The microbial community structure of different ... - Wiley Online Library

Blackwell Science, LtdOxford, UKEMIEnvironmental Microbiology 1462-2912Society for Applied Microbiology and Blackwell Publishing Ltd, 200472281293Original ArticleDistribution of bacterial fatty acids and FISHS. I. Bühring, M. Elvert and U. Witte

Environmental Microbiology (2005) 7(2), 281–293

doi:10.1111/j.1462-2920.2004.00710.x

The microbial community structure of different permeable sandy sediments characterized by the investigation of bacterial fatty acids and fluorescence in situ hybridization S. I. Bühring,*† M. Elvert† and U. Witte Max Planck Institute for Marine Microbiology, Celsiusstr. 1, 28359 Bremen, Germany. Summary This study describes the microbial community structure of three sandy sediment stations that differed with respect to median grain size and permeability in the German Bight of the Southern North Sea. The microbial community was investigated using lipid biomarker analyses and fluorescence in situ hybridization. For further characterization we determined the stable carbon isotope composition of the biomarkers. Biomarkers identified belong to different bacterial groups such as members of the CytophagaFlavobacterium cluster and sulfate-reducing bacteria (SRB). To support these findings, investigations using different fluorescent in situ hybridization probes were performed, specifically targeting CytophagaFlavobacterium, g-Proteobacteria and different members of the SRB. Depth profiles of bacterial fatty acid relative abundances revealed elevated subsurface peaks for the fine sediment, whereas at the other sandy sediment stations the concentrations were less variable with depth. Although oxygen penetrates deeper into the coarser and more permeable sediments, the SRB biomarkers are similarly abundant, indicating suboxic to anoxic niches in these environments. We detected SRB in all sediment types as well as in the surface and at greater depth, which suggests that SRB play a more important role in oxygenated marine sediments than previously thought. Introduction Sandy sediments cover large areas of the continental

Received 11 February, 2003; revised 8 July, 2004; accepted 15 July, 2004. *For correspondence. E-mail solveig.buehring@ uni-bremen.de; Tel. (+49) 421 2188934; Fax (+49) 421 2188664. † Present address: Research Center Ocean Margins, University of Bremen, Am Fallturm 1, 28359 Bremen, Germany.

© 2005 Society for Applied Microbiology and Blackwell Publishing Ltd

shelf. During the last decade much scientific effort has been directed at understanding this environment (e.g. Huettel et al., 1998; Jahnke et al., 2000; Rusch et al., 2001). Huettel and Gust (1992) were among the first to describe how the interaction of bottom currents with sediment topography gives rise to pressure gradients which induce advective porewater flow, an exchange process exclusively found in permeable sediments. Sandy sediments with a permeability exceeding 10-12 m2 are thought to allow advective porewater transport (Huettel and Gust, 1992) and thereby support aerobic and suboxic metabolism in deeper sediment layers. This process often homogenizes otherwise separated microbial populations by inducing changes in oxygen distribution and penetration. Despite the low abundance of bacteria in sandy sediments compared to finer compartments (Llobet-Brossa et al., 1998), bacterial assemblages in sandy sediments are characterized by high organic carbon turnover rates (Huettel and Rusch, 2000; Rusch et al., 2000) as well as elevated rates of oxygen consumption (F. Janssen, M. Huettel and U. Witte, submitted). To date, little attention has been paid to the composition of microbial communities inhabiting permeable sediments. Rusch and colleagues (2003) studied the spatio-temporal variation and metabolic activity of the microbial community in a coarse-grained sediment and found that members of different aerobic and anaerobic bacteria generally had a low abundance, but high organic matter turnover rates. For the investigation of the composition of the bacterial community, fluorescence in situ hybridization (FISH) was applied, using group-specific fluorescently labelled rRNAtargeted oligonucleotides. This new and promising approach of in situ detection has the advantage that it does not rely on isolation and cultivation. Since the first descriptions (Amann et al., 1990) and applications of this method (Llobet-Brossa et al., 1998), FISH has often been used to quantify microbial communities in marine (e.g. Ravenschlag et al., 2001; M. Mussmann, K. Ishii, R. Rabus and R. Amann, submitted) and freshwater sediments (Altmann et al., 2003). The advantage of FISH compared with PCR-based approaches to study marine sediments (e.g. Cifuentes et al., 2000) is the possibility to quantitatively determine morphology and in situ spatial

