Spatial-Temporal Pattern and Influencing Factors of Land Ecological Carrying Capacity in The National Pilot Zones for Ecological Conservation in China

Abstract

Improving land ecological carrying capacity (LECC) is important in accelerating the realization of national ecological civilization construction goals. Based on the panel data of the first batch of prefecture-level cities in the National Pilot Zones for Ecological Conservation initiative from 2005 to 2019, this study analyzes the spatial–temporal pattern of LECC using the improved ecological footprint model, Theil–Sen’s slope estimator and Mann–Kendall test, and investigates the influencing factors of LECC using the geodetector. Results show that the overall land ecological carrying status of each province tends to improve but also shows remarkable interprovincial differences in development trend, with Guizhou outperforming Jiangxi and Fujian in general. The pattern of LECC security has apparent regional heterogeneity. Most prefecture-level cities have high ecological pressure and uneven spatial distribution but slowly improve overall. The influencing factor of forest land coverage and population density has strong explanatory power on the LECC, and the interactions among the factors are enhanced. The four aspects of land ecological construction should be carried out. A first step is to strengthen land ecological management and optimize the land use practices actively. Second, modern technology is used to establish real-time monitoring and early warning systems for LECC security. Third, the two key factors of forest land coverage and population density should be focused on, and enhancing their positive interaction with industrial structure and arable land utilization rate. Finally, the experience of model construction should be promoted in the Non-national Pilot Zones for Ecological Conservation in China. The aim is to enhance the effectiveness of land ecology measures further and promote the construction of national ecological civilization in China.

1. Introduction

As a special natural synthesis, land plays the role of carrying all natural things and is the material basis of all means of production. However, along with China’s rapid economic development, a series of land ecological problems has emerged, such as the results of the Fifth National Monitoring of Desertification and Sanding, which showed that the national desertification and sand-affected land areas reached 261.16 and 172.12 million ha, respectively, accounting for 27.20% and 17.93% of the national land area [1]. In 2016, China issued Opinions on the Establishment of Unified and Standardized National Pilot Zones for Ecological Conservation. Jiangxi, Fujian, and Guizhou, which have a better ecological foundation and stronger resource and environmental carrying capacity, are among the first batch of the National Pilot Zones for Ecological Conservation, aiming to form a construction experience that can be replicated and promoted nationwide. Efficiently using and protecting land resources, enhancing the LECC, and achieving intra-generational and inter-generational equity in the use of land resources have become important issues in the current construction of ecological civilization in China.

The concept of carrying capacity was originally applied to engineering geology, referring to the strength of the foundation’s load-bearing capacity for buildings. It was subsequently introduced to the field of ecology, forming the concept of ecological carrying capacity [2]. In 1921, Park [3] extended the carrying capacity to human ecology, and considered the ecological carrying capacity as the maximum number of individual organisms that can survive under specific conditions. In 1922, Hadwen and Palmer [4] considered the subject based on Park’s view, where the ecological carrying capacity is the maximum number of organisms that can be accommodated without destroying the ecosystem; to some extent, this anticipated the concept of sustainable development. Since 1930, with the development of agricultural productivity and medical technology, the food supply capacity, number of people, and human life expectancy have increased, leading to environmental pollution and ecological imbalance. Research related to ecological carrying capacity has gradually shifted from biological populations to resource and environmental problems faced by human society [5,6], especially the problem of the ecological carrying capacity of land resources. Currently, the research on land ecological carrying capacity (LECC) has become a hot spot in ecology, economics, geography, and other fields [7,8,9,10]. In terms of research content, scholars mainly focus on the comprehensive measurement and evaluation of regional LECC from the concept of sustainable development [11,12] and the analysis of the influence of various factors on regional LECC [13,14]. In recent years, some scholars have also explored LECC in terms of the flow and stock of natural capital [15]. The nonrenewable characteristic of fossil fuel land is still less considered, and analysis of LECC security from the perspective of land type is limited. In terms of research scales, the existing studies cover a wide range, including global scale [16,17], national scale [18,19], urban scale [20,21], and watershed scale [22,23]. Among them, the research achievements on the LECC of a single city at urban scale are abundant, and relevant studies currently shift from the single-city scale to the urban-cluster scale. The main research methods are the ecological footprint method [24,25], system model method [26,27], and comprehensive evaluation method [28,29]; of these, the ecological footprint method has the advantages of a comprehensive evaluation index, calculation and wide application, and has become the mainstream method to measure LECC.

National Pilot Zones for Ecological Conservation constitute the testing field for the construction of ecological civilization in China, carrying the critical mission of accumulating experience, gathering reform power, and promoting economic green development for the construction of ecological civilization, thereby directly affecting the sustainable and harmonious development of the national ecological civilization. However, studies on the LECC of the National Pilot Zones for Ecological Conservation are still limited. Therefore, this study considers the prefecture-level cities of Jiangxi, Fujian, and Guizhou, the first batch of the National Pilot Zones for Ecological Conservation in China, as the study areas. The improved ecological footprint model, Theil–Sen’s slope estimator and Mann–Kendall test were used to measure the ecological carrying capacity of various types of real land and analyze its spatial–temporal pattern; a geodetector approach was used to explore the influencing factors of LECC.

