The Impact of Sediment–Water Ratio and Hydraulic Residence Time on the Release of Inorganic Nitrogen from Sediments in the Pearl River Delta

Abstract

Black-odorous water bodies in the Pearl River Delta have been treated. However, the re-release of nitrogen (N)-containing compounds in sediment can cause a relapse of black-odorous water bodies. Sediment–water ratio (SWR) and hydraulic residence time (HRT) influence pollutant release. Therefore, how to control SWR and HRT during the treatment process has become an urgent problem. This study focuses on the dynamic release of endogenous inorganic N from sediments into overlying water in a river channel of Dongguan City, Guangdong Province. Physicochemical parameters (dissolved inorganic nitrogen (DIN), NH₄⁺-N, NO₃⁻-N, NO₂⁻-N, dissolved oxygen (DO), pH, oxidation-reduction potential (ORP), chemical oxygen demand (COD), Fe and total phosphorus (TP)) of overlying water were monitored under different SWRs (0.71, 0.38, and 0.16) and HRTs (13 days and 6.5 days), and the nitrogen release flux under different conditions was compared. Finally, the correlation and influence pathways among environmental factors were analyzed. The results showed that SWR significantly affected DO, pH, ORP, and sediment N release fluxes while prolonging HRT-promoted denitrification. DIN → NO₂⁻-N → DO pathway had a total effect of 19.6%, and DIN may promote low DO concentration via NO₂⁻ oxidation. Maintaining reasonable SWR and HRT can reduce the release of inorganic N from sediment into the overlying water. This study provides a theoretical basis for controlling black-odorous water bodies.

1. Introduction

The Pearl River Delta is one of China’s most developed regions. In recent years, rapid development in the Pearl River Delta has caused a significant increase in pollution from excessive nitrogen (N) discharge due to fertilizer and fossil fuel use [1]. The discharge of wastewater from various sources—such as industrial emissions, agricultural production, urbanization, and human activities—combined with high water temperatures ranging from 13.9 °C in winter to 31.5 °C in summer has resulted in black-odorous water bodies, unique to southern China [2,3]. The Pearl River Delta region—part of Guangdong Province—records the highest number of black-odorous water bodies in China. According to data collected from the National Urban Black-odorous Waterbodies Governance and Supervision Platform (Chinese Ministry of Housing and Urban-rural Development, 2020), there were almost 250 black-odorous water bodies in Guangdong Province, with 41 still untreated. Algae and bacteria activity in Guangzhou and Hong Kong have also transformed some riverbed sediment into black-odorous sediment [4,5]. Therefore, measures have been put in place to treat black-odorous water bodies. However, pollutants can still desorb from sediment and diffuse into the overlying water under certain conditions, leading to the recurrence of black-odorous water bodies under certain conditions [6,7]. For example, disturbance or changes in oxidation-reduction potential (ORP) and pH values of sediment can cause pollutants to migrate into the overlying water, leading to a relapse of black-odorous water bodies [8,9]. High levels of dissolved inorganic N can also cause blue-green algae growth and eutrophication [10,11]. In addition, the process of dissimilatory nitrate reduction to ammonium (DNRA) can transform NO₃⁻ into ammonia (NH₃) [12]. Increased organic matter (OM) deposition and oxygen consumption on the surface of nutrient-rich sediment systems result in a more reduced state, affecting the biogeochemistry and nutrient cycling of sediment [13,14,15,16]. As shown by the document “Letter on the notification of the national water environment situation in the first half of 2022” from the General Office of the Ministry of Ecology and Environment, in the first half of 2022, 10 cities in China with completed black-odorous water body remediation reported significant relapse, with two of them located in the Pearl River Delta region. Thus, it is necessary to study environmental factors affecting sediment pollutant release to control water quality.

