Assessment results of ES
Spatial and temporal variation in ES
Carbon storage
Regarding spatial distribution, the overall carbon storage in Giant Panda National Park shows the spatial distribution characteristics of high in the north and south and low in the middle. Regarding time, the variation of carbon storage in Giant Panda National Park from 2005 to 2020 was not obvious, as shown in Fig. 4. So the carbon storage in different years was calculated by using ArcGIS, the carbon storage in Giant Panda National Park in 2005, 2010, 2015, 2020 were 78.28 × 107t, 78.77 × 107t, 79.03 × 107t, 79.45 × 107t, with an overall increase of 1.17 × 107t from 2005 to 2020, including an increase of 0.49 × 107t from 2005 to 2010, 0.26 × 107t from 2010 to 2015, and 0.42 × 107t from 2015 to 2020, as shown in Table 2. This is mainly due to the increasing area of forest and grassland in the region, increasing forest cover, and increased carbon storage services.
Spatial distribution of carbon storage from 2005 to 2020. (The map was generated by ArcGIS 10.8 and does not require any permission from anywhere).
Habitat quality
Regarding spatial distribution, the habitat quality of the Giant Panda National Park showed a spatial distribution pattern of low in the middle and high around, with little spatial interannual variation, as shown in Fig. 5. The average habitat quality indices of different years were calculated using ArcGIS, as shown in Table 2, the average habitat quality indices of giant panda national park in 2005, 2010, 2015 and 2020 were 0.89, 0.90, 0.90 and 0.84, respectively, and the average habitat quality index was generally high, but decreased in 2020. Human activities and disturbance of the various site types are the root causes of changes in habitat quality.
Spatial distribution of habitat quality from 2005 to 2020. (The map was generated by ArcGIS 10.8 and does not require any permission from anywhere).
Water conservation
Regarding spatial distribution, the water conservation content of Giant Panda National Park spatially showed the spatial distribution characteristics of high in the west and low in the east. Regarding time, the spatial variation of water content in Giant Panda National Park from 2005 to 2020 is not apparent, as shown in Fig. 6. So the water conservation content of different years was calculated by using ArcGIS, as shown in Table 2, the water conservation content of Giant Panda National Park in 2005, 2010, 2015 and 2020 were 52.157 billion cubic meters (bcm), 62.093 bcm, 57.315 billion cubic meters and 69.617 bcm, with an overall increase of 17.46 bcm from 2005 to 2020. Among them, 9.936 bcm increased from 2005 to 2010, 4.778 bcm decreased from 2010 to 2015, precipitation fluctuations are the main cause of water conservation phase changes, and 12.302 bcm increased from 2015 to 2020.
Spatial distribution of water conservation from 2005 to 2020. (The map was generated by ArcGIS 10.8 and does not require any permission from anywhere).
Soil conservation
In terms of spatial distribution, the soil conservation high value area of Giant Panda National Park mainly appeared in the center of the study area. Regarding time, the spatial variation of soil conservation content in Giant Panda National Park from 2005 to 2020 is not obvious, as shown in Fig. 7. So the soil conservation content of different years was calculated by using ArcGIS, as shown in Table 2. The overall trend of soil conservation in Giant Panda National Park from 2005 to 2020 is increasing, with a total increase of 6.22 × 109 t. The soil conservation in Giant Panda National Park in 2005, 2010, 2015, and 2020 is 37.49 × 109 t, 40.82 × 109 t, 41.75 × 109 t, and 43.71 × 109 t, respectively, of which 3.33 × 109 t from 2005 to 2010, 0.93 × 109 t from 2010 to 2015, and 1.96 × 109 t from 2015 to 2020. The small increase in soil conservation from 2010 to 2015 over 2005–2010 is primarily related to the rate of increase in forest cover, which increased faster in 2010–2015 than in 2005–2010.
Spatial distribution of soil conservation from 2005 to 2020. (The map was generated by ArcGIS 10.8 and does not require any permission from anywhere).
