The North Canal water system originates from the Southern foot of Yanshan Mountain in Changping District, Beijing, and flows through Beijing, Langfang, Hebei Province and Tianjin, successively (Yang et al., 2012). Shahe Reservoir is an important node located in the source area of the NorthCanal (Figure 1). The drainage area of Shahe Reservoir is about 1,125 km², of which the mountain area accountsfor about 75%. Shahe Reservoir is a typical channel-type reservoir in the upper reaches of the North Canal. The water in the Shahe Reservoir area has a long residence time and is a stagnant water body (Yu et al., 2012). The issue of eutrophication in reservoirs has garnered significant attention. Nutrients stored within reservoir sediments posed potential risks of release, thereby serving as a source of endogenous pollution within the reservoir. (Sun et al., 2021).

Based on their topographical characteristics, three sediment sampling points were set up within the Shahe Reservoir study area (Figure 1). In September 2018, Three columnar sediment samples were collected at sampling points S1,S2, and S3 using a mud core sampler to analyze the vertical distribution characteristics of nutrients and pathogens. (r = 50 mm, h = 60 cm).

The sediment columns were layered at 2 cm intervals, and the layered samples were freeze-dried (Model FD-1A-50 freeze dryer, Beijing Boyikang Experimental Instrument Co., Ltd.), crushed with a glass rod to remove impurities such as gravel, shells, and animal and plant residues, ground with a mortar, and passed through a 100-mesh sieve before analysis. Meanwhile, the samples obtained by the Peterson mud harvester were mixed and put into a 50 mL centrifuge tube, centrifuged at 4,000 rpm for 20 min to obtain interstitial water, and stored at −4 °C. The overlying water was sucked from the gravity column mud extractor by siphon method, put into a 50 mL centrifuge tube, and stored at −4°C (Sun et al., 2022).

Determination of nutrient indicators in interstitial water includes ammonia nitrogen (NH₄ ⁺-N), total nitrogen (TN), orthophosphate (PO₄ ^3--P) and total phosphorus (TP) (Sun et al., 2021): where TN: interstitial water is diluted to below 4 mg L⁻¹, and after potassium sulfate digestion (Alfa Aesar,United Kingdom), UV spectrophotometer colorimetry is used; TP:The interstitial water is diluted with raw water to below 1.2 mg L⁻¹ and after potassium sulfate digestion, molybdenum antimony anti-spectrophotometry is used. NH₄ ⁺-N: After the interstitial water was filtered through a 0.45 µm filter membrane and diluted to diluted to below 2 mg L⁻¹, the Nessler reagent spectrophotometry was used; PO₄ ^3--P: After the interstitial water was filtered through a 0.45 µm filter membrane and diluted to below 1.2 mg L⁻¹, the molybdenum antimony was used to Anti-spectrophotometry; the instruments used for the measurement are UV-Vis spectrophotometers (TU-1901, Beijing Puxi General Instrument Co., Ltd.)

Based on metagenomic classification and sequencing, the PCR products of 16S rRNA V4 regions were determined, and the microbial community structure in each sample was analyzed. The PCR for each sample was repeated three times before they were mixed. For this, the PCR products were recovered using gel, and the PCR products from different samples were mixed in equal amounts for library construction and sequencing; library construction and sequencing were completed by related sequencing companies, and the sequencing platform was Illumina Miseq × 250. For Miseq paired-end sequencing data, the primer adapter sequence (TGG AAT TCT CGG GTG CCA AGG AAC TC) needed to be removed first, and then the paired reads were merged into a sequence according to the overlap relationship between paired-end reads. Samples were then identified and distinguished according to the barcode tag sequence. Finally, quality control filtering was performed on the samples to ensure valid data for each sample (Pawlowski et al., 2022).

Following guidance on relevant standards for pathogens in surface waters from the United States Environmental Protection Agency, the European Union, and the World Health Organization, this study selected the typical pathogens Escherichia coli and ENT for analysis, using their gene copy numbers (DNA copies·g⁻¹) to represent their corresponding content in the sediment, and their proportion (%) in 16S rRNA to represent their abundance.

