Editor: Riks Laanbroek
Structure and function of denitrifying and nitrifying bacterial communities in relation to the plant species in a constructed wetland
Article first published online: 19 NOV 2008
© 2008 Federation of European Microbiological Societies. Published by Blackwell Publishing Ltd. All rights reserved
FEMS Microbiology Ecology
Volume 67, Issue 2, pages 308–319, February 2009
How to Cite
Ruiz-Rueda, O., Hallin, S. and Bañeras, L. (2009), Structure and function of denitrifying and nitrifying bacterial communities in relation to the plant species in a constructed wetland. FEMS Microbiology Ecology, 67: 308–319. doi: 10.1111/j.1574-6941.2008.00615.x
- Issue published online: 22 DEC 2008
- Article first published online: 19 NOV 2008
- Received 15 May 2008; revised 5 September 2008; accepted 15 September 2008.First published online 19 November 2008.
- denitrifying bacteria;
- nosZ gene;
- constructed wetland
The community structure and potential activities of nitrifying and denitrifying bacteria were studied in the rhizosphere of Typha latifolia and Phragmites australis present in a free water system constructed wetland (CW). Potential nitrate reduction and nitrification activities were shown to be significantly higher in the rhizosphere when compared with the nonvegetated sediment. Higher rates were generally obtained for P. australis. The community structure of denitrifying bacteria in the rhizosphere differed from that found at the bulk sediment, as revealed by PCR-denaturing gradient gel electrophoresis (DGGE) of the nitrous oxide reductase encoding gene nosZ. Results also show a greater nosZ genotype diversification and suggest a plant species effect in rhizosphere samples obtained during events of low hydraulic retention times. Ammonia-oxidizing communities were less complex on the basis of PCR-DGGE analysis of the 16S rRNA gene. Retrieved sequences were all related to Nitrosomonas marina and Nitrosomonas ureae, being both present in rhizosphere and bulk sediment regardless of environmental changes. The results demonstrate the effect of vegetation on the functioning and structure of bacterial communities involved in the removal of nitrogen in the treatment cells of a CW and point to the use of vegetation coverage to promote nitrification or denitrification in particular areas.
Constructed wetlands (CWs) have been extensively developed in the last decades as alternatives to on-site treatment methods for diffuse or nonpoint nitrogen pollution of water, and are used for biological treatment of wastewater from industries or medium size municipalities (Leonard & Swanson, 2001). CWs are particularly suited to promoting water reuse in areas of reduced freshwater resources, such as the Mediterranean region (Mandi et al., 1998; Angelakis et al., 1999). Nitrogen removal in wetlands occurs, apart from direct assimilation by plants, through microbial nitrification and denitrification activities. Nitrification includes two aerobic processes; oxidation of ammonia to nitrite, which is further oxidized to nitrate. The first process is performed by a phylogenetically constrained group of ammonia-oxidizing bacteria in addition to crenarchaea, whereas the last step is carried out by the nitrite-oxidizing bacteria. Denitrification is an anaerobic respiration process, in which nitrate is reduced stepwise to dinitrogen gas by a diverse range of bacteria and archaea. In wetlands, these processes are mainly associated with the sediment or other submersed surfaces (Kallner Bastviken et al., 2003).
Vegetation coverage is a key factor controlling nitrogen removal in wetlands both directly and indirectly. It slows down the water flow rate, which prolongs the residence time and thereby the reaction time, provides organic matter and promotes growth of microorganisms in the sediment (Cooper et al., 1996; Ibekwe et al., 2007). Submerged and emergent plants may also represent an additional oxygen source for bacteria growing as biofilms on plant surfaces, in the rhizosphere or the adjacent sediment, explaining differences in coupled nitrification and denitrification in the rhizosphere compared with the bulk sediment (Risgaard-Petersen & Jensen, 1997; Ottosen et al., 1999). Despite the importance of vegetation, plant and plant-species effects on nitrification and denitrification and the microbial communities involved have not been well studied in CWs. In soil ecosystems, plants have been shown to induce and stimulate the growth of specific bacterial groups and create well-defined bacterial communities around the rhizosphere due to multiple and interwoven factors, for example root exudation or rhizodeposition at the root surface (Sørensen, 1997; Jaeger et al., 1999; Duineveld et al., 2001; Smalla et al., 2001; Philippot et al., 2002; Sharma et al., 2005).
