Introduction

Ecosystem services (ESs) encompass the natural environmental conditions and effects involved in the formation and maintenance of ecosystems, representing all the benefits derived by humans from these ecosystems (Daily et al., 2000). While protected areas now cover 18% of China’s land area due to sustained efforts, ESs exist in a degraded state. Structural disruptions in ecosystems have evolved into functional disturbances, with the capacity and quality of services per unit area of forests, grasslands, and other ecosystems falling below global averages (Fu, 2013). The key difficulty in managing ESs lies in addressing the interdependencies among them, which are often characterized by trade-offs and synergies. ESs also cater to a diverse array of stakeholders (Rall et al., 2015; Cowling et al., 2008; Li et al., 2013). For instance, the over-emphasis of food provision from grasslands in western China can lead to desertification or soil erosion, which impacts sand stabilization efforts and increases the frequency of dust storms in northern China. Although the precise mechanisms of energy, material, and information flows remain unclear, over-stressing any one ES inevitably damages others (Peng et al., 2023; Zhang et al., 2007). As such, it is crucial to optimize the natural resource management of different ESs while reconciling the tensions between them (Turner et al., 2000; Bryan, 2013).

National parks, characterized by large protected areas, relatively intact ecological processes, and high public recognition, occupy a dominant position in the nature reserve system (Hou et al., 2023). The current development model of nature reserves in China, oriented toward quantitative growth, is unsustainable, and must be shifted toward a quality-priority pathway centered on optimizing spatial structures and improving ESs (Wang, 2017). In 2013, the Third Plenary Session of the 18th CPC Central Committee proposed the establishment of a national park system. In 2015, thirteen ministries, including the National Development and Reform Commission, jointly issued the Pilot Program for Establishing a National Park System. The CPC Central Committee and the State Council then released the Overall Plan for Reform of the Ecological Civilization System, further clarifying specific requirements for establishing a national park system. This was followed by the promulgation of top-level documents such as the Guiding Opinions on Establishing a Nature Reserve System Centered on National Parks. In 2022, four ministries, including the National Forestry and Grassland Administration (National Park Service), jointly issued the Spatial Layout Plan for National Parks, which designated 49 candidate national park areas across 28 provinces, covering over 1.1 million km2. Once completed, this will become the world’s largest national park system.

Historically, protected areas have faced several challenges prior to their official designation as national parks, including fragmented protection types, unclear baselines for asset valuation, prominent community conflicts, and difficulties transitioning to a public welfare-oriented model (Yang and Cao, 2018; Huang et al., 2018). Consequently, the scientific assessment of ecosystem service value (ESV) has become an important foundation for the effective natural resource management of national parks (Mulongoy and Gidda, 2008). Current trends in ES assessment have moved beyond individual or isolated measurements toward the examination of interactions among multiple ESs, aiming to avoid compromising one benefit in the pursuit of another (Daily and Matson, 2008; Tallis et al., 2008; Su and Liu, 2023). Among academics, there is a predominant focus on researching the trade-offs and synergies between ESs. Existing studies recommend first grouping ESs by relevance, and then analyzing the relationships between groups to reveal interactions (Kareiva et al., 2007; Bennett et al., 2009; Raudsepp-Hearne et al., 2010).

Since China first proposed its national park system in 2013, Chinese scholars have explored the ESs in 10 officially established pilot national parks from various perspectives. For example, Ran et al. (2023) used an InVEST model to analyze the changes in ES patterns in the Qianjiangyuan National Park (QNP) pilot area over 30 years and explored the relationships between them. Cui et al. (2024) constructed an ecological security framework for Wuyishan National Park based on ES importance. Du et al. (2019b) measured the spatiotemporal variations and driving factors of hydrological regulation in Tropical Rainforest National Park. Despite these and other valuable contributions, research on ES trade-offs and synergies in China’s nascent national park system remains limited. In general, the research on ESV, both domestically and internationally, is relatively well-developed, and a preliminary measurement index system has been established. However, because the development of national parks in China began relatively recently—only about 10 years ago—the theoretical frameworks and evaluation methods applicable to these parks require further investigation. In addition, the research on the trade-offs and synergy of ESVs at the national park scale is very limited, and their multi-scale spatiotemporal dynamics and impact mechanisms have not been revealed. Although the concept of ESs is a valuable tool for the practical management of national parks, its application in the construction of China’s national park system is in its infancy. Understanding how ESs have changed, identifying trade-offs and synergies between ESs, and fully realizing the ecosystem functions of national parks have become pressing issues for ensuring successful ongoing national park development.

In 2015, the Chinese government selected 12 provinces and cities to pilot its national park system, and the Qianjiangyuan area in Zhejiang Province was identified as one of the pilot areas. In 2016, the National Development and Reform Commission officially approved the Implementation Plan for the Pilot Area of Qianjiangyuan National Park System. The plan clearly proposed that the construction of the QNP system should aim to establish a unified, standardized, and efficient management system, enhance public welfare, strengthen ecological protection, and achieve harmonious coexistence between humans and nature. Before the implementation of the plan, the area was characterized by fragmented and uncoordinated management, limited funding, and pressures from community development (Huang et al., 2018; Hou et al., 2023). The QNP system pilot project will solve the contradiction between people and land and the dilemma of regional industrial transformation via the innovative development of ecological economic transformation. It is expected to play a pioneering and demonstrative role in exploring industrial structure transformation and regional sustainable development in densely populated and economically developed areas in eastern China. Although the ESV of the QNP pilot area has been evaluated in recent years, the trade-offs and synergies between different ESs, especially under specific socioeconomic conditions, are not well understood (Duan et al., 2021).

Based on this context, China’s QNP pilot area was selected as a case study to evaluate the spatiotemporal changes and spatial distribution characteristics of ESV, identify the influencing factors of ES evolution under specific socioeconomic conditions, and measure the complex relationships between ESVs. This study aims to develop a scalable and replicable research framework for national parks established in the economically developed, densely populated southern Yangtze River Delta region, whose collective forests account for a large proportion of land. The research results also provide important theoretical and practical support for the management of national park ecosystems.

