Spatiotemporal Dynamics and Drivers of Ecosystem Service Value in Coastal China, 1980–2020

Abstract

In response to the widespread decline in ecosystem service value (ESV) caused by rapid industrialization and urbanization-driven land-use transitions in Coastal China—characterized by shrinking farmland and expanding built-up land and crystallized in the “core-city sprawl and surrounding-farmland encroachment” pattern—this study integrated land-use and socioeconomic data from 1980 to 2020. Employing the equivalent-factor method and Geodetector model, we quantified the spatiotemporal evolution of ESV and its driving mechanisms across the entire coastal region. The results show that (i) the total ESV experienced a fluctuating increase. (ii) Spatially, the ESV exhibited a “high in the south, low in the north, and higher inland than along the immediate coast” pattern, with mountain–hill belts and estuarine wetlands in the south forming high-value clusters, whereas the Bohai Rim in the north emerged as a low-value zone. (iii) Socioeconomic factors increasingly dominated the driving forces, while NDVI became the most influential natural factor; the interactions between the drivers consistently produced bi-factor enhancement effects. These findings provide a scientific basis for implementing the “Two-Mountains Theory” and optimizing coastal territorial spatial planning.

1. Introduction

Ecosystem services refer to ecological characteristics, functions, or processes that directly or indirectly enhance human well-being through multiple pathways, and they represent various benefits that humans derive from functionally intact ecosystems [1]. Ecosystem service value (ESV) is an economic estimate of ecosystem services and natural capital based on market principles [2], and its variation directly mirrors the level of regional ecological security. A sound, healthy ecosystem harbours immense economic worth, continuously generating comprehensive benefits that underpin sustainable development. Amid rapid industrialization and urbanization, land-use transitions have profoundly reshaped ecosystem structures and functional patterns [3], causing severe losses in ecosystem service value and posing a formidable challenge to regional sustainable development. These transitions are strongly influenced by urbanization policies, demographic shifts, and spatial planning strategies, which collectively reconfigure land systems and alter ecological functions. Taking Coastal China as an example, construction land expansion from 1985 to 2021 reduced the farmland area by 13.8%, whereas the impervious surface area increased by more than 200%, directly impairing the region’s hydrological regulation and carbon sequestration capacities [4,5,6]. Concurrently, the ecological fragmentation induced by urban agglomeration expansion has intensified the spatial heterogeneity of ESV [7,8,9]. Such changes are not only ecological but also socio-institutional, driven by national urbanization policies, regulatory frameworks, and rural–urban integration strategies that shape land development trajectories. Since 2012, the ecological civilization concept of “harmonious coexistence between humanity and nature” has been established as a core national strategy. In 2022, China further declared “green development as an essential requirement of Chinese modernization” [10] and incorporated dual carbon goals into the global climate governance framework. In this context, there is an urgent need for scientific theories and technical methods to optimize the ecological service functions of territorial spaces—which are the materials for ecological civilization construction [11,12]—to support the rational assessment and optimization of their ecosystem service value.

As critical zones of land-sea interaction, coastal regions exhibit high ecological sensitivity and economic concentration [13]. According to 2023 data, Coastal China—accounting for just 13.8% of the nation’s land area—supports 46.7% of its population and generates 64.6% of its gross domestic product (GDP), underscoring its dual role as both an “economic engine” and an “ecological barrier” [14]. This dense concentration of population and economic activity is closely linked to national urbanization policies and regional development strategies, which drive infrastructure expansion, industrial clustering, and residential distribution—factors that significantly influence the supply and demand of ecosystem services. Globally, coastal urbanization is a pressing concern due to its adverse effects on marine and coastal ecosystem health, stemming from resource exploitation, pollution, and “ocean sprawl”—the widespread transformation of natural habitats into artificial structures [15]. However, intensive development has subjected these regions to multiple ecological challenges, such as a reduced carbon storage capacity due to forest degradation in karst-coastal zones [16]; a loss of aquatic ecosystem services from land reclamation in port cities [17]; wetland area shrinkage [18]; ecological corridor disruption [14]; and aggravated carbon deficits [19,20]. These challenges mirror global patterns where coastal urbanization leads to habitat loss, pollution, overfishing, and biodiversity decline, threatening the provisioning, regulating, cultural, and supporting services these ecosystems provide [15]. With the advancement of national strategies, such as the Beijing–Tianjin–Hebei integration and Yangtze River Delta coordination, the urban expansion and economic growth are further exacerbating supply-demand imbalances in ecosystem services [21]. These regional strategies, embedded in broader urbanization and territorial governance frameworks, reshape population mobility, land use intensity, and habitat connectivity, thereby modulating ESV distribution. These imbalances exhibit distinct spatial differences, where island cities show a slow ESV decline due to land constraints [22] while disorderly hinterland expansion in mainland coastal cities is causing precipitous drops in ESV [23]. The Guangdong–Hong Kong–Macao Greater Bay Area (Greater Bay Area) exhibited a decelerating ESV decline during the 2000–2015 period, with losses concentrated in the rapid urban expansion zones [8]. High-value ESV zones in the Yangtze River Delta Urban Agglomeration shrank into core ecological patches, such as major lakes and coastal wetlands [5]. Finally, grain supply functionality has notably weakened in coastal Liaoning due to persistent farmland contraction [24]. These changes underscore the urgent need to clarify the long-term evolutionary patterns and driving mechanisms of ESV in Coastal China [25], particularly in relation to urbanization governance, demographic dynamics, and spatial policy interventions.

The existing national-scale, long-term studies include Bao et al.’s analysis of how ESV in China responded to land-cover dynamics from 1992 to 2020 [26] and Yang et al.’s application of a Bayesian spatiotemporal hierarchical model to identify the drivers of Chinese ESV between 1992 and 2018 [27], provide valuable insights, they often lack a specific focus on the unique processes characteristic of coastal zones. Coastal regions are distinct due to intense land–sea interactions, high population density, concentrated economic activities, and specific threats like sea-level rise and saltwater intrusion. Furthermore, global comparative perspectives highlight that coastal ecosystems worldwide face similar pressures from urbanization, making studies in the Chinese context relevant for understanding and managing these challenges in other rapidly developing coastal regions, particularly in the Global South. However, few scholars have focused on China’s coastal provinces—a unique and ecologically sensitive region—to conduct systematic investigations of ESV and its determinants, especially incorporating the roles of urbanization policy, institutional reform, and population distribution in shaping land-system dynamics and subsequent ESV outcomes.