Sediment–water ratio (SWR)—the ratio of sediment thickness to overlying water depth—is a critical factor affecting pollutant distribution in sediments and overlying water. Previous studies have shown that changes in SWR induced by rainfall and human activities—such as dredging, rainfall, and wastewater discharge events—can significantly alter pollutant loadings in aquatic systems [17,18,19]. Typically, pore water in sediment serves as a carrier of sediment pollutants, while overlying water is the endpoint for pollutant migration. Thus, sediment thickness directly affects the total amount of pollutants, while overlying water volume affects pollutant concentration [20]. To date, most research in SWR has been focused on coastal and marine systems [21,22] with little attention paid to freshwater ecosystems in South China. Hydraulic residence time (HRT)—on the other hand—is an important parameter in water environmental engineering, representing the average time that overlying water stays in the system [23]. HRT is mainly determined by the flow rate and the cross-sectional area of the river. A reduction in HRT can trigger hydrodynamic shear force, turbulence, and sediment resuspension at the sediment–water interface (SWI) [24,25], which can alter nutrient and oxygen transport processes, thus affecting the microbial habitats crucial for the survival of aquatic organisms and plants [26,27]. The initial occurrence of low dissolved oxygen (DO) in the water column takes place at the SWI [28], where iron oxides dissolve and release iron into the pore water under reducing conditions [29,30]. Therefore, the process of blackening originates from the SWI and spreads towards the water body [31]. Previous studies have reported that controlling SWR and reducing HRT can effectively reduce pollution from the SWI in estuaries, lakes, coastal regions, and marine systems [32,33,34,35]. Nevertheless, the impact of different SWRs and HRTs on water quality and pollutant distribution in freshwater ecosystems in South China still needs to be further investigated, and a statistical analysis of the effects of SWR and HRT on water quality is currently lacking.

To investigate the dynamic sediment release rules under different SWRs and HRTs, we conducted a 53-day experiment where black-odorous sediment from the Pearl River Delta was added to three dynamic overlying water sediment (DOWS) systems, simulating and analyzing N (dissolved inorganic nitrogen (DIN), NH₄⁺-N, NO₃⁻-N and NO₂⁻-N) release flux and patterns of sediment into overlying water under different SWRs and HRTs. We monitored changes in physicochemical parameters—including DO, pH, ORP, chemical oxygen demand (COD), Fe, and total phosphorus (TP)—and analyzed their correlations with preset and water quality parameters. This study aims to provide theoretical support for controlling the relapse of black-odorous water bodies.

2. Materials and Methods

2.1. Sediments

As shown in Figure 1, the experiment selected sediment from the Xiacao River (23°1′17″, 113°37′51″), located in Wangniudun Town, Dongguan City, Guangdong Province, China. This river has formed black-odorous water bodies due to excessive discharge of metal and electronic wastewater, with a large amount of iron oxide visible on the water surface and a strong odor. A box corer was used during sediment collection, and sediment was collected up to a depth of 50 cm from the surface and immediately transported back to the laboratory within 4 h. To ensure comparability, debris such as garbage and branches in the sediment were removed before starting the experiment.

2.2. Set-Up

As shown in Figure 2, three DOWS systems were constructed in the laboratory using three organic glass columns (1.5 m high and 0.14 m inner diameter), a water tank, and a peristaltic pump (BT100-2J, Longer Precision Pump Co., Ltd., Baoding, China). Homogenized sediment was added to the bottom of each DOWS with controlled sediment thicknesses of 60 cm, 40 cm, and 20 cm, respectively, named DOWS1, DOWS2, and DOWS3. Deionized water was added before the start of the experiment, the water inlet of each DOWS was placed 3 cm above the SWI, and an overflow outlet was set at a height of 145 cm in each DOWS to maintain the depth of the overlying water at 85 cm, 105 cm, and 125 cm, respectively, resulting in SWRs of 0.71, 0.38 and 0.16. The peristaltic pump operated at a constant rate to ensure a continuous flow of the overlying water and compensate for water loss due to evaporation and sampling.

