Introduction

Urbanization has entered a new phase across the globe, with its impacts extending far beyond local urban areas (Wu 2022). Rural and urban territories are increasingly interconnected and interdependent due to globalization, digitalization, and environmental challenges (Kratzer and Kister 2020). Urban systems rely on rural systems for essential resources such as food, energy, water, and other materials, while rural systems also depend on urban systems for markets, manufactured goods, and medical resources (National Science Foundation 2021). The importance of rural areas in fostering regional sustainable development of the coupled urban-rural system has become increasingly evident as the mutual connections between urban and rural areas have deepened (He et al. 2020).

The pursuit of regional sustainable development requires the common prosperity of urban and rural areas. Urban-rural integration is a process of viewing cities and the countryside as an organic entities, facilitated by the flow of elements between them, the optimization of their spaces, the complementarity of their functions, and the innovations within their institutions. This process aims to promote the integration and harmonious coexistence of urban and rural economies, societies, spaces, public services, and social management. Ultimately, this process seeks to bridge the urban-rural gap and achieve common prosperity (He et al. 2017). Urban-rural integration reshapes the socioenvironmental relationships within these areas through the free flow and rational allocation of resources such as population, capital, and technology (He et al. 2022). This integration supports mutual benefits and sustainable development for both urban and rural areas by fostering economic growth, alleviating poverty, enhancing social cohesion, promoting cultural diversity, and conserving ecological systems (Organisation for Economic Co-operation and Development 2013; Akkoyunlu 2015; Gebre and Gebremedhin 2019; Forster et al. 2021; He et al. 2021). However, it may also cause pressure on natural resources such as water and land because of the increased population density in some areas and the emergence of suburban regions (Kapfudzaruwa et al. 2018). Consequently, prioritizing urban-rural integration is crucial for tackling regional and global sustainable development challenges.

Landscapes are geospatially heterogeneous, socioeconomically driven, regional human-environment systems (Wu 2013). They are considered the comprehensive spatial and functional entities of natural and cultural living spaces (Naveh 1991). As coupled human-environmental systems, landscapes offer an effective avenue for understanding the impacts of urban-rural integration on regional spatial structure and the eco-environment. Landscape patterns, as spatial representations of the interactions between people and the environment, play crucial roles in connecting ecosystems and social systems (Wu 2021). These interactions are always driven by various socioeconomic, political, technological, natural, and cultural forces (Bürgi et al. 2004; Wu 2021). Urban-rural integration involves societal convergence, economic interaction, technology sharing, and spatial interlinkage, which occur within specific environmental and institutional contexts (He et al. 2021). The transformations brought about by urban-rural integration have direct or indirect impacts on landscape patterns, affecting ecosystem services, human well-being, cultural identity, and regional competitiveness (Wu 2013), thereby influencing landscape sustainability. Therefore, exploring how urban-rural integration influences landscape patterns is essential for understanding the role of urban-rural integration in regional sustainable development.

However, there is currently limited direct research on the impacts of urban-rural integration on landscape patterns. Existing studies have primarily focused on the effects of urbanization on landscape patterns and ecological processes (Luck and Wu 2002; Grimm et al. 2008), with severe negative impacts observed on resources, the environment, and biodiversity (Bai et al. 2017; McDonald et al. 2013). Although landscape gradient analysis has detected landscape sequences along the urban-suburban-cultivated-managed-natural gradient, focusing on differences between urban and rural landscapes (Forman and Godron 1986; Luck and Wu 2002; Hahs and McDonnell 2006), this issue has not been thoroughly explored from the perspective of coupled urban-rural systems. Additionally, the ecological landscape degradation and fragmentation caused by the expansion of urban-rural transition zones reflect the landscape changes resulting from urban-rural spatial integration (Tian 2020). While recent years have seen increased recognition of the significant impacts of urban-rural interactions on landscapes (Chang et al. 2022; Benoît et al. 2012), the impacts of urban-rural integration on landscape dynamics and sustainability remain understudied.

Therefore, our study aims to examine the impacts of urban-rural integration on landscape patterns and explore its implications for landscape sustainability. Metropolitan areas, recognized as a suitable decision-making scales for promoting sustainability, provide a holistic view of the interaction between urban and rural areas (He et al. 2019) and serve as a pivotal scales for studying landscape patterns (Luck and Wu 2002). Thus, we conduct a case study of the Changsha metropolitan area to address the following questions: (1) What are the spatiotemporal patterns of urban-rural integration and landscape pattern dynamics within the study area? (2) What are the main drivers and processes in urban-rural integration responsible for the observed changes in landscape patterns? (3) What are the potential implications of landscape pattern changes with urban-rural integration for landscape sustainability?

Materials and methods

We examine the impacts of urban-rural integration on landscape patterns by constructing an assessment framework, mapping spatiotemporal patterns, and conducting regression analysis, using the Changsha metropolitan area as the research area (Fig. 1). The technical roadmap is shown in Fig. 2.

Fig. 1
figure 1

Location of the Changsha metropolitan area, in northeastern Hunan Province of south-central China

Full size image
Fig. 2
figure 2

Technical roadmap of this study

Full size image

Study area

The study area, denoted as the Changsha metropolitan area within Fig. 1, is demarcated by a one-hour commuting radius. It comprises 15 districts and 9 county-level administrative units, including the entire municipal areas of Changsha and Xiangtan, as well as parts of Zhuzhou, Yueyang, and Yiyang, in the Hunan Province of central China, covering an area of approximately 14,071 square kilometers. It spans longitudes 111°53'E to 114°15'E and latitudes 27°12'N to 29°03'N, gradually transitioning from a low-lying alluvial plain to hilly areas outward from the Xiangjiang River, with an average elevation of approximately 92 meters. The climate falls within the subtropical monsoon humid zone, characterized by an annual mean temperature of 18°C and an average yearly precipitation of 1,412 mm. The prevailing vegetation type is mid-subtropical evergreen broad-leaved forest.

