Introduction

The ongoing urbanization has allowed an increase in the exploitation intensity of the land surface and a growing complexity in the land functional use (Foley et al. 2005; Sun et al. 2020; Manisha and Artem 2022). The conflict between human production and living functions and natural ecosystem functions within the limited territory space (Long et al. 2014; Carlos et al. 2022) has resulted in aggravated pollution, ecological degradation, and resource shortages (Liu et al. 2018; Salhi et al. 2021). The territory space, as the vehicle of the coupled human–earth system, embodies land-use production, living, and ecological functions (Pagliarin 2018; Fan et al. 2019). The concept of the production, living, and ecological spaces (PLES) has become essential in understanding the human–earth relationship and promoting sustainable development (Zhao et al. 2021).

PLES are functional spaces that provide various products and services for human beings (Jiang and Liu 2020; Yang et al. 2020). However, identifying PLES is challenging due to spatial scale variability, functional diversity, and scope dynamics. At the macro-scale, PLES can be evaluated based on social, economic, and natural proxies (Zong et al. 2018), or the development suitability of different spatial types (Xia et al. 2021). However, this approach is complicated and subjective, making it difficult to identify PLES variability within a region. At the micro-scale, scholars mainly rely on the existing land-use classification system to identify PLES types by categorizing land sites or assigning them functional scores (Liu et al. 2017; Zhao et al. 2021). This approach is simple and fast to identify PLES, but it cannot fully characterize the spatial functional differences of the same land-use type. Recently, some scholars have delineated PLES based on land-use functional intensity or dominance, but this cannot fully capture the complexity of PLES. PLES is a functional space of a coupled human-land system; thus, it is necessary to interpret PLES from the LUF symbiosis perspective deeply.

Evaluating the spatial patterns of PLES and proposing differentiated measures is an important part of PLES optimization. Scholars have examined the spatial–temporal variation of PLES at the county, watershed, provincial, and national scales and found that PLES exhibit regional differences and gradient effects (Liu et al. 2017; Jiao et al. 2021). Land-use patterns in urban and rural areas are distinct (Chang and Brada 2006), reflecting the complexity of territorial space development across urban–rural gradients (Shi et al. 2020). However, the spatially heterogeneous effects of urbanization on the PLES landscape in fast-growing cities still require a detailed investigation, particularly across urban–rural gradients. PLES optimization is often based on the principle of functional dominance and also fails to take into account regional variability (Li et al. 2023). Therefore, it is essential to comprehensively assess the impact of urbanization on the spatial heterogeneity of PLES at the urban–rural levels.

Rapid urbanization and economic development in China have brought severe environmental problems such as dust storms, water pollution, and land degradation (Qiao and Huang 2022). This has resulted in a fierce conflict between PLES. Xiangyang City is a crucial national agricultural production base, an essential ecological function area, and a core city of the Han River Ecological and Economic Belt. However, PLES in Xiangyang City is under pressure with the ongoing social–economic development, the natural ecological constraints, and imbalanced urban–rural development. To this end, this paper proposes a creative methodology for identifying and optimizing PLES based on LUFs symbiosis to respond to the following research issues: (1) How PLES types and statuses are identified using LUFs symbiosis? (2) Are there any differences in landscape patterns of PLES, and does the spatial distribution vary from urban to non-urban areas? (3) How could PLES be optimized in the context of uneven urbanization? Addressing these questions can improve our understanding of the evolutionary mechanism of PLES and provide feasible adaptation strategies for sustainable urban–rural development.

Study area and data

Study area

Xiangyang City in northwestern Hubei Province, China (110°45′–113°06′E, 31°13′–32°37′N), subordinating the County-level cities of Zaoyang, Yicheng, and Laohekou, the Districts of Xiangzhou, Xiangcheng, Fancheng, and Counties of Nanzhang, Baokang, and Gucheng with a total area of 1.97 × 104 km2 (Fig. 1). The elevation ranges from 90 to 1962 m. The overall terrain is generally surrounded by mountains on the west, a hilly plain on the south, and a Hanjiang river alluvial plain in the central.

