Spatial Planning Strategies for Wetlands Based on a Multimethod Approach: The Example of Tianjin in China

Abstract

Wetlands form a crucial component of ecosystems, and wetland restoration serves as an effective strategy for promoting sustainable urban development. Spatial support is essential for wetland restoration, meaning that research on wetland spatial planning is of considerable importance. Existing studies on wetland spatial planning primarily focus on the analysis of wetland spatial distribution characteristics, with limited exploration of wetland spatial relationships. This paper aims to explore the potential of utilizing both spatial distribution characteristics and spatial relationships to identify wetland spatial issues, thereby facilitating the formulation of wetland spatial planning strategies. Using Tianjin City as a case study, this research applies nearest neighbor analysis, the geographic concentration index, the Gini index, and kernel density analysis to identify the spatial distribution characteristics of wetlands in Tianjin. Additionally, spatial autocorrelation analysis and connectivity analysis are employed to identify the interrelationships among wetlands in Tianjin. Based on the results derived from the analysis of spatial distribution characteristics and spatial relationships, wetland spatial planning strategies are proposed. The effectiveness of these strategies is validated using methods that consider both spatial distribution characteristics and spatial relationships. The findings reveal that, although wetlands in Tianjin are widely distributed, large wetland patches are primarily concentrated in areas with abundant water resources, while the six districts within the city have few or no large patches of wetlands. The spatial distribution of wetlands is highly uneven, exhibiting patterns of high–high aggregation and low–low aggregation. The number of connecting paths between wetland patches is relatively low, indicating a generally low overall connectivity. While medium-sized and larger wetland patches maintain the connectivity of existing wetlands in Tianjin, small wetlands that serve as stepping stones are lacking. Following the implementation of planning strategies, there would be an increase in the wetland area in Tianjin, accompanied by significant improvements in the spatial distribution pattern and spatial relationships of the wetlands.

1. Introduction

Wetlands are intricate natural assemblages resulting from the reciprocal interaction between water and land [1]. Sustaining their position as fundamental constituents of significant ecological systems, wetlands perform a myriad of multifaceted functions [2]. These encompass the regulation of climatic conditions, the provision of freshwater resources, and the preservation of biodiversity [3]. Notably, wetlands assume a pivotal role in augmenting the caliber of urban ecosystems, upholding ecological equilibrium, enhancing ecological attributes, and fostering the advancement of urban sustainability [4,5]. Since the inception of the Ramsar Convention on Wetlands, there has been a growing global acknowledgment of the paramount importance and pressing need to strengthen the ecological optimization, restoration, and sustainable utilization of marsh wetlands [6]. Furthermore, the international focus on marsh wetlands has transcended the singular emphasis on their role as waterfowl habitats, growing to encompass diverse areas, including wetlands’ optimization, restoration, conservation, and judicious utilization [7]. This expanding perspective reflects the collective recognition of the multifaceted value and potential of marsh wetlands in fostering ecological integrity and the harmonious coexistence between human activities and natural systems [8].

During the process of rapid urbanization, wetland resources in China have been excessively exploited and encroached upon [9], leading to a significant degradation of ecosystem services. Pressing issues, such as fragmented wetland patches, reduced freshwater storage, and diminished flood regulation capacity, have emerged [10], rendering wetlands incapable of meeting the requirements for sustainable urban development [11]. Since China’s accession to the Ramsar Convention in 1992, research and conservation efforts pertaining to wetland ecosystems have gradually gained attention [12,13]. However, owing to the inherent complexity and dynamic nature of wetland systems, further exploration is necessary to establish a scientifically sound approach towards their protection [14]. In light of these circumstances, China has undertaken a series of systematic planning initiatives for wetland regions, encompassing strategic planning, overall planning, detailed planning, and special planning [15]. In practical terms, the establishment of wetland parks has preceded wetland-planning theory, a testament to the Chinese government’s renowned commitment to wetland conservation [16,17]. Consequently, the creation of national wetland parks in China has yielded remarkable achievements. At present, the predominant model for constructing national wetland parks prioritizes protection as the fundamental objective, while also promoting rational utilization and placing a significant emphasis on the ecological service value that surpasses mere economic gains [18,19]. From a planning standpoint, conducting wetland research that is rooted in spatial planning enables the allocation of physical space for wetland restoration purposes during urban development, thereby safeguarding wetland areas from encroachment by urbanization [20,21]. Nevertheless, the prevailing challenges in wetland planning encompass issues such as singular planning approaches [22], spatial functional conflicts [23], and inadequate coordination [24]. The absence of a comprehensive mechanism to harmonize and coordinate spatial planning severely impedes the optimization of wetland spaces.

Wetland restoration plays a critical role in achieving ecological security and promoting sustainability. Pioneering the integration of macroscopic ecological principles with the optimized spatial allocation of land use, I. McHarg provided a comprehensive exploration of the application of spatial optimization to marsh wetlands in his seminal work Design with Nature [25]. Since the 1960s, numerous researchers and relevant governmental organizations have further advanced this theoretical framework, conducting practical investigations into marsh-wetland spatial planning using suitability analysis methods [26,27]. Notably, Forman’s 1995 publication Land Mosaic exemplified a holistic optimization of marsh-wetland landscape patterns while systematically summarizing the approaches for the spatial optimization of these landscapes [28]. Early studies primarily focused on optimizing artificial marsh wetlands to enhance their ecological environments [29]. In recent years, scholars have also employed various optimization algorithms to spatially optimize small to medium-scale marsh wetlands from a landscape-pattern perspective. Connolly et al. explored the spatial optimization and restoration of marsh wetlands from a landscape-pattern perspective, using a simple weighted approach based on the principles of holistic planning [30]. Qasaimeh et al. employed fuzzy mathematical theory to optimize the spatial design of artificial marsh wetlands, demonstrating the effectiveness of this method in artificial marsh-wetland optimization [31]. Using a multiobjective approach, Meghna et al. optimized the spatial distribution of marsh wetland restoration results, with marsh-wetland landscape patterns as the objective function [32]. By integrating landscape-ecology theory, regulating hydrological processes, and optimizing landscape spatial patterns and configurations based on the principles of landscape structure–process–function integration, Cai et al. enhanced landscape functions, such as water purification and biodiversity, in wetland parks [33].

The insufficient recognition of ecological interdependencies has hampered the implementation of targeted measures and posed challenges to the conservation and optimization of wetlands. The accurate scientific identification and demarcation of wetland resources at the regional level are essential for effective wetland conservation [34]. Furthermore, the rapid pace of urban expansion exacerbates the conflict between regional ecological conservation and development. Thus, it becomes crucial to proactively anticipate future urban-development scenarios, prioritize the protection of ecological spaces, and establish a harmonious relationship between wetland conservation and development through strategic planning [35]. In summary, the current research on wetland spatial optimization largely centers around wetland evolution [36], kernel density analysis [22], distribution types [37], and influencing factors [38] to propose strategies for wetland spatial planning based on the analysis of spatial distribution characteristics. However, limited attention has been paid to the spatial relationships between wetlands, encompassing their interconnectivity, correlation, and a number of interconnected paths. This study aims to address this gap by conducting a comprehensive analysis of the spatial relationships between wetlands, utilizing Tianjin City as a case study, and leveraging the existing knowledge on wetland spatial distribution patterns. The objective is to present wetland spatial planning strategies and attempt to provide methodological insights for a useful exploration of spatial planning decision support for urban wetland systems.

