Study Area

Bijie City is located in the northwest of Guizhou province (105°36′–106°43′E, 26°21′–27°46′N). It covers an area of 26,900 km² and is divided into 8 county-level administrative units (Fig. 1). The city has a subtropical monsoon humid climate, with an average annual temperature of 13.4℃ and an annual average precipitation ranging from 849 to 1399 mm. The region has a total surface water resource of 13.487 billion m³, with the rivers in the area belonging to the Yangtze and Pearl River basins. The Yangtze River basin alone covers an area of 25,600 km², accounting for 95.3% of the city’s land area [40]. The terrain of Bijie City slopes from west to east, characterized by complex topography and abundant Karst landforms. It is known for its vulnerable agricultural ecosystem, often referred to as the “eight mountains of water or sub-fields”. As of 2022, the total population of Bijie City is 6.816 million, with a population density of over 250 people per square kilometer, which is twice the national average. It is the most populous and densely populated area in Guizhou Province. Bijie City holds the distinction of being the first central comprehensive experimental zone in China focused on poverty alleviation through development and ecological construction. This makes it a typical city in China’s western development strategy.

Fig. 1.

Location map of the study area.

Data Sources

The land use data and vector boundaries are obtained from the Bureau of Natural Resources and Planning of Bijie City. The NDVI (Normalized Difference Vegetation Index) data are sourced from the National Ecological Data Science Center, accessible at http://www.nesdc.org.cn/. The Digital Elevation Model (DEM) is acquired from the Geographic Spatial Data Cloud, available at http:// www.gscloud.cn/search. The data on National Nature Reserves are obtained from the China National Nature Reserve Specimen Resource Sharing Platform, accessible at http://www.papc.cn/. The river and road data are sourced from the Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, accessible at http://www.resdc.cn. All the above-mentioned data are projected in the WGS_1984_UTM_Zone_48N coordinate system. However, since the land use data are point data with insufficient continuity, the raster data resolution is corrected to 30m using bilinear interpolation resampling.

Research Framework

In this study, Bijie City is the focus area. The research methodology includes several steps. Firstly, core areas exceeding 10 km² are extracted for landscape connectivity analysis using the Morphological Spatial Pattern Analysis (MSPA) method. Ecological source areas are then identified based on the patch importance index. Secondly, six resistance factors, namely land use, vegetation coverage, terrain ruggedness, slope, distance to rivers, and distance to roads, are selected. The Analytic Hierarchy Process (AHP) and the ArcGIS platform are used to establish a comprehensive resistance surface. Thirdly, ecological corridors are extracted using the circuit method. This is based on the results of the analysis of the ecological source areas and resistance surfaces. The ecological corridors are then graded according to the ratio of cost-weighted distance to least cost path (CWD/ LCP). Fourthly, a circuit-theory based analysis method is employed to identify ecological pinch points and barrier points. These critical areas are considered significant for ecological restoration. Finally, based on the findings from the previous steps and considering the natural ecological characteristics of the study area, ecological restoration zoning is conducted (Fig. 2). This zoning helps in determining the areas that require ecological restoration efforts.

Identification of Ecological-Source-Lands

Ecological-source-land is the core element of the ecological security pattern, as it provides important patches for the exchange and dispersal of ecological flow [41]. The selection of ecological-source-land should consider important principles, including high ecosystem service value, favorable habitat quality, and landscape connectivity [42]. National-level nature reserves, which serve functions such as species protection and ecological process maintenance, play a vital role in providing habitats for species and promoting their interaction. These reserves possess high ecological functionality and can be directly chosen as ecological-source-lands [43]. MSPA (Multiscale Patch Analysis) is a method based on digital morphology principles that measures, identifies, and segments the spatial pattern of raster images. It can identify areas that contribute significantly to landscape connectivity at the pixel level, making the selection of ecological-source-lands more scientifically grounded [44, 45]. Therefore, this study considers both ecological functionality and spatial structure [46]. By selecting national-level nature reserves and employing MSPA and landscape connectivity methods, ecological-source-lands can be identified more accurately.