The main objectives of this study are: (1) to scientifically evaluate the changing characteristics of ESV in the QNP pilot area and identify the spatial clustering characteristics of ESV; (2) to integrate various indicators and evaluation criteria to measure the socioeconomic development of national parks, with the goal of identifying key factors affecting the evolution of ESV in national park pilot areas; and (3) to scientifically identify the trade-offs and synergies among various ESVs in national park pilot areas, and promote the application of ES research in national park management decisions. The novelty of this research lies in the 30-year multi-domain ES study on the only national park pilot area in eastern China. This research includes both value assessment and spatial measurement analysis, as well as an investigation of the interrelationships between ESVs and their impact mechanisms under specific socioeconomic conditions. It ensures the comparability, generalizability, and replicability of research methods and results among multiple national parks, while also demonstrating practical feasibility.

Materials and methods

Overview of the study area

The QNP pilot area (herein referred to as “the pilot area”) is located in the western part of Zhejiang Province, at the junction of Zhejiang, Jiangxi, and Anhui provinces (Fig. 1). Established in 2015 as one of China’s first 10 national park pilot units, the core values of the pilot area include the protection of natural resources, scientific research, ecological services, demonstration, and promotion. These values focus on protecting nationally representative, intact, low-altitude subtropical evergreen broadleaf forest ecosystems, safeguarding ancient geological heritage, and promoting a unique, harmonious balance between humans and nature.

Fig. 1
figure 1

Basic overview of the research area.

Full size image

Covering ~252 km2, the pilot area serves as a vital connectivity node for fundamentally improving the ecological environment quality in central and eastern China. It integrates three natural protected areas: the Gutianshan National Nature Reserve, Qianjiangyuan National Forest Park, and the Qianjiangyuan Provincial Scenic Area. The area comprises two management divisions–the core protected area and the general controlled area–along with four functional divisions: the core protected area, an ecological conservation area, a recreation/exhibition area, and a traditional utilization area.

Property rights within the pilot area are highly complex. The state-owned land totals 48.64 km2, mainly comprising the Qixi Sub-farm of Kaihua Linchang, the Suzhuang Sub-farm, and the Gutianshan National Nature Reserve, which constitute 19.30% of the total area. The remaining 80.70% of the land is collective land, which covers 203.52 km2. The industrial structure is relatively homogeneous; resident income is derived primarily from agriculture, forestry, and migrant work. The pilot area spans four towns and 19 administrative villages with a total population of 9744.

Located in the economically developed Yangtze River Delta region, the pilot area is under substantial pressure concerning ecological protection. Conducting an ESV assessment and analyzing trade-off relationships in this densely populated area with complex collective property rights provides a valuable demonstration for exploring mutually beneficial pathways to balance ecological protection and economic development in developed regions. This issue, which underlies the core values and system construction effects of the pilot area, holds critical significance.

The selection of this region as the research object was mainly based on the two aspects of theory and application. At the theoretical research level, given the diverse types of natural protected areas, unclear ESVs, and fragmented management rights involved in the pilot area, it is crucial to systematically conduct ESV assessments and research to balance synergies within the area, while enhancing the theoretical framework for research on national park ESs. At the application level, the pilot area is located in China’s Yangtze River Delta, where the regional socioeconomic development level is relatively high, resulting in significant ecological protection pressure. Conducting assessments of ESVs in this area and their interrelationships plays a key demonstrative role in exploring a win-win path for collaborative ecological protection and economic development in economically developed, densely populated areas with complex collective land ownership. This is not only a crucial issue facing the research area, but is also central to its institutional construction.

Data sources

Based on the evolutionary characteristics of China’s ecological policies and considering data availability and comparability, 1990, 2000, 2010, 2015, and 2020 were selected as the time nodes for data collection. Except for the milestone year of 2015, when the pilot program for establishing a national park system was launched, the selection of other time points was based on 10-year increments. The Chinese government issued the Decision of the State Council on Further Strengthening Environmental Protection Work in 1990, indicating the strengthened role of ecological protection in resource development and utilization and the designation of sustainable development as the basic national strategy. By the year 2000, China had begun to embrace the concept that ecological environment protection is as equally important as economic and social development, recognizing its significance to national environmental security. In 2010, the Chinese government issued the National Ecological Environment Protection Outline and, in the following years, the construction of ecological civilization was comprehensively deployed nationwide. In 2020, China completed the initial pilot of the national park system, marking the beginning of the formal construction phase of China’s national parks.

Land-use remote sensing classification data were obtained from 1:100,000 land-use status remote sensing datasets covering the pilot area, which were provided by the Institute of Remote Sensing and Digital Earth, Chinese Academy of Sciences. Landsat TM/ETM+ images with a 30-m spatial resolution from 1990 to 2020 were used as the primary data source, and were generated through visual interpretation. Based on the national standard Land Use Status Classification (GB/T 21010–2017), in accordance with the actual conditions of the study area, and considering reasonable interpretability, the land-use types were divided into five categories: forest land, grassland, cultivated land, water area, and unused land.

National annual total grain output and sowing area data were sourced from the China Statistical Yearbook, while multi-year average grain yield data for the study area were obtained from the Kaihua Statistical Yearbook compiled by the Kaihua County Bureau of Statistics. The grain purchase prices in Zhejiang Province were provided by the Zhejiang Provincial Bureau of Grain and Material Reserves.

Research methods

Evaluation method using per-unit-area value equivalent factors

Methods for ESV accounting generally rely on either unit service function pricing or per-unit-area value equivalent factors. Compared to the former, which requires more parameters and cannot be calculated by readily standardized processes, the latter primarily relies on the land areas of different ecosystem types and their quantifiable value equivalencies to estimate ESV. Hence, the equivalent factor approach is better suited to assessing ESV at larger scales, like national parks, offering less costly data acquisition processes and simpler calculations (Xie et al., 2015).

