Vertical distribution of sulfate-reducing bacteria at the oxic-anoxic interface in sediments of the oligotrophic Lake Stechlin


  • Henrik Sass,

    1. Institute of Freshwater Ecology and Inland Fisheries, Department of Limnology of Stratified Lakes, Alte Fischerhütte 2, D-16775 Neuglobsow, Germany
    Search for more papers by this author
  • Heribert Cypionka,

    1. Institute for Chemistry and Biology of the Marine Environment, University of Oldenburg, P.O. Box 2503, D-26111 Oldenburg, Germany
    Search for more papers by this author
  • Hans-Dietrich Babenzien

    Corresponding author
    1. Institute of Freshwater Ecology and Inland Fisheries, Department of Limnology of Stratified Lakes, Alte Fischerhütte 2, D-16775 Neuglobsow, Germany
      Corresponding author. Tel.: +49 (33082) 69918; fax: +49 (33082) 69917; e-mail:
    Search for more papers by this author

Corresponding author. Tel.: +49 (33082) 69918; fax: +49 (33082) 69917; e-mail:


Vertical distribution of sulfate-reducing bacteria (SRB) and sulfate reduction rates were investigated in littoral and profundal sediments of the oligotrophic Lake Stechlin. SRB counted by the most probable number (MPN) technique showed two peaks, one at the oxic-anoxic interface, and the second deeper in the reduced sediment layer. Sulfate reduction rates determined as potential rates in anoxic sediment slurries showed a similar vertical distribution with the highest values in the anoxic zones. The highest cell numbers observed were 1.3×105 cells cm−3 with lactate as substrate. SRB were also found in oxic sediment layers in numbers similar to those detected in anoxic zones. From the highest positive MPN tubes 34 pure cultures were isolated. Physiological characterization including nutritional behaviour, tests for desulfoviridin and spore formation turned out that nearly all isolates belonged to the genera Desulfovibrio, Desulfomicrobium, and Desulfotomaculum. SRB from the oxic sediment layers revealed a higher oxygen tolerance and capacity of oxygen respiration than isolates from the anoxic sediment layers. However, no sulfate reduction was observed in the presence of oxygen, since oxygen was preferentially reduced.


Mineralization processes in sediments show a vertical sequence that follows the decreasing redox potentials of the electron acceptors available [1, 2]. Aerobic respiration takes place in the oxidized zones at the sediment surface, while sulfate reduction is thought to be characteristic for deeper reduced layers [3, 4]. While in marine sediments the dominating role of sulfate reduction in the anaerobic mineralization of organic matter is well established [5] its role in freshwater habitats is considered to be less important [3]. Freshwater usually has low sulfate concentrations (between 10 μM and 750 μM [6, 7]), and sulfate depletion occurs at higher sediment depths. For this reason, and since electron donors reach the sediment surface by sedimentation, SRB are forced to live near the oxic sediment layer.

It has been shown that sulfate-reducing bacteria are not as strict anaerobes as believed formerly. From several studies, evidence has been obtained that SRB are able to survive oxygen stress for several hours [8, 9], or even respire with oxygen [10–12]. This accords with recent reports that SRB can be isolated from oxic surface layers of marine and estuarine sediments [13–16] and that sulfate reduction proceeds in microbial mats at oxygen tensions near saturation [17–19]. However, no pure culture of sulfate-reducing bacteria has been isolated which carries out sulfate reduction in the presence of oxygen. Also only weak growth of pure cultures at low oxygen tensions has been demonstrated [20, 21].

While numerous studies revealed the presence of sulfate-reducing bacteria in the oxic zones of microbial mats [18, 19], biofilms [22], marine sediments [14, 16] or even in a brackish water column [23], less is known about this in limnic sediments [24]. Although sulfate reduction is more intense in habitats with a high nutrient supply [25], the present investigation was carried out in the sediments of an oligotrophic lake. This offered the advantage of less steep chemical gradients than in microbial mats or eutrophic lake sediments [3, 19, 22] and allowed us to isolate a number of different sulfate reducers.

2Materials and methods

2.1Site description and sample collection

The oligotrophic Lake Stechlin [26] is situated about 100 km north of Berlin. It has a surface area of 4.25 km2, a maximum depth of 68 m, and is stratified in summer with little hypolimnetic oxygen decrease.

Sediment samples were collected between October 1993 and April 1994 at two permanent sampling sites. The littoral sampling site was free of macrophytes and located about 15 m from the shore at a water depth of about 50 cm. The profundal sampling site was located in the middle of the southern bight at a water depth of 32 m. The temperature at the littoral sampling site ranged between 4°C (February 1994) and 10°C (October 1993), at the profundal site the temperature was 4°C during the whole period.