This study integrated theoretical frameworks from geography, environmental economics, and landscape ecology [28,29]. Using a modified equivalence factor method and Geodetector, we systematically assessed the spatial and temporal characteristics of ESV across Coastal China from 1980 to 2020 and identified the underlying driving factors. The study’s aims included (1) analysing the four-decade evolutionary trajectories of ESV and its components; (2) investigating the multi-scale spatial differentiation patterns; and (3) quantifying the interaction intensities and directional effects between natural [30] and social factors [31,32], with particular attention to urbanization policies, demographic variables, and territorial management schemes as critical social drivers. Theoretically, this work fills the knowledge gap concerning the long-term ESV patterns of Coastal China and deepens our understanding of coupled human–environment systems. From a practical perspective, the findings provide scientific support for optimizing territorial spatial layouts [33], accelerating the transformation of “lucid waters and lush mountains” into “invaluable assets,” and promoting a green, low-carbon transition toward socially, economically, and ecologically coordinated sustainable development in Coastal China, thereby contributing to the “Beautiful China” initiative and the realization of global sustainable development goals.

2. Materials and Methods

2.1. Research Area

This study focused on mainland China’s coastal regions, encompassing 1068 county-level units across 11 provinces (Beijing, Liaoning, Hebei, Tianjin, Shandong, Jiangsu, Shanghai, Zhejiang, Fujian, Guangdong, and Guangxi). The study area features a coastline stretching a total of approximately 15,780 km (Figure 1). As core hubs of national-level urban agglomerations, such as the Beijing–Tianjin–Hebei region, the Yangtze River Delta (Shanghai, Jiangsu, Zhejiang, and Anhui), and the Greater Bay Area (Guangdong, Hong Kong, and Macao), Coastal China represents the country’s most economically developed and densely populated zone. The region covers 12.59% of the nation’s land area and, according to the 2020 Seventh National Population Census of China, hosts a permanent population of 646 million, accounting for 46.81% of China’s total population. High-intensity human activities, including urbanization, population concentration, port construction, and mariculture, have significantly altered the coastal morphology and patterns of tidal dynamics [34]. These changes have not only disrupted the original ecosystem equilibrium but also profoundly impacted the regional ESV.

2.2. Data Sources

Land-use data were obtained from the Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences (https://www.resdc.cn, accessed on 1 September 2024), which cover 1980, 1990, 2000, 2010, and 2020, with a spatial resolution of 30 m × 30 m. In line with the China Land Use/Cover Change Remote Sensing Monitoring Classification System, land uses were categorized into six primary classes: farmland, forestland, grassland, water areas, construction land, and unused land. The baseline data required for ESV accounting—namely, the yields and cultivated areas of rice, wheat, and maize within the study region—were obtained from the China Statistical Yearbook and the provincial (regional/municipal) statistical yearbooks. Administrative boundary data were sourced from the National Geomatics Center of China (https://www.ngcc.cn/, accessed on 1 September 2024) and updated according to prefecture-level city boundary adjustments for each study year. Eight key influencing factors (

Table 1. Data sources.

Data	Abbreviation	Resolution (m)	Time Range	Source and Processing Method
Elevation	DEM	30	2020	Geospatial Data Cloud
Slope	SLOPE	30	2020	Derived from DEM Data
Normalized difference vegetation index	NDVI	1000	2000–2020	https://www.resdc.cn/, accessed on 1 September 2024
Mean annual temperature	TEM	1000	2000–2020
Mean annual precipitation	PRE	1000	2000–2020
Annual sunshine duration	SSD	1000	2000–2020
Population density	POP	1000	2000–2020
Gross domestic product	GDP	1000	2000–2020

) were selected: the natural factors elevation, slope, normalized difference vegetation index (NDVI), mean annual temperature, mean annual precipitation, and annual sunshine duration, and the social factors population density and GDP.

2.3. Research Methods

2.3.1. Land-Use Dynamic Degree

In this study, we employed the land-use dynamic degree to quantitatively analyse the trends and rates of land-use change in Coastal China [35]. The formula was defined as follows:

$$K={\displaystyle \frac{L_b-L_a}{L_a}}\times {\displaystyle \frac{1}{T}}\times 100\%$$

(1)

where $K$ represents the dynamic degree of a specific land-use category (%); $L_a$ represents the area of the land-use category during the initial period (km²); $L_b$ represents the area of the land-use category during the final period (km²); and $T$ represents the study time span (years).

2.3.2. ESV Evaluation

The equivalent-factor method is the most widely used approach for assessing ESV. It first differentiates the various functions provided by ecosystems; establishes standardized, quantifiable equivalence factors for each service type within each ecosystem category, and finally integrates these factors with the spatial extent of the corresponding ecosystems to derive an overall valuation. Drawing on the research by Xie et al. [36] and Xu et al. [19], we employed a modified equivalence factor method [37] to establish an ESV estimation model for Coastal China. In this study, ecosystem services were classified into eleven categories according to the equivalent-factor table [36]. The standard equivalent coefficient was determined to be CNY 2181.78 per hectare based on the average net profit per unit area of three staple crops (rice, wheat, and corn) from 2000 to 2015 [38]. This enabled the development of a unit-area ESV table for different land-use categories (

Table 2. Coefficients of the value of ecosystem services of various types of land uses (CNY/ha/yr).

