Multi-Scale Analysis of Ecosystem Service Trade-Offs/Synergies in the Yangtze River Delta

Abstract

The transformation of ecosystem structure leads to changes in ecosystem services (ESs) and their relationship. However, most research in this area has focused on particular scales and timeframes, often overlooking the significance of spatial and temporal variations. Therefore, we used the equivalent value method to evaluate seven typical ESs in the Yangtze River Delta (YRD) between 2000 and 2020: food production (FP), water supply (WS), climate regulation (CR), environmental purification (EP), soil conservation (SC), biodiversity maintenance (BM), and aesthetic landscape (AL). We further employed the Spearman correlation coefficient and bivariate Moran’s I to evaluate the relationship of ESs and their spatial heterogeneity at grid, township, county and city scales. Our results show that (1) All ESs except AL exhibited a fluctuating upward trend from 2000 to 2020, resulting in a total increase in ecosystem service (ES) value of RMB 650.63 billion. (2) Approximately 70% of the ES pairs demonstrated a synergistic relationship, with the exception of FP and other ESs, which mainly showed a trade-off. (3) With the scale increased from grid to city level, the degree of trade-off between FP and other ESs strengthened at different levels, while the synergy degree of among other ESs gradually decreased. (4) The relationship between ESs demonstrated strong spatial heterogeneity, with FP and other ESs exhibiting trade-offs primarily in the northern and southern YRD, while other ES pairs exhibited mostly synergy in these regions. This study provides scientific information for governments to optimize land use distribution and improve ESs.

1. Introduction

Nature provides fundamental resources for human survival. Ecosystem services (ESs) encompass various benefits that humans derive directly or indirectly from nature [1]. As the concept of sustainable development and ES gains more attention, people are increasingly recognizing that the ultimate research goal on ESs is incorporating this concept into urban planning and management [2,3], thereby promoting the coordinated development of natural ecological systems and socio-economic systems. However, due to human activities and global climate change, significant changes in ecosystem structure have negatively impacted the provisioning of ecosystem services (ESs) [4,5]. Relevant studies have demonstrated that 15 out of the 24 global ESs have been degraded to varying degrees [6]. Thus, different countries have taken different measures aimed at improving the level of ESs and human well-being based on the related studies.

The evaluation of ES values (ESVs) primarily relies on the service function price per unit area and value equivalent factor per unit area [7]. The former one is primarily concerned with value accounting that is predicated on the evaluation of the physical quantity of ESs, which needs many ecological parameters. The latter one obtains the equivalent value based on meta-analysis and the area of each ecosystem, which is more effective and widely used in research. On the basis of evaluation of ES material or value, scholars have conducted extensive research across a multitude of domains, such as driving mechanisms, ES relationships, human welfare, landscape planning, and ecosystem management [8]. A large number of studies have shown that the impacts of drivers vary across different ESs [9]. The fluctuating dynamics of ESs leads to evolving trade-offs and synergies among them, which are crucial for sustainable ecosystem management [10]. Based on their different changing trends, the relationship can be classified into synergy, trade-off, and neutrality, which means that one ES may have positive, negative, and negligible effects on another ES [11]. Trade-off/synergy analysis methods mainly include correlation analysis [12], scenario analysis [13], and ES bundles [14]. Compared with other methods, correlation analysis has the advantages of being simple and intuitive, requiring a small amount of data and having high flexibility. Based on the previous studies, this study will use correlation analysis to analyze the ES trade-offs/synergies based on the ES value which is calculated by value equivalent factor per unit area.

The relationship between ESs is not static but evolves with changes in time, spatial scale, and surrounding environment [15]. Different studies have analyzed trade-offs/synergies at global [16], national [17], regional, and landscape scales [18], where ESs exhibit complex interactions [19]. Drivers, including land use, socio-economic conditions, and meteorological factors, evolve over time and vary across spatial scales [20], potentially leading to diverse change trends among various ESs and affecting the relationship between ESs. However, most studies have researched the relationship between ESs at a single scale, neglecting their scale effects. The relationship between ESs also changes over time due to changes in key factors [21]. Studies in the Baota Mountain area by Hou [22] found that the trade-off and synergy degrees between grain and aquatic production varied with time at grid scale. However, most existing research has only examined their relationship at a single point in time. In addition, spatial heterogeneity of the relationship between ESs (spatial trade-offs) arises from different ecological process depending on the diverse natural environment and socio-economic situation. Therefore, the relationships between the same pairs of ESs may vary due to regional environmental differences. Many studies have used Moran’s Index to analyze the spatial distribution of individual ESs, which does not reflect the relationships between ESs as it pertains to only one ES. Furthermore, few studies have explored the spatial trade-offs in the relationships between ESs. Consequently, we have introduced the bivariate Moran’s Index, which considers two ESs, to assess the spatial trade-offs and synergies between them. A limited number of studies have explored the temporal changes and scale effects in the relationships between ESs, leading to an incomplete understanding of them [22,23,24,25,26]. Current studies have only revealed the average situation of ES trade-offs/synergies in a specific research area [27], without identifying specific locations. Understanding the temporal and spatial characteristics of the relationships between ESs is fundamental for urban development planning and is increasingly incorporated into decision-making by governments and environmental organizations [28]. Therefore, there has been a growing body of research on the scale effects of ES trade-offs/synergies, and some progress has been achieved; however, a considerable gap still exists in the depth and breadth of the related research. Firstly, the range of ESs addressed in current research is not comprehensive enough, and the temporal scope of the studies is also inadequate. Secondly, the prevailing research has a strong emphasis on the scale effects of the relationship between ESs, frequently neglecting the examination of their spatial distribution patterns.