Analysis of ecosystem services influence factors
Determinants of the spatial heterogeneity of ecosystem services q values for each influence factor for 2005–2020 were obtained by geographic detector. Lucc was the main factor influencing the spatial heterogeneity of carbon storage from 2005 to 2020. The q values of lucc from 2005 to 2020 were 0.98, 0.99, 0.98, and 0.97, respectively. Lucc was the main factor influencing the spatial heterogeneity of habitat quality from 2005 to 2020. The q values of lucc from 2005 to 2020 were 0.48, 0.42, 0.42, and 0.44, respectively. Evapotranspiration (ETO) was the main factor influencing the spatial heterogeneity of water conservation from 2005 to 2020. The q values of Evapotranspiration (ETO) from 2005 to 2020 were 0.68, 0.66, 0.65, and 0.64, respectively. Rainfall was the main factor influencing the spatial heterogeneity of soil conservation from 2005 to 2020. The q values of Evapotranspiration (ETO) from 2005 to 2020 were 0.47, 0.51, 0.55, and 0.60, respectively. As shown in Fig. 8.
Determinants of force q for different ecosystem service influences. (The map was generated by Microsoft Office PowerPoint and does not require any permission from anywhere).
Land use scenario simulation
Analysis of land use change
From 2005 to 2010, cultivated land and grassland in the Giant Panda National Park decreased by 57.3 km2 and 477.58 km2, respectively, while forest land, water area, construction land and unused land increased by 441.26 km2, 20.14 km2, 50.64 km2 and 22.84 km2, respectively. From 2010 to 2015, cultivated land, forest land, grassland and unused land decreased by 9.58 km2, 7.51 km2, 7.32 km2 and 0.36 km2, respectively, while water area and construction land increased by 10.34 km2 and 14.43 km2, respectively. From 2015 to 2020, cultivated land, forest land, construction land and unused land decreased by 1.6 km2, 50.25 km2, 0.01 km2 and 1.98 km2, respectively, while grassland and water area increased by 53.44 km2 and 0.4 km2, respectively. The trend in land-use type change is mainly from cultivated land to forest land and grassland. This is mainly because after the national park area has been demarcated, the area is mainly for the protection of giant panda habitats, and it is necessary to return cultivated land to forest land and grassland in order to restore the local ecological functions, so part of the cultivated land will be converted into forest land and grassland. As shown in Fig. 9.
Different land use change maps.
Land use scenario simulation results
Figure 10 and Table 3 show that the land use structure of Giant Panda National Park is dominated by forest land and grassland. Comparing the area of land use types in 2035 with that in 2020, under the natural development scenarios (NDS), the area of cultivated land decreases by 5.58 km2, the area of forest land increases by 9.23 km2, the area of grassland decreases by 30.21 km2, the area of water increases by 0.91 km2, the area of construction land increases by 0.69 km2, and the area of unused land decreases by 0.14 km2. Under the ecological protection scenarios (EPS), the cultivated land area decreased by 5.58 km2, the forest land area increased by 326.06 km2, the grassland area decreased by 335.89 km2, the water area increased by 0.91 km2, the construction land area decreased by 10.47 km2, and the unused land area decreased by 0.14 km2.
Spatial distribution of land use under different scenarios. (The map was generated by ArcGIS 10.8 and does not require any permission from anywhere).
Ecological space management and control zoning results
Conditional probability values for each node
The nodes extracted after the discretization process were fishnetted to get 38,595 sample data, and the “Incorp Case File” tool of Netica was used for parameter learning to get the bayes belief network (BBN) of Giant Panda National Park in 2020. The arrow pointing between the nodes represents the causal link, the node led by the arrow line is the cause, and the node introduced by the arrow line is the effect, the results are shown in Fig. 11.
In this study, the discrete data of land use under the natural development scenarios (NDS) and ecological protection scenarios (EPS) were input into the BBN-based ecosystem services model, and the probability values of habitat quality (HQ), water conservation (WC), soil conservation SC), and carbon storage (CS) under the natural development scenarios (NDS) and ecological protection scenarios (EPS) of Giant Panda National Park in 2035 were obtained, as shown in Table 4.
