The Effects of Concentration Gradient of Nitrogen Compounds on the Ammonium-nitrogen and Nitrate-nitrogen Fluxes at Sediment-Water Interface

The incubation experiments focused on altering concentration gradients of nitrogen between sediment and overlying water to examine the diffusion ux of ammonium-nitrogen (NH 4+ ) and nitrate-nitrogen (NO 3- ) at sediment-water interface. In this study, the diffusion ux can be estimated based on calculating the average of the net change rate of nutrient concentrations in the overlying water. For the incubation experiment of different TN concentrations in the sediment, the results showed that the diffusion ux of ammonia at sediment-water interface is -52.57~84.57 mg·m -2 ·d -1 , and for nitrate diffusion ux, the changing range during the incubation experiment is -110.13~143.25 mg·m -2 ·d -1 . For the incubation experiment of different nitrogen concentrations in the overlying water, the results of NH 4+ -N diffusion ux in L, M, H treatment were 3.37, -4.94, -3.84 mg·m -2 ·d -1 , respectively. And the average diffusion ux of nitrate in L (0 mg NO 3- -N, 0 mg NH 4+ -N), M (0.5 mg NO 3- -N, 1.5 mg NH 4+ -N) and H (1 mg NO 3- -N, 2.5 mg NH 4 + -N) treatment were 12.30, 10.39 and 7.11 mg·m -2 ·d -1 . Results highlighted that concentrations gradient of nutrients were indeed an important factor affecting the diffusion ux at sediment-water interface. In addition, the diffusion of nutrients at sediment-water interface in aquatic ecosystem is not only controlled by concentration gradients, some other factors such as incoming water, hydrodynamics, dissolved oxygen content, sediment structure, biological disturbance, horizontal migration and diffusion of nutrients and turbulent diffusion caused by wind and wave, are equally important.


Introduction
The continuous exchanges of solutes and particles between the sediment and its overlying water play a critical role in material cycling in the aquatic ecosystem, which affect the biological activity, chemical composition and ecosystem structure as well. In the biogeochemical cycles, there are three main processes should be considered. Firstly, the particulate organic matter in the overlying water is deposited into the sediment layers, which referred to as the depositional process. And then the particulate organic matter is mineralized in the sediment to form soluble intermediates. Lastly, some dissolved chemicals can transfer to the overlying water by diffusion or buried via sedimentation. The global nitrogen cycle has been dramatically changed due to the human activities, some practices such as agricultural fertilization increase the nitrogen level in aquatic ecosystems (Galloway et al.,2004;Li et al.,2021). There are plenty of negative impacts on increasing nitrogen loads such as bottom water hypoxia, nitrogen is a key factor leading to eutrophication, besides, nitrate is known to contribute to aquatic area acidi cation (Kelly et al.,1990). Phosphorus (P) enrichment is also a cause of eutrophication, many researchers have stated that it should be controlled as well (Paerl et al., 2011).
However, nitrogen biogeochemical cycling is more complicated. Therefore, controlling N is often a more challenging approach when dealing with aquatic ecosystem eutrophication (Queiroz et al.,2020). External nutrients sources are often considered the main cause of eutrophication, thus the primary method of eutrophication control is to mitigate external input, which include sewage discharge, non-point source pollution and fertilizer application, etc. ( that even if external input is greatly controlled, the aquatic ecological environment is not enough to restore to the ideal state as nitrogen will be continuously released from sediments to overlying water Different nitrogen reactions are occurred in the sediment, overlying water and the sediment-water interface. Within the sediments, reactive nitrogen deposition as organic N which can produces ammonium (NH4 + -N) by mineralization, NH4 + -N is further converted into nitrate (NO 3 − -N) by nitri cation under aerobic condition. Below the oxycline, NO 3 − -N (via denitri cation) and NH4 + -N (via anammox) are convert to N 2 in the absence of O 2 . The dissolved NO 3 − -N and NH4 + -N also diffused between the sediment and overlying water under the driving force of concentration difference. These reactions control the nitrogen net exchange uxes across the sediment-water interface, which may alter the availability of nitrogen in the aquatic ecosystem. Controlling the presence of nitrogen in aquatic system is a critical management issue due to the high inputs both from non-point source pollution and inner release, resulting in the problem of eutrophication and hypoxia (Mulholland et al., 2008).
Therefore, it's signi cant to recognize the nitrogen exchange process in the sediment-water interface for nitrogen ux control. Reactive nitrogen is released from sediments and maintained in the pore water, in the static aquatic ecosystem, the nitrogen transferred into the overlying water across the sediment-water estimated based on the concentration difference between sediment (pore water) and overlying water. The rates and magnitudes of uxes depend on the nitrogen content of sediment and dissolved nitrogen concentrations in overlying water. Despite the importance of these two factors on controlling the rates of N uxes in the sediment-water ecosystem, there have been no comparisons of in uence degree of the two factors on the diffusion ux. By identifying which factor makes the diffusion more e cient, the relevant results have certain signi cance in the management strategy of controlling internal release.
The aim of this study was to examine and determine how concentration gradient between pore water and surface water in uences nitrogen dynamics at sediment-water interface. We hypothesized that concentration gradient is a major factor affecting N migration at sediment-water interface. In this paper, we designed two groups of indoor incubation experiments which were different sediment nitrogen concentrations and different overlying water nitrogen concentration, respectively. We further quantify the diffusion uxes under these two factors and their migration direction, which de ned as either deposit from overlying water into sediment or release from sediment to water. Finally, we relate these ndings to their impact on ecosystem processes and some suggestions on controlling internal release were put forward.