Primary Type	Secondary Type	Farmland	Forestland	Grassland	Water Area	Unused Land
Provision Services	Food Production	2421.78	545.45	501.81	1745.42	21.82
	Raw Material Production	545.45	1265.43	741.81	501.81	43.64
	Water Resource Supply	−2858.13	654.53	414.54	18086.96	21.82
Regulatory Services	Gas Regulation	1941.78	4167.20	2639.95	1679.97	152.72
	Climate Regulation	1025.44	12457.96	6959.88	4996.28	109.09
	Environmental Purification	305.45	3643.57	2290.87	12108.88	458.17
	Hydrological Regulation	3272.67	8159.86	5105.37	223065.19	261.81
Support Services	Soil Conservation	1134.53	5061.73	3207.22	2029.06	174.57
	Nutrient Cycling Maintenance	349.08	392.72	240.00	152.72	21.82
	Biodiversity	370.90	4625.37	2923.59	5563.54	152.72
Cultural Services	Aesthetic Landscape	174.54	2029.06	1287.25	4123.56	65.45

), where a service value coefficient of 0 was assigned for construction land. The ESV calculation formula was defined as follows:

$$ESV=\sum _{i=1}^n{A}_i\times {VC}_i$$

(2)

where ESV is the total ecosystem service value; A_i is the area of the land-use category i; VC_i is the unit-area ESV coefficient for the land-use category i; and n is the number of land-use categories.

2.3.3. ESV Sensitivity Assessment

To validate the accuracy of the ESV accounting and evaluate the reliability of the ESV coefficients, we calculated the coefficient of sensitivity (CS). The sensitivity analysis involved ±50% adjustments to the unit-area ESV equivalence coefficient (VC) and measures the responsiveness of ESV to VC variations [31]. The calculation formula was defined as follows:

$$\left.CS=\left|{\displaystyle \frac{\left({ESV}_a-{ESV}_b\right)/{ESV}_b}{\left({VC}_{an}-{VC}_{bn}\right)/{VC}_{bn}}}\right.\right|$$

(3)

where CS is the sensitivity index (dimensionless); ESV_a is the ESV after coefficient adjustment; ESV_b is the ESV value before coefficient adjustment; VC_an is the adjusted ESV coefficient for the land-use category n; VC_bn is the original ESV coefficient for the land-use category n; and n is the specific land-use category.

2.3.4. Geodetector

Geodetector is a statistical method for detecting spatial differences and their driving forces [38]. In this study, the factor detector and interaction detector modules were employed primarily to identify the factors influencing ESV in the research area. The factor detector quantifies the explanatory power of individual influencing factors for ESV spatial differences through the q value (range: [0, 1]), where higher values indicate a stronger explanatory power [39]. The interaction detector evaluates the interactive effects between multiple influencing factors by comparing single-factor q values with combined q values to determine the interaction type: nonlinear weakening, single-factor nonlinear weakening, bilinear enhancement, or independent [40]. The core computational formula is expressed as follows:

$$q=1-{\displaystyle \frac{1}{N{\sigma }^2}}\sum _{h=1}^n{N}_h{\sigma }_h^2$$

(4)

where q represents the explanatory power of a factor for ESV spatial heterogeneity; N represents the total number of sample units; h represents the stratum; n represents the number of strata (classification levels of influencing factor X); ${\sigma }^2$ represents the total variance in ESV; N_h represents the number of samples within stratum h; and ${\sigma }_h^2$ represents the variance in ESV within stratum h.

3. Results and Analysis

3.1. Analysis of the Land-Use Dynamic Degree

From 1980 to 2020, farmland and forestland collectively dominated China’s coastal land use, accounting for approximately 80% of the total area, and were primarily distributed across the North China Plain—the middle and lower valley of the Yangtze River, the southeastern coastal zones, and Liaoning Province (Figure 2). Temporally, the proportion of farmland continuously declined from 40.35% in 1980 to 37.04% in 2020, representing a net reduction of 3.30%, whereas the proportion of construction land significantly increased from 5.25% to 9.95% (+4.71%). Among the other land-use categories, forestland, grassland, and unused land decreased by 0.29%, 1.42%, and 0.24%, respectively, whereas water areas experienced a modest increase of 0.54%.

Spatially, farmland was concentrated in the plains of Jiangsu, Shandong, and Hebei; forestland prevailed in the mountainous/hilly terrain in the northeast (Liaoning) and south (Zhejiang, Fujian, Guangdong, and Guangxi); grassland clustered in the ecologically fragile northern and western zones; water areas displayed linear distributions along rivers, lakes, and coastlines; and construction land agglomerated in core urban clusters, particularly the Yangtze River Delta, Greater Bay Area, and Bohai Sea Rim. Notably, construction land expansion was most intensive in prime farmland zones, such as the North China Plain and Yangtze River Delta, forming a distinct spatial pattern of “core urban encroachment–peripheral farmland loss”.

The land-use transition matrices (Figure 3) revealed farmland-to-construction land conversion as the dominant cross-category transfer across all provinces, except the island administrative regions, and was concentrated in provincial capitals and urban agglomeration cores. Land-use dynamic degree analysis (

Table A1. Land-use dynamics in Coastal China from 1980 to 2020.

Land-Use Type		1980–1990	1990–2000	2000–2010	2010–2020	1980–2020
Individual Land-Use Dynamic Degree (%)	Farmland	−0.03	−0.17	−0.45	−0.16	−0.79
	Forestland	−0.02	0.07	0.04	−0.13	−0.04
	Grassland	0.08	−0.52	−1.18	0.04	−1.54
	Water area	−0.43	0.46	0.66	0.77	1.50
	Construction land	0.65	1.51	4.04	1.05	9.02
	Unused land	−0.47	−1.69	−3.24	1.38	−3.91
Comprehensive Land Use Dynamic Degree (%)		0.00	0.00	0.00	0.01	0.00

) was conducted to quantify the changes: farmland decreased by 40,560.80 km² (dynamic degree: −0.79%), forestland decreased by 2257.31 km² (−0.04%), grassland decreased by 17,711.89 km² (−1.54%), and unused land decreased by 2971.99 km² (−3.91%), whereas water areas and construction land increased by 7014.05 km² (+1.50%) and 60,034.05 km² (+9.02%), respectively, with the latter exhibiting the highest expansion intensity.

The transition scale analysis (Figure A1) revealed that farmland was the largest source category (67,196 km² transferred out), with 64,336 km² converted to construction land—the dominant sink category. Forestland and grassland transfers ranked second and third (18,690 km² and 18,018 km², respectively), confirming urbanization’s predominant consumption of farmland, forestland, and grassland. This transition was concentrated in the Bohai Rim, the Yangtze River Delta, and the Greater Bay Area, and exhibited strong coupling with regional economic agglomeration and population growth.

