Construction and Optimization of Ecological Security Patterns Based on Ecosystem Services in the Wuhan Metropolitan Area

Abstract

Rapid urbanization has affected ecosystem stability, and the construction of ecological security patterns (ESPs) can rationally allocate resources and achieve ecological protection. Priority evaluation of critical areas can maximize the benefits of ecological protection, which is crucial for sustainable urban development. However, most prior studies have focused on assessing individual elements of the ESP, rarely considering both the protection priority of ecological sources and corridors. We constructed ESPs for the Wuhan Metropolitan Area (WMA) from 2000 to 2020 and evaluated the priority of ecological sources and corridors for protection. The findings indicated that high-level ecological sources exhibited higher overall landscape connectivity and ecosystem service values with lower patch fragmentation. The average area proportions of primary, secondary, and tertiary ecological sources in 2000, 2010, and 2020 were 41.11%, 23.03%, and 29.86%, respectively. High-level ecological corridors had shorter lengths and offered higher comprehensive ecosystem service values. The total length of secondary corridors exceeded that of primary corridors by 1951.19 km, 650.39 km, and 2238.18 km in 2000, 2010, and 2020, respectively. Primary corridors, which connected fragmented and isolated sources, should have their ecological land percentage increased to enhance connectivity. Secondary corridors connected two independent and distant sources, providing the basis for ecological protection in the intervening area, whose surrounding habitats should be protected. This study identifies the ecological protection priority and offers a theoretical basis and practical reference for balancing urban development with ecological protection.

1. Introduction

Rapid urbanization and irrational human activity have affected ecosystem stability and seriously threatened regional ecological security [1]. Ecological problems such as environmental degradation, soil erosion, and increased landscape fragmentation are becoming increasingly prominent [2,3]. Constructing an ecological security pattern (ESP) and enhancing ecological management and protection have remained pivotal approaches to balancing regional development and ecological security [4,5]. However, inevitable conflict exists between urban development and ecological protection, and prioritizing ecological security at the expense of economic growth may not represent an optimal solution. Therefore, effectively allocating resources to maximize ecological security within limited constraints has become a focus in the discourse on urban sustainable development.

ESP refers to a spatial configuration consisting of patches and key locations that are essential for critical ecological processes and ecosystem services [6,7]. Its primary goal is to identify critical areas for ecological security and to optimize the spatial arrangement of ecosystem components through targeted human interventions to enhance regional ecosystem resilience [8]. Constructing and optimizing ESPs can effectively promote ecological connectivity and maintain ecosystem stability, playing a crucial role in regional environmental protection [9,10]. Currently, the widely accepted framework for regional ESP identification involves “ecological sources determination—resistance surface creation—corridors identification” [11,12].

Ecological sources represent the key sites essential to regional ecological functions and processes [13,14]. Two main techniques exist for identifying ecological sources, the first step of ESP construction. Large-scale habitat patches like forests and grasslands, or nature reserves and scenic spots, are chosen as ecological sources by the first technique, direct identification. Despite being relatively convenient, this approach ignores the internal differences in land-use types and changes in ecosystem services, greatly influenced by human subjectivity [15,16]. Another is the indirect identification method, which creates an extensive index system of evaluation considering factors such as ecosystem service function and ecological sensitivity [17,18]. This approach is more comprehensive and objective but more complex and may be affected by trade-offs in ecosystem services [19,20]. Each of these methods has its advantages and disadvantages and should be chosen based on the specific conditions of the research area.

The foundation for determining ecological corridors is the resistance surface construction, which reflects the resistance to species migration between different ecological patches. Most early studies assigned ecological resistance coefficients based on land cover and expert experience. This approach often disregarded the internal distinctions among land-use types and failed to account for human factors [21,22]. Therefore, many studies have considered economic, demographic, and transportation factors, introducing indicators of human activity intensity to modify the basic ecological resistance surface [23,24]. Other studies have used various indicators to construct ecological resistance surfaces. Gao et al. [25] used the habitat quality method to construct resistance surfaces and identified the ESP of Changzhou City. Zhang et al. [26] integrated the ecological environment and economic characteristics, including vegetation cover, distance to water bodies, and GDP, into resistance surface construction. In general, the methods for constructing ecological resistance surfaces lack standardization. The widely used method is modifying the resistance surface based on specific factors.