Screening Ecological-Source-Lands Based on Landscape Connectivity

Landscape connectivity refers to the extent to which the structure of terrestrial landscapes facilitates or impedes biotic flows, and its strength can be used to judge the degree of connectivity between patches as well as to reflect the spatial structure of ecosystem landscape elements [50–52]. This study uses the probable connectivity index (PC) and the patch importance index (dPC) to analyze the level of connectivity of core areas and their importance to landscape connectivity. Utilizing Conefor Sensinode 2.6 software, 2000m is selected as the distance threshold based on the core patch size and distribution in the study area, and the connectivity probability is set to 0.5. The formulas for calculating PC and dPC are as follows:

Fig. 2.

Research framework.

Table 1.

Meaning of MSPA landscape types.

Landscape types	Ecological implications
Core Zone	Larger habitat patches in the foreground image provide larger habitats for species and are ecological-source-lands in ecological networks.
Island	Isolated, fragmented and poorly connected small patches that are not connected to each other, with less potential for internal material and energy exchange and transfer.
Hole	Transition areas between core and non-green landscape patches, i.e. internal patch edges (edge effects).
Edge	Transition area between the core area and the main non-green landscape areas.
Circle	Corridors connecting the same core area, shortcuts for species migration within the same core area.
Bridge	Narrow areas connecting to the core, corridors connecting patches in the ecological network.
Branch-line	Areas connected to edge, bridge, circle or pores at one end only.

Table 2.

Grading and weight assignment of ecological resistance factors.

Resistance factors	Resistance value					Weight	References
	1	3	5	7	9
Land use type	forest,water body	grassland	cultivated land	bare land	construction land	0.276	[58]
Vegetation cover	> 0.72	0.58–0.72	0.42–0.58	0.20–0.42	< 0.20	0.138	[59]
Terrain ruggedness	< 18	18–30	30–46	46–71	> 71	0.109	[60]
Slope	< 4.27	4.27–12.64	12.64–23.55	23.55–38.48	> 38.48	0.087	[61]
Distance from rivers	< 500	500–1000	1000–1500	1500–2000	> 2000	0.195	[62]
Distance from roads	> 1000	800–1000	500–800	500–200	< 200	0.195	[58]

(1)

PC=∑i=1n∑j=1nai·aj·pij*AL2j

(2)

dPC=PC−PCremovePC×100%

In this formula, n stands for the total number of patches in the study zone, aⁱ and a^j corresponds to the areas of patches i and j respectively, while p*_ij represents the highest path connectivity between patches i and j. PC_remove is the value of the likelihood connectivity index for a given patch removal. The larger the dPC value, the greater the connectivity importance of the patch.

Construction of the Ecological Resistance Surface

The ecological resistance surface, a virtual resistance surface mirrors the challenges and impediments encountered by species during migration in reality [53]. The construction process unfolds as follows: Initially, comprehensively considering the impacts of the natural environment, topography, and anthropogenic activities in the study area, six indicators, namely, land use type, vegetation cover, terrain ruggedness, slope, distance from rivers, and distance from roads, are selected as the resistance factors for the ecological-source-lands. Subsequently, the range of values for these factors is set to be between 1 and 9 [54–57], where higher values indicate greater resistance and vice versa. Finally, the Analytic Hierarchy Process (AHP) and leveraging the natural breaks method available in the ArcGIS platform are used to grade and assign weights to each factor (Table 2), and the weighted summation is calculated to obtain the integrated resistance surface.

Extraction of Ecological Corridors

The ecological corridor plays a crucial role in facilitating the smooth movement of matter, information, and energy between ecological source areas [63]. It also serves as a low-resistance pathway for ecological flows between neighboring ecological source areas [64]. By combining the connectivity model with the random-walk theory in the electrical resistive method, this study provides a more accurate evaluation of least-cost paths. Areas with higher current density indicate better connectivity and lower ecological resistance in the region [65]. Therefore, the electrical resistive method is employed to extract the ecological corridors in the study area. The Linkage Mapper plugin in ArcGIS 10.6 is used to extract the least costly path that connects the two ecological source areas, which serves as the ecological corridor.