2. Materials and Methods

2.1. Study Area

Tianjin (116°43′E–118′°04′E, 38°34′–40°15′N) is located in North China and is one of the core cities of the Beijing–Tianjin–Hebei urban agglomeration, covering an area of 11,966.45 km² (Figure 1). Situated in the warm temperate zone, Tianjin experiences a semihumid monsoon climate. With an annual average temperature of around 14 °C, and an annual average precipitation ranging from 360 to 970 mm, the region predominantly receives rainfall during the summer, often resulting in the occurrence of floods and waterlogging. The topography of Tianjin mainly comprises plains and basins, interspersed with low mountains and hills in its northern region. The elevation gradually decreases from north to south, with the highest point reaching 1052 m in the north, while the lowest point lies at 3.5 m above sea level in the southeast. The convergence of significant rivers, namely the Daqing River, Nanyun River, Beiyun River, Ziya River, and Yongding River, contributes to the complex water system and relatively abundant wetland resources present in Tianjin. Unfortunately, due to ongoing urban development and construction activities, the wetland area has diminished considerably over time. Wetland coverage, which stood at approximately 45.9% in the early 20th century, drastically declined to a mere 8.5% by the 1970s [39]. With increasing attention being paid to ecosystem conservation, wetland coverage has progressively rebounded, reaching 24.70% during the second National Wetland Inventory in 2013. However, a substantial disparity persists between the initial and current states. Thus, as a research site, Tianjin exhibits significant representativeness and exemplification.

2.2. Research Process

The research approach employed in this paper encompasses two main components, analysis and strategy, which correspond to problem analysis and problem-solving, respectively, conforming to the essential steps of spatial planning [40,41]. Expanding on the foundation laid by the analysis and strategy, this study introduces a validation element aimed at ensuring the scientific rigor of the planning strategies. The research approach can be summarized as follows: (1) by establishing a comprehensive database for the study area in 2018, the current state of the wetlands in Tianjin is examined through the lenses of spatial distribution characteristics and spatial relationships. (2) Based on the findings from the analysis, wetland spatial planning strategies are formulated, emphasizing wetland restoration, creation, and categorization. (3) Subsequent to the development of the spatial planning strategies, a postplanning database is constructed to assess alterations in the spatial distribution characteristics and spatial relationships of wetlands in Tianjin, thereby validating the scientific underpinnings of the planning strategies. The research approach, illustrated in Figure 2, can be readily applied to other urban contexts as well.

2.3. Data Sources and Database

The total amounts of water resources and precipitation were obtained from the National Bureau of Statistics of the People’s Republic of China (http://www.stats.gov.cn/, accessed on 7 May 2023) and the Bureau of Water Resources of Tianjin (http://swj.tj.gov.cn/, accessed on 7 May 2023). National wetland inventory data were obtained from the Forest Knowledge Service System (http://lygc.lknet.ac.cn/, accessed on 7 May 2023). Landsat 8 OLI_TIRS data of the study area and Digital Elevation Model (DEM) data with a resolution of 30 m were obtained from the Geospatial Data Cloud (http://www.gscloud.cn/, accessed on 8 May 2023). ENVI 5.3 software (Exelis Visual Information Solutions, Broomfield, CO, USA) was employed to conduct a meticulous analysis of the Landsat 8 satellite imagery of Tianjin in 2018, with the objective of extracting precise land-use information. Through a rigorous process of segmentation and elimination to exclude minute and invalid patches, a comprehensive inventory comprising 73,334 distinct wetland patches was acquired. These patches collectively covered an estimated area of approximately 2768.84 km². Notably, the derived extent of delineated wetlands accounted for approximately 23.13% of the overall land area of Tianjin, which spans 11,966.45 km². Interestingly, this calculated percentage closely approximates the wetland ratio of 24.74% reported in the official 2018 statistical data. DEM and Landuse data were compiled using ArcGIS 10.2 for Desktop software (Environmental Systems Research Institute, Inc. Redlands, CA, USA). Spatial autocorrelation was performed with GeoDa 1.18.0 software (Luc Anselin. Chicago, IL, USA). Spatial connectivity was performed using the Conefor Sensinode 2.6 software package (Saura and Torne. Madrid and Catalonia, Spain).

To facilitate computations pertaining to distribution patterns, concentration levels, kernel density, spatial autocorrelation, and spatial connectivity, a spatial database of Tianjin’s wetlands was carefully constructed utilizing data from the year 2018. First, a 2 × 2 km grid of Tianjin was created using the fishnet tool via ArcGIS 10.2 (Data Management Tools—Feature Classes—Created Fishnet). The total number of valid units within the Tianjin area is 3259. The wetland spatial data were then overlaid with the grid (Analysis Tools—Overlay—Intersect) (Figure 3a). The total number of valid wetland units within the Tianjin area is 2077; at the same time, the polygon features were converted to point features for the distribution pattern, geographic concentration, and kernel density analysis (Data Management Tools—Features—Feature To Point) (Figure 3b), and the wetland coverage within each unit was taken into account for the spatial autocorrelation analysis (Figure 3c).

2.4. Methods

The main purpose of this research was to examine the spatial strategies employed in wetland management in Tianjin. By analyzing the spatial distribution characteristics of wetlands and their interrelationships, this study aimed to establish a basis for developing effective spatial strategies. To initiate the spatial analysis of wetlands in Tianjin, we suggest employing various techniques, including distribution pattern analysis, geographic concentration analysis, and kernel density estimation. Additionally, the relationships between wetlands in Tianjin can be identified by utilizing spatial autocorrelation and spatial connectivity methods.

2.4.1. Spatial Distribution Characteristics

• Distribution pattern

Point-pattern analyses often focus on spatial distribution patterns [42]. Among the various analytical approaches used, nearest neighbor analysis (NNA) is widely employed to investigate the spatial distribution of point features and classify their distribution types within a study [43]. The assessment of point feature distribution characteristics, encompassing randomness, regularity, and clustering, is commonly achieved through the utilization of the nearest neighbor index (NNI) [44]. The formula for the NNI is as follows:

$$R=\frac{{\overline{r}}_o}{{\overline{r}}_e}; {\overline{r}}_o=\frac{\sum _{i=1}^n{d}_i}{n}; {\overline{r}}_e=\frac{0.5}{\sqrt{n/A}}$$

where $R$ is the NNI, ${\overline{r}}_o$ represents the average distance between wetland point features and their nearest neighbors, and ${\overline{r}}_e$ denotes the expected average distance between wetland point features in an ideal scenario; $n$ and $A$ correspond to the total number of point features and the territorial area of Tianjin, respectively. If $R$ equals or approaches 1, it indicates that wetlands in Tianjin have a random spatial distribution type. When $R > 1$, it suggests a dispersed distribution type. Conversely, when $R < 1$, it signifies an aggregated distribution type. In the case where $R=0$, it indicates a completely concentrated distribution type of wetlands in Tianjin. The NNI was calculated using Arcgis 10.2 for Desktop.

