Identifying and Optimizing the Ecological Security Pattern of the Beijing–Tianjin–Hebei Urban Agglomeration from 2000 to 2030

Abstract

The conflict between economic development and ecological protection continues to intensify, highlighting the necessity for constructing regional ecological security patterns (ESPs) to reconcile the relationship between development and protection effectively. This study used the GMOP and PLUS model to simulate future land use changes by 2030 under the following three scenarios: natural development (ND), ecological protection (EP), and economic development (ED). Employing the MSPA model and circuit theory, it identified ecological source areas and constructed the ESP for the Beijing–Tianjin–Hebei urban agglomeration (BTH) from 2000 to 2030. The results indicate that the proportion of ecological source areas increased from 22.24% in 2000 to 23.09% in 2020, the EP scenario showing the highest proportion of ecological source areas compared with the other two scenarios. These areas are densely distributed in the northern and western mountainous regions, with sparse distributions in the southern plains. The number of ecological corridors grew from 603 in 2000 to 616 in 2020, with the EP scenario having more corridors than the other two scenarios. From 2000 to 2030, corridors in the northern and western mountainous areas were denser, shorter, and more variable, while those in the southern plains were less dense, longer, and relatively stable. Over two decades, habitat areas for species in BTH increased, while landscape connectivity decreased. Compared with 2020 and the other two scenarios, the EP scenario saw an increase in habitat areas and improved landscape connectivity. The impact on ecological corridors and improvement areas primarily arose from a combination of socio-ecological drivers (e.g., elevation, slope, population), while the influence on restoration and key areas mainly stemmed from ecological factors (e.g., elevation, temperature, NDVI, precipitation). The findings demonstrate that distinguishing different geomorphological units to improve and restore the regional environment, while considering socio-ecological drivers, is crucial for restoring the overall ESP and landscape connectivity of BTH.

1. Introduction

The excessive exploitation of natural resources by humans often leads to a decrease in the quantity, quality, and connectivity of ecological lands such as green spaces, water bodies, and wetlands [1,2]. This diminishes the original regulatory mechanisms of ecosystems, leading to a decline in ecosystem functions and an increase in ecological risks, thereby posing a serious threat to ecological safety [3,4,5]. Ecosystem functions are primarily manifested through ecosystem services, which provide various material and resource guarantees for humanity, including supply, regulation, support, and cultural benefits [6]. Ecosystem services are not only crucial for ensuring regional sustainable development and balancing human well-being but also provide the capacity for regional ecological development and economic growth [7]. They are closely related to regional ecological security [8].

Ecological security represents a stable state where a healthy ecosystem provides a resource and environmental foundation for the long-term development of human society [9]. It is often linked to the goals of sustainable development [10]. The realization of ecological security relies on the planning of ecological security patterns (ESPs). Such planning can effectively connect fragmented habitat patches to form a complete ESP, thereby enhancing the quality of ecosystems [11,12]. Therefore, in the process of constructing regional ESPs, it is essential to consider fully the changes in ecosystem services and land use in the region. Besides advancing the construction of regional ESP, revealing how different socio-ecological driving factors affect these patterns is also crucial. This understanding is vital for identifying the ecological and environmental driving mechanisms in specific areas and designing effective protective measures to achieve sustainable development in the region [13].

Research on ESPs has undergone significant evolution, moving from initial analyses focused on landscape components [14,15] and static configurations of patches to embracing dynamic simulations and functional optimizations that highlight ecological processes [16,17]. Additionally, the field has broadened its scope beyond mere adjustments in land-use structures to include a wide array of aspects. These aspects range from the dynamics among ESPs to enhancing regional ecological functions [18,19]. The methodologies for constructing and optimizing these patterns have seen continuous refinement, leading to the establishment of a coherent framework. This framework comprises identifying ecological sources, constructing ecological resistance surfaces, extracting ecological corridors, and pinpointing key ecological nodes [11].

Ecological corridors in ESPs serve as linear or strip areas facilitating the transfer of energy and information between ecological sources and species. Currently, the minimum cumulative resistance (MCR) method [20,21] is commonly used for identification, clearly displaying trends in ecological corridors but failing to recognize key nodes within these corridors, such as “pinch points” or “obstacle points” [22]. In recent years, scholars have attempted to identify ecological sources through Morphological Spatial Pattern Analysis (MSPA) [23]. This method emphasizes structural connectivity, using land use data and the value of ecosystem services to identify landscape types crucial for maintaining connectivity, thus enhancing the scientific basis for selecting ecological sources and corridors [24]. On the other hand, circuit theory has been applied to the study of ecological corridors [25,26]. It uses the path of electronic random traversal to simulate the diffusion patterns of ecosystem services or species in a region [27], which better reflects the flow of matter or energy through patches and corridors and can identify key nodes in the flow [28]. Therefore, integrating the MSPA model with circuit theory, and incorporating changes in ecosystem services and land use into the framework for constructing regional ESPs, addresses the limitations of the MCR model in identifying ecological sources, corridors, and key node regions. This approach more accurately constructs regional ESPs from a comprehensive point–line–plane perspective.

In recent years, research on identifying national spatial restoration regions based on ESPs has received increasing attention. A series of innovative achievements have been made in arid areas [29], tropical rainforests [30], karst mountainous areas [31], landscape resource-based cities [32], and coal resource-based cities [33]. At the same time, many research achievements have been made in the optimization mode of ESPs. For example, Jiang et al. (2024) [34] optimized ESPs by evaluating plaque stability and network connectivity. Kim et al. (2023) [35] evaluated the ecological connectivity at the island scale based on ecosystem service functions and identified ecological pinch points to determine areas that need protection and restoration. Wang et al. (2023) [36] evaluated and optimized regional ESPs based on habitat quality, ecosystem service value, and soil and water conservation function. There have also been many research findings in the construction of ESPs for urban agglomerations, such as the Yangtze River Delta urban agglomeration [37], the middle reaches of the Yangtze River urban agglomeration [38], the Guangdong–Hong Kong–Macao Greater Bay Area urban agglomeration [39], the Chengdu–Chongqing urban agglomeration [40], and so on. These studies are mainly focused on a specific year or multiple years in time, but fewer studies explore the evolution of ESPs in urban agglomerations and make predictions for the future under various scenarios. In addition, the rapid development of urban agglomerations has encroached on ecological land, weakened ecosystem service functions, and posed a serious threat to regional ESPs, which is a concern that cannot be ignored. Therefore, in-depth research on the past, present, and future ESPs of large urban agglomerations is of great practical significance for national spatial planning and regional socio-economic sustainable development.

