Water Research 36 (2002) 638–646

Stability of the bacterial communities supported by a
seven-stage biological process treating pharmaceutical
wastewater as revealed by PCR-DGGE
Timothy M. LaParaa,*, Cindy H. Nakatsub, Lisa M. Panteac, James E. Allemana

School of Civil Engineering, Purdue University, West Lafayette, IN 47907-1284, USA
Department of Agronomy, Purdue University, West Lafayette, IN 47907, USA
Eli Lilly and Company (Tippecanoe Laboratories), Lafayette, IN 47902, USA
Received 1 January 2001; accepted 1 May 2001

The stabilities of the bacterial community structures supported by seven full-scale biological reactors treating
pharmaceutical wastewater were investigated by denaturing gradient gel electrophoresis (DGGE) of polymerase chain
reaction (PCR) amplified 16S rRNA gene fragments. Effluent quality from this treatment process was consistently high
with respect to BOD5 (o30 mg l1), soluble COD (o500 mg l1), and total ammonia (o 5 mg l1 as N) concentrations.
Long-term community structure stability was studied by comparing the similarity of PCR-DGGE fingerprints from
samples collected 87 days apart between which the influent wastewater characteristics were relatively stable. The Dice
index (Cs ) of similarity was moderately high for the first four reactors (Cs ¼ 0:6120:77) and very high for the last three
reactors (Cs ¼ 0:8920:91). Short-term community structure stability was studied by comparing PCR-DGGE
fingerprints from samples collected 15 days apart between which the influent wastewater characteristics changed
significantly, while the effluent quality remained consistently high. The bacterial community composition of each of the
seven bioreactors showed a moderate community shift (Cs ¼ 0:7020:76). Short-term variability in influent wastewater
composition, therefore, affected a greater community shift than did long-term operation treating a wastewater of
relatively consistent composition. These results indicate that functionally stable wastewater treatment bioreactors have
stable microbial community structures under normal operating conditions but are able to adapt in response to
perturbations to sustain high effluent quality. r 2002 Elsevier Science Ltd. All rights reserved.
Keywords: 16S rRNA; Community analysis; DGGE; PCR; Treatment; Wastewater

1. Introduction
The goal of every biologically-mediated wastewater
treatment process is to consistently produce a high
quality effluent such that surface waters are not harmed.
The microbial communities supported by these processes must accommodate variability in influent
*Corresponding author. Department of Civil Engineering,
University of Minnesota, 500 Pillsbury Drive SE, Minneapolis,
MN 55455-0116, USA. Tel.: +1-612-624-6028.
E-mail address: lapar001@tc.umn.edu (T.M. LaPara).

wastewater composition (temperature, pH, etc.) and
flow rate that could lead to fluctuating effluent quality.
To reduce the variability of reactor conditions, some
means of flow equalization is often included in the
design of biological wastewater treatment facilities.
Nonetheless, process upsets can still occur when the
composition of the influent wastewater changes, especially if a harmful or otherwise toxic substance is
Historically, bacterial community structure analysis
of biological wastewater treatment reactors has been
performed using cultivation-based approaches [1].

0043-1354/02/$ - see front matter r 2002 Elsevier Science Ltd. All rights reserved.