Ecological corridors are vital channels for materials and energy to move, influencing the integrity of ecosystem service functions and the connectivity of regional landscapes. Circuit theory and the minimum cumulative resistance (MCR) model are the main methods for identifying ecological corridors [27,28]. The MCR model can identify important nodes of corridors but fails to recognize corridor width [29,30]. By analogizing ecological flow to electric currents with stochastic random walk characteristics, circuit theory overcomes these shortcomings [24,31]. Because of its ability to forecast intricate landscape movement patterns and identify critical ecological corridors and key nodes, circuit theory has been extensively utilized in ESP construction [24,32].

The Wuhan Metropolitan Area (WMA) is an inland urban agglomeration located in the middle reaches of the Yangtze River, characterized by its numerous rivers, lakes, and abundant wetland resources [33,34]. However, with the acceleration of urbanization in WMA, ecological issues such as cropland erosion, lake reclamation, and soil degradation are being exacerbated [35,36]. These problems exacerbate the imbalance between urban development and ecological security in WMA, and how to maximize the benefits of ecological protection by using limited resources is an urgent problem. Prioritized management of key areas can effectively enhance the overall function of regional ecosystems at lower costs, thereby maximizing the integrated ecological, economic, and social benefits [37,38]. Lu et al. [39] prioritized the construction and protection of ecological corridors in WMA based on the quality, function, and structure of ecological networks. Zeng et al. [35] identified prioritized protected areas and different development plans for WMA based on landscape connectivity and protected area effectiveness. However, most previous studies have typically focused on grading individual elements within the ESP, rarely comprehensively considering the protection priority of ecological sources and corridors. Such approaches often fail to fully capture the complexity and multidimensionality of ecological protection needs. Therefore, this study improved the existing ESP identification framework and optimized and constructed the ESP of WMA in 2000, 2010, and 2020. We systematically evaluated the ecological protection priority of ecological sources and corridors and deeply explored the multidimensional perspectives of ecological protection.

2. Materials and Methods

2.1. Study Area

The Wuhan Metropolitan Area (WMA) consists of 9 cities, including Wuhan, with a total area of 57,947 km² and a population of more than 31.8 million, situated in the eastern part of Hubei province, China (Figure 1) [40]. As a strategic region for the rise of central China, WMA is a comprehensive reform pilot for building an environmentally friendly society and resource-saving society in China [41]. Its landforms are mainly plains and hills, ranging from the Dabie Mountain Range in the northeast to the Mufu Mountain Range in the south (Figure 1c). WMA has a subtropical monsoon climate, with high temperatures, frequent rainfall, and distinct seasons, with an average annual temperature of 16.37 °C and an average annual rainfall of 1102 mm in 2020. Land-use types in the area are dominated by forest and cropland, accounting for 49.13% and 25.20% of the total area. As urbanization progresses, problems such as cropland degradation, reduction of ecological space, and the heat island effect are posing many challenges to regional ecological security [33]. Therefore, it is necessary to construct and optimize the ESP and explore the ecological protection strategy that maximizes benefits at minimal cost, seeking the optimal solution to balance urban economic development and ecological security.

2.2. Data Source

Relevant data for the research area in 2000, 2010, and 2020 were gathered (

Table 1. Pertinent data for the research area.

Data	Resolution or Scale	Source
Land use	30 m	GlobeLand30 (https://www.tianditu.gov.cn, accessed on 12 November 2023)
Meteorological data	-	China Meteorological Data Service Center (http://data.cma.cn/, accessed on 15 November 2023)
DEM	30 m	ASTER Global DEM (http://lpdaac.usgs.gov/, accessed on 18 November 2023)
Soil data	1:1,000,000	Soil Database of China for Land Surface Modeling (http://vdb3.soil.csdb.cn/, accessed on 19 November 2023)
Nighttime light data	1000 m	An improved time-series DMSP-OLS-like data (1992–2022) in China by integrating DMSP-OLS and SNPP-VIIRS [42]
NDVI	250 m	Land Processes Distributed Active Archive Center (http://lpdaac.usgs.gov/, accessed on 20 November 2023)
Statistics data	-	Hubei Province Statistical Yearbook (https://tjj.hubei.gov.cn/, accessed on 20 November 2023)

). Land use was categorized into cropland, forest, shrubland, grassland, water, wetland, construction land, and bare land. All spatial datasets were projected to the WGS_1984_Albers and resampled to a 30 m resolution. Meteorological data were obtained by spatial interpolation using the Anusplin package in RStudio 1.1.463.

