Environmental Science and Pollution Research

, Volume 26, Issue 3, pp 2898–2907 | Cite as

Dissolved oxygen stratification changes nitrogen speciation and transformation in a stratified lake

  • Xiaoxuan Su
  • Qiang HeEmail author
  • Yufeng Mao
  • Yi Chen
  • Zhi Hu
Research Article


Dissolved oxygen (DO) stratification is a natural phenomenon in lakes, which potentially influences nitrogen (N) biogeochemical cycle. However, the specific effects of DO stratification on N speciation and transformation behaviors in different water layers are still poorly understood. Here, we reported that DO stratification remarkably influenced N species and transformation pathways in different water columns by high frequency sampling during summers in Longjing Lake, China. Results showed that DO stratification in the lake created three water layers: epilimnion (1–3 m), oxycline (4–11 m), and hypolimnion (12–20 m). In the epilimnion, N speciation was mainly controlled by phytoplankton assimilation and organic N dominated in this layer. Oxycline was the major place for N transformations and had the most notable N removal capacity (714 kg N from June to August). In the hypolimnion, \( {\mathrm{NH}}_4^{+} \) was the major N species, and sediment release contributed nearly 85% hypolimnetic \( {\mathrm{NH}}_4^{+} \). Furthermore, approximately 8 kg of dissolved N2O was also accumulated in the hypolimnion, contributing ~ 70% of N2O in the whole lake. Overall, our results indicated that DO stratification caused the shifts in N speciation and transformation behaviors among different water columns, which may have a great implication for lake managements for providing separated protection strategies from different water depths.


Longjing Lake Dissolved oxygen stratification Thermal stratification Nitrogen speciation Nutrient 


Lakes are important ecological system and resource for drinking water, agricultural irrigation, and industrial use (Torres et al. 2016). Dissolved oxygen (DO) stratification in lakes is a natural physical feature and creates lager DO gradients in lakes, which affects its ecosystem structure and function (Zhang et al. 2015). Previous studies have indicated that DO concentrations could control the release of phosphorus and nitrogen from lake sediments (Gantzer et al. 2009; Small et al. 2014), alter microbial community structures of lakes in different water columns (Yu et al. 2014), and impact fish physiological features and distribution patterns in benthic water (Nurnberg 2004). Hence, the limnologist Hutchinson (1957) stated that the nature of lakes could be better learnt by determining oxygen concentration than other chemical data. DO stratification can create three water layers: epilimnion, oxycline, and hypolimnion. The epilimnion is known as the surface mixed layer and is also saturated with oxygen (Woolway et al. 2014). The oxycline forms a sharp oxygen gradient providing aerobic and anoxic conditions for biochemical reactions (Kalff 2002). The hypolimnion is filled with the cool and dense water, and is often characterized by anoxic conditions, especially in eutrophic lakes. These characteristics suggest that DO stratification could provide various physical and chemical environments at different depths and thus drive elements biogeochemical cycle in lakes.

Nitrogen (N), a common element, is a major driver for eutrophication. Sources of N input to lakes include point source pollution, non-point source pollution (Li and Davis 2014; Taylor et al. 2005), atmospheric deposition (Miyazako et al. 2015), and endogenous pollution (Small et al. 2014). N exists in various oxidized and reduced forms that allow it to act as an electron donor or acceptor in oxidation-reduction reactions (Kalff 2002). N presents several chemical forms in lakes, including particulate organic N (PON), dissolved organic N (DON), ammonium (\( {\mathrm{NH}}_4^{+} \)), nitrate (\( {\mathrm{NO}}_3^{-} \)), nitrite (\( {\mathrm{NO}}_2^{-} \)), dissolved molecular N2, and nitrous oxide (N2O). Different N species play the different roles in bioavailability and N cycle in lakes (Elser et al. 2007; Paerl et al. 2011). Water columns with high DO concentrations often have the oxidized N forms, whereas the reduced N forms mainly exist in water columns with low DO concentrations. Because not all N species have the equal bioactivity (higher for inorganic N, and lower for organic N), knowledge of distribution behaviors of all N species in whole lakes is necessary to establish inland lake management and guide water policy in the future. However, how N speciation distributions in water different layers respond to DO stratification is not fully understood.

Unlike other nutrients elements (phosphorus, sulfur, silicon), N transformations in waters are mainly controlled by microbial-mediated processes (such as nitrification, denitrification, Anammox, dissimilatory \( {\mathrm{NO}}_3^{-} \) reduction to \( {\mathrm{NH}}_4^{+} \), and mineralization) (Kalff 2002), which are closely related to DO concentration. In a stratified lake, DO stratification creates various DO concentrations in waters, probably impacting N transformation behaviors in different water columns. However, the impacts of DO stratification on N transformations in the epilimnion, oxycline, and hypolimnion have rarely been reported.

With these considerations, Longjing Lake, an inland water in Chongqing, China, was selected to explore the influences of DO stratification on N species. Few non-point source pollutants enter the water body, and also no point source pollutions discharge. High frequency water samples were taken from Longjing Lake during summer in 2015 and 2016 due to stable stratification and hypoxic conditions in the hypolimnion. Samples for physicochemical variables (water temperature, DO, chlorophyll-a, pH) were measured in-situ at 0.2-m interval. Samples for N species were taken at 1-m interval and analyzed in lab. The specific aims of the study were to (1) investigate the formation of DO and thermal stratification during stable stratification, (2) reveal the effects of DO stratification on N speciation and transformation; and (3) assess N transformation pathways in the epilimnion, oxycline, and hypolimnion.

