Identifying Key Locations of the Ecological-Barrier System to Support Conservation Planning: A Study of the Sanjiangyuan National Park

Abstract

The establishment of the Sanjiangyuan National Park (SNP) system indicates a higher requirement for refining management practices, especially for protecting the ecological barrier system (EBS) that supports national ecological security and biodiversity. However, it is still understudied how planners can identify the key areas for maintaining EBS in addition to functional zoning. This study took the SNP as an example and built a comprehensive analytical framework, including fragmentation analysis, landscape morphology analysis, and connectivity analysis based on graph theory. The study found that the ecological patches of the Lancang River Source sub-park are relatively complete, while those of the Yangtze River Source sub-park and the Yellow River Source sub-park are more fragmented according to different indicators. The study then identified key nodes and edges of sources for maintaining the EBS. These areas are located mostly near core zones of habitat patches. Furthermore, the study analyzed key patches for maintaining landscape connectivity using two indicators DN (degree of nodes) and dIIC (the delta integral index of connectivity), which respectively quantify the number of neighbors of a habitat and its impact on the whole connecting EBS. Last, the study identified areas with dense landscape corridors in the EBS for suggesting key protection areas.

1. Background

In recent years, the protection and management of large-scale natural protected areas have received increased attention for sustainable development and climate change mitigation [1,2]. The ecological barrier system (EBS) plays a role in maintaining regional ecological security for natural protected areas [3,4]. EBS represents natural or human-planned ecological belts that serve as barriers to help humans prevent negative ecological effects, such as sandstorms, the invasion of agriculture, and the loss of biodiversity [4]. These functions bring benefits to many ecological processes, such as animal migration, plant seed spreading, and regional species diversity, and thus, play an important role in the conservation of natural ecosystems on large scales [4]. Depending on specific local context, EBS often consists of connected natural reserves, types of vegetation, rivers and wetlands, and green corridors. They can help connect individual ecological source areas into a system for maintaining ecosystem integrity, minimizing landscape fragmentation, and reducing disturbances [3,5]. However, the sustainable management of EBS in large natural protected areas frequently faces several practical challenges, such as limited management resources, complex dynamics of ecosystems, conflicts between conservation and development, and governance over administrative boundaries [6,7].

Globally, different countries have developed their own systems for managing large natural protected areas, which facilitate maintaining the integrity of ecosystems and environments [8,9]. For example, in Europe, landscape corridors play a vital role in maintaining ecological function and services in various environmental settings, including the natural protected areas [10]. They help constitute the cultural landscape and locality of a place, highlighting the spiritual value of sustainable management. The U.S. National Park Service has established specific criteria for setting up national parks, highlighting ecological conservation and public benefits [11]. The International Union for Conservation of Nature (IUCN) defines protected areas, including national parks and other wilderness areas, for their value in conserving biodiversity, tackling environmental challenges, and providing public goods [12,13]. National parks often serve as key components of ecological networks, facilitating connections at different spatial scales, including international, national, and regional levels. In recent years, China has announced its first batch of five national parks, aiming at the protection of landscape resources, biodiversity, and natural heritage. Although different countries have varying objectives, methods, and priorities in terms of management, those practices all highlight regional ecological security and call for attention to EBS [14,15]. Drawing on those management practices, planners can incorporate proven strategies, such as zoning and public engagement into the framework of planning EBS for National Park.

National parks are usually large and include areas that fall within different administrative units [16]. Those EBS within the parks may be adjacent to settlements and wilderness areas of various characteristics; therefore, they face various pressures and reflect different ecological conditions. The complex nature of national parks has raised challenges in managing EBS and preserving ecological integrity. However, the management of EBS within national parks has not been closely examined and remains challenging, particularly in terms of identifying key priority areas [2,17,18,19].

2. Literature Review

2.1. Overview of the Studies on Planning for EBS

Across contexts, EBS refers to types of ecological landscapes that prevent the degradation of natural systems and mitigate negative environmental incidents [20]. EBS itself often maintains stable landscape patterns and ecosystem functions; therefore, EBS can play a role in protecting adjacent lands, including both natural lands and built lands [21]. Regional EBS requires a set of connected ecological corridors or belts around large core areas of habitats, acting as either a physical or functional basis for the landscape. This type of system usually has the advantage of better environmental resilience, resistance to disturbances, and the ability to function as biological corridors [1,22,23].