Sediment samples from the littoral site were taken by hand with plexiglass tubes (i.d. 50 mm, length 50 cm), while at the profundal site a Jenkin sediment sampler [27] was used. The sediment cores were sliced into 5 or 10 mm thick segments, and the samples were processed immediately. Porewater was obtained by centrifugation for 10 min with 14000×g at 4°C [28]. The supernatant was filtered (0.2 μm) and stored at −20°C until dissolved anions were analyzed.

Sediment density and porosity were determined by measuring the weight loss by drying sediment samples of known volumes and weights [28].

2.2Chemical determinations

Porewater concentrations of sulfate and thiosulfate were measured by ion chromatography with conductivity detection (Sykam, Gilching, FRG) [29]. Free sulfide in intact sediment cores was measured by a needle electrode with a tip diameter of 100 μm as described by Marschall et al. [20]. Oxygen and pH gradients were determined using a combined oxygen/pH-needle electrode (Diamond General, Ann Arbor, Michigan) connected with a picoammeter (Keithley 485, Germering, FRG) and a pH/mV-meter (WTW, Weilheim, FRG), respectively [30].

2.3Determination of sulfate reduction rates

Sulfate reduction rates were determined in slurries containing 10 cm3 of sediment diluted 1:2 with autoclaved anoxic water from the sampling site. Three or four replicates from each sediment depth were performed at in situ temperature. The experiments were carried out using 35SO42− (Amersham Buchler, Braunschweig, FRG) as described by Rosser and Hamilton [31]. Since only a comparison between the different layers was intended, the acid volatile sulfide was analyzed. Control experiments with the chromium reduction method [32] revealed that typically less than 50% of the reduced sulfate ended up as pyrite or elemental sulfur (Meier and Voigt, personal communication).

Sulfate reduction rates of oxic and anoxic sites were both determined in anoxic sediment slurries. Thus potential anaerobic activities were obtained, but not sulfate reduction in the presence of oxygen.

2.4Enumeration and isolation procedures

Enumerations of culturable sulfate reducers were performed using the most-probable-number (MPN) technique with five tubes in parallel per tenfold dilution step. Four different MPN series were made. For lactate-utilizing SRB, the bicarbonate-buffered mineral medium described by Widdel and Bak [33] with 20 mM lactate, 10 mM sulfate and resazurin as redox indicator, and in addition Postgate's medium B, containing phosphate buffer, yeast extract, ascorbate, and thioglycollate [33] were used. Enumerations of acetate-utilizing SRB were carried out with the mineral medium [33] containing 20 mM acetate and 10 mM sulfate. Thiosulfate-disproportionating bacteria were counted with 15 mM thiosulfate and 2 mM acetate as a carbon source.

The MPN tubes were incubated at 28°C in the dark for five weeks (lactate-utilizing SRB) or three months (acetate-utilizing SRB and thiosulfate-disproportionating bacteria).

MPN tubes were considered to be positive if beneath an increase of turbidity by bacterial growth the formation of sulfide was detected [34]. In enumerations of thiosulfate-disproportionating bacteria the formation of sulfide and sulfate [34, 35] was analyzed.

In order to quantify spores of lactate-utilizing SRB, a separate series of MPN tubes filled with the medium of Widdel and Bak [33] were pasteurized for 15 min at 65°C in a water bath after inoculation and incubated as described above.

For comparative studies on the physiology of SRB dominating in the different sediment layers, pure cultures were obtained from the highest diluted but positive MPN tubes by repeated application of the deep agar dilution method [36]. From each sediment depth at least two isolates were obtained, originating from different MPN tubes.

2.5Physiological tests

Microscopic observations were performed using a Leica DMRB microscope. The formation of spores was analyzed in growth tests after pasteurization for 15 min at 65°C.

Substrate utilization was tested in completely filled screw cap tubes using the medium of Widdel and Bak [33]. Electron donors were used in concentrations between 5 mM and 20 mM, while 10 mM of the different electron acceptors were added. In tests for lithotrophic growth on H2, formate or by disproportionation of thiosulfate and sulfite, 2 mM acetate served as carbon source.

Growth rates of SRB were determined as increase in turbidity using a turbidimeter (Hach, Loveland, CO, USA). Growth was determined at 28°C using the medium described by Widdel and Bak [33] with lactate and sulfate as substrates.

Aerobic respiration and sulfate reduction were measured in a multi-electrode-chamber [38] with washed cells in 150 mM KCl [39]. Catalase activity was tested by use of H2O2[39].