2.3. Operation

According to the different HRT of the overlying water, the experimental period was divided into two stages, named stage 1 (days 1–35) and stage 2 (days 37–53). In stage 1, the HRT was 13 days, and the flow rates of DOWS1, DWOS2, and DOWS3 were 0.70, 0.86, and 1.03 mL·min⁻¹, respectively; in stage 2, the HRT was 6.5 days, and the flow rates of DOWS1, DWOS2, and DOWS3 were 1.40, 1.72, and 2.06 mL·min⁻¹, respectively. The flow rate of the overlying water was controlled by a precision peristaltic pump.

The relationship between HRT and overlying water flow rate is expressed as:

$$\begin{array}{c}\mathrm{HRT}=\frac{{10}^5\times H\times \mathsf{\pi }{d}^2}{576\times Q}\end{array}$$

(1)

where H represents the depth of the overlying water (m), d represents the inner diameter of the dynamic overlying water sediment (DOWS) (m) and Q represents the inlet flow rate of the DOWS (mL·min⁻¹).

2.4. Physicochemical Indexes

In our experiment, the Chinese National Standard methods were used to detect the concentrations of NH₄⁺-N, NO₃⁻-N, NO₂⁻-N, COD, TP, and Fe in the water samples. The methods are described as follows: NH₄⁺-N in water samples was detected according to the method specified in GB 7481-87, using Nessler’s reagent colorimetry. NO₃⁻-N was determined according to the procedures outlined in GB 11894-89 using reduction by cadmium and colorimetric detection with salicylic acid. NO₂⁻-N was detected by the method prescribed in GB 748-87, using sulfanilic acid and N-(1-Naphthyl)ethylenediamine dihydrochloride diazo chromogenic reaction. COD was analyzed with the potassium dichromate method outlined in GB 11914-89. TP was detected according to the procedures specified in GB 11893-89, using H₂SO₄ digestion and ammonium molybdate spectrophotometric analysis. The concentration of Fe was detected according to the procedures specified in GB 11892-89, using the 1,10-phenanthroline method.

Water samples below 20 cm from the water surface were collected periodically and filtered through a 0.45 μm filter. The concentrations of NH₄⁺-N, NO₃⁻-N, NO₂⁻-N, COD, and TP were analyzed using a UV-visible spectrophotometer (EPOCH2 microplate reader, BioTek Instruments, Santa Clara, CA, USA). DIN was determined as the sum of NH₄⁺-N, NO₃⁻-N, and NO₂⁻-N. ORP was measured using an ORP electrode (501-ORP, Leici Instrument Branch of Shanghai Yidian Scientific Instrument Co., Ltd., Shanghai, China). DO measurements were conducted in situ using a portable multi-parameter analyzer (JPSJ-605F, Leici Instrument Branch of Shanghai Yidian Scientific Instrument Co., Ltd., Shanghai, China), while pH was determined using a pH meter (PHS-3E, INESA Scientific Instrument Co., Ltd., Shanghai, China). Each sample was measured three times.

2.5. Data Statistics and Analysis

Statistical analysis was performed using IBM SPSS Statistics 22 (IBM, New York, NY, USA) and R version 4.2.1 (www.r-project.org accessed on 23 June 2022), and charts were generated using Origin 2023. The non-parametric Friedman test was used to determine the mean and statistical differences among the experimental results of each DOWS to check for differences between SWRs. When a statistically significant p-value (<0.05) was obtained from the Friedman test, Wilcoxon signed-rank test was conducted, whose advantage is that it does not require distributional assumptions, can handle small sample sizes and outliers, and is useful when the data are not normally distributed [36,37]. To assess whether the changes in HRT significantly impacted the physicochemical parameters, we conducted Wilcoxon signed-rank tests on each DOWS across various HRT conditions. Spearman rank correlation analysis was performed using R to investigate the correlation between the experimental results of each DOWS. To further explore the relative importance of DO, a best-fit model was established using a multiple-stepwise regression approach to test the overall goodness-of-fit of the model. Path analysis is a structural equation modeling method used to measure the comprehensive influence of multiple factors on multiple variables. Its advantage is that it allows for direct and indirect effects between variables, considers multiple factors’ influence, assesses the overall fit of a model, and can provide a high level of transparency in structural comparisons [38]. During multiple-stepwise regression analysis and path analysis, multicollinearity among environmental variables was diagnosed, and no multicollinearity was found (variance inflation factor, VIF < 5) [39].