Researchers have revealed that EBS can strengthen the structural integrity of habitats and enhance their ecosystem services within a certain spatial range, providing a basis for expansive conservation planning. From a practical perspective, effective planning of EBS plays a significant role for the following reasons. First, focusing on the prioritization of EBS is cost-effective in preserving large-scale nature reserves [24]. By leveraging natural ecological processes and self-sustaining characteristics, EBS can reduce reliance on expensive human intervention for nature conservation by prioritizing areas. Second, since EBS is multi-functional, developing EBS can target different environmental needs simultaneously, allowing for comprehensive planning and management [25,26]. Third, as an interconnected system, EBS can facilitate the movement of wildlife, promoting environmental adaptation to climate change and maintaining ecosystem health [23].

2.2. Studies on Identification of Key Areas of EBS for Conservation

There are numerous studies on national parks or large natural protected areas that aim to explore conservation priorities for sustainable development [14,19,27,28]. For example, one study in China developed a theoretical framework for monitoring ecological systems, and it proposed various decision-making strategies for preserving national parks by protecting habitat systems. The criteria should consider the distribution of key species, natural resource backgrounds, and management feasibility [29]. A recent trend in the region is that research teams have examined migration corridors in sensitive environments using least-cost path (LCP) models by defining the sources of habitats and resistance layers [27,30,31]. They have also applied connectivity and graph theory approaches to determine the significance of different nodes within ecological networks to identify areas critical for maintaining ecological barriers. Recent studies have used the spectral index of the annual mean normalized difference vegetation index (NDVI) at some time points or in a period of time to study the source areas of the ecological networks [32].

Existing studies have offered advances in understanding EBS, but there are still knowledge gaps in the practice within the contexts of large natural protected areas with various ecological conditions and strong ecological seasonality. In China, particularly, within the context of recent system reform for natural protected areas, management feasibility is an increasingly important topic in conservation planning. However, it is still not yet sufficiently studied in a spatially explicit way. Secondly, many existing studies focus on a single ecosystem type (for example, forests or wetlands) regarding EBS, while ecosystems with variability in vegetation types and seasonality are often ignored. Finally, many relevant studies, for different research purposes, usually designed methods for that objective but did not conduct a comprehensive analytical framework to evaluate EBS. To address these issues, our study took the SNP as an example and built a comprehensive analytical framework, including fragmentation analysis, landscape morphology analysis, and connectivity analysis based on graph theory. This paper applies the framework and provides planning suggestions.

Thus, the objective of this study is to take the SNP as a case study and build a comprehensive analytical framework for locating key areas for maintaining EBS. The analytical framework includes fragmentation analysis, landscape morphology analysis, and connectivity analysis based on graph theory.

3. Methodology

3.1. Study Area

The SNP is the region of the Tibetan Plateau where the Yangtze River, the Yellow River, and the Lancang River originate. Due to its vital ecosystem condition and the challenges in managing such a large area across several sparsely populated regions, the Chinese government reformed the administration and established SNP in 2021. It integrates all related technical, administrative, and legal services into one unified administrative system to enhance manageability. The park contains three sub-parks: the Yangtze River Source, the Yellow River Source, and the Lancang River Source (Figure 1, Figure A1, Figure A2 and Figure A3). These sub-parks are key to regional environmental conservation, watershed protection, water management, and biodiversity conservation. The SNP covers a vast area of Qinghai Province and spans over 10 townships with an average elevation of 4500 m above sea level. The park consists mainly of plateau mountains and gorges, with grasslands as the primary ecosystem, along with lakes, wetlands, and forests. One of the main purposes of establishing the SNP is to develop and maintain an EBS within this area, in response to biodiversity loss and landscape fragmentation due to human activities in this sensitive and vulnerable environment.

3.2. Identification of Key Conservation Areas of the EBS

3.2.1. Overview of the Analytical Framework and Data Processing

Based on the research objectives and a review of relevant literature, this study developed a spatially explicit framework with three subsequent modules (Figure 2). The three modules include spatial analysis of habitat fragmentation, shape-based landscape connectivity analysis, and graph-theory-based network analysis. Specifically, these three modules quantify different functionalities of the EBS in the SNP: (1) the degree of disturbance and fragmentation of the landscape, (2) the spatial composition of ecological source areas, and (3) the contribution of ecological source areas to the overall connectivity of ecological barrier systems.