Material and methods

Site description

Longjing Lake (29° 41′ N, 106° 32′ E), an inland lake, is located in a temperate zone of Chongqing, a mountain city in southwest China (Fig. 1). Summers in this region are characterized by relatively low wind speed and high air temperature. The lake is a typically warm monomictic lake exhibiting thermal stratification throughout most of the whole year (March to October, with particularly stable stratification in June, July, and August), except for a short complete mixing period in winter (November to February). Longjing Lake has a water area of 0.53 km2 with a mean and maximum depth of 11 m and 20 m, respectively.
Fig. 1

Location of the sampling site in Longjing Lake during 2015–2016

Monitoring and sampling methodology

Biweekly field investigations occurred at two sampling sites (Fig. 1) from July 2015 to September 2016. However, sampling frequency was increased to every 6–7 days during July and August. In this study, the data from 16 sampling campaigns (7 for 2015, and 9 for 2016) in 5 months (July to August 2015 and June to August 2016) were mainly used because stable stratification occurred in this period. Global positioning system (GPSmap629sc, Garmin, USA) was applied to position the sampling sites with an accuracy of 3–5 m.

Sampling campaigns for recording vertical profiles of water temperature, DO, pH, and chlorophyll-a (Chl-a) were conducted in situ using a multiparameter water quality profiler (Hydrolab DS5, HACH, USA) since October 2015. Data were collected from surface (0 m) to the maximum depth (0.5–0.7 m above sediments) at 0.2-m interval. We made labels every 0.2 m on the rope connected with the multiparameter profiler to ascertain the accurate sampling depth. Samples for N species and dissolved organic carbon (DOC) analysis were taken at 1-m increments at the sampling site from 1 m below surface to the maximum depth using a 2.5-L sampler. For N2O, samples were collected on July 12 and 29, and August 9 and 26, 2016 by filling 50-mL glass serum bottles from the sampler without air bubbles, and then preserved with 1 mL mercuric chloride immediately before sealing to inhibit microbial activity. After the investigations of thermocline and oxycline, water was collected in pre-processed 2-L high-density polyethylene bottles. All bottles were stored at 4 °C before testing and all tests were completed within 12 h.

Analytical procedures

Detection criterions of thermocline and oxycline are provided in the Supporting information (SI). Water is considered to be anoxic if DO concentrations are less than 1 mg L−1 (Perron et al. 2014). Water quality parameters, such as total N (TN), NO\( {\mathrm{NO}}_3^{-} \), \( {\mathrm{NO}}_2^{-} \), and \( {\mathrm{NH}}_4^{+} \), were analyzed in the laboratory using standard methods (APHA 1998). DOC was measured by a LiquiTOC II analyzer (Elementar, Germany) according to Li et al. (2008). Although particulate nitrogen (PN) comprises PON and adsorbed \( {\mathrm{NH}}_4^{+} \)and DON (Wetzel 2001), the concentration of \( {\mathrm{NH}}_4^{+} \) is below detectability, and DON readily dissolves in water. Therefore, PON was used to represent PN in this study. In addition, PON (PON = TN-dissolved total N (DN)) and DON (DON = DN-\( {\mathrm{NO}}_3^{-} \)-\( {\mathrm{NO}}_2^{-} \)-\( {\mathrm{NH}}_4^{+} \)) were calculated from direct N measurements (Li and Davis 2014; Taylor et al. 2005). DN samples were first filtered through a 0.45-μm membrane filter (same as for dissolved inorganic N (DIN)), then analyzed using methods identical to TN analysis. If the sum of \( {\mathrm{NO}}_3^{-} \), \( {\mathrm{NO}}_2^{-} \), and \( {\mathrm{NH}}_4^{+} \) exceeded the DN value, the DON concentration was recorded as zero. Dissolved N2O concentrations were determined by gas chromatography (GC-ECD, Shimadzu, Japan), using the headspace equilibrium technique (Walker et al. 2010; Wang et al. 2009). Additional details for handling and calculating N2O are shown in SI.

N species transformation estimation

Longjing Lake has two inlet rivers: Zhaojia Xi and Longjing Gou and has only one outlet. The corresponding inflow and outflow volumes during the summer, and the N concentrations were provided in SI (Table S1 and S2). In Longjing Lake, N loads from non-point source pollutions contributed less than 5% of N in the lake. Endogenesis pollutions provided the significant source of N in the lake. Based on our previous results (data not shown), the horizontal influence on N variations could be ignored.

