You are on page 1of 10

e c o l o g i c a l e n g i n e e r i n g 3 5 ( 2 0 0 9 ) 744–753

available at

journal homepage:

Substrate effect on bacterial communities from constructed

wetlands planted with Typha latifolia treating
industrial wastewater

Cristina S.C. Calheiros a , Anouk F. Duque a , Alexandra Moura b , Isabel S. Henriques b ,

António Correia b , António O.S.S. Rangel a , Paula M.L. Castro a,∗
a Universidade Católica Portuguesa, Escola Superior de Biotecnologia, Rua Dr. António Bernardino de Almeida, 4200-072 Porto, Portugal
b CESAM & Department of Biology, University of Aveiro, 3810-193 Aveiro, Portugal

a r t i c l e i n f o a b s t r a c t

Article history: Constructed wetlands (CWs) have been recognized as being able to effectively treat
Received 2 September 2008 wastewater from municipal and industrial sources. This study focused on the effect of
Received in revised form different substrates and long-term operation of horizontal subsurface flow CWs treating tan-
31 October 2008 nery wastewater on the bacterial communities. The CWs were planted with Typha latifolia in
Accepted 23 November 2008 three types of substrate: two units with different types of expanded clay aggregates and one
unit with fine gravel. Another unit with expanded clay was left unvegetated. Changes in the
bacterial community related to the type of substrate, different hydraulic loading rates and
Keywords: along CW operation were examined using denaturating gradient gel electrophoresis (DGGE).
Bacterial communities Bacterial enumeration was also performed and several bacterial isolates were retrieved from
Constructed wetland the CWs. Phylogenetic affiliations of those isolates were obtained on the basis of 16S rRNA
DGGE gene sequences and revealed that they were closely related to the genera Bacillus (TM1S1,
Typha latifolia TM1R3, TNR1 and TAR1), Paracoccus (TM1R2), Pseudomonas (TM1R1) and Halomonas (TM1S2).
Industrial wastewater The type of substrate and the presence of T. latifolia had a major effect on the species
richness and the structure of bacterial communities as inferred by numerical analysis of
DGGE profiles.
© 2008 Elsevier B.V. All rights reserved.

1. Introduction often encountered imposing treatment needs in order to avoid

negative impacts on the environment (COTANCE, 2002).
The tannery industry converts rawhide or skin, a putrescible Constructed wetlands (CWs) are an effective method for
material, into leather, a stable material, so that it can be used wastewater treatment (Vymazal, 2005), inclusively for several
in the manufacture of a wide range of consumer products. types of industrial wastewater (Kadlec et al., 2000). In the case
Most of the steps of the tannery operations are carried out of the tannery industry these systems have already shown
in water. Consequently, the wastewater treatment is of major favorable performance (Calheiros et al., 2007, 2008a). The treat-
concern. Tannery wastewater composition varies considerably ment mechanisms in CWs encompass a mix of physical,
with the production process that is engaged by the specifi- chemical, and biological processes. The vegetation is essential
cation of the final product. Problems related to high organic in wetland treatment systems (Kadlec et al., 2000); however,
content, and the presence of sulphides and chromium are the main role in the transformation and mineralization of

Corresponding author. Tel.: +351 22 5580059; fax: +351 22 5090351.
E-mail address: (P.M.L. Castro).
0925-8574/$ – see front matter © 2008 Elsevier B.V. All rights reserved.
e c o l o g i c a l e n g i n e e r i n g 3 5 ( 2 0 0 9 ) 744–753 745