Fig. 1
figure 1

Region of study

Full size image

In 2021, the area’s population was 526.10 × 104, ranking third in Hubei Province. Approximately 60% of the population was distributed in Zaoyang, Xiangzhou, Fancheng, and Gucheng. With the implementation of national strategies (e.g., the rise of central China and the Yangtze River Economic Zone) and regional strategies (e.g., Han River Ecological and Economic Belt and Western Hubei Eco-cultural Circle), the local economy developed rapidly. The area’s GDP was 601.97 billion in 2020, ranking second in Hubei Province. The built-up area increased more than five times from 59 km2 in 2000 to 363 km2 in 2020. Rapid urbanization and industrialization have resulted in urban space expansion, natural space shrinkage, and agricultural space reconfiguration. Therefore, assessing PLES and their urban–rural gradient characteristics in this region can provide a scientific reference for territorial spatial planning, ecological management, and integrated urban–rural development.

Data source and processing

Spatial data and statistical data were used in this paper. Spatial data included: (1) Land use data in 2000 and 2020 were collected at a spatial resolution of 30 m from the Data Center for Resource and Environment Science of the Chinese Academy of Sciences(https://www.resdc.cn). The data were categorized into eight land use types, i.e., arable land, forest, grassland, water area, urban construction land, rural residential land, other built-up lands, and unused land (Deng et al. 2006). (2) Digital elevation model (DEM) with a spatial resolution of 30 m was derived from Geospatial Data Cloud (http://www.gscloud.cn/). (3) Soil dataset at a spatial resolution of 1 km was collected from National Earth System Science Data Center (http://www.geodata.cn/). (4) Meteorological station data coving Xiangyang City were extracted from the China Meteorological Science Data Sharing Service (http://cdc.cma.gov.cn). The data, including precipitation, temperature, and sunshine hours, were interpolated using the thin plate splines method to convert point data into raster data with a spatial resolution of 30 m. (5) Nighttime light data were downloaded from the results in Li et al. (2020). (6) Vector data, including the administrative division, the transport map, and the Bureau of Natural Resources and Planning of Xiangyang, supplied the scenic spot. Statistical data, including secondary industry, tertiary industry, grain yield, population, and bus passenger turnover, were collected from Xiangyang (Xiangfan) Statistical Yearbook (2001 and 2021). All data were first converted to the standard spatial reference (WGS1984, UTM Zone 49 N) and then resampled to raster at a spatial resolution of 30 m. Lastly, all these datasets were calculated within 1 km × 1 km grids.

Methodology

Systematic framework understanding PLES from LUF symbiosis perspective

Symbiosis theory derives from biology and has been widely adopted in various subject fields because of its theoretical universality (Sokka et al. 2015; Aanen and Eggleton 2017). Symbiosis is a state in which different species mutually reinforce and coexist harmoniously under specific circumstances (Silvana et al. 2019). The symbiosis system is composed of symbiotic unit, substrate, media, environment, and mode (Munzi et al. 2019; Xu et al. 2022). The symbiotic unit is the primary material condition for forming a symbiotic body, and its information, matter, and energy define the symbiotic substrate. A symbiotic environment refers to all external factors that affect the symbiotic unit. Symbiotic media stands for the different ways of the connections between symbiotic units. Symbiotic mode refers to the interactions or combinations between symbiotic units. Overall, symbiosis theory emphasizes that symbiotic units shape the structure and function of symbiosis systems, which are influenced by the external environment and inner substrate.

It is widely accepted that PLES are composite functional zones that serve multiple purposes. For instance, construction land has both production and living functions, while arable land has both ecological and production functions (Liu et al. 2023). The formation of PLES reflects the carrying capacity and feedback mechanism of the natural system to human activities, as well as the spatial occupation and adaptation of human activities on the natural system. Therefore, PLES possess the symbiotic quality of the human-land system and can be seen as a LUFs symbiotic system, which are formed by interdependent LUFs on a specific spatial and temporal scale.

In the symbiotic system of PLES, the three functions of production, living, and ecology are referred to as symbiotic units, which are important components of the symbiotic system (Fig. 2). The interactions between production–living–ecological functions are called symbiotic relationships, which create necessary conditions for the PLES system. Land resources, population, industry, and transportation are the symbiotic substrates of LUFs, and land-use types or landscapes serve as carriers for the generation and interconversion of LUFs, i.e., symbiotic media, following the cascade relationship of element–structure–function. In our work, symbiotic environment refers to factors such as regional development policies that affect LUFs. Urbanization is a critical indicator of the regional development status. The gradient difference between urban and non-urban areas often results in the transfer of materials and energy between LUFs and changes in the structure and pattern of PLES.