The presence of desulfoviridin was proven by the fluorescence test of Postgate [37].


3.1Physical sediment parameters

The littoral sediment consisted of coarse-grained sand. A light-brown layer of about 20 mm to 30 mm thickness at the surface was followed by a black layer extending to about 100 mm depth. The black color indicated the presence of FeS. The porosity of the littoral sediment varied between 0.4 below 60 mm depth and 0.6 in the uppermost 10 mm. Sediment density ranged from 1.65 g cm−3 at the surface to 1.85 g cm−3 below 10 mm depth. The profundal sediment was uniformly dark brown over the upper 100 mm. The mean porosity was 0.91 while the average density was determined as 1.07 g cm−3. The physical parameters of the profundal sediment showed only little changes in the upper 100 mm.

3.2Chemical sediment profiles

In December 1993 at the profundal sampling site, oxygen penetrated up to 10 mm into the sediment (Fig. 1). Just below the oxic layer, concentrations of free sulfide increased from 0 to 300 μM, and decreased steeply again below 30 mm. The sulfate concentrations declined rapidly below 25 mm to depletion at about 80 mm (Fig. 1a). Thiosulfate was not detectable (detection limit 5 μM). The pH was in the range of 7.2 to 7.4 over the upper 80 mm (data not shown). In March 1994 the profiles were similar to those observed in December 1993, with an exception for sulfide. The sulfide concentrations were at least tenfold lower and free sulfide was detectable only 10 mm below the oxic/anoxic interface. At the littoral sampling site, the profiles of oxygen and sulfide were similar to those in the profundal sediment (Fig. 2). In October 1993 sulfide was detectable just below the oxic layer, while in February 1994 a distance of about 10 mm was observed between the oxic and the sulfide-containing sediment layers. However, caused by weather-induced water movement, the oxygen penetration depths varied between 6 mm and 14 mm (data not shown). After these events the sulfide-containing layer was shifted to higher sediment depths. The sulfate concentrations did not decrease below 150 μM in the upper 80 mm (Fig. 2). As in the profundal sediment, thiosulfate was not detectable.

Figure 1.

Depth profiles of oxygen (•), sulfate (□), and sulfide (◯), vertical distribution of lactate-utilizing sulfate-reducing bacteria (SRB), and sulfate reduction rates (SRR, bars indicating standard deviation) in the upper sediment layer at the profundal site of Lake Stechlin. (a) Recorded in December 1993. (b) Recorded in March 1994.

Figure 2.

Depth profiles of oxygen (•), sulfate (□), and sulfide (◯), vertical distribution of lactate-utilizing sulfate-reducing bacteria (SRB) and sulfate reduction rates (SRR, bars indicating standard deviation) in the upper sediment layer at the littoral site of Lake Stechlin. (a) Recorded in October 1993. (b) Recorded in February 1994.

3.3Sulfate reduction rates

At the profundal sampling site, the sulfate reduction rates were low (up to 1.66 nmol SO42− d−1 cm−3, Fig. 1). Even lower capacities were detected in the oxic layers (tested under anoxic conditions). There was little difference between the maximum values determined in December 1993 and in March 1994 (Fig. 1), while the depth integrated rates were about twofold higher in March.

At the littoral sampling site the highest sulfate reduction rates were observed in October 1993 (3.4 nmol SO42− d−1 cm−3, Fig. 2). In the oxic surface layer the potential sulfate reduction rates were similar to those in the anoxic zone. In February 1994, the rates were far lower, with a maximum of 0.45 nmol SO42− d−1 cm−3 (Fig. 2) and depth integrated rates being 15-fold lower than in October.

3.4Enumeration of sulfate-reducing bacteria

In all enumerations from both sampling sites (Figs. 1 and 2) the vertical distribution of lactate-utilizing, sulfate-reducing bacteria showed two peaks. The first peak occurred near the oxic-anoxic interface, whereas a second peak was located in the permanently anoxic layer from 20 to 30 mm depth.

The MPN counts obtained with Postgate's medium B and the defined medium described by Widdel and Bak gave differences of less than 20%. The highest number of lactate-utilizing SRB found was 1.3×105 cells cm−3 (Fig. 1a).

Pasteurization of profundal sediment samples for 15 min at 65°C lead to a drastic decrease of the viable cell counts to less than 0.1% in the upper sediment layers, and about 1% below 50 mm depth.