PII: S 0 0 4 3 - 1 3 5 4 ( 0 1 ) 0 0 2 7 7 - 9

4. Cells were lysed by performing a 75 min incubation at 701C followed by two consecutive freeze–thaw cycles. Materials and methods 2. its organic constituents. USA). Fernandez et al. acrylamide : bisacrylamide) in 0. Analytical methods Wastewater samples were collected over a 24-h period and composited. Partial 16S rRNA genes were amplified from the extracted genomic DNA by PCR using a PTC thermal cycler (MJ Research. 2U Taq polymerase and B1 ng of template DNA. Vista.5  TAE buffer (20 mM Tris-acetate. initially at 20 V (15 min) and then at 200 V (300 min). resistance to and recovery from perturbations) and have higher productivity than species-poor communities [6. PCR samples were then stored at 41C until processed further. These cultivation-independent techniques for bacterial community structure analysis have revealed extremely complex communities within wastewater treatment bioreactors [1]. In this study. DGGE was performed on a D-gene apparatus (BioRad. variability in process schedules lead to fluctuations of the wastewater flow rate. 4 nmol deoxynucleoside triphosphates. followed by 7 min at 721C for final extension.2. and stored frozen at 201C within 1-h of sample collection.1. the gel was stained with SYBR Green I (Molecular Probes. USA).5 mM Na2EDTA) using a denaturing gradient ranging from 30% to 60% (100% denaturant contains 7 M urea and 40% (vol/vol) formamide). Eichner et al. pH 7. [9] observed that a more diverse community structure more rapidly recovered from a glucose pulse than a less diverse and less flexible community in anaerobic bioreactors. Community analysis Biomass samples were collected from each of the seven full-scale biological treatment reactors during 30min periods on May 8 (day 0). 2% bovine serum albumin. 639 2. Genomic DNA was then purified from the solution using the FastDNA Spin Kit per manufacturer’s instructions (BIO 101. are biased towards those organisms that can be isolated on microbiological media. The importance of the diverse microbial communities supported by wastewater treatment bioreactors remains unknown. These waste streams contain relatively high concentrations of suspended solids. see [2–5]). 2. and 721C for 30 s (extension). 1998 (day 102). The successful biological treatment of pharmaceutical wastewater presents a significant challenge. [8] observed that the presence of a genetically engineered microorganism (GEM) within an activated sludge community provided better resiliency to shock loads of chlorinated and methylated phenols than control communities that did not include the GEM. August 3 (day 87). 5% sodium dodecyl sulfate). 25 pmol of forward and reverse primers.5 : 1. 175 mmol MgCl2.T. 30 cycles of 921C for 30 s (denaturation). pH 8. USA). we have investigated the bacterial community structure of its seven reactors with regard to longterm process operation under ‘‘normal’’ conditions and in response to short-term variability in influent wastewater composition. All analytical assays (COD. Eugene. The variable V3 region of the domain bacteria was amplified using the PRBA338F (50 -ACT CCT ACG GGA GGC AGC AG-30 ) [14] and PRUN518R (50 -ATT ACC GCG GCT GCT GG-30 ) primers with a GC clamp attached to the forward primer [15]. and August 18.0. Samples containing approximately equal amounts of PCR amplicons were loaded onto 8% (wt/vol) polyacrylamide gels (37. 551C for 30 s (annealing). LaPara et al. 0.000 g). whereas others [8. seven-stage process treating pharmaceutical wastewater.e. Inc.. Hercules. The limited data published thus far regarding community structure and function within wastewater treatment bioreactors appears to be consistent with this ecological theory. CA. .5 ml aliquots.. For wastewater treatment reactors. there is increasing evidence that species-rich ecosystems are more stable (i. this cultivable fraction is typically only 15–20% of the cells that can be counted directly [2]. soluble organics (many of which are recalcitrant). and ammonia [10].12] involved lab scale systems with artificially imposed perturbations. Additionally.M. Following electrophoresis.9. and their relative biodegradability [11]. 10. CA. 10 mM sodium acetate. split into 1. Electrophoresis was performed at 601C. / Water Research 36 (2002) 638–646 Cultivation-based bacterial community analyses. At the wastewater treatment facility of interest to this study. however. This study is unique compared to previous wastewater community analysis studies as it involves a full-scale treatment system with genuine process variability. we have used denaturing gradient gel electrophoresis (DGGE) of polymerase chain reaction (PCR) amplified 16S ribosomal RNA (rRNA) gene fragments to analyze the bacterial community structure of a full-scale. Watertown. BOD. MA. Madison.7]. The PCR protocol included a 5 min initial denaturation at 941C. WI). Numerous cultivation-independent techniques have therefore been developed over the last two decades to analyze bacterial community structures without the inherent biases of cultivation (for reviews. The final 50 ml reaction mixture contained: 1  PCR buffer (Promega. and NH3–N) were performed according to Standard Methods for the Examination of Water and Wastewater [13]. In terrestrial systems. The pellet was resuspended in 10 ml of lysis buffer (120 mM sodium phosphate buffer. Approximately 40 ml of wellmixed reactor contents from each reactor were collected and centrifuged (10 min.