The Changsha Metropolitan Area is a significant growth pole in Hunan Province and across China. It is a crucial component of the midstream Yangtze River Urban Agglomeration and is located at the junction of the Beijing-Guangzhou Economic Belt and the Yangtze River Economic Belt. In the national regional coordinated development pattern, it plays a strategic role as a hub that connects east to west and links north to south. In recent years, the area has witnessed a significant acceleration in urban-rural integration, accompanied by rapid economic and social development and strengthened urban linkages. By 2020, the total population had reached 17.868 million, with an urbanization rate of 75.49% and the per capita GDP exceeding 15,000 dollars. The urban-rural disposable income ratio was 1.25. The uniqueness of its geographical location, regional status, and stage of development has shaped its distinctive dynamics of urban-rural transition and contributed to landscape alterations. Due to its characteristics such as multi-centeredness, high density, strong mobility, network structure, and high openness (He et al. 2019), the landscape response process in the Changsha metropolitan area is also relatively evident and complex.

Assessment of urban-rural integration

Construction of a multidimensional indicator system for measuring urban-rural integration

To specifically explore the impacts of human social activities on landscapes, this study mainly focused on assessing the socioeconomic transformations during urban-rural integration. Urban-rural integration is a comprehensive and multidimensional process aimed at achieving coordinated development between urban and rural areas. The mobility of socioeconomic factors such as population, technology, and capital between urban and rural areas, along with the occurrence of related socioeconomic activities, is a key process in promoting the integrated development of urban and rural areas. It primarily involves the process of achieving social equity, economic coordination, technology sharing, and spatial connectivity between urban and rural areas through mutual interaction (He et al. 2019). Although the integration of urban and rural ecological environments is also an important part of it, it is always be attributed to the influences of urban-rural socioeconomic activities. Accordingly, we have selected indicators that reflect socioeconomic interactive processes to objectively describe the patterns and dynamic changes in urban-rural integration in a metropolitan context, such as reducing disparities in the urban-rural income and employment, optimizing industrial structures, improving production efficiency, sharing technology, and enhancing spatial connectivity. According to existing studies, these have always been characterized by indicators such as those reflecting equity in urban-rural employment and public service facilities, the urban-rural labor productivity ratio, the urban-rural industrial added value ratio, agricultural mechanization, internet coverage, and the gap in urban and rural population density, and so on (Ma et al. 2022; Zheng and Long 2023). A set of eight indicators was chosen for our research taking into account both representativeness and data availability (Table 1).

Table 1 Indicator system for assessing the level of urban-rural integration. The attribute of each indicator indicates whether it is positively or negatively related to the level of urban-rural integration, with ‘+’ denoting a positive indicator
Full size table

Integrated urban-rural development hinges on securing fair and equitable progress in both urban and rural areas (Chen and Long 2019). Reducing income disparity between urban and rural areas is essential to promoting social equity and achieving shared prosperity (Liu et al. 2013). Non-agricultural employment as a means of diversifying rural household income significantly influences income levels in rural areas (De Janvry and Sadoulet 2001; Kapfudzaruwa et al. 2018). The flow of labor between urban and rural areas is one of the most dramatic factors defining the rural–urban connection (Sietchiping et al. 2014). Therefore, two indicators were chosen to represent income fairness and balanced employment opportunities between urban and rural areas: the rural-urban per capita disposable income ratio and the proportion of rural labor forces engaged in non-agricultural employment.

Economic integration is a key component in achieving urban-rural development. This paper primarily utilized three indicators: the ratio of fiscal expenditure for agriculture to agricultural GDP, the proportion of high-value agriculture in gross agricultural production, and the rural-urban labor productivity ratio. By increasing investments in rural areas, it is possible to promote rural economic development, optimize rural industrial structures and value chains, and create more employment opportunities (Fan et al. 2002). Fiscal policies supporting agriculture are crucial for industrial sectors to contribute to agricultural development and rural areas (Kuzman et al. 2017). The ratio of fiscal expenditure on agriculture to agricultural GDP reflects both the potential for agricultural economic growth and the commitment of local governments to support such growth financially. Moreover, transforming the agricultural economic structure to meet evolving food consumption needs is vital for coordinated urban and rural economic development (Huang 2022). The change in the proportion of high-value agriculture within overall agricultural production can serve as an effective reflection of this transformation. Additionally, the productivity of agricultural labor is another critical indicator that reflects the extent of agricultural economic transformation. The progressive narrowing of the urban-rural labor productivity gap is a fundamental aspect that characterizes urban-rural integration. This is because only when agricultural labor productivity converges with that of the industrial/service sectors (or labor income), can the disparities in income between urban and rural areas, and between agricultural and non-agricultural sectors, be completely eliminated (Huang 2022).

The technological divide is another significant factor contributing to the urban-rural gap. With the extension of digitalization and informatization to rural areas, the advancement of digital villages and smart agriculture will undoubtedly accelerate significant changes in rural development. This transformation will also serve as a crucial means to narrow the urban-rural gap. However, due to the challenges in obtaining data, this study only utilizes the level of agricultural mechanization to assess the degree of technology sharing between urban and rural areas. Nevertheless, this indicator bears some relevance in regions where the development of rural informatization and intelligentization is still in its early stages.

Spatial connectivity, facilitated by transportation and information networks, promotes the unrestricted movement of elements between urban and rural areas (Yang et al. 2021). Small towns play a crucial role as intermediaries in facilitating urban and rural commerce, logistics, and distribution (He et al. 2019). Therefore, urban-rural road network density and small town density were selected as indicators to assess spatial integration between urban and rural areas.

