Evaluation and Analysis of the Gross Ecosystem Product towards the Sustainable Development Goals: A Case Study of Fujian Province, China

Abstract

Achieving sustainable development is an issue of global concern. Accounting for the gross ecosystem product (GEP) value can specifically quantify the value of ecosystems for people, which is conducive to the formulation of sustainable eco-management decisions. Multi-source data, including remote sensing images, geospatial data, and statistical bulletin information, were used to quantify the GEP value of material products, regulating services, and cultural services for Fujian Province, China, during 2000–2020. On this basis, the spatio-temporal characteristics of GEP and the coupling relationship between GEP and GDP were analyzed. The results showed that: (1) the value of GEP in Fujian Province increased by 27.9% from CNY 3589.04 billion in 2000 to CNY 4590.25 billion in 2020. Among the service values, the contribution rate of regulating services to GEP was always the highest during the study period. (2) The spatial distribution pattern of GEP in Fujian Province was higher in the west and lower in the east. Comparing prefecture-level cities, Nanping maintained its GEP at the maximum value level over the past 21 years, while Xiamen and Putian maintained their GEP at the minimum value level. (3) GDP grew faster relative to GEP over the past 21 years, and the difference between GEP and GDP decreased. GEP had a long-term positive effect on GDP, while GDP had a smaller effect on GEP in the short term. The research was not only enriched in relation to GEP accounting, but also the policy recommendations for improving the mechanisms related to the optimization of sustainable development goals have some practical significance.

1. Introduction

Gross ecosystem product (GEP) is the total economic value of the final material products and services provided by the ecosystem of a region for human welfare and sustainable economic development within a given period of time, including the value of material products, regulating services, and cultural services [1]. The idea of accounting for GEP stems from the concept of gross domestic product (GDP) and addresses the fact that GDP does not fully reflect the contribution of nature to economic activities and human welfare. GEP is a set of evaluation systems that operates in parallel with GDP [2]. In the process of human socio-economic development, GDP comes at the cost of overexploitation and consumption of ecological resources and the ecological environment, leading to the impairment of ecological services. The pressure of increasingly serious ecological and environmental problems has led to studies involving GEP accounting becoming an inherent part of global sustainable development. GEP accounting not only provides monetary evaluation for the construction of an ecological compensation value for the provision of ecological services but also serves as an important reference index for assessing the performance of ecological civilization construction.

Research on GEP can be traced to the concept of ecosystem services introduced by Ehrlich of Stanford University in 1981 [3]. Subsequently, Ehrlich and Mooney systematically explained this concept [4]. In 1997, Daily et al. defined ecosystem service functions as the natural environmental conditions and utilities created by ecosystems and ecological processes that sustain human survival [5]. In the same year, Costanza measured the value of ecosystem service functions globally using ecosystem goods and services [6]. In 2002, de Groot et al. defined ecosystem functions as the ability of natural processes and their components to provide goods and services that meet direct or indirect human needs [7]. In 2005, the United Nations Millennium Assessment considered the above definitions together and concluded that ecosystem service functions were the various benefits that humans derive from ecosystems [8,9]. In 2013, Ouyang et al. introduced the concept of GEP for the first time and started a research boom in this field in China when accounting for the 2010 provincial GEP in Guizhou Province [1]. A wide range of scholars have successively conducted GEP accounting studies on different ecosystems such as forests, wetlands, grasslands, and farmlands at the county scale [10], municipal scale [11,12,13], provincial scale [14,15,16], and other regional scales [17]. However, although the current research on GEP is constantly enriched, there are still some shortcomings in the current studies. For example, many studies focus on accounting for a single-period GEP or comparing accounting for longer interval years. The influence of ecological policies and decisions related to the accounting results is easily ignored, resulting in the possible specificity of the accounting results. This means that the findings cannot fully and objectively reflect the ecological potential of a study area. In addition, the current accounting of GEP mainly considers large-scale regions and individual ecosystems such as forests and grasslands. The number of GEP accounting studies at smaller scales, such as provinces, cities, districts, and counties, needs to be increased.

As the first demonstration area of ecological civilization identified by China, Fujian Province possesses rich natural and human resources, with among the highest quantities of freshwater and forest resources in the country. The regional location of the harbor is outstanding. Overall, it plays an important role in the process of regional sustainable development in China. Therefore, performing GEP assessment and analysis for Fujian Province would not only provide a concrete scientific basis for achieving its own sustainable development goals but also serve as a reference for GEP accounting studies in other regions. Considering the importance and existing limitations of GEP accounting, this study used multi-source data, i.e., remote sensing images, geospatial data, and statistical bulletins, to assess and analyze the GEP of Fujian Province under the sustainable development goals. The main objects of the study are to: (1) construct a comprehensive accounting system covering three aspects, i.e., product supply, regulating service, and cultural service values, to calculate the GEP of Fujian Province during 2000–2020; (2) quantify the spatio-temporal characteristics of the GEP in Fujian Province during the study period by spatial analysis and explore the coupling relationship between GEP and GDP; and (3) propose feasible policy recommendations for regional sustainable development.

