Monitoring and Evaluation of Ecological Restoration Effectiveness: A Case Study of the Liaohe River Estuary Wetland

Abstract

The Liaohe River Estuary Wetland, located in Panjin City, plays a critical role in reducing pollution loads, maintaining biodiversity, and ensuring ecological security in China’s coastal regions, contributing significantly to the implementation of the land–sea coordination strategy. As key components of ecological restoration projects, monitoring and evaluating restoration effectiveness provide a reliable basis for decision-making and ecosystem management. This study established an innovative three-dimensional integrated monitoring and evaluation system combining satellite imagery, UAV aerial photography, and field sampling surveys, addressing the technical gaps in multi-scale and multi-dimensional dynamic ecological monitoring. Through systematic monitoring and the assessment of key indicators, including water environment, soil environment, biodiversity, water conservation capacity, and carbon sequestration capacity, we comprehensively evaluated the enhancement effects of ecological restoration projects on regional ecosystem structure, quality, and service functions. The findings demonstrated that the satellite–airborne–ground integrated monitoring technology significantly improved water quality and soil properties, enhanced soil–water conservation capabilities, and increased biodiversity indices and carbon sequestration potential. These results validate the scientific validity of ecological protection measures and the comprehensive benefits of restoration outcomes. The primary contributions of this research lie in the following: developing a novel monitoring framework that provides critical data support for decision-making, project acceptance, effectiveness evaluation, and adaptive management in ecological restoration; establishing transferable methodologies applicable not only to the Liaohe River Estuary wetlands, but also to similar ecosystems globally, showcasing broad applicability in ecological governance.

1. Introduction

With the increasing severity of ecological and environmental challenges, ecological restoration has been elevated to a national strategy and has become a central approach to advancing the development of ecological civilization [1]. The primary objective of ecological restoration is to rehabilitate ecosystems, restoring their ecological functions and enhancing human well-being, while addressing the interconnectedness of social and ecological systems [2]. In China, the General Office of the Ministry of Natural Resources, the General Office of the Ministry of Finance, and the General Office of the Ministry of Ecology and Environment jointly issued the “Guidelines for Ecological Protection and Restoration Engineering of Mountains, Rivers, Forests, Fields, Lakes, and Grasses (Trial)”. This document represents the first standardized framework with general principles to systematically guide ecological protection and restoration practices, grounded in the concept of mountains, rivers, forests, fields, lakes, and grasses as an integrated life community.

As one of the largest wetland reserves in China, the Liaohe Estuary Wetland not only plays an important role in ecological protection, but also provides a vital habitat for bird migration and plays a positive role in protecting regional biodiversity. Beyond its ecological functions, the reed industry and wetland tourism in the Liaohe Estuary have generated substantial economic benefits for local communities, becoming a key pillar of regional sustainable development [3,4]. However, human activities such as urban construction, agricultural development, fishery resource utilization, and industrialization, as well as natural factors such as climate change, have a significant impact on the water sources and area of wetlands, resulting in the gradual fragmentation of wetland landscapes [5,6]. With rapid socio-economic growth and population increase, wetland degradation has become increasingly severe, posing a major challenge to ecological sustainability [7,8]. Human activities not only have a profound impact on the provision of wetland ecosystem services, but also heavily rely on the diverse ecosystem services provided by wetlands to maintain socio-economic development [9,10,11]. Studies have shown that in the past 300 years, the wetland area in Europe, the United States, and China has been reduced by half, and some countries, such as the United Kingdom, Ireland, and Germany, have a wetland loss rate of more than 75%. The total loss area of global wetlands is equivalent to the land area of India [12]. With the increasing importance of wetland ecological functions, countries around the world have gradually increased their emphasis on wetland protection. Ecological restoration has become one of the key measures for wetland protection and has gradually become a hot research field [13]. The assessment of wetland ecological quality changes and their driving factors is crucial for wetland protection and sustainable management. Borja et al. [14] believe that it is necessary to integrate the knowledge of different components and select appropriate physical, chemical, and biological indicators when assessing the status of ecosystems.

