The influence of nitrate addition and the presence of Glyceria maxima (reed sweetgrass) on the composition and nitrate metabolism of the dissimilatory nitrate-reducing bacterial community was investigated. Anoxic freshwater sediment was incubated in pots with or without G. maxima and with or without addition of nitrate. After incubation the sediments were sampled. Dilution series of these sediment samples were incubated in deep agar tubes and almost all colonies from the most diluted tubes were isolated and purified. When the nitrate concentration was low, 80% of the nitrate-reducing community in the rhizosphere of G. maxima consisted of NO2−-accumulating or NH4+-producing Bacillus strains. In bulk sediment with low nitrate concentrations, denitrifying Pseudomonas and Acinetobacter strains were dominant. The difference in the composition of the nitrate-reducing community between sediment with or without plants disappeared when nitrate was added. Denitrifying pseudomonads then made up 78–86% of the nitrate-reducing community. This shows that when the nitrate concentration was low, the presence of G. maxima greatly influenced the nitrate-reducing community. However, when nitrate was present and available, the composition and nitrate metabolism of the dominant nitrate-reducing community was probably not affected by G. maxima.
There are two known dissimilatory nitrate-reducing pathways: denitrification and the dissimilatory nitrate reduction to ammonia (DNRA). Denitrification is the reduction of nitrate (NO3−) to nitric oxide (NO), nitrous oxide (N2O) and, in most cases, dinitrogen (N2). By this process mineral nitrogen concentrations in soils and sediments may be reduced. In contrast, the dissimilatory reduction of NO3− to ammonia (NH4+) leads to conservation of mineral nitrogen in soils and sediments. Denitrification is the dominant process in soils and in sediments [1–4], whereas DNRA can be considerable in marine and freshwater sediments and sometimes in soils [5–7, 2]. Knowledge of the factors that determine the composition of the dissimilatory nitrate-reducing community would help to understand and predict temporal and spatial variation in denitrification and DNRA. The two processes are executed by different bacterial groups. The denitrifying population consists to a large extent of obligate oxidative organisms . Most of the denitrifying bacteria found in soil, freshwater or estuarine sediments belong to the genera Pseudomonas, Alcaligenes, Paracoccus[4, 1]. Bacteria dissimilating NO3− to NH4+ consists of facultative anaerobic, fermentative organisms, such as Klebsiella, Enterobacter, Vibrio, Citrobacter, Achromobacter and Bacillus[6, 8, 9, 4]. When there is sufficient NO3−, the most important factor determining which of the two nitrate-reducing processes will be favoured is probably the ability to compete for organic substrates . Another, important factor is the ratio of electron donors (i.e. organic substrates) and electron acceptors (i.e. nitrate and oxygen). Oxygen may also be an important environmental parameter in determining the composition of the nitrate-reducing community. In an anoxic environment without NO3− only the bacteria reducing NO3− to NH4+ will be able to grow, due to their fermentative abilities. A special environment in largely anoxic sediments is the root zone of aerenchymatous plant species, as these plants will release oxygen into the root zone [10–12]. The oxygen released may stimulate NO3− production and hence may lead to higher nitrate-reducing activities [13, 14]. Little is known about the influence of wetland plant species, like Glyceria maxima (Hartm.) Holmb., on the end-products of bacterial nitrate reduction, even though wetlands contribute substantially to global N2O production [15, 16].
The aim of the research presented was to examine the influence of G. maxima and the addition of nitrate on the composition and nitrate metabolism of the nitrate-reducing bacterial community.
