Vol. 53, No. 1APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Jan. 1987, p. 27-320099-2240/87/010027-06$02.00/0Copyright © 1987, American Society for Microbiology

Biofilm Dynamics and Kinetics during High-Rate Sulfate Reductionunder Anaerobic Conditions

PER HALKJAER NIELSEN

Environmental Engineering Laboratory, University ofAalborg, DK-9000 Aalborg, Denmark

Received 9 July 1986/Accepted 14 October 1986

The sulfate kinetics in an anaerobic, sulfate-reducing biofilm were investigated with an annular biofilmreactor. Bioflm growth, sulfide production, and kinetic constants (Km and Vmax) for the bacterial sulfate uptakewithin the biofilm were determined. These parameters were used to model the biofilm kinetics, and theexperimental results were in good agreement with the model predictions. Typical zero-order volume rateconstants for sulfate reduction in a biofilm without substrate limitation ranged from 56 to 93 ,umol of SO4cm-3 h-1 at 20°C. The temperature dependence (Qlo) of sulfate reduction was equivalent to 3.4 at between 9and 20°C. The measured rates of sulfate reduction could explain the relatively high sulfide levels found insewers and wastewater treatment systems. Furthermore, it has been shown that sulfate reduction in biofilmsjust a few hundred micrometers thick is limited by sulfate diffusion into biofilm at concentrations below 0.5mM. This observation might, in some cases, be an explanation for the relatively poor capacity of thesulfate-reducing bacteria to compete with the methanogenic bacteria in anaerobic wastewater treatment insubmerged filters.

Biofilms are used in a number of fixed-film wastewatertreatment systems in which substrate removal is a complexcombination of mass transfer and microbiological processes.However, during anaerobic treatment of wastewater in sub-merged filters, hydrogen sulfide produced by microbial sul-fate reduction is often an undesirable by-product. Removalof sulfide is expensive, and the gas is highly toxic. Further-more, in the presence of oxygen, sulfide is chemically orbiologically oxidized to the very corrosive sulfuric acid (25,31). Besides the production of sulfide, the activity of thesulfate reducers is also responsible for anaerobic metalcorrosion, which is well known, e.g., in the petrochemicalindustry (6). Also, in sewer systems transporting wastewaterunder anoxic conditions over long distances, the sulfideproduction from "slimes" or biofilm results in health andeconomic problems (3, 24, 31). However, in some casessulfate reduction in biofilm is used to remove sulfate fromsulfate-rich mine water or industrial effluents (20).

Despite the fact that sulfate reduction is a widespread andwell-known phenomenon, no studies of the biofilm kineticsin high-rate sulfate-reducing biofilms have been published,although biofilm kinetics in methanogenic reactors havebeen described (30, 33). Sulfate reduction is well describedonly in marine and fresh water sediments, in which theprocess has been shown to be very important for theanaerobic mineralization of organic matter (12, 15). Severalmodels exist that describe the fundamental biofilm processes(e.g., see references 2, 26, and 35), but none of these modelsdeals directly with sulfate reduction in biofilms. Some em-pirical models to explain the sulfide build-up in sewersystems have been published (3, 24, 31), but the biofilmkinetics are not directly taken into account. Studies with atubular biofilm reactor have resulted in sulfide productionrates comparable to those found in actual sewer systems (9).The aim of this work was to study the dynamics and

kinetics of an anaerobic, sulfate-reducing biofilm. The mea-surements were made by using a completely mixed rotatingannular reactor (4) with biofilm growth on the outer cylinder.The substrate uptake of the biofilm was modeled and com-

pared with experimental results at low sulfate concentra-tions.

MATERIALS AND METHODSExperimental system. All experiments except the determi-

nation of Km were performed in an annular reactor system,which consisted of a drum of polyvinylchloride enclosed in atransparent polyvinylchloride cylinder (Fig. 1). The drumrotated at 190 rpm (80 cm s-1). Two magnets in the drum andone fastened to a motor outside the reactor served forrotation. At this speed the stagnant liquid layer and, there-fore, the external mass transfer resistance were insignificant(16). A circulation circuit was connected to the reactor toprevent secondary streamings in the reactor. To minimizegas transfer through the reactor, tubes, fittings and thecirculation system were made from low-permeability mate-rials including nylon, Teflon (E. I. du Pont de Nemours &Co., Inc.), and polyvinylchloride. Samples for analyses weretaken with a syringe through butyl rubber stoppers placed inthe lid of the reactor. Biofilm formation was restricted to theouter cylinder by cleaning the lids and drum regularly understrict anoxic conditions. The combined volume of the reac-tor and circulation circuit was 562 ml, and the biofilmsurface-to-volume ratio was 90 m2m-3.Media and biofilm growth conditions. Bacteria were grown