2.3. Methods

Ecosystem services are crucial indicators for assessing the ecological environment and protection, contributing to the sustainable development of urban areas [43,44]. We applied the entropy weight method to integrate ecosystem services and used the landscape connectivity index to identify ecological sources. By integrating circuit theory, the gravity model, and Centrality Mapper, we propose an ESP optimization method to comprehensively evaluate the ecological protection priority of ecological sources and corridors. The research framework for ESP construction and optimization consists of the following five steps (Figure 2): (1) assessing ecosystem services; (2) identifying the ecological sources through the entropy weight method and landscape connectivity analysis; (3) constructing resistance surface and modifying it; (4) using circuit theory to construct the ESP; and (5) optimizing the ESP using the gravity model and Centrality Mapper.

2.3.1. Ecosystem Services Assessment

WMA has been experiencing rapid urbanization in recent years, resulting in significant urban expansion encroaching on cropland, forest, shrubland, grassland, water, and wetland [45]. Six ecological services were chosen for this research in response to these urbanization trends, including water conservation, soil conservation, carbon sequestration, habitat quality, food supply, and ecological recreation.

Table 2. Methods of six ecological services assessment.

Ecological Service	Method	Reference
Water conservation	InVEST’s Seasonal Water Yield Module	Sahle et al. [46]
Soil conservation	Revised Universal Soil Loss Equation (RUSLE)	Liu et al. [47]
Carbon sequestration	Carnegie-Ames-Stanford Approach (CASA)	Peng et al. [7]
Habitat quality	InVEST’s Habitat Quality Module	Li et al. [48]
Food supply	Statistical data obtained by spatial allocation using NDVI	Zhang et al. [49]
Ecological recreation	Recreation potential index based on landscape indicators	Schirpke et al. [50]

shows the methods and references for evaluating the six ecosystem services. Methods for assessing the different ecosystem services are presented in the Supplementary Materials.

2.3.2. Ecological Source Identification

The key ecological patches that supply ecosystem services and sustain ecological processes are known as ecological sources [51]. Ecological sources were determined by assessing ecosystem services and landscape connectivity. The entropy weight method is a technique independent of evaluator preferences, relying only on the entropy of the original index [52]. Therefore, the entropy weight method was used to weigh and combine six ecosystem service functions and quantify their relative importance. This importance was ranked across five levels, with the highest-ranked patches selected as candidate ecological sources [15].

The ecological sources were identified by considering landscape connectivity to maintain their integrity. Indicating how much the landscape promotes or impedes ecological flow, landscape connectivity is a crucial factor affecting biodiversity, ecosystem stability, and integrity [53]. Possible connectivity (PC) is an ideal index for evaluating landscape connectivity, with values ranging from 0 to 1, increasing as connectivity improves [16]. The dPC (%), derived from the PC value, represents the contribution of patches to overall landscape connectivity. The specific calculation formulas are as follows:

$$PC={\displaystyle \frac{\sum _{i=1}^n\sum _{j=1}^n{a}_i\times a_j\times p_{ij}}{A_L^2}}$$

(1)

$${dPC}_i=100\times {\displaystyle \frac{PC-{PC}_{i-remove}}{PC}}$$

(2)

where $n$ indicates the number of nodes overall, $A_L$ denotes the entire study area, $a_i$ and $a_j$ indicate patch $i$ and patch $j$ areas, respectively; $p_{ij}$ denotes all path final connectivity maximum between patch $i$ and patch $j$, and ${PC}_{i-remove}$ indicates the potential connectivity of the remaining patches following the patch $i$ removal.

ArcGIS 10.4.1 and Conefor Sensinode 2.6.0 were used to calculate PC and dPC, and ecological patches whose area was less than the minimum patch area threshold were eliminated. Ultimately, we chose patches as ecological sources that were bigger than 2 km² and had dPC > 0.5.

2.3.3. Resistance Surface Construction

Ecological resistance surface represents the difficulty faced by species migrating between different patches [7]. Using expert experience in assigning resistance values to the corresponding land-use type is a common approach [54]. However, this approach does not adequately describe the resistance since it ignores variations in the same land-use type [55]. Therefore, human and natural factors should be included for comprehensive consideration.

In this study, the indicators used to construct the basic resistance surface—land-use type, altitude, and slope—were selected based on the conditions of WMA (

Table 3. Resistance factors of ecological security pattern in WMA.