2.

Geographic concentration

①

Determination of Equilibrium

The geographic concentration index (GCI) serves as a valuable index within the field of point-pattern analysis, enabling the assessment of homogeneity in spatial distributions. Diverging from the nearest neighbor index (NNI), the GCI disregards distribution types and focuses solely on discerning the presence of a balanced spatial arrangement among the point features [45]. The formula for the NNI is as follows:

$$G=\sqrt{\sum _{i=1}^m{\left(x_i/n\right)}^2}\times 100$$

where $G$ represents the geographic concentration index (GCI) of wetlands in Tianjin. A GCI with a higher value indicates a stronger degree of spatial concentration in the distribution of wetlands within the city. $x_i$ represents the number of wetland point features within the $i$th grid cell, $n$ represents the total count of wetland point features, and $m$ represents the total number of grid cells.

②

Recognition of Equilibrium Degree

Identification of the degree of equilibrium: the Gini index (GI), initially widely employed as a metric to gauge income disparity among residents within a country or region, has subsequently been utilized to examine disparities in spatial distribution [46]. While the GCI provides insights into the equilibrium of wetland spatial distribution, the GI enables a more nuanced evaluation of equilibrium levels. The formula for the GI is as follows:

$$\begin{gathered} H=-\sum _{i=1}^N{P}_{i}ln{P}_i; H_m=lnN \\ Gini=H/H_m; C=1-Gini \end{gathered}$$

where $P_i$ represents the proportion of wetland point features within the $i$th grid cell, $N$ is the total count of wetland point features, and $C$ denotes the degree of spatial distribution equilibrium of the wetlands in Tianjin. The GI, a measure ranging from 0 to 1, serves as an indicator of the concentration level in the spatial distribution of wetlands in Tianjin. A higher GI implies a greater concentration. A GI approaching 0 indicates a tendency towards absolute equilibrium in the spatial distribution of wetlands in Tianjin, with values below 0.2 commonly recognized as indicative of absolute equilibrium. In the range of 0.2 to 0.3, the GI suggests a relatively balanced distribution, while a coefficient between 0.3 and 0.4 signifies the comparatively balanced spatial distribution of wetlands in Tianjin. A GI between 0.4 and 0.5 indicates an unbalanced distribution, whereas a GI above 0.5 signifies the significantly uneven spatial distribution of wetlands in Tianjin.

3.

Kernel density

Kernel density estimation (KDE) is a method used in point-pattern analysis. While NNI, GCI, and GI analysis provide results in numerical form, KDE allows for the visualization of the clustering patterns of point features, illustrating the intensity of their influence on the surrounding areas [47]. This, in turn, provides spatial guidance for planning strategies. The formula for KDE is as follows:

$$f\left(x\right) =\frac{1}{wh}\sum _{i=1}^{w}k\left(\frac{x-X_i}{h}\right)$$

where $f\left(x\right)$ represents the calculated value of kernel density and where higher values indicate a denser distribution of wetland point features in Tianjin. $k\left(\frac{x-X_i}{h}\right)$ refers to the kernel function, $h$ represents the analysis bandwidth, denoting a value greater than 0. $w$ represents the count of wetland point features within the bandwidth range, and $x-X_i$ represents the distance between points $x$ and $X_i$.

2.4.2. Spatial Relation Analysis

4.

Spatial autocorrelation

Spatial autocorrelation analysis plays a crucial role in investigating the spatial patterns and specific attribute information pertaining to the study subject. It provides an accurate depiction of the spatial distribution patterns exhibited by the subject and sheds light on the spatial interactions among its various features [48]. Global spatial autocorrelation and local spatial autocorrelation are two common approaches employed in spatial autocorrelation analysis.

Global spatial autocorrelation analysis is employed to characterize the overall spatial characteristics of the features within the study area. This can be achieved by computing the global Moran’s I index, which provides an intuitive measure of the interdependence of feature variables within the study region. The calculation formula for the global Moran’s I index is as follows:

$$I=\frac{N\sum _{i=1}^N\sum _{j=1}^N{w}_{ij}\left(x_i-\overline{x}\right)\left(x_j-\overline{x}\right)}{\left(\sum _{i=1}^N\sum _{j=1}^N{w}_{ij}\right)\sum _{i=1}^N{\left(x_i-\overline{x}\right)}^2}$$

where $I$ represents the global spatial autocorrelation coefficient. $x_i$ and $x_j$ denote the observed values of the same attribute information for objects $i$ and $j$, respectively, within the study area. In this research, the wetland ratio is adopted as the analyzed attribute information. $N$ refers to the sample size, $\overline{x}$ denotes the mean value of the attribute, and $w_{ij}$ represents the spatial weight matrix. In this study, the Euclidean distance is utilized as the weight metric. The range of the global Moran’s I index is [−1, 1]. If $I<0$, it indicates the presence of negative spatial correlation, signifying an opposite trend in the wetland ratio between adjacent wetland areas. Conversely, if $I>0$, it indicates a positive spatial correlation, implying a similar trend in the wetland ratio among neighboring wetland areas. If $I=0$, it suggests the absence of spatial correlation.

The global Moran’s I index can identify the mutual relationships among the elements of the research subject, but it cannot identify the specific spatial distribution of hot and cold areas [49]. Therefore, it is necessary to utilize local spatial autocorrelation for hotspot analysis. The local indicators of the spatial association (LISA) technique are used to assess the specific spatial correlation characteristics at the regional scale, identifying the spatial clustering or spatial anomaly of the research subject elements. The formula for calculating local spatial autocorrelation is as follows:

$$I_i=\frac{x_i-\overline{x}}{{\sigma }^2}\sum _{j=1}^N{w}_{ij}(x_j-\overline{x})$$

where $I_i$ represents the local spatial autocorrelation coefficient at the location of object i within the study area. $x_i$ and $x_j$ are the observed values of the same attribute information for objects $i$ and $j$ located within the study area, respectively. In this study, the wetland ratio is used as the attribute information for analysis. $\overline{x}$ denotes the mean of $x$, and ${\sigma }^2$ represents the variance of $x$. $w_{ij}$ represents the spatial weight matrix. Additionally, the Euclidean distance is used as the weight. The range and definition of positive and negative values for $I_i$ are the same as for the global Moran’s I index.

The association patterns in the local spatial autocorrelation analysis can be categorized according to four types: high–high, low–high, low–low, and high–low. The high–high pattern or low–low pattern indicates the spatial clustering of high or low values, respectively, representing positive spatial correlation. On the other hand, the low–high pattern and high–low pattern indicate that high or low values are surrounded by low or high values, respectively, which suggests spatial anomalies and represents negative spatial correlation.

5.