Acetate-utilizing and thiosulfate-disproportionating SRB in the profundal sediment revealed culturable population sizes of 10- to 100- fold lower than lactate-utilizing SRB (Fig. 3). The highest numbers achieved were 6×103 cells cm−3 and 1.3×103 cells cm−3, respectively (Fig. 3). The maximum counts were found near the oxic-anoxic interface (Fig. 3).

Figure 3.

Vertical distribution of acetate-utilizing sulfate-reducing bacteria (left panel) and thiosulfate-disproportionating bacteria (right panel) at the profundal site of Lake Stechlin (recorded December 1993).

3.5Characterization of isolates

From the highest dilutions of the MPN series 34 sulfate-reducing bacteria were isolated in pure culture. Genomic fingerprinting based on ERIC-PCR (enterobacterial repetitive intergenic consensus sequences) [40] indicated that among these 31 different strains were present (Sass et al., in preparation).

Twelve of thirteen strains isolated from the littoral site had characteristics typical for the genus Desulfovibrio, e.g. the presence of desulfoviridin, motile cells and the capacity to utilize hydrogen and lactate (Table 1). Similarly, in the upper layers of the profundal sediment only Desulfovibrio-like isolates were obtained. Below 2 cm depth spore-forming isolates became dominant. With one exception these were all able to utilize methanol, a capacity which relates them to Desulfotomaculum orientis[41].

Table 1.  Physiological properties of SRB isolated from littoral and profundal sediments of Lake Stechlin
 Origin (cm) (h)Motil.Electron donorsElectron acceptorsDisproportionat.
  1. Origin: depth of sediment layer from which isolates were obtained. Spor.: ability of forming spores. Dvir.: Desulfovridin. td: doubling time. Motil.: motility. (+): activity very weak or present only in part of the experiments. n.d.: not determined.

Littoral strains
STL20–1+19.5+++++ ++++
STL30–1+22.8+++++ ++
STL40–1+22.1++++ ++
STL51–1.5+12.4++++++ +++
STL62–3+14.6++++++ ++
STL72–3+34.6++++++ +++++
STL83–413.1+++++ ++
STL93–4+9.3++++ ++
STL104–5+18.6+++++ ++++
STL127–8+65.1++++++Malate, Succinate++++
STL137–8+31++++++ ++
Profundal strains
STP40–0.5+5.5++++ ++
STP70–0.5+18.4++++++ +++(+)
STP50.5–1+30.1++++ ++++
STP80.5–1+31.4+++ +++
STP91–1.5+29.9++++++ ++
STP101.5–224.3+++++ ++
STP162–317.0++++ +
STP123–4+20.9++++Methanol, Butyrate+
STP153–4+23.6++++ ++
STP145–6+12.5++(+)++Methanol, Alanine
STP137–8+13.2+++Methanol, Butyrate++
STP38–12+14.2+++++Methanol, Malate
Acetate-utilizing SRB
STP231.5–232.0+++++Acetate, Propionate+++
Thiosulfate-disproportionating SRB
STP310.5–1+20.8++++ +++++
STP341.5–251.6+++ +(+)

Most of all isolates were able to grow with sulfite or thiosulfate as an electron acceptor, while nitrate was used only by five Gram-negative strains. Growth by disproportionation of sulfite or thiosulfate was rare and found only among Desulfovibrio-like strains (Table 1).

Generally, the isolates grew slowly. Only with three strains doubling times below ten hours were observed (Table 1).

Most of the strains from the littoral site exhibited catalase activity and were able to respire with oxygen as an electron acceptor. Hydrogen, formate and lactate were the substrates most commonly used (Table 2). Similar results were obtained with isolates from the profundal sediments, though the number of strains able to respire with O2 was smaller and a lower number of substrates was used. Isolates from the permanently anoxic sediment layers generally oxidized a smaller range of substrates with oxygen or were unable to respire at all.

Table 2.  Substrates oxidized with O2 by SRB from littoral and profundal sediments of Lake Stechlin
 Origin (cm)Catalase activitySubstrates oxidized with O2
  1. Origin: depth of sediment layer from which isolates were obtained. (+): Activity very weak or present only in part of the experiments.

Littoral strains
Profundal strains

None of the isolates was able to grow in the presence of oxygen. As reported elsewhere [39] strains isolated from the deeper layers were killed if oxygen (above 0.1% of the headspace volume) was added to growing cultures. Strains from the oxic layers immediately consumed oxygen (up to 2.5% of 400 ml headspace above 100 ml culture) and continued growing after oxygen depletion.

No sulfate reduction was observed in the presence of oxygen and oxygen was the preferred electron acceptor.


