Introduction

The Earth system is entering a new era characterized by significant global environmental changes primarily driven by human activities, known as the "Anthropocene" (Steffen et al. 2011; Keys et al. 2019). Since entering the Anthropocene, the scope, intensity, and magnitude of human impacts on the Earth's land surface have been continuously expanding. These impacts have resulted in a series of global challenges, such as water scarcity, biodiversity loss, climate change, and environmental pollution, posing significant challenges to the sustainable development of human society (Cardinale et al. 2012; Stringer et al. 2021). Landscapes are the most suitable spatial scale for sustainable development research and practice, and achieving landscape sustainability is essential for global sustainability (Wu 2021; Fu et al. 2022; Wang et al. 2023).

Landscape sustainability assessment is a powerful tool to assist decision-makers in taking appropriate measures and promoting the transition of landscapes toward sustainability (Jenny et al. 2004; Wu 2013). While there is a strong consensus on the necessity of landscape sustainability, there is almost no consensus on which methods or indicators should be used to measure it. Scholars have proposed various landscape sustainability assessment methods from different perspectives. For instance, Wang et al. (2020) constructed a regional landscape sustainability assessment framework based on the paradigm of a safe operating space, selecting indicators from the dimensions of "maintenance" and "development." They used Liaoning, a coastal province in China, as an example to analyze the interaction trends between safe and just space status and the social-environmental system. Fang et al. (2015) argued that landscape sustainability assessment should evaluate landscape service capacity, landscape service flow, landscape service demand, and the dynamic relationships among them. They proposed a conceptual framework but did not conduct further empirical research. Eichler et al. (2020) considered environmental, social, and economic dimensions as well as stakeholder interests The indicators most relevant to agricultural landscape sustainability were then selected to assess the sustainability of the landscape in Mexico's Yaqui Valley. Dale et al. (2019) introduced a landscape sustainability assessment method that involves stakeholder participation, including six specific assessment steps, and applied this method to bioenergy production in the United States and agricultural production in the Yaqui Valley of Mexico. This method fully considers stakeholder participation principles but requires significant resources (personnel, time, staff time, money) to implement for assessing landscape sustainability, and the required indicators are not easily obtainable. These methods have made significant contributions to landscape sustainability assessment by providing different research perspectives and approaches, as well as having different scopes of applicability.

ESs have an important relationship with landscape sustainability, and many definitions of landscape sustainability involve ESs (Wu 2013; Fang et al. 2015; Peng et al. 2021). Wu (2021), building upon prior research, defined landscape sustainability as the "integrated capacity of a specific landscape to provide long-term stable ESs that maintain and enhance the well-being of the local population." Its core lies in the landscape's supply capacity of ESs and the demand for ESs required to maintain and improve human well-being within that landscape. It can be seen that ESs have an important relationship with landscape sustainability and can effectively represent and reflect the core content of landscape sustainability. In view of this, we propose a conceptual framework and steps for quantitatively assessing landscape sustainability from the perspective of ES supply, flow, and demand. This framework aims to support actions to enhance landscape sustainability. This study has two main objectives. First, through a literature review and theoretical analysis, we present a conceptual framework for assessing landscape sustainability based on ES supply, flow, and demand and introduce specific assessment steps. Second, we apply the proposed framework to the township landscape in Yixing and discuss the contributions, limitations, and shortcomings of the framework. The remainder of this paper is organized as follows (Fig. 1).

Fig.1
figure 1

The organization of this paper

Conceptual framework for assessing landscape sustainability based on ESs supply-flow-demand

The core of landscape sustainability includes "landscape pattern," "ecosystem services," and "human well-being." ESs are closely related to landscape sustainability, and many definitions of landscape sustainability involve ESs (Fang et al. 2015; Peng et al. 2021). Some studies argue that the relationship between the supply and demand of ESs can effectively characterize and reflect the crucial concepts of landscape sustainability, such as "long-term stable provision of ecological environments" and "maintenance and improvement of human well-being" (Wu 2013; Zhou et al. 2019). Some research suggests to using ESs as indicators for assessing landscape sustainability and points out the potential feasibility of using ecosystem supply and demand to represent landscape sustainability (Fang et al. 2015). This has provided a suitable entry point for landscape sustainability research and has drawn attention from the academic community. Therefore, we have developed a conceptual framework diagram that analyzes the connections between ESs supply, flow, demand, landscape patterns, and human well-being (Fig. 2). This framework aids in understanding why it is valuable to conduct landscape sustainability assessments from the perspective of ES supply-flow-demand. We try to explain several core elements of this framework and their theoretical relationships in the following six sections, which will help to understand the framework more clearly.

