OUP user menu

Bacterial activity and community composition in stream water and biofilm from an urban river determined by fluorescent in situ hybridization and DGGE analysis

Ruben Araya , Katsuji Tani , Tatsuya Takagi , Nobuyasu Yamaguchi , Masao Nasu
DOI: http://dx.doi.org/10.1111/j.1574-6941.2003.tb01050.x 111-119 First published online: 1 February 2003


Physiologic activity and community structure of planktonic and biofilm microbial communities in an urban river were analyzed using 5-cyano-2,3-ditolyl tetrazolium chloride (CTC) staining, fluorescent in situ hybridization (FISH) and denaturing gradient gel electrophoresis (DGGE) analysis of polymerase chain reaction (PCR)-amplified 16S rDNA fragments. Respiring bacteria estimated by CTC reduction were higher in biofilms (20%) than in stream water samples (12%). FISH analysis revealed that bacterial populations in both stream water and biofilms were dominated by ß-Proteobacteria and Cytophaga—Flavobacterium cluster. Microbial community changes determined by multidimensional scaling analysis from DGGE patterns showed that microbial community structures in biofilms matured within 3–7 days of their formation and did not change further, while those in stream water changed continuously.

  • River
  • Biofilm
  • Planktonic bacterium
  • Physiological activity
  • Microbial community structure

1 Introduction

In aquatic ecosystems, bacterial populations play an essential role in the transformation and demineralization of nutrients [1,2] in order to maintain the energy flux. Surface-associated bacterial populations, particularly in rivers, play an important part in the biodegradation of allochthonous substances [3] such as pollutants derived from human activities. Rivers are important to society, providing water for consumption, agriculture and carrying away human wastes. In contrast to our knowledge of marine and fresh water lake microbiology, the ecological impact of planktonic and attached bacteria within river ecosystems is poorly understood [4].

The increased recognition of the importance of surface colonization and improved methods for studying these communities have led to benefits such as the accurate assessment of population density, a better understanding between structure and function in biofilm communities and the optimization of biodegradation processes. Studies on biofilms have generally been performed using artificial biofilm models [58]. Few in situ studies endeavoring to explore the phylogenetic diversity of river biofilms, and considering seasonal fluctuations typical of rivers, have been described. This problem was recently addressed by Brümmer and collaborators [9]. They studied the biofilm community structure in polluted rivers over a complete annual cycle, finding clear seasonal peaks of abundance among major phylogenetic groups. Recent studies exploring the microbial communities associated with biofilms in situ, did not however include the analysis of planktonic bacterial populations. Thus, in order to understand the dynamics of bacterial populations in river ecosystems, ecological studies should account the different planktonic and attached bacterial populations, considering the bacterial exchange between the water column and biofilm as an important component in bacterial dispersion [10].

A considerable handicap in the investigation of water microbes has been the inability to cultivate many environmental bacteria by conventional laboratory techniques [1113]. The development of vital staining and new molecular biology techniques allowed studies to focus on the characterization and assessment of microbial metabolic potential in situ, without prior cultivation, to predict the responses of ecosystems to environmental perturbations. The objectives of the current study were to analyze (i) the physiologic activity and (ii) community structure changes of both planktonic and biofilm microbial communities in an urban river in Japan, using culture-independent methods including 5-cyano-2,3-ditolyl tetrazolium chloride (CTC) staining [8], fluorescent in situ hybridization (FISH) with rRNA-targeted oligonucleotide probes [14] and denaturing gradient gel electrophoresis (DGGE) [15].

2 Materials and methods

2.1 Study site

Biofilms and planktonic bacterial cells were collected from Kanzaki river (34°44′91″N, 135°29′81″E) located in an industrial area of Osaka city, Japan, during August, October and December 2000. This river is a eutrophic system due to the high level of total organic carbon that exceeds 10 mg l−1.

2.2 Analysis of physicochemical parameters

Water temperature, ambient temperature, conductivity, pH, oxygen and conductivity values were measured in situ using specific electrodes. Ambient temperature was determined with an Iuchi TM-150 thermometer. Conductivity values were estimated with a Horiba ES-14 conductimeter. pH values were determined with a Horiba D-13 electrode. Oxygen and water temperature were determined with a Horiba OM-14 oxymeter.