Data and processing

We analyzed the dynamic change characteristics of urban-rural integration development in the Changsha metropolitan area from 2005 to 2020 using the selected indicators. Previous studies have often employed a multi-indicator composite evaluation method to assess the comprehensive level and change characteristics of urban-rural integration development (Li 2012; Yang et al. 2021; Zheng and long 2023). In fact, the composite values often conceal the information reflected by indicators across different dimensions, such as social, economic, and spatial aspects. Therefore, to prevent potential loss of information during indicator integration, we conducted an independent analysis on different urban-rural integration indicators. This approach allows us to more comprehensively and accurately analyze the impacts of the urban-rural integration process across different dimensions on landscape patterns.

The period from 2005 to 2020 was specifically chosen for the spatial mapping assessment, primarily because China initiated the construction of the New Socialist Countryside in 2005. This initiative, aimed at promoting industry feeding back into agriculture and urban areas supporting rural areas, has accelerated the development of rural production within the Changsha Metropolitan Area, increased farmers' income, and facilitated a substantive transformation in urban-rural relations. Subsequent analyses for the years 2010, 2015, and 2020 not only considered the availability of data but also aligned with the national and local economic and social development planning cycles, thereby reflecting, to some extent, the policy impacts on urban-rural integration development.

County (district)-level administrative units were selected as the analysis unit to reveal spatial heterogeneity. This decision was influenced by several crucial factors. The urban-rural system is inherently a multiscale complex system, encompassing global, national, regional, and local dimensions, with functions and characteristics varying across these scales. At the county (district) level, the urban-rural system represents a fundamental and pivotal component of the broader urban-rural framework. Counties (districts) serve as geographical entities endowed with complete political, economic, social, and cultural functions, as well as abundant resources. They possess a degree of autonomy in fiscal management and resource allocation in China, making them critical links in the ongoing process of equitable exchange and bidirectional movement of urban and rural elements. Consequently, conducting comparative analyses at the county (district) level allows for a more accurate reflection of the impacts of economic and social contexts, coupled with policy backgrounds, on urban-rural integration. This approach also facilitates a deeper understanding of their corresponding landscape responses, contributing to the advancement of both urban-rural social governance and landscape management.

The social-economic data required for the analysis were collected from various sources, including China County Statistical Yearbook, China Agricultural Product Price Survey Yearbook, Hunan Statistical Yearbook, Hunan Rural Statistical Yearbook, and Hunan Survey Yearbook. In instances of missing values, the data were estimated via interpolation based on the annual average growth rate of the available data. The data for the Furong District in 2020 were excluded from the analysis because its urbanization level had already reached 100%.

Investigation of landscape patterns

Landscape classification and identification

In academic research, landscape classification is typically based on land use types, landscape functions, the appearance of macro-landforms, and the degree of human interference in natural landscapes (Forman 1995; Bastian 2000; Brabyn 2009; Wu 2013). To comprehensively assess dominant functions and human disturbances, particularly in distinguishing between urban and rural spatial categories, we classified landscapes into three types: natural landscape (NL), agricultural landscape (AL), and constructed landscape (CL). However, this classification may not fully capture the multifunctional and complex nature of landscapes, particularly those in transitional states. To address this limitation and enrich the understanding of landscapes, we further divided each primary category based on specific land use types. Natural landscape includes woodland, shrubland, sparse woodland, other woodland, grassland, and water body. Agricultural landscape was divided into paddy and dryland. Constructed landscape encompasses urban construted landscape, rural constructed landscape, industrial land, and traffic facility land. These subcategories were used to calculate the Shannon diversity index at the landscape level. Moreover, discussions and analyses concerning landscape sustainability are grounded in acknowledging the interconnectedness and multifunctionality inherent in landscapes.

Landscape type identification is primarily based on land use and land cover data obtained from China's National Land Use and Cover Change (CNLUCC) dataset (Xu et al. 2018). Four periods of LUCC data were selected, namely 2005, 2010, 2015, and 2020, with a resolution of 30 meters. It is important to note that this dataset's granularity is insufficient for accurately representing narrow, low-level roads accurately. To determine the significant impacts of road traffic on landscape patterns (Fu et al. 2010; Mo et al. 2017), especially as an indicator of urban-rural spatial connectivity, graded road data (http://www.openstreetmap.org) were integrated with land use data using ArcGIS software. To retain more detailed landscape pattern information and investigate the influence of road networks, the merged land use type data were converted into a grid map with a resolution of 10 meters (Fig. 3).

Fig. 3
figure 3

(a) 2005, (b) 2010, (c) 2015 and (d) 2020 land-use/land-cover maps of the Changsha metropolitan area

Full size image

Selection of landscape metrics

Landscape metrics offer insights into the spatial configuration and structural composition of the landscapes (Wu 2007; Fu et al. 2021). The selection of these metrics is typically guided by specific research objectives (Abdullah and Nakagoshi 2006; Aguilera et al. 2011). In this study, our objective goes beyond simply quantifying dynamics in landscape patterns. Instead, we aim to investigate the impacts of urban-rural integration on the landscape.

The effects of the urban-rural integration process on natural, agricultural, and constructed landscapes vary. To better elucidate the variations in the patterns of different landscape types during the urban-rural integration process, we employed a set of landscape metrics to examine class-level properties (Table 2). The dynamics of percentage of landscape (PLAND), patch density (PD), mean shape index (MSI), and connectance index (CONNECT) were used to evaluate the responses of agricultural and natural landscape patterns to urban-rural integration. Changes in rural constructed landscape (RCL) were mainly observed through patch density (PD) and the mean shape index (MSI). To measure the changes in connectivity of the urban and rural constructed landscape during the process of urban-rural integration, the urban constructed landscape and rural constructed landscape were combined as urban and rural constructed landscape (URCL) to calculate the connectance index (CONNECT). At the landscape level, we employed the Shannon diversity index (SHDI).