The identification of the existing EBS is determined using NDVI for the following reasons. For high-altitude and sensitive environments, the vegetation may present significant seasonal variations. The annual mean NDVI is a stable measure of vegetation over seasons. Additionally, large areas may have different ecosystem types ranging from dense to sparse vegetation. NDVI can effectively capture the overall presence of various landscapes. For the purpose of maintaining biodiversity and resilience, it is important to comprehensively understand an EBS that often consists of mixed ecosystems. Based on existing studies, this study labeled areas with an annual mean NDVI above 0.5 as the source patches of the ecological network. All spatial analysis was conducted using open-source tools, including QGIS and R, especially the sf and terra packages. The research data are from the Chinese Annual 1 km NDVI Spatial Distribution Dataset [33] and the Basic Geographic Database of The Ministry of Natural Resources in China (https://www.webmap.cn/commres.do?method=result100W (accessed on 1 May 2021)).

3.2.2. Fragmentation Analysis of the EBS Source Patches

The concept of fragmentation refers to a process in which large and continuous natural lands turn into smaller, disconnected patches, often due to human activities and ecosystem degradation. The degree of fragmentation serves as a key indicator to measure the conditions of the EBS. This study applies several fragmentation indicators (

Table 1. Fragmentation indicators for analyzing the spatial composition and fragmentation of source patches of the EBS [34,35,36,37,38,39].

Landscape Metrics	Formula	Description and Interpretation
Patch area (PA)	-	PA is the total area of the source patches of the EBS.
Mean patch size (MPS)	$\mathrm{MPS}=\frac{{\displaystyle \sum _{j=1}^{\mathrm{n}}{\mathrm{a}}_{ij}}}{{\mathrm{n}}_i}\left(\frac{1}{\mathrm{10,100}}\right)$	MPS is the average area of the source patches of the EBS.
Number of patches (NP)	$\mathrm{NP}={\mathrm{n}}_i$	NP is the number of the source patches of the EBS.
Edge density (ED)	$\mathrm{ED}=\frac{{\displaystyle \sum _{k=1}^{m\prime }{\mathrm{e}}_{ik}}}{\mathrm{A}}\left(\mathrm{10,100}\right)$	ED equals the total length of all patch edges per unit area within the EBS source areas.
Total edge (TE)	$\mathrm{TE}={\displaystyle \sum _{k=1}^{m\prime }{\mathrm{e}}_{ik}}$	TE equals the sum of the lengths (m) of all edge segments within the EBS source areas. Together with the total patch area, TE can be used to compute edge density.
Total core area (km2) (TCA)	$\mathrm{TCA}={\displaystyle \sum _{j=1}^n{\mathrm{a}}_{ij}{ }^c\left(\frac{1}{\mathrm{10,100}}\right)}$	TCA equals the sum of the core areas of each patch of the EBS source areas.
Shape index (SI)	$\mathrm{SI}=\frac{0.25{\mathrm{p}}_{ij}}{\sqrt{{\mathrm{a}}_{ij}}}$	SI is an indicator describing the compactness or looseness of the shape of the source patch, ranging from 1 (extremely compact, e.g., round) to positive infinity (extremely loose).
Fractal dimension (FD)	$\mathrm{FD}=\frac{2In\left(0.25{\mathrm{p}}_{ij}\right)}{In{\mathrm{a}}_{ij}}$	FD is a metric that describes the complexity of the morphology and the fragmentation; the larger the value, the rougher and more fragmented the morphology.
Effective mesh size (EMS)	$\mathrm{EMS}=\frac{\mathrm{A}t}{\mathrm{S}}=\frac{1}{\mathrm{A}t}{\displaystyle \sum _{i=1}^n{\mathrm{A}}_i^2}$	EMS is a metric that describes landscape fragmentation and is calculated based on the probability that two randomly placed points in space fall within the same patch. EMS represents the average size of an area when the study area is divided into S segments, each of which is the same size At (total area of the study regions/S).

