Dynamic Assessment and Change Analysis of Ecosystem Service Value Based on Physical Assessment Method in Cili County, China

Abstract

The effective implementation of ecological protection policies requires the adequate assessment of temporal and spatial changes in the environment. To understand how ecosystem services can be used to track environmental changes, we carried out a study which focused on assessing the changes in the ecosystem services over time and space in Cili County, which is an important forestry region. The method we used was to evaluate the value of the regional ecological services based on the method for evaluating the value equivalent factor per unit area; then, by introducing multi-source data, the InVEST model was optimized to accurately assess the function of the regional ecosystem services, such as carbon storage, soil conservation, and water production. In addition, the values of the ecosystem services and the function assessment results were compared. Finally, based on the results of the optimized InVEST model, the value of the ecosystem services in the study area was regulated and re-evaluated. After re-evaluation, the total value of the ecosystem services in Cili county between 2000 and 2020 still showed an upward trend, increasing from CNY 26.136 billion to CNY 35.444 billion, with a total increase of CNY 9.308 billion. Compared to before the re-evaluation, the total value of the ecosystem services only increased from CNY 32.243 billion to CNY 32.473 billion, with a total increase of CNY 0.23 billion; the change in the value of the ecosystem services was more obvious, with a stronger spatial heterogeneity. The areas with high ecosystem service value in Cili County are mainly concentrated in the eastern parts, as well as the northwestern and southern parts, while the areas with low value are mainly concentrated in the central part of Cili County. The value of the central, southern, and northwestern parts of Cili County increased significantly. Such changes are closely related to China’s implementation of ecological protection policies in this region since 2000, such as returning farmland to forest and natural forest protection. The evaluation results of the ecosystem services and the method for evaluating the value equivalent factor in this study are more consistent with the changes in the ecosystem services in the study area. The dynamic assessment method of ecosystem service value proposed in this study is helpful in achieving accurate assessments of the regional ecosystem services and thus provides a useful reference for the formulation of more reasonable regional ecological protection policies.

1. Introduction

Ecosystem services refer to the supporting products and services that human beings obtain, directly or indirectly, from the processes, structure, and function of the ecosystem. With the general recognition that “ecosystem service is the cornerstone of human existence”, more and more attention has been paid to the research of ecosystem services. The reasonable and effective assessment of ecosystem services is conducive to the scientific conservation of ecosystems and the appreciation of ecological assets and helps human beings to decide on more effective ecological protection policies for restoring degraded ecosystems. Performing these assessments is thus of great practical significance for the realization of sustainable development and the construction of an ecological civilization.

In view of the diversity of ecosystem services, the evaluation methods can be roughly divided into the physical assessment methods, the method for evaluating the value equivalent factor, and the energy evaluation methods [1]. At present, the physical assessment methods and the method for evaluating the value equivalent factor are the most important evaluation methods. The method for evaluating the value equivalent factor is mainly adopted to evaluate the value equivalent factor of an ecosystem service. Daily and Costanza et al. first conducted a systematic assessment of the service functions and values of different ecosystem types in 1997 [2,3]. Subsequently, with more and more ecosystem services, the method for evaluating the value equivalent factor has entered this field of research, and many scholars have carried out value assessments of various ecosystems [4,5,6,7,8,9,10,11]. On this basis, Xie Gaodi et al. constructed and improved the ecosystem service value equivalent method for mainland China from 2005 to 2015, and this method has been widely applied in China [12,13,14,15,16,17,18,19]; most of the current evaluation methods are based on their research methods. However, for global- and national-scale ecosystem services, when evaluating the value equivalent factor, the same value equivalent method is often applied to the same type of ecosystem in different time sequences. When applied to ecosystem service value assessments in smaller regions, the spatial–temporal heterogeneity is often difficult to reflect, and the assessment’s accuracy remains to be discussed. The physical assessment method usually adopts the ecosystem service function model to estimate the quantity of the product. As the most widely used model in the field of ecosystem service function assessment at present, the InVEST model has been used in a number of studies in China and other regions [20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38]. However, when the InVEST model is used to calculate the quality of some ecosystems, it also has the problem that the estimation parameters do not fully consider the spatial heterogeneity, which also affects the accuracy of the quality estimation to some extent. How to combine the two methods of ecosystem service assessment effectively to obtain an accurate assessment of ecosystem services is a topic that requires further research.

Before 2000, due to excessive deforestation, including that which was for farmland, serious soil and water loss was seen in many places in China; the ecological environment was deteriorating and natural disasters occurred frequently, causing huge loss of life and property [39]. In order to fundamentally curb the deterioration of the ecological environment, protect biodiversity, and promote sustainable social and economic development, China began to implement the natural forest protection project on a pilot basis in 1998. In 2002, China began to fully implement the conversion of farmland to forest. Cili County, located in the northwest of Hunan Province, is a typical forestry county with rich forest resources and a proportion of forest land making up more than half of its total area. In 2000, Cili County began to implement the natural forest protection project, the project of returning farmland to forest, and other policies for the purpose of protecting and restoring the local ecology. However, since the implementation of these policies, there is still a lack of detailed analysis of the temporal and spatial characteristics of the ecosystem services in Cili County.

Taking Cili County of Hunan Province as the study area, this paper discusses a reasonable ecosystem service assessment method, accurately evaluates the ecosystem services in the study area, and defines the characteristics of the temporal and spatial changes so as to provide a scientific basis for the conservation and restoration of the regional ecosystem. This study has a certain theoretical and practical significance for the realization of the stability and sustainability of the regional ecosystem.

2. Materials and Methods

2.1. Study Area

Cili County belongs to Zhangjiajie City, Hunan Province, China. The terrain of Cili County is high in the northwest and low in the southeast, with various geomorphological types, mainly mountains, mountain plains, and rivers, forming three mountains and two valleys, which account for 64% of the total area (Figure 1). The altitude is 75–1409.8 m. Cili County is located in the subtropical monsoon humid climate zone. The climate is warm and humid; the heat is abundant; the rainfall is rich; the light is abundant; the frost-free period is long; the cold period is short; and the four seasons change obviously. There are various landforms, among which woodland occupies the highest proportion in Cili County [40]. The main forest vegetation types in Cili County are evergreen broadleaved forest, deciduous broadleaved forest, mixed evergreen and deciduous broadleaved forest, evergreen coniferous forest, etc. Since the implementation of the natural forest protection project in Cili County in 2000, the planting of a public welfare forest in Cili County has been carried out in an orderly manner. In 2014, the area of the public welfare forest in Cili County reached 128,618.8 ha; the area of the key public welfare forest was 77,157.9 ha, and the area of the general public welfare forest was 51,458.9 ha [41].

2.2. Data Sources

The data used in this study include the land use and land cover data, remote sensing data, meteorological data, soil data, and DEM data.

In the Earth big Data science engineering data sharing service system, I downloaded the global 30m land cover product based on fine classification and cut the boundary vector data. (https://data.casearth.cn (accessed on 1 September 2022)). Based on the 30 m products from 2020, Landsat satellite data of a 1984–2020 time series were used to produce dynamic monitoring products of a global 30 m fine land cover for the 1985–2020 period by combining the coupled change detection and dynamic update technology. The overall accuracy of the remote sensing interpretation was 82.5% [42]. In order to keep the data consistent between the value assessment and the functional assessment, five periods of data from 2000, 2005, 2010, 2015, and 2020 were selected as data sources and reclassified according to the actual land use distribution in the study area and according to the research’s data requirements. The land was divided into eight kinds of land, as follows: dry land, irrigated cropland, grassland, broadleaf forest, coniferous forest, shrubbery, water, and land for construction.

