Human–Land Coupling Relationship in Lushan National Park and Its Surrounding Areas: From an Integrated Ecological and Social Perspective

Abstract

The interaction of human activity in national parks and their surrounding areas with natural landscapes is an important factor in the formulation of strategies for the protection and management of protected areas. From an integrated ecological and social perspective, this study aims to develop a human–land coupling model to reveal the ecological threats of human activities to national park and their surrounding areas under rapid urbanization. We first established a four-quadrant model based on the coupling relationship between human activity intensity (HAI) and the landscape pattern index (LPI). On this basis, we analyzed the spatial coupling characteristics of these indices from 2000 to 2020, preliminarily assessed the human–land relationship in Lushan National Park and its surrounding areas (LNPSA), and then explored the driving factors of the coupling relationship. The results show the following. (1) The proportion of regions with high and extremely high human activity intensity increased from 6.02% to 16.41% over the past two decades. These regions are mainly distributed in the surroundings of Lushan National Park, showing a gradually spreading trend to the core protected area. (2) The surroundings had higher landscape fragmentation, landscape diversity, and total variation in the landscape pattern compared with the core protected area. (3) Vegetation coverage and distance to Lushan have the most significant effect on the human–land coupling relationship in LNPSA, and human activity and natural evolution together shape the characteristics of this relationship in the study area. (4) Utilizing administrative divisions as the fundamental framework, the study area is delineated into four distinct zones based on the results of the human–land coupling analysis—harmonious development zones, stable transformation zones, environmental regulation zones, and risk prevention zones—with tailored optimization strategies proposed for each zone’s characteristics. These findings can facilitate the optimized formulation of schemes for different regions and provide a comprehensive methodology to guide the planning and management of natural protected areas.

1. Introduction

Human activity has become a major force driving the variations in the global ecological environment, whose impact has exceeded that of natural evolution [1]. Ongoing urbanization has exacerbated the exploitation of natural resources and the destruction of the ecological environment at an erosion rate far beyond the self-repair capacity of ecosystems. With the onset of the Anthropocene in the 21st century, the influence and disturbance of human activity on natural processes have become a hotspot for research [2]. In China, coordination of the human–land relationship is a critical issue in current social development due to its large population and scarce land resources [3]. This issue is crucial for maintaining ecological balance [4], protecting biodiversity [5], and mitigating climate change [6], while further highlighting the importance and urgency of coordinated development among the population, the economy, and the environment under the condition of finite land resources. Over the past century, the increase and expansion of the population have resulted in over-exploitation of resources and destruction of habitats, posing severe threats to ecosystems [7], the deterioration of which would, in turn, restrict the sustainable development of human society. Therefore, it is imperative to accurately understand the interdependent relationship between humans and the environment and take effective measures to alleviate human–land conflict. Facing the relentless expansion of human activities, even those strictly protected areas are confronted with certain pressure from humans [8]. National parks are the most important natural protected areas, playing a key role in the natural protection system of the world [9]. National parks are important ecological functional areas in China, and their planning and management have great impacts on sustainable socioeconomic development and ecological integrity in surrounding areas, which is regarded as the most important means to protect the species diversity in natural habitats [10]. However, due to inevitable overlaps between human activity spaces and ecological protection areas, national parks are currently under immense pressure from the conflicts between nature conservation and social development [11]. They face conflicts with surrounding human communities [12], as well as encroachment on wildlife habitats because of urban expansion [13], among other pressing issues.

To date, some progress has been achieved in research on the human–land relationship for different national parks, which mainly involves quantitative analysis of the effects of human activity on key ecosystem service functions [14] and quantification of the stress intensity of human activity [15] and the disturbance intensity on the environment [16]. Although current research has revealed the human–land relationship from certain perspectives, it has largely ignored the complexity of the interaction between human activity and the environment, and there is a lack of unified evaluation standards [17]. Therefore, there is an urgent need for a set of comprehensive methodologies for better understanding and managing the interactive relationship between human activity and the natural environment.

In terms of the protection and management of national parks, recent research is mostly related to the legislation system [18], the management system [19], and biodiversity protection [20], mostly focusing on the protection and development planning of the national parks themselves, while neglecting the changes in habitat quality and protection of the integrity of the surrounding areas [21]. With the formulation of the European Landscape Convention (ELC) and the development of landscape character theory, a certain consensus has been reached in the protection and management of comprehensive landscapes, and it has been increasingly recognized that the surrounding areas also play an important role in maintaining the ecological functions of the protected areas [22]. The research scope has been expanded from the protected area to the peripheral regions, and the research perspective has been extended from the single dimension of ecological protection to a multi-dimensional analysis covering society and the economy [23]. Changes in land use [24], the development of communities [25], variations in landscape patterns [26], and the well-being of residents [27] have become key topics in research on the surrounding areas of protected areas. Promotion of the internal–external coordinated governance of national parks is gradually becoming a new frontier in research to achieve the ecological balance and sustainable development of protected areas [28].

