Seasonal variability in lateral carbon exchange and dissolved organic matter exports from a salt-marsh tidal creek of the Yellow River Delta

Abstract

Coastal wetlands are key components of the global coastal carbon cycle. However, their carbon sink capacity cannot be fully assessed without considering lateral carbon exchange through tidal creeks. Here, we quantified seasonal and within−tide lateral carbon fluxes in a salt−marsh tidal creek in the Yellow River Delta, and characterized the sources and composition of exported dissolved organic matter (DOM) using optical indices, EEM−PARAFAC, and stable carbon isotopes. The annual carbon budget indicated an overall net lateral export of -662.06 g C m⁻² yr⁻¹, but both the magnitude and direction of exchange varied markedly among seasons. Net export was strongest in summer and autumn, whereas winter showed net import under some tidal conditions, particularly for DIC, indicating pronounced seasonal shifts in hydrodynamic connectivity and biogeochemical processing. Exported DOM was strongly influenced by terrestrially derived, highly humified material: low fluorescence index (FI), high humification index (HIX), and predominance of humic−like PARAFAC components pointed to a substantial marsh contribution, especially during warm seasons. Stable carbon isotopes further supported a greater contribution of marsh−derived DOC in summer and autumn. High−frequency observations revealed pronounced low−tide carbon pulses, with elevated DIC, dissolved CO₂, POC, and (to a lesser extent) DOC during ebb tide and at low water. Their relationships with turbidity and conductivity suggest that sediment resuspension and porewater drainage were important drivers of these short−lived export events. Overall, lateral exchange through tidal creeks is a major component of carbon cycling in this Yellow River Delta salt marsh; seasonal variability, DOM source characteristics, and low−tide pulsed export should therefore be explicitly incorporated into blue carbon assessments and coastal wetland carbon budgets.

1 Introduction

Coastal wetlands are integral to the global coastal carbon cycle. Although their total area is limited, salt marshes, mangroves, and tidal flats account for a disproportionately large share of carbon sequestration and storage in nearshore ecosystems (McLeod et al., 2011; Duarte et al., 2013). Their role in the blue carbon framework is largely based on their ability to accumulate organic matter and preserve it in waterlogged sediments over long periods. In tidal wetlands, frequent inundation and low oxygen availability in soils can slow aerobic decomposition and promote carbon retention in sediments (Pezeshki, 2001; Kirwan and Megonigal, 2013). This has contributed to the common view that tidal wetlands function as efficient long-term carbon sinks and should be included in regional and global carbon budgets (Windham-Myers et al., 2018). However, sediment burial represents only one component of wetland carbon cycling. Carbon in these systems is not only stored in sediments, but also continuously exchanged with adjacent waters (Bogard et al., 2020). Studies in tidal marshes have shown that greenhouse gas balance, ecosystem metabolism, and sediment carbon storage are closely linked to hydrological exchange across the marsh-estuary boundary (Weston et al., 2014). A wetland may accumulate carbon in its soils while exporting considerable amounts of dissolved and particulate carbon to nearby tidal creeks and estuaries. Burial alone therefore cannot fully describe its carbon function.

This lateral exchange has received increasing attention in recent years. Tidal flooding and ebbing, porewater drainage, groundwater discharge, and creek water exchange all transport carbon between marsh platforms and adjacent estuarine waters (Bauer et al., 2013; Tamborski et al., 2017; Ward et al., 2017). A growing number of studies suggest that lateral export can represent a substantial fraction of marsh carbon processing and, in some cases, may be comparable to or even greater than sediment burial (Santos et al., 2019). Dissolved inorganic carbon is often the dominant exported form, reflecting strong sediment respiration, carbonate-system reactions, and the flushing of carbon-rich porewaters. By contrast, dissolved organic carbon and particulate organic carbon tend to respond more strongly to vegetation, hydrology, sediment resuspension, and creek morphology (Wang et al., 2017; Schiebel et al., 2018). Studies in mangrove creeks have also shown that groundwater-derived carbon can make a substantial contribution to lateral fluxes, reinforcing the view that coastal wetlands are not simply passive storage sites, but active processors and exporters of carbon (Maher et al., 2013). This has important implications for blue carbon assessments. If lateral transfer is ignored, the net carbon sink strength of wetlands may be overstated, or at least incompletely characterized.

The significance of lateral fluxes is also linked to the broader question of the fate of terrestrial and wetland-derived carbon in estuaries. Large amounts of carbon are transported from land to the coastal ocean, yet the amount of clearly terrestrial organic matter preserved offshore is often lower than expected (Bianchi, 2011). This mismatch has been discussed as part of the estuarine and coastal carbon paradox, in which coastal waters can simultaneously support carbon burial, rapid mineralization, and CO₂ release (Cai, 2011). Salt marshes lie directly within this transition zone, so the carbon they export is likely to be substantially altered before it reaches the coastal ocean. Marsh-derived dissolved organic matter can differ markedly from riverine material in both its optical and chemical properties (Tzortziou et al., 2008). Once mobilized, it may be altered by microbial processing, photochemical reactions, and flocculation during salinity mixing (Schiebel et al., 2014). Sediments themselves are also active sources of dissolved carbon, because decomposition, desorption, and porewater exchange can all release carbon from the marsh surface and shallow subsurface into creek water. However, most of what we know about these processes comes from relatively mature marshes with stable geomorphology, dense vegetation, and moderate suspended-sediment conditions (Valiela, 2015). Those systems have been invaluable for understanding marsh carbon budgets, but they do not necessarily capture the full range of conditions found in rapidly accreting deltaic wetlands. In pioneer marshes, vegetation is often sparse, surface elevation is still evolving, and hydrodynamic forcing may play a stronger role in carbon exchange than litter production and decomposition alone (Li et al., 2021). High suspended-sediment concentrations may further influence carbon transport by reducing light penetration, modifying photochemical processing, and shifting the partitioning between dissolved and particulate phases (Schiebel et al., 2014).

These issues are particularly relevant in the Yellow River Delta, where coastal wetlands develop under conditions of strong sediment supply, rapid geomorphic change, and ongoing ecological succession (Tan et al., 2023). Pioneer salt marshes are widespread in this region and differ substantially from the mature marshes that dominate the existing literature (Lei et al., 2019; Tu et al., 2021). Vegetation cover is patchy, suspended-sediment concentrations are high, and water exchange between marsh surfaces and tidal creeks is strongly shaped by sediment dynamics and local topography (Gu et al., 2025). Under such conditions, the magnitude, composition, and sources of laterally transported carbon remain poorly constrained. In particular, it remains unclear how strongly marsh production and sediment processes contribute to exported carbon, how these contributions vary over complete tidal cycles and across seasons, and to what extent suspended sediments regulate carbon transport and transformation in this highly turbid setting. To address these questions, we investigated lateral carbon exchange in a pioneer salt-marsh tidal creek in the Yellow River Delta using continuous 25-h observations in four seasons. We combined hydrological measurements with stable carbon isotope analysis and excitation-emission matrix (EEM) spectroscopy coupled with parallel factor analysis (PARAFAC) to characterize both the magnitude of lateral carbon transport and the source characteristics of exported DOM. Specifically, we aimed to (1) quantify seasonal fluxes of DIC, DOC, and POC; (2) identify the source characteristics of exported DOM; and (3) evaluate the influence of suspended-sediment conditions on carbon transport and transformation. By focusing on a sediment-rich pioneer marsh, this study extends current understanding of lateral carbon exchange in coastal wetlands beyond the mature marsh systems that have received the most attention to date.

