1 Introduction

Denitrification, a microbial process by which nitrate is reduced stepwise into atmospheric terminal products such as N2O and/or N2 (Lam and Kuypers 2011; Silva et al. 2018), has ecological and geochemical consequences in freshwater systems and wastewater treatment plants (Dunne et al. 2013; Nizzoli et al. 2018; Zhang et al. 2019). Denitrification mostly occurs in the sediment and acts as the prevailing sink for nitrogen in the water column or sediments under anaerobic conditions. Denitrification plays an important role in controlling nitrogen transformation, accounting for more than 70% of natural nitrogen loss (Babbin et al. 2014; Gao et al. 2016).

Organic carbon is a common electron donor for denitrification (Kesserű et al. 2002; Wunderlich et al. 2012; Wu et al. 2013), whereas inorganic substrates, including reduced sulphur (Yang et al. 2012) and iron compounds (Melton et al. 2014), can support denitrification as well. Sulphide oxidation coupled with nitrate reduction has been intensively studied (Jørgensen et al. 2009) and identified as an important pathway of nitrogen cycling in anoxic water (Canfield et al. 2010; Wenk et al. 2013) and potentially in sediments (Sayama et al. 2005; Burgin and Hamilton 2007). However, the possibility of Fe(II) as an electron donor for nitrate reduction in the natural environment had not been investigated until recent decades (Robertson and Thamdrup 2017).

Iron is an abundant redox-active element in sediments, and the transition between Fe(II) and Fe(III) plays an important role on watershed biogeochemistry (Weber et al. 2006a; Melton et al. 2014). Iron can be utilised as an electron donor by many various microorganisms that gain energy from that reaction, and the iron cycle is associated with many other elementary cycles, such as the nitrogen cycle (Laufer et al. 2016). Microbial and partly abiotic mechanisms by which Fe(II) is oxidised with nitrate are summarised as “nitrate-dependent iron oxidation” (Scholz et al. 2016). Water-soluble Fe(II) is generally less reactive, but it is readily utilised by microorganisms as an electron donor (Straub et al. 1996). These microorganisms include aerobic acidophilic (Bonnefoy and Holmes 2012), micro-aerobic neutrophilic (Krepski et al. 2012), and anaerobic bacteria using light (Hegler et al. 2008) or nitrate-dependent Fe(II)-oxidising bacteria (NDFOB) (Straub et al. 1996), which have been found in many different environments including soils, freshwater sediments, marine sediment, and hypersaline sediments (Shelobolina et al. 2012; Rowe et al. 2014; Laufer et al. 2015). Most NDFOB can reduce nitrate to nitrite, followed by further reduction to NO, N2O, and N2 (Chakraborty and Picardal 2013; Melton et al. 2014). The identification of NDFOB strains and investigation of functional enzymes involved in nitrate-dependent iron oxidation could be performed using molecular sequencing and transcription techniques (Beller et al. 2013; Dubinina and Sorokina 2014; Li et al. 2018; Hu et al. 2019). In addition, the critical metabolic components of NDFOB could be identified by whole-genome sequencing (Hu et al. 2017).

Various enzymes involved in the main reaction phase of nitrogen cycling have been discovered in lake sediments, and some scholars have conducted in-depth research on denitrifying enzymes (Cao et al. 2008; Gardner and White 2010). The conversion process of nitrogen compounds to nitrate can be carried out not only in microbial cells but also under enzymatic reactions. A specific enzyme, especially in anaerobic conditions, was involved in each stage of the nitrate transformation process. In contrast to our knowledge of NDFOB, it is largely unknown whether the reaction concerning nitrate reduction coupled with Fe(II) oxidation is related to denitrification enzyme activity (DEA) responsible for the process of nitrate reduction.

With the increase in nitrogen input to a lake, the likelihood of eutrophication increases, resulting in more frequent anoxic conditions in the lake. Hence, more transformation of Fe(III) to Fe(II) occurs, and the role of nitrate-dependent iron oxidation becomes more important. However, an understanding of the internal relationship between nitrogen transformation and nitrate-dependent iron oxidation under increasing nitrate and Fe(II) is still lacking. In this study, a 30-day incubation experiment with the sediments and overlying water from a freshwater lake was carried out. The nitrogen cycle, including water and gas, and its response to Fe(II) oxidation were investigated. We also assessed the possible effect of nitrate-dependent Fe(II) oxidation on denitrification enzyme activity and NDFOB numbers in sediments. Our aim was to better understand the connection between Fe(II) oxidation and the nitrogen cycle through denitrification.

