Spatial and Temporal Dynamic Evolution and Correlation of Ecological Quality and Ecosystem Service Value in Fujian Province

Abstract

To promote sustainable development and improve human well-being, understanding how ecological quality influences ecosystem service value is essential. In this study, we quantified the ecological quality and ecosystem service value in Fujian Province from 2000 to 2021 using the Remote-Sensing Ecological Index (RSEI) and the Equivalent Factor Approach, respectively. This analysis aimed to examine their spatial and temporal dynamic evolution and to explore the correlation between the two. The results indicate that the temporal and spatial patterns of ecological quality and ecosystem service value in Fujian Province from 2000 to 2021 were similar, with both showing fluctuating upward trends over time. The spatial distributions were high in central and northern Fujian and low in the southeastern coastal areas. The Pearson correlation analysis revealed reveals a significant positive correlation between the RSEI and ecosystem service value (r = 0.673, p < 0.01), suggesting a synergistic relationship. The highest correlation was observed between the supporting service value and the RSEI indicators (|r| = 0.449–0.815, p < 0.01), and between the NDVI and the supporting service value (r = 0.815, p < 0.01). The spatial autocorrelation analysis also showed that ecological quality and ecosystem service value were positively correlated spatially, with high–high agglomerations concentrated in northwest and central Fujian, and low–low agglomerations in the coastal area of south Fujian.

1. Introduction

Ecological quality (EQ) is a comprehensive evaluative measure of ecological health, stability, and functionality [1,2]. In EQ evaluations, we explore the relationship between ecosystems and human activities by describing both positive and negative attributes of a given ecosystem [3,4]. Therefore, evaluating and understanding the regional ecological status is essential for conserving and managing environmental resources [2,5]. In addition to providing comprehensive, rapid, and extensive information about the Earth’s surface, remote sensing is widely used [6,7]. Xu Hanqiu suggested a remote sensing-based Ecological Index (RSEI) in 2013, which provides a new approach to evaluating EQ [8]. In addition to integrating four easily accessible ecological evaluation indicators, namely green, dry, wet and heat, the RSEI also uses a Principal Component Analysis (PCA) to automatically and objectively weight each indicator. There have been numerous studies using this method to quantify and accurately evaluate EQ [9], such as those conducted in watersheds [10,11,12], mining areas [13,14], wetlands [15,16], cities [17,18,19,20], etc.

Ecosystem services refer to the variety of direct or indirect benefits and functions provided by ecosystems to humans, which can be categorized into supporting services, regulating services, provisioning services and cultural services [21]. Currently, the assessment methods for ecosystem services can be divided into physical quality assessment methods, value assessment methods and energy assessment methods [22]. In contrast, Xie Gao Di et al. proposed an equivalent factor method for calculating the integrated ecosystem service value (ESV) using the initial ESV per unit area [23]. This method allows for comparison between individual ecosystem service functions, as well as with calculations of the gross regional product. Furthermore, it has been widely used to describe changes in ecosystem services under anthropogenic disturbances in the region after locational corrections due to its ease of data collection and suitability for studies with large areas and long periods [24,25].

Ecosystem services are intimately connected to EQ, and numerous scholars have examined EQ and ESV independently. For example, Menghao Yang et al. investigated the spatial and temporal evolution and correlation between ecosystem services and urbanization in the Yellow River Basin [26]. Jianwei Geng et al. analyzed the spatial and temporal changes in EQ and the driving factors behind these changes in Fuzhou City, utilizing the RSEI index [27]. However, the relationship between EQ and ESV across multiple time scales remains insufficiently explored. Costanza posited that ecosystem health is a critical ecological indicator and that robust ecosystems can sustain ecosystem services [28]. Maes et al. noted that the condition of ecosystems influences the delivery of multiple services, thereby enhancing human well-being [29]. Studies have indicated that higher quality ecological conditions lead to overall improvements in ESV [4,30]. Nonetheless, not all service types are correlated with the quality attributes of ecosystems. Qing Zhu et al. found that the spatial distribution of carbon storage, soil conservation and the Ecosystem Service Index (ESI) in the Pingjiang watershed mirrored that of EQ, while the pattern for food supply was counter to this trend [31]. Thus, comprehending the link between ecosystem EQ and ESV is essential for fostering the integrated natural and socioeconomic development of regional ecosystems.

In the southern region of China, Fujian Province is recognized as China’s first ecological civilization pilot area [32]. The strategic initiative to build an ecological province was launched in 2000. It was selected as the study area due to its rapid urbanization and advancements in eco-civilization construction. The specific objectives of this study were to (i) examine the characteristics of the spatial and temporal changes in EQ and major ESV across multiple time scales; (ii) explore the correlation and the spatial interaction stress relationship between EQ and ESV; and (iii) offer references and recommendations for the coordinated development of the ecological environment in Fujian Province.