A comparative study of functional groups involved in nitrogen removal in the rhizosphere of wetland vegetation may help to better understand the distribution of these microorganisms and to infer their activity in distinct areas of treatment wetlands, especially when compared with potential sediment metabolic capabilities (Kallner Bastviken et al., 2003; Dollhopf et al., 2005). Cattail (Typha latifolia and Typha angustifolia), reed (Phragmites australis and Phragmites communis) and bulrush (Scirpus sp. and Schoenoplectus lacustris) are among the most commonly used species for artificial vegetation coverage of CWs worldwide. Typha sp. and Phragmites sp. dominate in the Spanish Mediterranean area and are often planted in single or mixed stands. Knowledge about plant species effects on nitrogen removal is crucial when designing CWs to optimize and maybe compartmentalize the nitrification and denitrification processes.
The objective was to determine the impact of Typha sp. and Phragmites sp. on community structure and activity of nitrifying and denitrifying bacteria in the sediment of a free water-surface CW (FWS-CW) at two different operational conditions. The FWS-CW is located in the Costa Brava area in Spain and has to cope with highly varying influent loadings due to the many tourists during summer vacation compared with other periods of the year. The two sampling occasions were chosen to evaluate plant effects under low and high water flows and nitrogen loadings.
Materials and methods
Site description and sampling of sediment and rhizosphere
The FWS-CW system is located downstream the wastewater treatment plant of EDAR-Empuriabrava, which is situated outside Girona in the Costa Brava area in Spain. The wetland consists of three parallel cells with a total surface area of 8000 m2 and an average depth of 0.5 m. The aquatic vegetation is composed of P. australis and Typha sp., which mainly form separate stands. Large seasonal variations in the influent wastewater flow are common due to tourism in the region in summer. To assure that effluent ammonium concentrations never exceed 7 mg L−1, the load is restricted by bypassing the water flow to the nearby located Muga River. This action is regularly performed to prevent the development of severe anoxic conditions in the CW, which have been related to the development of avian botulism outbreaks in the natural reserve Els Aiguamolls de l'Empordà located downstream of the CW.
Sediment and rhizosphere were sampled June 16 and September 28, 2006. The June samples were representative for a restricted water-flow period and the CW was almost drained (Table 1). The September period was characterized by short water residence times in the wetland and a high nitrogen load, mainly in the form of ammonia. Three to six replicate rhizosphere samples of P. australis (P1–P8) and T. latifolia (T1–T8), as well as sediment devoid of vegetation (S1–S12) were collected in the third cell of the Empuriabrava CW. Replicates from each sample type were collected randomly in a 1 m2 area. Reed and cattail stands were located 40 m apart at a central point in the wetland. Because of difficulties accessing the CW, bulk sediment was not sampled from exactly the same site in June and September. Sampling was done using a 7-cm-diameter plexiglass tube mounted in a manual core sampler. Roots and root fragments were carefully removed from sediments by hand. Rhizosphere samples were considered as the portion of sediment that remained attached to plant roots after washing twice with sterile isotonic solution. Sediments were manually homogenized using a sterile spatula and roots were cut into c. 0.5-cm pieces and homogenized. Aliquots of all sediment replicates were maintained at −20 °C for further analysis.
|WWTP effluent||FWS-CW effluent||Flow (m3 day−1)||Total N load FWS-CW (kg)||N removal FWS-CW (%)|
|N-NH4+ (mg L−1)||N-NO3− (mg L−1)||N-NH4+ (mg L−1)||N-NO3− (mg L−1)|
|June mean (SE)||18.43 (1.92)||1.36 (0.41)||2.65 (0.88)||0.48 (0.13)||54.8 (44.9)||39||99|
|September mean (SE)||9.83 (4.24)||4.91 (1.54)||1.26 (0.30)||2.55 (0.29)||1074.1 (288.5)||591||94|
Potential activity assays
Potential nitrate+nitrite reduction rates were determined as an indication of denitrification activity in all samples. For the September samples, potential denitrification activity was also measured using the acetylene inhibition technique, according to Pell et al. (1996) with some modifications. Preparation of sediment slurries was similar for both methods. Approximately 10 g of fresh sediment or roots were placed in 250-mL flasks and diluted in 49 mL of sterile isotonic solution. The atmosphere in the flasks was exchanged after five evacuation and filling cycles with N2. After 30-min agitation, incubations were supplemented with KNO3 (28 mg L−1 NO3−-N, final concentration) and ammonium acetate (0.5 g L−1), ethanol (0.2 g L−1) and propionic acid (0.3 g L−1). When measuring denitrification with the acetylene method, 30 mL acetylene were added to reach a 0.1 atm partial pressure before substrate addition. Samples of the liquid phase for nitrite and nitrate determination were taken every 30–45 min for the first 2 h of incubation and every hour for the next 4 h. Gas samples were taken every 12 min during the first hour and then every half hour for 5 h.