seven biological reactors.19]. whereas the temperature of the final three reactors (MLE1–MLE3) is typically 25–351C. Vaudreuil-Dorion. COD. Process description and performance The wastewater treatment facility consists of an equalization tank. The final three reactors are operated to achieve biological nitrification and denitrification through the manipulation of air application and hydraulic recycling of the overflow from the final reactor (MLE3) to the first mesophilic reactor (MLE1) (Fig. / Water Research 36 (2002) 638–646 OR. and a gravitational clarifier (Fig. Because these fermentation processes are batchoperated. Relative tank sizes are shown as hydraulic residence time. Fig. LaPara et al. independent of intensity. The seven-stage reactor system was able to consistently produce a high quality effluent with respect to BOD5 (o30 mg l1). 1). 4. and NH3 were not significantly affected. The four thermophilic reactors are aerated continuously to maintain aerobic conditions.1 software (Lynnon Biosoft. wastewater is produced intermittently. the concentrations of BOD5. 2). Dendrograms were constructed using the unweighted pair group method with arithmetic averages (UPGMA) [17] using DNAMAN ver. During this study normal operation occurred between day 0 and day 87. Results 3. while atypical operation occurred between day 87 and day 102 as the result of a scheduled shutdown of one of the three fermentation processes. This index ranges from 0 (no common bands) to 1 (identical band patterns). DGGE fingerprints were scored at least three times to ensure consistent results. DGGE fingerprints were manually scored by the presence or absence of co-migrating bands. Equalization is used to help stabilize the wastewater flow rate into the treatment facility.M. This process shutdown caused a shift in the specific organic constituents and flow rate of the wastewater. 1. where j is the number of bands common to samples A and B. then visualized with a UV transilluminator and photographed. . 1). respectively [16]. 3.1.640 T. Normal operation at this biological wastewater treatment facility generally depends on the production of wastewater from three different fermentation processes. however. Overall treatment removal efficiencies were excellent as BOD5. More details of this treatment process have been presented elsewhere [18. USA). soluble COD (o500 mg l1) and NH3 (o5 mg l1 as N) concentrations throughout both of these operating conditions (Fig. soluble COD and NH3–N levels were all reduced by >95%. The first four bioreactors (TBR1–TBR4) are operated at thermophilic temperatures (T > 451C). Pairwise community similarities were quantified using the Dice index ðCs Þ as: Cs ¼ 2j=ða þ bÞ. Binary sequences were generated for individual DGGE lanes by determining the number and position of bands compared to the total number of band positions detected. and a and b are the number of bands in samples A and B. Quebec). Process flow schematic of the industrial wastewater treatment facility studied.