Benefiting from unique terrain, climate, and other natural factors, the Yangtze River Delta (YRD) has formed a unique ecosystem which has a great influence on the ecological security of China. Additionally, the YRD is one of China’s fastest developing and most developed regions, which is very significant to China’s economic growth [29]. But, a large amount of ecological land has been replaced by gradually expanding cities, and the area of ecological land has decreased significantly. Additionally, after years of efficient and rapid development, the YRD is facing many ecological issues, such as inadequate water supply, air pollution, and water and soil loss. These issues have made the ecosystem structure more vulnerable and altered the trade-offs/synergies among ESs [30]. Therefore, we selected the YRD as our study area with the objectives to (1) examine the spatial–temporal change trends for seven key ESs (Food Production (FP), Water Regulation (WR), Climate Regulation (CR), Soil Conservation (SC), Environmental Purification (EP), Biodiversity Maintenance (BM), and Aesthetic Landscape (AL)) from the year 2000 to 2020; (2) elucidate the dynamics of the relationship between ESs over time and their spatial scale effects, ranging from grid to township, county, and city levels; and (3) ascertain the spatial heterogeneity in the relationships between ESs. This research is intended to inform decision-makers in the development of strategies for land management and the promotion of sustainable ecosystem development.

2. Materials and Methods

2.1. Study Area

The YRD is situated in the lower reaches of the Yangtze River and proximal to the East China Sea and encompasses 26 cities (Figure 1a) [31]. The YRD spans an area of 210 thousand km², accounting for 2.3% of China’s total area. Its topography chiefly entails lowlands in the center and highlands around the periphery. Notably, the altitude is mostly below ten meters (Figure 1b). The YRD has a subtropical monsoon climate and concurrent rainy and warm periods. The average temperature remains above 0 °C, and subtropical evergreen broad-leaved forests dominate the vegetation. Furthermore, the YRD is the Chinese region with the densest river network.

The YRD had a population of 227 million and a high level of urbanization in 2020. During the study period, the Gross Domestic Product (GDP) of the YRD grew significantly from RMB 1.86 trillion to RMB 24.47 trillion (in comparison to 2000), with an average annual growth rate of 12.8%, which was higher than the national average. As such, it is an essential contributor to China’s overall economic development. However, the expanding cities have replaced a substantial amount of arable and forested land, resulting in significant damage to the ecosystem. Remarkably, the area of arable land (irrigated cropland and dry land), forest land (broad-leaved forest and shrubbery), grassland (grassland, open and dense), and wetlands in the YRD declined by 12,800 km², 650 km², 330 km², and 573 km², respectively, between 2000 and 2020 (Figure 1c,d)).

2.2. Data Sources and Processing

The primary datasets utilized in this study encompass both land use and socio-economic information. We obtained land use data with a spatial resolution of 1000 × 1000 m from the Resource and Environmental Science Data Platform of the Chinese Academy of Sciences and refined the land use classification system (

Table 1. The adjustment of land use classification in this study.

Classification in the Land Use Data	Land Use Classification
	Xie et al. [32]	This Study
Paddy field	Paddy field	Irrigated cropland
Dry land	Dry land	Dry land
Marsh, saline–alkaline land, beach, bottom land	Wetland	Wetland
Canals, lakes, reservoirs and ponds	River system	Water bodies
Woodland, sparse woodland, other woodlands	Broad-leaved forest	Broad-leaved forest
Shrubbery	Shrubbery	Shrubbery
High0cover grassland	Bush	Grassland, dense
Low- and medium-cover grassland	Meadow	Grassland, open
Sandy land, bare land, bare rock stony land	Bare land	Barren or sparsely vegetated

) in alignment with the methodologies established by Xie et al. [32], taking into account the unique regional characteristics of the YRD. Due to the minimal ESs provided by construction land or barren or sparsely vegetated regions, we assigned a value of zero to these areas in this study. Xie et al. [32] introduced the initial equivalent value for ESs per unit area, as detailed in

Table 2. Ecosystem service equivalent value per unit area of land use in China.

Land Use		FP	WR	CR	EP	SC	BM	AL
Arable land	Dry land	0.85	0.02	0.36	0.1	1.03	0.13	0.06
	Irrigated cropland	1.36	−2.63	0.57	0.17	0.01	0.21	0.09
Forest land	Broad-leaved forest	0.29	0.34	6.5	1.93	2.65	2.41	1.06
	Shrubbery	0.19	0.22	4.23	1.28	1.72	1.57	0.69
Grassland	Dense grassland	0.38	0.31	5.21	1.72	2.4	2.18	0.96
	Open grassland	0.22	0.18	3.02	1	1.39	1.27	0.56
Water bodies	Wetland	0.51	2.59	3.6	3.6	2.31	7.87	4.73
	Water bodies	0.8	8.29	2.29	5.55	0.93	2.55	1.89

Note: FP: Food Production, WR: Water Resources Supply, CR: Climate Regulation, EP: Environmental Purification, SC: Soil Conservation, BM: Biodiversity Maintenance, AL: Aesthetic Landscape. Source: Xie et al. [32].

, which represented the average ES levels across China. However, the intensity of ESs can vary considerably depending on the ecological conditions of different regions. This variability necessitates the adjustment of the equivalent value per unit area when assessing the ESV for a particular area. The detailed exposition of the methods and processes for equivalent value correction is shown in Section 3.1.

In this study, socio-economic data are sourced from the Statistical Yearbooks of various cities and provinces in the YRD and China from 2000 to 2020. Key indicators mainly included the agricultural, forestry, animal husbandry, fishery production value, forest coverage, tourism income, and demographic data for both China and the YRD. Additionally, we considered the prices, yields, cultivation areas, and producer price indexes (PPI) of the three principal cereal crops: wheat, corn, and rice. The PPI for agricultural products is also derived from the National Bureau of Statistics.