An ESP is often influenced by many social and ecological factors because of its continuous and dynamic characteristics [34]. As a type of spatial basic information reflecting comprehensive interventions of regional geography, climate, and socio-economic factors [41], the formulation and implementation of regional ecological policies often need to be based on long-term monitoring of an ESP and its driving factors. Wei et al. (2023) [29] found that precipitation, slope, elevation, evapotranspiration, and NDVI are the main driving factors affecting the ecological source of urban agglomerations in the arid regions of Xinjiang. Liu et al. (2023) [42] confirmed that the combined effects of natural and human factors have an impact on the ESP of the Fenhe River Basin, forming typical characteristics of mountainous watershed units and basin administrative units. Therefore, accurately grasping the main driving factors that affect the evolution of ESPs has important indicative significance for taking targeted ecological protection measures to promote an improvement in regional ESPs in the future. At the same time, it also enhances the ecological contribution of the construction of ESPs and the identification of influencing factors in national ecological planning, ecological protection red line delineation, and other construction work [43,44].

Predicting land use changes can help optimize ESPs and improve regional ecological security [45]. Various models have been developed and applied to land use simulation and prediction, such as the cellular automata (CA) model [46], the SD-CA model [47], the conversion of land use and its effects (CLUE) model [48], and the CA-Markov model [49,50]. Each has its advantages, but none have simultaneously considered land use optimization and policy scenario simulation. The GMOP-PLUS model [51] has advantages in simulating the dynamic changes in natural land use types and their conflict boundaries [52]. In addition, it fills the gap in the existing research that only considers the restrictive role of planning (e.g., protected areas, no-build zones) and not the driving and guiding roles of planning policies [53]. Therefore, it can describe the land use changes generated by the spatial dynamic interaction among economic, ecological, and social systems, accurately depict land use patterns in different future scenarios, help understand and simulate complex land use systems, and play an important role in improving and optimizing ESPs [52].

The Beijing–Tianjin–Hebei (BTH) urban agglomeration, a pivotal region in China, is anchored by the capital city, Beijing, and its surrounding urban areas. Over the past two decades, BTH has faced prominent ecological issues such as a decline in ecosystem services, increased ecological risks, and ecological security degradation [54,55]. Currently, BTH is encountering unique ecological management challenges [56,57]. Inappropriate ecological management will hinder the region’s sustainable development process [58,59]. Therefore, research on the ESP of BTH warrants significant attention.

This study explores scientific methods to identify ecological sources, corridors, and critical nodes, including pinch points and barrier zones, to construct a regional ecological security framework. It integrates ecological service value (ESV) accounting with MSPA and land use change analysis to identify ecological sources accurately. Using circuit theory, this research aims to establish ecological corridors and define critical ecological structures, thus creating an ecological security model for BTH. It assesses ESPs from 2000 to 2020 and projects scenarios for 2030, analyzing their evolution. The GeoDetector tool helps identify key factors influencing BTH’s ESP. The results aim to provide insights for identifying essential restoration areas and supporting the development of environmental policies.

2. Materials and Methods

2.1. Study Area

BTH includes the entire territory of Beijing, Tianjin, and Hebei Province (Figure 1). The land area covers 218,000 km², accounting for about 2.3% of China’s total area. The region is hot and rainy in the summer, cold and dry in the winter, and has a typical temperate continental climate. The terrain slopes from northwest to southeast, with the northwestern part bordering Inner Mongolia and featuring extensive grasslands. The western region includes the Taihang Mountains and the northern region contains the Yan Mountains; both areas are rich in forest resources. The eastern region lies adjacent to the Bohai Sea, while the south is part of the North China Plain, characterized by its flat terrain.

2.2. Data Source

The primary datasets used in this study include (1) land use data of the Beijing–Tianjin–Hebei region for the years 2000, 2005, 2010, 2015, and 2020, which encompass seven major land use categories, i.e., farmland, forestland, shrubbery, grassland, water body, bare land, and construction land; (2) climate and environmental data including topography, climate, soil types, the hydrological network, and slope data; and (3) socio-economic data including the spatial distribution of population (POP), GDP, and the road network, as well as data on grain sowing area, yield, and price. Detailed data sources are listed in

Table 1. Data sources.

Data Type	Indicators	Temporal Coverage	Spatial Resolution	Data Source
Land use	——	2000, 2005, 2010, 2015, 2020	30 m	http://www.resdc.cn (accessed on 10 May 2023)
Climate and environment	DEM	2009	30 m	http://www.gscloud.cn (accessed on 8 May 2023)
	Slope	——	30 m	Converted from DEM data in ArcGIS 10.5
	Precipitation	2010, 2020	1 km	http://www.resdc.cn (accessed on 29 June 2023)
	Temperature	2010, 2020	1 km	http://www.resdc.cn (accessed on 29 June 2023)
	Soil types	1995	1:1 million	http://www.resdc.cn (accessed on 29 June 2023)
	River system	2020	1:1 million	http://www.webmap.cn (accessed on 12 July 2023)
Socio-economic	POP	2010, 2020	1 km	http://www.resdc.cn (accessed on 29 June 2023)
	GDP	2010, 2020	1 km	http://www.resdc.cn (accessed on 29 June 2023)
	Road network	2020	——	http://www.webmap.cn (accessed on 12 July 2023)
	Grain sowing area and yield	2010–2020	——	Statistical Yearbook of China
	Grain price	2010–2020	——	Compilation of national agricultural cost–benefit data

.

2.3. Methods

The logical framework for constructing the ESP in the BTH region is shown in Figure 2, with specific methods described in detail in the following subsections.

2.3.1. ESV Estimation

The ESV was quantified with the equivalent factor method [60]. The economic value of major grain crops in the Beijing–Tianjin–Hebei area (wheat, corn, and rice) was calculated to estimate the value of the standard equivalent factors. The formula is as follows:

$$E={\displaystyle \frac{1}{7}}{\sum }_{i=1}^n{\displaystyle \frac{m_i{p}_i{q}_i}{M}}$$

(1)

where E represents the economic value of per unit ecosystem service (CNY·ha⁻¹), i denotes the type of grain crop, m_i is the average price of the ith grain crop (CNY·kg⁻¹), p_i is the yield per hectare of the ith grain crop (CNY·ha⁻¹), q_i is the planting area of the ith grain crop (ha), and M is the total planting area of grain crops (ha).

The formula for the ESV is as follows:

$$ESV={\sum }_{i=1}^n{\sum }_{j=1}^n{e}_{ij}\times E$$

(2)

where ESV represents the total ESV(10⁸ CNY), i is the ecosystem type, j is the ecosystem service function type, and e_ij is the equivalent factor of the i-type ecological service function in the j-type ecosystem.

2.3.2. Simulation of Land Use Scenarios for 2030

By optimizing the land use structure through the GMOP model [61,62], this study estimated the land use area under different scenarios. The PLUS model [49,52] predicts the land use distribution for 2030 under these scenarios, providing a basis for quantitatively assessing the ESV during different periods.

Considering land use characteristics and the territorial spatial planning (2021–2035) and main functional zoning of Beijing, Tianjin, and Hebei Province, this study set the following three scenarios for simulating land use in 2030: natural development (ND), ecological protection (EP), and economic development (ED). ND follows the historical trend in development without strict restrictions on land conversion, ignoring policy and planning impacts. EP prioritizes ecological protection to maximize ecological benefits, while ED focuses on economic growth to maximize economic returns. The areas of different classes in different scenarios were calculated by Lingo18.0 [42], and the constraints are shown in

Table 2. Constraints of BTH land use prediction.