Identifying key variables
In this study, the top three influence factors for each ecosystem services variance reduction value were selected as key variables, as shown in Table 5, where the key variables for habitat quality were lucc, rainfall erosion force (R), and population density (PD), the key variables for water conservation were evapotranspiration (ETO), lucc, and rainfall, the key variables for soil conservation were lucc, normalized difference vegetation index (NDVI), and rainfall erosion force (R), and the key variables for carbon storage were lucc, population density (PD), and normalized difference vegetation index (NDVI).
Spatial distribution of key state subsets
In this study, Netica software was used to obtain the key variable key state subsets (KSS) corresponding to different levels of each ecosystem services in Giant Panda National Park, as shown in Table 6.
Based on the identified key variables and key state subsets (KSS), the raster layers of the key variables and the key state subsets (KSS) were intersected using the Con function of ArcGIS software raster calculator. The spatial distribution of key state subsets (KSS) under the natural development scenarios (NDS) and ecological protection scenarios (EPS) in 2020 and 2035 was obtained for Giant Panda National Park. As shown in Fig. 12.
Subset of key states of ecosystem services under different scenarios. (The map was generated by ArcGIS 10.8 and does not require any permission from anywhere).
The key state subsets I (KSS I) of carbon storage is the set of high probability states when the carbon storage is low, corresponding to the lucc is unused land, high population density (PD), and low normalized difference vegetation index (NDVI), and the area is distributed in strips in each area. The key state subsets II (KSS II) is the set of high probability states when the carbon storage is medium, which corresponds to the lucc is cultivated land, and the population density (PD) and the normalized difference vegetation index (NDVI) are high. The key state subsets III (KSS III) is the set of high probability states when the carbon storage is high, which corresponds to the lucc is grassland, low population density (PD) and high normalized difference vegetation index (NDVI). The key state subsets IV (KSS IV) is the set of high probability states when carbon storage is the highest, corresponding to the lucc is forest land, with low population density (PD) and the highest normalized difference vegetation index (NDVI), and the region has a wide spatial distribution and the largest area.
The key state subsets I (KSS I) of habitat quality is the set of high probability states when the habitat quality is low, corresponding to the lucc is forest land, rainfall erosion force(R) is medium and the population density (PD) is high. The key state subsets II (KSS II) is the set of high probability states when the habitat quality is medium, corresponding to the lucc is forest land, R is low and the population density (PD) is low. The key state subsets III (KSS III) is the set of high probability states when the habitat quality is high, corresponding to the lucc is grassland, rainfall erosion force(R) is medium and population density (PD) is low. The key state subsets IV (KSS IV) is the set of high probability states when the habitat quality is the highest, corresponding to the lucc is forest land, rainfall erosion force (R) is medium and population density (PD) is low, the ecological condition of this area is better, less affected by human activities, and the spatial distribution area is wider.
The key state subsets I (KSS I) of water conservation is the set of high probability states when the water conservation is low, which corresponds to the area where the lucc is forest land, the Evapotranspiration (ETO) is high and the rainfall is low. This area is mainly distributed in the Qinling area and Baishuijiang area, while other areas are scattered. Key state subsets II (KSS II) is the set of high probability states when the water conservation is medium, which corresponds to the lucc is grassland and the area with medium Evapotranspiration (ETO) and rainfall. The spatial distribution of this area is fragmented and has no obvious spatial characteristics. The key state subsets III (KSS III) is the set of high probability states when water conservation is high, corresponding to the lucc is grassland, low Evapotranspiration (ETO) and medium rainfall, which is mainly distributed in the northern part of Qinling Mountains, the western part of Minshan Area and the central part of Qionglai Mountain-Daxiangling Area, and also distributed in the edge of subset IV. The key state subsets IV (KSS IV) is the set of high probability states when water conservation is the highest, corresponding to the lucc is grassland, with low Evapotranspiration (ETO) and high rainfall, which is mainly distributed in the northern part of Minshan Area and Qionglai Mountain-Daxiangling Area, with more concentrated distribution.