2.3. Methods

2.3.1. Equivalent Value Correction

The correction coefficient was introduced in this study to carry out the first and second corrections to the equivalent values per unit area. First, we revised the equivalent value of ESs according to the economic characteristics in the YRD. The adjustment coefficients for FP were recalibrated for various land use types, based on the per-unit-area output values of agriculture, forestry, animal husbandry, and fisheries within the YRD and the national averages [33]. This recalibration is detailed in

Table 3. The functional adjustment index of equivalent value factors of different ecosystem services in Yangtze River Delta between 2000 and 2020.

Correction Factor	2000	2005	2010	2015	2020
FP of arable land	3.50	2.91	2.67	2.55	2.22
FP of forest land	3.40	3.20	2.90	2.78	2.53
FP of grassland	3.04	2.33	2.26	2.16	1.65
FP of water bodies	6.49	5.86	5.54	5.27	5.58
CR, EP, SC, BM	1.86	1.85	1.84	1.84	1.82
AL	1.17	1.67	2.14	0.99	0.87

. The equivalent value for CR, EP, SC, and BM was modified based on forest coverage [34]. Since tourism can indirectly reflect the aesthetic value, the value equivalent of AL can be revised according to the total tourism revenue and the total number of people. We have corrected the equivalent value of AL based on the tourism revenue per person [33]. The functional adjustment index of equivalent value factors of ESs in the YRD between 2000 and 2020 is shown in Table 3.

The magnitude of various ESs is intrinsically linked to the ecosystem’s biomass; an increase in biomass generally corresponds to a higher capacity for ES provision [35]. In accordance with this principle, we conducted a secondary correction of the equivalent value factors, which was based on the biomass factors and the respective proportions of the provinces and municipality within the YRD, namely Jiangsu, Zhejiang, Anhui, and Shanghai. The biomass factor for the YRD was determined to be 1.52 and was utilized as the secondary correction coefficient. The ES equivalent value underwent both primary and secondary revisions in accordance with the method outlined in Formula 1.

$${\mathrm{b}}_{\mathrm{i}\mathrm{j}}={\mathrm{n}}_{\mathrm{f}}\times {\mathrm{B}}_{\mathrm{i}\mathrm{j}}$$

(1)

where ${\mathrm{b}}_{\mathrm{i}\mathrm{j}}$ represents the corrected unit equivalent value, i = 1, 2… 6 represents different types of ESs, and j = 1, 2… 8 represents different land use types; ${\mathrm{B}}_{\mathrm{i}\mathrm{j}}$ represents the equivalent value prior to correction; ${\mathrm{n}}_{\mathrm{f}}$ represents the first and second correction coefficients.

2.3.2. Evaluation of Ecosystem Services

The equivalent value for ESs per unit area was ascertained by calculating one-seventh of the income generated from grain production per unit area (Formula (2)).

$${\mathrm{C}}_{\mathrm{t}}={\displaystyle \frac{1}{7}}{\sum }_{\mathrm{i}=1}^{\mathrm{i}}{\displaystyle \frac{{\mathrm{S}}_{\mathrm{i}}\times {\mathrm{P}}_{\mathrm{i}}\times {\mathrm{Q}}_{\mathrm{i}}}{\mathrm{S}}}$$

(2)

where ${\mathrm{C}}_{\mathrm{t}}$ represents ESV of unit value equivalent (RMB 10,000/km²); i represents wheat, corn and rice. ${\mathrm{S}}_{\mathrm{i}}$ and $\mathrm{S}$ refer to the sown area of the ith grain and total sown area of three grains, respectively (km²); ${\mathrm{P}}_{\mathrm{i}}$ and ${\mathrm{Q}}_{\mathrm{i}}$ represent the price (RMB 10,000/kg) and the yield per unit land use area of the ith grain (kg/km²), respectively.

To ensure the comparability of ESVs across various time periods, we adjusted for price fluctuations by employing the PPI. This allowed us to calculate the equivalent prices of crops for different years, normalized to the year 2000. Between 2000 and 2020, the equivalent value factor for ESs in the YRD was RMB 7.12, 10.90, 17.27, 22.68, and 21.30, respectively, as detailed in

Table A1. Ecosystem service value per unit land area of YRD (unit: RMB 10,000/km²).