Table 2 Landscape metrics employed in this study (McGarigal and Marks 1995; McGarigal 2015). All the landscape metrics are calculated by FRAGSTATS 4.2. The threshold distance for the connectance index was established at 400 m, as it roughly corresponds to the radius of the rural community life circle and the previous studies have demonstrated that when the threshold distance is set at 400 m, the calculated results for landscape connectance are highly representative (Shanghai Municipal Planning and Natural Resources Bureau 2021; Du et al. 2018; Liu et al. 2021)
Full size table

Time-fixed effects model for examining impacts of urban-rural integration on landscape patterns

The fixed effects model is a commonly used method for inferring causality in panel data analyses. Scholars always utilize this model to adjust for unobserved inherent individual or temporal attributes that have confounding effects (Imai and Kim 2021). By introducing dummy variables and control variables, the fixed effects model helps eliminate the remaining factors that affect dependent variables, enabling more accurate estimation of causality for the focal variables (Hu 2020). In this paper, we used the fixed effects model to elucidate the causal relationship between urban-rural integration and changes in landscape patterns. The time-fixed effects model was used to control time-influencing factors that do not change with individuals, with dummy variables introduced to capture the changes in these factors. In addition, factors such as urban-rural integration, geographical location, topography, urbanization, and economic development have been known to influence landscape patterns (Turner 1989; Weng 2007; Li et al. 2017). Therefore, in exploring the causal relationship between urban-rural integration and landscape patterns, we incorporated elevation, slope, geographic proximity to the central urban area of Changsha, urbanization rate, and GDP as control variables to mitigate interference from these factors.

The panel data regression model was established as follows:

$${Landscape\, Pattern\, Index}_{it}={\beta }_{0}+{\sum }_{j=1}^{n}{\beta }_{j}*{Urban-Rural\, Integration \,Indicators}_{jit}+{\sum }_{k=1}^{p}{\gamma }_{k}*{control\, variable}_{kit}+{\mu }_{t}+{\varepsilon }_{it}$$

where i and t refer to the county and year, respectively, j represents the jth urban-rural integration indicator, with a total of n such indicators, k represents the kth control indicator, with a total of p control indicators, \({\beta }_{0}\) and \({\beta }_{j}\) represent the intercept term and the coefficient value respectively, \({\mu }_{t}\) captures the time-fixed effects and \({\varepsilon }_{it}\) is the random error.

Research has demonstrated that panel unit root tests exhibit reduced power when the time dimension (T) is small (Gengenbach et al. 2009). Moreover, a significant risk exists that the entire panel might be incorrectly classified as nonstationary, even if a considerable number of series within the panel are stationary (Karlsson and Löthgren 2000; Chen 2014). Given that the data we studied are short panel data (T=4), we refrained from utilizing the unit root test. To address the heteroscedasticity and serial correlation in the data, we utilized clustering robust standard errors to replace the original standard errors.

Results

Dynamic changes in urban-rural integration from 2005 to 2020

Based on the dynamic analysis of urban-rural integration indicators, it was evident that the Changsha metropolitan area experienced a significant increase in its level of urban-rural integration from 2005 to 2020. This increase was observed across society, economic, technology, and space dimensions (Fig. 4). Most indicators demonstrated significant growth, with both the mean and median values increasing, except for the rural-urban labor productivity ratio. The mean and median values of the rural-urban labor productivity ratio decreased from 0.30 and 0.27 in 2005 to 0.21 and 0.19 in 2020, indicating an expanding overall disparity in labor productivity between urban and rural areas. This trend was attributed to the growth rate of urban productivity significantly outpacing that in rural regions.

Fig. 4
figure 4

Spatiotemporal changes in urban-rural integration, represented by boxplots, which illustrate the characteristics of the data distribution and the temporal changes of each indicator. The boxplots display the median (indicated by the line in the blue box), the average (represented as a square), the upper and lower quartiles (shown within the box), the range between the maximum and minimum values (illustrated by whiskers), and any outliers (depicted as individual points). The abbreviations used in the figure are consistent with Table 1

Full size image

Upon analyzing the pattern maps (Fig. 5), a center-periphery distribution pattern was quite distinct for most indicators. It became evident that certain indicators demonstrated notably higher levels within the municipal districts of Changsha, Zhuzhou, and Xiangtan, such as the ratio of rural-urban per capita disposable income, the proportion of high-value agriculture in gross agricultural production, and the rural labor forces' non-agricultural employment proportion. Conversely, indicators such as urban-rural labor productivity and agricultural mechanization degree exhibited opposite trend. Moreover, the urban-rural road network density was evolving from a single-center pattern to a multicenter pattern. As opposed to the center-periphery, areas with high values of the ratio of fiscal expenditure for agriculture to the agricultural GDP and small town density showed a distinct distribution pattern along the Xiangjiang River. Overall, the process of urban-rural integration exhibited a significant core-periphery spatial pattern as a whole.

Fig. 5
figure 5

Spatiotemporal changes in urban-rural integration, represented by pattern diagrams for (a) 2005, (b) 2010, (c) 2015 and (d) 2020. These diagrams illustrate the heterogeneity of the spatial patterns for each urban-rural integration indicator and the changes in these patterns from 2005 to 2020. The abbreviations used in the figure are consistent with Table 1

Full size image

Dynacmic changes in urban and rural landscape patterns from 2005 to 2020

Based on the calculation results of the landscape metrics the percentage of natural and agricultural landscapes in the Changsha metropolitan area had decreased significantly, accompanied by a noticeable decline in the mean shape index and connectance index, while patch density has increased. Specifically, the percentage of agricultural and natural landscapes decreased, with average values declining from 38.80 and 51.61 in 2005 to 33.28 and 48.50 in 2020, respectively, with a particularly pronounced decrease in the central area. Except for Furong District, all other districts and counties experienced varying degrees of increases in agricultural and natural landscape patch density, while the mean shape index and connectance index have generally decreased. Specifically, the average patch density for agricultural and natural landscapes increased from 0.32 and 0.40 in 2005 to 0.45 and 0.51 in 2020, respectively. The average mean shape index for agricultural and natural landscapes decreased from 2.21 and 1.93 in 2005 to 2.06 and 1.82 in 2020, respectively. The average connectance index for natural landscape decreased from 2.51 in 2005 to 1.98 in 2020, while the connectance index for agricultural landscape increased slightly due to rapid growth in Furong District.