SPSS 25.0 software was used to analyze the correlation between the pathogens and Nutrients in interstitial water. The vertical spatial distribution of nutrients in the interstitial waterand the absolute content of pathogens(copies·g⁻¹) were analyzed by Origin 2017. The heat map of the microbial community structure in the sediment was constructed using Heml 1.0 (http://hemi.biocuckoo.org/down.php). The R language ade4 package (https:// www.R-project.org/.12c) was used to perform noise reduction analysis on operational taxonomic units (OTUs) in the community structure. The average abundance of OTUs in all samples were required to be higher than 0.01%. The OTUs after noise reduction analysis were used for subsequent analysis.

The spatial distribution characteristics of NH₄ ⁺-N and TN at the SWI are shown in Figure 2. At the upstream of the reservoir (S1), the central area of the reservoir (S2), and the downstream of the reservoir (S3) in the overlying water, NH₄ ⁺-N had the same vertical change trend as TN, and its concentration did not change. The concentrations of NH₄ ⁺-N and TN in S1 were 3.20–90.35 and 5.85–95.75 mg·L⁻¹, respectively, with average values of 41.77 ± 23.21 and 44.60 ± 22.86 mg·L⁻¹, respectively. The concentrations of NH₄ ⁺-N and TN in S2 were 9.11–86.35 and 9.32–86.40 mg·L⁻¹, respectively, with average values of 53.18 ± 25.22 and 54.02 ± 25.19 mg·L⁻¹, respectively. The concentrations of NH₄ ⁺-N and TN in S3 were 4.45–57.45 and 6.83–77.30 mg·L⁻¹, respectively, with mean values of 43.09 ± 16.25 and 47.49 ± 17.71 mg·L⁻¹, respectively. With the increase in depth from the SWI downward, the oxygen content in the sediment decreased, the anaerobic environment became conducive to the ammonification of organic nitrogen, and ammonia consumption weakened, resulting in the accumulation of NH₄ ⁺-N in sediments (Shen et al., 2020), These results showed that the concentration of NH₄ ⁺-N increases with depth.

The spatial distribution characteristics of PO₄ ³⁻-P and TP at the SWI are shown in Figure 2. In the water overlying S1, S2, and S3, PO₄ ³⁻-P and TP exhibited the same vertical change trend and their concentration did not change. However, the concentration of PO₄ ³⁻-P and TP in the interstitial water of the sedimentary column at each sampling point gradually increased with the increase in depth. The concentrations of PO₄ ³⁻-P and TP in S1 were 0.15–6.65 and 1.31–17.10 mg·L⁻¹, respectively, with the average values of 2.60 ± 2.33 and 10.27 ± 4.42 mg·L⁻¹, respectively. The concentrations of PO₄ ³⁻-P and TP in S2 were between 1.10–7.35 and 1.48–17.50 mg·L⁻¹, respectively, with the average values of 3.19 ± 1.57 8.60 ± 4.26 mg·L⁻¹, respectively. The PO₄ ³⁻-P and TP concentrations in S3 were 0.26–13.45 and 1.00–26.90 mg·L⁻¹, respectively, with the mean values of 8.50 ± 3.79 and 17.05 ± 7.19 mg·L⁻¹, respectively. Related research (Gong et al., 2017) found that with the strengthening of the reducing environment inside the sediment, Fe³⁺ is continuously reduced into Fe²⁺, and iron-bound phosphorus is also released such that the release of phosphorus at the SWI increases. Under reducing conditions within sediment, Fe3+ is converted to Fe2+ through reductive dissolution. This process is driven by microbial activity and anaerobic conditions. As Fe3+ is reduced to Fe2+, iron-bound phosphorus is released into the surrounding water. (Ding et al., 2016). At the same time, related works discovered that the secretion of nitrate bacteria (like Nitrosomonas and Nitrobacter) can accelerate the dissolution of Fe³⁺, releasing the phosphorus adsorbed by Fe(OH)₃ (Jansson, 1986). Therefore, the orthophosphate content at 0–3 cm in the SWI environment showed a gradually decreasing distribution.

The release flux (F) of NH₄ ⁺-N at the SWI is shown in Table 2. Its correlation coefficient in S1, S2, and S3 exceeded 0.80. The exponential fitting curve was ideal. As can be concluded from the value of F, NH₄ ⁺-N at the SWI is released from the sediment in the interstitial water to the overlying water body and sediment is the source of NH₄ ⁺-N. The F of NH₄ ⁺-N ranged from 7.26 mg·m²·day⁻¹ to 12.72 mg·m²·day⁻¹. NH₄ ⁺-N in the sediment interstitial water enters the overlying water body with the concentration gradient under the action of molecular diffusion; therefore, NH₄ ⁺-N is potentially released from the sediment in the interstitial water to the overlying water (Nicholls and Trimmer, 2009). NH₄ ⁺-N had high F in the upper reaches because aquatic plants and phytoplankton utilize NH₄ ⁺-N first when using nitrogen nutrients (Melzer and Melzer, 2004). Most of the NH₄ ⁺-N in interstitial water originates from organic matter decomposition, which regenerates NH₄ ⁺-N (Meng et al., 2020). Moreover, human activities and industrial production in the upper reaches contribute abundant nitrogen nutrients to the upper reaches.