282 S. I. Bühring, M. Elvert and U. Witte distribution of the microbial community in their natural habitat (Amann et al., 1995). A second important approach to investigate bacterial communities in situ is the analysis of specific lipid biomarkers (Rajendran et al., 1992; Hinrichs et al., 1999; Rütters et al., 2002b). The extraction and identification of fatty acids is a useful tool, because bacteria synthesize highly specific components, such as branched chain fatty acids (Kaneda, 1991). Pioneering work on fatty acid composition of several bacterial strains was carried out by Dowling and colleagues (1986), further studies then transferred the knowledge from culture analysis to the interpretation of natural prokaryotic communities (e.g. Rajendran et al., 1992; Oude Elferink et al., 1998; Orphan et al., 2001; Elvert et al., 2003; Wakeham et al., 2003). The information gained from biomarker studies can be further increased by simultaneous determination of the stable carbon isotope composition. Natural abundance studies use the small difference in stable carbon isotope ratios which can be used to identify specific biogeochemical processes (Freeman et al., 1990; Collister et al., 1992; Elvert et al., 1999; Hinrichs et al., 1999; Elvert et al., 2000) as well as different trophic levels in the food web structure (Werne et al., 2002). The carbon source as well as the discrimination between 13C and 12C caused by enzymatic reactions affects the stable carbon isotope composition (Peterson, 1999; Hayes, 2001). So far, few studies have used the combination of molecular and biomarker approaches (e.g. Orphan et al., 2001; Elvert et al., 2003), which provide an independent verification of the bacterial distributions in situ. In this study, we combined lipid biomarker analysis with investigations targeting bacterial taxa using FISH probes to study microbial communities on three different permeable sands in the southern North Sea. The study specifically investigates the vertical colonization patterns of bacteria in sandy sediments and its relationship to sediment permeability. Results and discussion General observations Bacterial biomass varied distinctly during the investigation of the three sands. During April and June, 169.6 ± 36 mg ml-1 and 63 ± 10 mg ml-1, respectively, were found on the fine sand. In the medium sand the bacterial biomass accounted for 16.7 ± 16 mg ml-1 and on the coarse sand for 6.5 mg ml-1. The higher biomass on the finer sands can be explained by the larger specific surface area (Dale, 1974) and higher porewater content. Decreasing abundances of bacteria in coarser grained sediments have also been documented by Llobet-Brossa and colleagues (1998). Despite their low abundances, bacteria in permeable sediments show high activities, as illustrated,

for example, by high ammonification rates in the porewater (Ehrenhauss et al., 2004). Advective porewater exchange occurring in sediments with a permeability exceeding 10-12 m2 (Huettel and Gust, 1992) has been suggested as the cause of this high activity. Advective porewater movements provide a fast transport mechanism for the exchange of substances between the water column and the upper sediment layers. Investigations by Janssen and co-workers (F. Janssen, M. Huettel and U. Witte, submitted) on porewater exchange at the same three sediment stations revealed nearly no advective exchange processes for the fine sand, but greatly increased values for the medium and the coarse sands. This circulation of water through the sediment matrix leads to enhanced solute transport processes. Fatty acid distribution patterns The investigation of the fatty acid composition revealed a complex suite of different fatty acids at the three sand stations. The depth profiles were dominated by the evenchained fatty acids C16:1w7, C16:0, C18:1w7 and C18:1w9. The fatty acids C14:0, aiC15:0, iC15:0 and 10Me-C16:0 followed in abundance. All other fatty acids could only be detected in minor amounts. No cyclic fatty acids or archaeal-derived lipids were observed. The most abundant fatty acids, C16:1w7, C16:0 and C18:1w7, are generally common in marine sediments (e.g. Boschker et al., 2001; Rütters et al., 2002a,b). C16:0 is ubiquitous in marine life forms and is extensively biosynthesized de novo, and therefore not suitable for detailed biomarker studies (Sargent and Whittle, 1981). In contrast, monounsaturation at the w7 position is generally considered as typical for bacteria (Sargent and Whittle, 1981; Lechevalier and Lechevalier, 1988). C16:1w7 and C18:1w7 are constituents of marine phytoplankton (e.g. Birgel et al., 2004), but are also commonly associated with bacteria (Abraham et al., 1998; Boschker and Middelburg, 2002) and the depleted d13C values of down to -27 and -30‰, respectively, make an exclusive phytoplankton source unlikely (Table 2). Branched-chain fatty acids such as iC15:0, aiC15:0 and iC16:0 are major constituents of Gram-positive bacteria (Lechevalier and Lechevalier, 1988) but are also present in other anaerobic bacteria (Findlay and Dobbs, 1993). Our study revealed higher relative abundances for aiC15:0 compared with iC15:0 on the fine sand during both seasons and in the upper and lower part of the sediment (Fig. 1). With increasing grain size the relative amount of aiC15:0 subsequently decreased, with equal concentrations of both isomers on the coarse sediment. Nevertheless, the stable carbon isotope ratios of iC15:0 and aiC15:0 of approximately -20‰ indicate a marine source of these fatty acids, suggesting that the source bacteria are not chemoautotrophs (Degens, 1969; Ruby et al., 1987).

