Next Article in Journal
Study on the Evolution of Groundwater Level in Hebei Plain to the South of Beijing and Tianjin Based on LSTM Model
Previous Article in Journal
AI-Enabled Digital Twin Framework for Safe and Sustainable Intelligent Transportation
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Sustainability
Volume 17
Issue 10
10.3390/su17104393
sustainability-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Spatio-Temporal Dynamics and Driving Forces of Ecosystem Service Value at Multiple Scales in the Shandong Peninsula Urban Agglomeration, China

by
Yongwei Liu
1,* and
Tianping Zhang
2
1
School of Business, Ludong University, Yantai 264025, China
2
School of Resources and Environmental Engineering, Ludong University, Yantai 264025, China
*
Author to whom correspondence should be addressed.
Sustainability 2025, 17(10), 4393; https://doi.org/10.3390/su17104393
Submission received: 14 January 2025 / Revised: 2 May 2025 / Accepted: 9 May 2025 / Published: 12 May 2025
(This article belongs to the Section Environmental Sustainability and Applications)
Browse Figures
Versions Notes

Abstract

:
The analysis of ecosystem service value (ESV) dynamics across space and time, along with their driving factors, is essential for informed ecosystem service administration and policy development. The Shandong Peninsula Urban Agglomeration (SPUA) is an important, highly efficient eco-economic zone in China. Leveraging land use/land cover datasets covering the period 2000–2020, spatial autocorrelation analysis and geographical detector were used to examine the spatial distribution characteristics and driving forces of the ESV. The results indicated the following: (1) From 2000 to 2020, the ESV of SPUA exhibited an overall trend of “increase—decrease—increase”. Cropland, forest, and water bodies were the primary sources of ESV, with significant variations in the changes of ESV across different land-use types. (2) As the spatial scale increased, the level of spatial autocorrelation of the per-unit ESV gradually decreased, and no spatial autocorrelation was observed at the urban scale. Analysis revealed that the clustering trend was more pronounced at the township scale, and its stability over the years was higher than that at the county scale. (3) The per-unit ESV was driven primarily by socio-economic factors, and the relative importance of these driving forces was minimally affected by the spatial scale, indicating a certain degree of similarity across different scales. (4) The spatial distribution pattern of per-unit ESV was not driven by a single factor but by the interaction of multiple factors. These interactions were significantly influenced by spatial scale, with more complex interaction effects observed at the county scale. Slope, in particular, played a crucial role in the interactions. This research contributes valuable scientific knowledge for developing environmental conservation frameworks in the SPUA while supporting the region’s sustainable growth initiatives.

1. Introduction

Ecosystem services are collective terms for all the benefits that people receive from ecosystems [1]. Ecosystem services are naturally closely linked to humans, ensuring regional ecological security and economic development [2,3], and are important for human survival, well-being and health [4,5]. However, ecosystem services are currently under pressure because of anthropogenic activities and climate change [6]. Among the factors that cause changes in ecosystem services are rapid economic development [7], land use/cover change (LUCC) [8,9], and population growth [10]. In particular, LUCC affects the ecological function and structure of ecosystems and is constrained by various environmental and societal factors [11,12]. The mechanism of ecosystem service changes is still unclear because the feedback between these factors and ecosystems is highly complicated [11]. Therefore, exploring the spatio-temporal dynamics and driving forces of ecosystem services is highly important for land management and regional sustainable development.
Ecosystem service value (ESV) is an approach used to quantify and assign economic value to ecosystem goods and services and their functions [13] as a significant indicator of ecological quality [14]. ESV can improve people’s understanding of the importance, well-being, and value of natural assets [2]; thus, a growing acknowledgment of the urgency to evaluate ecosystem service value has emerged [15]. Methods for assessing ESV can be divided into two types: data-based methods and proxy-based methods [16]. The data-based method usually employs ecological models to calculate the ESV because of its comprehensive consideration of various factors, resulting in relatively reliable outcomes [17]. However, this method increases the cost of model implementation and limits its widespread utilization [18]. It is often applied in small-scale studies [16] that combine ecological models and primary data to quantify the ecosystem processes and functions that underlie ecosystem services, and then convert the derived ecosystem services into market prices [19]. The proxy-based method relies on benefits transfer using secondary data, such as land use data, and is widely used in ESV assessment because these methods are easier to use and more intuitive [20,21]. Among these, the equivalent coefficients table method, proposed by Costanza [1] and improved by Xie [22,23], is regarded as one of the representative methods for ESV assessment in China, and the estimated value for each land use type can be transferred from one location to another with similar conditions. This method obtains the equivalent value based on meta-analysis and the area of each ecosystem to obtain the regional ESV [24]. It is more commonly used with LULC data because of their widespread availability [16]. Therefore, the equivalent coefficients table method was adopted in the ESV assessment.
Research on the ESV has gradually shifted from evaluating values to analyzing driving factors. Many scholars have studied the factors influencing ESV, which can generally be categorized into two types: natural factors and socio-economic factors. Indicators used for natural factors include elements such as the Digital Elevation Model (DEM) [14,25], slope [14,26], temperature [14,25,27], precipitation [14,25,27], snow [25], Net Primary Productivity (NPP) [28,29,30,31], leaf area index [32], Normalized Difference Vegetation Index (NDVI) [28,30,33], and soil characteristics (such as soil organic matter, type, properties, erosion and pH) [34]. These factors influence ecosystem structure and function by affecting atmospheric circulation, water cycles, geological processes, and biological cycles, ultimately impacting regional ESV. Socio-economic factors are typically represented by indicators such as the human activity index [33], population [32,35,36], Gross Domestic Product (GDP) [34,35,36], night light [27,32], roads [11,26], land use [37,38], rivers [26,36,39], infrastructure [40], and energy [40]. In the same region, improvements in the socio-economic development level have changed regional consumption patterns and industrial structures [41], subsequently affecting the ESV. Generally, both natural and socioeconomic elements jointly affect ecosystem dynamics. Hence, examining the combined influence of natural and socio-economic drivers is crucial for thoroughly understanding and precisely interpreting the intricate variations in ESVs. Furthermore, identifying the dominant factors influencing ESV can help elucidate the root causes of the spatio-temporal variations in regional ESV.
Ecosystem service capabilities depend on ecological and geographic mechanisms operating across various spatial and temporal dimensions [42]. Understanding ecosystem service values necessitates consideration of scale, as the intricate nature of ecological processes results in varying patterns and characteristics across different dimensions. This leads to the possibility of conducting ESV studies at different spatial scales. Numerous scholars have conducted research on ecosystem services across various scales, including macro scale studies, such as those at the global [1] and continental levels [43]; meso-scale studies, such as those focused on river basins [44] and urban agglomerations [45]; and micro-scale studies, such as those examining county-level administrative units [9] and different grid units [46]. As the basis of all ecological studies, scale has always been a focus and challenge in ecosystem services research [46,47,48], and the key driving forces of ESV, as well as the relative weight of these determinant factors, can differ depending on the scale of analysis. While certain researchers have initiated studies on the scale-dependent effects of ESV influencing factors [28,30,31], research examining the spatiotemporal dynamics and driving mechanisms of ESV across multiple administrative scales remains insufficient. This research seeks to fill this knowledge gap by examining ESV across three administrative levels: township, county, and city.
Urban agglomerations are areas with rich ecological resources and significant economic value [49]. However, because human disturbances disrupt the ability of ecosystems, some urban agglomerations are becoming increasingly ecologically vulnerable and sensitive to environmental change [50,51]. Therefore, conducting research on the ESV of urban agglomerations can facilitate the scientific assessment of ecological security patterns in these regions, providing valuable insights for promoting high-quality regional development. As a crucial high-performance eco-economic zone in China, the ecological significance of the Shandong Peninsula Urban Agglomeration (SPUA) has grown substantially. However, with economic growth and population increases, the region has also begun to face a series of significant challenges, such as soil erosion, local ecological space degradation, and poor ecosystem stability. Accordingly, this study pursues the following main objectives: (1) calculate the ESV for five different periods—2000, 2005, 2010, 2015, and 2020—using the equivalent coefficients table method; (2) assess the spatio-temporal distribution characteristics of ESV during 2000–2020 using spatial analytical approaches; (3) uncover the driving mechanisms behind the spatial differentiation of ESV through geographical detector analysis and establish the dominant factors. By studying the spatio-temporal differentiation characteristics and driving factors of ESV in the SPUA, this research contributes to the efficient management of regional resources, the optimization of ecological structures, and the sustainable protection of ecosystems.

