Effect of water temperature on internal nitrogen release from sediments in the Pearl River Delta region, China

The rapid population growth and the expansion of industrial and agricultural economy have resulted in the excessive discharge of nitrogen (N) into receiving waters (Kirschke et al. 2019; Kazakis et al. 2020; Luo et al. 2022). This discharge is commonly attributed to the widespread use of fertilizers and the intensive discharge of untreated municipal and industrial wastewater (Zhang et al. 2017; Bai et al. 2018; Martínez-Santos et al. 2018). As a nutrient, nitrogenous compounds are essential for the survival of all aquatic life. However, when present in excess, they can cause severe ecological and environmental impacts such as eutrophication, hypoxia of water bodies, and toxicity to benthic organisms (Oviatt et al. 2017; Shayo & Limbu 2018; Zhu et al. 2019b; Shahmohamadloo et al. 2020). The reduction of exogenous N disposal into natural receiving waters is a positive trend in China (Wang et al. 2016; Wu & Yu 2021). However, despite this, decades of accumulated nitrogen in river sediments continue to be released into surface waters, hindering anticipated improvements in water quality (Zhou 2020a; Yao 2023). Moreover, the gradual warming of the climate environment in recent years has led to changes in the nitrogen content of water bodies (Greaver et al. 2016), which, although subtle, may result in severe consequences on water quality over time (Zhou et al. 2020b; Canfield et al. 2021). Therefore, it is crucial to obtain information about the rate of nitrogen release from sediments, particularly at high water temperatures, and to know the basic process of N pollution diffusion which is essential for further study.

Located in southern China, the Pearl River is the largest in southern China and exhibits a typical subtropical lotic ecosystem (Xu et al. 2019; Wang et al. 2020a). However, rapid commercialization, urbanization, and large-scale industrialization have led to serious water pollution in the Pearl River Delta (PRD), with eutrophication of the river and estuary becoming the dominant environmental concern in the region (Wang et al. 2021). As China's most densely populated and economically developed area, the PRD region faces severe pollution threats. As a result, extensive research has been conducted on water quality issues in this region (Chen & Chen 2020; Rong et al. 2020; Chen et al. 2022). Among the various sources of water pollution, endogenous pollution remains a major contributor, posing continuous threats to water bodies. Endogenous N release studies have been conducted to investigate the reaction of N and other substances (Zhang et al. 2014a; Huang et al. 2020, 2021). However, despite numerous studies focusing on the environmental behavior of N in water bodies (Martinelli et al. 2018; Nikolenko et al. 2018; Letizia Costantini et al. 2021; Zhang et al. 2021), most of these assessments are short-term and ignore seasonal effects on temperature and water quality. While field surveys can certainly shed light on the variations of pollutants across different regions, indoor model experiments with controlled environmental conditions allow for a comprehensive investigation into the mechanisms of pollutant dispersion within a selected timeframe. Such experiments provide a better understanding of the diffusion process.

Nitrification and denitrification in aquatic environments are closely interconnected, and the rate of nitrate diffusion at the sediment–water interface (SWI) is believed to limit the role of sediment denitrification (Olde Venterink et al. 2003; O'Connor & Hondzo 2008). Denitrification is a process whereby N (predominantly nitrate-nitrogen) is reduced, leading to the production of N₂ and N₂O gases. The effect of temperature on N transformation has been extensively studied, and it has been found that benthic dissolved Inorganic Nitrogen (DIN, sum of NH₄-N, NO₂-N, and NO₃-N) concentrations play a key role in the transformation process (Hardison et al. 2015; Wei et al. 2020; Huang et al. 2021). The rate of nitrification and denitrification in the sediment and water column (WC) becomes faster at higher temperatures, resulting in unique distributions of nitrogenous compounds (Young et al. 2017; Chen et al. 2018). Previous studies have examined N removal and the impact of temperature on microorganism growth and N behavior in various environments such as soil, lakes, rivers, and basins under different temperature conditions related to climate change, based on site-specific observations (Liu et al. 2018; Wu et al. 2019; Wang et al. 2020b). For the PRD region, access constraints have resulted in limited measurements of the rate of N release from sediments. Furthermore, the nutrient load resulting from N release in each season cannot be accurately quantified using traditional field methods. Physical model simulation is a significant and reliable method for analyzing N diffusion (Xu et al. 2016; Liu et al. 2018). In past studies on endogenous release (Zhang et al. 2014b; Wang et al. 2017; Sima et al. 2020), columnar simulators were often used to simulate the endogenous release process and analyze the longitudinal distribution of nutrients. However, this analog device cannot control the temperature of the entire environment. Unlike a columnar incubator, a physical model can simulate a more realistic culture environment.