During the stable stratification, water columns were divided into three layers: epilimnion (1–3 m), oxycline (4–11 m), and hypolimnion (12–20 m), based on the results of DO concentrations. Hypolimnion is a highly stable and consistent environment during summer stratification. It is not significantly influenced by rainfall and runoff. Epilimnion and oxycline, however, are readily disturbed by hydrological and meteorological conditions. Therefore, it is difficult to estimate N species transformation in these water layers. In this study, only N transformations in the hypolimnion were assessed, using the following equation:
$$ \Delta {\mathrm{N}}_{\mathrm{i}}={\mathrm{C}}_{\mathrm{i}\mathrm{ni}-\mathrm{i}}\times {V}_{\mathrm{h}}-{\mathrm{C}}_{\mathrm{end}-\mathrm{i}}\times {V}_{\mathrm{h}} $$
where i denotes N forms (DON, N2O, \( {\mathrm{NO}}_2^{-} \), and \( {\mathrm{NO}}_3^{-} \)); Ni (mg) denotes the mass variations of N forms i during summer (from the start of July to the end of August). Cini-i and Cend-i (mg L−1) are the measured initial and final concentrations of N forms i, respectively. Vh (assumed to be 70 × 104 m3) represents the volume of the hypolimnion. In addition, Ve and Vo are assumed to be 10 × 104 and 80 × 104 m3 for the epilimnion and oxycline, respectively.
To assess N transformations in the hypolimnion, half-reactions involved in microbial nitrification and denitrification were used in this study (Table S3). They were combined according to the following mass balance equation based on thermodynamic electron equivalents model during bacterial metabolism (Rittmann and McCarty 2001; Sawyer et al. 1994):
$$ R={f}_{\mathrm{s}}{R}_{\mathrm{c}}+{f}_{\mathrm{e}}{R}_{\mathrm{a}}-{R}_{\mathrm{d}} $$
Rc is the half reaction for synthesis of bacterial cells, Ra, Rd represents the half reaction for electron acceptor and donor, respectively. The values fs and fe are the portion of the electron donors used for synthesis and energy, respectively, and fs + fe = 1. In this study, fs is 0.60 for denitrification (Rittmann and McCarty 2001). Species of organic substances in lakes are difficulty to ascertain; therefore, CH2O is used as a proxy to balance reactions (Yan et al. 2016). Half reactions involved in N2 production are not mentioned. Additional details are presented in Table S3. Reactions for \( {\mathrm{NO}}_3^{-} \) reduction to \( {\mathrm{NO}}_2^{-} \), and then to N2O were listed below (adapted from Yan et al. (2016)):
$$ 0.25\ \mathrm{C}{\mathrm{H}}_2\mathrm{O}+0.221\ \mathrm{N}{\mathrm{O}}_3^{-}\to $$
$$ 0.2\ \mathrm{N}{\mathrm{O}}_2^{-}+0.021\ {\mathrm{C}}_5{\mathrm{H}}_7{\mathrm{O}}_2\mathrm{N}+0.125\ \mathrm{HC}{\mathrm{O}}_3^{-}+0.018\ \mathrm{C}{\mathrm{O}}_2+0.104\ {\mathrm{H}}^{+}+0.061\ {\mathrm{H}}_2\mathrm{O} $$
$$ 0.25\ {\mathrm{C}\mathrm{H}}_2\mathrm{O}+0.223\ {\mathrm{N}\mathrm{O}}_2^{-}+0.098\ {\mathrm{H}}^{+}\to 0.023\ {\mathrm{C}}_5{\mathrm{H}}_7{\mathrm{O}}_2\mathrm{N}+0.1\ {\mathrm{N}}_2\mathrm{O}+0.125\ {\mathrm{H}\mathrm{CO}}_3^{-}+0.010\ {\mathrm{C}\mathrm{O}}_2+0.156\ {\mathrm{H}}_2\mathrm{O} $$
Additionally, we also identified the possible pathways of \( {\mathrm{NH}}_4^{+} \) production and \( {\mathrm{NO}}_3^{-} \) consumption in the hypolimnion:
$$ {\mathrm{C}}_{\mathrm{N}{\mathrm{H}}_{4\ \mathrm{end}}^{+}}\times {V}_h={\mathrm{C}}_{\mathrm{N}{\mathrm{H}}_{4\ \mathrm{ini}}^{+}}\times {V}_h+{\mathrm{N}}_{\mathrm{N}{\mathrm{O}}_3^{-}\to \mathrm{N}{\mathrm{H}}_4^{+}}+{\mathrm{N}}_{\mathrm{DON}\to \mathrm{N}{\mathrm{H}}_4^{+}}+{\mathrm{N}}_{\mathrm{sediments}} $$
$$ \Delta {\mathrm{N}}_{\mathrm{N}{\mathrm{O}}_3^{-}}=\mathrm{N}{\mathrm{O}}_{3\ \mathrm{DNRA}}^{-}+\mathrm{N}{\mathrm{O}}_{3\ \mathrm{denitrification}}^{-} $$
where \( {\mathrm{C}}_{\mathrm{N}{\mathrm{H}}_{4\ \mathrm{end}}^{+}} \) and \( {\mathrm{C}}_{\mathrm{N}{\mathrm{H}}_{4\ \mathrm{ini}}^{+}} \) (mg L−1) are the measured initial and final concentrations of NH4+ during stable stratification, respectively.\( {\mathrm{N}}_{\mathrm{N}{\mathrm{O}}_3^{-}\to \mathrm{N}{\mathrm{H}}_4^{+}} \) (mg), representing dissimilatory NO3 reduction to NH4+ (DNRA), is assessed using NO3~1.3NH4+ (Lam et al. 2009). \( {\mathrm{N}}_{\mathrm{DON}\to \mathrm{N}{\mathrm{H}}_4^{+}} \)(mg) denotes DON mineralization to \( {\mathrm{NH}}_4^{+} \) and was assessed using (CH2O)106(NH3)16H3PO4~16\( {\mathrm{NH}}_4^{+} \) (Lam et al. 2009). N sediments is the release of \( {\mathrm{NH}}_4^{+} \) from sediments during summer. Detailed equations for DNRA and DON mineralization are provided in SI. \( \Delta {\mathrm{N}}_{\mathrm{N}{\mathrm{O}}_3^{-}} \) (mg) represents the transformation of \( {\mathrm{NO}}_3^{-} \) in the hypolimnion and is calculated by Eq. (1). \( \mathrm{N}{\mathrm{O}}_{3\ \mathrm{denitrification}}^{-} \) (mg) refers to \( {\mathrm{NO}}_3^{-} \) consumed by denitrification and is assessed by reactions (3) and (4). \( \mathrm{N}{\mathrm{O}}_{3\ \mathrm{DNRA}}^{-} \) (mg) refers to \( {\mathrm{NO}}_3^{-} \) consumed through DNRA and is calculated by the differences between \( \Delta {\mathrm{N}}_{\mathrm{N}{\mathrm{O}}_3^{-}} \) and \( \mathrm{N}{\mathrm{O}}_{3\ \mathrm{denitrification}}^{-} \).

Data analysis

Statistical analysis was performed using the Statistical Program for Social Sciences (SPSS 19.0), including the calculation of mean values and standard deviations, linear fitting, and t tests. Network analysis for Pearson correlations between N species and physicochemical variables were determined by Gephi (Bastian et al. 2009). Significance levels (p < 0.05) were reported.