Identification of Key Areas for Ecological Restoration

Ecological pinch points (PinchPoints) are critical locations for biological migration that, when degraded or obstructed, have a significant impact on ecological stability. From the perspective of electrical current theory, pinch points are landscape patches characterized by high current density and low resistance. They can be considered “bottleneck zones” that influence landscape connectivity, and their degradation or loss can potentially disrupt habitat connectivity [66]. Ecological barriers, on the other hand, are resistance points that impede normal ecological flow. Restoring high barrier points can enhance the connectivity of the natural landscape and ensure the smoothness and integrity of biological migration processes [67]. Therefore, this study focuses on identifying and analyzing ecological pinch points and barriers to identify key areas for ecological restoration in the study area [68]. The Pitchpoint Mapper mode of the Circuitscape plugin is employed for identifying ecological pinch points, while the Barrier Mapper module is used for identifying ecological barrier points.

Fig. 3.

Distribution of MSPA landscape types in Bijie.

Identification of Ecological-Source-Lands

Selection of Ecological-Source-Lands

After conducting the MSPA analysis, patches with a core area larger than 10 km² are selected and imported into Conefor2.6 for landscape connectivity calculation. By combining the results of the patch importance analysis presented in Table 4 with the current national-level nature reserves and the actual situation of the study area, a total of 62 patches are identified as ecological source lands (Fig. 4), with a combined area of 3944.37 km². Among these patches, those with an area larger than 20 km² and a dPC (patch importance index) greater than 4 are selected as key ecological source lands, in addition to the national-level nature reserve patches. This results in a total of 26 source areas with a combined area of 2643.35 km², accounting for 67% of the total source area. These key source areas are mainly distributed in the QXG district, DF, and NY counties in the northern, central, and southern regions of Bijie City. Patches 61 and 62 correspond to the Rare and Endemic Fish Nature Reserve and the Caohai Nature Reserve in the Bijie segment of the upper reaches of the Yangtze River, respectively, with a combined area of 229.72 km², representing 5.82% of the total source area.

Table 3.

Statistics on the area and percentage of MSPA landscape types in 2022.

Landscape types	Area (km2)	Percentage of foreground landscape(%)	Percentage of overall landscape(%)
Core Zone	9138.16	60.98	34.04
Island	334.39	2.23	1.25
Hole	364.29	2.43	1.36
Edge	3119.03	20.81	11.62
Circle	305.90	2.04	1.14
Bridge	745.91	4.98	2.78
Branch-line	977.21	6.52	3.64
Total	14984.88	100.00	55.82

Table 4.

Landscape connectivity of ecological-source-lands.

Patch number	dPC	Area (km2)	Patch number	dPC	Area (km2)	Patch number	dPC	Area (km2)
1	35.10	431.26	20	4.53	41.14	39	1.97	21.72
2	29.17	96.27	21	4.33	25.57	40	1.94	58.40
3	28.79	168.90	22	4.05	75.73	41	1.93	10.55
4	24.69	240.78	23	4.03	20.29	42	1.76	31.34
5	23.94	291.29	24	4.02	19.07	43	1.49	27.68
6	21.00	366.92	25	3.86	15.64	44	1.43	32.75
7	20.23	55.53	26	3.84	105.09	45	1.43	29.31
8	19.12	181.87	27	3.82	36.70	46	1.36	14.82
9	11.38	18.52	28	3.46	17.01	47	1.33	39.58
10	11.08	31.97	29	3.26	11.12	48	1.32	26.80
11	10.67	10.17	30	3.22	50.97	49	1.24	38.87
12	10.21	31.85	31	3.13	57.52	50	1.24	25.89
13	9.04	59.42	32	2.82	13.90	51	1.11	24.94
14	8.03	20.59	33	2.68	105.02	52	1.08	14.06
15	7.12	103.37	34	2.59	15.43	53	1.08	13.14
16	7.09	34.55	35	2.49	19.11	54	1.07	19.39
17	6.47	19.25	36	2.36	50.54	55	1.02	21.76
18	5.77	31.82	37	2.30	46.38	56	1.02	29.95
19	5.60	37.34	38	1.98	43.87	57	1.01	123.02

Additionally, patches with an area larger than 10 km² and 4 > dPC >1 are identified as important ecological source areas, resulting in 33 source areas with a combined area of 1191.49 km², accounting for 30.2% of the total source area. These areas are primarily located around the key source sites, with a few patches also found in the western region of Bijie City. Notably, patches 58, 59, and 60 in the western region of Bijie City hold significant positions, each with an individual area exceeding 10 km². Considering the spatial distribution structure of the landscape pattern in the western region, patches 58, 59, and 60 are chosen as important ecological source lands, with a combined area of 111.40 km², representing 2.83% of the total source area.