In this study, we constructed a model consisting of five tanks in the laboratory, filled with sediment and distilled water. We measured the temporal changes in ammonia, nitrite, nitrate, dissolved oxygen (DO), and pH at various positions in the sediment and WC of the model. The objectives of this experimental study were (1) to assess the effect of temperature on N release from the sediment, (2) to evaluate the rate of N release under different temperatures, and (3) to provide some data support for early warning and prevention of N pollution in the PRD region. The results of this research can help to determine the acceptable rate of N release under idealized conditions, serve as a basis for modeling water quality in water body, and provide a means to explore potential treatments.

Sediment was taken from a site located in the main waterway of the Pearl River, in the southern portion of Guangzhou City (113°16′ E, 23°4′ N), NH₄-N of less than 1 mg·L⁻¹ as N and DO of approximately 5 mg·L⁻¹ in the water sample. A grab-type sampler was used to collect the sediment from a depth of 2–3 m below the water surface, and the sediment was then transferred to a sealed polyethylene plastic barrel and transported back to the laboratory. The sediment was thoroughly stirred and sieved to remove large stones, debris, leaves, and benthos before being added to the tank. The Rhizon pipes (2.5 mm inner diameter, Rhizosphere Research Products, Wageningen, Netherlands) were inserted into the sediment to extract pore water. The use of Rhizon pipes has many benefits, including low cost, minimal disturbance, avoiding sample contact with air, and preserving the sample's chemical components (Seeberg-Elverfeldt et al. 2005; Shotbolt 2010). The sediment surface was scraped smoothly to guarantee it had a uniform thickness of 10 cm. Previous studies (Bao et al. 2021; Yao et al. 2022) have shown that N transformation is faster in the top 5 cm of the surface sediment and the N levels are relatively stable below a sediment depth of 10 cm; thus, a sediment depth of 10 cm was chosen. The distilled water was gradually added to the tank to reduce sediment suspension as much as possible until the water depth reached 18 cm. One sampling position was located at the SWI, and there were five sampling positions in the sediment at depths of 1, 2, 3, 5, and 7 cm below the SWI, in this study, we divided three parts at different depths: WC, shallow sediment (SS, 0–3 cm below the SWI), and deep sediment (DS, 5–7 cm below the SWI). Three sampling positions were situated at a distance of 5, 10, and 15 cm above the SWI.

Five identical water tanks were used to control the water temperature to 10, 15, 20, 25, and 30 °C, respectively, representing the typical temperature range in the PRD region. Based on previous experimental results, the release typically reaches a steady state after around 10 days. However, considering the influence of different temperatures, in this study, each test at a specific water temperature was carried out for 15 days under the same initial conditions.

Three water samples were taken at each predetermined sampling position in the WC and pore water using a vacuum tube at the same time each day during the experiment. The NH₄-N, NO₂-N, and NO₃-N concentrations of the samples were measured using a microplate spectrophotometer (Epoch 2, BioTek Instruments) that applied the indophenol method, Griess chromogenic reagent (Chr-R) method, and Copper-Cadmium (Cu-Cd) reduction method, respectively (Laskov et al. 2007; Tu et al. 2010; Park et al. 2018). First, a pre-configured N solution was used, and reagents were added to induce a color reaction. Absorbance values were measured at different wavelengths (NH₄-N: 660 nm, NO₂-N: 540 nm, NO₃-N: 540 nm) to establish standard curves with a correlation coefficient greater than 99.99% between absorbance values and concentration values. Subsequently, the obtained absorbance values from the analysis of actual water samples were used to calculate the corresponding concentrations.

In addition, the pH and DO of the water at a depth of 9 cm were measured using a calibrated pH meter (PH-801/902, ADVICS, Changzhou, China) and a YSI ODO meter (model # 605404, YSI Incorporated, Yellow Springs, Ohio, USA).

Analysis of Variance (ANOVA) is a statistical formula used to compare variances across the means (or average) of different groups. In this study, we use this method to test N differences at different temperatures. The WC and pore water were separated, and 10 groups of sampling data were collected from five tanks with different temperatures. The differences between the five groups of data were calculated. The null hypothesis in this study is H₀: there is no difference between different groups. If the p-value is less than the indicated significance level (0.05), the null hypothesis can be rejected, and the samples are likely distinct. Before the analysis, Shapiro–Wilk test, Bartlett's test, and Levene's test were used to analyze outlier, normality, homogeneity of variances of the residuals.