2 Materials and methods

2.1 Study area

The study was conducted in the Yellow River Delta National Nature Reserve near the estuary of the Yellow River (37°43′10″ N, 119°15′14″ E), Dongying City, Shandong Province, China. The study site is a mesohaline tidal-creek system draining a pioneer salt marsh dominated by the invasive species Spartina alterniflora. This species is widely distributed in the tidal wetlands of the Yellow River Delta and commonly forms the initial vegetation cover on newly accreted marsh surfaces. Mean salinity in creek water ranges from 5 to 15 PSU. Here, the pioneer salt marsh refers to a newly accreted marsh area colonized by early-stage S. alterniflora vegetation. The marsh surface was low and recently formed, with patchy vegetation cover and active sediment deposition, and had not yet developed the dense vegetation and stable soil profile typical of mature marshes in the region. To define the spatial extent and watershed boundary of the target tidal-creek system, high-resolution airborne LiDAR data (vertical accuracy: ± 10 cm) were acquired in May 2024 using a DJI M300 RTK unmanned aerial vehicle equipped with a Livox Mid-40 sensor. A 1-m-resolution digital elevation model (DEM) was generated and used to delineate the tidal-creek watershed. Because no major geomorphic disturbance or channel migration was observed during the study period, the LiDAR-derived DEM was assumed to be representative of the seasonal sampling campaign.

The creek boundary was set at the 2.0-m contour relative to the 1985 National Height Datum of China, corresponding to the mean high-water level during spring tides, in accordance with established approaches to tidal marsh delineation (Wang et al., 2016; Bogard et al., 2020). The 2.0-m contour was further confirmed to approximate the mean spring high-water level based on long-term records from the Dongying Tidal Gauge Station. Field observations during spring tides also showed that this contour approximated the regular inundation limit of the creek-marsh system. The watershed area defined by this boundary was 4,132 m². Real-time tidal height data from the Dongying Tidal Gauge Station were referenced to the same vertical datum to ensure consistency in hydrodynamic forcing (Figure 1).

Figure 1

2.2 Field sampling

Field observations and water sampling were carried out at the mouth of the primary tidal creek in four seasons: spring (April), summer (July), autumn (October), and winter (January). In each season, one representative spring tide and one representative neap tide were selected based on local tidal-amplitude indices. On each selected day, intensive time-series sampling covered one full lunar day (24 h 50 min), including two consecutive semidiurnal flood-ebb cycles and thus capturing the full range of tidal stages. Surface water samples were collected manually at hourly intervals from a fixed platform at a depth of 20 cm. This depth was chosen to represent the near-surface, well-mixed water layer during most tidal stages while minimizing direct interference from near-bed sediment resuspension. However, some influence of sediment resuspension during extreme low-tide conditions cannot be excluded. Concurrent in situ measurements included water depth, salinity, temperature, pH, turbidity, dissolved oxygen (DO), electrical conductivity (EC), and oxidation-reduction potential (ORP), all recorded at each sampling time point using a multiparameter water-quality monitoring system. Salinity, temperature, and EC were measured using an SX836 portable conductivity meter (Sanxin, China), and DO, pH, turbidity, and ORP were measured using a YSI EXO2 sonde. Sampling began at local low tide (± 15 min) to ensure balanced representation of both flood and ebb phases over the tidal cycle.

2.3 Laboratory analyses

Water samples for particulate organic carbon (POC) analysis were filtered under low vacuum (<50 kPa) through pre-combusted Whatman GF/F filters (450 °C for 5 h; nominal pore size 0.7 μm). The filters were acid-fumigated with 6 mol L^-1 HCl vapor for 24 h to remove inorganic carbon. After fumigation, the filters were re-dried at 60 °C for 48 h to remove residual HCl and then packed into pre-combusted silver capsules. POC concentrations were determined using a Vario MAX CN elemental analyzer (Elementar, Germany), with a detection limit of 0.1 μmol C and analytical precision of ±1.2% (n = 10).

For dissolved organic carbon (DOC) concentration analysis, 50-mL water samples were immediately filtered through pre-combusted 0.45-μm polycarbonate membranes, preserved with saturated HgCl₂ at approximately 0.02% (v/v), and stored at -20 °C until analysis. HgCl₂ was added to inhibit microbial alteration during storage. Its effect on DOC concentration was considered negligible because it contains no organic carbon and was added at a very low volume ratio. DOC concentrations were quantified using a Shimadzu TOC-VCPH analyzer in high-temperature catalytic oxidation mode (680 °C), with a detection limit of 0.1 mg L^-1 and analytical precision of ±1.5% (n = 12). For δ¹³C-DOC analysis, filtered water samples were processed separately from DOC concentration and optical samples. The samples were acidified to remove dissolved inorganic carbon before freeze-drying, and the recovered organic residues were then combusted and analyzed using an elemental analyzer coupled to an isotope ratio mass spectrometer (EA-IRMS). Because the creek water was brackish, the freeze-dried residues may still have contained dissolved salts. No separate desalting step was applied. Potential salt-related matrix effects therefore cannot be fully excluded. To reduce analytical variability, the carbon mass loaded into tin capsules was kept within the calibrated range of the EA-IRMS system. Procedural blanks and laboratory reference materials were analyzed together with the samples. The δ¹³C-DOC data were therefore used mainly to indicate relative seasonal changes in source contribution. Isotopic values are reported relative to the Vienna Pee Dee Belemnite (VPDB) standard. Dissolved inorganic carbon (DIC) was measured by coulometric titration using a Marianda VINDTA 3C system, with an analytical accuracy of ±0.5%. Dissolved CO₂ was determined by headspace equilibration followed by gas chromatographic analysis (Agilent 7890B, FID/TCD). Briefly, 11 mL of unfiltered water was introduced into a 22-mL glass vial, and the remaining headspace was filled with N₂. The vial was mechanically shaken for 30 min to achieve gas–liquid equilibrium. The equilibrated headspace gas was then withdrawn with a gas-tight syringe and injected into the gas chromatograph for analysis. The detection limit was 0.5 μmol L^-1, and analytical precision was ±2.1% (Maher et al., 2013). Water samples for CDOM and FDOM analyses were collected separately from the HgCl₂-preserved DOC concentration samples. Samples were filtered through pre-combusted GF/F filters and then through 0.22-μm membrane filters into pre-cleaned amber glass bottles. The filtrates were stored in the dark at 4 °C and analyzed generally within 48 h after collection. No HgCl₂ was added to samples used for UV-visible absorbance or EEM fluorescence measurements.