Accordingly, the possible contributions of this study are as follows: first, to a certain extent, it fills the shortage of existing literature on the LECC in the National Pilot Zones for Ecological Conservation in China. Second, the ecological footprint model is improved and made more suitable for the National Pilot Zones for Ecological Conservation in China. Third, this study examines the evolution characteristics of the spatial pattern of LECC security from the perspective of land types, achieving a comprehensive research analysis of the LECC of the National Pilot Zones for Ecological Conservation in China.

The paper is organized as follows. Section 2 introduces the research materials and methods. Section 3 presents the study results. Section 4 discusses the results, policy recommendations, limitations, and future directions. Section 5 draws conclusion from the results.

2. Materials and Methods

2.1. Study Area

The critical national strategy of the National Pilot Zones for Ecological Conservation was initially proposed by the Fifth Plenary Session of the 18th CPC Central Committee. The first batch of National Pilot Zones for Ecological Conservation was established in Fujian, Jiangxi, and Guizhou from 2016 to 2018 to form a new ecological civilization construction path that can be promoted and replicated nationwide through the construction of pilot zones. The approach has great practical and theoretical importance for transforming China’s economic development mode, improving the ecological environment, and realizing the sustainable development concept of lucid waters and lush mountains as invaluable assets [30]. Therefore, this study analyzes and evaluates the LECC of 26 prefecture-level cities (excluding autonomous prefectures) in the first batch of the National Pilot Zones for Ecological Conservation (Figure 1).

2.2. Index System and Data Sources

Based on the 2D ecological footprint evaluation index system proposed by Rees [31] and Wackernagel [16] and considering the actual situation of the National Pilot Zones for Ecological Conservation, the index of contents of the biological resource account and the construction land account are determined (

Table 1. Index System of LECC.

Type of Account	Evaluation Indicators	Type of Land
Biological resources account	Grain, Oilseeds, Vegetables, Cotton, Sugarcane, Tobacco	Arable land
	Meat, Milk, Poultry and Eggs	Grassland
	Tea, Wood, Fruit	Forest land
	Aquatic products	Waters
Construction land account	Total electricity consumption	Construction land
Pollution account	Industrial solid wastes	Arable land
	Industrial sulfur dioxide, Industrial dust	Forest land
	Industrial wastewater	Waters

). At the same time, the pollution account is added, and the fossil fuel account is deleted for two main reasons, as follows: first, considering ecological pollution can measure the ecological footprint more comprehensively and accurately. Second, fossil fuel land is virtual land without corresponding land-carrying capacity; it is essentially different from real lands, such as arable land and forest land [32]. In addition, forest land can absorb multiple pollutants simultaneously; thus, when measuring the ecological footprint in the pollution account, only the pollutant with the largest ecological footprint value is calculated. The original use of arable land and waters is destroyed after absorbing the pollutants; thus, the sum of the ecological footprint of the biological resource account and the pollution account is obtained.

Data on the study area included socioeconomic data from 2005 to 2019, conversion parameters, and land use change data in 2005, 2010, 2015, and 2018. Among them, the socioeconomic data of population, land area, production, and consumption were obtained from provincial and municipal statistical yearbooks, statistical bulletins, Chinese city statistical yearbooks, and government websites. The land use change data were obtained from the Resources and Environment Science and Data Center of the Institute of Geographic Sciences and Natural Resources Research. The equivalence and yield factors were obtained from the research results of Liu [33,34]. The average production capacity was mainly calculated from the database of FAO and its related data, except for the average production capacity of aquatic products which was drawn from the research of Xie and Ye [35]. The pollutant conversion parameters were drawn from the related literature [32,36].

In addition, as long as biological resources are produced in the country, the domestic ecologically productive land is occupied, thereby affecting the domestic ecological environment; thus, the biological resource data use production data [37]. The unit of wood and total electricity consumption are cubic meters and kilowatt–hours respectively, and the units of other products and pollutants are kilograms. For random missing data or abnormal data in the panel data, linear interpolation was used to complete or replace that data. Data of each province were obtained by averaging the data of prefecture-level cities each year. Software used for data processing were SPSS, Matlab, Arc GIS, and Geodetector Software in Excel.