had many of the same bands as TBR1–TBR3. 3). also operated at thermophilic temperatures. and (c) NH3–N concentrations. day 87. 3. the community fingerprints of the seven-stage reactor system followed a consistent pattern. Each of these communities from these thermophilic reactors (TBR1– TBR4) were only moderately similar (Cs ¼ 0:42) to the three mesophilic reactors (MLE1. Samples were collected for community analysis on day 0. The first three reactors (TBR1–TBR3) operated at thermophilic temperatures and had virtually all bands common to each of these three reactors. 3). and 102. Mixed liquor samples were collected for community analysis from all seven biological treatment reactors on days 0. / Water Research 36 (2002) 638–646 3. and day 102. Reactor performance data with respect to (a) BOD5. (K)=influent wastewater (measured in the equalization tank). The fourth reactor (TBR4). day 87. Figs. For example. Fig. 87. Bacterial community fingerprints for each of these seven reactors on each of the sample dates were then generated by DGGE of PCR-amplified 16S rRNA gene fragments (Fig. MLE2.M. (’)=wastewater following treatment by TBR1–TBR4. see solid black arrows. LaPara et al. Community analysis Fig. and that the TBR4 fingerprint was Table 1 Individual reactor temperatures (1C) of a seven-stage biological treatment process on day 0. (m)=effluent wastewater.2. From a qualitative perspective. 3b and c). Examples of bands common to TBR1–TBR3 are identified by solid white arrows in Fig. each of these mesophilic reactor communities were identical to each other (Cs ¼ 1). TBR2. Key differences from these samples are that TBR1–TBR3 and MLE1–MLE3 were highly similar (Cs > 0:85) but not identical. In the final effluent. see open white arrows. However. but often had several additional bands (for examples of additional bands. Similar patterns for Cs were found between the seven reactor communities from day 87 and day 102 (Tables 3 and 4). however. the fingerprints from day 0 (Table 2) were identical for TBR1. The thermophilic reactors. and TBR3 (Cs ¼ 1). Reactor temperatures at the time of sampling are included in Table 1. and day 102 TBR1 TBR2 TBR3 TBR4 MLE1 MLE2 MLE3 day 0 day 87 day 102 53 56 53 55 28 32 33 54 57 58 50 28 32 32 53 53 54 51 27 30 31 . (b) soluble COD. although treatment performance was highly variable. 2. and MLE3). The Dice index (Cs ) was used to quantify the similarity of these community fingerprints between each of the reactors on individual sample dates in a pairwise fashion. There was a slight community shift from TBR1–TBR3 to TBR4 (Cs ¼ 0:89). The thermophilic reactors (TBR1–TBR4) were able to achieve a large fraction of these BOD5 (mean=68%) and soluble COD (mean=61%) removals. The last three reactors (MLE1–MLE3) operated at mesophilic temperatures and had nearly identical band patterns (for examples of common bands. NH3–N levels did increase to levels >30 mg l1 (maximum=55 mg l1) from days 18–27 indicating a slight process upset. no period of reactor upset was detected for BOD5 and soluble COD removal. were only capable of poor NH3 removal (mean=28%). however.641 T.

98 0.g. TBR1 on day 87) (Table 5).36 0.89 0.42 0.42 0.28 0.642 T.98 1 not as similar to the TBR1–TBR3 communities (Cs ¼ 0:5920:72).42 0. 3.42 0. (6) MLE2. Long-term stability of the bacterial communities supported by each of the seven reactors was assessed MLE3 by calculating Cs for each of the seven reactors between day 0 and day 87 (e. LaPara et al. and solid black arrows identify representative bands common to MLE1–MLE3. PCR-DGGE fingerprints from a seven-stage biological wastewater treatment process on (a) day 0 (b) day 87 and (c) day 102. Table 2 Dice coefficients (Cs ) comparing the similarities of PCR-DGGE fingerprints (Fig. / Water Research 36 (2002) 638–646 Fig. 3a) from a seven-stage biological wastewater treatment process on day 0 TBR1 TBR2 TBR3 TBR4 MLE1 MLE2 MLE3 TBR1 TBR2 TBR3 TBR4 MLE1 MLE2 1 1 0.33 0.42 0.29 0. comparing TBR1 on day 0 vs. Lanes flanking the numbered lanes contain a reference fingerprint [22] that is used to confirm consistent gradient of denaturants across a gel. Numbered gel lanes contain PCR-amplified 16S rRNA gene fragments from reactors: (1) TBR1.94 0.42 0.97 0.42 0.42 1 1 1 MLE3 Table 3 Dice coefficients (Cs ) comparing the similarities of PCR-DGGE fingerprints (Fig.74 0. (3) TBR3.97 0.M.33 0.76 0. (4) TBR4.32 0.42 0.44 0. (7) MLE3.40 0. Solid white arrows identify representative bands common to TBR1–TBR3. open white arrows identify representative bands unique to TBR4. (2) TBR2.89 0..42 1 0. 3b) from a seven-stage biological wastewater treatment process on day 87 TBR1 TBR2 TBR3 TBR4 MLE1 MLE2 MLE3 TBR1 TBR2 TBR3 TBR4 MLE1 MLE2 0.29 0.89 0.42 0.36 0. (5) MLE1.72 0. This time-period corresponded with normal operation in which all three .42 0.42 0.42 0.