2.3 Sampling device and sampling procedure for disrupted analysis

A collector was designed to raise a sufficient amount of natural river biofilm. It comprised three parallel PVC tubes (1 m in length) carrying 24 polycarbonate slides [5] (75 mm×25 mm×1.5 mm; previously cleaned and sterilized by autoclaving), which were submerged about 0.5 m below the river water surface. For sampling, six polycarbonate slides were carefully collected from the study site and introduced into sterile plastic tubes (50 ml) filled with river water at each sampling. Biofilms on slides were washed carefully with sterilized deionized water, filtered through a 0.2 μm pore size filter (cellulose acetate membrane; Advantec, Japan) to remove loosely attached cells from the surface. Bacteria were detached from the slide surfaces with a sterile blade razor and dispersed aseptically by sonication for 2 min at 125 W, 400 kHz (JUS-S01 bath-sonicator, JEOL, Japan). The slides were washed again with sterilized deionized water to remove the majority of cells and finally suspended in two sterile plastic tubes at a final volume of 30 ml. Recovery rates after biofilm dispersion were determined by comparing the total cell number after the dispersion and washing treatment with the number of microbes remaining on the slides, staining directly with 4,6-diamino-2-phenylindole (DAPI) for triplicate samples. More than 90% of the bacterial cells were recovered from the polycarbonate slides.

Samples of stream water were collected with a sterile Duran Shott bottle (500 ml) placed 20–30 cm below the water surface. Biofilm and surface water samples were refrigerated and transported to the laboratory within 1 h of collection for total direct count (TDC) by DAPI staining, vital staining with CTC, FISH analysis and DNA extraction for polymerase chain reaction (PCR).

2.4 CTC-DAPI double staining

Biofilm and stream water samples were stained in triplicate with DAPI (final concentration 1 μg ml−1; Sigma) and CTC (final concentration 2 mM; Polysciences Inc., USA), supplemented with R2A broth [16] at 10% (v/v), and incubated for 1 h at 25°C under dark conditions [17]. After incubation, bacteria from biofilm and stream water were trapped onto black polycarbonate filters (pore size 0.2 μm; Advantec) and observed using an E-400 epifluorescence microscope (Nikon, Japan). UV-2A and G-2A filter sets were used for the direct counting of DAPI-stained and CTC-stained cells, respectively. The microscopic observations were in all cases obtained from three parallel samples. For enumeration of bacterial cells, 30 microscopic fields and more than 1000 cells were counted per filter.

2.5 Oligonucleotide probes and stringency conditions

Probes EUB338, ALF1b, BET42a, GAM42a, CF319a+b and HGC69a were used for the specific detection of targeted bacterial cells (Table 1). Non-fluorescent competitor probes were used in an equal amount with probes ALF1b, BET42a, GAM42a and HGC69a to obtain optimal stringency conditions [9,18,19]. BET42a and GAM42a served as competitors for each other. An NHGC probe was used as the competitor for HGC69a [19]. Two competitors, NEP and NLHGC, were used for the ALF1b probe as described by Brümmer et al. [9]. Non-specific binding was checked with the NON probe (Table 1). The 5′-ends of the oligonucleotides (except for the competitor probes) were labeled with the indocarbocyanine dye Cy3.

View this table:

Oligonucleotides used in this study

ProbeSpecificityTarget positionaFAbNaClcReference
EUB338domain bacteria16S, 338–3550%0.9 M[14]
NONnegative control16S, 338–3550%0.9 M[18]
ALF1bα subclass of Proteobacteria16S, 19–3520%0.225 M[18]
BET42aβ subclass of Proteobacteria23S, 1027–104335%80 mM[18]
GAM42aγ subclass of Proteobacteria23S, 1027–104335%80 mM[18]
CF319a+bCytophaga–Flavobacterium cluster of CFB-phylum16S, 319–33615%80 mM[18]
HCG69aGram-positive bacteria with high GC content of DNA23S, 1901–191835%80 mM[19]
EUBf933dbacteria, V6,7,8, region16S, 933–955not used for FISH[23]
EUBr1387bacteria, V6,7,8, region16S, 1387–1368not used for FISH[23]
  • a Position in the rRNA of Escherichia coli.