Potential nitrification activity measurements were determined as nitrate+nitrite production rates. Incubations were performed in a similar way as those for nitrate+nitrite reduction, but under aerobic conditions. The slurries were amended with (NH4)2SO2 to a final concentration of 210 mg L−1 NH4+-N and liquid samples were collected over a 24-h period. All activity measurements were done at 25 °C with continuous agitation (150 r.p.m.). Controls with sterilized sediments and roots were included.
Samples for the analysis of nitrate and nitrite concentrations (1 mL) were collected directly from slurries, centrifuged for 2 min at 12 000 g and filtered (0.2 μM). Samples were kept at −20 °C until analyzed by HPLC (Waters Corporation, Barcelona, Spain) using a 4.6 × 200 mm Waters Spherisorb® 5 μM SAX as analytical column (Smith et al., 2002). Nitrous oxide in gas vials was analyzed on a gas chromatograph (model CP 9000, Chrompack, Rotterdam, the Netherlands) equipped with a 63Ni electron capture detector. Initial and final ammonium concentrations in the nitrate+nitrite reduction activity experiments were determined in a destillator BÜCHI B-324, using an automatic titrator (Titrino 719S Metrohm). Five milliliter samples of either sediment or rhizosphere slurries were centrifuged at 3000 g for 5 min and the supernatant was collected in a sterile tube and acidified with 1N HCl. Samples were kept at −20 °C until analyzed.
DNA extraction and PCR conditions
DNA was extracted using the Ultraclean Soil DNA kit (MO BIO Laboratories Inc.) following the manufacturer's instructions and the concentration was measured using a NanoDrop® ND-1000 spectrophotometer (NanoDrop Technologies Inc., Wilmington, DE). PCR for the amplification of nosZ gene fragments were performed with primers nosZ-F (5′-CGY TGT TCM TCG ACA GCC AG-3′) (Kloos et al., 2001) and nosZ 1622R (5′-CGS ACC TTS TTG CCS TYG CG-3′), or nosZ1622R-GC before denaturing gradient gel electrophoresis (DGGE) (Throbäck et al., 2004). A touchdown PCR was performed in a minicycler (MJ Research) according to PCR conditions presented by Enwall et al. (2005).
Partial 16S rRNA genes of the β-subgroup of ammonia oxidizers were obtained by PCR amplification using the primer pair CTO189F AB/C-GC and CTO654R (5′-CTA GCY TTG TAG TTT CAA ACG C-3′) (Kowalchuk et al., 1997). Amplification reactions were run in a Geneamp 9700 thermal cycler according to PCR conditions described by (Kowalchuk et al., 1997).
DGGE analysis of nosZ and 16S rRNA genes
Three hundred nanograms of nosZ PCR products were loaded on 7% (v/v) acrylamide–bis-acrylamide gels with a 40–70% urea–formamide denaturing gradient. DGGE was performed in a DCode system from Bio-Rad Laboratories Inc. according to Enwall et al. (2005). For DGGE analysis of the 16S rRNA gene of the ammonia-oxidising bacteria (AOB), DGGE was carried out in an Ingeny phorU system (Ingeny, the Netherlands) using 6% acrylamide–bis-acrylamide gels and a 35–70% urea–formamide denaturing gradient (Bäckman et al., 2003). In both DGGE assays, gels were run for 17 h at 120 V and stained for 30 min with Sybr® Gold (Molecular Probes Europe, Invitrogen Corporation), for visualization under UV excitation. Images were captured with GelDoc 2000 system. For AOB gels, representative bands of every position in the gel were excised using a sterile scalpel. The DNA fragment was recovered by elution in Tris/HCl at 65 °C during 1 h and reamplified as described above.
Based on the DGGE fingerprinting results, rhizosphere samples from June and September were selected for construction of four nosZ clone libraries. PCR products from sample triplicates were pooled, separated by electrophoresis on agarose gels, and purified with the MiniElute Gel Extraction Kit (Qiagen). Amplicons were then cloned with the TOPO TA Cloning Kit for Sequencing (Invitrogen) according to the manufacturer's specifications. Transformants were checked for correct insert size by transferring cells with a toothpick into 25 μL PCR reactions prepared as described above. Restriction fragment length polymorphism (RFLP) was used to screen 90, 87, 63, and 84 positive clones for Typha sp. September (TS), Typha sp. June (TJ), Phragmites sp. September (PS) and Phragmites sp. June (PJ) samples. PCR products were digested overnight with AluI at 37 °C, and then loaded onto 2% agarose gels, electrophoresed for 2 h at 90 V. Gel image analysis was carried out using Gelcompar II (Applied Maths, NV) to identify different RFLP patterns and group clones accordingly. Coverage was calculated according to C=1−nN−1, where n is the number of different RFLP types encountered only once, and N is the total number of clones analyzed in the library. Clones from RFLP pattern types represented by at least two clones were selected from each of the four libraries and grown overnight in Luria–Bertani tubes at 37 °C before plasmid isolation with the QIAprep Spin Miniprep kit (Qiagen).