© 2004 Society for Applied Microbiology and Blackwell Publishing Ltd, Environmental Microbiology, 7, 281–293

Distribution of bacterial fatty acids and FISH 283 Table 1. Station characteristics.

Cruise

Date

Position

HE 145

08.-18.04.01

HE 148

07.-15.06.01

HE 154

24.-30.09.01

53∞51¢N, 007∞44¢E 53∞51¢N, 007∞44¢E 53∞50¢N, 007∞45¢E 53∞49.5¢N, 007∞44.5¢E

Permeability k (10-12 m2)

Average water depth (m)

Water temperature (∞C)

Sediment POC (% dry mass)

3.02 (±1.66)

19

9

3.02 (±1.66)

19

13

0.114 (±0.014)

26.27 (±3.26) 77.24 (±14.36)

16 14

16 16

0.023 (±0.003) 0.032 (±0.003)

The mesophilic, Gram-negative sulfate-reducing bacteria (SRB) form coherent groups within the d-subdivision of the proteobacteria and C17-fatty acids are generally described as typical for these bacteria (Boschker and Middelburg, 2002). Findlay and Dobbs (1993) stated that 10Me-C16:0, iC17:0 and aiC17:0 are typical membrane components of SRB. These fatty acids were present with relative abundances of 4.0–5.1% in the upper layer and of 4.1–6.5% in the deeper sediment layer at all investigated sand stations (Fig. 1). Absolute concentrations of 10MeC16:0 revealed subsurface peaks in the fine and medium sand (Fig. 2). 10Me-C16:0 is considered to indicate Desulfobacter species (Dowling et al., 1986) and more generally the family Desulfobacteriaceae (Kuever et al., 2001; Rütters et al., 2002b). The concentrations in the coarse sediment seemed to be less variable with depth, suggesting little sediment stratification, and probably arose as a consequence of strong advective flushing. The d13C measurements of 10Me-C16:0 reflect an origin dominated by marine sources (d-values around -22.8‰) which showed little variation between the sand type and seasons. IC17:1w7 and aiC17:0 accounted for approximately 1% of all fatty acids present in our study, but showed slightly elevated concentrations in deeper sediment layers (Fig. 1). These fatty acids are common cellular constituents in SRB, such as Desulfosarcina species (Rütters et al., 2001). Down-

n.a.

Notes 6 cores (Ø200 mm) +2 cores (Ø36 mm) 2 cores (Ø200 mm) +2 cores (Ø36 mm) 5 cores (Ø200 mm) +2 cores 1 cores (Ø200 mm) +1 core (Ø36 mm)

core profiles are comparable and show a subsurface peak at 9.5 cm (Fig. 2). Moreover, both fatty acids displayed the most 13C-enriched carbon isotope values of all investigated bacterial fatty acids with mean values of all stations of -17.1 and -17.6‰ for iC17:1w7 and aiC17:0 respectively (Table 2). IC17:1w7 is also described as a marker fatty acid diagnostic for Desulfovibrio (Taylor and Parkes, 1983; Coleman et al., 1993; Llobet-Brossa et al., 2002; Londry et al. 2004). Members of the genus Desulfovibrio have been shown to be able to respire with oxygen (Dilling and Cypionka, 1990), which would provide an explanation for the higher concentrations of iC17:1w7 in upper sediment layers (Fig. 2), although they do not seem to grow under oxic conditions in the laboratory (Dannenberg et al., 1992). Microbial ecology of sandy sediments Figure 3 displays the percentages of bacteria detected with probe EUB 338 in comparison with all bacteria detected by 4¢,6¢-diamidino-2-phenylindole (DAPI) counts in selected sediment horizons. Up to 65% of the DAPIstained cells hybridized with EUB 338. Only in the upper layers of the fine and coarse sediments was the proportion of EUB 338 higher than 50%, whereas in deeper layers of the coarse sediment the fraction of detectable bacteria