2 Materials and methods

2.1 Sampling

Lake Moshui, a freshwater lake with an area of 3.14 km2 and a mean depth of approximately 1.2 m, is located in Wuhan, Hubei Province, China. The sediments and overlying water of Lake Moshui were taken from eight sites on April 9, 2018 (Fig. 1). The sediments at a depth of 10–15 cm were collected using a Petersen grab dredge and stored in a bag, and 10 L overlying water at each site was stored in a bucket. The pH, EC, dissolved oxygen (DO), and water temperature of the overlying water were measured in situ using a water quality analyser (EXO3, YSI international Co., Ltd., USA). In addition, several 200-mL glass bottles pre-filled with 4 mL hydrochloric acid solution were filled with overlying water and sealed with a rubber stopper for Fe(II) and Fe(III) analysis. The parameter properties in the overlying water and sediments are listed in Tables 1 and 2. All samples were stored at 4 °C before the incubation experiments.

Fig. 1
figure 1

Location of the sampling sites in Lake Moshui (site 1, 30°33′06.4″N 114°12′45.7″E; site 2, 30°32′44.9″N 114°12′44.0″E; site 3, 30°32′32.1″N 114°13′08.4″E; site 4, 30°32′44.4″N 114°13′09.2″E; site 5, 30°32′46.6″N 114°13′30.1″E; site 6, 30°32′38.9″N 114°13′50.9″E; site 7, 30°32′22.7″N 114°14′05.6″E; site 8, 30°32′09.6″N 114°13′38.3″E)

Full size image
Table 1 Physicochemical properties and nitrogen contents of the overlying water from Lake Moshui
Full size table
Table 2 Contents of different parameters in the sediments from Lake Moshui
Full size table

2.2 Sediment slurry incubations

The sediments and overlying water were prepared for the incubation experiment by mixing an equal amount of sample from the eight sites. Then, 100 g of fresh sediments and 400 mL of overlying water were added to 500-mL plastic bottles. One hundred eight bottles were prepared and split into three groups: a control group (CK), with original sediments and overlying water, and two treatment groups, to which NaNO3 was added to obtain a concentration of 32.5 μM nitrate (N group) and NaNO3+FeSO4·7H2O was added simultaneously to obtain the concentrations of 32.5 μM nitrate and 71.25 μM Fe(II) (N+F group). All incubation bottles were then sealed with an inner lid and outer lid to achieve anoxic conditions. Two needles fitted with a three-way valve were inserted into the inner lid for gas and solution sampling. The three groups were then placed on a shaking table (50 r min−1) at 25 °C for 10 min, followed by incubation in a dark incubator at 25 °C.

At each sampling time, 0, 0.5, 1, 3, 5, 7, 10, 14, 18, 22, 26, and 30 days, the incubation bottles were removed from the three groups. Three replicates were used in each group. The gas was sampled before the determination of other parameters. The pH and DO were measured in situ. Similarly, the sediment samples were freeze-dried (− 80 °C) for the following determinations.

An additional experiment was conducted to justify that the dominant reaction was biotic or abiotic. Sterilised (121 °C for 20 min) and unsterilised samples of the sediments and overlying water taken from Lake Moshui were prepared. The subsequent procedure was as described above for the N+F group, and the samples were prepared in triplicate. Gas for N2O flux determination after 1 day and water for nitrate determination after 18 days were taken with the aforementioned syringes.

2.3 Analysis and calculation

Overlying water was filtered (0.45 μm) for the measurements of nitrate, nitrite, and ammonium with standard photometric techniques (State Environmental Protection Administration of the People’s Republic of China 2002), and total nitrogen (TN) concentrations were measured using the oxidation method of alkaline potassium persulfate (Valderrama 1981). Dissolved iron in the overlying water and sediments was measured photometrically using the Ferrozine assay (Viollier et al. 2000). The pH and DO concentrations of the overlying water were determined using a water quality analyser (EXO3, YSI international Co., Ltd., USA). N2O and CO2 concentrations were analysed by a gas chromatograph (7890A, Agilent technologies Co., Ltd., USA) equipped with an electron capture detector. The concentrations of N2O and CO2 in the headspace gas were determined, and the N2O and CO2 fluxes were calculated using the following Eq. 1 (Hayakawa et al. 2013) and Eq. 2 (Wang et al. 2016), respectively.