Material And Method
Page 4/17

Description of the study sites
The experiments were conducted on three types of sediments with different total nitrogen (TN) and organic matter concentrations in the hydro-uctuation belt of Danjiangkou reservoir. According to the previous research (Wang et al., 2020), the sediment with highest TN concentration (sample 1) located in Gaozhuang (sandy soil; 32°28′15″N 111°24′10″E), the sediment with median TN concentration (sample 2) located in Zhuyuan village (tidal soil; 33°0′47″N 111°17′37″E) and the sediment with lowest TN concentration (sample 3) located in Nanchengang (loam; 32°26′38″N 111°24′12″E). The top 10 cm sediment cores were sampled in March 2021. Apart from these sediment cores, three extra sediment samples were collected at each site to test sediment properties before the incubation experiment. The collected sediment samples were put in a sealed plastic bag and kept at 4℃ for further analysis. Table 1 shows the sediment properties of surface soils (0-10 cm) at the three sampling sites.

Incubation experiment design
The experiments were carried out in April 2021 in the laboratory at North China University of Water Resources and Electric Power. The square aquaria were prepared for incubation experiment. The dimension of the aquarium was 15 cm×12 cm×30 cm. The aquaria were lled with 10 cm sediments and the sediments were inundated with 15 cm self-made solution, which was similar to the property of natural river water and the speci c compositions were shown in Table 2, the solution was equilibrated by pumping air for 48 hours, the pH, Eh, TDS and conductivity of the solution were 7.07, 72 mv, 0.049 g/L and 0.081 ms/cm, respectively. The pore water sampler (Rhizon SMS-10 cm, 0.15 µm, Female luer lock, Rhizosphere Research Products bv.) was placed 1 cm below the sediment-water interface ( Figure 1). To explore the effects of different TN concentrations in the sediment on diffusion ux of nitrogen, we selected 3 types of sediment samples (Table 1), the determination of sampling points was based on the soil type, the land use mode before inundation and the measured value in the hydro-uctuation belt in Danjiangkou reservoir (Incubation experiment 1). For different nitrogen concentrations in the overlying water simulation, the sediment collected at S1 were added to overlying water with different nitrogen concentrations. Three concentration gradients are proposed to be designed: (1) 0 mg/L ammonia nitrogen, 0 mg/L nitrate nitrogen; (2) 0.5 mg/L ammonia nitrogen, 1.5 mg/L nitrate nitrogen; (3) 1.0 mg/L ammonia nitrogen, 2.5 mg/L nitrate nitrogen (Incubation experiment 2). All incubations were carried out with three replicates per treatment and utilizing distilled water as overlying water was considered as controls. In the expressions, S refers to arti cial solutions and D refers to distilled water.