Table 2. Carbon isotope ratios of different investigated bacterial fatty acids on the fine and medium sediment (medium values from measurements over the whole sediment sampling depth are presented, numbers in brackets indicate S.D.). Bacterial fatty acid

Fine sediment (April)

Fine sediment (June)

Medium sediment (September)

iC14:0 iC15:0 aiC15:0 iC16:0 C16:1w7 C16:1w5 iC17:1w7 10Me-C16:0 iC17:0 aiC17:0 C17:1w8 C17:1w6 C18:1w7

-23.3 -22.6 -21.6 -23.4 -27.1 -35.1 -14.6 -23.7 -21.8 -18.6 -24.6 -24.8 -30.0

-20.8 -20.1 -19.6 -25.6 -24.9 -30.7 -19.0 -22.2 -21.3 -16.2 -21.1 -22.1 -25.5

-20.5 -18.4 -21.4 -19.0 -21.5 -24.1 -17.6 -22.5 -21.1 -17.9 -24.4 -23.0 -24.4

(±1.0) (±1.9) (±1.4) (±1.8) (±1.0) (±4.7) (±1.3) (±1.0) (±1.3) (±2.5) (±1.9) (±1.4) (±1.1)

(±2.2) (±0.6) (±0.6) (±2.7) (±0.8) (±1.6) (±0.8) (±0.8) (±2.2) (±1.8) (±0.9) (±0.9) (±0.6)


(±1.9) (±1.0) (±0.4) (±1.8) (±1.7) (±1.5) (±0.9) (±1.1) (±1.4) (±1.6) (±1.9) (±1.2) (±1.1)

284 S. I. Bühring, M. Elvert and U. Witte 0-3 cm

7-10 cm 30

Relative abundance [%]


30 25



20

20

15

15

10

10

5

5

0 30

0 30

25




25

20

20

15

15

10

10

5

5 0 30

0 30 25



25

20

20

15

15

10

10

5

5 0 30

0 30 25

Coarse sediment (September)

Coarse sediment (September)

25 20

15

15

10

10

5

5

0

0 iC14:0 C14:0 iC15:0 aiC15:0 C15:0 iC16:0 C16:1w9 C16:1w7 C16:1w5 C16:0 iC17:1w7 10Me-C16:0 iC17:0 aiC17:0 C17:1w8 C17:1w6 C17:0 C18:1w9 C18:1w7 C18:0 C20:4w3 C20:5w3 C22:6w3

20

iC14:0 C14:0 iC15:0 aiC15:0 C15:0 iC16:0 C16:1w9 C16:1w7 C16:1w5 C16:0 iC17:1w7 10Me-C16:0 iC17:0 aiC17:0 C17:1w8 C17:1w6 C17:0 C18:1w9 C18:1w7 C18:0 C20:4w3 C20:5w3 C22:6w3


25

Fig. 1. Relative abundance [%] of fatty acids of the three sands. Grey bars indicate fatty acids of presumed bacterial origin. The fine sand was investigated during April and June and the medium and coarse sands during September The distribution in 0–3 cm sediment depth is plotted on the left and in 7–10 cm sediment depth on the right. The vertical bars indicate S.D.

was very low. The cells that are not labelled with EUB 338 could be either Archaea or bacteria with a weak fluorescence signal. Nevertheless, the amount of Archaea is described as low in sandy sediments (K. Ishii, M. Mussmann, B. J. MacGregor and R. Amann, submitted). These

authors found less than 1% of DAPI-stained cells accounting for Archaea in a sandy tidal flat sediment of the German Wadden Sea. With a set of six probes for mayor phyla within the domain bacteria, we were able to affiliate 19–70% of the