2. Materials and Methods

2.1. Study Area

Fujian Province is located on the southeast coast of China, stretching from 23°33′ N to 28°20′ N and from 23°33′ E to 28°20′ E. It borders Zhejiang Province, Jiangxi Province, and Guangdong Province to the north, west, and south, respectively. The eastern region is separated from Taiwan Province by the Taiwan Strait. Fujian Province has nine prefecture-level cities, i.e., Fuzhou, Longyan, Nanping, Ningde, Putian, Quanzhou, Sanming, Xiamen, and Zhangzhou, with a total area of 1.24 × 10⁵ km² (Figure 1). The region has a subtropical maritime monsoon climate with abundant light, heat, and water resources. It has an average annual temperature of 17–21 °C and an average rainfall of 1400–2000 mm. There are rich ecological resources in the region. Specifically, the forest coverage rate has been the highest in China for 40 consecutive years. There is a dense network of waterways with many rivers, and the provincial coastline is the second longest in China. The ecological landscape is complex with a wide variety of wildlife, making it one of the most biodiverse provinces in China.

2.2. Data Sources and Pre-Processing

The Fujian GEP was calculated for the years 2000–2020, and multiple data sources were integrated into the study, including statistical information, price parameters, and geospatial data.

Statistical data: the output value of agriculture, forestry, animal husbandry, and fishery products; water consumption; reservoir and pond capacity; offshore and coastal wetland area; tourism revenue; etc. were obtained from statistical survey information or bulletins of industry departments such as the Bureau of Statistics, the Department of Water Resources, the Department of Natural Resources, the Forestry Bureau, and the Bureau of Cultural Tourism.

Price parameter data: the water resources trading price, water quality pollutant treatment cost, air pollutant treatment cost, pollutant emission price, reservoir construction cost, reservoir dredging cost, carbon sequestration price, market oxygen trading price, wind and wave control price, forest and grassland artificial pest control cost, and other price parameters were obtained through price departments, market research, and literature references.

Geospatial data: land cover data with a spatial resolution of 30 m were obtained from the Zenodo platform (https://zenodo.org/record/5210928#.Y2cpbnZBxPa, accessed on 22 September 2022). The normalized difference vegetation index and net primary productivity were MODIS data products from the National Aeronautics and Space Administration (https://ladsweb.modaps.eosdis.nasa.gov/, accessed on 4 October 2022) with a spatial resolution of 1 km and 500 m, respectively. Digital elevation data were obtained from the geospatial data cloud platform (http://www.gscloud.cn/sources, accessed on 24 September 2022) with a spatial resolution of 30 m. To ensure the consistency of the data processing units, all the geospatial data were resampled to 1 km.

2.3. Accounting Methods

Considering the meaning of GEP and the characteristics of different ecosystems in the study area, an accounting index system was constructed to evaluate the GEP of Fujian Province from 2000 to 2020, which covered three functional categories and 13 accounting items. Specifically, the material product supply value included agriculture, forestry, animal husbandry, fishery products, and water supply. The ecosystem regulating service value included water connotation, soil conservation, flood storage, coastal zone protection, carbon sequestration, oxygen supply, air purification, water purification, climate regulation, and pest control. The ecosystem cultural service value included the value of leisure tourism. The specific information for the GEP accounting index system is shown in Table 1.

In this study, the concept of calculating the value based on the functional quantity was used to account for each GEP item. The functional quantity is the physical quantity of final products or services that humans obtain directly or indirectly from the ecosystem. Due to the different measurement units among the functional quantities, the functional quantities of different ecological products and services cannot be directly summed. Prices were used to convert the functional quantities of different ecological products and services into monetary units, providing the value of the various types of service functions, which were then summed to obtain the GEP (Table 1). In this research, the main methods of price determination are the market value method [10,18,19], the shadow engineering method [18], the alternative cost method [10,20], the expert appraisal method [11], and the travel cost method [12,21]. Moreover, in order to eliminate the impact of price changes due to time, the price index was used to adjust the price during the accounting process.

Table 1. Accounting index system and accounting method of GEP in Fujian Province.

