Lateral Responses of Coastal Intertidal Meta-Ecosystems to Sea-Level Rise: Lessons from the Yangtze Estuary

Abstract

Understanding the spatiotemporal dynamics of coastal intertidal meta-ecosystems in response to sea-level rise (SLR) is essential for understanding the interactions between terrestrial and aquatic meta-ecosystems. However, given that annual SLR changes are typically measured in millimeters, ecosystems may take decades to exhibit noticeable shifts. As a result, the extent of lateral responses at a single point is constrained by the fragmented temporal and spatial scales. We integrated the tidal inundation gradient of a coastal meta-ecosystem—comprising a high-elevation flat (H), low-elevation flat (L), and mudflat—to quantify the potential application of inferring the spatiotemporal impact of environmental features, using China’s Yangtze Estuary, which is one of the largest and most dynamic estuaries in the world. We employed both flood ratio data and tidal elevation modeling, underscoring the utility of spatial modeling of the role of SLR. Our results show that along the tidal inundation gradient, SLR alters hydrological dynamics, leading to environmental changes such as reduced aboveground biomass, increased plant diversity, decreased total soil, carbon, and nitrogen, and a lower leaf area index (LAI). Furthermore, composite indices combining the enhanced vegetation index (EVI) and the land surface water index (LSWI) were used to characterize the rapid responses of vegetation and soil between sites to predict future ecosystem shifts in environmental properties over time due to SLR. To effectively capture both vegetation characteristics and the soil surface water content, we propose the use of the ratio and difference between the EVI and LSWI as a composite indicator (ELR), which effectively reflects vegetation responses to SLR, with high-elevation sites driven by tides and high ELRs. The EVI-LSWI difference (ELD) was also found to be effective for detecting flood dynamics and vegetation along the tidal inundation gradient. Our findings offer a heuristic scenario of the response of coastal intertidal meta-ecosystems in the Yangtze Estuary to SLR and provide valuable insights for conservation strategies in the context of climate change.

1. Introduction

Coastal ecosystems play pivotal roles in maintaining biodiversity, supporting fisheries, and providing essential ecosystem services, such as carbon sequestration and coastal protection, but they are highly sensitive and increasingly threatened by sea-level rise (SLR), which is a consequence of global climate change [1]. In intertidal environments, coastal estuarine wetlands are characterized by complex interactions between the intensities and frequencies of surface water movements (e.g., tidal activity) [2]. SLR would alter the hydrology, vegetation, land cover, and other ecosystem structures in coastal regions, causing many ecological and societal consequences due to hydrological and sediment inputs from rivers [3]. Specifically, SLR has the potential to alter lateral material fluxes in coastal and estuarine ecosystems by changing the frequency, duration, and depth of seawater inundation [4]. These changes would in turn impact the biota and the soils, which can lead to habitat loss, changes in species composition, and alterations in ecosystem functions [5]. Here, SLR dynamics and their impacts on coastal/estuarine ecosystems shall be explored in terms of the lateral flows of materials, species, and energy [6]. For example, it is crucial to understand and predict lateral mass exchanges (e.g., nutrient transportation) under different tidal activities and SLR [7], which serve as a critical knowledge base for assessing the impacts of global climate change in coastal/estuarine areas [8].

Coastal estuarine wetlands are important interfaces between marine, freshwater, and terrestrial ecosystems and are often regarded as transitional zones, providing us with an opportunity to assess the lateral impacts of SLR [9]. At the landscape scale, the meta-ecosystem framework has been proposed as a powerful concept for understanding spatial flows of energy, materials, and organisms across coupled ecosystem boundaries [10]. It offers an alternative method for indirectly exploring the interaction between terrestrial and aquatic systems that connect changes in sea levels for monitoring frequent tidal floods [11]. Importantly, the lateral response of coastal intertidal meta-ecosystems to SLR is reflected in several factors, such as variations in water, energy, carbon fluxes, and vegetation features [12]. For example, there was a significant correlation between the differences in remote sensing-based estimates of gross primary production (GPP) (∆GPP_RS) and the differences in tower-based GPP (∆GPP_EC) at both high- and low-elevation sites [13]. Previously, we demonstrated that the estimated results of hydrological lateral fluxes were in good agreement with field measurements of reciprocal lateral carbon and nitrogen flows in a Yangtze River coastal marsh [14]. These studies have demonstrated the advantages of estimating hydrological lateral fluxes, making them suitable for evaluating the spatiotemporal changes in coastal intertidal meta-ecosystems due to SLR [15]. However, the extent of lateral responses at a single point appears to be limited by fragmented temporal and spatial scales, potentially leading to a misinterpretation of actual landscape connectivity when the entire ecological process is detected [16,17].