LaPara et al.M.76 0. A3=day 87 (August 3). The first three reactors (TBR1– TBR3) had highly similar community fingerprints between day 0 and day 87 as Cs ranged between 0. 4. 4).26 0.71 0. 3c) day 0 vs.77 0.61 0. Cluster analysis was performed with the UPGMA algorithm (see Section 2) to study general patterns of community similarity among the different reactors over the three sampling dates (Fig. In this case. This detail was therefore further elucidated by cluster analysis of all three PCR-DGGE fingerprints.33 0.76 (Table 5). Samples are identified by reactor and sample date. Short-term stability of the bacterial communities supported by each of the seven reactors was assessed by calculating Cs for each of the seven reactors from the day 87 and day 102 samples.70 0.74 0.86 0.91 0.41 0. Dendrogram revealing the relatedness of PCR-DGGE fingerprints from seven different biological wastewater treatment reactors from three different sample dates. The three mesophilic reactors had very high similarities (Cs ¼ 0:8920:91).76 0.97 0. M8= day 0 (May 8). day 102 (Fig.31 0. day 87 (Fig. 3c) from a seven-stage biological wastewater treatment process on day 102 TBR1 TBR2 TBR3 TBR4 MLE1 MLE2 MLE3 TBR1 TBR2 TBR3 TBR4 MLE1 MLE2 0. the bacterial communities from all seven reactors were relatively stable as Cs ranged between 0.74 and 0. The bacterial communities from the first three thermophilic reactors (TBR1–TBR3) and from the three MLE reactors (MLE1–MLE3) grouped closely Fig. The number of bands in an individual lane ranged from 17 to 28 (mean=21. corresponding to a total of 47 unique band positions.27 0. The fingerprints from TBR4 had relatively high community similarities (Cs ¼ 0:61).59 0.74 0.85 0.45 0.71 0. day 87 day 87 vs.37 0.70 and 0.d. This pairwise comparison of community structures suggested a larger community shift in response to the change in influent wastewater characteristics than observed for long-term operation under stable conditions.94 0. .89 0.28 0.28 0. This time period corresponded with abnormal operation in which only two of the three fermentation processes supplied wastewater to the treatment facility.70 0.97 1 MLE3 Table 5 Dice coefficients (Cs ) comparing the similarities of PCR-DGGE fingerprints of individual reactors from day 0 (Fig.73 fermentation processes supplied wastewater to the treatment facility. 3a) vs.6. day 102 TBR1 TBR2 TBR3 TBR4 MLE1 MLE2 MLE3 0. / Water Research 36 (2002) 638–646 Table 4 Dice coefficients (Cs ) comparing the similarities of PCR-DGGE fingerprints (Fig.33 0.77.1).26 0.91 0.67 0.73 0.=4.643 T. 3b) and day 87 vs. A18=day 102 (August 18). s.41 0.