2. Materials and Methods

2.1. Study Area

Fujian Province occupies a land area of 124,000 km² in southeast China, between 23°33′ N and 28°20′ N and 115°50′ E and 120°40′ E (Figure 1). Mountainous and hilly terrain dominates 80% of the province’s total area. As one of China’s six major forest regions, it boasts a forest coverage rate of over 66.80%, ranking first among the country’s provinces for 40 consecutive years. The province features a dense network of water systems, including four major rivers: the Jin River, the Min River, the Ting River and the Jiulong River. The region experiences an average annual temperature range of 17–21 °C, moderate humidity, and an average annual rainfall of 1400–2000 mm, which is characteristic of a subtropical maritime monsoon climate. There are significant regional differences in the climate, with the southern part of Fujian experiencing a southern subtropical climate and the northeastern, northern, and western parts experiencing a middle subtropical climate.

Fujian Province has high forest cover, dominated by Masson pine and fir. However, there are problems such as a tendency toward simplistic forest ecosystems and a decline in the function of biodiversity maintenance. Additionally, Fujian Province is a typical southern hilly area with red soil. Due to the high and concentrated precipitation, the underlying granite rock is relatively soft, making it very susceptible to soil erosion. It is the second most serious soil erosion area in China, second only to the Loess Plateau. Strong soil erosion and wind erosion lead to a serious loss of soil, including vital elements like aluminum and iron, resulting in land sandification and other environmental problems [33]. As the first national ecological civilization advanced demonstration area, studying the importance of ecological protection in Fujian Province is of great significance. It can provide a theoretical basis for implementing targeted ecological protection and achieving sustainable development [34,35].

2.2. Data Source

Considering the timing of ecological construction policies and the availability of research data, the study has chosen the period from 2000 to 2021 as the research timeframe. Based on preliminary experimental analysis, EQ assessment data for each year from 2000 to 2021 were obtained. We selected six years characterized by minimal cloud cover and a first principal component contribution rate greater than 70% following the PCA, ensuring similar time gaps between each selection. This strategy is conducive to enhancing the accuracy of the ecological quality study. This study utilized multi-source data, including spatial remote-sensing information and the statistical yearbook of Fujian Province, as detailed below:

(1)

Land cover data. The study used land cover data from Fujian Province for the years 2000, 2004, 2008, 2012, 2016, and 2021, which were produced by the team of Jie Yang and Xin Huang from Wuhan University based on Google Earth Engine using Landsat images with a spatial resolution of 30 m × 30 m. The data were used for the study [36].

(2)

The MOD09A1, MOD11A2, and MOD13A1 datasets provided by NASA were selected for this study to quantify the RSEI metrics, with a spatial resolution of 500 × 500 m. The datasets were used for the quantification of the RSEI metrics.

(3)

Correction of data related to factors such as ESV. Provided by the National Earth System Science Data Centre (http://www.geodata.cn/, accessed on 11 January 2024) and the Resource Environmental Science and Data Centre (http://www.resdc.cn/, accessed on 11 January 2024).

(4)

Other socio-economic data were from the Statistical Yearbook of Fujian Province (2000–2021) and the Statistical Yearbook of China (2000–2021).

2.3. Research Methods

2.3.1. Remote-Sensing Ecological Index Construction

The RSEI is a composite index that is used for the rapid monitoring of the ecological status based on remote-sensing data [37]. It is defined as the product of greenness, wetness, heat and dryness.

$$RSEI=f(Greenness,Wetness,Heat,Dryness)$$

(1)

where Greenness is the greenness component; in this study, we use the NDVI index from the MOD13A1 imagery. Heat is the thermal component and is characterized in this study using the Land Surface Temperature (LST) from the MOD11A2 imagery. Wetness and Dryness, on the other hand, are the humidity and dryness components, respectively. In this case, the humidity component was characterized using the third component, WET, of the MOD09A1 multispectral image after tassel-cap transformation [38,39]; the dryness component was then portrayed using the Normalized Difference Built-Up and Soil Index (NDBSI) constructed by Hu and Xu [3].

After calculating the four ecological components based on the MODIS data, PCA was utilized to synthesize these indicators, thereby avoiding biases from human subjective factors in the weight setting [40]. Given the non-uniform scale of these four components, it was necessary to standardize the indicators before performing PCA and to construct the RSEI using the first principal component. Furthermore, the first principal component values were standardized again to facilitate cross-sectional comparisons over the study period. This process is considered objective and reliable, as it does not require human intervention when calculating the RSEI [1,41]. Due to space limitations, the detailed calculation method can be found in the literature [8].

2.3.2. Quantification of Ecosystem Services

In this study, ecosystem services were selected according to the categories proposed by Costanza et al., with comprehensiveness (multiple types of ecosystem services), feasibility, and dominant function (natural environmental characteristics of the study area) serving as the selection principles [21]. Referring to the results from the eastern coastal cities of China and combined with the geographic environment characteristics, we selected a total of 4 first-level ecosystem services and 10 subtypes of ecosystem services [42,43,44].