nutrients and organic pollutants is played by microorganisms

(Baptista, 2003; Stottmeister et al., 2003). The functional com-
position of the different bacterial groups in subsurface flow
CWs has not been fully explored (Ibekwe et al., 2003). Accord-
ing to Baptista (2003) these functional bacterial groups may be
largely determined by the environmental conditions and the
nature of the system feed.
Culture-dependent methods are known to be insufficient
for studying natural microbial community composition since
numerous microorganisms are uncultured under laboratory
conditions (Ward et al., 1990). The denaturating gradient gel
electrophoresis (DGGE) analysis of 16S rRNA genes amplified
by polymerase chain reaction (PCR) is useful for profiling com-
plex microbial populations (Muyzer et al., 1993). Examples
of studies of microbial communities in CWs using DGGE are
those of Ibekwe et al. (2003), who characterized the micro-
bial communities composition in CWs for dairy wastewater Fig. 1 – Schematic representation of the constructed
treatment, of Truu et al. (2005), who investigated the varia- wetlands (CWs).
tion of microbiological parameters within a CW for domestic
wastewater treatment, of Baptista et al. (2003), who analyzed
the microbial mechanisms of carbon removal in subsurface
flow wetlands, and of Nicomrat et al. (2006), who assessed of the systems was carried out according to Calheiros et al.
the microbial community in a CW receiving acid coal mine (2008a).
drainage. Wastewater samples were collected periodically from the
The aim of this study was to analyze the effect of inlet and outlet of the CW units and physico-chemical param-
different substrates and long-term operation of horizon- eters were determined based on Standard Methods (APHA,
tal subsurface flow CWs treating tannery wastewater on 1998): chemical oxygen demand (COD; Closed Reflux, Titri-
the dynamics of the bacterial communities. The bacte- metric Method), biochemical oxygen demand (BOD5 ; 5-day
rial abundance was evaluated by plate counting and the BOD Test), total suspended solids (total suspended solids, TSS;
dynamics of the bacterial communities was evaluated by dried at 103–105 ◦ C method), Kjeldahl nitrogen (TKN; Kjeldahl
DGGE electrophoresis of 16S rRNA gene amplification prod- method), nitrate nitrogen (NO3 − –N; nitrate electrode method),
ucts. Numerical analysis of DGGE profiles allowed the ammonia nitrogen (NH3 –N; phenate method), total phospho-
estimation of species richness along the operation of each rus (total P; manual digestion and flow injection analysis
CW. for total phosphorus) and pH. The sulphate determination
(SO4 2− ; turbidimetric method) was done based on the method
of the Association of Official Analytical Chemists (AOAC,
2. Methods 1995). The analyses were done immediately after sample col-
lection otherwise samples were properly stored. Dissolved
2.1. CWs set-up and physico-chemical analysis oxygen (DO) and conductivity were registered with a WTW
handheld multi-parameter instrument 340i at the inlet and
The setup conditions of the CWs, located at a tannery wastew- outlet of the units. The substrates used in the units were
ater treatment plant in Portugal, are described in Calheiros et analyzed for organic matter content, based on Houba et al.
al. (2008a). Briefly, the CWs consisted of three units planted (1995).
with Typha latifolia in different substrates: unit one (U1) and
unit three (U3) had expanded clay aggregates being respec- 2.2. Bacterial enumeration
tively Filtralite® MR 3–8 (FMR) and Filtralite® NR 3–8 (FNR) (from
maxit, Argilas Expandidas, SA, Portugal), and unit two (U2) had Microbiological analyses were performed simultaneously with
fine gravel: AGH 4–8 (FG) (from Areipor-Areias Portuguesas, physico-chemical analysis. Colony forming units (CFUs) were
Lda, Portugal). An unvegetated unit (Uc), with the expanded determined based on the surface-plate counting procedure.
clay aggregate FMR as substrate, was also included (Fig. 1). The Briefly, three subsamples were pooled to form two compos-
CWs operated with horizontal subsurface flow and had a sur- ite samples (10 g) of plant roots and substrate (from a depth
face area of 1.2 m2 . They received tannery wastewater after between 10 and 15 cm) of each CW, being placed separately
primary treatment under different hydraulic loading rates. in sterile tubes with 10 mL of saline solution (0.15 mol L−1
The two units with FMR, U1 and Uc, had been in operation NaCl) and shaken on a vortex mixer for 1 min at room tem-
for 17 months during which two hydraulic regimes, 3 and perature. Serial dilutions were made in duplicate and 0.1 mL
6 cm d−1 , were tested (Calheiros et al., 2007). The units U2 of each dilution was spread onto nutrient agar (LABM, UK).
and U3 had an acclimation period of 3 weeks before the tan- Plates were incubated at 25 ◦ C for 4 days after which CFU
nery wastewater was applied. After that, all the systems were were counted. The same procedure was used for bacterial
operated for 31 months under different hydraulic loadings enumeration of the wastewater at the inlet and outlet of the
and interruptions in feed (18, 6 and 8 cm d−1 ). Maintenance CWs.
746 e c o l o g i c a l e n g i n e e r i n g 3 5 ( 2 0 0 9 ) 744–753

2.3. Bacteria isolation, DNA extraction and RAPD buffer using a denaturing gradient ranging from 35% to 60%
typing (100% denaturant solution is defined as 7 M urea and 40% (v/v)
formamide (Muyzer et al., 1993)). A standard marker was also
Different bacterial colonies were isolated based on size, mor- included in all gels, to serve as an indicator of the analysis
phology and pigmentation, from nutrient agar plates using quality. The standard marker was constructed using bacte-
a streak-plate procedure. DNA of each isolate was obtained rial isolates (obtained as described above) selected to cover
by picking a colony with a sterile toothpick, suspending the an adequate range of bands. Electrophoresis conditions and
cells in 20 ␮L sterile water and incubating for 10 min at 100 ◦ C image acquisition were as described previously (Henriques et
(Henriques et al., 2006b). Molecular typing of bacterial isolates al., 2006b).
was performed by random amplified polymorphic DNA (RAPD) DGGE profiles, concerning the presence and intensity of the
analysis. Amplification was performed in 25 ␮L reaction mix- bands, were analyzed using GelCompar® II software (Version
tures containing: 0.75 U Taq polymerase, 1.5 mM MgCl2 , 0.2 mM 4.6; Applied Maths, Sint-Martens-Latem, Belgium). Detected
of each dNTP, 1.0 ␮M primer M13 (MWG-Biotech AG) and band patterns were transferred to an absence/presence
0.5 ␮L of crude cell lysates. The thermal cycling profile was matrix. The binary matrix was transformed into a similar-
as follows: initial denaturation (94 ◦ C for 5 min); 45 cycles of ity matrix using the Bray-Curtis measure. Dendrograms were
denaturation (94 ◦ C for 1 min), annealing (34 ◦ C for 2 min), and generated by unweight pair group mean average (UPGMA)
extension (72 ◦ C for 2 min); and a final extension (72 ◦ C for cluster analysis. Cluster analysis and multidimensional scal-
10 min) (Silva et al., 2006). The reactions were carried out ing diagram (MDS) were performed using PRIMER 5 for
in a Bio-Rad iCycler Thermal Cycler (Bio-Rad Laboratories, Windows (Version 5.2, 2001, PRIMER-E Ltd.) (Clarke and Gorley,
Richmond, CA, USA) using Taq polymerase and nucleotides 2001).
purchased from MBI Fermentas (Vilnius, Lithuania). Polymor- DGGE banding data were used to estimate diversity, H
phic DNA fragments were analyzed by electrophoresis in a (Shannon and Weaver, 1963) and equitability, E (Pielou, 1975)
1% agarose gel in Tris–acetate–EDTA (TAE) buffer, after stain- indexes to describe possible changes in the dominance among
ing with ethidium bromide. Gel image was acquired using phylotypes (Fromin et al., 2002).
a Molecular Image FX apparatus (Bio-Rad Laboratories, Her-
cules, CA, USA). 2.6. Nucleotide sequence accession numbers