Nutrient fluxes were calculated using Equation (2). Positive values indicate nutrient release from sediment to overlying water, while negative values indicate nutrient sorption by sediment.

$$\begin{array}{c}F=\frac{1000\times \left(c_m-c_n\right)\times V}{\pi {\left(\frac{d}{2}\right)}^2\times \left(m-n\right)}\end{array}$$

(2)

In Equation (2), F represents nutrient flux ((mg·(m²·d)⁻¹); m and n represent specific days on which sample collection was conducted during the experiment (d); c_m and c_n represent nutrient concentrations of samples on day m and day n, respectively (mg·L⁻¹); V represents the volume of the overlying water (m³); and d represents the inner diameter of DOWS (m).

3. Results and Discussion

3.1. Significance of Physicochemical Parameter Differences under Different SWR and HRT

The Friedman test results for various physicochemical parameters among different SWRs and the Wilcoxon signed-rank test results for various physicochemical parameters among different HRTs are shown in

Table 1. Test of differences in environmental variables among overlying water depths at the four DOWS of the experiment. Results based on the Friedman test and Wilcoxon Signed Ranks test.

		SWR a			HRT b
		Stage 1	Stage 2		DOWS1	DOWS2	DOWS3
DO	Asymp. Sig.	0.001 **	<0.001 ***	Z	−2.578 c	−1.023 c	−1.156 c
	χ2	13.300	16.545	Asymp. Sig.	0.010 **	0.306	0.248
pH	Asymp. Sig.	0.019 *	0.004 **	Z	−1.423 c	−0.356 d	−1.245 c
	χ2	7.895	10.857	Asymp. Sig.	0.155	0.722	0.213
ORP	Asymp. Sig.	0.086	0.004 **	Z	−2.934 c	−2.490 c	−2.936 c
	χ2	4.900	11.091	Asymp. Sig.	0.003 **	0.013 *	0.003 **
COD	Asymp. Sig.	0.329	0.406	Z	−0.889 d	−0.652 c	−2.380 c
	χ2	2.225	1.805	Asymp. Sig.	0.374	0.515	0.017 *
DIN	Asymp. Sig.	0.058	0.695	Z	−1.689 c	−1.867 c	−1.956 c
	χ2	5.700	0.727	Asymp. Sig.	0.091	0.062	0.050
NH4+-N	Asymp. Sig.	0.411	0.070	Z	−1.334 d	−1.156 c	−1.580 c
	χ2	1.788	5.317	Asymp. Sig.	0.182	0.248	0.114
NO3−-N	Asymp. Sig.	0.058	0.307	Z	−1.867 c	−1.689 c	−2.045 c
	χ2	5.700	2.364	Asymp. Sig.	0.062	0.091	0.041 *
NO2−-N	Asymp. Sig.	0.045 *	0.074	Z	−1.868 c	−1.201 c	−1.601 c
	χ2	6.200	5.200	Asymp. Sig.	0.062	0.230	0.109
Fe	Asymp. Sig.	0.782	0.148	Z	−0.089 c	−0.178 c	−0.533 d
	χ2	0.492	3.818	Asymp. Sig.	0.929	0.859	0.594
TP	Asymp. Sig.	0.235	0.241	Z	−0.178 c	−1.125 c	−1.960 c
	χ2	2.896	2.850	Asymp. Sig.	0.859	0.260	0.50 *