UV-visible absorbance spectra (200–800 nm, 1-nm resolution) were measured using a Shimadzu UV-2600 spectrophotometer with a 1-cm quartz cuvette. Excitation–emission matrices (EEMs) were acquired using a Horiba Aqualog fluorometer over excitation wavelengths of 220–600 nm and emission wavelengths of 280–650 nm at 5-nm intervals. Raw fluorescence data were corrected for Raman scatter, inner-filter effects, and instrumental bias following standard procedures (Murphy et al., 2013). The humification index (HIX) and biological index (BIX) were calculated using established methods (Huguet et al., 2009).

Parallel factor analysis (PARAFAC) was applied to the corrected EEM dataset using the DOM Fluor toolbox in MATLAB R2018b. A total of 192 EEM samples were used for model construction. Models with two to six components were tested. Model robustness was evaluated using split-half validation and core consistency diagnostics. The four-component model was retained because it provided the most interpretable representation of the main fluorescence features. However, the core consistency of this model was below 60%, indicating deviation from ideal trilinear structure. Therefore, the PARAFAC components were treated as optical descriptors of DOM composition rather than source-specific tracers. Fluorescent components were identified by comparing their excitation–emission peak positions with previously reported DOM components in aquatic and coastal wetland environments (Coble, 1996; Stedmon and Bro, 2008). No independent protein-like component was resolved by the PARAFAC model. This does not necessarily indicate the absence of protein-like fluorophores, but suggests that such signals were weak relative to the dominant humic-like and fulvic-like fluorescence in this dataset.

To estimate the relative contribution of marsh-associated DOC, a two-end-member stable isotope mixing model was applied based on the natural abundance difference between marine DOC and marsh-associated organic matter:

where f is the percentage contribution of marsh-associated DOC in tidal-creek water, δ¹³C_creek is the measured isotopic composition of DOC in creek water, δ¹³C_marine is the marine DOC end member (-23.7‰), and δ¹³C_marsh is the marsh-associated organic matter end member (-27.6‰). Both end-member values were determined from samples collected in this study. The marine DOC end member was obtained from adjacent coastal water samples. The marsh-associated end member was obtained from marsh-associated organic matter samples collected within the study area. A separate Yellow River terrestrial end member was not included because direct river-water influence at the sampling site was limited during the tidal-cycle observations, and the isotopic signal of river-derived terrestrial organic matter may overlap with that of marsh-associated organic matter. Without an additional tracer, a three-end-member model would be poorly constrained. The mixing model was therefore used to distinguish marine and non-marine/marsh-associated contributions. Because the estimated fractions are sensitive to end-member selection and no additional independent tracer was available to separate Yellow River-derived terrestrial DOC from marsh-associated DOC, the mixing-model results were interpreted only as relative seasonal indicators rather than exact source fractions. To more clearly evaluate the uncertainty associated with end-member selection, a brief sensitivity analysis was conducted by varying the marine and marsh δ¹³C end members within a plausible range and recalculating the corresponding marsh-associated DOC contributions.

2.4 Flux calculations

Lateral carbon flux in this study refers to horizontal carbon exchange between the tidal-creek marsh system and adjacent coastal waters. Negative values indicate net export from the marsh to coastal waters, and positive values indicate net import from coastal waters to the marsh. Hourly tidal prism volume (Vt, m³) was calculated from the DEM-derived topography and real-time tidal-height data. The tidal prism volume included both the inundated marsh platform and the tidal-creek channel within the delineated watershed boundary. First, both the DEM-derived marsh-surface elevation and the real-time tidal height were referenced to the 1985 National Height Datum. Second, for each 1-m² grid cell within the delineated creek boundary, inundation was defined when the instantaneous tidal height exceeded the local marsh elevation. Third, instantaneous tidal volume was calculated as the sum of the depth differences between tidal height and marsh elevation across all inundated cells. Hourly water flux (ΔV, m³ h^-1) was calculated as the difference between consecutive hourly tidal volumes, with positive values indicating net inflow and negative values indicating net outflow (Wang et al., 2016; Yao et al., 2022). This tidal prism approach has been widely applied in small tidal-creek systems with clear topographic boundaries and is appropriate for estimating lateral water and carbon exchange in marsh watersheds (Bogard et al., 2020). Lateral carbon flux (FC, g C m^-2 d^-1) was estimated as the product of hourly water flux and the corresponding carbon concentration, normalized to a 24-h basis to enable cross-seasonal comparison:

where S is the watershed area (4,132 m²), T is the sampling duration, and C_i is the carbon concentration at time i, expressed as g C m^-3 after conversion from the measured mg C L^-1 values. Fluxes for unmeasured tidal cycles were linearly interpolated between the measured spring-tide and neap-tide values. Annual lateral carbon fluxes were estimated by linearly interpolating fluxes between adjacent seasonal observations. This approach assumes gradual seasonal changes in lateral carbon exchange between the four sampling campaigns. Because the annual estimate was derived from seasonal observations rather than continuous year-round measurements, it was interpreted as a first-order annual estimate. No separate porewater flux term was included in the mass balance because flux estimates were derived from net tidal water exchange measured at the creek mouth. Therefore, any dissolved carbon entering creek water through porewater drainage and subsequently transported during ebb tide was implicitly included in the calculated lateral fluxes. The reported lateral fluxes therefore represent integrated tidal-creek exchange. To evaluate the uncertainty associated with tidal-state dependence and annual upscaling from discrete seasonal observations, annual lateral carbon fluxes were recalculated using spring-tide and neap-tide observations as two end-member seasonal scenarios. For each scenario, annual flux was estimated by linearly interpolating fluxes between adjacent seasonal observations using the same procedure as for the central estimate. The resulting spring-tide-only and neap-tide-only annual estimates were used to define a scenario-based annual range. Because the annual estimate was derived from four seasonal campaigns rather than continuous year-round measurements, it was interpreted as a first-order, observation-based estimate.

2.5 Statistical analysis

All statistical analyses were performed using IBM SPSS Statistics 25.0. Data normality and homogeneity of variance were evaluated using the Shapiro-Wilk test and Levene’s test, respectively. Non-normal datasets were log- or square-root transformed before analysis. If the transformed data still failed to meet the assumptions of parametric tests, the non-parametric Kruskal--Wallis test was used to assess differences among groups. One-way analysis of variance (ANOVA) with least significant difference (LSD) post hoc testing was used to evaluate differences in carbon concentrations and isotopic signatures among seasons and tidal phases. Pearson correlation analysis was used to examine relationships among salinity, dissolved oxygen, turbidity, and carbon fluxes.

