2. Methods

2.1 Study area and sampling sites

The Yangtze River (also called Changjiang River), originating from the Qinghai-Tibet Plateau, is the third longest river in the world, with a total mainstream length of about 6300 km, a basin area of 1.8×106 km2 and an average annual discharge of 892 km3 (Yan et al. , 2010)(Fig. 1) . In China, the river is divided based on watershed boundaries into three reaches, the upper reaches (upstream of Yichang), middle reaches (between Yichang and Hukou) and lower reaches (downstream of Hukou) (Wang et al. , 2008). The lower reaches areas of Yangtze River have dense river networks, numerous lakes, extensive plains and dense cities, forming a prosperous industrial belt, but also facing ecological environment problems such as the increasing concentrations of nutrient in the watershed, frequent occurrence of harmful algal blooms in lakes and severe red tide in estuaries (Yi et al. , 2011; Qinet al. , 2010; Tang et al. , 2006; Liu et al. , 2018). In addition, the lower reaches of Yangtze River can be subdivided into three parts, the Anhui section, Jiangsu section and estuary, according to the provinces and tidal (Fig. 1) .
In this study, 44 surface water samples (0.5 m below the water surface) were collected from the river-estuary continuum of the Yangtze River coastal zone in July, 2018, to investigate the spatial changes of REGs and Upots (Fig. 1) . There were 19 (CJ1 ~ CJ19), 15 (CJ20 ~ CJ34) and 10 (CJ35 ~ CJ44) sites belonging to the Anhui section, Jiangsu section and estuary, respectively.

2.2 Sample collection and analysis

Surface water for NH4+ recycling experiments were collected in 1 L carboys. Water samples for nutrient concentrations (NO3, NO2, and NH4+) analyses were filtered through 0.7 μm fiberglass filters (Whatman GF/F) immediately following collection in the field. Water temperature, dissolved oxygen (DO), and pH were measured in situ using a multi-parameter water quality analyzer (YSI Professional Plus, 6600V2, USA). All samples were collected in triplicate and were immediately stored in a dark cooler. Total dissolved nitrogen (TDN), total dissolved phosphorus (TDP), ammonium (NH4+), nitrate (NO3), nitrite (NO2), and phosphate (PO43−) concentrations were analyzed in filtered samples (Jin and Tu, 1990). Concentrations of TDN and TDP were determined using the potassium persulfate digestion and spectrophotometric method (detection limits of 4 μmol N L−1 and 1 μmol P L−1 for TDN and TDP, respectively). NH4+ was determined using the nesslerization colorimetric method (detection limit 1 μmol N L−1). NO3and NO2 were determined using the phenol acid ultraviolet colorimetric method (detection limit 3 μmol N L−1) and N-(1-naphthyl)-ethylenediamine colorimetric method (detection limit 0.2 μmol N L−1), respectively. Total nitrogen (TN) and total phosphorus (TP) were determined on unfiltered water samples using the potassium persulfate digestion and spectrophotometric method (Jin and Tu, 1990). Particulate nitrogen (PN) was calculated as the difference between TN and TDN, and the standard deviation for PN was obtained using a propagation of error analysis. NOx was the sum of NO3 and NO2. Urea was measured using diacetylmonoxime reagent, and the detection limit was 0.04 μmol Urea-N L−1 (Mulvenna and Graham, 1992). Chl-aconcentrations, chemical oxygen demand (COD), and suspended solid (SS) were determined using standard methods (Jin and Tu, 1990). Dissolved organic carbon (DOC) concentrations were determined using TOC-V CPN (Shimadzu, Tokyo, Japan) analyzer at high temperature (680 °C) after being acidified with 10 μL of 85% H3PO4.

2.3 NH4+regeneration and uptake experiment

Water column REGs and Upots were determined using isotope dilution methods. Isotope dilution experiments are usually conducted with low trace amendment level (about 10% of ambient) (Glibert et al. , 1982). However, low NH4+ concentrations and fast NH4+ recycling rates in summer flood season of Yangtze River may lead to depleted NH4+ pool before the end of incubation (Blackburn, 1979), excess15NH4+(approximately 20 μmol N L−1), as the reaction product rather than as a potentially limiting substrate, was added at the beginning of incubation (Mccarthy et al. , 2007a). Excessive15NH4+ addition can promote NH4+ uptake rates, so the NH4+ uptakes obtained in this study were potential rates. On the other hand, because NH4+ is the end product rather than the substrate, excess additions will not affect regeneration rates (Blackburn, 1979).
Water from each site was enriched with 98%15NH4Cl and decanted into duplicate clear polystyrene culture bottles (70 ml; Coring) after thoroughly mixed. Initial samples were filtered through a rinsed 0.2 μm syringe filter immediately after enrichment and mixing for total NH4+ concentrations and NH4+-15N analysis. Bottles were incubated in a transparent bucket containing Yangtze River water to provide near-ambient light and temperature for 24 h. After incubation, final samples were collected in the same way as the initial samples. NH4+-15N was measured using NH4+ oxidation membrane inlet mass spectrometry (OX/MIMS) (Yin et al. , 2014).
Water column REGs and Upots were calculated using a modified isotope dilution method (Glibert et al. , 1982; Blackburn, 1979). The relative abundance of NH4+-15N (R ) is required to calculate the REGs and Upots, which can be calculated as:
\begin{equation} R=^{15}N/(^{15}N+^{14}N)\nonumber \\ \end{equation}
where 15N and 14N are the concentrations of NH4+-15N and NH4+-14N (μmol N L−1), respectively.
The REGs can be calculated as follows (Bruesewitz et al. , 2015), which was derived from the logarithmic equations of Blackburn (1979):
\begin{equation} \text{REG}=(R_{0}-R_{t})/t\ \times(C_{0}/R_{t})\nonumber \\ \end{equation}
where R0 and Rt are the relative abundances of NH4+-15N at the initial and finial point, respectively, and t is the incubation time (h). C0 is the initial concentrations of NH4+ (μmol N L−1). REGs (μmol N L−1 h−1) are absolutely positive values, indicating the actual regeneration rates of NH4+.
The Upots (μmol N L−1h−1) were calculated with the NH4+ concentrations change and regeneration rates (Bruesewitz et al. , 2015):
\begin{equation} \ U_{\text{pot}}=(C_{t}-C_{0}-REG\ \times t)/t\nonumber \\ \end{equation}
where C0 and Ct are the initial and finial concentrations of NH4+ (μmol N L−1).
CBAD characterizes internal NH4+recycling by representing the difference between measured potential NH4+ uptake rates and actual NH4+ regeneration rates in aquatic systems (Gardner et al. , 2017). Therefore, CBAD was calculated as:
\begin{equation} \text{CBAD}=\ \ U_{\text{pot}}-REG\nonumber \\ \end{equation}

2.4 Statistical analysis

Statistical analyses were performed using SPSS 22.0. One-way analysis of variance (one-way ANOVA) combined with the Independent-Samples T-test were used to evaluate statistically significant differences between group average values. Pearson correlation analyses were applied to analyze the relationship between NH4+recycling rates and environmental factors. Differences and correlations were considered statistically significant at p < 0.05.