The interaction between humans and the natural environment forms a complex system with society–ecology interdependency [29]. The surrounding areas of national parks are typical social–ecological systems, and their human and natural landscape interactions can well reflect the dual demands of ecological protection and social development. Due to the multiple and complex intersections between the internal and external spaces of national parks in China, there are frequent interactions between the natural ecological systems and socioeconomic activities, resulting in great environmental pressure and ecological risks [30]. These negative impacts have also aggravated the ecological degradation of the core protected area, posing serious challenges to the future development of protected areas. Therefore, this study proposes a method integrating the perspectives of ecology (landscape pattern) and society (human activity) for more effective and sustainable protection of the protected areas. The method can facilitate more comprehensive understanding of the complex relationship between humans and nature to meet human demands while maintaining the ecological balance of the protected areas [31].

Lushan National Park (LNP) has a rich natural and cultural heritage and high biodiversity with a great value of ecological protection. It is also a typical protected area with prominent human–land contradiction, and it is faced with the pressure caused by the expansion of surrounding cities and towns. Human production and living compete with ecology for space, and cities and towns have gradually permeated into the boundary of LNP, causing serious damage to the natural ecological system and posing great challenges to the management and protection of the landscape. In recent years, the government of Lushan City has positively promoted the integration and optimization of the space in LNP, but there are still various problems, such as unclear management and protection targets. Particularly in the marginal areas with complex intersections, the management measures have become a mere formality due to the lack of quantitative methods for zoning and boundary definition, which has led to weakened protection of the core area of the park. Hence, it is necessary to further explore more comprehensive and fine strategies for the protection and management of LNP and realize quantitative boundary definition and functional zoning [32].

This study aims to (1) quantitatively assess the human activity intensity and landscape pattern index in Lushan National Park and its surrounding areas (LNPSA) from 2000 to 2020 and investigate their spatiotemporal distribution characteristics; (2) construct a four-quadrant model to analyze the coupling relationship between human activity intensity and the landscape pattern index, reveal the changes in this coupling relationship, and further identify the key factors influencing the relationship; (3) reveal the differences between the internal and external spaces of the national park from the ecological and social perspectives by comparing the core and surrounding areas, and the potential influence of these differences on ecological protection and sustainable development; and (4) formulate zoning strategies for ecological protection and optimizing the protection and development pattern of the park and surrounding areas with administrative boundaries as the reference framework and the analysis results of the coupling relationship. This study introduces a four-quadrant model to conduct an in-depth analysis of the human–land coupling relationship in LNPSA. Based on this analysis, a zoning strategy for ecological protection based on administrative boundaries is proposed. This strategy aims to provide novel approaches and methodologies for optimizing the protection and regional development models of national parks, making a significant contribution to expanding the theoretical framework of human–land system coupling and formulating effective conservation management methods.

2. Materials and Methods

2.1. Study Area

Lushan (115°51′ E, 29°30′ N) to (116°07′ E, 29°41′ N) is located in the intersection area of Poyang Lake and the Yangzi River (Figure 1). LNP was inscribed on the World Heritage List by the UNESCO in 1996. For thousands of years, Lushan has been a typical example of harmonious coexistence of humans and nature in China with its unique natural resources and solid history of humanity and culture. However, with the rapid urbanization, human activities, such as urban expansion, tourism, and the construction of residential buildings, have posed serious threats to the ecological safety of LNPSA in the past decades, and management and protection are faced with realistic difficulties.

This study selects the first-level administrative regions attached to towns and streets in the surrounding areas of LNP as the research boundary, with a comprehensive consideration of the regions covered by the current protection policies, aiming to make a systematic integrated analysis of LNPSA. The study covers a total area of 906.28 km², including three administrative units at the district or county level, and 22 first-level administrative units at the town or street level.

2.2. Research Methods

2.2.1. Construction of Human–Land Coupling Model

The four-quadrant model was initially considered as a tool to measure the changes in the real estate market [33]. With the integration of multiple disciplines, it has also been applied to the field of landscape ecology as it can clearly show the relationship among multiple variables with intuitive and easy operation [34]. This study uses a four-quadrant model to assess the spatial coupling of human activity intensity and the landscape pattern index from 2000 to 2020 and analyze the distribution characteristics of the coupling in different regions. In order to calculate these indicators, we utilized the data presented in

Table 1. Sources of data for the study.

Data	Source	Year of Data Access	Spatial Resolution (m)
Digital Elevation Model (DEM)	Shuttle Radar Topography Mission	2000–2020	30
Land Use/Land Cover (LULC)	National Cryosphere Desert Data Center	2000–2020	30
Population density	AI Earth	2000–2020	100
Boundary of the Lushan National Park	Comprehensive Plan of Lushan Scenic and Historic Interest Area	2020	-
Nighttime lights	Resource and Environmental Science and Data Center of the Chinese Academy of Sciences	2000–2020	1000
GDP density	AI Earth	2000–2020	100
Township administrative boundaries	Bigmap	2020	-
Village-level administrative boundary	National Earth System Science Data Center	2020	-

. The model takes human activity intensity as the X axis and the landscape pattern index as the Y axis, with different quadrants representing different types of human–land coupling (Figure 2).

