Ecosystem Services Supply–Demand Matching and Its Driving Factors: A Case Study of the Shanxi Section of the Yellow River Basin, China

: Understanding the supply–demand relationships and driving mechanisms of ecosystem services (ES) provides a theoretical foundation for sustainable ecosystem management. This study utilized Integrated Valuation of Ecosystem Services and Tradeoffs (InVEST) models and geographical detectors to quantify the spatial–temporal patterns of the supply, demand, and supply–demand ratio of ESs such as water yield, soil conservation, and carbon sequestration, along with their driving factors, in the Shanxi section of the Yellow River Basin. The results show that: (1) From the year 2000 to 2020, although the supply and demand of water yield, soil conservation, and carbon sequestration ﬂuctuated, they generally increased during this period of time. In comparison to ecosystem services from the year 2000 to 2020, the supply of water yield exceeded the demand in 2020. The supply, demand, and supply–demand ratio of ESs exhibited notable spatial heterogeneity. (2) The most notable factors inﬂuencing the supply–demand ratio of water yield varied between 2000 and 2020. In 2000, construction land was the most important factor, while in 2020, cropland had the greatest impact. However, the primary factors affecting the supply–demand ratio of soil conservation and carbon sequestration remained the same in 2000 and 2020. Forestland was the primary factor in 2000, while construction land was the primary factor in 2020. (3) Considering interaction factors, the interaction factors between construction land and precipitation had the greatest impact on the supply–demand ratio of water yield in 2000, while the interaction between forestland and cropland had the greatest impact in 2020. The interaction between cropland and shrubland had the greatest impact on the supply–demand ratio of soil conservation in 2000, whereas the interaction factors between construction land and forestland had the greatest impact in 2020. The interaction between construction land and shrubland had the greatest impact on the supply–demand ratio of carbon sequestration in 2000, while the interaction between construction land and cropland had the greatest impact in 2020. Overall, the interaction between construction land and various land-use factors had the strongest explanation for the supply–demand ratio of ecosystem services. This study can serve as a reference for the comprehensive development and utilization of the Shanxi section of the Yellow River Basin.


Introduction
Ecosystem services refer to the various benefits, including provisioning, regulating, cultural, and supporting services, that humans directly or indirectly obtain from ecosystems [1][2][3][4].Ecosystem service supply refers to the ability of an ecosystem to provide particular goods and services under specific spatial and temporal conditions.Ecosystem service demand refers to the total amount of ecological goods and services that humans Sustainability 2023, 15, 11016 2 of 18 use or consume within a particular area or time frame [5].The Millennium Ecosystem Assessment (MEA) revealed that the loss of ecosystems in recent decades poses a direct threat to both regional and global ecological security [6].However, the demand for ecosystem services in human society is increasing [7], resulting in an imbalance between the supply and demand of ecosystem services.Clarifying the relationship between the demand and supply of ecosystem services, as well as identifying the driving factors, is of great significance for maintaining the coordinated development of the social economy and the ecosystem.
The Yellow River Basin serves as an essential ecological barrier and economic development region, providing crucial ecosystem services [8].Numerous studies conducted by both domestic and international scholars have focused on the ecosystem services in the Yellow River Basin [9][10][11][12][13].Yang et al. utilized scenario analysis to clarify the impacts of changes in precipitation and land-use cover on water yield in the Yellow River Basin [14].Zhai et al. evaluated carbon sequestration in the Yellow River Basin using the InVEST model and discussed the discrepancy between supply and demand [15].An et al. used soil-loss equations (USLE) to investigate soil conservation in the Yellow River Basin [16].Zhi et al. evaluated the supply and demand of ecosystem services and analyzed the matching relationship between supply and demand of ecosystem services in the Yellow River Basin [17].Hu et al. employed models such as GeoDetector to analyze the effects of climate, human activities, and vegetation cover on ecosystem services [18].Long et al. employed GeoDetector to analyze the factors that contribute to the spatial heterogeneity of ecosystem services in the Yellow River Basin [19].Previous studies have mostly focused on a single type of ecosystem service in their research, and have not adequately considered multiple ecosystem services [20].Similarly, there are many static studies that examine the spatial and quantitative relationships between supply and demand of ecosystem services at a specific point in time.However, there is a lack of research on the temporal and spatial dynamics between supply and demand over time.
The Shanxi section of the Yellow River Basin is a fragile ecologic system, facing severe issues of soil erosion and energy pollution [21].The Grain-for-Green Program was implemented in 1999 with the aim of protecting the ecological environment [22,23].Wu et al. have pointed out that the implementation of the Grain-for-Green Program has increased vegetation cover, resulting in an increase in soil conservation and carbon sequestration [24].However, some scholars have argued that the Grain-for-Green Program may have negative effects.Increasing vegetation cover could enhance evapotranspiration, which may lead to a decrease in regional water yield [25].Therefore, studying of the ecosystem services in the Yellow River Basin is of great importance for regional ecological protection and sustainable development.This paper aimed to investigate the supply-demand ratios of ecosystem services from the year 2000 to 2020, utilizing the InVEST model and Geographic Detectors.The study aimed to answer the following questions: (1) What were the spatial and temporal distribution patterns of the amount of supply and demand, as well as the supply-demand ratios, for water yield, soil conservation, and carbon sequestration from the year 2000 to 2020? (2) What are the spatial and temporal matching characteristics of the supply and demand relationships of ecosystem services?(3) What are the mechanisms that underlie the supply and demand relationships of ecosystem services?