Fig.2
figure 2

Conceptual framework for assessing landscape sustainability based on ESs supply-flow-demand. Adapted from (Fang et al. 2015; Wu , 2013, 2021; Assis et al. 2023)

(1) ES supply refers to the potential of a given ecosystem to generate services based on its processes or functions. this is equivalent to "ES capacity" (Villamagna et al. 2013; Metzger et al. 2021). Some research defines it as the "potential ES supply" to distinguish it from the "actual ES supply" (Baró et al. 2016; Zeng et al. 2023). ES demand refers to the quantity of services that society needs or expects and can be expressed through the direct use or consumption of goods and services (Villamagna et al. 2013; Wolff et al. 2017; Assis et al. 2023). ES flow is the actual process that connects supply and demand (the flow of people, organisms, or spatial materials) (Metzger et al. 2021). However, for some authors, it is also understood as "realized services," (Villamagna et al. 2013) "actual ES supply," (Serna-Chavez et al. 2014) or "the transfer of services from ecosystems to humans" (Wang et al. 2022; Perschke et al. 2023). As a process of transferring ES supply to demand, ES flow depends on mechanisms that connect supply and demand, including spatial distribution and landscape structure, and each type of ES flow has unique characteristics (Serna-Chavez et al. 2014; Metzger et al. 2021; Assis et al. 2023).

(2) As a crucial link between human society and nature, ESs have direct or indirect impacts on human well-being (Qiu et al. 2022). When humans actually consume or demand, they transform potential ESs partially or entirely into actual ESs, which are the ultimate ESs that humans actually enjoy (Villamagna et al. 2013; Bagstad et al. 2020; Wang et al. 2022). Factors such as climate change, invasive species, and human disturbances directly or indirectly drive changes in land use/land cover, affecting the structure, processes, and functions of ecosystems and thus influencing the supply of ESs (Branco et al. 2015; Dong et al. 2021). The fluctuations in supply, which are the responses of ESs, directly feed back to human well-being (Pires de Souza Araujo et al. 2021; Qiu et al. 2022). ESs maintain and enhance human well-being through a cascading process of "supply-flow-demand-human benefits" (Haines-Young and Potschin 2010; Assis et al. 2023). Therefore, this study reflects the "maintenance and enhancement of human well-being by ESs" through the state of ES supply-flow-demand in the landscape.

(3) Landscape patterns, including composition and configuration, can directly influence the condition of ecosystems (e.g. their quantity or the quality of ecological processes within the landscape), thereby affecting their capacity to provide ESs (Knoke et al. 2016; Chen et al. 2022). Furthermore, landscape patterns also have a significant impact on the flow of ESs. Mitchell et al. (2015b) argued that fragmentation has a critical influence on ES supply and flow. Research has shown that landscape structure affects service provisioning and service flow through multiple landscape-level processes, such as fragmentation, edge effects, and connectivity (Villard and Metzger 2014; Metzger et al. 2021). Additionally, (Assis et al. 2023) proposed a theoretical framework to specifically explore how the spatial flow of ESs varies based on landscape structure (i.e. composition and configuration).

(4) While ES supply and flow are particularly influenced by landscape structure, the demand for services is a driving factor affecting landscape patterns (Willemen et al. 2012; Neyret et al. 2023). There is evidence that the demand for ESs stems from societal development and human life, often manifesting as economic pressures from crop and livestock production and industrial land use changes (Baró et al. 2015, 2016). The demand for multiple ESs directly or indirectly drives changes in landscape patterns (both composition and configuration) (Willemen et al. 2012; Neyret et al. 2023). Moreover, drivers such as climate change and urban expansion more directly lead to changes in landscape patterns (Assis et al. 2023).

(5) From this perspective, landscape patterns influence the supply and flow capacity of ESs, while the demand for ESs drives changes in landscape patterns. In this framework, the states of ES supply, flow, and demand can characterize the operational state of this framework, and the states of ES supply, flow, and demand reflect the degree to which ESs meet human well-being. Based on this analysis, we attempt to characterize and quantify landscape sustainability through the states of ES supply, flow, and demand.