The first quadrant represents relatively higher human activity intensity, and there are more patch types, complex patch shapes, and high patch density and a low largest patch index and patch aggregation degree. These regions are characterized by relatively higher human activity intensity and more diverse types of land use, and they reflect the effective use of natural environment by humans, such as new urban districts, urban ecological spaces, and tourist towns.

The second quadrant indicates regions with relatively lower human activity intensity, and the characteristics of the landscape pattern index are similar to those of the first quadrant. However, these regions generally show fragmented, diversified, and dispersed landscape structures and a development state of land abandonment and hollowing, such as abandoned land, slope cropland, and mining land.

The third quadrant represents regions with relatively lower human activity intensity, and in the landscape pattern, there are single patch types, high integrity of patch shapes, low patch density, obvious dominance of the largest patch index, and high patch aggregation degrees. These regions feature lower human disturbance and a good ecological environment and undertake the functions of an ecological barrier with contiguous forest land and large areas of lake waters.

The fourth quadrant indicates relatively high human activity intensity, and the landscape pattern index is similar to that of the third quadrant. These regions are characterized by relatively higher human activity but single and homogenized landscape types, such as contiguous city and town agglomeration and large-scale industrial parks.

To further explore the internal evolution law of the human–land relationship, according to the current situation of the site and relevant research, the first and third quadrants are considered a coordinated state of human–land relationships, and the second and fourth quadrants are regarded as an antagonistic state of human–land relationships. Finally, a four-quadrant model is constructed, which represents four types of evolution of human–land relationships, including the coordination type, the antagonism type, the gradually coordinated type, and the gradually antagonistic type. The evolution characteristics of four-quadrant human–land coupling are shown in

Table 2. Coupling types of human activity intensity and landscape pattern.

Type	Time 1	Time 2	Human Activity Intensity	Landscape Pattern	Situation
Coordination type	I	I	No obvious change	No obvious change	Always coordinated
	III	III
	I	III	↓	↓
	III	I	↑	↑
Antagonism type	II	II	No obvious change	No obvious change	Always antagonistic
	IV	IV
	II	IV	↑	↓
	IV	II	↓	↑
Gradually antagonistic type	I	II	↓	No obvious change	From coordination to antagonism
	III	IV	↑
	I	IV	No obvious change	↓
	III	II		↑
Gradually coordinated type	II	I	↑	No obvious change	From antagonism to coordination
	IV	III	↓
	II	III	No obvious change	↑
	IV	I		↓

.

2.2.2. Analysis of the Landscape Pattern

As an important and intuitive reflection of the human–land relationship, the landscape pattern provides an ecological perspective to assess the response and adaptability of the ecosystem to human activity. Therefore, the landscape pattern is taken as an index of the ecological dimension [35]. To comprehensively evaluate the structural characteristics and functional state of the ecosystem and reveal the response mechanism of the natural landscape pattern to human activity, five landscape pattern indices were selected, including the aggregation index (AI), largest patch index (LPI), patch density (PD), landscape shape index (LSI), and Shannon diversity index (SHDI), according to the ecological significance and actual situation of the regions [36,37,38]. These indices can better reflect the characteristics of individual landscape units, the spatial composition of landscape components, and the diversity of the overall landscape.

The moving window method in the Fragstats4.2.1 software was chosen to calculate the above indices. The radius of the moving window was determined by using the abrupt inflection point with easy operation. The study took 300 m as the initial radius, and calculated the landscape pattern index at the radius of 300, 500, 700, 900, 1100, 1300, 1500, and 1700 m with intervals of 200 m. Specifically, 200 random sampling sites were first generated in the study area; then, the size of the landscape pattern index of the sampling sites was determined and the average was calculated; finally, the curve of changes in the average value for different scales of landscape patterns was constructed to observe the inflection point of the average value. Finally, 800 m was determined as the radius of the moving window, and the five indices of the landscape pattern from 2000 to 2020 were generated with the moving window method, which were then input into ArcGIS10.8 for further processing. In addition, to more clearly reflect the research information, the variation of the landscape pattern in the internal and external regions of LNP was quantified to provide a reference for further analysis.