Temperature, DO, and Chl-a profiles

Over summer periods, the lake was stably stratified (Fig. 2 and S1). Surface water temperature increased gradually, whereas variations of water temperature in the hypolimnion were limited (Fig. 2a). Average temperature difference between surface and bottom was 20 °C. A stable thermocline was formed between 1.2 and 19.0 m. Furthermore, the epilimnion was statured with DO (7.20 mg L−1), and the hypolimnion was characterized by anoxic condition with DO concentrations below 0.1 mg L−1 (Fig. 2b). The oxycline existed between 3.6 and 10.8 m. In this layer, DO concentrations varied from 0.17 to 3.90 mg L−1. As for Chl-a, the concentrations were declined from surface (33.90 μg L−1) to bottom (1.81 μg L−1) (Fig. 2c). Clearly, our results suggested that lake stratification could provide various physical and chemical environments at different depths, which could drive the accumulations of organisms and phytoplankton as well as the N-related biochemical reactions.
Fig. 2

Temporal evolution of the water columns for a temperature, b DO, and c Chl-a during summer in 2016. Based on our results, the lake was divided into three layers in this study: epilimnion (1–3 m), oxycline (4–11 m), and hypolimnion (12–20 m)

N species and variations in the epilimnion, oxycline, and hypolimnion

From depth-scale, concentrations of N species in different water columns during summer were given in Fig. 3 and Table S4. In detail, \( {\mathrm{NH}}_4^{+} \), \( {\mathrm{NO}}_2^{-} \), and \( {\mathrm{NO}}_2^{-} \) concentrations increased with the depth and reached the highest in the hypolimnion. Concentrations of \( {\mathrm{NO}}_3^{-} \) first increased to the maximum 0.52 mg L−1 in the oxycline, and then declined gradually. In contrast to other N species, PON and DON followed the same pattern as Chl-a: reaching the highest in the epilimnion, and dropping gradually with depth. Additionally, major N speciation in different water columns also exhibited the significant differences (Fig. 4). In the epilimnion, organic N (PON and DON) was the major N species, accounting for 74% of TN. Among oxycline, NO3 was most noticeable in N species. In the hypolimnion, NH4+ (59% of TN) was the primary species.
Fig. 3

Heat maps of N species among water columns during summer in 2015 and 2016. Database covered the 16 sampling campaigns. Sampling dates between 2015 and 2016 were separated by the white line. Dashed line box refers to oxycline (4–11 m). (For interpretation of the references to color in this figure legend, the reader is referred to the online version of this article)

Fig. 4

N species concentrations in the epilimnion, oxycline, and hypolimnion

From time-scale, PON and \( {\mathrm{NO}}_3^{-} \) in the epilimnion followed the similar trends with TN and DN: declining with time (Fig. 3). In the oxycline, the significant declines of \( {\mathrm{NO}}_3^{-} \) and \( {\mathrm{NO}}_2^{-} \) during summer period were observed. In the hypolimnion, the obvious increase of \( {\mathrm{NH}}_4^{+} \) and the decrease of \( {\mathrm{NO}}_3^{-} \) as well as N2O were recorded. Moreover, \( {\mathrm{NH}}_4^{+} \) accumulation rates and \( {\mathrm{NO}}_3^{-} \) removal rates in epilimnion, oxycline, and hypolimnion during summer were calculated in this study, and shown in Table 1. The maximum \( {\mathrm{NH}}_4^{+} \) accumulation rate occurred in the hypolimnion with 0.011 mg L−1 day−1. No significant variations of \( {\mathrm{NH}}_4^{+} \) were observed in the epilimnion. \( {\mathrm{NO}}_3^{-} \) removal rates in the oxycline and hypolimnion (0.016 and 0.020 mg L−1 day−1) were higher than that in the epilimnion (0.010 mg L−1 day−1). These results indicated that the hypolimnion of a stratified lake could act as the important source of \( {\mathrm{NH}}_4^{+} \) and sink of \( {\mathrm{NO}}_3^{-} \).
Table 1

NO– 3 removal and NH+ 4accumulation rates in Longjing Lake during the summer of 2015–2016





NO3 removal rate (mg L−1 day−1)




NH4+ accumulation rate (mg L−1 day−1)

~ 0

− 0.003


Increase and decrease of N in the epilimnion, oxycline, and hypolimnion

Table 2 showed the mass variations of all N species in different water layers during summer. The increases or decreases of N mass exhibited considerable differences. In detail, the epilimnetic DON increased by 30 kg during summer, whereas PON were declined by 6 kg. In the oxycline, large amounts of TN and DN were removed by 714 and 754 kg, respectively (Table 2). Also, approximately 80 kg \( {\mathrm{NO}}_3^{-} \) were reduced, and a slight increase (40 kg) of \( {\mathrm{NH}}_4^{+} \) were observed in this layer, suggesting that oxycline might be the major place for N transformations. Although hypolimnetic \( {\mathrm{NH}}_4^{+} \) increased by 390 kg, the variations of TN and DN followed the same pattern with \( {\mathrm{NO}}_3^{-} \) removal: declining with time.
Table 2

Mass variations of all N species in epilimnion, oxycline, and hypolimnion during summer

N species





− 36

− 714

− 163


− 30

− 754

− 100


− 62

− 802

− 518



− 16

− 35






− 6


− 63








− 2

Unit of variations in N species is presented as kg. “−” and “+” denotes the decrease and increase of N species mass during summer, respectively. Epilimnion is 1–3 m; oxycline is 4–11 m; hypolimnion is 12–20 m