Overview of the Study Area
The Shanxi section of the Yellow River Basin is located between 34 • 33 and 40 • 19 N and 110 • 12 and 113 • 38 E, covering an area of approximately 98,500 km 2 in China.The area accounts for 12.3% of the entire Yellow River Basin.The Shanxi section of the Yellow River Basin consists mainly of the Loess Plateau Gully Area, the Rocky Mountainous Area, and the Plain Region (Figure 1).The Loess Plateau Gully Area is divided into two subareas, B1 and B2, with the Weifen River serving as the boundary between them.The study area has a warm temperate continental monsoon climate.The Shanxi section of the Yellow River Basin experiences low annual precipitation, making it mostly suitable for rainfed agriculture.The region is dry during the winter and receives more precipitation and higher temperatures during the summer.Cropland and grassland cover the largest area, followed by forestland and shrubland.Due to rapid economic development, carbon dioxide emissions and soil erosion have become increasingly severe in this region, making it a highly fragile ecological environment area.This area has become a site of intensified conflict between human society and the natural environment.areas, B1 and B2, with the Weifen River serving as the boundary between them.The study area has a warm temperate continental monsoon climate.The Shanxi section of the Yellow River Basin experiences low annual precipitation, making it mostly suitable for rainfed agriculture.The region is dry during the winter and receives more precipitation and higher temperatures during the summer.Cropland and grassland cover the largest area, followed by forestland and shrubland.Due to rapid economic development, carbon dioxide emissions and soil erosion have become increasingly severe in this region, making it a highly fragile ecological environment area.This area has become a site of intensified conflict between human society and the natural environment.

Data Sources
This study utilized land use/cover, meteorological, soil, and statistical datasets, as well as relevant auxiliary datasets, to assess the ecosystem services listed in Table 1.

Data Sources
This study utilized land use/cover, meteorological, soil, and statistical datasets, as well as relevant auxiliary datasets, to assess the ecosystem services listed in Table 1.

Research Methods
Given the water shortage and soil erosion in the Shanxi section of the Yellow River Basin, this study focused on three crucial ecosystem services: water yield (WY), soil conservation (SC), and carbon sequestration (CS).The InVEST model was used for the evaluation.This model has several advantages, including a small number of parameters, low data requirements, and widespread use in evaluating ecosystem services.

Water Yield
The supply for water yield was calculated through the water-yield module of the InVEST model.This model is based on the principle of water balance and assesses the regional supply of water yield under the influence of various natural factors.The difference between precipitation and actual evapotranspiration in a grid is the supply, which is calculated as follows: where Y sum is the annual supply of water yield (mm), AET i is the average annual actual evapotranspiration (mm), and P i is the average annual precipitation (mm).The demand of water yield is the amount of industrial, agricultural, domestic, and ecological water consumption.The water consumption in the Shanxi Province Water Resources Bulletin was categorized by land-use types to estimate the water-yield demand.ArcGIS was used to extract each land-use type and allocate the industrial water use to industrial and mining land, agricultural water use to cropland, domestic water use to residential land, and ecological water use to forestland, grassland, and water areas.
where D wat denotes the total demand for water supply services, while W dom , W agr , W ind , and W eco represent domestic water, agricultural irrigation water, industrial production water, and ecological water demand, respectively.

Soil Conservation
The supply of soil conservation is calculated by the SDR module in the InVEST model.The specific calculation equation is as follows: where AER x is the soil conservation in raster x, RKLS x is the potential soil erosion in raster x, USLE x is the actual soil erosion in grid x, SDR x is the sediment transport ratio, T x is the amount of sediment run-off from the upper slope intercepted by raster x, R is the precipitation erosivity index, K is the soil erodibility, LS is the slope length-slope index, C is the vegetation cover index, and P is the soil conservation index.The demand for soil conservation is calculated by the SDR module in the InVEST model.The soil conservation demand refers to the actual amount of soil loss that humans expect to be able to replenish and is calculated as follows: The parameters in the formula are defined above.

Carbon Sequestration
The supply of carbon sequestration is based on the land-use data and carbon density of each category (Table 2), using ArcGIS 10.8 to calculate the carbon stock.ArcGIS processed and analyzed the data by utilizing image elements.The software's raster calculator was used to multiply the population density and carbon emissions per capita for each year, resulting in the calculation of the demand for carbon sequestration in 2000, 2005, 2010, 2015, and 2020.The formula is as follows: (7) where C tot is the carbon sequestration, C above is the aboveground biogenic carbon stock, C below is the belowground biogenic carbon stock, C soil is the soil carbon stock, and C dead is the dead organic carbon stock.Using carbon emissions as an indicator, the carbon emissions of coal, oil, natural gas, and other energy sources were calculated by the carbon emission index of IPCC for raw coal, which is the amount of demand for carbon sequestration.This was spatialized using population raster data to plot the carbon emission raster data per capita, and the calculation formula is as follows: where D cx is the demand for carbon sequestration in grid x, D percarbon is the carbon emissions per capita, D sum is the total carbon emissions in each municipality, and P is the residential population in each municipality.

Supply-Demand Ratio for Ecosystem Services
The supply-demand ratio is an indicator used to measure the match between the supply and demand of ecosystem services and has the following formula: where ESDR is the ecological supply-demand ratio, S is the supply of ecosystem services, D indicates the demand for ecosystem services, S max indicates the maximum value of ecosystem service supply, D max indicates the maximum value of ecosystem service demand, and ESDR > 0 indicates that the supply of ecosystem services exceeds the demand.An ESDR value equal to 0 indicates that the supply and demand are in balance, and an ESDR value of less than 0 indicates that the supply does not meet the demand.