Constraint Types	Constraints (km2)	Instructions
Land area (xi)	$S={\sum }_{i=1}^7{x}_i$	The total land use area remains unchanged in each scenario.
Cultivated area (x1)	92,372.04 ≤ x1 ≤ 94,446.99	The cultivated land area in each scenario is based on the upper limit of the research area in 2020, and the lower limit is based on the area predicted by Markov Chain.
Forest area (x2)	53,338.5 ≤ x2 ≤ 66,645.95	During the research period, forest land showed an increasing trend, with the area of the study area in 2020 as the lower limit and the Markov Chain prediction area of 120% as the upper limit.
Shrub land area (x3)	642.33 ≤ x3 ≤ 781.92	During the research period, shrub land showed an increasing trend, with the lower limit being the area of the study area in 2020 and the upper limit being 120% of the Markov Chain predicted area.
Grassland area (x4)	x4 ≥ 32,045.94	The grassland area in each scenario shall not be lower than the current value.
Water area (x5)	x5 ≥ 2852.64	The water area in each scenario shall not be lower than the current value.
Wasteland area (x6)	40.05 ≤ x6 ≤ 47.88	As the demand for land increases, the intensity of unused land development will also increase, with the research area in 2020 as the upper limit and the Markov Chain predicted area as the lower limit.
Construction land area (x7)	32,585.67 ≤ x7 ≤ 42,361.37	Because of the development of urban construction, the construction land area is not lower than the existing area, and the increase is limited to the construction land increase indicator planned in the 2020 research area.
Ecological service value	$\sum _{i=1}^7{g}_i\times x_i$ ≥ $\sum _{i=1}^7{g}_i\times x_i$2020	The overall ecosystem service value will be higher in 2030 than in 2020. gi: Ecological service value in different land types.
Economic benefit value	$\sum _{i=1}^7{c}_i\times x_i$ ≥ $\sum _{i=1}^7{c}_i\times x_i$2020	The overall economic benefits in 2030 will be higher than in 2020. xi: Economic benefit value in different land types.

.

2.3.3. Identifying Ecological Source Areas

Ecological source areas play a crucial role in maintaining ecosystem stability, providing various ecological services, and ensuring regional ecological security because of their high habitat quality [63,64]. The size of ecological sources directly affects the flow and cycle of matter and energy in the region and the overall performance of ecosystem services [65]. Based on previous studies [66], this research calculated the annual ESV for the study area, classified it into five categories, selected the top two categories, and overlapped them with core areas identified using the MSPA model to determine significant ecological sources larger than 10 km².

This study considered forestlands, shrubberies, grasslands, and water bodies with good environmental quality as foreground values and farmlands, construction land, and bare land with significant human disturbance as background values. Using Guidos Toolbox 3.2 software, it identified seven landscape types within the study area, including core areas, ring zones, gaps, islets, edges, and connectors, by distinguishing between foreground and background values. These landscape types are crucial for further identifying ecological sources and understanding the role of various patches in the BTH region’s ecosystem [67].

2.3.4. Build Resistance Surfaces

Before determining the ecological corridor, it is necessary to first construct a resistance surface. Human activities and related natural factors often hinder the cycle of matter and energy within ecosystems. This study identified seven critical factors for the development of resistance surfaces including the following: type of land use, NDVI, distance from residential area, distance from water source, elevation, slope, and the degree of soil erosion [68] (

Table 3. The composition, weight, and score of the ecological resistance coefficient in BTH.

Resistance Factor	Weight	Score Assignment
		5	4	3	2	1
Land use type	0.2203	construction land, bare land	farmland	grassland	forestland, shrubbery	water body
NDVI	0.1839	<0.3	[0.3, 0.5)	[0.5, 0.6)	[0.6, 0.8)	>0.8
Distance from water source (m)	0.1912	>2000	[1000, 2000)	[500, 1000)	[100, 500)	[0, 100)
Distance from residential area (m)	0.1130	[0, 200)	[200, 500)	[500, 1000)	[1000, 2000)	>2000
Altitude (m)	0.0802	Divide into 5 categories from large to small
Slope (°)	0.1004
Soil erosion degree	0.1110

). These elements were then systematically weighted and integrated using the “Raster Calculator” function in ArcGIS 10.5, culminating in a detailed resistance surface for the study region. The “Raster calculator” in ArcGIS can realize the mathematical operation of the pixel value corresponding to two or more raster layers. We used the “Raster calculator“ of ArcGIS to multiply the grid layers corresponding to the 7 resistance factors in Table 3 by their respective weight values and add them up, finally obtaining the face value of ecological resistance in different periods of BTH.

2.3.5. Construction of the ESP

An ESP is conceptualized as a circuit structure, utilizing ecological resistance surfaces as resistances, and simulating the flow of current through the ecosystem under a constant voltage. Current density, representing the magnitude of current passing through an individual pixel, signifies the likelihood of matter and energy traversing a region among ecological sources. Regions exhibiting high current density are designated as ecological pinch points, often marking critical conduits for the transmission of matter and energy. This stochastic movement of matter and energy follows the electrical current formula (I = U/R) [22], where R denotes the resistance offered by landscape patches against the movement and exchange of matter and energy, U represents the likelihood of matter and energy transitioning from one node to another, and I quantifies the probability of matter and energy migrating along a given pathway.

Utilizing ArcGIS 10.5 in conjunction with the Linkage Mapper 2.0.0 toolbox (https://circuitscape.org/linkagemapper (accessed on 16 August 2023)), the Linkage Pathways Tool module identifies ecological corridors and ecological corridors in a study area [66]. The Pinch Point Mapper module identifies regions of ecological pinch points, areas requiring priority protection because of their high ecosystem service migration density. The Barrier Mapper module, with a maximum and minimum search radius of 3000 and 600, respectively, and a step size of 1200, identifies ecological barrier point areas, where higher barrier point values indicate a greater increase in landscape connectivity upon removal [69].

Ecological sources are designated as ecological protection zones, and the flow density of ecosystem services is classified into three categories [66]. The top two categories are designated as ecological improvement zones, while first- and second-level ecological barrier areas are marked as ecological restoration zones. First- and second-level ecological pinch point areas are designated as critical ecological zones, and ecological corridors longer than 10,000 m are identified as protected corridors. Potential ecological corridors longer than 10,000 m are designated as restoration corridors, constructing the ESP.

2.3.6. Stability Analysis

The coefficient of variation not only reflects the degree of fluctuation in a variable but also tests whether the variable is stable in the time series. This study used the coefficient of variation to measure the stability of various components of the ESP. The formula is as follows:

$$CV={\displaystyle \frac{\sigma }{\mu }}\times 100\%$$

(3)

where σ is the standard deviation and μ is the average value. The CV values were divided into four categories as follows: lowest fluctuation (CV ≤ 10%); lower volatility (10% < CV ≤ 20%); medium fluctuation (20% < CV ≤ 30%); high volatility (CV > 30%) [70].