The remote sensing data included Landsat series images, Global Ecosystem Dynamics Investigation (GEDI) L4A data, and MOD17A3, a MODIS-based Net Primary Productivity (NPP) data product. The Landsat image data comes from the Landsat satellite series, which is jointly managed by NASA and the United States Geological Survey (https://earthexplorer.usgs.gov/ (accessed on 4 September 2022)), that included Landsat-5 with thematic mapper (TM) in 2000, 2005, and 2010 and LandSAT-8 with Land Imager (OLI) and thermal infrared sensor (TIRS) in 2015 and 2020. Images acquisition time of April to May of every five years of 2000–2020 were used, and each image acquisition time was almost the same. GEDI L4A data comes from LP DACC, that were used to obtain the footprint-level observations of the ground biomass products (https://doi.org/10.3334/ORNLDAAC/1986, product data date for 17 April 2019 solstice on 5 August 2021.(accessed on 13 September 2022)). The remote sensing data products from MOD17A3 provided annual net primary production (NPP) information at a 500 m pixel resolution (data download address https://e4ftl01.cr.usgs.gov/MOLT/MOD17A3HGF.006/ (accessed on 13 September 2022))

The meteorological data include annual precipitation and annual potential evapotranspiration. The rainfall data were obtained from the Data Center for Resources and Environmental Sciences, Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences (https://www.resdc.cn (accessed on 13 September 2022)). The potential evaporation data for the research area were obtained from the Global Aridity and PET Database (https://cgiarcsi.community/data/global-aridity-and-pet-database/ (accessed on 3 October 2022)).

The soil data were derived from the Harmonized World Soil Database version 1.1 (HWSD), established by the Food and Agriculture Organization of the United Nations (FAO) and the International Institute for Applied Systems (IIASA) in Vienna.

DEM data provided the surface elevation data of the study area and were derived from the Geospatial data cloud (http://www.gscloud.cn/ (accessed on 13 October 2022)).

2.3. Methods

2.3.1. A Flowchart of the Methods

The procedure of the methods we use is shown in Figure 2. Firstly, the 11 ecological service values were evaluated based on the value equivalent factor method. Then, by introducing the multi-source data, the InVEST model was optimized to accurately estimate the values of the ecosystem services. Finally, based on the results of the optimized InVEST model, the ecosystem service values of the study area were regulated and re-evaluated. A consistency analysis and an analysis of the spatial–temporal variation characteristics of the ecosystem services were carried out before and after the fusion of the two evaluation results.

2.3.2. Ecosystem Service Assessment Based on the Method for Evaluating the Value Equivalent Factor

The method for evaluating the value equivalent factor proposed by Costanza et al., is applicable to a large area of China, but it cannot assess the value of the ecosystem services in some small areas; so, we used only the net profit of food production per unit area as the medium for the value of the ecosystem service capacity [13]. The value equivalent conversion method of Xie et al., was therefore adopted to evaluate the land service value in Cili County [43], and the net profit of the grain production per unit area of the farmland ecosystem was taken as the value of the ecosystem services of one standard equivalent factor [44]. In 2010, the value of the ecosystem services per unit area was 3406.50 CNY/ha. The grain output per unit area of Zhangjiajie was taken as the grain output per unit area of Cili County, and the grain output per unit area of Zhangjiajie in 2010 was 4501 CNY/ha, which can be compared with the national grain output per unit area of 4973.58 CNY/ha in 2010. The correction coefficient was determined to be 0.905. It can thus be concluded that the equivalent of the ecosystem service value per unit area in Cili County is 3082.88 CNY/ha. The equivalent ecosystem service value in the study area was modified regionally [45] according to the table of farmland ecosystem biomass factors in different regions determined by Xie et al. (

Table 1. Table of equivalent factors of unit area value.

Ecosystem Classification		Supply Services			Conditioning Services					Support Services		Cultural Services
First-Level Classification	Secondary Classification	Food Production	Material Production	Water Supply	Gas Regulation	Climate Regulation	Environment Purification	Hydrological Adjusting	Soil Conservation	Nutrient Cycling	Biodiversity	Landscape
Cultivated land	Rainfed cropland	0.85	0.4	0.02	0.67	0.36	0.1	0.27	1.03	0.12	0.13	0.06
	Irrigated cropland	1.36	0.09	−2.63	1.11	0.57	0.17	2.72	0.01	0.19	0.21	0.09
Forest	Coniferous forest	0.22	0.52	0.27	1.70	5.07	1.49	3.34	2.06	0.16	1.88	0.82
	Mixed leaf forest	0.31	0.71	0.37	2.35	7.03	1.99	3.51	2.86	0.22	2.60	1.14
	Broadleaved forest	0.29	0.66	0.34	2.17	6.50	1.93	4.74	2.65	0.20	2.41	1.06
	Shrubland	0.19	0.43	0.22	1.41	4.23	1.28	3.35	1.72	0.13	1.57	0.69
Grassland	Grassland	0.10	0.14	0.08	0.51	1.34	0.44	0.98	0.62	0.05	0.56	0.25
	Shrub–grassland	0.38	0.56	0.31	1.97	5.21	1.72	3.82	2.40	0.18	2.18	0.96
	Meadow	0.22	0.33	0.18	1.14	3.02	1.00	2.21	1.39	0.11	1.27	0.56
Wetland	Wetlands	0.51	0.50	2.59	1.90	3.60	3.60	24.23	2.31	0.18	7.87	4.73
Desert	Desert	0.01	0.03	0.02	0.11	0.10	0.31	0.21	0.13	0.01	0.12	0.05
	Bare areas	0.00	0.00	0.00	0.02	0.00	0.1	0.03	0.02	0.00	0.02	0.01
Waters	Water body	0.80	0.23	8.29	0.77	2.29	5.55	102.24	0.93	0.07	2.55	1.89
	Permanent ice and snow	0.00	0.00	2.16	0.18	0.54	0.16	7.13	0.00	0.00	0.01	0.09

). The correction coefficient of Hunan Province was 1.95.

The formula for calculating the ecosystem service value is as follows:

$$V={\sum }_{i=1}^n{S}_i\times V{C}_i,$$

(1)

$$V_f={\sum }_{i=1}^n{S}_i\times V{C}_{f,i},$$

(2)

where $V$ is the total value of the ecosystem services; $S_i$ is the area of the $i^{\prime }$th land use type, ha; $V{C}_i$ is the yield of land use type $i$ per unit area in the ecosystem service value of CNY/ha; $V_f$ is the value of the ecosystem services in item $f$, in CNY; $V{C}_{f,i}$, is the value equivalent of a unit area of ecosystem services for item $f$ of the $i^{\prime }$th land use type, CNY/ha; and n is the land use type.

2.3.3. Ecosystem Service Assessment Based on the InVEST Model

Optimization of the carbon storage evaluation method of the InVEST model based on multi-source data fusion

The carbon storage module in the InVEST model divides the ecosystem’s carbon storage into four basic carbon pools: aboveground biomass carbon, underground biomass carbon, soil carbon, and dead organic carbon. According to the classification of the land use/cover, the average carbon density of the four carbon pools in the different land classes was counted and then multiplied and summed with the area of each land class to obtain the total carbon storage in the study area. Its calculation formula is as follows:

$$C_{total}=C_{above}+C_{below}+C_{soil}+C_{dead},$$

(3)

where $C_{total}$ is the total carbon storage; $C_{above}$ is the carbon storage of the aboveground biomass; $C_{below}$ is the carbon storage of the underground vegetation; $C_{soil}$ is the carbon storage of the soil; and $C_{dead}$ is the carbon storage of the dead organic matter.