Fig. 2
figure 2

Conceptual model of symbiotic system of production–living–ecological spaces

Full size image

Spatialization of LUFs

Xiangyang’s territorial space is vital to provide the material basis for urban development and the supply of agricultural and by-products for food security. Moreover, it also provides environmental protection for regional ecological conservation. In this paper, LUFs in Xiangyang City were categorized into production, living, and ecological functions from the perspective of production intensification, living comfort, and ecological security. On this basis, considering the accessibility of data and the stability of indicators, this paper constructed an indicator system for calculating production, living, and ecological functions in Xiangyang City. Among them, agricultural and non-agricultural production indicators reflected the capacity of food supply and economical driving in production space, respectively, which were calculated by the methods in Liu et al. (2021). Residential carrying, travel guarantee, and recreation indicators demonstrated the attributes of security, convenience, and comfort in living space, respectively. The calculation methods of the above indicators were based on Yu et al. (2020) and Liu et al. (2023) findings. Climate regulation, water connotation, soil retention, and flood storage indicators represented the capacity of self-regulation in ecological space, while habitat maintenance represented its biodiversity. Of these, climate regulation, water connotation, and soil retention indicators were estimated by InVEST model (Tallis et al. 2011; Peng et al. 2017), and flood storage indicator was calculated by Pan et al. (2018) research. Individual land-use function was computed by weighted summation of the functional indicators (Zhang et al. 2019), where the weights were obtained by analytic hierarchy process (Zeng et al. 2018) (Table 1).

Table 1 Indicator weight of LUFs in Xiangyang City
Full size table

Construction of LUFs symbiosis degree model

Quantification of symbiotic degree

The symbiotic degree model reveals the correlation degree of qualitative parameters and mutual influence between two symbiotic units (Sokka et al. 2015; Lan et al. 2019). PLES are the integrated result of multiple LUFs interacting with each other in territorial development and protection. Therefore, intending to examine the symbiotic degree between LUFs, this paper took production, living, and ecological functions in each grid as symbiotic units and selected production, living, and ecological function scores as qualitative covariates. Thus, the symbiotic degree of function A to function B was calculated as follows.

$${\delta }_{AB}=\left(\frac{d{Z}_{A}}{{Z}_{A}}\right)/\left(\frac{d{Z}_{B}}{{Z}_{B}}\right)$$

And then, the symbiotic degree of function B to function A was calculated as follows.

$${\delta }_{BA}=\left(\frac{d{Z}_{B}}{{Z}_{B}}\right)/\left(\frac{d{Z}_{A}}{{Z}_{A}}\right)$$

where \({\delta }_{AB}\) was the symbiotic degree of function A to function B, which reflected that the change rate of the highest quality parameter of function B was caused by that of the main quality parameter of function A. Likewise, \({\delta }_{BA}\) was the symbiosis degree of function B to function A. ZA and ZB were the qualitative parameters of function A and function B, respectively. dZA and dZB were the change rate of function A and function B, respectively.

Measurement of symbiotic coefficient

The symbiotic coefficient is the interaction degree between function A and function B. The calculation method was as follows.

$${\theta }_{AB}=\frac{\left|{\delta }_{AB}\right|}{\left|{\delta }_{AB}\right|+\left|{\delta }_{BA}\right|}$$
$${\theta }_{BA}=\frac{\left|{\delta }_{BA}\right|}{\left|{\delta }_{BA}\right|+\left|{\delta }_{AB}\right|}$$
$${\theta }_{AB}+{\theta }_{BA}=1$$

where \({\theta }_{AB}\) denoted symbiotic coefficient of function A to function B, and \({\theta }_{BA}\) denoted symbiotic coefficient of function B to function A, respectively. If \({\theta }_{AB}\) = 0 and \({\theta }_{BA}\) = 1, function A had no effect on function B while only function B affected function A. If \({\theta }_{AB}\) = 1 and \({\theta }_{BA}\) = 0, function B had no effect on function A, while only function A affected function B. If 0 < \({\theta }_{AB}\)  < 0.5 and 0.5 < \({\theta }_{BA}\)  < 1, the effect degree of function B on function A was greater than that of function A on function B. If \({\theta }_{AB}\) = \({\theta }_{BA}\)  = 0.5, the interaction degree between function A and function B was the same. If 0.5 < \({\theta }_{BA}\)  < 1 and 0.5 < \({\theta }_{AB}\)  < 1, the effect degree of function A on function B was greater than that of function B on function A.