Using the correction method based on farmland [45], a correction factor of 1.07 for the ecosystem service equivalence in Fujian Province was calculated. This calculation corrected the ecosystem value equivalence per unit area from the table provided by Xie Gao Di et al. [46], taking into account the average grain yields in the study area and the national average grain yields from 2000 to 2021. Concurrently, the equivalent factor of ecosystem service value was adjusted, using the grain yield and purchase price per unit area of arable land in Fujian Province in 2021 as a benchmark. Given that the economic value provided by the natural ecosystem is 1/7 of the economic value of food production from the unit area of cropland [47], and checking the official website of Fujian Food and Material Reserve Bureau about the purchase price of grain in 2021 is 2.48 CNY/kg, and given that the grain output per unit area of arable land in Fujian Province in 2021 was 5289.29 kg/hm², the calculation obtains the ecological service value factor of the economic value of the ecological service value factor, which was calculated to be 1873.92 CNY/hm². Notably, the ecosystem service value for construction land was set to 0. This approach allowed us to determine the ecosystem service value per unit area for different ecosystems in Fujian Province (

Table 1. Ecosystem service value per unit area in Fujian Province (CNY/hm²).

ESs		Cropland	Forest	Grassland	Water	Barren
Provision service (PSV)	Food production (FP)	2038.83	449.74	562.18	1499.14	0.00
	Raw materials (RM)	134.92	1021.29	833.90	431.00	0.00
	Subtotal	2173.75	1471.03	1396.07	1930.14	0.00
Regulating service (RSV)	Gas regulation (GR)	1664.04	3354.32	2913.95	1442.92	37.48
	Climate regulation (CR)	854.51	10,053.59	7711.19	4291.28	0.00
	Waste disposal (WD)	254.85	3007.64	2548.53	10,400.27	187.39
	Water regulation (WR)	4077.65	7580.01	5649.87	191,589.76	56.22
	Subtotal	6851.06	23,995.57	18,823.54	207,724.22	281.09
Support service (SSV)	Soil retention (SR)	14.99	4094.52	3551.08	1742.75	37.48
	Biodiversity protection (BP)	284.84	309.20	271.72	131.17	0.00
	Biological diversity (BD)	314.82	3729.10	3232.51	4778.50	37.48
	Subtotal	614.65	8132.82	7055.32	6652.42	74.96
Cultural service (CSV)	Aesthetic landscape (AL)	134.92	1639.68	1424.18	3541.71	18.74
Total ESV		9774.37	35,239.10	28,699.11	219,848.50	374.78

). The formula for valuing ecosystem services is:

$$ES{V}_i={\displaystyle \sum _{k=1}^n{A}_k}\times V{C}_i$$

(2)

$$E=\frac{1}{7}\times \sum _{f=1}^S\frac{m_f{q_{f}p}_f}{M}$$

(3)

$$ESV={\displaystyle \sum _{i=1}^{m}ES{V}_i}$$

(4)

where E is the economic value per unit of ecosystem service value quantity factor, f is the crop variety, p_f is the average price of crop f, q_f is the yield of crop f, m_f is the area of crop f, m is the total area of all crop types, s is the number of crop types, ESV_i and ESV are the individual ESV_s and total ESV_s, respectively, A_k is the area of land-use type k, and VC_i is the coefficient of ecosystem service value of the i-th land-use type.

2.3.3. Correlation Analysis

• Pearson correlation analysis

In this study, Pearson correlation analysis was used to reveal the relationship between different ESV indicators and RSEI indicators. When analyzing correlations between continuous data variables, Pearson correlation analysis tends to provide more consistent and better results than other correlation analysis methods (e.g., Spearman correlation analysis, Kendall correlation analysis) [48]. Therefore, it was used in this study to infer the interaction between ESV and EQ. The formula is as follows:

$${\mathrm{R}}_{xy}=\frac{{\displaystyle \sum _{i=1}^n(x_i-\overline{x})(y-\overline{y})}}{\sqrt{{\displaystyle \sum _{i=1}^n{(x_i-\overline{x})}^2}}\sqrt{{\displaystyle \sum _{i=1}^n{(y_i-\overline{y})}^2}}}$$

(5)

where R_xy is the correlation coefficient; n is the number of samples; x_i and y_i are the i-th values of x and y, respectively; and $\overline{x}$ and $\overline{y}$ are the mean values of the variables x, y, respectively.

2.

Spatial autocorrelation analysis

In addition, spatial autocorrelation analysis methods, including univariate spatial autocorrelation analysis and bivariate spatial autocorrelation analysis (global and local), were used in this study to express the degree of spatial correlation, spatial pattern, and significance of ESV and EQ in the study area. The global spatial autocorrelation analysis used Moran’s I to characterize the spatial characteristics of the variables on a global scale. Local spatial autocorrelation analysis used LISA cluster maps to describe the spatial correlation of attributes within different spatial units [49]. The formula is as follows:

$$I=\frac{n{\displaystyle \sum _{i=1}^n{\displaystyle \sum _{j=1}^n{W}_{ij}(y_i-\overline{y})(y_j-\overline{y})}}}{({\displaystyle \sum _{i=1}^n{\displaystyle \sum _{j=1}^n{W}_{ij}){\displaystyle \sum _{i=1}^n{(y_i-\overline{y})}^2}}}}$$

(6)

where I is the global Moran’s I index; y_i and y_j are the attribute values of units i and j, respectively; n is the number of spatial units in the study area; and W_ij is the weight matrix established based on spatial adjacencies.