2Materials and methods
2.1Sampling location and procedure
Bulk sediment (0–15 cm upper layer) was sampled in August 1990 at Junner Koeland, (52°30′N, 06°30′E) in a shallow, isolated meander of the river Vecht, in the Netherlands. G. maxima was the dominant aerenchymatous plant at this location. The sediment was transported to the laboratory in airtight buckets at 4°C, thoroughly mixed and the initial sediment parameters were determined. Six pots (500 ml) were filled with sediment, to study the effects of NO3− additions and the presence of G. maxima, on the species composition and nitrate metabolism of the nitrate-reducing community. In four of the six pots three non-sterile seedlings were planted. Pots with plants received either 0 or 531 μg NO3−-N (g dry sediment)−1, pots without plants received 0 or 235 μg NO3−-N (g dry sediment)−1. NO3− addition between sediments with and without plants differs because of differences in evapotranspiration and evaporation. The experiment was conducted in a growth chamber (Heraeus Vötsch, HPS-1500, Heraeus, Wijk bij Duurstede, Netherlands), air humidity was 80%, temperature 20°C–15°C with a light-dark regimen of 16–8 h and a light intensity of 215±10 μmol (m2·s)−1. After 9–12 weeks incubation, the roots had spread throughout the sediment and all sediment could be regarded as rhizosphere sediment. Rhizosphere sediment from the pots with plants and bulk sediment from the pots without plants were collected and the composition and nitrate metabolism of the nitrate-reducing community was determined.
The pH (H2O) of the sediments was determined by shaking 5 g of moist sediment with 10 ml water for 2 h. The pH (KCl) and the mineral nitrogen concentrations were determined by shaking 5 g of moist sediment with 10 ml 1 M KCl also for 2 h. After shaking, samples were taken and centrifuged at 15 000×g in a Biofuge A (Heraeus Christ, Dijkstra Verenigde, Almere, Netherlands) bench centrifuge for 10 min. The supernatants were analyzed for NO3−, NO2− and NH4+ using a Technicon Traacs 800 autoanalyzer (Technicon Instruments Corp., Tarrytown, NY, USA) with a detection level of 10 μM for all three compounds. The organic matter content was determined by analysis of weight losses after heating of about 5 g of dry sediment at 550°C for 4 h.
2.3Composition of the dominant nitrate-reducing community
Three dilution series of sediment samples were made in deep agar tubes. The medium consisted of tryptic soy broth (TSB) (per litre: 19 g tryptone, 3 g soy peptone, 2.5 g dextrose, 5.0 g chloride and 2.5 g phosphate) with nitrate (4 mM) and 1% (w/v) agar, pH 7.1–7.3 . After incubation for 10 days at 20°C, as many as possible visually different colonies were picked from the most diluted tubes and subcultured aerobically until pure cultures were obtained. The isolated potentially nitrate-reducing bacteria were tentatively divided into different groups by four tests: the Gram test , the oxidase test with 1% (w/v) tetramethyl-p-phenylene-diamine-HCl, the catalase test with 10% (v/v) H2O2 and the oxidation-fermentation test with glucose as substrate . The Gram-positive strains were examined microscopically for the possession of endospores. The Gram-negative strains were further identified with API-20NE and API-20E tests (Api System, S.A., Montalieu-Vercieu, France), to genus level.
2.4Nitrate metabolism of the dominant nitrate-reducing bacteria
The cultures were tested on different media under different conditions. The first test was under anoxic conditions in deep agar tubes. The medium consisted of TSB with and without nitrate (10 mM). The tubes were incubated at 20°C. Based on this test the initially isolated strains were divided in three groups. The first group of isolates was non-culturable under the tested conditions. The second group of isolates was able to grow but did not consume NO3−: the non-nitrate reducing strains. The third group of isolates grew in the tubes and was also able to consume NO3−: the nitrate-reducing strains. The isolates producing gas in the presence of NO3− and without growth in the absence of NO3− were considered to be potentially denitrifying strains. When growth was observed in the tubes with and without NO3−, the strains were tentatively classified as bacteria able to dissimilate NO3− to NH4+. To determine the metabolism of the nitrate-reducing strains more precisely these bacteria were again tested on TSB (100%) with and without NO3− in 100 ml serum bottles. The bottles were flushed with N2 for 10 min and subsequently 10 kPa C2H2 was injected, to inhibit reduction of N2O to N2. The bottles were incubated for 2 weeks at 20°C. After the incubation period, the bottles were tested for N2O production, NO3−, NO2− and NH4+ concentration. N2O concentrations were measured using a gas chromatograph (6000 VEGA series 2, Carlo-Erba Instruments, Milan, Italy) equipped with a hot wire detector and a Porapack Q column. N2 was used as carrier gas (flow rate 30 ml per min). The column, injector and detector temperatures were 80°C, 120°C and 119°C, respectively. Peak area was computed by an integrator (model CR3A, Shimadzu, Interscience, Breda, Netherlands). The nitrate-reducing strains were divided into three groups based on the major end-products that were formed, group I: N2O producers, reducing >80% of the added NO3− to N2O; group II: NO2− accumulators, reducing >50% of the added NO3− to NO2− and N2O; group III: NH4+ producers, reducing about 50% of the added NO3− to NH4+, and no NO3−, NO2− or N2O were detectable.