Cheng et al. [15] assessed the health status of river ecosystems in the Haihe River Basin, China, using physical, chemical, nutritional, and macroinvertebrate indicators. Chen et al. [16] evaluated the ecological health of wetlands in the Beijing–Tianjin–Hebei region by selecting integrated comprehensive indicators, including water, soil, biology, landscape, and social factors. Su et al. [17] evaluated the ecosystem value and sustainability of the Liaohe Estuary Wetland by using emergy analysis and emergy indicators. Based on the emergy analysis of the LEW environment and economic system, some suggestions are provided for its sustainable development. However, these studies primarily focused on local or regional scales [18,19,20], or specific ecosystem types and characteristics [21,22,23], leaving a gap in comprehensive evaluations at the national scale. With the rapid development of remote sensing technology and geographic information systems [24,25], satellite remote sensing is widely used in monitoring interventions in large-scale and remote areas [26]. Studies have shown that the application of remote sensing technology in environmental research can provide accurate and reliable results at a lower cost and in a shorter time [27]. For wetland that has been repaired, the combination of long-term monitoring and comprehensive data analysis is usually used to scientifically evaluate the ecological effects and functional recovery of the restoration [28]. The normalized difference vegetation index (NDVI) can reflect the growth of vegetation. It is considered to be the best indicator for evaluating vegetation coverage and monitoring the ecological environment. It plays a critical role in assessing the impact of full closure on ecology [29,30]. In the Ebinur Lake basin, the quantitative evaluation of vegetation dynamics is used to analyze the evolution of drought, the changes in lake surface area, and the sustainable changes in vegetation, and then characterize the dynamic changes to the surface [31]. In the study of wetland vegetation change monitoring in the Qinghai–Tibet Plateau, scholars used the BFAST (breaks for additive seasonal and trend) algorithm to detect abrupt changes in the NDVI, and discussed the NDVI changes in dynamic wetland and their correlation with temperature, precipitation, and solar radiation for the first time [32]. In the study of wetland vegetation change and migratory bird distribution in Poyang Lake, scholars obtained the correlation coefficient between NDVI time series and different types of migratory birds by comparing and analyzing the spatial and temporal variation trend of NDVI time series and the distribution of wetland migratory birds from 1999 to 2009 [33]. Satellite remote sensing technology provides important support for vegetation growth monitoring and climate and anthropogenic change assessment by virtue of the advantages of multi-scale data. Remote sensing data, such as Sentinel-2, Landsat-8, and MODIS, with their differentiated spatial, temporal, and spectral resolutions, can more comprehensively reflect vegetation dynamics and provide a range of information for ecological environment monitoring [34].

In recent years, escalating global ecological challenges have driven the widespread adoption of air–space–ground integrated monitoring technology in ecological conservation and restoration research [35]. This comprehensive approach synergizes multi-source data from satellite remote sensing (space), unmanned aerial vehicles (UAVs), aerial platforms (air), and ground-based surveys (ground), enabling multi-scale and holistic ecosystem monitoring [36,37]. While such technology has been increasingly applied to wetland ecological assessments, significant bottlenecks persist in multi-modal data fusion and dynamic evaluation mechanisms. Empirical studies reveal that although satellite systems (e.g., GF-5 and Sentinel series) provide large-scale monitoring capabilities, optical image quality in cloudy or rainy regions (e.g., the Guangdong–Hong Kong–Macao Greater Bay Area) suffers from temporal discontinuity, necessitating UAVs and ground sensors for high-frequency data supplementation [38]. However, inherent heterogeneity in multi-source data—spanning spatial resolution (e.g., Landsat’s 30 m vs. Sentinel-2’s 10 m), temporal alignment, and spectral characteristics—severely compromises fusion efficacy, with reported 28% error rates in wetland inundation boundary delineation [39,40]. Furthermore, current inversion algorithms for critical wetland parameters (e.g., soil moisture and vegetation biomass) predominantly rely on empirical models. Sensitivity variations among parameters (e.g., leaf area index and chlorophyll content) and error accumulation mechanisms substantially degrade accuracy, as evidenced by the PROSAIL model’s limited performance in coastal wetland biomass estimation (R² = 0.55–0.71) [41].

Building on previous studies, this paper contributes to the field by integrating multiple approaches and indices, including water environment, soil environment, carbon sequestration capacity, biodiversity, and ecological remote sensing interpretation, and proposes a regional ecological restoration effect. A comprehensive “sky–ground–space” three-dimensional monitoring and evaluation system is proposed to determine whether the pre-restoration goals have been achieved. Specifically, remote sensing evaluation indices are calculated using remote sensing imagery and land cover data, enabling the assessment of ecological restoration projects from three key perspectives: water resources, water environment, and ecological functions.

The remainder of this paper is structured as follows: The Section 2 introduces the general situation of Liaohe Estuary wetland and the collection of monitoring data. In Section 3, experimental data processing and the selection and calculation of evaluation indexes are introduced. The Section 4 verifies the feasibility and superiority of the target method through the discussion of the evaluation results of the study area. The Section 5 makes a horizontal comparison with other similar studies, and clarifies the scientific significance and shortcomings of this paper. Finally, Section 6 concludes this study with a summary of the main findings.