Function Category	Accounting Indicators	Accounting Formula	Formula Description	References
Ecosystem product supply	Agriculture, forestry, livestock, and fishery products	Value quantity according to the statistical research survey data		[15,22,23]
	Fresh water supply	$V_m={\displaystyle \sum _{i=1}^4\left(E_i\times P_i\right)}$	$V_m$ is the value of fresh water supplied (Yuan/year), $E_i$ is the water consumption of the ith type, $P_i$ is the market unit price of water consumption for the ith type.	[18]
Ecosystem regulation service	Water conservation	$\begin{array}{l}{Q}_{wr}=\left(U{Q}_w-T{Q}_w\right)+\\ \left(L{Q}_w-E{Q}_w\right)\left(1-\delta \right)\\ V_{wr}=Q_{wr}\times C_{we}\end{array}$	$V_{wr}$ is the water conservation value (Yuan/year), $Q_{wr}$ is the volume of water conservation (m3/year), $C_{we}$ is the construction cost per unit of reservoir capacity (Yuan/m3), $U{Q}_w$ is the volume of water used within the accounting area (m3/year), $T{Q}_w$ is the volume of net water transfer across the study area (m3/year), $L{Q}_w$ is the water volume flowing out of the region (m3/year), $E{Q}_w$ is the water volume inflowing into the region (m3/year), $\delta$ is the runoff coefficient of the study area.	[18,24,25]
	Soil conservation	$\begin{array}{c}{Q}_{sr}=R\times K\times L\times S\times \left(1-F\times P\right)\\ V_{dpd}={\displaystyle \sum _{i=1}^2{Q}_{sr}\times C_i\times P_i}\\ V_{sd}=\lambda \times \frac{Q_{s\mathrm{r}}}{{\rho }_1}\times C\\ V_{sr}=V_{sd}+V_{dpd}\end{array}$	$V_{sr}$ is the value of ecosystem soil conservation (Yuan/year), $V_{sd}$ is the value of reducing sedimentation (Yuan/year) $V_{dpd}$ is the value of reducing surface source pollution (Yuan/year), $\lambda$ is the coefficient of sedimentation, $Q_{sr}$ is the weight of soil conservation (ton/year), ${\rho }_1$ is the soil capacity (ton/m3), $C$ is the project cost of dredging the unit reservoir (Yuan/m3), $C_i$ is the pure content of a pollutant (nitrogen or phosphorus) in the soil (%), $P_i$ is the unit treatment cost of a pollutant (nitrogen or phosphorus) (Yuan/ton), $R$ is the rainfall erosion force factor, $K$ is the soil erodibility factor, $L$ is the slope length factor, $S$ is the slope factor, $F$ is the vegetation cover factor, $P$ is the soil conservation factor.	[20,26,27]
	Flood control	$\begin{array}{c}{C}_{fm}=C_{rfm}+C_{lfm}\\ C_{rfm}=C_t\times 0.35\\ C_{lfm}=e^{4.924}\times A^{1.128}\times 3.19\\ V_{fm}=C_{fm}\times C_{we}\end{array}$	$V_{fm}$ is the value of ecosystem flood control (Yuan/year), $C_{fm}$ is the volume of flood water control (m3/year), $C_{we}$ is the construction cost per unit of reservoir capacity (Yuan/m3), $C_{rfm}$ is the flood control volume of reservoirs (m3/year), $C_{lfm}$ is the flood control volume of lakes (m3/year), $C_t$ is the total reservoir capacity (10,000 m3), $A$ is the area of the lakes (km2).	[20,28]
	Coastal zone protection	$V_{cl}=D_{cl}\times C_{cl}$	$V_{cl}$ is the value of coastal zone protection (Yuan/year), $D_{cl}$ is the area of offshore and coastal wetlands (ha/year);$C_{cl}$ is the unit price of wave elimination and siltation promotion for offshore and coastal wetlands (Yuan/ha).	[11,29,30]
	Carbon sequestration	$\begin{array}{c}{V}_c=V_{c_1}+V_{c_2}\\ V_c{}_{{}_1}=A\times K_s\times C_c\\ V_{c_2}={\displaystyle \sum _{i=1}^{n}NP{P}_i}\times \left(1.62\times C_c\right)\end{array}$	$V_c$ is the carbon sequestration value (Yuan/year), $V_{c_1}$ is the value of carbon sequestration on farmlands (Yuan/year), $V_{c_2}$ is the value of carbon sequestered by each ecosystem types in the study (Yuan/year), $A$ is the total area of the farmlands (ha), $K_s$ is the rate of carbon sequestration in agricultural soils (ton/(ha year)), $C_c$ is the market price of carbon sequestration (Yuan/ton), $NP{P}_i$ is the net primary productivity of the ith ecosystem type.	[19,31]
	Oxygen supply	$V_o={\displaystyle \sum _{i=1}^{n}NP{P}_i\times \left(1.2\times C_o\right)}$	$V_o$ is the value of oxygen supply (Yuan/year), $C_o$ is the market oxygen transaction price (Yuan/ton).	[19]
	Air purification	$\begin{array}{c}{Q}_{api}={\displaystyle \sum _{i=1}^n{\displaystyle \sum _{j=1}^m{Q}_{ij}\times A_j}}\\ V_{ap}={\displaystyle \sum _{i=1}^n{Q}_{api}\times P_i}\end{array}$	$V_{ap}$ is the value of ecosystem air purification (Yuan/year), $Q_{api}$ is the quantity of the ith air pollutant cleaned (ton/year), $P_i$ is the cost required to treat the ith air pollutant (Yuan/ton), $Q_{ij}$ is the quantity of the ith atmospheric pollutant absorbed by the jth type of ecosystem per unit area (ton/(km2·year)), $A_j$ is the area of the jth ecosystem type (km2).	[10,32,33,34,35,36,37,38,39,40]
	Water purification	$\begin{array}{c}{Q}_{wpi}={\displaystyle \sum _{i=1}^2{Q}_i\times A}\\ V_{wp}={\displaystyle \sum _{i=1}^2{Q}_{wpi}}\times P_i\end{array}$	$V_{wp}$ is the value of ecosystem water purification (Yuan/year), $Q_{wpi}$ is the quantity of the ith water pollutant cleaned (ton/year), $P_i$ is the cost required to treat the ith water pollutant (Yuan/ton), $Q_i$ is the amount of the ith water pollutant purified by the ecological land per unit area (ton/km2), $A$ is the water area of the ecosystems (km2).	[10,41,42]
	Climate regulation	$\begin{array}{c}{E}_{tt}=E_{pt}+E_{we}\\ E_{pt}=\frac{{\displaystyle \sum _{i=1}^{3}GPP\times S_i\times d}}{3600\times R}\\ E_{we}=\frac{E_w\times q\times {10}^3}{3600}+E_w\times \gamma \\ V_{tt}=E_{tt}\times C_{tt}\end{array}$	$V_{tt}$ is the value of ecosystem climate regulation (Yuan/year), $E_{tt}$ is the total amount of energy consumed by the ecosystem during transpiration and evaporation (kWh/year), $C_{tt}$ is the average electricity price in 2020 (Yuan/kWh), $E_{pt}$ is the amount of energy consumed by the plant during transpiration (kWh/year), $E_{we}$ is the total amount of energy consumed in the evaporation process of water (kWh/year), $GPP$ is the amount of energy consumed by transpiration per unit area of each type of ecosystem (kJ/(ha day)), $S_i$ is the area of the ith ecosystem type (m2), $d$ is the number of days in a year that the air conditioner is turned on (day), $R$ is the ratio of the energy conversion efficiency of the air conditioner, $E_w$ is the volume of water consumed by evaporation (m3), $q$ is the water evaporation potential (J/g), $\gamma$ is the amount of energy required for the humidifier to convert 1 m3 of water into steam.	[10,43,44]
	Pest and disease control	$\begin{array}{c}{V}_{pc}=V_{fpc}+V_{gpc}\\ V_{fpc}=S_f+C_{fpc}\\ V_{gpc}=S_g+C_{gpc}\end{array}$	$V_{pc}$ is the value of pest and disease control (Yuan/year), $V_{fpc}$ is the value of forest pest and disease control (Yuan/year), $V_{gpc}$ is the value of grassland pest and disease control (Yuan/year), $S_f$ is the forest area (km2), $S_g$ is the area of the grasslands (km2), $C_{fpc}$ is the cost of forest pest and disease control per unit area (Yuan/km2), $C_{gpc}$ is the cost of grassland pest and disease control per unit area (Yuan/km2).	[20]
Ecosystem culture service	Leisure travel	$\begin{array}{c}{V}_r=V_{cc}+V_{cs}\\ V_{cc}=V_{tc}+V_{tv}\\ V_{cs}=40\%\times V_{tc}\\ V_{tc}=H\times W\end{array}$	$V_r$ is the value of leisure travel (Yuan/year), $V_{cc}$ is consumer expenditure (Yuan/year), $V_{cs}$ is consumer surplus (Yuan/year), $V_{tc}$ is the cost of travel (Yuan/year), $V_{tv}$ is the value of travel time (Yuan/year), $H$ is the length of the tourist tour (hour), $W$ is the wage rate (Yuan/hour).	[12,20]