2.4. DNA sequence and phylogenetic analysis The 16S rRNA gene sequences of bacterial isolates obtained
in this study were deposited in GenBank under the
Isolates displaying unique RAPD profiles were subsequently accession numbers TM1R1: EU430693, TM1R2: EU430694,
identified by 16S rRNA gene sequencing analysis. Amplifica- TM1R3: EU430695, TM1S1: EU430696, TM1S2: EU430697, TNR1:
tion was performed with universal bacterial primers 27F and EU430700, TAR1: EU430701.
1492R, as described by Lane (1991). PCR products were puri-
fied with Jetquick PCR Product Purification Spin Kit (Genomed, 2.7. Data analysis
Löhne, Germany). DNA sequencing was conducted under
BigDyeTM terminator cycling conditions, and analyzed using Statistical analysis was performed using the software SPSS
an automatic sequencer 3730xl (Macrogen Inc., Seoul, Korea). (SPSS Inc., Chicago, IL, USA; Version 12.0). When applicable,
To determine the phylogenetic affiliation, similarity searches the data were analyzed through one-way analysis of variance
were performed using the BLAST program (Altschul et al., (ANOVA) and Student’s t-test. To detect the statistical signifi-
1997). cance of differences (p < 0.05) between means of observation,
the Duncan test was performed. When applicable, values were
2.5. Analysis of bacterial communities of substrate presented as the mean ± standard error.
and roots from CWs

Genomic DNA from substrate and root samples (six subsam- 3. Results
ples were pooled to form one composite sample of plant roots
and substrate for each CW) was extracted using the Ultra 3.1. Physico-chemical characterization
CleanTM Soil DNA Isolation Kit (MO BIO Laboratories, Inc.,
USA), according to the manufacturer’s protocol. PCR amplifi- In Fig. 2 the removal of several wastewater components by
cation of bacterial 16S rRNA gene fragments was performed the CWs is illustrated. COD, BOD5 and TSS inlet concentration
using primers 338F GC and 518R, as described previously varied between 835–2261, 430–850 and 43–110 mg L−1 , respec-
(Henriques et al., 2006a). Nested PCR amplifications were per- tively. In general, the FMR unit U1 presented higher removal
formed using as template 1 ␮L of the DNA amplicon obtained levels of organic matter, followed by the FNR unit U3, the fine
after the first amplification round and using the same primers gravel unit U2 and the FMR unvegetated unit Uc.
and conditions applied in the first PCR amplification. The inlet concentrations of TKN, NH3 and SO4 2− ranged
DGGE analysis was performed on a DCodeTM Universal between 102–160, 60–98 and 78–1070 mg L−1 , respectively.
Mutation Detection System (Bio-Rad Laboratories, Hercules, The outlet concentrations of TKN, NH3 and SO4 2− ranged
CA, USA). Samples containing approximately equal amounts between 57–115, 35–65 and 32–1005 mg L−1 , respectively. For
of nested-PCR amplicons were loaded onto 8% (w/v) poly- NO3 − the inlet varied between 17–59 mg L−1 and the out-
acrylamide gels (37.5:1, acrylamide/bis-acrylamide) in 1× TAE let between 9–47 mg L−1 . Total phosphorus ranged between
e c o l o g i c a l e n g i n e e r i n g 3 5 ( 2 0 0 9 ) 744–753 747

The number of aerobic bacteria present in the wastewa-

ter at the inlet and outlet of the CWs was also determined.
Samples collected periodically (February, July, September,
December 2005 and February, June, October 2006) showed a
variation between 3.9 × 101 and 4.2 × 105 for the inlet, between
3.2 × 103 and 2.2 × 106 for the outlet of the vegetated units (U1,
U2 and U3) and between 2.6 × 104 and 1.2 × 106 for the outlet
of the unvegetated unit (Uc).
Dominant bacterial isolates obtained from the plates were
further molecular typed by RAPD-PCR. Seven different RAPD
types were then characterized through sequencing of the 16S
rRNA encoding gene. According to BLAST results, 2 strains
were affiliated with ␥-Proteobacteria, 4 with Firmicutes and 1
with ␣-Proteobacteria. These strains were isolated from plant
roots (TM1R1, TM1R2, and TM1R3) and substrate (TM1S1 and
TM1S2) of unit U1 and from plant roots of unit U2 (TAR1) and
unit U3 (TNR1) (Table 1).