4. and NH3 levels. However. the adaptation of the bioreactor bacterial communities was permissible only because of the presence of low numbers of bacterial populations that were not originally detectable by PCR-DGGE. the TBR4 community structures were only distantly related to the other thermophilic reactor communities. In this study. This analysis revealed that during normal operating conditions (from day 0 to day 87). extremely high temperature or highly saline). Bacterial community structures in these reactors were demonstrated to be relatively stable over time by pairwise similarity analysis using the Dice coefficient. Presumably. affected a greater shift in bacterial community composition in only 15 days than that occurred over a period of 87 days of normal operation. Cluster analysis also showed that the communities supported by reactors TBR1–TBR3 and reactors MLE1–MLE3 on day 0 and day 87 branched more closely to each other than to the communities supported by these reactors on day 102. Both pairwise similarity analysis and clustering analysis (Fig. The PCR-DGGE technique was used in this study to assess bacterial community stability because traditional cultivation-based techniques typically include less than 15% of the cells observed by direct counts [2]. These results are consistent with previous research using PCR-DGGE on bench-scale bioreactors in which temperature served as a selective factor for bacterial community structure development during biological wastewater treatment [19.21. several recent investigations have discovered the presence of high biomass densities of Archaea in more hospitable locations such as seawater [29] and agricultural soils [30]. 4) demonstrated that the community structures from day 0 and day 87 were more similar than those from day 87 and day 102. which demonstrated that retention time could serve as a selective factor for bacterial community development [22]. Similarity analysis on the PCR-DGGE fingerprints from each sample date suggests that the bacterial communities supported by the seven bioreactors at least partially developed in response to reactor temperature.644 T.22]. Although Archaea are often assumed to proliferate only in ‘‘extreme’’ environments (e.. but also flexibility is needed to adapt the community structure in response to changing conditions. remained consistent with respect to effluent BOD5. stable reactor performance requires at least some degree of stability among the individual populations that comprise the microbial community supported by such bioreactors. PCR-DGGE does not necessarily provide a completely accurate and unbiased fingerprint of the microbial community. While this is a significant improvement compared to cultivation-based approaches. The fourth thermophilic reactor (TBR4) had slightly lower Cs values compared to either of the other thermophilic reactors (Cs ¼ 0:5920:89) depending on the sampling date. DGGE of PCR-amplified 16S rRNA gene fragments revealed stable bacterial community structures in a seven-stage biological wastewater process over a period spanning almost three months in which operating conditions were reasonably consistent. The fourth thermophilic reactor (TBR4) was closely related to the other thermophilic reactors only on day 0. Fern!andez et al. Similarity analysis also showed that the seven bioreactors were less stable during the abnormal operating conditions between day 87 and day 102 (Cs ¼ 0:7020:76). This suggests stable reactor performance can be maintained by bioreactors capable of adapting their community structures to accommodate varying reactor conditions.g. These results suggest that operationally stable full-scale biological wastewater treatment reactors support stable bacterial community structures. [9] have shown that flexibility with respect to bacterial community structure leads to more stable process performance. reactors TBR1–TBR3 (Cs ¼ 0:7420:77) and MLE1–MLE3 (Cs ¼ 0:8920:91) maintained stable reactor community structures. These Cs values are quite high compared to those reported by two recent studies using PCR-DGGE to compare microbial community structures in bioreactors (Cs ¼ 0:1020:93 [23]. Discussion The primary goal of biological wastewater treatment is to sustain optimum reactor performance over longterm operationFeven under varying influent wastewater characteristics. as did the mesophilic reactors (MLE1– MLE3) (Cs > 0:95). soluble COD. / Water Research 36 (2002) 638–646 according to sample date.20]. This was possibly due to slightly lower reactor temperatures combined with a more than 2-fold increase in reactor size. Our results also show that the microbial community could adapt to changing environmental conditions (in this case. The first three thermophilic reactors (TBR1–TBR3) had very similar community structures to each other (Cs > 0:85). therefore. LaPara et al. a change in influent wastewater characteristics). An additional limitation of our research approach is that prokaryotic life from the domain Archaea was not studied.M. This is also consistent with previous research results. however. Cs ¼ 020:10 [21]). . This result is consistent with that of other researchers who have shown analogous levels of bacterial community stability in bench-scale systems as long as bioreactor function also remained stable [8. Otherwise. Reactor performance. These biases have been discussed elsewhere in detail [24–27]. PCR-DGGE is also limited in that it is capable of detecting only 95–99% of the bacterial community [28]. The change in wastewater composition.