2.2.3. Evaluation of Human Activity Intensity

Human activity intensity provides a social perspective for measuring the impacts and modification of humans on the natural environment, and it was taken as the index of the social dimension in this study [39]. Due to the special attributes of the study area, such as the co-existence of multiple interest bodies and the interlacing of multiple functions, there are dominant and recessive manifestation forms of human activity. Based on the comprehensive index measurement method of human activity proposed by Sanderson [40], indicators are selected from both dominant and recessive aspects in accordance with the actual situation of the study area. Previous scholarly discussions have explored the quantification of human activity intensity [41]. The intensity of land use development is a direct reflection of human activities, reflecting the extent to which land resources are utilized in human endeavors. The density of impervious surfaces is a key metric reflecting the degree of artificiality and urbanization. Concurrently, the nighttime light index can, to a certain degree, directly reflect human activity levels. Furthermore, within the LNPSA, population growth and economic development are mutually reinforcing, collectively transforming the landscape both within LNP and its surroundings. Consequently, population density and economic intensity, as recessive indicators of human activity, have been incorporated into the purview of this research. After synthesizing the aforementioned analyses, this study selected five indicators to evaluate the spatiotemporal variations in human activity intensity in the LNPSA: impervious surface density (ISD), the nighttime light index (NTLI), economic density (ED), population density (PD), and land use intensity (LUI). An analytic hierarchy process was adopted to determine the weight of each index, and the weight was modified by referring to existing research [42]. Finally, a composite index method was used for weight superposition to obtain the map for human activity intensity in LNPSA in 2000, 2010, and 2020 [43]. The weights of the indices are shown in

Table 3. Weight classification of human activity intensity factors.

Human Activity Intensity Factor	Weight of Each Factor
Nighttime light index (NTLI)	0.2213
Population density	0.1721
Economic density (ED)	0.2465
Land use intensity (LUI)	0.2035
Impervious surface density (ISD)	0.1566

.

$$\mathrm{H}\mathrm{A}\mathrm{I}={\sum }_{l=1}^n{P}_l\times A_l,$$

(1)

where $P_l$ is the grid number after standardization of the index $l$ and $A_l$ is the weight that the index $l$ accounts for in the overall human activity intensity.

To further explore the spatiotemporal changes in human activity intensity in the study area, the human activity intensity from 2000 to 2020 was divided into five grades on the above basis, including regions with extremely high, high, general, low, and extremely low human activity intensity. The proportions of each grade in each period were calculated to quantify the intensity and spatial differences of human activity.

2.2.4. Correlation Analysis of Human Activity and the Landscape Pattern

To analyze the evolution and spatial distribution of the human–land relationship in LNPSA from 2000 to 2020, the Geoda1.18 software was used for bivariate local spatial autocorrelation analysis [44]. The equation is as follows:

$$I_{lm}^a=Z_l^a{\sum }_{q=1}^n{W}_{pq}{Z}_m^q,$$

(2)

In the equation, $Z_l^a=\frac{U_l^a-{\mathrm{U}}_{l\mathrm{m}\mathrm{e}\mathrm{a}\mathrm{n}}}{{\mathrm{e}}_l}$; $Z_m^a=\frac{U_m^a-U_{m\hspace{0.17em}\mathrm{m}\mathrm{e}\mathrm{a}\mathrm{n}}}{e_m}$; $X_l^p$ is the attribute $l$ value of the spatial unit $p$; $U_m^a$ is the attribute m value of the spatial unit $p$; $U_{l\hspace{0.17em}\mathrm{m}\mathrm{e}\mathrm{a}\mathrm{n}} \mathrm{a}\mathrm{n}\mathrm{d} e_l$ are the average value and variance of $l$; and $U_{m\hspace{0.17em}\mathrm{m}\mathrm{e}\mathrm{a}\mathrm{n}} \mathrm{a}\mathrm{n}\mathrm{d} e_m$ are the average value and variance of $m$, respectively.

2.2.5. Analysis of Driving Factors Based on Geodetector

Considering that the four types of human–land coupling relationships have significant differences in spatial distribution and proportion, Geodetector was employed to perform a spatial differentiation analysis of the driving factors of the four types of relationships [45]. The selected driving factors included three major types (human activity, natural conditions, and landscape pattern) and covered 14 specific indices, such as distance to the LNP, distance to the landscape resource points, vegetation coverage, and elevation (

Table 4. Driving factors code table.

Type	Driving Factors (Unit)	Serial Number
Human activity	Population density (people/km2)	X1
	NTLI (%)	X2
	ISD (%)	X3
	Gross domestic product (CNY ten thousand·km2)	X4
	Distance to the landscape resource points (km)	X5
	Distance to Lushan National Park (km)	X6
Natural conditions	Elevation (m)	X7
	Slope (degrees)	X8
	Vegetation cover (%)	X9
Landscape pattern	Shannon’s Diversity Index (SHDI)	X10
	Patch density (PD)	X11
	Landscape shape index (LSI)	X12
	Largest patch index (LPI)	X13
	Aggregation index (AI)	X14

). The four types of relationships were subjected to binarization processing and analyzed using the three modules in Geodetector after unified sampling. According to differences in the results of the q-value, the driving factors were divided into weak (0–0.1), moderate (0.1–0.2), and strong (>0.2) factors.

3. Results

3.1. Spatiotemporal Distribution of Human Activity Intensity

From 2000 to 2020, human activity in LNPSA gradually increased year by year and showed an expanding trend from cities to the peripheral areas (Figure 3). The differences in human activity intensity among different regions became increasingly prominent and showed a spreading trend from the edge areas to the core area of the park. Most areas in LNP had relatively low human activity intensity as the terrain, slope, and constraints from the protection of local government on the core area restricted the human activity in these areas to some extent. However, human activity intensity gradually rose in the edge areas of the park, forming a belt of high human activity from the mountain peak to the mountain foot, which poses potential threats to the natural landscape and ecological safety of LNP (

Table 5. Changes in human activity intensity.