  • b Formamide concentration in hybridization buffer.

  • c Sodium chloride concentration in washing buffer.

  • d When used for DGGE, the primer has the following GC-clamp at its 5′-end: 5′-CGCCCGCCGCGCGCGGCGGGCGGGGCGGGGGCACGGGGGG.

2.6 Fluorescent in situ hybridization

FISH analysis was performed as described by Amann et al. [20] and Manz et al. [18]. For fixation, 12% (w/v) paraformaldehyde in phosphate-buffered saline was added to the disrupted biofilms and stream water samples to a final concentration of 4% (w/v) and incubated at 4°C for 12 h. Triplicate samples were concentrated onto black polycarbonate filters (diameter 25 mm, pore size 0.2 μm; Advantec). The filters were subsequently washed with 3 ml of sterilized water, dried under vacuum, and dehydrated with ethanol (50, 80 and 100%, 3 min each). Each filter section was placed on a glass slide and covered with 50 μl of prewarmed (46°C) hybridization solution containing 0.9 M NaCl, 20 mM Tris–HCl (pH 7.2), 5 mM ethylenediamine tetraacetic acid (EDTA), 0.01% (w/v) sodium dodecyl sulfate (SDS), a variable concentration of formamide (Table 1), and 100 ng of each oligonucleotide probe. The filters were incubated at 46°C for 3 h in an equilibrated chamber for hybridization, then transferred to a well chamber containing prewarmed (48°C) washing solution (a variable concentration of NaCl, 20 mM Tris–HCl (pH 7.5), 5 mM EDTA, 0.01% (w/v) SDS) for incubation without shaking at 48°C for 30 min. The filters were dried on paper and stained with DAPI solution (1 μg ml−1 in sterilized water; Sigma) for 5 min at room temperature in the dark. Subsequently, they were gently washed in sterilized water, dried on paper and mounted on glass slides with immersion oil. Bacterial cells on the filter sections were observed with an E-400 epifluorescent microscope (Nikon) equipped with optimal filter sets (UV-2A for DAPI, HQ:CY3 for Cy3). Probe-specific cell counts are presented as the percentage of cells visualized by DAPI, and the mean abundance and standard deviations were calculated. All counts were corrected by subtracting the counts obtained with the negative control probe NON. Bacterial counts were estimated using the protocol described above. During all experimental steps, the losses of bacterial cells from the polycarbonate filters were estimated by comparison of the cell number obtained by microscopy before and after all FISH procedures of triplicate samples. More than 90% of the DAPI-stained cells remained on the polycarbonate filters.

2.7 Statistical analysis

Percentages of respiring activity and abundance from in situ hybridization in relation to total direct counts were arcsin transformed before analysis. All counts reported for this study are the means of three determinations, and the coefficients of variation between replicate aliquots were less than 12%. Differences among planktonic and biofilm bacterial samples during the different months were analyzed by one-way analysis of variance (ANOVA) followed by multiple mean comparisons. Probabilities lower than or equal to 0.05 were considered significant. The computer program used for this analysis was STATGRAPHICS (version 2.1, Statistical Graphics Corporation, USA).