2.3.7. Impact of Socio-Ecological Drivers on the ESP

We employed the Geodetector method to explore the impact of driving factors on the ESP. The Geodetector method, proposed by Wang et al. (2016) [71], is a statistical method used to identify potential driving forces in space. It can examine the spatial stratified heterogeneity in a single variable and detect possible causal relationships between two variables by examining the coupling of their spatial distributions [72]. The model quantifies the explanatory power of different driving factors on geographical data, with results represented by the q statistic. The formula is as follows:

$$q=1-{\displaystyle \frac{1}{N{\sigma }^2}}{\sum }_{h=1}^M{N}_h{\sigma }_h^2$$

(4)

where M represents the variable hierarchy; N is the total sample count; N_h is the sample count for variable h; ${\sigma }^2$ is the variance for the entire region; and ${\sigma }_h^2$ denotes the variance of dispersion for variable h within a sub-region.

Ecological factors such as topography, geomorphology, and climate, as well as socio-economic factors like economic development, influence the spatial distribution of an ESP. We listed the independent variable factors, as shown in

Table 4. Social-ecological drivers on ecological security patterns.

Category	Indicator	Abbreviation
Ecological factors	Land use type	land
	NDVI	ndvi
	Altitude	dem
	Slope	slope
	Soil erosion degree	soil
	Temperature	tem
	Precipitation	pre
	Distance from water source	dws
Socio-economic factors	Distance from residential area	dra
	Distance from main transportation roads	dtr
	Gross domestic product	gdp
	Population density	pop

, and executed the Geodetector analysis in RStudio using the “GD” package [73,74] (https://www.r-project.org (accessed on 2 September 2023)).

3. Results

3.1. Analysis of Spatiotemporal Changes in Ecological Sources

In 2000, we identified a total of 216 ecological source areas, which accounted for 22.24% of the study area. By 2005, the number of ecological sources increased to 222, raising their proportion of the study area to 23.05%. In 2010, the number of ecological sources decreased to 206, yet their share of the study area rose to 24.18%. During 2015 and 2020, the number of ecological sources was on the rise, reaching 211 and 214, respectively, while their percentage of the study area decreased, at 23.78% and 23.09% (Figure 3). From 2000 to 2020, the proportion of ecological sources within the study area first increased and then decreased, peaking in 2010. During this period, the total number of ecological sources saw a marginal decrease of two, but their overall proportion of the study area increased by 0.85%.

Under the three different scenarios for 2030, the ecological sources within the study area showed varying outcomes. Under the EP scenario, the number of ecological sources was the lowest at 199, yet they occupied the highest proportion of the study area at 24.20%. Under the ED scenario, the ecological sources were the most numerous at 204 but occupied the lowest proportion of the study area at 23.44%. Under the ND scenario, the number and proportion of ecological sources fell between those of the ED and EP scenarios.

From 2000 to 2005, slight increases in ecological sources were observed in both the northwest and southeast. In 2010, a slight reduction occurred in the southeastern ecological sources. Between 2015 and 2020, the spatial distributions of changes in ecological sources were relatively small. By 2030, under all three scenarios, the spatial distribution of ecological sources was basically the same as in 2020, maintaining stability (Figure 4). In summary, from 2000 to 2030, the BTH region’s ecological sources were densely distributed in the northern and western mountainous areas, with sparser distributions in the Zhangjiakou Bashang area in the northwest and the southern plains. The landscape connectivity in the north and west primarily relied on mountainous areas to establish links.

3.2. Analysis of Spatiotemporal Changes in Ecological Resistance Surfaces

From 2000 to 2020, the range of changes in the ecological resistance value of BTH was 1.3333–4.5984. Comparing the maximum values of ecological resistance surfaces, there was an increasing trend from 2000 to 2015 (from 4.4954 to 4.5984), followed by a decrease from 2015 to 2020 (from 4.5984 to 4.5149). However, overall, it increased from 4.4954 in 2000 to 4.5149 in 2020. During the same period, the minimum values of ecological resistance surface decreased from 1.5570 in 2000 to 1.4135 in 2020. In 2030, under the EP scenario, both the highest and lowest values of ecological resistance surfaces were between those of the ND and ED scenarios. Compared with 2020, both the highest and lowest values of ecological resistance surfaces under the EP scenario showed a decline. The ND scenario saw a stable highest value with a decreased minimum value; the ED scenario experienced an increase in the highest value while the minimum value remained stable (Figure 5).

From 2000 to 2030, the distribution of ecological resistance values in BTH showed notable trends. Initially, high resistance values were predominantly found in urban built-up areas and their peripheries, including the Bashang area in the northwest, while lower values were clustered and distributed in the western and northern mountains. Over the years, these high-value zones expanded, especially around the main urban centers of Beijing city and Tianjin city, reflecting a broadening influence of urbanization (Figure 5).

By 2015, the expansion of high-value ecological resistance areas became more pronounced, not only encompassing the urban cores of Shijiazhuang city, Xingtai city, and Handan city but also extending to towns with extensive farmland in the southeast. This trend persisted into 2020, with significant increases in high-value resistance areas particularly noted around southern urban centers like Baoding city, further emphasizing the impact of urban expansion and agricultural land use on ecological resistance (Figure 5).

By 2030, projections under all scenarios suggest that the spatial pattern of ecological resistance—characterized by high values in urban and extensive farmlands in the south and southeast and lower values in less developed, mountainous regions—will largely remain stable compared with 2020 (Figure 5).

3.3. Analysis of Spatiotemporal Changes in Ecological Corridors

In 2000, there were 603 ecological corridors and 48 potential ecological corridors, with a total length of 8042.60 km and 1817.82 km, respectively (Figure 6a,c). By 2005, the number of ecological corridors increased to 633, with a total length of 8551.69 km, while the number of potential ecological corridors rose by one, although their total length decreased to 1648.42 km. In 2010, the number of ecological corridors was reduced to 582, and the total length was reduced to 6934.02 km, but there was an increase in potential ecological corridors to 53, with a total length of 2405.26 km. By 2015, both the number and total length of ecological corridors increased to 596 and 7396.97 km, respectively, and potential ecological corridors also saw an increase to 62, totaling 2467.88 km in length. In 2020, the increasing numbers and total lengths of ecological corridors continued to 616 and 7774.99 km, respectively, while the numbers and total lengths of potential ecological corridors decreased to 48 and 2388.39 km.

Overall, from 2000 to 2020, the number of ecological corridors in BTH showed an increasing trend, but the total length decreased, while the number of potential ecological corridors remained stable with an increase in total length. By 2030, under the EP scenario, there were more ecological corridors (593) compared with the ND (585) and ED (589) scenarios. However, the total length of ecological corridors under the EP scenario (7324.03 km) was shorter than under the ND (7457.95 km) and ED (7508.47 km) scenarios. The number and total length of potential ecological corridors were highest under the ED scenario (50 corridors, 2400.38 km) compared with the EP (45 corridors, 2090.61 km) and ND (45 corridors, 2320.20 km) scenarios (Figure 6b).