Intensity Level of Human Activity	2000		2010		2020
	Area (km2)	Percentage (%)	Area (km2)	Percentage (%)	Area (km2)	Percentage (%)
Extremely low intensity of human activity	345.28	27.89	389.1	31.43	544.1	43.95
Low intensity of human activity	653.17	52.76	543.48	43.9	354.81	28.66
General intensity of human activity	165.15	13.34	180.5	14.58	135.93	10.98
High intensity of human activity	52.99	4.28	87.15	7.04	118.48	9.57
Extremely high intensity of human activity	21.54	1.74	37.76	3.05	84.68	6.84

).

3.2. Analysis of Landscape Pattern

This study explored the laws of spatiotemporal evolution of the landscape pattern in LNP and the surrounding areas from 2000 to 2020 (Figure 4). The results reveal that the SHDI generally showed a trend of positive growth in the study area over the 20 years, and LNP showed relatively lower SHDI compared with the surrounding areas, exhibiting significant differences in distribution. In general, the PD and LSI increased with time, indicating further fragmentation of the habitats and an increase in the complexity of the landscape’s internal structure, respectively. The LPI and AI showed a generally decreasing trend, reflecting the gradually reduced connectivity of landscape patches and landscape aggregation as well as the fragmentation of landscape patches [46]. Generally, there were increases in the complexity and dynamics of the overall landscape pattern, the promotion of landscape heterogeneity, and increases in human disturbance year by year in the study area [47]. Additionally, this study conducted a thorough calculation of the landscape pattern indices for both the internal and external spaces of LNP, with specific values presented in

Table 6. Changes in the landscape pattern in Lushan National Park and its surrounding areas.

	Lushan National Park			Surrounding Areas of Lushan National Park
	2000	2010	2020	2000	2010	2020
Aggregation index	96.3039	96.4634	96.5010	91.9661	90.9078	88.3806
Largest patch index	82.3288	83.2958	86.6097	26.8496	26.2077	24.0915
Landscape shape index	12.9804	12.4804	12.3626	35.4318	39.7723	41.9435
Patch density	4.0176	3.6587	3.7160	7.6431	8.0905	8.3379
Shannon diversity index	0.4560	0.4437	0.4325	1.0981	1.1409	1.3982

. A comprehensive calculation of the landscape pattern change quantities was performed. The results show that the total variation of the landscape pattern in the internal space of LNP is 5.4209, and that of the surrounding areas is 13.8502, which to some extent reveals the significant differences in the dynamics of the landscape pattern between the protected area and the surrounding areas.

3.3. Spatial Autocorrelation Analysis of Human–Land Coupling

The bivariate spatial autocorrelation analysis shows that in 2000 the regions belonging to the first quadrant were mainly concentrated in the areas of cities and towns with a relatively flat terrain, indicating that human activity causes a significant disturbance in the landscape during this period (Figure 5); in 2020, the area of regions falling into the fourth quadrant showed an expanding trend, reflecting a certain degree of deceleration in the encroachment of human activities on natural spaces. This trend may be related to the continuous implementation of China’s large-scale policy of grain for green over the past two decades.

Notably, both the density and quantity of high–high (high human activity and high landscape pattern) cluster areas in the surrounding areas showed obvious increases, indicating the occurrence of great changes in the human–land relationship in these areas, with increases in human activity intensity and significant changes in the landscape pattern.

• Correlation analysis of the coordination relationship

The coupling between human activity and the landscape pattern shows consistency, indicating a positively correlated coupling relationship, which can be classified into the following categories. The first category maintained the initial high–high or low–low clustering from 2000 to 2020, which is represented by the Taohuayuan Village. The village landscape in this period maintains a relatively primitive and natural state and showed no obvious change in structure caused by human activity, exhibiting a relatively low coupling degree. The second category was mainly distributed in the regions with a grain for green policy in the mountain areas of LNP and the “field for lake” regions in the Poyang Lake. In these regions, high human activity intensity gradually became low human activity intensity, and the landscape pattern was somewhat restored from the fragmented state. These phenomena indicate that reduction of human activity and the adoption of certain ecological measures can effectively improve the landscape pattern to promote the recovery of the ecosystem. The third category showed a transformation from low–low to high–high clustering. Despite relatively large variations, the landscape pattern developed towards a positive direction because the economic benefits created by human activity promoted regional development.

2.

Correlation analysis of the antagonistic relationship

The first category consistently showed an antagonistic state in the study period with low human activity and was faced with ecological pressure caused by overexploitation, such as abandoned land and slope cropland. The second category was represented by the excessive spreading of city and town space in the edge areas of the park, and the rising human activity intensity caused obvious negative impacts on the authenticity of LNP. The third category mainly included hollowed village agglomeration and abandoned mining land, where the reduction of human activity led to the further degradation of the local landscape pattern.

3.