According to the principle of the InVEST carbon storage model given above, the carbon storage service function of Cili County during the fifth phase from 2000 to 2020 was evaluated. The data required for the assessment of the carbon storage services mainly include the land cover map of the study area, as described above, and the data table composed of the four carbon pools (

Table 2. Carbon pool table of the InVEST model carbon storage module (unit: tons/ha).

Lucode	LULC_Name	c_above	c_below	c_Soil	c_Dead
1	Rainfed cropland	38.9	7.3	89	0
2	Broadleaf forest	48	8	118	1.75
3	Shrubland	9.6	1.6	54	0
4	Coniferous forest	35	6	91	1.75
5	Irrigated cropland	38.9	7.3	89	0
6	Land for construction	0	0	78	0
7	Waters	0	0	0	0
8	Grassland	3.6	24.4	80	0

) in .CSV format, in tons/ha.

In order to improve the accuracy of the carbon storage evaluation, this paper improved the module by introducing remote sensing data and GEDI biomass data to invert the aboveground biomass. Based on the inversion results of the biomass, the evaluation results of the vegetation carbon storage were obtained.

The topography of the study area is complex, and the distribution of the forests is seriously fragmented. In order to improve the accuracy of the data, the original GEDI spots were screened. Then, based on the GEDI biomass data and Landsat images, the aboveground forest biomass inversion was carried out, and various parameters were extracted as variable factors to establish the relationship between them and the aboveground forest biomass. On this basis, the aboveground forest biomass was estimated.

The different features of the forests are represented by different spectral and textural, features in the remote sensing images.The image features images of the location of GEDI spots were extracted into the GEDI spot dataset using the geostatistical platform ArcGIS10.2. The forest GEDI spot data were classified into broadleaf forest and coniferous forest according to the land use classification, with the sample numbers 6543 and 11,450, respectively. Next, 20% of the samples of each type were selected as the test set and the remaining 80% were selected as the training set. The estimation of aboveground forest biomass was realized by establishing the mathematical relationship model between the biomass of the sample points and the corresponding characteristic factors. Firstly, SPSS software was used to conduct a correlation analysis on the biomass of all the points and its corresponding characteristic parameter data. Due to the large spatial heterogeneity of forests, the saturated spectral information, and the large number of GEDI samples, the correlation coefficient between the biomass and the characteristic parameters was generally low. Therefore, a simple linear regression model could not meet the modeling requirements. We needed to find a more suitable model.

This paper attempts to construct a remote sensing estimation model of the forest biomass based on random forest regression and geographically weighted regression.

Table 3. Results of random forest regression.

Type of Forest	$\mathbf{Training} \mathbf{Sample} {\mathbf{R}}^2$	Training Sample RMSE (tons/ha)	Sample for $\mathbf{Verification} {\mathbf{R}}^2$	Sample for Verification RMSE (tons/ha)
Broadleaved forest	0.73	29.87	0.53	36.98
Coniferous forest	0.66	28.12	0.38	34.66

summarizes the accuracy of the different forest types obtained by running the random forest model.

The test R² of the two stand types under random forest regression is much smaller than that of the training R², which indicates that the biomass inversion factor using the random forest model has a certain degree of overfitting. In the biomass inversion modeling, the study area was large and prone to spatial instability. The ArcGIS spatial analysis tool was used to carry out spatial autocorrelation analysis on the biomass of all the light spot data. The results showed that the p value was <0.1, the z value was 371.45, and the Moreland index was 0.2845. In other words, based on the reliable analysis results, spatial autocorrelation existed in the biomass point data; so, it was necessary to select a local regional regression model. The geographically weighted regression model is an extension of the ordinary linear regression model and embeds the geographical location of the data into the regression parameters, as follows:

$$y_i={\beta }_0\left(u_i,v_i\right)+{\Sigma }_k{\beta }_k\left(u_i,v_i\right)x_{ik}+{\epsilon }_i,$$

(4)

where $\left(u_i,v_i\right)$ is the coordinate of the $i$th sample point (such as latitude and longitude), which is the $k$th regression parameter ${\beta }_k\left(u_i,v_i\right)$ on the $i$th sample point and is a function of geographic location, $x_{ik}$ is the argument at the $i$th sampling point, ${\beta }_0\left(u_i,v_i\right)$ is the intercept term, ${\epsilon }_i$ is the random error term. The core of the geographically weighted regression model is the spatial weight matrix, and different spatial weight functions express different spatial relationships between the data. Several commonly used spatial weight functions include the distance threshold method, the inverse distance method, the Gauss function method, and the truncated function method.

According to the variable factors selected by the above forests and the coefficient raster images obtained by the geographical weighted regression, the biomass inversion raters of the two forest types were calculated through the geostatistical platform raster calculator tool and then the land use map of each forest type was used as the salt film to cut out the biomass of each forest type. The biomass density of the shrubs was 19.76 tons/ha [46], and the biomass of the shrubs in the different years was modified according to the NDVI vegetation index of the study area. The total forest biomass of Cili County was synthesized from the biomass of the broadleaf forest, coniferous forest, and shrubland.

The non-forest biomass was obtained from the MOD17A3 NPP data. After removing the outliers, it was mosaicked with forest biomass raster images from the corresponding periods.

Based on the aboveground biomass obtained by inversion, the carbon content ratio (0.5) recommended by the IPCC national greenhouse gas inventory guidelines was used to calculate the vegetation carbon storage in the study area. In addition, the rhizome-to-stem ratio (6.23) [47] and the ratio of biomass carbon to dead organic matter carbon (2.25%) [48] were obtained by referring to the relevant literature.

Different carbon sequestration mechanisms within the vegetation carbon pool and the soil carbon pool led to the various factors influencing carbon density. The soil carbon density was assigned by referring to the relevant literature (

Table 4. Soil carbon density of different land cover types.

Type of Land Use	Soil Carbon Density (tons/ha)
Broadleaf forest	118
Coniferous forest	91
Shrubland	54
Rainfed cropland	89
Irrigated cropland	89
The grass	80
Land for construction	78
Waters	0

) [49,50]. According to the research in this literature, the carbon storage density of forest soil per unit area in Hunan Province increased by 0.35 tons C/ha per year, and the forest soil’s carbon density was modified based on this. In addition, since the soil carbon density data in the study area were not measured, this study chose to average the data from multiple sources. Figure 3 shows the distribution map of the soil carbon density obtained according to the soil types [51], and the average data of the soil carbon density from the two sources were taken as the final distribution of the soil carbon density. Finally, the carbon storage distribution map, the carbon storage per unit area, and the total carbon storage of the different geographical classes were obtained.