Data of sediment parameters were analyzed by one-way ANOVA. Differences between means were tested for significance using the least significant difference (LSD) procedure.
3Results and discussion
The sediment parameters of Junner Koeland are given in Table 1. After incubation, the NO3− concentration only decreased significantly, compared to the initial value, in the unamended rhizosphere sediment. In contrast the NH4+ concentrations, especially in the rhizosphere, decreased significantly during incubation compared to the initial NH4+ concentration. Only in bulk sediment to which 235 μg NO3−-N (g dry sediment)−1 had been added, the NH4+ concentration did not show significant decrease compared to the initial concentration. With the addition of NO3−, the pH(H2O) and the pH(KCl) increased significantly.
Table 1. Sediment parameters of Junner Koeland sediment before the experiment (initial) and after harvest
% Organic matter
Concentrations are given in μg N (g dry sediment)−1. *Significant differences are indicated with different letters (P<0.05). N.D.=not determined.
3.2Composition of the dominant nitrate-reducing bacteria
In Table 2, the effect of the different NO3− additions and the presence of G. maxima on the percentage non-culturable, non-nitrate-reducing and nitrate-reducing strains are given. The percentage of nitrate-reducing strains in the rhizosphere increased when NO3− was added. In bulk sediment with NO3− addition the percentage of nitrate-reducing bacteria was low, due to the high proportion of non-culturable strains.
Table 2. The effect of different NO3− additions on the percentage non-culturable, non-nitrate-reducing and nitrate-reducing strains in the rhizosphere of Glyceria maxima and in bulk sediment
Concentration added nitrate given in μg N (g dry sediment)−1.
% Non-nitrate reduction
% Nitrate reduction
These percentages agreed well with the percentage non-culturable and non-nitrate-reducing strains initially isolated from various soils and sediments [19, 1]. From the initial isolated bacteria 37% to 41% were non-culturable when transferred to liquid media and 31% of the isolated strains were unable to reduce nitrate .
The culturable strains were identified to genus level regardless the ability to reduce NO3−. In this way the composition of the culturable strains could be compared to the composition of the culturable, nitrate-reducing strains. Although the percentage of strains unable to reduce NO3− was rather high, the genus composition was not affected by excluding the non-nitrate-reducing strains from the total number culturable, nitrate-reducing strains. Data presented in Fig. 1 show the genus composition of the culturable nitrate-reducing community present. The results show that different genera were present in the rhizosphere and bulk sediment when no NO3− was added. Since the tested strains were not consistently isolated from the most diluted tubes, to obtain as many as possible visually different strains (n=105), the genus composition was also determined based on colonies isolated only from the most diluted tubes (10−8, 10−9, n=34). The composition of the nitrate-reducing community based on strains isolated from these tubes was similar (data not shown) to the data presented in Fig. 1. So, the composition represents the dominant culturable nitrate-reducing genera (Fig. 1). From the deep agar tubes it was not possible, due to gas production, to determine the total number of potentially nitrate-reducing bacteria. Nonetheless, information concerning the number of gas-producing strains could be gained from these tubes. The number of gas-producing strains was the same in rhizosphere and bulk sediment without NO3− addition (data not shown). In the rhizosphere sediment without NO3− addition, the nitrate-reducing community consisted of 80% of Gram-positive strains. In the bulk sediment Gram-negative Pseudomonas and Acinetobacter strains were more abundant. When NO3− was added the difference between rhizosphere and bulk sediment almost disappeared. In both cases Gram-negative strains became dominant. The number of gas-producing strains increased by one (rhizosphere) or two (bulk) orders of magnitude, compared with no NO3− addition.