3.3. DGGE analysis of bacterial community

Analysis of bacterial communities from substrate and root

samples, collected along the CWs operation, was performed
Fig. 2 – Removal efficiency of (A) COD, BOD5 and TSS and (B) by PCR–DGGE. Two examples of banding patterns for the 16S
TKN, NH3 and SO4 2 , by the constructed wetlands (CWs). rRNA gene PCR–DGGE amplicons are presented in Fig. 4, corre-
Dispersion bars represent standard error of the mean. U1: sponding to units U1 and U2. Each gel presents the variability
CW with Typha latifolia planted in Filtralite® MR 3–8; U2: CW within each CW at different sampling periods correspond-
with T. latifolia planted in fine gravel, AGH 4–8; U3: CW with ing to different hydraulic conditions. Reproducibility of PCR
T. latifolia planted in Filtralite® NR 3–8; Uc: unvegetated unit amplification and DGGE was confirmed in replicates.
with Filtralite® MR 3–8. One-way ANOVA was performed in Table 2 shows, for each CW, the number of bands in each
order to compare the performance in terms of removal lane and the Shannon and equitability indexes (calculated by
efficiency between units for each parameter. Columns using the DGGE profiles) correspondent to the different sam-
marked with different letters differed significantly pling times within each gel. The total number of different band
according to Duncan’s multiple range tests at of p < 0.05. positions in the four gels was 68 (58 in gel U1, 50 in gel U2, 51 in
gel U3 and 43 in gel Uc). The Shannon’s diversity index (H) was
in average of 1.10 ± 0.07 for U1, 1.08 ± 0.03 for U2, 1.14 ± 0.05
0.13–0.83 mg L−1 at the inlet and between 0.06–0.77 mg L−1 at for U3 and 1.01 ± 0.07 for Uc. The equitability index (E) was in
the outlet. average of 0.77 ± 0.05 for U1, 0.82 ± 0.02 for U2, 0.80 ± 0.02 for
DO at the inlet and outlet was low, varying between U3 and 0.73 ± 0.05 for Uc.
0.17–0.99 mg L−1 and 0.09–0.88 mg L−1 , respectively. The pH Fig. 5 shows clustering analysis of DGGE banding patterns.
values were quite stable at the outlet (7.23–8.68) although the In general, samples clustered mostly according to the place of
inlet values varied between 4.76 and 8.26. Conductivity varied collection (substrate or root) and the type of substrate, being
between 4.74–9.64 mS cm−1 at the inlet and 5.00–9.61 mS cm−1 the temporal factor less relevant in bacterial assemblage com-
at the outlet. position.
The organic matter content was determined for the three Multidimensional scaling diagrams of similarity matrices
substrates as being 0.30% for FMR, 0.30% for FNR and 0.04% for (Fig. 6), calculated from the DGGE patterns of samples, showed
FG. At the inlet and outlet zone of each unit the organic matter that the bacterial community changed according to the differ-
content was also determined as being, respectively: 0.86 and ent substrates. Concerning the analysis between the FMR units
0.63% for U1, 0.37 and 0.17% for U2, 1.00 and 0.71% for U3 and U1 and the unvegetated control (Uc), differences between
0.89 and 0.71% for Uc. community profiles within each unit are clear, with a higher
dispersion for the vegetated unit U1.
3.2. Plate-counting and bacterial isolation

The enumeration of bacteria in the units, including roots and 4. Discussion

substrate, is shown in Fig. 3. One-way ANOVA was applied
to compare the CFUs between the planted units and no sig- This study intended to investigate the bacterial communi-
nificant differences were found. When Student’s t-test was ties in CWs established with different substrates, namely
applied to compare bacterial counts from roots and substrate expanded clay aggregates and fine gravel. Close attention was
of each unit significant differences were seen for units U2 and given to bacteria, although it is known that fungi may also play
U3, although for U1 no significant differences were found. an important role in the function of wetlands (Baptista, 2003).
748 e c o l o g i c a l e n g i n e e r i n g 3 5 ( 2 0 0 9 ) 744–753

Fig. 3 – Bacterial enumeration analyzed by plate counts and expressed in CFUs g−1 for substrate and plant roots of
constructed wetlands (CWs). Dispersion bars represent standard error of the mean. U1: CW with T. latifolia planted in
Filtralite® MR 3–8; U2: CW with T. latifolia planted in fine gravel, AGH 4–8; U3: CW with T. latifolia planted in Filtralite® NR 3–8;
Uc: unvegetated unit with Filtralite® MR 3–8.

The complex physico-chemical composition of the tannery propagation for the expanded clay aggregates (Calheiros et al.,
wastewater passing through the wetland is of major impor- 2008a). The macrophytes are important in the wetlands since
tance since it can have a relevant effect on the vegetation they provide structure and a source of reduced carbon for the
(Calheiros et al., 2007, 2008a,b) and in the bacterial commu- microbes that mediate most of the pollutant transformations
nities that inhabit these ecosystems. The wastewater content occurring in the wetlands (Kadlec et al., 2000). Collins et al.
varies in terms of nutrients and toxic substances and can (2004) concluded that plants do have an effect on water qual-
limit or promote bacterial activity and growth. The main role ity, in part because they affect bacterial assemblages. Higher
concerning the direct degradation of organic chemicals in pollutant removals, in terms of COD and BOD5 , were achieved
wastewater treatment is played by microorganisms despite in expanded clay planted units after long-term operation. The
the capacity of plants to detoxify xenobiotics (Stottmeister similar behavior of the expanded clay systems (U1 and U3)
et al., 2003). The organic compounds degradation in the hori- concerning the pollutant removal may be attributed to the fact
zontal subsurface flow wetlands is carried out aerobically and that they may have similar functional group of microorgan-
anaerobically at different extents by bacteria attached to plant isms.
roots and substrate surfaces (Kadlec et al., 2000). The substrate is an important wetland component since
The three tested substrates have proven to be adequate it supports plant growth, establishment of microbial biofilms
for T. latifolia development although there was higher plant and influences the hydraulic processes (Stottmeister et al.,