Evaluation of water production based on the InVEST model

The principle of the water production module in the InVEST model is that the water balance is taken as the theoretical basis for the calculation of the water production, that is, the rainfall minus the actual evapotranspiration. The main algorithms of the model are as follows:

$$Y_{xj}=\left[1-\frac{AE{T}_{xj}}{P_x}\right]\times P_x,$$

(5)

where $Y_{xj}$ is the annual water quantity of pixel $x$ in ecosystem type $j$; $AE{T}_{xj}$ is the annual actual evapotranspiration of pixel $x$; and $P_x$ is the annual precipitation of pixel $x$.

$$\frac{AE{T}_{xj}}{P_x}=\frac{1+{\omega }_x+R_{xj}}{1+{\omega }_x+R_{xj}+1/R_{xj}},$$

(6)

where $R_{xj}$ is the ratio of the potential evapotranspiration of pixel $x$ in ecosystem $j$ to rainfall (Budyko dryness index), and ${\omega }_x$ is the ratio of the annual water demand of the vegetation to the annual precipitation.

$${\omega }_x=Z\times \left(AW{C}_x/P_x\right),$$

(7)

where $Z$ is the empirical constant adjusted according to the seasonal characteristics of the regional precipitation, usually derived by referring to the InVEST model guide, and $AW{C}_x$ is the plant effective water content of pixel $x$.

$$R_{xj}=\left(k_{xj}\times et{o}_x\right)/P_x,$$

(8)

where $et{o}_x$ is the potential evapotranspiration of pixel $x$, and $k_{xj}$ is the evapotranspiration coefficient of the vegetation, namely the ratio of the evapotranspiration to the potential evapotranspiration.

Soil conservation assessment based on the InVEST model

Soil conservation is the difference between potential soil erosion and actual soil erosion. Potential soil loss is the soil loss based on nature and vegetation protection (RKLS), and actual soil loss is the soil loss under artificial management and conservation measures (USLE). The calculation formulas of potential soil erosion and actual soil erosion are as follows:

$$USLE=R\times K\times LS\times C\times P,$$

(9)

$$RKLS=R\times K\times LS,$$

(10)

$$SD=RKLS-USLE,$$

(11)

where $R$ is the rainfall erosivity factor; $K$ is the soil erodibility factor; $LS$ is the slope length factor; $C$ is the vegetation as cover and management factor; and $P$ is the soil conservation measure factor.

2.4. The Regulating Factor Calculation of Ecological Service Value per Unit Area Based on Results of the Optimized InVEST Model

According to the study by Xie et al. [52], among the 11 ecosystem services, 8 functions, namely food production, material production, gas regulation, climate regulation, environment purification, nutrient cycling, biodiversity and landscape, are positively correlated with vegetation biomass. The supply of water resources and the hydrological adjustments are closely related to precipitation, which is consistent with the evaluation results of the water yield [52]. The soil conservation service fits perfectly with the soil conservation quantity in the quality assessment [52]. Based on the above understanding, it is theoretically feasible to regulate the evaluation results of the value quantity based on the quality evaluation results. Biomass, water yield, and soil conservation were used as regulating factors of the spatial–temporal dynamics of the 11 ecosystem service values, and the equivalent of the ecological services per unit area was calculated by combining the ecosystem service value as the scale.

The equivalent regulating formula of the spatial–temporal dynamic changes in value based on the material mass is:

$$F_{nij}=Q_{ij}\times F_n,$$

(12)

$$Q_{ij}=Bij/\stackrel{-}{B},$$

(13)

where $F_{nij}$ represents the unit area equivalent value of the $n$th class ecosystem service of the $i$th pixel in the $j$th period; $F_n$ refers to the equivalent value per unit area of the $n$th ecosystem service; $Q_{ij}$ refers to the spatial–temporal regulating factor of the mass of the $i$th pixel in the $j$ period; $Bij$ refers to the mass of the $i$th pixel in the $j$ period; and $\stackrel{-}{B}$ refers to the average mass of the land use type corresponding to the pixels over the years.

3. Results

3.1. Evaluation Results and Analysis of Ecosystem Services Based on the Value Equivalent Method

3.1.1. Evaluation Results and Analysis of Ecosystem Services Based on the Value Equivalent Method

From 2000 to 2020, the total value of the ecosystem services in Cili County increased from CNY 32.243 billion to CNY 32.473 billion, a total increase of CNY 230 million, and the individual values of the ecosystem services changed significantly (

Table 5. Value changes in individual ecosystem service functions in Cili County from 2000 to 2020 (unit: CNY 100 million).

Types of Ecosystem Services	Food Production	Material Production	Water Supply	Gas Regulation	Climate Regulation	Environment Purification	Hydrological Adjustment	Soil Conservation	Nutrient Cycling	Biodiversity	Landscape	Total
2000	9.47	9.39	0.73	30.19	78.60	24.72	88.64	34.72	3.19	29.60	13.19	322.43
2005	9.38	9.41	0.91	30.20	78.84	24.80	88.70	34.81	3.19	29.69	13.23	323.15
2010	9.33	9.41	1.04	30.17	78.86	24.83	88.88	34.83	3.18	29.71	13.24	323.48
2015	9.27	9.41	1.22	30.15	78.94	24.89	89.45	34.86	3.17	29.75	13.26	324.36
2020	9.20	9.41	1.36	30.13	79.02	24.93	89.54	34.90	3.17	29.79	13.28	324.73

). Among them, food production showed a decreasing trend from 2000, with a total decrease of CNY 27 million. The function of material production fluctuated slightly over 20 years, and the functions of gas regulation and nutrient cycling remained basically stable. The values of the other service functions showed an upward trend and increased year by year, especially the value of hydrological adjustment, which increased by a total of CNY 90 million from 2000 to 2020. Secondly, the supply of water resources also increased significantly, with a total increase of CNY 63 million.

From 2000 to 2020, the values of the ecosystem services in Cili County were, from the highest to the lowest, shrubland > coniferous forest > broadleaved forest > waters > rainfed cropland > irrigated cropland > land for construction > grassland. Due to the decrease of 758 ha in the rainfed cropland and 3284 ha in the irrigated cropland between 2000 and 2020, the value of the cultivated land was always in a state of decline. The value of the rainfed cropland decreased by CNY 19 million and that of the irrigated cropland decreased by CNY 77 million, which was four times that of the rainfed cropland. However, the ecological value of the forest land increased year by year, with a total increase of CNY 180 million from 2000 to 2020. The value of the broadleaf forest fluctuated slightly in the past two decades, and the final total value decreased by CNY 42 million, while the values of the coniferous forest and shrubland showed a significant increase, with the total values being increased by CNY 144 million and CNY 78 million, respectively. The value of the water resources increased slowly from 2000 to 2010 and sharply increased by CNY 76 million from 2015 to 2020, with a total increase of CNY 132 million. The value of the land for construction also increased, with an increase of CNY 13 million, or 69.26%, compared with 2000. It can be seen that the land for construction has expanded significantly in the past 20 years and that the urbanization level has increased rapidly. By 2020, the urbanization level of Cili County had reached 46.97%. The area of grassland was very small, and the value of its ecosystem services was almost negligible. The change rate of total ecological service value showed that the largest changes occurred during 2010–2015, followed by 2000–2005, and the two periods of 2005–2010 and 2015–2010 had small change ranges (

Table 6. Changes in the values of ecosystem services in Cili County from 2000 to 2020 (unit: CNY 100 million).

Land Use Type	Broadleaf Forest	Coniferous Forest	Shrubland	Rainfed Cropland	Irrigated Cropland	Grassland	Land for Construction	Waters	Total	Rate of Change (%)
2000	88.43	86.99	90.96	14.47	8.15	0.00	0.19	33.26	322.43	——
2005	88.09	88.27	90.97	14.38	7.90	0.00	0.21	33.33	323.15	0.22
2010	88.00	88.29	91.20	14.36	7.75	0.00	0.25	33.64	323.48	0.10
2015	87.97	88.38	91.42	14.33	7.57	0.00	0.28	34.40	324.36	0.27
2020	88.02	88.42	91.74	14.28	7.38	0.00	0.32	34.58	324.73	0.11

).