Notes: Asymp. Sig., asymptotic significance; Z, Z-score; χ², Chi-Square; DO, dissolved oxygen (mg·L⁻¹); ORP, oxidation-reduction potential (mV); COD, chemical oxygen demand (mg·L⁻¹); DIN, dissolved inorganic nitrogen (mg·L⁻¹); NH₄⁺-N, ammonia nitrogen (mg·L⁻¹); NO₃⁻-N, nitrate nitrogen (mg·L⁻¹); NO₂⁻-N, nitrite nitrogen (mg·L⁻¹); Fe (mg·L⁻¹); TP, total phosphorus (mg·L⁻¹). Significant correlations in bold; * indicates a statistical difference, p < 0.05; ** indicates a statistically significant difference, p < 0.01; *** indicates a highly statistically significant difference, p < 0.001. Positive or negative ranks in the table indicate higher or lower values detected in the low-HRT than in the high-HRT. ^a Friedman test. ^b Wilcoxon Signed Ranks test. ^c Based on positive ranks. ^d Based on negative ranks.

. As shown in the table, during stage 1, significant concentration differences were observed for DO, pH, and NO₂⁻-N among different SWRs (p < 0.05). During stage 2, significant differences above the level of significance were observed for DO, pH, and ORP among different SWRs (p < 0.01). The differences in DO and pH were most significant across different SWRs, which can be attributed to variations in microbial abundance within sediments under different SWRs. DO and pH can affect the ability of microorganisms to adapt and metabolize within sediments, therefore influencing oxygen consumption rates and organic matter degradation in sediment–water systems [40,41]. Physicochemical parameters, shown in Figure 3 and Figure 4, typically had lower average values in the second stage than in the first stage. Notably, exceptions to this trend were observed for COD and NH₄⁺-N in DOWS1, as well as Fe in DOWS3. Significant differences were observed in DO in DOWS1, COD, NO₃⁻-N, and TP in DOWS3, as well as ORP values across all DOWS, under different HRT conditions. The response of ORP to changes in HRT was most significant, as longer HRT favors bio-oxidation [42].

3.2. Changes and Wilcoxon Signed-Rank Test of DO, pH, COD, ORP, Fe, and TP

The changes in DO, pH, COD, ORP, Fe, and TP in the overlying water of each DOWS, as well as the differences between stage 1 and stage 2, are shown in Figure 3.

DO showed a significant concentration gradient and reaeration process under different SWRs, with significantly lower values in DOWS1 than that in DWOS2 and 3 (Figure 3a). After the start of the experiment, the DO concentrations in all DOWS rapidly decreased, with all DOWS maintaining around 6.5 mg·L⁻¹ on the first day. However, from the 2nd to the 23rd day, DO rapidly decreased to around 0.1 mg·L⁻¹ in DOWS1, while it dropped to around 1.5 mg·L⁻¹ in DOWS2 and 3. After the 23rd day, DO did not change significantly in DOWS1, whereas it increased rapidly to around 4.5 mg·L⁻¹ in DOWS2 and 3, exhibiting a phenomenon of initial decrease followed by an increase after HRT was shortened. These data indicate that the reaeration capacity of the overlying water is inhibited when the SWR is 0.71, and the DO concentration in the overlying water increases as the SWR decreases [43]. The pH value of the overlying water was higher at an HRT of 13 days than at an HRT of 6.5 days, indicating that the longer HRT promotes the production of alkaline substances, which dissolve into the overlying water. In addition, although COD does not exhibit an obvious pattern of change, a decrease was observed after the HRT was shortened. These may be due to a higher proportion of organic matter degradation as Stage 2 DO was lower than Stage 1 DO [44]. Moreover, the concentration of COD in DOWS1 was statistically significantly higher than that in DOWS2 and 3, which may be due to its highest SWR. Moreover, ORP and Fe decreased overall as the HRT was shortened. However, ORP remained at a strong oxidative level throughout the experiment (Figure 3d,e). According to some previous studies, the release of phosphorus (P) from sediments is mainly derived from surface sediments [45], and the amount of P released under alkaline conditions is greater than that released under acidic conditions [41]. These may explain the reason the concentration of TP was initially high but decreased after day 8.