Sequencing in both directions of cloned partial nosZ gene fragments and AOB 16S rRNA gene fragments obtained from reamplification of DGGE excised bands, was performed by the Macrogen service (Macrogen, Korea). Sequences were manually checked and aligned using the clustalw software (European Bioinformatics Institute, http://www.ebi.ac.uk).
Statistical and phylogenetic analyses
Differences in potential activity rates were tested for the effects of planted vs. nonplanted areas, plant species, sampling time and the corresponding interactions by anova using the spss software (SPSS Inc.). Differences between means of either potential nitrification or nitrate+nitrite reduction were analyzed using Tukey's honestly significant difference (HSD) post hoc test at P≤0.05. The analyses were performed with the General Linear Model (GLM) procedure of the spss software (SPSS Inc.).
DGGE patterns of nosZ genes were analyzed using Gelcompar II (Applied Maths). Known samples were loaded in all DGGE gels and used as internal standards for the comparison between fingerprints. For sample comparison, a presence–absence matrix was used to calculate similarities between patterns and statistical analysis based on hierarchical cluster analysis was performed with the Jaccard distance and the unweighted pair group method with arithmetic mean grouping algorithm. The sequences were compared with those deposited in the GeneBank (NCBI) database using the blastn software (http://www.ncbi.nlm.nih.gov/BLAST/). Phylogenetic analyses of the deduced amino acid sequences were conducted using the software mega version 4 (Tamura et al., 2007). Neighbour-joining trees were reconstructed by Amino Poisson correction and pairwise deletion. Tree topology was evaluated by bootstrap analysis using 1000 replicates.
The nitrate+nitrite reduction rates were significantly (P<0.05) higher in the P. australis and Typha sp. rhizosphere samples compared with the sediment samples on both sampling occasions (Fig. 1a). Not only the plant, but also the plant species affected the reduction rate. Potential rates measured in the P. australis rhizosphere were slightly higher than in that of Typha sp., although differences were not significant, due to the variation within replicates at each rhizosphere location. No significant differences in nitrite+nitrate reduction potential rates between any of the locations were observed at both sampling occasions. Ammonium was not detected (<0.3 mg NH4+-N L−1) during the incubation, which indicates neglectable dissimilatory nitrate reduction to ammonia under these conditions. In the September samples, potential denitrification rates were also determined from N2O production after acetylene inhibition. Denitrification rates were 19.9±3.6 and 14.1±1.1 in the P. australis and Typha sp. rhizosphere, respectively, whereas the rate in the sediment was 2.9±0.4 μg N2O-N g−1 DW h−1 (data not shown). Thus, the plant effect on denitrification rates were in agreement with those determined as nitrate+nitrite reduction, and the two methods correlated (R2=0.8). Nevertheless, the denitrification rates were two to three times lower compared with the nitrate+nitrite reduction rates. A similar relationship between the two techniques has been reported previously when applied to CW sediments and related to an incomplete blockage of the reduction of nitrous oxide (Seitzinger et al., 1993; Kallner Bastviken et al., 2003).
The potential nitrification rates were significantly higher in the planted areas compared with nonplanted sediments (P<0.01). Sampling time was not a significant variable for nitrification potential in the bulk sediment or the P. australis rhizosphere according to the results of a one-way anova test, but a more detailed analysis using Tukey's HSD test indicated a significant shift in the potential rate found in Typha sp. rhizosphere. Here, the rate was similar to the rates in the rhizosphere of P. australis during June, but comparable to that obtained for the bulk sediment during September (Fig. 1b). However, the interaction between plant species and time was not significant.
nosZ community structure
The DGGE patterns of nosZ genotypes from the rhizosphere samples were different from the sediment samples, although some bands were shared between most DGGE profiles (Fig. 2). Thus, vegetation was an important factor controlling the denitrifier community structure in the wetland. Besides the rhizosphere effect, a temporal effect was detected as the June and September samples separated from each other within the sediment and rhizosphere cluster, respectively. A plant species-specific separation was seen in the September samples, but this was not as obvious in the June samples of the two rhizospheres. In addition, DGGE fingerprints corresponding to rhizospheres obtained from Typha sp. showed higher small-scale variation between replicates whereas the P. australis rhizosphere fingerprints were more similar between replicates, indicating a less heterogeneous environment in the proximity of the roots of this plant species.