ES Type		FP	WR	CR	EP	SC	BM	AL	Total
2000	Dry land	51.51	−99.62	11.47	3.42	0.2	4.23	1.14	−27.65
	Irrigated cropland	32.2	0.76	7.25	2.01	20.73	2.62	0.76	66.33
	Broad-leaved forest	10.67	12.51	130.84	38.85	53.34	48.51	13.42	308.14
	Shrubbery	6.99	8.1	85.15	25.77	34.62	31.6	8.74	200.97
	Grassland, dense	12.5	10.2	104.88	34.62	48.31	43.88	12.16	266.55
	Grassland, open	7.24	5.92	60.79	20.13	27.98	25.56	7.09	154.71
	Wetland	5.52	28.03	72.47	72.47	46.5	158.42	59.89	443.30
	Water bodies	56.19	582.27	46.1	111.72	18.72	51.33	23.93	890.26
2005	Dry land	65.57	−126.8	17.47	5.21	0.31	6.44	2.49	−29.31
	Irrigated cropland	40.98	0.96	11.03	3.07	31.57	3.98	1.66	93.25
	Broad-leaved forest	15.38	18.03	199.23	59.16	81.22	73.87	29.33	476.22
	Shrubbery	10.07	11.66	129.65	39.23	52.72	48.12	19.09	310.54
	Grassland, dense	14.67	11.97	159.69	52.72	73.56	66.82	26.56	405.99
	Grassland, open	8.49	6.95	92.57	30.65	42.6	38.93	15.49	235.68
	Wetland	8.45	42.91	110.34	110.34	70.8	241.22	130.87	714.93
	Water bodies	77.67	804.86	70.19	170.11	28.51	78.16	52.29	1281.79
2010	Dry land	95.32	−184.33	27.53	8.21	0.48	10.14	5.06	−37.59
	Irrigated cropland	59.58	1.4	17.39	4.83	49.75	6.28	3.37	142.60
	Broad-leaved forest	22.08	25.88	313.95	93.22	128	116.4	59.55	759.08
	Shrubbery	14.46	16.75	204.31	61.82	83.08	75.83	38.76	495.01
	Grassland, dense	22.54	18.39	251.65	83.08	115.92	105.3	53.93	650.81
	Grassland, open	13.05	10.68	145.87	48.3	67.14	61.34	31.46	377.84
	Wetland	13.39	67.99	173.88	173.88	111.57	380.13	265.71	1186.55
	Water bodies	116.34	1205.59	110.61	268.07	44.92	123.17	106.17	1974.87
2015	Dry land	119.55	−231.2	36.16	10.78	0.63	13.32	3.07	−47.69
	Irrigated cropland	74.72	1.76	22.84	6.34	65.33	8.25	2.05	181.29
	Broad-leaved forest	27.79	32.58	412.3	122.42	168.09	152.87	36.18	952.23
	Shrubbery	18.21	21.08	268.31	81.19	109.1	99.59	23.55	621.03
	Grassland, dense	28.3	23.08	330.48	109.1	152.24	138.28	32.76	814.24
	Grassland, open	16.38	13.4	191.56	63.43	88.17	80.56	19.11	472.61
	Wetland	17.58	89.29	228.35	228.35	146.53	499.21	161.43	1370.74
	Water bodies	145.34	1506.09	145.26	352.04	58.99	161.75	64.5	2433.97
2020	Dry land	97.75	−189.03	33.59	10.02	0.59	12.37	2.54	−32.17
	Irrigated cropland	61.09	1.44	21.21	5.89	60.69	7.66	1.69	159.67
	Broad-leaved forest	23.75	27.85	383.01	113.72	156.15	142.01	29.86	876.35
	Shrubbery	15.56	18.02	249.25	75.42	101.35	92.51	19.44	571.55
	Grassland, dense	20.3	16.56	307	101.35	141.42	128.46	27.04	742.13
	Grassland, open	11.75	9.62	177.95	58.92	81.9	74.83	15.77	430.74
	Wetland	16.51	83.85	212.13	212.13	136.12	463.73	133.23	1257.70
	Water bodies	144.53	1497.66	134.94	327.03	54.8	150.26	53.24	2362.46

Note: FP: Food Production, WR: Water Resources Supply, CR: Climate Regulation, EP: Environmental Purification, SC: Soil Conservation, BM: Biodiversity Maintenance, AL: Aesthetic Landscape.

.

2.3.3. Ecosystem Services Trade-Off/Synergy Analysis

This study employed GIS technology and correlation analysis to assess the trade-off and synergy intensity of seven key ESs across different spatial scales, including 1000 × 1000 m grid, township, county, and city levels, for the years 2000, 2005, 2010, 2015, and 2020. The correlation analysis methods mainly include Pearson and Spearman correlations. The selection between these methods was contingent upon the data’s adherence to a normal distribution. The Pearson correlation was applied in cases where data exhibited normality, while the Spearman correlation was utilized for non-normally distributed data [36]. The Kolmogorov–Smirnov test was utilized to assess the normality of the data distribution, which indicated that 64% of the ESs exhibited non-normal distributions. Therefore, Spearman correlation analysis was selected to examine the relationships among the ESs in this study. The calculation of the Spearman correlation coefficient is presented in Formula (3).

$$\mathsf{\rho }=1-{\displaystyle \frac{6\sum _{\mathrm{i}=1}^{\mathrm{n}}{\mathrm{d}}_{\mathrm{i}}^2}{\mathrm{n}({\mathrm{n}}^2-1)}}$$

(3)

where $\mathsf{\rho }$ represents the correlation coefficient; ${\mathrm{d}}_{\mathrm{i}}$ denotes the rank difference in variable values; i is the rank index of the variable value; n is the total number of variable values.

2.3.4. Spatial Heterogeneity of Ecosystem Service Trade-Offs/Synergies

Spatial autocorrelation refers to the possible interconnections between variables within a specific area, which is typically assessed using Moran’s I index. This statistical measure has been widely employed to investigate the spatial relationships of ESs [37]. The bivariate global Moran’s I (BGM) is a tool that evaluates the overall spatial correlation between two variables, reflecting the extent to which their spatial distributions are interrelated. On the other hand, the bivariate local Moran’s I (BLM) delves into the local spatial correlation between the two variables, highlighting the degree of clustering within proximate regions. The BLM identifies four distinct clustering patterns: High–High (H-H), Low–Low (L-L), High–Low (H-L), and Low–High (L-H), each representing different spatial associations. In this study, we employed the queen adjacency approach to calculate the BGM, which allowed us to evaluate the overall spatial correlation between ESs. Additionally, we utilized the BLM to quantify the local spatial agglomeration among these services. The specific methodologies for calculating BGM and BLM are delineated in Formulas 4 and 5, respectively. This analytical framework provides a comprehensive understanding of the spatial dynamics within the ESs under investigation.