Discernment of symbiotic mode

The symbiotic modes of function A-B were judged by comparing the symbiotic degree and coefficient between function A and function B (Table 2). Specifically, the symbiosis modes of function A and function B included mutual and biased benefit symbiosis. Mutual benefit symbiosis meant that the symbiotic degrees between the two LUFs were all greater than 0, while biased benefit symbiosis meant that the symbiotic degree of one function was greater than 0 and the other was equal to 0.

Table 2 Discernment of symbiotic modes among LUFs in Xiangyang City
Full size table

The non-symbiosis modes of function A and function B included mutual, single, and biased damage. Of these, mutual damage mode was defined when the symbiotic degrees were all less than 0, single damage mode was defined when the symbiotic degree of one function was less than 0, and the other one was greater than 0. Biased damage was defined when the symbiotic degree of one function was equal to 0, and the other one was less than 0. In addition, the independence mode was identified when the symbiotic degree of the two functions were all equal to 0, indicating that there was an unaffected mutual relationship of function A-B.

On this basis, symbiotic modes of function A-B could be further classified into symmetry and first modes according to different symbiotic coefficients. Of them, symmetry mode was identified as the symbiotic coefficient was equal, indicating that two functions are beneficial or damage to the same extent. The first mode was identified as the symbiotic coefficient differs, indicating that one function was more beneficial or damage.

Identification procedure of PLES

The PLES identification system including the functional combination and ranking was constructed based on LUFs’ symbiotic modes and coefficients. First, the functional combination types of PLES were determined by comparatively analyzing the symbiotic modes of production–living functions, production–ecology functions, and living–ecology functions (Fig. 3).

Fig. 3
figure 3

Integrated identification tree of production–living–ecological spaces based on symbiotic modes of LUFs

Full size image

Specifically, (1) if the symbiotic mode of three pairs of LUFs were mutual benefit, mutual damage, and independence relationships, correspondingly, PLES were defined as production–living–ecology benefit space, production–living–ecology damage space and production–living–ecology independence space, respectively. (2) If the three pairs of LUFs were biased benefit, single harm, and biased harm relationships, respectively, they were defined as a function A benefit–function B damage–function C independence space or function A damage–function B benefit–function C independence space. (3) If the two pairs of LUFs were mutual benefit relationships when a pair of LUFs was a biased benefit relationship, it was defined as a function A-B benefit–function C independence space; when a pair of LUFs was a single damage relationship, it was defined as function A-B benefit–function C damage space or function A damage–function B-C benefit space. (4) If the two pairs of LUFs were mutual damage relationships when a pair of LUFs was a biased damage relationship, it was defined as function A-B damage–function C independence space; when a pair of LUFs was a single damage relationship, it was defined as function A-B damage–function C benefit space or function A benefit–function B-C damage space. 5) If two pairs of LUFs were independence relationships when a pair of LUFs was a biased benefit relationship, it was defined as a function A benefit–function B-C independence space; when a pair of LUFs was a biased damage relationship, it was defined as a function A damage–function B-C independence space.

Last, the ranking of the production, living, and ecology functions, i.e., the space’s dominant, secondary, and tertiary functions in PLES, was judged based on the symbiotic coefficient. Therefore, PLES in this paper were functional spaces with an orderly combination of production, living, and ecological functions.

Landscape pattern of urban–rural gradient

Landscape pattern indices

Landscape pattern indices captured the landscape’s pattern composition and spatial characteristics. They are often used to assess differences in structural characteristics between landscapes or to compare variability presented by the same type of landscape at various gradients. This paper selected the indices of patch, class, and landscape levels to characterize the individual units, component configurations, and overall diversity of the landscape (Table 3) (Cong et al. 2016). The ecological implication of these landscape pattern indices could be found in Masoudi and Tan (2019) findings, and their values were calculated in Fragstats 4.2.

Table 3 Landscape pattern indexes and their ecological significance
Full size table

Defining urban–rural gradient

Rapid urbanization leads to spatial expansion from metropolitan area to fringes and built-up intensification in urban regions (Zheng et al. 2022). The urban–rural gradient analysis has been approved as a helpful tool to analyze urban spatial expansion and its impact on landscapes (Stephan et al. 2021). Therefore, this method was used to reveal the evolution and optimization of PLES in different regions along the urban–rural gradient.