3.3. Changes in DIN, NH₄⁺-N, NO₃⁻-N and NO₂⁻-N

Sediment is a crucial N storage reservoir that can release nutrients to overlying water under certain conditions. Inorganic N serves as an essential nutrient for aquatic organisms; thus, investigating the N flux at the SWI has significant ecological implications. The changes in DIN, NH₄⁺-N, NO₃⁻-N, and NO₂⁻-N concentrations in each DOWS and the differences between stage 1 and stage 2 are presented in Figure 4.

The results indicated that there were no statistically significant differences in the concentration of each parameter among different DOWS or stages (p > 0.05). During stage 1, the concentrations of DIN, NH₄⁺-N, and NO₃⁻-N were slightly higher in DOWS2 than in DOWS1 and 3. Although the Friedman test for NO₂⁻-N during stage 1 was significant, the p-values between DOWS for NO₂⁻-N were all greater than 0.05 after the Wilcoxon signed-rank test was performed.

The characteristics of the changes at different stages showed that the average concentrations of DIN, NH₄⁺-N, NO₃⁻-N, and NO₂⁻-N were highest during stage 1 and lowest during stage 2. The average values of DIN, NH₄⁺-N, NO₃⁻-N, and NO₂⁻-N in DOWS2 during stage 1 were 1.81, 1.65, 1.80, and 2.27 times those during stage 2, respectively. The increase in NO₂⁻-N during stage 1 was accompanied by a decrease in DO and ORP, while during stage 2, NO₂⁻-N continued to decrease despite lower DO, indicating the occurrence of anaerobic denitrification [46]. The sustained decrease in NO₂⁻-N during stage 2 may also be due to the shortened HRT and decreased COD, resulting in insufficient denitrification strength [47]. During stage 1, the peak and average values of DIN, NH₄⁺-N, and NO₃⁻-N in DOWS2 were higher than those in DOWS1 and 3. In the entire experiment, the concentrations of DIN and NO₃⁻-N in the overlying water under two different HRTs have decreased, which may be attributed to the presence of denitrification. In addition, previous studies have indicated that increased organic carbon and nitrogen content can promote ammonium release [48,49]. These suggested that sediment—with its rich functional microorganisms—may play a dual role as a source of pollutants and a site for microbial removal. Therefore, SWR and HRT can lead to differences in N concentrations in the overlying water.

In addition, the fluctuation in N release flux during stage 1 was more significant than during stage 2 (Figure 5). However, there was no significant difference in the mean and median N release flux between the two HRTs. The variation of DIN flux is more sensitive to HRT changes relative to NH₄⁺-N and NO₃⁻-N. N flux in this study holds great significance for practical applications. However, the simulation experiment yielded nitrogen flux values higher than those observed in reality, possibly due to greater sediment contamination levels in the simulation [50].

3.4. The Relationship between Nutrients and Potential Influencing Factors

To illustrate the relationship between nutrients and potential influencing factors, we performed Spearman rank correlation analysis, and the results are shown in Figure 6. As shown in the figure, during stage 1, DIN was significantly correlated with NH₄⁺-N in all DOWS and extremely significantly correlated with NO₃⁻-N. DO was negatively correlated only with NO₂⁻-N. The dissolved oxygen (DO) concentration in the overlying water has been demonstrated to affect denitrification in aquatic systems. Molecular oxygen can inhibit the activity of denitrifying enzymes, especially nitrite reductase [51].

DO is an important parameter that affects microbial metabolic function, and it is usually a key parameter in controlling nitrification, denitrification processes, and OM degradation efficiency in wastewater treatment [52]. The stepwise multiple regression results between DO concentration and potential influencing variables are shown in

Table 2. Multiple-stepwise regression of DO with the potential influencing variables (PIN(0.050), POUT(0.100)).