in a mineral salt medium with the following components(grams per liter of distilled water unless otherwise specified):KH2PO4 (0.5), NH4Cl (1.0), CaCl2 .2H20 (0.06),MgCl2- 4H20 (0.05); trace elements (1 ml) (34), resazurin(0.01), sodium citrate 2H20 (0.3), and sodium lactate (2 mlof a 50% solution). The pH was adjusted to 7.2 with NaOH,and the solution was flushed with pure N2 and reduced byadding Na2S 9H20 from an anoxic sterile stock solution.Various concentrations of sulfate were added from an anoxicsterile stock solution, and the concentration was, in somecases, controled by ion chromatography.The inoculum of sulfate-reducing bacteria originated from

raw sewage. A stock culture was grown as a batch and fedwith anoxic domestic wastewater. During the first 3 to 4days, the reactor was run as a batch reactor until some

27

on October 18, 2020 by guest

http://aem.asm

.org/D

ownloaded from

APPL. ENVIRON. MICROBIOL.

0

.<I

C

p

T 4cm

FIG. 1. Schematic diagram of the annular biofilm reactor. Com-ponents: R, Rotating drum; C, circulation circuit; P, gas-tight pump;M, magnets; I, inlet for media; 0, outlet.

surface colonization had occurred. Thereafter, substrate wascontinuously fed by a peristaltic pump. The sulfate concen-tration in the reactor was kept at around 1 mM, which is a

common value for the sulfate concentration in domesticwastewater. All experiments were carried out in the dark at15 or 20°C. Biofilm thickness was determined microscopi-cally (4) on three plugs, each with a diameter of 6 mm, drilledout from the reactor wall. Ten measurements were made oneach plug, and the thickness was determined as the meanwith a standard deviation of approximately 10 to 15%. Dryweight was determined by removing the biofilm from a

known area of the reactor wall. The sample was filtered ontoa preweighed filter (pore size, 0.45 ptm) and dried to a

constant weight at 600C.Sulfide production in the biofilm. Sulfide production in the

biofilm was calculated from the initial velocity of sulfideformation in the reactor after the flow through the reactorwas stopped. The flow was stopped when the actual sulfatesteady-state concentration was obtained, normally after 4 to5 residence times. The duration of each rate measurementvaried from 20 to 30 min.The sulfide production rate was also measured as a steady-

state balance in the continuously fed culture. The sulfideproduction rate was calculated from the difference in sulfideconcentrations in the inlet and outlet and the flow ratethrough the reactor. Sulfide levels were determined photo-metrically by the methylene blue method (5).Measurements of Km and Vmax. The kinetic parameters for

sulfate uptake were determined by using radiotracer tech-niques to produce a progress curve of sulfate depletionversus time in a suspended culture. A part of the biofilmfrom the annular reactor was gently homogenized understrict anoxic conditions and washed in anoxic sulfate-freemedium. An adequate amount of the homogenized biofilmwas transferred into 120-ml serum bottles which each con-

tained 50 ml of sterile medium. The bottles were sealed withblack butyl rubber stoppers and incubated in the dark at 20°Cin a shaking-water bath. After 2 h of incubation, unlabelledsulfate (final concentration, 5 to 20 FiM) and 10 to 20 ,uCi ofcarrier-free 35so2- (Risoe, Roskilde, Denmark) were in-

jected into the bottles. Samples of 0.5 ml each were injectedinto 5-ml portions of a 1% (wt/vol) zinc acetate solution toprecipitate the sulfide and terminate bacterial activity. Thezinc sulfide was removed by centrifugation, and the radio-activity of the sulfate was measured in a subsample of thesupernatant by using a liquid scintillation counter (PackardInstrument, Co., Inc.).At least three progress curves of sulfate depletion were

used to determine the kinetic constants calculated from thelinearization of the integrated Michaelis-Menten equation(27):

tl(So- S) = (KmlVmax) ln (SoIS)/(So - S) + 1/Vmax (1)

where S0 is the substrate concentration at time zero, S is thesubstrate concentration, Vmax is the maximal specific rate ofsubstrate uptake, Km is the apparent half-saturation con-

stant, and t is time.Modeling biofilm kinetics. The substrate uptake of bacteria

is normally described by Michaelis-Menten kinetics, and thediffusion into biofilm is normally described by Fick's secondlaw of diffusion (35). At steady state the mass balance forone limiting substrate with no external mass transfer resist-ance can be described by the following equation (35):