From 2000 to 2030, the spatial distribution of ecological corridors in BTH changed relatively little, with a predominance of north–south- and northwest–southeast-oriented corridors. In the northern and western mountainous areas, corridors were denser and shorter, while in the southern plains, corridors were less dense and longer. Potential ecological corridors were mainly found in the northern and western mountains in 2000 and 2005, but by 2010, 2015, and 2020, and under all three scenarios for 2030, they were also present in the southern plains in addition to the northern and western mountains. Between 2000 and 2030, areas with frequent changes in ecological corridors were primarily in the northern and western mountains, with the southern plains being relatively stable (Figure 7).

3.4. Analysis of Spatiotemporal Changes in Ecological Pinch Point Areas

The proportion of ecological pinch point areas within the study region showed a declining trend. In 2000, pinch point areas comprised 34.91% of the study region, with first-, second-, and third-level pinch point areas covering 4232.34 km², 24,540.12 km², and 46,609.56 km², respectively. By 2005, the proportion of pinch point areas decreased to 34.63%, with an increase of 1735.20 km² in the combined area of first- and second-level pinch points compared with 2000. In 2010, the proportion further decreased to 29.52%, with a reduction in the total area of first- and second-level pinch points from 2005, although the area of first-level pinch points increased by 1289.16 km². By 2015, the proportion dropped to 28.83%, with a decrease of 352.71 km² in the total area of first- and second-level pinch points from 2010 (Figure 8a,b). In 2020, the proportion fell to 27.41%, but the areas of first- and second-level pinch points increased by 1525.50 km² and 40.23 km², respectively, from 2015.

Overall, from 2000 to 2020, BTH saw a continuous decrease in the area of ecological pinch point areas, indicating a gradual reduction in the transition sites for the flow of matter and energy among different ecological source areas, with human activities expanding their interference in ecological lands. By 2030, under the EP scenario, the proportion of pinch point areas (26.53%) was higher than under the ND (25.87%) and ED (26.01%) scenarios. The total area of first- and second-level pinch points under the EP scenario (32,073.66 km²) exceeded those in 2020 (31,068.00 km²) and the ND (31,033.62 km²) and ED (31,659.84 km²) scenarios, suggesting a positive impact on regional ecological environmental improvement, with a higher number of critical areas for matter and energy flow (Figure 8a,b).

From 2000 to 2020, the spatial distribution of ecological pinch point areas remained relatively stable, with dense distributions in the northwest, northeast, and southwest of the study area, and sparse distributions in the southeast. First- and second-level pinch points were primarily located in the northern and western parts of Zhangjiakou city, western and eastern parts of Chengde city, Qinhuangdao city, Tangshan city, the coastal areas of the northern and eastern Tianjin city, and the western areas adjacent to the Taihang Mountains in the southwestern cities of Hebei Province (Baoding city, Shijiazhuang city, Xingtai city, Handan city). Third-level pinch points were mainly near ecological corridors, mostly in the central and southern plains of the study area, with a sparser distribution. By 2030, under all three scenarios, the spatial distribution of first-, second-, and third-level ecological pinch point areas was similar to that in 2020 (Figure 9).

3.5. Analysis of Spatiotemporal Changes in Ecological Barrier Areas

In 2000, the first- to third-level barrier areas were 3919.32 km², 9512.37 km², and 109,185.39 km², respectively; these areas accounted for 56.78% of the research area. By 2005, the proportion of barrier areas slightly decreased to 56.57%, with a reduction in the area of first- and second-level barriers to 2716.29 km² and 6467.31 km², respectively, while the area of third-level barriers increased to 112,974.66 km². In 2010, the proportion of barrier areas fell to 49.78%, with an increase in the area of first- and second-level barriers to 3573.36 km² and 10,390.23 km², respectively, but a decrease in third-level barriers to 93,536.91 km². By 2015, the proportion of barrier areas rose to 52.41%, with increases in the area of first- and third-level barriers to 3697.56 km² and 100,256.76 km², respectively, and a decrease in second-level barriers to 9228.69 km². In 2020, the proportion decreased to 51.12%, with an increase in the area of first- and second-level barriers to 4333.68 km² and 11,202.66 km², respectively, while the area of third-level barriers reduced to 94,871.34 km² (Figure 10a,b).

From 2000 to 2020, the overall trend for ecological barrier areas in the study region was a downward fluctuation, with an increase in the area of first- and second-level barriers and a decrease in third-level barriers. Although the overall resistance to the flow of matter and energy decreased, indicating some improvement in ecological quality, the degree of resistance to these flows actually increased. By 2030, under the EP scenario, the proportion of barrier areas (48.00%) was lower than in the ND (49.63%) and ED (49.40%) scenarios. The area of first- to third-level barrier areas under the EP scenario was lower than in 2020, as well as in the ND and ED scenarios, indicating a reduction in resistance to the flow of matter and energy and a gradual improvement in ecological environmental quality.

From 2000 to 2020, the spatial distribution of ecological barrier areas in the study region generally showed a decreasing trend (Figure 11). First-level barrier areas were mainly concentrated in the northern part of Zhangjiakou city, with scattered distributions in Tianjin city and its eastern coastal areas. Additionally, first-level barrier areas appeared in Beijing city, Shijiazhuang city, Xingtai city, and Handan city during 2010, 2015, and 2020. Second-level barrier areas were clustered around Zhangjiakou city and the border with Chengde city, with scattered distributions in Beijing city, Tianjin city, Tangshan city, Baoding city, Shijiazhuang city, Xingtai city, and Handan city. Third-level barrier areas had a wide coverage in the study area, mainly in the eastern and southern plains. Overall, from 2000 to 2020, barrier areas were more densely distributed in the northern part of the study area, while the southern region had a more dispersed distribution. By 2030, under all three scenarios, the spatial distribution of first- to third-level barrier areas was similar to that in 2020.

3.6. Coefficient of Variation of ESP Components

The coefficient of variation of ESP components in the study area from 2000 to 2020 was statistically analyzed (Figure 12). The research results show that in the past 20 years, the coefficient of variation of ESP components was below 20%. Among them, the total pinch point area, first-level barrier area, potential economic corridors, first-level pinch point area, and second-level barrier area have lower volatility (10% < CV ≤ 20%). The second-level pinch point area, ecological resistance surfaces, total economic barrier areas, and third-level barrier areas have the lowest volatility (CV ≤ 10%).

3.7. Analysis of Spatiotemporal Changes in ESP

In 2000, ecological conservation zones, improvement zones, restoration zones, and critical zones occupied 22.24%, 25.75%, 6.22%, and 13.32% of the study area, respectively. There were 148 protection corridors totaling 6869.63 km in length and 35 restoration corridors spanning 1733.83 km. By 2005, the proportions of conservation and improvement zones increased to 23.05% and 29.28%, respectively, while the restoration zone proportion decreased to 4.25%, and the critical zone proportion increased to 14.13%. The number of protection corridors rose to 162 and restoration corridors rose to 37. In 2010, the conservation zone proportion increased to 24.18%, while the improvement zone proportion decreased to 24.46%, the restoration zone proportion rose to 6.47%, and the critical zone proportion fell to 13.82%. The number of protection corridors decreased to 126, but restoration corridors increased to 42. By 2015, the conservation zone proportion decreased to 23.78%, while the improvement zone proportion increased to 25.21%, the restoration zone proportion fell to 5.99%, and the critical zone proportion to 13.66%. The number of protection corridors rose to 143, and restoration corridors to 52. In 2020, the conservation zone proportion further decreased to 23.09%, but the improvement zone proportion increased to 26.15%, the restoration zone to 7.19%, and the critical zone to 14.39%. The number of protection corridors rose to 157, while restoration corridors decreased to 38.