Table 1 – Phylogenetic affiliation of bacterial strains isolated from the constructed wetlands.
Isolate NCBI Phylogenetic Closest relative (accession no.) Similarity Origin
accession no. affiliation (%)

Pseudomonas stutzeri ST27MN3 (U26419) 99 Sea

TM1R1 EU430693 ␥-Proteobacteria
Pseudomonas stutzeri 19smn4 (U22426) 99 Sea
Paracoccus sp. G30 (DQ667122) 98 Antarctic seawater
TM1R2 EU430694 ␣-Proteobacteria
Paracoccus sp. G29 (DQ667121) 98 Antarctic seawater
Bacillus pumilus MLA14 (EF462914) 99 n.d.
TM1R3 EU430695 Firmicutes
Bacillus sp. NS88 (EF633174) 99 Phyllostachys edulis
Bacillus sp. BCw063 (DQ492812) 99 Arctic sea water
TM1S1 EU430696 Firmicutes
Bacillus sp. GB02-29 (DQ079002) 99 Hydrothermal
sediments and plumes
Halomonas sp. 3014 (AM110971) 99 Deep sea sediment
TM1S2 EU430697 ␥-Proteobacteria
Halomonas sp. A3-3 (AB219125) 99 Fermented foods
Bacillus sp. BCw063 (DQ492812) 99 Arctic sea water
TNR1 EU430700 Firmicutes
Bacillus sp. GB02-29 (DQ079002) 99 Hydrothermal
sediments and plumes
Bacterium 8-gw3-5 (DQ990048) 99 River waters and
TAR1 EU430701 Firmicutes
shallow groundwater
Bacillus pumilus NUC-F (DQ833752) 99 Agricultural soil
e c o l o g i c a l e n g i n e e r i n g 3 5 ( 2 0 0 9 ) 744–753 749

Fig. 4 – DGGE analysis of 16 rRNA gene fragments of total bacterial population from samples of the root and substrate of
constructed wetlands planted with T. latifolia in different matrixes. (A) Gel image of root and substrate samples collected in
U1 in September 2004 (lanes 1 and 2), February 2005 (substrate, lane 4), June 2005 (lanes 5 and 6), July 2005 (lanes 7 and 8),
July 2006 (lanes 9 and 10), October 2006 (lanes 11 and 12) and (B) gel image of root and substrate samples collected in U2 in
February 2005 (lanes 19 and 20), June 2005 (lanes 21 and 22), July 2005 (lanes 23 and 24), July 2006 (lanes 25 and 26), October
2006 (lanes 27 and 28). A DNA marker (M) was included in all the gels to serve as control.

2003). A porous matrix, such as expanded clay, provides a at this point the wastewater organic loading was higher than
greater surface area for treatment contact and biofilm devel- at the outlet.
opment. Each substrate used here has different characteristics The number of CFU found in the vegetated units is within
in terms of pH, electrical conductivity, porosity, and organic the range of what has been published by Truu et al. (2005)
matter content. The higher organic matter content verified at for aerobic heterotrophic bacteria in horizontal subsurface
the inlet substrate of all units was attributed to the fact that flow CW for domestic wastewater treatment, filled with coarse

Table 2 – Number of bands and Shannon diversity (H) and equitability (E) indexes, calculated for the constructed
wetlands (CWs). U1: CW with Typha latifolia planted in Filtralite® MR 3–8; U2: CW with T. latifolia planted in fine gravel,
AGH 4–8; U3: CW with T. latifolia planted in Filtralite® NR 3–8; Uc: unvegetated unit with Filtralite® MR 3–8.
Samplesa U1 U2 U3 Uc

Nr bands H E Nr bands H E Nr bands H E Nr bands H E

Set R 04 35 1.40 0.91 n.d. n.d. n.d. n.d. n.d. n.d. n.a. n.a. n.a.
Set S 04 36 1.18 0.76 n.d. n.d. n.d. n.d. n.d. n.d. 17 0.82 0.66
Feb R 05 n.d. n.d. n.d. 21 0.94 0.71 17 0.81 0.66 n.a. n.a. n.a.
Feb S 05 26 0.88 0.62 14 0.92 0.80 21 1.05 0.79 25 1.06 0.76
Jun R 05 29 1.02 0.70 22 1.18 0.88 24 1.09 0.79 n.a. n.a. n.a.
Jun S 05 30 1.01 0.68 27 1.20 0.84 22 1.05 0.78 23 1.00 0.74
Jul R 05 26 1.64 1.16 18 1.05 0.84 29 1.18 0.81 n.a. n.a. n.a.
Jul S 05 24 0.76 0.55 19 1.12 0.87 25 1.21 0.87 23 1.29 0.94
Sep S 05 n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. 21 0.64 0.49
Jun S 06 n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. 30 1.04 0.71
Jul R 06 22 1.13 0.84 22 1.16 0.86 35 1.30 0.84 n.a. n.a. n.a.
Jul S 06 21 0.94 0.71 25 1.13 0.80 35 1.23 0.80 30 1.02 0.69
Oct R 06 22 1.09 0.81 25 1.10 0.79 33 1.26 0.83 n.a. n.a. n.a.
Oct S 06 22 1.02 0.76 22 1.03 0.77 26 1.19 0.84 28 1.18 0.82

n.a., not applicable; n.d., not determined.