2.8 DNA extraction and purification

DNA was extracted following the method of Tsai and Olson [21], with slight modification. 100 ml of stream water or 30 ml of biofilm suspension were filtered onto sterile cellulose acetate filters (pore size 0.2 μm, Advantec) to concentrate bacterial cells. Then the filters were added to 2 ml of lysis solution (0.15 M NaCl, 0.1 M EDTA, pH 8.0) containing 15 mg ml−1 of lysozyme and incubated at 37°C for 2 h with agitation (100 rpm). 2 ml of buffer (0.1 M NaCl, 0.5 M Tris–HCl (pH 8.0), 10% SDS (w/v)) were added and followed by three freeze (in liquid nitrogen) and thaw (under 65°C) cycles. Then 2 ml of buffered phenol (pH 8.0):chloroform:isoamylalcohol (25:24:1; phenol-CIAA) were added and vortexed to obtain an emulsion, and these samples were centrifuged for 15 min at 3500×g. 3 ml of the top aqueous layer were collected and then mixed with 1.5 ml of phenol-CIAA and 1.5 ml of CIAA (chloroform:isoamylalcohol, 24:1). 2.5 ml of the resulting supernatant was further extracted with an equal volume of CIAA. Finally, nucleic acids in the extracted supernatant (2 ml) were precipitated with 2 ml of 100% ethanol at −20°C overnight. The pellet of crude nucleic acids was obtained by centrifugation at 12 000×g for 10 min at 4°C, then vacuum-dried and dissolved in 100 μl of TE buffer (20 mM Tris–HCl, 1 mM EDTA, pH 8.0). RNA in the crude extract was removed by incubation with heat-treated pancreatic RNase A (final concentration 0.2 μg μl−1) for 2 h at 37°C. The RNA-free DNA was then purified for PCR amplification with an Elutip-d column (Schleicher and Schuell, Keen, NH, USA) attached to the cellulose acetate filter (Schleider and Schuell NA010/27; 0.45 μm pore size). DNA was recovered from the column as instructed by the manufacturer.

2.9 Primers and PCR amplification

16S rDNA fragments were amplified using EUBf933-GC-clamp and EUBr1387 primers (Table 1), which are specific for universally conserved bacterial 16S rDNA sequences. A 40 bp GC-rich sequence (GC-clamp) was attached to the 5′-end of the EUBf933 primer to prevent complete melting of the DNA fragments during the DGGE analysis. PCR amplification reactions were carried out using Amplitaq Gold (Applied Biosystem) reagents in 50 μl of PCR mixture with 3 mM MgCl2 and 20 pmol of each primer. Hot start PCR was performed at 95°C for 9 min and a touchdown PCR was performed as follows: the annealing temperature was initially set at 65°C and then decreased by 0.5°C every cycle until 55°C, followed by primer extension at 72°C for 3 min. Next, 15 additional cycles were carried out at 55°C for 1 min (primer annealing), followed by denaturation at 94°C for 1 min and primer extension at 72°C for 3 min. Finally, an extension step was carried out at 72°C for 7 min.

2.10 DGGE analysis

Approximately 300 ng of PCR products were loaded onto a 6.5% (w/v) polyacrylamide gel cast in 1×TAE (40 mM Tris, 20 mM acetic acid, 1 mM EDTA, pH 8.0). The polyacrylamide gels (acrylamide:bisacrylamide, 37.5:1) were made with denaturing gradients ranging from 45 to 65%. 100% denaturant contained 7 M urea and 40% formamide. Electrophoresis was initially at 55°C for 10 min at 20 V, and thereafter for 12 h at 100 V. After electrophoresis, the gels were stained for 20 min with SYBR Gold nucleic acid gel stain (Molecular Probes) as specified by the manufacturer. DGGE gels were scanned with a FluorImager (Molecular Dynamics) for digitalization using a 488 nm argon laser. Images were analyzed by Image QuaNT (version 4–2-J) to generate a densitometric profile. Bands were detected when the relative peak height to total peak height exceeded 1%. To avoid the effect of non-reproducible PCR biases, parallel PCR amplifications from the same sample were compared, obtaining identical DGGE profiles.

2.11 Multidimensional scaling analysis of DGGE banding patterns

To assess changes in the genetic diversity of bacterial communities during the study period, DGGE banding patterns were analyzed by multidimensional scaling (MDS) analysis [22] as described by Iwamoto et al. [23]. For this purpose, the presence and absence of DGGE bands and their intensity over all profiles were recorded in a binary matrix, which was then analyzed with the program SPSS 9.0 J for Windows (SPSS Japan) using a stress value of <0.1. The resulting graphical representation (MDS map) showed every band pattern as one plot where relative changes in community structure can be visualized and interpreted as the distances between the points. The closer the points are to each other, the more similar are the DGGE banding patterns.