$${\mathrm{I}}_{\mathrm{k}\mathrm{l}}={\displaystyle \frac{{\sum }_{\mathrm{i}=1}^{\mathrm{n}}{\sum }_{\mathrm{j}=1}^{\mathrm{n}}{\mathrm{W}}_{\mathrm{i}\mathrm{j}}({\mathrm{x}}_{\mathrm{k}}^{\mathrm{i}}-{\overline{\mathrm{x}}}_{\mathrm{k}})({\mathrm{x}}_{\mathrm{l}}^{\mathrm{i}}-{\overline{\mathrm{x}}}_{\mathrm{l}})}{{\mathrm{S}}^2\sum _{\mathrm{i}=1}^{\mathrm{n}}\sum _{\mathrm{j}=1}^{\mathrm{n}}{\mathrm{W}}_{\mathrm{i}\mathrm{j}}}}$$

(4)

$${{\mathrm{I}}^{\mathrm{\prime }}}_{\mathrm{k}\mathrm{l}}={\displaystyle \frac{{\mathrm{x}}_{\mathrm{k}}^{\mathrm{i}}-{\overline{\mathrm{x}}}_{\mathrm{k}}}{{\mathrm{S}}_{\mathrm{k}}^2}}{\sum }_{\mathrm{j}=1}^{\mathrm{n}}{\mathrm{W}}_{\mathrm{i}\mathrm{j}}{\displaystyle \frac{{\mathrm{x}}_{\mathrm{l}}^{\mathrm{j}}-{\overline{\mathrm{x}}}_{\mathrm{l}}}{{\mathrm{S}}_{\mathrm{l}}^2}}$$

(5)

where ${\mathrm{I}}_{\mathrm{k}\mathrm{l}}$ and ${{\mathrm{I}}^{\mathrm{\prime }}}_{\mathrm{k}\mathrm{l}}$ represent the BGM and BLM, respectively; ${\mathrm{x}}_{\mathrm{k}}^{\mathrm{i}}$(${\mathrm{x}}_{\mathrm{l}}^{\mathrm{j}}$) is the observed value of k (l) variable on space element I (j); ${\mathrm{x}}_{\mathrm{l}}^{\mathrm{i}}$ is the observed value of l variable on space element i; ${\overline{\mathrm{x}}}_{\mathrm{k}},{\overline{\mathrm{x}}}_{\mathrm{l}}$ are the average of the observed values of k and l variables; ${\mathrm{W}}_{\mathrm{i}\mathrm{j}}$ is the spatial weight matrix; ${\mathrm{S}}_{\mathrm{k}}^2$, ${\mathrm{S}}_{\mathrm{l}}^2$ are the variance of the observed values of k and l variables, respectively; n is the total number of geographical units in the study area.

3. Results

3.1. Spatial and Temporal Evolution of Ecosystem Services

3.1.1. Temporal Evolution of Ecosystem Services

The ESV per unit area generally displayed a trend of fluctuating ascent, characterized by an initial increase followed by a decline in the YRD (Table A1). By synthesizing the various land use areas within the YRD across different time periods, we have generated the ESV fluctuations for the seven distinct ESs over five distinct periods, as illustrated in Figure 2. Upon a comparative analysis of Figure 2 and Table 1, it was evident that the ESVs exhibited a trend of fluctuating increase. Generally, the ESVs witnessed a substantial increase from 2000 to 2020, amounting to a total rise of RMB 650.625 billion. The zenith of this growth was reached in 2015, with the total ESV peaking at RMB 1017.05 billion. This was followed by the year 2020, with the total ESV stood at RMB 982.31 billion, a slight decrease from the peak but still significantly higher than the initial value in 2000, which was recorded at RMB 331.68 billion. The most remarkable surge in the ESV occurred between the years 2005 and 2010, during which the value increased by RMB 304.469 trillion, marking a growth rate of 59%. When examining the change rates of the seven different ESVs from 2000 to 2020, they were ranked from highest to lowest as follows: CR, WR, EP, BM, SC, FP, and AL. Notably, the value of CR consistently emerged as the most significant contributor to the ESV, with its value peaking annually and constituting nearly one-third of the total ESV.

3.1.2. Spatial Evolution of Ecosystem Services

The ESV density across various land use types showed a distinct geographical pattern within the study area. It was observed that the ESV in the northern regions was relatively lower when compared to the southern regions. This disparity could be attributed to differences in land use practices, environmental conditions, and the inherent capacity of the ecosystems in these areas to provide services. Furthermore, the ESV associated with aquatic ecosystems, including water bodies, rivers, lakes, wetlands, and estuaries of the Yangtze River and Qiantang River, was notably higher than in their adjacent areas. This elevated ESV could be attributed to the significant ecological functions performed by these water bodies, such as WR, EP, BM, and AL. The higher ESV in these water-related land use types underscores the importance of preserving and enhancing the ecological integrity of these areas, as they contribute substantially to the overall environmental health and economic value of the region. Furthermore, the northeastern inland region of the YRD demonstrated a markedly lower ESV in comparison to the coastal zones of the YRD.