Df (d2SB/dZ2) = XVmaxSBI(Km + SB) (2)

where SB is the substrate concentration at a point within thebiofilm, Df is the molecular diffusion coefficient within thebiofilm, X is the biomass, and z is the distance normal to thebiofilm surface.The second-order nonlinear, ordinary differential equation

does not have an explicit solution. It is, however, possible touse some analytical approximations (7, 26, 30). In this studythe equation was solved by a numerical method, and thesurface flux through the biofilm surface was calculated forvarious sulfate concentrations.

RESULTS

Biofilm growth. An increase in biofilm thickness in thereactor, caused by the growth of a mixed population ofanaerobic bacteria without substrate limitation, is shown inFig. 2. After an induction period for the development of a

primary layer of biofilm, a nearly exponential increase in thebiofilm thickness was observed. When the thickness of thebiofilm was over 300 to 400 ,um, it was no longer fullypenetrated by sulfate at an actual concentration of 1 mM inthe reactor (see Fig. 5). At higher biofilm thicknesses,increasing hydraulic stress caused the rate of accumulationof bacteria in the biofilm to decrease and finally to stop (datanot shown). The activity of the sulfate reducers in the biofilmwas measured as the sulfide production rate for fully pene-trated biofilm. This activity showed a nearly exponentialincrease with time, and the doubling time for sulfide produc-tion was around 30 to 31 h. This means that the doubling timefor the sulfate reducers was less than 30 h, because a partof the population was continuously washed out of thechemostat.

In other experiments in which biofilm thicknesses rangedfrom 100 to 300 ,um, the measured zero-order volume rateconstant (kof) for fully penetrated biofilm ranged from 56 to93 p.mol of SO4- cm-3 h-1 at 200C. The biofilm dry weightswere between 40 and 91 mg cm-The temperature dependence of the bacterial activity in

the biofilm was measured at temperatures between 9 and200C. The sulfide production in the fully penetrated biofilm

II

I III I

R

28 NIELSEN

I 1._ ..

on October 18, 2020 by guest

http://aem.asm

.org/D

ownloaded from

BIOFILM DYNAMICS AND KINETICS 29

showed a linear increase in an Arrhenius plot (Fig. 3). Theactivation energy of the process was found to be 84.9 kJmol' which yields a Qlo of 3.4 for kof in the measuredtemperature interval.

Sulfate uptake kinetics of the bacteria. A typical progresscurve for sulfate use in the homogenized biofilm is shown inFig. 4A. These data were used to calculate the kineticconstants with a linearized form of the integrated Michaelis-Menten equation (Fig. 4B).The initial sulfate concentration in the experiments varied

from 5 to 20-fold the Km. The Km for a mixed population ofsulfate reducers in the biofilm was 1.4 + 0.2 (standard errorof the mean) ,uM So3-, and the specific sulfate uptake Vmn,was 1.4 0.1 mmol of so2 g-1 h- (n = 3).

Bioffilm kinetics. The sulfide production rates at differentsulfate concentrations from two biofilms with different thick-nesses are shown in Fig. 5. The experiments were done at15°C to minimize the effect of biofilm growth during theexperiment, which normally lasted 6 to 8 h. For each biofilmthe sulfide production rate from the biofilm surface did notincrease when the sulfate concentration exceeded a certainvalue. Consequently, beyond these values the biofilms werefully penetrated with sulfate, and all the bacteria within thebiofilms were active. The addition of more lactate did notincrease the sulfide production rate, which showed thatlactate was always in surplus.

Since sulfate was the only source for sulfide production,the zero-order constant for sulfide production was assumedto be identical to the sulfate reduction rate. This assumptionwas confirmed for the thin biofilm, in which the rate ofsulfate consumption was measured by radiotracer tech-niques. At four different sulfate concentrations the sulfateconsumption and sulfide production were in good agreement(Fig. 5A). kof was calculated based on the sulfide productionrate in the fully penetrated biofilms and found to be 44.7 x103 and 36.2 x 103 mmol m-3 h-1 at biofilm thicknesses of112 and 239 ,um, respectively.