Resistance Factors	Classification	Resistance Value	Weights
Land use	Forest	1	0.6
	Shrubland	10
	Grassland	10
	Cropland	30
	Water	50
	Wetland	50
	Bare land	300
	Construction land	500
DEM (m)	<100	1	0.2
	100–200	40
	200–400	60
	400–700	80
	>700	100
Slope (°)	<5	1	0.2
	5–10	40
	10–20	60
	20–30	80
	>30	100

). The weightings for each resistance factor were determined based on relevant literature [56]. In addition, Nighttime Light Intensity (NLI) was incorporated to optimize the ecological resistance surface, as it better reflects the impact of human activities on species movement [57]. The calculation method is as follows:

$$R_i={\displaystyle \frac{{NL}_i}{{NL}_a}}\times R$$

(3)

where $R_i$ indicates the pixel $i$’s value of modified resistance, ${NL}_a$ represents land-use type a’s average NLI, ${NL}_i$ denotes the pixel $i$’s NLI, while $R$ denotes the basic resistance coefficient.

2.3.4. ESP Construction

An ESP is composed of ecological sources, corridors, and nodes, serving as essential for biodiversity conservation and integrity of the ecosystem, which is advantageous for maintaining ecosystem service functions and safeguarding ecological security [10]. High-frequency areas of ecological flow are referred to as pinch points, having a high possibility for species migration. Pinch points often form when surrounding ecological resistance compresses the corridors in a relatively narrow area [58]. Preservation and restoration in this region should be prioritized, as the loss or deterioration of pinch points will decrease landscape connectivity. Areas where species are prevented from migrating between ecological sources are known as barriers. The removal of barriers is essential to enhance connectivity and protect and restore ecosystems [59].

This study identified corridors, pinch points, and barriers based on circuit theory using the Linkage Mapper 2.0.0 toolbox in ArcGIS. Circuit theory is based on the principle that electrons in a circuit exhibit random walks [60]. Circuit theory allows researchers to integrate all potential routes between sources and predict species migration probabilities by analyzing ecological processes [61]. In circuit theory, ecological sources and landscapes are represented as circuit nodes and conductive surfaces, respectively, while material or energy is analogous to electrons. In this framework, the potential, probability, and degree of obstruction in the flow of ecological information between sources are represented by voltage, current, and resistance, respectively.

2.3.5. Optimization of the ESP

Ecological sources and corridors are the crucial components of an ESP, essential to sustaining ecological processes. We optimized the ESP by determining their relative importance. The Centrality Mapper in Linkage Mapper can be utilized to calculate the centrality of ecological sources, thereby determining their significance and contribution to maintaining the connectivity of the regional ecological network [62]. Thus, we assessed ecological sources by Centrality Mapper and referred to relevant studies [26] to classify them utilizing the natural break point method into three levels: primary, secondary, and tertiary. The interaction strength across ecological corridors, directly proportional to their importance, was assessed using the gravity model [63]. Based on the gravity model, we subdivided ecological corridors into two levels: primary (interaction force ≥ 100) and secondary (interaction force < 100) [64]. The following is the computation method:

$$G_{ij}={\displaystyle \frac{\left(\frac{1}{P_i}\times {lnS}_i\right)\left(\frac{1}{P_j}\times {lnS}_j\right)}{{\left(\frac{L_{ij}}{L_{max}}\right)}^2}}={\displaystyle \frac{{\left(L_{max}\right)}^2{lnS}_i\times {lnS}_j}{{\left(L_{ij}\right)}^2{P}_i\times P_j}}$$

(4)

where $G_{ij}$ and $L_{ij}$ represent the interaction and the corridor cumulative resistance value between source $i$ and $j$; $P_i$, $P_j$, $S_i$, and $S_j$ denote the value of resistance and area of the sources, respectively; while $L_{max}$ represents the total ecological corridors’ maximum cumulative resistance.

3. Results

3.1. Land-Use Change

The land-use changes in WMA are depicted in Figure 3. From 2000 to 2020, cropland, forest, and water were the dominant land-use types, comprising around 85% of the region. Cropland was primarily located in the central and western plains. Water was distributed along the Yangtze River through the plains, primarily in the central region. Forest was predominantly found in the northeast and south, notably in the Dabie and Mufu Mountains. Construction land was concentrated in Wuhan City, with the remainder scattered across other urban areas.