2.3. Methods

2.3.1. Improved Ecological Footprint Model

The 2D ecological footprint model is a quantitative research method that converts the earth’s resources consumed and waste produced by humans to sustain their survival and development to equivalent biologically productive land areas. The model usually includes ecological footprint, biocapacity and ecological pressure. The ecological footprint accounts for the degree of human impact on natural ecosystems. Biocapacity accounts for the maximum amount of resources that can be provided to satisfy the requirements of human survival and development. Ecological pressure is an important indicator reflecting the LECC security pattern, showing the degree of human interference with the natural system; a larger value indicates greater ecological pressure [38]. The calculation formula for the indicators is as follows:

$$EF=r_j\ast {\displaystyle \sum _{\mathrm{i}=1}^{n}a{a}_i}=r_j\ast {\displaystyle \sum _{\mathrm{i}=1}^n(c_i}/p_i)$$

(1)

$$BC=(1-12\%)\ast r_j\ast y_j\ast {\displaystyle \sum _{\mathrm{j}=1}^n{a}_j}$$

(2)

$$EP=EF/{\rm B}C$$

(3)

where EF is ecological footprint per capita; r_j is the equivalence factor of class j biologically productive land; aa_i is the converted biologically productive land area per capita of class i traded goods; c_i is the consumption per capita of class i traded goods; p_i is the average production capacity of class i traded goods; BC is biocapacity per capita; y_j is the yield factor of class j biologically productive land; a_j is the area of class j biologically productive land per capita; EP is ecological stress index.

The 3D ecological footprint model is based on the 2D ecological footprint model, adding the concepts of footprint depth and footprint breadth. The footprint depth has the property of time, which indicates the degree of human consumption of natural resources over time, the multiple land areas occupied to satisfy human survival and development, or the time when the consumed natural resources are reproduced. The footprint breadth has the property of space, which indicates the area of human occupation of land resources and is closely related to the stock of land resources in the study area. The 3D ecological footprint model expands the depth of the longitudinal study of land ecology from natural capital stock and flow; it overcomes the problem that the 2D ecological footprint model cannot reflect the role of natural capital stock on the ecological balance of the earth [39]. The calculation formula is as follows:

$$E{F}_{depth}=1+{\displaystyle \sum _{i=1}^n\max ({\rm E}{F}_i-{\rm B}{C}_i,0)}/{\displaystyle \sum _{i=1}^{n}B{C}_{\mathrm{i}}}$$

(4)

$$E{F}_{size}={\displaystyle \sum _{i=1}^n\min ({\rm E}{F}_{\mathrm{i}},{\rm B}{C}_{\mathrm{i}})}$$

(5)

$$E{F}_{3\mathrm{d}}=E{F}_{depth}\ast E{F}_{size}$$

(6)

where EF_depth is the depth of ecological footprint per capita, EF_i is the ecological footprint per capita of class i land; BC_i is the biocapacity per capita of class i land; EF_size is the footprint breadth per capita; EF_3d is the 3D footprint per capita.

2.3.2. Time Series Analysis Methods

In this study, Theil–Sen’s slope estimator and Mann–Kendall test were used in turn for time series feature analysis. Theil–Sen’s slope estimator is a robust and efficient trend analysis method often used for trend analysis of long time series data [40], usually in combination with the Mann–Kendall test. The Mann–Kendall test is a nonparametric statistical method based on the rank order of the original time series [41], which does not require the original data to obey a specific distribution and is not affected by a few outliers; it can further analyze the significance of the data change trend on the basis of the trend analysis of Theil–Sen’s slope estimator. The formula is as follows [42,43]:

$$S_R=\mathrm{median}(\frac{x_j-x_i}{j-i})$$

(7)

$$Z=\left\{\begin{array}{c}\frac{S-1}{\sqrt{{}_S(\mathrm{S})}},S>0\\ 0,S=0\\ \frac{S+1}{\sqrt{{}_S(\mathrm{S})}},S<0\end{array}\right.$$

(8)

$$S={\displaystyle \sum _{\mathrm{j}=1}^{n-1}{\displaystyle \sum _{\mathrm{i}=\mathrm{j}+1}^n\mathrm{sgn}(x_j}}-x_i)$$

(9)

$${}_S(S)=\frac{n(n-1)(2n+5)}{18}$$

(10)

$$\mathrm{sgn}(x_j-x_i)=\left\{\begin{array}{c}1,x_j-x_i>0\\ 0,x_j-x_i=0\\ -1,x_j-x_i<0\end{array}\right.$$

(11)

where S_R is Sen’s slope, x_j and x_i are the data at time j and i (j > i), Z is significance level, n is the number of data, and sgn is symbolic function.

When data are increasing, S_R > 0, and vice versa; |Z| > 2.58 shows an extremely significant change in trend; 1.96 < |Z| ≤ 2.58 shows a significant change in trend; |Z| ≤ 1.96 shows no significant change in trend.