The key state subsets I (KSS I) of soil conservation is the set of high probability states when soil conservation is low, corresponding to the lucc is cultivated land, and the normalized difference vegetation index (NDVI) and rainfall erosion force (R) are the highest. This area is mainly located in the Qinling Mountains and the southern part of the Minshan Area, with a relatively large area. The key state subsets II (KSS II) is the set of high probability states when soil conservation is medium, corresponding to the lucc is cultivated land, the highest normalized difference vegetation index (NDVI) and high rainfall erosion force (R), and the distribution of this area is more scattered, and the area is most widely distributed in the Minshan area and Qionglai Mountain-Daxiangling area. The key state subsets III (KSS III) is the set of high probability states when the soil conservation is high, corresponding to the lucc is forest land, with the highest normalized difference vegetation index (NDVI) and medium rainfall erosion force(R). The key state subsets IV (KSS IV) is the set of high probability states when soil conservation is the highest, corresponding to the lucc is forest land, with the highest normalized difference vegetation index (NDVI) and low rainfall erosion force (R).
Ecological space management and control zoning
This paper uses the ecological space management and control zoning (ESMCZ) framework to spatially overlay different key state subsets of ecosystem services in Giant Panda National Park under two scenarios of natural development and ecological protection in 2020 and 2035, and obtains the comprehensive key state subsets (CKSS) of ecosystem services in Giant Panda National Park under different scenarios in 2020 and 2035, as shown in Fig. 13, and overlaying the comprehensive key state subsets (CKSS) for different scenarios. Meanwhile, with reference to the General Plan of Giant Panda National Park and the Interim Measures for National Park Management, the Giant Panda National Park is divided into strictly protected zone, ecological buffer zone, ecological control zone and control development zone, as shown in Fig. 14.
Comprehensive key state subset spatial distribution. (The map was generated by ArcGIS 10.8 and does not require any permission from anywhere).
Ecological spatial management and control zoning of Giant Panda National Park. (The map was generated by ArcGIS 10.8 and does not require any permission from anywhere).
Strictly protected zone
The strictly protected zone has a good ecological and natural background condition and a large potential demand, mainly concentrated in the southern part of Qinling Area and Baishuijiang Area, the western and northern part of Minshan Area, and the north and south ends of Qionglai Mountain-Daxiangling Area, which is dominated by forest land and grassland, a natural ecological barrier and the area with the highest ecological value. The strictly protected zone is 19,370.9km2, accounting for 50% of the total area of the study area, with the widest distribution area.
Ecological buffer zone
The ecological buffer zone has a medium ecological background and low potential demand, mainly concentrated in the southern and central part of the Qinling Area, the northern part of the Baishuijiang Area, and the central part of the Qionglai Mountain-Daxiangling Area, which is dominated by grassland and forest land. The control objective is to improve the productivity and conservation of ecosystem services based on the optimization of ecosystem services in the area. The ecological buffer zone covers an area of 13,247.52km2, accounting for 34% of the total area of the study area.
Ecological control zone
The ecological control zone is mainly forest land and cultivated land, with a small portion of land for construction, distributed primarily on the northern part of the Qinling Area, the southern fringe of the Baishuijiang Area, the eastern part of the Qionglai Mountain-Daxiangling Area, and the southeastern part of the Minshan Area. Most of the existing developed areas of the Giant Panda National Park sightseeing spots are in the ecological control zone, which covers an area of 5646.86 km2, accounting for 15% of the total area of the study area.
Control development zone
The control development zone is mainly cultivated land and construction land. The ecological environment carrying capacity is high, and this area is mainly set up to meet the needs of ecotourism services. The spatial distribution of this area is relatively scattered, concentrated primarily on the intersection zone between the ecological control zone and other areas. Reasonable development can be carried out without damaging the original ecological environment. The controlled development zone is 352 km2, accounting for 1% of the total area of the research area.
link