The genus composition in the rhizosphere and non-rhizosphere of T. angustifolia was different from that of G. maxima. In the rhizosphere of T. angustifolia many Enteriobacteriaceae and Aeromonas/Vibrio bacteria were present . In soil the composition of the nitrate-reducing community was investigated by Smith and Zimmerman . The nitrate-reducing community consisted of denitrifiers: Pseudomonas, Flavobacterium and Alcaligenes species and NO2− accumulators: Bacillus, Enterobacter, Flavobacterium and Citrobacter species. In an estuarine sediment a large percentage of the nitrate-reducing community belonged to the fermentative Aeromonas/Vibrio group [6, 9]. Most strains isolated from a soil were Gram-positive bacteria, showing endospores, suggesting that they were Bacillus strains .
3.3Nitrate metabolism of the dominant nitrate-reducing bacteria
Fig. 2 shows the nitrate metabolism of the dominant bacteria belonging to the nitrate-reducing community. In the rhizosphere without NO3− addition, the percentage of N2O producers was only 7%, whereas the percentage of NO2− accumulators was 40% and NH4+ producers 53%. In the bulk sediment 71% of the isolated strains produced N2O. When NO3− was added the nitrate-reducing community in the rhizosphere and bulk sediment consisted of 78% and 86% of N2O producers, respectively. In the rhizosphere sediment only 4% of the Bacillus strains produced NH4+. Our results, concerning the nitrate metabolism of the nitrate-reducing community, did not differ much from the nitrate metabolism of the nitrate-reducing community isolated from the rhizosphere and non-rhizosphere of Typha angustifolia, taking the vague difference between NO2− accumulators and assumed NH4+ producers into account and excluding the unidentified bacteria .
Apparently the presence of the flooded aerenchymatous plant G. maxima was of great importance for the composition of the nitrate-reducing bacterial community under conditions of NO3− depletion. However, when NO3− was added, the presence of G. maxima had little influence on the composition and nitrate metabolism of the dominant nitrate-reducing community, provided nitrogen enters the environment as NO3−. The opposite may be true when only NH4+ is present. The presence of G. maxima is then probably important, due to the coupling between nitrification and denitrification in sediments , especially in the rhizosphere of aerenchymatous plants [13, 12, 4]. So, under NO3− limiting conditions, NO3− produced in the rhizosphere of G. maxima will subsequently be converted to NH4+ and the production of N2O will be restricted. Only under conditions of sufficient NO3− supply denitrification and consequently N2O production may occur but this is not affected by G. maxima. Thus G. maxima has no effect on the production of the greenhouse gas N2O, either with or without the addition of NO3−.
The hypothesis is put forward that under NO3− limiting conditions the influence of the aerenchymatous plant G. maxima on the composition and nitrate metabolism of the nitrate-reducing community is large. When sufficient NO3− is available, the nitrate-reducing bacteria are less dependent on G. maxima and consequently the influence of G. maxima is minimal.
However, Brunel et al.  stated that the roots of T. angustifolia had no distinct effect on the dominance of denitrifiers or fermentative bacteria reducing NO3− to NH4+. The results of this study do differ from our results, also regarding the composition of the nitrate-reducing community. This could be due to the different aerenchymatous plant or to the fact that in the sediment of T. angustifolia NO3− was not limited.
In future, further research is necessary to verify the hypothesis, and to clarify if the sediment type, the aerenchymatous plant species or the nitrate concentration is an important parameter controlling the composition of the nitrate-reducing bacterial community. Also, further research is necessary to prove that NO3− availability quantitative leads to higher nitrate-reducing activity and dominance of denitrifying bacteria, regardless the presence of an aerenchymatous plant.
We thank dr. W. de Boer and Prof. Dr. J.W. Woldendorp for the valuable comments on the manuscript.