Spatial connectivity

①

Integral Index of Connectivity

Connectivity serves as a crucial indicator for evaluating ecosystem quality and plays a significant role in preserving the overall integrity of an ecosystem. Various metrics, such as the connectivity probability index, the spatial connectivity index, and the movement connectivity index, are commonly employed to quantify connectivity. In this study, the integral index of connectivity (IIC) [50] was employed to assess the connectivity of the wetland ecosystem in Tianjin City. The computational formula for the IIC is given as follows:

$$IIC=\frac{\sum _{i=1}^n\sum _{j=1}^n\frac{a_i\mathrm{*}{a}_j}{1+n{l}_{ij}}}{A_L^2}$$

where n represents the total number of wetland patches present in the region under consideration. The parameters $a_i$ and $a_j$ correspond to the areas of the $i$th and $j$th wetland patches, respectively. The quantity $n{l}_{ij}$ denotes the count of pathways, constrained by a predefined distance threshold, connecting patch $i$ with patch $j$. $A_L$ signifies the cumulative area encompassed by all ecological patches within the region. Notably, the integral index of connectivity (IIC) satisfies the inequality 0 ≤ IIC ≤ 1. A value of IIC = 0 signifies the absence of interconnectivity between the ecological patches. Conversely, when IIC = 1, it indicates a substantial level of interconnectedness among all ecological patches, implying that each independent patch can be regarded as a constituent part of a larger unified ecological patch.

②

Delta integral index of connectivity

Based on the IIC, it is feasible to derive the patch importance index (delta integral index of connectivity, dIIC) [51]. The calculation formula for this index is as follows:

$$dIIC=100\times \frac{IIC-{IIC}_{remove}}{IIC}$$

where, $IIC$ represents the Integral Index of Connectivity, and ${IIC}_{remove}$ represents the modified connectivity index resulting from the removal of any wetland patch.

This study employed the IIC to calculate the wetland ecosystem connectivity. Initially, four distance thresholds were established at intervals of 2.5, 5, 7.5, and 10 km. Subsequently, utilizing the Conefor_Inputs_10 plugin within ArcGIS, data pertaining to patch area, topological distance between patches, and the number of pathways connecting patches were extracted, with calculations originating from the edges of each patch. Moreover, using ArcGIS, the spatial distribution of the pathways among the wetland patches in Tianjin City was generated across various distance thresholds. Thereafter, the obtained data was imported into the Conefor Sensinode 2.6 software package to compute the Integral Index of Connectivity (IIC) and delta Integral Index of Connectivity (dIIC) for each of the designated distance thresholds [52]. Finally, utilizing ArcGIS, the distribution of the wetland patch importance index at different distance thresholds within Tianjin City was extracted and visualized.

3. Results

3.1. Wetland Spatial Distribution Characteristics

3.1.1. Wetland Patch Size

The wetland patches are classified into four distinct categories: super-large wetland patches (10 km² and above), large wetland patches (5–10 km², excluding 10 km²), medium wetland patches (1–5 km², excluding 5 km²), and small wetland patches (below 1 km²) [53]. Through an analysis of the land-use map of Tianjin in 2018, the wetland patch types prevailing in the region were determined (

Table 1. Wetland-patch types in Tianjin, 2018.

Patch Type	Quantity	Percentage of Quantity	Area/km2	Percentage of Area
Super-large wetland patches	29	0.04%	999.61	36.10%
Large wetland patches	36	0.05%	240.79	8.70%
Medium wetland patches	251	0.34%	526.84	19.03%
Small wetland patches	73,018	99.57%	1001.59	36.17%

). Small wetland patches constitute the majority, with a count of 73,018 patches, accounting for 99.57% of the total count. Nevertheless, their collective area encompasses just 36.17% of the entire wetland area. Conversely, super-large, large, and medium wetland patches account for a mere 0.43% of the count, but they span 63.83% of the total wetland area. This demonstrates a highly imbalanced distribution of wetland patch sizes in Tianjin, with small wetland patches dominating the landscape. The fragmentation of wetland patches poses a notable threat to the ecological functionality of wetland ecosystems, resulting in a depletion of biodiversity within the patches, diminished resistance to invasive species, and an overall decline in wetland ecosystem quality. Consequently, the preservation and restoration of wetlands face formidable challenges in Tianjin.

While wetland sites are widely distributed across Tianjin, they exhibit a notable concentration in specific coastal, riverine, and reservoir-adjacent areas, resulting in discernible north–south disparities. Analyzing the spatial distribution of super-large and large wetland patches in Tianjin (Figure 4) reveals the uneven availability of these wetland resources, characterized by a pronounced differentiation between the southern and northern regions. Remarkably, the six central districts and their surrounding areas exhibit a significant deficiency of large and super-large wetland patches. Conversely, significant occurrences of such wetland resources are predominantly observed alongside major water systems, including the Yuqiao Reservoir, Huanggang Reservoir, Tuandao Lake, and Beidagang Wetland. It is worth noting that both banks of the Haihe River suffer from a scarcity of large and super-large wetland patches, a phenomenon attributable to the extensive development and construction endeavors along the river corridor.

3.1.2. Distribution Pattern

By utilizing the average nearest neighbor analysis tool in ArcGIS (Spatial Statistics Tool–Analyzing Patterns–Average Nearest Neighbor), we computed the nearest neighbor index of wetlands in Tianjin City. The analysis revealed that the average distance between wetland point features and their nearest neighboring point features was 1868.8567 m, whereas the expected average distance was 1562.5760 m. Consequently, the ratio of these values denoted as the nearest neighbor R, measured 1.1960, surpassing 1. This outcome suggests the dispersed spatial distribution pattern of wetlands in Tianjin City (Figure 5a). This dispersion can be attributed to the extensive water network present in the region. However, it is important to note that the existing spatial distribution pattern solely confirms the presence or absence of wetland distribution within the boundaries of Tianjin City. Further analysis is required to examine the quantity of wetland features within the Tianjin City region.

3.1.3. Geographic Concentration

The GCI of wetlands in Tianjin City was determined to be $G$ = 2.86. In a hypothetical scenario where all wetland sites in Tianjin City are uniformly distributed across the grid, the average GCI $\overline{G}$, would amount to 1.76. The fact that $G>\overline{G}$ indicates a concentration of wetlands in Tianjin City, thereby underscoring the presence of spatial imbalances. Furthermore, the Gini index, computed for the wetlands in Tianjin City, yields a value of $Gini$ = 0.6256. As $Gini$ > 0.5, it signifies a highly uneven spatial distribution of wetlands in Tianjin City, which is indicative of low uniformity.

3.1.4. Kernel Density

The spatial kernel density analysis of wetlands in Tianjin City was conducted using ArcGIS’s kernel density analysis tool (Spatial Analysis Tool—Density Analysis—Kernel Density). With an output cell size of 1000 m and a search radius of 5000 m, the analysis employed the natural breaks (Jenks) classification method to categorize the wetland kernel density into five levels: high, relatively high, moderate, relatively low, and low (Figure 5b). The spatial distribution of wetlands in Tianjin City depicts a dispersed pattern, characterized by a “clustered network distribution” at the overall level but with notable variations between the northern and southern regions. This finding further substantiates the previous conclusions regarding the dispersed and uneven distribution of wetlands in Tianjin City. Notably, the administrative districts of Binhai New Area, Jinghai, Xiqing, Baodi, and Jizhou emerge as the primary hotspots of wetland concentration. This can be attributed to the presence of water systems that provide a favorable ecological foundation for nurturing wetlands in these regions. In contrast, the central districts of Heping, Nankai, Hebei, and Hedong, which have undergone extensive urban development, exhibit minimal wetland kernel densities above the moderate level. These areas are important for future wetland spatial planning strategies.