Sulfate reduction in the biofilm was modeled by solvingequation 2 by numerical methods with the values shown intable 1. The diffusion coefficient for sulfate in pure water (Do)is 8.28 x 10-6 cm2 s-1 at 15°C (18), and the diffusioncoefficient within the biofilm is assumed to be 0.8 x Do (36).The predicted sulfate reduction in the biofilm based on the

-

EE

C%

0-v

-1

3.4 3.51/T, 103 (K-1)

3.6

FIG. 3. Arrhenius plot of sulfide production rates in a fullypenetrated biofilm. The temperature characteristic was equivalent to84.9 kJ mol-P or a Qlo of 3.4.

model showed good agreement with the measured sulfideproduction rates at different sulfate concentrations for bothbiofilm thicknesses (112 and 239 ,um) (Fig. 5).

In these experiments sulfate reduction in rather thinbiofilms (<300 ,um) without limitations on organic matterwere observed to be limited by sulfate diffusion into thebiofilm at the typical sulfate concentrations (0.1 to 0.5 mM)found in both fresh water and domestic wastewater.

DISCUSSION

It has been shown that sulfide production from a sulfate-reducing biofilm could be described by a well-known, rela-tively simple model when kinetic constants were available.Furthermore, it was seen that diffusional limitation of sulfatereduction might be an important phenomenon in manyhigh-rate anaerobic biofilm systems. The biofilm used in thisstudy gave the same growth and activity patterns with timeas have been reported for other heterotropic biofilms (8, 32).

120 160 200 240 280TIME (hours)

FIG. 2. Biofilm growth and sulfide production. The experimentswere carried out in an annular reactor at 200C.

12

210-

z 8-0

6-

;- 4-n

2

Ln6

4--- 2~

0 20 40 60 80TIME (minutes)

0.1 0.2 0.3 0.4

In (S0/S) / (S0 - S)

100

FIG. 4. (A) Time course of sulfate consumption in a homoge-nized biofilm. The experiments were conducted at 20°C. (B) Kineticconstants calculated from the data in Fig. 4A by a linearization ofthe integrated Michaelis-Menten equation (27). The maximum spe-cific rate of sulfate uptake was 1.4 mmol SO}- g (dry weight)-' h-'.

-I

E

EE

z0

:D0cr

uJ0

DUl)

()Km = 1.4 p MVmax = 13.3 pM hV'

0

VOL. 53, 1987

on October 18, 2020 by guest

http://aem.asm

.org/D

ownloaded from

APPL. ENVIRON. MICROBIOL.

c-I

E

E

C 0.2 0.4 0.6 0.8 1.0SULFATE CONC. (mM)

FIG. 5. Biofilm surface sulfide production rate at various concentrations of sulfate. Temperature, 15°C. Curves: A, experimental data forsulfide production (O) and sulfate reduction (U) in 112-p.m-thick biofilm; B, experimental data for sulfide production in 239-,um-thick biofilm(A). The two curves are predictions from the model for sulfate uptake in the biofilm, based on the parameters listed in Table 1.

When the necessary kinetic constants are obtained formodeling biofilm kinetics, it is important to distinguishbetween the kinetic parameters for individual bacteria insidethe biofilm (Km and Vmax) and the biofilm kinetic parameters(kofand Df). kof includes the bacterial density and the specificactivity of the bacteria (Vmax).The kinetic constants for the population of sulfate-

reducing bacteria in the homogenized biofilm were compa-rable to the constants reported in other studies (10, 22). TheKm of 1.4 pFM was far below normal fresh-water sulfateconcentrations. This value was slightly lower than the levelof 4 to 10 ,.M reported for different fresh-water strains. TheKm (30 to 300 FM) for marine strains indicates that thesulfate reducers in nature are adapted to ambient sulfateconcentrations (10, 11). This probably means that sulfatereducers (and maybe other types of bacteria) in a thickbiofilm have different affinities (Kin) for the substrate, de-pending on their locations within the biofilm. In this studythe biofilm was always fully penetrated with sulfate exceptfor rather short experimental periods. Therefore, variationsin Km for the sulfate reducers were probably very small. TheVmax for sulfate reduction was around 1.4 mmol SO4 g-1h-1, which is in agreement with earlier findings for purecultures (10).The volumetric kof is very important for the amount of

sulfate which is reduced in the biofilm and therefore is also

TABLE 1. Values of kinetic parameters of the model

Biofilm Value of indicated kinetic parameter (units)a:thickness Xb Vmax K,,, Df

(tim) (g m-3) (mmol g-1 h-1) (mmol m-3) (Mi2 h-1)