3.2. Assessment of ESV and Its Spatiotemporal Characteristics

3.2.1. Temporal Variation Characteristics of ESV

As shown in

Table 3. ESV of different land-use types in Coastal China from 1980 to 2020.

Land-Use Type	ESV (CNY Billion)
	1980	1990	2000	2010	2020
Farmland	444.32	443.00	435.32	415.94	409.10
Forestland	2238.62	2234.64	2249.49	2259.38	2228.92
Grassland	302.95	305.27	289.39	255.29	256.35
Water Area	1279.01	1224.08	1280.68	1365.84	1471.24
Unused Land	1.13	1.07	0.89	0.60	0.69
Total	4266.04	4208.08	4255.77	4297.06	4366.29

, the total ESV in Coastal China exhibited fluctuating growth from 1980 to 2020, rising from CNY 4266.04 to 4366.29 billion, a net increase of CNY 100.25 billion. This trend closely correlates with the structural land-use changes: although the ESV from farmland, forestland, grassland, and unused land declined due to area reductions, the ESV from water areas increased significantly over the four decades, offsetting the losses from the other categories. Further analysis revealed that the ESV trends strongly aligned with the dynamic degree of water areas, confirming that water bodies were the key driver of the regional ESV changes.

The detailed results of the phase-specific analysis (

Table 4. Value of individual ecosystem services in Coastal China from 1980 to 2020.

Primary Type	Secondary Type	ESV (CNY Billion)
		1980	1990	2000	2010	2020
Provision Services	Food Production	166.25	165.53	163.63	158.24	156.64
	Raw Material Production	104.70	104.47	104.07	102.33	101.23
	Water Resource Supply	−22.97	−26.19	−19.95	−8.34	0.42
Regulatory Services	Gas Regulation	354.65	353.85	352.31	346.01	342.29
	Climate Regulation	804.54	802.84	803.05	796.14	788.71
	Environmental Purification	288.54	285.92	287.97	288.83	290.78
	Hydrological Regulation	1692.27	1646.75	1689.62	1746.85	1824.50
Support Services	Soil Conservation	368.08	367.31	366.52	361.59	358.03
	Nutrient Cycling Maintenance	41.80	41.70	41.41	40.45	39.97
	Biodiversity	319.51	318.16	318.79	316.94	315.64
Cultural Services	Aesthetic Landscape	148.67	147.75	148.36	148.04	148.10

and

Table A2. ESV change rates of different land-use types in Coastal China from 1980 to 2020.

Land-Use Type	ESV Change Rate (%)
	1980–1990	1990–2000	2000–2010	2010–2020	1980–2020
Farmland	−0.30	−1.74	−4.45	−1.64	−7.93
Forestland	−0.18	0.66	0.44	−1.35	−0.43
Grassland	0.76	−5.20	−11.78	0.41	−15.38
Water Area	−4.29	4.62	6.65	7.72	15.03
Unused Land	−4.68	−16.92	−32.42	13.76	−39.11

) are as follows. From 1980 to 1990, the total ESV decreased by CNY 57.97 billion. While the grassland ESV increased by CNY 2.32 billion due to localized ecological restoration, declines in the other categories were driven by sporadic construction land expansion encroaching onto farmland and areas dominated by forests. From 1990 to 2000, ESV increased by CNY 47.69 billion, driven primarily by water areas (+CNY 56.60 billion, +4.62%) through coastal wetland conservation and aquaculture development, and forests (+CNY 14.84 billion) from afforestation in the southern hills. However, intensified development reduced the ESV of farmland, grassland, and unused land by CNY 7.69, 15.88, and 0.18 billion, respectively. From 2000 to 2010, ESV continued to grow (+CNY41.29 billion) and was sustained by water areas (+CNY 85.16 billion) and forests (+CNY 9.90 billion). The farmland ESV decreased by CNY 19.38 billion (−4.45%) due to industrial park expansion in the Yangtze River Delta and Greater Bay Area, whereas the grassland ESV fell by CNY 34.09 billion (−11.78%) due to coastal salinization. From 2010 to 2020, the total ESV in the study area increased to CNY 4366.29 billion, with a net increase of CNY 69.23 billion and a growth rate of 1.61%. This overall gain was primarily driven by a significant increase in the ESV of water areas (+CNY 105.40 billion) and unused land (+CNY 0.08 billion), which partially offset the decline in the ESV of farmland and forest land. This pattern of change reflects the characteristics of land-use transition in the study area: a reduction in farmland and forests due to urbanization, while ecological protection measures may have facilitated the restoration of water areas and grassland.

From the perspective of secondary service types (Table 4 and

Table A3. The change rate of individual ESVs in Coastal China from 1980 to 2020.

Primary Type	Secondary Type	ESV Change Rate (%)
		1980–1990	1990–2000	2000–2010	2010–2020	1980–2020
Provision Services	Food Production	−0.44	−1.15	−3.29	−1.01	−5.78
	Raw Material Production	−0.23	−0.38	−1.67	−1.08	−3.32
	Water Resource Supply	14.00	−23.82	−58.18	−105.02	101.82
Regulatory Services	Gas Regulation	−0.22	−0.44	−1.79	−1.08	−3.49
	Climate Regulation	−0.21	0.03	−0.86	−0.93	−1.97
	Environmental Purification	−0.91	0.72	0.30	0.68	0.78
	Hydrological Regulation	−2.69	2.60	3.39	4.45	7.81
Support Services	Soil Conservation	−0.21	−0.22	−1.34	−0.98	−2.73
	Nutrient Cycling Maintenance	−0.24	−0.69	−2.31	−1.19	−4.38
	Biodiversity	−0.42	0.20	−0.58	−0.41	−1.21
Cultural Services	Aesthetic Landscape	−0.62	0.41	−0.22	0.04	−0.39

), hydrological regulation was the most valuable ecosystem service, followed by climate regulation, whereas water supply services consistently had the lowest value. Over the past 40 years, there has been a slight decline in most service values, except for water resource supply, environmental purification, and hydrological regulation, which present varying degrees of growth. Notably, the value of water supply services underwent a remarkable transformation: it surged from CNY −22.97 billion in 1980 to CNY 0.42 billion in 2020, representing a 101.82% change and marking a critical transition from a negative value to a positive value.