From the clone libraries of the rhizosphere samples, in total 37 RFLP types could be assigned and 31 of them were found in more than one clone. Based on the RFLP types, the coverage was estimated to be 93%, 93%, 85% and 97% in the TJ, TS, PJ and PS samples, respectively. A few types dominated and seven RFLP types comprised almost 70% of the clones. Whereas some types were detected in all libraries, the majority of RFLP types, including the most abundant, seemed to be site or time specific (Fig. 3). As an example, types 25 and 19 as well as the less abundant types 16 and 27, were only found in clones obtained from Typha sp. rhizosphere, while types 32, 28 and 24 appeared only in clone libraries from P. australis. In general, September libraries contained a higher number of different RFLP types.
A number of clones from all RFLP types represented by at least two clones were selected for sequencing. The sequences showed similarities with previously reported nosZ sequences from environmental clones retrieved from natural river sediments, agricultural soils and activated sludge. The phylogram of the deduced amino acid sequences separated the clones into two main groups (Fig. 4). Sequences present in cluster I corresponded to the RFLP types frequent in samples from June, whereas cluster II was more heterogeneous and dominated by clones retrieved from the September samples. Because the dominant RFLP types were represented by all four clone libraries, sequences obtained from either Typha sp. or P. australis libraries appeared randomly distributed over the phylogram.
AOB community structure
The AOB community pattern had a rather low complexity (Fig. 5). Sediment and rhizosphere AOB community fingerprints were similar in June, but the September samples showed higher variation between sample types. In addition, the DGGE patterns indicate a season-related change for all sample types. Sequencing confirmed that all bands were AOB, except bands 1, 2 and 18. These were related to Methylobacillus sp. and only detected in the September profiles. All AOB sequences were phylogenetically affiliated to Nitrosomonas species, and showed at least 98% homology with sequences from environmental clones retrieved from wastewater treatment systems. The majority of sequenced AOB bands showed high similarities to each other and the closest cultured match was Nitrosomonas marina, while the fuzzier bands (nos 7, 19 and 30) were related to Nitrosomonas ureae (Fig. 6).
Both potential nitrification and nitrate reduction rates in the sediment were in general higher in the rhizosphere compared with bulk samples, showing that vegetation was an important factor for the nitrogen removal capacity in the CW. Plant surfaces in wetlands have been pointed out as significant sites for nitrifying bacteria (Eriksson & Weisner, 1999; Körner, 1999; Kallner Bastviken et al., 2003) and this could be part of the explanation for the higher nitrification rates observed in the rhizosphere. Emergent plants also control dissolved oxygen concentrations in the rhizosphere, by root respiration and transport of oxygen from the air down through the aerenchyma to the water-saturated sediment. The plant effect on nitrification rates was probably attributed to the oxygen supplied by the plants (Jespersen et al., 1998; Armstrong & Armstrong, 2001). In contrast, effects on nitrate reduction can be a mix of various factors. It has been argued that the ability to respire anaerobically using nitrogenous compounds could be a selective advantage for denitrifiers in the rhizosphere. Mutants unable to synthesize the membrane-bound nitrate reductase, the cd1 nitrite reductase or the copper nitrite reductase, were outcompeted by the wild-type strains in maize rhizosphere, which demonstrates that denitrification itself could be an advantage for root colonization (Philippot et al., 1995; Ghiglione et al., 2000). In agreement, von Berg & Bothe (1992) found that the ratio of denitrifiers to other heterotrophic bacteria increased near the roots.
The results indicated that P. australis supported nitrification and nitrate reduction in the sediment better than T. latifolia did, which points toward a plant species dependence of the measured potential activities. The greater development of the P. australis root zone in deep sediment layers would increase the oxygen supply, which favors the activity of AOB. Ammonium removal in CW have been shown to be higher in Phragmites sp. stands than in those planted with Typha sp. (Gersberg et al., 1986), but it remains unclear whether this is mainly due to higher nitrification rates in Phragmites sp. rhizosphere, or due to a more efficient denitrification in sediments covered with this plant species. The finding that potential denitrification activities in intact cores with Phragmites sp. or Typha sp. shoots were not significantly different, as reported by Kallner Bastviken et al. (2003), supports the theory that Phragmites sp. mainly plays a role in nitrogen removal by enhancing nitrification activity. In our study, potential nitrification rates were low and accounted for <10% of the potential nitrate reduction rates. Because nitrogen removal in treatment wetlands predominantly loaded with ammonia depend on an efficient nitrification step, management practices to favor nitrification are important. Single-stage CWs seldom achieve high removal of total nitrogen due to their inability to provide both aerobic and anaerobic conditions at the same time (Vymazal, 2007). The results obtained in the present work indicate the potential use of Phragmites sp. as the best candidate plant species to support nitrogen removal along the water flow in CWs.