GeoDetector Models
GeoDetector models, which are statistical methods, are key factors in detecting spatial differentiation and have been widely used in recent infectious diseases and geological research [26][27][28].GeoDetector models include four detection methods: spatial-variability and factor detection, interaction detection, ecological detection, and risk-area detection.
Sustainability 2023, 15, 11016 where h = 1, 2, . . ., h, L represents the stratification of variable Y or Factor X, N h and N represent the number of cells in Stratum h and the whole region, respectively, σ 2 h and σ 2 represent variances of the Y values in Layer h and the entire region, respectively, and SSW and SST are the sum of the variances within the layer and the total variance of the entire region, respectively.q has a value range of [0, 1], which indicates the degree to which it explains the spatial heterogeneity of the dependent variable Y by the independent variable X.The greater its value, the stronger its explanation.
In this paper, we used factor detection and interaction detection to synthesize the driving mechanism of the supply and demand ratio of ecosystem services.The forestland ratio, grassland ratio, shrubland ratio, cropland ratio, construction-land ratio, bare-land ratio, NDVI, evapotranspiration, precipitation, elevation, GDP, population density, per capita consumption, and urbanization rate were selected as detection factors., the demand of water yield in the Shanxi section of the Yellow River Basin was 36.72 mm, 38.17 mm, 46.74 mm, 52.59 mm, and 52.39 mm, respectively.Overall, the demand for water yield in this area exhibited an increasing trend, with a 42.7% increase during the specified period.Furthermore, there was a notable variation in demand in the Plain Region when compared to other regions during the same time frame (Figure 2).

SST = N (13)
where h = 1, 2,..., h, L represents the stratification of variable Y or Factor X, N and N represent the number of cells in Stratum h and the whole region, respectively, σ and σ represent variances of the Y values in Layer h and the entire region, respectively, and SSW and SST are the sum of the variances within the layer and the total variance of the entire region, respectively.q has a value range of [0, 1], which indicates the degree to which it explains the spatial heterogeneity of the dependent variable Y by the independent variable X.The greater its value, the stronger its explanation.
In this paper, we used factor detection and interaction detection to synthesize the driving mechanism of the supply and demand ratio of ecosystem services.The forestland ratio, grassland ratio, shrubland ratio, cropland ratio, construction-land ratio, bare-land ratio, NDVI, evapotranspiration, precipitation, elevation, GDP, population density, per capita consumption, and urbanization rate were selected as detection factors.

Spatial and Temporal Patterns of the Water-Yield Supply and Demand
The supply of water yield in the Shanxi section of the Yellow River Basin was 36.42 mm, 35.83 mm, 42.88 mm, 42.27 mm, and 73.28 mm in 2000, 2005, 2010, 2015, and 2020, respectively.Although the supply of water yield fluctuated, it increased by 101% over a 20-year period starting from the year 2000.The supply of water yield fluctuated sharply in the Loess Plateau Gully Area and Plain Region.In 2000, 2005, 2010, 2015, and 2020, the demand of water yield in the Shanxi section of the Yellow River Basin was 36.72 mm, 38.17 mm, 46.74 mm, 52.59 mm, and 52.39 mm, respectively.Overall, the demand for water yield in this area exhibited an increasing trend, with a 42.7% increase during the specified period.Furthermore, there was a notable variation in demand in the Plain Region when compared to other regions during the same time frame (Figure 2).In terms of spatial scale, there were notable variations in different sub-catchments.The low-supply and low-demand areas for water yield were located in the Loess Plateau Gully Area.On the contrary, the high-supply and high-demand areas for water yield were distributed in the Rocky Mountainous Area and Plain Region.Notably, in this section, the high-supply areas of water yield continuously increased, while the low-supply areas decreased.The demand for water yield increased in the original low-value areas, while there was no notable change in the high-value areas (Figures 3 and 4).In terms of spatial scale, there were notable variations in different sub-catchments.The low-supply and low-demand areas for water yield were located in the Loess Plateau Gully Area.On the contrary, the high-supply and high-demand areas for water yield were distributed in the Rocky Mountainous Area and Plain Region.Notably, in this section, the high-supply areas of water yield continuously increased, while the low-supply areas decreased.The demand for water yield increased in the original low-value areas, while there was no notable change in the high-value areas (Figures 3 and 4).

Spatial and Temporal Patterns of Soil Conservation Supply and Demand
The supply of soil conservation in the Shanxi section of the Yellow River Basin was 125.09 t/ha, 121.63 t/ha, 158.42 t/ha, 153.08 t/ha, and 197.57t/ha in 2000, 2005, 2010, 2015, and 2020, respectively.Overall, the supply of soil conservation fluctuated but increased by 57.9% during this period.Furthermore, the demand for soil conservation was 134.54 t/ha, 130.90 t/ha, 170.22 t/ha, 164.68 t/ha, and 212.02 t/ha in 2000, 2005, 2010, 2015, and 2020, respectively (Figure 5).The demand for soil conservation fluctuated but ultimately increased by 57.6%.The fluctuations in the supply and demand of soil conservation were evident in the Loess Plateau Gully Area during this period.

Spatial and Temporal Patterns of Soil Conservation Supply and Demand
The supply of soil conservation in the Shanxi section of the Yellow River Basin was 125.09 t/ha, 121.63 t/ha, 158.42 t/ha, 153.08 t/ha, and 197.57t/ha in 2000, 2005, 2010, 2015, and 2020, respectively.Overall, the supply of soil conservation fluctuated but increased by 57.9% during this period.Furthermore, the demand for soil conservation was 134.54 t/ha, 130.90 t/ha, 170.22 t/ha, 164.68 t/ha, and 212.02 t/ha in 2000, 2005, 2010, 2015, and 2020, respectively (Figure 5).The demand for soil conservation fluctuated but ultimately increased by 57.6%.The fluctuations in the supply and demand of soil conservation were evident in the Loess Plateau Gully Area during this period.In terms of spatial distribution, the high-supply areas for the soil conservation were located in the Loess Plateau Gully Area, which is predominantly mountainous.There were two main reasons for the high-supply areas of soil conservation in this area.Firstly, the slopes of the hills are steep, leading to a high potential erosion in this area.Secondly, due to the dense vegetation coverage, the amount of actual erosion was minimal.The notable disparity between potential and actual erosion has led to a high supply for soil conservation.As for demand, the low-demand area for soil conservation was distributed in the Plain Region.This region is characterized by gentle slopes and sparse vegetation cover, resulting in a relatively low demand for soil conservation.From the year 2000 to 2020, the high-demand area of soil conservation expanded, while the low-value area contracted.The demand for soil conservation exhibited a spatial distribution, with low values in the northeast and high values in the southwest.The Loess Plateau Gully Area had low demand for soil conservation, while the Plain Region had high demand (Figures 3 and 4).