This shift was closely linked to the interactive effects of farmland and water areas. The water supply service value of farmland was CNY −2858.13 per hectare, which was attributed to the ecological cost, where agricultural irrigation consumption surpassed the recharge capacity. Thus, the continuous reduction in farmland area directly diminished its negative contribution and steered the ESV of the water supply towards a positive value. Conversely, the expansion of water areas enhanced natural water body recharge, directly increasing the ESV of water supply services. Consequently, the ESV of interlinked services—hydrological regulation and environmental purification—was lifted by 7.81% and 0.78%, respectively. These structural changes mirror the trade-off relationships among coastal ecosystem services. While farmland reduction leads to some loss in regulatory service values, the strengthened ecological functions of water areas partially offset the negative effects on water supply, giving rise to a service value rebalancing mechanism centred on “reducing water-consuming land use and increasing water-storing land use”.

3.2.2. ESV Spatial Variation Characteristics

We visualized the ESV of the coastal regions from 1980 to 2020 at the county level and classified the values into five categories using natural breaks: very low (≤CNY 2.45 billion), low (CNY 2.45 to 6.47 billion), medium (CNY 6.47 to 11.45 billion), high (CNY 11.45 to 19.36 billion), and very high (≥CNY 19.36 billion).

The spatial pattern showed a significant gradient (Figure 4): there was a north–south gradient, with the southern region generally having a higher ESV than the northern region. The southern coastal area, with its subtropical humid climate, promoted forest growth (15–20% greater vegetation coverage than the northern coastal area) and a dense distribution of high-efficiency ecosystems, such as estuarine wetlands and mangroves, had a continuous high-value ESV zone in mountainous areas, such as Fujian, Guangdong, and Guangxi. The northern Bohai Sea area, which is affected by the climate and coastal salinization, and industrialization, had a lower ESV than the southern region, except for wetland-rich areas such as the Yellow River Delta and Liao River Estuary. There was also a land–sea gradient, where inland areas tended to have higher values than coastal areas. Due to intense human activities, such as port construction and land reclamation (e.g., a 230 km² increase in land reclamation in the Pearl River Estuary over 30 years), coastal counties suffered from coastal wetland degradation and ecosystem fragmentation. Their ESV was 12–15% lower than that of inland counties on average. In contrast, inland counties, such as the mountainous areas in Zhejiang, Fujian, and northern Guangdong, which have better preserved natural ecosystems, maintained high ESV levels.

The spatial autocorrelation analysis (Figure 5) showed a significant clustering pattern for ESV. High-value clusters (H-H clusters) corresponding to areas with excellent natural conditions were mainly found in the Nanling Mountains, Zhejiang–Fujian hills, and estuarine wetlands, such as the Pearl River Estuary and Jiulong River Estuary, and they were highly stable. Low-value clusters (L-L clusters), coupled with urban agglomerations, started from core areas, such as the Beijing–Tianjin–Hebei region, the Yangtze River Delta, and the Greater Bay Area, and spread inward along transport corridors. Most anomalies were high–low anomalies (high-value areas surrounding low-value counties), which appeared mainly in “ecological island” areas around large cities (e.g., counties around Chongming Island, Shanghai). Low–high anomalies (low-value areas surrounding high-value counties) were mostly around coastal wetland enclaves (e.g., the Red-Crowned Crane Reserve in Yancheng, Jiangsu).

3.2.3. Sensitivity Analysis

The sensitivity analysis results (

Table 5. Results of sensitivity analysis of ESV.

Land-Use Type	1980	1990	2000	2010	2020
Farmland	0.10	0.11	0.10	0.09	0.09
Forestland	0.52	0.53	0.53	0.53	0.51
Grassland	0.07	0.07	0.07	0.06	0.06
Water Area	0.30	0.29	0.30	0.32	0.34
Unused Land	0.00	0.00	0.00	0.00	0.00

) showed that from 1980 to 2020, the sensitivity indices of all the land-use types in Coastal China were less than 1, indicating an inelastic response of ESV to land-use changes. Forestland had the highest average sensitivity index of 0.52, indicating that its changes in per-unit area ecological service value had the most significant effect on the regional ESV. Water areas ranked second, with an index of 0.3093. In contrast, the sensitivity indices of cultivated land, grassland, and unused land were less than 0.10, which indicates that changes in their ecological service value weakly influenced the overall ESV. These results meet the sensitivity analysis criterion (a sensitivity index <1 indicates stable valuation results) and confirmed the reliability of the per-unit area ecological service value parameters used in this study.

3.3. Analysis of ESV Influencing Factors at the County Level

3.3.1. Factor Detector Analysis

Next, we examined Coastal China at the administrative county level. We used ESV as the dependent variable, and factors such as annual mean temperature (X1), annual mean precipitation (X2), slope (X3), elevation (X4), NDVI (X5), annual mean sunshine duration (SSD, X6), population density (X7), and GDP (X8) were the independent variables. The data for each factor were classified into five categories using the natural-breaks (Jenks) method [41] to explore the drivers of spatial differences in ESV. Given that Geodetector is not a statistical method based on probability distribution models, we used Bootstrap resampling to estimate the confidence interval of the q value and calculated its 95% confidence interval.

The factor detector results from Geodetector (

Table 6. Interaction detector analysis for ESV influencing factors in Coastal China.