Because the Phragmites sp. effect on nitrification and nitrate reduction potential was only significant in the September samples, we suggest that the observed differences might depend on changes in environmental conditions over time in addition to changes in plant physiology. The total nitrogen load and the ammonium to nitrate ratio in the influent to the CWs changed over the studied period, mainly because of fluctuations in the mean water residence time. The wastewater loading to the CW during June was very low due to the bypass flow used when ammonium concentration in the effluent from the wastewater treatment plant is higher than 7 mg L−1 This situation caused a volume loss of c. 30% in the CW, which limited the nutrient load for a long period. The condition in September was the opposite and sampling coincided with high nutrient loads and low water retention time.
The number of AOB genotypes detected by DGGE showed only minor differences among samples. In June, both rhizosphere fingerprints were comparable to those from the bulk sediment, but, in September, the bulk sediment was more complex than the rhizosphere samples. However, no clear difference either between sample type or occasion for the AOB was observed when analyzing the sequences derived from the DGGE. All retrieved AOB-like sequences revealed homologies to yet uncultivated bacteria related to N. ureae and Nitrosomonas oligotropha (cluster 6a, three sequences) and to N. marina (cluster 6b, 14 sequences). The former group has been shown to dominate in some activated sludge plants treating wastewater (Dionisi et al., 2002; Hallin et al., 2005) and in soil and freshwater sediments contaminated with wastewater (Purkhold et al., 2000; Cebron et al., 2003). Cluster 6b-related sequences appear to be dominating the AOB community in the wetland and this cluster is also numerically important in activated sludge processes (Purkhold et al., 2000; Wagner et al., 2002). In our study, it may also be related to seawater infiltration to the wastewater pipes in the area. In fact, relatively high conductivity values are routinely measured in the water (1400–4 500 μS cm−1; L. Sala, Consorci de la Costa Brava, Girona, pers. commun.). A few sequences obtained from the bulk sediment in September clustered with Methylobacillus sp. and Methylophilus sp. It has previously been shown that the CTO primers can amplify non-AOB sequences (Purkhold et al., 2000; Bäckman et al., 2003), which implies that misleading conclusions can be drawn if only based on DGGE fingerprints.
In contrast to the AOB community, a clear plant effect on the structure of the denitrifier community was detected by the nosZ PCR-DGGE fingerprints. Similarly, Philippot et al. (2002), focusing on narG genes coding for nitrate reductase, found that the community structure of nitrate-reducing bacteria in unplanted and maize-planted soils was linked to plant distribution, supposedly due to selection for some strains, which might possess better competitive traits. In our study, the DGGE fingerprints of nosZ furthermore suggested plant species-specific effects on the composition of nosZ genotypes in the CW, especially in the September samples. A similar effect was observed in another wetland, where the denitrifying community structure, targeted by nirS genes coding for the heme cd1-nitrite reductase, differed in the sediment with an invasive cattail hybrid Typha×glauca compared with the sediment with native plant species, for example Scirpus sp. (Angeloni et al., 2006). Also in soil ecosystems, plant species have been shown to influence the composition of denitrifiers (Patra et al., 2006; Bremer et al., 2007), although when investigating the effect of Lolium perenne and Trifolium repens on the nitrate reducer community, plant species had no effect (Deiglmayr et al., 2004). In studies assessing the total bacterial community, specific community patterns were associated with different plants in agricultural soils (Smalla et al., 2001; Costa et al., 2006), but Ravit et al. (2006) suggested that individual plant species have a less pronounced effect in wetlands, especially if subject to disturbances.
In addition to the plant effects, we found clear seasonal variations in the nosZ genotype community structure in the DGGE patterns, both within bulk sediment and rhizosphere sites. The cloned nosZ fragments from rhizosphere samples revealed differences in the community structure mainly according to sampling time and only partly according to plant species. Seasonal variations in denitrifying communities are most likely explained by a combination of plant effects and wetland management-related factors. These include water residence time and nutrient load, which have been identified as key regulators of denitrifier community activity and composition in treatment wetlands (Kjellin et al., 2007). In agreement, ammonium and nitrate loadings have previously been shown to correlate to changes in the community structure of nirS-type denitrifiers in the bulk sediment of the Empuriabrava FWS-CW in areas devoid of vegetation (Ruiz-Rueda et al., 2007), but the effect on the rhizosphere community was not investigated. Sequences retrieved from June showed fewer polymorphisms and mainly related to nosZ in Betaproteobacteria, which contrasted with the greater heterogeneity encountered in sequences obtained from the September rhizosphere samples. Most of the nosZ sequences showed homologies to those from uncultivated bacteria obtained from environmental samples, such as estuarine sediments, agricultural soil and activated sludge.
