A social–ecological approach for identifying and mapping ecosystem service trade-offs and conservation priorities in peri-urban areas

Abstract

Considering both ecological and social dimensions in the assessment of ecosystem services (ESs) can facilitate acceptable and inclusive management strategies, especially in peri-urban areas characterized by intricate human–ecosystem interactions. A limited body of research, however, has mapped the plural values of ESs and their different types of trade-offs in such areas. This research aimed to execute an interdisciplinary analysis of the biophysical and social values of ESs in peri-urban Shanghai, China, through a social–ecological approach that integrates spatial biophysical assessment with participatory mapping. Trade-off analysis in both ES types and ES valuations were then conducted, and multicriteria decision-making was applied for conservation. Our results reveal that trade-off intensities were lower within the social values compared to the biophysical values. Within both value dimensions, relatively stronger trade-offs were found between food production and other ESs. Areas with both high biophysical and social values were infrequently observed across ESs. Based on the characteristics of diverse values, our study identified priority conservation areas and provided management implications. We argue that adopting the integrated social–ecological perspective in sustainable environmental management contributes to the realization of harmonious coexistence between people and nature in peri-urban areas.

Supplementary Information

The online version contains supplementary material available at 10.1007/s13280-024-02031-6.

Introduction

Ecosystem services (ESs) refer to the direct and indirect benefits people derive from ecosystems (Millennium Ecosystem Assessment 2005; Costanza et al. 2017). Peri-urban areas, characterized as complex and dynamic social–ecological systems between urban and rural areas with diverse land use (Spyra et al. 2020; Filyushkina et al. 2022; Marshall et al. 2024), provide a variety of ESs for human well-being. These include peri-urban agricultural production (Lee et al. 2015), air and water purification (Wangai et al. 2019), and recreational opportunities (Li 2020). Yet, rapid urbanization has exerted profound impacts on peri-urban ecosystems and their capacity to provide ESs, particularly in the Global South (Peng et al. 2017; Hutchings et al. 2022). Currently, peri-urban areas are undergoing challenges of food insecurity (Follmann et al. 2021), air and water pollution (Shih 2017), soil erosion and degradation (Seifollahi-Aghmiuni et al. 2022), and biodiversity loss (Chaves et al. 2022), thereby pushing these areas toward unsustainable development trajectories. Moreover, the intricate ecological and social dynamics and the diverse human–ecosystem interactions in peri-urban areas amplify these sustainability challenges (Žlender 2021; Seifollahi-Aghmiuni et al. 2022). Consequently, there is an imperative necessity for the comprehensive assessment of ESs in peri-urban areas and their integration into conservation and management (Spyra et al. 2020; Chen et al. 2023).

For the identification and mapping of ESs within peri-urban areas, previous studies have predominantly centered on biophysical values from the ecological dimension (Peng et al. 2017; Sylla et al. 2020). Biophysical values refer to the ability of ecosystems to provide and maintain ESs (Fontaine et al. 2014), which are informed by scientific knowledge (Raymond et al. 2023b). These values focus on the biophysical capacity of ecosystems to provide ESs and are distinguished from the actual use as well as preferences of ESs (Schröter et al. 2014; Crouzat et al. 2022). Studies within this dimension have primarily used biophysical data and ecological modeling (Peng et al. 2017; Harrison et al. 2018). In contrast, a limited number of studies have focused on the mapping of social values of similar ESs within peri-urban areas (Escobedo et al. 2020; Filyushkina et al. 2022). Here, social values represent perceptions held by stakeholders (Sherrouse et al. 2011), which can be assessed through subjective data and participatory approaches (De Vreese et al. 2016). It is worth noting that the social values of ESs are conceptually different from, and not limited to, cultural ESs, as people can perceive or ascribe values to ESs beyond the cultural ones. Social valuation thereby facilitates connections with all types of ESs (Felipe-Lucia et al. 2015; Scholte et al. 2015; Ruiz-Frau et al. 2018). While the rise of participatory mapping and the Public Participation Geographic Information System (PPGIS) has advanced the spatially explicit social valuation (Brown 2013; Fagerholm et al. 2019), studies employing these approaches for social valuation within peri-urban areas of the Global South, such as China, remain relatively infrequent (Escobedo et al. 2020; Xu et al. 2020). Relying only on biophysical values while disregarding social values in ES analysis may result in the lack of social acceptance and inclusiveness of proposed conservation and management strategies (Raymond et al. 2023b), which can be detrimental to environmental justice (Loos et al. 2023).

Considering peri-urban areas as coupled social–ecological systems where seminatural ecosystems constantly interact with growing populations (Hutchings et al. 2022; Chen et al. 2023), it is necessary to carry out the integrated valuation of ESs from the social–ecological perspective (Quintas-Soriano et al. 2021; Cusens et al. 2023). Stemming from research on social–ecological systems (Herrero-Jáuregui et al. 2018), this perspective is rooted in value pluralism and adopts an interdisciplinary methodology for assessing ESs (Jacobs et al. 2018; Kronenberg and Andersson 2019). For instance, Martín-López et al. (2014) and Castro et al. (2014) previously introduced an ES valuation framework that explicitly considered both biophysical and social values, along with economic values influenced by the latter. More recently, building upon the ES framework, the Intergovernmental Science–Policy Platform on Biodiversity and Ecosystem Services (IPBES) has presented a more refined and inclusive concept of nature’s contributions to people (NCP) (Diaz et al. 2018). This concept embraces the social–ecological perspective, focusing on the role of stakeholders and the integration of indigenous and local knowledge in the valuation process (Diaz et al. 2018; McElwee et al. 2020). In its values typology, biophysical and social values are emphasized within the layer of “value indicators” (Pascual et al. 2023; Raymond et al. 2023a). Recent reviews by Martín-López et al. (2019) and Schutter and Hicks (2021) indicated that social valuation and pluralistic approaches have been gradually increasing in ES assessment. Social–ecological assessments can be applied across various ES categories, existing studies, however, still tend to quantify them separately. In other words, the biophysical approach has been predominantly adopted for provisioning, regulating, and supporting ESs, while the social approach has primarily focused on cultural ESs without considering other types (e.g., Lin et al. 2017; Chen et al. 2023; Cusens et al. 2023).

Furthermore, it is important to identify and map trade-offs in ESs using a social–ecological approach, for example, by incorporating both biophysical supply and social perceptions in the trade-off analysis (Chicago et al. 2022; Cusens et al. 2024). Typically, two types of trade-offs in ESs can be found within social–ecological systems (Lu et al. 2021). The first type involves trade-offs in ES types (i.e., trade-offs within the same biophysical or social value dimensions), which arise when an ES increases while another decreases (Rodriguez et al. 2006). The opposite status is synergies in ES types where both ESs increase or decrease together. The second type concerns trade-offs between ES valuations (i.e., trade-offs between different value dimensions), where a particular ES has a higher biophysical value but a lower social value, or vice versa (Martín-López et al. 2014; Lu et al. 2021). In terms of mapping, scholars have worked on identifying both spatial interactions among different ES types (Sylla et al. 2020; Zhao and Pan 2022) and areas with varying ES values (hot and cold spots) (Bagstad et al. 2016; Schwartz et al. 2022). It is also crucial to incorporate these diverse trade-offs into the conservation strategies of peri-urban areas (Cord et al. 2017; Spyra et al. 2020), by using integrated approaches such as (spatial) multicriteria decision analysis (Harrison et al. 2018; Lourdes et al. 2022). Recently, the transition of ES trade-off studies to the social–ecological perspective has been recognized as an emerging trend (Chicago et al. 2022). A systematic review by Aryal et al. (2022), however, revealed that most trade-off analyses still primarily rely on biophysical constraints without considering the beneficiaries, and multicriterion analysis of these ecological and social trade-offs for informing ES conservation strategies has been largely overlooked.

In summary, relatively few peri-urban studies have employed participatory approaches for social valuation for various ES categories and combined them with biophysical valuation. Fewer have applied these plural valuations to ES trade-offs and decision-making. The present study aimed to fill these gaps by developing an integrated social–ecological approach for a comprehensive and interdisciplinary analysis of four sets of ESs in Qingpu District, a peri-urban area of Shanghai, China. Based on this, we further adopted trade-off analysis and multicriteria decision-making to guide ES conservation and management. Specifically, our study aims to (1) quantify the biophysical and social values of ESs in a spatially explicit manner, (2) identify trade-offs both within and between biophysical and social values, and (3) determine priority conservation areas for ESs based on these values.

Materials and methods

Study Area

Qingpu District is situated in western Shanghai, China (120° 53′–121° 17′ E, 30° 59′–31° 16′ N, Fig. 1). Located in the Yangtze River Delta, it covers an area of 668.54 km² and comprises three subdistricts and eight towns. The climate is subtropical maritime monsoon, with an annual precipitation of 1493 mm and an average temperature of 17.8 °C. Land cover within the district predominantly consists of cropland (36.4%), built-up areas (27.1%), forest (15.9%), and water bodies (6.6%). Spatially, the western and eastern areas of Qingpu District differ in characteristics. The western areas of Qingpu District serve as water conservation areas in Shanghai, characterized by abundant water bodies, lush forests, and historical agricultural systems. Conversely, the eastern areas have been profoundly influenced by the proximity of central urban area of Shanghai, undergoing rapid urbanization over the past two decades (Xia et al. 2023). Consequently, the eastern areas feature a higher proportion of built-up areas. This urbanization process has given rise to environmental challenges within the district, including landscape fragmentation and ecological degradation (Wang et al. 2022). In light of these circumstances, it becomes imperative to embrace suitable spatial management approaches for conservation and to facilitate the harmonization of environmental and social development.

Fig. 1.

Study area

It is worth noting that different qualitative and quantitative methods have been used for the selection of peri-urban boundaries (Tian et al. 2017; Sahana et al. 2023). In this study, we still adopted the administrative boundary as the basis for delimiting peri-urban areas following the precedent of Tian et al. (2017) and Xia et al. (2024), i.e., the whole extent of Qingpu District was regarded as a peri-urban area. The district as a peri-urban area is surrounded by the central urban area of Shanghai to the east and the rural areas of neighboring provinces to the west. This peri-urban selection by the administrative boundary is in line with the current Shanghai Master Plan (2017–2035) (SMPG 2018) and allows our results to serve the existing planning system. It can also facilitate the calculation of ESs that require statistical data, as well as the understanding of geographic extent by respondents during the participatory mapping process.

Analytical framework

Figure 2 illustrates the analytical framework and approach flow for identifying and mapping ES trade-offs and conservation priorities. We first selected four sets of ESs and quantified both their biophysical values and social values in a spatially explicit way. Then, we identified and mapped the trade-offs in ES types (i.e., within biophysical or social value dimensions) as well as the trade-offs in ES valuations (i.e., across biophysical and social value dimensions) (Lu et al. 2021). Finally, we integrated the biophysical and social values of ESs through spatial multicriteria decision-making. By determining the highest ES conservation efficiency across various decision risk scenarios, we delineated conservation priorities for ESs in peri-urban areas.

Fig. 2.

Flowchart of the approach

ES typology

For the ES typology, we tried to cover all four ES categories listed in the Millennium Ecosystem Assessment (2005) and selected ESs from these. We then considered data availability for the study area from the biophysical perspective while also pretesting with local experts and residents, to ensure that the selected ESs could be quantified in our study area from both value dimensions. We finally identified six types of ESs from the biophysical dimension and nine types of ESs from the social dimension. To correspond the two value dimensions and to investigate the trade-offs between valuations, the ESs in both value dimensions were grouped into four sets of ESs (Table 1). These were: food production (FP), environmental capacities (EC), habitat maintenance (HM), and cultural ecosystem services (CES), which fall into the categories of provisioning, regulating, supporting, and cultural ESs, respectively.

Table 1.

Selection and quantification methods of ES

Biophysical valuation			Social valuation
ES type		Method	ES type		Method
Food production		NDVI allocation method (Zhao and Pan 2022)	Food production		PPGIS survey and SolVES model for all ESs (Sherrouse et al. 2011, 2014)
Environmental capacities	Air pollution removal	Per unit area method (Sun et al. 2022)	Environmental capacities
	Water purification	InVEST nutrient delivery ratio (NDR) model (Natural Capital Project 2023)
	Soil retention	RUSLE (Renard et al. 1997)
Habitat maintenance		InVEST habitat quality model (Natural Capital Project 2023)	Habitat maintenance
Cultural ecosystem services		Indicators impacting human perception and enjoyment of landscapes, including distance to water, presence of natural and cultural features, landscape heterogeneity, and relief heterogeneity (adapted from Crouzat et al. 2022)	Cultural ecosystem services	Outdoor recreation
				Aesthetic
				Cultural heritage
				Education
				Spiritual
				Social interaction

It is essential to highlight that the term “environmental capacities” was employed to denote a set of regulating ESs, following empirical studies of Fagerholm et al. (2019) and Plieninger et al. (2019) from the social value perspective. In this context, EC embodies “the environmental capacity to produce, preserve, clean, and renew air, soil, and/or water”. This term was chosen based on the pretest, which revealed that residents have difficulty in making specific distinctions between the regulation of various environmental elements, but rather considered them as the overall environment and landscapes. In correspondence to this social value dimension, EC within the biophysical value dimension was represented through the calculation of three regulating ESs: air pollution removal (APR), water purification (WP), and soil retention (SR).

It is also worth noting that general CES within the biophysical value dimension was calculated using four landscape biophysical indicators. These indicators were selected and adapted from the study of Crouzat et al. (2022), which are all existing indicators from CES modeling studies and are “expected to impact human perception and enjoyment of landscapes”. In correspondence to this, CES within the social value dimension was represented by calculating the average of six cultural ESs: outdoor recreation, aesthetic, cultural heritage, education, spiritual, and social interaction. For more information on these selected social values of CES please refer to Xia et al. (2024).

ES valuation and mapping

The biophysical valuation of ESs was conducted using GIS mapping and ecological modeling (Harrison et al. 2018, Table 1). The former relied on land cover mapping (for APR and CES) and incorporated additional information such as the Normalized Difference Vegetation Index (NDVI, for FP) and the digital elevation model (DEM, for CES). The latter involved ES modeling, notably the Integrated Valuation of Ecosystem Services and Tradeoffs (InVEST) model (for WP and HM) and the Revised Universal Soil Loss Equation (RUSLE, for SR). Details regarding the data sources and processing for the biophysical valuation are provided in Appendix S1. Subsequently, the three ESs encompassing APR, WP, and SR were normalized and averaged within each grid to derive EC, thereby corresponding to the social value dimension. All these ESs were further normalized to a uniform range of 0 to 1 to enable comparison.

The social valuation of ESs, on the other hand, was conducted using PPGIS and the Social Values for Ecosystem Services (SolVES) model (Table 1). First, face-to-face questionnaire surveys incorporating PPGIS were carried out in August 2022 to capture the local respondents’ perceived supply of the ESs. Each ES was represented by a specific question formulated in accordance with the PPGIS and ES research conducted by Brown (2013) and Fagerholm et al. (2019). Based on the question descriptions, respondents were asked to mark up to five points on a satellite image paper map, similar to the study of Plieninger et al. (2013; up to three sites) as well as studies in the Chinese context (e.g., Meng et al. 2020; Zhao et al. 2023; up to three or four points). We limited the number of points since respondents in Chinese studies marked fewer points, usually more than one or two per ES, even for those studies that did not limit the number (e.g., Xu et al. 2020; Zhang et al. 2020). In our study, the average number of points for different ESs marked by respondents ranged from 0.543 (for spiritual) to 2.211 (for aesthetic) (Table S10). This can be attributed to residents’ less familiarity with participatory mapping, with little resident representation under China’s current top-down management compared to well-established public participation systems in the Global North (Xu et al. 2020). In total, 223 questionnaires were collected within the study area. Detailed question descriptions for ESs, along with respondent characteristics, can be found in Appendix S2.

Following the PPGIS surveys, the points mapped by the respondents were digitized and imported into the SolVES model, which integrates normalized kernel density analysis and the maximum entropy (MaxEnt) model to generate social value index maps ranging from 0 to 10 for these ESs (Sherrouse et al. 2014). Details regarding the data sources and processing for the model can be found in Appendix S3. Finally, the value index corresponding to social values was scaled by dividing it by 10, thereby rendering it within the range of 0 to 1, facilitating direct comparison with the biophysical values (Pan et al. 2022). The biophysical and social values of all ESs were mapped and analyzed with a grid cell size of 10 × 10 m.

Trade-off analysis

Regarding trade-offs in ES types, the root mean square deviation (RMSD) was applied to quantify the degree of dispersion of a given ES in relation to the average of ESs (Bradford and D'Amato 2012; Yang et al. 2018). The RMSD value shows the distance between the ES and the synergy (no trade-off) state (represented by the 1:1 line). A larger RMSD value indicates a greater distance, signifying a higher degree of trade-off and a lower degree of synergy (Yang et al. 2018). In our study, RMSD was calculated to identify the intensity of trade-offs and synergies between each pair of ESs across the entire study area, as well as at the grid scale.

$$\text{RMSD}=\sqrt{\frac{1}{n-1}\times \sum _{i=1}^n{\left({\text{ES}}_i-\overline{\text{ES}}\right)}^2}$$ 1

where $i$ is a specific ES, $n$ is the number of ES, which is equal to 2 in this study, representing the two ESs for the trade-off identification, ${\text{ES}}_i$ is the normalized ES, and $\overline{\text{ES}}$ is the average of normalized $n$ ESs.

Regarding trade-offs in ES valuations, the quadrant division approach was used to identify and map the mismatching relationship between the two value types at the grid scale (Ding et al. 2023). Biophysical and social values for each grid were first standardized using Z-score standardization, and thus both values were adjusted to the distribution with an average of 0 and a standard deviation of 1. Then, the standardized biophysical and social values were plotted on a two-dimensional coordinate system, while the x-axis and y-axis represent these two value types, respectively. The origin of the two-dimensional coordinate system is the cut-off between “high” and “low” for these two value types. Therefore, four quadrants were generated, namely, high biophysical values—high social values (Quadrant I), low biophysical values—high social values (Quadrant II), low biophysical values—low social values (Quadrant III), and high biophysical values—low social values (Quadrant IV). Z-score standardization was calculated as follows:

$$x=\frac{x_j-\overline{x}}{s}$$ 2

where $x$ is the standardized biophysical or social value of ESs, $x_j$ is the biophysical or social value in grid $j$, $\overline{x}$ is the average biophysical or social value over all grids in the study area, and $s$ is the standard deviation of biophysical or social value over all grids in the study area. All calculations were performed using ArcGIS 10.8 software.

Conservation priorities determination

For the determination of conservation priorities for ESs, the biophysical and social values of ESs were integrated into the ordered weighted averaging (OWA) method. Proposed by Yager (1988) as a multicriteria decision-making procedure, OWA can handle trade-offs in the process of prioritizing conservation areas and has been used in ES studies (Zhang et al. 2022a, b). The OWA method first sorts the different criterion (ES values) maps in descending order based on their grid values (Gong et al. 2023). Then, it sets diverse decision scenarios by altering the decision risks and further calculates and assigns ordered weights (Wang et al. 2021).

$${\omega }_x={\left(\frac{x}{n}\right)}^{\propto }-{\left(\frac{x-1}{n}\right)}^{\propto },x=1,2,3,\dots ,n,\alpha \in \left(0,\infty \right)$$ 3

$$E=1-\sqrt{\frac{n{\sum }_x^n{\left({\omega }_x-\frac{1}{n}\right)}^2}{n-1}},\left(0\le T\le 1\right)$$ 4

where ${\omega }_x$ is the ordered weight of the sorted ordered layer $x$, $n$ is the number of criterion (ES values) raster layers, $\alpha$ is the decision risk coefficient, and $E$ is the evenness degree of the ordered weights under different risks. When $\alpha$ < 1, higher weights are assigned to the higher-ordered ES values. When $a$ = 1, each ES value is given the same weight, and the maximum evenness value of 1 is obtained. When $\alpha$ > 1, higher weights are assigned to the lower-ordered ES values (Gong et al. 2023). In our study, seven different decision scenarios (decision risk coefficients of 0.0001, 0.1, 0.5, 1, 2, 10, and 10 000) were established to calculate the ordered weights, according to the study of Zhang et al. (2022a).

By applying ordered weighting to different ES maps, the OWA finally prioritizes ES conservation areas under different decision risk scenarios (Wang et al. 2021).

$$\text{OWA}\left(a_{\mathit{xj}}\right)=\sum _x^n{\omega }_x{S}_{\mathit{xj}},\left({\omega }_x\in \left[0,1\right],\sum _x^n{\omega }_x=1\right)$$ 5

where $\text{OWA}\left(a_{\mathit{xj}}\right)$ is the OWA value at grid $j$ of layer $x$, $n$ is the number of criterion (ES values) raster layers, ${\omega }_x$ is the ordered weight, and $S_{\mathit{xj}}$ is the ordered layer obtained by normalizing the input layer in descending order.

Following the study of Zhang et al. (2022b), the top 20% of the OWA value areas in each scenario were regarded as priority conservation areas to protect the biophysical and social values of ESs. Furthermore, conservation efficiencies of these priority conservation areas across different scenarios were calculated and compared. The scenario that provides higher conservation efficiency for the biophysical and social values of ESs was selected as the optimal scenario for conservation priorities (Zhang et al. 2022b).

$$\text{CE}=\frac{\overline{{\text{ES}}_c}}{\overline{{\text{ES}}_t}}$$ 6

where $\text{CE}$ is the conservation efficiency, $\overline{{\text{ES}}_c}$ is the average of specific ES in the priority conservation areas, and $\overline{{\text{ES}}_t}$ is the average of specific ES in the total study area. Notably, among the seven scenarios considered, scenarios 1 and 7 (α = 0.0001 and 10 000, respectively) were two extreme scenarios with unbalanced weights and a single ES dominating, which did not provide practical references (Zhang et al. 2022a). Consequently, these two scenarios were excluded from our selection of conservation priorities. The OWA analysis was performed using IDRISI TerrSet software.

Results

Valuation and spatial distributions of ESs

Table 2 presents the average biophysical and social values of ESs across the study area. The ES with the highest average biophysical value was EC, followed by CES and HM, and the lowest was observed for FP. The social values of ESs presented lower averages compared to biophysical values, where the highest was found in FP, followed by HM, EC, and CES.

Table 2.

Average biophysical and social values of ESs in the study area

ES	FP	EC	HM	CES
Biophysical values	0.249	0.426	0.318	0.353
Social values	0.086	0.053	0.061	0.025

Figure 3 displays the spatial distributions of both biophysical and social values. In general, the spatial pattern of biophysical values exhibited a trend of higher in the west and lower in the east (Fig. 3a). Specifically, the biophysical values of FP and HM were higher in the western, southwestern, and northeastern areas. The high-value areas of EC and CES were primarily located in the western part of the study area, characterized by a greater presence of forests and water bodies exist. Conversely, the spatial distribution of social values did not entirely conform to the high-west and low-east pattern (Fig. 3b). Areas with higher social values for FP exhibited a relatively scattered distribution in the central parts of the study area. Higher values of EC, HM, and CES were concentrated in wetlands and aquatic areas, predominantly situated in the western and central areas.

Fig. 3.

Spatial distributions of a biophysical and b social values of four sets of ESs

Trade-offs in ES types and ES valuations

Table 3 provides an overview of the trade-offs in ES types across the entire study area. Concerning biophysical values, the trade-off intensities between different ESs were FP–EC > EC–HM > FP–CES > EC–CES > FP–HM > HM–CES. Within social values, the trade-off intensities between different ESs were FP–CES > HM–CES > FP–EC > EC–CES > FP–HM > EC–HM. In comparison to the biophysical values, the trade-off intensities were generally lower and the synergy intensities were higher within the social values. Notably, relatively stronger trade-offs were consistently observed between FP and other ESs within both the biophysical and social value dimensions, particularly for trade-offs between FP and EC, as well as FP and CES.

Table 3.

The intensity of trade-offs in ES types for the total study area calculated using RMSD

ES	FP	EC	HM	CES
FP	–	0.024	0.018	0.043
EC	0.126	–	0.006	0.019
HM	0.049	0.077	–	0.025
CES	0.074	0.052	0.025	–

Biophysical value trade-offs are displayed in the lower left corner and social value trade-offs are displayed in the upper right corner

Figure 4 illustrates the spatial distributions of trade-offs in ES types at the grid scale. In terms of biophysical values, FP–EC, FP–HM, and EC–CES trade-offs exhibited higher intensities in the western areas than in the eastern areas (Fig. 4a). Conversely, the trade-off intensity of FP–CES, EC–HM, and HM–CES was relatively evenly distributed throughout the study area. In terms of social values, areas with higher trade-off intensities between FP and other ESs (EC, HM, and CES) displayed a scattered pattern, manifesting in various points and patches across the study area. In contrast, trade-offs among other social values of ESs were primarily characterized by low-intensity distributions (Fig. 4b).

Fig. 4.

Trade-offs in ES types: Spatial distributions of trade-offs within a biophysical and b social values of ESs

Figure 5 depicts the trade-offs in ES valuations. The characteristics for the four types of relationships (quadrants) were as follows: (1) areas with high biophysical and high social values were less common for all ESs. In the case of FP, these areas were dispersed throughout the study area. In EC, HM, and CES, they were primarily situated within forests, water bodies, and wetlands. (2) Areas with high biophysical and low social values were more common in FP, HM, and CES, while less in EC. In the former, they were predominantly found in the northeast and southwest areas, where cropland was more abundant. In the latter, they were distributed in patches around forests in the western areas. (3) Areas with low biophysical and high social values were more prevalent in FP and less in EC, HM, and CES. These areas were largely situated around built-up areas and along rivers in the central part of the study area. Additionally, they were present in several water bodies in the western areas, particularly in the case of EC. (4) Areas with low biophysical and low social values were the most common relationships in EC and CES, exhibiting a broad distribution across the study area. In the cases of FP and HM, these areas were primarily concentrated and distributed in the eastern part of the study area.

Fig. 5.

Trade-offs in ES valuations: Spatial distributions of biophysical and social value matching of four sets of ESs

Identification of priority conservation areas

Priority conservation areas for both the biophysical and social values of ESs, determined through the OWA method under various decision risk scenarios, are presented in Fig. S1, Appendix S4. As the decision risk coefficient α increased, the decision-making gradually changed, leading to an expansion in the conservation areas across the decision scenarios. The conservation efficiencies of these priority areas under different scenarios are detailed in Table S12. After excluding extreme scenarios 1 and 7 with single ES dominating the weight assignment, scenario 2 emerged with the highest average conservation efficiency (2.155) among the remaining scenarios. In addition, the conservation efficiency of each ES was greater than 1 in scenario 2, ensuring the representation of all ESs (Table S12). Consequently, scenario 2 was selected as the most reliable scenario, and its associated conservation areas were considered priority conservation areas.

Figure 6 illustrates the priority conservation areas of ESs, encompassing 16.83% of the total study area. These areas were primarily distributed across most parts of Jinze Town, southern and northern Liantang Town, and northern Zhujiajiao Town in the western part of the study area, with some portions extending into the eastern and northeastern part (Fig. 6a). Land cover types within these priority areas predominantly consisted of forests, cropland, water bodies, and grassland (Fig. 6b). In terms of different land cover types within the study area, a substantial proportion of forest and wetland were designated priority conservation areas (Fig. 6c).

Fig. 6.

Characteristics of priority conservation areas in Qingpu District. a Spatial distributions of priority conservation areas, b composition of land cover types in priority conservation areas, and c conservation area and non-conservation area proportions for different land cover types

Discussion

Integrating biophysical and social values in ES mapping

In this study, we adopted an interdisciplinary social–ecological approach that integrates biophysical and social values to provide a spatially explicit assessment of ESs in peri-urban areas. Analyzing only biophysical values, without considering social values, does not facilitate equitable ES management (Xia et al. 2023). Conversely, examining only social values, while neglecting biophysical values, may result in unbalanced information and potentially leading to detrimental outcomes (Ruiz-Frau et al. 2018). Therefore, these two valuation approaches built on distinct knowledge systems complement each other, aligning with the call for value pluralism in sustainable environmental management (Arias-Arévalo et al. 2018; IPBES 2022; Pascual et al. 2023). Furthermore, embracing this systematic and integrated social–ecological approach in management can be viewed as an endeavor to foster harmonious coexistence between people and nature (Zhang and Fu 2023), which is the foundation of the “ecological civilization” currently advocated in China (Hansen et al. 2018; Xue et al. 2023).

In light of these considerations, our findings reveal that ESs exhibited the highest average biophysical value for EC, while FP demonstrated the highest average social value, highlighting their disparities. Meanwhile, the spatial distribution of these two value dimensions also varied. The biophysical values of ESs generally exhibited a pattern of being high in the west and low in the east, consistent with the landscape characteristics and urbanization trends within the study area. The western part of Qingpu District features seminatural landscapes, such as paddy fields and water ponds, whereas the eastern part has seen an expansion of built-up areas due to central subdistrict and transportation infrastructure development, impacting the biophysical values of ESs. The social values of ESs, however, did not conform to a similar spatial pattern, instead primarily clustered near specific ecosystems or landscape features, such as water bodies and wetlands within the study area, consistent with the findings of Garcia-Martin et al. (2017) and Garau et al. (2023). In addition, the studies of Fagerholm et al. (2019) and Cusens et al. (2022) revealed that social values were linked to accessibility, where places closer to residences and with easy accessibility tend to be mapped by residents. This can be also attributed to geographical or spatial discounting, i.e., people wish to be close to the things they value and to be distant from what they dislike (Brown and Kyttä, 2014). These suggest that residents’ perceived social values can be more closely linked to their awareness, preferences, and use of specific landscape features in these areas. These disparities in the spatial distribution of values offer more comprehensive information for identifying priority conservation areas.

Peri-urban ES trade-offs: From biophysical to social dimensions

Our study incorporates trade-off analysis to elucidate the relationships between different ES values. Within both value dimensions, the intensity of the trade-offs between FP and other ESs was higher. Such trade-off relationships can be commonly found in studies focused on the biophysical dimension, as a systematic review by Aryal et al. (2022) found that the highest trade-offs were observed between agricultural production and other ESs. Studies focused on peri-urban ESs have also shown the trade-offs occurring between provisioning ESs and regulating ESs (Chen et al. 2023), as well as between provisioning ESs and cultural ESs (Sylla et al. 2020). These trade-offs are objective phenomena of complex social–ecological systems, frequently stemming from the prioritization of a single land use objective (Meyfroidt et al. 2022).

Simultaneously, the trade-off relationships between FP and other ESs exhibited spatial heterogeneity, featuring distinct patterns of high trade-off intensities in both biophysical and social values. These spatial variations can be explained by the land-sparing measures, which dominate land use practices in China, including Qingpu District (Ren et al. 2021). According to the Draft Master Plan and Land Use Master Plan of Qingpu District (2017–2035) (PGQDSM 2020), the district has designated ecological spaces and ecological red lines, permanent basic farmland, and urban development boundaries for spatial zoning control and land utilization. On the one hand, strict land use policies are implemented within local ecological spaces, encompassing measures such as the prohibition of wildlife hunting, the control of fish farming in natural waters, and the gradual reduction of agricultural land (Wang et al. 2022). On the other hand, contiguous cropland areas have been delimited as permanent basic farmland to supply essential grains and vegetables for Shanghai. These intensively cropped areas cannot arbitrarily implement alternative land uses, such as afforestation (Wu et al. 2017). These land-sparing measures that divide landscapes into multiple monofunctional zones account for the emergence of ES trade-offs in specific spatial locations rather than across the entire peri-urban areas.

Comparing biophysical and social value dimensions, our findings reveal that ES trade-off intensities were relatively lower in the social values compared to the biophysical values. This is in line with the findings of social valuation studies (Plieninger et al. 2019; Karimi et al. 2020), which identified more synergies between the social values of ESs. One possible explanation is that despite the rapid urbanization in peri-urban areas, several types of human–nature connections persist in these agriculturally dominated landscapes (Plieninger et al. 2019). Local residents may perceive these diverse ESs as interdependent (Yang et al. 2019). Furthermore, respondents’ daily lives and interactions may contribute to positive perspectives of ESs, while assessments grounded in technical and ecological expertise tend to unveil more pronounced ES trade-offs (Hossu et al. 2019). Therefore, ESs are not mutually constraining in the societal dimension.

When integrating biophysical and social values, our findings indicate the presence of trade-offs between different valuations of ESs. Among the two types of mismatches observed, areas with high biophysical and low social values were situated in the southwest and northeast areas, particularly for FP and HM, which were primarily located in the agricultural landscapes of these areas. Respondents were less likely to map the related multiple ESs in these areas. These social value gaps may be attributed to the uniform land cover and cropping patterns and the visually homogeneous appearance in these peri-urban agricultural areas (Schwartz et al. 2022). Conversely, areas with low biophysical and high social values were predominantly clustered around water bodies or built-up areas in the central core, characterized by higher urbanization levels and population densities. These areas were less biophysically valued but were more identified and mapped by respondents. This can be attributed to the accessibility to more local residents (Fagerholm et al. 2019; Raymond et al. 2023b). It is worth noting that areas with high biophysical and high social values were less commonly found across all ESs, highlighting the limited instances where a social–ecological win–win scenario is achieved, necessitating attention in conservation and management efforts.

Recommendations for conservation and management

Through the integration of biophysical and social value layers within multicriteria decision-making, our study simulated the influence of risk attitudes on ES conservation and its overall efficiency. For the priority conservation areas in the chosen scenario 2 obtained by the OWA method, comprehensive protection of their environment and culture is needed, as well as control of land development and construction activities to mitigate the impacts of rapid urbanization. Meanwhile, this multicriteria decision-making for conservation and management needs to be complemented by the considerations of ES interactions, including ES trade-offs in our study, to better understand the results of conservation priorities and to obtain more complete and detailed information for ES synthesis and further strategies (Cortinovis et al. 2021). We suggest that sustainable environmental management and policy-making entail not only the combination of complementary scientific and local knowledge to identify priority areas, but also the implementation of diverse strategies and measures to manage ES trade-offs identified in our study, encompassing both trade-offs in ES types and ES values within the local context. This can complement the effective identification results of conservation priorities and allow them to be better operationalized for management. The implementation of this integrated governance of ES trade-offs needs to be oriented toward pluralistic and adaptive (Spyra et al. 2020), by combining different land use strategies as well as different forms of knowledge. In this regard, the following two specific recommendations are proposed to address the two types of ES trade-offs.

In addressing trade-offs in ES types, we emphasize promoting the synergies of ES types and mitigating the trade-offs identified in our study, i.e., the prominent trade-offs between FP and other ESs. To address this issue, we do not wish to choose between the common land‐sparing or land-sharing strategies dealing with agriculture and natural ecosystems, but rather emphasize integrated approaches that combine and balance these two strategies (Grass et al. 2019). On the one hand, the current land‐sparing strategy can enhance the efficiency of land policy implementation under China’s public land ownership (Feng et al. 2016). On the other hand, land‐sparing measures also face challenges, as implied in our study. Intensively cultivated areas with homogeneous visual characteristics may diminish residents’ recognition of social values (Schwartz et al. 2022), and such landscape simplification can erode the diverse relationships between people and nature (Riechers et al. 2020), which are considered to foster synergies in perceived social values (Plieninger et al. 2019). Hence, the key is to create peri-urban areas that integrate both land‐sparing and land-sharing strategies. For example, a small amount of ecological space can be supplemented within existing intensive cropland (Chen et al. 2023), as well as the fostering of agrotourism. Another example is the appropriate use of shaded environments within forested areas to cultivate understory products. Moreover, peri-urban areas encompass not only agriculture and nature conservation but also sprawling urban landscapes requiring rational development patterns (Collas et al. 2017). In management practice, it is necessary to prudently delineate and optimize land uses, and to carry out multidimensional governance to improve the comprehensive benefits in peri-urban areas.

In addressing trade-offs in ES valuations, despite the delineation of priority conservation areas, there are still areas outside that require improvement through management tools. This necessitates targeted management of high and low matching relationships among different ES values. Specifically, areas with high biophysical and high social values represent “win–win” areas with highly matching values that deserve strong management support (Bagstad et al. 2016). In such areas, it is imperative to continue leveraging existing social and ecological strengths while adopting a socially inclusive approach to environmental conservation. Conversely, areas with low biophysical and low social values constitute “lose–lose” areas with a low value match. In such areas, prioritization in environmental management could be reduced, and alternative development options need to be explored (Raymond et al. 2023b). For the two types of value mismatches, areas with high biophysical and low social values are scientifically justified for ES conservation but lack social acceptance. With this regard, educational programs for scientific knowledge can be implemented in these areas to raise local awareness of protecting related ESs (Bryan et al. 2011; Whitehead et al. 2014). Last, for areas with low biophysical and high social values, ESs hold significant social importance but are not scientifically applicable. In such instances, fostering community engagement in such areas is essential (Whitehead et al. 2014; Raymond et al. 2023b). This entails listening to the views and perspectives of residents and facilitating their participation in the management process (Xia et al. 2023). These management measures for addressing ES value trade-offs complement those aimed at addressing ES type trade-offs, and together serve as an underpinning for the conservation of peri-urban areas.

Limitations and prospects

Our study has several limitations that warrant attention in future research. First, the selection of peri-urban area boundaries involves a variety of methods, and it is difficult to make a general delimitation of diverse peri-urban areas across the globe (Mortoja et al. 2020; Žlender 2021; Sahana et al. 2023). There are uncertainties in our delimitation of the peri-urban area based on the administrative boundary, which may lead to different spatial distributions in ES mapping results. For example, the eastern part of the study area features more built-up areas and can exhibit more urban characteristics, compared to the southwestern part. Considering the eastern part as urban and excluding it from the peri-urban area would no longer result in a distribution of high in the west and low in the east for the biophysical values. The delimitation in our study was supported to some degree since higher density built-up is a common feature in peri-urban areas in the Global South, such as China, compared to the Global North (Sahana et al. 2023). Nevertheless, future studies could examine multiple quantitative methods for delimitating peri-urban boundaries and determine the method based on the specific context of the study area. For example, the demarcation of interest areas as peri-urban, through population, land cover, and night-light data (Wangai et al. 2019; Mortoja et al. 2020). These possible reclassifications would allow the results to more prominently reflect the distinctive characteristics of peri-urban areas in the Global South.

Second, the ES analysis in our study needs further reflection. On the one hand, the number of ESs needs to be increased in future social–ecological assessments. Our study only selected ESs that could be valued and mapped from both the biophysical and social dimensions in the study area, and used them as a methodological example to demonstrate the analytical framework. Further studies need to apply it to a wider range of ESs, including those not covered in this study (e.g., carbon sequestration). This would avoid potential oversight and to better inform environmental management. On the other hand, we conducted both biophysical and social valuations for each ES, yet the representativeness and precision of these valuation approaches varied across different ESs. For instance, biophysical valuation proved to be appropriate for provisioning, regulating, and supporting ESs, whereas the PPGIS approach might be more applicable for cultural ESs and inadequate for mapping other ESs (Bagstad et al. 2016; Cusens et al. 2023).

Last but not least, although our study addressed the globally relevant topic of plural valuation and conservation of ESs, the findings relate to the specific context. For instance, specific local landscape characteristics, socio-demographic characteristics of residents, and their relationships with landscape can influence the perceived social values of ESs (Fagerholm et al. 2019). These factors vary between sites, regions, and countries. Our case study aimed to provide a methodological basis and lessons for developing management strategies in peri-urban areas with similar contexts. Future studies need to consider the above factors and incorporate cross-site comparisons of different cases, thus providing further insights into the characteristics of diverse values.

Conclusions

In the conservation and management of peri-urban ESs, it is vital to strike a balance between leveraging scientific knowledge and incorporating residents’ voices. Using a spatially explicit social–ecological approach, our study provided both biophysical and social valuations of ESs in peri-urban areas, thoroughly analyzed the trade-offs within ES types and ES valuations, and integrated these plural values into the identification of priority conservation areas. We conclude that the biophysical and social values of ESs exhibit disparities in their intensities and spatial distributions. The trade-offs and synergies within different values also manifest varying degrees of intensity, with biophysical values generally exhibiting higher trade-offs compared to social values. Furthermore, multicriteria decision-making informed by the relationships between different values effectively guides the incorporation of land cover, biophysical functions, resident perceptions, and social well-being into conservation prioritization and management practices. Given the outcomes of the trade-off analysis, we believe that integrated management strategies are needed to promote the synergies among multiple ESs of peri-urban areas as a whole. It is also essential to achieve a win–win situation that encompasses both scientific justification and social acceptance through educational initiatives and resident participation. We call for increased interdisciplinary research collaborations between environmental and social scientists to comprehensively identify and map the diverse values of ESs while gaining a deeper understanding of their dynamics and drivers. We also encourage future transdisciplinary collaborations involving scientists, managers, decision-makers, and local communities to facilitate knowledge integration, thereby enabling the practical application of ES valuation in policy-making for peri-urban sustainability.

Supplementary Information

Below is the link to the electronic supplementary material.

Biographies

Zheyi Xia

is a Ph.D. student at the College of Land Science and Technology, China Agricultural University. His research interests include ecosystem services valuation and environmental management.

Jiasi Huang

is a master’s student at the College of Land Science and Technology, China Agricultural University. His research interests include ecosystem services mapping and resilient landscape planning.

Yuwen Huang

is a Ph.D. student at the Center for Historical Geographical Studies, Fudan University. Her research interests include cultural landscapes and environmental planning.

Kui Liu

is a Ph.D. student at the School of Public Policy and Administration, Xi’an Jiaotong University. His research interests include environmental economics and urban policies.

Runmiao Zhu

is a Ph.D. student at the College of Land Science and Technology, China Agricultural University. Her research interests include landscape multifunctionality and ecosystem services valuation.

Zhen Shen

is a Ph.D. student at the College of Land Science and Technology, China Agricultural University. His research interests include social–ecological systems and land use modeling.

Chengcheng Yuan

is an associate professor at the College of Land Science and Technology, China Agricultural University. His research interests include sustainable land use and management.

Liming Liu

is a professor at the College of Land Science and Technology, China Agricultural University. His research interests include land use management and landscape planning.

Author contributions

Zheyi Xia contributed to conceptualization and writing—original draft preparation; Zheyi Xia, Zhen Shen, and Chengcheng Yuan helped in methodology; Zheyi Xia, Jiasi Huang, and Yuwen Huang helped in formal analysis and investigation; Kui Liu, Runmiao Zhu, and Liming Liu done writing—review and editing; Chengcheng Yuan and Liming Liu helped in funding acquisition; Liming Liu done supervision.

Declarations

Conflict of interest

The authors have no competing interests to declare that are relevant to the content of this article.