In samples collumn is presented the month, location (R: root or S: substrate) and year, that each sample was collected.
750 e c o l o g i c a l e n g i n e e r i n g 3 5 ( 2 0 0 9 ) 744–753

Fig. 5 – Cluster analysis of microbial communities generated by the analysis of DGGE 16S rRNA gene patterns representing
the genetic similarity of the microbial profiles acquired. Similarities were calculated using the Bray-Curtis measure. (A)
Cluster analysis of bacterial communities from U1 (Filtralite® MR 3–8 substrate), U2 (fine gravel: AGH 4–8 substrate) and U3
(Filtralite® NR 3–8 substrate). (B) Cluster analysis of bacterial communities from U1 and Uc (unvegetated control with
Filtralite® MR 3–8 substrate).

sand and dominated by Scirpus sylvaticus, Urtica dioica and in the future, to undertake anaerobic enumeration consid-
Epilobium hirsutum. The fact that there were no significant dif- ering the fact that the organic degradation can occur both
ferences in bacterial numbers between the substrate and the aerobically and anaerobically in the CW systems. Attempts
roots in unit U1 was attributed to the diffuse propagation of T. to evaluate the relative importance of different microbial
latifolia, since this unit was established for units longer than reactions on organic matter removal (in terms of COD), in hor-
U2 and U3. Significantly higher numbers of CFUs were fre- izontal subsurface flow CWs treating urban wastewater, have
quently found in the wastewater samples collected from the been carried out using a 2D simulation model (Ojeda et al.,
CWs outlet when compared to the inlet, which can be due to 2008). They indicated that the microbial anaerobic reactions
the fact that the water passing through the wetland washes involved in organic matter removal (methanogenesis and sul-
out bacteria from the substrate. Although in this study an aer- phate reduction) occurred over larger areas of the wetlands
obic plate count method was followed it would be interesting, than anoxic (denitrification) and aerobic reactions.
e c o l o g i c a l e n g i n e e r i n g 3 5 ( 2 0 0 9 ) 744–753 751

by the hydraulic alterations and wastewater contaminations,

no direct relationship could be found.
In this study the community composition was not deep-
ened in relation to the identification of bacterial groups.
Despite that, species richness and the structure of bacterial
communities have been inferred by numerical analysis of
DGGE profiles (Moura et al., 2007). DGGE profiles showed a
high variability among bacterial assemblages. The presence
of vegetation seemed to have a major effect on the commu-
nity diversity, as lower bacterial diversity (lower H values) was
observed in the unvegetated unit (Uc). Similarly, Baptista et
al. (2003) reported that microbial communities from inlet and
outlet of horizontal subsurface flow CW with plants are dif-
ferent when compared to an unplanted wetland, treating a
solution of filtered beer. In this study, species richness was
higher in the unit with the expanded clay FNR (U3), followed
by FMR (U1) and FG (U2). The differences observed might be
explained in part by the substrates that differ in characteris-
tics such as pH, conductivity and porosity (Calheiros et al.,
2008a,b). In general, all units showed higher diversity near
root comparing to substrate, which can be related to rhi-
zosphere effects on bacterial communities. Previous studies
have reported the importance of plant metabolites excreted to
the rhizosphere, such as vitamins, that may stimulate micro-
bial growth (Stottmeister et al., 2003). Additionally, Marschner
et al. (2002) have studied the microbial community structure in
the rhizosphere of cluster roots of Lupinus albus and reported
that changes in the community structure may be related with
organic acid plant exudates.
Differences in microbial community structure were
detected by using the equitability index (Pielou, 1975). The
Fig. 6 – Multidimensional scaling diagram of DGGE patterns equitability was higher in the unit with the sand FG (U2) fol-
of samples collected in the constructed wetlands (CWs) of lowed by the units with expanded clay FNR (U3) and then FMR
(A) U2 and U3 and (B) U1 and Uc. U1: CW with T. latifolia (U1), and finally the unvegetated unit (Uc). Vacca et al. (2005)
planted in Filtralite® MR 3–8; U2: CW with T. latifolia planted have reported that when using CWs for domestic wastewater
in fine gravel, AGH 4–8; U3: CW with T. latifolia planted in treatment, the rhizosphere of P. australis is colonized by dis-
Filtralite® NR 3–8; Uc: unvegetated unit with Filtralite® MR tinct communities depending on the filter material (sand and
3–8. expanded clay) and have concluded that the technology, filter
material and plant have a selective influence on the microbial
community within the CW. The functional diversity may be
linked to certain extents to the ecophysiological roles played
The microbial diversity in the substrate and rhizosphere by functional groups although the present study did not allow
is of great importance since it may be influenced by the type such conclusions. In general, equitability was higher in root
and amount of plants. Most of the bacterial isolates retrieved samples comparing to substrate samples. Again, this may be
from the roots and substrate of the CWs presented here were related to the rhizosphere effect on microbial communities
similar to environmental isolates reported from sources such that may promote the growth of certain bacterial groups.
as river and sea waters, sediments and soil. Two isolates affil- The assessment of microbial communities in CWs has been
iated with ␥-Proteobacteria were recovered from substrate and addressed by several authors (Baptista et al., 2003; Ibekwe
roots from the CWs. Franco et al. (2005), based on enrichment et al., 2003; Truu et al., 2005; Nicomrat et al., 2006). Ibekwe
cultures with soil collected nearby the same CWs, have iso- et al. (2003) have characterized the microbial communities
lated bacteria able to degrade polyphenols used in the tannery and composition in CW for dairy wastewater and reported
process being all of them members of ␥-Proteobacteria. that these systems are dependent on microbial communi-
The dynamics of the bacterial community in the CWs over 3 ties for optimal wastewater treatment. Since little information
years of operation were analyzed by comparison of 16S rDNA is available regarding the bacterial communities that inhabit
PCR–DGGE profiles. Data from this study showed that there these ecosystems, the PCR–DGGE technique employed here
was a diverse community of bacteria in the CWs, and that was of great importance in obtaining new data concerning
might influence to different extents the final effluent quality. the microbial structure and dynamics of CWs for tannery
Although the FMR units U1 and Uc had been in operation for a wastewater treatment. The study of dynamics of microbial
bit longer than U3 and U2, which could indicate that these sys- communities from CWs is thus essential in understanding
tems would be more stable and resistant to the stress caused these treatment systems, being a valuable tool for improv-
752 e c o l o g i c a l e n g i n e e r i n g 3 5 ( 2 0 0 9 ) 744–753