Overall, from 2000 to 2020, there was an increase in ecological conservation zones, improvement zones, restoration zones, critical zones, protection corridors, and restoration corridors. By 2030, under the EP scenario, the proportions of conservation and critical zones (24.20% and 14.85%, respectively) were higher than in 2020 (23.09% and 14.39%) and the ND (23.65% and 14.37%) and ED (23.44% and 14.66%) scenarios. However, the improvement and restoration zone proportions under the EP scenario (23.69% and 6.55%, respectively) were lower than in 2020 (26.15% and 7.19%) and the ND (25.27% and 6.60%) and ED (25.10% and 7.16%) scenarios.

The spatial distribution of the ESP largely remained consistent from 2000 to 2020, with a higher density in the northern part of the study area. Conservation zones were primarily concentrated in the northern Yan Mountains and the western Taihang Mountains, while improvement zones were distributed in bands across the central and southern plains. Restoration zones were mainly located in Zhangjiakou city, western Chengde city, northern Beijing city, central and eastern Tianjin city, and parts of Tangshan city and Qinhuangdao city, as well as the southwestern cities of Baoding, Shijiazhuang, Xingtai, and Handan where they meet the Taihang Mountains. Critical zones were primarily found in Zhangjiakou city, eastern Chengde city, Tangshan city, Tianjin city, western Baoding city, Shijiazhuang city, Xingtai city, and Handan city. By 2030, under all three scenarios, the ESP was similar to that in 2020, with the spatial distribution of restoration zones under the EP scenario being smaller than in 2020 and the ND and ED scenarios. In summary, the density of ESP in the research area was higher in the north than in the south, with Zhangjiakou city, northern Beijing city, central and eastern Tianjin city, parts of Tangshan city and Qinhuangdao city, and the southwestern cities meeting the Taihang Mountains identified as key areas for ESP optimization (Figure 13).

3.8. Socio-Ecological Drivers of the ESP

In 2000, the ESP was primarily driven by slope, elevation, precipitation, and land use types. After 2005, in addition to slope, elevation, precipitation, and land use types, the driving forces of temperature and soil became significantly more influential. Moreover, over the past two decades, the driving force of the population generally showed an increasing trend, while the influence of distance to roads decreased.

For the ecological impact assessment, in 2000, the ESP was mainly driven by elevation, slope, distance from main transportation roads, and land. After 2005, alongside elevation, slope, distance from main transportation roads, and land use types, the driving forces of temperature and population fluctuated but increased, and by 2020, temperature and population replaced slope and distance from main transportation roads as the second and third most significant drivers, respectively.

The ecological restoration areas were consistently driven by precipitation, NDVI, temperature, distance to water sources, and elevation from 2000 to 2020. However, over time, the driving forces of precipitation and NDVI slightly decreased. The ecological key areas were constantly influenced by temperature, elevation, precipitation, NDVI, and slope from 2000 to 2020 (Figure 14).

The dominant driving factors of BTH’s ESP change correspondingly over time. The main drivers for some elements of the ESP, such as ecological restoration areas and ecological key areas, remained relatively stable across different years, while for other elements, like restoration corridor and ecological improvement areas, the main drivers changed over time. Overall, the impact on ecological corridors and ecological improvement areas was primarily due to a combination of socio-ecological driving factors (such as elevation, slope, distance from main transportation roads, and population), whereas the influence on ecological restoration areas and ecological key areas was mainly from ecological environmental factors (such as elevation, temperature, NDVI, and precipitation).

4. Discussion

4.1. Identifying the ESP

The foundation for constructing an ESP is the scientific identification of precise ecological source areas within a region. Existing studies have assessed ecosystem services [75] and used the results of comprehensive evaluations of ecological significance to identify ecological source areas [76]. Utilizing multi-source data can effectively highlight high-value ecological areas, but relying solely on these areas may overlook the connectivity of landscape patches that define ecological source areas. This research considers the importance of landscape connectivity through the MSPA model, coupling differences in ESVs and regional variations in land use types for a relatively comprehensive evaluation and determination. This approach addresses the limitations of previous methods that focused on connectivity or ecological value in isolation. Furthermore, the MSPA model follows the basic principles of graph theory by conducting topological analyses on the characteristics of linear elements or patches and their spatial accessibility, resulting in a set of morphological analysis variables. Compared with other methods, MSPA offers faster computation, more systematic selection of indicators, and clearer topological relationships, providing valuable assistance in prioritizing ecological source areas and corridors.

Previous studies based on the MCR model [77] identified and confirmed ecological corridors by determining the least-cost paths among ecological source areas and eliminating redundant corridors to preserve unique minimum-cost pathways. This suggests that the number of connecting corridors among different ecological source areas correlates positively with the number of source areas. However, in this study, although more ecological source areas were selected in 2000 (216) than in 2015 (214), fewer ecological corridors were identified in 2000 (603) compared with 2015 (616). In 2010, there was one fewer ecological source area (296) than in 2015 (297), yet there were 14 fewer corridors. Therefore, the flow of matter and energy does not necessarily follow the proximity principle, as adjacent ecological source areas might not be connected by corridors. Circuit theory appears to model and express the rationality of ecological corridors more effectively than the MCR model.

By leveraging circuit theory and considering the stochastic nature of species migration, numerous potential ecological corridors were identified between 2000 and 2020. These potential corridors offer alternative solutions under restricted terrain and conservation construction conditions, significantly enhancing the efficiency of protected resource utilization and holding substantial ecological significance. By analyzing current density, ecological pinch points and barrier areas were identified, enhancing the comprehensiveness of the ESP in BTH. This represents a refinement over traditional construction methods that rely on water bodies or road locations and empirical node determination, enriching the practical methods for quantitatively identifying key landscape pattern areas and making conservation construction levels more scientifically sound.

4.2. Drivers of the ESP

In mountainous terrains, influenced by topographical differences, ecological source areas are predominantly located within forestlands and grasslands, with dense ecological corridors contributing to a high degree of network closure [77,78,79]. Conversely, in plain landscapes, ecological source areas mainly consist of green spaces such as forest parks or water bodies such as lakes, and ecological corridors mainly form closed loops around these areas. However, ecological corridors are sparse in urban centers and their surroundings, where urban development and human activities significantly impact the environment [80].