$$ {F}_{{\mathrm{N}}_2\mathrm{O}}=\left[\rho \times C\times \left({V}_g+\alpha {V}_L\right)\times 273/\left(273+T\right)\right]/\left(M\times \varDelta t\right) $$
(1)
$$ {F}_{{\mathrm{CO}}_2}=\left[\rho \times C\times {V}_g\times 273/\left(273+T\right)\right]/\left(M\times \Delta t\right)\times 12/44 $$
(2)

where FN2O and FCO2 are the N2O and CO2 fluxes (μg N kg−1 day−1), respectively, M is the dry weight of the sediment (kg), Δt is the incubation time (day), ρ is the density of N2O and CO2 under standard temperature and pressure (1.25 g L−1 and 1.977 g L−1, respectively), C is the concentration of N2O and CO2 (ppmv), Vg is the headspace volume of the incubation bottle (L), α is the Bunsen coefficient of N2O (0.549 at 25 °C) (Tiedje 1994), VL is the liquid volume (L), 273 is the conversion factor between degrees Celsius and kelvin, T is the temperature (°C), and 12/44 is the conversion factor from CO2 to C.

The C2H2 inhibition method (Smith and Tiedje 1979) was used to determine the DEA of the sediment. NDFOB inoculation was conducted at the UV sterilisation station. The sediment samples were diluted to five gradients, 10−3, 10−4, 10−5, 10−6, and 10−7, and each dilution had five replicate series. Following the preparation of a serially diluted suspension, vitamins and trace metals (10 mL L−1 each) from stock solutions (Bruce et al. 1999) were added into the incubation tubes containing 1 mL serially diluted suspension. Sodium pyrophosphate (0.1%) was added to the incubation tubes sterilised prior to use. The incubation tubes were then filled with three different bicarbonate-buffered freshwater basal media (Weber et al. 2006c). For incubation of dilution, all incubation tubes were closed with butyl rubber stoppers to achieve anoxic conditions and incubated in a dark biochemical incubator at 25 °C. After 6 weeks of incubation, the growth was detected by the presence of a brownish-red or brownish-green precipitate in the incubation tubes (Weber et al. 2006c). The NDFOB in the incubation tubes were counted according to the MPN method (Cochran 1950).

2.4 Statistical analyses

The correlation between parameters was analysed by Pearson correlation analysis. F statistics were generated in SPSS 20.0 to evaluate the relationship of Fe(II) oxidation and N2O flux. Data from the sediment slurry incubation experiment were analysed for significant differences in N2O cumulative emission of representative days by comparison using a t test in SPSS 20.0.

3 Results

3.1 DO, pH, and nitrogen in the overlying water

The DO concentration in the three groups showed a sharp decrease after 1 day of incubation, followed by a plateau range of 0.6–1.0 mg L−1 (Fig. 2). A decrease during the early days was observed for the pH of all the groups, but little change in pH was found after 10 days of incubation. In general, the pH of the N and the N+F groups showed lower levels than that of the CK group, indicating that more H+ was produced after nitrate input during the incubation.

Fig. 2
figure 2

DO concentrations and pH in the overlying water during the incubation

Full size image

No sharp decrease in nitrate concentrations was observed in the overlying water of the CK group, whereas an obvious decrease in nitrate concentrations occurred in the N group and N+F group mainly due to denitrification (Fig. 3a). The reduction rate of nitrate in the N group (0.071 day−1, R2 = 0.9300) and N+F group (0.089 day−1, R2 = 0.9797) followed the pseudo-first-order reaction in the first-week incubation, consistent with a previous study by Pyzola (2013). After the 30-day incubation, the nitrate removal efficiencies in the N group and N+F group were 75.7% and 81.9%, respectively.

Fig. 3
figure 3

Nitrogen concentrations in the overlying water during the incubation. NO3--N (a), NO2--N (b), NH4+-N (c) and TN (d)

Full size image

Contrary to the nitrate decrease, the nitrite concentrations in the N and N+F groups illustrated a rising profile before 22 days (Fig. 3b). Nitrite concentrations for CK remained relatively stable during the incubation. Ammonium concentrations continued to be released from the sediment to the overlying water (Fig. 3c), probably attributed to the ammonification by the microorganisms in the sediments. Ammonium concentrations were higher in the CK group than those in the other two groups from 7 to 30 days. Drastic decreases in TN concentrations after 1 day of incubation were observed in all the samples (Fig. 3d), consistent with the decrease in DO. Despite the initial nitrate input, TN concentrations in the N group and the N+F group were lower than those in the CK group in most cases, suggesting much more rapid denitrification in the sediments.