In conclusion, vegetation influenced potential nitrification and nitrate reduction as well as the community structure of ammonia-oxidizing and denitrifying bacteria in the sediment of the CW. T. latifolia and P. australis rhizospheres exhibited significantly different nitrification and nitrate reduction rates, which may be of particular interest in the management of CWs at high nutrient loadings. A plant species effect on the microbial community was only seen for denitrifying bacteria and the differences between Typha sp. and Phragmites sp. associated denitrifiers were more evident in periods of high nutrient load, indicating that environmental conditions, plant development and plant species together changed the bacterial community in the rhizosphere of emergent plants in CWs.
O.R.-R. is the recipient of a PhD scholarship from the Generalitat de Catalunya. This work has been financially supported by the Spanish Ministerio de Educación y Ciencia (CGL2006-02382/HID) and the Formas financed Uppsala Microbiomics Center. Lluís Sala and Anna Huguet are gratefully acknowledged for providing environmental data and valuable comments on the functioning of CWs. Magali Martí helped in field sampling and activity measurements.
- 1999) The status of wastewater reuse practice in the Mediterranean basin: need for guidelines. Water Res 33: 2201–2217. , , & (
- 2006) Effects of an invasive cattail species (Typha×glauca) on sediment nitrogen and microbial community composition in a freshwater wetland. FEMS Microbiol Lett 263: 86–92. , , & (
- 2001) Rice and Phragmites: effects of organic acids on growth, root permeability, and radial oxygen loss to the rhizosphere. Am J Bot 88: 1359–1370. & (
- 2003) Liming induces growth of a diverse flora of ammonia-oxidising bacteria in acid spruce forest soil as determined by SSCP and DGGE. Soil Biol Biochem 35: 1337–1347. , , & (
- 2007) Impact of plant functional group, plant species, and sampling time on the composition of nirK-type denitrifier communities in soil. Appl Environ Microb 73: 6876–6884. , , , , & (
- 2003) Nitrification and nitrifying bacteria in the lower Seine River and estuary (France). Appl Environ Microb 69: 7091–7100. , & (
- 1996) Reed Beds and Constructed Wetlands for Wastewater Treatment. WRc Publications, Swindon, Wiltshire, UK. , , & (
- 2006) Effects of site and plant species on rhizosphere community structure as revealed by molecular analysis of microbial guilds. FEMS Microbiol Ecol 56: 236–249. , , , , & (
- 2004) Structure and activity of the nitrate-reducing community in the rhizosphere of Lolium perenne and Trifolium repens under long-term elevated atmospheric pCO2. FEMS Microbiol Ecol 49: 445–454. , , & (
- 2002) Quantification of Nitrosomonas oligotropha-like ammonia-oxidizing bacteria and Nitrospira spp. from full-scale wastewater treatment plants by competitive PCR. Appl Environ Microb 68: 245–253. , , , , & (
- 2005) Quantification of ammonia-oxidizing bacteria and factors controlling nitrification in salt marsh sediments. Appl Environ Microb 71: 240–246. , , , , & (
- 2001) Analysis of bacterial communities in the rhizosphere of chrysanthemum via denaturing gradient gel electrophoresis of PCR-amplified 16S rRNA as well as DNA fragments coding for 16S rRNA. Appl Environ Microb 67: 172–178. , , , & (
- 2005) Activity and composition of the denitrifying bacterial community respond differently to long-term fertilization. Appl Environ Microb 71: 8335–8343. , & (
- 1999) An experimental study on effects of submersed macrophytes on nitrification and denitrification in ammonium-rich aquatic systems. Limnol Oceanogr 44: 1993–1999. & (
- 1986) Role of aquatic plants in wastewater treatment by artificial wetlands. Water Res 20: 363–368. , & (
- 2000) Role of respiratory nitrate reductase in ability of Pseudomonas fluorescens YT101 to colonize the rhizosphere of maize. Appl Environ Microb 66: 4012–4016. , , , & (
- 2005) Community survey of ammonia-oxidizing bacteria in full-scale activated sludge processes with different solids retention time. J Appl Microbiol 99: 629–640. , , , , , & (
- 2007) Impact of plant density and microbial composition on water quality from a free water surface constructed wetland. J Appl Microbiol 102: 921–936. , , & (
- 1999) Mapping of sugar and amino acid availability in soil around roots with bacterial sensors of sucrose and tryptophan. Appl Environ Microb 65: 2685–2690. , , , & (