From the perspective of growth on urban land coverage, urban–rural gradients were measured using built-up land (Hu et al. 2017; Cao et al. 2021). Specifically, we first defined urbanized areas by aggregating all the grids with more than 50% built-up land in 2000 and 2020, respectively. Second, the urbanized area in 2000 was identified as an old urban area (OUA, 480 km2), and the expanded urbanized area during 2000–2020 was defined as a new urban area (NUA, 1621 km2). Third, the non-urbanized area in 2020 was defined as a rural area (RA, 17840 km2) (Fig. 4).

Fig. 4
figure 4

Delineation range of urban–rural gradient in Xiangyang City in 2020

Full size image

Results

Spatio-temporal variations in LUF symbiosis

For production and living functions (Fig. 5a), the area proportions of production-first biased symbiosis and production-first biased damage relationships were about 26.80% and 25.32%, respectively, while the least living-first biased benefit symbiosis at 0.03%. Spatially, production-first biased symbiosis primarily occurred in the eastern Zaoyang and Yicheng cities, and production-first bias damage relationship was mainly distributed in the southern Baokang and Nanzhang counties. Mutual benefit symbiosis relationships, including production-first and living-first mutual benefit symbiosis, were mainly distributed in a point-like pattern in the built-up areas of each county.

Fig. 5
figure 5

Spatial distribution of symbiotic modes for a production and living functions, b production and ecology functions, and c living and ecology functions in Xiangyang City in 2020

Full size image

In terms of production and ecology functions (Fig. 5b), the area of production-first single damage and production-first mutual benefit symbiosis relationships accounted for 29.87% and 26.82%, respectively. The smallest area proportion of ecology-first biased damage relationship was 0.05%. Spatially, mutual benefit symbiosis relationships, including production-first and ecology-first mutual benefit symbiosis relationships, were mainly distributed in Zaoyang, Yicheng, and Laohekou cities owing to rich farmland and forest land. Production single damage relationship was mainly located in Xiangzhou and Xiangcheng districts, and ecology single damage relationship occurred in areas flowing through the Han River.

For living and ecology functions (Fig. 5c), the area of ecology-first biased benefit symbiosis relationship accounted for 50.60%, while living-first biased damage relationship accounted for the smallest percentage of 0.02%. Spatially, ecology-first biased benefit symbiosis relationship was distributed facultatively in Nanzhang County and Yicheng and Zaoyang cities. Ecology-first mutual benefit symbiosis relationship was mainly distributed in the west of Xiangzhou District, the north of Laohekou City. Ecology-first single damage was mainly distributed in the built-up areas of the counties, while living-first single damage was mainly distributed around the built-up areas.

Spatial characteristics of PLES

In terms of the number of PLES types, there were 25 combination types in Xiangyang City in 2020, including 22 types in Gucheng County; 21 types in Baokang County; 20 types in Zaoyang City; 19 types in Yicheng City, Xiangzhou District, and Nanzhang County; and 18 types in Xiangcheng District and Laohekou City (Fig. 6).

Fig. 6
figure 6

Spatial distribution of production–living–ecological spaces in Xiangyang City in 2020. Note: the space of (1) production–living–ecology independence, (2) production–ecology benefit–living independence, (3) production–ecology benefit–living damage, (4) production damage–living–ecology independence, (5) production damage–living–ecology benefit, (6) production damage–living benefit–ecology independence, (7) production damage–ecology benefit–living independence, (8) production benefit–living–ecology independence, (9) production benefit–living–ecology damage, (10) production benefit–living damage–ecology independence, (11) production benefit–ecology damage–living independence, (12) production–living–ecology damage, (13) living–ecology damage–production independence, (14) living–ecology benefit–production independence, (15) living damage–ecology benefit–production independence, (16) living benefit–ecology damage–production independence, (17) ecology damage–production–living independence, (18) ecology benefit–production–living independence, (19) production–living–ecology benefit, (20) production–living damage–ecology independence, (21) production–living damage–ecology benefit, (22) production–living benefit–ecology independence, (23) production–living benefit–ecology damage, (24) production–ecology damage–living independence, (and 25) production–ecology damage–living benefit

Full size image

The area percentage of each PLES type was below 22% and varies significantly. Among them, the area of production–ecology benefit–living independence was 4358 km2, accounting for 21.85% of the city’s area, mainly distributed in the eastern hilly areas, such as Zaoyang City and Yicheng City. The area of production damage–ecology benefit–living independence is 4206 km2 (21.09%), which was scattered in the southwestern mountainous areas, especially in Nanzhang County and Baokang County. The area proportion of ecology benefit–production–living independence and production–living–ecology benefit spaces were 8.55% and 8.14%, respectively, mainly distributed in the northern hills, such as Xiangzhou District and Laohekou City. The percentage of other space types ranged from 0.01% (production–living damage–ecology independence space) to 7.65% (production–ecology benefit–living damage space). In addition, production–living damage–ecology independence space was only found in Xiangcheng District.