Although the response of coastal areas to climate change has been extensively studied via experimental observations and satellite-based approaches, the role of lateral shifts associated with SLR in regulating these responses has rarely been explored through the integration of temporal and spatial scales [18]. A major challenge here is that the lateral response to SLR may only become detectable or fully manifest on the timescale of landscape connectivity, which is typically measured in decades or even centuries within coupled ecological processes [19]. For this reason, modeling the response over a short period cannot fully capture the long-term impacts of SLR on coastal intertidal dynamics. Furthermore, over the long term, SLR can induce a series of ecological and environmental shifts (Figure 1), such as the following: (1) changes in the hydrological environment, including spatiotemporal variations in the extent of tidal flooding and the effects of flood inundation; (2) succession processes in plant communities, along with alterations in microbial and animal community composition and functionality; and (3) impacts on carbon and nutrient cycling (e.g., nitrogen) [20]. Overall, the effects of SLR may directly or indirectly influence local environmental connectivity, depending on long-term lateral dynamics [21].

Vegetation responses to SLR in coastal environments are mediated by many factors, such as variations in chlorophyll content, nutrient levels, water content, and underlying soil characteristics [22,23]. These factors affect surface reflectance in coastal wetlands through various mechanisms. Consequently, vegetation indices (VIs) provide simple yet practical measurements of vegetation features and development stages using different band combinations [24]. An increasing number of satellite-based observations and field investigations provide evidence, emphasizing the validation and scaling-up of lateral responses to SLR in coastal wetlands, which have become a subject of greater focus [25]. Owing to its dynamic geomorphology, minimal anthropogenic interference, and clear spatial transitions in elevation and vegetation, Dongtan in the Yangtze River Estuary offers a heuristic setting to investigate the lateral responses of intertidal meta-ecosystems to sea-level rise. Therefore, this study was designed with a hybrid approach by combining GIS techniques to develop a novel method for detecting spatiotemporal changes in the extent of tidal flooding and assessing the rapid response of coastal wetlands to SLR [26]. Rather than presenting a calibrated projection, a pair of eddy-covariance (EC) flux towers (H: high-elevation site; L: low-elevation site), located along a tidal inundation gradient with contrasting inundation frequencies, were installed to represent stages of inundation-driven ecological transformation. Although spatial gradients cannot fully substitute for long-term temporal observations, the natural sedimentation regime and vegetation zonation in Dongtan—shaped by variations in flooding frequency, salinity, and anaerobic conditions—offer a valuable analog for approximating lateral ecosystem shifts in response to SLR. Accordingly, time series moderate-resolution imaging spectroradiometer (MODIS) datasets have enhanced our understanding of global dynamics and processes and play a crucial role in developing validated, global, interactive models to track and interpret the effects of SLR [27]. By coupling tower-based measurements and remotely sensed data, we sought to perform the following: (1) construct a composite index that combines vegetation and water properties to quantify coastal environmental changes; (2) estimate the lateral response to SLR based on this composite index; and (3) explore the potential of the novel method to contribute to the development of a meta-ecosystem framework [28]. This study further refines the application of meta-ecosystem theory in coastal ecosystems, enhancing our understanding of long-term interactions between terrestrial and aquatic meta-ecosystems under the changing climate [29].