In statistics, the two-way ANOVA is an extension of the one-way ANOVA that examines the influence of two different categorical independent variables on one continuous dependent variable. The two-way ANOVA not only aims at assessing the main effect of each independent variable but also if there is any interaction between them. In this study, due to the influence between samples in the group, the two-way ANOVA method was adopted.

The ANOVA results of the differences in N concentrations for different temperatures are presented in Table 1. For clarity, the data collected at the same temperature and location are treated as the same group (e.g., data collected from the WC on different days at 10 °C are assigned to group 1). Because of the relationship between N at different depths, the effects at different depths were also analyzed within the group. The WC results indicate significant differences (p < 0.001) with pronounced differences observed among groups, and there was no intra-group difference between the different depths. The differences in nitrate concentrations of pore water are also evident (p < 0.001). Similarly, different depths in the group also had a significant impact on the results, and NH₄-N, NO₃-N and DIN showed statistical results of mutual influence between temperature and depth, so a separate effect analysis was conducted on them, as shown in Table 2. Except for NO₃-N at 5 and 7 cm depths, the other indicators were all affected by temperature at different depths. Also, there was significant statistical difference.

Figure 2(d) shows that the DIN release has a sudden change at 20 °C where it increases sharply. Although the diffusion process increases at higher temperatures, it is bidirectional. At the start of the experiment, due to the extremely low N content in the WC, N in the sediments also exists in the form of NH₄-N, the release of which is driven by the concentration difference and temperature, gradually accelerating with increasing temperature. During the experiment, nitrification gradually increases with increasing temperature, leading to a decrease in NH₄-N and an increase in NO₃-N in the WC. The nitrification rate at 30 °C is not significantly different from that at 25 °C, but the diffusion rate is higher. At this point, the difference in the rate of N diffusion up or down is less than that at 25 °C. Therefore, both NH₄-N, NO₃-N, and DIN are relatively low in concentration. The temperature of 20 °C stands out as relatively abnormal, with the NH₄-N content far exceeding that of other temperatures, resulting in a greater concentration of DIN at 20 °C. The results demonstrate that the net rate of NH₄-N released from internal sources of sediments is highest at 20 °C. Temperature affects the internal release of NH₄-N, an increase in temperature accelerates the release rate of endogenous NH₄-N, and it also enhances the nitrification rate (Painter & Loveless 1983). As a result, the net growth quantity derived from the difference between release and reaction rates reaches a relatively high level at 20 °C. However, the presence of NH₄-N in the water body can inhibit the release process. Therefore, as the temperature continues to rise, the NH₄-N content gradually decreases.

Due to the concentration difference at the SWI, NO₂-N and NO₃-N enter the pore water, leading to NO₂-N and NO₃-N concentrations in the SS that are greater compared to the DS. Because of the diffusion action, NO₂-N and NO₃-N contents at a temperature of 20–25 °C below the SWI are greater than those in the WC at lower temperatures. When the temperature increases from 15 to 20 °C, average NO₂-N concentration in WC increases from 14 to 122 μg·L⁻¹, indicating that NO₂-N can exist in the WC more stably at higher temperatures.

DIN in the WC comes from the sediment, affected by physical diffusion. Compared to the changing concentrations of NH₄-N, NO₂-N, and NO₃-N related to DO, DIN is mainly constant in the longitudinal direction (Figure 3(d)). It is seen that the DIN variation is similar to that of NH₄-N, which is highest at 20 °C and decreases with temperature up to 30 °C. From the concentration distribution of DIN in the pore water, the DIN in the WC approximates that of the pore water in the shallow sediments at 20 °C, a large amount of DIN is released from the sediment to the WC, and gradually reaches an equilibrium state under this temperature condition. With higher or lower temperatures, there is obvious DIN stratification in the vertical direction (Figure 3(d)). The average DIN concentration at 1 cm below the SWI is 1.64 and 2.46 times that of the SWI at 10 and 30 °C, respectively.

It should be noted that NO₂-N and NO₃-N were not present in the WC at the beginning of the experiment. These nitrogenous compounds diffuse from the sediment to the WC, and their flux increases at higher temperatures (Figure 4(b) and 4(c)). Under anaerobic conditions in the sediment, both NO₂-N and NO₃-N concentrations are relatively low. When NH₄-N in the WC is converted to NO₂-N and NO₃-N under aerobic conditions, their concentrations become greater than those in the sediment. These nitrogenous compounds further diffuse from the WC to the sediment. Simultaneously, NO₂-N and NO₃-N in the sediment are consumed by denitrification, which maintains the concentration gradient near the SWI. Although the concentration of NO₂-N is low (all lower than 0.6 mg·L⁻¹), there is still a significant difference between the NO₂-N concentration at 10 °C (average 3.9 μg·L⁻¹) and 15 °C (average 14.1 μg·L⁻¹); and at higher temperatures, NO₂-N release and concentration maintain an increase from 15 to 20 °C.