Landscape pattern analysis of PLES

In terms of individual landscape unit characteristics, CA ranged from 1 to 4358 km2, with nine types more prominent than the mean value (797.64 km2) (Fig. 7). LSI ranged from 1 to 35, with the value of 12 PLES types greater than the mean value (13.485). As for landscape component configuration, PD ranged from 0 to 0.032, with the value of 11 PLES types greater than the mean value (0.009), and COHESION ranged from 0 to 96.019, with the value of 13 PLES types greater than the mean value (47.123). LPI ranged from 0.005 to 10.9172, with greater than average LPI (1.112) for production–living benefit–ecology damage, production–living–ecology benefit, production–ecology benefit–living independence, and production damage–ecology benefit–living independence spaces.

Fig. 7
figure 7

The overall landscape features of production–living–ecological spaces in Xiangyang City in 2020. Note: the space of (1) production–living–ecology independence, (2) production–ecology benefit–living independence, (3) production–ecology benefit–living damage, (4) production damage–living–ecology independence, (5) production damage–living–ecology benefit, (6) production damage–living benefit–ecology independence, (7) production damage–ecology benefit–living independence, (8) production benefit–living–ecology independence, (9) production benefit–living–ecology damage, (10) production benefit–living damage–ecology independence, (11) production benefit–ecology damage–living independence, (12) production–living–ecology damage, (13) living–ecology damage–production independence, (14) living–ecology benefit–production independence, (15) living damage–ecology benefit–production independence, (16) living benefit–ecology damage–production independence, (17) ecology damage–production–living independence, (18) ecology benefit–production–living independence, (19) production–living–ecology benefit, (20) production–living damage–ecology independence, (21) production–living damage–ecology benefit, (22) production–living benefit–ecology independence, (23) production–living benefit–ecology damage, (24) production–ecology damage–living independence, (and 25) production–ecology damage–living benefit

Full size image

It should be noted that PD, LSI, and COHESION were the smallest for production–living damage–ecology independence space, indicating that this space has the most concentrated distribution, the simplest shape, and the least connectivity. LSI, COHESION, and LPI of production damage–ecology benefit–living independence space were the largest, suggesting that this space had most complex shape, the strongest connectivity, and the greatest influence on the overall landscape pattern. As a result, production damage–ecology benefit–living independence space was the dominant landscape type. PD of production–ecology benefit–living independence space was the largest, but its CA was also the largest, indicating that the space is more fragmented despite its larger area.

Urban–rural gradient analysis of PLES

Overall landscape characteristics of PLES

Throughout the whole city, SHDI of OUA, NUA, and RA were 2.281, 2.367, and 2.316, respectively, which were all below the average level of the whole city (TWC, 2.385) (Fig. 8). SHEI of OUB, NUB, and RU were 0.823, 0.777, and 0.729, respectively, while the city’s average level is 0.741. Specifically, in OUA, NUA, and RA, SHDI showed a trend of increasing first and then decreasing, and SHEI decreased from 0.82 to 0.72. The changes indicated that in OUA, built-up land was the dominant landscape type due to the slight difference in land use patterns. While NUB was rapidly expanding, with the increased intensity of human activities, PLES’s landscape types became rich, and landscape fragmentation increased. RA developed slowly characterized by the weak intensity of human activities. However, RA was mainly in the hilly mountainous area, where various landscape types were interspersed, and the diversity and heterogeneity of patches were substantial.

Fig. 8
figure 8

The overall landscape diversity of production–living–ecological spaces in OUB, NUB, RU, and TWC of Xiangyang City in 2020. Note: OUA represents old urban area, NUA represents new urban area, RUB represents rural area, and TWC represents the whole city

Full size image

Urban–rural landscape characteristics of PLES

As shown in Table 4, the number of PLES types in OUA, NUA, and RUA were 16, 21, and 24, respectively. Specifically, CA, LPI, and COHESION of production–living benefit–ecology damage space were the largest in OUA, while LSI and PD of production–ecology benefit–living damage were the largest. This was because construction land was the primary land-use type in the downtown area of Xiangyang City, which was densely populated. The fact resulted in the strongest disturbance to the ecological environment from production and living activities. Hence, production–living benefit–ecology damage space became the dominant landscape type, while production–ecology benefit–living damage space was the most fragmented.