2.3.3. Geodetector

The geodetector is a statistical method to detect spatial heterogeneity and reveal the driving force behind it. The model calculation results include four parts, namely, factorial, interactive, ecological, and risk detection, and its mathematical principles are referred from the related literature [44]. Among them, the results of factorial detection include p–value and q–statistic. The p–value reflects the significance of the influence of the independent variable on the dependent variable. The q-value reflects the explanatory power of the independent variable on the dependent variable. Interactive detection reveals the degree of influence on the dependent variable under the interaction of multiple independent variables. Compared with the principal component analysis model, traditional multiple linear regression, and other research methods, geodetector has three characteristics. First, it is still applicable when the sample size is less than 30. Second, it can detect the influence of the interaction of independent variables on the dependent variable without being limited to the prespecified multiplicative interaction. Third, the geodetector can be immune to the covariance of multiple explanatory variables [44]. The geodetector is widely used in the natural sciences, social sciences, human health, and other fields, given these advantages.

LECC is the basis of natural resource endowment [45] and is closely related to the socioeconomic level. The mutual coordination and development of LECC, socioeconomic level, and resource endowment will be conducive to the sustainable development of intra-generational equity and inter-generational equity in land resource utilization. To explore the influencing factors of LECC, the 3D ecological footprint is considered a dependent variable, the previous research results [46,47,48,49] are combined, and eight influencing factors from three dimensions are selected, including social, economic and resource dimensions, to establish the factor system (

Table 2. Evaluation index system of influencing factors of LECC.

Index Types	Index Name	Index Meaning
Society	Urbanization level (UL)	Urban resident population/Resident population
	Population density (PD)	Resident population/Administrative area
Economy	Regional gross domestic product (RGDP)	Regional gross domestic product
	Total retail sales of social consumer goods (TOG)	Total retail sales of social consumer goods
	Industrial structure (IS)	Value added of secondary industry/Regional gross domestic product
	Investment intensity (II)	Fixed Assets Investment/Regional gross domestic product
Resource	Arable land utilization rate (ALUR)	Food production/Arable land area
	Forest land coverage (FLC)	Forest land area/Administrative area

). First used is K-means clustering, which is a kind of nonsystematic clustering with fast operation speed and is one of the most widely used algorithms in cluster analysis; this serves to discrete factors into type variables. Then, the influencing factors of LECC are explored using Geodetector Software in Excel. In addition, the specific factor descriptions are as follows:

(1)

Urbanization level. Population urbanization is the core of urbanization, and land urbanization is the carrier of urbanization. The development of population urbanization has greatly increased the land urbanization level. The ratio of urban resident population to resident population in the area was selected to characterize the urbanization level.

(2)

Population density. Population density reflects the degree of spatial concentration of regional population; a too high spatial concentration of population exerts great pressure on land ecology. The ratio of resident population to administrative area was selected to characterize population density.

(3)

Regional gross domestic product. The regional gross domestic product reflects the level of regional economic development; the higher the level of economic development, the more evident is its influence on land use.

(4)

Total retail sales of social consumer goods. Total retail sales of consumer goods is the most direct data showing domestic consumption demand, and many consumer goods are directly or indirectly provided by land.

(5)

Industrial structure. Numerous high-pollution and high-consumption industries exist in the secondary industry, and these industries have a great impact on the land ecological environment. The ratio of the value added of secondary industry to regional gross domestic product was selected to characterize the industrial structure.

(6)

Investment intensity. Part of the fixed asset investment occupies or pollutes the land, and the land itself can be used as fixed assets for investment. The ratio of fixed assets investment to regional gross domestic product was selected to characterize the investment intensity.

(7)

Arable land utilization rate. The higher the degree of arable land utilization, the greater the possible impact to the ecology of arable land resources. The ratio of food production to arable land area was selected to characterize the arable land utilization rate.

(8)

Forest land coverage. Forest land has the functions of soil and water conservation, wind breaking, and sand fixing, etc. and has a good protection effect on land resources. The ratio of forest land area to administrative area was selected to characterize the degree of forest land coverage.

3. Results

3.1. Temporal Dynamic Evolution Characteristics

3.1.1. Evolutionary Characteristics of Ecological Footprint

The ecological footprints per capita in Jiangxi, Fujian, and Guizhou from 2005 to 2019 were calculated by Equation (1), and the ecological footprints per capita in the three provinces exhibited the overall characteristics of two declines and one increase (Figure 2).

The two declining trends suggest a decline in the ecological footprint per capita in Guizhou and Jiangxi from 2005 to 2019 (Guizhou’s S_R = −0.03, Jiangxi’s S_R = −0.01). The overall ecological footprint in Guizhou showed an extremely significant decline trend (Z = −2.97), but the decline trend of ecological footprint in Jiangxi was not significant (Z = −1.19). In terms of partial trends, both provinces showed a significant decline trend starting around 2015, indicating that LECC in the two provinces had improved to varying degrees. The decline of forest land ecological footprint in Guizhou is the main factor leading to the decline of its total ecological footprint, with an average annual decline rate of 10.36%. The influencing factors of the decline in ecological footprint in Jiangxi are waters and forest land, among which forest land is the dominant influence. The average annual decline rate of forest land ecological footprint in Jiangxi was 9.18%, the average annual decline rate of water ecological footprint was 2.15%, and the comprehensive decline rate of the two reached 8.46%. Although the ecological footprint of arable land, grassland, and construction land in Jiangxi showed a significant upward trend (Z > 1.96), the comprehensive average annual growth rate of the three was only 1.39%. Therefore, the overall ecological footprint in Jiangxi showed a decline trend, but the trend was not considerable.