2. Materials and Methods

2.1. Study Area

The SPUA covers all 16 cities within Shandong Province [9], located in the downstream region of the Yellow River. Geographically, the land area spans from 34°23′ to 38°17′ N latitude and from 114°48′ to 122°42′ E longitude, extending 721.03 km from east to west and 437.28 km from north to south (Figure 1). Stretching from the northern to southern regions, this area shares boundaries with the Hebei, Henan, Anhui, and Jiangsu Provinces, encompassing a terrestrial territory of 154,300 km2 and an offshore region of approximately 158,600 km2. The terrain of this region is complex and can be broadly categorized into plains, terraces, hills, and mountainous areas. Flatlands constitute 65.56% of the entire territory, with their primary distribution spanning across the northwestern district and selected southwestern zones. Terraces account for 4.46%, primarily located in the eastern region. The topography includes hilly terrain comprising 15.39% of the area, primarily situated in the eastern and southwestern sectors, whereas mountainous zones constitute 14.59% of the area, largely found in the central sector and portions of the southwest. The SPUA is situated along the coasts of the Bohai Sea and the Yellow Sea. The total length of the mainland coastline is 3504.74 km, and there are 589 islands within the jurisdictional waters.

2.2. Data Source

The data used in this study include the following: (1) Land use/cover (LULC) data: The LULC data were downloaded from the China Land Cover Dataset (available online: https://zenodo.org/records/8176941 (accessed on 1 July 2024)) [52]. (2) Revised data of ESV: NDVI data for revising the ESV were obtained from the National Ecosystem Science Data Center, National Science & Technology Infrastructure of China (available online: http://www.nesdc.org.cn (accessed on 1 July 2024)). (3) Driving factor data: Following a comprehensive review of published literature concerning ESV influence parameters and considering the specific context of the SPUA, this study identified 12 driving factors from both natural and socio-economic dimensions, as summarized in Table 1.

2.3. Research Methods

2.3.1. Estimation of ESV Method

Based on the research by Costanza [1], Xie constructed the China Ecosystem Service Value Equivalence Table [22,53]. The ESV of the SPUA was assessed using the equivalent coefficients table method.
There is a significant impact of vegetation coverage on the level of ecosystem services. To improve the accuracy and rationality of ESV assessment, many scholars have revised ESV using the NDVI [25,54,55]. Therefore, in this paper, the NDVI is introduced to revise the ESV calculation results. The formula is as follows:
C i k = N D V I i k N D V I i ¯
E S V = ∑ i = 1 a ∑ j = 1 b ∑ k = 1 c A i j B i k C i k
where Cik represents the NDVI revision coefficient for land use type i in the research unit k; NDVIik represents the NDVI value for land use type i in unit k; N D V I i ¯ represents the average NDVI value for land use type I; ESV represents the total value of ecosystem services; Aij represents the value coefficient of the j-th type of ecosystem service for the i-th type of land use; Bik represents the area of the i-th land use type in the k-th research unit. The indices i, j, and k represent land use types, ecosystem service types, and research unit numbers, respectively.

2.3.2. Spatial Autocorrelation Analysis

Spatial autocorrelation analysis can be divided into positive and negative correlation [56]. The presence of a positive correlation is marked by synchronous changes in attribute values between spatial units and their neighbors, whereas a negative correlation reflects opposing value movements between adjacent areas. Spatial autocorrelation includes both global spatial autocorrelation and local spatial autocorrelation [57].
(1)
Global Spatial Autocorrelation
Moran’s I index is used to measure global spatial autocorrelation, and its formula is as follows:
I = n ∑ i n ∑ j n w i j y i − y ¯ y j − y ¯ ∑ i n ∑ j n w i j ∑ i n y i − y ¯ 2
where n represents the total number of units, yi and yj represents the attribute values of points i and j, respectively, y ¯ is the average value of all attribute values in the area, and w i j represents the spatial weight. The value of Moran’s I falls within the range of [−1, 1], indicating that there is a negative correlation when the observed value is less than 0, an independent random distribution when it is equal to 0, and a positive correlation when it is greater than 0.
(2)
Local Spatial Autocorrelation
The local Moran’s I was proposed by Anselin [58] and is used to represent local spatial autocorrelation; the formula is as follows:
I 1 = z i ∑ i w i j z j  
where z i represents the standard amount of the mean value, z j represents the standardized quantity of the standard deviation, z i = x i − x ¯ δ , and δ represents the standard deviation of x i .

2.3.3. Geographical Detector

The geographical detector (Geodetector) was first proposed by Wang. The Geodetector is a spatial statistical method used to measure and attribute spatially stratified heterogeneity and can test the relationships between geographical phenomena and their potential driving factors. The Geodetector includes the factor detector, interaction detector, ecological detector, and risk detector [59,60,61]. However, the spatial data discretization and spatial scale effects are generally determined by experience and lack accurate quantitative assessment. To address this issue, an optimal parameters-based geographical detector (OPGD) model was developed by Song for more accurate spatial analysis [62]. In order to analyze the key driving forces of ESV in the SPUA, this study focuses on factor detection and interaction detection using the OPGD model. The explanation of relevant indicators is derived from Wang’s literature [63].

3. Results

3.1. Overall Evolution Trend of ESV in the SPUA

The ESV of the SPUA was assessed using the equivalent coefficients table method. The coefficients, drawn from prior research studies, are presented in Table 2 [9]. The ESV of the SPUA was calculated on the basis of Table 2, the NDVI, and the land use/land cover data.
The results indicate that from 2000 to 2020, the total ESV of the SPUA exhibited a fluctuating trend, characterized by a “rise–fall–rise” pattern with distinct phases (Figure 2). The turning points occurred in 2010 and 2015. Specifically, the ESV increased from CNY 278.2005 billion in 2000 to CNY 282.1774 billion in 2010, then decreased to CNY 278.9791 billion in 2015, and subsequently increased again to CNY 280.6286 billion in 2020. Overall, the ESV of the SPUA increased by CNY 2.4281 billion between 2000 and 2020, representing a growth of approximately 0.87%. However, a significant decline occurred between 2010 and 2015, with the ESV decreasing by CNY 3.1983 billion, a reduction of about 1.13%. The trend of ESV was mainly directly influenced by land use/cover changes; for example, from 2000 to 2010, the continuous increase in water body area led to an increasing trend in ESV. In addition, environmental factors and ecological management policies also played indirect roles.
Further analysis of the ESV across different land use/cover types, including cropland, forest, grassland, unused land, and water bodies, revealed that cropland had a much higher ESV than other land types, though it experienced a continuous decline. The ESV of cropland decreased from CNY 199.5575 billion in 2000 to CNY 179.3107 billion in 2020, a reduction of CNY 20.2468 billion, or 10.15%. In contrast, the ESV of forest land followed a “decline–rise” trend, with a turning point in 2010. The ESV of forest land increased from CNY 40.65 billion in 2000 to CNY 47.6041 billion in 2020, an increase of CNY 6.9541 billion, or 17.11%. The ESV of water bodies exhibited a “rise–fall–rise” trend, with turning points in 2010 and 2015. Overall, the ESV of water bodies increased from CNY 28.5808 billion in 2000 to CNY 48.2582 billion in 2020, representing an increase of CNY 19.6774 billion, or 68.85%. Grassland, on the other hand, has shown a continuous decline, with its ESV decreasing from CNY 8.9109 billion in 2000 to CNY 5.374 billion in 2020, representing a reduction of 39.69%. Unused land also followed a downward trend, with its ESV decreasing significantly from CNY 0.5013 billion in 2000 to CNY 0.0815 billion in 2020. Overall, the evolution of the ESV in the SPUA exhibited complexity and diversity in its trends.
For a more comprehensive understanding of how ESV distributes spatially within the SPUA, the ESV was calculated at three spatial scales: township units, county units, and city units. The average ESV per hectare of land (AESV) was derived at each of these scales (Figure 3, Figure 4 and Figure 5). The overall distribution pattern of the AESV varied greatly across the three spatial scales, with cities such as Dongying, Binzhou, and Weifang increasing faster, while Qingdao, Rizhao, and Weihai decreased, and Dezhou and Liaocheng remained stable at the city scale. The differences were more pronounced at the county and township scales, with counties in different cities and townships in different counties showing more pronounced differences. Over time, a basic stable distribution pattern has formed. Therefore, further analysis of the distribution characteristics of the average ESV via spatial autocorrelation analysis was necessary.