Table 1. Accounting index system and accounting method of GEP in Fujian Province.

Function Category	Accounting Indicators	Accounting Formula	Formula Description	References
Ecosystem product supply	Agriculture, forestry, livestock, and fishery products	Value quantity according to the statistical research survey data		[15,22,23]
	Fresh water supply	$V_m={\displaystyle \sum _{i=1}^4\left(E_i\times P_i\right)}$	$V_m$ is the value of fresh water supplied (Yuan/year), $E_i$ is the water consumption of the ith type, $P_i$ is the market unit price of water consumption for the ith type.	[18]
Ecosystem regulation service	Water conservation	$\begin{array}{l}{Q}_{wr}=\left(U{Q}_w-T{Q}_w\right)+\\ \left(L{Q}_w-E{Q}_w\right)\left(1-\delta \right)\\ V_{wr}=Q_{wr}\times C_{we}\end{array}$	$V_{wr}$ is the water conservation value (Yuan/year), $Q_{wr}$ is the volume of water conservation (m3/year), $C_{we}$ is the construction cost per unit of reservoir capacity (Yuan/m3), $U{Q}_w$ is the volume of water used within the accounting area (m3/year), $T{Q}_w$ is the volume of net water transfer across the study area (m3/year), $L{Q}_w$ is the water volume flowing out of the region (m3/year), $E{Q}_w$ is the water volume inflowing into the region (m3/year), $\delta$ is the runoff coefficient of the study area.	[18,24,25]
	Soil conservation	$\begin{array}{c}{Q}_{sr}=R\times K\times L\times S\times \left(1-F\times P\right)\\ V_{dpd}={\displaystyle \sum _{i=1}^2{Q}_{sr}\times C_i\times P_i}\\ V_{sd}=\lambda \times \frac{Q_{s\mathrm{r}}}{{\rho }_1}\times C\\ V_{sr}=V_{sd}+V_{dpd}\end{array}$	$V_{sr}$ is the value of ecosystem soil conservation (Yuan/year), $V_{sd}$ is the value of reducing sedimentation (Yuan/year) $V_{dpd}$ is the value of reducing surface source pollution (Yuan/year), $\lambda$ is the coefficient of sedimentation, $Q_{sr}$ is the weight of soil conservation (ton/year), ${\rho }_1$ is the soil capacity (ton/m3), $C$ is the project cost of dredging the unit reservoir (Yuan/m3), $C_i$ is the pure content of a pollutant (nitrogen or phosphorus) in the soil (%), $P_i$ is the unit treatment cost of a pollutant (nitrogen or phosphorus) (Yuan/ton), $R$ is the rainfall erosion force factor, $K$ is the soil erodibility factor, $L$ is the slope length factor, $S$ is the slope factor, $F$ is the vegetation cover factor, $P$ is the soil conservation factor.	[20,26,27]
	Flood control	$\begin{array}{c}{C}_{fm}=C_{rfm}+C_{lfm}\\ C_{rfm}=C_t\times 0.35\\ C_{lfm}=e^{4.924}\times A^{1.128}\times 3.19\\ V_{fm}=C_{fm}\times C_{we}\end{array}$	$V_{fm}$ is the value of ecosystem flood control (Yuan/year), $C_{fm}$ is the volume of flood water control (m3/year), $C_{we}$ is the construction cost per unit of reservoir capacity (Yuan/m3), $C_{rfm}$ is the flood control volume of reservoirs (m3/year), $C_{lfm}$ is the flood control volume of lakes (m3/year), $C_t$ is the total reservoir capacity (10,000 m3), $A$ is the area of the lakes (km2).	[20,28]
	Coastal zone protection	$V_{cl}=D_{cl}\times C_{cl}$	$V_{cl}$ is the value of coastal zone protection (Yuan/year), $D_{cl}$ is the area of offshore and coastal wetlands (ha/year);$C_{cl}$ is the unit price of wave elimination and siltation promotion for offshore and coastal wetlands (Yuan/ha).	[11,29,30]
	Carbon sequestration	$\begin{array}{c}{V}_c=V_{c_1}+V_{c_2}\\ V_c{}_{{}_1}=A\times K_s\times C_c\\ V_{c_2}={\displaystyle \sum _{i=1}^{n}NP{P}_i}\times \left(1.62\times C_c\right)\end{array}$	$V_c$ is the carbon sequestration value (Yuan/year), $V_{c_1}$ is the value of carbon sequestration on farmlands (Yuan/year), $V_{c_2}$ is the value of carbon sequestered by each ecosystem types in the study (Yuan/year), $A$ is the total area of the farmlands (ha), $K_s$ is the rate of carbon sequestration in agricultural soils (ton/(ha year)), $C_c$ is the market price of carbon sequestration (Yuan/ton), $NP{P}_i$ is the net primary productivity of the ith ecosystem type.	[19,31]