3.2. Wetland Spatial Relation

3.2.1. Spatial Autocorrelation

By employing Geoda 1.18.0 software, it was ascertained that the minimum Euclidean distance between individual wetland patches and their neighboring patches in the wetland database amounted to 10,262.31 m. Accordingly, a bandwidth of 11,000 m was specified when establishing the spatial weights. The global Moran’s I index analysis, determined by using the Spatial Analysis—Univariate Moran’s I function, facilitated the computation of the Moran’s I index using wetland coverage as the variable within each grid. Consequently, a scatterplot illustrating the global Moran’s I index for wetland coverage in Tianjin City was generated (Figure 6a). Notably, the calculated global Moran’s I index value for wetland coverage in Tianjin City was determined to be 0.413, exhibiting a confidence level of 99%. This positive value of the global Moran’s I index signifies a pronounced spatial positive correlation among the wetland coverage in Tianjin City. The scatterplot effectively delineates the clustering of wetland coverage, predominantly falling within the “high-high” and “low-low” categories, thereby indicating a proclivity for spatial aggregation in both regions with high and low wetland coverage within the city.

Utilizing the local Moran’s I index analysis tool (Spatial Analysis—Univariate Local Moran’s I) in the Geoda 1.18.0 software, a comprehensive examination of the spatial autocorrelation pertaining to wetland coverage in Tianjin City was conducted. This meticulous analysis yielded a LISA (Local Indicators of Spatial Association) cluster map depicting the distribution of wetland coverage (Figure 6b). Notably, the outcomes derived from the local spatial autocorrelation analysis for wetland coverage in Tianjin City evince an elevated confidence level of 95% or higher, thus attesting to the thoroughness and reliability of this analytical endeavor. The LISA cluster map effectively elucidates a discernible trend of marked spatial clustering within the geographical expanse of wetland coverage in Tianjin City. The “high-high” mode, indicative of concentrated wetland distribution hotspots, is primarily observed in the Binhai New Area, Jinghai District, and Jizhou District of Tianjin City. Remarkably, these regions align with the areas characterized by ample water resources within the city. The presence of wetland hotspots implies a dispersed allocation of wetlands throughout the urban landscape. Conversely, no substantial wetland hotspots are found in the remaining administrative districts, thereby reaffirming the inherent unevenness of wetland distribution. The amalgamation of diverse spatial analysis techniques provides compelling evidence of the dispersion, imbalance, and pronounced nonuniformity of the spatial arrangement of wetlands in Tianjin City. Notably, the watercourses in Tianjin City, including the Hai River, demonstrate a prevailing “low-low” mode (Figure 6c). These rivers suffer from the intense degradation of the associated wetlands, which is primarily attributed to extensive human activities and dwindling upstream water resources. Consequently, the ecological functionality of riverine wetland ecosystems is undergoing a progressive decline [54].

3.2.2. Spatial Connectivity

The study employed the IIC to evaluate the connectivity of wetland ecosystems. Initially, four distance thresholds (2.5 km, 5 km, 7.5 km, and 10 km) were established. Subsequently, by using the Conefor_Inputs_10 plugin in ArcGIS, data encompassing the patch area, the topological distance between patches, and the number of pathways connecting patches were extracted, with the calculations commencing from the edges of the patches. Moreover, ArcGIS was utilized to generate the distribution of pathways between wetland patches in Tianjin under different distance thresholds. Lastly, the extracted data were imported into Conefor Sensinode 2.6 software to compute the IIC and the dIIC for each of the four distance thresholds. Furthermore, ArcGIS was employed to export the spatial distribution of the patch importance index for the diverse distance thresholds observed within Tianjin’s wetland patches.

From the perspective of pathway quantities (Figure 7), it can be observed that, under the 2.5 km distance threshold, there are distant pathways between wetland patches in Tianjin. This scarcity indicates inadequate connectivity among the patches, impeding biotic exchange. As the distance threshold increases, the number of pathways between wetland patches also increases. However, regardless of the distance threshold, the distribution of pathway quantities between wetland patches in Tianjin consistently reveals a “multi-clustered” pattern, with large and super-large patches serving as central hubs. Although some connections exist between clustered patches, their number is relatively limited. The protection and restoration of Tianjin’s wetland ecosystem are currently characterized by fragmented efforts, resulting in overall weak connectivity.

Based on an analysis of the IIC (

Table 2. IIC of wetland ecosystems in Tianjin under different distance thresholds in 2018.

Distance Thresholds/km	Number of Path	IIC
2.5	1699	0.0301
5	5347	0.0638
7.5	10,519	0.1056
10	16,709	0.1450

), it is evident that, for distance thresholds ranging from 2.5 km to 5 km, the connectivity among wetland patches in Tianjin approaches zero. Within this distance range, the wetland ecosystem in Tianjin exhibits scarce internal connectivity. Only when the distance threshold reaches 7.5 km does the IIC surpass 0.1, indicating a relatively high degree of connectivity. However, despite this, the IIC remains low, accentuating the weak connectivity between clusters of wetland patches. As the distance thresholds increase, the observed growth in connectivity within Tianjin’s wetland ecosystem is not pronounced. This finding further underscores the uneven distribution of wetland resources and the inadequate quantity of patches present in Tianjin.

Through an examination of the dIIC, it is evident that the average dIIC of wetland patches in Tianjin experiences a substantial increase as the distance thresholds increase (

Table 3. dIIC of wetland ecosystem patches in Tianjin under different distance thresholds in 2018.

Distance Thresholds/km	dIIC
2.5	0.1337
5	0.1953
7.5	0.2671
10	0.3232

), thereby underscoring their role as crucial stepping stones. However, it is noteworthy that, regardless of the distance threshold, highly important wetland patches tend to be concentrated in specific regions along the coast, rivers, and adjacent to reservoirs (Figure 8). Consequently, these regions exhibit robust connectivity within their respective wetland ecosystems. In contrast, the urban districts and surrounding areas, comprising six urban districts and four hugging the city, lack significant wetland patches, leading to diminished connectivity within their regional wetland ecosystems. Furthermore, a detailed analysis presents statistical data on the type and quantity of wetland patches ranked within the top 200 based on their differential importance index (dIIC) for different distance thresholds (

Table 4. Number of wetland patches in the top 200 dIIC at different distance thresholds in 2018.

Patch Type	Distance Thresholds of 2.5 km	Distance Thresholds of 5 km	Distance Thresholds of 7.5 km	Distance Thresholds of 10 km
Super-large wetland patches	23 (79.31%)	23 (79.31%)	23 (79.31%)	23 (79.31%)
Large wetland patches	24 (66.67%)	21 (58.33%)	24 (66.67%)	24 (66.67%)
Medium wetland patches	92 (36.65%)	118 (47.01%)	133 (52.98%)	144 (57.37%)
Small wetland patches	61 (0.08%)	38 (0.05)	20 (0.03%)	9 (0.01%)

Note: Percentages indicate the proportion of the top 200 dIIC patch types to the total number of that patch type in the study area.