Table 4 Landscape characteristics of PLES types of urban–rural gradient in Xiangyang City
Full size table

CA, LSI, and PD of production–ecology benefit–living damage were max in NUA, indicating that although the space was large-scale, its fragmentation caused by diverse land use activities was also significant. Consequently, production–ecology benefit–living damage space failed to become a dominant landscape type, which was also confirmed by its maximum LPI and COHESION.

Nevertheless, in RA, production–ecology benefit–living independence space had the largest CA and PD, which was closely associated with the diminished anthropogenic disturbance and dominance of forest and cropland. RA was also the main region for implementing green for grain program, which might weaken the agricultural production function while improving ecological function. Therefore, LSI, LPI, and COHESION of production damage–ecology benefit–living independence were the largest.

Discussions

Impact of urbanization on PLES

Cities are complex, open systems in which the interaction between cities, semi-urbanized areas, and the rural hinterland are fundamental to the process of urbanization (Rodríguez-Pose 2008). The urban–rural gradient resulted from urbanization is characterized by the uneven spatial distribution of population aggregation, economic development, and transportation location from urban to rural areas (Ilkwon et al. 2020). The findings of urban–rural gradient analysis in this paper indicated that the landscape pattern of PLES exhibited different characteristics with the change of distance to OUA. In general, the degree of heterogeneity and fragmentation of PLES landscape pattern was highest in NUA, followed by RU, and slightly slower in OUA, influenced by natural conditions and land development intensity. Thus, the variation in the landscape pattern of PLES from urban to rural regions reflects the flow of population, resources, goods, information, technology, capital, and cultural concepts.

Urbanization is often characterized by the expansion of densely populated built-up areas. Since 2000, Xiangyang City had undergone rapid urbanization, resulting in a 304 km2 increase in built-up areas by 2020. OUA and NUA, with urban space as the primary component, were experiencing the most significant impact of human activity due to their high population density and socioeconomic investment. These areas had higher production and living functions and lower ecological functions. Thus, OUA exhibited the largest production–living benefit–ecology damage space due to the continuous encroachment of urban built-up land on arable and ecological land. However, the NUA still maintained a significant proportion of agricultural and ecological spaces, resulting in a more spatially dispersed PLES pattern than that of OUA. To improve the urban ecosystem, Xiangyang City had developed several ecological parks such as Yuliangzhou Central Ecological Park, Lianshan Lake Ecological Park, and Fancheng Circular Greenway in NUA. Obviously, these efforts led to the improvement of ecological functions in NUA, making it an essential aspect of the new urbanization process.

CA and PD of production–ecology benefit–living independence were the largest because RA was a coupled human–nature system with agricultural space and ecological space as the mainstay. Thus, the landscape pattern of PLES was greatly influenced by agricultural development and ecological protection. RA mainly was mountainous and hilly areas with a forest land proportion of 65% and an arable land proportion of 60%. Pursuing spiritual enjoyment and getting close to nature had been popular in urbanization to enhance people’s living standards (Emborg and Gamborg 2016). Moreover, Xiangyang City is the core city of the Western Hubei Eco-cultural Circle. Rural areas have become important tourist destinations (Li et al. 2022), leading to an improvement in living functions. However, rural settlements in the western mountainous and southern hilly areas of Xiangyang City, such as Nanzhang, Baokang, and Gucheng counties, had reduced in size through rural consolidation (Cui et al. 2018). Therefore, the production and ecological functions of RA were enhanced, while the living functions remained unchanged.

Optimization strategy of PLES

A scientific and reasonable arrangement of PLES enhances economic efficiency and improves the ecological environment in urbanization. The bucket effect emphasizes the limiting roles of crucial factors on specific development goals, while the comparative advantage theory emphasizes the potential capacity of key factors to determine the differentiated development direction. PLES have the potential to exploit advantageous resources to improve LUFs due to its functional dominance and geographical differences (Ji et al. 2020). Accordingly, PLES optimization should aim to repair damaged functions and strengthen beneficial functions. Healthy ecosystems are the basis for sustained urbanization, so ecological functions were prioritized over production and living functions. Therefore, we propose PLES optimization paths for OUA, NUA, and RA (as shown in Table 5) for land-use management in urban–rural integration development.