The study area is primarily composed of two geomorphological units including the Yan-Tai mountainous region and the North China Plain. In the northern and western parts, the Yan Mountains and Taihang Mountains host ecological source areas mainly comprising forestlands and grasslands, especially in the Bashang area, which is predominantly grassland. These areas feature dense ecological corridors forming a network with high closure (Figure 7). Because of the higher elevation (averaging over 2000 m), significant soil erosion, and severe grassland fragmentation in the northwest and Bashang area, coupled with dense human settlements, the ecological resistance values in these regions are high (average above 4) and negatively affect the ESP. Consequently, the spatial distribution of ecological restoration areas is dense and extensive in these regions.

Ecological improvement areas are influenced not only by environmental factors (such as elevation, slope, and land use type) but also by socio-economic factors (such as distance from main transportation roads, and population). In the southern plains, ecological source areas mainly consist of sizable lakes or water bodies, with dispersed source areas and sparse ecological corridors that generally follow the direction of rivers in the region. Because of dense human settlements, high soil erosion, and a dense river network, the ecological resistance values are high, with high-value areas spatially resembling the distribution of towns. Furthermore, there were frequent land use changes, with urbanization leading to a 5.2% decrease in farmland and a 2.3% increase in urban land from 2001 to 2020 [81]. The population density in nine cities of the southern plains (Beijing, Tianjin, Tangshan, Langfang, Baoding, Cangzhou, Hengshui, Shijiazhuang, Xingtai, Handan) increased from 619 people·km⁻² in 2000 to 781 people·km⁻² in 2020. This analysis also found that the population’s influence on ecological improvement areas rose from fifth place in 2000 to third in 2020, indicating increased ecological and environmental pressures in the region, with scattered ecological restoration areas and larger band-shaped ecological recovery areas.

This study revealed that ecological conservation areas are more numerous and larger in the west and north, while less frequent and smaller in the southern plains. Ecological recovery areas are mainly located in the vicinity of towns and the northwest of the research area in the Bashang area. Protection corridors are mainly concentrated in the western and northern mountainous areas of the research area but are sparse in the plains. Ecological key areas are primarily situated around protection corridors and conservation areas. Therefore, distinguishing different geomorphological units and identifying socio-ecological drivers affecting the ESP, as well as protecting and increasing the area and number of ecological source areas to improve and restore the regional environment, are crucial for restoring the overall ESP and landscape connectivity of BTH.

Furthermore, changes in land use and expansion of urban built-up areas cannot be ignored in terms of their impact on the ESP. The establishment of Xiong’an New Area (decided on 1 April 2017) has profound historical significance in adjusting and optimizing the urban spatial structure of BTH [82]. The current landscape substrate of Xiong’an New Area, which is mainly dry land, will inevitably develop towards the expected massive urban development and ecological construction direction as the new area develops [64], which will certainly have an impact on the ESP in BTH. In the prediction results of future scenarios, one ecological corridor passing through the area disappeared compared with 2015 and 2020 (Figure 7), which is more obvious evidence. Therefore, considering the concept of “ecological benchmark” construction, Xiong’an New Area needs to focus on the construction of ecological source areas in the future. Based on farmland as the substrate, new ecological land will be added, and ecological greening construction along key construction corridors will be increased to enhance the connectivity among source areas, improve habitat quality, and ultimately form an ecological livable urban development pattern with ecological source areas as the green heart and blue-green ecological corridors connected. Ultimately, it is expected to achieve coordination with BTH’s national spatial planning and ecological protection in the future urban development of Xiong’an New Area.

4.3. Prediction in 2030

It is found that under the EP scenario, the proportion of ecological conservation areas and critical areas is the highest. This indicates that the space available for species habitation is greater than in the other three periods, suggesting that the space for species to live, act, and migrate will expand under this scenario. Additionally, the EP scenario shows the lowest proportions of ecological improvement and restoration areas, suggesting higher landscape connectivity than in the other periods, thereby further reducing the spaces that hinder species survival and migration. Hence, under policies prioritizing ecological protection and aiming to maximize ecological benefits, effective protection of ecological resources and rational control of urban development can optimize the future ESP of the BTH region.

4.4. ESP in the Same or Different Regions

According to previous ESP studies on BTH, Zhang et al. (2017) [65] and Wang et al. (2022) [83] confirmed that the main ecological sources of BTH urban agglomeration are distributed in the Yanshan–Taihang mountain area, consistent with the key areas of ecosystem services, which is similar to the distribution of ecological sources in the results of this study. Hu et al. (2016) [84] used the minimal-cost path model to extract a total of 579 ecological corridors in BTH urban agglomerations, which was basically close to the 582–633 ecological corridors in different periods in this study. Zhang et al. (2022) [85] found that the ecological corridors in the BTH urban agglomeration were intertwined with ecological source areas and distributed across the Yanshan and Taihang mountain areas, while there were no ecological source areas or corridors in the Jizhongnan Plain. The 2030 forecast results show that the construction land in the area will increase significantly and expand in a ring shape, threatening the healthy development of surrounding ecological source areas and corridors. This is different from the distribution of ecological sources and ecological corridors in the Jizhongnan Plain in this study, mainly because of the difference in the selection of the areas and numbers of ecological sources in the ESP construction process. Compared with the above studies, this study not only identified ecological sources and constructed ecological corridors, but also identified ecological pinch point areas and ecological barrier areas so that the ecologically vulnerable areas and areas in urgent need of protection were more accurately identified.

In recent years, research on ESP in urban agglomerations has been a hot topic. Guo et al.’s (2019) [86] study on the ESP of the Harbin–Changchun urban agglomeration showed that some ecological source areas were occupied by unreasonable construction and unused land had ecological development potential. Because of urban construction and location reasons, the ecological corridor formed a central circular network and a peripheral tree-like structure. Jiang et al. (2021) [39] constructed the ESP of the Guangdong–Hong Kong–Macao Greater Bay Area urban agglomeration from the perspective of regional ecological protection cooperation, and their study showed that the ecological source area was mainly distributed in the north, the ecological corridor was mainly distributed in the periphery of the study area, and the entire ESP in the study area showed a circular space. Zhang et al. (2022) [87] conducted ESP research on the Yangtze River Delta urban agglomeration, showing that the urban development speed was higher in the north than in the south, and the ecological security level was lower in the north than in the south. Based on the spatiotemporal changes in the ecosystem, an ecosystem optimization model was proposed to determine the ecological functional zones. According to the nature of each functional area, urban construction and management plans were formulated to promote sustainable urban development. Jia et al. (2023) [38] conducted a study on the ESP construction of urban agglomerations in the middle reaches of the Yangtze River and confirmed that developed cities had a high overall imbalance of ecosystems and can adjust their ecosystem supply and demand by appropriately expanding the area of ecological source areas. At the same time, priority protected areas should be appropriately allocated in areas with low supply and high demand to reduce the risk of ecosystem imbalance. Wei et al. (2023) [29] studied the ESP construction of arid urban agglomerations on the north slope of the Tianshan Mountains and found that ecological strategic points were located in the urban fringe and areas with frequent human activities. They recommended that these “points” must be protected in the future construction and planning process. Human activities, transport networks, and fragile ecosystems hinder the flow of matter and energy, and in future planning, the policy of returning farmland to grassland should be continuously promoted to improve regional ecological connectivity. Obviously, the ESPs of different urban agglomerations have their own characteristics, which are closely related to the pattern of an urban agglomeration itself and the number, area, and location of the ecological source areas in the region. It is worth noting that a consensus has been reached on the idea of providing targeted ecological planning for regional development based on the characteristics of ESP in urban agglomerations. This can be accomplished mainly by connecting ecological corridors to enhance the circulation of materials and energy, implementing ecological protection to improve ecologically fragile areas, and promoting the sustainable development of urban agglomerations in terms of ecology, society, and economy. Based on this consensus, this study explores the evolution of the ESP in BTH and its structure in different future scenarios, identifies the concentration areas and ecologically fragile areas of the source area, predicts the future changes in these areas, and recommends targeted management measures for these areas with different ecological conditions, ultimately achieving the goal of ESP optimization and stability.