However, indicators such as patch density, the mean patch shape index, and connectance have also revealed increasing intensification in urban and rural constructed landscape, and landscape diversity has gradually strengthened. More than half of the counties and districts have experienced a decrease in patch density and the mean shape index for rural constructed landscape, with their average values decreasing from 0.24 and 1.54 in 2005 to 0.22 and 1.52 in 2020, respectively. Three-quarters of the counties and districts experienced an increase in the connectance index for urban and rural constructed landscape, with their average value increasing from 2.18 in 2005 to 3.50 in 2020. Furthermore, the Shannon diversity index has steadily increased, with the average value increasing from 1.55 in 2005 to 1.67 in 2020 (Figs. 6 and 7).

Fig. 6
figure 6

Spatiotemporal changes in landscape patterns, represented by the boxplots, which illustrate the characteristics of the data distribution and the temporal changes in each indicator. The boxplots display the median (indicated by the line in the blue box), the average (represented as a square), the upper and lower quartiles (shown within the box), the range between the maximum and minimum values (illustrated by whiskers), and any outliers (depicted as individual points). The abbreviations used in the figure are consistent with Table 2

Full size image
Fig. 7
figure 7

Spatiotemporal changes in landscape patterns, as depicted by pattern diagrams for (a) 2005, (b) 2010, (c) 2015 and (d) 2020. These diagrams illustrate the heterogeneity of the spatial patterns for each landscape metric and the changes in these patterns from 2005 to 2020. The abbreviations used in the figure are consistent with Table 2

Full size image

Influence of urban-rural integration on landscape patterns

By employing the time-fixed effects model, we revealed the impacts of urban-rural integration indicators on landscape metrics in the Changsha metropolitan area, with detailed outcomes presented in Fig. 8 The results indicated a significant influence of urban-rural integration indicators on landscape metrics. Moreover, different urban-rural integration indicators exhibited varying effects on various landscape metrics. A detailed analysis is provided below.

Fig. 8
figure 8

Estimation results of the time-fixed effects model using Stata 16.0, based on panel data from 24 counties (districts) and 4-time points, with urban-rural integration indicators as independent variables and landscape metrics as dependent variables. The regression coefficients of the urban-rural integration indicators to a significant landscape metric, based on the t-test, are presented in each column, with significance levels denoted as follows: *p<0.1, **p<0.05, ***p<0.01. Red-toned colors signify a significant positive correlation, blue-toned colors denote a significant negative correlation, and white indicates no significant correlation. The abbreviations used in the figure are consistent with Tables 1 and 2

Full size image

Urban-rural integration had a significant impact on the percentage of landscape, especially influenced by factors such as the ratio of fiscal expenditure for agriculture to agricultural GDP, the proportion of high-value agriculture in gross agricultural production, the degree of agricultural mechanization, and urban-rural road network density. The analytical findings demonstrated that most urban-rural integration indicators, such as the ratio of fiscal expenditure for agriculture to agricultural GDP, the degree of agricultural mechanization, the rural-urban per capita disposable income ratio and urban-rural road network density, had a significantly negative effect on the percentage of agricultural and natural landscapes. However, the proportion of high-value agriculture had a significantly positive effect on the percentage of agricultural and natural landscapes. Specifically, a one-unit increase in the proportion of high-value agriculture was associated with a 13.29-unit increase in the percentage of the agricultural landscape and a 9.16-unit increase in the percentage of the natural landscape. Notably, a one-unit increase in small town density was linked to a remarkable 132.72-unit increase in the percentage of the agricultural landscape.

The effects of urban-rural integration on the mean shape index of natural and agricultural landscapes were also evident, primarily in relation to increased rural investments, rural road construction, and small town growth. The analysis revealed that the ratio of fiscal expenditure for agriculture to agricultural GDP had a significant positive impact on the mean shape index of agricultural landscape, as did small town density on agricultural and natural landscapes. For example, an increase of one unit in the ratio of fiscal expenditure for agriculture to agricultural GDP or small town density resulted in a corresponding increase of 0.17 units or 6.07 units in the mean shape index of agricultural landscape. In contrast, urban-rural road network density and the rural-urban per capita disposable income ratio exhibited a significant negative impact on the mean shape index of natural landscape.

Most of the urban-rural integration metrics had significant impacts on the landscape connectance indexes, and they generally exhibited a positive correlation. For the natural landscape connectance index, the small town density, the proportion of high-value agriculture, the ratio of fiscal expenditure for agriculture to agricultural GDP, and the rural-urban labor productivity ratio played significant roles. Particularly, for every one-unit increase in small town density, the natural landscape connectance index increased by 20.68 units. For every one-unit increase in the proportion of high-value agriculture, the natural landscape connectance index increased by 1.59 units. The agricultural landscape connectance index was notably impacted at the 1% significance level by both the ratio of fiscal expenditure for agriculture to agricultural GDP and the degree of agricultural mechanization, while its influence from small town density was observed at the 5% significance level. The urban and rural constructed landscape connectance index was primarily impacted by the ratio of fiscal expenditure for agriculture to agricultural GDP, the level of agricultural mechanization, the rural-urban per capita disposable income ratio, and the urban-rural road network density.

Urban-rural integration significantly affected the fragmentation of rural constructed landscape, agricultural landscape, and natural landscape, with the most notable impact observed in the case of natural landscape. The data revealed that an increase in the rural labor forces' non-agricultural employment proportion and the rural-urban per capita disposable income ratio leads to a more pronounced trend of fragmentation in natural and agricultural landscapes. For instance, a rise of 1 unit in urban-rural road network density was associated with an increase of 0.65 units in natural landscape fragmentation, as indicated by its significance at the 1% level. On the other hand, the ratio of fiscal expenditure for agriculture to agricultural GDP and the rural-urban labor productivity ratio exerted a significant negative influence on fragmentation in natural, agricultural, and rural constructed landscapes. Furthermore, it is important to note that agricultural mechanization demonstrated a significant and negative correlation with fragmentation in both agricultural and rural constructed landscapes. Additionally, small town density was significantly negatively associated with fragmentation in natural landscape, showing a negative relationship.