2. Study Area and Data

2.1. Overview of the Study Area

The study area is located within the Liaohe River Estuary National Nature Reserve, approximately 30 km southwest of Panjin City, Liaoning Province. It spans a geographic range from 121°28′24.58″ to 121°58′27.49″ east longitude and from 40°45′00″ to 41°05′54.13″ north latitude, covering a total area of 80,000 hectares (Figure 1). The Liaohe River Estuary National Nature Reserve is a sedimentary plain at the lower reaches of five rivers: Liaohe, Hunhe, Taizihe, Raoyanghe, and Dalinghe. The terrain is flat, with an elevation ranging from 0 to 6.5 m. Situated at the confluence of the Liaohe River and Liaodong Bay, the area consists of tidal wetlands, where freshwater, rich in nutrients, mixes with seawater, creating an ideal environment for diverse biological reproduction. The dominant ecosystem in this region is wetland, primarily comprising reed marshes, river water bodies, and mudflats. Coastal wetland vegetation, such as reeds (Phragmites australis) and Suaeda salsa, is widely distributed. The reserve serves as a crucial habitat and migratory corridor for numerous rare species and supports a rich diversity of marine, terrestrial, and aquatic life. Among its endangered inhabitants are the red-crowned crane (Grus japonensis), black-headed gull (Chroicocephalus saundersi), and harbor seal (Phoca largha). These ecological attributes contribute to the area’s distinctive “Red Beach” landscape, a unique and vibrant coastal wetland ecosystem. The extraction of oil and gas resources, along with the construction of internal roads, not only occupies natural wetlands and destroys surface vegetation and wetland landscapes, but also poses a risk of oil and gas leakage during development. Such leaks can contaminate wetland soils and, through surface runoff, infiltrate nearby water sources, exacerbating pollution and severely impacting the wetland ecosystem.

2.2. Data Sources

This study utilized remote sensing interpretation and water environment monitoring data from two different time periods in the study area. The baseline period was set in 2020, prior to the implementation of the restoration project, while the monitoring periods covered 2021 and 2022, following the project’s execution. At the ecological restoration unit scale, remote sensing interpretation focused on three key aspects: ecosystem structure, ecosystem quality, and ecosystem service functions. The monitoring indicators are shown in

Table 1. Remote sensing monitoring data sources at the ecosystem restoration unit scale.

Data Type	Data Source	Remote Sensing Monitoring Index for Ecological Restoration Units
Sentinel-2 land use/land cover	ESRI	Landscape stability (spread, patch density, total edge contrast), habitat quality index, habitat connectivity index, landscape richness, landscape fragmentation, habitat irreplaceability index, water cultivation, net primary productivity, autotrophic respiration, total primary productivity
Sentinel-2 reflectance data	European Space Agency (ESA)	Leaf area index, biodiversity index, vegetation cover, water retention, soil conservation
Species distribution data	GBIF	Habitat irreplaceability index
Temperature data	ERA5-Land	Gross primary productivity, net primary productivity, autotrophic respiration
Rainfall data	ERA5-Land	Water conservation, soil conservation, net primary productivity, autotrophic respiration
Soil data	HWSD v1.2	Soil conservation, gross primary productivity, net primary productivity, autotrophic respiration
DEM elevation data	SRTM1	Soil conservation

. A combination of monitoring techniques, including satellite imagery, UAV aerial photography, and field surveys, was employed to conduct periodic assessments of the study area, forming a comprehensive multi-dimensional monitoring system integrating aerial, terrestrial, and satellite observations. Water environmental monitoring serves to evaluate aquatic ecosystem restoration effectiveness through the systematic assessment of 24 fundamental water quality parameters, including turbidity, conductivity, transparency, flow rate, and velocity, with triannual sampling cycles. Monitoring sections were strategically established in line with the following criteria: (1) at the hydrological boundaries of the study area, (2) within key conservation zones, and (3) in regions exhibiting abrupt water quality variation (e.g., wastewater discharge outlets). Stagnant zones and backflow areas were systematically excluded, with the preferential selection of sampling sites demonstrating stable hydrodynamic conditions and undisturbed substrate morphology.

3. Methods

Figure 2 presents the technical workflow of the proposed method. Grounded in regional ecological challenges, the characteristics of restoration measures, and ecological objectives, this study first gathered comprehensive data on the ecological environment and socioeconomic conditions of the study area. Building on this foundation, monitoring points were strategically established to assess water quality, soil conditions, biodiversity, and carbon sequestration capacity. Ecological environment surveys were conducted both before and after the restoration project, collecting plot survey data alongside meteorological and hydrological conditions, topography, hydrogeology, soil properties, vegetation characteristics, and land use information. To evaluate the effectiveness of ecological restoration at the unit level, analytical tools such as the InVEST model and landscape pattern analysis were employed.

3.1. Water Environment Evaluation Method

To evaluate the quality of the water environment, this study employs a combination of the single-factor water quality index method and the comprehensive water quality index method.

The formula for the single-factor water quality labeling index is expressed as follows:

$$P_i=K_i+{\displaystyle \frac{C_i-S_{ik}^{lower}}{S_{ik}^{upper}-S_{ik}^{lower}}}$$

(1)

$$P_{DO}=K_{DO}+{\displaystyle \frac{S_{DO,K}^{upper}-C_{DO}}{S_{DO,K}^{upper}-S_{DO,K}^{lower}}}$$

(2)

In the formula, P_i represents the single-factor water quality index of the i-th parameter; C_i is the measured concentration of the i-th water quality parameter; and K_i denotes the water quality category of the i-th parameter, determined by comparing the measured C_i with the “Environmental Quality Standards for Surface Water” (GB 3838-2002) [42]. The possible values for K_i are 1, 2, …, 6. $S_{ik}^{upper}$ and $S_{ik}^{lower}$ indicate the upper and lower limits of the concentration of the ith index in the water quality of the K_i category of the Environmental Quality Standard for Surface Water (GB 3838-2002) [42], respectively. Taking dissolved oxygen as an example, when the measured concentration of C_i = 5.2 mg/L, its K_i category = III ($S_{ik}^{lower}$ = 5.0, $S_{ik}^{upper}$ = 6.0); at this time, P_i = 3 + (5.2 − 5.0)/(6.0 − 5.0) = 3.2, which characterizes the water quality in the middle of Class III.