To measure the spatial correlation among multivariate variables, based on this, Anselin proposed bivariate local spatial autocorrelation [50]. It can determine whether and to what extent different spatial variables are correlated with their neighboring regions in contrast to traditional spatial autocorrelation with only one variable. The formula is:

$$I_{\mathrm{mn}}=Z_m^i{\displaystyle \sum _{j=1}^n{W}_{\mathrm{ij}}{Z}_{\mathrm{n}}^{\mathrm{j}}}$$

(7)

where I_mn is a bivariate localized Moran’s I index; $Z_{\mathrm{m}}^p=\frac{X_m^{\mathrm{p}}-\overline{X}}{e_m}$, $Z_{\mathrm{n}}^p=\frac{X_n^{\mathrm{p}}-\overline{X}}{e_n}$, $X_{\mathrm{m}}^p$ is the value of attribute m of spatial unit p; $X_{\mathrm{n}}^p$ is the value of attribute n of spatial unit p; $\overline{X}$, e_m are the mean and variance of attribute m; and $\overline{Y}$, e_n are the mean and variance of attribute n, respectively.

2.4. Research Framework

(1)

The RSEI model was constructed using the GEE (Google Earth Engine) platform to quantify the ecological quality of Fujian Province from 2000 to 2021. The four indicators—greenness, heat, dryness, and humidity—were calculated separately, and the RSEI was derived using principal component analysis.

(2)

The equivalent factor method was employed to quantify the ecosystem services in Fujian Province from 2000 to 2021. The four indicators—supporting services, regulating services, provisioning services, and cultural services—were calculated and spatially visualized through spatial mapping with ArcGIS 10.2.

(3)

The dynamic evolution of land cover, EQ, and ESV in Fujian Province was analyzed both temporally and spatially. The interannual trends of ESV and EQ, as well as the land cover transfer matrix, in Fujian Province from 2000 to 2021 were analyzed on a temporal scale. Additionally, the spatial clustering characteristics of both were examined on a spatial scale using global spatial autocorrelation and local spatial autocorrelation.

(4)

Pearson correlation analysis was conducted using IBM SPSS Statistics 25 to clarify the relationship between ESV and EQ in the provincial area. Bivariate spatial autocorrelation analysis and spatial mapping were performed using GeoDA 1.14.0 to study the spatial heterogeneity between ESV and EQ in terms of the spatial patterns at the county level (Figure 2).

3. Results

3.1. Spatial and Temporal Land Cover Change

Taking into account the physical and geographic characteristics of the study area and the objectives of the study, this study reorganized and integrated the land cover data and reclassified them into six types: cropland, forest, grassland, water, construction land and barren. The study ultimately obtained the land cover classification maps of Fujian Province for six periods: 2000, 2004, 2008, 2012, 2016, and 2021.

Combined with

Table 2. Land cover type area table of Fujian Province from 2000 to 2021 (km²).

Year	2000	2004	2008	2012	2016	2021
Cropland	15,643.65	17,000.59	14,434.23	14,109.43	13,819.41	14,037.71
Forest	102,502.49	100,709.2	102,997.24	102,870.62	103,170.61	102,950.26
Grassland	96.75	84.11	107.90	75.06	76.21	80.29
Water	1916.45	1580.38	2006.55	1958.78	1949.86	1947.67
Barren	24.32	20.99	17.26	23.78	20.77	18.18
Construction land	2220.79	3008.79	2841.27	3366.78	3367.59	3370.34

and

Table 3. Land cover type change table of Fujian Province from 2000 to 2021.

Year	2000–2004	2004–2008	2008–2012	2012–2016	2016–2021	2000–2021
Cropland	8.67%	−15.10%	−2.25%	−2.06%	1.58%	−10.27%
Forest	−1.75%	2.27%	−0.12%	0.29%	−0.21%	0.44%
Grassland	−13.06%	28.28%	−30.44%	1.53%	5.35%	−17.01%
Water	−17.54%	26.97%	−2.38%	−0.46%	−0.11%	1.63%
Barren	−13.69%	−17.77%	37.78%	−12.66%	−12.47%	−25.25%
Construction land	35.48%	−5.57%	18.50%	0.02%	0.08%	51.76%

, it can be seen that the area ranking of the land cover types in Fujian Province is as follows: forest > cropland > construction land > water > grassland > barren. The land cover area as of 2021 is 102,950.26 km² (84.11%), 14,037.71 km² (11.47%), 3370.34 km² (2.75%), 1947.67 km² (1.59%), 80.29 km² (0.07%), and 18.18 km² (0.01%), respectively. Fujian Province is predominantly forested, accounting for more than 80% of the total area, which aligns with the saying “eight mountains, one water, one field”. In terms of the spatial distribution, construction land and cropland are mainly concentrated along the southeastern coast, while forested land is largely concentrated inland.