3 Results

3.1 Physiologic activity

CTC staining revealed a higher ratio of metabolic activity in bacterial cells associated with biofilms (average 19%) than free-living bacterial populations (average 12%) (P=0.001) during the study period (Fig. 1). In stream water samples, the CTC(+)/TDC ratio increased from 12% in August to 16% in October and declined to 8% in December. In biofilm samples, however, the CTC(+)/TDC ratio decreased from 22% in August to 19% in October and 15% in December.


Percentage of respiring bacteria determined by CTC in planktonic bacteria (PB) and biofilm-associated bacteria (BB) in August, October and December. Data represent means±S.D. of triplicate samples.

3.2 Phylogenetic analysis by FISH

The percentage of DAPI-stained bacterial cells that could be enumerated microscopically with the domain Bacteria EUB338 probe was 55 and 73% for planktonic bacterial and surface-associated cells, respectively (Fig. 2). FISH with probes for the major phylogenetic groups revealed that bacterial populations in stream water and biofilms were dominated by β-Proteobacteria and Cytophaga–Flavobacterium cluster, even though slight differences by month were observed. During sampling periods within each month, bacterial community structure determined by FISH did not show significant differences for planktonic bacterial cells in stream water (P>0.05) and during biofilm formation (P>0.05). For that reason, we summarized Fig. 2 as the monthly average after 1, 3, 7 and 14 days.


Box plot of the monthly distribution of the main phylogenetic groups in planktonic bacteria (PB) and biofilm-associated bacteria (BB) in August (A), October (B) and December (C) determined by FISH using the EUB338 (EUB), CF319a+b (CF), ALF1b (ALPHA), BET42a (BETA) and GAM42a (GAMMA) probes. The boxes present the values of cell counts in triplicate: the ends of boxes give the minimum and maximum values and the line within the box gives the third value.

FISH analysis in August showed that β-Proteobacteria was the dominant phylogenetic group in stream water and biofilms. In biofilms, β-Proteobacteria was followed by α-Proteobacteria and the CF319 cluster. However, in flowing water, the second most dominant group was Cytophaga–Flavobacterium cluster. Although in October the community structure on biofilms was also dominated by the β subclass of Proteobacteria, stream water was dominated by members belonging to the Cytophaga–Flavobacterium cluster (23%) and β subclass of Proteobacteria (22%). More than 70% of the total hybridized cells from biofilms and stream water samples were affiliated with the β-Proteobacteria and Cytophaga–Flavobacterium cluster in December. In biofilms, β-Proteobacteria was the dominant group in each month.

Members of the γ-Proteobacteria did not constitute a numerically dominant phylogenetic group in biofilms or stream water, although their abundance in biofilms was about twice that in stream water. Gram-positive bacteria with a high G+C content, as detected with the HGC69a probe, were quantified at less than 1% for both surface-associated and free-living bacterial populations (data not shown).

3.3 DGGE patterns and MDS analysis

For DGGE analysis we ran two gels covering a total of 24 samples taken from stream water and from biofilm during its formation (Fig. 3). The analysis of both gels after Image QuaNT scan analysis resulted in complex DGGE profiles in which a total of 437 bands were detected (233 from planktonic bacteria and 204 for biofilm bacteria). All bands were detected by the software when the relative intensities were above 1% of the total peak height in a lane. Weaker bands could not be distinguished in Fig. 3, and were only detectable after generation of a densitometric profile by the Image QuaNT software. The mean of DGGE bands varied between 18.5 and 19.4 per sample for planktonic and biofilm microbial communities, respectively. In general, for both microbial communities about 40% of the bands (four to 10 bands) accounted for about 70% of the relative intensity in each lane.


DGGE of PCR-amplified 16S rDNA fragments from planktonic bacteria (A) and biofilm bacteria (B). The time course (days) of the sample analysis is indicated for August, October and December.