- 1998) Growth and root oxygen release by Typha latifolia and its effects on sediment methanogenesis. Aquat Bot 61: 165–180. , & (
- 2003) Potential nitrification and denitrification on different surfaces in a constructed treatment wetland. J Environ Qual 32: 2414–2420. , , , , & (
- 2007) Spatial variations in denitrification activity in wetland sediments explained by hydrology and denitrifying community structure. Water Res 41: 4710–4720. , & (
- 2001) Denitrification within the genus Azospirillium and other associative bacteria. Aust J Plant Phyisiol 28: 991–998. , , & (
- 1999) Nitrifying and denitrifying bacteria in epiphytic communities of submerged macrophytes in a treated sewage channel. Acta Hydroch Hydrob 27: 27–31. (
- 1997) Analysis of ammonia-oxidizing bacteria of the beta subdivision of the class Proteobacteria in coastal sand dunes by denaturing gradient gel electrophoresis and sequencing of PCR-amplified 16S ribosomal DNA fragments. Appl Environ Microb 63: 1489–1497. , , , , & (
- 2001) Comparison of operational design criteria for subsurface flow constructed wetlands for wastewater treatment. Water Sci Technol 43: 301–307. & (
- 1998) Application of constructed wetlands for domestic wastewater treatment in an arid climate. Water Sci Technol 38: 379–387. , & (
- 1999) Direct and indirect measurements of nitrification and denitrification in the rhizosphere of aquatic macrophytes. Aquat Microb Ecol 19: 81–89. , & (
- 2006) Effects of management regime and plant species on the enzyme activity and genetic structure of N-fixing, denitrifying, and nitrifying bacterial communities in grassland soils. Environ Microbiol 8: 1005–1016. , & (
- 1996) Potential denitrification activity assay in soil-with or without cloramphenicol. Soil Biol Biochem 28: 393–398. , , & (
- 1995) Use of Tn5 mutants to assess the role of the dissimilative nitrite reductase in the competitive abilities of two Pseudomonas strains in soil. Appl Environ Microb 61: 1426–1430. , & (
- 2002) Molecular analysis of the nitrate-reducing community from unplanted and maize-planted soils. Appl Environ Microb 68: 6121–6128. , , , & (
- 2000) Phylogeny of all recognized species of ammonia oxidizers based on comparative 16S rRNA and amoA sequence analysis: implications for molecular diversity surveys. Appl Environ Microb 66: 5368–5382. , , , , & (
- 2006) Salt Marsh vegetation affects rhizosphere microbial community structure and function. Soil Biol Biochem 38: 2359–2371. , & (
- 1997) Nitrification and denitrification in the rhizosphere of the aquatic macrophyte Lobelia dortmanna L. Limnol Oceanogr 42: 529–537. & (
- 2007) Diversity of the nitrite reductase gene nirS in the sediment of a free-water surface constructed wetland. Int Microbiol 10: 253–260. , , & (
- 1993) Denitrification measurements in aquatic sediments: a comparison of three methods. Biogeochemistry 23: 147–167. , , & (
- 2005) Characterization of bacterial community structure in rhizosphere soil of grain legumes. Microb Ecol 49: 407–415. , , , & (
- 2001) Bulk and rhizosphere soil bacterial communities studied by denaturing gradient gel electrophoresis: plant-dependent enrichment and seasonal shifts revealed. Appl Environ Microb 67: 4742–4751. , , , , , , , & (
- 2002) Evaluation of methods for the extraction of nitrite and nitrate in biological fluids employing high-performance anion-exchange liquid chromatography for their determination. J Chromatogr B 779: 201–209. , & (
- 1997) The rhizosphere as a habitat for soil microorganisms. Modern Soil Microbiology (Van ElsasJD, TrevorsJT & WellingtonEMH, eds), pp. 21–45. Marcel Dekker, New York, NY. (
- 2007) MEGA4: molecular evolutionary genetics analysis (MEGA) software version 4.0. Mol Biol Evol 24: 1596–1599. , , & (
- 2004) Reassessing PCR primers targeting nirS, nirK and nosZ genes for community surveys of denitrifying bacteria with DGGE. FEMS Microbiol Ecol 49: 401–417. , , & (
- 1992) The distribution of denitrifying bacteria in soil monitored by DNA-probing. FEMS Microbiol Ecol 86: 331–340. & (
- 2007) Removal of nutrients in various types of constructed wetlands. Sci Total Environ 380: 48–65. (
- 2002) Microbial community composition and function in wastewater treatment plants. Antonie Van Leeuwenhoek 81: 665–680. , , , , & (