2. Materials and Methods

2.1. Description of the Study Area

This study was conducted in the largest wetland (ca. 230 km²) of the Yangtze River Estuary (31°25′–31°38′N, 121°50′–122°05′E) on the east shore of Chongming Island, also known as Dongtan (Figure 2) [30]. The marsh’s hydrology is characterized by a well-developed creek system and tidal water influx from the adjacent coast [31]. The tides follow a mixed, predominantly semidiurnal pattern, leading to frequent seawater inundation across most areas [32]. Additionally, spring and neap tides occur in cycles of approximately 15 days [33]. Tidal heights range from 4.6 to 6.0 m above sea level during extreme spring tides [34]. Owing to the extensive creek network, tidal flow rates are generally less than 1.0 m s⁻¹ but can reach 2.0 m s⁻¹ in the main channels, with a peak flow of 2.45 m s⁻¹ [35]. During our study, the vegetation exhibited clear zonation of dominant native plant communities, primarily Scirpus mariqueter and Phragmites [36]. Spartina, an invasive species, was introduced to the tidal flats and offshore sands of the Yangtze River Estuary in 1997 to accelerate natural land formation, successfully increasing sediment accretion [37]. These vegetation distributions align well with the spatial resolution of MODIS 500-m pixel data [38].

Two eddy flux towers were installed in August 2004 along the elevation gradient to provide continuous records of carbon and water for understanding the effects of SRL on the estuarine wetlands in the Yangtze River Delta and for validating remote sensing models [39]. In this study, the low-elevation site (L) was located 1.6 km from the sea wall and 1.1 km and 0.9 km away from the high-elevation site (H) and the mudflat, respectively. The relative elevation at the H site is 0.4 m higher than that at the L site and 1.2 m higher than that of the mudflat, representing a clear elevation gradient. These sites along an east–west belt were selected to capture the variability in tidal influence and its impact on ecological processes across different elevation zones (Figure 2).

2.2. Data

2.2.1. Eddy Covariance (EC) Data

Eddy covariance observation instruments were installed at each site to record data at a half-hourly resolution [40]. This setup included an open-circuit CO₂/H₂O infrared gas analyzer (LI7500A, LI-COR, Inc., Lincoln, NE, USA) [41]. These EC data were processed using EddyPro 7.0.9 software (www.licor.com/support/EddyPro/software.html (accessed on 13 February 2025)) following the workflow described in Reichstein et al. [42] and Chu et al. [43]. Additionally, a water content reflectometer (CS616, Campbell Scientific, Logan, UT, USA) was used to monitor the volumetric water content (VWC) of the soil at a depth of 5 cm [44]. Half-hourly carbon dioxide (CO₂) fluxes (F_CO2) were aggregated into daily values, which were subsequently averaged into 8-day means to align with the composited MODIS data [45].

2.2.2. MODIS Products

Since its launch in March 2000, MODIS has been widely used for estimating and evaluating vegetation characteristics and land surfaces [46]. NASA’s Land Processes Distributed Active Archive Center (LP DAAC) (https://lpdaac.usgs.gov/data_access (accessed on 26 October 2022)) offers a suite of standard MODIS data products, providing single-pixel datasets based on geographic location for specified periods [47]. This study focused on the period from 2001 to 2010, during which the selected cells overlapped with the footprints of our two EC tower stations and the mudflat platform. To estimate satellite-based vegetation indices, we used 8-day composite data (representing the best-quality daily reflectance within each 8-day period, coded with the first day of the period) and 500-m surface reflectance data (MOD09A1) to match the footprints of our two EC towers and the mudflat platform [48]. In our previous work, the Enhanced Vegetation Index (EVI) was selected due to its improved sensitivity to high biomass regions and reduced susceptibility to soil background noise, particularly in salt marshes with dense vegetation [36]. Multitemporal profiles at a 500 m resolution from MODIS pixels (e.g., the EVI) are commonly used to analyze vegetation phenology responses [49]. EVI is calculated as follows:

$$\mathrm{E}\mathrm{V}\mathrm{I}=2.5\times {\displaystyle \frac{{\mathsf{\rho }}_{\mathrm{N}\mathrm{I}\mathrm{R}}-{\mathsf{\rho }}_{\mathrm{R}\mathrm{e}\mathrm{d}}}{1+{\mathsf{\rho }}_{\mathrm{N}\mathrm{I}\mathrm{R}}+6\times {\mathsf{\rho }}_{\mathrm{R}\mathrm{e}\mathrm{d}}-7.5\times {\mathsf{\rho }}_{\mathrm{B}\mathrm{l}\mathrm{u}\mathrm{e}}}}$$

where ${\mathsf{\rho }}_{\mathrm{N}\mathrm{I}\mathrm{R}}$, ${\mathsf{\rho }}_{\mathrm{R}\mathrm{e}\mathrm{d}}$ and ${\mathsf{\rho }}_{\mathrm{B}\mathrm{l}\mathrm{u}\mathrm{e}}$ represent the partially atmospherically corrected (accounting for Rayleigh scattering and ozone absorption) surface reflectances for the MODIS near-infrared, red, and blue bands, respectively. The coefficients used in this study, such as the canopy background adjustment and the aerosol resistance term (which utilizes the 500 m blue band of MODIS to correct for aerosol influences in the red band), were assumed to be the same as those proposed by Huete et al. [50].

To assess the water content in a normalized manner, shortwave infrared (SWIR) spectroscopy is highly sensitive to moisture in both water bodies and plant tissues [51]. To detect water-column influences on the SWIR band, our previous study addressed that the Land Surface Water Index (LSWI), which is responsive to changes in soil and surface water content [45]. Building on this concept, LSWI was selected to evaluate water content responses resulting from tidal activities and is calculated as follows:

$$\mathrm{L}\mathrm{S}\mathrm{W}\mathrm{I}={\displaystyle \frac{{\mathsf{\rho }}_{\mathrm{N}\mathrm{I}\mathrm{R}}-{\mathsf{\rho }}_{\mathrm{S}\mathrm{W}\mathrm{I}\mathrm{R}}}{{\mathsf{\rho }}_{\mathrm{N}\mathrm{I}\mathrm{R}}+{\mathsf{\rho }}_{\mathrm{S}\mathrm{W}\mathrm{I}\mathrm{R}}}}$$

where ρ_NIR and ρ_SWIR are the reflectance values for MODIS bands 2 and 5, respectively.

Because of the rapid changes in vegetation cover and surface water content under SLR, a parameter combining vegetation characteristics and background water content was developed to represent the lateral responses of wetland ecosystems [52]. Therefore, two composite indices were selected to evaluate the integrated response to SLR. The ELD is defined as the difference between the EVI and LSWI (ELD = EVI − LSWI), whereas the ELR represents the ratio of the EVI to the LSWI [53].

$$\mathrm{E}\mathrm{L}\mathrm{R}={\displaystyle \frac{\sum \mathrm{E}\mathrm{V}\mathrm{I}}{\sum \mathrm{L}\mathrm{S}\mathrm{W}\mathrm{I}}}$$

Predictably, both the ELD and ELR values decrease with increasing SLR in coastal wetlands. Because daily index values are influenced by various uncertainties and environmental factors, the annual sum of each index was calculated and then averaged (arithmetic mean). This approach enhances the cumulative effect at the annual scale, making it easier to identify interannual variations in the indices. When the ELR and ELD are used as criteria to indicate the scenarios of successive SLR-driven stages, the ELR and ELD values from different years or along different elevation gradients can be compared. Approximately equal ELR and ELD values may represent similar ecological scenarios in terms of inundation and vegetation dynamics. Given that each MODIS pixel represents a 500 m × 500 m area, once full cross-comparisons are conducted, the spatial extent of each scenario and the temporal duration required for ecological transitions can be estimated. This approach does not aim to detect precise temporal trends but to infer potential SLR-induced transformation scenarios in time and space.