Year	2000		2010		2020
Detected Factor	q Value	p Value	q Value	p Value	q Value	p Value
X1	0.20	0.00	0.19	0.00	0.20	0.00
X2	0.19	0.00	0.27	0.00	0.13	0.00
X3	0.29	0.00	0.30	0.00	0.29	0.00
X4	0.28	0.00	0.27	0.00	0.27	0.00
X5	0.24	0.00	0.28	0.00	0.35	0.00
X6	0.13	0.00	0.20	0.00	0.09	0.00
X7	0.27	0.00	0.31	0.00	0.29	0.00
X8	0.27	0.00	0.34	0.00	0.29	0.00

) showed that from 2000 to 2020, the explanatory power of the socioeconomic factors for ESV significantly increased and remained above that of most of the natural factors. The explanatory power of population density increased from 0.27 to 0.31 and then slightly decreased to 0.29. The explanatory power of GDP increased from 0.27 to 0.34 and then marginally decreased to 0.29. This finding indicates a growing impact of human activities on ecosystem value, with a slight easing in 2020. Among the natural factors, the explanatory power of NDVI rose remarkably, whereas that of the SSD fluctuated downward. Notably, terrain factors, such as slope and elevation, maintained a stable explanatory power, reflecting their fundamental long-term role in ecological services. Overall, population density, GDP, and NDVI were the key drivers of the ESV changes. The spatial differences in ESV in Coastal China resulted from the combined effects of natural and socioeconomic factors.

3.3.2. Interaction Detector Analysis

Based on the factor interaction detector results from Geodetector (Figure 6), the interactions between all the influencing factors showed a significant enhancement compared to the effects of the individual factors alone. Among the 48 interaction detector datasets, all factor combinations exhibited a bivariate enhancement relationship. Specifically, the interaction q values in different years were as follows. In 2000, the interaction between annual mean precipitation (X2) and GDP (X8) had the highest q value of 0.50. In 2010, the interaction between annual mean precipitation (X2) and population density (X7) reached a q value of 0.55. In 2020, the interaction between annual mean temperature (X1) and population density (X7) had a q value of 0.55. Further analysis indicated that the explanatory power of certain factor combinations, such as annual mean temperature (X1) with GDP (X8), annual mean temperature (X1) with population density (X7), annual mean temperature (X1) with NDVI (X5), annual mean precipitation (X2) with NDVI (X5), annual mean precipitation (X2) with elevation (X4), annual mean precipitation (X2) with slope (X3), slope (X3) with GDP (X8), slope (X3) with population density (X7), and slope (X3) with NDVI (X5), remained above 0.40 from 2000 to 2020.

The results showed that natural factors (especially annual mean temperature, annual mean precipitation, slope, and NDVI) significantly interact with the other factors. They shaped the environmental baseline of Coastal China and influenced human activity and regional development. The synergistic effects exerted by socioeconomic drivers—population density and GDP—on all other factors intensified markedly throughout the study horizon, while the mutual interactions within natural factors continued to hover at a comparatively steady level. This finding reflects the growing human impact on the natural environment. Most factor interactions peaked in 2010, likely due to rapid economic growth and intensified human activities. By 2020, some interaction strengths had declined, possibly because of the implementation of ecological protection policies. Overall, both natural and socioeconomic factors drove the spatial differences in ESV. The interaction of any two factors enhanced the spatial heterogeneity of ESV. The dominant role of the socioeconomic factors gradually strengthened, whereas the interactions among the natural factors remained relatively stable.

3.3.3. Analysis of ESV Influencing Factors in Different Provinces

The detection results for the ESV influencing factors across different provinces (Figure A2) revealed that the ESV formation mechanisms varied significantly, both temporally and spatially. In 2000, ESV was mainly influenced by the natural factors. However, with the advancement of urbanization, the impact of the socioeconomic factors on ESV became considerably stronger by 2010, and had formed complex interactions with the natural factors. By 2020, the dominance of the socioeconomic factors had further solidified, and the interactions between the natural factors and human activities had become more complex, leading to a diverse range of impacts on ESV.

Beijing’s ESV in 2000 was primarily influenced by natural factors. Annual mean temperature (X1) and annual mean precipitation (X2) had high q values, indicating that the climate played a key role in shaping ESV. Notably, annual mean precipitation (X2) and NDVI (X5) showed a significant synergistic enhancement effect, jointly driving up the ESV.

In 2000, Fujian’s ESV was also dominated by the natural factors. Annual mean precipitation (X2) and NDVI (X5) had high q values, reflecting the significant contributions of the favourable climate and abundant vegetation resources in Fujian to its ESV. Fujian’s unique geographical features, namely, its slope (X3) and elevation (X4), also had a significant impact on ESV. In particular, in mountainous areas, these topographical factors, together with the climate and vegetation, shaped the spatial distribution and functional characteristics of the ecosystem.

In China’s coastal economic hubs, such as Zhejiang, Jiangsu, and Guangdong, the factors influencing ESV evolved similarly. In 2000, both natural and socioeconomic factors mattered, as reflected by the high q values for annual mean temperature (X1), annual mean precipitation (X2), population density (X7), and GDP (X8), highlighting the close natural-socioeconomic links. By 2010, the socioeconomic factors had grown more influential, especially GDP, indicating that economic development had an increasing impact on ESV. Specifically, Zhejiang’s topography [slope (X3) and elevation (X4)] affected ESV; Jiangsu’s flat terrain made the climate and socioeconomic factors more prominent; and Guangdong presented a nonlinear slope (X3)–GDP (X8) interaction and a synergistic annual mean temperature (X1)–population density (X7) effect.

In 2000, Guangxi’s ESV was mainly shaped by natural factors, such as annual mean precipitation (X2) and NDVI (X5), with its karst landforms (elevation (X4) and slope (X3)) also playing a key role in sustaining ESV. In 2000, the ESV in Hebei was driven by natural factors, such as annual mean temperature (X1), annual mean precipitation (X2), and slope (X3). By 2010, population density (X7) and GDP (X8) had gained influence, likely due to Hebei’s role in the Beijing–Tianjin–Hebei region. There was a synergistic annual mean precipitation (X2)–NDVI (X5) effect and a potential antagonistic population density (X7)–slope (X3) interaction.

The ESV in Shanghai in 2000 was dominated by socioeconomic factors, such as population density (X7) and GDP (X8), reflecting its status as an economic centre. By 2020, both the natural and socioeconomic factors had significant q values, with more complex interactions. The ESV in Tianjin in 2000 was influenced equally by the natural and socioeconomic factors, including annual mean temperature (X1), annual mean precipitation (X2), population density (X7), and GDP (X8). In 2020, the q values of these factors remained high. Owing to its location on the North China Plain, elevation (X4) and slope (X3) had less impact, whereas the climate and human activities were more influential.