The data suggests that Fujian’s ecological footprint per capita presents an overall upward trend (S_R = 0.005), but the upward trend is not significant (Z = 1.29). In terms of land type, Fujian is in the coastal area, the supply of aquatic products is relatively stable, and the overall change of its waters’ ecological footprint is small. The ecological footprint of forest land and arable land showed a decreasing trend, but the trend was not significant, whereas the ecological footprint of grassland and construction land showed an extremely significant upward trend (Z > 2.58), with an average annual comprehensive growth rate of 3.24%, which determined the rising characteristics of ecological footprint per capita in Fujian. Although Fujian’s ecological footprint is generally high, its ecological footprint has also shown a rapid downward trend similar to the other two provinces around 2015, indicating that Fujian’s ecological civilization construction still has had a certain effect.

3.1.2. Evolution Characteristics of Biocapacity

The per capita biocapacity of Guizhou, Fujian, and Jiangxi was calculated by Equation (2). The results showed that the evolution trend of land biocapacity per capita in each province was stable (Jiangxi’s Z = 0.00, Fujian’s Z = 0.50, Guizhou’s Z = 0.40), and no significant change trend was found (Figure 3). On the one hand, limited by the natural environment, the level of science and technology development and other realistic conditions, changes to the land available for human use are limited; thus, the biocapacity did not change dramatically. On the other hand, different types of land vary in production capacity; land use change leads to changes in regional land biocapacity. The land biocapacity in Jiangxi, Guizhou, and Fujian is mainly based on arable land and forest land, which largely determine the land biocapacity of the province. Under the increasingly strict protection policies of arable land and forest land, such as the arable land red line and the comprehensive logging ban of nature forests, the regional biocapacity will reach the optimum and then become increasingly stable.

3.1.3. Evolutionary Characteristics of Natural Capital

The per capita footprint depth and per capita footprint breadth of Jiangxi, Fujian, and Guizhou were calculated by Equations (4) and (5), respectively (Figure 4).

From 2005 to 2019, the footprint depth per capita of each province presented decline characteristics (Jiangxi’s S_R = −0.04, Fujian’s S_R = −0.05, Guizhou’s S_R = −0.06) and showed an extremely significant decline trend (Jiangxi’s Z = −2.77, Fujian’s Z = −3.07, Guizhou’s Z = −2.97), indicating that the consumption of natural capital stock in each province showed a decreasing trend, which greatly changed the past mode of sacrificing ecological environment for social and economic development. Fujian’s footprint depth ranks first among the three provinces, with an average annual decline rate of 2.22%; the footprint depth of prefecture-level cities is basically in the range of 1.50–4.00, except for Xiamen. Jiangxi has the second-largest footprint depth among the three provinces, with an average annual decline rate of 1.87%. The footprint depth of most prefecture-level cities is basically in the range of 1.50–3.00, among which the footprint depth of Pingxiang, Xinyu, and Nanchang shows evident stages characteristic of decline–rise–decline. The footprint depth of Guizhou is the smallest among the three provinces, with an average annual decline rate of 3.80%, which was reduced to 1.50 by 2019. Among them, the footprint depth of Liupanshui and Anshun fluctuated greatly and began to decline after reaching a peak around 2011. In terms of footprint depth value, the current footprint depth of each province and prefecture-level city is still greater than 1, indicating that the current natural capital flow still cannot meet the human demand for resources and needs to consume the natural capital stock or call on resources from other regions to meet the development needs.

The footprint breadth per capita in each province from 2005 to 2019 generally shows a decline characteristic (S_R < 0), among which the footprint breadth of Jiangxi and Fujian shows a significant decline trend (Z = −2.38), whereas that of Guizhou shows an insignificant decline trend (Z = −0.20). These results indicate that Jiangxi and Fujian show a decreasing trend in the occupation of natural capital flows in the process of social and economic development, whereas the footprint depth was at a high level, which also reflected the current low utilization level of natural capital flows. From the perspective of prefecture-level cities, except for Ganzhou, which shows a non-significant upward trend in footprint breadth, all other prefecture-level cities in Jiangxi show a decline trend in footprint breadth, especially Ji’an and Shangrao, which have an extremely significant decreasing trend in footprint breadth. All the prefecture-level cities in Fujian show extremely significant decreasing trends, except for Sanming, Ningde, and Putian, where the changes of footprint breadth are not significant, suggesting that Jiangxi and Fujian prefecture-level cities generally over-rely on natural capital stocks and underutilize natural capital flows. Anshun, Bijie, Zunyi, and Liupanshui in Guizhou show an increasing trend in footprint breadth, whereas Guiyang and Tongren are in an extremely significant decreasing trend in comparison with the gradual decline of footprint depth. This result indicates that most of the prefecture-level cities in Guizhou have improved the efficiency of natural capital flow utilization and optimized the natural capital utilization.