112 59 x 103 0.75 1.4 2.39 x 10-6239 48 x 103 0.75 1.4 2.39 x 10-6a Temperature was kept at 15°C.b Calculated with kof of 44.7 x 103 and 36.2 x 103 mmol m-3 h-1 in biofilms

of 112- and 239-p.m thicknesses, respectively, and with a Vmax of 1.40 mmolg-' h-1 at 20°C, assuming a Qlo of 3.4.

important for modeling biofilm kinetics. A kof of 56 to 93mol So2- cm-3 h-' at 20°C was rather high compared with

results from other studies of the activity of sulfate reducersin mixed microbial communities. In marine sediments,where sulfate reduction is a very important mineralizationprocess, rates around 100- to 1,000-fold lower than ourresults have been found (15). The specific activity of thebacteria is probably in the same range as that found in thisstudy, but the density of sulfate reducers is much lower. Thepresence of excess lactate throughout our study caused abuildup of a high density of sulfate reducers compared withother high-rate systems. Only in areas rich in easily degrad-able organic matter can comparably high bacterial densitiesand rates be expected, e.g., in connection with the dischargeof wastewater or in wastewater treatment systems. Sulfideproduction from biofilms of unknown thicknesses has beenmeasured in sewer systems, and reported values are as highas 7.3 to 32 mM S2- m-2 h-1 (31). Assuming an active biofilmthickness of 250 p.m will give a kof of around 30 to 120 p.molcm-3 h-1, which is in the same range as the values reportedin this study. To my knowledge only one other study hasreported the measurement of sulfate uptake rates in biofilms(9). Using an inoculum from a creek sediment in a tubularbiofilm reactor at high sulfate concentrations, these re-searchers found kof of 200 to 500 ,umol So2- cm-3 h-1 at410C.The activity of the sulfate reducers was highly dependent

on temperature, with a Qlo of around 3.4. Comparably highvalues for Qlo of 3.4 to 3.5 for sulfate reduction have beenfound in marine coastal sediments and in salt marsh sedi-ments (1, 15), and a Qlo of 2.9 was found in eutrophicfresh-water lake sediment (12). This implies that the temper-ature effect is rather high on a fully penetrated biofilm.Another important parameter for the transport of sulfate

into the biofilm is the molecular diffusion coefficient, whichis often assumed to be about 80% of the value in pure water(36). This value was used in this study for modeling thebiofilm kinetics. Recently, molecular diffusion coefficientswithin aerobic heterotrophic biofilms have been reportedranging from 60 to nearly 100% of the value in pure water

30 NIELSEN

on October 18, 2020 by guest

http://aem.asm

.org/D

ownloaded from

BIOFILM DYNAMICS AND KINETICS 31

(23, 29). Therefore, determination of the diffusion coefficientwould improve the prediction accuracy of the model.The effect of internal diffusional resistance is seen very

clearly from the experimental results and the prediction fromthe model (Fig. 5). Despite the very high affinity for sulfateshown by the sulfate-reducing bacteria in the biofilms with aKm of around 1.4 ,uM, diffusional resistance would limitsulfate reduction in biofilms at much higher sulfate concen-trations, depending on the biofilm thickness. Therefore, it isalso very important to avoid any diffusional resistance whenKm for single bacteria in aggregates or flocs is determined.Microscopic investigations of the homogenized biofilrm inthis study showed only single bacteria and aggregates below5 to 10 ,um, so the diffusional resistance was not expected tobe significant in the determination of Km.The model can be used to predict sulfate reduction in

biofilm under different conditions. For instance, an impor-tant parameter such as temperature often affects the pro-cesses in biofilm. The overall effect of temperature on sulfatereduction in biofilm depends on the degree of penetration ofsulfate. If the biofilm is fully penetrated, a high temperaturedependence with a Qlo of 3.4 can be expected, arising onlyfrom the temperature dependence of the bacterial activity.However, if sulfate reduction in biofilms is limited bydiffusion, the amount of sulfate reduced is determined partlyby kof and partly by the diffusion coefficient. The tempera-ture dependence of the diffusion coefficient is much lowerthan was observed for kof, namely an increase of around 35%at temperatures of 10 to 20°C (18). Therefore, the apparentQlo for sulfate uptake and flux into the biofilm would beconsiderably lower than 3.4. For instance, the model pre-dicts that the apparent Qlo for sulfate uptake in thick biofilm(Fig. SB) would be only 1.8 at a sulfate concentration of 50,uM.