Additionally, urban-rural integration significantly influenced landscape diversity. On the one hand, the non-agriculturalization of rural labor and increased incomes among households contributed to enhancing landscape diversity. To illustrate it, a one-unit increase in the rural-urban per capita disposable income ratio corresponded to a 0.13-unit increase in landscape diversity, while a one-unit increase in the proportion of rural labor forces engaged in non-agricultural employment resulted in a 0.14-unit increase in landscape diversity. On the other hand, agricultural mechanization, the development of high-value agriculture, and increased rural investments limited landscape diversity.

Discussion

How urban-rural integration affects the landscape patterns

Urban-rural integration constitutes a transformative process with profound implications for landscape composition and structure. Our research in the Changsha metropolitan area has shed light on the diverse urban-rural integration indicators and their distinct influences on various landscape metrics. These findings, together with other previous studies, revealed the complex factors and processes behind the observed alterations in landscape patterns, which may vary in different regional settings.

By analyzing the impacts of urban-rural integration on the percentage of landscape within the study area, it was discovered that the development of high-value agriculture plays a crucial role in stabilizing agricultural and natural landscapes. Additionally, an increase in small town density was identified as a potent deterrent against the reduction of agricultural landscape. In contrast, factors such as increased investments in agriculture, enhanced traffic, improved agricultural mechanization, and a narrowed urban-rural income gap were associated with a decrease in both agricultural and natural landscapes. Existing researches have also indicated that increased mechanization and the subsidized intensification of agriculture contribute to the loss of semi-natural habitats in Britain (Robinson and Sutherland 2002), and declines in vegetated areas and forest coverage occur alongside an increase in income in eastern China (Su et al. 2014).

The analysis of urban-rural integration's impacts on landscape fragmentation has revealed several key insights in our study. Increased investments in agriculture, enhanced rural labor productivity, and progress in agricultural mechanization have collectively served as effective measures against excessive landscape fragmentation. Moreover, an increase in small town density contributed to a reduction in fragmentation in natural landscape. However, the non-agriculturalization of labor and the intensification of urban-rural road networks, while fostering higher rural residents' income, exacerbate the fragmentation of both agricultural and natural landscapes. Additional papers have also confirmed that mechanization often leads to the consolidation of small, fragmented farmland into large-scale farms (Liao et al. 2022). Conversely, when households become less reliant on agriculture, farmland is more likely to become fragmented (Wang et al. 2020). Additionally, road have been identified as a major contributor to habitat loss, fragmentation, and degradation (Van Der Ree et al. 2011).

The impacts of urban-rural integration on the mean shape index in our research underscored that an increase in rural household incomes has led to a diminishing complexity of agricultural and natural landscape patches. Conversely, the growth of small towns  and rural investments has been beneficial for preserving the complexity of habitat patch shapes. This is consistent with the view that landscapes altered by resource exploitation or development tend to exhibit patches with simpler shapes (Krummel et al. 1987; Forman 1995). Additionally, the influence of urban-rural integration on the connectance of landscapes was analyzed, concluding that factors such as rural investment, enhanced rural labor productivity, agricultural mechanization, development in high-value agriculture, and small town growth have played crucial roles in promoting the connectance of agricultural and natural landscapes. Simultaneously, agricultural mechanization, traffic improvement, rural residents' income enhancement and rural investments contribute to enhancing the connectance of urban and rural constructed landscape. Previous studies have noted that urban-rural connectivity in Asia is often linked to increases in overall agricultural inputs and outputs. Enhancing rural-urban connectivity could theoretically influence agricultural sustainability by boosting farm productivity in urbanizing regions (Boudet et al. 2020).

Furthermore, the influences of urban-rural integration on landscape diversity in the studied region revealed that enhancing non-agricultural employment and increasing incomes contributed to an increase in landscape diversity. However, factors such as agricultural mechanization, the development of high-value agriculture, and increased rural investments limited landscape diversity. This finding has been confirmed by other studies on the relationships between jobs, incomes, social diversity, agricultural mechanization, and crop diversity (Garibaldi and Pérez-Méndez 2019; Daum et al. 2023; Palacios et al. 2013).

Mechanism of how urban-rural integration impacts landscape pattern

Understanding the mechanisms through which urban-rural integration affects landscape patterns is paramount. During the process of advancing urban-rural integration, various factors and processes, including societal convergence, economic interaction, technology sharing, and spatial interlinkage, may induce diverse changes in landscape patterns.

First, urban-rural economic coordination, encompassing initiatives such as increasing rural investment, reducing the urban-rural productivity gap, and developing high-value agriculture, played a crucial role in controlling reduction, enhancing connectivity, mitigating fragmentation, maintaining the complexity of agricultural and natural landscapes, and preventing the fragmentation of constructed landscapes, despite a decrease in landscape diversity. Investments in agriculture and improvements in agricultural productivity contributed to the expansion of agricultural production, enhancement of farmland quality, and promotion of agricultural specialization and intensification (Frison and International Panel of Experts on Sustainable Food Systems 2016). To some extent, this protected agricultural landscape, improving their regularity and connectivity (Boudet et al. 2020). Landscape changes resulting from adjustments in agricultural structure were also evident under specific investment and production conditions. For instance, focusing on high-value crops allows farmers to alleviate pressure on converting natural landscape, such as forests or wetlands, into agricultural land. Additionally, diversifying crop types and cultivating agroforestry systems have positively impacts on the complexity of agricultural landscape (Asbjornsen et al. 2014; Wilson and Lovell 2016; Sirami et al. 2019). Moreover, when more investments are directed towards village revitalization and infrastructure development, it simultaneously leads to increased intensity of urban and rural construction (Jiang et al. 2013), while also controlling the occupation of construction on agricultural and natural landscapes.