Note: a_ij (Area (m²) of patch ij), p_ij (Perimeter (m) of patch ij), e_ik (Total length (m) of edge in landscape between patch types (classes) i and k, and includes landscape boundary segments representing tree edges only involving patch type i), a_ij^c (Core area (m²) of patch ij based on specified buffer width (m)), n = n_i (Number of patches in the landscape of patch type (class) i), A(Total landscape area (m²), m’(Number of patch types (classes) present in the landscape, including the landscape border if present.

) to evaluate the comprehensive vulnerability of EBS.

3.2.3. Morphological Spatial Pattern Analysis

This section focuses on the morphological configuration of EBS using the morphological spatial pattern analysis (MSPA) tool. In many studies in landscape ecology, MSPA is used to analyze the geometry and connectivity of habitats and then spatially identify the morphological classification such as core, edge, bridge, and loop. The study applied the Guidos Toolbox developed by the European Commission. Addressing the fragmenting effects of roads and built environments on vegetated habitats, this study uses four-directional connectivity and sets 1000 m as the edge distance.

3.2.4. Connectivity Analysis from the Perspective of Graph Theory

In nature conservation, graph theory is commonly used to analyze the biological interactions and energy flow between different habitat patches. By treating each patch as a node and the link between two patches as an edge, graph theory provides a mathematical understanding of evaluating each node’s importance in the network system. Some important metrics applied in this study include the degree of nodes, the centrality of edges, the integral index of connectivity (IIC) for the system, and the effects of node removal (dIIC) for quantifying individual relative importance. The parameter for connectivity probability was set to 0.5.

The formulas for IIC and dIIC are as follows:

$$\mathrm{I}\mathrm{I}\mathrm{C}=\frac{\sum _{i=1}^n{\sum }_{j=1}^n({\mathrm{a}}_i\ast {\mathrm{a}}_j)/(1+{\mathrm{n}\mathrm{l}}_{ij})}{{\mathrm{A}}_L^2}$$

$$\mathrm{d}\mathrm{I}\mathrm{I}\mathrm{C}=\frac{\mathrm{I}\mathrm{I}\mathrm{C}–{\mathrm{I}\mathrm{I}\mathrm{C}}_{\mathrm{r}\mathrm{e}\mathrm{m}\mathrm{o}\mathrm{v}\mathrm{e}}}{\mathrm{I}\mathrm{I}\mathrm{C}}\ast 100\mathrm{\%}$$

Specifically, symbol a denotes the area of each source patch, and nl_ij is the count of links along the shortest connectivity path between source patches i and j. The symbol A represents the total landscape area (study area). The IIC values range from 0 to 1, where higher values reflect stronger connectivity. For an individual node, dIIC measures the change in landscape connectivity for the whole system when an individual node is hypothetically removed.

On this basis, this study added the calculation of several connectivity metrics (

Table 2. Connectivity metrics for evaluating the overall and local connectivity of the EBS [40,41].

Connectivity Metrics	Formula	Description
Equivalent Connected Area (ECA)	$\mathrm{ECA}=\sqrt{{\displaystyle \sum _{i=1}^n{\displaystyle \sum _{j=1}^n{\mathrm{a}}_i{\mathrm{a}}_j{\mathrm{p}}_{ij}{ }^{*}}}}$	ECA is defined as the size of a single habitat patch that provides the same probability value of connectivity (maximum connectivity) as the actual habitat pattern in the landscape.
Probability of Connectivity (PC)	$\mathrm{PC}=\frac{{\displaystyle \sum _{i=1}^n{\displaystyle \sum _{j=1}^n{\mathrm{a}}_i{\mathrm{a}}_j{\mathrm{p}}_{ij}{ }^{*}}}}{{\mathrm{A}}_L^2}$	PC is defined as the probability that two randomly placed points in the landscape will fall into an area of interconnected habitat given n habitat patches and direct connections between them. It is a network-based habitat availability index that quantifies functional connectivity.
Protected Connected Land (ProtConn)	$\mathrm{Protconn}=100\times \frac{\mathrm{ECA}}{{\mathrm{A}}_L}$	ProtConn is defined as the percentage of connected protected land in the study area.
PA Coverage/Protected Land (Prot)	$\mathrm{Prot}=100\times \frac{{\displaystyle {\sum }_{i=1}^n{\mathrm{a}}_i}}{{\mathrm{A}}_L}$	Prot is defined as the percentage of the study area covered by protected land.
Protected Not Connected Land (ProtUnconn)	$\mathrm{ProtUnconn}=\mathrm{Prot}-\mathrm{ProtConn}$	ProtUnconn is defined as the percentage of the study area covered by protected lands that are isolated.