3 Results

3.1 Seasonal dynamics of aqueous carbon concentrations

The concentrations of the measured carbon pools in the tidal creek varied significantly among seasons (p < 0.05; Table 1). Among these carbon pools, POC showed the strongest seasonal variability, reaching its highest concentration in autumn (5.36 ± 0.22 mg C L^-1) and its lowest in spring (2.73 ± 0.31 mg C L^-1). DOC varied within a relatively narrow range (2.22-2.79 mg C L^-1) and was generally higher in summer and spring than in autumn and winter. In contrast, DIC peaked in winter (9.33 ± 1.06 mg C L^-1), whereas dissolved CO₂ was highest in summer (2.30 ± 0.42 mg C L^-1). Hourly observations further revealed distinct tidal variations in carbon concentrations across seasons (Figure 2). DIC and dissolved CO₂ generally increased during ebb tide, and this pattern was more evident in summer and autumn than in winter and spring. By comparison, DOC varied over a relatively small range throughout the tidal cycle. POC showed stronger short-term fluctuations than DOC, particularly in summer and autumn, whereas its variability was weaker in spring.

Figure 2

Correlation analysis showed different relationships between carbon components and environmental variables (Figure 3). POC was positively correlated with water temperature and turbidity, whereas DIC showed the strongest positive correlation with electrical conductivity. DOC exhibited weaker correlations with the measured environmental variables than POC and DIC. Overall, these results indicate that the seasonal and tidal variability of particulate and inorganic carbon was more pronounced than that of DOC in tidal creek water.

Figure 3

3.2 Seasonal and tidal modulation of lateral carbon fluxes

Tidal heights and per-tide exchange volumes shown in Figure 4 were combined with instantaneous concentration measurements to calculate the per-cycle lateral carbon fluxes. Lateral carbon exchange between the tidal creek system and adjacent coastal waters varied markedly among seasons and tidal conditions (Figure 5). Negative values indicate net export from the marsh-creek system to adjacent coastal waters, whereas positive values indicate net import. Summer and autumn were characterized by persistent net carbon export under both spring and neap tides, whereas winter and spring showed shifts in exchange direction depending on tidal condition. The strongest total carbon export occurred during autumn neap tide (-40.39 g C m^-2 cycl^-1), while the largest net import was observed during spring neap tide (13.25 g C m^-2 cycl^-1). Component-specific fluxes differed among seasons and tidal states, POC and DIC dominated lateral carbon exchange during summer and autumn, whereas DIC largely controlled the direction of net exchange during winter and spring. Dissolved CO₂ consistently represented only a minor fraction of total lateral carbon flux. Annual integration of the seasonal observations yielded a first-order net lateral carbon flux of -662.06 g C m^-2 yr^-1. Using spring-tide and neap-tide observations as two end-member seasonal scenarios, the annual flux was estimated to range from -691.31 to -632.79 g C m^-2 yr^-1. This result indicates overall net carbon export from the tidal creek system to adjacent coastal waters, with DIC contributing the largest proportion of the annual integrated flux. Because this value was derived from four seasonal campaigns and linear interpolation between observations, the scenario-based range reflects uncertainty associated with tidal-state dependence and annual upscaling from discrete observations, but it does not fully capture uncertainty associated with unobserved short-term events or concentration and water-flux measurement errors.

Figure 4

Figure 5

3.3 Characteristics of dissolved organic matter

The fluorescence characteristics of DOM showed relatively limited variation among seasons and tidal states (Figure 6). FI values ranged from 1.42 to 1.70, with an overall mean of approximately 1.52. BIX values were generally between 0.7 and 0.9, whereas HIX values were mostly between 4 and 8 and showed a wider spread than FI and BIX. PARAFAC analysis resolved four fluorescent components from the EEM dataset (Figure 7; Table 2). Based on their excitation and emission peak positions, Cp1 (Ex/Em = 250/450 nm) and Cp2 (Ex/Em = 230/420 nm) were identified as humic-like and fulvic-like components, respectively, Cp3 (Ex/Em = 270/490 nm) was assigned as a long-wavelength humic-like component, and Cp4 (Ex/Em = 220/400 nm) as a terrestrially derived humic-like component. The fluorescence intensity results showed that Cp1 and Cp2 were the dominant components throughout the observation periods, whereas Cp3 and Cp4 were comparatively lower (Figure 8). Despite the relatively limited variation in bulk fluorescence indices, humic-like fluorescence signals were generally more pronounced during summer and autumn than during winter and spring, indicating a stronger contribution of humified DOM during the warm seasons.

Figure 6

Figure 7

Figure 8

Stable carbon isotope measurements of DOC also showed clear seasonal differences (Figure 9). Mean δ¹³C-DOC values were -25.21‰ in summer, -25.34‰ in autumn, -24.37‰ in winter, and −24.35‰ in spring. Based on a two-end-member mixing model using marine DOC (-23.7‰) and marsh-associated organic matter (-27.6‰) as end members, the estimated relative contribution of non-marine or marsh-associated DOC was higher in autumn (42.1%) and summer (38.7%) than in winter (17.3%) and spring (16.6%). These results indicate a seasonal shift in the relative isotopic signal of DOC, with stronger non-marine or marsh-associated influence during the warm seasons. This seasonal isotopic pattern was broadly consistent with the optical DOM results, as the more depleted δ¹³C-DOC values in summer and autumn coincided with stronger humic-like and terrestrially influenced fluorescence characteristics. A brief sensitivity analysis showed that perturbation of the end-member δ¹³C values changed the absolute source estimates by several percentage points, but the seasonal pattern remained unchanged, with higher non-marine or marsh-associated contributions in summer and autumn than in winter and spring.