When the water body is classified as worse than Class V, the increasing trends of water quality parameters and DO (dissolved oxygen) are calculated using the following formulas:

$$P_i=6+{\displaystyle \frac{C_i-S_{i5}^{upper}}{S_{i5}^{upper}}}$$

(3)

$$P_{DO}=6+{\displaystyle \frac{S_{DO,5}^{lower}-C_{DO}}{S_{DO,5}^{lower}}}$$

(4)

The formula for the composite water quality labeling index is as follows:

$$P=\theta \overline{P}+(1-\theta )P_{max}$$

(5)

$$\overline{P}={\displaystyle \frac{{\displaystyle \sum _{i=1}^n{P}_i}}{n}}$$

(6)

In the formula, P represents the comprehensive water quality index; $\overline{P}$ is the average value of the single-factor water quality indices; and $\mathsf{\theta }$ is the weight controlling the influence of each water quality index, with a range of [0, 1]. In this study, $\mathsf{\theta }$ is set to 0.5; P_max is the maximum value among the n single-factor water quality indices. The value of P can be used to determine the water quality category and compare pollution levels within the same category. According to the classification criteria, when 1.0 ≤ P ≤ 2, the water body is classified as Class I; when 2.0 < P ≤ 3.0, the water body is classified as Class II; when 3.0 < P ≤ 4.0, the water body is classified as Class III; when 4.0 < P ≤5.0, the water body is classified as Class IV; when 5.0 < P ≤ 6.0, the water body is classified as Class V; and when P > 6.0, the water body is classified as worse than Class V.

3.2. Ecosystem Structure Evaluation Method

The evaluation of ecosystem landscape structure is based on indicators derived from the ESA WorldCover land cover dataset, produced by Impact Observatory. These maps are based on Sentinel-2 imagery from the European Space Agency, with a resolution of 10 m. Each map is a composite of LULC (land use and land cover) predictions for the entire year, classified into nine categories, to generate a representative snapshot for each year. The dataset achieves an average accuracy of over 75% (Figure 3).

The habitat quality index measures the rate of improvement in biological habitat quality within the study area, reflecting the suitability of the regional ecosystem for sustaining wildlife. Landscape fragmentation is primarily determined by the Patch Number (NP), which represents the total number of patches of a specific land use or land cover type in a landscape image. A higher NP value indicates greater fragmentation or dispersion of an ecosystem type, necessitating a combined analysis with the average patch area index for a more comprehensive assessment. Natural Habitat Connectivity evaluates the overall connectivity between key ecological spatial patches within the study area, serving as an indicator of ecosystem stability and ecological security. Landscape stability is assessed using three key metrics: landscape spread, patch density, and total edge contrast. A higher landscape spread, coupled with lower patch density and reduced total edge contrast, indicates greater landscape stability. Landscape richness is quantified using the Shannon diversity index, which captures the diversity of landscape elements or ecosystems in terms of structure, function, and temporal dynamics, reflecting the complexity and heterogeneity of landscape types.

3.3. Ecological Environment Quality Evaluation Method

The data used for ecosystem quality evaluation were derived from Sentinel-2 imagery collected between July and October in 2020 and 2022, with a temporal resolution of 5 days and a spatial resolution of 10 m. These data were obtained from the Google Earth Engine (GEE) remote sensing cloud platform. To assess ecosystem quality, three core indicators were utilized: vegetation coverage (FVC), leaf area index (LAI), and gross primary productivity (GPP). The specific process is as follows: the maximum value of ecological parameters for four vegetation types—forest, shrubland, grassland, and farmland—within each ecological function zone is taken as the reference value. Then, the ratio of the ecological parameter value for each vegetation type of ecosystem within the zone to its reference value is calculated, yielding the relative density of that ecological parameter for the zone. The closer the relative density is to 1, the more similar the ecological parameter for that pixel is to the reference value. The calculation formula is as follows:

$${\mathrm{RVI}}_{\mathrm{i},\mathrm{j},\mathrm{k}}={\displaystyle \frac{{\mathrm{F}}_{\mathrm{i},\mathrm{j},\mathrm{k}}}{{\mathrm{F}}_{\mathrm{maxi},\mathrm{j},\mathrm{k}}}}$$

(7)

In the formula, ${\mathrm{R}\mathrm{V}\mathrm{I}}_{\mathrm{i},\mathrm{j},\mathrm{k}}$ represents the relative density of the ecological parameter of the k-th vegetation ecosystem in the j-th region of the i-th year; ${\mathrm{F}}_{\mathrm{i},\mathrm{j},\mathrm{k}}$ represents the value of the ecological parameter of the k-th vegetation ecosystem in the j-th region of the i-th year; and ${\mathrm{F}}_{\mathrm{m}\mathrm{a}\mathrm{x}\mathrm{i},\mathrm{j},\mathrm{k}}$ represents the maximum value of the ecological parameter of the k-th vegetation type in the j-th region of the i-th year.