3.2 Fe(II) and Fe(III) concentrations in the overlying water

The oxidation rate of Fe(II) followed the pseudo-first-order reaction in the N group and the N+F group in the first 5-day incubation (Fig. 4), and the rates were 0.064 day−1 (R2 = 0.8892) and 0.079 day−1 (R2 = 0.9670), respectively. No significant change in Fe(II) concentration was observed for the CK group (P > 0.05). After the 30-day incubation, the decrease in the percentages of Fe(II) concentration in the N+F group (47%) and the N group (43%) was similar but higher than that in the CK group (23%). As the initial Fe(II) concentration in the N+F group was the same as that in the CK group, a lower concentration at the end of incubation meant that more rapid transformation of Fe(II) occurred in the N+F group.

Fig. 4
figure 4

Fe(II) (a) and Fe(III) (b) concentrations in the overlying water during the incubation

Full size image

A drastic increase in Fe(III) in all groups was observed after a 1-day incubation, resulting from Fe(II) transformation. For the N+F group, the 30-day decrease in Fe(II) concentration was approximately 57 μmol L−1, whereas only a 42 μmol L−1 increase in Fe(III) concentration was found. The increase in Fe(III) concentration was not equal to the decrease in Fe(II) oxidation, probably attributed to the production of poorly soluble Fe(III) minerals.

3.3 N2O emissions

The anoxic conditions promoted denitrification, leading to a peak of N2O fluxes in all groups after 1 day (Fig. 5). A slow decline was illustrated during the latter incubation period. The N2O fluxes were significantly higher in the N and N+F groups than those in the CK group (P < 0.05), indicating that N2O emissions were promoted with nitrate addition. The N2O fluxes in the N and N+F groups ranged from 15.7 to 23.5 μg N kg−1 day−1 after 21 days and were higher than those in the CK group (5.4–7.3 μg N kg−1 day−1). In addition, the N2O emission in the N+F group was higher than that in the N group, and the difference could be mediated by the Fe(II) addition. During the whole incubation period, no significant differences in CO2 fluxes among the N, N+F, and CK groups (P > 0.05) were observed.

Fig. 5
figure 5

Variations in N2O (a) and CO2 (b) fluxes in the overlying water during the incubation

Full size image

3.4 DEA and NDFOB numbers in the sediment

The denitrification in the sediments was accelerated as the DO decreased drastically during 1 day of incubation, which could be reflected by the DEA (Fig. 6). Similar to the N2O emission, the DEA in the three groups showed a high level after 1 day, indicating an internal connection between N2O production and DEA, and then fluctuated between 6 and 9 mg N kg−1 h−1. Although no significant differences (P > 0.05) in DEA were observed among the three groups, the DEA of the N+F group was generally higher than that of the CK group and the N group.

Fig. 6
figure 6

DEA and NDFOB numbers in the sediments during the incubation

Full size image

The NDFOB in the sediments was approximately 1 × 105 cells g−1 wet sediment for all the three groups at the beginning of the incubation (Fig. 6). Comparing cell numbers in the three groups, we observed a similar development of cell numbers within the first 3 days: the exponential growth phase appeared shortly after incubation, and all samples reached a cell number of 9.0 × 105 cells g−1 wet sediment. After 7 days, the growth of NDFOB in the N and N+F groups was faster than that in the CK group. The cell numbers in the three groups were then kept stable after 10 days. The cell numbers of NDFOB at the end of incubation in the CK group, N group, and N+F group were 9.5 × 105, 1.1 × 106, and 1.4 × 106 cell g−1 wet sediment, respectively.

4 Discussion

4.1 Transformation of nitrogen coupled with iron oxidation in the overlying water

Nitrate coupled with Fe(II) oxidation results in the production of nitrite, N2O, N2, or ammonium, related to the physicochemical properties of water and the authigenic Fe minerals (Weber et al. 2006b; Scholz et al. 2016; Robertson and Thamdrup 2017). Compared with the N group, the N+F group showed higher nitrate reduction and nitrite production in the overlying water. This finding could be attributed to the following reaction (Scholz et al. 2016):

$$ {2\mathrm{Fe}}^{2+}+{{\mathrm{NO}}_3}^{-}+{5\mathrm{H}}_2\mathrm{O}\to 2\mathrm{Fe}{\left(\mathrm{OH}\right)}_3+{{\mathrm{NO}}_2}^{-}+{4\mathrm{H}}^{+} $$
(3)