In most natural environments and in some high-rateanaerobic systems, the methanogenic bacteria are poorcompetitors with the sulfate reducers (17, 19, 28). However,some authors have reported the opposite in high-rateanaerobic reactors (14, 21). Several explanations have beenproposed for this observation, e.g., different substrate up-take kinetics for the bacteria, a more effective adhesion ofthe methanogenic bacteria to the filter carrier in the reactor(13, 14), and lack of a suitable energy-rich organic substratefor the sulfate reducers to outcompete the methanogenicbacteria (21). However, this study has shown that the masstransfer of sulfate may also be a very important factor forlimiting sulfate reduction in high-rate biofilm, especially atthe normally rather small sulfate concentrations found indomestic wastewater.The data and the model presented here show that in

high-rate systems sulfate reduction, like other processessuch as oxygen consumption and nitrate reduction (7), veryoften was limited by mass transfer of the electron acceptorinto a thick biofilm. The results presented here are based onlaboratory experiments under optimal nutritional conditions;therefore, the sulfate reduction rates found were probablyhigher than in most natural high-rate systems. However,with the calibration of a given system by using an actualzero-order volume constant, this model could be used tocalculate sulfate flux and penetration depth into a sulfate-reducing biofilm or, in some cases, into sediment.

ACKNOWLEDGMENTS

I thank T. Hvitved-Jacobsen and N. Iversen for helpful discus-sions and critical review of the manuscript. I also thank T. Larsen,

who contributed valuable help with the numerical solution of themathematical model.

This study was supported in part by Danish National ScienceFoundation grant 16-3677.H-747.

LITERATURE CITED1. Abdollahi, H., and D. B. Nedwell. 1979. Seasonal temperature as

a factor influencing bacterial sulfate reduction in a saltmarshsediment. Microb. Ecol. 5:73-79.

2. Benefield, L., and F. Molz. 1985. Mathematical simulation of abiofilm process. Biotechnol. Bioeng. 27:921-931.

3. Boon, A. G., and A. R. Lister. 1975. Formation of sulphide inrising main sewers and its prevention by injection of oxygen.Prog. Water Technol. 7:289-300.

4. Characklis, W. G., M. G. Trulear, J. D. Bryers, and N. Zelver.1982. Dynamics of biofilm processes: methods. Water Res.16:1207-1216.

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

6. Hamilton, W. A. 1985. Sulphate-reducing bacteria andanaerobic corrosion. Annu. Rev. Microbiol. 39:195-217.

7. Harramoes, P. 1978. Biofilm kinetics, p. 71-109. In R. Mitchell(ed.), Water pollution microbiology. John Wiley & Sons, Inc.,New York.

8. Hoehn, R. C., and A. D. Ray. 1973. Effects of thickness onbacterial film. J. Water Pollut. Control Fed. 45:2302-2320.

9. Holder, G. A., G. Vaughan, and W. Drew. 1984. Kinetic studiesof the microbiological conversion of sulphate to hydrogensulphide and their relevance to sulphide generation withinsewers. Water Sci. Technol. 17:183-196.

10. Ingvorsen, K., and B. B. J0rgensen. 1984. Kinetics of sulfateuptake by freshwater and marine species of Desulfovibrio.Arch. Microbiol. 139:61-66.

11. Ingvorsen, K., A. J. B. Zehnder, and B. B. J0rgensen. 1984.Kinetics of sulfate and acetate uptake by Desulfobacterpostgatei. Appl. Environ. Microbiol. 47:403-408.

12. Ingvorsen, K., J. G. Zeikus, and T. D. Brock. 1981. Dynamics ofbacterial sulfate reduction in a eutrophic lake. Appl. Environ.Microbiol. 42:1029-1036.

13. Isa, Z., S. Grusenmeyer, and W. Verstraete. 1986. Sulfatereduction relative to methane production in high-rate anaerobicdigestion: technical aspects. Appl. Environ. Microbiol.51:572-579.