3.1.2. Spatial Differentiation of Ecosystem Service Values

The ecosystem service values of Cili County showed a certain spatial heterogeneity (Figure 4). In general, the highest ecosystem service values are distributed in the middle of the Lihue River region, in some small water bodies on the southeastern edge, and in the Jiangnan Reservoir in the northwest of Cili County. The areas with high ecosystem service values are mainly distributed in the north, the central east, and the south, and these areas are mainly broadleaved forests. The middle value areas are mainly concentrated in the northwest and south, and the main land types there are coniferous forest and shrubland. The areas with low ecosystem values are mainly concentrated in the central part of the county, which is mostly composed of land for construction and cultivated land. Unlike the ranking of the total ecosystem service value of the different regions, the supply of water resources has the highest ecosystem service value per unit area. However, since Cili County is a forestry county, the woodland area is 54 times greater than that of the waters; so, the contribution rate of the woodland to the total ecosystem service value of Cili County is higher than that of the other land types.

The temporal changes in the spatial distribution of the ecosystem service values in Cili County were not obvious (Figure 5). Relatively speaking, the changes in the ecoservice values in Cili County were obvious from 2000 to 2005. In the northwestern, central, and southern parts of Cili County, the ecosystem service values increased significantly, while the area of the ecosystem service value decreased less and there was no obvious aggregation. However, the changes in the ecosystem service values during 2005–2010, 2010–2015, and 2015–2020 were relatively weak, which was related to the low change in land use distribution in Cili County in recent years. In general, the ecological service value of Cili County showed an increasing trend from 2000 to 2020; the increasing regions were distributed in the middle, south, and northwest but were not concentrated.

3.2. InVEST Model Evaluation Results and Analysis

The InVEST model was run to obtain the assessment result of Cili County’s carbon storage in the fifth phase. Figure 6 shows the distribution map of the carbon storage in Cili County from 2000 to 2020. It can be seen that the distribution of carbon stocks in Cili County had no obvious change from 2000 to 2020; the spatial heterogeneity is low and the carbon stocks in the same region show no change, which is obviously inconsistent with the facts.

3.2.1. Analysis of Temporal and Spatial Variation Characteristics of Carbon Stocks

The biomass maps of Cili County in 2000, 2005, 2010, 2015, and 2020 were obtained by using the improved multi-source data (Figure 7). Then, the evaluation results of the vegetation carbon storage were obtained. From 2000 to 2020, the total carbon storage of the study area showed an upward trend. The carbon storage of the study area was the highest in 2020 (41.18 TgC), while the carbon storage of the study area was the lowest in 2000 (37.30 TgC). In addition, studies have shown that from 1983 to 2004, the forest area and vegetation carbon storage in Zhangjiajie City also showed an increasing trend, but the rate of increase in vegetation carbon density during this period was significantly lower than during the 2000–2020 period [51].

Table 7. Carbon storage assessment results of different land cover types.

LULC	2000		2005		2010		2015		2020
	MEAN (tons/ha)	SUM (tons)	MEAN (tons/ha)	SUM (tons)	MEAN (tons/ha)	SUM (tons)	MEAN (tons/ha)	SUM (tons)	MEAN (tons/ha)	SUM (tons)
Broadleaf forest	143.85	9,058,371	146.64	9,205,132	145.70	9,268,238	155.62	9,899,982	164.21	10,454,065
Coniferous forest	128.31	10,478,249	125.59	10,410,444	129.82	10,845,586	138.41	11,579,586	146.74	12,287,717
Shrubland	86.02	8,152,590	87.38	8,295,101	87.30	8,648,218	88.64	8,807,501	89.91	8,970,374
Rainfed cropland	98.44	5,967,130	98.60	5,929,222	99.02	5,957,023	99.62	5,970,309	99.49	5,935,426
Irrigated cropland	96.67	3,386,447	96.74	3,284,002	97.02	3,239,908	97.62	3,180,918	97.50	3,093,479
Grassland	95.07	291	95.61	361	95.77	405	97.35	412	96.91	454
Land for construction	88.48	261,108	88.56	297,364	88.86	341,754	89.34	388,823	89.32	439,897
Waters	0.00	0	0.00	0	0.00	0	0.00	0	0.00	0

shows the carbon storage per unit area and the total carbon storage of the different land cover/use types in Cili County. The results showed that the carbon storage per unit area of different land use types in Cili County was in the order of broadleaf forest > coniferous forest > rainfed cropland > irrigated cropland > grassland > land for construction > shrubland > waters. The carbon storage per unit area of cultivated land was second only to that of the forestland, indicating that the forestland and cultivated land have strong carbon sequestration capacities and that the grassland has the same carbon storage per unit area as the cultivated land. According to the results of the total carbon storage, the carbon storage of the different land types in descending order was coniferous forest > broadleaved forest > shrubland > rainfed cropland > irrigated cropland > land for construction > grassland > waters. Although the carbon storage per unit area of broadleaved forest was higher than that of the coniferous forest, the area of coniferous forest is greater than that of broadleaved forest in Cili County; so, the total carbon storage of the broadleaved forest was higher than that of the coniferous forest. The grassland area is small, and its total carbon storage is almost negligible. The urban land and the waters can basically be regarded as land use types without carbon sequestration ability. In this paper, only the soil carbon of land for construction is included in the calculation, but due to its small footprint, the total carbon storage is also small. The carbon storage density of the same land use/cover type also changed greatly over the study period and was mainly influenced by the external environment.

The distribution of carbon storage in the study area changed to a certain extent from 2000 to 2020, but the overall pattern remained basically stable (Figure 8). From the spatial characteristic perspective, the areas with large carbon reserves in Cili County are mainly distributed in the north, central, southwest, and eastern parts of Cili County, and the land use/cover types in these areas are mostly woodland and cultivated land; in particular, the forest land in the scenic area in the southwest has better quality and higher carbon sequestration capacity. The areas with low carbon storage are mainly concentrated in the northwestern and central areas of Cili County, and the land use/cover types in these areas are mainly water bodies and urban land. In general, the carbon sequestration capacity of the different land use/cover types is significantly different. The carbon sequestration capacity of the forest land is the strongest, followed by the cultivated land and grassland; the land for construction and the water have almost no carbon sequestration capacity.

3.2.2. Analysis of Temporal and Spatial Variation Characteristics of Water Yield

The land for construction has little capacity to produce water and is generally not considered in the assessment of the water supply capacity. The total water yields of the different land types in descending order are shrubland > rainfed cropland > broadleaf forest > coniferous forest > irrigated cropland > waters > grassland. Cili County is dominated by forestry, with woodland area accounting for 60% of the total area, within which shrubland occupies the largest area; so, the total water supply of the shrubland is the largest. Furthermore, the area of water and grassland is small; so, the water yield is also small. During the 20 years from 2000 to 2020, the water production in the study area showed a trend of first decreasing, then increasing, and then decreasing again (

Table 8. Water yield assessment results of different land cover types.