(6) It is important to note that, like many concepts in landscape ecology, landscape sustainability is a multiscale concept, spanning decades to centuries in terms of time scales and influenced spatially by multiple scales from local to global (Wu 2021). Therefore, in the assessment process of landscape sustainability, we need to consider assessments and analyses spanning multiple years. Additionally, Wu (2013) emphasized that landscape sustainability should be a strong sustainability concept, meaning that we should strive to enhance key ESs so that they can meet local needs and improve human well-being.

Steps for assessing landscape sustainability based on ESs supply-flow-demand

After a theoretical analysis of the conceptual framework for landscape sustainability assessment based on the supply-flow-demand of ESs, here are the steps for assessing landscape sustainability (Fig. 3).

Fig.3
figure 3

Steps for assessing landscape sustainability based on ES supply-flow-demand

Step 1: Select Key ES Types Based on Resource Endowment and Human Development Needs.

The Millennium Ecosystem Assessment (MEA) categorized ecosystems into four types: provisioning services (e.g. food, fiber, fuel, genetic resources), regulating services (e.g. water purification and regulation, climate regulation, disease control), supporting services (e.g. primary production, nutrient cycling), and cultural services (e.g. ecotourism and recreation, aesthetics, and spiritual values) (Chapin et al. 2005; Haines-Young and Potschin 2010). In this study, we refer to the MEA classification and select key ESs based on these categories as the foundation for assessment. However, assessing all types of ESs as outlined in the MEA can be challenging due to data and workload constraints. Therefore, a selection of key ESs for the target area is necessary. Methods for selecting these services could involve surveys, interviews with experts, stakeholders (local residents), and government officials, or by reviewing government policy documents, relevant literature, and research reports.

Step 2: Quantify the Supply-Flow-Demand Levels of Selected Key ESs.

The number of methods and tools available for assessing ESs in specific situations is continuously increasing. These include biophysical models, integrated mapping-modeling methods (such as InVEST and ESTIMAP), land-use scoring methods, monetary methods, and more (Harrison et al. 2018). (Wood et al. 2018) conducted a literature review, compiled and compared commonly used assessment models for different types of ESs, and analyzed their applicability. In this step, you need to select the appropriate assessment method to quantify ESs based on the selected key ES types and available data. Sometimes, referencing relevant literature can help determine the method and model parameters required for quantification. The potential supply of ESs is mainly based on existing biophysical models, while the demand for ESs is mainly based on socioeconomic statistics, such as per capita food consumption, per capita water demand, and per capita carbon emissions. The ES flow (the actual supply of ESs) needs to be calculated based on the relationship between the potential supply and demand of ESs or biophysical models (Baró et al. 2015, 2016). In this process, it is necessary to unify the measurement unit of supply, flow and demand assessment results, such as using a unified physical quantity unit or a unified percentage.

Step 3: Integrate the Supply, Flow, and Demand States of Different ESs to Characterize Landscape Sustainability.

In this step, integrate the states of supply, flow, and demand of ESs to characterize landscape sustainability. You can start by simultaneously quantifying the ratio of ES supply to demand and the ratio of flow to demand for each type of ES. Then, you sum the supply–demand ratio and flow-demand ratio to determine the level of ES supply-flow-demand. Finally, combine the levels of supply-flow-demand across different ESs to represent landscape sustainability comprehensively. We should strive to enhance key ESs so that they can meet local needs and improve human well-being.

Step 4: Analyze and Discuss Landscape Sustainability over Longer Time Scales.

Sustainability, including landscape sustainability, is a long-term issue, with a time scale of at least several decades (Wu 2021). Therefore, assessing landscape sustainability for a single time period is insufficient. To obtain more reliable results, it is necessary to consider longer time frames and conduct assessments for multiple time periods, creating a time dimension for landscape sustainability (Banamar and Smet 2018; Urli et al. 2019). This allows for comprehensive analysis and discussion of trends in landscape sustainability over an extended period, whether it is declining, stable, or improving. However, during this process, you should consider the consistency and availability of data sources for long-term assessments.

Case study

Study area and indicator selection

In this study, we treated each township in Yixing as a landscape. We focused on the "township landscape" of Yixing as our research object. The "township landscape" refers to a part of the Earth's surface within the boundaries of a township, formed by the spatial arrangement of multiple ecosystems and human elements. Here, we emphasize the physical and spatial characteristics of the landscape (Bastian et al. 2014). Yixing city is located in Jiangsu Province, China, with geographical coordinates ranging from 31°07' to 31°37'N and 119°31' to 120°03'E (Fig. 4). It encompasses 18 towns and covers a total area of 1996.6 square kilometers, including 242.29 square kilometers of the Taihu Lake water area. Yixing city is situated in the upper reaches of the Yangtze River Delta and is characterized by rapid urbanization. It plays a vital role in both social and ecological aspects. Yixing city boasts abundant agricultural resources and is a significant grain-producing region in the Yangtze River Delta. It is renowned as a cultural and ecotourism international city and is often referred to as the "Garden City of China" due to its rich leisure and entertainment resources. The city also has an important role in water provision, with numerous rivers, lakes, reservoirs, and its proximity to Taihu Lake. However, the rapid development of industry and urbanization has led to severe pollution of rivers and lakes from various sources, including industrial, agricultural, and domestic activities. This has resulted in elevated nitrogen and phosphorus concentrations, posing significant challenges to water quality. Given Yixing city's natural resource endowment, regional development policies (e.g. "Yixing City Land Spatial Planning 2021–2025" and "Yixing City National Economic and Social Development 'Fourteenth Five-Year' Plan and Vision Outline"), data availability, and previous research (Li et al. 2016; Bai et al. 2020; Li 2022), five categories of ESs were selected for landscape sustainability assessment: carbon sequestration, food production, nature recreation, water provision, and water purification. The aim is to provide scientific support for regional landscape governance and socioecological sustainable development.

Fig.4
figure 4

The location of the case area

Data sources

This study utilized six main categories of data: (1) Land use classification data, rainfall data, population density grid data, and NDVI data for the years 2000, 2010, and 2020 were obtained from the Data Center for Resources and Environmental Sciences of the Chinese Academy of Sciences (http://www.resdc.cn). The data have a resolution of 30 m. (2) Digital elevation model (DEM) data with a resolution of 30 m and soil data with a resolution of 1 km were sourced from the Chinese Geospatial Data Cloud (www.gscloud.cn) and the Harmonized World Soil Database (HWSD), respectively. (3) Meteorological data such as precipitation and evapotranspiration were obtained from the "China Surface Climate Data Set" provided by the China Meteorological Administration (http://data.cma.cn/). (4) Per capita comprehensive water consumption data were sourced from the "Water Resources Bulletin of Jiangsu Province." (5) Per capita carbon emissions data were calculated using the China Emission Accounts and Datasets, a database for carbon accounting in China. (6) Population data were derived from the "Statistical Yearbook of Yixing City," and per capita food consumption data were obtained from the "Statistical Yearbook of China" and the "Statistical Yearbook of Jiangsu Province" for the per capita consumption of major food items in Jiangsu Province.

Assessment methods

Quantification of ES supply, demand, and flow

For the assessment of ES supply and demand, various existing models and methods are mature and effective, and data for the case study area is available. In this study, the assessment of the potential supply of ESs mainly utilizes models such as NPP and InVEST. The assessment of demand for ESs is primarily based on residents' actual consumption and demand for ESs. The assessment of ES flow (actual supply) is crucial, and it is challenging to comprehensively assess multiple ES flows. To address this, we refer to the ES supply-flow-demand matrix proposed by (Burkhard et al. 2014) to assess various ES flows. This matrix was constructed through literature analysis combined with expert ratings, and it includes ES-S ES-F, and ES-D matrices for 28 ESs and 7 land use types. It clearly maps the relative scores (ranging from 0 to 5) of ES potential supply, flow, and demand under different land cover types. The matrix is characterized by its comprehensiveness, flexibility, operability, and has been effectively applied (Li et al. 2016; Zeng et al. 2023). Of course, this method is not perfect, but it has some applicability in this study. We hope to explore more diverse and effective methods for assessing ecosystem service flows in future research, which is also an important direction for future studies (Peng et al. 2023).

In addition, when selecting the methods and models for ES assessment, one consideration is that these models are mature and widely used. Another consideration is data requirements. We obtained relatively complete resource and environmental data for Yixing city, such as land use classification data, rainfall data, as well as socioeconomic data such as population density grid data, per capita carbon emissions, and per capita food consumption. These data can effectively support the application of the selected methods and models in the case study area. Specific quantification methods are as follows (Table 1).

Table 1 Description of ESs supply-flow-demand index and its evaluation method

Characterizing landscape sustainability based on ES supply-flow-demand

After clarifying the theoretical logic of landscape sustainability assessment based on the supply-flow-demand of ESs, this study introduces two indicators, the supply–demand ratio and the flow-demand ratio, to represent the relationships or states between ES supply, flow, and demand. These indicators are then used to comprehensively characterize landscape sustainability based on various ESs. The supply–demand ratio represents the balance between the potential supply capacity of ESs in the landscape and human demand (Eq. (1)). The supply–demand ratios of various ESs are not bound to specific measurement units. A higher value indicates a greater potential for ESs of that type to meet human needs. If the ratio is greater than 1, it means that the potential supply of that ES can meet human demand.

$${Ratio}_{S-D}=\frac{{ES}_{Potential Supply}}{{ES}_{Demand}}$$
(1)

The flow-demand ratio represents the actual ES supply to human demand (Eq. (2)). The flow-demand ratios of various ESs are not bound to specific measurement units. A higher value indicates a higher level of actual satisfaction of human demand for that type of ES. If the ratio is greater than 1, it means that the flow (actual supply) of that ES can meet human demand.

$${Ratio}_{F-D}=\frac{{ES}_{Flow} }{{ES}_{Demand}}$$
(2)

Next, the combined value of the supply–demand ratio and the flow-demand ratio is used to characterize the supply-flow-demand levels of a particular type of ES (Eq. (3)). The supply-flow-demand levels of various ESs are not bound to specific measurement units. This indicator aims to reflect both the relationship between the potential supply capacity of ESs and human needs, as well as the relationship between the actual realization of ESs and human needs. This is because even when potential supply meets the demand, the flow (actual supply) may not necessarily meet the demand. A higher value indicates a higher level of the ES flowing from the natural system to human demand, which contributes more to human well-being. If the combined value is greater than 2, it indicates that the supply-flow-demand levels of that ES can meet human demand, contributing to better maintenance and improvement of human well-being.

$${Level}_{S-F-D}={Ratio}_{S-D}+{Ratio}_{F-D}$$
(3)

Finally, landscape sustainability is characterized by considering the supply-flow-demand levels of various ESs. From the perspective of strong sustainability, it is not appropriate to simply add the levels of supply-flow-demand of various ESs, as this would allow for substitution. Therefore, we attempt to represent different dimensions of landscape sustainability by considering the supply-flow-demand levels of various services.

Assessment results

Temporal and spatial characteristics of carbon sequestration service supply-flow-demand

In terms of time (Table 2), from 2000 to 2020, the overall supply and flow of carbon sequestration services in Yixing city were far from sufficient to meet local demand. Furthermore, the supply-flow-demand levels of carbon sequestration services have been declining year by year, decreasing from 0.47 in 2000 to 0.16 in 2020. The potential supply and flow of carbon sequestration services have slightly increased, but the demand for carbon sequestration has increased significantly, leading to a reduction in the supply-flow-demand levels of carbon sequestration services in Yixing city.

Table 2 The supply-flow-demand levels of various ESs in Yixing from 2000 to 2020

In terms of space (Fig. 5), the supply-flow-demand levels of carbon sequestration in various townships in Yixing city have all experienced varying degrees of decline. The most significant declines occurred in the eastern townships of Fangqiao, Zhoutie, and Xinzhuang, as well as the northwestern township of Yangxiang. However, the townships of Taihua and Hufu in the southwest showed significantly higher supply-flow-demand levels compared to other townships.

Fig.5
figure 5

The supply-flow-demand level of ESs in Yixing city

Temporal and spatial characteristics of food production service supply-flow-demand

Over the period from 2000 to 2020 (Table 2), the overall supply of food production services in Yixing city was able to meet local demand, with the supply-flow-demand level shows a continuous rising trend.

The potential supply and flow of food production services have shown a continuous growth trend. However, the demand for food production experienced a significant decline in 2010, mainly due to changes in residents' food consumption patterns, leading to a substantial decrease in the demand for staple foods.

In terms of space (Fig. 5), most townships in Yixing city have maintained stable supply-flow-demand levels for food production services, with some townships in the central and southern regions having lower supply-flow-demand levels than others. However, there has been some improvement, such as in the central townships of Xinjie and Qiting, as well as the southern townships of Taihua, Hufu, and Dingshu.

Temporal and spatial characteristics of nature-based recreation service supply-flow-demand

In terms of time (Table 2), the overall supply-flow-demand levels of nature-based recreation services in Yixing city consistently remained relatively high, despite a slight decrease. Potential supply and flow greatly exceeds demand. The supply of nature-based recreation services initially increased and then decreased, while the demand for these services continued to increase.

In terms of space (Fig. 5), except for some townships in the central and northern regions, the supply-flow-demand levels of nature-based recreation services in Yixing city generally remained at high levels and stable. There have been varying degrees of decline in the supply-flow-demand levels of nature-based recreation services in central townships such as Qiting and Fangqiao, as well as in the northeastern township of Wanshi. However, the supply-flow-demand levels of nature-based recreation services in the newly established township of Xinjian in the north improved to some extent.

Temporal and spatial characteristics of water provision service supply-flow-demand

In terms of time (Table 2), the supply-flow-demand levels of water provision services in Yixing city have seen some improvement. Potential supply and flow has generally been able to meet local demand. Specifically, the potential supply and flow of water provision services steadily increased over the 20-year period, but the demand for water provision services initially increased and then decreased, with the demand in 2020 being lower than that in 2000.

In terms of space (Fig. 5), the supply-flow-demand levels of water provision services in the central, western, and northern townships of Yixing city are lower than those in the southwestern and eastern regions but have shown varying degrees of improvement. For example, the supply-flow-demand levels of water provision services in Heqiao, Wanshi, and Qiting have improved significantly, while in Xizhu, Xushe, Guanlin, Xinjie, and Gaocheng, the improvement in the supply-flow-demand levels of water provision services has been relatively small.

Temporal and spatial characteristics of water purification service supply-flow-demand

In terms of time (Table 2), the supply-flow-demand levels of water purification services in Yixing city have gradually improved, but the potential supply and flow still cannot meet local demand. Specifically, the supply of water purification services has been gradually decreasing, and the demand for water purification has been gradually decreasing.

In terms of space (Fig. 5), the supply-flow-demand levels of water purification services in the southwestern townships of Yixing city are generally higher than those in the central and eastern regions. The townships of Taihua and Hufu in the southwest have the highest levels of supply-flow-demand for water purification services and have seen some improvement. The townships of Yangxiang and Xinjian and the central townships of Gaocheng, Xinjie, and Yicheng have relatively stable levels of supply-flow-demand for water purification services. In the northeastern and eastern regions, the townships of Heqiao, Wanshi, Qiting, Zhoutie, and Xinzhuang have all experienced improvements in the supply-flow-demand levels of water purification services.

Analysis of the landscape sustainability level in various townships of Yixing City

Due to space limitations, we selected only one township from eastern, central, and western Yixing city to showcase and analyze the results of township landscape sustainability (Fig. 6). The selected townships include Zhoutie, Yicheng, and Xushe. From 2000 to 2020, landscape sustainability in terms of carbon sequestration and water purification in Zhoutie was significantly at the lowest level and below the threshold value of 2. However, landscape sustainability in terms of Nature-based recreation and food production was at a higher level, far exceeding the threshold value of 2. The landscape sustainability in terms of the water provision remained relatively stable around the threshold value, with some improvement.

Fig.6
figure 6

The landscape sustainability level in various townships of Yixing city

From 2000 to 2020, Yicheng had a relatively low level of landscape sustainability in all ESs, with only the Nature-based recreation exceeding the threshold value. The landscape sustainability in other dimensions was below the threshold value, but there was some improvement in the water provision and water purification.

From 2000 to 2020, Xushe had a relatively high level of landscape sustainability in food production, Nature-based recreation, and water provision. However, the landscape sustainability in terms of carbon sequestration and water purification was significantly below the threshold value.

Discussion

Operability and contributions of the framework

This study characterizes landscape sustainability through the relationships between ES supply, flow, and demand, applying the framework to the landscape of townships in Yixing city as a case study. The sustainability of township landscapes comprises multiple dimensions, meaning that landscape sustainability is a comprehensive reflection of various states of ES supply, flow, and demand. Low levels of supply-flow-demand for carbon sequestration and water purification services impose the most significant constraints on landscape sustainability. Although the supply-flow-demand levels for recreational services are the highest, they cannot compensate for or replace the negative impact of carbon sequestration and water purification services on landscape sustainability. We should strive to enhance key ESs so that they can meet local needs and improve human well-being.

The final results of landscape sustainability levels represent the state relationship between the supply, flow, and demand of ESs, which reflects the relative contribution of ESs to human well-being. Of course, the level of landscape sustainability is also influenced by landscape structure and ecological processes. After assessing the level of landscape sustainability, we can take relevant management measures to optimize the landscape pattern and enhance the degree to which ESs meet human well-being, thereby improving landscape sustainability. Furthermore, it is crucial to exercise caution when selecting key ESs in Step 1 of the assessment. Because the level of landscape sustainability assessed and the subsequent measures taken to enhance landscape sustainability are both influenced by the selected ESs.

Compared to other landscape sustainability assessment methods, this study's assessment is based on an analysis of the essence of landscape sustainability, more fully reflecting the important content of ESs, human needs, and well-being in landscape sustainability (Wu , 2013, 2021). Compared to single-type landscape sustainability assessments (such as agricultural landscape sustainability in Mexico or coastal urban landscape sustainability in China) (Eichler et al. 2020; Wang et al. 2020), this study's landscape sustainability assessment is suitable for comprehensive regional landscape sustainability. The framework is also applicable to landscape sustainability assessments at different scales, with scale size being determined based on the research area (Metzger et al. 2021), making it highly replicable. In comparison to sustainability assessments based on sustainability development indicators (Huang et al. 2016), this framework requires indicators for critical ES types in the study area, and the assessment of ESs can make use of many existing mature assessment models (Wood et al. 2018) with readily available data, making it highly operational.

Limitations of the framework and future research directions

The landscape sustainability assessment framework proposed in this study aims to enrich the theory and methods of landscape sustainability assessment, but it still has several limitations. First, the definition of landscape boundaries is not sufficiently clear. In the case study, township landscapes were considered relatively independent entities, primarily for the sake of assessment convenience and data availability. However, in reality, landscape boundaries are often difficult to define clearly and influence the material and energy exchange between the interior and exterior of the landscape (Bluemling et al. 2021). Second, the framework did not fully consider the impact of landscape structure and processes on flows, and it only considered potential ES flows within the landscape toward local human demands (Assis et al. 2023). It overlooks the exchange of materials and energy between the landscape and the external environment, simplifying the quantification and implementation of flows, which could potentially lead to overestimations in assessment results (Wang et al. 2022). Third, the case study only selected the township landscape scale, neglecting other scales. In practice, the choice of scale significantly affects landscape patterns, processes, and the supply of ESs (Mitchell et al. 2015a; Metzger et al. 2021). Fourth, the framework did not consider the trade-off/synergy effects between key ESs, which also have an important impact on landscape sustainability (Karimi et al. 2020; Jafarzadeh et al. 2021), and the quantitative indicators and formulas for landscape sustainability are also not perfect and require further consideration. Therefore, in future research, we will further focus on the influence of landscape structure and processes on the supply, flow, and demand of ESs, giving due consideration to the impact of trade-off/synergy effects between key ESs on landscape sustainability. Clear delineation of landscape boundaries and the assessment of landscape sustainability at different scales will also be necessary.

Conclusion

This study introduced a landscape sustainability assessment framework based on the supply-flow-demand of ESs and applied it in Yixing city. The main conclusions are as follows: (1) Landscape sustainability in various townships of Yixing city is primarily constrained by carbon sequestration services and water purification services. To achieve comprehensive landscape sustainability, it is essential to maintain landscape sustainability in all dimensions above a critical level. (2) The landscape sustainability assessment method based on the supply-flow-demand of ESs focuses more on the essence of landscape sustainability and its core concepts. It is mainly applicable to regional landscape sustainability assessments rather than specific types of landscapes. This assessment framework can be applied across various temporal and spatial scales, utilizing mature models and methods with readily available data, making it highly operational. (3) The framework presented in this study still has certain limitations. Future landscape sustainability assessments should pay more attention to defining landscape boundaries, understanding the influence of landscape structure and processes on ES supply-flow-demand, considering assessments at multiple scales, and accounting for the trade-off/synergy effects between different types of ESs. In conclusion, this research contributes to a deeper understanding of landscape sustainability through the lens of ESs and provides a versatile framework for assessing landscape sustainability in various contexts. It highlights the importance of considering multiple dimensions of sustainability and the need for further research to address the identified limitations and enhance the applicability of such frameworks.