Figure 9

4 Discussion

4.1 Seasonal variability in lateral carbon exchange

Coastal wetlands are widely recognized as important blue carbon ecosystems because of their capacity to fix, transform, and store substantial amounts of organic carbon at the land-sea interface (Duarte et al., 2005; McLeod et al., 2011). However, these systems do not function simply as closed carbon sinks. A considerable fraction of marsh-processed carbon can be transported laterally through tidal creeks and exported to adjacent estuarine and coastal waters, where it may be further transformed, remineralized, buried, or temporarily retained (Cai, 2011; Bauer et al., 2013). In this study, the Yellow River Delta salt marsh exhibited pronounced seasonal variability in lateral carbon exchange, with substantial shifts in both flux magnitude and net exchange direction. Although the annual integrated balance indicated overall net lateral carbon export, this net value masked strong seasonal contrasts. Net export was concentrated in summer and autumn, whereas winter showed a tendency toward import under some tidal conditions. This pattern indicates that carbon exchange between the marsh platform and adjacent creek water was highly dynamic rather than constant over the annual cycle. The observed seasonal differences were broadly associated with changes in hydrodynamic exchange and tidal connectivity. Seasons with stronger water exchange generally showed larger carbon fluxes, suggesting that lateral carbon exchange was influenced by the intensity of water movement through the creek network. This interpretation is consistent with the view that tidal inundation, creek drainage, and tidal pumping strongly influence material exchange in vegetated coastal wetlands (Wang et al., 2016; Santos et al., 2019). In the present study, the strongest net export occurred during the warm seasons, when water exchange was also more active (Figures 4, 5). These results suggest that carbon export was controlled not only by carbon concentration itself, but also by the extent to which tidal forcing could mobilize and transport dissolved and particulate materials from the marsh platform and creek margins. The positive relationship of POC with turbidity and temperature (Figure 3) further supports this interpretation, indicating that sediment resuspension and detrital transport were particularly important during periods of intensified exchange. Similar responses have been reported from other tidal-creek systems, where stronger tidal exchange enhances suspended sediment mobilization and particulate carbon export (Call et al., 2015; Schiebel et al., 2018).

The stronger export in summer and autumn was likely driven not only by hydrodynamic forcing, but also by seasonal shifts in marsh biogeochemical activity. Higher temperature generally stimulates plant productivity, belowground turnover, microbial decomposition, and the production of dissolved and particulate carbon substrates in vegetated coastal wetlands (Alongi, 2014; Ouyang et al., 2017). In addition, repeated tidal flooding and drainage can enhance flushing of porewater and organic-rich interstitial water from marsh soils into tidal creeks, thereby increasing the availability of dissolved carbon for lateral exchange (Maher et al., 2013; Chen et al., 2022). These coupled biological and physical processes provide a plausible explanation for the larger carbon export observed in the warm seasons. By contrast, winter was characterized by weaker exchange overall and net import under some tidal conditions. This seasonal reversal may reflect reduced marsh-associated carbon production and release during the cold season, allowing the influence of coastal water intrusion and mixing to become more pronounced in the creek system. The winter pattern of DIC is consistent with stronger marine influence on carbon dynamics during periods of reduced marsh activity. Low temperature can suppress microbial respiration and decomposition in marsh soils, thereby reducing in situ production and release of marsh-derived carbon. At the same time, if saline coastal water penetrates further into the creek system during winter tidal exchange, DIC concentrations in creek water may be increasingly influenced by mixing with saline coastal water. The positive correlation between DIC and conductivity observed in this study (Figure 3) is consistent with this mechanism and suggests that conductivity may have served, at least partly, as an indicator of saline-water intrusion. Similar associations between salinity, porewater exchange, and inorganic carbon transport have been documented in marsh environments elsewhere (Raymond and Hopkinson, 2003; Wang et al., 2016). Taken together, these results indicate that lateral carbon exchange in the Yellow River Delta cannot be represented by a single constant annual term. Instead, both the sign and magnitude of net exchange varied seasonally in response to the combined effects of tidal forcing, hydrological connectivity, and seasonal marsh metabolism.

These findings are important for evaluating blue carbon budgets in coastal wetlands. Current assessments still often emphasize vertical carbon burial and gaseous exchange, while lateral export is omitted or simplified (McLeod et al., 2011; Regnier et al., 2013). Our results show that such an approach may overlook a major component of ecosystem carbon function. In the Yellow River Delta salt marsh, the tidal-creek network acted as an active pathway for carbon redistribution between the marsh platform and adjacent coastal waters, and this pathway varied strongly through the year. Ignoring seasonal lateral fluxes could therefore overestimate the degree to which carbon is retained within the marsh itself. Because the annual estimate in this study was derived from seasonal spring-tide and neap-tide observations rather than continuous year-round measurements, it should be interpreted as a first-order, observation-based annual estimate. The scenario-based annual range derived from spring-tide-only and neap-tide-only interpolation further indicates that uncertainty in the annual budget is influenced by tidal-state dependence and sparse temporal coverage. Therefore, the annual value is most useful for indicating the direction and approximate magnitude of net lateral exchange rather than as a fully constrained annual carbon budget. Future work should include higher-frequency concentration and velocity measurements to better constrain short-duration peaks, temporal interpolation uncertainty, and measurement-related uncertainty in annual flux estimates. Nevertheless, the seasonal contrasts were sufficiently strong to demonstrate that lateral exchange represents a substantial and dynamic component of marsh carbon cycling.

4.2 Optical and isotopic signatures of exported DOM

The optical and isotopic characteristics of dissolved organic matter indicate that DOM exported from the Yellow River Delta salt marsh was strongly influenced by terrestrially derived, humified material rather than by freshly produced autochthonous compounds in creek water. This interpretation is supported by several lines of evidence. Across all seasons, fluorescence index (FI) values remained relatively low, suggesting a limited contribution from microbial or planktonic sources and a stronger influence of terrestrial organic matter. Humification index (HIX) values were comparatively high, implying that the DOM pool had undergone substantial prior humification and diagenetic alteration (Ohno, 2002; Huguet et al., 2009). PARAFAC analysis further showed that humic-like and fulvic-like optical components dominated the fluorescence signal, whereas the remaining components were comparatively lower in intensity (Figures 7, 8; Table 2). Given the relatively low core consistency of the PARAFAC model, these components are interpreted here as optical descriptors of DOM composition rather than definitive source-specific tracers. Together, these characteristics are consistent with DOM influenced by marsh vegetation, soils, and sediment-associated organic matter enriched in relatively refractory compounds (Coble, 1996; Fellman et al., 2010). The predominance of humic-like optical signals suggests that a substantial fraction of the exported DOM may have undergone prior transformation within marsh soils or sedimentary environments before entering the creek. In tidal wetland environments, vascular plant detritus is commonly subjected to prolonged anaerobic decomposition, selective preservation of resistant fractions, and repeated redox shifts associated with inundation and drainage. These processes tend to favor the formation and retention of humified dissolved and particulate carbon pools (Alongi, 2014). Marsh sediments are therefore not simply passive repositories of carbon, but active zones of carbon transformation that can release optically distinctive DOM during physical disturbance or tidal flushing. Previous work has shown that sediment resuspension and sediment-water interactions can liberate CDOM and DOC with strong humic-like signatures from marsh sediments (Tzortziou et al., 2008; Schiebel et al., 2014). In our study, humic-like fluorescence signals were stronger in summer and autumn, which also corresponded to the period of strongest lateral exchange. This seasonal coincidence suggests that warm-season tidal flushing likely mobilized a previously processed, terrestrially influenced DOM pool associated with marsh vegetation, soils, and sedimentary organic matter, rather than simply exporting newly produced dissolved organic matter from the water column.