Based on the pH, nitrate, nitrite, and Fe(II) concentrations of the incubation experiment, the Gibbs free energy was calculated, ΔG = −98.8 kJ/mol, indicating that the oxidation of Fe(II) was favourable in the anoxic sediment slurry incubation and provided powerful support for the continued increase in nitrite concentration in the N and the N+F groups. Lower concentrations of ammonium in the N and N+F groups than in the CK group were probably caused by the anaerobic ammonium oxidation (anammox), as an increase in nitrite production appeared in the two groups. However, anammox, one of the pathways for nitrogen removal in the watershed (Yu et al. 2016), usually played a minor role compared to the denitrification process (Schubert et al. 2006; Han et al. 2014). There was a significant relation between TN and nitrite (R = − 0.791, P < 0.01). An apparent decline in total nitrogen in the N and N+F groups (Fig. 7) meant that a reducing gas such as N2O and N2 was formed, by which nitrogen left the aqueous phase for the gas phase.

Fig. 7
figure 7

Values of NDFOB and DEA in relation to the concentrations of nitrate, nitrite, ammonium, TN, Fe(II), Fe(III), and the cumulative emission of N2O during the incubation. Numbers after three groups represent different incubation days

Full size image

The consumption of Fe(II) in the N+F group was high during the incubation, especially in the first 3 days, at approximately 8 μmol L−1 day−1. Fe(II) oxidation with oxygen, either abiotically or microbially induced, is a viable explanation for rapid Fe(II) consumption (Scholz et al. 2016). The increased Fe(III) concentration in the overlying water was not equal to the decreased Fe(II) concentration, probably because both soluble and insoluble Fe(II) were oxidised by bacterium to produce various iron minerals (Chaudhuri et al. 2001; Weber et al. 2006a). In the anoxic water, most iron is rapidly re-precipitated (Scholz et al. 2016). There was a negative relation between Fe(III) and TN (R = − 0.589, P < 0.01), whereas a positive correlation between Fe(III) and nitrite (R = 0.549, P < 0.01) was found. Nitrogen transformation and iron cycle are potentially coupled in anoxic environments (Kampschreur et al. 2011), and a positive relation was observed between the Fe(II) and nitrate concentrations (R = 0.483, P < 0.05). The reaction rates of Fe(II) and nitrate were used to establish the slopes of the linear least squares fit of Fe(II) and nitrate, thus estimating the contribution of Fe(II) oxidation to nitrate reduction via the following equation (Eq. 4):

$$ {5\mathrm{Fe}}^{2+}+{{\mathrm{N}\mathrm{O}}_3}^{-}+{7\mathrm{H}}_2\mathrm{O}\to \mathrm{FeO}\left(\mathrm{OH}\right)+1/2\ {\mathrm{N}}_2+{9\mathrm{H}}^{+}\ \Delta G{}^{\circ}=-96.23\ \mathrm{kJ}/\mathrm{mol} $$
(4)

The contribution of Fe(II) oxidation to the reduction of nitrate was estimated to be 27.7% according to the method by Pyzola (2013).

4.2 N2O production driven by iron oxidation

As nitrite accumulation was known to greatly increase N2O emission, N2O fluxes in the N and N+F groups during the incubation were higher than those in the CK group. The sharp increase in N2O emission after 1 day indicated that denitrification occurred rapidly in the anoxic conditions (Hibiya et al. 2003). Compared with the cumulative N2O emissions in the CK group, those in the N group and the N+F group at the end of incubation increased by 157% and 231%, respectively, indicating that nitrate reduction may proceed in a required environment with relatively high concentrations of nitrate related to the high rate of denitrification. A significant positive correlation (R = 0.757, P < 0.01, t test) between N2O fluxes and nitrate concentrations indicated that denitrification was most likely the main source of N2O.

The N2O fluxes were positively correlated with Fe(II) concentrations during incubation (R = 0.560, P < 0.01). N2O cumulative emission in the N+F group was generally higher than that in the N group, indicating that denitrification was stimulated with Fe(II) addition. Organic carbon is often used as the main electron donor for denitrification (Cayuela et al. 2014); however, Fe(II) can also support denitrification by abiotic and biotic pathways (Melton et al. 2014). Hence, the aspect of environmentally unwanted N2O production enhanced by this pathway should be realised when removing nitrogen from nitrogen-rich and iron-rich waters, especially in anoxic conditions (Kampschreur et al. 2011).