With the promotion of ecological construction and urbanization, the area of cropland has been significantly reduced over the past 21 years, with a decrease of 1605.94 km², and the area ratio has dropped by 10.27%. In contrast, there was a slight increase in the area of forest, but the increase was smaller, less than 3%. The grassland area decreased more from 2004 to 2012, but then increased, and overall, there was a decrease of 16.46 km² in 2021 from 2000, a decrease of 17.01%. The area of the water increased significantly between 2004 and 2008, with an increase of 31.22 km² or 1.63% over the 21 years. Except for the period 2008–2012, the area of barren showed an overall decreasing trend, with a total overall decrease of 6.14 km² (25.25%). Except for the period from 2004 to 2008, there has been a continuous expansion of construction land with a significant increase, with an overall increase of 1149.55 km² (51.76%). Overall, the land cover changes in Fujian Province during the 21 years were characterized by a continuous and dramatic expansion of construction land and a gradual decrease in cropland, grassland, and barren.

3.2. Spatial and Temporal Changes in Ecological Quality

As shown in

Table 4. RSEI principal component analysis results in Fujian Province.

Year	RSEI Mean	PCA	NDVI	LST	WET	NDBSI	Eigenvalue	Percent Eigenvalue
2000	0.723	PC1	0.9876	−0.1183	0.0054	−0.1025	0.0242	76.51%
		PC2	0.1396	0.9506	−0.1367	0.2411	0.0045	14.10%
		PC3	−0.0591	0.2858	0.3759	−0.8795	0.0024	7.62%
		PC4	−0.0392	−0.0253	−0.9165	−0.3973	0.0006	1.77%
2004	0.649	PC1	0.9926	−0.1199	0.0017	−0.0206	0.0340	85.24%
		PC2	0.1161	−0.9797	0.1150	−0.1164	0.0042	10.63%
		PC3	0.0264	−0.1538	−0.8776	0.4533	0.0014	3.40%
		PC4	−0.0248	0.0474	−0.4654	−0.8835	0.0003	0.72%
2008	0.725	PC1	0.9686	−0.2344	0.0215	−0.0792	0.0225	87.08%
		PC2	−0.2413	−0.9642	0.0811	−0.0745	0.0021	8.22%
		PC3	−0.0358	0.1214	0.8059	−0.5783	0.0010	3.78%
		PC4	−0.0470	0.0249	−0.5860	−0.8085	0.0002	0.92%
2012	0.714	PC1	0.9840	−0.0596	0.0870	−0.1438	0.0271	74.87%
		PC2	0.1242	0.8904	−0.3417	0.2740	0.0044	12.18%
		PC3	−0.0951	0.4337	0.8349	−0.3251	0.0039	10.82%
		PC4	0.0857	−0.1248	0.4225	0.8936	0.0008	2.13%
2016	0.757	PC1	0.9624	−0.2348	0.0946	−0.0984	0.0188	78.04%
		PC2	−0.1854	−0.3297	0.9138	−0.1478	0.0027	11.16%
		PC3	−0.1860	−0.9119	−0.3659	0.0057	0.0025	10.29%
		PC4	−0.0695	0.0677	−0.1488	−0.9841	0.0001	0.50%
2021	0.736	PC1	0.9193	−0.3254	0.1008	−0.1972	0.0211	76.55%
		PC2	−0.3553	−0.9254	0.1090	−0.0735	0.0043	15.43%
		PC3	0.1411	−0.1921	−0.7825	0.5752	0.0018	6.35%
		PC4	−0.0935	0.0275	−0.6047	−0.7905	0.0005	1.67%

, the contribution of principal component one to the six images from 2000 to 2021 was 76.51%, 85.24%, 87.08%, 74.87%, 78.04% and 76.55%, respectively. The results show that the four indicators are dominated by PC1, with LST and NDBSI as negative indicators, and NDVI and WET as positive indicators, which is consistent with the results of related studies [6,27]. Among them, the contribution of NDVI to PC1 was the highest among the four sub-indices, reflecting the important role of vegetation in EQ, consistent with previous views [1,2,5].

To facilitate the statistical analysis of the EQ in Fujian Province, the RSEI values were categorized into five grades based on previous studies [9]: excellent (0.8–1.0), good (0.6–0.8), moderate (0.4–0.6), poor (0.2–0.4), and very poor (0–0.2). As can be seen from Figure 3, areas with excellent and good ecological quality in Fujian Province are mainly concentrated in the west and central parts of the province, where dense vegetation is present, while those with poor EQ are found in the central urban areas of the southeastern coastal cities and counties—namely, areas with more construction land and barren. The main reason for this distribution is the high level of urbanization, economic development, and human activities in the central city and the southeast coast. Figure 4 shows that the EQ of the vast majority of Fujian Province declined significantly from 2000 to 2004, became significantly better from 2004 to 2008, changed more significantly in the central region from 2008 to 2016, and changed little in most areas from 2016 to 2021. There has been no significant change in the EQ of the southeastern coastal region over the past 21 years, while all the other regions have experienced varying degrees of change.

Figure 5 shows the changes in the area share of ecological quality and the mean value of the RSEI in Fujian Province from 2000 to 2021. Over the past 21 years, the average value of the RSEI in Fujian Province has shown an overall fluctuating upward trend. The RSEI mean declined significantly between 2000 and 2004, then rose to return to its original level, before declining slightly. It rose significantly between 2012 and 2016 and then declined slightly by 2021, but overall, it showed a slight increase. Overall, the EQ of Fujian Province, rated as excellent and good, accounts for more than 60% of the total, with the overall proportion of the change first decreasing and then increasing. The percentage of areas with very poor ecological quality is less than 1%.