Table 5 Optimization paths for PLES types in OUA, NUA and RA
Full size table

The optimization of PLES aims to transform the competitive relationships between LUFs into collaborative ones by establishing a symbiotic system, ultimately achieving the sustainable development of the city as a whole. Damaged functions are crucial determinants for PLES harmonization, while beneficial functions determine PLES superiority. Thus, the primary consideration for PLES optimization is to restore damaged functions, followed by leveraging the advantageous role of beneficial functions to reinforce and improve the system. Specifically, in areas where production, living, or ecological functions have been damaged, the primary task is to restore these functions to resolve spatial incompatibilities between socio-economic development and environmental protection. In contrast, for areas that offer benefits or independence in terms of production, living, or ecology, the optimization direction should involve reinforcing, improving, or maintaining these functions due to their mutual or biased benefit symbiosis. Additionally, production–living–ecological coordination areas should be developed for independent areas of production, living, and ecology to resolve the issue of separate development between them.

Moreover, OUA, NUA, and RA pursued different goals and had varying effects on symbiotic integration, emphasizing the need to consider regional heterogeneity in production–living–ecological optimization policies. For OUA, the focus was on improving service convenience and livability, so priority should be given to enhancing living functions while also improving production and ecological functions. This could be achieved through improving land resource utilization intensity and reasonably controlling population density in downtown areas, as well as strengthening ecological restoration and expanding urban green spaces to improve habitat eco-environment. NUA aimed to achieve intensive land use and efficient output by improving industrial production functions. Thus, priority was given to improving production and living functions through the orderly use of land resources, while protecting ecology functions in large-scale land development. RA was a key food supply and ecological protection area, so PLES optimization focused on maintaining food production and environmental protection functions. Ecological functions were given priority, and the supply capacity of ecological products was reinforced.

Advantage and future direction

Symbiosis theory is an emerging science that investigates the formation of synergistic evolution of complex systems (Silvana et al. 2019). However, literature on the symbiotic interaction between LUFs in the PLES is scarce. Some scholars have tried to analyze LUFs coordination to explore the internal connection between LUFs. The inclusive, holistic, and systematic nature of PLES indicates that the elements and functions are interconnected and affect each other (Liu et al. 2021). Nevertheless, the coordination emphasizes the orderly cooperation between different functions, while symbiosis places more emphasis on the interdependence and reciprocity between different functions. Therefore, this paper introduces symbiosis theory to establish integrated assessment criteria for PLES based on the symbiotic relationships between LUFs. Limited by the complex linkages between LUFs and the inadequate methods, this paper measured the symbiotic relationship between pairwise LUFs instead of three LUFs. Although this approach also clearly and indirectly reveals the dependency characteristics of three LUFs, it is critical to establish a new integrated model to directly and simultaneously depict the symbiotic relationship of three or more LUFs in the future.

Moreover, PLES is optimized by considering the comparative advantages and shortcomings of LUFs, and optimal strategies are proposed by incorporating urban–rural differentiation. It is important to note that PLES optimization not only is about coordination among the three functions of production, living, and ecology but also involves interactions with different regions, such as urban and rural areas. Social network analysis theory can be applied to clarify the flow of elements (information, capital, population, etc.) between urban and rural systems, thereby achieving a comprehensive understanding of the PLES symbiosis system at the regional scale.

Conclusions

The methodology developed in this paper, based on symbiosis theory, utilized the symbiosis degree model and landscape pattern indices to measure the variability of PLES along the gradient belt from urban to rural areas. The proposed idea of urban–rural differentiation optimization for PLES was also explored. The results showed that our model was effective in measuring LUFs symbiosis and identifying PLES. PLES in Xiangyang City revealed 25 combinations types with varying spatial distributions. Of these, production–ecology benefit–living independence space had the largest area but was fragmented and scattered. Production damage–ecology benefit–living independence space had the most complex shape, strongest connectivity, and largest landscape pattern. The landscape richness and fragmentation of PLES increased gradually from OUA to RA, reflecting the integrated effect of uneven spatial population aggregation and economic development caused by urbanization.

This study offers valuable insights into the diverse relationships between LUFs, PLES, and urban–rural systems, which deepens and concretizes the study of symbiosis theory and provides a new perspective for integrated urban–rural development. It also informs the improvement of territorial space planning in Xiangyang City and beyond. With accelerating global climate change and urbanization, drastic changes in LUFs will easily lead to an increase in the risk of PLES conflicts. Therefore, risk prediction and early warning of PLES conflicts will become an essential part of land-use change science and an invaluable basis for territorial space optimization.