4. Discussion

4.1. Comparison with Existing Research

This study focused on the coastal zones of China—a region with the dual functions of an “economic engine” and “ecological barrier” and intense human–environment coupling—filling the gap in long-term ESV assessments of this area. Although national-scale studies [26,27] have analysed changes in ESV in China, they failed to reveal the unique processes of land–sea interactions. We found that coastal ESV is shaped not only by universal drivers such as urbanization but also by region-specific factors like wetland loss due to reclamation and salinization, which is significantly different from inland or national patterns. The increase in the water area ESV compensated for value losses in the other land-use types, which is consistent with the equivalent-factor method framework proposed by Xie et al. [36]. Moreover, this study identified a unique “reduced water-consuming land use and increased water-storing land use” balance mechanism in coastal areas. For example, the water area ESV in the Pearl River Estuary increased by CNY 133.10 billion due to wetland protection and large-scale aquaculture development from 1990 to 2010. This echoes the findings of Xu et al. [19] in the Bohai Bay region and demonstrates the universal applicability of the mechanism along coastlines.

The analysis of the driving mechanisms showed that the explanatory power of socioeconomic factors (population density and GDP) for ESV increased from 0.27/0.27 in 2000 to 0.29/0.29 in 2020, supporting Liu et al. [8], who showed that urban agglomeration expansion increased ESV heterogeneity. However, in this study, we further explored nonlinear interactions using Geodetector and found that the q value for the interaction between annual mean temperature and population density reached 0.55. This finding reflects a combined effect of climate and human activity, aligning with the coastal region’s “climate-sensitive and high-intensity development” characteristics.

Compared with single-urban agglomeration or watershed studies [7,35], this study revealed the spatial differentiation pattern of ESV in Coastal China, with high values in southern and inland areas and low values in northern and coastal areas. With a subtropical climate and mangrove wetlands, southern coastal areas accounted for 28.6% of the high-value ESV zones. In contrast, 41.2% of the counties in the northern Bohai Sea region, which are affected by salinization and heavy industry, are low-value ESV zones. This finding corresponds with the results of local studies on the decline in farmland and grain-supply functions in the Liaodong coastal area [24].

Furthermore, our findings gain additional global relevance when contrasted with international experiences in integrating ecosystem assessments into spatial planning. The methodology proposed by Córdoba and Camerin (2024) demonstrates how directly incorporating ecosystem attributes and territorial pressures into planning instruments can enhance environmental protection decisions [42]. Similarly to the MAES-based approach applied in the Spanish context, our study underscores that ESV dynamics in Coastal China are profoundly influenced by land-use planning decisions, particularly the conversion of ecologically vital areas for development. The successful integration of ecosystem service assessments into land use planning, as seen in European practices [42], highlights a viable pathway for China to operationalize its “ecological civilization” and “Beautiful China” initiatives. Key to this is translating ESV evaluations into legally binding protection schemes within territorial spatial planning, such as establishing ecological protection redlines specifically informed by ESV hotspots and vulnerability assessments.

4.2. Ecological Implications of ESV Trends

The observed ESV changes not only reflect land-use transitions but also signify deeper adjustments in ecosystem functioning across Coastal China. The most remarkable shift is the transformation of water-supply services from negative to positive values, declining from −CNY 22.97 billion in 1980 to +CNY 0.42 billion in 2020. This indicates a reduction in agricultural water-consumption pressure and a gradual recovery of natural recharge capacity, largely driven by wetland conservation and aquaculture expansion. It demonstrates that the strategy of “reducing water-consuming land use and increasing water-storing land use” has begun to rebalance hydrological ecosystem services effectively.

Simultaneously, the steady increase in hydrological regulation services (+7.81% over 40 years) implies enhanced flood mitigation and coastal resilience, particularly in estuarine wetlands such as the Pearl River Estuary. Conversely, the decline in farmland ESV (−7.93%) reflects not only land-use conversion but also a reduction in agroecosystem pressure, potentially benefiting water quality and soil health. However, the slight decrease in forest ESV (−0.43%) raises concerns about carbon sequestration capacity and habitat quality, especially in southern hilly regions where urban encroachment is intensifying. These findings underscore the urgent need to protect high-value ecological patches (e.g., mangroves and forest belts) while guiding urban expansion toward low-ecological-impact zones, ensuring sustainable coastal development in the context of rapid urbanization.

4.3. Policy Implications

4.3.1. Region-Specific Ecological Protection Strategies

For high-value ESV areas, such as Fujian and Guangdong (e.g., mangroves), the nature reserve system should be upgraded, ESV metrics should be incorporated into local government performance evaluations [33], and land reclamation projects should be strictly limited. In the salt-affected Bohai Sea region, a “farmland fallow–wetland restoration” crop rotation model should be promoted. Drawing on the Yellow River Delta’s ecological water replenishment experience [43], hydrological regulation should be enhanced through the coordinated governance of river basins, with the aim of reducing the proportion of salt-affected farmland by 15–20%.

4.3.2. Ecological Management Mechanisms for Urban Agglomerations

In urbanization cores (e.g., the Yangtze River Delta and Greater Bay Area), an “ESV loss compensation system” should be established. Ecological restoration in proportion to urban land expansion should be required, and the vertical spatial orientation development of industrial land should be encouraged. In the Liaodong–Shandong Peninsula urban agglomeration, port and industrial land reclamation should be strictly controlled. The Tianjin Binhai New Area’s “eco-land reclamation” model should be followed [34], aquaculture ponds should be transformed into ecological wetlands, and the water area ESV should be increased.

4.3.3. Land-Use Planning Optimization Paths

Based on ESV sensitivity indices, “forest–water” core protection grids should be designated during national land space planning. The annual forested area reduction in regions such as the Yangtze River Delta and Greater Bay Area should be kept at ≤0.1%. A “water-consuming land-use (farmland, construction land) reduction–water-storing land-use (water-area, wetland) increase” conversion model should be established. Ecological-compensation funds should be used to incentivize local governments to convert 10% of new construction land into wetland restoration areas.