Water samples
During the experiment, water temperature (T), pH, dissolved oxygen (DO) and total dissolved solid (TDS) were measured daily by a portable multi-functional water quality analyzer (HORIBA, Japan) at 10 AM. The surface water was collected at 0d, 3d, 6d, 9d, 12d and 14d by using a 60 ml syringe at 5 cm below the water surface for incubation experiment 1, and the surface water was collected at 0d, 4d, 8d, 12d and 15d for incubation experiment 2. Pore water was taken anaerobically at the same frequency by connecting 50ml syringe to the rhizon sampler. After each sampling, the aquaria were replenished with 100 ml selfmade solution to compensate for the loss of the sampling volume. Besides, water losses by evaporation etc. were compensated with demineralized water adjusted according to water level. For water samples, TN, NH 4 + -N, NO 3 − -N, NO 2 − -N and total organic carbon (TOC) were measured.

Sediment samples
Before and after the incubation experiment, amount of sediment samples was collected to compare the changes of each treatment. The sediment samples need pretreatment before analysis which include airdrying, homogenization and grinding. For sediment samples, TDN, NH 4 + -N, NO 3 − -N, NO 2 − -N and OM were measured.

Analysis methods
Analysis methods (Wang, 2020) were as follows: TN was determined by the alkaline potassium persulfate digestion-UV spectrophotometry method. Samples for analysis of NH 4 + -N, NO 3 − -N and NO 2 − -N were ltered (0.45 µm cellulose acetate lters), NH 4 + -N was determined by Nessler's reagent colorimetric method, NO 3 − -N and NO 2 − -N was determined by a UV spectrophotometry method. For sediment OM measurement, a part of sediment was dried at 100°C for 48h until a constant weight was achieved, the difference between the pre and post weights were regarded as the organic fractions. OM and porosity were measured before and after the incubation experiments. TOC was determined by potassium dichromate-volumetric method.

Diffusion ux estimation
The nutrients diffusion ux at the sediment-water interface can be estimated based on the average of the net change rate of nutrient concentrations in the overlying water at each sampling interval during incubation experiment, and then normalized by sediment -water interface area. The equations are as follows: Since the overlying water needs to be replenished during the incubation experiment, which will affect the initial concentration in the overlying water, thus the results should be corrected by Equation 4.

Quality control and statistical analysis
For all samples, triplicate analyses were measured and the results expressed as the average of these. In the experiment, analytical pure reagent and ultrapure water were used, and the reagent used for standard curve is the standard solution prepared by Jinnong Science and Technology Co., Ltd, Zhengzhou. All statistical were carried out with SPSS for windows.

( )
In different sediment TN concentrations incubation experiment group, the variation range of NH 4 + -N concentration in overlying water is relatively consistent (

Effects of different sediment TN concentrations on diffusion ux of nitrogen at sediment-water interface
Sediment-water interface is the most signi cant place for material exchange between the bottom of aquatic ecosystem and overlying water. Under the solo or synergistic effects of various physical, chemical and biological processes such as concentration gradients, microbial metabolism and benthic fauna disturbance, dissolved particles can migrate between sediment pore water and overlying water through sediment-water interface. Therefore, sediment usually play a role of "source" or "sink" for nutrients and pollutants migration.
The diffusion uxes of TN at sediment-water interface from different sampling sites are shown in Fig. 1, the diffusion uxes of TN gradually increase with incubation time. At the beginning of the experiment the sediment act as the accumulation sinks of nitrogen and then transformed into release source. The average diffusion ux of TN at sampling points S1, S2 and S3 were 72.45, 66.35 and 66.30 mg·m −2 ·d −1 , respectively. During the incubation experiment, TN is released from sediment to the overlying water, and the higher TN concentration in the sediment, the greater the diffusion ux was shown.
Across the entire incubation experiment, both ammonia and nitrate showed deposition and release process, and there was no regular pattern in magnitude and direction amongst the treatments. The diffusion ux of ammonia at sediment-water interface is -52.57~84.57 mg·m −2 ·d −1 , and for nitrate diffusion ux, the changing range during the incubation experiment is -110.13~143.25 mg·m −2 ·d −1 , the results indicated that the diffusion ux of nitrate was slightly larger than ammonia (Fig. 2-3). This can be explained by the aerobic condition in the overlying water, in which the dissolved oxygen concentration is 6.93~10.81 mg·L −1 , the NH 4 + -N was transformed into NO 3 − -N by nitri cation process. Thus, there was an increase in NO 3 − N concentrations, and its diffusion ux also increased. In the early stage of the incubation experiment, NH 4 + -N dominated the exchange process at sediment-water interface. While in the late stage of the experiment, the exchange process of NO 3 − N was relatively signi cant. The average diffusion ux of ammonia at sediment-water interface in S1, S2 and S3 were 16.19, 15.63 and 12.85 mg·m −2 ·d −1 , respectively, and the nitrate average diffusion ux were 3.08, 7.73 and -1.71 mg·m −2 ·d −1 . The previous analysis showed that the diffusion ux of NO 3 − N was slightly greater than that of NH 4 + -N, however, the results of average diffusion ux was just the opposite, the average diffusion ux of NO 3 − N was less than that of NH 4 + -N, indicating that the migration direction of NO 3 − N at sediment-water interface varied frequently, the accumulation process counteracts the release process.