FP exhibited a spatial distribution that was relatively low in the southern part of the YRD but notably high in the northern areas (Figure 3). This region was characterized by its flat topography, abundance of rivers and lakes, and extensive tracts of arable land, which contributed to its status as the principal area for grain production within the YRD. The favorable conditions for agriculture in the northern plains have made it a vital contributor to the region’s food security. WR in the northern region of the YRD was comparatively lower than in the south. This difference could be attributed to the predominance of shrubbery and broad-leaved forests in the southern region, which were instrumental in water conservation due to their capacity to retain moisture and regulate water flow. Conversely, the northern YRD was characterized by extensive arable land, which necessitates significant water usage for irrigation and other agricultural practices. The high demand for water in agriculture can lead to a relative depletion of available water resources. Furthermore, the northeast coastal areas of the YRD also showed high values of WR. The distribution of CR mirrored that of SC, with lower values observed in the northern YRD and higher values in the south. This pattern primarily stemmed from the predominance of shrubbery and broad-leaved forests in the south, which were highly effective in regulating various environmental factors such as soil quality, temperature, humidity, wind speed, and precipitation. These forests played a crucial role in enhancing physical and mental comfort and conserving soil, thereby contributing significantly to the high values of CR and SC in the southern areas. The distribution pattern of EP, BM, and AL exhibited similarities across the region. Broadly speaking, the southern forested areas demonstrated higher values for these services compared to the northern ones, which were predominantly characterized by arable land.

3.2. Temporal Evolution of Ecosystem Services Trade-Offs/Synergies

The visual representation of these relationships is conveyed through a series of “bubbles”, where the size of each bubble corresponds to the magnitude of the correlation coefficient, indicating the strength of the relationship between the ESs (Figure 4). Additionally, the color of the bubbles signifies the direction of the correlation: the darker the color, the more pronounced the correlation coefficient. Red and blue bubbles indicate a positive and negative correlation, respectively. In the YRD, our findings revealed that a substantial majority, 70%, of the ES pairs exhibited synergistic interactions.

3.2.1. Spatial Scale Effects

The relationships between ESs observed at the grid, township, county, and city scales exhibited a notable degree of similarity. Although the correlation between ESs at grid scale was roughly similar to that of the three administrative regional scales, the correlation value and change trend are still quite different due to different research scales. Compared with the other three scales, the trade-off between FP and other ESs at the grid scale was weak, and most of them even showed weak synergy. The synergy between WR-CR, WR-EP, WR-SC, WR-BM, and WR-AL was higher than that of the other three scales. CR-EP, CR-SC, CR-BM, CR-AL, EP-SC, EP-BM, EP-AL, SC-BM, SC-AL, and BM-AL also showed medium- to high-intensity synergy.

At the township and county scales, the relationships between WR and the other ESs (CR, EP, SC, BM, and AL) exhibited a consistent and moderate level of synergy. This stability suggested a balanced and harmonious interaction among these services within these administrative scales, contributing to a resilient and sustainable ecological framework. However, the degree of synergy observed was more pronounced at the township scale compared to the county and city scales. At the city scale, the interactions between FP and the other ESs (CR, SC, and BM) were characterized by strong trade-off relationships, while other ES pairs demonstrated synergy.

As the scale expanded from the township to the city level, the synergistic relationships of CR-EP, CR-BM, CR-Al, CR-SC, EP-SC, EP-BM, EP-AL, SC-BM, SC-AL, and BM-AL exhibited a gradual decline. Conversely, the trade-offs of FP-CR, FP-EP, FP-SC, FP-BM, and FP-AL were most pronounced at the city scale. Interestingly, the strongest trade-off between FP and WR was observed at the township scale. From the township to city scales, the relationship of FP-WR underwent a transition from a weak trade-off to a weak synergy. The synergy of WR-CR, WR-EP, WR-SC, WR-BM, and WR-AL experienced a rapid decline as the scale expanded. This reduction was particularly notable, with the initially weak synergies of WR-CR and WR-SC eventually turning into weak trade-offs at larger scales. Nevertheless, it persisted, with a consistent degree of robustness despite fluctuations in its intensity. Despite the shifts observed in other relationships between ESs, the synergy of CR-EP, CR-SC, CR-BM, CR-AL, EP-SC, EP-BM, EP-AL, SC-BM, SC-AL, and BM-AL remained resilient. These relationships demonstrated a consistent level of robustness, even though their intensity may have varied.

3.2.2. Temporal Evolution

Compared with the other three scales, from 2000 to 2020, the trade-off between FP and other ESs at the grid scale gradually weakened and most of them turned into synergy. Similar to other scales, the synergy of WR-CR, WR-EP, WR-SC, WR-BM, and WR-AL also showed a decreasing trend from 2000 to 2020. The synergistic relationship of CR-EP, CR-SC, CR-BM, CR-AL, EP-SC, EP-BM, EP-AL, SC-BM, SC-AL, and BM-AL showed an increasing trend

At the township, county, and city scales, the trade-off degree of FP-WR, FP-CR, FP-EP, FP-SC, FP-BM, and FP-AL experienced a decline, and FP-WR even turned into weak synergy from 2000 to 2020. Moreover, the synergistic relationships of WR-CR, WR-EP, WR-SC, WR-BM, and WR-AL also declined. Although there was a strong and significant synergistic effect of CR-EP, CR-SC, CR-BM, CR-AL, EP-SC, EP-BM, EP-AL, SC-BM, SC-AL, and BM-AL, they generally showed a weakening trend at the township scale and an increasing trend at the county and city scales, but neither the strengthening nor the weakening trend is obvious.

Over time, the trade-off degrees of FP-WR, FP-CR, FP-SC, FP-BM, and FP-AL gradually diminished, and even partly turned into a weak synergy. Similarly, the level of synergy and its significance for WR-CR, WR-EP, WR-SC, WR-BM, and WR-AL declined. Furthermore, the relationships of WR-CR and WR-SC evolved from synergy into weak trade-offs. In contrast, the ES pairs of CR-EP, CR-SC, CR-BM, CR-AL, EP-SC, EP-BM, EP-AL, SC-BM, SC-AL, and BM-AL showed a significant synergy.