4.1Vertical distribution of sulfate-reducing bacteria

Our study has revealed that sulfate-reducing bacteria have a distinct vertical distribution pattern in Lake Stechlin sediments and that they are present also in the oxic layers of an oligotrophic environment. Enumerations carried out at different seasons and with two different growth media gave similar results. The existence of two peaks in cell counts suggests the existence of different populations of culturable sulfate-reducing bacteria along the vertical profile. This is comparable to results obtained in other environments. Risatti et al. [42] found a layered distribution of different SRB in a microbial mat by the use of oligonucleotide probes. The Desulfovibrio probe indicated a peak in the uppermost layer of the mat containing the oxic/anoxic interface and a second peak in the permanently anoxic zone. By MPN techniques and oligonucleotide probes Teske et al. [23] showed that in the water column of a fjord SRB are present in oxic layers with an increase in cell numbers at the oxycline.

We have found mainly Desulfovibrio-like SRB (Table 1) in the upper and lower peak. However, a molecular analysis of 34 isolates (Sass et al., in preparation) indicated a high genetic diversity of the isolates. Although the most isolates are relatives of the genus Desulfovibrio, among all strains only three were found which were identical to another. The capacity of aerobic respiration, catalase activity and resistance against oxygen stress during growth [39] indicated that the SRB in the upper layer are especially well adapted to live near the oxic sediment layer. This is particularly the case at the littoral sampling site where advective porewater transport [43] induced by wind may cause enhanced oxygen penetration.

In the profundal sediments of Lake Stechlin in deeper sediment layers spore-forming sulfate reducers of the genus Desulfotomaculum were dominating. Considering the low nutrient supply at this habitat this is not surprising as spores can overcome long periods of starvation. On the other hand, our finding that only a small fraction of these bacteria was sporulated (and survived pasteurization prior to dilution in MPN series) is difficult to explain. Bak and Pfennig [24] reported that in littoral sediments of Lake Constance about 50% of the acetate-utilizing Desulfotomaculum cells were present as spores, hundredfold more than found in Lake Stechlin.

Acetate-utilizing SRB were relatively less abundant in Lake Stechlin sediments than in marine habitats [16]. In this oligotrophic freshwater habitat their numbers increased to about 30% of that of lactate-utilizing SRB at the sediment surface and below 50 mm depth (Fig. 3). The relative increase in cell numbers of acetate-utilizing SRB with depth seems to be a common feature of different habitats as microbial mats [19, 41] or sediments [16]. Bak and Pfennig [24] determined even far lower numbers of acetate-utilizing SRB in Lake Constance sediments. However, they found bacteria able to carry out thiosulfate disproportionation to be as abundant as H2-utilizing SRB, while in Lake Stechlin this type of bacteria seems to play a less important role.

4.2Sulfate reduction rates

Although we found relatively high numbers of SRB in the permanently oxic zones of the sediments we do not know whether these bacteria carry out sulfate reduction in the presence of oxygen or use O2 as electron acceptor. All sulfate reduction rates were measured in the absence of O2 and thus represent potential rates. Furthermore, since a slurry method was used (which gives higher rates than core injection experiments) [44], and only acid-volatile sulfur compounds were analyzed (which may result in an underestimation of 50% in Lake Stechlin sediments, Meier and Voigt, pers. communication) our rates should be regarded only for internal comparison. The highest rates were found below the oxic-anoxic interface. The littoral sediments showed pronounced differences in sulfate reduction rates between October and March. On the one hand this can be explained by the higher temperature in October. But since Bak and Pfennig [28] found a Q10 of 2.25 in littoral sediments of Lake Constance and Babenzien [45] stated a Q10 value of 3.0 for profundal sediment of Lake Stechlin, this could hardly serve as the sole explanation for the differences obtained. Therefore the higher nutrient supply caused by the decay of macrophytes and infallen leaves from the forests surrounding the lake during autumn has to be taken into account.

At the profundal site sedimentation is the only source of organic carbon, which is relatively constant throughout the year. So it is not surprising to find only small differences in the activities. These differences however might be considered to be caused by an enhanced input of organic matter due to the spring bloom.

Although the sediments of an oligotrophic lake seem not to be very typical in terms of microbial activities they offer some advantages for the examination of microbial populations as they exhibit a high diversity of organisms. The less steep gradients with a permanent oxic layer make it possible to investigate layers characterized by different chemical conditions separately. For this reasons we were able to reveal that in oxic layers of freshwater habitats SRB are present and that oxygen tolerance is a feature that has an strong influence on the distribution of SRB in freshwater sediments.


The authors wish to thank Monika Degebrodt for technical assistance. This work was supported by a grant of the Deutsche Forschungsgemeinschaft.