Spatial and Temporal Patterns of Carbon Sequestration Supply and Demand
The supply of carbon sequestration in the Shanxi section of the Yellow River Basin was 0.406 tC/ha, 0.407 tC/ha, 0.407 tC/ha, 0.407 tC/ha, and 0.403 tC/ha in 2000, 2005, 2010, 2015, and 2020, respectively.The supply of carbon sequestration fluctuated but decreased overall during this period.However, the change in the supply was not evident in every sub-catchment.The demand for carbon sequestration in the Shanxi section of the Yellow River Basin was 4.74 tC/ha, 8.47 tC/ha, 10.63 tC/ha, 11.13 tC/ha, and 13.14 tC/ha from the year 2000 to 2020.The demand of carbon sequestration showed steady growth, increasing from 4.74 tC/ha in 2000 to 13.14 tC/ha in 2020.Furthermore, the growth rate was approximately 177% (Figure 6).The change was more pronounced in the Plain Region compared to other regions.In terms of the spatial distribution, the high-supply areas of carbon sequestration were mainly located in the Rocky Mountainous Area and the Loess Plateau Gully Area.Forestland and grassland in these regions absorb and store carbon dioxide (CO2), resulting In terms of spatial distribution, the high-supply areas for the soil conservation were located in the Loess Plateau Gully Area, which is predominantly mountainous.There were two main reasons for the high-supply areas of soil conservation in this area.Firstly, the slopes of the hills are steep, leading to a high potential erosion in this area.Secondly, due to the dense vegetation coverage, the amount of actual erosion was minimal.The notable disparity between potential and actual erosion has led to a high supply for soil conservation.As for demand, the low-demand area for soil conservation was distributed in the Plain Region.This region is characterized by gentle slopes and sparse vegetation cover, resulting in a relatively low demand for soil conservation.From the year 2000 to 2020, the high-demand area of soil conservation expanded, while the low-value area contracted.The demand for soil conservation exhibited a spatial distribution, with low values in the northeast and high values in the southwest.The Loess Plateau Gully Area had low demand for soil conservation, while the Plain Region had high demand (Figures 3 and 4).

Spatial and Temporal Patterns of Carbon Sequestration Supply and Demand
The supply of carbon sequestration in the Shanxi section of the Yellow River Basin was 0.406 tC/ha, 0.407 tC/ha, 0.407 tC/ha, 0.407 tC/ha, and 0.403 tC/ha in 2000, 2005, 2010, 2015, and 2020, respectively.The supply of carbon sequestration fluctuated but decreased overall during this period.However, the change in the supply was not evident in every sub-catchment.The demand for carbon sequestration in the Shanxi section of the Yellow River Basin was 4.74 tC/ha, 8.47 tC/ha, 10.63 tC/ha, 11.13 tC/ha, and 13.14 tC/ha from the year 2000 to 2020.The demand of carbon sequestration showed steady growth, increasing from 4.74 tC/ha in 2000 to 13.14 tC/ha in 2020.Furthermore, the growth rate was approximately 177% (Figure 6).The change was more pronounced in the Plain Region compared to other regions.In terms of spatial distribution, the high-supply areas for the soil conservation were located in the Loess Plateau Gully Area, which is predominantly mountainous.There were two main reasons for the high-supply areas of soil conservation in this area.Firstly, the slopes of the hills are steep, leading to a high potential erosion in this area.Secondly, due to the dense vegetation coverage, the amount of actual erosion was minimal.The notable disparity between potential and actual erosion has led to a high supply for soil conservation.As for demand, the low-demand area for soil conservation was distributed in the Plain Region.This region is characterized by gentle slopes and sparse vegetation cover, resulting in a relatively low demand for soil conservation.From the year 2000 to 2020, the high-demand area of soil conservation expanded, while the low-value area contracted.The demand for soil conservation exhibited a spatial distribution, with low values in the northeast and high values in the southwest.The Loess Plateau Gully Area had low demand for soil conservation, while the Plain Region had high demand (Figures 3 and 4).