2.2.3. Relative Sea-Level Rise

As absolute elevation is not required for this study, relative elevation measurements are used instead. Relative sea-level rise (RSLR) rates referenced to a fixed local point on land were obtained from the National Oceanic and Atmospheric Administration (NOAA), as opposed to reflecting a combination of SLR and local vertical land motion [54]. Additionally, the University of Colorado’s Sea Level Research Group compares global sea level rates reported by different research organizations and examines key issues in the data (https://tidesandcurrents.noaa.gov/sltrends/sltrends.html (accessed on 17 November 2024)). In this study, reference data were obtained from the nearest tide gauge in Lusi, China (Figure 2). Furthermore, we calculated the linear trend for the study period from 2001 to 2010, following Saintilan et al. [55]. In addition, a field survey was conducted to measure the water depth at the mudflat platform during each tidal cycle in 2010 using a wooden ruler, with the depth recorded every 15 min. However, owing to the lack of field records on tidal elevation prior to the establishment of EC towers at various elevations across the wetland in 2004, reference data of tidal elevations were obtained from the tide tables of the nearest tidal station, Hengsha Station (Figure 2). The water tables have been measured at both EC towers since 2005. Additionally, eight-day average tidal elevations were calculated from data recorded at 0.5–1.0 h intervals to match the MODIS satellite pass schedule.

2.3. Identifying the Hydrological Status

Because of the distances (~1000 m) between the two EC towers, as well as between the low-elevation site (L) and the mudflat platform, the time required to complete a stage change and the average response rate could be estimated once all cross-comparisons were completed [56]. In addition, since the establishment of the EC flux towers in August 2004, vegetation surveys, including phenological observations, aboveground biomass measurements, and spectral reflectance assessments, were conducted every 2–4 weeks (Figure 3). Moreover, topographical surveys were conducted in 2005 and 2008 to measure elevation variations near the towers.

As transitional zones between marine and terrestrial ecosystems, coastal intertidal wetlands rely on surface water content as a key indicator for assessing their SLR [57]. Real-time tidal elevation data are essential for evaluating the validity and feasibility of using MODIS time series to detect hydrological dynamics [58]. Finally, we analyzed the temporal and spatial dynamics of tidal flood effects and empirically estimated the hydrological rate based on flood periods at each EC site and mudflat [59]. The hydrological status was defined as the ratio of flood points to the total number of points at each pixel for each year, as described by Zhao et al. (2009) [36] and Yan et al. (2010) [45].

2.4. Data Processing and Analysis

Figure S1 presents the comprehensive analytical framework of this study, which aims to investigate the lateral responses of coastal intertidal meta-ecosystems to SLR. All environmental characteristic data (e.g., aboveground biomass) were tested for homogeneity of variance and normality using Levene’s test and the Kolmogorov–Smirnov test, respectively. No data transformation was required to meet the assumptions of normality or homogeneity. Statistical analyses and model fitting were performed using R (R Development Core Team, 2017, version 3.4.3). Unless otherwise specified, the significance level was set at 0.01, and uncertainty (±) represents 99% confidence or quantile intervals. Differences between sites were assessed using two-tailed t tests, whereas pairwise mean comparisons were conducted using Duncan’s multiple range test. When significant differences were detected, rank ordering was determined using Tukey’s studentized range test. Linear regressions were performed using the ‘lm’ function to compare the observed diurnal maximum tide elevation in the mudflat creek with the predicted diurnal maximum tide elevation based on Hengsha tidal station data from 2010.

3. Results

3.1. Environmental Characteristics

The observed differences between the two EC tower sites provide valuable insights into how ecosystem characteristics may change over time due to SLR (

Table 1. Comparison of major ecosystem characteristics at the two study sites (H: high-elevation site; L: low-elevation site). The significance levels are represented by asterisks; **: p < 0.01 (n = 30).

Site	H	L
Reed
Relative elevation (m)	1.20	0.80
Elevation above mean sea level (m)	3.2	2.8
Distance to the sea wall (km)	0.5	1.6
Aboveground biomass (g m−2)	1170.0 ± 103.1	400.7 ± 86.8 **
Dominated species	Spartina alterniflora	Spartina alterniflora
	Phragmites australis	Phragmites australis
		Scirpus mariqueter
Soil total C (kg C m−3)	18.24	14.77
Soil total N (kg N m−3)	0.84	0.59
Leaf area index (m2 m−2)	4.70	1.59