). The findings reveal that a majority of the super-large and large wetland patches fall within the top 200 rankings, with nearly half of the medium-sized patches also occupying this range. In contrast, only a limited number of small-sized wetland patches are represented within the top 200. These results emphasize the critical importance of conserving and restoring super-large, large, and medium-sized wetland patches to ensure the connectivity of Tianjin’s wetland ecosystem, warranting targeted conservation efforts. Additionally, efforts should also focus on enhancing the stepping-stone functionality of small-sized wetland patches.

4. Discussion

Building on the preceding analysis of the spatial distribution characteristics and spatial relationships of wetlands in Tianjin, this study seeks to explore viable strategies for wetland spatial planning and assess their feasibility. The primary objective is to facilitate the sustainable development of Tianjin.

4.1. Wetland Restoration

Extensive human intervention and the widespread hardening of riverbanks have led to the degradation of river wetlands. To facilitate their restoration, we recommend transforming the riverbanks into either permeable solid riverbanks or ecologically soft riverbanks. As elucidated in the Tianjin Water System Planning (2008–2020) program, the city’s river channels are classified into two hierarchical levels: primary channels (flood channels) and secondary channels (drainage channels). Given the imperative to uphold flood-control functionality, it is not feasible to implement a uniform transformation approach of soft riverbanks for all river channels. Therefore, prior to proposing strategies for the transformation of waterfront spaces, this study classifies the primary channels into two distinct types based on the rainfall distribution pattern and the spatial distribution of flood-storage zones, as represented in Figure 9a,b, respectively. These types encompass the rapid flow section and the slow flow section, with the former pertaining to river channels exhibiting higher flow rates within Tianjin, while the latter pertains to channels characterized by lower flow rates. Furthermore, considering the precipitation and flood flow distribution in Tianjin, along with the definition of secondary channels provided in the Water System Planning program, the secondary channels are further divided into two categories: those equipped with embankments (rapid flow section) and those lacking embankments (slow flow section). Finally, drawing on the specific classification of river channels depicted in Figure 9.c, this study presents a set of distinct transformation methods for waterfront spaces (

Table 5. Classification and transformation of river courses in Tianjin.

River Grade	River Classification	Classification Basis	Transformation Method
Primary river course	Rapid flow section	Heavy rainfall and flood-storage areas	Permeable hard hydrophilic embankments
	Slow flow section	Low rainfall and flooding	Ecological soft hydrophilic embankment
Secondary river course	Rapid flow section	Heavy rainfall and flood-storage areas	Permeable hard hydrophilic embankments
	Slow flow section	Low rainfall and flooding	Ecological soft hydrophilic embankment

).

Transformation of permeable hard hydrophilic embankments: river sections characterized by high rainfall and located within flood-storage areas are designated as rapid flow segments. These specific river channels require measures to ensure their capacity to effectively manage floodwaters. During the renovation of waterfront spaces, it is essential to preserve the structural integrity of the existing durable hydrophilic embankments. Additionally, the riverbed component necessitates engineering modifications aimed at surface hardening in order to enhance its capability to convey peak flood flows (Figure 10a). For primary rivers, a comprehensive ecological control zone is established on both sides of the rapid flow segments, spanning a minimum width of 100 m, with 50 m specifically designated as the absolute ecological control area. Correspondingly, for secondary rivers, an ecological control zone is established on both sides of the rapid flow segments, covering a minimum width of 20 m, with 10 m designated as the absolute ecological control area. In cases where hard paving is essential for waterfront spaces, the existing impervious surfaces should be transformed into permeable pavements with functionalities, such as water permeability and filtration.

Ecological transformation of soft and hydrophilic riverbanks: river sections characterized by low rainfall and flood flow rates are demarcated as gentle flow segments. These particular river segments do not bear the primary responsibility for flood control, thereby presenting an opportunity to reconfigure their waterfront spaces into soft and hydrophilic ecological riverbanks (Figure 10b). For primary rivers, an ecological control zone is established on both sides of the gentle flow segments, spanning a minimum width of 100 m. Within this zone, 50 m is designated as the absolute ecological control area. To achieve this, surface hardening engineering modifications are implemented on the riverbed, while concurrently ensuring the retention of the primary river’s fundamental flood-management capacities. The riparian zones of the low flow sections in secondary river channels are designated as ecological redlines, with a minimum width of 20 m on both sides. Among them, a 10 m strip is designated as an absolute ecological control zone, where the ecological soft riverbed is preserved, and the already hardened riverbed is restored to its natural soft state. In the planning of ecological and soft waterfront embankments, the previous approach of engineering hardening should be abandoned; instead, an ecological transformation approach should be adopted to restore the original ecological shoreline of the water body and enhance the water-holding capacity of the waterfront space during periods of high water levels (Figure 10c).

4.2. Wetland Creation

Based on the current situation of uneven patch sizes, a high degree of fragmentation, highly imbalanced resource distribution, and a low IIC in the Tianjin wetlands, increasing the number of wetlands can enhance the quality of the wetland ecosystem. With a focus on wetland blank areas in Nankai District and other regions, potential spaces for wetland construction can be identified by analyzing unused land from the 2018 Tianjin land-use map (Figure 1). This selection process should be conducted in conjunction with the publicly available planning proposals, including the “Tianjin Urban Master Plan (2005–2020)”, the “Draft Plan for Urban Parks and Surrounding Areas in the Outer Ring of Tianjin City”, and “Planning for the Green Ecological Barrier Zone between the Twin Cities of Tianjin (2018–2035)”. The green ecological barrier zone proposes a three-tiered control system for the green ecological barrier. Specifically, within the first-tier control zone, taking into account the spatial features of lakes, ponds, rivers, salt marshes, and other water bodies, the nature, location, and specific scope of wetland planning are determined, along with the overall development goals for wetlands. In the second-tier control zone, the focus is on addressing the structure and layout of the wetland system within the ecological barrier zone, specifying the intensity of wetland resource utilization, and controlling the scale of construction and development intensity around the wetlands. In the third-tier control zone, the emphasis is on intrinsic development, the reasonable division of wetland functional zones, the formulation of zone-specific construction guidelines, the clarification of various wetland characteristics, and the orderly promotion of regional ecological renewal. These plans, together with the spatial distribution of water bodies, such as lakes, ponds, rivers, and salt fields, serve as the key criteria for determining the layout of additional wetland spaces (Figure 11).

Due to the presence of spatial autocorrelation among wetlands, wetlands can facilitate their own growth. When expanding the existing wetlands, it is advisable to extend the area around them by incorporating other sections of unused land. For instance, this can be implemented in wetlands such as the Zhouhe Wetland and the Beidagang Wetland. Regarding the creation of new wetlands, the primary focus should be on utilizing water bodies such as lakes, ponds, rivers, and salt fields as the foundation. This involves establishing large-scale wetland parks and wetland patches that serve as stepping stones. Examples of these large-scale wetland parks include the Ziyahe Wetland Park, the Nanyunhe Wetland Park, the Jiyunhe Wetland Park, the Chaobaixinhe Wetland Park, and the Haihe Wetland Park (Figure 11). Situated upstream along the inbound rivers, these parks not only provide recreational functions but also contribute to purifying the water quality of downstream areas. Moreover, we recommend establishing three types of wetland conservation zones, namely core, buffer, and experimental zones, around the vicinity of Tianjin’s wetland areas.