3.2. Spatial Evolution Characteristics

Ecological pressure is an important indicator reflecting the regional LECC security pattern. WWF (World Wildlife Fund International) made a standard ecological pressure index classification in 2004, dividing ecological pressure on arable land, forest land, grassland, water, and construction land into six classes (

Table 3. Ecological pressure index grade.

Pressure Grade	1	2	3	4	5	6
Pressure Index	(0, 0.5)	(0.5, 0.8)	(0.8, 1)	(1, 1.5)	(1.5, 2)	(2, +∞)
Pressure state	Very	Relatively	Slightly	Less	Very	Extremely
	secure	secure	insecure	secure	insecure	insecure

). The ecological stress index was calculated by Equation (3), and the calculation results showed that each province’s overall ecological stress index was large, and the LECC security situation was severe. From 2005 to 2019, the average ecological stress index of 26 prefecture-level cities in Jiangxi, Fujian, and Guizhou was greater than 1.0. Among them, the ecological pressure index of five cities is 1.0–1.5, including Ganzhou, Shangrao, Ningde, Zunyi, and Bijie (accounting for 19.23% of the total number of cities), which are in a relatively insecure state; The stress index of seven cities was 1.5–2.0, including Jiujiang, Ji’an, Fuzhou, Nanping, Longyan, Anshun, and Tongren (accounting for 26.92% of the total number of cities), which is a very insecure state. The 14 remaining cities have an ecological pressure index greater than 2.0, including Nanchang, Fuzhou, Guiyang, etc., and account for 53.85% of the total number of cities, which are in a very insecure state.

Although the ecological pressure is generally high in the three provinces, the analysis of ecological pressure in each prefecture-level city in 2005, 2010, 2015, and 2018 by ArcGis software revealed that the spatial evolution of ecological pressure in each province showed a moderating feature (Figure 5).

The ecological pressure in Jiangxi shows the differentiation characteristics of high in the north and low in the south, tending to be small in general. In 2005, the ecological pressure of seven cities in northern Jiangxi, including Nanchang, Yichun, Pingxiang, Jingdezhen, and Yingtan, was mainly at fifth to sixth grade in a very insecure or extremely insecure state, whereas the ecological pressure in southern Jiangxi was mainly at first grade, in a very secure state. From 2005 to 2018, the ecological pressure in northern Jiangxi decreased yearly. In 2018, the number of cities, which were very insecure or extremely insecure in northern Jiangxi declined from seven to two, and the five remaining cities generally returned to a very secure or relatively secure state, indicating that the ecological security in northern Jiangxi has been significantly improved. From 2005 to 2018, the ecological pressure in southern Jiangxi changed steadily and was generally in a very secure state, but a small area was very insecure. The ecological security in southern Jiangxi, as the green ecological barrier of Jiangxi, should attract further attention.

The ecological pressure in Fujian shows the differentiation characteristics of zonal clustering and overall improvement. In 2005, the ecological pressure in Fujian was geographically clustered in seven zoning areas. Among these, Zhangzhou and Nanping are mainly in the less secure states, and Longyan is mainly in the slightly insecure state, Ningde is mainly in the very secure state, Sanming and Quanzhou are in the very insecure state, Xiamen, Putian, and Fuzhou are in the extremely insecure state. The pattern of ecological security in Fujian changed dramatically from 2005 to 2018. The ecological pressure in 2018 showed the characteristics of nine zoning areas. The ecological pressure differentiation among cities was further highlighted, but the overall ecological pressure decreased. Except for Zhangzhou and Ningde, the ecological pressure in all other cities decreased to different degrees. For example, Putian has changed from the extremely insecure state to the relatively secure state, and Nanping has changed from the less secure state to the slightly insecure state, indicating that Fujian has made some achievements in ecological security construction.

The ecological pressure in the study area of Guizhou generally shows the characteristics of high in the south, low in the north, and fluctuating and slowing down. The ecological pressure in the southern region of the study area of Guizhou in 2018 is significantly higher than that in the north, and the ecological pressure in Liupanshui and Anshun is at the fifth to sixth grade, in the very insecure or extremely insecure state. However, that in Zunyi and Tongren in the northern region is mainly at first to third grade, in the very secure state, relatively secure state or slightly insecure state. In addition, compared with 2005, the ecological pressure of the study area tends to be smaller in 2018, among which the ecological pressure in Zunyi in the northern region fluctuates within the range of first to third grade, whereas the ecological pressure in Tongren changes from extremely insecure to relatively secure in general. The ecological pressure in Anshun and Guiyang in the southern region changes from extremely insecure to very insecure and less secure respectively in general, indicating that the ecological security pattern has been optimized.