Fig. 4.

Distribution map of ecological-source-lands in Bijie.

Fig. 5.

Comprehensive resistance surface.

Fig. 6.

Resistance surface of each indicators.

Establishment Results of Resistance Surface

The different resistance factors are combined and overlaid using weighted summation analysis, resulting in the creation of the Comprehensive Resistance Surface for Bijie City (Fig. 5) and the individual indicator resistance surfaces (Fig. 6). Fig. 5 illustrates that the comprehensive ecological resistance surface in Bijie City is generally high, with localized areas reaching very strong levels of resistance. At the same time, there is pronounced spatial heterogeneity, with the highest resistance values above moderate levels observed in the southwestern region, while the northern region exhibits lower resistance. It is worth noting that the crimson-colored regions representing the highest resistance values in both figures are intricately intertwined with both natural and anthropogenic factors. Fig. 6 demonstrates that terrain ruggedness has a significantly greater influence on the resistance surface compared to slope. Terrain ruggedness shows a more than moderate resistance influence throughout the city, while slope does not have a discernible impact on resistance. Vegetation cover exhibits concentrated pockets of moderate-to-high resistance zones in the southwestern and southeastern parts of Bijie City. However, in extensive areas spanning the central, northern, and southern-central regions, resistance remains predominantly minimal. The influence of land use is distributed ubiquitously and is particularly pronounced in the western zone, where the distribution of resistance is denser at the upper end of the medium range. However, the areas with the highest red resistance values are more distributed in the northern region. The most significant influence on the comprehensive resistance surface is the distribution of roads and towns, with areas of high resistance values highly correlated with the extent of roads and urbanization in each county. At the county level, the highest concentration of high resistance values is found in the QXG district adjacent to DF county, followed by the three counties or municipalities of NY, ZJ, and QX, which are located in the southern part of Bijie City. In these areas, high resistance values are primarily associated with roads and construction land. The western WN county, known for having the most cultivated land in Bijie City, exhibits a higher distribution of surrounding high resistance values compared to other counties and cities, except for the construction of land and roads. Low resistance values in Bijie City are mainly found in the northern areas of DF and JS counties, which are dominated by forests and have fewer construction sites for land and roads, resulting in lower resistance values compared to other areas.

Fig. 7.

Distribution of ecological corridors.

Extraction Results of Ecological Corridors

Based on the analysis results of ecological-source-lands and the resistance surface, ecological corridors are identified by calculating the least-cost pathway using the Linkage Mapper tool (Fig. 7). A total of 147 corridors are extracted, with corridor lengths ranging from 0.05 to 72.63 km, resulting in a total length of 1333.99 km. Spatially, the western region of Bijie City, which has a scarce and diffuse distribution of ecological-source-lands, predominantly consists of long-distance corridors. These corridors facilitate the exchange of species and information between the western and eastern source sites. On the other hand, the QXG district and DF county, where source sites are abundant and concentrated, exhibit a higher concentration of medium and short-distance corridors, ensuring strong connectivity. By calculating the ratio of Cost-Weighted Distance (CWD) to the Least Cost Path (LCP), the corridors are classified into three grades using the natural breakpoints method. A smaller ratio indicates less resistance encountered by species along the path, and vice versa. Consequently, the corridors are categorized into critical ecological corridors, important ecological corridors, and general ecological corridors based on the CWD-to-LCP ratio. Among these corridors, there are 45 critical ecological corridors with a total length of 353.44 km. The longest corridor spans 52.53 km, connecting source sites 59 and 60. Additionally, there are 65 important ecological corridors with a total length of 869.15 km. The longest corridor extends for 72.63 km, making it the longest ecological corridor in Bijie City, connecting source sites 31 and 58. Lastly, there are 37 general ecological corridors with a total length of 111.39 km. The longest corridor measures 22.65 km, connecting source sites 44 and 47.

Fig. 8.

Distribution of ecological pinch points.