4.3. Wetland Classification

On 28 August 2020, the Ramsar Convention released the most recent edition of “The List of Wetlands of International Importance,” revealing that China now boasts a total of 64 internationally significant wetlands (Ramsar, 2020) [55]. This marks an increase of seven sites, including the inclusion of the Beidagang Wetland in Tianjin, compared to the previous iteration of the list published on July 14. Additionally, on 16 March 2020, the National Forestry and Grassland Administration unveiled the “National List of Important Wetlands for 2020,” which encompasses the designation of the Beidagang Nationally Important Wetland and the Qilihai Nationally Important Wetland in Tianjin. This recognition serves as a testament to the achievements made in wetland protection and restoration efforts in Tianjin, which are acknowledged at the national level. The dual classification of the Beidagang Wetland in Tianjin as both an internationally significant wetland and a nationally important wetland underscores the efficacy of Tianjin’s endeavors in wetland conservation and restoration.

As of 2017, wetland classification systems have been established through legal and regulatory measures in 19 provinces and municipalities across China. Notably, certain provinces, such as Zhejiang, Liaoning, and Sichuan, have implemented a four-tiered classification system for wetlands, encompassing levels of international importance, national importance, municipal importance, and general importance. This classification framework primarily accounts for the presence of internationally significant wetlands within these regions. In contrast, prior to the inclusion of the Beidagang Wetland in the Ramsar List, Tianjin did not possess any internationally important wetlands. The “Regulations on Wetland Protection in Tianjin,” enacted in 2016, instead classified wetlands into three levels: national importance, municipal importance, and general importance. With the fortunate inclusion of the Beidagang Wetland as an internationally important wetland, a unique opportunity arises to optimize the wetland classification system in Tianjin. By doing so, the alignment and harmonization between Tianjin’s classification system and those of other provinces and municipalities at the national level can be enhanced. Ultimately, this adjustment will effectively underscore the pivotal role of wetland conservation and restoration, specifically for both internationally and nationally important wetlands, within Tianjin’s wetland classification system.

Under the guiding principles of wetland expansion strategies and the updated wetland list, the wetland classification system in Tianjin has undergone a process of optimization. The revised classification system incorporates four tiers: international importance, national importance, municipal importance, and general importance, as illustrated in Figure 12. Specifically, the Beidagang Wetland represents the internationally important category, while the nationally important wetlands consist of extensive and contiguous wetland areas, including the Yuqiao Reservoir Wetland, the Dahuangbao Wetland, the Qilihai Wetland, the Northern Coastal Saline Wetland, the Central Coastal Saline Wetland, the Southern Coastal Saline Wetland, and the Tuanbo Lake Wetland. Municipally important wetlands comprise linear wetland formations such as the Zhou River Wetland, the Chaobai Wetland, the Xinhe Wetland, and the Haihe River Wetland. The general importance classification encompasses lakes, ponds, and reservoir wetlands, excluding the aforementioned categories. Furthermore, the study follows the principle of nearest neighbors, connecting adjacent wetland patches to construct a new wetland network. Within the ecological network, patches that are relatively clustered or adjacent are most likely to aggregate into complete ecological patches, thereby further forming an ecological network. The optimization of the wetland classification system in Tianjin is intended to establish a comprehensive wetland framework that promotes the development and preservation of higher level wetlands in conjunction with lower level wetlands. This strategic approach addresses the existing spatial disparities in the distribution of wetlands and the limited availability of wetland resources within Tianjin’s current system. While sustaining the positive momentum of wetland restoration efforts in Tianjin, it is crucial to dedicate further endeavors toward enhancing the overall quality of wetland ecosystems in the region.

4.4. Strategy Validation

The effectiveness of the wetland planning strategy was assessed by investigating the spatial distribution pattern of the wetlands using a range of analytical methods, including the nearest neighbor index, the geographic concentration index, the Gini index, kernel density analysis, spatial autocorrelation analysis, and connectivity analysis. These analytical approaches were applied to examine the impact of the wetland planning strategy. Utilizing the wetland spatial database establishment approach discussed earlier in this paper, an analysis grid depicting the distribution of wetlands in Tianjin after the implementation of the planning strategy was generated (Figure 13a). Additionally, a point distribution map illustrating the geographic distribution of wetlands in Tianjin (Figure 13b), as well as a grid-based wetland ratio for Tianjin (Figure 13c), were derived from the analysis.

4.4.1. Increase in Wetland Area

The expansion of existing wetlands and the establishment of new wetland areas have yielded substantial advancements in patch diversity, spatial distribution, and connectivity. Specifically, a total of 294 wetland patches were implemented in Tianjin City, comprising 13 large-sized, 15 medium-sized, 109 small-sized, and 157 minor patches. Therefore, the collective wetland area in Tianjin City underwent an approximate augmentation of 553.97 km², resulting in a notable increase of 4.63 percentage points in terms of wetland coverage. Projections suggest that the wetland coverage will ultimately encompass 27.76% of the total area.

4.4.2. Optimization of the Wetland Spatial Distribution Pattern

The ArcGIS average nearest neighbor analysis tool (Spatial Statistics Tools—Analyzing Patterns—Average Nearest Neighbor) was employed to evaluate the enhancements in the spatial distribution of wetlands in Tianjin City resulting from the implementation of the planning strategies. The analysis focused on computing the nearest neighbor index for the postplanning wetland spatial distribution. It was determined that the average distance between wetland point features and their closest neighboring point features was 1406.4668 m, with an expected average distance of 1142.386 m. The resulting ratio of these distances, known as the nearest neighbor ratio (R), was found to be 1.231. This indicates that the spatial distribution type of wetlands in Tianjin City, after implementing the planning strategies, can be classified as a dispersed distribution (Figure 14a). Notably, the dispersion of wetland spatial distribution in Tianjin City has improved further compared to the current situation, where the R value is 1.1960. This enhancement successfully addresses the issue of insufficient wetland coverage in specific areas of Tianjin City.

After the implementation of the planning strategies, the geographical concentration index (G) of wetlands in Tianjin City was determined to be 2.57. Under the assumption of an even distribution of all wetland point features within the grid, the expected geographical concentration index (G̅) would be 1.79, where G surpasses G̅. Although some spatial imbalances in wetland distribution persist when following the planning strategies, the disparity diminishes significantly compared to the current conditions, where G stands at 2.87 and G̅ at 1.76. This indicates a notable improvement in the concentration level of wetlands in Tianjin City resulting from the planning strategies. The Gini index, calculated as 0.3835 for wetlands in Tianjin City after the planning strategies, falls within the range of 0.3 to 0.4, highlighting a transition from a highly uneven to a relatively balanced spatial distribution of wetlands in Tianjin City as a result of the planning strategies.