	Oxygen supply	$V_o={\displaystyle \sum _{i=1}^{n}NP{P}_i\times \left(1.2\times C_o\right)}$	$V_o$ is the value of oxygen supply (Yuan/year), $C_o$ is the market oxygen transaction price (Yuan/ton).	[19]
	Air purification	$\begin{array}{c}{Q}_{api}={\displaystyle \sum _{i=1}^n{\displaystyle \sum _{j=1}^m{Q}_{ij}\times A_j}}\\ V_{ap}={\displaystyle \sum _{i=1}^n{Q}_{api}\times P_i}\end{array}$	$V_{ap}$ is the value of ecosystem air purification (Yuan/year), $Q_{api}$ is the quantity of the ith air pollutant cleaned (ton/year), $P_i$ is the cost required to treat the ith air pollutant (Yuan/ton), $Q_{ij}$ is the quantity of the ith atmospheric pollutant absorbed by the jth type of ecosystem per unit area (ton/(km2·year)), $A_j$ is the area of the jth ecosystem type (km2).	[10,32,33,34,35,36,37,38,39,40]
	Water purification	$\begin{array}{c}{Q}_{wpi}={\displaystyle \sum _{i=1}^2{Q}_i\times A}\\ V_{wp}={\displaystyle \sum _{i=1}^2{Q}_{wpi}}\times P_i\end{array}$	$V_{wp}$ is the value of ecosystem water purification (Yuan/year), $Q_{wpi}$ is the quantity of the ith water pollutant cleaned (ton/year), $P_i$ is the cost required to treat the ith water pollutant (Yuan/ton), $Q_i$ is the amount of the ith water pollutant purified by the ecological land per unit area (ton/km2), $A$ is the water area of the ecosystems (km2).	[10,41,42]
	Climate regulation	$\begin{array}{c}{E}_{tt}=E_{pt}+E_{we}\\ E_{pt}=\frac{{\displaystyle \sum _{i=1}^{3}GPP\times S_i\times d}}{3600\times R}\\ E_{we}=\frac{E_w\times q\times {10}^3}{3600}+E_w\times \gamma \\ V_{tt}=E_{tt}\times C_{tt}\end{array}$	$V_{tt}$ is the value of ecosystem climate regulation (Yuan/year), $E_{tt}$ is the total amount of energy consumed by the ecosystem during transpiration and evaporation (kWh/year), $C_{tt}$ is the average electricity price in 2020 (Yuan/kWh), $E_{pt}$ is the amount of energy consumed by the plant during transpiration (kWh/year), $E_{we}$ is the total amount of energy consumed in the evaporation process of water (kWh/year), $GPP$ is the amount of energy consumed by transpiration per unit area of each type of ecosystem (kJ/(ha day)), $S_i$ is the area of the ith ecosystem type (m2), $d$ is the number of days in a year that the air conditioner is turned on (day), $R$ is the ratio of the energy conversion efficiency of the air conditioner, $E_w$ is the volume of water consumed by evaporation (m3), $q$ is the water evaporation potential (J/g), $\gamma$ is the amount of energy required for the humidifier to convert 1 m3 of water into steam.	[10,43,44]
	Pest and disease control	$\begin{array}{c}{V}_{pc}=V_{fpc}+V_{gpc}\\ V_{fpc}=S_f+C_{fpc}\\ V_{gpc}=S_g+C_{gpc}\end{array}$	$V_{pc}$ is the value of pest and disease control (Yuan/year), $V_{fpc}$ is the value of forest pest and disease control (Yuan/year), $V_{gpc}$ is the value of grassland pest and disease control (Yuan/year), $S_f$ is the forest area (km2), $S_g$ is the area of the grasslands (km2), $C_{fpc}$ is the cost of forest pest and disease control per unit area (Yuan/km2), $C_{gpc}$ is the cost of grassland pest and disease control per unit area (Yuan/km2).	[20]
Ecosystem culture service	Leisure travel	$\begin{array}{c}{V}_r=V_{cc}+V_{cs}\\ V_{cc}=V_{tc}+V_{tv}\\ V_{cs}=40\%\times V_{tc}\\ V_{tc}=H\times W\end{array}$	$V_r$ is the value of leisure travel (Yuan/year), $V_{cc}$ is consumer expenditure (Yuan/year), $V_{cs}$ is consumer surplus (Yuan/year), $V_{tc}$ is the cost of travel (Yuan/year), $V_{tv}$ is the value of travel time (Yuan/year), $H$ is the length of the tourist tour (hour), $W$ is the wage rate (Yuan/hour).	[12,20]