Over the past 20 years, cropland decreased the most significantly by 2.5%, and construction land increased the most by 2.3% (Figure 4). Most new construction land was converted from cropland (

Table A1. Land-use transfer matrix for WMA from 2000 to 2020 (km²).

Year	Land-Use Type	2020
		Cropland	Forest	Shrubland	Grassland	Water	Wetland	Construction Land	Bare Land	Total
2000	Cropland	27,999.92	200.63	36.64	11.78	449.58	39.54	1173.82	7.15	29,919.06
	Forest	179.37	14,277.79	44.93	31.50	28.63	7.45	153.47	4.67	14,727.81
	Shrubland	37.78	54.27	2698.55	4.45	5.06	1.34	19.35	1.22	2822.02
	Grassland	11.23	36.40	4.84	1353.85	13.50	2.29	21.43	1.85	1445.38
	Water	91.67	16.72	3.06	5.82	4731.23	107.46	97.20	23.65	5076.81
	Wetland	9.63	1.01	0.23	0.24	189.29	426.16	16.39	1.08	644.02
	Construction land	113.03	8.02	1.11	0.65	10.09	6.99	2926.85	1.98	3068.72
	Bare land	3.60	0.36	0.01	1.57	62.92	3.35	7.73	163.67	243.20
	Total	28,446.23	14,595.19	2789.37	1409.87	5490.30	594.58	4416.23	205.27	57,947.03

). Forest, shrubland, grassland, wetland, and bare land have generally declined, mainly due to the conversion of forest to cropland, shrubland and grassland to forest, and wetland and bare land to water. Water increased from 2000 to 2010, primarily converted from cropland and wetland (

Table A2. Land-use transfer matrix for WMA from 2000 to 2010 (km²).

Year	Land-Use Type	2010
		Cropland	Forest	Shrubland	Grassland	Water	Wetland	Construction Land	Bare Land	Total
2000	Cropland	27,826.37	274.21	51.97	22.46	724.23	139.80	862.86	17.17	29,919.06
	Forest	155.97	14,341.07	42.27	24.88	30.12	4.04	128.54	0.92	14,727.81
	Shrubland	28.95	48.51	2717.05	6.93	4.72	0.56	15.07	0.23	2822.02
	Grassland	10.31	38.30	13.28	1344.13	14.04	6.00	18.37	0.94	1445.38
	Water	182.32	15.35	3.10	5.73	4485.73	272.27	82.61	29.70	5076.81
	Wetland	86.01	2.42	0.40	0.93	289.56	237.93	16.77	10.02	644.02
	Construction land	84.77	10.11	2.21	1.59	29.82	3.29	2934.87	2.05	3068.72
	Bare land	11.01	1.28	0.02	1.75	51.85	51.56	8.02	117.71	243.20
	Total	28,385.70	14,731.26	2830.31	1408.39	5630.08	715.46	4067.12	178.73	57,947.03

), and decreased from 2010 to 2020, mostly converted to cropland (

Table A3. Land-use transfer matrix for WMA from 2010 to 2020 (km²).

Year	Land-Use Type	2020
		Cropland	Forest	Shrubland	Grassland	Water	Wetland	Construction Land	Bare Land	Total
2010	Cropland	26,737.00	265.64	51.15	16.01	293.50	87.81	920.55	14.04	28,385.70
	Forest	365.11	14,079.27	69.60	46.96	31.87	7.53	124.76	6.17	14,731.26
	Shrubland	73.86	72.67	2640.60	15.83	7.37	1.43	17.52	1.02	2830.31
	Grassland	26.45	39.54	10.12	1306.63	6.01	1.63	15.85	2.16	1408.39
	Water	476.01	33.37	5.73	7.46	4834.90	158.20	84.48	29.92	5630.08
	Wetland	112.85	3.23	0.59	3.94	251.40	314.72	12.87	15.85	715.46
	Construction land	638.10	100.51	11.50	12.16	50.00	13.74	3235.35	5.77	4067.12
	Barel and	17.12	1.08	0.01	0.79	15.20	9.42	4.82	130.28	178.73
	Total	28,446.51	14,595.31	2789.30	1409.78	5490.25	594.50	4416.18	205.20	57,947.03

).