3.3. LECC Influencing Factors

The results of factorial detection indicate that, except for industrial structure and investment intensity, the influencing factors passed the significance test. Urbanization level, population density, regional gross domestic product, total retail sales of social consumer goods, arable land utilization rate and forest land coverage passed the 1% significance test (

Table 4. Geodetector factorial detection results.

Index name	UL	PD	RGDP	TOG	IS	II	ALUR	FLC
q–statistic	0.117 ***	0.544 ***	0.128 ***	0.152 ***	0.055	0.038	0.175 ***	0.549 ***

Note: *** p < 0.01.

). In terms of resource factors, forest land resources and arable land resources occupy more than 80% of the productive land area, and China is a large grain-producing country; forest land coverage and arable land utilization rate have a strong explanatory power for LECC, with the q–statistic of 0.549 and 0.175, respectively, ranking in the top three among all factors. In terms of social factors, along with the increase in urbanization level and the growth of population, which exerts great pressure on LECC, the population density has a very strong explanatory power on the LECC, with the q–statistic of 0.544. It ranks second among all factors, next to the forest land coverage. In terms of economic factors, the regional gross domestic product and the total retail sales of social consumer goods directly or indirectly affect the production and consumption of biological resource products or energy products, and the production and consumption of products maintain a relative balance generally. Hence, they have certain explanatory power to the LECC, and the explanatory power is comparable, with the q–statistic of 0.128 and 0.152.

The results of interactive detection indicate that the interactive effects of the influencing factors on the LECC show an enhanced relationship, including bifactor enhancement and nonlinear enhancement, without nonlinear weakening, single-factor weakening, or independence (

Table 5. Geodetector interactive detection results.

	UL	PD	RGDP	TOG	IS	II	ALUR	FLC
UL	0.117	EB	EN	EN	EN	EN	EN	EB
PD	0.632	0.544	EB	EB	EN	EN	EN	EB
RGDP	0.250	0.610	0.128	EB	EN	EN	EN	EB
TOG	0.357	0.653	0.194	0.152	EN	EN	EN	EB
IS	0.339	0.763	0.272	0.267	0.055	EN	EN	EN
II	0.284	0.644	0.263	0.287	0.288	0.038	EN	EN
ALUR	0.326	0.729	0.373	0.423	0.585	0.377	0.175	EN
FLC	0.651	0.683	0.653	0.688	0.721	0.638	0.742	0.549

Note: EB is the bifactor enhancement, and EN is the non-linear enhancement.

). The larger value of interaction factors indicates that the interaction of influencing factors is stronger. Compared with the influence of a single factor on LECC, the influence of most bifactor interactions on the LECC is enhanced to a large extent. Among them, the total retail sales of social consumer goods have interactive synergistic effects with regional gross domestic product; forest land coverage and population density have interactive synergistic effects with most influencing factors, indicating that the effects of these factors on LECC are mutually reinforcing. The interactions of other factors show nonlinear enhancement, and the factors with the most evident interaction enhancement were ranked as follows: population density and industrial structure, arable land utilization rate and forest land coverage, arable land utilization rate and population density, forest land coverage and industrial structure. The q-statistic of these factors exceeded 0.700 after interaction, showing strong explanatory power, indicating a close relationship between them.

4. Discussion

4.1. Result Analysis

The study found that the first batch of National Pilot Zones for Ecological Conservation, namely Jiangxi, Fujian and Guizhou, has a relatively low biocapacity per capita and a high ecological footprint per capita, and the ecological footprint per capita showed a significant decline trend around 2015. On the one hand, land area constraints and China’s previous crude economic development model of seeking economic development at the expense of the ecological environment have placed great pressure on the land ecosystem [50,51], resulting in the reduction of productive land quantity and quality degradation [52], with values for biocapacity that have been relatively low for a long time. Along with the accelerated industrialization process, the pollution of productive land by industrial waste has become increasingly serious and greatly increased the land ecological footprint [53], leading to a serious imbalance between land ecological supply and land ecological demand; humans have had to consume the natural capital stock to meet the needs of socioeconomic development. However, since 2015, China has issued Opinions on the Establishment of Unified and Standardized National Pilot Zones for Ecological Conservation, National Forest Farm Reform Scheme and other series of policy documents, which emphasize the green development concept that lucid waters and lush mountains are invaluable assets, point out the direction, and provide the system with safeguards for the land ecological construction. On these bases, land ecological civilization construction and protection have been effectively enhanced, and the ecological footprint has shown more significant decreasing characteristics. However, the problem of the low level of land natural capital flow utilization and the high consumption of land natural capital stock remain and have become a current norm for economic and social development [54]. Thus, these problems need to be further optimized and solved.