Stable isotope evidence provides a complementary line of evidence for seasonal changes in the relative contribution of non-marine or marsh-associated DOC. Using marine DOC and marsh-associated organic matter as end members, the two-end-member mixing model suggested that the relative non-marine or marsh-associated DOC contribution was higher in autumn and summer than in winter and spring (Figure 9). A brief sensitivity analysis further indicated that although the absolute source estimates were influenced by end-member selection, the inferred seasonal shift toward stronger non-marine or marsh-associated influence during the warm seasons remained robust. This pattern is consistent with the fluorescence results and suggests that warm-season DOC export reflected not only a larger absolute flux, but also a stronger contribution from the marsh itself. In other words, seasonal enhancement of lateral DOC export was associated with a compositional shift toward more terrestrially influenced and marsh-derived material. Similar source shifts have been reported in other tidal wetlands where export intensity increases with tidal flushing and subsurface exchange (Santos et al., 2019; Webb et al., 2019). The agreement between isotopic and optical indicators strengthens confidence in this source interpretation and makes it less likely that the observed seasonal changes were driven solely by in-channel processing. Nevertheless, the relationship between δ13C-DOC and optical DOM properties should not be interpreted as a strict one-to-one source diagnostic, because both isotopic composition and fluorescence characteristics can be modified by mixing, degradation, and in situ processing during transport. Therefore, these proxies are best interpreted as complementary indicators of seasonal DOM source variation.

The present study did not directly track the fate of exported DOM after it entered adjacent coastal waters, and therefore cannot determine whether exported marsh-associated DOM was rapidly mineralized, temporarily retained, or incorporated into longer-lived coastal carbon pools. Even so, the optical evidence suggests that at least part of this material may have been relatively resistant to immediate degradation. Humified DOM is often less bioavailable than freshly produced DOM, although its persistence ultimately depends on multiple interacting processes, including microbial utilization, photochemical alteration, flocculation, oxidation, and burial (Jiao et al., 2010; Bianchi, 2011). In the Bohai coastal zone, high turbidity and reduced light penetration may limit photochemical breakdown and thereby prolong the residence time of humified DOM after export. This does not demonstrate long-term sequestration, but it suggests that lateral DOM export should not automatically be treated as an instantaneous carbon loss from the broader coastal system. From a blue carbon perspective, these findings imply that the carbon sink function of marshes may extend beyond sediment accumulation on the marsh platform itself (Bauer et al., 2013). Still, the most defensible conclusion from the present data is that exported DOC in this salt marsh was strongly influenced by pre-processed, humified, marsh-derived material, and that this compositional feature should be incorporated into future assessments of marsh carbon fate and connectivity with adjacent coastal waters.

4.3 Low-tide carbon release and implications for flux estimates

A notable feature of this study was the pronounced sub-tidal pulsing of carbon release during ebb and low-tide conditions. Time-series observations showed that DIC and dissolved CO₂ generally increased during ebb tide, while POC also displayed clear short-term variability, especially in summer and autumn (Figure 2). By comparison, DOC varied over a relatively narrow range throughout the tidal cycle. This pattern indicates that lateral carbon export was not governed solely by the total volume of water exchanged over a tidal cycle. Instead, short-duration tidal-stage dynamics played an important role in determining when carbon was released most intensely, consistent with previous observations of strong semi-diurnal and tidal-stage-dependent variability in tidal-creek carbon exchange (Call et al., 2015; Santos et al., 2019). Carbon export was therefore temporally concentrated, with low-tide drainage representing a disproportionately important window of carbon mobilization.

One likely explanation for this low-tide pulse is the combined influence of sediment resuspension and porewater drainage. As tidal water level fell, shallow-water conditions and drainage through the creek channel may have favored the entrainment of fine sediments and associated particulate organic matter. At the same time, falling water levels can alter hydraulic gradients between marsh soils and adjacent surface water, thereby promoting the release of porewater and shallow subsurface water enriched in DIC, DOC, and other dissolved constituents from creek banks and marsh-edge sediments. However, because porewater fluxes were not measured separately, their contribution cannot be quantitatively separated from the integrated tidal-creek exchange measured at the creek mouth. Similar porewater- and groundwater-mediated export pathways have been documented in salt marshes and mangrove creeks, where intertidal sediments function as major sources of dissolved carbon to tidal waters (Maher et al., 2013; Tamborski et al., 2017). In the present study, the relationships among carbon pools, turbidity, and conductivity (Figure 3), together with the coincidence between ebb-stage drainage and concentration increases (Figures 2, 4), are consistent with this interpretation and suggest that both physical disturbance and subsurface exchange contributed to the low-tide concentration peaks. The role of sediment resuspension appears to have been especially important for the particulate carbon pool. POC was positively correlated with turbidity, indicating that particulate carbon export was closely tied to the mobilization of suspended sediment and detrital particles. In salt marsh environments, resuspension can export not only particle-bound organic carbon, but also promote the release of dissolved compounds through leaching, desorption, or disturbance of surface sediment matrices (Schiebel et al., 2014, 2018). This mechanism may also contribute to short-term variability in dissolved organic signals, although DOC varied within a narrower range than POC in this study. In this sense, ebb-stage export reflected a suite of short-lived mobilization processes that intensified carbon transfer during marsh drainage. Comparable sub-tidal variability has been reported in other tidal-creek systems, where carbon concentrations and exchange rates vary strongly over semi-diurnal, spring-neap, and episodic drainage timescales (Santos et al., 2019; Yao et al., 2022). Our results show that low-water stages can act as focal points of carbon release in the Yellow River Delta marsh system, particularly during summer and autumn when within-cycle variability was most pronounced.

Recognition of this sub-tidal pulsing behavior has direct methodological implications. Sampling based on a single tidal stage, or at low temporal resolution, may fail to capture these short-duration but important export events. As a result, annual lateral carbon fluxes may be underestimated, particularly in marshes where a substantial fraction of export is concentrated during ebb or low-water periods (Bogard et al., 2020). This issue is especially relevant for hydrologically dynamic systems, where tidal asymmetry, spring-neap variability, and episodic drainage can strongly influence flux estimates (Windham-Myers et al., 2018; Yao et al., 2022). For studies aimed at annual carbon budgeting, these findings support the use of high-frequency observations or, at minimum, strategically distributed sampling across different tidal stages. Without such temporal coverage, flux estimates derived from sparse observations may smooth out the very peaks that contribute disproportionately to cumulative export. Beyond sampling design, the observed low-tide carbon pulses also have broader conceptual implications for carbon accounting in tidal wetlands. If substantial carbon release occurs during short, recurring drainage phases, then carbon exchange cannot be adequately represented by static water-quality measurements or by models that assume relatively uniform concentrations over time. Instead, lateral exchange must be understood as a process tightly coupled to tidal hydrodynamics and marsh drainage mechanics (Wang et al., 2017). This reinforces the argument that blue carbon budgets should extend beyond vertical sediment burial and atmospheric gas exchange to include physically mediated horizontal transport across ecosystem boundaries (Ward et al., 2017). In the Yellow River Delta salt marsh, the creek system functioned not simply as a passive drainage network, but as an active biogeochemical interface regulating carbon redistribution between land and sea.