Correlation analysis of the gradually antagonistic relationship

The spatial coupling showed a trend of shifting from coordination to antagonism. The first category turned from high–low to high–high clustering, just like the evolution of farmland around LNP to city and town agglomeration. The second category shifted from low–low to high–low clustering, which is mainly reflected by the extension of residential zones and large-scale construction of residential houses.

4.

Correlation analysis of the gradually coordinated relationship

The spatial coupling showed a trend of shifting from antagonism to coordination, such as scenic spots with low development degrees and agriculture tourist regions with a grain for green policy or transformation from farmland to tea cultivation in LNPSA. In these regions, the transformation and upgrading of industry as well as the introduction of modern technologies have promoted both environmental protection and economic benefits.

3.4. Assessment of Coupling between Human Activity Intensity and Landscape Pattern

A map of the human–land relationship for LNPSA areas was obtained by superposing the spatial autocorrelations of human activity intensity and the landscape pattern in 2000 and 2020 according to the transformation rules of the human–land coupling model (Figure 6). According to the analysis results of spatial distribution, there are regions with large-scale antagonistic and gradually antagonistic types of relationships in the surrounding areas, particularly in those regions adjacent to LNP, which may be attributed to the unfavorable ecological impacts caused by the boundary effects of the protected area and the continuous growth of tourist activities. In addition, the regions of cities generally showed a gradually antagonistic response, shifting from the original high–high state to a high–low state, indicating that these regions are undergoing homogenization. The country or near city regions with relatively lower terrain and certain degrees of topographic relief commonly showed antagonistic responses, which may be ascribed to the sharp contrast between the superior inherent ecological conditions and relatively lower human activities in these regions.

Further observation of the changing trend in the map, particularly the intersection areas between the protected area and cities, shows that the continuous expansion of urban built-up areas has caused significant threats to LNP. There is an urgent need for a series of feasible planning and management strategies to mitigate the pressure caused by rapid urbanization on the ecological environment in LNPSA.

3.5. Analysis of Factors Driving the Changes in Coupling between Human Activity and the Landscape Pattern

• Correlation analysis of the coordination relationship

All of the 14 driving factors in the coordinated regions passed the significance test, indicating that these factors are correlated with the spatial distribution of coordinated regions (Figure 7). Among them, the nightlight index (NTLI), impervious surface density (ISD), and elevation had relatively more powerful explanatory power regarding spatial differentiation, while population density, the distance to LNP, and the Shannon diversity index (SHDI) had relatively lower explanatory power. In the bivariate interaction analysis, GDP alone had limited explanatory power, but it showed significantly upregulated explanatory power when combined with other factors. NTLI and ISD also showed stronger explanatory power when interacting with other factors. These results indicate that in coordinated regions, the human activity intensity has great impacts on the regulation of the human–land relationship.

2.

Geographic detection analysis of antagonistic regions

In antagonistic regions, all the factors had p-values lower than 0.01, suggesting that these factors have significant correlations with the distribution of antagonistic regions (Figure 8). Among the factors of human activity, NTLI, ISD, and GDP had relatively higher q-values. The risk detection reveals that the explanatory power of spatial distribution decreased with increasing values of the factors related to human activity. Among the factors of natural conditions, elevation had the strongest explanatory power, and topographic relief and vegetation coverage also showed relatively high spatial explanatory power. In the bivariate interaction analysis, GDP and elevation showed the strongest interaction effect, and elevation generally showed strong interaction effects with other human activity factors, suggesting that the interaction between human activity and natural conditions has great impacts on the spatial distribution of antagonistic regions.

3.

Geographic detection analysis of gradually antagonistic regions

Among the correlation factors of gradually antagonistic regions, although the factors related to natural conditions and the landscape pattern had relatively low explanatory power in terms of spatial differentiation, the factors related to human activity, especially ISD, NTLI, GDP, and vegetation coverage, showed relatively high q-values (Figure 9). Among them, ISD had the highest q-value, indicating that factors related to human activity have significant correlations with the spatial distribution of gradually antagonistic regions. Risk detection reveals that the explanatory power of human activity factors increased significantly with increasing division levels. For instance, the T-value of ISD rose from 0.11 at the first stage to 0.86 at the seventh stage, indicating that human activity intensity has positive correlations with the distribution probability of gradually antagonistic regions.

4.

Geographic detection analysis of gradually coordinated regions

In gradually coordinated regions, all of the 14 factors were correlated with the spatial distribution (Figure 10). Although the X1–X6 human activity factors had relatively low explanatory power, other human activity factors except for population density and distance to LNP showed more significant correlations with the landscape pattern when the analyzed area was further divided. The factors related to natural conditions and the landscape pattern had relatively high explanatory power in terms of spatial distribution, particularly vegetation coverage. However, the explanatory power decreased with increasing vegetation coverage in the region, and in regions with low vegetation coverage, vegetation coverage had higher correlations with the spatial distribution of gradually coordinated regions. Notably, the explanatory power of the other five factors related to the landscape pattern increased with the enhancement of landscape continuity and the reduction of landscape fragmentation, indicating that integrity and continuity of the landscape pattern are a key to promote the harmony between humans and land in LNPSA.