The F value (−0.26 mg m^-2·day⁻¹) of PO₄ ³⁻-P at the SWI was different from that of NH₄ ⁺-N. PO₄ ³⁻-P at the SWI in the upper reaches (S1) is absorbed by sediments. As can be concluded from the F values in S2 and S3, PO₄ ³⁻-P at the SWI is released from the sediment in the interstitial water to the overlying water body. The F values in S2 and S3 were 0.03 and 0.85 mg m^-2·day⁻¹, respectively. The migration and release of PO₄ ³⁻-P at the SWI is affected by many factors (Liu et al., 2018). Li found that under aerobic conditions, the millimeter-scale aerobic layer on the sediment surface and the diffusion boundary layer below the SWI prevent the release of phosphorus in the interstitial water to the overlying water (Li et al., 2016), Moreover, when the dissolved oxygen content in the overlying water decreases, the diffusion boundary layer or aerobic layer becomes thin or disappears, and phosphorus in the interstitial water is released to the overlying water body along the concentration gradient under the action of molecular diffusion (Wang et al., 2010).

The distribution of nutrient content at the SWI mainly shows that the nutrient content in the sediment was significantly greater than that in the overlying water body. The calculation of F in accordance with Fick’s first law revealed that only the F value of PO₄ ³⁻-P in S1 was negative (−0.26 mg/m²·day), that is, PO₄ ³⁻-P is deposited from the overlying water to the sediment. The release rates of NH₄ ⁺-N in S1, S2, and S3 (12.72, 7.26, and 11.97 mg/m²·day) were significantly greater than those of PO₄ ³⁻-P in S2 and S3 (0.03 and 0.85 mg/m²·day) because when aquatic plants and phytoplankton use nitrogen nutrients, they first use NH₄ ⁺-N. Moreover, most of the NH₄ ⁺-N in interstitial water originates from the decomposition of organic matter, which regenerates NH₄ ⁺-N. Meanwhile, for PO₄ ³⁻-P, the diffusion boundary and aerobic layer in the lower layer of the interface prevent the release of phosphorus in the interstitial water. Moreover, overlying water or the peak concentration of PO₄ ³⁻-P in the sediment surface inhibits the migration and release of PO₄ ³⁻-P in the lower layer to the upper layer. The F of PO₄ ³⁻-P was significantly lower than that of NH₄ ⁺-N.