2.4. Standard Deviational Ellipse

The standard deviational ellipse (SDE) is a spatial statistical method used to quantitatively describe the overall characteristics of the spatial distribution of geographic elements based on the spatial location and spatial structure of the study object [45]. The center of gravity migration trajectory model can express spatio-temporal changes based on the weighted geographic elements [46]. In this paper, we used the SDE generated by the first-level standard deviation to analyze the directional distribution characteristics of GEP. The long and short axes of the ellipse were used to show the distribution range of GEP, and the center point and directional angle were used to identify the center of gravity and distribution direction of GEP, respectively. A larger difference between the long and short axes reflects stronger directionality in the GEP distribution, and vice versa. The elliptical center of gravity, standard deviation of long and short semi-axes, and azimuth were calculated as follows [47]:

$$N\left(X,Y\right)=\left({\displaystyle {\sum }_{i=1}^n{W}_i{X}_i}/{\displaystyle {\sum }_{i=1}^n{W}_i},{\displaystyle {\sum }_{i=1}^n{W}_i{Y}_i}/{\displaystyle {\sum }_{i=1}^n{W}_i}\right)$$

(1)

$${\delta }_x=\sqrt{\frac{{\displaystyle {\sum }_{i=1}^n(W_i\overline{X_i}\cos \theta }-W_i\overline{Y_i}\sin \theta {)}^2}{{\displaystyle {\sum }_{i=1}^n{W}_i^2}}}$$

(2)

$${\delta }_y=\sqrt{\frac{{\displaystyle {\sum }_{i=1}^n(W_i\overline{X_i}\sin \theta }-W_i\overline{Y_i}\cos \theta {)}^2}{{\displaystyle {\sum }_{i=1}^n{W}_i^2}}}$$

(3)

$${\theta }_{i-j}=n\pi /2+\arctan \left[\left(Y_i-Y_j\right)/\left(X_i-X_j\right)\right]$$

(4)

where, n is the number of prefecture-level cities, (X, Y) are the weighted average center of gravity coordinates, $W_i$ is the weight; ${\delta }_x$ and ${\delta }_y$ represent the standard deviation on the x-axis and y-axis, respectively; $\theta$ is the azimuth of the SDE, i.e., the angle between the clockwise rotation in the due north direction and the long axis of the SDE.

2.5. The Panel Vector Auto Regression Model

Because of the complex interrelationships and action mechanisms of the factors contained in GEP and GDP, this study selected a panel vector auto regression (PVAR) model to explore the interactive response relationship between GEP and GDP. The PVAR model is an organic combination of a vector auto regression model and a time series model, which combines the advantages of both and can effectively avoid the endogeneity problem [48]. On the basis of passing the smoothness test, a generalized moments estimation was used as the first method to estimate the model parameters, and then the interaction strength of the two was analyzed by impulse response. Finally, variance decomposition was used to examine the contribution of each variable to the prediction variance to clarify the degree of influence and effectiveness of the two [49]. The specific model setup was as follows.

$$Y_{it}={\gamma }_0+{\displaystyle {\sum }_{j=1}^k{\gamma }_j}{Y}_{it-j}+{\alpha }_i+{\beta }_i+{\mu }_{it}$$

(5)

where i represents the cities in the Yangtze River Delta region, t represents the year, $k$ is the lag order, ${\gamma }_0$ denotes the intercept term vector, $Y_{it}$ is the model explanatory variables, including GEP and GDP, and ${\gamma }_j$ denotes the parameter matrix at lag order j. ${\alpha }_i$ denotes the individual effect, ${\beta }_i$ denotes the time effect, and ${\mu }_{it}$ is the random disturbance term.

3. Results

3.1. Changes in GEP

The GEP of Fujian Province and the changes in the value of product supply, the regulating services, and the cultural services in the period of 2000 to 2020 are shown in Figure 2. The GEP of Fujian Province showed an upward trend from 2000 to 2019 and a decline in 2020 owing to the COVID-19 pandemic, with an overall growth rate of 27.90% (CNY 1001.205 billion). The results indicated that the ecological products and services provided by ecosystems in Fujian Province were valuable and had the potential to be transformed from ecological to economic advantages.

In comparing the three major indicators in the graph with the change line, it can be seen that the product supply value showed a stable growth trend from 2000 to 2020, with an increase from CNY 174.464 billion in 2000 to CNY 551.143 billion in 2020 (a growth rate of 215.91%). The value of regulating services presented irregular changes with an overall decline of 1.91% (CNY 64.292 billion), and the smallest and largest values occurred in 2020 (CNY 3307.864 billion) and 2006 (CNY 3451.691 billion), respectively. Benefiting from the positive influence of the global economic development level and the continuous improvement of local eco-culture, the number of tourists increased year by year until 2020, with the value of cultural services growing from CNY 47.237 billion in 2000 to a peak of CNY 1116.59 billion in 2019. Unfortunately, the COVID-19 pandemic has had a huge negative impact on the global economy since 2020, directly leading to a reduction in the value of cultural services. This is the reason why the total amount of GEP reached its peak in 2019. During the period from 2000 to 2020, the largest contribution to GEP was always the value of regulating services. From 2000 to 2013, the contribution of the product supply value to GEP was greater than the contribution of the cultural services value, but the value of cultural services showed a high growth trend relative to the value of product supply. Since 2014, the contribution of the value of cultural services to GEP was larger than the contribution of the value of product supply.