Distribution of bacterial fatty acids and FISH 285 iC17:0 0

2

2

4

4

6

6

8

8

0.0

0.2

0.4

1.2

1.4

0.0

0.2

0.4

12 0.8 2

0.6

aiC17:0

4 0 6 2

8 4

8 4

10 6

10 6

12 8

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

0.0

0.2

0.4

0.5

1.0

1.5

2.0

0.0

0.1

iC17:1w7

4 0

0.8

aiC17:0

aiC15:0

0.0

0.6

12 8 10 0

0.2

0.3

0.4

C17:1w8

12 2 4 0

6 2

6 2

8 4

8 4

10 6

10 6

12 8

0.0

0.5

1.5

2.0

0.0

0.1

iC17:1w7

10 0 0.0

0.1

0.2

0.3

0.4

C17:1w8

0.3

0.4

0.5

0.0

0.1

0.2

10Me–C16:0

4 0

0.2

0.3

12 8 10 0

0.4

0.5

0.6

C17:1w6

12 2 4 0

6 2

6 2

8 4

8 4

10 6

10 6

12 8 –0.1

0.0

0.1

8

0.4

0.5

0.0

0.1

0.2

0.2

0.4

0.6

0.8

0.3

0.4

0.5

0.6

C17:1w6

1.0

1.2

1.4

Abundance [µg g–1 sediment dry weight]

1.6

0.0

0.1

0.2

12 8 10 0

0.3

Abundance [µg g–1 sediment dry weight] Fine sediment (April) Fine sediment (June) Medium sediment (September) Coarse sediment (September)

12 0.4 2 4 6 8

Sediment depth [cm

12 2–0.2 0.0

6

0.3

10Me–C16:0

10 0

4

0.2

Sediment depth Sediment [cm] depth [cm]

12 2–0.1

1.0



1.0

6 2

12 2


0.8

aiC15:0

4 0

10 0

ediment depth [cm]

0.6

10 0



12 2

iC17:0

iC15:0

10 0

Sediment depth [cm]

Sediment depth [cm]

iC15:0 0

Fig. 2. Depth distribution of different bacterial fatty acids on the fine sand in April and June and on the medium and coarse sands during September. The horizontal bars indicate S.D. © 2004 Society for Applied Microbiology and Blackwell Publishing Ltd, Environmental Microbiology, 7, 281–293

286 S. I. Bühring, M. Elvert and U. Witte Fine sediment (June) 0-0.5 cm

Fine sediment (June) 9-10 cm

Unknown

39%

EUB 338

Fig. 3. Share of EUB 338 targeted cells from all DAPI counts of the three different sands in 0–0.5 cm depth on the left and in 9–10 cm depth on the right.

49%

51%

61%

Medium sediment (September) 0-0.5 cm

Medium sediment (September) 9-10 cm

49%

46%

54%

51%

Coarse sediment (September) 0-0.5 cm

Coarse sediment (September) 9-10 cm

35% 87% 13% 65%

EUB 338 counts with known bacterial groups. In order to characterize the microbial community composition, these probes were chosen according to hybridization results gained earlier on a nearby sandy site (M. Mussmann, pers. comm.). Our specific aim was to further distinguish the SRB, as already indicated by SRB lipid biomarker findings. The relative amounts of the different bacterial groups targeted are given in Fig. 4 (see also Table 3 for probe specificity). Members of the Cytophaga-Flavobacterium cluster were detected in the fine sediment in both sediment layers analysed. In the medium sand they could only be detected in the lower layer and they were negligible in the coarse sediment. Cytophaga were found to be the largest

fraction in sandy sediments of the nearby Jadebusen Bay (Llobet-Brossa et al., 1998). Moreover, Rusch et al. (2003) obtained the same result from a coarse-grained Middle Atlantic Bight shelf sediment. Members of the CytophagaFlavobacterium cluster (taxonomically reclassified into the subgroup Bacteroidetes) are Gram-negative bacteria which are specialized in the degradation of complex macromolecules (Reichenbach, 1992) and adapted to low nutrient levels (Höfle, 1983; Stoeck et al., 2002). Flavobacteria are also described as obligate anaerobic, fermenting sugars to primarily acetate and succinate. Members of the g-Proteobacteria detected with the probe GAM 42a were found at all stations and investigated


Distribution of bacterial fatty acids and FISH 287 Fine sediment (June) 0-0.5 cm >1% 33%

>1%

26%

10%

11%

Fig. 4. Relative share of the different FISH probes to all EUB 338 counts.

Fine sediment (June) 9-10 cm 4%

8%

4%

8% 40%

55%

Medium sediment (September) 9-10 cm 2% 3%

Medium sediment (September) 0-0.5 cm 7% 4% 9%

9%

18%

5%

81%

62%

Coarse sediment (September) 9-10 cm

Coarse sediment (September) 0-0.5 cm 1% 3%

10% 32% 16%