Microbial Ecol 1998.70(4. Microbiol Rev 1995. Raskin L. Fernandez AS. and reactor performance. Criddle C. Conclusions The DGGE of PCR-amplified 16S rRNA genes was used to assess the stability of bacterial communities supported by a seven-stage. [3] Olsen GJ. Ecology 1996.35(1):1–21. TN. Xing J. DC. Teramoto M. Fern!andez A. Hickey R. Uitterlinden AG. Water Environment Federation and American Water Works Association. Cloete TE. Microbial ecology and evolution: a ribosomal RNA approach.80(5):1455–74.):637–42. 19th ed. Dewaal EC.65(8):3697–704. Kans Univ Sci Bull 1958. Functionally stable wastewater treatment reactors were maintained in response to varying influent wastewater characteristics by a microbial community that adapted its structure to sustain high effluent quality.26(1):37–57. [10] [11] [12] [13] [14] [15] Acknowledgements This research was supported by Eli Lilly and Company (Tippecanoe Laboratories). Schro. Glagoleva O. Fermentation industry. PCR-DGGE is a useful tool to evaluate microbial community structures in biological wastewater treatment reactors. Washington. LaPara TM. Nakatsu CH. Nucleic acid techniques in bacterial systematics. Hashsham SA. [5] [6] [7] [8] [9] 5. Sokal RR.. Criddle CS. Theron J.77(2):350–63. LaPara TM. Monitoring the community structure of wastewater treatment plants: a comparison of old and new techniques. Shepherd HL. Appl Environ Microbiol 1993.59(3):695–700. 1991. A statistical method for evaluating systematic relationships. 1995.59(1): 143–69. The results from this work lead to the following conclusions: * * * Operationally stable wastewater treatment reactors supported stable microbial communities over longterm operation. 1998. Timmis KN. Appl Environ Microbiol 1999.66(9):3951–60. Michener CD. Grismer ME. American Public Health Association. Flexible community structure correlates with stable community function in methanogenic bioreactor communities perturbed by glucose. Pantea L. Nashville. Huang S. LaPara et al. FEMS Microbiol Ecol 1998.M. An outbreak of nonflocculating catabolic populations caused the breakdown of a phenol-digesting activated-sludge process.38:1409–38. diversity and ecology: a decade of ribosomal RNA [18] [19] [20] 645 analysis of uncultivated microorganisms. Wagner-D. Tiedje J. Standard methods for the examination of water and wastewater. Pace NR. Substance-specific detection and pursuit of non-eliminable compounds during biological treatment of waste water from the pharmaceutical industry. Lane DJ. [2] Amann RI. Tiedje JM. Goodfellow M. p.19(2):111–23. Harayama S. Appl Environ Microbiol 1999. Ludwig W. Water Environ Res 1998. Annu Rev Microbiol 1986. In: Stackebrandt E. Tilman D.65(7):2813–9. We believe that PCR-DGGE and other nucleic acid-based techniques for cultivation-independent community analysis (e. Microbial evolution. Molecular techniques for determining microbial diversity and community structure in natural environments. Wagner M. Erb RW. [16] [17] References [1] Amann R. Phylogenetic identification and in situ detection of individual microbial cells without cultivation. Dice LR. Dazzo FB. Phylogenetic analysis of bacterial communities in mesophilic and thermophilic bioreactors treating pharmaceutical wastewater. 