3.2. Spatiotemporal Dynamics of ESV in the SPUA

GeoDa 1.22.0.2 software was used to perform spatial autocorrelation analysis. Using GeoDa software, the establishment of a spatial weight matrix serves as the foundational step, with this study adopting the Rook contiguity approach for matrix formation. Once the spatial weight matrix was generated, the global spatial autocorrelation index, Moran’s I, values for the five years were calculated, and significance was tested using the Monte Carlo simulation method with 999 permutations (Table 3). At the township-level scale, Moran’s I > 0.6, with little variation across the years. The p-values were all less than 0.01, indicating significance at the 99% confidence level. The analysis reveals strong positive spatial autocorrelation globally at the township scale, whereby areas exhibiting elevated ESV measurements typically neighbor other high-ESV regions, whereas localities with reduced values commonly adjoin similar low-ESV territories. At the county-level scale, Moran’s I values were positive and exhibited a “decline–rise” trend. With the exception of 2015, where the p-value was 0.058, the p values for the other years were all within 0.01. This indicates that the distribution of average ESV was not completely random but showed significant positive global spatial autocorrelation, where counties with higher ESV tended to cluster with other high-ESV counties, and low-ESV counties tended to cluster with other low-ESV counties. However, this clustering was strongest in 2000, gradually weakened to its lowest point in 2010, and then began to strengthen again. During the study period, the average ESV at the county scale exhibited a general trend of “weakening–strengthening” in terms of clustering. At the city level, the Moran’s I values were relatively small and not significant, indicating that there was no obvious clustering of the average ESV at the city level, and that its distribution appeared random.
Local spatial autocorrelation analysis results are often presented in the form of Local Indicators of Spatial Association (LISA) maps. LISA maps classify research units based on their clustering characteristics into four categories: High–-High, Low–Low, High–Low, and Low–High. Additionally, due to the selection of different significance levels when generating LISA maps, some areas in the maps may appear as non-significant. High–High: This category refers to areas where both their own spatial autocorrelation and that of their neighbors are positive. Both the county-level and township-level average ESV exhibited significant spatial clustering. To better compare the homogeneity and heterogeneity of local features between spatial units and their neighboring units, the local Moran’s I values for the five target years were calculated. The analysis focused on areas with a higher significance level (this study applied a 5% significance filter, with a significance threshold of 0.05) and created LISA maps (Figure 6 and Figure 7), along with clustering type statistics (Tables S1 and S2). At the township-level scale, HH clustering and LL clustering were dominant. In the SPUA, HH clustered towns were chiefly found in the eastern areas of Yantai and Weihai, as well as in the central Jinan–Tai’an–Zibo–Linyi area, southern Jining, and southeastern Rizhao. From 2000 to 2020, the overall number of HH clustered towns showed a decreasing trend, although the distribution pattern remained largely consistent. Towns with LL clustering were predominantly located in cities such as Weifang, Qingdao, and Jinan, along with the central urban regions of Heze, Jining, and Zaozhuang. The number of LL clustered towns showed an increasing trend, with the scope of clustering gradually expanding, though the overall spatial distribution pattern remained relatively stable. The LH and HL clustering types were less common. The number of LH clusters remained stable at 16–18, primarily concentrated in Jining and Zaozhuang, with scattered occurrences in cities such as Zibo, Jinan, Qingdao, and Weihai. The number of HL clusters was consistently around 4–6, mainly distributed in Jinan and Liaocheng, with significant variations over the years.
At the county-level scale, HH clustering was dominant, primarily concentrated in the contiguous Jinan–Tai’an–Zibo area and the Yantai–Weihai region. Except for Yutai County in Jining, which transitioned from HH clustering to LH clustering in 2020, the stability of HH clustered counties persisted throughout the study period. The number of LL clustered counties decreased rapidly, from seven in 2000 to three in 2005, and by 2015, only one remained. LL clustering gradually decreased from the contiguous Dongying–Weifang area to Huantai County in Zibo by 2020.
The quantity of LH clustered areas exhibited a general upward trend, growing from 5 in 2000 to 7 in 2020. These clusters were predominantly located in the adjacent Jining–Zaozhuang region and Lijin County within Dongying. The number of HL clusters remained constant at 2 between 2000 and 2015, located in Fangzi District of Weifang and Laoshan District of Qingdao. By 2020, only the Laoshan District in Qingdao remained as an HL cluster.

3.3. Exploring the Driving Factors of ESV in the SPUA

3.3.1. Drive Factor Processing

A mathematical assessment of the impacts of various factors on ESV is fundamental. Since spatial autocorrelation of ESV per unit area is evident only at the county and township scales, the study uses GeoDetector to investigate ESV driving factors at the county and township scales, with a focus on 2020 as the reference year. The explained variable is the ESV per unit area, and the explanatory variables are the 12 driving factors listed in Table 1. The OPGD model is employed to conduct both factor detection and interaction detection at the county and township scales. The classification methods are set as equal, natural, quantile, geometric, and standard deviation (sd), with the number of categories ranging from four to eight. The OPGD model calculates the q value of each driving factor under different classification methods and category numbers, and the combination (classification method and category number) that yields the highest q value for each driving factor is selected for discretization. This combination represents the optimal spatial discretization of the driving factor, as the one with the highest q value is considered the best [62].

3.3.2. ESV Factor Detector Analysis

The factor detection module interprets the driving force of each factor through q values. The value of q lies within the range of (0, 1), with larger values indicating more significant driving forces [26]. The results are shown in Figure 8.
Figure 8 reveals that at the county scale, the six factors with the greatest influence on ESV per unit area are ROD, PB, SLO, POP, NIG, and ELE, all with q-values greater than 0.3. RAD and PRE have the least influence, with q-values both below 0.07. The q-values for GDP, SOM, TEM, and NPP range from 0.15 to 0.19. Socio-economic factors have become the primary driving forces influencing the distribution of ESV per unit area at the county scale in the SPUA. Specifically, the q-value for ROD is 0.56277, and for PB, it is 0.47805, making them the dominant factors affecting the distribution of ESV per unit area. Natural factors also exert a relatively strong influence, with SLO having a q-value of 0.43489, making it the most important natural driving factor. ELE, with a q-value of 0.30622, and SOM, with a q-value of 0.1684, are also crucial natural driving factors.
At the township scale, the six factors with the greatest influence on ESV per unit area are PB, ROD, SLO, NIG, POP, and ELF, all with q-values greater than 0.23. RAD and PRE have the least influence, with q-values both below 0.038. The q-values for GDP, NPP, TEM, and SOM range from 0.06 to 0.16. Similar to the county scale, socio-economic factors are the main driving forces influencing the distribution of ESV per unit area at the township scale. The q-value for PB is 0.51023, making it the most critical factor driving ESV distribution at this scale. Natural factors also play a significant role, with SLO having a q-value of 0.35393 and ELE having a q-value of 0.23524, indicating that these are important natural driving factors.