Note: Where a_i and a_j are the areas of habitat patches i and j, respectively, n is the number of habitat patches, A_L is the maximum landscape attribute, and p_ij is the strength of each link, p_ij^∗ is the maximum product probability of all possible paths between patches i and j, including direct diffusion between the two patches. See [40,41] for the formula and detailed explanation.

) to determine the connectivity priority of the region. These metrics provide a comprehensive understanding of the integrity of EBS. Together with the fragmentation analysis, this section highlights the resilience of the ecological network system and the relative importance of each source patch.

4. Results

4.1. Overview of Fragmentation Analysis of the EBS in the SNP

According to the analysis of annual mean NDVI, the vegetation-rich areas of SNP are mainly near the eastern boundary of its Yangtze River Source sub-park, the southern part of the Yellow River Source sub-park, and across almost the entire Lancang River sub-park. In contrast, large areas in the western and central parts of the Yangtze River Source sub-park have lower NDVI values because of geographical and climatic factors. Different indicators demonstrate various characteristics of fragmentation (

Table 3. Landscape metrics of the source patches of the EBS in SNP and each of its sub-parks.

Landscape Metrics	The SNP	The Lancang River Source Sub-Park	The Yangtze River Source Sub-Park	The Yellow River Source Sub-Park
Patch area (km2)	27,784.5437	7749.8225	11,188.4614	8710.8436
Number of patches	1225	222	674	337
Size (mean)	22.6813	34.9091	16.6001	25.8482
Patches < minimum patch area	707	124	407	182
Patches < minimum patch area (%)	3.146	1.798	4.4704	2.5077
Total edge	26,323.784	6981.469	11,823.117	7482.382
Edge density	0.9474	0.9009	1.0567	0.859
Total Core Area (km2)	16,169.2662	4594.9692	6122.8223	5370.502
Cority	0.3347	0.3243	0.3101	0.3887
Shape Index (mean)	46.0948	77.399	32.0057	31.6949
FRAC (mean)	1.1668	1.1674	1.1648	1.1609
MESH (km2)	1830.817	1324.662	1615.1	788

). With an average area of 22.68 km², the source patches in the Yangtze River Source sub-park are relatively small compared to other parks. However, the edge distance of this sub-park is the highest with a value of 1.057 m/ha, indicating a highly fragmented state. In the Yangtze River Source sub-park, the Cority index and shape index are the lowest, but the effective mesh size is the largest, meaning that the source patches are compact with greater connectivity. In the Yellow River Source sub-park, the edge density and average fractal dimension are the lowest, representing smooth and compact edge patches; however, the core areas may have higher fragmentation according to the Cority index. The Lancang River Source sub-park presented better connectivity and integrity through indicators such as the mean patch size (MPS) and the effective mesh size (MESH).

4.2. Morphological Spatial Patterns of Source Areas of the EBS

MSPA can reveal the spatial composition of the EBS in terms of core areas, edge areas, and other morphological elements of an interconnected system (

Table 4. Area and proportion of different landscape morphology types according to MSPA analysis.

Landscape Morphology Type	Area (km2)	Percentage of the Area of Ecological Barrier (%)	Percentage of Study Area (%)
Core	10,197.36	36.68	8.22
Edge	9366.19	33.68	7.55
Perforation	136.46	0.5	0.11
Islet	3275.06	11.8	2.64
Loop	186.08	0.67	0.15
Branch	3361.90	12.08	2.71
Bridge	1277.77	4.59	1.03

). The core areas are far from external disturbances and relatively stable internally. They are vital to preserving habitat resilience and biodiversity. In the Yangtze River Source sub-park, the core areas cover 10,197 km², which is 36.68% of the source area and 8.22% of the study extent. Notably, major core areas are located on the eastern side of the Yangtze River Source and Lancang River Source sub-parks, and on the southern side of the Yellow River Source sub-park. Therefore, those ecological bridges and loops around them play key roles in serving as corridors in achieving landscape connectivity at the regional scale. When ecological bridges and loops are spatially dense and connected, the EBS has great potential to maintain its functionalities (Figure 3).