ing its design and operation. A more extensive investigation Calheiros, C.S.C., Rangel, A.O.S.S., Castro, P.M.L., 2008a. Evaluation
may also be undertaken at other levels, mainly those related of different substrates to support the growth of Typha latifolia
to the detection of microbial genes encoding enzymes linked in constructed wetlands treating tannery wastewater over
long-term operation. Bioresour. Tecnhol. 99, 6866–6877.
to degradation pathways, and analysis of their quantitative
Calheiros, C.S.C., Rangel, A.O.S.S., Castro, P.M.L., 2008b. The
expression within each CW. effects of tannery wastewater on the development of different
plant species and chromium accumulation in Phragmites
australis. Arch. Environ. Con. Tox. 55 (3), 404–414.
5. Conclusions Clarke, K.R., Gorley, R.N., 2001. PRIMER v5: User Manual/Tutorial.
PRIMER-E, Plymouth, UK.
(1) The type of substrate and the presence of T. latifolia seemed Collins, B., McArthur, J.V., Sharitz, R.R., 2004. Plant effects on
microbial assemblages and remediation of acidic coal pile
to have a major effect on the dynamics and diversity of the
runoff in mesocosm treatment wetlands. Ecol. Eng. 23,
bacterial community. 107–115.
(2) The variations introduced in the systems in terms of COTANCE, 2002. The European Tanning Industry Sustainability
hydraulic regimes and wastewater contamination did not Review. Confederation of Tanning Industries of the European
result in substantial changes in the diversity of the micro- Union Brussels, Belgium.
bial communities along the systems operation. Franco, A.R., Calheiros, C.S.C., Pacheco, C.C., De Marco, P., Manaia,
C.M., Castro, P.M.L., 2005. Isolation and characterization of
(3) A high diversity of bacterial populations was found in the
polymeric galloyl-ester-degrading bacteria from a tannery
CW units (according to DGGE profiles) and that could con-
discharge place. Microbial Ecol. 50, 550–556.
tribute to the resilience and resistance of the CWs to stress Fromin, N., Hamelin, J., Tarnawski, S., Roesti, D.,
created by the wastewater loadings applied. Jourdain-Miserez, K., Forestier, N., Teyssier-Cuvelle, S., Gillet,
(4) The type of substrate was more relevant in determining F., Aragno, M., Rossi, P., 2002. Statistical analysis of denaturing
the bacterial composition than the temporal variability, gel electrophoresis (DGE) fingerprinting patterns. Environ.
with the planted CWs units presenting a high similarity Microbiol. 4 (11), 634–643.
Henriques, I.S., Alves, A., Tacão, M., Almeida, A., Cunha, A.,
within the same year for root and substrate.
Correia, A., 2006a. Seasonal and spatial variability of
free-living bacterial community composition along an
estuarine gradient (Ria de Aveiro, Portugal). Estuar. Coast
Acknowledgements Shelf S 68, 139–148.
Henriques, I.S., Moura, A., Alves, A., Saavedra, M.J., Correia, A.,
The authors thank Dias Ruivo, Curtumes e Produtos Indus- 2006b. Analysing diversity among ␤-lactamase encoding
triais, Lda for making possible the establishment of this genes in aquatic environments. FEMS Microbiol. Ecol. 56,
project and to maxit, Argilas Expandidas, SA that kindly
Houba, V.J.G., Van Der Lee, J.J., Novozamsky, I., 1995. Soil Analysis
offered Filtralite® . Cristina S.C. Calheiros, Isabel Hen- Procedures—Other Procedures (Soil and Plant Analysis. Part
riques and Alexandra Moura wish to thank the research 5B). Department of Soil Science and Plant Nutrition. Sixth
grants from Fundação para a Ciência e Tecnologia (FCT), edition. Wageningen Agricultural University. Syllabus.
Portugal (SFRH/BDE/15507/2004, SFRH/BPD/21384/2005 and Wageningen, The Netherlands.
SFRH/BD/19433/2004). The work was supported by the project Ibekwe, A.M., Grieve, C.M., Lyon, S.R., 2003. Characterization of
microbial communities and composition in constructed dairy
wetland wastewater effluent. Appl. Environ. Microbiol. 69 (9),
references Kadlec, R.H., Knight, R.L., Vymazal, J., Brix, H., Cooper, P., Haberl,
R., 2000. Constructed wetlands for pollution
control—processes, performance, design and operation. IWA
Scientific and Technical Report No. 8. IWA Publishing. London,
Altschul, S.F., Madden, T.L., Schäffer, A.A., Zhang, J., Zhang, Z., UK.
Miller, W., Lipman, D.J., 1997. Gapped BLAST and PSI-BLAST: a Lane, D.J., 1991. 16S/23S rRNA sequencing. In: Stackebrandt, E.,
new generation of protein database search programs. Nucleic Goodfellow, M. (Eds.), Nucleic Acid Techniques in Bacterial
Acids Res. 25, 3389–3402. Systematics. John Wiley and Sons Ltd., West Sussex, England.
AOAC, 1995. Official Methods of Analysis, 16th ed. Association of Marschner, P., Neumann, G., Kania, A., Weiskopf, L., Lieberei, R.,
Official Analytical Chemists Inc., Arlington, VA, USA. 2002. Spatial and temporal dynamics of the microbial
APHA, 1998. Standard Methods for the Examination of Water and community structure in the rhizosphere of cluster roots of
Wastewater, 20th ed. American Public Health white lupin (Lupinus albus L.). Plant Soil 246, 167–174.
Association/American Water Works Association/Water Moura, A., Tacão, M., Henriques, I., Dias, J., Ferreira, P., Correia, A.,
Environment Federation, Washington, DC, USA. 2007. Characterization of bacterial diversity in two aerated
Baptista, J.C., 2003. Microbial communities in subsurface flow lagoons of a wastewater treatment plant using PCR–DGGE
wetlands. In: Proceedings of the 1st international seminar on analysis. Microbiol. Res., doi:10.1016/j.micres.2007.06.005.
the use of aquatic macrophytes for wastewater treatment in Muyzer, G., De Waal, E.C., Uitterlinden, A.G., 1993. Profiling of
constructed wetlands, Lisbon, Portugal, pp. 265–276. complex microbial populations by denaturating gradient gel
Baptista, J.D.C., Donnelly, T., Rayne, D., Davenport, R.J., 2003. electrophoresis analysis of polymerase chain
Microbial mechanisms of carbon removal in subsurface flow reaction-amplified genes coding for 16S rRNA. Appl. Environ.
wetlands. Water Sci. Technol. 48 (5), 127–134. Microb. 59 (3), 695–700.
Calheiros, C.S.C., Rangel, A.O.S.S., Castro, P.M.L., 2007. Nicomrat, D., Dick, W., Tuovinen, 2006. Assessment of the
Constructed wetland systems vegetated with different plants microbial community in a constructed wetland that receives
applied to the treatment of tannery wastewater. Water Res. acid coal mine drainage. Microbial Ecol. 51, 83–89.
41, 1790–1798.
e c o l o g i c a l e n g i n e e r i n g 3 5 ( 2 0 0 9 ) 744–753 753