4.3.4. Policy Coordination and Implementation Suggestions

ESV assessment should be incorporated into the “Beautiful China” development evaluation system. Cross-regional ecological service value compensation should be piloted in the coordinated development of the Beijing–Tianjin–Hebei region. For example, Beijing pays ESV compensation to Chengde, Hebei’s water source conservation area (leveraging the synergy between annual precipitation and NDVI). Additionally, resource and environmental science data platforms [30] should be integrated with Geodetector to establish a dynamic early warning system for the ESV of coastal regions. Red alert thresholds should be set for factors with q values > 0.3 (e.g., GDP and population density) to dynamically balance development intensity with ecosystem services.

4.4. Discussion of Research Methods

Future research could enhance the methodological precision of this study in the following ways. First, high-resolution remote-sensing data (e.g., Sentinel-2) could be integrated with the InVEST model to dynamically simulate wetland ecosystem services at the patch level. Second, nighttime light [43] and industrial structure data could be incorporated to construct a two-dimensional “economic–ecological” model. Third, a coupled analysis of climate factors (e.g., extreme precipitation frequency) and ESV could be conducted to examine the interactive stress effects of climate change and human activities. Cross-scale comparative studies are also recommended, such as research that compares the evolutionary paths of ESV in the Bohai Sea (industry-dominated) and Greater Bay Area (service industry-dominated) at the provincial level, extends the time dimension to 2030 to simulate ESV responses under the “dual-carbon” goals, and includes data across the Taiwan Strait to establish a complete sea-land interaction research system. Additionally, methods for constructing ecological corridors, such as those proposed by Liang et al. [28], could be used to explore connectivity protection strategies for high-value ESV areas.

4.5. Limitations and Future Directions

This study has several limitations. First, the 30 m resolution of the land-use data—although adequate for county-level analysis—cannot capture patch-scale changes in coastal wetlands (e.g., mangrove fragmentation). Second, only population density and GDP were included as socioeconomic variables; sector-specific indicators such as the share of energy-intensive industries were not considered. Third, the use of county-level rather than grid-level analysis, while facilitating direct comparison of counties, may mask within-county heterogeneity. Fourth, the sensitivity analysis only examined the response of ESV to land-use changes and did not address the compounding effects of climate change (e.g., sea-level rise). Finally, inflation or price adjustments over the 40-year period were not taken into account. These shortcomings are common to most existing coastal ESV studies [14,20] and urgently need to be addressed in future research.

5. Conclusions

In this study, we analysed land-use and socioeconomic data from 1980 to 2020 and used the equivalent-factor method and Geodetector to examine the spatial–temporal differentiation and driving mechanisms of ESV in Coastal China. The key findings are given below.

From 1980 to 2020, the land-use structure in Coastal China underwent significant changes. Cultivated land and forestland accounted for approximately 80% of the land use, but the proportion of cultivated land declined from 40.35% to 37.04%, whereas construction land increased from 5.25% to 9.95%. Spatially, this change formed a typical “core city expansion–peripheral farmland encroachment” pattern, which was most pronounced in the Yangtze River Delta, Greater Bay Area, and Bohai Rim urban agglomerations.

The total ESV showed a fluctuating upward trend, increasing from CNY 4266.04 to 4366.29 billion. This increase was mainly attributed to a significant increase in the ESV of water areas (15.03%). Spatially, cultivated land was concentrated in plains, such as Jiangsu and Shandong; forestland was located in southern mountainous regions and the coastal areas of Liaoning; and construction land was highly concentrated in the Yangtze River Delta, Greater Bay Area, and Bohai Rim. The primary land-use transition was from cultivated land to construction land, especially in provincial capitals and urban agglomeration cores, forming a typical “core city expansion–peripheral farmland reduction” pattern.

The total ESV in Coastal China showed a fluctuating upward trend, increasing from CNY 4266.04 to 4366.29 billion, with a net increase of CNY 100.25 billion. This increase was mainly due to a significant increase in the water area ESV (6.49%), which offset the value loss of other land-use types, such as farmland and forestland. The ESV exhibited significant spatial heterogeneity, presenting gradient patterns of “high in the south, low in the north” and “high in inland areas, low in coastal areas”. In southern coastal areas, due to favourable climatic conditions and abundant wetland resources, high-value ESV zones formed in mountainous and hilly regions, such as Fujian and Guangdong. In contrast, in the northern Bohai Sea area, the ESV was generally low due to salinization and industrialization, but estuarine wetlands still maintained local high-value zones. The ESV in coastal counties was 12–15% lower than that in inland adjacent counties because of human activities, such as port construction and land reclamation, resulting in a distinct land-sea gradient.

The explanatory power of socioeconomic factors for ESV significantly increased over the study period. The q values of population density and GDP rose from 0.27 and 0.27 to 0.29 and 0.29, respectively, making them the core drivers of changes in ESV. Among the natural factors, the explanatory power of NDVI increased from 0.24 to 0.35, highlighting the importance of vegetation cover for ecosystem services. The interaction detector revealed that both natural factors (e.g., annual mean temperature and precipitation) and socioeconomic factors (e.g., population density and GDP) had a bivariate enhancement effect. In 2020, the q value for the interaction between annual mean temperature and population density reached 0.55, revealing the combined impact of human activities and the natural environment. The sensitivity analysis indicated that forestland (0.52) and water areas (0.31) had the most significant impacts on ESV and that the accounting method showed high reliability.

This study provides insights into optimizing coastal land-use planning. For example, we suggest strengthening the protection of southern mountainous areas and estuarine wetlands, addressing salinization and industrial pollution in the northern Bohai Sea area, adjusting the pace of construction land expansion in urban agglomeration low-value zones, and establishing a value-balancing mechanism with “reduced water-consuming land use and increased water-storing land use”, in line with the “dual-carbon” goals to promote a green and low-carbon transition. This study fills the gap in long-term ESV assessments for coastal regions and provides theoretical support and a decision-making reference for the “Beautiful China” initiative.