4.3. Graph-Based Network Analysis of the EBS

Using the graph theory approach, each ecological source area can be represented as a node, and the link between them as edges [42,43]. For the study site, the EBS is conceptualized as a network of nodes and edges. By evaluating different attributes of the nodes and edges, researchers can quantify each source area’s contribution to the overall landscape connectivity. The analysis used DN (degree of nodes) to count how many neighboring nodes surround each source area to identify nodes that play a central role in the EBS. The dIIC was used to understand each node’s impact on the network’s overall connectivity by hypothetically removing it. The results show that one source area in the Yellow River Source sub-park is dominant, with about 51 neighboring source areas. The results also identify a series of critical nodes whose DN ranges from 20 to 30. These form several south-north oriented links in the Yangtze River Source and Lancang River Source sub-parks, which represent potential corridors.

The comparison of various indicators in terms of connectivity is effective for conservation planning (Figure 4). Though those important nodes determined by dIIC in the Yellow River Source sub-park are similar to those determined by DN, this consistency is not seen in the Yangtze River Source sub-park. Generally, the relative contribution of a source area is impacted by its size, as a larger source area can influence a wider range of adjacent source areas, leading to greater landscape connectivity. This effect of area size can be reflected by various indicators. However, this study highlights the importance of small patches that are located in strategic geographic positions, as they act as key stepping stones. In conservation planning, large source patches with substantial core areas are invaluable for EBS because of natural integrity and species richness. However, some small patches also play significant roles in maintaining connectivity within the EBS.

5. Discussion

5.1. Identification of Key Areas in the EBS based on Fragmentation, Spatial Composition, and Landscape Connectivity Analysis

This study presents a multi-aspect framework to evaluate and prioritize conservation areas in EBS. According to the functional characteristics of the EBS, we conducted (1) landscape metrics and fragmentation assessment to quantify the degree of disturbance and fragmentation; (2) MSPA to understand the spatial composition of the source areas; and (3) graph-theory-based connectivity analysis to measure the relative importance of each source area for prioritizing key locations. On the one hand, the framework pays attention to the structural stability of EBS exposed to human activities and other environmental factors. On the other hand, it concerns the functionality of EBS as biological corridors [44,45,46]. The study, therefore, identified key source areas as stepping stones and a few east-west extending corridors near the eastern sides of the Yangtze River Source and Yellow River Source sub-parks (Figure 5). The average corridor density index in Figure 5 reflects the concentration of ecological corridors within different sampling areas. Higher density values indicate areas with more densely packed corridors, calling for attention to nature conservation.

5.2. Research Features and Technical Overview

This analytical framework aligns with similar studies but is innovative in several ways. First, we develop specialized modules for different attributes and functionalities of the EBS. It is important to avoid the oversimplification of the EBS and the use of a single metric. Second, the study applies spatially explicit methods to help planners locate key positions for conservation planning. Third, instead of using certain land use types to identify ecological source areas, such as grassland or forest, this study uses mean annual NDVI to reflect stable and mixed vegetation over seasons. In addition, when identifying the stable vegetation as source areas of the EBS, this study incorporates detailed road network and built environment data to better simulate how human activities impact source area fragmentation. Fourth, from the methodological perspective, this study builds the network presentation of EBS using all the source patches, instead of only a few large source areas. This method advances some existing studies and provides a more accurate understanding of regional EBS, thanks to the programming approach using R language for effective computing.

5.3. Recommendations for the Conservation Planning of EBS

5.3.1. Refining the Zoning Plan for More Effective Management

In accordance with the zoning and management requirements issued by the Chinese government, this national park is spatially divided into core conservation areas and general management areas. Studies of the EBS are closely related to both zones. The core conservation area plays a dominant role in preserving ecosystem integrity and the natural environment and in order to conserve and restore key habitats to maintain structural stability and functional connectivity. Conversely, general management zones, where human activities and developments occur, call for proper risk management measures and surveillance to protect the EBS. This study provides a spatially explicit evaluation of the source areas of the EBS, and it identifies key locations at a fine scale for supporting sustainable management. In light of sustainable management needs, key source areas were selected for their critical roles in maintaining ecosystem integrity and connectivity. The study confirms some important spatial locations revealed by existing studies but also identifies new key areas for conserving the EBS, which could help improve the zoning and management of national parks. Some planning implications are listed below.