3.2. Spatial Pattern of GEP

On the basis of the GEP accounting results of Fujian Province from 2000 to 2020, the custom grading method of ArcGIS was used to divide the GEP values of each prefecture-level city into eight grades over a period of 21 years to show the spatial distribution characteristics of GEP. The years 2000, 2005, 2010, 2015, and 2020 were selected as time points, and the SDE model was used to analyze the spatial and temporal changes in GEP in Fujian Province over 21 years. The results are shown in Figure 3 and

Table 2. Standard deviation elliptic parameters of GEP distribution in Fujian Province.

Years	CenterX/(°)	CenterY/(°)	XStdDist/km	YStdDist/km	Rotation/(°)
2000	117.981	26.083	0.966	1.564	39.361
2005	117.985	26.055	0.971	1.568	38.7551
2010	118.004	26.041	0.979	1.562	38.478
2015	118.028	26.015	0.988	1.557	37.409
2020	118.047	26.002	0.994	1.555	37.321

Note: CenterX denotes the longitudinal center of SDE; CenterY denotes the latitudinale center of SDE; XStdDist denotes the major axis of SDE; YStdDist denotes the minor axis of SDE; Area denotes the area of SDE; Rotation denotes the rotation of SDE.

.

From Figure 3, it can be seen that from 2000 to 2020, GEP in Fujian Province was generally higher in the northwest region than that in the southeast region, and, except for 2020, the GEP of each prefecture-level city had an overall growth trend. The GEP of each prefecture-level city was compared and, over 21 years, the GEP values were ranked as follows: Nanping > Sanming > Longyan > Fuzhou, Ningde > Zhangzhou > Quanzhou > Putian and Xiamen. The GEP value for Ningde was greater than that for Fuzhou until 2011, and the GEP value for Putian was greater than that for Xiamen until 2014. Figure 3 and Table 2 also show that the center of gravity of the GEP value distribution tended to move from north to south and from west to east over the 21 years of the study, but the center of gravity was still within Sanming city.

3.3. The Interactive Response Relationship between GEP and GDP

Both the impulse response function and variance decomposition methods were used to analyze the dynamic relationship between GEP and GDP based on the PVAR model. To avoid the pseudo-regression phenomenon, the Levin–Lin–Chu test was first used to perform a unit root test for GEP and GDP, and the model passed the 5% significance level test to determine that the data were smooth. Then, the optimal lag order of the model was determined to be 1 for all criteria, Akaike Information Criterion (AIC), Bayesian Information Criterion (BIC), and Hannan–Quinn Criterion (HQIC) (

Table 3. Results of hysteresis order selection.

Lagging Order	AIC	BIC	HQIC
1	26.0472 *	26.4514 *	26.2112 *
2	26.5477	27.0433	26.7489
3	30.4625	31.0567	30.7039
4	26.3887	27.0899	26.6737
5	26.6773	27.4951	27.0096

Note: * denotes the optimal lag order chosen in this criterion. AIC, BIC, and HQIC are all measures of goodness of fit for statistical models, providing a trade-off between the complexity of the estimated model and the goodness of fit data.

).

The systematic generalized method of moments approach of the PVAR model was used to estimate the interaction between GEP and GDP. The results showed that GEP with a 1-year lag had a positive effect on itself and on GDP, while GDP with a 1-year lag had a positive effect on itself but had no significant effect on GEP (

Table 4. GMM estimation results of the Panel Vector Auto Regression model.

Explanatory Variables	Explained Variables
	GEP	GDP
L GEP	1.115841 ***	0.6878061 ***
L GDP	−0.0584083	0.7088464 ***

Note: The superscript *** represents significant at 1% significance level.

).

Further analysis of the interaction response relationship between GEP and GDP, and the results are shown by the impulse response function plots (Figure 4). Specifically, GEP had a relatively stable positive response to its own shock; GDP had a positive response to GEP’s shock, and the value of the response became larger with a time lag. GDP had an initial positive response to its own shock, which became negative over time. Overall, the long-term contribution of GEP to GDP was positive.

Variance decomposition evaluates the relative importance of variables through the contributions of the structural shocks of different variables. The results of the variance decomposition of the GEP and GDP models are shown in

Table 5. Variance decomposition of prediction errors.

Lagging Order	Impact Variables
	Forecast Period	GEP	GDP
GEP	5	0.978	0.022
GDP	5	0.816	0.184
GEP	10	0.948	0.052
GDP	10	0.915	0.085
GEP	15	0.929	0.071
GDP	15	0.916	0.084
GEP	20	0.916	0.084
GDP	20	0.907	0.093

. From period 10 onward, the variance decomposition values of GEP and GDP were similar, which implied that the mutual contributions of the variables were stable after period 10. The prediction results in period 20 showed that in the panel variable GEP equation, the explanation of GEP for itself was 91.6%, and the explanation strength of GDP was only 8.4%. This indicated that the improvement of GEP was a long-term process and that GDP had less influence on GEP in the short term. In the panel variable GDP equation, the explanation strength of GEP to GDP was 90.7%, which indicated that GEP was very important for GDP changes.

4. Discussion

4.1. Determinant Factors Affecting GEP

GEP is an integrated value representing ecosystem services, and, therefore, the magnitude of GEP changes with the evolution of ecosystem services. The evolution of ecosystem services is a complex process driven by multiple factors, which can be divided into two categories—natural and anthropogenic factors—that affect ecosystem services by changing the structure and function of ecosystems [50,51]. Typically, natural factors have a more stable influence on ecosystem services, while anthropogenic factors have a more dynamic effect [52]. Changes in ecosystem services are the result of changes in ecological processes owing to the reconfiguration of ecological resources, while at certain scales, ecological resources are reconfigured through changes in land use practices [53]. Ecosystem services are also influenced by land use patterns. The distribution of ecological resources under the influence of different land use patterns varies significantly [54,55], and thus ecological services under complex land use patterns can show significant spatial differences owing to different environmental conditions [56].