4.2. Policy Recommendations

The research results and analytical conclusions indicate that the differences in the land ecological foundation and socioeconomic development level of the first batch of the National Pilot Zones for Ecological Conservation in China are the main reasons for the overall performance of LECC, as Guizhou outperforms Jiangxi and Fujian. Based on these findings, to further make up for the shortcomings and improve the LECC, the following policy recommendations are proposed:

(1)

Strengthening land ecological management and optimizing the practice of land use. Local government departments should focus on strengthening the ecological management of arable land, forest land, and grassland; conduct ecological evaluation and grading; and conduct fine land ecological restoration and protection activities accordingly. Moreover, they should actively adjust the practice of land use, pay more attention to the use of natural capital flows, improve the level of use of natural capital flows, and gradually reduce the consumption of natural capital stock.

(2)

A security warning system should be established to protect the land ecological security. The development intensity should be controlled according to the main function orientation, with a focus on the land ecological protection in very insecure and extremely insecure areas. GIS, RS, GPS, drones, and other modern technical means should be used synthetically to obtain ecological alert data for various types of land. On this basis, real-time monitoring and early warning systems for land ecological security should be established, land ecological security risk information should be reported on a timely basis, and targeted measures should be developed to solve different problems that arise in different parcels of land.

(3)

Capturing the key influencing factor and enhancing positive interactions between them. On the one hand, on the basis of ensuring the quality and quantity of existing forest land, unused land could be converted into forest land through afforestation and other measures to increase forest land coverage. On the other hand, the spatial distribution pattern of population could be optimized by adjusting the industrial layout and improving road traffic to alleviate the excessive regional spatial concentration of population. At the same time, by optimizing the industrial structure and reasonably controlling the arable land utilization rate, the positive interactions of forest land coverage and population density in support of LECC will be enhanced.

(4)

The experience of model construction should be promoted to stimulate other regions’ LECC to reach a high level. The construction of land ecological civilization has been effective for Guizhou. On the one hand, full play should be continuously given to the measures to improve the LECC further. On the other hand, full play should be given to the role of a model to guide other regions to transform and upgrade their industrial structures and promote economic green development, while strengthening key ecosystem service functions, completing the spatial integration and optimization of forest land, grassland, waters, and agricultural production space, and stimulating other regions’ LECC to improve steadily.

4.3. Limitations and Future Directions of This Study

This study measures the LECC of the first batch of the National Pilot Zones for Ecological Conservation in China based on the improved ecological footprint model, and then analyzes the spatial–temporal pattern and evolution trends. The following two points should be further examined:

Fossil fuel land is virtual land without corresponding land carrying capacity and is essentially different from other lands, such as forest land and arable land [32]. This study does not investigate the fossil fuel land together with the real land. In the follow-up research, a thematic study on land ecology aspects of fossil fuel land is necessary.

This study focuses on the spatial–temporal pattern and influencing factors of LECC in the first batch of the National Pilot Zones for Ecological Conservation, without considering the National Pilot Zones for Ecological Conservation that were established subsequently and the Non-national Pilot Zones for Ecological Conservation. The subsequent study could conduct horizontal and vertical comparative research on the LECC of the National Pilot Zones for Ecological Conservation and Non-national Pilot Zones for Ecological Conservation. This condition is conducive to more targeted determination of the shortcomings and future directions of land ecological construction in Non-national Pilot Zones for Ecological Conservation in China in a well-targeted manner to form a common scheme for national land ecological construction in the country.

5. Conclusions

In this study, a total of 26 prefecture-level cities in Jiangxi, Fujian, and Guizhou of the first batch of Ecological National Pilot Zones for Ecological Conservation in China were taken as the study area. The spatial–temporal pattern of LECC and influencing factors were studied using an improved ecological footprint model, Theil–Sen’s slope estimator, Mann–Kendall test, and geodetector. To a certain extent, the study fills the shortage of the existing literature on LECC in the Ecological National Pilot Zones for Ecological Conservation in China and provides a reference for the future construction of China’s land ecological civilization. Based on the study results, the following conclusions are obtained:

(1)

In terms of temporal evolution, the overall land ecological carrying status of each province tends to improve but also shows remarkable interprovincial differences in development trends; the overall performance of Guizhou is better than Jiangxi and Fujian. Focusing on protecting arable land, forest land, and grassland and reducing the consumption of natural capital stock are key to maintaining LECC.

(2)

In terms of spatial evolution, the situation of land ecological security in most prefecture-level cities is more severe, and the spatial differentiation of land ecological security patterns is more evident. However, the land ecological pressure index is decreasing, and ecological security is being enhanced. In the long term, actively working on land ecological construction in very insecure, and extremely insecure areas can help optimize the overall land ecological security patterns in the region.

(3)

In terms of influencing factors, forest land coverage and population density have the strongest explanatory power for LECC. The interactions of population density with industrial structure and arable land utilization and that of forest land coverage with arable land utilization rate and industrial structure have significant enhancing effects on the interpretation of LECC. Paying attention to the main influencing factors and strengthening their positive interactions are key to improving LECC.