Spatial and Temporal Patterns of Carbon Sequestration Supply and Demand
The supply of carbon sequestration in the Shanxi section of the Yellow River Basin was 0.406 tC/ha, 0.407 tC/ha, 0.407 tC/ha, 0.407 tC/ha, and 0.403 tC/ha in 2000, 2005, 2010, 2015, and 2020, respectively.The supply of carbon sequestration fluctuated but decreased overall during this period.However, the change in the supply was not evident in every sub-catchment.The demand for carbon sequestration in the Shanxi section of the Yellow River Basin was 4.74 tC/ha, 8.47 tC/ha, 10.63 tC/ha, 11.13 tC/ha, and 13.14 tC/ha from the year 2000 to 2020.The demand of carbon sequestration showed steady growth, increasing from 4.74 tC/ha in 2000 to 13.14 tC/ha in 2020.Furthermore, the growth rate was approximately 177% (Figure 6).The change was more pronounced in the Plain Region compared to other regions.In terms of the spatial distribution, the high-supply areas of carbon sequestration were mainly located in the Rocky Mountainous Area and the Loess Plateau Gully Area.Forestland and grassland in these regions absorb and store carbon dioxide (CO2), resulting In terms of the spatial distribution, the high-supply areas of carbon sequestration were mainly located in the Rocky Mountainous Area and the Loess Plateau Gully Area.Forestland and grassland in these regions absorb and store carbon dioxide (CO 2 ), resulting in a large carbon sequestration supply.In contrast, the low-supply areas of carbon sequestration were located in the Plain Region, which is primarily cropland with a limited capacity for storing CO 2 .Furthermore, the demand for carbon sequestration showed a spatial distribution, with lower values in the northwest and higher values in the southeast.The majority of high-demand areas for carbon sequestration were located in the Plain Region, due to the population increase from the year 2000 to 2020 and the rapid development of the secondary and tertiary industries in the Plain Region.The demand for carbon sequestration in the Shanxi section of the Yellow River Basin had also notably increased (Figures 3 and 4).
Overall, the supply and demand of water yield, soil conservation, and carbon sequestration fluctuated and increased over time.The high-supply areas for water yield and the low-supply areas for soil conservation and carbon sequestration were located in the Plain Region, while the low-supply areas for water yield and the high-supply areas for soil conservation and carbon sequestration were located in the Loess Plateau Gully Area.The high-demand areas for water yield, soil conservation, and carbon sequestration were located in the Plain Region, whereas the low-value areas were distributed in the Loess Plateau Gully Area.In 2000, 2005, 2010, 2015, and 2020, the supply-demand ratios of water yield were −0.0004, −0.0034, −0.0039, −0.0111, and 0.0238, respectively.Overall, the supply-demand ratio for water yield fluctuated but ultimately increased over time.In 2020, there was an oversupply of water yield and the supply-demand ratio reached its highest value.The supply for water yield far exceeded the local production and living needs.From 2000 to 2015, the supply for water yield in the Yellow River Basin was insufficient to meet the daily production and living needs of the local population (Figure 7).During this period, notable changes occurred in the Loess Plateau Gully Area and Rocky Mountainous Area.
Sustainability 2023, 15, x FOR PEER REVIEW 10 of 18 in a large carbon sequestration supply.In contrast, the low-supply areas of carbon sequestration were located in the Plain Region, which is primarily cropland with a limited capacity for storing CO2.Furthermore, the demand for carbon sequestration showed a spatial distribution, with lower values in the northwest and higher values in the southeast.The majority of high-demand areas for carbon sequestration were located in the Plain Region, due to the population increase from the year 2000 to 2020 and the rapid development of the secondary and tertiary industries in the Plain Region.The demand for carbon sequestration in the Shanxi section of the Yellow River Basin had also notably increased (Figures 3 and 4).Overall, the supply and demand of water yield, soil conservation, and carbon sequestration fluctuated and increased over time.The high-supply areas for water yield and the low-supply areas for soil conservation and carbon sequestration were located in the Plain Region, while the low-supply areas for water yield and the high-supply areas for soil conservation and carbon sequestration were located in the Loess Plateau Gully Area.The high-demand areas for water yield, soil conservation, and carbon sequestration were located in the Plain Region, whereas the low-value areas were distributed in the Loess Plateau Gully Area.

Spatial and Temporal Patterns of the Supply and Demand Ratio for Water Yield
In 2000, 2005, 2010, 2015, and 2020, the supply-demand ratios of water yield were −0.0004, −0.0034, −0.0039, −0.0111, and 0.0238, respectively.Overall, the supply-demand ratio for water yield fluctuated but ultimately increased over time.In 2020, there was an oversupply of water yield and the supply-demand ratio reached its highest value.The supply for water yield far exceeded the local production and living needs.From 2000 to 2015, the supply for water yield in the Yellow River Basin was insufficient to meet the daily production and living needs of the local population (Figure 7).During this period, notable changes occurred in the Loess Plateau Gully Area and Rocky Mountainous Area.There were notable differences in the matching of supply and demand of the water yield in various sub-catchments.From the year 2000 to 2020, the supply-demand ratio of water yield in the Loess Plateau Gully Area remained above 0, indicating a surplus of water yield.From the year 2000 to 2015, the supply-demand ratio in the Plain Region was less than 0, but it had improved by 2020.When precipitation was abundant, the supply of water yield could meet the local demand for water, but when precipitation was insufficient, it threatened regional water security (Figure 8).There were notable differences in the matching of supply and demand of the water yield in various sub-catchments.From the year 2000 to 2020, the supply-demand ratio of water yield in the Loess Plateau Gully Area remained above 0, indicating a surplus of water yield.From the year 2000 to 2015, the supply-demand ratio in the Plain Region was less than 0, but it had improved by 2020.When precipitation was abundant, the supply of water yield could meet the local demand for water, but when precipitation was insufficient, it threatened regional water security (Figure 8).

Spatial and Temporal Patterns of the Supply and Demand Ratio of Soil Conservation
The supply-demand ratio for soil conservation was −0.0073, −0.0071, −0.0085, −0.0079, and −0.0077 in 2000, 2005, 2010, 2015, and 2020, respectively.The supply for soil conservation was insufficient to meet the demand, and the supply-demand ratio fluctuated and decreased, with notable changes in the Loess Plateau Gully Area (Figure 9).