In contrast, societal convergence, characterized by an increase in non-agricultural employment among rural residents and a reduction in the urban-rural income gap, led to a decrease in the proportion, connectivity, and complexity of agricultural and natural landscapes, as well as an increase in landscape fragmentation, while promoting landscape diversity. This occurred because rural households engaged in non-agricultural activities not only directly altered agricultural production methods, such as the abandonment of cultivated land, but also boosted rural household incomes (Li et al. 2021). These changes increased the likelihood of landscape-disturbing activities, such as altering land usage, transitioning it to non-agricultural purposes, and constructing new residences (Wang et al. 2020; Zhang et al. 2016). Additionally, this shift could lead to the adoption of new lifestyles and values, influencing the landscape through changes in consumption patterns, community development, and various other aspects.

Furthermore, the role of agricultural mechanization in reducing landscape fragmentation and enhancing landscape connectivity, coupled with its contribution to the decline in landscape diversity, exemplified the bidirectional influence of technology sharing on landscapes during urban-rural integration. The agricultural mechanization process often requires an increase in the concentration and contiguity of farmland through comprehensive improvements (Daum et al. 2023), thus reducing landscape fragmentation and enhancing landscape connectivity. However, it also contributed to a decline in landscape diversity, mainly due to shifts towards a more uniform agricultural planting structure. In addition to mechanization, information technology has also brought about changes in product production, trade methods, population mobility, and social interactions. In summary, the transfer of knowledge, technology, and information from urban to rural areas during the urban-rural integration process inevitably alters agricultural production methods, fosters innovation, and expedites the transformation of rural society and the economy (de Bruin et al. 2021). This has significant implications for the structure of urban and rural regions, thereby influencing the overall landscape patterns.

Additionally, the development of small towns and the improvement of road networks effectively reflected the landscape response process of urban-rural spatial linkage. Roads, serving as vital bridges connecting urban and rural spaces, significantly enhanced the connectivity of urban and rural constructed landscape. However, the construction of road networks resulted in conspicuous incisions in the landscape along their routes, leading to landscape fragmentation and a decrease in habitat landscape shape complexity. Conversely, the development of small towns fostered the concentration of urban and rural populations and spaces. This concentration effectively curbed the expansion of urban-rural construction and reduced human activities' interference in agricultural and natural landscapes. This concentration enhanced habitat landscape complexity and connectivity. Moreover, some studies have suggested that urban-rural spatial linkages often accompany the emergence and resolution of spatial conflicts (He et al. 2017). In this process, farmlands, forests, and natural areas may be converted into urban and rural development spaces, potentially leading to habitat fragmentation (Liu et al. 2016). This type of landscape pattern change due to urban-rural spatial conflicts was particularly pronounced in peri-urban areas (Ma et al. 2020). These studies further confirmed the complexity of the impacts of urban-rural spatial linkages on landscape pattern changes (Fig. 9).

Fig. 9
figure 9

Mechanism of how urban-rural integration impacts landscape patterns. Urban-rural integration exerts a multifaceted influence on the dynamics of landscape patterns through economic integration, technology sharing, space interlinkage, and societal convergence. These influences encompass two parts: the positive part potentially contributing to landscape sustainability and the negative part potentially diverging from landscape sustainability. The dynamics of landscape patterns toward sustainability can be facilitated through integrated governance and landscape planning. The abbreviations used in the figure are consistent with Table 2

Full size image

Potential implications for landscape sustainability

From the perspective of promoting landscape sustainability, gaining deeper insights into the impacts of urban-rural integration on landscape patterns can significantly enhance its practical value. Landscape sustainability is defined as "the capacity of a landscape to consistently provide long-term, landscape-specific ecosystem services essential for maintaining and improving human well-being" (Wu 2013). Landscape patterns, which connect nature-society relationships to the landscape, play a crucial role in these processes, for better or worse (Wu 2021). However, the discussion on which landscape patterns are more sustainable still needs exploration (Peng et al. 2021). Here, we attempted to infer the potential implications for landscape (un)sustainability, taking into account changes in landscape metrics. We found that the multifaceted urban-rural integration has complex and intertwined impacts on landscape sustainability.

Initially, the results of our study indicated that urban-rural economic coordination appears to be beneficial for landscape sustainability. Increased investments in rural areas, advancements in high-value agriculture, and a narrowed productivity gap between urban and rural settings helped preserve natural habitats by reducing encroachment while improving connectivity and complexity. These enhancements aided in the provision of ecosystem services such as carbon storage, flood control, and climate regulation (Mitchell et al. 2013; Harrison et al. 2014; Li et al. 2022) They also mitigated landscape fragmentation, which otherwise could lead to habitat loss and diminished species diversity (Collinge 1996; Haddad et al. 2015). Complex patch shapes in natural and agricultural landscapes help enhance species richness and facilitate animal migration, thus contributing to landscape resilience and ecosystem balance (Buechner 1989). However, our findings indicated that these economic changes do not substantially promote landscape diversity to further improve ecosystem balance.

Our findings also suggested that urban-rural societal convergence currently does not favor the development of landscape sustainability. This has led to a reduced proportion of habitat landscapes and increased landscape fragmentation, posing significant challenges to biodiversity conservation and ecosystem stability.

The impacts of urban-rural spatial connections and technological exchanges are complex. Although agricultural mechanization help to create larger habitat patches by decreasing fragmentation (Daum et al. 2023), it may also lead to crop homogenization (Macfadyen and Bohan 2010), thereby increasing ecosystem vulnerability(Sims and Kienzle 2016). Improved road networks between urban and rural areas enhance landscape connectivity but can also increase the fragmentation of agricultural and natural landscapes, exacerbating biodiversity loss (Bennett 2017; Moser et al. 2002). Conversely, a rise in small town density could preserve diverse agricultural and natural landscapes by minimizing disruptions and enhancing habitat complexity and connectivity, thus positively influencing landscape sustainability.