Moreover, potential N transformation in the hypolimnion of Longjing Lake was assessed and the pathways from epilimnion, oxycline, and hypolimnion were depicted in Fig. 5. In the epilimnion, inorganic N uptake by phytoplankton assimilation (stored as organic N) was the major pathway for N transformation. Among the oxycline, nitrification, denitrification, and DNRA might take place. In the hypolimnion, N release from sediments, organic mineralization, denitrification, and DNRA could occur simultaneously. \( {\mathrm{NO}}_3^{-} \) removal through denitrification contributed approximately 70%, whereas DNRA was 30%. NH4+ from sediment release contributed the most part of \( {\mathrm{NH}}_4^{+} \) contents (85%) in the hypolimnion.
Fig. 5

N cycling in the epilimnion, oxycline, and hypolimnion. N transformations in the hypolimnion were assessed. Data in the figure are from July and August, 2016, and it is assumed that Ve, Vo, and Vh are 0.1, 0.8, and 0.7 million m3, respectively. Data near N forms represent the initial and remaining mass of N in the water at the beginning and end of summer in 2016. The wide of lines denote the strength of main function

Relationships between environment variables and N species

Network analysis was conducted in this study to reveal the correlations between environment variables and N species in different water columns (Fig.6). In the epilimnion, Chl-a had the significant and positive relationship with PON (Fig.6a), indicating that phytoplankton contributed a large part of PON. No significant correlations were found between DO and N species in the epilimnion. As for oxycline, DO was negatively correlated with \( {\mathrm{NH}}_4^{+} \) and positively with Chl-a (Fig.6b). Additionally, we also found the positive correlation between TN and \( {\mathrm{NO}}_3^{-} \). In the hypolimnion, DO had the significant and negative relationships with \( {\mathrm{NH}}_4^{+} \), DON, and N2O (Fig. 6c). Different with oxycline, TN was found to have the positive relationship with \( {\mathrm{NH}}_4^{+} \).
Fig. 6

Network analysis of correlations between temperature, DO, pH, Chl-a, and N species in different water layers (a epilimnion, b oxycline, and c hypolimnion). Green nodes denote physicochemical parameters measured in the sampling sites, and red nodes denote N composition analyzed in the lab. Connection lines denote strong significant (p < 0.01, Pearson’s r > 0.65) correlations (positive (blue) and negative (yellow)). Sizes of each node stand for the number of connections. The width of lines denotes the significance (thick line: p < 0.01; thin line: p < 0.05)


Thermal and DO stratification in Longjing Lake

It is generally accepted that oxycline is closely related to thermocline in deep and stratified lakes. However, Fig. 2 showed that oxycline in this study (3.6–10.8 m) was not completely linked to thermocline (1.2–19.0 m), which is not in line with previous studies (Foley et al. 2012; Perron et al. 2014). Zhang et al. (2015) found that the DO stratification followed the similar trend with thermal stratification in Qiandaohu Lake, China. The trophic state of lakes may be responsible for the different observations. In oligotrophic lakes, DO distribution is mainly controlled by water temperature (Kalff 2002), and thus oxyline is well-matched with thermocline. In eutrophic lakes, however, the vertical distribution of DO is not only governed by temperature but by other factors (such as phytoplankton, and organic matter). For example, DO productions could be increased by phytoplankton photosynthesis in the eutrophic lakes (Pearce et al. 2017; Zhang et al. 2015). Oxygen consumptions via organic matter decomposition also affect the DO distributions in the eutrophic lakes. Therefore, these results have indicated that DO stratification in the eutrophic lakes is a multifactor-controlled process and potentially provides obvious DO gradients, which could drive N transformations.

N species in different water layers

Epilimnion: phytoplankton assimilation is a major process controlling N composition

The epilimnetic TN in Longjing Lake was 1.41 ± 0.28 mg L−1, higher than the mean TN concentration in global lakes (0.611 ± 2.5 mg L−1) (Chen et al. 2015). In Longjing Lake, the epilimnion during the summer was saturated with oxygen and had a warm water temperature (Fig. 2). which provides the ideal conditions for phytoplankton growth. A strong and positive relationship between phytoplankton biomass (measured as Chl-a) and PON indicated that phytoplankton proliferation contributed to large amounts of organic N in the epilimnion (Fig. 6a). Our results also showed that PON and DON comprised over 70% of TN pool in the epilimnion (Fig. 3). Additionally, it has been estimated that 0.1 mg 6a). Our results also showed that PON and DON comprised over 70% of TN pool in the epilimnion (Fig. 3). Additionally, it has been estimated that 0.1 mg \( {\mathrm{NH}}_4^{+} \) could produce ~ 7 mg of algae cells (Tuantet et al. 2014). In this eutrophic epilimnion, large amounts of inorganic N (especially \( {\mathrm{NH}}_4^{+} \) and \( {\mathrm{NO}}_3^{-} \)) can be assimilated by phytoplankton and are excreted extracellularly as large amounts of simple and complex PON and DON (Meffert and Zimmermann-Telschow 1979), which also contribute to organic N (PON and DON) productions. These results thus explained the higher TN concentration in the epilimnion.

In the epilimnion of Longjing Lake, inorganic N is uptake by phytoplankton and is stored inside as organic N, implying that phytoplankton assimilation is an important process controlling the composition and concentrations of N in the epilimnion. In our study, there was no significant correlation between epilimnetic DO concentration and N species (Fig. 6a). Such response suggested that DO might not be the important factor in governing N speciation and concentrations in the epilimnion.