The equivalent factor denotes the relative importance of a particular ecological service provided by an ecosystem compared to the food production services of cultivated land (Xie et al., 2008). To account for regional variation, the cultivated land-based revision method proposed by Xu et al. (2012) was applied to adjust the national average equivalent factors reported by Xie et al. (2003). The crop yields per unit of the study area and nationwide in 2020 were respectively 5853 and 5734 kg/hm2. Therefore, the revision coefficient of the study area was calculated as 1.021. The economic value per unit area of ecological services provided by natural systems, without human input, is estimated to be one-seventh of the value generated by food production services on existing cultivated land (Zhang et al., 2020a, 2020b). Based on the average crop yield of 5090 kg/hm2 from 1990 to 2020 and the 2020 provincial average crop price of 3.5 CNY/kg in the study area, the economic value equivalent of one unit area was 2597.69 CNY/hm2. These results facilitated the derivation of the table of ESV coefficients per unit area for the study area (Table 1).

Table 1 The value coefficients of various types of land ESs in the pilot area (unit: CNY hm−2 a−1).
Full size table

The ESV calculation method employed in this study (Yang et al., 2021; Liu et al., 2020) is expressed as follows:

$${\rm{ESV}}=\sum {A}_{k}\times {{VC}}_{k},$$
(1)

where ESV is the national park ESV (CNY); Ak is the area (hm2) of land cover type k; and VCk is the ESV coefficient per unit area (CNY hm−2 a−1) for land cover type k.

Sensitivity model

Previous studies have confirmed some uncertainty in the value coefficients (VCs) of ESs (Cao et al., 2020). A sensitivity analysis model was employed in this study to examine the accuracy of the VCs to enhance the scientific rigor of ESV assessments. Utilizing the concept of elasticity in economics, a coefficient of sensitivity (CS) was calculated to quantify the sensitivity of ESV changes to VC changes, thereby validating if the ESV coefficients align with the study area (Zhang et al., 2020a, 2020b; Su et al., 2020). If CS > 1, the elasticity of the ESV relative to the VC indicates a significant influence of the VC on the ESV, reflecting the relatively low accuracy of the calculation results. Alternatively, CS < 1 indicates inelasticity and hence more reliable calculation results. The equation is expressed as follows (Das and Das, 2019):

$${\rm{CS}}=\left|\frac{\left({{ESV}}_{j}-{{ESV}}_{i}\right)/{{ESV}}_{i}}{\left({{VC}}_{{jk}}-{{VC}}_{{ik}}\right)/{{VC}}_{{ik}}}\right|,$$
(2)

where CS denotes the coefficient of sensitivity, ESVi and ESVj respectively signify the initial and adjusted ESVs, and VCik and VCjk respectively represent the initial and adjusted (±50%) ESV coefficients.

Spatial autocorrelation analysis

Spatial autocorrelation (Global Moran’s I) was utilized to delineate the spatial distributions of ESVs in the pilot area, for which the Spatial Statistics tools and Getis-Ord Gi* index analyses in ArcGIS 10.2 software were employed. This process revealed the spatial clustering features of the ESVs. Moran’s I ranges from −1 to 1: values greater than 0 indicate positive spatial autocorrelation, with stronger spatial clustering as the value increases; values less than 0 indicate negative spatial autocorrelation, reflecting greater spatial dispersion as the value decreases; and a value of 0 signifies spatial randomness. By comparing statistical Z-scores and P values, the Getis-Ord Gi* tool was used to generate cluster maps to spatially locate high- or low-value feature clusters. The formula for calculating spatial autocorrelation is expressed as follows (Yang et al., 2021):

$$I=\frac{{\rm{n}}\mathop{\sum }\nolimits_{{\rm{i}}=1}^{{\rm{n}}}\mathop{\sum }\nolimits_{{\rm{j}}=1}^{{\rm{n}}}{{\rm{W}}}_{{\rm{ij}}}\left({{\rm{x}}}_{{\rm{i}}}-{{\rm{x}}}_{{\rm{a}}}\right)\left({{\rm{x}}}_{{\rm{j}}}-{{\rm{x}}}_{{\rm{a}}}\right)}{\left(\mathop{\sum }\nolimits_{{\rm{i}}=1}^{{\rm{n}}}\mathop{\sum }\nolimits_{{\rm{j}}=1}^{{\rm{n}}}{{\rm{W}}}_{{\rm{ij}}}\right)\mathop{\sum }\nolimits_{{\rm{i}}=1}^{{\rm{n}}}{\left({{\rm{x}}}_{{\rm{i}}}-{{\rm{x}}}_{{\rm{a}}}\right)}^{2}},$$
(3)

where I is the global spatial autocorrelation Moran’s index; xi and xj respectively represent the ESV observations of evaluation units i and j; xa is the mean value of the observations; n is the total number of ESV evaluation units on the research scale; and Wij is the spatial weight matrix between evaluation units i and j.

Partial least squares regression model

Given that this study centers on the influencing factors of ESV evolution in the study area across five time points over 30 years, the data volume of the driving factors also corresponds to five time points. The limited data capacity renders traditional linear regression unsuitable for this analysis. Therefore, partial least squares regression (PLSR) was employed to investigate the socioeconomic drivers of EVS evolution. As a multivariate statistical regression model, PLSR effectively addresses multicollinearity issues between predictor variables and accommodates smaller sample sizes (Qin et al., 2003; Luo et al., 2002).