4. Discussion

4.1. Comparative Analysis of the Internal and External Spaces of Luhsan National Park

In the protection of national parks, the surrounding landscape should also be given full consideration, and the biodiversity and ecosystem functions in the protected area are also influenced by the external landscape environment [48]. In recent years, there has been increasing research on the relationship between national parks and the surrounding environment, and integrated protection and management of the internal and external spaces of national parks as well as the promotion of internal–external coordinated development have become increasingly popular practices [49]. For the large number of protected areas in China, there are various problems, such as unclear boundaries and conflicts between protection and development, particularly in the surrounding areas with relatively more rapid economic development and weak protection. This situation requires the adoption of comprehensive management strategies that consider the internal and external spaces of national parks as a unified whole, thereby achieving integrated protection of the ecosystem of the national park and the formation of scientific and effective networks of protected areas.

The surrounding areas of LNP have a large population and frequent development and construction activities, which have led to sharp conflicts between protection and development and significant negative impacts on the overall value of natural resources in LNP. Therefore, it is important to comparatively analyze the effects of changes in human activity intensity and the landscape pattern on LNPSA against the background of urbanization from a perspective of coordinated management of the national park and the surrounding areas, which can further reveal the impacts of the surrounding areas on LNP from a perspective of human–land coupling. The results indicate the following:

(1)

The human activity intensity in the surrounding areas of LNP gradually increases with time and shows a trend of expanding to the core protected area. In sharp contrast, the human activity intensity in the internal space of LNP is relatively lower and displays an evident declining trend in recent years, indicating that the current protection measures have achieved certain effects.

(2)

Compared with its surrounding areas, LNP has relatively lower landscape diversity, patch density, and shape index, indicating that more rigid protection measures have been implemented in the national park, which have maintained relatively higher ecological integrity and landscape continuity in the short run. In the surrounding areas, there are relatively higher degrees of landscape fragmentation, and both the continuity and integrity of the ecosystem are affected. These phenomena indicate that although a relatively good ecological status is maintained in LNP, the protection strategies for the surrounding areas still require certain adjustments to strengthen supervision so as to reduce the negative impacts on LNP.

(3)

Combined with the analysis results of human–land coupling, it can be indicated that the antagonistic and potentially antagonistic regions are geographically expanding towards the internal space of the park, which may be closely related to the intensification of human activity and changes in land use during the urbanization process. These changes intensify the contradiction between humans and nature, posing significant challenges to the protection and ecological safety of LNP. Hence, more effective ecological protection strategies should be taken to alleviate the expansion of antagonistic regions, so as to maintain the ecological balance and sustainable development of LNPSA.

4.2. Optimization of the Zoning Plan of Lushan National Park and Its Surrounding Areas

Zoning planning is an important means of spatial differential management and the achievement of multiple management purposes for national parks [50]. It is compulsory to explore zoning planning strategies with higher adaptability, rationality, and efficiency under the emerging sharp human–land contradiction in national parks and the surrounding areas. Currently, relevant research and practices have considered the needs of zoning plans in protected areas, but there has been a lack of in-depth analysis of the zoning patterns with different research scopes and different management goals [51]. Therefore, this study further explored the interaction relationship between human activity and the landscape pattern in the study area for more specific functional zoning by comprehensively considering the ecological and social attributes of LNP [52].

Because the human–land coupling assessment map takes the assessment units with grids as the carriers, and there is no specific spatial boundary, this study adopted administrative boundaries as the basic framework and combined the systematic human–land coupling analysis, aiming to achieve more specific spatial zoning. The study divided the study area into four functional zones, including harmonious development zones, stable transformation zones, environmental regulation zones, and risk prevention zones (Figure 11). This zoning scheme can facilitate the fine management and formulation of targeted strategies, providing a scientific foundation for the differential management and protection of protected areas.

For the above zoning pattern, “one policy for one zone” landscape management and protection strategies can be proposed. Harmonious development zones are located within the boundary of LNP, where humans and the land have a coordinated relationship owing to the support of the government and strict management and protection measures. In these zones, more positive protection strategies should be adopted to encourage green development so as to ensure a harmonious human–land relationship and long-term stable development. Stable transformation zones are mostly located in the internal space of the national park and cities and towns adjacent to the protected area, where the human–land relationship is being continuously improved. It is suggested that relevant sectors and responsible parties should pay more attention and provide more supports for the targeted adjustment of land use and strengthen environmental monitoring and management so as to further improve the ecological conditions. In environmental regulation zones, there are obvious human–land contradictions, and therefore there is a need for emergent risk assessment and management measures, including strict environmental legislation and a strengthened delimitation of ecological protected areas, so as to reduce the human–land contradiction and relieve ecological pressure. Risk prevention zones indicate that there are potential contradictions between humans and land in the zones. Therefore, more importance should be attached to these zones, and more prospective planning and management strategies should be formulated and advanced monitoring techniques should be employed for early warning and alleviating the potential ecological threats so as to enhance the ecological resilience of these zones and promote the coordination of socioeconomic development and ecological and environmental protection.