Spatial and Temporal Patterns of the Supply and Demand Ratio of Soil Conservation
The supply-demand ratio for soil conservation was −0.0073, −0.0071, −0.0085, −0.0079, and −0.0077 in 2000, 2005, 2010, 2015, and 2020, respectively.The supply for soil conservation was insufficient to meet the demand, and the supply-demand ratio fluctuated and decreased, with notable changes in the Loess Plateau Gully Area (Figure 9).From a spatial perspective, the regions with high supply-demand ratios were concentrated in the Loess Plateau Gully Area.The area with a supply-demand ratio greater than −0.005 increased from 2000 to 2005, primarily due to the implementation of the Grainfor-Green Program.Before 1999, the forestland and grassland were drastically reduced due to unreasonable human exploitation practices.These practices caused serious damage to the environment and resulted in severe soil erosion.After implementing the program, afforestation was carried out based on local conditions to restore forest vegetation.This effort successfully alleviated the soil erosion in the Loess Plateau Gully Area, improving the ecological environment of the Yellow River Basin (Figure 8).

Spatial and Temporal Patterns of the Supply-Demand Ratio for Carbon Sequestration
The supply-demand ratio for carbon sequestration was −0.0568, −0.0518, −0.0387, −0.0385, and −0.0377 in 2000, 2005, 2010, 2015, and 2020, respectively.The supply-demand ratio of carbon sequestration in the Shanxi section of the Yellow River Basin had been in deficit, with the supply being lower than the demand.This deficit was more notable in the Plain Region than in other regions (Figure 10).From a spatial perspective, the regions with high supply-demand ratios were concentrated in the Loess Plateau Gully Area.The area with a supply-demand ratio greater than −0.005 increased from 2000 to 2005, primarily due to the implementation of the Grainfor-Green Program.Before 1999, the forestland and grassland were drastically reduced due to unreasonable human exploitation practices.These practices caused serious damage to the environment and resulted in severe soil erosion.After implementing the program, afforestation was carried out based on local conditions to restore forest vegetation.This effort successfully alleviated the soil erosion in the Loess Plateau Gully Area, improving the ecological environment of the Yellow River Basin (Figure 8).

Spatial and Temporal Patterns of the Supply-Demand Ratio for Carbon Sequestration
The supply-demand ratio for carbon sequestration was −0.0568, −0.0518, −0.0387, −0.0385, and −0.0377 in 2000, 2005, 2010, 2015, and 2020, respectively.The supply-demand ratio of carbon sequestration in the Shanxi section of the Yellow River Basin had been in deficit, with the supply being lower than the demand.This deficit was more notable in the Plain Region than in other regions (Figure 10).From a spatial perspective, the regions with high supply-demand ratios were concentrated in the Loess Plateau Gully Area.The area with a supply-demand ratio greater than −0.005 increased from 2000 to 2005, primarily due to the implementation of the Grainfor-Green Program.Before 1999, the forestland and grassland were drastically reduced due to unreasonable human exploitation practices.These practices caused serious damage to the environment and resulted in severe soil erosion.After implementing the program, afforestation was carried out based on local conditions to restore forest vegetation.This effort successfully alleviated the soil erosion in the Loess Plateau Gully Area, improving the ecological environment of the Yellow River Basin (Figure 8).

Spatial and Temporal Patterns of the Supply-Demand Ratio for Carbon Sequestration
The supply-demand ratio for carbon sequestration was −0.0568, −0.0518, −0.0387, −0.0385, and −0.0377 in 2000, 2005, 2010, 2015, and 2020, respectively.The supply-demand ratio of carbon sequestration in the Shanxi section of the Yellow River Basin had been in deficit, with the supply being lower than the demand.This deficit was more notable in the Plain Region than in other regions (Figure 10).The supply-demand ratio of carbon sequestration in the Yellow River Basin showed low values below −0.2 in the southeast and high values in the northwest.The low-value areas were located in the Plain Region.Due to human activities related to production, large-scale agricultural development, and increased industrialization, there was a growing demand for carbon sequestration.At the same time, the vegetation coverage was low, resulting in a decrease in the supply of carbon sequestration.Due to this, the demand exceeded the supply, resulting in a notable deficit in carbon sequestration.The supplydemand ratio in the Plain Region improved from the year 2005 to 2010, and the supply of carbon sequestration increased due to afforestation and forest conservation.As a result, the deficit situation has improved in recent years (Figure 8).The GeoDetector was utilized to obtain the driving factors of the demand-supply ratio of water yield in 2000 and 2020 (Table 3).In 2000, the demand-supply ratio of water yield was primarily influenced by construction land, forestland, bare land, GDP, and urbanization rate.Socioeconomic factors played a dominant role.In 2020, the demand-supply ratio for water yield was influenced by several factors, including cropland, precipitation, per capita consumption, urbanization rate, and grassland.Natural factors played a dominant role in the supply-demand ratio for water yield.It was evident that the impact of precipitation on the supply-demand ratio of water yield was increasing.However, population growth, as well as industrial and agricultural development, has led to the notable consumption of water resources, resulting in increased ecological pressure in the Yellow River Basin.
Forestland, shrubland, cropland, and construction land had a notable impact on the supply-demand ratio for soil conservation in 2000 and 2020.Land use was the primary factor contributing to spatial heterogeneity in the supply-demand ratio.The impact of socioeconomic factors on the supply-demand ratio remained unchanged (Table 4).However, natural factors had a notable effect on the supply-demand ratio for soil conservation.The supply-demand ratio of carbon sequestration was primarily influenced by construction land, GDP, and population density in 2000 and 2020 (Table 5).Socioeconomic factors provided the greatest explanation for the supply-demand ratio for carbon sequestration.This was due to population growth and rapid economic development.Another reason for this is that the energy, industrial, and mining industries release a notable amount of carbon dioxide, and the supply of carbon sequestration cannot offset the carbon emissions.The results of the interaction detector indicate that environmental factors have varied degrees of influence on the supply-demand ratio of ecosystem services in the Yellow River Basin.A higher q value for interaction indicates a greater impact of the interaction between the two factors on the supply-demand ratio of ecosystem services.
The primary factors affecting the supply-demand ratio of water yield in 2000 was construction land and precipitation.The q-value of the interaction between construction land and other factors was high (q > 0.5), indicating that construction land was a strong explanation as a single factor.This also indicated that the expansion of construction land has resulted in a higher demand for water, especially for industrial and agricultural development.This explained the highest supply-demand ratio for water yield.The interaction factors that had the greatest influence on the supply-demand ratio of water yield in 2020 were forestland and cropland.This implies that the type of land use had a greater impact on the supply for water yield.For instance, an increase in the forestland can lead to a rise in evapotranspiration, which can cause a reduction in supply of water yield (Table 6).Over the 20-year period, the driving factors had shifted from the demand side to the supply side.This shift is mainly attributed to an increase in the supply of water yield, which was caused by a rise in precipitation in 2020.The dominant interaction factors affecting the supply-demand ratio of soil conservation in 2000 were cropland and shrubland (Table 7).However, in 2020, the dominant interaction factors affecting the supply-demand ratio for soil conservation were construction land and forestland.This was related to the greater effect of the forestland as a single factor.The grassland, shrubland, and forestland all had a positive effect on soil conservation, improving the supply for soil conservation during the 20-year period.The construction land led to more severe soil erosion, which in turn affected the supply-demand ratio of soil conservation.
In 2000, the main influencing interaction factors affecting the supply-demand ratio of carbon sequestration were construction land and shrubland.However, in 2020, the dominant interaction factors that influenced the ratio were construction land and cropland.The q values for the interaction between construction land and all other factors were high (q > 0.9) (Table 8), which can be attributed to the greater effect of construction land as a single factor.This indicated that the expansion of construction land and GDP growth have resulted in a rise in carbon emissions.Human economic activities, such as urbanization, rapid industrial development, mining, and construction of transportation routes, have notably altered the land use.These activities have had the greatest impact on the supplydemand ratio of carbon sequestration.Shrubland, as carbon sinks, absorb CO 2 and fix it in the soil.This increases the supply of carbon sequestration, which in turn affects the supply-demand ratio of carbon sequestration.We observed fluctuations and increases in the supply, demand, and supply-demand ratio of water yield in the Yellow River Basin from the year 2000 to 2020.The research conducted by Wang supports this conclusion [29].This study found that from the year 2000 to 2020, there was an increase in the supply, demand, and supply-demand ratio of soil conservation.The results of Zhao are consistent with the findings of this study [30].From the year 2000 to 2020, the supply, demand, and supply-demand ratio of carbon sequestration services increased.The research conducted by Zhao and Fu is consistent with the findings presented in this study [31,32].