Consistent with previous research (Li et al., 2019), SIMCA-P software was used to conduct PLSR analysis. The dependent variable of this study is ESV, and the independent variables are influencing factors such as the economy, production, population, and society. Matrices for the dependent and independent variables were obtained via standardization, respectively denoted as S0 = Y and F0 = X. The first principal component \({t}_{1}\) was then extracted. The equations of matrices \({S}_{0}\) and \({F}_{0}\) for the first principal component \({t}_{1}\) are as follows:

$${F}_{0}={t}_{1}{n}_{1}^{T}+{F}_{1},$$
(4)
$${S}_{0}={t}_{1}{r}_{1}+{S}_{1},$$
(5)

where r1 and n1 are vector values, respectively denoted as\({r}_{1}={S}_{0}^{T}{t}_{1}/||{t}_{1}|{|}^{2}\) and \({n}_{1}={F}_{0}^{T}{t}_{1}/||{t}_{1}|{|}^{2}\), and \({S}_{1}\) and \({F}_{1}\) are the residual matrices of PLSR. The calculation and extraction processes for the second, third, fourth, and n th principal components t2, t3, t4, and tm are the same as those for t1.

The verification of the accuracy of the ESV evolution driving mechanism is required to assess the rationality of the regression. The PRESS variable in SIMCA-P software was applied to test the fitting effect. A cross-validity value above 0.097 signifies a sufficiently robust PLSR model and the reasonable extraction of the principal components. The explanatory power of the regression model, also known as the goodness of fit, is represented by R2Y. An R2Y value exceeding 0.50 indicates strong predictive ability and a Q2 value over 0.097 denotes robustness. The variable importance of projection (VIP) value reflects the extent to which each independent variable contributes to explaining the dependent variable. Variables with VIP values above 1 have relatively significant explanatory abilities; values between 0.5 and 1 indicate that the variable explains the dependent variable sufficiently; and values below 0.5 represent unimportance for the dependent variable (Zhang et al., 2016; Li et al., 2019).

Model of the degree of trade-off and synergy between ESs

Recent studies (Zhang et al., 2020a, 2020b; Li et al., 2023) have affirmed the effectiveness of the ES trade-off degree (ESTD) in elucidating the relationships between ESVs. Building on the method proposed by Gong et al. (2019), where ESTDij = ESTDji, the calculated ESTD remains consistent regardless of the order of the ESs considered, ensuring an accurate reflection of conditions in the study area. The calculation formula is as follows:

$$ESC{I}_{i}=\frac{E{S}_{i{\rm{a}}}-E{S}_{ib}}{E{S}_{ib}},$$
(6)
$$EST{D}_{ij}=\frac{1}{2}\left(\frac{ESC{I}_{i}}{ESC{I}_{j}}+\frac{ESC{I}_{j}}{ESC{I}_{i}}\right),$$
(7)

where ESia and ESib respectively denote the value of the ith ESV at time nodes a (final state) and b (initial state); ESCIi represents the ES change index of the ith ESV; and ESTDij signifies the ESTD between ESVs i and j. If ESTDij > 0 (or ESTDij < 0), a synergy (or trade-off) exists between the two ESVs, and the absolute value indicates the level.

Results and analysis

Spatiotemporal changes in ESV in the pilot area

Evaluation of ESV in the pilot area

The ESV of the pilot area exhibited relative stability over the 30-year study period, with values of CNY 170.63 × 107, 169.65 × 107, 171.89 × 107, 171.95 × 107, and 172.11 × 107 in 1990, 2000, 2010, 2015, and 2020, respectively. ESV increased throughout this period, apart from a minor 0.57% decrease from 1990 to 2000. This suggests that the ecological environmental quality of QNP, formed by integrating multiple protected areas, has remained stable. Table 2 reports the detailed changes in ESV across different land-use types. Throughout 1990–2020, forest land contributed the most substantially to total ESV, followed by cultivated land and grassland. From 1990 to 2000, the contributions of forest land and unused land to ESV declined, while the contributions of other land-use types remained unchanged or increased. Notably, grassland ESV rose 48.65% due to the expansion of the area of this land type. Grassland ESV then dropped from 2000 to 2010 and from 2015 to 2020 by 42.86% and 24.23%, respectively.

Table 2 The ESV changes of various land-use types in the pilot area from 1990 to 2020.
Full size table

During 2010–2015, the ESV changes of all land-use types were minimal or even unchanged, likely due to the brevity of the observation period. The year 2015 was selected as a time node, as this was the year in which China’s national park pilot system was constructed. Because the study area was one of the first 10 pilot areas, the data from 2015 arguably hold marked significance for follow-up research on the ESVs of national parks. Over the 30-year period, the water area and unused land—despite their relatively small area proportions—showed the greatest changes in ESV, likely reflecting considerable anthropogenic influence. Increases in forest land and water area suggest the need to improve the ecological quality within the pilot area.

As shown in Fig. 2, regulating services constituted the majority of ESVs from 1990 to 2020, followed by supporting, provisioning, and cultural services. Specifically, regulating service contributions originated predominantly from gas, climate, and hydrological regulation, respectively accounting for 15.11%, 14.35%, and 14.57% of the total ESV. Throughout the observation period, key ESV contributions from forest land included biodiversity conservation, gas regulation, and hydrological regulation, accounting for nearly 45.95% of the total ESV provided by forest land each year. Soil conservation services predominated for grassland, while farmland was characterized by soil conservation and waste treatment. Hydrological regulation and waste treatment were the primary services for water areas, accounting for nearly 74.14% of the total ESV. ESV from unused land remained negligible, indicating a strong correlation between land-use changes in the national park and shifting ESs over 30 years (Fu and Zhang, 2014).

Fig. 2
figure 2

Changes in individual ESVs in the pilot area from 1990 to 2020.