Figure 6 illustrates the transfer of different levels of the RSEI in Fujian Province from 2000 to 2021. In addition to transfers within the same category, regions of very poor EQ were reclassified as poorer regions, medium regions were reclassified as both poorer and good regions, and good regions were reclassified as excellent regions. Among these transfers, a total of 941.75 km² was shifted from very poor areas, with most being reclassified as poorer areas, representing about 25.25% of the total area transferred. The total area transferred from the poorer regions was 24,818 km², and 6149.75 km² was reclassified as medium regions, which accounts for 24.78% of the total area transferred. Medium areas were reclassified into various categories, with the largest area, accounting for 38.21%, shifting to good areas. This indicates that the ecological quality in Fujian Province has generally improved.

3.3. Spatial and Temporal Changes in Ecosystem Services

In order to easily visualize the spatial distribution characteristics of the ESV in Fujian Province, this study calculated the average ecological service value of 68 county-level cities (districts), as shown in Figure 7. The spatial distribution of the ESV in Fujian Province is characterized by high values in the middle and low values in the surroundings, with areas of high ESV mainly found in the western and central cities. In contrast, areas with lower ESV are distributed in the southeastern coastal cities, which have a high level of urbanization and more construction land and barren.

Combined with

Table 5. Temporal change in the ecosystem service value in Fujian Province from 2000 to 2021.

ESs		ESV (Million CNY)
		2000	2004	2008	2012	2016	2021
Provision service (PSV)	Food production (FP)	8092	8237	7882	7801	7754	7789
	Raw materials (RM)	10,770	10,590	10,809	10,787	10,814	10,794
	Subtotal	18,862	18,827	18,691	18,588	18,568	18,583
Regulating service (RSV)	Gas regulation (GR)	37,291	36,863	37,272	37,159	37,210	37,173
	Climate regulation (CR)	105,286	103,445	105,727	105,526	105,800	105,599
	Waste disposal (WD)	33,246	32,388	33,460	33,356	33,430	33,368
	Water regulation (WR)	120,848	113,596	122,462	121,300	121,239	121,121
	Subtotal	296,670	286,292	298,921	297,341	297,679	297,262
Support service (SSV)	Soil retention (SR)	42,362	41,566	42,582	42,510	42,631	42,542
	Biodiversity protection (BP)	3643	3621	3625	3610	3611	3611
	Biological diversity (BD)	39,664	38,873	39,857	39,766	39,865	39,790
	Subtotal	85,668	84,061	86,064	85,886	86,107	85,943
Cultural service (CSV)	Aesthetic landscape (AL)	17,711	17,314	17,809	17,762	17,805	17,771
Total ESV		418,912	406,494	421,486	419,578	420,158	419,559

and

Table 6. Growth rate of the ecosystem service value in Fujian Province from 2000 to 2021.

ESs	PSV	RSV	SSV	CSV	ESV
2000–2004	−0.25%	−2.79%	−2.08%	−2.62%	−2.50%
2004–2008	−0.59%	3.45%	2.65%	3.37%	3.07%
2008–2012	−0.53%	−0.43%	−0.23%	−0.31%	−0.39%
2012–2016	−0.09%	0.14%	0.25%	0.22%	0.16%
2016–2021	0.07%	−0.14%	−0.19%	−0.19%	−0.15%
2000–2021	−1.40%	0.13%	0.34%	0.38%	0.12%

, it can be seen that from 2000 to 2021, with the continuous advancement of urbanization, the ESV in Fujian Province generally followed a fluctuating upward trend. The value of all the services showed a significant decline from 2000 to 2004, and the ESV of Fujian Province was CNY 406,494 million in 2004, down 2.96% compared with 2000. The RSV, SSV, CSV, and ESV all increased significantly from 2004 to 2008, with a slight decrease in the PSV, which is related to the increase in the area of green space and the concurrent decrease in the area of cultivated land. The ESV reached CNY 421,486 million in 2008, an increase of 3.69%, and has fluctuated with little change since then. Overall, the value of all the services except the PSV increased slightly from 2000 to 2021, with the decrease in the PSV being the result of a reduction in the cropland acreage.

3.4. Relevance Analysis

3.4.1. Overall Relationship

To explore the interrelationship between the EQ and ESV indicators, a PCA was conducted on the relevant data in 2021, and the results are displayed in Figure 8. The analysis revealed a significant correlation between the EQ and ESV indicators. Specifically, the RSEI demonstrated a significant positive correlation with the ESV (r = 0.673, p < 0.01). The RSEI exhibited significant positive correlations with the ESV, cultural service value, supporting service value, and provisioning service value, with the highest correlation coefficient observed with the supporting service value (r = 0.718). The WET showed a significant positive correlation with all the ESV indicators. The NDVI showed a significant positive correlation with the ESV, cultural service value, supporting service value, and provisioning service value, with the highest correlation coefficient between the NDVI and supporting service value (r = 0.815). The NDBSI showed a significant negative correlation with all four ESV indicators. The Land Surface Temperature was significantly negatively correlated with the ESV, cultural service value, supporting service value, and regulating service value, and it was significantly positively correlated with the provisioning service value. In contrast, the correlation between the regulating service value and each of the EQ indicators was relatively low.