Oxycline: oxycline is a major place for N transformations

In the oxycline (4–11 m), total N removal was 714 kg over summer period, higher than 36 kg in the epilimnion and 163 kg in the hypolimnion. This indicated that oxycline had the most notable capacity to remove N loads. Generally, oxycline could create large gradients of DO, thus offering aerobic and anoxic conditions for N transformations (Woolway et al. 2014). In detail, \( {\mathrm{NO}}_3^{-} \) increased with depth due to nitrification in this layer, but decreased with time as a result of denitrification and DNRA (Brezonikl and Lee 1968; Downes 1988; Kalff 2002). Since \( {\mathrm{NO}}_3^{-} \) was major N speciation in the oxycline, NO3 reduction was responsible for TN removal from the layer. In this study, removal of \( {\mathrm{NO}}_3^{-} \) from oxycline exceeded 800 kg (Table 2), which could explain the 714 kg of TN removal as mentioned above. Furthermore, we also calculated the \( {\mathrm{NO}}_3^{-} \) removal rates in different water columns during summer (Table 1). \( {\mathrm{NO}}_3^{-} \) removal rate in the oxycline was 0.020 mg L−1 day−1, which was a slightly higher than 0.012–0.018 mg L−1 day−1 conducted in the oxycline of Mendota Lake (Brezonikl and Lee 1968).

For \( {\mathrm{NH}}_4^{+} \), it increased gradually with depth among the oxycline (Fig.4). Such variation is probably because phytoplankton biomass decrease with depth caused by light attenuation and temperature decline restrains their uptake for \( {\mathrm{NH}}_4^{+} \). The negative correlations between \( {\mathrm{NH}}_4^{+} \) and PON or Chl-a further supported this observation (Fig. 6b). Another reason for \( {\mathrm{NH}}_4^{+} \) increase is DNRA process by reducing \( {\mathrm{NO}}_3^{-} \) to produce \( {\mathrm{NH}}_4^{+} \). Previous studies have shown that DNRA could be an important pathway to generate \( {\mathrm{NH}}_4^{+} \) (Lam et al. 2009). N2O is a potent greenhouse gas and can be produced in lakes, reservoirs, rivers, and oceans (Brezonikl and Lee 1968; Walker et al. 2010). In this study, dissolved N2O in the oxycline was enhanced by 1.7-fold compared to the epilimnion. Coupled nitrification-denitrification could contribute N2O in this layer (Gao et al. 2014; Wetzel 2001). Average concentration of dissolved N2O was 4.48 μg L−1 (0.51% mol N2O/mol \( {\mathrm{NH}}_4^{+} \)), which was higher than the 0.04 to 0.42% reported by other studies (Wang et al. 2009). This difference existed because Wang et al. (2009) estimated N2O emission from nitrification process without considering the contribution of microbial denitrification. Also, the upwelling of N2O produced in the hypolimnion also improved N2O in the oxycline, despite the low concentrations (Beaulieu et al. 2014). Because the aerobic and anoxic conditions co-existed in the oxycline, the layer is a major place for N transformations.

Hypolimnion: \( {\mathrm{NH}}_4^{+} \) is the major N species and most derive from sediment release

In contrast with oxycline, \( {\mathrm{NH}}_4^{+} \) dominated in all kinds of N species in the hypolimnion and was increased by 390 kg over summer period (Table 2). \( {\mathrm{NH}}_4^{+} \) accumulation rate in the hypolimnion was 0.011 mg L−1 day−1, lower than 0.013, and 0.027 mg L−1 day−1 observed in Shibianyu Reservoir (a stratified eutrophic reservoir) (Li et al. 2015). Generally, hypolimnetic \( {\mathrm{NH}}_4^{+} \) pool is commonly produced by three ways: organic N mineralization, DNRA, and sediment release (Lam et al. 2009; Small et al. 2014). Organic N mineralization (DON to \( {\mathrm{NH}}_4^{+} \)) is the potential process to contribute to hypolimnetic \( {\mathrm{NH}}_4^{+} \). In this study, however, a total of 63 kg DON were reduced during the summer period, and less than 7 kg of \( {\mathrm{NH}}_4^{+} \) could be produced if all 63 kg of DON are fully mineralized (details in SI). This suggested that DON mineralization could not be the major pathway in contributing \( {\mathrm{NH}}_4^{+} \) in the hypolimnion. In contrast, DNRA process might be an important source of hypolimnetic \( {\mathrm{NH}}_4^{+} \). Based on Eqs. (5) and (6) (details in SI), the amounts of \( {\mathrm{NH}}_4^{+} \) from DNRA during summer were assessed to be ~ 50 kg, which is higher than the contribution of DON mineralization. In the hypolimnion, sediment release is another important way to produce \( {\mathrm{NH}}_4^{+} \). It is generally accepted that release of \( {\mathrm{NH}}_4^{+} \) from sediments is closely linked to the available DO (Small et al. 2014). This was also evidenced by the significant negative correlations between DO and NH4+ in the hypolimnion (Fig. 6c). In our study, a release of ~ 330 kg \( {\mathrm{NH}}_4^{+} \) from the sediments during summer was assessed based on Eq. (5). Thus, these results showed that release of \( {\mathrm{NH}}_4^{+} \) from the sediments seems to be the primary pathway in contributing the hypolimnetic \( {\mathrm{NH}}_4^{+} \).

In the hypolimnion, a total of 518 kg \( {\mathrm{NO}}_3^{-} \) were removed during the summer, slightly lower than 802 kg in the oxycline. Such reduction indicated that the hypolimnion of stratified lake could be regarded as an important sink of \( {\mathrm{NO}}_3^{-} \), which is similar with the role of oxycline. \( {\mathrm{NO}}_3^{-} \) removal in this layer is mainly through denitrification and DNRA processes. Based on the half-reactions (Eqs. 3 and 4), \( {\mathrm{NO}}_3^{-} \) removal through denitrification pathway might contribute over 70%, and DNRA process holds nearly 30%, a similar proportion from earlier results of De Brabandere et al. (2015). Potent greenhouse gas N2O is an intermediate product in \( {\mathrm{NO}}_3^{-} \) removal by denitrification. Previous study showed that the maximal N2O yield occurred at DO concentrations of ~ 0.14 mg L−1 (Yoh et al. 1983). In the hypolimnion, DO maintained at 0–0.1 mg L−1 during the whole summer period, providing a better condition for N2O production. N2O concentration in the layer was 4.8-fold higher than that in the epilimnion, contributing ~ 70% of N2O in the whole lake. Overall, such the shifts in N of the stratified lake reveal the dynamic changes and active transformation in N species among different water columns.