Table 3 shows the average diffusive flux of NH₄-N and nitrogenous compounds at different temperatures. At 10 °C, the average diffusive flux of NH₄-N is 2.04 mg·m⁻²·d⁻¹, which is 2.6 times greater than that at 30 °C. The NO₂-N flux increases rapidly from −0.05 μg·m⁻²·d⁻¹ at 10 °C to −40.68 μg·m⁻²·d⁻¹ at 20 °C, and reaches the maximum of −49.86 μg·m⁻²·d⁻¹ at 25 °C (a positive value indicates that nitrogenous compounds are released from the sediment into the WC). Both NO₃-N and NO₂-N have similar trends, with a maximum flux at 25 °C. D_s increases with temperature. Both NO₃-N and NO₂-N flux initially increase and then decrease, indicating a threshold for diffusive flux between 20 and 30 °C. The daily flux in Figure 4(a) shows that NH₄-N can stabilize at a higher level in the WC at 20 °C, resulting in a lower average flux.

The main factors driving diffusion are the differences in concentration and temperature, affecting both mass transfer flux and ongoing reactions. Overall, the diffusive flux is not a linear function of temperature. To obtain the average diffusive flux at the SWI and quantify the effect of temperature on N release from sediment, curve fitting was conducted based on the vertical distribution of N concentrations. Based on data distribution and model distribution, a Polynomial2D model was chosen for NH₄-N and NO₃-N, and a Lorentz2D model for NO₂-N. Numerous functional fits were assessed based on their R² value, and for brevity, only the best relationships are shown. is the average diffusive flux of the first t days at T°C. Other variables in the equations are simple fitting parameters chosen to match the results best. Most of the best-fit curves have good fit results (R² > 0.83), as shown in Figure A1 and Table B1. Based on the fitted equations, the rate of diffusion of N from sediment to WC at different temperatures (10–30 °C) and its change trend can be estimated.

There is a correlation between N forms with depth. The absolute value of the Pearson correlation coefficient (r) was classified into four groups: negligible correlation (r < 0.30), low correlation (0.30 < r < 0.50), moderate correlation (0.50 < r < 0.70), and high correlation (r > 0.70).

In this study, we collected sediment in situ and brought it back to the laboratory for simulation after thorough mixing to maintain sediment consistency under different working conditions as much as possible. The water environment maintains a constant temperature during the experiment, which accurately simulates the corresponding conditions and thus obtains valid results.

The biogeochemical transformation of N is an important factor affecting the N content of a water body. The nitrification process was enhanced with increasing temperature at the beginning of the experiment. With a large increase in NO₃-N, the denitrification rate was also gradually enhanced. As the NO₃-N substrate was reduced, NO₃-N concentration in the sediment or the WC became an effective environmental factor in the NO₃-N reduction process (Tan et al. 2019). At 25 and 30 °C, N is primarily in the form of NO₃-N in water bodies. When the temperature rises, there is an increase in NO₃-N concentrations, diffusion rates, and denitrification rates. As an intermediate product of nitrification and denitrification, NO₂-N not only reflects the nitrification process but also does serious harm to water bodies (Wu et al. 2020). Additionally, NO₂-N is less than 1% in this temperature range (10–15 °C). As NO₂-N is produced in a similar way to NO₃-N, it is more active at higher temperatures. NO₂-N has the greatest NNR value at 20 °C and shows the most significant ratio at 25 °C. It was also concluded by Zhang et al. (2014c) that NH₄-N and NO₂-N oxidation rates are highly dependent on temperature. Furthermore, the experimental findings indicate that the change rates of NH₄-N and NO₃-N have little effect when temperatures are changed between 24 and 28 °C, and there is a significant difference between 10 and 24 °C, which is consistent with the fact that the highest NO₃-N content occurred at 25 °C.