3.4.2. Spatial Interaction Stressor Relationships

In this study, 68 county-level administrative districts in Fujian Province were used as the basic unit for analysis. The aim was to obtain the global spatial autocorrelation Moran’s I index (

Table 7. Statistical table of the Moran’s I of the EQ and ESV spatial autocorrelation in Fujian Province from 2000 to 2021.

Year		EQ	ESV	EQ-ESV
2000	Moran’s I	0.538	0.347	0.387
	Z-score	7.314	4.782	6.499
	p-value	0.001	0.001	0.001
2004	Moran’s I	0.471	0.394	0.258
	Z-score	6.503	5.426	4.463
	p-value	0.001	0.001	0.002
2008	Moran’s I	0.607	0.467	0.438
	Z-score	8.180	6.446	6.971
	p-value	0.001	0.001	0.001
2012	Moran’s I	0.543	0.512	0.474
	Z-score	7.282	7.039	7.355
	p-value	0.001	0.001	0.001
2016	Moran’s I	0.645	0.528	0.545
	Z-score	8.564	7.245	8.264
	p-value	0.001	0.001	0.001
2021	Moran’s I	0.629	0.582	0.573
	Z-score	8.573	7.994	8.353
	p-value	0.001	0.001	0.001

), the local spatial agglomeration distribution LISA map, and the bivariate spatial agglomeration distribution LISA map (Figure 9) for the EQ and ESV from 2000 to 2021. Table 6 indicates that Moran’s I indexes are consistently positive, with a p-value < 0.01 and a z-score > 2.58, which corresponds to a 99% confidence level. This suggests a strong positive spatial correlation between both the EQ and ESV, as characterized by significant clustering. The Moran’s I index generally exhibits an upward trend throughout the years from 2000 to 2021, indicating that the spatial correlation has become increasingly pronounced, with larger values.

The results of the local autocorrelation analysis, as depicted in Figure 9, indicate that the high–high and low–low agglomeration areas are the more dominant relationship types within the study area. Analysis of the EQ results reveals that the low–low agglomeration area increased significantly from 2000 to 2021 and has remained stable since 2016. This area is primarily located in the municipal districts of Xiamen, Fuzhou, and Quanzhou—regions of Fujian Province characterized by a high level of urbanization and a more developed economy. In contrast, the high–high agglomeration area initially decreased and then increased. It is predominantly situated in Nanping, Sanming, Longyan, and most counties and cities in Ningde. These areas are predominantly mountainous and wooded, with high vegetation cover and good ecological quality. Examination of the ESV results demonstrates that the low–low agglomeration area also increased from 2000 to 2021, with regional distribution roughly consistent with the EQ findings. Meanwhile, the high–high agglomeration area, after a decrease, showed an increase and is mainly found in the central region of Fujian Province, specifically in most counties and cities in Nanping, Sanming, and Ningde. Bivariate LISA clustering and significance have exposed various types of spatial correlations between the EQ and ESV. There is a substantial distribution area of high–high agglomeration, along with a small number of areas that exhibit high–low and low–high agglomeration.

4. Discussion

4.1. Evolution of EQ and ESV Spatio-Temporal Dynamics

From the analysis of the evolution of the time dynamics, the patterns of change in the ecological quality and ecosystem service value in Fujian Province from 2000 to 2021 are essentially the same, with both showing a fluctuating upward trend. This indicates that the ecological status of Fujian Province as a whole is improving. This improvement is closely related to the province’s emphasis on ecological protection and its longstanding policies of afforestation and forest closure. These trends match the findings in the existing literature and confirm the effectiveness of ecological protection policies [37]. However, there was a significant decline in the ecological status from 2000 to 2004, mainly due to the continued rapid economic development of Fujian Province during this period. This development led to the expansion of land used for construction, a continued reduction in the area of other land-use types, and a significant increase in the level of urbanization. Yang et al. showed that there is a negative correlation between urbanization and the ecological environment, where large-scale and rapid urbanization leads to the deterioration of the ecological environment [26]. The ecological downward trend during this phase also reflects the contradiction between economic development and ecological protection.

In response to this issue, the Fujian Provincial Government has continued to strengthen its emphasis on ecological environment construction and has issued a series of relevant policies and opinions. It is stipulated that by 2015, the main objectives of ecological province construction should be essentially realized, with most cities with districts reaching or being close to the standards for creating ecological cities. The aim is to lead in building a resource-saving and environmentally friendly society, placing the construction of ecological civilization at the forefront of the country. The implementation of these policies has resulted in a significant improvement in the ecological status over the period 2004–2021, albeit with small fluctuations.