3.2. Ecosystem Service Change

The spatial patterns of the six ecosystem services differed significantly over time (Figure 5). From 2000 to 2020, the high values of soil conservation, carbon sequestration, ecological recreation, and habitat quality services were predominantly concentrated in the southern and northeastern areas, with high vegetation cover and low human disturbance. The soil conservation value in 2020 reached 813.88 t/hm², representing a 17.4% decrease compared to 2000. Carbon sequestration in 2020 was 568.61 gC/m², a 22.2% increase over the past 20 years. Ecological recreation and habitat quality services had average values of 0.41 and 0.62 in 2020, increasing by 2.5% and decreasing by 1.6%, respectively. The total water conservation in 2020 was 3.45 × 10¹⁰ m³, an increase of 15.8% from 2000. The high-value areas were mainly located in the eastern and southern parts of the study area, where vegetation is rich and water storage capacity is high. Differences in climatic conditions have contributed to slight shifts in the spatial distribution of the high-value regions across different years. Food supply increased by approximately 436% over the past 20 years and reached 3.31 × 10¹¹ RMB in 2020, with the high-value areas mainly in the central and western croplands.

3.3. Identification and Change of Ecological Sources

The ecological patches were determined through a comprehensive analysis of ecosystem services and landscape connectivity. Figure 6 shows that both the area and number of patches below the threshold initially increase rapidly and then stabilize as the minimum area threshold for ecological patches increases. From 2000 to 2020, the turning points of total patch area and number in different periods were close to the threshold of 2 km². When the 2 km² threshold was applied, numerous unselected patches emerged. However, these patches were small and had a negligible impact on the spatial distribution of ecological sources (

Table 4. Patches with the area below the threshold between 2000 and 2020.

Year	Evaluation Index	Ecological Patches (Area < 2 km2)	All Ecological Patches	The Proportion of Ecological Patches (Area < 2 km2)
2000	Total area (km2)	987.04	6917.86	14.27%
	Number	29,005	29,170	99.43%
2010	Total area (km2)	847.09	7836.10	10.81%
	Number	13,752	13,945	98.62%
2020	Total area (km2)	1697.34	11,942.42	14.21%
	Number	29,422	29,843	98.59%

). Therefore, we chose 2 km² as the threshold in this research.

Figure 7 displays the spatial patterns of ecological sources during the three periods. In 2000, 58 ecological sources covered an area of 5179.98 km². In 2010, 49 ecological sources spanned 6037.47 km². By 2020, the total area and number had increased to 7711.83 km² and 78 in 2020. These sources were primarily located in the forested areas of the northeast and south. These regions had high vegetation cover, a well-preserved ecological environment, and high biodiversity. They also exhibited superior soil conservation, carbon sequestration, and water conservation capabilities.

3.4. Resistance Surface Change

Figure 8 shows that the basic resistance surface values ranged from 0 to 340, with average values of 37.34, 42.16, and 43.86 for the three respective periods. Considering the influence of human disturbances on the ecological resistance surface, we optimized the resistance surface using NLI. The average values of the modified resistance surface for 2000, 2010, and 2020 were 39.72, 47.41, and 50.44, respectively.

The modified resistance surface was generally similar to the basic resistance surface but showed local variations, particularly in the expansion of the middle and high resistance value areas in the central. High-resistance regions were predominantly concentrated in the central urban area of Wuhan and surrounding urban centers, where construction land was the dominant land-use type. Although the northeastern and southern regions generally had good ecological conditions, higher resistance values were observed due to elevation and slope.

3.5. Spatio-Temporal Change of the ESP

The ESP of WMA is depicted in Figure 9. In 2000 and 2010, there were 154 and 122 ecological corridors, with a total length of 2914.71 km and 1281.96 km, respectively. By 2020, the corridor numbers had increased to 178, while the total length had increased to 3475.39 km. They were primarily found in the east and south and later extended to the north. Predominant land-use types in those regions were forest, shrubland, and grassland.

From 2000 to 2020, we identified 244.9 km², 157.69 km², and 170.46 km² of ecological barriers, respectively. These barriers were primarily located on the fringes of ecological sources and corridors. The area of the pinch points was 236.58 km², 97.33 km², and 197.42 km² in 2000, 2010, and 2020, respectively. They were mainly found in the middle of long ecological corridors and close to the junction of ecological sources and corridors. Generally, the distributions of barriers and pinch points were spatially aligned with the corridors, primarily in the northeast and southeast of WMA.