LULC	2000		2005		2010		2015		2020
	MEAN (mm)	SUM (103 m3)	MEAN (mm)	SUM (103 m3)	MEAN (mm)	SUM (103 m3)	MEAN (mm)	SUM (103 m3)	MEAN (mm)	SUM (103 m3)
Broad leaved forest	498	312,722	364	227,426	789	492,665	504	315,031	504	315,225
Coniferous forest	955	931,056	793	774,785	1261	1,234,730	973	956,068	973	959,949
Brush forest	383	309,309	254	207,951	638	523,069	391	320,601	391	320,853
Irrigated cropland	709	247,500	532	179,906	993	329,496	735	237,907	735	231,686
Rainfed cropland	963	580,953	787	471,147	1257	750,736	984	585,724	984	583,002
Waters	239	10,364	54	2354	510	22,433	264	11,887	264	11,953
Grassland	580	18	430	16	889	38	615	26	616	29
Total	—	2,391,921	—	1,863,587	—	3,353,166	—	2,427,244	—	2,422,696

). The amount of water produced in 2005 was the lowest, while the amount of water produced in 2010 was the highest in the study area. The amount of water produced was closely related to rainfall. The annual output of water decreased sharply in 2005 and increased sharply in 2010. The reason for this is that the precipitation in the study area was 984.1mm in 2005, and it increased to 1582mm in 2010, with an increase of 60.76%. The water yield density of different land use types in Cili County was in the order of rainfed cropland > shrubland > irrigated cropland > grassland > broadleaf forest > coniferous forest > water. Cili County is located in the south of China and has sufficient rainfall; so, the cultivated land has stronger water conservation capacity because it may have stronger water retention capacity than the forest land under the conditions of sufficient rainfall. The water yield per unit area of forest land is similar to that of the cultivated land and grassland. In fact, the water conservation capacity of the forest land and grassland is also very strong. However, the evapotranspiration of the forest land is larger, including not only soil evaporation, but also the evaporation from the vegetation itself. Water produces the smallest amount of water per unit area, except for the land for construction, because water has no soil to retain water and evapotranspiration occurs directly on its surface, the speed of which is affected by climatic conditions.

From the perspective of the spatial distribution characteristics, the water yield in the central north, southwest, and central east of Cili County is larger, and these areas are mainly areas with concentrated distributions of shrubland, coniferous forest, and cultivated land (Figure 9). Although the broadleaved forests belong to the same woodland, the water yield of the concentrated broadleaved forest areas in the north, southwest, and central east of Cili County is in the middle. The lowest water yield is mainly distributed in the northwestern and central areas of Cili County, which is mainly water and land for construction. In general, the spatial heterogeneity of the water production in Cili County was strong, and the water supply capacity of the different land use types was significantly different. The water supply capacity of the cultivated land and woodland was the strongest, followed by the grassland. The water supply capacity of the land for construction was almost negligible. The temporal variation in the water yield of the same land use type is mainly related to the external natural environment. For example, the change in rainfall is mainly affected by various climatic conditions, and global warming often leads to an increase in evapotranspiration.

3.2.3. Analysis of Spatial–Temporal Variation Characteristics of Soil Conservation

During the 20 years from 2000 to 2020, soil conservation in the study area decreased first, then increased, and then decreased again. In 2000, 2005, 2010, 2015, and 2020, the total amount of soil conservation in the study area was 7.74 × 10⁷, 6.27 × 10⁷, 10.60 × 10⁷, 7.80 × 10⁷, and 7.80 × 10⁷ tons, respectively (

Table 9. Soil conservation assessment results of different land cover types.

LULC	2000		2005		2010		2015		2020
	MEAN (tons/ha)	SUM (tons)	MEAN (tons/ha)	SUM (tons)	MEAN (tons/ha)	SUM (tons)	MEAN (tons/ha)	SUM (tons)	MEAN (tons/ha)	SUM (tons)
Rainfed cropland	116.57	7,043,816	91.23	5,468,583	154.44	9,238,802	114.74	6,837,220	114.30	6,779,358
Broadleaved forest	377.76	23,795,535	313.09	19,653,230	529.30	33,191,893	383.96	24,082,638	383.96	24,100,406
Shrubland	230.17	22,516,095	187.26	18,338,273	317.86	31,217,988	232.83	22,936,740	232.49	22,997,759
Coniferous forest	279.11	22,628,083	220.97	18,191,559	371.18	30,570,898	276.23	22,782,613	275.98	22,781,580
Irrigated cropland	37.68	1,317,627	28.28	958,391	48.42	1,609,167	35.65	1,156,215	34.89	1,101,803
Land for construction	26.11	76,867	21.74	72,822	37.67	144,366	29.54	128,106	33.31	163,538
Waters	10.00	43,495	8.03	35,066	13.98	61,607	10.68	48,190	10.79	48,956
Grassland	136.12	417	101.83	385	156.70	663	114.16	483	104.04	487

). Soil retention capacity is related to rainfall erosivity, and rainfall erosivity is based on total rainfall. Rainfall erosivity can reflect the amount of soil quality lost by rainfall in different regions or at different periods in the same region. Therefore, soil conservation decreased sharply in 2005 and increased sharply in 2010. The soil conservation per unit area of different land use types in Cili County from largest to smallest is in the order of broadleaf forest > coniferous forest > shrubbery > grassland > rainfed cropland > irrigated cropland > land for construction > water. The amount of soil conservation per unit area of forest land was much higher than that of the other land cover types, indicating that forest can effectively reduce soil loss and that tree planting is the most effective way to improve soil conservation. In addition, the contribution of arable land to soil conservation is also considerable. From the results of the total soil conservation quantity, the soil conservation quantity of the different land types was in the order of broadleaf forest > coniferous forest > shrubland > rainfed cropland > irrigated cropland > land for construction > waters > grassland. Because the waters, the land for construction, and the grass in Cili County were small areas, the soil conservation effect could almost be ignored.

From the perspective of spatial characteristics, the distribution pattern of the soil conservation in Cili County was basically stable from 2000 to 2020 and was similar to the carbon storage and water supply. The areas with large soil conservation are mainly concentrated in the central north, southwest, and central east of Cili County, where the land cover is also mainly woodland and cultivated land. In the northwestern and central areas of Cili County, soil conservation is low, and the land use/cover types are mainly waters and land for construction. In general, the soil conservation abilities of the different land use types were significantly different, and the spatial heterogeneity was strong. The soil conservation ability of the forest land was the strongest, followed by grassland and rainfed cropland, and the soil conservation ability of the irrigated cropland, land for construction, and waters was weak. Soil conservation is related to various factors and is mainly affected by human activities and the natural environment. Studies have found that soil conservation ability is greatly affected by terrain, vegetation cover, and meteorological conditions. Due to the drastic precipitation changes in Cili County during the study period, the soil retention per unit area of the same land use type also showed significant differences in different periods (Figure 10).

3.3. Comparison of Three Ecosystem Service Assessment Results with the Value Equivalent Method of Ecosystem Service Value

3.3.1. Comparison of Ecosystem Service Assessment Results in Spatial Distribution

In order to compare the results between the ecosystem service value assessments, three corresponding spatial distribution maps of the ecosystem service values from 2000 to 2020 were generated based on the value equivalent factor method. The spatial consistency across the three ecosystem service assessment results of carbon storage (climate regulation), water production (water supply), and soil conservation was analyzed.

Figure 11 shows the distribution of the climate regulation values in Cili County from 2000 to 2020. Compared with the carbon storage spatial distribution map for 2000–2020 in Section 3.2.1, the assessment results of the Cili County carbon storage and climate regulation values have good spatial consistency overall. The spatial variation in the carbon storage in Cili County is significantly different from that of the climate regulation values; this is related to the use of the same ecological type and the same value equivalent factor in the estimation of the climate regulation values.

Figure 12 shows the distribution of the water resources supply values in Cili County from 2000 to 2020. By comparison with the spatial distribution map of the annual water quantity from 2000 to 2020 in Section 3.2.2, it was found that although the spatial distribution of the water resource supply values and water production were consistent across the different grades, the grades in the same region were inconsistent.

Figure 13 shows the distribution of the soil conservation values in Cili County from 2000 to 2020. Compared with the spatial distribution map of the soil conservation quantity from 2000 to 2020 in Section 3.2.3, it was found that the evaluation results of the soil conservation values in Cili County were spatially consistent with those of the soil conservation quantity.