Driving Factors of the Spatial Distribution of Ecosystem Services
Zhang believes that an increase in precipitation will lead to an increase in supply of water yield [33].The demand for water yield has been increasing year by year due to population growth, agricultural irrigation, and industrial production.This finding is consistent with the results of the research.This study found that forestland plays a crucial role in the supply-demand ratio of soil conservation.Wang et al. believe that the high vegetation coverage in the Shanxi section of the Loess Plateau helps to prevent soil erosion, thereby promoting a balance between the supply and demand of soil conservation [34].This conclusion is consistent with his findings.Moreover, the Yellow River Basin is located in the Loess Plateau, which is known for its loose soil and seasonal precipitation that carries a substantial amount of sediment.With the rapid development of urbanization and population growth, the area of land used for construction has increased, leading to serious problems of water and soil erosion.This study found that the availability of construction land has a notable impact on the supply-demand ratio of carbon sequestration services.Fu argues that increasing the amount of land used for construction will lead to a notable reduction in the supply-demand ratio of carbon sequestration services.

Recommendations for Land Use Management
In regions with low supply of water yield, such as the Plain Region, it is crucial to protect forestlands and grasslands to maintain sustainable ecosystem development.In addition, accelerating the transition of the industrial structure towards the tertiary sector, optimizing the industrial layout, and adopting new irrigation techniques such as drip and spray irrigation in cultivation can effectively address the issue of water resource scarcity.Encouraging the recycling of water resources in residential households can promote sustainable water usage.Grasslands and forestlands play a notable role in improving soil retention capacity.In the future, we should continue to consolidate the achievements of the Grain-for-Green Program.These projects promote the restoration of vegetation, prevent soil erosion, maintain soil nutrients, improve soil quality, and enhance soil conservation.Forestlands are the largest contributors to carbon sequestration.Implementing afforestation plans is a crucial measure for carbon sequestration.These initiatives aim to improve the environmental quality and enhance the carbon storage of forests.To achieve the target of reaching carbon peak by 2030 and carbon neutrality by 2060, it is essential to boost investment in science and technology.Furthermore, we should enhance our carbon sequestration capacity by utilizing ecological engineering, biological carbon capture, ecological carbon storage, and the utilization of biological carbon.Furthermore, we must increase our efforts to reduce carbon emissions by optimizing industrial structures and improving transportation methods.

Limitations
This study evaluates three ecosystem services of the Yellow River Basin, focusing on water yield, soil conservation, and carbon sequestration.The reserve also provides various services, including food production, wind and sand stabilization, climate regulation, habitat preservation, and cultural recreation.Future research could conduct a comprehensive evaluation of the ecosystem in the Yellow River Basin.Such an evaluation is crucial for the integrated management of the basin and for improving the ecological environment in the Shanxi section of the Yellow River Basin.