It is noteworthy that the zoning management strategy proposed in this paper is a targeted solution to the human–land contradiction in the research area based on a comprehensive consideration of the unique geographical environment, ecosystem, and socioeconomic conditions of LNPSA. Given the diversity of human activities and natural environmental conditions in different regions, this strategy requires adaptive adjustments when promoted to other areas. It is recommended to establish a comprehensive assessment system based on the specific scale characteristics, landscape types, and research objectives of the region to determine the protection priorities of different areas, thereby ensuring its effectiveness and applicability to the greatest extent. Although the zoning management strategy proposed in this study is aimed at a specific research object, its main idea, that is, to adopt differentiated management measures according to the characteristics of human–land relations in different regions, still has a certain degree of foresight and universality, providing a flexible strategic framework for a wider range of areas. It provides a scientific basis for the protection of various landscape types, including rural landscapes, wetland landscapes, and polder landscapes, in the future.

4.3. Limitations and Future Work

Current research has explored the response of the landscape against the background of changes in human activities and human–land interaction [53,54]. However, these theoretical studies are not associated with specific spatial planning and management, which may have led to the poor connection between research and practice, and they are not conducive to the management of dynamic changes and sustainable development of the landscape. This study attempts to overcome the limitations and application obstacles of current research and combines the administrative division and coupling analysis results to propose zoning optimization strategies on the basis of existing research. This combination of academic research and current administrative division can provide meaningful guidance for actual landscape management [55]. However, administrative division faces a lack of in-depth understanding of the natural ecological process and human activity patterns, which tends to cause the neglect of the uniqueness and heterogeneity of the landscape in the study area, thereby influencing the precision and implementation effect of the zoning strategies. Moreover, administrative division usually does not completely match natural boundaries, which may limit the protection of continuity and integrity of the protected area.

With the development and improvement of conceptions in landscape management and protection, landscape character assessment (LCA) has become an effective tool to understand the content, manage the changes, and determine the value of the landscape, which has been widely used in research and practice worldwide [56]. In ample research on LCA, some research has included LCA into the decision-making framework for the optimized zoning of protected areas, providing different protection strategies for each zone with different landscape characteristics [57,58]. With characteristic landscape units as the carrier, the assessment of the human–land coupling for each unit can realize the differential management and protection of different landscape units. Certainly, this method also has certain limitations. For instance, it may require more complex data and technical deduction, and it is more difficult to be realized in some situations. In addition, because administrative boundaries always affect the planning and management of protected areas in the dominant dimension, it is highly necessary to consider the coordination and connection with administrative boundaries when implementing zoning management strategies based on landscape units with different characteristics. The zoning schemes singly depending on landscape characteristics are not necessarily applicable in real planning of protected areas. Therefore, to achieve more effective management and protection of the regional landscape, it is necessary to compare the two zoning strategies and fully utilize the advantages of administrative boundaries in organization and management in the future, and, at the same time, understand the unique value of the landscape by integrating the LCA method, thereby achieving effective management and protection of the protected areas.

5. Conclusions

Under rapid urbanization, there have been constant human–land contractions in the surrounding areas of LNP, which have posed serious threats to the integrity and authenticity of the national park. Under this context, this study selects LNPSA as the research object and constructs a spatiotemporal coupling model of human activity and the landscape pattern from 2000 to 2020 in the study area. The study further uses bivariate spatial local autocorrelation analysis to reveal the spatial coupling characteristics between human activity intensity and the landscape pattern index and analyzes the driving factors for the four types of coupling relationships. The findings may have implications for the targeted management and protection of national parks. The study preliminarily explores the reasonable quantification of human activity and landscape variations and the dissection of the underlying cause-and-effect mechanism underlying complex human–land interactions so as to realize the effective protection and sustainable development of the ecological environment of LNPSA. The meaning of this exploration lies in the provision of a multi-dimensional analytical method that integrates human activity and natural environmental factors for understanding the mutual and dynamic relationship between humans and nature. In addition, the study highlights the significance of the surrounding areas of national parks and gives full consideration to the impacts of human activity in the surrounding areas on the protected area. It also proposes relatively unified zoning principles for the internal and external spaces of national parks, thereby better supporting the formulation and implementation of relevant policies. In general, human activity and landscape patterns are, respectively, the social and ecological characteristics, and their interaction has a great influence on regional ecological protection and economic development. Effective ecological restoration measures can promote the coordinated development of human–land relationships, while irrational urban expansion and land exploitation may lead to conflicts between protection and development and ecological degradation. The coupling model and the zoning mode proposed by this study fill the gap in research on the response of human–land coupling in national parks and provide a basis for formulating effective protection and management measures, demonstrating a certain degree of foresight. The research findings can guide relevant departments in formulating land use policies, ecological compensation mechanisms, and environmental supervision measures, to more scientifically balance the needs of protection and development and mitigate the increasingly prominent negative impacts of human–land conflicts on natural landscapes.