14. Isa, Z., S. Grusenmeyer, and W. Verstraete. 1986. Sulfatereduction relative to methane production in high-rate anaerobicdigestion: microbiological aspects. Appl. Environ. Microbiol.51:580-587.

15. J0rgensen, B. B. 1977. The sulfur cycle of a coastal marinesediment (Limfjorden, Denmark). Limnol. Oceanogr. 22:814-832.

16. La Motta, E. J. 1976. External mass transfer in a biological filmreactor. Biotechnol. Bioeng. 18:1359-1370.

17. Lawrence, A. W., P. L. McCarty, and F. J. A. Guerin. 1966. Theeffects of sulfide on anaerobic treatment. Int. J. Air WaterPollut. 10:207-221.

18. Li, Y.-H., and S. Gregory. 1974. Diffusion of ions in sea waterand in deep-sea sediments. Geochim. Cosmochim. Acta38:703-714.

19. Lovley, D. R., and M. J. Klug. 1983. Sulfate reducers canoutcompete methanogens at freshwater sulfate concentrations.Appl. Environ. Microbiol. 45:187-192.

20. Maree, J. P., and W. F. Strydom. 1985. Biological sulphateremoval in an upflow packed bed reactor. Water Res.19:1101-1106.

21. Nanninga, H. J., and J. C. Gottschal. 1986. Anaerobic purifica-tion of waste water from a potato-starch producing factory.Water Res. 20:97-103.

22. Nethe-Jaenchen, R., and R. K. Thauer. 1984. Growth yields andsaturation constant of Desulfovibrio vulgaris in chemostat cul-ture. Arch. Microbiol. 137:236-240.

23. Onuma, M., T. Omura, T. Umita, and J. Aizawa. 1985. Diffusioncoefficient and its dependency on some biochemical factors.

VOL. 53, 1987

on October 18, 2020 by guest

http://aem.asm

.org/D

ownloaded from

32 NIELSEN APPL. ENVIRON. MICROBIOL.

Biotechnol. Bioeng. 27:1533-1539.24. Pomeroy, R. D., and J. D. Parkhurst. 1977. The forecasting of

sulfide buildup rates in sewers. Prog. Water Technol. 9:621-628.25. Postgate, J. R. 1984. The sulphate reducing bacteria, 2nd ed.

Cambridge University Press, Cambridge.26. Rittmann, B. E., and P. L. McCarty. 1978. Variable-order model

of bacterial-film kinetics. J. Env. Eng. Div. Am. Soc. Civ. Eng.104:889-900.

27. Robinson, J. A., and W. G. Characklis. 1984. Simultaneousestimation of Vmax, Ki,m and the rate of endogenous substrateproduction (R) from substrate depletion data. Microb. Ecol.10:165-178.

28. Schonheit, P., J. K. Kristjansson, and R. K. Thauer. 1982.Kinetic mechanism for the ability of sulfate reducers to out-compete methanogens for acetate. Arch. Microbiol. 132:285-288.

29. Siegrist, H., and W. Gujer. 1985. Mass transfer mechanisms ina heterotrophic biofilm. Water Res. 19:1369-1378.

30. Suidan, M. T., and Y.-T. Wang. 1985. Unified analysis of biofilm

kinetics. J. Env. Eng. Div. Am. Soc. Civ. Eng. 111:634-646.31. Thistlethwayte, D. K. B. (ed.). 1972. The control of sulphides in

sewerage systems. Butterworth, Sydney.32. Trulear, M. G., and W. G. Characklis. 1982. Dynamics of

biofilm processes. J. Water Pollut. Control Fed. 54:1288-1301.33. Wang, Y.-T., M. T. Suidan, and B. E. Rittman. 1986. Kinetics of

methanogens in an expanded-bed reactor. J. Env. Eng.- Div.Am. Soc. Civ. Eng. 112:155-170.

34. 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.

35. Williamson, K., and P. L. McCarty. 1976. A model of substrateutilization by bacterial films. J. Water Pollut. Control Fed.48:9-24.

36. Williamson, K., and P. L. McCarty. 1976. Verification studies ofthe biofilm model for bacterial substrate utilization. J. WaterPollut. Control Fed. 48:281-296.

on October 18, 2020 by guest

http://aem.asm

.org/D

ownloaded from