Identification of Key Areas

The analysis of ecological corridors includes the identification of ecological pinch points using the Pinchpoint Mapper tool in “All to One” mode. The current density is classified into four levels using the natural break point method, and the highest current density level is selected as the alternative pinch point. Redundant patches with an area smaller than 0.1 km² are removed. In total, 23 key ecological pinch points are extracted, covering an area of 10.85 km². The results indicate that areas with higher current density are mostly located on both sides of the corridors. The land use types at the pinch points primarily consist of forests, cultivated lands, water bodies, and grasslands (Fig. 8, 10). The spatial distribution analysis of ecological pinch points reveals that not every corridor has pinch points along its route, and the distribution of pinch points is more concentrated in the northern and central regions of Bijie City. In the QXG district, which has abundant and compact source sites and shorter corridors, there is a higher current density and concentration of ecological pinch points, facilitating frequent species exchange and migration. At the boundary junction of DF and JS counties, a string of ecological pinch points forms along a natural river corridor. These pinch points serve as stepping stones and should be prioritized for protection. In the western region, which has the only key pinch point, there is intensive species exchange and vital interaction with eastern source areas.

Using the Barrier Mapper module, the ecological obstacle areas in Bijie City are identified with a search radius of 200m. They are divided into five levels using the natural break point method, and the areas with the highest cumulative current density are extracted as alternative ecological barrier sites. Fragmented and redundant patches are eliminated, and areas larger than 0.1 km² are selected as critical ecological barrier points in Bijie City. In total, 31 barrier points were extracted, covering a total area of 25.70 km². The spatial distribution of ecological barrier points is predominantly concentrated in the central and northern regions of Bijie City, with fewer occurrences in the western and southern areas, and no distribution in the eastern part (Fig. 9, 10). Among these, the distribution of barrier points is most concentrated in the QXG district, DF county, and HZ county. Multiple barrier points can be found along a single corridor, possibly due to transportation infrastructure such as railways and national highways hindering species exchange along the corridors. Barrier points are predominantly concentrated in areas designated for construction and agricultural land use, which are influenced by human activities and lead to increased cumulative current values.

Fig. 9.

Distribution of ecological barrier points.

Fig. 10.

Key ecological pinch points and barrier points.

Fig. 11.

Ecological restoration zoning.

Ecological Restoration Zoning

Based on the ecological security needs and the characteristics of natural ecosystems in Bijie City, and considering the requirements of territorial spatial sustainability for ecosystem integrity, systematicity, and continuity, the land cover types in Bijie City are divided into five areas: ecological controlling zone, buffer zone, conservation zone, optimization zone, and corridor restoration zone. This division is done to establish the potential ecological restoration pattern of the territorial spatial space in Bijie City (Fig. 11). Ecological-source-lands, which have high ecological service values and good environmental quality, and ecological pinch points, which are important nodes with high current density and frequent species exchange, are considered priority areas for ecological protection. The edge areas of ecological-source-lands and pinch points play a crucial role in maintaining the healthy functioning of ecosystems and promoting continuous species exchange. Therefore, a 1 km buffer zone is established between ecological-source-lands and pinch points as an ecological radiation zone to enhance connectivity and integrity between source sites and pinch points. Ecological barrier points hinder species exchange and migration and disrupt landscape connectivity. Thus, high-value areas of ecological barriers are designated as key areas for ecological restoration. The comprehensive resistance surface map indicates that high resistance value areas are mostly associated with construction areas and roads. However, socio-economic development should not come at the expense of the ecological environment. Therefore, based on the distribution characteristics of the resistance surface, the comprehensive resistance surface is divided into five levels using the natural break method. The top two levels are selected as ecological control zones, while the remaining levels are designated as ecological enhancement zones.