3.3. Spatial Heterogeneity of Ecosystem Services Trade-Offs/Synergies

One purpose of our study was to show the spatial distribution characteristics of the interaction between ESs in the YRD so as to better formulate and adjust the regional development plan. At the grid scale, the correlation between FP and other ESs was relatively weak. Therefore, we did not consider this particular scale in this part. Among the other three scales, the township scale had the smallest area and could more finely show the spatial heterogeneity of ESs. Attached Figure A1 and Figure A2 are the spatial distribution of the interaction between ESs at the county and city scale in 2020, respectively. By comparing the research results at city, county, and township scales, we found that the spatial distribution characteristics of the interaction between the ESs they reflect were very similar. Compared with the township scale, the county and city scales were the simplified or even more simplified versions of their characteristics. Only the township scale can reflect the spatial distribution characteristics of the interaction between ESs in the most detailed way.

Finally, considering time validity, we no longer focused on the spatial distribution characteristics of the interaction between ESs before 2020 and selected the township scale in 2020 to conduct spatial heterogeneity analysis of ES relationships and obtained the Local Indicator of Spatial Association (LISA) diagram between ESs that passed the significance test at the level of 0.05. Statistical analysis confirmed that there was a significant correlation between different ESs (0.05 level), as shown in Figure 5. The results of BGM are consistent with the previous analysis introduced in Section 3.2.1 and Section 3.2.2.

Utilizing the spatial distribution patterns of the relationships between ESs, we categorized the ES pairs into three distinct groups to better understand their interactions and spatial dynamics. The first group included pairs that involve FP with other ESs: FP-WR, FP-CR, FP-EP, FP-SC, FP-BM, and FP-AL. The second group consisted of pairs that involve WR with other services, including WR-CR, WR-EP, WR-SC, WR-BM, and WR-AL. The final group encompassed ES pairs that did not include FP or WR: CR-EP, CR-SC, CR-BM, CR-AL, EP-SC, EP-BM, EP-AL, SC-BM, SC-AL, and BM-AL.

The first group was characterized by a distinctive spatial distribution pattern: H-H and L-L associations were sparse, indicating that areas of high or low ES provision tended not to cluster together. Conversely, L-H and H-L associations were widespread, suggesting ES pairs in the first group mostly exhibited trade-off relationships in the YRD (Figure 6). The areas exhibiting trade-offs (H-L and L-H) were predominantly situated in the northern and southern parts, while the central region remained largely insignificant. L-L areas were observed to be dispersed around the areas of H-L, while H-H areas tended to be distributed around the L-L areas. Notably, the FP-WR pair did not exhibit a significant relationship in the northern of the YRD.

In the second group, the relationship between WR and other ESs showed two different patterns in the north and south of the YRD. The northern part was dominated by the synergistic area, which was mainly characterized by the distribution characteristics of L-L with a small amount of H-H, indicating that the supply of water and other ESs in this area was complementary. The south was mainly dominated by trade-offs, with a small number of synergistic areas. The distribution of L-H areas surrounded H-H areas. The northern and southern parts of the YRD presented a spatial distribution pattern in which high (low) service areas and low (high) service areas were interspersed with each other. In contrast, the central part of the YRD showed an insignificant presence of these synergistic patterns.

In the third group, the relationships between ESs in the study area predominantly demonstrated synergistic patterns, with only minor variations observed among different ES pairs. Areas characterized by L-L and H-H were notably concentrated in the northern and southern parts of the YRD, respectively. The central part of the YRD, however, remained an area of insignificance in terms of significant ES relationships. Regions with H-L and L-H patterns were found widespread around the L-L and H-H areas. Notably, the SC-BM and SC-AL exhibited large areas of H-L in the northern YRD. In contrast, the EP- SC had extensive areas of H-L distributed around the L-L regions, which may suggest a more complex interplay between these services.

4. Discussion

4.1. Spatio-Temporal Evolution and Multi-Scale Analysis of Ecosystem Service Trade-Offs/Synergies

Land use is a pivotal determinant in shaping the dynamics of trade-offs and synergies between ESs [38]. The existence of trade-offs is primarily due to the inherent incompatibility between certain ESs and particular types of land use. Grasping the ES relationship is crucial for enhancing total ES level and human well-being. In the YRD, the majority of ES pairs demonstrated synergistic relationships, similar to ESs in other regions [39]. The relationship between FP and other ESs exhibited a trade-off situation, which was predominantly attributed to the spatial competition among different land uses such as arable land, forests, grasslands, and water bodies. The need to allocate space for agricultural production, which is essential for FP, often leads to the conversion of ecological land, thereby affecting other ESs like WR, CR, EP, SC, and BM. The majority of ES pairs between WR and other ESs showed moderate synergies; however, not all exhibited synergistic relationships. The synergies were typically observed among regulation, support, and cultural services, which is consistent with the findings reported in the study by Li [40].

When integrating ESs into regional planning, it is imperative to take into account the interrelationships of ESs across spatial and temporal scales. This multi-dimensional perspective is essential for a comprehensive understanding of how ESs interact within and across different landscapes and over time, enabling more effective and sustainable land use strategies. Previous research has predominantly focused on examining the relationships between ESs at a particular scale, often neglecting the influence of spatial scale and temporal dynamics on these relationships [41]. This study revealed that the synergies of most ES pairs, except CR with other ESs and SC with other ESs, weakened with increasing scale. In 2020, a notable transition occurred in the dynamics of ES interactions as the spatial scale expanded from grid to city level. Specifically, the relationships of WR-CR and WR-SC shifted from weak synergies to weak trade-offs. Concurrently, the trade-offs between FP and other ESs, with the exception of the FP-WR, became more pronounced. Additionally, this study noted that the significance of the relationships between ESs tended to diminish as the spatial scale increased. This observation is in line with the findings reported in the research conducted by Pan [27].