Conclusions
From the year 2000 to 2020, the supply and demand of water yield in the Shanxi section of the Yellow River Basin showed continued growth, and the distribution of the supply and demand showed a pattern of low values in the northwest and high values in the southeast.The soil conservation supply and demand showed continued growth from the year 2000 to 2020.Specifically, the high-supply areas of soil conservation and carbon sequestration were concentrated in the Loess Plateau Gully Area, while the high-demand areas of soil conservation and carbon sequestration were located in the Plain Region.
The water-yield supply could basically meet the demands of local daily production and living needs during the 20-year period.The soil conservation supply and carbon sequestration could not meet the demand, but the supply-demand ratio showed an increasing trend each year, and the supply and demand were improving.
There was notable spatial heterogeneity in the explanation of each driver for the supply-demand ratio of ecosystem services.Overall, the interaction between construction land and other land use types had the highest explanation for the demand-supply ratio of ecosystem services, which verifies that the ecosystem service function is influenced by the interaction between natural factors and socioeconomic development factors.

Figure 1 .
Figure 1.Location of the study area in China.

Figure 1 .
Figure 1.Location of the study area in China.
Patterns of Supply and Demand of Ecosystem Services 3.1.Spatial and Temporal Patterns of the Water-Yield Supply and Demand The supply of water yield in the Shanxi section of the Yellow River Basin was 36.42 mm, 35.83 mm, 42.88 mm, 42.27 mm, and 73.28 mm in 2000, 2005, 2010, 2015, and 2020, respectively.Although the supply of water yield fluctuated, it increased by 101% over a 20-year period starting from the year 2000.The supply of water yield fluctuated sharply in the Loess Plateau Gully Area and Plain Region.In 2000, 2005, 2010, 2015, and 2020

Figure 2 .
Figure 2. Trends in water-yield supply and demand from the year 2000 to 2020.(a)The supply of water-yield.(b)The demand of water-yield.

Figure 2 .
Figure 2. Trends in water-yield supply and demand from the year 2000 to 2020.(a)The supply of water-yield.(b)The demand of water-yield.

Figure 3 .
Figure 3. Spatial pattern of ecosystem service supply from the year 2000 to 2020.Figure 3. Spatial pattern of ecosystem service supply from the year 2000 to 2020.

Figure 3 .
Figure 3. Spatial pattern of ecosystem service supply from the year 2000 to 2020.Figure 3. Spatial pattern of ecosystem service supply from the year 2000 to 2020.

Figure 4 .
Figure 4. Spatial pattern of ecosystem services demand from the year 2000 to 2020.

Figure 4 .
Figure 4. Spatial pattern of ecosystem services demand from the year 2000 to 2020.

Figure 5 .
Figure 5. Trend of soil conservation supply and demand from the year 2000 to 2020.(a) The supply of soil conservation.(b) The demand of soil conservation.

Figure 6 .
Figure 6.Trend of carbon sequestration supply and demand from the year 2000 to 2020.(a) The supply of carbon sequestration.(b) The demand of carbon sequestration.

Figure 5 .
Figure 5. Trend of soil conservation supply and demand from the year 2000 to 2020.(a) The supply of soil conservation.(b) The demand of soil conservation.

Figure 5 .
Figure 5. Trend of soil conservation supply and demand from the year 2000 to 2020.(a) The supply of soil conservation.(b) The demand of soil conservation.

Figure 6 .
Figure 6.Trend of carbon sequestration supply and demand from the year 2000 to 2020.(a) The supply of carbon sequestration.(b) The demand of carbon sequestration.

Figure 6 .
Figure 6.Trend of carbon sequestration supply and demand from the year 2000 to 2020.(a) The supply of carbon sequestration.(b) The demand of carbon sequestration.

3. 4 .
Spatial and Temporal Patterns of Supply-Demand Ratios for Ecosystem Services 3.4.1.Spatial and Temporal Patterns of the Supply and Demand Ratio for Water Yield

Figure 7 .
Figure 7. Trends in the supply-demand ratio for water yield from the year 2000 to 2020.

Figure 7 .
Figure 7. Trends in the supply-demand ratio for water yield from the year 2000 to 2020.

Figure 8 .
Figure 8. Spatial pattern of the supply-demand ratio for ecosystem services from the year 2000 to 2020.

Figure 8 .
Figure 8. Spatial pattern of the supply-demand ratio for ecosystem services from the year 2000 to 2020.

Figure 9 .
Figure 9. Trend of the supply-demand ratio for soil conservation from the year 2000 to 2020.

Figure 10 .
Figure 10.Trend of the supply-demand ratio for carbon sequestration from the year 2000 to 2020.The supply-demand ratio of carbon sequestration in the Yellow River Basin showed low values below −0.2 in the southeast and high values in the northwest.The low-value

Figure 9 .
Figure 9. Trend of the supply-demand ratio for soil conservation from the year 2000 to 2020.

Sustainability 2023 , 18 Figure 9 .
Figure 9. Trend of the supply-demand ratio for soil conservation from the year 2000 to 2020.

Figure 10 .Figure 10 .
Figure 10.Trend of the supply-demand ratio for carbon sequestration from the year 2000 to 2020.The supply-demand ratio of carbon sequestration in the Yellow River Basin showed low values below −0.2 in the southeast and high values in the northwest.The low-value Figure 10.Trend of the supply-demand ratio for carbon sequestration from the year 2000 to 2020.

Table 1 .
Summary of main data types and data sources.

Table 1 .
Summary of main data types and data sources.

Table 2 .
Carbon stock of each component in different land-use types (t/ha).

Table 3 .
Single-factor detection of water yield in 2000 and 2020.

Table 4 .
Single factor detection of soil conservation in 2000 and 2020.

Table 5 .
Single-factor detection of carbon sequestration in 2000 and 2020.

Table 6 .
Interaction factor detection for water yield in 2000 and 2020.

Table 7 .
Interaction factor detection of soil conservation in 2000 and 2020.

Table 8 .
Interaction factor detection of carbon sequestration in 2000 and 2020.