3.6. Spatio-Temporal Change of Optimized ESP

Figure 10 shows the optimized ESP for WMA. In 2000, there were 8 primary, 20 secondary, and 30 tertiary ecological sources. By 2010, these numbers decreased to 5, 16, and 28, respectively. They increased to 9 primary, 21 secondary, and 48 tertiary sources in 2020. The three-year average area proportions of primary, secondary, and tertiary ecological resources were 41.11%, 23.03%, and 29.86%, respectively. From 2000 to 2010, primary and secondary corridors decreased from 107 and 47 to 91 and 31. However, from 2010 to 2020, they increased again to 141 and 37. The total length of secondary corridors exceeded that of primary corridors by 1951.19 km, 650.39 km, and 2238.18 km in 2000, 2010, and 2020, respectively.

Spatially, primary ecological sources were widespread and primarily located in the southern part of WMA. Secondary ecological sources consisted mainly of small fragmented sources between primary ecological sources or large sources far from the primary ones. Tertiary ecological sources were scattered primarily in peripheral areas, such as the northern and eastern boundaries, exhibiting significant fragmentation. Primary corridors were concentrated in the south and extended north and were characterized by a short and dense configuration. Secondary corridors were predominantly situated in the central and eastern regions, appearing long and dispersed.

4. Discussion

4.1. Priority Assessment of Ecological Protection

Constructing and optimizing the ESP is crucial for maintaining ecosystem functions, ensuring ecological security, and coordinating ecosystem protection with regional high-quality development [53,65]. Implementing ecological protection requires the rational allocation of resources, and it is more efficient to prioritize protection in key areas [22]. However, earlier research has primarily focused on prioritizing single elements within ESPs and has rarely considered both ecological sources and corridors simultaneously [35,39,66]. In contrast, this study integrates the priority of ecological sources and corridors in constructing the ESP, thus addressing ecological protection from multiple dimensions.

The high-level ecological sources in WMA were mainly distributed in the northeast and south. In contrast, low-level ecological sources were primarily in the north and surrounding these sources. This distribution characteristic is consistent with earlier research [35]. Compared to previous studies, this research also emphasizes the prioritization of ecological corridors for comprehensive ecological protection. The high-level corridors were mainly distributed in the northeast and south, while the low-level corridors were concentrated in the central region. This distribution is consistent with previous studies [39]. As a result of incorporating ecosystem services in ecological sources identification, the ecological corridors are not as widely distributed across WMA as in that study.

By counting the ecological sources in 2000, 2010, and 2020, we found that the higher-level ecological sources were characterized by a larger average area and dPC, except that secondary sources had a smaller average area than tertiary sources in 2000. Higher-level ecological sources exhibited greater landscape connectivity, lower levels of fragmentation, and higher ecosystem service value [26,67]. These findings are consistent with the conclusions obtained from the previous study [35]. Between 2000 and 2020, the primary land-use types within ecological corridors were cropland, forest, and water. The proportion of cropland in primary corridors was smaller than in secondary corridors, while the proportion of forests was higher. Moreover, the average integrated ecosystem service value in primary corridors was higher than in secondary corridors. Therefore, higher-level ecological corridors demonstrated better environmental conditions, lower ecological resistance, and more efficient ecological flows [68,69].

4.2. Analysis and Recommendations Based on Optimized ESP

We determined the ecological protection priority according to various priorities and proposed corresponding protection measures. Ecological sources are patches that provide essential ecological functions and maintain ecological security, which are crucial for promoting regional sustainable development [26]. Primary ecological sources were mainly intact forest patches, providing abundant resources and critical ecological services. It is necessary to prioritize their protection and preservation, establish nature reserves, and prohibit over-exploitation to ensure their connectivity and integrity [19,64]. Secondary and tertiary ecological sources were more fragmented and had the potential for expansion. Ecological restoration efforts should be strengthened to enhance the overall quality of these patches, improving their structure and connectivity [9]. In the central region with scarce ecological resources, parks and green spaces of appropriate scale could be constructed to enhance the connectivity between sources and cope with the ecological pressure brought by urbanization [70].

Ecological corridors are vital channels for connecting ecological sources, and strengthening their construction can help increase landscape connectivity and ecological protection [71,72]. Primary corridors were mainly short-distance corridors that connected fragmented and isolated sources, reducing fragmentation and enhancing ecological network integrity. Therefore, the protection and construction of these corridors should be prioritized to minimize anthropogenic interference and safeguard the stability of ecosystems [73,74]. Secondary corridors were long-distance corridors that connected two independent sources far from each other, providing a basis for ecological safety protection in the intermediate areas. The land-use composition should be optimized to ensure habitat quality and to reduce the negative impacts of nearby cropland and construction land [75]. Additionally, it is necessary to connect corridors to urban green spaces in urban centers and install stepping stones to reduce the resistance of long-distance corridors and improve landscape connectivity [26,76].