The kernel density of the wetlands in Tianjin City, following the implementation of the planning strategies, was computed using the ArcGIS kernel density analysis tool (Spatial Analysis Tools—Density Analysis—Kernel Density). Employing the natural breaks (Jenks) classification method, the wetland kernel density was categorized into five levels: high, relatively high, moderate, relatively low, and low (Figure 14b). Notably, after the application of the planning strategies, the wetlands in Tianjin City exhibit a comprehensive north-to-south coverage pattern. This observed pattern signifies a significant reduction in the spatial disparity of wetland distribution between the northern and southern regions of Tianjin City compared to the current state. These findings further validate the previously outlined enhancements, in terms of both wetland distribution types and concentration levels, resulting from the planning strategies in Tianjin City.

4.4.3. Improvements in Wetland Spatial Relationships

Upon analyzing the wetland database using Geoda 1.18.0 software, it was discovered that the minimum Euclidean distance between neighboring patches of wetlands, after the implementation of the planning strategies, amounted to 5883.56 m. Consequently, a bandwidth of 6000 m was employed when constructing the spatial weights. The global Moran’s I index analysis tool (Spatial Analysis—Univariate Moran’s I) and the local Moran’s I index analysis tool (Spatial Analysis—Univariate Local Moran’s I) were then applied to examine both the global and local spatial autocorrelations of wetland coverage in Tianjin City subsequent to the planning strategies. The findings are depicted in a scatterplot illustrating the global Moran’s I index for wetland coverage in Tianjin City after the implementation of the planning strategies (Figure 15a), as well as the LISA cluster map illustrating the spatial distribution of wetland coverage in Tianjin City after the planning strategies (Figure 15b). These figures clearly demonstrate an enhanced spatial autocorrelation in wetland coverage across Tianjin City following the implementation of planning strategies, with a conspicuous increase in the aggregation of areas with high wetland coverage and the expansion of wetland coverage hotspots.

The spatial connectivity analysis revealed notable improvements in the connectivity of the wetland ecosystem in Tianjin City across four distance thresholds: 2.5 km, 5 km, 7.5 km, and 10 km. Specifically, the increases in IIC were measured as 0.0228, 0.0315, 0.0395, and 0.0531, respectively (Figure 16) (

Table 6. Changes in the quality of the wetland ecosystem in Tianjin before and after planning.

	Number of Wetland Patch			Connectivity				Area
	Super Large	Large	Medium	2.5 km	5 km	7.5 km	10 km	Area	Wetland rate
Status quo	29	36	251	0.0301	0.0638	0.1056	0.1450	2768.84	23.13%
Postoptimization	42	51	360	0.0529	0.0953	0.1451	0.1981	3322.81	27.76%

). This comprehensive enhancement in connectivity has significant implications for the conservation and restoration of the wetland ecosystem in Tianjin City, affirming its positive impact on ecological sustainability.

5. Conclusions

In this study, we employed various methods to analyze the spatial issues related to wetlands in Tianjin City and proposed corresponding planning strategies to facilitate sustainable urban development in Tianjin. (1) Different analytical approaches were utilized to deepen the analysis beyond a single method. For instance, the nearest neighbor analysis identified the spatial distribution type of wetlands as dispersed, while the geographic concentration index revealed the uneven distribution of wetlands in Tianjin. Moreover, the Gini index indicated a high degree of imbalance in the distribution of wetlands within the city. These methods provided quantitative insights. Additionally, kernel density analysis visually represented the clustering patterns, offering spatial guidance for planning strategies. (2) Wetlands are not isolated entities in spatial terms; they exhibit varying degrees of interdependence. Spatial autocorrelation analysis can be employed to uncover the spatial relationships among wetlands across the entirety of Tianjin City, as well as to identify hotspots and cold spots in terms of wetland distribution. Such an analysis aids in identifying focal areas for wetland spatial planning. For instance, if there exists a spatial correlation among wetlands within the entire Tianjin City region, wetlands tend to exhibit either a “high–high” or a “low–low” clustering pattern. Expanding wetlands in proximity to existing ones can be considered by leveraging the spatial autocorrelation of wetlands. Furthermore, connectivity analysis can be used to assess the integrity of the wetland network in Tianjin City and evaluate the shortcomings associated with different wetland types. This analytical approach is instrumental in formulating spatial planning strategies tailored to the diverse wetland categories. Notably, medium-to-large-sized wetland patches in Tianjin City assume a pivotal role in preserving wetland connectivity and, therefore, necessitate targeted conservation initiatives. Consequently, wetland classification efforts should underscore the conservation significance attributed to medium-to-large wetland patches. (3) Subsequent to the implementation of the spatial planning strategies, the optimized spatial distribution of wetlands in Tianjin City was substantiated through the employment of spatial distribution analysis and spatial relationship analysis methods. This validation process resulted in the addition of a total of 294 wetland patches, leading to a significant increase of 4.63 percentage points in terms of wetland coverage. Notably, the previously clearly prevalent inequitable distribution of wetlands underwent a transformative shift towards a comparatively balanced configuration. The spatial extent of “high–high” wetland aggregation zones expanded, thereby amplifying wetland interconnectivity. Moreover, when considering the distance thresholds of 2.5 km, 5 km, 7.5 km, and 10 km, the wetland ecosystem in Tianjin City exhibited respective enhancements in the IIC of 0.0228, 0.0315, 0.0395, and 0.0531, respectively. These cumulative improvements collectively contributed to an overall enhancement in wetland connectivity.

In the broader context of wetland conservation research, it is evident that many studies have overlooked the recognition of ecological relationships among wetlands, resulting in limited spatial applicability in their planning efforts. This study, in contrast, incorporates spatial distribution correlations, connectivity metrics, and pathways among wetlands into its framework. This comprehensive approach enhances the basis for strategic decision-making and augments the practical guidance for wetland planning. It is important to note that previous research [17] has underscored the close linkage between human wellbeing and the surrounding environment, emphasizing the significant impact of wetland quality on the physical and mental health of nearby residents. While this study did not consider the ecosystem services provided by wetlands and their implications for human wellbeing, it acknowledges the need for future investigations into the ecological valuation of wetland resources and the development of planning methodologies aligned with value-conversion pathways. This stands as a key agenda for future research endeavors. In the future, we also plan to conduct research on multitemporal data using the future multiscenario simulation method based on historical data validation to develop more comprehensive spatial planning strategies for wetlands. Furthermore, the validation of these planning strategies is crucial to ensuring their accuracy and scientific rigor. Overall, this study’s findings can be extrapolated to other cities or regions and can serve as a foundation for investigating spatial planning strategies pertaining to additional ecological elements, such as forest ecosystems.

Some limitations of this study should be acknowledged. The primary focus of this paper lies in examining the ecological attributes of wetland spaces and analyzing their ecological spatial distribution and relationships. However, the concept of spatial planning has evolved to encompass a holistic approach that incorporates multiple elements, including social, economic, and ecological factors. To address these limitations in future research, we recommend incorporating additional attributes, such as the social characteristics of wetlands. Concurrently, network analysis techniques can be employed to assess the accessibility of wetlands for the population. By integrating these diverse aspects, a comprehensive wetland spatial planning strategy can be developed that not only maximizes ecological benefits but also takes into account social advantages, thus achieving a win–win outcome through nature-based solutions.