(1)

This study highlights source areas that are vital to maintaining landscape connectivity at the regional level. Managers and practitioners should pay special attention to them. For example, in the Yangtze River Source sub-park, the east-west ecological corridors interconnect several large source areas in this region. It is important to protect those corridors from disturbance and fragmentation to maintain a larger ecological network.

(2)

This study also identifies small and fragmented areas scattered over the core source regions that are of great importance in preventing further landscape fragmentation and isolation, especially in the areas where native settlements live. It is vital to connect patches of these source areas.

5.3.2. Be Aware of Key Corridors and Nodes across Different Administrative Regions

The EBS reflects a continuous natural process that has no obvious boundary. However, it is essential to understand that for management purposes, these corridors and key source patches often span across different administrative units. These crucial key positions may encounter different situations. Some are in wild environments, some are near built environments, and some are under the pressure of environmental disturbance. While it is important to keep the integrity and continuity of the EBS across administrative boundaries, it should be recognized that key source areas are often in different geographical environments. This implies that the regional management of natural resources should consider linking ecological restoration to local conditions, especially regarding industrial planning for settlements with intense human activities. National park managers can consider using policy tools such as ecological compensation and industrial restructuring for sustainable management. These policy tools may offer promising opportunities due to the reform of the national park system. Finally, the study integrates the results of the spatial analysis with relevant policy considerations to offer planning implications for the EBS. The focus on corridors and key nodes across administrative boundaries helps preserve the continuity and functionality of the EBS.

5.4. Limitations of the Study

The study has two main limitations. First, the study identifies the source areas of the EBS using multi-source data, especially the NDVI indicator. This approach is more from a structural perspective than a functional one; therefore, it is difficult to measure the corresponding vegetation accurately. This challenge partly lies in the significant seasonal variation in vegetation, as well as the mixed types of vegetation in this region. Second, the spatial recommendations for protecting the EBS are based on modeling methods and a spatial analysis. However, the results have not been verified in the field. Although field studies concerning the practical constraints in management are beyond the scope of our specific research objective, this constitutes a major limitation of our research. To enhance robustness, future studies could consider field studies or more available multi-source datasets for cross-validation.

It should be noted that the new boundaries and area of SNP, as officially established in October 2021, differ from those published in the “Master Plan for Sanjiangyuan National Park” in 2018. Since this study was planned and conducted in late 2020 and mid-2021 and the accurate new boundaries of SNP were not available, this study used the boundaries of the 2018 version. This situation is in line with numerous similar studies. It may present limitations in understanding the latest conditions of SNP; however, the methods demonstrated in this study are transferable and the results can be referenced. If the accurate and new boundaries can be obtained from official sources, the results can be updated in subsequent studies.

6. Conclusions

For spatially identifying key conservation areas of the EBS within the SNP, this study developed a multi-faceted analytical framework. It analyzes the fragmentation of source areas, landscape morphology, and ecological connectivity based on graph theory in sequence. In general, the Lancang River Source sub-park shows better connectivity and integrity, as reflected by higher indicators such as the mean patch size and the MESH. However, analysis of the Yangtze River Source and Yellow River Source sub-parks reveals a greater degree of fragmentation through various indicators, demonstrating a complex situation of disturbances and divisions. The findings highlight the necessity of a comprehensive assessment framework to evaluate source areas of EBS from different perspectives. The MSPA has further identified important spatial components that help maintain or expand the connectivity around the core of source areas. Those spatial components include branches and bridges. Moreover, this study evaluates the characteristics of network-based connectivity using a set of indicators. This helps not only quantify how much each individual source area can impact the overall connectivity of EBS but also reveals where those key locations are. Finally, this study applies a minimum cumulative resistance model that projects a highly connected pathway between the sources indicating the key conservation areas. The analytical framework and developed implications can be adapted for broader application, especially with regard to its consideration of maintaining ecosystem integrity and enhancing connectivity. The proposed management system can guide conservation planning in similar ecological settings.