB cause the release of hydrogen byhydrogen transfer from the preceding bacteriawith the concomitant formation of acetate. Ithas been demonstrated with rumen microorga-nisms that hydrogen transfer stoichiometricallyincreases the proportion of acetate in the meta-bolic products of the fermentative-hydrogen-transferring bacteria preceding the hydrogen-scavenging organism (8, 10, 18), whereas in theabsence of hydrogen-scavenging bacteria thereis a stoichiometric increase in the proportion ofreduced fermentative products such as lactate,propionate, succinate, etc., at the expense ofacetate, because hydrogen transfer is inhibitedby the accumulation of hydrogen. The mainte-nance of low sedimentary hydrogen concentra-tions by H2-scavenging bacteria presumably hasthe same effect in anoxic sediments.Such a scheme suggests two possible mecha-

nisms for competition among the SRB andmethanogenic bacteria, namely, between the hy-drogen-scavenging SRB and the methanogensusing the H2 plus CO2 pathway to methane andamong the acetate-oxidizing SRB and the ace-tate-dissimilating methanogens (Fig. 5). The sit-uation apparently existing in the surface saltmarsh sediment at these two sites of potentialcompetition is indicated in Fig. 5, the SRB ap-parently outcompeting the methanogens forboth substrates. In comparison, a number ofauthors (9, 11, 23) have shown that in sulfate-depleted sediment a much greater proportion ofthe acetate is metabolized to methane than tocarbon dioxide, presumably because lack of sul-fate inhibits the SRB and prevents them fromcompeting either for acetate or for hydrogen.

ORGANIC INTERMEDIATESH| transferringorganotrophs

CHjCOOH H2

SOi, ) 3 CH4

co2 H20FIG. 5. Illustration of the proposed role ofSRB in

electron flow in salt marsh sediment. Abbreviations:SRBI, acetate-oxidizing SRB; SRB2, H2-oxidizingSRB; MBI, acetate-dissimilating methanogenic bac-teria; MB2, H2-oxidizing methanogenic bacteria.

VOL. 42, 1981

on January 16, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

APPL. ENVIRON. MICROBIOL.

ACKNOWLEDGMENTS

This work was supported in part by grant GR3/3517 toD.B.N. from the Natural Environment Research Council,United Kingdom.

LITERATURE CITED

1. Abram, J. W., and D. B. Nedwell. 1978. Inhibition ofmethanogenesis by sulfate-reducing bacteria competingfor transferred hydrogen. Arch. Microbiol. 117:89-92.

2. Abram, J. W., and D. B. Nedwell. 1978. Hydrogen as asubstrate for methanogenesis and sulfate reduction inanaerobic saltmarsh sediment. Arch. Microbiol. 117:93-98.

3. American Public Health Association. 1974. Standardmethods for the examination of water and wastewater.American Public Health Association, Washington, D.C.

4. Badziong, W., R. K. Thauer, and J. G. Zeikus. 1978.Isolation and characterization of Desulfovibrio growingon hydrogen plus sulfate as the sole energy sources.Arch. Microbiol. 116:41-50.

5. Banat, L. M., D. B. Nedwell, E. Senior, and E. B.Lindstrom. 1980. Interactions between organotrophic,methanogenic and sulphate-reducing bacteria in anaer-obic saltmarsh sediment. Soc. Gen. Microbiol. Quarterly7:79.

6. Bryant, M. P., E. A. Wolin, M. J. Wolin, and R. S.Wolfe. 1967. MethanobaciUus omelianskii, a symbioticassociation of two species of bacteria. Arch. Mikrobiol.59:20-31.

7. Cappenberg, T. E. 1975. Relationships between sulfate-reducing and methane-producing bacteria. Plant Soil43:123-129.

8. lannotti, E. L., D. Kafkewitz, M. J. Wolin, and M. P.Bryant. 1973. Glucose fermentation products of Rum-inococcus albus grown in continuous culture with Vib-rio succinogenes. Changes caused by interspecies trans-fer of H2. J. Bacteriol. 114:1231-1240.

9. King, G. M., and W. J. Wiebe. 1980. Tracer analysis ofmethanogenesis in salt marsh soil. Appl. Environ. Mi-crobiol. 39:877-881.

9a.Laanbroek, H. J., and N. Pfennig. 1981. Oxidation ofshort-chain fatty acids by sulfate-reducing bacteria infreshwater and marine sediments. Arch. Microbiol. 128:330-335.

10. Latham, M. J., and M. J. Wolin. 1977. Fermentation ofcellulose by Ruminococcus flavefaciens in the presenceand absence ofMethanobacterium ruminantium. Appl.

Environ. Microbiol. 34:297-301.11. Mountfort, D. 01., R. A. Asher, E. L. Mays, and J. M.

Tiedje. 1980. Carbon and electron flow in mud andsandflat intertidal sediments at Delaware Inlet, Nelson,New Zealand. Appl. Environ. Microbiol. 39:686-694.

12. Nedwell, D. B., and J. W. Abram. 1978. Bacterialsulphate reduction in relation to sulphur geochemistryin two contrasting areas of saltmarsh sediment. Estua-rine Coastal Mar. Sci. 6:341-351.

13. Nedwell, D. B., and J. W. Abram. 1979. Relative influ-ence of temperature and electron donor and electronacceptor concentrations on bacterial sulfate reductionin saltmarsh sediment. Microbial. Ecol. 5:67-72.

14. Oremland, R. S., and M. P. Silverman. 1979. Microbialsulfate reduction measured by an automated electricalimpedance technique. Geomicrobiol. J. 1:355-372.

15. Oremland, R. S., and B. F. Taylor. 1978. Sulfate reduc-tion and methanogenesis in marine sediments. Geo-chim. Cosmochim. Acta 42:209-214.

16. Peck, H. D. 1959. The ATP-dependent reduction of sul-fate with hydrogen in extracts of Desulfovibrio desul-furicans. Proc. Natl. Acad. Sci. U. S. A. 45:701-708.

17. Postgate, J. R. 1979. The sulphate-reducing bacteria.Cambridge University Press, Cambridge.

18. Scheifinger, C. C., B. Lineham, and M. J. Wolin. 1975.H2 production by Selenomonas ruminantium in theabsence and presence of methanogenic bacteria. Appl.Microbiol. 29:480-483.

19. Sorokin, Y. I. 1962. Experimental investigation of bac-terial sulfate reduction in the Black Sea. Microbiology32:320-335.

20. Sorokin, Y. I. 1966. Role of carbon dioxide and acetate inbiosynthesis by sulphate reducing bacteria. Nature(London) 210:551-552.

21. Strickland, J. D. H., and T. R. Parsons. 1972. A prac-tical handbook of seawater analysis, Fisheries ResearchBoard of Canada, Ottawa.

22. Widdel, F., and N. Pfennig. 1977. A new anaerobicsporing, acetate oxidizing, sulfate-reducing bacterium,Desulfotomaculum (emend.) acetoxidans. Arch. Micro-biol. 112:119-122.

23. Winfrey, M. R., and J. G. Zeikus. 1977. Effect of sulfateon carbon and electron flow during microbial methan-ogenesis in freshwater sediments. Appl. Environ. Micro-biol. 33:275-281.

24. Wolin, M. J., and T. L. Miller. 1980. Molybdate andsulfide inhibit H2 and increase formate production fromglucose by Ruminococcus albus. Arch. Microbiol. 124:137-142.

992 BANAT ET AL.

on January 16, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from