3.3.2. Comparison of Assessment Results of Ecosystem Services in Different Genotypes

According to the above results of the ecosystem service value and function assessment, the consistency across the three ecosystem service assessment results of carbon storage (climate regulation), water yield (water supply), and soil conservation under different land use types was analyzed.

Figure 14 shows the comparison between the change in the climate regulation value and the carbon stock of the different land use types. There is a big difference between the assessment results of the climate regulation value and the carbon storage of the water. Due to different assessment principles, water’s carbon storage is zero and is not comparable with the climate regulation value. The climate adjustment value of the land for construction fits the trend of the carbon storage completely. The main reason for this is that both the climate adjustment value and the carbon storage of the land for construction were calculated by the assignment of land use type, and the change in land use area was consistent over time. The increase in the climate regulation value of the grassland was obvious, and the climate regulation value of the grassland fit well with the trend of the carbon storage. The climate regulation value and the carbon storage of the irrigated cropland showed a decreasing trend because the irrigated cropland had a stable water supply capacity and was less affected by the external environment. The rainfed cropland’s climate regulation value and carbon storage showed the same trend during the 2005–2015 period, but there was a big difference in the overall fit of the trend because the rainfed cropland’s output was greatly affected by the climate, environment, and other factors. Both the shrubland and coniferous forest showed an increasing trend in climate regulation values and carbon storage. The coniferous forest was dominated by artificial forest, which refers to forests formed by artificial measures; artificial forests are younger in age and have a relatively small amount of cutting, but their carbon storage capacity increases at a faster rate and their climate regulation value increases at a slower rate, which was determined by different assessment systems. There was a big difference between the climate regulation value of the broadleaf forest and its changed carbon storage trend. The carbon storage of the broadleaved forest showed an increasing trend, while the climate regulation value showed a decreasing trend. The distribution of the broadleaved forest, as the main cutting forest type in Cili County, changed greatly during the study period; so, the evaluation results of the broadleaved forest were not consistent. In general, the assessment results of the climate regulation values and carbon storage under different land use types were basically the same, but there were some differences.

Figure 15 shows the comparison of the water supply value and water yield of the different land use types. The evaluation results of the water resources supply value and the water production were not consistent from 2000 to 2010 but showed a consistent trend from 2010 to 2020. The water yield of the land for construction was 0, which was determined by the evaluation principle of water yield. The value of the water resource supply was on the rise because the change in the water resource supply value was only reflected by the change in the land for construction area. The area of land for construction in Cili County has been expanding continuously for the past 20 years. The variation trend of the water supply value of the grassland was completely consistent with that of the water yield, and the increase in the water supply value of the grassland was more obvious because of the small proportion of the grassland area. The variation trends of the water supply value and water yield of the irrigated cropland were basically the same after 2005. The supply value of water resources in the rainfed cropland has been increasing, but the variation trend of the water yield is volatile. The water supply value and water production of the shrubland and coniferous forest were significantly different in 2005 and 2010, but both showed an increasing trend. The variation trends of the water supply value and water production of the broadleaf forest were consistent in the other periods, except for 2005. In general, there is a certain degree of consistency between the evaluation results of the water supply value and water production of the different land use types, and the inconsistency between the two results is mainly reflected in the differences in the evaluation systems.

Figure 16 shows the comparison between the soil conservation value and the amount of soil conservation of the different land use types. The change trends of soil conservation value and soil conservation quantity were basically the same, except for in 2005. The change trend of the soil conservation value of the land for construction was completely consistent with that of the soil conservation quantity. The change in the soil conservation capacity of the land for construction was not great. The increase in the soil conservation quantity and the value of land for construction was mainly reflected in the increase in the land for construction area. The soil conservation value of the grassland has been on the rise, and the soil conservation quantity for the grassland reached its maximum value in 2010, but the overall trend was the same. The change trends of the irrigated cropland’s soil conservation value and soil conservation quantity were inconsistent from 2005 to 2010, and the change trends in the other time periods were basically the same. The overall trend showed a consistent decreasing trend. The reason for this was that the soil conservation ability of the irrigated cropland was not strong, and its total area in Cili County was reducing over time. The soil conservation value of the rainfed cropland and grassland has been on the rise, and the amount soil conservation reached a maximum value in 2010, with poor consistency. However, from the general trend, while the amount of soil conservation of the rainfed cropland was basically stable, the area of rainfed cropland was decreasing. Indirectly, the soil conservation ability of the rainfed cropland was also increasing, which was consistent with the value of the soil conservation to a certain extent. The change trends of the soil conservation value and soil conservation quantity of the shrubland and coniferous forest were significantly different between 2000, 2005, and 2010 to 2015, but they showed an increasing trend from 2000 to 2020. The soil conservation quantity and soil conservation value of the two forest types showed an increasing trend, and there was a consistency in the trend change. The change trend of the soil conservation value and soil conservation quantity of the broadleaved forest was consistent in the other time periods, except for 2010, but the overall trend was quite different. The reason why the soil conservation value of the broadleaved forest showed a declining trend was that a large amount of the broadleaved forest was felled, and thus, the broadleaved forest reduced in area and decreased in value, while the soil conservation value was more focused on the reflection of the soil conservation ability. Although the area of broadleaved forest in Cili County decreased, the soil conservation capacity of the broadleaved forest also increased with time; so, the soil conservation capacity of the broadleaved forest did not decrease. There was a good agreement between the soil conservation value and soil conservation quantity under the different land use types.

3.4. Temporal and Spatial Changes in Ecological Service Value Based on Dynamic Assessment

Based on the quality assessment results, three spatial–temporal dynamic regulation factors were constructed to adjust the original ecosystem service value equivalent, and the assessment results of the ecosystem service value in Cili County from 2000 to 2020 after adjustment were obtained. The ecosystem service values of the different land cover/use types were calculated, and the spatial distribution map of the ecosystem service values after adjustment was obtained.

After adjustment, the total value of the ecosystem services in Cili County from 2000 to 2020 still showed an upward trend, increasing from CNY 26.136 billion to CNY 35.444 billion, with a total increase of CNY 9.308 billion, which was more obvious than that before adjustment (

Table 10. Changes in the value of ecosystem services after adjustment in Cili County from 2000 to 2020 (unit: CNY 100 million).

LULC	2000	2005	2010	2015	2020
Rainfed cropland	65.67	66.43	73.61	69.65	78.03
Broadleaved forest	54.15	47.92	62.84	65.81	82.45
Shrubland	81.99	78.21	85.85	91.71	103.27
Coniferous forest	52.13	44.43	59.11	63.37	82.73
Irrigated cropland	7.41	6.93	7.48	7.54	7.97
Grassland	7.40 × 10−4	8.19 × 10−4	12.44 × 10−4	11.60 × 10−4	15.60 × 10−4
Total	261.36	243.92	288.91	298.08	354.44

). The total values of the ecosystem services in Cili County from 2000 to 2020 were in the order of shrubbery > rainfed cropland > broadleaf forest > coniferous forest > irrigated cropland > grassland. Compared with 2000, the ecosystem service value of Cili County had a slight decrease in 2005. Since 2010, the ecosystem service value of Cili County has increased significantly, and the rate of increase is becoming faster.