The distribution characteristics of the microbial community structure at the SWI are shown in Figure 3A, the results of principal component analysis are shown in Figure 3B, which shows that the similarities and differences of the microbial community structure at the SWI are reflected in the overlying water bodies and sediments, respectively. Methyloparacoccus is a strictly respiratory Gram-negative aerobic genus that uses oxygen as the terminal electron acceptor, and its average relative abundance in the SWI of each sampling point was the highest (13.88% ± 6.30%). Its relative abundance was the highest (23.42%) at the 20 cm layer of the overlying water downstream of the reservoir (3# sedimentation column), and its relative abundance was the lowest (1.58%) at the −2 cm layer of the sediment. The genus Methylomonas has a wide distribution and is a chemoorganotrophic bacterium without strict requirements for inorganic nutrients. It may grow on natural substrates containing methane (Tikhonova et al., 2023). Most of its members are symbiotic or associated with heterotrophic bacteria that cannot oxidize methane, and isolating, purifying, and maintaining pure species of this genus are extremely difficult. Trace amounts of organic substances, such as amino acids and polypeptides, can inhibit their growth, and even some methane-oxidizing bacteria are inhibited by very small amounts of methanol. Some bacteria are strictly aerobic, using oxygen molecules as electron acceptors to oxidize methane gradually into alcohols, aldehydes, acids, and finally into carbon dioxide, and obtain energy through the monophosphate ribose or serine pathway (Khalifa et al., 2015). The average relative abundance of Methylomonas at the SWI was 13.60% ± 8.55% and was second only to that of Methyloparacoccus. It was highest (28.24%) at the 5 cm layer of the overlying water in S2 and lowest (2.12%) at the −2 cm layer of S1. The average abundances of Arenimonas and Methylophilus at the SWI at each sampling point were also relatively high, reaching 9.57% ± 3.46% and 7.94% ± 2.26%, respectively. The maximum values appeared at 10 (14.54%) and 15 cm (9.66%) of S1, and the smallest values were distributed at 5 cm of the upper water in S2 (5.79%) and the −2 cm layer of sediment in S3 (1.96%). The average abundances of Polynucleobacter, Phaeodactylibacter, Sulfurisoma, and Subdivision3_genera_incertae_sedis at the SWI at each sampling point exceeded 3.00%. The maximum abundances of Phaeodactylibacter and Sulfurisoma were 7.20% and 8.84%, respectively, and appeared at 20 cm above S1. The maximum abundance of Subdivision3_genera_incertae_sedis was 9.44% and was found at the −2 cm sediment layer in S3. The minimum abundances of Polynucleobacter and Phaeodactylibacter were 0.55% and 1.32%, respectively, and appeared at the −2 cm layer in S3. The minimum abundance of Sulfurisoma was 0.64% and appeared at the 10 cm layer of the overlying water of S3. The minimum value of Subdivision3_genera_incertae_sedis was 1.69% and was found at the 25 cm layer of the overlying water of S3.

The absolute abundance distribution of E. coli and Enterococcus at the SWI is shown in Figure 4. Given that the sediment layer in S3 was relatively high, the overlying water body was only sampled to 10 cm above the surface sediment. The figure shows that the absolute abundance ranges of E. coli in S1, S2, and S3 were 2.03 × 10⁵–4.72×10⁶, 2.41 × 10⁵–7.39×10⁶ and 2.20 × 10⁵–2.47 × 10⁷copies·g⁻¹, respectively. The absolute abundance of E. coli at the SWI negligibly differed. Notably, the absolute abundance of E. coli in the surface layer of the SWI at the −2 cm sediment layer in the three sedimentary columns was one order of magnitude higher than that in the overlying water body. Although the absolute abundance of E. coli at the SWI in the overlying water showed little difference, that at 20 cm above the SWI in S1 and S3 was slightly higher than that at other depths of overlying water.

The distribution characteristics of the absolute abundances of Enterococcus and E. coli at the SWI were not considerably different and were both higher in sediment than in the overlying water body. However, the absolute abundance of Enterococcus at the SWI was approximately two orders of magnitude higher than that of E. coli. Enterococcus species are known for their resilience and ability to survive in harsh environmental conditions, including low nutrient environments and fluctuating temperatures. They can form biofilms and persist in sediments for extended periods, enhancing their abundance at the SWI compared to E. coli, which may be less adapted to such conditions. The absolute abundance ranges of Enterococcus in S1, S2, and S3 were 6.30 × 10⁶–1.23×10⁹, 6.94 × 10⁶–8.81×10⁸ and 8.28 × 10⁶–1.43×10⁹, respectively. The abundance of Enterococcus in S3 was approximately 1.60 and 1.37 times that in the sedimentation columns in S2 and S1, respectively. The absolute abundance of Enterococcus at the SWI was not significantly different. Similar to that of E. coli, the absolute abundance of Enterococcus in the three sediment columns was approximately one order of magnitude higher than that in the surface −2 cm sediment in the SWI than in the overlying water.

Table 3 shows that E. coli, ENT, TN, and TP in the SWI in S1 and S3 had extremely significant positive correlations (p < 0.01) or significant positive correlations (p < 0 .05). Specifically, E. coli had a positive correlation with TN and TP (p < 0.05) at the SWI in S1, whereas ENT at the SWI in S1 had extremely significant positive correlations with TN and TP (p < 0.01). However, the correlation of Enterococcus with TN and TP in the SWI in S2 was not significant (p > 0.05). Enterococcus had extremely significant positive correlations with TN and TP (p < 0.01) at the SWI in S3. Relevant studies (Xiang et al., 2023) have found that nitrite can effectively improve denitrification efficiency and enrich bacteria during the nitrification and denitrification processes. Therefore, ammonia can be regarded as one of the pathogenic bacteria pollution source parameters.