Ojeda, E., Caldentey, J., Saaltink, M.W., García, J., 2008. Evaluation Truu, J., Nurk, K., Juhanson, J., Mander, Ü., 2005. Variation of
of relative importance of different microbial reactions on microbiological parameters within planted soil filter for
organic matter removal in horizontal subsurface-flow domestic wastewater treatment. Environ. Sci. Health 40,
constructed wetlands using a 2D simulation model. Ecol. Eng. 1191–1200.
34, 65–75. Vacca, G., Wand, H., Nikolausz, M., Kuschk, P., Kästner, M., 2005.
Pielou, E.C., 1975. Ecological Diversity. Wiley, New York. Effect of plants and filter materials on bacteria removal in
Shannon, C.E., Weaver, W., 1963. The Mathematical Theory of pilot-scale constructed wetlands. Water Res. 39, 1361–
Communication. University of Illinois Press, Urbana, IL. 1373.
Silva, M.F., Tiago, I., Veríssimo, A., Boaventura, A.R., Nunes, O.C., Vymazal, J., 2005. Horizontal sub-surface flow and hybrid
Manaia, C.M., 2006. Antibiotic resistance of enterococci and constructed wetlands systems for wastewater treatment.
related bacteria in an urban wastewater treatment plant. Ecol. Eng. 25, 478–490.
FEMS Microbiol. Ecol. 55 (2), 322–329. Ward, D.M., Weller, R., Bateson, M.M., 1990. 16S rRNA sequences
Stottmeister, U., Wießner, A., Kuschk, P., Kappelmeyer, U., reveal numerous uncultured microorganisms in a natural
Kästner, M., Bederski, O., Müller, R.A., Moormann, H., 2003. community. Nature 345, 63–65.
Effects of plants and microorganisms in constructed wetlands
for wastewater treatment. Biotechnol. Adv. 22, 93–117.