4.5. Recommendations

Ecological source areas within the study region are mainly concentrated in the western and northern mountainous areas, with fewer and smaller source areas in the southern plains. Therefore, forest management in the western and northern mountainous areas should be further improved to enhance forest quality and stabilize and improve the quality of these ecological source areas. In the southern plains, efforts should increase green spaces and water bodies, along with strengthening green urban development to expand urban green belt spaces. Relevant management departments should establish protected areas to reduce human disturbances, particularly promoting the flow of matter and energy among mid-to-long-distance ecological corridors in the southern plains, maintaining stability and connectivity among ecological source areas. In the Bashang area of the northwestern part of the study region, grassland management should be intensified to promote and restore grassland resources, enhancing the stability of ecological source areas in that region. Furthermore, it is recommended that future studies should build upon the scale of this research to explore and develop further multi-scale, hierarchical constructions of ESPs to address the multi-layered ecological security issues of different zones.

4.6. Limitations and Prospects

Increasing population pressure, limited spatial resources, fragile ecosystems, and large-scale urban construction activities have constrained the sustainable development of ecosystems and directly threatened the ecological security of national land, which showed significant resistance effects in this study (Figure 5). The structural differences in different ESP elements are the feedback of the quality of the regional ecological security core [88]. Unlike previous research, which paid attention to the construction of ESPs, this study explored the structure and constituent elements of ESPs, considering the spatiotemporal heterogeneity in basic ecosystem units. This not only focuses on the spatial distribution and overall morphological characteristics of ecosystems and their interrelationships [89] but also emphasizes the effective guarantee of ecosystem stability and health based on spatial planning [90,91]. Furthermore, this study combines multiple models to analyze the spatiotemporal dynamics of the constituent elements of the ESP in the past, present, and future scenarios of the study area and constructs a system of the ESP in the study area. This has a certain scientific indicative effect on the future national spatial planning of the study area and also has good international applicability.

This study introduces the ESV as an important indicator in identifying ecological source areas, but the accounting of ESVs involves many subcategory indicators, which inevitably leads to more subjectivity in the selection process. In order to objectively reflect the contribution of ecosystem services in the process of selecting ecological source areas, changes in different regional ecosystems can be determined spatially through ecosystem service flows, ecosystem service clusters, and ecosystem service supply and demand [92,93]. In addition, in the context of a larger research area, there is still uncertainty in the buffer width of the ecological corridor and the search radius of the ecological pinch points, and it is necessary to determine their reasonable range through on-site investigation. In the future, with a gradual improvement in the ESP, the ESP structural system that integrates multiple model technologies (such as CA, neural networks, and system dynamics) will gradually replace traditional ecological resource indicators based on quantity structure [94], thereby promoting the evaluation and optimization of ecological security construction [95].

5. Conclusions

This research utilized the GMOP-PLUS model to forecast land use alterations by 2030 across three distinct scenarios, employing the MSPA model in conjunction with variations in ESVs and diverse land use categories to pinpoint ecological source areas accurately. By integrating circuit theory, it further delineated ecological corridors, pinch points, and barrier zones. Subsequently, this study meticulously crafted the ESP for BTH spanning from 2000 to 2030, providing a comprehensive examination of its evolution over time and space. Key findings include the following:

(1)

The proportion of ecological source areas in BTH increased from 22.24% in 2000 to 23.09% in 2020. By 2030, under the EP scenario, the proportion of ecological source areas was predicted to reach the highest at 24.20%, with dense distributions in the western and northern mountainous areas, and sparser distributions in the northwestern Bashang area and the southern plains. The spatial extent of high-value ecological resistance areas expanded, mainly concentrating in the southern plains, with the northwestern Bashang area consistently being a high-resistance region.

(2)

From 2000 to 2020, the number of ecological corridors increased from 603 to 616. By 2030, the number of corridors under the EP scenario (593) was predicted to exceed those in the ND (585) and ED (589) scenarios. From 2000 to 2030, corridors in the northern and western mountains were denser, shorter, and more variable, while those in the southern plains were less dense, longer, and relatively stable.

(3)

The proportion of ecological pinch point areas decreased from 34.91% in 2000 to 27.41% in 2020. By 2030, under the EP scenario, the proportion of pinch point areas (26.53%) was predicted to be higher than in the ND (25.87%) and ED (26.01%) scenarios. From 2000 to 2030, pinch point areas were densely distributed in the northwest, northeast, and southwest, with sparse distributions in the southeast.

(4)

The proportion of ecological barrier areas dropped from 56.78% in 2000 to 51.12% in 2020. By 2030, the EP scenario was predicted to have a lower proportion of barrier areas (48.00%) compared with the ND (49.63%) and ED (49.40%) scenarios. From 2000 to 2030, barrier areas were more densely distributed in the northern part of the region, with more dispersed distributions in the south.

(5)

From 2000 to 2020, habitat areas for species in BTH increased, while landscape connectivity decreased. By 2030, the EP scenario was predicted to have increased habitat areas and improved landscape connectivity compared with 2020 and the ND and ED scenarios, leading to some optimization of the ESP. From 2000 to 2030, the density of the ESP was greater in the north than in the south, with conservation areas mainly concentrated in the northern and western mountains, improvement areas distributed in bands across the southern plains, and restoration and key areas mainly located in Zhangjiakou city, western Chengde city, and scattered throughout the southern plains.

(6)

The dominant driving factors of the ESP varied over time. The impact on ecological corridors and improvement areas mainly came from a combination of socio-ecological drivers (such as elevation, slope, distance from main transportation roads, and population), while the influence on restoration areas and key areas predominantly came from ecological environmental factors (such as elevation, temperature, NDVI, precipitation). Distinguishing different geomorphological units to improve and restore the regional environment, while considering socio-ecological drivers, is crucial for restoring the overall ESP and landscape connectivity of BTH.

This study used relevant models to identify ecological sources, corridors, pinch points, and obstacle areas and systematically constructed the ESP of BTH from 2000 to 2030. The social and ecological driving factors affecting the ESP in the region were explored, which can provide a scientific basis for the optimization of ESPs and the formulation of ecological protection policies in countries and other parts of the world.