		Step	R2	F	p	Equation
Stage 1	DOWS1	1	0.491	19.335	<0.001	DO = −4.855 (±1.104) NO2−-N + 5.457 (±0.864)
		2	0.712	24.464	<0.001	DO = −3.818 (±0.874) NO2−-N + 4.678 (±1.217) pH + 30.263 (±9.312)
		3	0.773	22.606	<0.001	DO = −3.765 (±0.775) NO2−-N + 4.046 (±1.112) pH + 0.572 (±0.241) NO3−-N − 26.734 (±8.392)
	DOWS2	1	0.646	35.735	<0.001	DO = −3.298 (±0.552) NO2−-N + 6.048 (±0.456)
	DOWS3	1	0.434	15.561	<0.001	DO = −1.724 (±0.489) NO2−-N + 2.266 (±0.990) pH − 12.404 (±7.572)
		2	0.544	12.352	<0.001	DO = −1.524 (±0.451) NO2−-N + 2.111 (±0.897) pH + 0.308 (±0.140) NH4+-N − 11.688 (±6.850)
Stage 2	DOWS1	No variables entered
	DWOS2	1	0.408	7.885	0.020	DO = 12.817 (±4.564) TP + 3.628 (±0.101)
	DOWS3	1	0.620	14.705	0.004	DO = 0.017 (±0.004) ORP + 1.572 (±0.575)
		2	0.836	20.360	<0.001	DO = 0.015 (±0.003) ORP − 0.967 (±0.299) NO2−-N + 1.607 (±0.401)

. During multiple-stepwise regression analysis, no multicollinearity was found among environmental variables (ORP, COD, pH, NH₄⁺-N, NO₃⁻-N, NO₂⁻-N, Fe, and TP). During stage 2, no variables entered the equation for DOWS1; DO was positively correlated with ORP in DOWS2, explaining 41% of the variance; DO was positively correlated with ORP but negatively correlated with NO₂⁻-N in DOWS3, explaining 84% of the variance. These models could well fit the DO concentration in DOWS3 during stage 2 and describe the potential influence well. These results further indicate that pH, ORP, DIN, NO₃⁻-N, and NO₂⁻-N have a significant impact on DO concentration.

The benthic release of DIN may stimulate primary production and oxygen consumption in aquatic systems [53]. Therefore, a pathway analysis was conducted to investigate the influence of DIN on DO. The results of the path analysis are shown in Figure 7. According to the results of χ², p, and CFI values, the path model for DOWS1 during stage 1 was acceptable. Overall, when pH, NO₃⁻-N, and NO₂⁻-N were included in the model, 42.6% of DIN variation was considered. When ORP and NO₂⁻-N were included in the model, 23.0% of DIN variation was considered. The path coefficients from DIN to NO₃⁻-N and from NO₃⁻-N to DO were positive, indicating that DIN indirectly promoted DO by increasing NO₃⁻-N (total effect = 20.6%). Based on the positive correlation between DIN and NO₂⁻-N combined with the negative correlation between NO₂⁻-N and DO, it can be inferred that DIN may promote the consumption of a large amount of DO by promoting NO₂⁻ oxidation, therefore maintaining low DO concentration (total effect = 19.6%) [54].

4. Conclusions

The main conclusions can be summarized as follows:

• In DOWS, sediment is a potential source and transformation site of pollutants. Sediment can not only adsorb and precipitate pollutants but also release and transform them.
• SWR is positively correlated with DO in the overlying water and negatively correlated with COD, directly affecting the N release flux from the sediment. N-containing compounds in sediment can be transformed into various forms of N through microbial reactions and released into the overlying water.
• Shortening HRT may lead to weakened denitrification capacity in DOWS.
• DIN promotes NO₂⁻ oxidation—which consumes DO in the overlying water—maintaining DO concentration at a lower level.
• Thus, the SWR and HRT jointly determine the outcome of the overlying water quality. In the future, greater attention should be paid to the roles of SWR and HRT in nutrient-release processes from sediments. This study can provide a theoretical basis for controlling the reversion of black-odorous water bodies.