Effects of different overlying water ammonia and nitrate concentrations on diffusion ux of nitrogen at sediment-water interface
The diffusion process of nutrients at sediment-water interface is determined by the concentration gradients of particles between pore water and surface water. When the nitrogen concentrations (NH 4 + -N, NO 3 − -N etc.) in the overlying water changes, it will inevitably lead to the changes of nitrogen concentration difference between sediment and overlying water, which will affect both the magnitude of diffusion and direction of migration.
The diffusion ux of TN at sediment-water interface under different overlying water are shown in Figure  4. The diffusion ux of ammonia at sediment-water interface decreased rst and then increased among all treatments, nally achieving the transformation from source to sink (Fig. 5). The calculated results of NH 4 + -N diffusion ux in L, M, H treatment were 3.37, -4.94, -3.84 mg·m −2 ·d −1 , respectively. In L treatment, the concentrations of NO 3 − -N and NH 4 + -N were relatively low in overlying water, NH 4 + -N was released from sediment to surface water. When the concentrations of NO 3 − -N and NH 4 + -N were higher (M and H treatment), it was absorbed by sediment.
The diffusion ux of nitrate at sediment-water interface was transformed from release source to accumulation sink, and the average diffusion ux in L, M and H treatment were 12.30, 10.39 and 7.11 mg·m −2 ·d −1 , and the NO 3 − -N was released from the sediment in all treatments. The results showed that the diffusion ux of nitrate gradually decreased with the increase of NO 3 − -N and NH 4 + -N concentrations in the overlying water, and nally reached equilibrium between release and accumulation (Fig. 6). On the whole, the diffusion ux of nitrate was numerically greater than that of ammonia, which can be explained by the strong absorption of ammonia by sediment. According to the data of DO concentrations (8.98-11.69 mg·L −1 ) in the overlying water, it was found that the aquatic system was in an aerobic state during the incubation experiment. The ammonia was further nitrated, thus the nitrate concentration in the overlying water was signi cantly higher than that of ammonia, resulting in the greater diffusion ux of nitrate.
The diffusion ux of dissolved inorganic nitrogen (DIN) was obtained by summing the uxes of NH 4 + -N, mg·m −2 ·d −1 , respectively. The release rate dominated the process of nitrogen release from sediment to water (Diamond at al., 1990), and the migration of dissolved nitrogen from pore water to sediment and then into surface water was determined by the concentration gradient between the two medias. Since there was no extra NH 4 + -N and NO 3 − -N in the overlying water in L treatment, the higher concentration gradient was the main reason for the signi cant release of DIN from sediment. Therefore, the concentration gradients of DIN between pore water and surface water were affected by controlling the concentration of inorganic nitrogen in the overlying water, resulting in the change of source-sink relationship and TDN diffusion ux at sediment-water interface.