In terms of the spatial distribution, the characteristics of the ecological quality and ecosystem service value in Fujian Province are similar, with high values concentrated in areas with high vegetation cover in northern and central Fujian, and a slight increase in the overall area. The low values are mainly located along the eastern coast of Fujian, with an increasing percentage of the area spreading to the west. This is consistent with previous findings on the subject [37]. The clear distinction between areas of high and low values reveals the variability in the influence of anthropogenic and natural factors on ecological quality. The eastern part of Min has put greater pressure on the ecosystem due to its high population density and high level of economic development, while other areas are in relatively good ecological condition due to their mountainous and hilly topography, high vegetation cover and relatively dispersed population.

In summary, to further enhance the ecological environment’s quality in Fujian Province, a series of measures can be implemented. These include artificially planting vegetation and protecting and restoring natural plant communities to increase vegetation cover, thereby promoting continuous ecological improvement. Additionally, it is essential to balance economic development with ecological protection. This balance ensures economic growth does not cause irreversible harm to the ecological environment. With these comprehensive measures, Fujian Province is poised to make significant strides in developing an ecological civilization.

4.2. Correlation Study between EQ and ESV

In terms of the overall relationship, the RSEI was significantly positively correlated with the ESV (r = 0.673, p < 0.01), indicating a high degree of coordination between the EQ and ESV. Increasing greenness and humidity can improve biodiversity, soil erosion resistance, and nutrient cycling capacity, thereby enhancing the value of all types of services in ecosystems [51]. Areas with a high surface temperature and dryness are primarily found in construction land and bare land, which are composed of artificial materials with rapid heat absorption and low heat capacity. The urban heat island effect exacerbates heat and dryness, which is detrimental to ecological environment improvement. The correlation between the NDVI and other indicators was significantly higher, suggesting that vegetation cover is a key factor influencing various ecological indicators in Fujian Province. The LST was significantly positively correlated with only the PSV. This correlation exists because the bare surface area of arable land contributes more to the sensible heat flux than forests and grasslands do, resulting in higher surface temperatures [52]. There are varying degrees of correlation between the different indicators of the EQ and ESV, and not all are correlated. Therefore, the process of promoting ecological governance in Fujian Province should integrate multiple considerations to achieve higher ecological value.

Through spatial autocorrelation analysis, it was found that the high–high clustering of the EQ and ESV in Fujian Province is concentrated in the northwestern and central regions, while the low–low clustering is concentrated in the coastal region of southern Fujian. The main reasons are similar to those influencing the spatial distribution. Bivariate spatial autocorrelation results indicate that the higher EQ between neighboring counties and cities increases the ESV supply capacity, consistent with the findings of Zhang et al. [53]. This is primarily because neighboring counties and cities share similar natural resources, environmental conditions, and socio-economic development levels, and they are constantly transferring various types of resources, leading to frequent exchanges and mutual influence [24,54]. However, there are also a small number of counties and cities with negative correlations locally, suggesting a complex interaction mechanism between the EQ and ESV. Other anthropogenic factors may be involved, causing changes in the regional ESV [55,56].

To summarize, promoting ecological governance and enhancing the value of ecosystem services in Fujian Province require comprehensive measures that take into account various factors. These measures include enhancing vegetation cover, optimizing urban planning and land use, and focusing on the interactions between different ecological indicators. Additionally, emphasis should be placed on interregional collaboration and resource sharing to achieve an overall improvement in ecological quality and to maximize the value of ecosystem services. The implementation of these strategies is expected to foster a more harmonious and sustainable ecological environment in Fujian Province.

4.3. Research Shortcomings and Prospects

This study has the following shortcomings. First, although this study has corrected the equivalent factor, the complex geography of Fujian Province and the differences between regions may lead to different values of the equivalent factor within the same service value, thus affecting the accuracy of the ESV evaluation. Therefore, future research should conduct field observations, measure actual data, enhance parameter localization and validation, and improve the accuracy of the basic data [57]. Second, the correlation analysis assumes a linear relationship between the EQ and ESV, ignoring the uncertainty associated with their nonlinear properties. Consequently, it remains challenging to fully elucidate the mechanism of the relationship between the EQ and ESV from the perspective of correlation analysis. Finally, while this study examined the EQ and ESV correlations qualitatively, future quantitative analyses could offer further insight.

5. Conclusions

This study analyzed the spatio-temporal characteristics of EQ and ESV in Fujian Province from a multi-timescale perspective based on multi-source data, and revealed the interaction relationship between the two, offering unique insights into a region that has been a frontrunner in eco-civilization construction. The results show a certain similarity in the pattern of change and spatial distribution of the EQ and ESV from 2000 to 2021, with both exhibiting a fluctuating upward trend. The overall spatial distribution is high in central and northern Fujian and low in the southeastern coastal region. Pearson correlation analysis yielded significant correlations between the indicators of EQ and ESV, with a notable positive correlation between the RSEI and ESV (r = 0.673, p < 0.01), indicating a synergistic relationship. Additionally, spatial autocorrelation analysis revealed a positive spatial correlation between EQ and ESV, with high–high agglomeration concentrated in northwestern and central Fujian, and low–low agglomeration concentrated in the coastal area of southern Fujian. This study thoroughly analyzed the coordinated relationship between EQ and ESV and concluded that ecological protection in Fujian Province should prioritize ecosystems, combat soil erosion, and strengthen forest protection and reforestation measures to promote synergistic nature–society–economy development.