4.2. Sustainable Development Recommendations for the Study Area

Our findings showed that among the three major service values of GEP in Fujian Province from 2000 to 2020, the value of regulation services has always remained the largest, and the province’s GEP as a whole showed a higher distribution characteristic in the northwest than in the southeast. We found that this is related to the distribution characteristics of land use types and ecosystem types. This is consistent with the findings of Wang et al. [57] and Zhou et al. [58]. In response to the following results, we make several suggestions for the sustainable development of Fujian Province.

For sustainable development to occur, it is important that rational planning of land use and coordination of economic development and ecological protection be considered. This requires government guidance and regulation, combined with the ecological protection red line, to establish the maximum boundary for the development area to avoid excessive damage to the environment. In addition, before development, a development method and construction content that are suitable for the area should be selected scientifically, and the original ecosystems should be protected as much as possible during the development and construction processes.

According to the differences in the available resources of each region, different development plans need to be formulated. For regions where the supply of ecological products is dominant, an efficient unified management platform for ecological resources can be established to improve the supply value of ecosystem products. For regions where the supply of ecological regulating services is dominant, reward and penalty policies related to environmental quality should be improved to raise awareness of the need to optimize ecological environment quality. Regions with a dominant supply of cultural services should shape brands with regional characteristics to improve the realization of ecological and cultural values. The comparative advantage of production from ecosystems in the region should be fully utilized to improve the overall GEP.

A better understanding of the value of ecological products is an effective way to manage the conflict between protection and development. The market value of products depends on the regulation of market mechanisms. For tangible ecological products or intangible services that are suitable for trading, market resources such as technology, capital, and management should be introduced to promote the diversification of ecological products and achieve a higher value-added supply. For ecological products that are not suitable for trading, better protection and restoration are emphasized to maintain the structural and functional integrity of the ecosystem and fully realize its service value.

Establishing a dual accounting and assessment mechanism for regional GEP and GDP will have a significant positive effect on the implementation of the policy. According to the differences in the natural endowments of each region, a differentiated assessment mechanism will provide important institutional support for sustainable development.

4.3. Contributions and Disadvantages of the Study

It is of great significance to study the accounting of the gross ecosystem product of Fujian Province for 21 consecutive years. The first is to deepen the relationship between the ecosystem and the economy and society and provide scientific data support for the development and protection of ecological resources in Fujian Province. Secondly, it not only excavates the overall ecosystem economic value of Fujian Province but also provides reference and thinking for the ecological management and ecological economic development of Fujian Province. Thirdly, this study not only enriches the management research on the ecological value of the ecosystem in Fujian Province but also provides new ideas and methods for ecosystem research in other regions.

However, there were some shortcomings in this study. Due to the availability of data, the accounting scope of this study was somewhat narrow, and the accounting results were relatively conservative. Therefore, the driving factors of GEP and GDP changes in Fujian Province were not studied in depth, and the effectiveness of relevant policies may not be accurately assessed. In addition, the current study used the static data measurement of ecosystem GDP at a single level and did not explore the dynamic transformation relationship between GEP and GDP.

4.4. Future Directions for GEP Research

GEP has received increasing attention worldwide as a comprehensive monetary indicator for evaluating final ecosystem services. In China, a series of regional economic accounting pilots have been actively pursued, and the first government economic accounting system has been established preliminarily, aiming to incorporate GEP accounting into real-world decision making. However, there are still a number of issues in current GEP accounting practice that prevent GEP from being measured accurately in the short term, including these five main areas: inconsistent understanding of ecosystem services, missing services in GEP, overestimation of service values, insufficient accuracy in physical quantification, and high uncertainty in valuation [59]. According to GEP’s accounting, which provides scientific guidance for regional and even global development, further exploration is still needed.

On the one hand, the accounting technology of GEP should be improved. On the basis of existing research results and practical experience, the GEP accounting method system and data collection standards need to be further unified to form a more standardized GEP data acquisition system and enhance the reliability of research. On the other hand, the driving factors of GEP dynamics should be deeply analyzed, and the interaction mechanism between GEP and GDP also should be explored further, which is critical to providing more scientific references for coordinating eco-economic development and regional sustainable development goals.

5. Conclusions

The spatial and temporal evolution of the GEP of Fujian Province from 2000 to 2020 was analyzed and the relationship between GEP and GDP changes over 21 years was compared, leading to the following conclusions.

(1) GEP in Fujian Province showed an overall growth trend with a growth rate of 27.9% (CNY 1001.205 billion). In terms of the trend of three major indicators, the value of product supply showed a continuous growth trend, the value of regulating services showed irregular changes, while the value of cultural services showed a growth trend until 2019 and a decline in 2020. The contribution of the value of regulating services to GEP was always the largest among the three indicators, and the contribution of the value of cultural services to GEP exceeded the value of product supply after 2014.

(2) The overall spatial distribution of GEP in Fujian Province was characterized by a higher distribution in the northwest than in the southeast. Comparing the prefecture-level cities, the GEP value of Nanping remained the largest, followed by Sanming and Longyan, while Putian and Xiamen remained the smallest. During the study period, the center of gravity of the distribution of GEP values tended to move from north to south and from west to east but remained within Sanming.

(3) The growth rate of GDP was relatively faster than the growth rate of GEP, and the difference between GEP and GDP decreased. In addition, the study found that GEP had a positive effect on GDP in the long run, while GDP had a smaller effect on GEP in the short run.