Therefore, to effectively promote landscape sustainability, integrated landscape governance that strategically leverages the advantages of urban-rural integration while precisely addressing its drawbacks is essential. Landscapes are intricately composed of land uses, people, and conflicting interests (Ros-Tonen et al. 2015). To promote landscape sustainability, it is crucial to prioritize the conflicting interests among the various stakeholders in both urban and rural areas to improve human well-being (Joan 2014). Establishing a community of interest that spans urban and rural areas proves to be an effective approach to resolving conflicts by bringing together stakeholders with vested interests in a specific landscape. This allows for the identification of individual and collective needs, objectives, and trade-offs and managing them accordingly (Ros-Tonen et al. 2018). This process can be facilitated through various policies related to land resource management, community governance, urban-rural planning, etc.

The policy design of the urban-rural community of interest should prioritize landscape preservation within the context of advancing urban-rural equity objectives. Additionally, it should focus on improving the accessibility of rural areas and expanding opportunities for farmers to access public services and jobs. Therefore, we can strengthen the development of small towns to better utilize the potential of small towns as nodes connecting rural and urban areas (He et al. 2019). This strategic approach aims to conserve natural landscapes and enhance connectivity based on the mechanistic understanding of how urban-rural integration impacts landscape patterns. Furthermore, policies that encourage the development of high-value agriculture through financial support, technical training, and other means are crucial for promoting urban-rural integration and landscape sustainability. Such measures, including small town construction, rural investment, industrial structure adjustment, agricultural technology improvement, create both upward and downward connections between communities, counties, and metropolitan areas. This facilitates the promotion of urban-rural synergy by guiding, regulating, and accommodating multi-stakeholder decision-making. In addition, to capitalize on the positive impacts of urban-rural interactions on landscapes, it is crucial to establish dependable promotional policies. These policies will better enable us to harness the role of the urban-rural community of interest in constructing a sustainable landscape pattern.

Additionally, sustainable and multifunctional urban and rural landscapes can be constructed through integrated landscape planning and management (Rauws and de Roo 2011; Scott et al. 2013). This approach not only provides protection for critical ecosystem services and a healthy living environment but can also be seen as an effective measure to control the threats sustainably imposed by anthropogenic activities (Peng et al. 2021). Adopting systematic, holistic, and comprehensive approaches to restoration and management enables us to mitigate the damage caused by urban-rural traffic and residential construction in the original natural ecological patches of the natural landscape. This includes preventing or repairing habitats that were disrupted during the urban-rural integration process. One viable solution may involve introducing corridors or installing more permeable matrix habitats (Metzger and Brancalion 2013). For instance, using rivers, mountain ridges, riverine green belts, and woodland ecological corridor networks to connect isolated ecological patches can form a complete regional ecological safety network. Furthermore, it is possible to establish agroforestry or commercial forestry systems in native forest areas by increasing vegetation types and structural complexity. This can enhance habitat permeability and improve the landscape's urban-rural continuity (Metzger and Brancalion 2013). Contributing to the restoration of degraded rural and urban ecosystems to resemble historical ecosystems in terms of structure, species composition, and ecosystem functioning is also crucial (Society for Ecological Restoration 2004). These measures not only address the cause impacts of urban-rural integration on landscape patterns but also have the potential to bring about positive social, economic, and environmental changes in both urban and rural areas. Additionally, they promote the coordination and complementarity of urban and rural landscape functions.

Conclusions

The multifaceted urban-rural integration exerted intertwined effects on landscape patterns. Particularly in densely populated and economically active regions, such as metropolitan areas, urban-rural integration has emerged as a pivotal force driving landscape transformations. In this research, we delved into the impacts of urban-rural integration on landscape patterns and explored its implications for landscape sustainability, using the Changsha metropolitan area as a case study. Our findings revealed a clear upward trend in the level of urban-rural integration across societal, economic, technological, and spatial dimensions in the Changsha metropolitan area from 2005 to 2020, characterized by a discernible core-periphery spatial pattern. Concurrently, the regional landscape had underwent notable shifts marked by the loss and fragmentation of natural and agricultural landscapes, as did heightened landscape diversity and connectivity.

We found that urban-rural economic cooperation emerges as a potential driver of landscape sustainability, fostering control over reduction, enhanced connectivity, mitigated fragmentation, and the preservation of complexity in agricultural and natural landscapes while simultaneously averting fragmentation in constructed landscapes. Conversely, societal convergence trends yielded less favorable outcomes, manifesting in decreased proportions, connectivity, and complexity of agricultural and natural landscapes alongside heightened landscape fragmentation. The impacts of urban-rural technological advances and spatial linkages on landscape sustainability were twofold. Thus, integrated governance should be embraced to address the adverse effects of urban-rural integration while strategically harnessing its positive aspects across diverse empirical contexts. This necessitates the cultivation of a community of interest bridging urban and rural areas, underpinned by robust policy frameworks encompassing a wide array of initiatives, from small town development and rural investments to industrial structure adjustments and agricultural technology enhancements.

Our research, primarily focusing on urban-rural integration as a driver of landscape changes, advances a mechanistic understanding of landscape patterns and processes. It contributes to a more nuanced assessment of the impacts of the economic, social, spatial, and technological dimensions of urban-rural integration on landscape sustainability. We have also preliminarily discussed the potential impacts of landscape pattern changes on landscape sustainability driven by urban-rural integration. However, establishing a quantitative relationship between them presents numerous challenges and necessitates further in-depth exploration because an effective method for evaluating sustainable landscape patterns remains elusive. Furthermore, the process by which urban-rural integration affects landscape patterns and their sustainability is complex and may vary due to regional socioeconomic and ecological characteristics. Therefore, whether the conclusions are consistent across broader or different regions remains to be further validated.