). For example, the aboveground biomass is significantly lower at the L site (400.7 ± 86.8 g m⁻²) than at the H site (1170.0 ± 103.1 g m⁻²) (p < 0.01). In addition, both sites are dominated by similar vegetation species, including Spartina alterniflora and Phragmites australis. However, Scirpus mariqueter at the L site is common. The soil properties also differ between the two sites, with the L site exhibiting lower total soil carbon (14.77 kg C m⁻³) and nitrogen (0.59 kg N m⁻³) contents than the H site (18.24 kg C m⁻³ and 0.84 kg N m⁻³, respectively). Additionally, the response of ecosystem productivity to SLR based on the leaf area index (LAI) is lower at the L site (1.59 m² m⁻²) than at the H site (4.70 m² m⁻²).

Overall, our data indicates that the L site retains lower soil C and N levels throughout the soil profile than the H site does, with both parameters decreasing with depth. To better illustrate the tidal effects on the C and N contents in the soil profiles, we present the mean (standard deviation) values at the two study sites (Figure 4). Compared with the H site, the L site presents lower soil C contents across all depths (0–10 cm, 10–20 cm, 20–30 cm, and 30–50 cm). At the surface layer (0–10 cm), the soil C content at the H site peaks at 20.87 kg C m⁻³, whereas that at the L site remains at 15.69 kg C m⁻³. The difference between the two sites narrows with increasing depth. Notably, the soil C content in the surface layer (0–10 cm) at the L site is comparable to that in the bottom layer (30–50 cm) at the H site (Figure 4a). Like the soil C site, the L site presents consistently lower N content than the H site. At the 0–10 cm depth, the N contents at the H and L sites are approximately 1.17 kg N m⁻³ and 0.73 kg N m⁻³, respectively. This difference decreases with increasing depth, reaching similar values at depths of 30–50 cm. Consistent with the soil C content, the soil N content in the surface layer (0–10 cm) at the L site is also comparable to that in the bottom layer (30–50 cm) at the H site (Figure 4b).

3.2. MODIS-Based Vegetation Characteristics

Vegetation indices derived from MODIS datasets (e.g., EVI and LSWI) were run at an 8-day time scale to capture vegetation dynamics effectively during the study period. The annual average EVI/LSWI ratio (ELR) increased over the study period at both sites, with the H site showing consistent ELR values (Figure 5a). The mudflat (M) presented the lowest ELR values, which can be utilized as a reference to identify the SLR stage. Additionally, temporal and spatial comparisons of ELR values provide an effective means to identify and differentiate between SLR stages. For example, the ELR value at the L site in 2005 was comparable to that at the H site in 2001, whereas the ELR value of the mudflat (M) in 2008 was similar to that at the L site in 2001 (Figure 5a). Notably, the H site presented a stronger vegetation signal than did the L site and the mudflat (M), highlighting the interconnected effects of vegetation and hydrology on wetland dynamics under SLR.

The EVI-LSWI difference (ELD) also decreased from the H site to the L site in the same year, whereas a temporal increase was observed at the same site from 2001 to 2010 (Figure 5b). The mudflat (M) exhibited a consistent negative ELD value of less than −0.15, indicating the dominance of water signals over vegetation signals in this environment. The increasing ELR and ELD at both the H and L sites suggest an expansion or densification of vegetation, potentially driven by adaptive responses to increase competitiveness or ensure survival under SLR conditions.

3.3. Estimation of the Lateral Response

The interaction between vegetation and tidal activities appears to drive the spatial and temporal variations in the flood ratios across elevations within our systems (Figure 6). This compares the flood ratios across three areas: the high-elevation site (H), the low-elevation site (L), and the mudflat (M) (Figure 6a). The flood ratio remains consistently zero for the high-elevation site (H), indicating that few flooding events occurred during this period. The low-elevation site (L) shows minor fluctuations in its flood ratio, reflecting a slightly greater vulnerability to tidal influences. In contrast, the mudflat (M) has a significantly higher flood ratio, peaking close to 1.0, emphasizing its susceptibility to tidal inundation. There is a strong correlation between the observed diurnal maximal tidal elevation (TE) at the mudflat creek (M) and the values predicted at the Hengsha tidal station in 2010 (Figure 6b). The linear relationship suggests the feasibility of using regional tidal data to predict local tidal dynamics. Based on the simple linear regression between the observed and estimated TE, this relationship demonstrates strong predictability for assessing the impacts of SLR on tidal-inundated wetlands (Figure 6b).