Barriers refer to areas that significantly impede ecological flow, while pinch points are critical regions where ecological flow is concentrated [24,72]. Removing barriers and protecting pinch points can significantly improve landscape connectivity. Barriers and pinch points in different land-use types should be managed with specific protective measures tailored to the unique ecological characteristics of each area [71]. In corresponding cropland areas, measures such as limiting agricultural cultivation, promoting sustainable agricultural practices, and encouraging eco-friendly agricultural tourism can help minimize ecological resistance [77]. In water and wetland areas, it is necessary to create buffer zones, plant protective forests along the shoreline, control pollutant discharges, and improve water quality [78,79]. In construction land areas, efforts should focus on strengthening ecological greening and establishing wildlife corridors [80].

4.3. Restrictions and Future Exploration

A scientifically effective approach to balancing urbanization with ecological preservation is through the construction and optimization of an ESP. This study evaluated the ecological protection priority of ESP critical areas and provided important ideas for maximizing the benefits of ecological conservation under sustainable urban development. However, several limitations in this study warrant further investigation and refinement.

First, the research on ecosystem services and their changes is insufficient. Although ecosystem services were selected according to the actual situation in this study, the accuracy was still affected by the limitations of data, models, and technologies. The changes in ecosystem services are also closely related to population growth, climate change [81], economic development, and ecological protection [82,83], and have a close relationship with each other. Under the influence of urbanization and environmental impacts, the regional effects of ecosystem services will not remain static but rather respond in complex ways. For instance, there are multi-scale response characteristics of environmental factors to ES constraints [84], and the early growth of ecosystem services may be offset by the negative impacts of urbanization [85], among other issues. These are directions we can explore in greater depth in the future.

Secondly, threshold values are an issue in ESP construction. Many studies on ESPs have categorized the evaluation results and classified ecological security levels with corresponding thresholds, ignoring the internal differences and complexity of the region, as well as the significant differences across various assessment regions [86]. Ecological security assessments also involve many threshold issues, such as the thresholds for high-value areas of ecosystem services, distance thresholds when using landscape connectivity indices to identify ecological sources [87], and specific cost-weighted distance thresholds for identifying ecological corridors [88]. Although this study set thresholds based on previous findings, it did not account for inter-regional variability, and future studies should select more appropriate thresholds based on field conditions.

Finally, ESPs need to be explored in depth. Analyzing the drivers affecting ESP changes can reveal the effects of anthropogenic interventions and the natural environment on the ESP, providing a basis for developing ecological optimization strategies [89,90]. Future land-use changes are also an important direction in exploring the dynamic changes of ESPs. Simulating future land-use change under multiple scenarios and constructing ESPs can assist in formulating land-use optimization policies under different socio-economic conditions [91,92]. In the future, we can conduct a more in-depth analysis of ESP changes, especially the drivers of changes in ESP elements of different priorities, or predict ESP changes by simulating future land-use changes. This would help identify ecological corridors, pinch points, and barriers that may be threatened in the future, enabling us to propose targeted conservation strategies.

5. Conclusions

The construction and optimization of ESPs can effectively promote the balance between urbanization and regional ecological protection. Prioritizing key protection areas is essential for efficient resource allocation to enhance ecological benefits. However, most previous studies have only considered the ecological conservation prioritization of a single element of the ESP with a limited focus on evaluating ecological sources and corridors together. Based on the ESP pattern of WMA, this study conducted a comprehensive assessment of the ecological conservation priorities of both ecological sources and corridors. The results show that higher ecological source grades indicate greater landscape connectivity, higher ecosystem service values, and lower fragmentation levels. Similarly, higher corridor grades correspond to shorter corridor lengths and higher ecosystem service values. Primary ecological corridors connect fragmented and isolated ecological sources, playing a vital role in maintaining ecological connectivity. Secondary corridors link two independent ecological sources far apart and provide a basis for ecological protection in the intervening area. Based on the optimized ESP, we proposed hierarchical protection and management strategies for critical areas. This study incorporates ecological priority conservation into the ESP framework, broadening its applicability. The research conclusions provide a theoretical basis and reference for balancing urban economic development and ecological protection.