Stable stratification of lakes inhibits the well-mixing of water columns and restrains oxygen transferred from surface to bottom. This accelerates inorganic N releasing from sediments, which potentially promotes a hypereutrophic hypolimnion. In this study, DO stratification led to large amount of \( {\mathrm{NH}}_4^{+} \) N2O, and DON accumulated in the hypolimnion. When autumn overturn occurs, these high concentrations of nutrients would be conveyed to surface, and the greenhouse gas N2O would be eventually emitted to atmosphere (Fernandez et al. 2014), potentially contributing the water eutrophication and global warming.

Lake is the important water resource for agricultural irrigation. Our results have indicated that lake water used for agricultural irrigation could be taken from the hypolimnion, because it contains high concentrations of \( {\mathrm{NH}}_4^{+} \), which is benefit for plant growth. For fish-farming, the waters in the epilimnion are preferred. Overall, our study has a great implication for managing and utilizing water resources from different water layers in the future.


Funding information

This study was supported by the National Natural Science Foundation of China (NO.51278508) and the National Water Pollution Control and Treatment Science and Technology Major Project (2012ZX07307-001).

Supplementary material

11356_2018_3716_MOESM1_ESM.docx (147 kb)
ESM 1 (DOCX 147 kb)


  1. APHA (1998) Standard methods for the examination of water and wastewater, twenty-first ed. In: American public health association. USA, Washington, DCGoogle Scholar
  2. Bastian M, Heymann S, Jacomy M (2009) Gephi: an open source software for exploring and manipulating networks. ICWSM 8:361–362Google Scholar
  3. Beaulieu JJ, Smolenski RL, Nietch CT, Townsend-Small A, Elovitz MS, Schubauer-Berigan JP (2014) Denitrification alternates between a source and sink of nitrous oxide in the hypolimnion of a thermally stratified reservoir. Limnol Oceanogr 59(2):495–506CrossRefGoogle Scholar
  4. Brezonikl PL, Lee GF (1968) Denitrification as a nitrogen sink in Lake Mendota. Wis Environ Sci Technol 2(2):120–125CrossRefGoogle Scholar
  5. Chen M, Zeng GM, Zhang JC, Xu P, Chen AW, Lu LH (2015) Global landscape of total organic carbon, nitrogen and phosphorus in lake water. Sci Rep 5:1–7Google Scholar
  6. De Brabandere L, Bonaglia S, Kononets MY, Viktorsson L, Stigebrandt A, Thamdrup B, Hall POJ (2015) Oxygenation of an anoxic fjord basin strongly stimulates benthic denitrification and DNRA. Biogeochemistry 126(1–2):131–152CrossRefGoogle Scholar
  7. Downes MT (1988) Aquatic nitrogen transformations at low oxygen concentrations. Appl Environ Microbiol 54(1):172–175Google Scholar
  8. Elser JJ, Bracken MES, Cleland EE, Gruner DS, Harpole WS, Hillebrand H, Ngai JT, Seabloom EW, Shurin JB, Smith JE (2007) Global analysis of nitrogen and phosphorus limitation of primary producers in freshwater, marine and terrestrial ecosystems. Ecol Lett 10(12):1135–1142CrossRefGoogle Scholar
  9. Fernandez JE, Peeters F, Hofmann H (2014) Importance of the autumn overturn and anoxic conditions in the hypolimnion for the annual methane emissions from a temperate lake. Environ Sci Technol 48(13):7297–7304CrossRefGoogle Scholar
  10. Foley B, Jones ID, Maberly SC, Rippey B (2012) Long-term changes in oxygen depletion in a small temperate lake: effects of climate change and eutrophication. Freshw Biol 57(2):278–289CrossRefGoogle Scholar
  11. Gantzer PA, Bryant LD, Little JC (2009) Effect of hypolimnetic oxygenation on oxygen depletion rates in two water-supply reservoirs. Water Res 43(6):1700–1710CrossRefGoogle Scholar
  12. Gao Y, Yi N, Wang Y, Ma T, Zhou Q, Zhang ZH, Yan SH (2014) Effect of Eichhornia crassipes on production of N2 by denitrification in eutrophic water. Ecol Eng 68:14–24CrossRefGoogle Scholar
  13. Hutchinson G (1957) A treatise on limnology. Geogr Phys Chem 1:1015Google Scholar
  14. Kalff J (2002) Limnology: inland water ecosystems. Prentice Hall, New YorkGoogle Scholar
  15. Lam P, Lavik G, Jensen MM, van de Vossenberg J, Schmid M, Woebken D, Dimitri G, Amann R, Jetten MSM, Kuypers MMM (2009) Revising the nitrogen cycle in the Peruvian oxygen minimum zone. Proc Natl Acad Sci U S A 106(12):4752–4757CrossRefGoogle Scholar
  16. Li LQ, Davis AP (2014) Urban stormwater runoff nitrogen composition and fate in bioretention systems. Environ Sci Technol 48(6):3403–3410CrossRefGoogle Scholar
  17. Li W, Wu FC, Liu CQ, Fu PQ, Wang J, Mei Y, Wang LY, Guo JY (2008) Temporal and spatial distributions of dissolved organic carbon and nitrogen in two small lakes on the Southwestern China Plateau. Limnology 9(2):163–171CrossRefGoogle Scholar