3.3.3. ESV Interaction Detector Analysis

The factors driving ESV are interconnected and demonstrate mutual interactions. Interactions between natural factors and socio-economic factors, as well as between natural and socio-economic factors, can either enhance or weaken their influence on ESV. The interaction detection results are shown in Figure 9.
Figure 9 shows that at the county scale, 95.45% of the factor interactions have q-values that exceed those of the individual factors. Among these, 48.48% exhibit nonlinear enhancement, while 46.97% show bivariate enhancement. Additionally, 4.55% show univariate weakening, specifically for the interactions of SOL∩ELF, NIG∩GDP, and RAD∩GDP. This indicates the complexity of how factor interactions influence ESV per unit area.
In terms of ELF’s interactions with other variables, one pair shows univariate weakening, four pairs show nonlinear enhancement, and six pairs show bivariate enhancement. For SLO’s interactions with other variables, one pair exhibits univariate weakening, four pairs show nonlinear enhancement, and six pairs show bivariate enhancement. The interaction of the TEM with other variables results in seven pairs showing nonlinear enhancement and four pairs showing bivariate enhancement. For the PRE’s interactions with other variables, ten pairs show nonlinear enhancement, and one pair exhibits bivariate enhancement. SOM’s interactions with other variables result in seven pairs showing nonlinear enhancement and four pairs showing bivariate enhancement. In the interaction between NPP and other variables, eight pairs show nonlinear enhancement, and three pairs show bivariate enhancement. For POP interactions with other variables, four pairs exhibit nonlinear enhancement, while seven pairs exhibit bivariate enhancement. With respect to the GDP’s interactions with other variables, four pairs demonstrate nonlinear enhancement, five pairs show bivariate enhancement, and two pairs show univariate weakening. For NIG’s interactions with other variables, three pairs exhibit nonlinear enhancement, seven pairs exhibit bivariate enhancement, and one pair shows univariate weakening. The PB’s interactions with other variables show one pair with nonlinear enhancement and ten pairs with bivariate enhancement. The interaction of RAD with other variables displays ten pairs with nonlinear enhancement and one pair with univariate weakening. In terms of ROD’s interactions with other variables, two pairs exhibit nonlinear enhancement, while nine pairs demonstrate bivariate enhancement. Among these, the interaction between ROD and SLO has the highest q-value at 0.7923, followed by the interaction between ROD and ELF, with a q-value of 0.7694. The interactions between SLO and POP, as well as SLO and NIG, also exhibit high q-values of 0.7585 and 0.724, respectively.
At the township scale, 98.48% of the factor interactions have q-values exceeding those of individual factors. Among these, 39.39% exhibit nonlinear enhancement, and 59.09% show bivariate enhancement. Additionally, 1.52% show univariate weakening, specifically for the interaction between RAD and the PB. In terms of ELF’s interactions with other variables, three pairs show nonlinear enhancement, and eight pairs exhibit bivariate enhancement. In the interaction between SLO and other variables, two pairs show nonlinear enhancement, while nine pairs show bivariate enhancement. For the interaction of the TEM with other variables, eight pairs demonstrate nonlinear enhancement, and three pairs show bivariate enhancement. With respect to the PRE’s interactions, ten pairs exhibit nonlinear enhancement, and one pair shows bivariate enhancement. In terms of the SOM, seven pairs show nonlinear enhancement, and four pairs show bivariate enhancement. For the NPP, four pairs demonstrate nonlinear enhancement, while seven pairs show bivariate enhancement. For POP’s interactions with other variables, three pairs display nonlinear enhancement, while eight pairs show bivariate enhancement. With respect to the GDP, five pairs demonstrate nonlinear enhancement, and six pairs show bivariate enhancement. In NIG’s interactions, three pairs show nonlinear enhancement, and eight pairs show bivariate enhancement. The PB’s interactions show one pair with nonlinear enhancement, nine pairs with bivariate enhancement, and one pair with univariate weakening. RAD’s interactions include five pairs that show nonlinear enhancement, five pairs that exhibit bivariate enhancement, and one pair with univariate weakening. For ROD’s interactions, two pairs demonstrate nonlinear enhancement, and nine pairs exhibit bivariate enhancement. Among these, the interaction between ROD and SLO has the highest q-value, at 0.6781, followed by the interaction between SLO and the PB, with a q-value of 0.6679. The interactions between ROD and the PB, as well as between ELF and ROD, also show significant q-values of 0.6229 and 0.621, respectively.

4. Discussion

4.1. ESV Evolutionary Features

Functioning as the economic center and principal driving force of the Yellow River Basin, as well as a strategic zone for coordinated ecological–economic development [64], the ecosystem services of the SPUA have received increasing attention [64,65]. ESV represents a fundamental measure for evaluating regional sustainable development, and ESV evaluation has become a significant research hotspot. This study applies the equivalent coefficients table method to evaluate and assess the ESV of SPUA. Since the equivalence factor table proposed by Xie was based on the average state values across China and did not account for spatial differences between regions, we corrected the ESV in this study using the NDVI, which better reflects the impact of vegetation growth in different regions on the ESV. The results are more reasonable and align with the correction methods and outcomes from the studies by Liu [55].
When examining the total ESV, due to changes in land use, the overall trend shows a fluctuation characterized by “increase–decrease–increase”, with distinct stage characteristics. This substantially corresponds with the findings reported by Quan [31], Pan [28] and Liu [30]. This trend is mainly due to land use/cover changes, environmental factors, and ecological management policies. At different spatial scales, the distribution characteristics of the ESV per unit area vary significantly. However, unlike previous studies that focused more on differences at various grid scales, this research investigates the ESV from the administrative divisions of the township, county, and city levels. Townships function as the basic units for spatial planning and management, representing the most basic administrative level for decision-making. Counties act as the primary units for policy implementation, while cities have greater autonomy in policy formulation. Therefore, examining the spatial distribution differences in ESV per unit area across different administrative scales holds significant policy guidance and practical value. At various spatial scales, the level of spatial autocorrelation per unit ESV exhibits distinct differences. The township-level scale shows the strongest positive spatial autocorrelation, and as the spatial scale increases, the level of spatial autocorrelation gradually decreases. Both township and county-level scales display clear positive spatial autocorrelation, with the LISA maps predominantly showing high–high clusters and a considerable number of LL cluster areas. However, at the city level, no spatial autocorrelation is observed, and the distribution of ESV per unit area exhibits a certain degree of randomness. This study provides valuable insights for sustainable development priorities at different administrative divisions. At the township and county levels, land use/cover types such as forests and water bodies should be protected, and a focus on overall regional sustainable development should be maintained. Special attention needs to be given to ecosystem development within the various cluster types revealed by the LISA maps, such as HH and HL, because these regions have a significant impact on the value of regional ecosystem services. At the city level, comprehensive approaches should be adopted to promote regional ecosystem construction and sustainable development, such as strengthening the construction and protection of ecological protection zones and strictly limiting the occupation of forest and cropland by urban construction.