After adjustment, the spatial heterogeneity of the ecosystem service value in Cili County was more significant, and the land use change was no longer the only cause of change (Figure 17). From the perspective of spatial distribution, the ecosystem service value of the broadleaf forest was the highest and was distributed in the central and western part, the central and eastern part, and the southwest part of Cili County. The lowest ecosystem service value was mainly concentrated in the north and central east and mainly in the waters and the land for construction; the service value per unit area of irrigated cropland was also low. The value of the coniferous forest, shrubland, and cultivated land was in the middle, and the distribution was wide. The spatial heterogeneity of the ecosystem service value of the cultivated land and forest land was no longer obvious after adjusting the value quantity by quality. The contribution rate of the forest land to the total value of the ecosystem services in Cili County was higher than that of the other land classes, which was consistent with the conclusion before adjustment. The spatial distribution of the ecosystem service value in Cili County was very obvious. The value of the ecological services increased significantly in some regions in northwestern, central, and southern China. In general, the overall ecological service value of Cili County increased significantly from 2000 to 2020.

4. Discussion

In the ecosystem service assessment based on the value equivalent method, the same value equivalent was applied to the same type of ecosystem across different time series, but it was difficult to reflect the spatial and temporal heterogeneity of the ecosystem. Based on the quality evaluation results, this paper adjusted the value evaluation results. The change in the ecosystem service value in Cili County after adjustment was more significant than that before adjustment, and the spatial heterogeneity was stronger. Theoretically, the assessment results are more consistent with the actual situation of the study area because the changes in the ecosystem services are not only reflected in the changes in land use type but are also reflected in the improvement in the quality of the ecological environment. Considering the rate of improvement of the ecosystem service value, the ecological protection policies implemented in Cili County since 2000, such as returning farmland to forest and natural forest protection, have been effective. According to the data released by the Forestry Bureau of Hunan Province, Cili County has implemented nine key afforestation projects and completed afforestation projects totaling 62,666.67 ha, with a total investment of CNY 459.69 million. We will strictly manage forest cutting quotas, ban and reduce forest cutting, vigorously implement ecological restoration projects such as the protection of public and natural commercial forests, standardize the management of occupied and expropriated forest land, and strictly observe the red line for forest land protection. We will strengthen the protection of biodiversity, strictly protect wild animals and plants and their habitats, make all-out efforts to prevent and control forest pests, strengthen forest fire prevention and control, innovate working mechanisms, and comprehensively implement grid management for forest fire prevention and control. Actively exploring the effects of implementing the “forest chief system” to promote the “forest chief management” [53] indicates that the ecosystem service quality of Cili County has indeed been greatly improved. However, compared with the average biomass of the forest ecosystem in Hunan Province, the forest biomass in Cili County is still at a low level and the forest carbon sequestration ability is weak; so, the forest’s ecological quality needs to be improved. In this regard, Cili County should continue to strengthen the forest conservation work, optimize the land use structure, adhere to the implementation of natural forest protection and the policy of returning farmland to forest, respond to the national call of “clear waters and lush mountains are gold and silver mountains”, and continue to promote the construction of an ecological civilization while still taking ecological protection and restoration as the focus of ecological work in the future.

However, people usually evaluate ecosystem services independently using quality or quantity of matter as the method for evaluating the value equivalent factor. For example, Zhang et al. [54] predicted the ecosystem service value of Dongfeng City in 2025 based on the ecosystem value equivalent factor method. Zhao et al. [55] corrected the five-unit area value equivalent factors and built and established a value evaluation model that varies with different regions to evaluate the value of the ecosystem services in Gansu Province. Li et al. [56] used ArcGIS and the InVEST model to evaluate the wetland ecosystem of Donating Lake. Sui Yusheng et al. [57] used the InVEST model to evaluate the evolution of the spatial–temporal pattern of blue carbon distribution and its service value in the coastal zone of Jiaozuo Bay and also obtained the carbon changes in different spatial–temporal periods so as to provide a reference for the assessment of the ecosystem service value of coastal zones.

However, the carbon storage module of the InVEST model takes land use type as a single influencing factor to estimate carbon storage, ignoring the influence of soil, terrain, climate, and other factors, which reduces the accuracy of carbon storage estimation to a certain extent. Therefore, this paper improved the method for estimating carbon storage and obtained the assessment results of vegetation carbon storage by inverting biomass through multi-source data. The biomass inversion results of 2015 were compared with the biomass prediction results of Dr. Li [58] in northwestern Hunan Province. The left side (a) of Figure 18 shows the biomass prediction results of different stand types in northwestern Hunan Province based on the PMV model in 2015, and the biomass distribution in Cili County is captured in the figure. On the right (b) is the corresponding forest biomass reclassified by the biomass inversion results of 2015 in this study. According to the comparison, it was found that the biomass intervals in Cili County in the two maps were basically the same, and the biomass distribution in most of the regions was between 0 and 120. The biomass in the northern, central, and southern parts of Cili County is concentrated, indicating that the biomass inversion results in this study are relatively reliable.

At the same time, when calculating the regulation factor of the ecological service value per unit area based on the quality of the object, the biomass, water yield, and soil conservation obtained from the assessment of the quality of the object were selected as the spatial–temporal dynamic regulation factors of the three ecosystem service values, which is theoretically feasible. In addition, according to 3.3, the changes in the ecosystem services in Cili County can be more realistically reflected after the adjustment of the water production and soil conservation, which further explains the rationality and scientific nature of the adjustment of the ecological service value per unit area based on quality.

5. Conclusions

In this work, Cili County was selected as the study area to evaluate the ecosystem service value of five periods of land cover/use data from 2000 to 2020 using the unit area value equivalent method. An InVEST model was used to evaluate the three ecosystem service functions of carbon storage, water yield, and soil conservation. In order to improve the evaluation accuracy of the carbon storage, remote sensing data and GEDI data were introduced to obtain the evaluation results of the carbon storage based on the inversion results of the biomass. Finally, the quantitative value was adjusted based on the quality assessment results. The main conclusions are as follows:

(1)

Before the adjustment of the value of the ecosystem services, the total value of the ecosystem services in Cili County increased from CNY 32.243 billion to CNY 32.473 billion during 2000–2020, with a total increase of CNY 230 million, and the individual values of the ecosystem services changed significantly. The contribution of forest land to the ecosystem service value was relatively high, and there was no spatial heterogeneity in the ecosystem service values within a single ecosystem. In Cili County, the areas with high ecosystem service values were mainly concentrated in the distribution areas of water and woodland in the eastern, central, northwestern, and southern parts of Cili County, while the areas with low ecosystem service values were mainly concentrated in the residential areas in the middle and eastern parts of Cili County. The value of the areas in the central, southern, and northwestern parts of Cili County increased significantly during the study period.

(2)

The total carbon storage in the study area showed an increasing trend from 2000 to 2020. Water yield and soil conservation decreased first, then increased, and then decreased again. From the spatial characteristic perspective, the spatial distribution of the three ecosystem service functions of carbon sequestration, water supply and soil conservation in the study area showed a high consistency. The northern, central, southwestern, and central eastern parts of Cili County had better ecological quality, with higher carbon storage, water production, and soil conservation, and the land use distribution was mainly woodland and cultivated land. The northwestern and central areas of Cili County had poor ecological quality and weak carbon sequestration capacity, water supply capacity, and soil conservation capacity. The land use distribution was mainly land for construction and water.

(3)

After adjustment, the ecosystem service value of Cili County increased from CNY 26.136 billion to CNY 35.444 billion, with a total increase of CNY 9.308 billion. The total value of the ecosystem services showed an upward trend, and the rate of increase became faster with time. From 2000 to 2020, the overall increase in the value of the ecosystem services in Cili County was obvious and was consistent with the actual changes in the ecosystem in Cili County.