4. Discussion

4.1. Uncertainties and Challenges in Lateral Ecosystem Shifts Under Rising Sea Levels

To better understand global ecosystem dynamics under future SLR, it is essential to develop innovative approaches that can capture the lateral responses of coastal intertidal ecosystems—particularly those driven by spatial elevation gradients and associated changes in hydrology, biogeochemical fluxes, and vegetation dynamics. Based on the relative sea level trend of 4.34 mm yr⁻¹ from 2001 to 2010, the L site can be regarded as an analog of the H site approximately one century (92 years) into the future. Similarly, the mudflat represents a more advanced stage—roughly two centuries (184 years) ahead—potentially reflecting the future condition of the L site under continued SLR. This spatial change allows us to predict how changes in relative elevation and proximity to the seawall may impact ecosystem properties, such as sediment accretion, vegetation dynamics, and hydrological alterations [60]. The vegetation density and composition in coastal wetlands are influenced primarily by hydrology [61]. Therefore, along the tidal inundation gradient, SLR alters hydrological dynamics, leading to changes in various ecosystem characteristics and, ultimately, shifts in ecophysiological processes [62]. These include reduced aboveground biomass, increased plant species diversity, decreased total soil carbon and nitrogen, and a lower LAI.

Elevation-induced variations in soil properties provide important insights into nutrient distribution patterns, forming the basis for assessing potential impacts of future SLR in coastal intertidal ecosystems [63]. The cross-comparison results (Figure 4) indicate that the elevation difference between the two sites aligns with the depth variation in the C and N contents observed in the soil profiles. Thus, we estimated the rate of SLR from the closest tide gauge, using spatial variations to assess future SLR impacts [64]. SLR is expected to alter ecological processes in coastal wetlands (e.g., the lateral flows of energy, materials, and organisms) and consequently affect ecosystem functions such as soil respiration and methane generation under the anaerobic conditions of wetland soils [65]. Sea waves carry significant amounts of C (e.g., dissolved and organic materials) [66]. There are many large modeling groups who explore lateral carbon exchange between sea and land [67,68,69]; however, effective methods for estimating the impact of water table fluctuations on C cycles have yet to be well developed [70]. Although methane measurements have been included in conventional eddy covariance techniques, the estimation of methane emissions from coastal wetlands remains limited in our study, as the CH₄ sensor (LI-7700) was only installed in 2010 [71,72,73]. Additionally, soil respiration—the second largest C flux in terrestrial ecosystems—is significantly influenced by soil moisture even in tide-influenced coastal wetlands. Moreover, the lateral N flux driven by tidal activities is a crucial component in quantifying the N budget in coastal wetlands [10]. Therefore, studying lateral C and N flows at terrestrial–aquatic interfaces is essential for understanding their impact on soil C and N dynamics under future SLR scenarios.

4.2. Extension and Limitations Along Elevational Gradients Under SLR

To better understand the ecological responses of coastal wetlands to SLR, we delineated three representative scenarios along the elevational gradient at the H site, L site, and adjacent mudflat (M). These scenarios illustrate different stages of inundation and ecological transitions: (1) High-elevation scenario: This scenario is characterized by relatively high elevation, which supports more complex plant communities and denser vegetation cover (Table 1). At this elevation, tidal influence is minimal, and the spectral response is predominantly governed by vegetation properties, with negligible contributions from soil background reflectance. (2) Low-elevation scenario: This scenario features a relatively lower elevation with frequent but short-duration tidal flooding. Consequently, surface water and soil moisture levels increase. The wetlands at this elevation support fewer plant species, with vegetation distributed sporadically. The area is marked by significant water coverage but lower vegetation density (e.g., LAI). (3) Mudflat scenario: This scenario is defined by mud-covered areas that are devoid of vegetation and experience frequent and prolonged submersion under seawater. Across these scenarios, a clear connection between vegetation cover and soil water content is evident.