The main factor controlling N transformation among environmental factors is DO (Yan et al. 2019). The DO content decreases over time at all temperatures for all tests. Higher temperatures increase the rate of various chemical reactions, causing more DO consumption correspondingly (Zhu et al. 2019a). During the experiment, the water tank is airtight. The DO content decreases rapidly and falls below 1.5 mg·L⁻¹ at 30 °C, while remaining greater than 3 mg·L⁻¹ at other temperatures. For example, at 10 and 15 °C, DO is greater than 6.4 mg·L⁻¹. With the consumption of a large amount of DO, NO₃-N content increases at higher temperatures. Despite low DO content at 30 °C, both nitrification and denitrification processes occur, leading to an increasing proportion of NO₃-N. Apart from the changes caused by chemical processes, temperature also affects the rate of diffusion, which, in turn, affects the nitrogen concentration in the WC and the pore water. Changes in concentration differences caused by chemical processes also affect diffusion. Furthermore, pH, as an important environmental indicator influencing N transformation, also exerts an impact on water pollution. In this study, pH remained between 7.1 and 7.2 during the temperature range of 10–20 °C. The influence of temperature variation was primarily observed at 25 and 30 °C, where the average pH values reached 7.34 and 7.58, respectively. This indicates that with increasing environmental temperature, the water gradually shifts from neutral to alkaline conditions. Under alkaline conditions, denitrification is enhanced, leading to increased consumption of NO₃-N (Zou et al. 2016). This finding also suggests that as the temperature continues to rise, the concentration of NO₃-N actually decreases.

Previous studies primarily focused on in situ sampling analyses (Zhang et al. 2014a; Hong et al. 2019; Tan et al. 2019) to estimate the effect of temperature on release fluxes by comparing the vertical distribution of N pollution across seasons. However, nutrients in water bodies mainly originate from upstream sources, particularly during summer when runoff is abundant (Zhang et al. 2013). Therefore, the in situ studies' results of temperature-dependent endogenous release may be inaccurate in the PRD region. It is also difficult to compare the results of this study to the in situ observations. Since the soil is fairly capable of adsorbing NH₄-N, farmland sources are unlikely to contribute to NH₄-N loading in the water (Rush 1915; Hong et al. 2019). As a result of controlling exogenous pollution, NH₄-N in the PRD area primarily comes from endogenous release from sediments, upstream inflows, and DIN transformation.

This study demonstrates that NH₄-N can easily be maintained at high levels at lower temperatures (10–20 °C). The average temperature in Guangzhou, a typical city in the PRD region, is estimated to be between 10 and 15 °C during December, January, and February, and between 15 and 20 °C during November, March, and April, while exceeding 25 °C from June to August. Water temperature maintains a good linear correlation with air temperature (Stefan & Preud'homme 1993) and tends to be lower. Thus, NH₄-N concentration in water bodies has a risk of increasing from November to the following April. During summer, there is a significant increase in the ratio of NO₂-N in DIN, increasing aeration or promoting denitrification using other means are the way to control the N. In summary, the endogenous nitrogen release in the PRD region is significantly influenced by temperature, and seasonal variations in water quality parameters should be considered when developing management strategies. Further research is needed to identify the temperature on nutrient delivery conditions scenario and develop appropriate management measures.

In this study, an indoor tank model experiment was used to investigate the effect of temperature on endogenous nitrogen release from sediments. The conclusions are as follows:

• (1)

  The effect of temperature on N in the WC and sediment was found to be nonlinear. NH₄-N concentration in the WC was more significantly impacted at lower temperatures, stabilized at a higher level at 20 °C. The rate of nitrification was found to be the most rapid at 25 °C. NO₂-N can exist more stably in the WC at higher temperatures. NO₃-N concentration was heavily affected at higher temperatures, and N was mainly found in the form of NO₃-N from 25 to 30 °C. The influence of temperature on N in WC and pore water is different. Nitrate concentration in the SS pore water was only affected by temperatures in the range of 20–25 °C.

• (2)

  NO₂-N and NO₃-N in the WC were found to be converted from NH₄-N and were released from the sediment. Due to the continuous release process, obtaining a solid correlation of NO₂-N with depth was difficult. Higher temperature was found to improve the internal release rate of NH₄-N, yet the presence of NH₄-N in the WC could inhibit the release process. The diffusive process could be described by Polynomial2D and Lorentz2D models, respectively.

• (3)

  It is recommended that the risk of increased NH₄-N should be noted from November to the following April. In the summer, the ratio of NO₂-N in DIN increases significantly, and measures such as increasing aeration or promoting denitrification by other means can be used to control it.

This work was supported by the Open Research Fund Program of the State Key Laboratory of Hydroscience and Engineering (sklhse-2021-B-02), the Fund of Science and Technology Program of Guangzhou (202102020254), Science and Technology Innovation Program from Water resources of Guangdong Province (2023-06), and the National Key R&D Program of China (2019YFD0900302).

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.