Changes in the community structure during August, October and December were addressed by MDS analysis of DGGE patterns from planktonic and surface-associated microbial communities. MDS is a mathematical technique that generates a spatial configuration map where the distance between data points reflects the relationship between individuals in the underlying data set. In this study, MDS was applied to the DGGE banding pattern to illustrate the similarity of all possible pairs of each gel track. The two-dimensional plots of MDS scores for planktonic and biofilm bacteria are shown in Fig. 4. The x and y axes of the figures are simply for plotting purposes. The relative scale can not be compared between Fig. 4A and B. MDS analysis from biofilm samples showed a large change in the microbial community structure between immature and mature biofilms (Fig. 4B). The bacterial community structure in August (after 1, 3 and 14 days of formation) revealed continual changes during biofilm formation. However, MDS analysis of the October sample revealed changes in microbial community structure from day 1 to 7 only; it then remained without further change until day 14. Results from December showed that the microbial community structure during biofilm formation underwent a large change from day 1 to 3 and then became stable from day 3 to 14 (Fig. 4B). In contrast, MDS analysis from stream water revealed a continual change in bacterial community structure in August, October and December (Fig. 4A). The DGGE profile comparison by MDS analysis between the planktonic and surface-associated microbial communities during the different months revealed that community structures from planktonic and attached bacterial populations were different within the individual months. Furthermore, a significantly different community structure was observed between immature biofilms and the surrounding planktonic bacterial community (data not shown).


Two-dimensional plots of MDS analysis from DGGE patterns of planktonic bacteria (A) and biofilm bacteria (B) in August (▪), October (●) and December (○).

4 Discussion

As has been described for drinking water bacterial biofilms [24], river biofilms [25], and river snow [26], the results of this study show that bacterial populations associated with biofilms have a significantly higher ratio of respiring activity (20%) than planktonic bacterial cells (12%). Yamaghuchi et al. [17] reported a higher ratio of physiologic activity in planktonic bacterial populations in eutrophic than in oligotrophic rivers. Carbon availability is critical, because a lack of this element limits in situ respiratory activity [27]. For that reason, the higher ratios of CTC(+) cells associated with biofilms would be influenced directly by the higher availability of carbon associated with surfaces rather than the water column, compared to planktonic cells. Interestingly, as described by Böckelmann and colleagues studying microbial communities of aggregates in the Elbe River [26], a metabolic activity reduction from summer (August) to winter (December) was observed for free-living and surface-associated bacterial populations (Fig. 1). We attributed this phenomenon to the entry of bacterial populations into a physiologically stressed state with low levels of enzymatic activity (i.e. injured or in a dormant state), or to a reduction in respiratory rates provoked by a decrease in system productivity associated with a drop in water temperature, which has been used as a predictor of respiratory activity across seasonal scales [26,28]. In this sampling period, a water temperature reduction of 18°C from August (average 29°C) to December (average 11°C) was observed. The increasing respiring bacteria ratio in stream water from August (12%) to October (16%) was the result of admixture between soil and surface water provoked by a heavy rain during the sampled period.

Increased physiologic activity was highly correlated with hybridization efficiencies. The percentage of hybridized cells with the domain Bacteria probe (EUB338) was higher for biofilm-associated cells (73%) than for free-living bacterial populations (55%) (Fig. 2). These ratios were within the range of values described for biofilms [5,9,29], river snow [26] and free-living bacteria [30,31]. We assume these represent a direct relationship between ribosome content and nutrient availability – a survival advantage that surfaces offer to prokaryotic cells. In general, biofilm bacterial populations are more active and are present at the highest densities associated with surfaces, because they tend to collect and concentrate nutrients in higher concentrations than surface water [3,29,3234]. For that reason, bacterial cells associated with biofilms grow faster than planktonic bacterial cells. In the light of this observation, we may presume that surface-associated bacterial populations will be playing a critical role in the ecosystem of the Kanzaki River, due to their higher activity in comparison to planktonic bacterial communities, as has been described for particle-attached bacteria in the Columbia River estuary [35].