  18. Li X, Huang TL, Ma WX, Sun X, Zhang HH (2015) Effects of rainfall patterns on water quality in a stratified reservoir subject to eutrophication: implications for management. Sci Total Environ 521:27–36CrossRefGoogle Scholar
  19. Meffert ME, Zimmermann-Telschow H (1979) Net release of nitrogenous compounds by axenic and bacteria-containing cultures of Oscillatoria-redekei (Cyanophyta). Arch Hydrobiol 87(2):125–138Google Scholar
  20. Miyazako T, Kamiya H, Godo T, Koyama Y, Nakashima Y, Sato S, Kishi M, Fujihara A, Tabayashi Y, Yamamuro M (2015) Long-term trends in nitrogen and phosphorus concentrations in the Hii River as influenced by atmospheric deposition from East Asia. Limnol Oceanogr 60(2):629–640CrossRefGoogle Scholar
  21. Nurnberg GK (2004) Quantified hypoxia and anoxia in lakes and reservoirs. Sci World J 4:42–54CrossRefGoogle Scholar
  22. Paerl HW, Xu H, McCarthy MJ, Zhu G, Qin B, Li Y, Gardner WS (2011) Controlling harmful cyanobacterial blooms in a hyper-eutrophic lake (Lake Taihu, China): the need for a dual nutrient (N & P) management strategy. Water Res 45(5):1973–1983CrossRefGoogle Scholar
  23. Pearce AR, Chambers LG, Hasenmueller EA (2017) Characterizing nutrient distributions and fluxes in a eutrophic reservoir, Midwestern United States. Sci Total Environ 581-582:589–600CrossRefGoogle Scholar
  24. Perron T, Chetelat J, Gunn J, Beisner BE, Amyot M (2014) Effects of experimental thermocline and oxycline deepening on methylmercury bioaccumulation in a Canadian Shield Lake. Environ Sci Technol 48(5):2626–2634CrossRefGoogle Scholar
  25. Rittmann BE, McCarty PL (2001) Environmental biotechnology: principles and applications. McGraw-Hill, New YorkGoogle Scholar
  26. Sawyer CN, McCarty PL, Parkin GF (1994) Chemistry for environmental engineering, four edn. McGraw-Hill Education, New YorkGoogle Scholar
  27. Small GE, Cotner JB, Finlay JC, Stark RA, Sterner RW (2014) Nitrogen transformations at the sediment-water interface across redox gradients in the Laurentian Great Lakes. Hydrobiologia 731(1):95–108CrossRefGoogle Scholar
  28. Taylor GD, Fletcher TD, Wong THF, Breen PF, Duncan HP (2005) Nitrogen composition in urban runoff - implications for stormwater management. Water Res 39(10):1982–1989CrossRefGoogle Scholar
  29. Torres E, Galvan L, Canovas CR, Soria-Piriz S, Arbat-Bofill M, Nardi A, Papaspyrou S, Ayora C (2016) Oxycline formation induced by Fe (II) oxidation in a water reservoir affected by acid mine drainage modeled using a 2D hydrodynamic and water quality model - CE-QUAL-W2. Sci Total Environ 562:1–12CrossRefGoogle Scholar
  30. Tuantet K, Temmink H, Zeeman G, Janssen M, Wijffels RH, Buisman CJN (2014) Nutrient removal and microalgal biomass production on urine in a short light-path photobioreactor. Water Res 55:162–174CrossRefGoogle Scholar
  31. Walker JT, Stow CA, Geron C (2010) Nitrous oxide emissions from the gulf of Mexico hypoxic zone. Environ Sci Technol 44(5):1617–1623CrossRefGoogle Scholar
  32. Wang SL, Liu CQ, Yeager KM, Wan GJ, Li J, Tao FX, Lue YC, Liu F, Fan CX (2009) The spatial distribution and emission of nitrous oxide (N2O) in a large eutrophic lake in eastern China: anthropogenic effects. Sci Total Environ 407(10):3330–3337CrossRefGoogle Scholar
  33. Wetzel RG (2001) Limnology: lake and river ecosystems, third edn. Academic Press, San Diego, CAGoogle Scholar
  34. Woolway RI, Maberly SC, Jones ID, Feuchtmayr H (2014) A novel method for estimating the onset of thermal stratification in lakes from surface water measurements. Water Resour Res 50(6):5131–5140CrossRefGoogle Scholar
  35. Yan S, Liu YY, Liu CX, Shi L, Shang JY, Shan HM, Zachara J, Fredrickson J, Kennedy D, Resch CT, Thompson C, Fansler S (2016) Nitrate bioreduction in redox-variable low permeability sediments. Sci Total Environ 539:185–195CrossRefGoogle Scholar
  36. Yoh M, Terai H, Saijo Y (1983) Accumulation of nitrous oxide in the oxygen deficient layer of freshwater lakes. Nature 301(5898):327–329CrossRefGoogle Scholar
  37. Yu Z, Yang J, Amalfitano S, Yu XQ, Liu LM (2014) Effects of water stratification and mixing on microbial community structure in a subtropical deep reservoir. Sci Rep 4:1–7Google Scholar
  38. Zhang YL, Wu ZX, Liu ML, He JB, Shi K, Zhou YQ, Wang MZ, Liu XH (2015) Dissolved oxygen stratification and response to thermal structure and long-term climate change in a large and deep subtropical reservoir (Lake Qiandaohu, China). Water Res 75:249–258CrossRefGoogle Scholar

Copyright information

© Springer-Verlag GmbH Germany, part of Springer Nature 2018

Authors and Affiliations

  1. 1.Key Laboratory of the Three Gorges Reservoir Region’s Eco-Environment, Ministry of EducationChongqing UniversityChongqingChina
  2. 2.College of Urban Construction & Environmental EngineeringChongqing UniversityChongqingChina

Personalised recommendations