4.2. ESV Drive Mechanism

Understanding the driving mechanisms of ESV is essential for the rational management of ecosystem services, the regulation and optimization of ecosystem functions, and informed decision-making. The results of the geographical detector analysis indicate that at the county level, the average q value of the six socio-economic factors is 0.3388, while the average q value of the six natural factors is 0.2104. Among the six factors with the greatest influence on per unit ESV, four are socio-economic factors. At the township level, the average q value of the six socio-economic factors is 0.3043, and the average q value of the six natural factors is 0.1561, with four socio-economic factors also among the top six factors affecting ESV. In contrast to other study areas, such as the water source area of the central line project of the South-to-North Water Diversion and the Chinese section of the “Silk Road Economic Belt” [14], where natural factors dominate, socio-economic factors have become the primary drivers of the distribution of per unit ESV in the SPUA. This indicates that in rapidly developing urban clusters, socio-economic factors have become the main force driving the spatial distribution pattern of ESV. This finding is consistent with the Yellow River floodplain area [27], where socio-economic factors also play a dominant role, mainly because, as an urban agglomeration area with rapid economic development, economic and social development-oriented urban construction plays a decisive role in the transformation of land use/cover types, and changes in land use/cover types directly influence ESV. However, the influence of natural factors on ESV cannot be ignored. Slope and elevation are also important driving forces shaping the spatial distribution of ESV. At the county level, their q values reached 0.4389 and 0.3062, respectively, and at the township level, they reached 0.3539 and 0.2352, respectively. These results demonstrate the significant impact of slope and elevation on the spatial distribution pattern of ESV, which is consistent with findings from studies conducted in regions such as the Tibet Autonomous Region [30], the Shule River Basin [28], and the Nanjing Metropolitan Area of China [66].
At both the county and township levels, the influencing factors were ranked by their importance based on q values as follows: at the county level: Rod > Pbu > Slo > Pop > Nig > Ele > GDP > SOM > Tem > NPP > Rad > Pre; at the township level: Pbu > Rod > Slo > Nig > Pop > Ele > GDP > NPP > Tem > SOM > Rad > Pre. Aside from the substantial variation in the importance of NPP and SOM across different scales, the importance of other factors affecting ESV remained relatively consistent.
At different spatial scales, the interaction effects between different factors displayed complexity. At both the county and township levels, more than 95% of factor interactions had a greater impact on ESV than single factors alone. However, at the county level, the interactions among SOL∩ELF, NIG∩GDP, and RAD∩GDP, as well as RAD∩PB at the township level, exhibited a univariable weakening effect. For different factors, the interaction effects of ROD, PB, SLO, and POP with other factors are significantly greater than those of PRE and GDP with other factors. Notably, for SLO, at the county level, the q values for its interaction with other factors are mostly above 0.55, while at the township level, they are mostly above 0.45. This demonstrates a significant enhancement effect on the ESV, indicating that SLO, as a critical natural factor, plays an important role in shaping the ESV distribution through its interaction with other factors. This is mainly due to the diverse topography of the SPUA, which includes plains, mountains, hills, and other terrains; the combination of terrain and other factors can significantly alter the original magnitude of influence. The spatial patterns of regional ESV, as revealed by interaction detection, emerge from the interplay of multiple driving forces rather than individual factors. At the county level, three factors exhibited a univariable weakening effect, whereas at the township level, only one factor showed this effect, indicating that the driving mechanisms vary across different scales. Overall, the driving mechanisms at the county level are more complex and diverse. Thus, when optimizing ecosystems and managing risks in the SPUA, it is essential to consider the interactions and synergies between different driving factors. Adopting targeted and differentiated development models is necessary, and strategic planning should account for the distinct characteristics at different spatial scales to enhance the effectiveness and sustainability of regional ecological management. At the regional and global levels, emphasis should also be placed on evaluating ESV and achieving sustainable development. To identify the key driving factors of ESV, we need to balance the relationships between protection and development. On the one hand, we need to eliminate all behaviors that damage the ecology, and on the other hand, we need to consider the development needs of the region. Through reasonable means, residents can benefit from protection and actively participate in protective work to achieve sustainable development. At the same time, targeted arrangements should be made in economic and social development planning, land planning, ecological protection planning, and other aspects, taking into account the influence and interaction of different factors, such as natural and socio-economic factors, to provide support for sustainable development at regional and global levels.

4.3. Limitations and Prospects for Future Work

There are various methods for evaluating ESV, among which the equivalent factor method proposed by Xie is the most widely applied in studies of relevant regions in China. On the basis of the characteristics of the SPUA, the accessibility of research data, and the operability of research methods, this study selected the equivalent coefficients table method combined with the NDVI to assess ESV. However, due to technical limitations, no secondary classification was conducted for the same land use type, and only the NDVI was used to adjust the ESV, which could potentially influence the precision of the ESV assessment results. Moving forward, it is essential to develop an ESV evaluation and correction model that more accurately represents the actual conditions of the SPUA. Moreover, this research used the geographic detector to assess individual factors and their interactions on ESV but failed to consider regional variations in factor impacts caused by differences in natural conditions and socio-economic development. Future research could incorporate the spatiotemporal heterogeneity of influencing factors into the research framework.

5. Conclusions

This investigation used the equivalent coefficients table method to assess the ESV of SPUA. At multiple spatial scales, including urban, county, and township levels, the spatio-temporal dynamics of ESV and their driving forces were analyzed using spatial analysis and the geographic detector method. The results indicate the following:
(1) From 2000 to 2020, the ESV of SPUA exhibited a trend of “increase-decrease-increase”, with distinct phase characteristics. The inflection points occurred in 2010 and 2015. The ESV of agricultural land was substantially higher than that of other land use types, yet it demonstrated a continuous decrease over time. Forestland and water bodies had relatively high ESV, with forestland showing a trend of “decrease—increase” and water bodies showing a trend of “increase—decrease—increase”. Grassland and unused land had relatively low ESV and showed a rapid decline. The overall evolutionary trend of ESV in the SPUA demonstrated complexity and diversity.
(2) Research on per-unit-area ESV at three spatial scales, including city, county, and township, revealed that as the spatial scale increased, the level of spatial autocorrelation gradually decreased. Among these, the township-level scale exhibited the strongest positive spatial autocorrelation, followed by the county-level scale, while the city scale did not show spatial autocorrelation, instead presenting a random distribution. The LISA map studies at the county and township scales indicated that at both scales, the high–high (HH) aggregation type was the most prevalent. Over the study period, both the number and spatial arrangement of different regional types underwent changes, reflecting the diverse ESV distributions at the county and township scales. A more distinct aggregation pattern emerged at the township level, and the stability across the years was higher than that at the county level. These findings provide guidance for ecosystem construction and sustainable development at different scales. At the township and county levels, the focus should be on the protection of land use/cover types such as forests and water bodies, whereas at the city level, the focus should be on macroplanning to promote regional ecosystem construction.
(3) At the county scale, roadway density, the proportion of built-up land, night light, and population density were the main socio-economic driving factors. At the township scale, the proportion of built-up land, roadway density, night light, and population density also emerged as the primary socio-economic driving forces. At both the county and township scales, slope and elevation were the most significant natural driving factors. Overall, the relative importance of each influencing factor remained consistent across different spatial scales, with socio-economic factors generally outweighing natural factors in terms of significance. Special attention needs to be given to the various impacts of different factors to provide a foundation for further ecological protection and the construction of an ecosystem service security pattern.
(4) Interaction detection analysis revealed that multiple driving factors working in combination, rather than any single factor, determine the spatial distribution of ESV in the region. Moreover, the scale effect of these interacting factors was highly pronounced, showing more complex interaction results at the county scale: nonlinear enhancement accounted for 48.48%, bivariate enhancement accounted for 46.97%, and univariate weakening accounted for 4.55%. With respect to different factors, the interactions between roadway density, the proportion of built-up land, slope, and population density with other factors were much stronger than those involving precipitation and the GDP. Notably, slope exhibited a significant enhancement effect at both township and county scales, highlighting the crucial role of natural factors in shaping ESV distribution, both through their direct influence and their interaction with other factors. Furthermore, on the basis of the impact of different factors on ESV, sustainable development in the region and the world can be achieved through the establishment, planning, and use of modern technology in ecological reserves.

Supplementary Materials

The following supporting information can be downloaded at https://www.mdpi.com/article/10.3390/su17104393/s1: Table S1: Statistics of Different Cluster Types in LISA Maps (by County); Table S2: Statistics of Different Cluster Types in LISA Maps (by Number of Towns).

Author Contributions

Conceptualization, methodology, software, formal analysis, investigation, resources, data curation, writing—original draft, visualization, and funding acquisition, Y.L.; software, writing—review and editing, and validation, T.Z. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by the Humanity and Social Science Foundation of the Ministry of Education of China, No. 20YJCZH104.