The observed changes in the ELR and ELD provide valuable insights into the lateral responses of coastal intertidal meta-ecosystems to SLR. The next step is to develop an approach by analyzing each site individually and establishing thresholds or empirical constants, which could increase the broader applicability of the method [74]. The differences in vegetation structure among the three sites further indicate the spatial variability in ecosystem responses to hydrological and elevation gradients. As SLR progresses, advanced stages intensify tidal effects in coastal wetlands, resulting in higher soil moisture and lower surface reflectance.

Our results demonstrate that spatiotemporal variability in MODIS-based vegetation indices corresponds to tidal inundations. SLR stages can be identified by comparing the ELR or ELD values across temporal and spatial scales. This finding underscores the importance of incorporating tidal flood levels into the reevaluation of the role of coastal intertidal meta-ecosystems when the tidal flood effects of SLR are considered. We utilized surface reflectance and the combined application of the ELR and ELD as explanatory drivers without accounting for other influencing factors, such as lateral mass fluxes and landscape connectivity. The spatial patterns and temporal dynamics of SLR effects are shaped by not only soil properties and vegetation performance but also lateral flows and landscape connectivity [28]. Landscape connectivity can be characterized by using remote sensing products; however, further studies on lateral matter exchange are needed to explore the tidal impact on the response of vegetation to SLR, particularly within the context of the meta-ecosystem framework.

4.3. Spatial Simulations and Environmental Drivers of Lateral Responses to SLR

The combination of flood ratio data and tidal elevation modeling underscores the utility of spatial assessments to simulate the impacts of SLR. The different flood ratios between the elevation sites (H, L) and the mudflat (M) reflect how elevation gradients modulate flood dynamics. Furthermore, the strong predictive capacity of regional tidal data for local conditions supports the feasibility of evaluating future SLR impacts.

Notably, each stage of SLR represents a complex interplay of factors, including the background matrix and vegetation succession. Zhao et al. (2009) and Yan et al. (2010) reported that tidal activity has substantial effects on rapid vegetation dynamics and detected spatiotemporal changes in tidal inundation [36,45]. This study focused on soil properties and vegetation indices as key variables for predicting the lateral response to SLR. Notably, MODIS-based vegetation indices for land surfaces without vegetation cover may be influenced by plankton photosynthesis, which is transported inland during tidal activities [75]. Moreover, other potential environmental factors, such as seawater temperature, salinity, biological invasions, and the extent of ‘outwelling’, may also influence lateral flow in coastal intertidal meta-ecosystems [76]. Because seawater tends to exhibit a planar distribution, landscape connectivity plays a crucial role in influencing the lateral transport of excess nutrients to distant estuaries or the open ocean through tidal activities [69]. In addition to tidal effects, quantifying soil characteristics poses significant challenges in terms of the time and labor required for acquiring ground-based measurements. This technique is highly effective, as it highlights the importance of integrating spatial variability with predictive models to potentially apply MODIS-based vegetation indicators to infer ground elevation in tidal-active ecosystems under future SLR scenarios [77]. More auxiliary data can help explore these drivers to gain a deeper understanding of tidal-induced lateral transport and landscape connectivity, supporting our hypothesis that the influence of tidal activity diminishes with increasing distance from the sea [78].

5. Conclusions

This study provides essential insights into the spatiotemporal responses of coastal intertidal meta-ecosystems to SLR. By integrating the tidal inundation gradient, we explored the potential long-term impacts of SLR on the environmental characteristics of the Yangtze River Estuary. Our findings indicate that elevation differences significantly influence hydrological dynamics, resulting in shifts in vegetation structure (e.g., LAI), plant species diversity, and soil nutrient composition. The application of composite indices, specifically the ELR and ELD, enhances our ability to monitor ecosystem responses to SLR and predict future environmental shifts. These insights underscore the importance of advanced modeling techniques in refining a space-for-time approximation and informing adaptive management strategies to mitigate the impacts of SLR on coastal ecosystems.