In general, most of the EUB338-detectable cells, both on biofilms and free-living bacteria, could be related to the five major phylogenetic groups used in this study. The microbial community structure of biofilm and surface water samples was dominated by β-Proteobacteria and Cytophaga–Flavobacterium cluster even though slight differences within months were observed (Fig. 2). During the different months sampled, bacteria on polycarbonate slides were dominated by the β subclass of Proteobacteria. This group of bacteria has been described as the most morphologically diverse group within biofilms [5], and a dominant group in fresh water systems such as oligotrophic lakes, drinking water biofilms [29,36,37], river snow [26] and eutrophic rivers [9,30]. It seems to be a common feature of most aquatic environments. β-Proteobacteria may attach more easily to surfaces during initial biofilm development than members of other bacterial groups [5] and may therefore dominate during biofilm succession, as seen in our study. In December, especially, a bimodal distribution between β-Proteobacteria and Cytophaga–Flavobacterium cluster was demonstrated where members of the β-Proteobacteria were only half as abundant as those of the Cytophaga–Flavobacterium cluster. Brümmer and coworkers [9] attributed the presence of a large fraction of the β subclass of Proteobacteria to their ability to oxidize ammonia and participate in the degradation of pollutants.

Although FISH revealed no differences in microbial community structure for planktonic populations within a given month and during biofilm formation at the level of the major phylogenetic groups, MDS analysis from complex DNA-derived DGGE profiles showed that the microbial community structure of the surface water changed continuously during the study period (Fig. 4A). We should note that the number and intensity of bands were not equal to the number and abundance of species within the bacterial community, due to possible PCR bias, differences in gene copy number between species or heterogeneity of 16S rDNA within species. Even though PCR presents some disadvantages, genetic fingerprinting techniques such as PCR-DGGE and PCR-temperature (T)GGE can clarify the change in microbial community without any cultivation processes [3846]. Although the bacterial community structure changed between initial and mature biofilms, MDS plots of 1 day-old biofilms were well separated from those which followed (Fig. 4B). A relative stability was observed in October and December after 7 and 3 days in biofilms, respectively (Fig. 4B). Day 7 plotted close to day 14 in October and December. These showed that the community structure in biofilm increased in stability after 7 days. We attribute the changes in microbial communities associated with immature biofilms to the higher respiratory activity registered in August and October compared with December (Fig. 1). A positive correlation between the abundance of CTC(+) cells and bacterial production has been described in lakes [47,48]. Similarly, the changes in microbial community structure during August and October are likely caused by a higher turnover capacity in the microbial community associated with biofilms. In contrast, the slower turnover observed for the microbial community of mature biofilms in December was the result of a reduction in system productivity due to the reduced water temperature as discussed above.

This study demonstrated that weighted MDS analysis (considering the relative intensity of each band) from DGGE gels was able to detect in more detail subtle differences between complex microbial assemblages that FISH analysis could not detect. Thus, DGGE analysis supported with MDS proved to be a powerful tool to monitor community structure changes, allowing the throughput which is required to study ecological questions.

The higher physiologic activity of microbes and the dominance of members related to the β-Proteobacteria group were shown throughout the study period in biofilms from a eutrophic urban river. That is, microbes in river ecosystems and specifically those associated with biofilms must play more important roles, such as the biodegradation of pollutants, than planktonic bacterial cells. Further research on river biofilms should be helpful in establishing a relationship between the structure and function of the microbial communities in river ecosystems.


This study was supported by the Science and Technology Agency, Japan (Promotion System for Intellectual Infrastructure of Research and Development, under special Coordination Funds for Promoting Science and Technology).


  1. [1]
  2. [2]
  3. [3]
  4. [4]
  5. [5]
  6. [6]
  7. [7]
  8. [8]
  9. [9]
  10. [10]
  11. [11]
  12. [12]
  13. [13]
  14. [14]
  15. [15]
  16. [16]
  17. [17]
  18. [18]
  19. [19]
  20. [20]
  21. [21]
  22. [22]
  23. [23]
  24. [24]
  25. [25]
  26. [26]
  27. [27]
  28. [28]
  29. [29]
  30. [30]
  31. [31]
  32. [32]
  33. [33]
  34. [34]
  35. [35]
  36. [36]
  37. [37]
  38. [38]
  39. [39]
  40. [40]
  41. [41]
  42. [42]
  43. [43]
  44. [44]
  45. [45]
  46. [46]
  47. [47]
  48. [48]
View Abstract