Discussion
Results from this study indicate that concentrations gradients of particles were indeed an important factor affecting the diffusion ux at sediment-water interface, including deposition or release rates from sediments and diffusion direction during our incubation experiment. In our study, we set up two schemes to alter the N concentration gradient. The rst scheme is to change the nitrogen concentrations in the sediment, the other one is to adjust the ammonia and nitrate concentrations in the overlying water. Berelson found that the production of endogenous of N and P nutrients accounts for more than 50% of exogenous nutrients in the study of nutrients biogeochemical cycling in Port Phillip Bay, Australia (Berelson et al.,1998). The nutrients ux from the sediment is mainly caused by the concentration gradients between pore water at sediment surface and overlying water, which directly showed the importance of sediment pore water for nutrients release. Additionally, as a critical sink for nutrients, the quantity of nutrients in the sediment could directly affect the quality of the surrounding environment as well as the survival, composition of benthos to a certain degree. Obviously, the content of nutrients in the sediment is quite important for the supply or supplement of nutrients to overlying water. Sediment is a "holding place" for nutrients such as N and P, the nutrients can be released into overlying water when the concentration gradients of nutrients are positive.
Sediment is a signi cant place for nutrients accumulation and intermittent regeneration. It can be seen from the results in Fig. 1 that the diffusion uxes of TN were positive except at the beginning of the experiment, indicating that sediment is one of the important input sources of nitrogen. Since the diffusion ux of NH 4 + -N and NO 3 − -N is affected by the relative rate of nitri cation and denitri cation, additionally, aerobic-anaerobic environment occurs alternatively and high activity of organic matter at sediment-water interface, resulting in the strong oxidation-reduction reactions including nitri cation and denitri cation.
The generated NH 4 + -N is di cult to be fully absorbed by microorganisms and NH 4 + -N cannot be completely transformed to NO 3 − -N under anaerobic condition, thus the diffusion ux of NH 4 + -N is relatively high (Fig. 2).
The results from Fig. 3 indicated the diffusion ux of NO 3 − N varied greatly, since the whole incubation experiment was completed under still water conditions, which had little disturbance on sediment. In the decomposition process, the organic matter consumes large amount of dissolved oxygen and reducibility of sediment is enhanced, which affects the nitri cation and denitri cation process, thus vary the distribution and diffusion of nitrate. The low DO state inhibits the nitri cation reaction, and nitrate was consumed during denitri cation process, resulting in the lower NO 3 − N concentration in the sediment than that in the overlying water. In this case, sediment act as the sink for NO 3 − N. As the intermediate product of nitri cation and denitri cation, the concentration of NO 2 − -N is low and unstable, thus the signi cance of its diffusion ux is unde ned (Hall et al., 1996).
In the incubation experiment of different overlying water nitrogen concentration, the concentrations of NH 4 + -N and NO 3 − N in the overlying water had little relationship with corresponding diffusion ux at sediment-water interface, and the correlation is quite low. However, this does not mean that the nutrient concentration in the overlying water has little effects on its diffusion ux. It only shows that when the nutrients concentration of sediment is much higher than that of overlying water, the nutrient concentration in the overlying water is not the decisive factor. The reason for this phenomenon may be that the migration of nitrogen at sediment-water interface is not only controlled by concentration gradient, but also related to other factors such as dissolved oxygen content, sediment structure and horizontal migration and diffusion of nutrients.

Conclusions
This study sought to determine whether concentration gradients could affect diffusion ux of ammonia and nitrate at sediment-water interface. Results highlighted that concentrations gradient of nutrients were indeed an important factor affecting the diffusion ux at sediment-water interface. With the increase of TN concentration in the sediment, the diffusion ux increases accordingly. The diffusion ux of NH 4 + -N and NO 3 − -N is affected by the relative rate of nitri cation and denitri cation, therefore, alternative aerobicanaerobic environment at sediment-water interface led to sediment act as the sink or source for nutrients at different cases. In addition, the diffusion of nutrients at sediment-water interface in aquatic ecosystem is not only controlled by concentration gradients, some other factors such as incoming water, hydrodynamics, dissolved oxygen content, sediment structure, biological disturbance, horizontal migration and diffusion of nutrients and turbulent diffusion caused by wind and wave, are equally important.

Declarations
Ethics approval and consent to participate Not applicable

Consent for publication
Not applicable Availability of data and materials The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request. The variation of diffusion ux of TDN at sediment-water interface under different sediment TN concentrations Figure 3 The variation of diffusion ux of ammonia at sediment-water interface under different TN sediment concentrations Figure 4 The variation of diffusion ux of nitrate at sediment-water interface under different TN sediment concentrations Figure 5 The variation of diffusion ux of TN at sediment-water interface in L, M, H treatment.

Figure 6
The variation of diffusion ux of NH 4 + -N at sediment-water interface in L, M, H treatment.

Figure 7
The variation of diffusion ux of NO 3 --N at sediment-water interface in L, M, H treatment.