Data Availability Statement

The data that support the findings of this study are available upon reasonable request from the corresponding author.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Costanza, R.; d’Arge, R.; de Groot, R.; Farber, S.; Grasso, M.; Hannon, B.; Limburg, K.; Naeem, S.; O’Neill, R.V.; Paruelo, J.; et al. The value of the world’s ecosystem services and natural capital. Nature 1997, 387, 253–260. [Google Scholar] [CrossRef]
  2. Wang, L.-J.; Ma, S.; Zhao, Y.-G.; Zhang, J.-C. Ecological restoration projects did not increase the value of all ecosystem services in Northeast China. For. Ecol. Manag. 2021, 495, 119340. [Google Scholar] [CrossRef]
  3. Wang, X.; Dong, X.; Liu, H.; Wei, H.; Fan, W.; Lu, N.; Xu, Z.; Ren, J.; Xing, K. Linking land use change, ecosystem services and human well-being: A case study of the Manas River Basin of Xinjiang, China. Ecosyst. Serv. 2017, 27, 113–123. [Google Scholar] [CrossRef]
  4. Reid, W.V.; Mooney, H.A.; Cropper, A.; Capistrano, D.; Carpenter, S.R.; Chopra, K.; Dasgupta, P.; Dietz, T.; Duraiappah, A.K.; Hassan, R.; et al. Ecosystems and Human Well-Being: Synthesis; Island Press: Washington, DC, USA, 2005. [Google Scholar]
  5. Song, W.; Deng, X. Land-use/land-cover change and ecosystem service provision in China. Sci. Total Environ. 2017, 576, 705–719. [Google Scholar] [CrossRef]
  6. Tolessa, T.; Senbeta, F.; Kidane, M. The impact of land use/land cover change on ecosystem services in the central highlands of Ethiopia. Ecosyst. Serv. 2017, 23, 47–54. [Google Scholar] [CrossRef]
  7. Ouyang, Z.; Zheng, H.; Xiao, Y.; Polasky, S.; Liu, J.; Xu, W.; Wang, Q.; Zhang, L.; Xiao, Y.; Rao, E.; et al. Improvements in ecosystem services from investments in natural capital. Science 2016, 352, 1455–1459. [Google Scholar] [CrossRef]
  8. Kindu, M.; Schneider, T.; Teketay, D.; Knoke, T. Changes of ecosystem service values in response to land use/land cover dynamics in Munessa-Shashemene landscape of the Ethiopian highlands. Sci. Total Environ. 2016, 547, 137–147. [Google Scholar] [CrossRef]
  9. Liu, Y.; Zhang, Y. Responses of Ecosystem Services to Land Use/Cover Changes in Rapidly Urbanizing Areas: A Case Study of the Shandong Peninsula Urban Agglomeration. Sustainability 2024, 16, 6100. [Google Scholar] [CrossRef]
  10. Jantz, C.A.; Manuel, J.J. Estimating impacts of population growth and land use policy on ecosystem services: A community-level case study in Virginia, USA. Ecosyst. Serv. 2013, 5, 110–123. [Google Scholar] [CrossRef]
  11. Gao, F.; Cui, J.; Zhang, S.; Xin, X.; Zhang, S.; Zhou, J.; Zhang, Y. Spatio-Temporal Distribution and Driving Factors of Ecosystem Service Value in a Fragile Hilly Area of North China. Land 2022, 11, 2242. [Google Scholar] [CrossRef]
  12. Baude, M.; Meyer, B.C.; Schindewolf, M. Land use change in an agricultural landscape causing degradation of soil based ecosystem services. Sci. Total. Environ. 2019, 659, 1526–1536. [Google Scholar] [CrossRef] [PubMed]
  13. Sannigrahi, S.; Chakraborti, S.; Joshi, P.K.; Keesstra, S.; Sen, S.; Paul, S.K.; Kreuter, U.; Sutton, P.C.; Jha, S.; Dang, K.B. Ecosystem service value assessment of a natural reserve region for strengthening protection and conservation. J. Environ. Manag. 2019, 244, 208–227. [Google Scholar] [CrossRef] [PubMed]
  14. Kang, L.; Jia, Y.; Zhang, S. Spatiotemporal distribution and driving forces of ecological service value in the Chinese section of the “Silk Road Economic Belt”. Ecol. Indic. 2022, 141, 109074. [Google Scholar] [CrossRef]
  15. Jiang, L.; Wang, Z.; Zuo, Q.; Du, H. Simulating the impact of land use change on ecosystem services in agricultural production areas with multiple scenarios considering ecosystem service richness. J. Clean. Prod. 2023, 397, 136485. [Google Scholar] [CrossRef]
  16. Wang, Y.; Li, X.; Zhang, Q.; Li, J.; Zhou, X. Projections of future land use changes: Multiple scenarios-based impacts analysis on ecosystem services for Wuhan city, China. Ecol. Indic. 2018, 94, 430–445. [Google Scholar] [CrossRef]
  17. Elmqvist, T.; Setälä, H.; Handel, S.N.; van der Ploeg, S.; Aronson, J.; Blignaut, J.N.; Gómez-Baggethun, E.; Nowak, D.J.; Kronenberg, J.; de Groot, R. Benefits of restoring ecosystem services in urban areas. Curr. Opin. Environ. Sustain. 2015, 14, 101–108. [Google Scholar] [CrossRef]
  18. Peng, K.; Jiang, W.; Ling, Z.; Hou, P.; Deng, Y. Evaluating the potential impacts of land use changes on ecosystem service value under multiple scenarios in support of SDG reporting: A case study of the Wuhan urban agglomeration. J. Clean. Prod. 2021, 307, 127321. [Google Scholar] [CrossRef]
  19. Martínez-Harms, M.J.; Balvanera, P. Methods for mapping ecosystem service supply: A review. Int. J. Biodivers. Sci. Ecosyst. Serv. Manag. 2012, 8, 17–25. [Google Scholar] [CrossRef]
  20. Eigenbrod, F.; Armsworth, P.R.; Anderson, B.J.; Heinemeyer, A.; Gillings, S.; Roy, D.B.; Thomas, C.D.; Gaston, K.J. The impact of proxy-based methods on mapping the distribution of ecosystem services. J. Appl. Ecol. 2010, 47, 377–385. [Google Scholar] [CrossRef]
  21. Plummer, M.L. Assessing benefit transfer for the valuation of ecosystem services. Front. Ecol. Environ. 2009, 7, 38–45. [Google Scholar] [CrossRef]
  22. Xie, G.-d.; Lu, C.-X.; Leng, Y.-F.; Zheng, D.; LI, S.-c. Ecological assets valuation of the Tibetan Plateau. J. Nat. Resour. 2003, 18, 189–196. (In Chinese) [Google Scholar]
  23. Xie, G.-d.; Zhang, C.-x.; Zhang, L.-m.; Chen, W.-h.; Li, S.-m. Improvement of the evaluation method for ecosystem service value based on per unit area. J. Nat. Resour. 2015, 30, 1243–1254. [Google Scholar]
  24. Zhao, X.; Wang, J.; Su, J.; Sun, W. Ecosystem service value evaluation method in a complex ecological environment: A case study of Gansu Province, China. PLoS ONE 2021, 16, e0240272. [Google Scholar] [CrossRef] [PubMed]
  25. Yang, Y.; Qin, T.; Yan, D.; Liu, S.; Feng, J.; Wang, Q.; Liu, H.; Gao, H. Analysis of the evolution of ecosystem service value and its driving factors in the Yellow River Source Area, China. Ecol. Indic. 2024, 158, 111344. [Google Scholar] [CrossRef]
  26. Zhang, X.; Zheng, Z.; Sun, S.; Wen, Y.; Chen, H. Study on the driving factors of ecosystem service value under the dual influence of natural environment and human activities. J. Clean. Prod. 2023, 420, 138408. [Google Scholar] [CrossRef]
  27. Zhao, H.; Xu, X.; Tang, J.; Wang, Z.; Miao, C. Understanding the key factors and future trends of ecosystem service value to support the decision management in the cluster cities around the Yellow River floodplain area. Ecol. Indic. 2023, 154, 110544. [Google Scholar] [CrossRef]
  28. Pan, N.; Guan, Q.; Wang, Q.; Sun, Y.; Li, H.; Ma, Y. Spatial Differentiation and Driving Mechanisms in Ecosystem Service Value of Arid Region: A case study in the middle and lower reaches of Shule River Basin, NW China. J. Clean. Prod. 2021, 319, 128718. [Google Scholar] [CrossRef]