Full size image

Sensitivity analysis of ESV coefficients

The CS calculation results are reported in Table 3. After adjusting the VC (±50%), the CS values for all land-use types (except forest land) were found to be less than 1 across the entire observation period. Forest land exhibited the highest CS, ranging from 0.9599 to 4.9599, indicating that a 10% increase in VC would lead to a 9.599–49.599% increase in the total ESV. The CS for cultivated land ranked second, ranging from 0.0221 to 0.0226, suggesting that a 10% increase in VC would cause a 0.221–0.226% increase in the total ESV. The CS was the lowest for unused land, approaching 0. Therefore, the VCs calculated after revising the equivalent factor table are reliable and reflect the actual conditions in the pilot area.

Table 3 The changes of ESV and CS after adjusting the VC.
Full size table

Spatial autocorrelation analysis of ESV

To clarify the spatial distribution characteristics of the values of various ESs in the national park pilot area, the study area was divided into 308 grids of 1 km × 1 km. This grid size was based on the total size of the study area, visualization effectiveness, and multiple comparisons of grid accuracy. A global spatial autocorrelation analysis was then conducted using the Spatial Statistics tool in ArcGIS 10.2. The results reported in Table 4 show that the Moran’s I values for various ESVs were greater than 0 and passed the significance test at the P < 0.01 level, indicating clustering patterns with strong positive correlations. Specifically, high-high and low-low value clusters were observed for the distribution of various ESs.

Table 4 The global autocorrelation analysis of ESV.
Full size table

As shown in Fig. 3, the Getis-Ord Gi* clusters indicate that during 1990–2020, high-value ESV clusters in the pilot area were mainly distributed in Suzhuang Town and Qixi Town, while low-value clusters were mainly distributed at the edges of the area—particularly at the border between Suzhuang Town and Changhong Town. The former exhibited a clustered spatial distribution and the latter had a linear distribution.

Fig. 3
figure 3

The hotspot distributions of ESVs in the pilot area.

Full size image

Overall, the ESV exhibited a sporadic distribution trend with small fluctuations. The clustering of similar services did not significantly change in different periods, and significant spatial heterogeneity was observed. The clusters of different services in the same year were more distinct. This suggests that the ESV spatial distributions are closely correlated with land use; high-value clusters were identified in places with the most abundant forest resources, mostly in core protected spaces, while low-value clusters were closer to communities. These phenomena further reflect the impact of anthropogenic activities on the ecosystems of the pilot area.

Socioeconomic influencing factors of ESV changes in the pilot area

The past 30 years have been an important period of rapid social and economic development in China. Zhejiang Province, where the pilot area is located, has been at the forefront of China’s reform and opening-up policies, instigating various changes in community livelihoods, labor force quantity, and industrial structures within the pilot area. Throughout this process, local governments have frequently adjusted land-use types to align with developmental needs, resulting in fluctuations in ESVs over both time and space alongside regional economic growth and societal changes. It is crucial to clarify the impacts of socioeconomic factors on the evolution of ESVs in the pilot area. Based on previous research (Yao et al., 2009), the socioeconomic drivers of ESV changes were categorized into four types: economic, production, demographic, and social. As shown in Table 5, the influencing factors were selected based on data availability and quantifiability, and included four primary indicators and 19 secondary indicators. The data were sourced from the Kaihua Statistical Yearbook.

Table 5 The ESV influencing factors.
Full size table

Table 6 shows that the Q2 values for ecosystem provisioning, regulating, supporting, cultural, and total services in the pilot area were respectively 0.14, 0.66, 0.12, 0.61, and 0.51, all of which are above 0.097. The model explanatory powers (R2Y) for regulating, cultural, and total services were greater than 50%, namely 0.78, 0.75, and 0.68, respectively, indicating that these socioeconomic factors respectively explain the evolution of regulating, cultural, and total ESVs.

Table 6 The PLSR results for ESVs and their influencing factors in the pilot area.
Full size table

It can be seen from Table 4 that the Q2 values of ecosystem provisioning services, regulating services, supporting services, cultural services, and total services in the pilot area were respectively 0.14, 0.66, 0.12, 0.61, and 0.51, all of which are greater than 0.097. The model interpretation ability (R2Y) values of regulating services, cultural services, and total services were respectively 0.78, 0.75, and 0.68, all of which are greater than 50%. These findings indicate that these socioeconomic factors can explain the evolution of the ESVs of regulating, cultural, and total services in the pilot area. After adjusting the PLSR of regulating services, cultural services, total services, and various influencing factors, the factors that significantly affect the value changes of these services (VIP > 1) were found to be EF-3, EF-2, EF-5, and EF-7. Factors PF-1, DF-2, SF-2, and PF-3 had certain impacts on regulating and cultural services (0.5 < VIP < 1), and their impact performance was consistent. Factors DF-3, PF-1, DF-2, SF-2, and PF-3 had certain impacts on the total service value (0.5 < VIP < 1). Factors PF-4, DF-4, DF-3, PF-1, DF-2, and SF-2 were negatively correlated with ESV in the pilot area. Finally, PF-2 was found to have no effect (VIP = 0.07, <0.5).

Overall, economic factors such as the output value of the primary industry, the output value of the secondary industry, the output value of agriculture, forestry, animal husbandry, and fishery, and the total fiscal revenue were found to have the greatest impacts on the ESV of the pilot area, with positive correlations. According to the Kaihua Municipal Bureau of Statistics, from 1990 to 2020, the value of secondary industries increased from 141.24 to 528.436 billion RMB; the value of primary industries rose from 273.61 to 141.53 billion RMB; and the value of agriculture, forestry, animal husbandry, and fishery grew from 273.61 to 143.28 billion RMB. Changes in the industrial structure have transformed land-use patterns and driven fluctuations in ESVs, especially for gas regulation, climate regulation, hydrological regulation, waste treatment, and landscape esthetics. All population factors (DF-1-4), as well as factors such as the crop sowing area and Engel’s coefficient of rural residents, were found to have negative correlations with the ESV of the pilot area, while tea production did not have an impact.

Trade-off/synergy relationships of ESV in the pilot area

Figure 4 displays the synergistic relationships among various ESs in the pilot area, as calculated by the ES trade-off model. The synergies among ESs were found to be more balanced and prevalent than trade-offs. Among the 144 pairs of ESs analyzed across four time periods (1990–2000, 2000–2010, 2010–2015, and 2015–2020), 96 pairs exhibited positive relationships, while 48 pairs had negative relationships, resulting in a synergy rate of 66.67%.

Fig. 4
figure 4

The trade-offs/synergies of ESs in the pilot area.

Full size image

Trade-offs were found to primarily exist between food production services and other services. Food production services demonstrated trade-off relationships with all other services in 2000–2010 and 2010–2015, and the degree of these trade-offs weakened in 2015–2020, during which the degrees of trade-offs with hydrological regulation, waste treatment, and landscape esthetics services were −1.02, −1.24 and −1.63, respectively.

Regarding synergies over time, the highest synergy occurred in 1990–2000 between food production and waste treatment services at a rate of 8.32. From 2000 to 2010, the highest synergy was that between raw material production and soil conservation services at a rate of 1.30. From 2010 to 2015, the highest synergy was that between waste treatment and landscape esthetics services at a rate of 2.70. From 2015 to 2018, the highest synergy was that between food production and raw material production services at a rate of 3.99. This analysis demonstrates fluctuations in the relationships among ESs in the pilot area over the study period (Wang et al., 2017).

The results show that the pilot area has preliminarily become a “highland” of ESs, has the conditions for the collaborative optimization of a variety of ESs, and can simultaneously achieve economic, ecological, and cultural benefits, as also confirmed by Duan et al. (2021). Academic circles generally believe that the formation of ES synergy is determined by regional vegetation and climate conditions. While the increase of vegetation coverage results in the improvement of carbon storage, water conservation, and soil conservation service values, it also leads to the decline of water production and soil moisture, which will lead to a trade-off conflict between ESs (Feng et al., 2020; Yang et al., 2018). In recent years, this phenomenon has often occurred in China’s arid and semi-arid regions. To this end, in 2021, the Chinese government issued the Guiding Opinions of the General Office of the State Council on Scientific Greening, proposing to adhere to the working principles of “determining by water, measuring water, green when appropriate, and waste when appropriate.” The study area is located in eastern China, which is characterized by rich climate resources and forest coverage of nearly 90%, inherently making it an efficient ES. However, the study area is also faced with a fierce contradiction between humans and land, the conflict between ecological protection and community development, and other issues. Therefore, the identification of the complex trade-offs and synergies between the ESs in the national park and the scientific guidance of national park management decisions remain important issues.

Discussion

The uncertainty of the ESV evaluation results in national parks

ES assessment methods generally fall into one of two categories—functional assessments or value assessments—which involve various technical methods such as ecological process modeling, statistical data conversion, expert scoring, and stated preference methods (Wang et al., 2019). However, none of the previously proposed methods have gained widespread academic and public acceptance, primarily due to their immaturity. The establishment of appropriate per-unit values aligned with the characteristics of China’s national parks is the key to scientifically assessing their ESs. While “The Value of the World’s Ecosystem Services and Natural Capital” by Costanza et al. (1997) published in Nature clarified ESV accounting and estimation methods, it was mainly based on the global scale and is not suitable for China’s current ecosystem conditions.

Aligned with the research of Costanza et al. (2014), Chinese scholars Xie et al. (2003; 2008; 2015) built and improved China’s ESV equivalents over the course of several study projects. Their techniques have been widely applied in research on various types of ecosystems in China (Xie et al., 2003; 2008; 2015). However, the value equivalent factors they proposed reflect the average level of China’s ES functions; whether these value equivalent factors are suitable for national parks, a special type of nature reserve, is debatable. With the smallest total area among the 10 national park system pilot areas in China and a relatively large population density, the QNP pilot area was further investigated using existing revision methods (Song et al., 2021; Qin and Fu, 2020; Xie et al., 2015), and grain yield was ultimately selected as the value equivalent factor revision method. A further sensitivity analysis verified the reliability of the VCs. A sensitivity analysis of ESV coefficients was also conducted, which confirmed that the VCs obtained from revising the equivalent factor table were reliable and in line with the actual situation of the pilot area, thus minimizing the uncertainty of the research to the greatest possible extent. Ran et al. (2023) used the InVEST model to calculate the ES functions in the pilot area from 1990 to 2020, and obtained ecological service change trends consistent with the findings of the present study. Although their research involved functional assessments, the identical time nodes verify, to some extent, the scientific validity of the results presented in this work.

In addition, according to land-use data, the construction land in the pilot area comprises 89.82 hm2, accounting for only 0.36% of the total area. As the pilot area is part of China’s national park system, characterized by the highest possible protection level, its ecosystems are predominantly natural. Therefore, construction land was not included in the ESV calculation. However, the proposed method is not necessarily inapplicable to other types of nature reserves, such as scenic areas, which contain more cultural (anthropogenic) elements within their nature reserves. Chinese scenic attractions often contain historical temples, academies, and rural dwellings, and their ESV calculations must include the impact of construction land on ESs and reflect the cultural attributes of construction land. Addressing this issue with key technologies is an important research direction for ES assessments in nature reserves.

Factors influencing the evolution of ESV in national parks under specific socioeconomic conditions

Studies have found that the primary factors influencing the rankings of ecosystem regulating services, cultural services, and total services in China’s national park pilot areas are the values of secondary industries, primary industries, and agriculture, forestry, animal husbandry, and fishery. These three economic factors are positively correlated with ESs, indicating that the development of agriculture and manufacturing will affect the ESVs in national parks by influencing land-use changes. In addition, the increase in total fiscal revenue will also change the infrastructure within the region, leading to changes in the ESVs in the pilot area. All population factors (DF-1–4), along with the cultivated land area and rural Engel’s coefficient factors, were found to exhibit negative correlations, suggesting that population growth and associated agricultural expansion have indirectly impacted urban and rural construction and agricultural development, as well as ecological regulation and cultural services. Rural per capita living costs in the pilot area have risen from the UN poverty line to a more affluent standard, as indicated by the decrease in the Engel coefficient from 63.4% in 1990 to 29.1% in 2020 (Chen and Chen, 2018). However, the decline in total ESV warrants attention from local governments and society. There is an evident need for further research on the interactions between national park ESs and human well-being (Qiu et al., 2019). Tea yields, for instance, do not significantly impact ESs due to local land-use practices and the historical Grain for Green policy (Huang et al., 2019).

In the pilot area, residents primarily rely on agriculture, cultivating crops such as rice, rape, corn, and Camellia oleifera, as well as other economic forests. Income from agriculture and migrant work accounts for more than 80% of total income. The terrain is not suitable for mechanized operation, and the degree of agricultural industrialization is very low. Additionally, the foundation for a modern service industry is weak. Although there are signs of tourism, residential services, and the commercial catering industry emerging as key sectors, industrial agglomeration has yet to take shape. Based on the research results, it is recommended that the national park management department gain a deeper understanding of the community’s natural, economic, and social structures. The department should also leverage the community’s strengths, actively guide industrial development, and shift from traditional, fragmented land-use and agricultural practices to more sustainable models. This involves developing ecological agriculture, forestry, and rural tourism while fostering industrial diversification and upgrading. Such efforts would help address the contradiction between community development and ecological protection.

Due to data availability, only socioeconomic drivers of ESV in the pilot area were examined, without considering natural or policy factors. ESs are often affected by socioeconomics through local efforts to increase yields at the cost of reduced regulation and supporting services (Kareiva et al., 2007). Recognizing and appropriately managing socioeconomic drivers are thus essential for maximizing regional ESs. Raudsepp-Hearne et al. (2010) also found that regional social factors significantly influence ES relationships. As the only national park pilot in the economically developed Yangtze River Delta region, balancing ecological conservation and socioeconomic development is a key issue for QNP. Li et al. (2022) studied nonlinear drivers of ES evolution in the Qiantang River Basin, concluding that socioeconomic changes have altered watershed ecosystem trajectories over time. Xiong et al. (2019) found that development models in eight Guangdong Province villages significantly impact local ES provision, maintenance, and utilization. These studies further validate the socioeconomic findings of the present research.

Trade-off/synergy relationships of ESs in national park system pilot areas

ES studies have garnered significant attention, particularly as they involve research objects within coupled human-land systems. Early investigations into the relationships between changes in ESs and land use/land cover (e.g., farmland, forests) revealed complex trade-off and synergy relationships, gradually attracting more attention (Barbier et al., 2008; Pretty et al., 2006; Naidoo and Ricketts, 2006). Research on national park ESs involves core topics relevant to human-land relationships and can offer practical solutions for ecosystem management and sustainable development (Kates, 2011; Wu, 2012; Wu et al., 2014). This study revealed the relationships between different ESs in national parks to be mainly synergistic (with a 66.67% synergy rate), while trade-offs were found to primarily exist between food production services and other services. This indicates some conflict between provisioning services and other ESs, aligning with the research conclusions on river basin ecosystems by Yang et al. (2021). McNally et al. (2011) also analyzed the trade-offs/synergies between provisioning services and other services in the mangrove ecosystem of Saadani National Park. They found that although the local community’s fish and shrimp economy would initially suffer from banning mangrove cutting for fuel wood, the community would gain greater long-term benefits from strengthening mangrove protection and improving mangrove ecological services. Specifically, they found that a 10% increase in mangrove cover would triple a community’s income from fish and shrimp. This further verifies the competitive relationship between cultivated land as production land and forest and grassland as ecological land.

Research into ES trade-offs and synergies holds important significance, as it can provide support for regional spatial planning initiatives and formulating ecological compensation and ecosystem protection/restoration policies. However, the specific mechanisms and processes for applying findings from research on the trade-offs/synergy of national park ESs into actual decision-making processes remain unclear. Wang et al. (2019) found that ES trade-offs mainly include four types: supply-demand trade-offs, supply mathematics trade-offs, stakeholder interest trade-offs, and spatial-temporal trade-offs. Previous studies on ES trade-off/synergy relationships have focused on the supply side. Future research should incorporate ES demand factors and socioeconomic and cognitive preference factors, especially the interests and preferences of communities within and surrounding national parks. The incorporation of these factors would allow for a more holistic understanding of ES trade-offs and synergies to better inform policy decisions.

Research limitations and prospects

In China, considering the typicality of ecological resources and the feasibility of management and other goals, national parks are generally set up in the vast and sparsely populated western region. This avoids the contradiction between people and land to the greatest possible extent, which is conducive to national park management. As a result, to date, most national park research sites have been located in large-scale western regions, such as Sanjiangyuan National Park (190,700 km2) and Giant Panda National Park (22,000 km2). The selection of indicators for ESs also focuses on specific ecological indicators such as carbon sequestration and water conservation. Among the first 10 national park pilot areas in China, QNP is the only one located in the economically developed eastern region, for which there have been few achievements in the application of related ES theory research to management decisions. Therefore, this research provides a scalable and replicable research framework for the study of ESs in national parks in economically developed, densely populated areas with complex land resource ownership. The results of this study provide an important theoretical reference for ecosystem management in national parks in eastern China to alleviate complex human-land contradictions. Moreover, they support the comparability, generalization, and replication of research methods and findings among multiple national parks.

The construction of national parks in China began relatively recently, less than a decade ago. At present, the research on the ESs of national parks mainly focuses on the accounting and functional evaluation of ESVs, and the influencing factors of the changes of ESVs have not yet been explored. Moreover, cultural service value evaluation data are mostly obtained by social methods, such as questionnaire surveys. Although this research method can take into account a variety of stakeholders, the time scale of the evaluation can only be limited to the year from which the data are obtained, causing the research on national park ESV to be mainly based on static evaluations. Additionally, given the significant variation in the natural resource attributes across national parks, selecting a scientifically sound and reasonable evaluation method, as well as increasing the comparability between the evaluation results, remains a key challenge in current ESV assessments. Moving forward, it will be imperative to embrace interdisciplinary approaches that consider ES responses to socioeconomic factors across multiple levels. This will entail linking ES dynamics across different spatiotemporal scales to ecosystem management processes at various administrative levels, balancing multiple ESs, and reconciling conflicts among different stakeholders (Raudsepp-Hearne et al., 2010; Wang et al., 2023; Yadav et al., 2024).

Furthermore, the natural ecosystems and heritage of national parks are more nationally representative than those of general nature reserves, featuring larger protected areas and more intact ecosystems. It is crucial to delve into the nonlinear relationships between national park ESs and various influencing factors. Previous research has revealed a failure to adequately account for the multiple drivers, characteristics, and stakeholders in ESs, which has raised questions about the objectivity of valuations and the practical utility of assessments. The inability to accurately measure ecological benefits has also resulted in a lack of social recognition, leading to the failure of management regimes and methods for many protected areas (Ouyang and Zheng, 2009; Du et al., 2019a, 2019b; Bennett et al., 2009). Therefore, building on scientifically assessed values of national park ESs, exploring how various natural and anthropogenic factors influence ES trade-offs and synergies, and elucidating the nonlinear responses of service trade-offs and synergies to multiple drivers will be critical for optimizing national park ecosystem management. These considerations should be prioritized in future research. In addition, key scientific challenges remain unresolved, including how to integrate interdisciplinary advantages, promote the practical application of ES research results in national park management decisions, and develop optimized, diversified ecological protection and compensation modes based on ESs. Another critical issue is how to assess the functions and values of national park ecosystems based on trade-offs and synergies. Addressing these questions is essential for coordinating multiple objectives and maximizing human well-being (Su and Liu, 2023).

Conclusions and policy recommendations

Research conclusions

A long-term, multi-scale, and comprehensive dynamic ESV study was conducted on the only national park system pilot area in eastern China. An evaluation index system for which data are easy to obtain and that can effectively measure the social and economic status of national parks was also developed, thus laying the foundation for the horizontal comparison of the evaluation results of multiple national parks. The key conclusions can be summarized as follows. (1) The ESVs of the QNP pilot area in 1990, 2000, 2010, 2015, and 2020 were respectively CNY 170.63 × 107, 169.65 × 107, 171.89 × 107, 171.95 × 107, and CNY 172.11 × 107, overall reflecting small changes and a discernible increasing trend. Forest land was the main contributor to ESV, followed by cultivated land and grassland. The ESV with the highest contribution rate was regulating services, mainly including gas, climate, and hydrological regulation. (2) Strong positive spatial autocorrelation was observed in the ESVs of the study area, accompanying a certain aggregation pattern. Areas with high aggregation were concentrated in Suzhuang Town and Qixi Town, while areas with low aggregation were situated near neighboring communities. (3) The added value of secondary industries, the added value of primary industries, and the added value of agriculture, forestry, animal husbandry, and fishery were the most important factors affecting ESV changes in the pilot area. Factors such as the crop sowing area, birth rate, newborn population, grain yield, rural population, and Engel’s coefficient of rural residents displayed negative correlations with ESV, while tea yield exhibited no effect. (4) Synergies between ESs in the pilot area appeared to be more stable than trade-offs, and dominated at a synergy rate of 66.67%. Trade-offs primarily existed between food production services and other services in both the 2000–2010 and 2010–2015 periods. However, the degree of this trade-off weakened between 2015 and 2020. The results of this study provide a comparative, generalizable, and replicable research framework for national parks established in economically developed, densely populated areas with complex land ownership. They also provide a reference for the study of ESs in scenic spots, forest parks, wetland parks, and other types of nature reserves.

Policy recommendations

By incorporating ESV as a key reference indicator in the formulation of national park protection and development policies–and by reflecting on the experience and limitations of existing policies–ESV can play a supporting role in national park management decisions. Based on the research conclusions and the actual conditions of the study area, the following policy recommendations are proposed.

First, the theoretical framework of national park ESV accounting should be improved. In response to key scientific challenges related to China’s ecological assets—such as ambiguous definitions, an unclear background, underdeveloped methodologies, and the absence of top-level design—an evaluation index system suitable for national parks in different regions was constructed in this study. A theoretical and methodological system was put forward, which operationalizes the statistical evaluation of ESV and improves the comparability between research results. By developing corresponding tools and establishing technical standards and specifications, the practical applicability of the results is also expanded, ultimately providing a technical basis for the establishment of a national green ESV accounting system.

Second, mechanisms for realizing the ESV of national parks should be explored. According to the specific conditions of the study area, this study analyzed the property rights structures and the evolution of ecological resource assets owned by the government and collectives. It explored approaches to separate ownership, contracting rights, and management rights, and proposed innovations in the paid-use system of natural resources. Based on protection costs, opportunity costs, and ESVs, a differentiated ecological compensation standard system was formulated and investigated. Moreover, pathways for government, market, and social organization compensation were examined, and a diversified ecological compensation model based on ESV and stakeholders was ultimately proposed.

Finally, the connection between ESs in national parks and human well-being should be strengthened. Participatory community management should be promoted, ecological management and protection posts should be set up according to ecological protection needs, and priority should be given to the original residents. Environment-friendly green industries should be developed, indigenous residents should be supported to participate in franchising activities, and mechanisms for natural resource owners to participate in franchising income distribution should be explored. Moreover, volunteer services in national parks should be strengthened, the social donation system should be improved, and multiple stakeholders should be encouraged to participate in national park protection and development.