5 Conclusion

This study shows that the tidal-creek system of the Yellow River Delta salt marsh is an active pathway for lateral carbon exchange, with strong variability across both seasonal and sub-tidal timescales. Although the first-order annual estimate indicated net lateral carbon export (-662.06 g C m^-2 yr^-1), the direction and magnitude of fluxes changed markedly with season. Net export was strongest in summer and autumn, whereas winter showed net import under some tidal conditions, especially for inorganic carbon. These patterns indicate that lateral carbon exchange in the marsh-creek system was jointly regulated by hydrodynamic connectivity, tidal forcing, and seasonal biogeochemical activity. The exported dissolved organic matter was strongly influenced by terrestrially derived and strongly humified material. Optical indices, PARAFAC components, and stable carbon isotopes consistently suggested a substantial marsh contribution, particularly during the warm seasons. High-frequency observations also revealed clear low-tide carbon pulses, with elevated DIC and dissolved CO₂ during ebb and low-water stages and pronounced short-term variability in POC. Overall, the results show that lateral exchange through tidal creeks is a significant component of carbon cycling in the Yellow River Delta salt marsh and should be explicitly considered in blue carbon assessments. Future work should further quantify porewater exchange, sediment mobilization, and the fate of exported carbon in adjacent coastal waters to better constrain the role of lateral fluxes in coastal carbon budgets.

References

• 1

  AlongiD. M. (2014). Carbon cycling and storage in mangrove forests. Annu. Rev. Mar. Sci.6, 195–219. doi: 10.1146/annurev-marine-010213-135020

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 2

  BauerJ. E.CaiW.-J.RaymondP. A.BianchiT. S.HopkinsonC. S.RegnierP. A. G. (2013). The changing carbon cycle of the coastal ocean. Nature504, 61–70. doi: 10.1038/nature12857

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 3

  BianchiT. S. (2011). The role of terrestrially derived organic carbon in the coastal ocean: A changing paradigm and the priming effect. Proc. Natl. Acad. Sci.108, 19473–19481. doi: 10.1073/pnas.1017982108

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 4

  BogardM. J.BergamaschiB. A.ButmanD. E.AndersonF.KnoxS. H.Windham-MyersL. (2020). Hydrologic export is a major component of coastal wetland carbon budgets. Glob. Biogeochem. Cycles34, e2019GB006430. doi: 10.1029/2019GB006430

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 5

  CaiW. J. (2011). Estuarine and coastal ocean carbon paradox: CO2 sinks or sites of terrestrial carbon incineration? Annu. Rev. Mar. Sci.3, 123–145. doi: 10.1146/annurev-marine-120709-142723

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 6

  CallM.MaherD. T.SantosI. R.Ruiz-HalpernS.MangionP.SandersC. J.et al. (2015). Spatial and temporal variability of carbon dioxide and methane fluxes over semi-diurnal and spring–neap–spring timescales in a mangrove creek. Geochim. Cosmochim. Acta150, 211–225. doi: 10.1016/j.gca.2014.11.023

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 7

  ChenX.SantosI. R.HuD.ZhanL.ZhangY.ZhaoZ.et al. (2022). Pore‐water exchange flushes blue carbon from intertidal saltmarsh sediments into the sea. Limnol. Oceanogr. Lett.7, 312–320. doi: 10.1002/lol2.10236

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 8

  CobleP. G. (1996). Characterization of marine and terrestrial DOM in seawater using excitation-emission matrix spectroscopy. Mar. Chem.51, 325–346. doi: 10.1016/0304-4203(95)00062-3

  • CrossRef
  • Google Scholar
  • View reference in article

• 9

  DuarteC. M.LosadaI. J.HendriksI. E.MazarrasaI.MarbàN. (2013). The role of coastal plant communities for climate change mitigation and adaptation. Nat. Clim. Change3, 961–968. doi: 10.1038/nclimate1970

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 10

  DuarteC. M.MiddelburgJ. J.CaracoN. (2005). Major role of marine vegetation on the oceanic carbon cycle. Biogeosciences2, 1–8. doi: 10.5194/bg-2-1-2005

  • CrossRef
  • Google Scholar
  • View reference in article

• 11

  FellmanJ. B.HoodE.SpencerR. G. M. (2010). Fluorescence spectroscopy opens new windows into dissolved organic matter dynamics in freshwater ecosystems: A review. Limnol. Oceanogr.55, 2452–2462. doi: 10.4319/lo.2010.55.6.2452

  • CrossRef
  • Google Scholar
  • View reference in article

• 12

  GuY.HanG.GengL.XieB.ZhangY.XieD.et al. (2025). Trade-offs between vegetation and sediment availability in shaping long-term tidal network morphodynamics. Front. Mar. Sci.12, 1574276. doi: 10.3389/fmars.2025.1574276

  • CrossRef
  • Google Scholar
  • View reference in article

• 13

  HuguetA.VacherL.RelexansS.SaubusseS.FroidefondJ. M.ParlantiE. (2009). Properties of fluorescent dissolved organic matter in the Gironde Estuary. Org. Geochem.40, 706–719. doi: 10.1016/j.orggeochem.2009.03.002

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 14

  JiaoN.HerndlG. J.HansellD. A.BennerR.KattnerG.WilhelmS. W.et al. (2010). Microbial production of recalcitrant dissolved organic matter: Long-term carbon storage in the global ocean. Nat. Rev. Microbiol.8, 593–599. doi: 10.1038/nrmicro2386

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 15

  KirwanM. L.MegonigalJ. P. (2013). Tidal wetland stability in the face of human impacts and sea-level rise. Nature504, 53–60. doi: 10.1038/nature12856

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 16

  LeiZ.ZhaoningG.QiweiW.DiandianJ.XingW.首都师范大学资源环境与旅游学院,北京 100048 College of Resources Environment & Tourism, Capital Normal University, Beijing 100048, Chinaet al. (2019). Wetland mapping of Yellow River Delta wetlands based on multi-feature optimization of Sentinel-2 images. Natl. Remote Sens. Bull.23, 313–326. doi: 10.11834/jrs.20198083

  • CrossRef
  • Google Scholar

• 17

  LiJ.ChenH.GuoK.LiW.FengX.LiuX. (2021). Changes in soil properties induced by pioneer vegetation patches in coastal ecosystem. CATENA204, 105393. doi: 10.1016/j.catena.2021.105393