The response of certain ESs to environmental changes has been identified as being relatively slow [37]. Consequently, it is imperative to take a long-term perspective when considering the evolution of the relationship between ESs. For our research, we selected the time frame from 2000 to 2020 for analysis. Within this two-decade span, we observed that the trade-offs of most ES pairs experienced a weakening trend, while synergies among them showed an increase. This suggests a more harmonious coexistence of these services over time.

4.2. Spatial Heterogeneity of Ecosystem Service Trade-Offs/Synergies for Policy Implications

In terms of the spatial heterogeneity of ES relationships, FP was found to be negatively correlated with other ESs. In contrast, WR demonstrated a weak positive correlation with other ESs, and the remaining ESs showed a moderate positive correlation. It was particularly noteworthy that the value of BGM for the relationship between FP and other ESs closely mirrored the results of the Spearman correlation coefficient. This concurrence lends robust support to the findings of our study, underscoring the reliability of the observed spatial patterns in ES interactions.

When devising a management plan of land use and ecosystems, pinpointing the spatial distribution of trade-offs and synergies between ESs is of paramount importance. This precise location identification of ES relationships is essential for targeted interventions and the optimization of land use strategies. Despite the significance of this task, a dearth of research exists that systematically analyzes the spatial heterogeneity of these ESs interactions [36]. Previous research on trade-offs and synergies in the YRD predominantly concentrated on average levels, neglecting the distinctive spatial distribution patterns of these interactions. This paper underscored that the dynamics of trade-offs and synergies for identical pairs of ESs were subject to geographical variation, a conclusion that aligns with the findings of Xue et al. [42]. For example, while there was a pronounced trade-off between FP and SC, the LISA diagram revealed instances of synergistic co-existence in specific regions (central areas), characterized by H-H and L-L associations. This highlighted the importance of considering the spatial dimension when assessing ES interactions to capture the full complexity of their relationships. By comparing the results related to ES relationships across three scales, we found that the interactions were quite similar in spatial distribution. As the research scale expands from the township to the city level, the spatial distribution becomes increasingly simplified, and hence the information reflected at the township scale is more abundant.

Various ESs exhibited distinctive spatial clustering characteristics, which were significantly correlated with the distribution of different land use types. Specifically, the YRD’s wide distribution of arable land in the north and forests in the south resulted in spatial aggregation of ES relationships exhibiting “upper, middle, and lower” characteristics. This spatial feature was manifested in that ESs in the northern and southern YRD showed different correlation aggregation characteristics, and most of the ESs in the north and south of the YRD were dominated by one correlation feature. However, in the central region of the YRD, the correlation of ESs showed insignificant characteristics.

The LISA maps of various ES pairs involving FP-WR, FP-CR, FP-EP, FP-SC, FP-BM, and FP-AL displayed extensive areas of L-H and H-L (trade-offs) relationships, consistent with the findings of the Spearman correlation coefficient and the BGM in this study. The results regarding the relationship between FP-WR and FP-AL at the township level were also consistent with those of Chen [15]. Furthermore, each ES pair was dominated by two aggregation characteristics, while the other two were dispersed around them, similar to the findings of research conducted in the source area of the three rivers [37].

4.3. Research Limitations and Prospects

This study has delivered notable insights into the scale effects, temporal variations, and spatial heterogeneity of ES trade-offs/synergies; however, certain limitations persist. Firstly, the method of value equivalent factor per unit area simplifies land into various types while ignoring the internal differences of land use types under different natural backgrounds. Although primary and secondary corrections were made in this study, differences in climate, vegetation, and soil quality within the same land use type could not be detailed. Additionally, this study roughly assumed that the ESV of construction land was zero, neglecting the effects of street greening and small urban parks in cities and towns. Secondly, while the Spearman correlation coefficient and bivariate Moran’s index were utilized to quantify trade-offs/synergies and spatial clustering of ESs, the driving mechanisms of interaction between these ESs were not explored [43]. Therefore, there is a need to explore more driving factors of the dependencies and link them to government decisions [44]. Thirdly, the limitation of the analysis is its concentration on the subjective valuation of ecosystems, overlooking their essential objective existence. It is essential to implement an objective assessment that recognizes the intrinsic significance of nature for human life and well-being and effectively conveys the intrinsic and functional value of nature. Lastly, future research should include more spatial scales to provide more comprehensive and scientific information for decision-makers to regulate ecosystems and land use management.

5. Conclusions

This study evaluated the values of seven typical ESs in the YRD using the equivalent value method based on land use data from 2000 to 2020. We examined the time variation, spatial scale effects, and spatial heterogeneity of the ES relationships. The findings revealed that the value of the seven ESs and the total ESV exhibited a fluctuating upward trend. The total ESV increased by RMB 650.625 billion, with the highest growth rate (59%) occurring between 2005 and 2010. Approximately 70% of the ES pairs showed synergies. The direction and degree of trade-offs/synergies between ESs were similar at the grid, township, county, and city scales, whereas the grid scale revealed slight differences. Moreover, the relationship between FP and other ESs showed trade-offs mainly in the northern and southern areas of the YRD, while the other ES pairs exhibited mainly synergy. The spatial distribution results of the relationships between ESs were similar to those of the Spearman correlation analysis and the spatial aggregation of the relationship between ESs exhibiting “upper, middle, and lower” characteristics. This study provides scientific evidence to support local governments in formulating sustainable land and ecosystem management policies.