  29. Liao, Q.; Li, T.; Wang, Q.; Liu, D. Exploring the ecosystem services bundles and influencing drivers at different scales in southern Jiangxi, China. Ecol. Indic. 2023, 148, 110089. [Google Scholar] [CrossRef]
  30. Liu, J.; Pei, X.; Liao, B.; Zhang, H.; Liu, W.; Jiao, J. Scale effects and spatial heterogeneity of driving factors in ecosystem services value interactions within the Tibet autonomous region. J. Environ. Manag. 2024, 351, 119871. [Google Scholar] [CrossRef]
  31. Quan, L.-a.; Jin, S.; Chen, J.; Li, T. Evolution and Driving Forces of Ecological Service Value in Anhui Based on Landsat Land Use and Land Cover Change. Remote Sens. 2024, 16, 269. [Google Scholar] [CrossRef]
  32. Hu, B.; Xie, M.; He, R.; Shi, Z.; Zhou, Y.; Ni, H.; Li, H. Spatio-temporal evolution of cropland ecosystem services value and its spatially varying dominate over the past two decades: A case study in Jiangxi Province in southern China. J. Clean. Prod. 2023, 427, 139228. [Google Scholar] [CrossRef]
  33. Wang, Y.; Shataer, R.; Zhang, Z.; Zhen, H.; Xia, T. Evaluation and Analysis of Influencing Factors of Ecosystem Service Value Change in Xinjiang under Different Land Use Types. Water 2022, 14, 1424. [Google Scholar] [CrossRef]
  34. Zhang, W.; Wang, F.-F.; Shan, W. Spatial and temporal changes in ecosystem services value and analysis of driving factors in the Yangtze River Delta Region. Open Geosci. 2024, 16, 20220604. [Google Scholar] [CrossRef]
  35. Song, F.; Su, F.; Mi, C.; Sun, D. Analysis of driving forces on wetland ecosystem services value change: A case in Northeast China. Sci. Total Environ. 2021, 751, 141778. [Google Scholar] [CrossRef]
  36. Ding, X.; Shu, Y.; Tang, X.; Ma, J. Identifying Driving Factors of Basin Ecosystem Service Value Based on Local Bivariate Spatial Correlation Patterns. Land 2022, 11, 1852. [Google Scholar] [CrossRef]
  37. Zhang, X.; Xie, B.; Li, J.; Yuan, C. Spatiotemporal Distribution and Driving Force Analysis of the Ecosystem Service Value in the Fujiang River Basin, China. Land 2023, 12, 449. [Google Scholar] [CrossRef]
  38. Shao, M.; Wu, L.; Li, F.; Lin, C. Spatiotemporal Dynamics of Ecosystem Services and the Driving Factors in Urban Agglomerations: Evidence From 12 National Urban Agglomerations in China. Front. Ecol. Evol. 2022, 10, 804969. [Google Scholar] [CrossRef]
  39. Xue, S.; Yao, L.; Xu, Y.; Li, C. Spatiotemporal Dynamics and Driving Factors of Ecosystem Services in the Yellow River Delta, China. Sustainability 2024, 16, 3466. [Google Scholar] [CrossRef]
  40. Yang, R.; Ren, F.; Xu, W.; Ma, X.; Zhang, H.; He, W. China’s ecosystem service value in 1992–2018: Pattern and anthropogenic driving factors detection using Bayesian spatiotemporal hierarchy model. J. Environ. Manag. 2022, 302, 114089. [Google Scholar] [CrossRef]
  41. Alexander, P.; Rounsevell, M.D.A.; Dislich, C.; Dodson, J.R.; Engström, K.; Moran, D. Drivers for global agricultural land use change: The nexus of diet, population, yield and bioenergy. Glob. Environ. Change 2015, 35, 138–147. [Google Scholar] [CrossRef]
  42. Kremen, C. Managing ecosystem services: What do we need to know about their ecology? Ecol. Lett. 2005, 8, 468–479. [Google Scholar] [CrossRef] [PubMed]
  43. Assessment, M.E. Ecosystems and Human Well-Being: Wetlands and Water; World Resources Institute: Washington, DC, USA, 2005. [Google Scholar]
  44. Ashrafi, S.; Kerachian, R.; Pourmoghim, P.; Behboudian, M.; Motlaghzadeh, K. Evaluating and improving the sustainability of ecosystem services in river basins under climate change. Sci Total Environ. 2022, 806 Pt 3, 150702. [Google Scholar] [CrossRef] [PubMed]
  45. Ouyang, X.; Tang, L.; Wei, X.; Li, Y. Spatial interaction between urbanization and ecosystem services in Chinese urban agglomerations. Land Use Policy 2021, 109, 105587. [Google Scholar] [CrossRef]
  46. Xiong, C.; Ren, H.; Xu, D.; Gao, Y. Spatial scale effects on the value of ecosystem services in China’s terrestrial area. J. Environ. Manag. 2024, 366, 121745. [Google Scholar] [CrossRef]
  47. Sun, W.; Li, D.; Wang, X.; Li, R.; Li, K.; Xie, Y. Exploring the scale effects, trade-offs and driving forces of the mismatch of ecosystem services. Ecol. Indic. 2019, 103, 617–629. [Google Scholar] [CrossRef]
  48. Bai, Y.; Chen, Y.; Alatalo, J.M.; Yang, Z.; Jiang, B. Scale effects on the relationships between land characteristics and ecosystem services- a case study in Taihu Lake Basin, China. Sci Total Environ. 2020, 716, 137083. [Google Scholar] [CrossRef]
  49. Wang, Y.; Niu, Y.; Li, M.; Yu, Q.; Chen, W. Spatial structure and carbon emission of urban agglomerations: Spatiotemporal characteristics and driving forces. Sustain. Cities Soc. 2022, 78, 103600. [Google Scholar] [CrossRef]
  50. Berihun, M.L.; Tsunekawa, A.; Haregeweyn, N.; Tsubo, M.; Fenta, A.A. Changes in ecosystem service values strongly influenced by human activities in contrasting agro-ecological environments. Ecol. Process. 2021, 10, 52. [Google Scholar] [CrossRef]
  51. Ai, M.; Chen, X.; Yu, Q. Spatial correlation analysis between human disturbance intensity (HDI) and ecosystem services value (ESV) in the Chengdu-Chongqing urban agglomeration. Ecol. Indic. 2024, 158, 111555. [Google Scholar] [CrossRef]
  52. Yang, J.; Huang, X. The 30 m annual land cover dataset and its dynamics in China from 1990 to 2019. Earth Syst. Sci. Data 2021, 13, 3907–3925. [Google Scholar] [CrossRef]
  53. XIE, G.-d.; Zhen, L.; LU, C.-x.; Xiao, Y.; Chen, C. Expert knowledge based valuation method of ecosystem services in China. J. Nat. Resour. 2008, 23, 911–919. (In Chinese) [Google Scholar]
  54. Liu, Z.; Wang, S.; Fang, C. Spatiotemporal evolution and influencing mechanism of ecosystem service value in the Guangdong-Hong Kong-Macao Greater Bay Area. J. Geogr. Sci. 2023, 33, 1226–1244. [Google Scholar] [CrossRef]
  55. Liu, Z.; Wu, R.; Chen, Y.; Fang, C.; Wang, S. Factors of ecosystem service values in a fast-developing region in China: Insights from the joint impacts of human activities and natural conditions. J. Clean. Prod. 2021, 297, 126588. [Google Scholar] [CrossRef]
  56. Hu, X.; Ma, C.; Huang, P.; Guo, X. Ecological vulnerability assessment based on AHP-PSR method and analysis of its single parameter sensitivity and spatial autocorrelation for ecological protection—A case of Weifang City, China. Ecol. Indic. 2021, 125, 107464. [Google Scholar] [CrossRef]
  57. Cao, X.; Liu, Y.; Li, T.; Liao, W. Analysis of Spatial Pattern Evolution and Influencing Factors of Regional Land Use Efficiency in China Based on ESDA-GWR. Sci. Rep. 2019, 9, 520. [Google Scholar] [CrossRef]
  58. Anselin, L. Local indicators of spatial association—LISA. Geogr. Anal. 1995, 27, 93–115. [Google Scholar] [CrossRef]
  59. Wang, J.F.; Li, X.H.; Christakos, G.; Liao, Y.L.; Zhang, T.; Gu, X.; Zheng, X.Y. Geographical Detectors–Based Health Risk Assessment and its Application in the Neural Tube Defects Study of the Heshun Region, China. Int. J. Geogr. Inf. Sci. 2010, 24, 107–127. [Google Scholar] [CrossRef]
  60. Wang, J.-F.; Zhang, T.-L.; Fu, B.-J. A measure of spatial stratified heterogeneity. Ecol. Indic. 2016, 67, 250–256. [Google Scholar] [CrossRef]