Following this method, reference values for vegetation coverage, leaf area index, and total primary productivity are determined for each region and vegetation type. The relative density values are then normalized to a range of 0 to 1 using the following formula:

$${\mathrm{x}}^{\prime }={\displaystyle \frac{\mathrm{x}-\min \left(\mathrm{x}\right)}{\max \left(\mathrm{x}\right)-\min \left(\mathrm{x}\right)}}$$

(8)

In the formula, ${\mathrm{x}}^{\prime }$ represents the normalized index after processing; $\mathrm{x}$ represents the original index.

Ecosystem quality, which reflects the overall health and condition of the regional ecosystem, is then computed using the relative densities of LAI, FVC, and GPP with the following equation:

$${\mathrm{EQI}}_{\mathrm{i},\mathrm{j}}={\displaystyle \frac{{\mathrm{LAI}}_{\mathrm{i},\mathrm{j}}{+\mathrm{FVC}}_{\mathrm{i},\mathrm{j}}{+\mathrm{GPP}}_{\mathrm{i},\mathrm{j}}}{3}}\times 100$$

(9)

In the formula, ${\mathrm{E}\mathrm{Q}\mathrm{I}}_{\mathrm{i},\mathrm{j}}$ represents the ecosystem quality in the j-th region of the i-th year; ${\mathrm{L}\mathrm{A}\mathrm{I}}_{\mathrm{i},\mathrm{j}}$ represents the relative density of the leaf area index in the j-th region of the i-th year; ${\mathrm{F}\mathrm{V}\mathrm{C}}_{\mathrm{i},\mathrm{j}}$ represents the relative density of vegetation coverage in the j-th region of the i-th year; and ${\mathrm{G}\mathrm{P}\mathrm{P}}_{\mathrm{i},\mathrm{j}}$ represents the relative density of total primary productivity in the j-th region of the i-th year.

According to the “National Ecological Status Survey and Evaluation Technical Specifications—Ecosystem Quality Assessment”, ecosystem quality is categorized into five levels, with the specific classification detailed in

Table 2. Ecosystem quality classification.

Level	Excellent	Good	Medium	Low	Poor
Ecosystem Quality	EQI ≥ 75	55 ≤ EQI < 75	35 ≤ EQI < 55	20 ≤ EQI < 35	EQI < 20

.

3.4. Ecosystem Service Function Evaluation Method

The data used for ecosystem function evaluation indicators include Sentinel-2 data, Esri 10m resolution land use type data, GBIF species distribution data, 30m Global Fine Surface Cover Product, ERA-5 precipitation data, and HWSD V1.2 soil data. In the ecosystem function assessment, four core indicators are selected: biodiversity maintenance, carbon sequestration, water conservation, and soil and water conservation.

For the evaluation of biodiversity maintenance, a comprehensive biodiversity maintenance index is used as the evaluation metric. First, the habitat irreplaceability index of the study area is calculated using the Marxan2.43 site selection model, and the protection priority and habitat irreplaceability index of each region are obtained through iterative calculations based on constraint conditions. The habitat irreplaceability index ranges from 0 to 100. By dividing this index by 100, it is normalized to a 0–1 scale. The final comprehensive biodiversity maintenance index is then obtained by adding the normalized habitat irreplaceability index to the biodiversity index.

The objective function for the iterative calculation in the Marxan model is expressed as follows:

$$\mathrm{F}={\displaystyle \sum _{PUs}Cost}+BLM{\displaystyle \sum _{PUs}B}oundary+{\displaystyle \sum _{ConVahue}S}PF+CostTherSholdPenalty(t)$$

(10)

In the formula, F is the overall objective function; ${\sum }_{PUs}Cost$ represents the total cost of the planning units; $BLM{\sum }_{PUs}Boundary$ denotes the total length correction value of the protected area boundary; ${\sum }_{ConValue}SPF$ refers to the compensation value for protection goals that have not been met; and $CostThreSholdPenalty\left(t\right)$ indicates the penalty value for exceeding the cost threshold.

For carbon sequestration evaluation, autotrophic respiration is used as the evaluation indicator. Autotrophic respiration is the difference between gross primary productivity (GPP) and net primary productivity (NPP). This study calculates GPP and NPP using the VPM model and the CASA model, respectively.

For water conservation assessment, the amount of water conservation is selected as the key evaluation indicator. The water conservation capacity is estimated using the precipitation storage method. The runoff coefficient is determined based on studies that establish a relationship between vegetation coverage (FVC) and runoff coefficient. A polynomial fitting approach is applied to derive the corresponding equation for calculating both the bare land precipitation runoff coefficient and the ecosystem precipitation runoff coefficient. The formula is

$$TQ=J_0\times K\times (R_0-R_g)$$

(11)

In the formula, TQ is the water conservation capacity; J₀ is the annual precipitation; K is the ratio of runoff-producing rainfall to total precipitation; R₀ is the runoff coefficient for bare land; and R_g is the runoff coefficient for the ecosystem.

For soil and water conservation evaluation, the modified Universal Soil Loss Equation (RUSLE) is used to calculate the soil and water conservation capacity. The driving parameters of this model include the rainfall erosivity factor, soil erodibility factor, slope length and steepness factor, vegetation cover and management factor, and conservation practice factor. These parameters are easily accessible and convenient to apply, offering reliable simulation results, which makes the model well suited for the study area. The formula is

$$Q_{sr}=R\times K\times L\times S\times (1-C\times P)$$

(12)

In the formula, Q_sr is the soil conservation amount (t/(year·km²)); R is the rainfall erosivity factor; K is the soil erodibility factor; L is the slope length factor; S is the slope steepness factor; C is the vegetation cover and management factor; and P is the conservation practice factor.

4. Results

4.1. Water Environment Evaluation Results

In 2021, post-restoration, and in 2022, before (April) and after the first batch of restoration (July), the water quality at all eight monitoring points was above Class III. However, in the second phase of post-restoration monitoring (September), four monitoring points maintained water quality above Class III, while the remaining four declined to Class IV, primarily due to elevated total phosphorus levels. Across the four monitoring stages, water quality at four monitoring points remained stable, consistently meeting good quality standards. However, at the other four points, water quality deteriorated, shifting from good status (2021 post-restoration, 2022 pre-restoration in April, and first-phase post-restoration in July) to Class IV in September following the second phase of restoration. This downward trend warrants further monitoring and targeted intervention (Figure 4).

4.2. Ecosystem Landscape Structure Evaluation Results

4.2.1. Habitat Quality Index

The habitat quality index values for the Liaohekou Wetland ecological restoration units were 93.16 at baseline and 93.37 in 2022, reflecting a change rate of 0.23%. As illustrated in Figure 5 and Figure 6, the total area experiencing an increase in habitat quality slightly exceeds the area experiencing a decline. Specifically, 4.47% of the region exhibited an increase in quality to Level III, whereas only 0.46% showed a decline at the same level. Regions where habitat quality improved significantly are primarily concentrated in the central wetland areas, where the expansion of water bodies has contributed to mitigating severe fragmentation in the Liaohekou Wetland. Meanwhile, habitat quality in the coastal and estuarine regions is exhibiting a trend toward dynamic stability.

4.2.2. Landscape Fragmentation

The average patch area index for the Liaohekou Wetland Ecological Restoration Unit was 6.16 ha in the baseline year and 6.38 ha in 2022, reflecting a change rate of 3.57%. As shown in Figure 7 and Figure 8, the 2022 average patch area index was predominantly classified as Class V, accounting for 49.28% of the area, followed by Class II (21.05%) and Class III (14.72%). Class I represented the smallest proportion at 5.46%. This shift is likely attributed to the implementation of wetland and hydrological restoration projects in the Liaohekou Wetland Ecological Restoration Unit, which have addressed the issues of area reduction and severe fragmentation. Consequently, significant portions of grassland, wetland, and shrubland have been converted into wetlands, resulting in reduced landscape fragmentation.

4.2.3. Natural Habitat Connectivity

The habitat connectivity index of the Liaohekou Wetland Ecological Restoration Unit was 24.48 during the baseline period and increased to 26.28 in 2022, reflecting a growth rate of 7.35% (Figure 9). Compared to the baseline, the 2022 index showed a modest improvement. Most areas, particularly in the central and southern regions, experienced no significant change in habitat connectivity. However, regions with a slight increase in the index covered a slightly larger area than those where the index declined (Figure 10).

4.2.4. Landscape Stability

The landscape stability index of the Liaohekou Wetland Ecological Restoration Unit was 0.038 during the baseline period and increased to 0.039 in 2022, reflecting a change rate of 2.63%. As shown in Figure 11 and Figure 12, landscape stability was predominantly concentrated in Level I, with the proportion of areas where Level I increased reaching 32.34%, while areas where Level I decreased accounted for 29.31%. Overall, the area experiencing increased stability was larger than that with decreased stability. The smallest proportion was observed in decreased Level II, at just 6.45%. The total proportion of areas with increased stability reached 54.52%, surpassing half of the total area. Over the two years, land cover types in the Liaohekou Wetland Ecological Restoration Unit primarily transitioned from other land uses to wetland ecosystems. The expansion of wetland areas was accompanied by a relatively slow degree of landscape growth during the early stages of restoration, with patch shapes remaining relatively complex.

4.2.5. Landscape Richness

The Shannon diversity index for the Liaohekou Wetland Ecological Restoration Unit was 1.31 in the baseline year and increased to 1.33 in 2022, reflecting a change rate of 1.53%. As shown in Figure 13 and Figure 14, 27.27% of the ecological restoration unit exhibited no change in the Shannon diversity index, primarily consisting of non-ecological land. Areas classified as reduced Level I and increased Level I together accounted for 23.09%, indicating minimal variation in the spatial distribution pattern of ecological land. Regions with a Level III increase in the Shannon diversity index comprised 16.17% of the total area, while those experiencing a Level III decrease accounted for 20.57%. In these regions, the Shannon diversity index demonstrated a noticeable upward trend.

4.3. Ecosystem Quality Evaluation Results

4.3.1. Vegetation Coverage

The vegetation coverage of the Liaohe Estuary Wetland Ecological Restoration Unit was 0.40 in both the baseline and monitoring years. As shown in Figure 15 and Figure 16, the highest proportion of vegetation coverage change occurred in the increased Class I range, accounting for 57.02%. The proportions of areas with increased Class II, decreased Class I, and decreased Class II were 0.34%, 40.90%, and 1.73%, respectively.

4.3.2. Leaf Area Index

The mean leaf area index (LAI) values for the Liaohe Estuary Wetland Ecological Restoration Unit were 1.31 in the baseline year and 1.58 in the monitoring year, reflecting a change rate of 20.61%. As shown in Figure 17 and Figure 18, the LAI change rate in the Liaohe Estuary Wetland Ecological Restoration Unit had the largest area proportion in the “Increase Level I” category, accounting for 23.46%, followed by 13.88% in “Increase Level II” and 17.56% in “Increase Level III”. The area proportions for “Decrease Level I”, “Decrease Level II”, and “Decrease Level III” were 19.61%, 11.78%, and 13.71%, respectively. The LAI change rate in the increasing levels significantly exceeded that in the decreasing levels, and this trend was consistent across the entire project area.

4.3.3. Gross Primary Productivity

The average gross primary productivity (GPP) for the baseline and monitoring years in the Liaohe Estuary Wetland Ecological Restoration Unit was 189.6 gC·m⁻²·a⁻¹ and 198.78 gC·m⁻²·a⁻¹, respectively, reflecting a mean change rate of 4.84%. As shown in Figure 19 and Figure 20, the areas with an increase in GPP change rate slightly outnumber those with a decrease, and this trend is consistent across the entire project area. The GPP change rate in the Liaohe Estuary Wetland Ecological Restoration Unit is primarily concentrated in the “Increase Level I” range, which accounts for 38.02%.

4.4. Ecosystem Service Function Evaluation Results

4.4.1. Biodiversity Conservation

In 2022, the average biodiversity conservation index for the Liaohe Estuary Wetland Ecological Restoration Unit was 0.79. As shown in Figure 21 and Figure 22, the largest area proportion was in Grade IV, accounting for 55.07%, followed by Grade III at 28.52%, while Grades I, II, and V accounted for 0.14%, 10.13%, and 6.14%, respectively. The distribution is generally clustered, with high-value areas concentrated along the river and coastal regions, while other areas are more evenly spread. During the 2022 monitoring period, the biodiversity conservation index change rate was primarily influenced by Grade I increase and Grade I decrease areas, with proportions of 37.74% and 51.61%, respectively. The area proportion of Grade I increase was higher than that of Grade I decrease, indicating an overall upward trend.

4.4.2. Carbon Sequestration

The autotrophic respiration values for the Liaohe Estuary Wetland Ecological Restoration Unit in the baseline year and 2022 were 129.20 g/m² and 138.51 g/m², respectively, reflecting a change rate of 7.21%. Autotrophic respiration was higher in the northern region and lower in the southern region. The largest area proportion was in Grade III, accounting for 39.28%, followed by Grade II at 22.88%. The smallest area proportions were observed in Grade I and Grade IV, accounting for 17.63% and 20.22%, respectively. As shown in Figure 23 and Figure 24, the autotrophic respiration change rate in the Liaohe Estuary Wetland Ecological Restoration Unit was primarily classified as an increase in Grade I, with an area proportion of 30.08%. The area proportions for Grade II, decrease in Grade I, and decrease in Grade II were 23.60%, 12.69%, and 4.52%, respectively. Based on land use changes, compared to the baseline year, the area of cultivated land in the Liaohe Estuary Wetland Ecological Restoration Unit expanded in 2022, leading to an increase in the carbon sequestration capacity of crops and, consequently, a rise in autotrophic respiration.

4.4.3. Water Conservation Capacity

In 2022, the average water conservation capacity per unit area of the Liaohe Estuary Wetland Ecological Restoration Unit was 61.94 mm. The largest proportion of areas fell within the 50–100 mm range, accounting for 47.02%, while the smallest proportion was found in the 150–200 mm range. Areas with higher water conservation capacity accounted for a relatively larger proportion. As shown in Figure 25 and Figure 26, most areas of the Liaohe Estuary Wetland Ecological Restoration Unit experienced a decrease in water conservation services, with the change rate primarily concentrated on reductions in Level I. The area proportion of reduction in Level I was 64.39%, while areas in other levels accounted for a smaller proportion. Overall, the water conservation capacity remained relatively stable, despite a significant reduction in area, and the ecological restoration effects were not pronounced.

4.4.4. Soil Conservation Amount

The average soil conservation amount for the Liaohekou Wetland Ecological Restoration Unit in the baseline period and 2022 was 9.07 t/km² and 10.25 t/km², respectively, reflecting a change rate of 13.01%. In 2022, the majority of the soil conservation amount was classified as Level I, with an area proportion of 89.24%. The area with a soil conservation amount at Level II occupied 5.85%, while areas at Levels III, IV, and above Level V accounted for 1.93%, 1.48%, and 1.52%, respectively. As shown in Figure 27 and Figure 28, the area proportions of soil conservation amounts in the Liaohekou Wetland Ecological Restoration Unit are as follows: 28.67% in decreasing Level I and 42.76% in increasing Level I, followed by increasing Level II with an area proportion of 19.57%. The area proportions of decreasing Level III, decreasing Level II, and increasing Level III are 2.70%, 0.45%, and 5.92%, respectively. The soil conservation function in most areas remained essentially unchanged, with a larger area proportion in increasing Level II, exceeding 15%.

5. Discussion

This study revealed a significant enhancement in landscape connectivity and a reduction in fragmentation (patch density decreased by 27.19%) in the Liaohe Estuary Wetland, aligning with the global wetland restoration strategy of “reducing anthropogenic disturbances”. For instance, studies in the Yellow River Delta have identified oilfield exploitation and land reclamation as primary anthropogenic drivers of wetland fragmentation [43]. In contrast, the Liaohe Estuary has effectively mitigated human pressures through initiatives such as the “returning aquaculture ponds to wetlands” project, which removed aquaculture facilities and restored natural shorelines. These findings resonate with research on Zhuhai Wetlands, where reduced aquaculture density similarly led to decreased patch density [44]. Additionally, natural factors, such as variations in sediment transport and runoff (accounting for 35.1% of the first principal component’s contribution), influenced restoration outcomes, consistent with the mechanisms of “synergistic natural and anthropogenic drivers” observed in the middle reaches of the Heihe River Wetlands [45]. Notably, spatial heterogeneity remains a common challenge in global wetland restoration. For example, Zhalong Wetland experienced localized fragmentation due to cropland expansion [46], while the heterogeneous water quality improvements in the Liaohe Estuary may stem from uneven regional water quality baselines, necessitating the further integration of zoning management strategies.

The 2.41% increase in the biodiversity maintenance index was primarily attributed to enhanced habitat connectivity and reduced pollutant inputs. As highlighted by IPBES (2019), land use changes (e.g., wetland restoration) and pollution control are direct drivers of biodiversity recovery, which aligns with the Liaohe Estuary’s water quality improvement (from Class IV to above Class III) and enhanced carbon sequestration capacity (12.8% increase in net primary productivity—NPP). Similarly, studies on the Ruoergai Wetland demonstrated a positive correlation between wetland area recovery and the population growth of black-necked cranes [47], a pattern corroborated by the observed increase in overwintering red-crowned cranes in the Liaohe Estuary. However, biodiversity recovery often exhibits time-lagged effects. For example, in the Dongzhai Harbor Mangrove restoration in Hainan, benthic communities required 3–5 years to stabilize [48]. Thus, long-term monitoring in the Liaohe Estuary should incorporate dynamic assessments of ecological succession.

This study integrated remote sensing, ground monitoring, and ecological modeling to dynamically quantify synergistic changes in water environments, landscape structures, and ecosystem services, comparable to the “multi-source remote sensing + entropy method” framework applied in the Wuliangsuhai Basin [49]. For instance, the simultaneous improvement in the leaf area index (LAI) and gross primary productivity (GPP) validated vegetation restoration’s role in enhancing carbon sink functions, consistent with conclusions from subtropical forest restoration studies emphasizing that “multi-source data fusion improves evaluation accuracy” [50]. Nevertheless, current methodologies have limited capacity to capture microscopic ecological processes, such as soil microbial community reconstruction. Future efforts could integrate molecular biology techniques (e.g., environmental DNA) to refine monitoring systems [51].

6. Conclusions

This study demonstrated through multi-dimensional assessments that the ecological restoration of the Liaohe Estuary Wetland has significantly improved water quality (100% compliance rate for Class III or higher standards), enhanced landscape connectivity (dynamic index of 27.19%), and elevated ecosystem service functions (12.8% increase in carbon sequestration capacity). By implementing the “returning aquaculture ponds to wetlands” initiative and tidal channel dredging projects, aquatic ecological corridors connecting native habitats were successfully established, offering a global coastal wetland restoration paradigm characterized by “anthropogenic disturbance mitigation coupled with natural process synergy”. The application of multi-source data integration methods provided robust technical support for the dynamic monitoring and precise regulation of ecological restoration efforts.

However, this study has limitations: (1) The evaluation period was relatively short (only one year post-restoration), limiting comprehensive insights into long-term ecosystem stability. (2) The mechanism underlying heterogeneous water quality improvement remains unclear and requires further elucidation through hydrological modeling. (3) The biodiversity assessments relied on macro-level indicators and lacked habitat suitability analyses for keystone species, such as spotted seals.

Future studies should prioritize prolonging the monitoring period to 5–10 years, incorporating micro-scale indicators (e.g., soil microbial communities and benthic fauna), and developing coupled “restoration effectiveness–socioeconomic” models to evaluate the contributions of ecological recovery to regional sustainable development.