Ecological Restoration Strategies

Strategies for five types of ecological areas are proposed as follows:

In the ecological conservation area, which consists primarily of forests, grasslands, and water bodies, strategies should focus on minimizing the expansion of construction land and reducing human interference. It is crucial to strictly prohibit any changes to the current land use status and strengthen monitoring and control measures. Respecting and adhering to the natural processes of plant and animal succession is essential, and activities that disrupt natural growth and life processes should be strictly forbidden. Illegal logging, deforestation, and pollution discharge must be strictly prohibited. Afforestation can be implemented in narrow ecological pinch points to widen the width of these areas, and any activities that may harm biodiversity within pinch point areas must be strictly prohibited. Establishing a comprehensive environmental monitoring system and conducting regular assessments in sensitive areas is necessary to address issues promptly. Additionally, considering the uneven distribution of ecological-source-lands in Bijie City, it is advisable to increase ecological source areas in the eastern and western regions. Restoration efforts in these areas should focus on expanding core areas and enhancing patch connectivity by restoring larger areas of forests and grasslands.

In the ecological buffer zone, which acts as subordinate areas for ecological-source-lands, strategies should encompass both ecological conservation and restoration. The key objective is to repair fragmented patches outside the source sites and establish connections between these patches through ecological corridors or man-made connective points based on local conditions. Environmental protection should be prioritized in human activities within this region, and efforts to protect and restore the environment should be strengthened. Strict control measures should be implemented to regulate the current land use status, and measures to control the expansion of construction facilities are necessary.

In the corridor restoration zone, which is primarily located in high-resistance value areas, restoration efforts should focus on enhancing connectivity between source areas and facilitating species exchange. This area consists mainly of construction land, cultivated land, and grassland. In construction land areas, it may not be feasible to remove dwellings or traffic roads, but the construction of ecological corridors such as culverts, tunnels, or flyovers in the vicinity can facilitate species migration and exchange. For cultivated land areas, planting forestry industries such as fruit trees around the cultivated land can maximize ecological and economic benefits. Additionally, fallow land can be returned to forests to restore and protect vegetation cover, thus reducing ecological resistance. In grassland areas, reasonable grazing management and rotational grazing systems should be implemented to control the intensity of grazing. Strengthening the control and management of invasive plants and harmful organisms is crucial to prevent damage to the grassland ecosystem and reduce ecological resistance.

In the ecological optimization area, which covers the largest areas in Bijie City and includes various land use types, strategies should be tailored to specific landscapes. For small-sized forests and low-connectivity landscapes, efforts should focus on conserving existing forests, planting appropriate trees in marginal areas, and improving connectivity. Pollution in cultivated lands should be reduced, and the promotion of eco-friendly agricultural technologies such as organic farming, precision agriculture, and water-saving agriculture can contribute to the ecological environment and improve the quality and safety of agricultural products. The ecological landscape of farmland should be rationally planned, with the establishment of forbidden lines and landscape green belts to recover ecological functions and provide better ecosystem services and biological habitats. Implementing rotational or rest grazing systems can help in the recovery and productivity growth of grassland ecosystems, reducing overuse pressure. Controlled application of fertilizers contributes to soil quality recovery and meets the needs of grassland production while reducing water and soil pollution. Rational management and utilization of water resources are essential for ensuring the abundance of water ecosystems and a stable supply of aquatic products. When necessary, water purification and pollution control measures should be adopted to control agricultural and industrial wastewater discharge. Protection strategies for wetlands can increase connectivity among water bodies and wetlands, provide better habitats, and promote biodiversity conservation, thereby stabilizing aquatic ecosystems.

In the ecologically controlling zone, which is characterized by intensive human activities and high resistance values, areas such as urban regions, rural residential areas, and transportation corridors are typically included. It is recognized that economic development is crucial for people’s livelihoods, especially in remote rural areas. However, this development should not come at the expense of unacceptable ecological degradation. Therefore, it is important to clearly define the scope and boundaries of the ecologically controlled areas. To regulate development activities and minimize ecological degradation, various measures can be implemented. Billboards, educational advertisement boards, and scientific or notice boards should be set up around the perimeter of the control zone to raise awareness and inform the public about the importance of ecological protection. Law enforcement and supervision over the controlling areas should be strengthened to effectively regulate development activities, including preventing illegal mining and unauthorized construction. Additionally, efforts should be made to repair and restore damaged habitats within the control zone. This can be achieved by providing economic and technical support, encouraging local residents and enterprises to participate in protection and restoration efforts, and offering reasonable compensation and sustainable development opportunities. By involving local stakeholders and providing incentives for conservation and restoration, it becomes more feasible to achieve ecological protection goals while also addressing economic development needs.