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 18

  MaherD. T.SantosI. R.Golsby-SmithL.GleesonJ.EyreB. D. (2013). Groundwater-derived dissolved inorganic and organic carbon exports from a mangrove tidal creek: The missing mangrove carbon sink? Limnol. Oceanogr.58, 475–488. doi: 10.4319/lo.2013.58.2.0475

  • CrossRef
  • Google Scholar

• 19

  McLeodE.ChmuraG. L.BouillonS.SalmR.BjörkM.DuarteC. M.et al. (2011). A blueprint for blue carbon: Toward an improved understanding of the role of vegetated coastal habitats in sequestering CO2. Front. Ecol. Environ.9, 552–560. doi: 10.1890/110004

  • CrossRef
  • Google Scholar

• 20

  MurphyK. R.StedmonC. A.GraeberD.BroR. (2013). Fluorescence spectroscopy and multi-way techniques. PARAFAC Anal. Methods5, 6557–6566. doi: 10.1039/c3ay41160e

  • CrossRef
  • Google Scholar

• 21

  OhnoT. (2002). Fluorescence inner-filtering correction for determining the humification index of dissolved organic matter. Environ. Sci. Technol.36, 742–746. doi: 10.1021/es0155276

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 22

  OuyangX.LeeS. Y.ConnollyR. M. (2017). The role of root decomposition in global mangrove and saltmarsh carbon budgets. Earth Sci. Rev.166, 53–63. doi: 10.1016/j.earscirev.2017.01.004

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 23

  PezeshkiS. R. (2001). Wetland plant responses to soil flooding. Environ. Exp. Bot.46, 299–312. doi: 10.1016/S0098-8472(01)00107-1

  • CrossRef
  • Google Scholar

• 24

  RaymondP. A.HopkinsonC. S. (2003). Ecosystem modulation of dissolved carbon age in a temperate marsh-dominated estuary. Ecosystems6, 694–705. doi: 10.1007/s10021-002-0213-6

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 25

  RegnierP.FriedlingsteinP.CiaisP.MackenzieF. T.GruberN.JanssensI. A.et al. (2013). Anthropogenic perturbation of the carbon fluxes from land to ocean. Nat. Geosci.6, 597–607. doi: 10.1038/ngeo1830

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 26

  SantosI. R.WebbJ. R.MaherD. T.SandersC. J.LarkinR. (2019). Carbon outwelling and outgassing vs. burial in an estuarine tidal creek surrounded by mangrove and saltmarsh wetlands. Limnol. Oceanogr.64, 996–1013. doi: 10.1002/lno.11090

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 27

  SchiebelH. N.GardnerG. B.WangX.PeriF.ChenR. F. (2018). Seasonal export of dissolved organic matter from a New England salt marsh. J. Coast. Res.34, 939–954. doi: 10.2112/JCOASTRES-D-16-00196.1

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 28

  SchiebelH. N.WangX.ChenR. F.PeriF. (2014). Photochemical release of dissolved organic matter from salt marsh sediments. Estuaries Coasts38, 1692–1705. doi: 10.1007/s12237-014-9893-3

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 29

  StedmonC. A.BroR. (2008). Characterizing dissolved organic matter fluorescence with parallel factor analysis: a tutorial. Limnol. Oceanogr. Methods6, 572–579. doi: 10.4319/lom.2008.6.572

  • CrossRef
  • Google Scholar

• 30

  TamborskiJ. J.CharetteM. A.AllenC. T. (2017). Investigation of pore water residence times and drainage velocities in salt marshes using short-lived radium isotopes. Mar. Chem.196, 107–115. doi: 10.1016/j.marchem.2017.08.007

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 31

  TanK.SunD.DouW.WangB.SunQ.LiuX.et al. (2023). Mapping coastal wetlands and their dynamics in the Yellow River Delta over last three decades: Based on a spectral endmember space. Remote Sens.15, 5003. doi: 10.3390/rs15205003

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 32

  TuC.LiP.LiZ.WangH.YinS.LiD.et al. (2021). Synergetic classification of coastal wetlands over the Yellow River Delta with GF-3 full-polarization SAR and Zhuhai-1 OHS hyperspectral remote sensing. Remote Sens.13, 4444. doi: 10.3390/rs13214444

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 33

  TzortziouM.NealeP. J.OsburnC. L.MegonigalJ. P.MaieN.JafféR. (2008). Tidal marshes as a source of optically and chemically distinctive colored dissolved organic matter in the Chesapeake Bay. Limnol. Oceanogr.53, 148–159. doi: 10.4319/lo.2008.53.1.0148

  • CrossRef
  • Google Scholar

• 34

  ValielaI. (2015). The Great Sippewissett salt marsh plots—Some history, highlights, and contrails from a long-term study. Estuaries Coasts38, 1099–1120. doi: 10.1007/s12237-015-9976-9

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 35

  WangS. R.IorioD.CaiW.-J.HopkinsonC. S. (2017). Inorganic carbon and oxygen dynamics in a marsh-dominated estuary. Limnol. Oceanogr.63, 47–71. doi: 10.1002/lno.10614

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 36

  WangZ. A.KroegerK. D.GanjuN. K.GonneeaM. E.ChuS. N. (2016). Intertidal salt marshes as an important source of inorganic carbon to the coastal ocean. Limnol. Oceanogr.61, 1916–1931. doi: 10.1002/lno.10347

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 37

  WardN. D.ButmanD.RaymondP. A. (2017). Where carbon goes when water flows: Carbon cycling across the aquatic continuum. Front. Mar. Sci.4, 7. doi: 10.3389/fmars.2017.00007

  • CrossRef
  • Google Scholar

• 38

  WebbJ. R.SantosI. R.MaherD. T.FinlayK. (2019). The importance of aquatic carbon fluxes in net ecosystem carbon budgets: A catchment-scale review. Ecosystems22, 508–527. doi: 10.1007/s10021-018-0284-7

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 39

  WestonN. B.NeubauerS. C.VelinskyJ. D.VileM. A. (2014). Net ecosystem carbon exchange and the greenhouse gas balance of tidal marshes along an estuarine salinity gradient. Biogeochemistry120, 163–189. doi: 10.1007/s10533-014-9989-7

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 40

  Windham-MyersL.CrooksS.Moseman-ValtierraS.MegonigalJ. P. (2018). “ Tidal wetlands and estuaries,” in The Second State of the Carbon Cycle Report (SOCCR2): A Sustained Assessment Report, (Washington, DC: U.S. Global Change Research Program), 596–648. doi: 10.7930/SOCCR2.2018.Ch15

  • CrossRef
  • Google Scholar

• 41

  YaoH.MontagnaP. A.WetzM. S.StarykC. J.HuX. (2022). Subtropical estuarine carbon budget under various hydrologic extremes and implications on the lateral carbon exchange from tidal wetlands. Water Res.217, 118436. doi: 10.1016/j.watres.2022.118436

  • Pubmed Abstract
  • CrossRef
  • Google Scholar