4.3 Microbial properties motivated by nitrate-dependent iron oxidation

The DEA and NAFOB numbers in sediments were important for the denitrification as the sediments were beneficial to nitrogen removal in watershed (Dai et al. 2018). The pH in the overlying water ranged from 7.2 to 7.8 during the incubation, providing a favourable environment for DEA. A previous study (Šimek et al. 2002) showed that the optimum pH for denitrifying bacteria growth and DEA is neutral or slightly alkaline. The DEA in the N+F group was higher than that in the CK group, indicating a high denitrification rate mediated by the Fe(II) addition. The DEA showed an apparent decline in all groups after a week incubation. This result was due to the nitrate-consuming reactions, and the lack of nitrate slowed the bacterial denitrification rates. A positive correlation between DEA and N2O emission (R = 0.711, P < 0.01) but a negative correlation between DEA and NDFOB (R = − 0.566, P < 0.01) was observed during the incubation (Fig. 7). Hence, the results did not indicate with certainty that the nitrate reduction coupled with Fe(II) oxidation was catalysed by DEA.

A significant relation between NAFOB and nitrite was found (R = 0.769, P < 0.01), and the relation between NAFOB and Fe(III) was also significant (R = 0.654, P < 0.01). The bacteria in the N group multiplied better than those in the CK group, indicating that NAFOB growth consumed nitrate as the nitrogen source (Straub et al. 1996; Straub and Buchholz-Cleven 1998). The NAFOB numbers in the N+F group were generally larger than those in the CK group, and the N group during incubation as NDFOB gained more energy with the additional Fe(II), and the activity of nitrate reduction coupled to Fe(II) oxidation was promoted to some extent. A previous study (Weber et al. 2009) has shown that Fe(II) acts as a “second substrate” that provides energy for microbial growth and that abundant energy is beneficial to NAFOB growth. A positive correlation between NDFOB numbers and N2O cumulative emission was observed (Fig. 7), illustrating that the nitrate reduction coupled with Fe(II) oxidation was driven by bacteria. NDFOB numbers ranging from 1 × 103 to 5 × 108 cell g−1 sediment (Hauck et al. 2001; Senn and Hemond 2002; Weber et al. 2006b) were reported, consistent with the number of NDFOB (1.1 × 105–1.4 × 106 cell g−1 wet sediment) observed in the sediments during the current incubation.

The nitrate concentration in the sterilised sediment was significantly higher than that in the unsterilised sediment on the 18th day of incubation (P < 0.01, t test) (Fig. 8). Furthermore, there was a significant difference (P < 0.01, t test) in the N2O fluxes and nitrate concentrations between sterilised and unsterilised sediment incubation (Fig. 8). A N2O flux of 55.4 μg N kg−1 day−1 was detected in the sterilised sediment slurry incubation, whereas an abundant N2O flux of 362.9 μg N kg−1 day−1 was observed in the unsterilised sediment after 1 day of incubation. Abiotic denitrification driven by Fe(II) can also occur in sediments, especially in the condition of nitrite accumulation (Kampschreur et al. 2011), and the chemical denitrification of nitrite mediated by Fe(II) oxidation can lead to the formation of N2O (Kampschreur et al. 2011). The sterilised experiment indicated that the abiotic pathway accounted for approximately 15% of the N2O production, and the biotic pathway played a major role in N2O production, i.e., denitrification.

Fig. 8
figure 8

Comparison of N2O (a) concentrations and nitrate (b) fluxes between sterilised and untreated samples of the sediment from Moshui Lake. The bars are labelled with the significant differences between the groups (P < 0.05, t test)

Full size image

5 Conclusions

The nitrate reduction was accelerated after Fe(II) addition, attributed to Fe(II) used as the second electron donor for denitrification. The contribution of Fe(II) oxidation to the nitrate reduction was 27.7% at the end of the incubation. The Fe(II) and nitrate addition led to an increase in N2O production, especially after 1 day of incubation. Sterilised experiments suggested that nitrate reduction and its subsequent N2O reduction were affected by the combination of biotic and abiotic reactions, and the biotic reaction was dominant. The numbers of NDFOB in the sediment during incubation were 1.1 × 105–1.4 × 106 cell g−1 wet sediment. The current study illustrated that anoxic nitrate-dependent iron oxidation played a non-negligible role in regulating nitrogen transformations, especially in iron-rich freshwater lake sediments.