1. Saunders JR. New York: Wiley. Eichner CA. Profiling of complex microbial populations by denaturing gradient gel electrophoresis of polymerase chain reaction-amplified genes coding for 16S rRNA.66(9):4058–67. Schleifer K-H. The ecological consequences of changes in biodiversity: a search for general principles. Alleman JE. Stahl DA. Proceedings of Water Environment Federation Industrial Wastewater Technical Conference. microbial community structure. Ecology 1999. Hickey RF. . Seston S. Tilman D. 451–59. ARDRA and in situ hybridization) can be used in future research to develop a mechanistic understanding of the relationships between reactor operational strategies. Pickup RW. / Water Research 36 (2002) 638–646 The PCR-DGGE results presented herein indicate that functionally stable wastewater treatment bioreactors also have stable microbial community structures under normal operating conditions but are capable of adapting in response to perturbations. Pantea LM. vol. Analysis of a fullscale thermophilic aerobic biological treatment facility.T. Giovannoni SJ. Appl Environ Microbiol 2000. [4] Head IM.g. 16S/23S rRNA sequencing. Measures of the amount of ecological association between species.25(3):205–15. Muyzer G.26:297–302. Dollhopf SL. How stable is stable? Function versus community composition. Waste Manage 1999. Watanabe K. der HFr.40:337–65. Crit Rev Microbiol 2000. Ecology 1945. Appl Environ Microbiol 1999.65(1):102–9. Biodiversity: population versus ecosystem stability.obler I. Lane DJ. Thermal gradient gel electrophoresis analysis of bioprotection from pollutant shocks in the activated sludge microbial community. full-scale biological process treating pharmaceutical wastewater. editors. Appl Environ Microbiol 2000. Lemmer H. Alleman JE.

Biased 16S rDNA PCR amplification caused by interference from DNA flanking the template region. Appl Environ Microbiol 1996. / Water Research 36 (2002) 638–646 [21] Konopka A.62(2): 625–30. Graber JR. Tolkernielsen T. Wickham GS. Differential amplication of rRNA genes by polymerase chain reaction. LaPara et al. Nature 2001. [23] LaPara TM. van Elsas JD.646 T. FEMS Microbiol Ecol 1998. Appl Environ Microbiol 1995. Givskov M.23(2):127–32. [24] Reysenbach A-L. [29] Karner MB.126(8):739–44. Appl Environ Microbiol 1992. [27] Hansen MC.38(1–2):1–15. Phylogenetic analysis of nonthermophilic members of the kingdom Crenarchaeota and their diversity and abundance in soils.26(2):141–9. Assessment of bacterial community structure in soil by polymerase chain reaction and denaturing gradient gel electrophoresis. Bacterial function and community structure in reactors treating biopolymers and surfactants at mesophilic and thermophilic temperatures. Nakatsu CH. J Ind Microbiol Biotechnol 1999. J Environ Eng 2000. Zakharova T. LaPara TM. [22] LaPara TM.24(2):140–5. DeLong EF. [30] Buckley DH. J Ind Microbiol Biotechnol 2000. Effect of genome size and rrn gene copy number on PCR amplification of 16S rRNA genes from a mixture of bacterial species. Archaeal dominance in the mesopelagic zone of the Pacific Ocean. . Stackebrandt E. Effects of elevated temperature on bacterial community structure and function in bioreactors treating a synthetic wastewater. Giver LJ. Cacco G. Nakatsu CH.64(11):4333–9. [28] Gelsomino A. [25] Farrelly V. Konopka A. Alleman JE. Giovannoni SJ. Pace NR. Alleman JE.58(10): 3417–8. Rainey FA. Thermophilic aerobic wastewater treatment in continuousflow bioreactors.M. Molin S. Konopka A. Karl DM. Bias caused by template annealing in the amplification of mixtures of 16S rRNA genes by PCR. J Microbiol Methods 1999. Schmidt TM. [26] Suzuki MT. Keijzer-Wolters AC. Appl Environ Microbiol 1998.61(7):2798–801.409(6819):507–10.