  61. Wang, J.; Haining, R.; Zhang, T.; Xu, C.; Hu, M.; Yin, Q.; Li, L.; Zhou, C.; Li, G.; Chen, H. Statistical Modeling of Spatially Stratified Heterogeneous Data. Ann. Am. Assoc. Geogr. 2024, 114, 499–519. [Google Scholar] [CrossRef]
  62. Song, Y.; Wang, J.; Ge, Y.; Xu, C. An optimal parameters-based geographical detector model enhances geographic characteristics of explanatory variables for spatial heterogeneity analysis: Cases with different types of spatial data. GIScience Remote Sens. 2020, 57, 593–610. [Google Scholar] [CrossRef]
  63. Wang, J.; Xu, C.D. Geodetector: Principle and prospective. Acta Geogr. Sin. 2017, 72, 116–134. (In Chinese) [Google Scholar]
  64. Yan, S.; Chen, H.; Quan, Q.; Liu, J. Evolution and coupled matching of ecosystem service supply and demand at different spatial scales in the Shandong Peninsula urban agglomeration, China. Ecol. Indic. 2023, 155, 111052. [Google Scholar] [CrossRef]
  65. Wang, C.; Wang, X.; Wang, Y.; Zhan, J.; Chu, X.; Teng, Y.; Liu, W.; Wang, H. Spatio-temporal analysis of human wellbeing and its coupling relationship with ecosystem services in Shandong province, China. J. Geogr. Sci. 2023, 33, 392–412. [Google Scholar] [CrossRef]
  66. Chen, S.; Liu, X.; Yang, L.; Zhu, Z. Variations in Ecosystem Service Value and Its Driving Factors in the Nanjing Metropolitan Area of China. Forests 2023, 14, 113. [Google Scholar] [CrossRef]
Sustainability 17 04393 g001
Figure 1. Geographical location of the study area.
Figure 1. Geographical location of the study area.
Sustainability 17 04393 g001
Sustainability 17 04393 g002
Figure 2. ESV of various land use/cover types in the SPUA from 2000 to 2020.
Figure 2. ESV of various land use/cover types in the SPUA from 2000 to 2020.
Sustainability 17 04393 g002
Sustainability 17 04393 g003
Figure 3. Average ESV map at the city level.
Figure 3. Average ESV map at the city level.
Sustainability 17 04393 g003
Sustainability 17 04393 g004
Figure 4. Average ESV map at the county level.
Figure 4. Average ESV map at the county level.
Sustainability 17 04393 g004
Sustainability 17 04393 g005
Figure 5. Average ESV map at the township level.
Figure 5. Average ESV map at the township level.
Sustainability 17 04393 g005
Sustainability 17 04393 g006
Figure 6. LISA map at the township level.
Figure 6. LISA map at the township level.
Sustainability 17 04393 g006
Sustainability 17 04393 g007
Figure 7. LISA map at the county level.
Figure 7. LISA map at the county level.
Sustainability 17 04393 g007
Sustainability 17 04393 g008
Figure 8. Map of factor detector results in the SPUA.
Figure 8. Map of factor detector results in the SPUA.
Sustainability 17 04393 g008
Sustainability 17 04393 g009
Figure 9. Map of interaction detector results in the SPUA (En, no indicates nonlinear enhancement, En, bi indicates bivariate enhancement, and We, un indicates univariate weakening).
Figure 9. Map of interaction detector results in the SPUA (En, no indicates nonlinear enhancement, En, bi indicates bivariate enhancement, and We, un indicates univariate weakening).
Sustainability 17 04393 g009
Table 1. This study’s data sources for driving factors.
Table 1. This study’s data sources for driving factors.
TypesDriving FactorsResolutionSource
Natural factorsElevation (ELE)30 mNASA dataset: Available online: www.earthdata.nasa.gov (accessed on 1 July 2024)
Slope (SLO)30 mNASA dataset: Available online: www.earthdata.nasa.gov (accessed on 1 July 2024)
Temperature (TEM)1 kmNational Earth System Science Data Center: Available online: https://www.geodata.cn (accessed on 1 July 2024)
Precipitation (PRE)1 kmNational Earth System Science Data Center: Available online: https://www.geodata.cn (accessed on 1 July 2024)
Soil organic matter (SOM)1 kmNational Tibetan Plateau/Third Pole Environment Data Center: Available online: https://data.tpdc.ac.cn (accessed on 1 July 2024)
Net Primary Productivity (NPP)500 mMODIS/Terra Net Primary Production Gap-Filled Yearly L4 Global 500 m SIN Grid: Available online: https://lpdaac.usgs.gov/products/mod17a3hgfv006/ (accessed on 1 July 2024)
Socio-economic factorsPopulation density (POP)1 kmResource and Environmental Science Data Platform: Available online: www.resdc.cn (accessed on 1 July 2024)
Gross Domestic Product per unit of land (GDP)1 kmResource and Environmental Science Data Platform: Available online: www.resdc.cn (accessed on 1 July 2024)
Night light (NIG)1 kmGlobal Change Research Data Publishing & Repository: Available online: ww.geodoi.ac.cn (accessed on 1 July 2024)
Proportion of bult-up land (PB)30 mZenodo: Available online: https://zenodo.org/records/8176941 (accessed on 1 July 2024)
Roadway density (ROD)--Gaode Maps: Available online: https://lbs.amap.com/(accessed on 1 July 2024)
Railway density (RAD)--Gaode Maps: Available online: https://lbs.amap.com/(accessed on 1 July 2024)
Table 2. ESV coefficients per unit area in the SPUA.
Table 2. ESV coefficients per unit area in the SPUA.
Ecosystem ServicesUnit Area Value Coefficient by Land Use/Cover Type (CNY·hm−2·a−1)
ForestGrasslandCroplandWater BodiesUnused Land
Provisioning services7068.791687.112968.461879.32128.14
Regulating services30,325.3212,599.968222.0177,286.861110.5
Supporting services18,216.558777.265317.618200.651217.28
Cultural services4442.021857.96363.059482512.54
Total60,052.6824,922.2916,871.1396,848.832968.46
Table 3. Moran’s I values at multiple scales.
Table 3. Moran’s I values at multiple scales.
YearScaleMoran’s Ip
2000Township0.6270.001 ***
County0.2010.001 ***
City0.0120.309
2005Township0.6310.001 ***
County0.1170.01 ***
City−0.1130.442
2010Township0.6210.001 ***
County0.0710.058 *
City−0.0480.419
2015Township0.6220.001 ***
County0.1330.005 ***
City−0.0230.369
2020Township0.6230.001 ***
County0.1580.001 ***
City0.0210.295
*** and * indicate significance levels of 1% and 10%, respectively.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Liu, Y.; Zhang, T. Spatio-Temporal Dynamics and Driving Forces of Ecosystem Service Value at Multiple Scales in the Shandong Peninsula Urban Agglomeration, China. Sustainability 2025, 17, 4393. https://doi.org/10.3390/su17104393

AMA Style

Liu Y, Zhang T. Spatio-Temporal Dynamics and Driving Forces of Ecosystem Service Value at Multiple Scales in the Shandong Peninsula Urban Agglomeration, China. Sustainability. 2025; 17(10):4393. https://doi.org/10.3390/su17104393

Chicago/Turabian Style

Liu, Yongwei, and Tianping Zhang. 2025. "Spatio-Temporal Dynamics and Driving Forces of Ecosystem Service Value at Multiple Scales in the Shandong Peninsula Urban Agglomeration, China" Sustainability 17, no. 10: 4393. https://doi.org/10.3390/su17104393

APA Style

Liu, Y., & Zhang, T. (2025). Spatio-Temporal Dynamics and Driving Forces of Ecosystem Service Value at Multiple Scales in the Shandong Peninsula Urban Agglomeration, China. Sustainability, 17(10), 4393. https://doi.org/10.3390/su17104393

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop