Next Article in Journal
Mapping and Quantification of Soil Erosion and Sediment Delivery in Poorly Developed Urban Areas: A Case Study
Previous Article in Journal
Perceived Destination Image Cohesion: A Comparison Study of Attractions on the Grand Canal, China
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Sustainability
Volume 15
Issue 18
10.3390/su151813680
sustainability-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Evaluation of Tourism Ecological Security Based on Driving Force–Pressure–State–Influence–Response Framework and Analysis of Its Dynamic Evolution Characteristics and Driving Factors in Chinese Province Territory

by
Yingchang Li
1,
Zhenzhen Liu
2 and
Gaifang Liu
1,*
1
School of History and Culture, Shanxi University, Taiyuan 030006, China
2
College of Urban and Rural Construction, Shanxi Agricultural University, Jinzhong 030801, China
*
Author to whom correspondence should be addressed.
Sustainability 2023, 15(18), 13680; https://doi.org/10.3390/su151813680
Submission received: 21 August 2023 / Revised: 9 September 2023 / Accepted: 12 September 2023 / Published: 13 September 2023
Browse Figures
Versions Notes

Abstract

:
Tourism ecological security (TES) is an important measure of the sustainable development of the tourism industry. It is also an important indicator for evaluating the balance between economic growth and the environmental load of tourism destinations. Therefore, the scientific measurement and examination of TES have important theoretical and practical value in promoting the coordinated and sustainable development of the regional tourism economy and ecological environment. From the perspective of systems theory, based on the driving force–pressure–state–influence–response model, the theoretical framework and index system of China’s provincial TES were constructed. The Technique for Order Preference by Similarity to an Ideal Solution method, spatial autocorrelation, traditional and spatial Markov chain, ordinary least squares regression, and geographically and temporally weighted regression model were used to analyze the dynamic evolution characteristics and driving factors of TES. The results indicated the following: (1) Regarding time series, the average value of TES was generally relatively stable with small fluctuations, the differences among provinces exhibited a converging trend, and a significant spatial correlation was observed between the TES of provinces. (2) In terms of dynamic evolution, the transfer of TES types exhibited “path dependence” and “self-locking” effects, meaning the probability of transfer to other types was low, and the status and transfer of TES types were closely related to their neighborhood status. (3) Regarding the driving factors, except for the negative inhibitory effect of environmental pollution on TES, all other variables had a positive promoting effect on TES; however, the effect of each variable in different provinces varied significantly. The results and methods used in this study can enrich the research on TES and provide a theoretical basis and decision-making reference for the healthy and sustainable development of the tourism industry in Chinese provinces.

1. Introduction

With the rapid development of China’s tourism industry and the advent of mass tourism, the tourism industry has become a driver of local economic growth and a new growth point for regional development [1,2]. However, the dual industrial attributes of tourism, environmental dependence and resource consumption, create a binary contradictory relationship between tourism and the ecological environment [3,4,5]. Due to problems in tourism development, such as insufficient innovation in tourism development models, imperfect tourism management mechanisms, prominent negative impacts on the tourism environment, and excessive tourism-carrying capacity, the contradiction between tourism activities and the ecological environment is becoming increasingly prominent, threatening the safety of the ecological system of tourism destinations and hindering the sustainable development of the tourism industry [6,7]. Sustainable tourism development is a crucial area of focus in international tourism geography, tourism ecology, and other disciplines and has gradually become an important policy goal of land space planning and tourism ecological security (TES) conservation at the global and regional levels [8,9,10,11]. TES is a deepening and expansion of the research on sustainable tourism development in the new era. It aligns with ecological security research and is the foundation and prerequisite for sustainable development, high-quality transformation, and upgrading of the regional tourism industry [12,13,14,15]. Therefore, TES has important research value.
Since the 1970s, many international conferences and conventions, such as the United Nations Conference on the Human Environment, United Nations Conference on Environment and Development, and United Nations Framework Convention on Climate Change, have played vital roles in promoting global environmental protection and the development of TES research. TES is an important component of ecological security theory; it intersects multiple disciplines [16,17,18,19,20,21]. Owing to differences in the academic backgrounds and research perspectives of researchers, the definitions of TES vary. Therefore, the concept of a TES has not yet been unified. However, summarizing the definitions by relevant researchers, TES has two meanings [14,22]: first, whether the natural resources and ecological environment system that the sustainable development of tourism destinations rely on are safe; second, whether the natural resources and ecological environment system of tourist destinations are safe for human production and life, and whether the services they provide can meet the needs of human survival and development. Therefore, a tourism destination is considered safe when the tourism ecosystem can operate in an orderly, balanced, and coordinated manner without threat and can meet the needs of human survival and sustainable development [7,14,23]. Existing research on TES can be roughly divided into three stages: (1) Embryonic stage (before 2004): In this stage, TES research mainly focused on the ecological imbalance pressure caused by tourism activities on tourist destinations, with less research on ecological security [8]. (2) Exploration stage (2005–2016): In this stage, TES research focused on tourism ecological risks [24], health [25], and security [26], and research was characterized by interdisciplinary integration. (3) Development stage (after 2017): In this stage, TES research is an important topic for sustainable tourism development, focusing on the interaction between the ecosystem and tourism system [27,28], and has gradually developed an internal logic of “evaluation-influencing factors-early warning-regulation” [8,29,30].
TES evaluation is a specific method for comprehensively diagnosing TES by qualitatively and quantitatively evaluating the ecological environment of tourist destinations [20,31]. Following the principles of science, systematicity, hierarchy, dynamism, and quantification, and considering the pressure of human activities on the ecological environment and the changes in ecological status, researchers have constructed many theoretical frameworks of index system applications to different scenarios, such as pressure–state–response (PSR) [32], driving force–pressure–state–influence–response (DPSIR) [7,33], and ecological footprint (EF) [34,35]. Establishing a complete evaluation index system is the foundation of TES research. TES evaluation indices have progressed from single to multiple indices, enabling quantitative research on TES. Quantitative research on TES is crucial for the economic development, rational utilization of resources, and ecological environment protection of tourism destinations; thus, researchers have constructed different models and methods to quantize TES based on theoretical frameworks, such as the composite index method [36] and system dynamics model [19,37]. Exploring the influencing factors of TES is crucial for the sustainable development of tourist destinations. The validation methods for influencing factors of TES involve quantitative methods, such as geographical detector [7,28], panel quantile regression [14,33,38], and the spatial quantitative model [7,37]. Alternatively, previous research results can be referenced to qualitatively summarize the influencing factors of tourism destinations. Analyzing the influencing factors of ecological security forms the research foundation for the early warning and regulation of tourist destinations [8]. However, current research lacks an in-depth exploration of impact mechanisms.
In 2012, China strategically decided to “vigorously promote the construction of an ecological civilization” and integrate it into political, economic, cultural, and social constructions [39]. In 2020, the National Overall Planning for Major Ecosystem Protection and Restoration Projects (2021–2035) proposed focusing engineering construction on developing and optimizing the national ecological security barrier system and promoting ecological security as a major national strategy [40]. TES is an important approach to help promote regional sustainable development and an important goal of ecological civilization construction. With the increasing importance of the ecological environment in developing the national economy in China and worldwide, research on TES has shifted from a single tourism destination to a regional unit, covering multiple levels of tourist attractions [41], cities [19,20], regions [7,42], and countries [5,38].
Previous research results on TES have laid a good foundation for this study and subsequent research; however, the following problems are worth exploring: (1) Existing research results have analyzed the spatial differentiation of TES from a geographical perspective; however, most have focused on the differences in the spatial distribution of TES, and few studies have focused on the spatial correlation characteristics and dynamic transfer laws of TES. For example, Ma et al. built an index system of regional TES evaluation based on the DPSIR model and used spatial autocorrelation to analyze the spatiotemporal patterns from 2009 to 2017 in the Yangtze River Delta, and the results showed that the TES follows an overall rising trend of “W” fluctuation with regional differences [13]; Li et al. applied the PSR model to establish an index system of TES evaluation from 2004 to 2010 for Wuhan, and the results showed that the TES status has increasingly improved from a “sensitive” to “safe” condition [16]. However, the results did not consider the spatiotemporal evolution characteristics of TES. (2) To reveal the factors influencing TES, most previous studies were based on models, such as geographical detectors and panel quantile regression. These methods can explore the key driving factors of TES; however, they ignore the spatiotemporal nonstationarity of explanatory variables and response variables and use an idealized model with mean regression, which leads to deviations in the research conclusions. For example, He et al. measured the driving mechanism of TES using a geographical detector from 2004 to 2019 in the Yellow River Basin of China, and the results showed that the factors related to tourism and the economy are the most important driving forces in the whole basin [33]. Mu et al. measured the driving factors of TES using the quantile regression panel method in the Yellow River Basin, and the results indicated that the factors had different effects in different quantiles [14]; however, the results did not reveal the differences in the effect of the driving factors across different regions.
Therefore, based on the issues raised in the previous paragraph, this study considers the provinces in China as research units to analyze the dynamic changes and influencing factors of TES from 2000 to 2021. First, based on the DPSIR model, a theoretical framework and comprehensive evaluation index system for measuring TES were systematically constructed, and the Technique for Order Preference by Similarity to an Ideal Solution (TOPSIS) method was used to evaluate its comprehensive evaluation index. Second, based on the spatial correlation law of geography, mathematical statistics and spatial analysis tools were used to describe the spatial agglomeration characteristics of the TES, and traditional and spatial Markov chain methods were used to explore its dynamic transfer characteristics. Finally, based on a geographically and temporally weighted regression (GTWR) model, we analyzed the spatial characteristics of the driving factors of TES. We explored the dynamic response of TES to various influencing variables. This study aims to comprehensively explain the dynamic evolution characteristics and driving factors of TES in Chinese province territory and provide scientific references for the coordinated development of the tourism industry and ecological environment at different TES levels and for promoting high-quality tourism development in various provinces.

2. Theoretical Framework

The systems theory perspective provides a framework that allows scholars to draw on theories from different disciplines to analyze the complexity of human interaction within a social environment [43]. The term system emerged in Durkheim’s early study of social systems [44]. However, within social work, systems theory has been heavily influenced by the theoretical biologist Bertalanffy, who viewed the system as a whole, with a basic assumption that “the whole is more than the sum of its parts” and its relationships and interactions with other systems as a mechanism of growth and change [43,45]. Therefore, according to systems theory, integrity, relevance, complexity, self-organization, hierarchical structure, dynamic equilibrium, and timing are common basic characteristics in all systems, which are the fundamental ideas and viewpoints of the system and the fundamental principles of the system methodology [43].
The DPSIR model was proposed by the European Environment Agency in 1993 after a systematic improvement in PSR and driving force–state–respond models to solve environmental and resource management problems [46,47]. The DPSIR model objectively reflects the interaction and impact between human activities and the ecological environment and scientifically explains the autonomous and positive feedback mechanism of human society [14,28]. Therefore, the DPSIR model is widely used in research fields, such as ecological environment management and sustainable development, because of its scientific and rational nature [13,48,49,50].
In this study, the DPSIR model, placed in a systematic framework from the perspective of systems theory, was used to express the information coupling relationship between various factors affecting the entire TES system to analyze the dynamic evolution characteristics of TES and provide a scientific basis for the rational utilization of tourism resources and the development and protection of TES in the region. The DPSIR in the TES system is an orderly, sustainable, and open circular system that emphasizes the symbiotic relationship between tourism and the environment. Under the principle of circular cumulative causality, developing social, economic, and tourism activities as potential variables becomes the driving force affecting TES. Improper activities, such as the extensive growth of the tourism economy and tourism pollution emissions exceeding the standard, will pressure the sustainable development of the tourist destination ecosystem and tourism industry and affect the state of the tourism and ecological binary system. The disturbance of the state and structure of tourism and ecological binary systems will change the level of TES and sustainability, resulting in positive or negative impacts, such as the contribution rate of the tourism economy and tourism consumption vitality. To reduce negative impacts and maintain positive effects to promote and maintain virtuous coordination between the tourism industry and the ecological environment, tourism stakeholders are bound to exert their subjective initiative and take appropriate measures to respond, such as enhancing expenditures on environmental protection and improving the technology of pollution treatment (Figure 1).
The DPSIR theoretical framework of TES is an open system with self-organization, high complexity, multilevel, and uncertainty, and the five subsystems are interdependent and interactive [14,51]. Therefore, the overall TES system will present a safe, stable, and spirally rising state only when all subsystems form a virtuous cycle organism through balanced coordination, orderly operation, and proper coordination. A healthy tourism ecosystem is safe, stable, sustainable, and capable of maintaining its organizational structure and autonomy over time and maintaining the resilience of the ecosystem under external stress. Conversely, an unhealthy tourism ecosystem is a confusing system with incomplete functions or uncoordinated subsystems, and the status of TES is unsafe when the tourism ecosystem is under threat [14,28].
In this study, the TES evaluation system, which includes 5 criteria (driving force, pressure, state, impact, and response), 14 elements, and 33 indices (Table 1), was established based on the DPSIR theoretical framework. These indices, frequently used in TES research, were determined based on the principles of systematicity, representativeness, and data availability by soliciting the opinions of tourism experts.
In addition, the EF approach is an important theoretical framework for establishing the index system of TES evaluation. As an important method for studying ecological security and sustainable development issues, EF analysis has continued to expand its application areas [52,53]. Based on the EF thinking method, Xiao et al. selected seven tourism indicators, including transportation, accommodation, sightseeing, entertainment, shopping, catering, and waste, to construct an island tourism EF model, and they then proposed an assessment framework and criteria for the ecological security and sustainable development of island tourism destinations [54]. Ali et al. selected five indicators, including internal travel and tourism consumption, government individual expenditures, capital investment, international tourism receipts, and international tourism expenditures, to build a new global tourism index to evaluate the impact of tourism, renewable energy, and economic growth on EF and natural resources in 128 countries from 1995 to 2019 [55]. Anser et al. evaluated the impact of four crucial factors on EF, including inbound tourism, population density, trade, and economic growth, in 130 countries for the period of 1995–2018 [56]. Comparing the indicators selected based on EF theory, this study adopts a systems theory perspective and follows the sustainable operation and open-loop development characteristics of “driving force-pressure-state-impact-response”, selecting 33 indices to provide a more comprehensive assessment of the status of TES.

3. Methods

3.1. Evaluation of TES

The TOPSIS method was used to evaluate TES in this study. The TOPSIS method, proposed by Hwang and Yoon in 1981, is a widely used and effective multi-objective system decision-making method, with no strict restrictions on data distribution, sample size, and indicator quantity and has the advantages of authenticity, intuitiveness, and reliability [28,57,58]. Please refer to other studies for the principle and calculation process of TOPSIS as this paper will not elaborate on it.
The TES threshold, which is the evaluation standard, was used to qualitatively evaluate the TES level. The determination of the TES threshold has a certain degree of subjectivity and arbitrariness; thus, no unified threshold or grading standard exists for TES. Combining the actual measurement results of TES in this study and the research results of other scholars, TES was divided into seven levels: deteriorative level (DL), risky level (RL), sensitive level (SL), critical security level (CL), general security, relative security, and extraordinary security levels (Figure 2).

3.2. Dynamic Evolution of TES

In this study, Markov chains and spatial Markov chains were used to analyze the dynamic characteristics of the TES with temporal evolution. Markov chains, proposed by Markov in 1906 as a mathematical model for studying natural processes, are a type of Markov process with discrete time and state [59].
Based on the Markov chain model, the continuous data of the TES are discretized into k types, and the probability distribution and transfer situation of the corresponding types are then calculated. Generally, the probability distribution of TES types in t year is expressed as a 1 × k state probability vector; thus, the transfer between different levels of TES types in different years can be represented by a k × k Markov transition probability matrix. If the level of TES of a province is i in the initial year and remains unchanged in the following year, its TES transfer type is stable; if the level of TES is increased, its transfer type is upward; otherwise, its transfer type is downward.
The likelihood ratio (LLR) test method was adopted to test whether the Markov transition results were statistically significant [60,61]. The statistic LLR is asymptotically distributed as χ2 with k × (k − 1) degrees of freedom. Furthermore, the degrees of freedom must be adjusted by removing the elements with a transfer probability of zero.
Traditional Markov chains assume that each region is independent of the others to study the changing characteristics of a certain region in the time dimension [62]. However, as an objective phenomenon, TES exhibits spatial correlation and dependence in geographical space, closely related to the nearest adjacent unit. Using the spatial weight matrix to express the connection between the regions to introduce the concept of “spatial lag”, a spatial Markov chains model is constructed by combining the traditional Markov chains with spatial statistical methods [63,64]. Spatial Markov chains fully consider the influence of neighboring units on the attribute or state of the target unit and can be used to analyze the transfer and change in the spatial state of geographical phenomena [65,66,67,68].
In contrast to the traditional Markov transition probability matrix of size k × k, the spatial Markov transition probability matrix comprises a k k × k transition matrix. The likelihood ratio test method was also adopted to test whether the spatial Markov transition results were statistically significant. Different from traditional Markov chains, the LLR of spatial Markov transition results is asymptotically distributed as χ2 with k × (k − 1)2 degrees of freedom.

3.3. Driving Factors of TES

In this study, ordinary least squares (OLS) and GTWR were used to analyze spatiotemporal changes in the driving factors of TES. Linear regression includes global and local linear regression models [69]. Thereinto, OLS is the best known global linear regression method. The global linear regression model assumes that the relationship between the variables has spatial homogeneity throughout the study area when analyzing spatial data and that the regression parameters are independent of the geographic location of the samples [70,71]. Therefore, the global linear regression model results are only a certain “mean” within the study area.
The local linear regression model assumes that the relationship between variables may vary depending on the geographical location of the observation points, manifested as spatial nonstationarity, implying that different regression coefficients in different regions change with spatial location [72]. Geographically weighted regression (GWR), a typical local linear regression model, was proposed based on the assumption of spatial nonstationarity. GWR incorporates the spatial coordinate parameters into the linear regression model to detect the nonstationary change in parameters and explain the spatial differences in the influence of variables. The GWR model is a classic model for studying spatial heterogeneity; however, it can only be used to study cross-sectional datasets that have a certain occasionality compared with panel datasets [71,73,74,75]. The GWR model considers the spatial nonstationarity of the data but ignores temporal nonstationarity, resulting in deviations in its results. Huang et al. introduced time-characteristic parameters into the GWR model to construct GTWR [71].
Considering the calculation results of the TES of this study and the recommendations of tourism experts, the representative indices with high weights were selected to construct the index system of the driving factors of TES from the environmental, economic, and social elements, giving them a new variable name (Table 2).

4. Data

For this study, various data were collected from 2000 to 2021. To guarantee the accuracy of the data, all original data were obtained from the China Statistical Yearbook, the Yearbook of China Tourism Statistics, statistical yearbook of each province, and national economic and social development bulletin of each province. However, data were not collected for Taiwan, Hong Kong, or Macao. The missing data were supplemented by the linear interpolation method. The statistical summaries of the TES evaluation variables are in Table S1 in the Supplementary Materials section. Administrative boundary data were downloaded from the website of the National Catalogue Service for Geographic Information of China (http://www.webmap.cn/; accessed on 1 May 2023).

5. Results

5.1. Spatial and Temporal Changes

Before exploring the dynamic evolution characteristics and driving factors of TES, its time-series characteristics must be analyzed. The boxplot illustrates the temporal change in TES from 2000 to 2021 (Figure 3). The average value of TES is relatively stable with little fluctuation, with the average value from 2000 to 2008 being slightly higher than that from 2009 to 2021. The differences in TES among provinces constantly narrowed, revealing a convergence trend. Second, the global Moran’s index was used to characterize the spatial relationship of the TES from 2000 to 2021 (Table 3). The p-values were all less than 0.1, and Moran’s index values were all greater than zero, indicating a significant positive spatial correlation of TES in the Chinese province territory. Third, we spatially visualized the TES data during the study period (Figure 4). The figure illustrates that the TES level in each province was between the deteriorative level ((0, 0.25]) and the critical security level ((0.45, 0.55]). TES was worrisome; however, a trend toward high-level development was observed. The main level of TES was RL ((0.25, 0.35]), accounting for 61.29% of all provinces in 2000. The RL area remains large nationwide; however, its percentage decreased to 38.71% in 2021, equal to the SL ((0.35, 0.45]). The count of DL increased from one in 2000 to three in 2012, 2013, 2014, and 2019, mainly concentrated in Hebei, Henan, Shandong, and Jiangsu, and decreased to 0 in 2021. The SL increased from 7 in 2000 to 12 in 2021, and CL increased from 4 in 2000 to 7 in 2021, indicating that the overall TES level significantly improved compared to the beginning of the research period.

5.2. Dynamic Evolution Characteristics

The spatial autocorrelation and distribution analyses indicate a spatial correlation feature in TES. Thus, whether a dynamic transfer law exists in the continuity of time and spatial processes should be evaluated by constructing a Markov transition matrix to reveal the dynamic transfer characteristics of the TES level.
According to the Markov transition probability matrix, elements on the diagonal represent the probability that the TES type has not changed, whereas elements on the nondiagonal represent the probability of transition between different types of TES. The results of the dynamic evolution characteristics of TES without considering the geographical spatial pattern revealed the following (Table 4): (1) The values on the diagonal were greater than the values on the nondiagonal, indicating that the club convergence of DL, RL, SL, and CL was stable with “path dependence” and “self-locking” effects. If a province belongs to DL, RL, SL, or CL at the initial period, the probability of belonging to this type in subsequent years is at least 55.00%, 91.30%, 79.59%, and 84.29%, respectively. (2) Values close to the diagonal are greater than those far from the diagonal, and values equal to zero are all far from the diagonal, indicating that the transformation of TES usually occurs between adjacent levels, and the possibility of cross-level transfer is relatively small. (3) The probability of transition from DL to RL is 0.4500, and the probabilities of transition from RL to SL and CL are 0.0604 and 0.0072, respectively, indicating that the probability of upward transfer is higher than that of downward transfer; thus, the DL and RL levels should be actively guided to a higher level. Conversely, the probabilities of transition from CL to SL and RL are 0.1429 and 0.0143, respectively, and the probability of transition from SL to RL is 0.1293, indicating that the probability of downward transfer is higher than that of upward transfer. Thus, the transfer of CL and SL levels to lower levels should be monitored to prevent the expansion of the scale of lower levels.
Moreover, the results of the likelihood ratio test for Markov chains indicated that the likelihood ratio LLR = 48.8003 and the degrees of freedom were adjusted from 4 × (4 − 1) to 8 after removing sample points with a transfer probability of 0. Under the significance level with α = 0.005, LLR > χ2 = 21.955, rejecting the null hypothesis that the transfer of TES types remains stable throughout the entire research period. However, the results of the Markov transition were effective.
A significant spatial correlation of TES was observed in the Chinese province territory, which has spatial agglomeration and spatial interaction effects, and the level of TES was affected by changes in the level of its surrounding provinces. Therefore, a spatial Markov transfer probability matrix for TES was constructed to further explore the impact of neighborhood background on the transfer of TES types.
Compared with the traditional Markov transfer matrix, the regional background is important in the dynamic change process of TES. Under the different neighborhood conditions, the transfer probability of TES types changes significantly (Table 5). For example, the probability of transition from DL to RL is 0.4500 without considering the spatial neighborhood, but the probabilities of transition are 0, 0.4118, 1, and 0 when adjacent to DL, RL, SL, and CL, respectively. (1) When the neighborhood type is DL, the probability of maintaining the DL type is 1, and the probability of other transfer types is 0, indicating that the DL type is difficult to transfer to other types when the TES is at DL; however, its number is very small. (2) When the neighborhood type is RL, the probability of maintaining DL, RL, SL, and CL types is 0.5882, 0.9281, 0.7612, and 0.7407, respectively, indicating that the probability of SL and CL types transferring to low-level types is increased compared with when the spatial neighborhood is not considered. (3) When the neighborhood type is SL, the probabilities of maintaining the DL, RL, SL, and CL types are 0, 0.8617, 0.8194, and 0.9070, respectively, indicating that the probability of maintaining a high level of TES increases when adjacent to the SL type. (4) When the neighborhood type is CL, the probability of maintaining the SL type is 0.8750, the probability of transition from SL to CL is 0.1250, and the probability of other transfer types is zero, indicating that transitioning to lower-level types when adjacent to higher-level types is difficult. Based on the spatial Markov transition probability matrix, the club convergence phenomenon of the TES type is more significant owing to the influence of the spatial neighborhood.
Moreover, the results of the likelihood ratio test for spatial Markov chains indicated that the likelihood ratio LLR = 88.8852 and the degrees of freedom were adjusted from 4 × (4 − 1)2 to 31 after removing sample points with a transfer probability of zero. Under the significance level with α = 0.005, LLR > χ2 = 55.003, rejecting the null hypothesis that the transfer of TES types remains stable throughout the entire research period and is unaffected by spatial neighborhood. The results of the spatial Markov transition were effective.
Figure 5 illustrates the transfer types of TES in the Chinese provincial territory. First, without considering the spatial neighborhood (local transfer), the downward transfer from 2000 to 2021 of TES types was distributed in Xinjiang, Inner Mongolia, and Shaanxi, accounting for 9.68% of all provinces (Figure 5a). The type of TES was transferred from SL to RL (Figure 5b). The upward transfer of TES types accounted for 45.16% of all the provinces, and the main transfer type was from RL to SL. Second, under the effect of spatial lag (considering the spatial neighborhood, neighbor transfer), no downward transfer of TES types was observed from 2000 to 2021. The upward transfer of TES types accounted for 41.94% of all provinces, and the main transfer type was from RL to SL (Figure 5c). This result contradicts the transfer type between local transfer and neighbor transfer for Xinjiang and Inner Mongolia because these two provinces have a large area and many neighboring provinces, and their neighboring transfer types are greatly affected by the surrounding provinces. Moreover, compared with the local transfer type of all provinces, the type of neighbor transfer is more concentrated in the spatial distribution.

5.3. Driving Factor Analysis

In the context of transforming new and old growth drivers of China’s economy and green development, the main driving factors of TES from 2000 to 2021 must be explored to promote high-quality and intensive development of the tourism industry and enhance the TES level.
The results of the Pearson correlation coefficients between the explanatory variables and the response variable, TES, indicated that seven variables had a significance level of 0.01 in relation to TES. ENP was significantly negatively correlated with TES, with a value of −0.572. The variable with the highest correlation coefficient was IDL, with a value of 0.645 (Table 6). The variable EL was not significantly related to TES. In addition, a significant correlation was observed between most of the explanatory variables at a significance level of 0.01.
According to R2, 90.19% of the variation in TES could be explained using the OLS model (Table 7). The OLS results reveal that ENP has a significant negative inhibitory effect on TES, while IDL, EL, ECP, TEL, TCV, LI, and GI had a significant promoting effect on TES in the OLS regression. The collinearity statistics of the explanatory variables were less than 5.50, indicating that these variables were effective; however, EL did not significantly interpret TES because of its p-value of 0.1216 for EL. Significant differences were observed in the impact of each variable on TES from the regression coefficients of each variable, while TCV and GI with a higher regression coefficient had a higher interpretation for TES, and TEL and LI had a lower interpretation for TES.
Comparing the diagnostic information of the OLS and GTWR models, the percentage of explanatory variance increased from 90.19% in the global OLS model to 97.80% in the GTWR model. The residual sum of squares and AICc values decreased, indicating that GTWR better fits the data than the global OLS model (Table 8). The GTWR results reveal that, except for ENP, which negatively affects TES, the other seven variables positively affect TES, similar to the OLS results. TCV and GI with higher regression coefficients had higher interpretations for TES, and TEL and ECP had lower interpretations of TES.
Figure 6 illustrates the different impacts of explanatory variables in different provinces: (1) ENP had a higher negative inhibitory effect on TES in the western and eastern coastal provinces, such as Xinjiang, Gansu, Zhejiang, and Fujian, with a lower effect in the northeast and middle provinces, such as Inner Mongolia, Heilongjiang, Shanxi, and Hunan. However, its impact coefficient was relatively low compared with the other variables. (2) ECP had a significant positive effect on TES in most provinces, except for Shandong and Tibet, and had a greater effect in the northwestern and southeastern provinces, but its impact coefficient was also relatively low compared with other variables. (3) IDL also had a significant positive effect on TES in most provinces, except Tibet, and had a greater effect in the northern and southern provinces, such as Inner Mongolia, Hebei, Guangdong, and Guangxi. (4) TEL had the lowest effect on TES, with a higher effect in the southeast and southwest provinces, such as Guangdong, Jiangxi, Qinghai, and Tibet, and a lower effect in the north and northeast provinces, such as Inner Mongolia, Hebei, Beijing, and Shanxi. (5) TCV had a significant effect on TES compared with other variables and had a higher effect in the southwestern and southeastern provinces, such as Yunan, Guizhou, Guangxi, and Fujian, and a lower effect in the western provinces, such as Xinjiang, Qinghai, and Tibet. (6) EL also had a significant positive effect on TES in most provinces, except for Hunan, with a higher effect in the western and southwestern provinces, such as Yunnan, Guangxi, Tibet, and Xinjiang, and a lower effect in the middle, northern, and eastern provinces, such as Hunan, Inner Mongolia, Hebei, and Shandong. (7) LI had a highly positive effect on TES, with a higher effect in the eastern provinces, such as Shandong, Jiangsu, Zhejiang, and Anhui, and a lower effect in the northeast and southwest provinces. (8) GI had the highest effect on TES and had a higher effect in the northeastern and northern provinces, such as Henan, Beijing, Jilin, and Inner Mongolia, and a lower effect in the western and southeastern provinces, such as Jiangsu, Zhejiang, Anhui, Tibet, and Xinjiang.

6. Discussion

In the new era, where ecological civilization construction has become a major national strategy in China, the tourism industry has shifted from scale expansion to quality improvement and the promotion of high-quality development of the tourism industry. We conducted a systematic and in-depth study of the dynamic evolution characteristics and driving factors of TES in China, which is crucial for promoting ecological protection and sustainable tourism development. This study integrates multiple geographical methods to analyze the spatiotemporal characteristics and geographical correlation laws of TES development, identify bottlenecks that constrain regional TES, and provide a scientific basis for formulating corresponding TES strategies.
First, this study combines the traditional and spatial Markov chain methods to study TES, which can effectively measure the process and laws of dynamic transfer of TES, intuitively reveal the heterogeneity characteristics and neighborhood background effects of the dynamic evolution of TES, compensate for the shortcomings of using only the traditional Markov chain method in previous studies, and clarify the future direction for improving TES. Regarding the dynamic evolution characteristics, the results of spatial and traditional Markov chains methods indicate that the transfer of TES types has serious “path dependence” and “self-locking” effects. The probability of maintaining the same type of TES is much greater than the probability of transferring to other types, and the transfer of TES types usually occurs between adjacent levels, with a lower probability of cross-level transfer. However, the situation of TES is not optimistic in the Chinese provincial territory, and a significant downward transfer risk exists at the SL and CL of TES. TES distribution has a significant spatial correlation, and the transfer of TES types is also greatly affected by neighborhood types, indicating that the regional tourism ecosystem is an organic whole. The results of this study reveal the dynamic evolution characteristics of TES in various provinces in China over the past 20 years, which is of great significance for establishing global awareness and strengthening regional cooperation.
Second, this study used the GTWR model to explore the driving factors of TES to obtain a more comprehensive and scientific explanation of the spatiotemporal changes in TES, thus enriching the research methods on the impact mechanism of TES. The results indicate that this method emphasizes the spatiotemporal heterogeneity trend of driving factors for TES and compares it with the idealized model of global mean regression, such as OLS regression, reflecting, in detail, the actual situation of the impact of each explanatory variable on TES in each province, providing a research method for a more comprehensive, systematic, and dynamic exploration of TES driving mechanisms in provinces in the future. The results of the GTWR model indicated that the explanatory variables had different effects in each province. Environmental pollution factors, such as SO2 emissions and sewage discharge, negatively impact regional TES. Environmental pollution is the main obstacle in low-level TES areas; thus, strengthening environmental protection and pollution control in TES research units is important for enhancing the resilience of tourism ecology. Factors, such as ECP, IDL, TEL, TCV, EL, LI, and GI, have a positive impact on regional TES, thus intensifying environmental protection, developing green technologies, optimizing industrial structure, improving the education level and environmental awareness of all people, improving the quality of tourism practitioners, increasing environmental protection investment, strengthening environmental publicity, integrating tourism resources, and promoting sustainable development models, which is crucial to affect regional TES. Compared to the driving factor effect measured using global regression methods, the results of this study using the GTWR model reveal, in more detail, the effects of various driving factors in different provinces, which is important for accurately grasping the most important driving factors in each province and formulating measures accordingly.
In addition, this study used Moran’s Index to analyze the spatial autocorrelation for spatial data. Spatial autocorrelation, an important feature of spatial data, is crucial in this study because it establishes the spatial relationship between the research units (provinces) through the spatial weight matrix, determines the neighborhood status of each research unit, and provides a basis for the quantitative analysis of the regional spatial layout for spatial Markov chains and GTWR. Neglecting spatial factors, such as spatial dependence or heterogeneity, when analyzing the dynamic evolution process and driving factors of regional TES may result in inappropriate model settings, weakened explanatory power of the model, and difficulty in accurately reflecting the dynamic changes and driving mechanisms of regional TES.
Regional TES is an open system, and the driving factors of TES, such as regional economic development, trade exchanges, tourist mobility, knowledge and technology diffusion, and dissemination, make regional TES interact and rely on each other. The TES comprehensively reflects and represents social, economic, and environmental factors in regional tourism development. Research on TES issues involves exploring the complexity, systematicity, sustainability, interaction, and integration between tourism activities and the tourism environment in the regional tourism system. This study adopts a systems theory perspective and selects 33 indices based on the DPSIR theoretical framework to establish a comprehensive evaluation index system and then evaluates the TES situation in the Chinese Province Territory from 2000 to 2021. The research results objectively reflect the evolutionary trend and driving mechanism of TES, providing an important reference for the ecological protection and high-quality development of the tourism industry. The foundation of a TES study is to construct a comprehensive and objective evaluation model and select scientific and reasonable research methods. This study is based on systems theory to construct a multidimensional framework for evaluating TES; however, further in-depth research is needed to reduce the impact of index selection and method application on the explanatory power of evaluation results, owing to the availability of data and the limitations of research methods. Big data should be applied in future research, such as remote sensing image data, land use data, pollution monitoring site data, scenic spot monitoring data, big data on tourism flows and enterprises, and tourist survey data, to build a scientific and comprehensive evaluation index system. In addition, in this study, all TES evaluation indices and data are derived from statistical data. Due to the highly integrated nature of the tourism industry and the narrow scope of tourism statistics, it was difficult to separate statistical data related to environmental pollution, energy consumption, and other aspects directly caused by tourism, which is a key point in improving the evaluation indicators for TES in the future. Considering the available data, this study reveals the spatiotemporal evolution and driving factors of TES in the province territory; thus, future research will target more micro levels, such as cities and counties, which may be more scientific and accurate. In addition, under the context of China’s tourism industry shifting toward high-quality development, ensuring comprehensive and reasonable TES, and achieving sustainable development of the tourism industry, scientifically handling the coordination relationship between tourism development and TES and constructing a regional coordinated mechanism for TES are also key research topics in the future.
As an important part of tourism sustainable development, TES is a hotspot in tourism geography and tourism ecology research. This study is guided by a systems theory combination with the DPSIR model to build a TES index system, exploring the spatiotemporal evolution characteristics and the spatial effects of the driving factors of TES. The results obtained are expected to provide scientific support and theoretical reference for implementing effective tourism ecological planning and sustainable development to a certain extent. In summary, this study provides two main inspirations: (1) Guiding the research of TES with a systems theory approach is of important theoretical significance. The tourism territorial system is a complex system consisting of multiple elements, including nature, resources, economy, society, and culture, and its TES is characterized by a cyclic transformation from disorder to order, formed by the intertwining, interaction, and integration of various element systems. Guided by systems theory, this study establishes an indicator system for studying the TES problem from the five aspects of driving force, pressure, state, influence, and response, thereby prompting the academic community to examine the issue of TES more comprehensively and in depth, which is of great significance for enriching the basic theory of ecotourism and ecological security. (2) The construction of TES evaluation indicators, analysis of spatiotemporal dynamics, and driving factor of TES enrich the research methods and research contents of TES as follows. First, the index system construction has long been a major challenge in comprehensive evaluation research, and the same applies to TES evaluation. In this study, we built a TES evaluation index system through theoretical preparation, goal decomposition, indicator selection, related analysis, and expert evaluation, which is of certain reference value for other researchers. Second, the spatial autocorrelation, spatial Markov chains, GTWR, and other methods employed in this study can be applied to other scales of TES research, providing methodological support for conducting related research. Finally, this study adopts a GTWR to analyze the spatial effects of driving factors for TES, parsing the role of each driving factor in different provinces; hence, the results are more detailed and can, therefore, provide a theoretical basis for governments and managers to formulate corresponding ecological security management policies and measures.

7. Conclusions

Studying the spatiotemporal dynamic evolution law and mechanism of TES is an essential aspect of constructing ecological civilization and realizing a “Beautiful China”. This study establishes an evaluation index system for TES based on the DPSIR model to evaluate the status of TES and then analyzes the dynamic evolution characteristics and driving factors from 2000 to 2021. The results indicate the following: (1) In terms of time series, the average value of TES remains relatively stable, with minor fluctuations. The differences among provinces demonstrate a convergence trend. A significant spatial correlation was observed between the TES of provinces. The TES level was relatively low, demonstrating a gradual upward trend. (2) Regarding dynamic evolution, the transfer of TES types exhibits “path dependence” and “self-locking” effects. The probability of transfer is low; however, the CL and SL of TES demonstrate downward transfer risks. The status and transfer of TES types are closely related to their neighborhood status. (3) Concerning the driving factors, except for the negative inhibitory effect of environmental pollution on TES, all other variables had a positive promoting effect on TES; however, the effect of each variable in different provinces differed significantly.

8. Suggestions

Based on the results and conclusions of the spatiotemporal evolution characteristics and driving factor analysis of TES in Chinese province territory, the following suggestions are proposed:
(1) Break down administrative barriers and strengthen regional cooperation. To comprehensively improve the status of TES, it should transcend administrative boundaries, establish the “integration” awareness and ecological community concept, strengthen the joint prevention and control of cross-regional and cross-sectoral tourism ecological protection, and strengthen the integrality, systematicity, and cooperativity for the regional tourism special planning, ecological environment protection, and land space development planning, achieve effective communication of regional tourism ecological security information to reduce the constraints and impacts of spatial effects on tourism ecological security, and then gradually break the “path dependence” effect and realize the coordinated development of regional TES.
(2) Adopt measures suited to local conditions at and in different TES levels and regions. Due to differences in TES status and resource endowment, while strengthening regional cooperation, the uniqueness of the province should be considered, the main obstacles and risks that affect improvements in TES should be identified, delicacy management should be adopted, and corresponding measures should be formulated based on the resources of the province. The eastern provinces should continue to leverage their regional capital and talent advantages, as well as focus on improving the level of environmental protection technology innovation, effectively support the key work of ecological environment protection and restoration in tourist areas, and fully demonstrate its leading role in environmental restoration; then, the TES level may be gradually improved in the region. The central and western provinces should actively introduce advanced environmental protection ideas and pollution treatment technology from the east and even abroad, as well as strengthening the publicity and education on tourism ecological environment protection, eliminating a batch of industry or tourism enterprises with high environmental risks, and resolutely guarding the bottom line of TES to prevent the downward translation of the TES level.
(3) Make scientific plans and establish a TES early warning system. In addition to launching a national ecological plan, each province should formulate scientific ecological protection plans based on its actual conditions, strengthen the construction of nature reserves, and create a good tourism ecological environment. TES early warning is a prediction and warning for the sustainable development of the tourism ecological environment and an important means for monitoring and regulating the ecological security of tourist destinations. Therefore, in areas with frequent tourism activities and fragile ecological environments, it is urgently needed to establish a TES early warning system to monitor and feedback changes in the ecological environment in real time in a timely, as well as to reduce the damage to the ecosystem during tourism, minimize tourism environmental pollution, and improve tourism ecological efficiency.
(4) Establish and improve the benefit compensation mechanism for tourism ecological construction. Establishing a benefit compensation mechanism for ecological construction plays an important role in the following: coordinating the distribution of ecological and economic benefits among stakeholders in ecological construction, resolving conflicts between ecological protection and economic interests, promoting equality and fairness among regions, urban and rural areas, and groups, strengthening regional cooperation, ultimately achieving harmonious social development among different interest groups and regions, and promoting the sustainable development of the tourism industry. Therefore, the principles of people-oriented ecological protection and harmony between man and nature should be followed, and the construction of an ecological compensation system should be carried out from the perspective of government and the market, so that productive enterprises, tourism service enterprises, and local residents can all participate, thereby gradually establishing a strategic, long-term, and overall benefit compensation mechanism for tourism ecological construction.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/su151813680/s1, Table S1: Statistics summaries of the variables.

Author Contributions

Y.L.: Conceptualization, methodology, software, data curation, formal analysis, validation, visualization, project administration, supervision, writing—original draft, writing—review and editing. Z.L.: methodology, data curation, formal analysis, validation, visualization, writing—original draft, writing—review and editing. G.L.: conceptualization, methodology, project administration, supervision, writing—review and editing. All authors have read and agreed to the published version of the manuscript.

Funding

This study was financially supported by the Doctor Foundation for Working in Shanxi Province (no. 102699901008).

Data Availability Statement

The data presented in this study are available on request from the corresponding author. The data are not publicly available due to the use of this data in other unpublished studies.

Acknowledgments

The authors would like to thank our editors, as well as the anonymous reviewers for their valuable comments.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Shi, F.; Weaver, D.; Zhao, Y.; Huang, M.-F.; Tang, C.; Liu, Y. Toward an ecological civilization: Mass comprehensive ecotourism indications among domestic visitors to a Chinese wetland protected area. Tour. Manag. 2019, 70, 59–68. [Google Scholar] [CrossRef]
  2. Lo, K. Ecological civilization, authoritarian environmentalism, and the eco-politics of extractive governance in China. Extr. Ind. Soc. 2020, 7, 1029–1035. [Google Scholar] [CrossRef]
  3. Weng, G.; Pan, Y.; Li, L. The Eco-security Grading and Spatial-temporal Evolution of Tourism Based on Improved DPSIR-DS Model: A Case Study of the Five Northwestern Provinces along the Silk Road. Tour. Sci. 2018, 32, 17–32. (In Chinese) [Google Scholar]
  4. Williams, P.W.; Ponsford, I.F. Confronting tourism’s environmental paradox: Transitioning for sustainable tourism. Futures 2009, 41, 396–404. [Google Scholar] [CrossRef]
  5. Liu, D.; Yin, Z. Spatial-temporal pattern evolution and mechanism model of tourism ecological security in China. Ecol. Indic. 2022, 139, 108933. [Google Scholar] [CrossRef]
  6. Li, X.; Wu, L.; Wu, Q.; Zhou, Y. Evaluation on Tourism Ecological Security and Diagnosis of Its Obstacle Indicators in China. Ecol. Econ. 2017, 33, 90–95. (In Chinese) [Google Scholar]
  7. Ruan, W.; Li, Y.; Zhang, S.; Liu, C.-H. Evaluation and drive mechanism of tourism ecological security based on the DPSIR-DEA model. Tour. Manag. 2019, 75, 609–625. [Google Scholar] [CrossRef]
  8. Han, Y.; Tang, C.; Zeng, R. Review of Tourism Ecological Security from the Perspective of Ecological Civilization Construction. J. Resour. Ecol. 2022, 13, 734–745. [Google Scholar]
  9. Liu, J.; Zhang, J.; Fu, Z. Tourism eco-efficiency of Chinese coastal cities—Analysis based on the DEA-Tobit model. Ocean Coast. Manag. 2017, 148, 164–170. [Google Scholar] [CrossRef]
  10. Buijtendijk, H.; Blom, J.; Vermeer, J.; van der Duim, R. Eco-innovation for sustainable tourism transitions as a process of collaborative co-production: The case of a carbon management calculator for the Dutch travel industry. J. Sustain. Tour. 2018, 26, 1222–1240. [Google Scholar] [CrossRef]
  11. Sharpley, R. Tourism, sustainable development and the theoretical divide: 20 years on. J. Sustain. Tour. 2020, 28, 1932–1946. [Google Scholar] [CrossRef]
  12. Qian, J.; Shen, H.; Law, R. Research in Sustainable Tourism: A Longitudinal Study of Articles between 2008 and 2017. Sustainability 2018, 10, 590. [Google Scholar] [CrossRef]
  13. Ma, X.; Sun, B.; Hou, G.; Zhong, X.; Li, L. Evaluation and spatial effects of tourism ecological security in the Yangtze River Delta. Ecol. Indic. 2021, 131, 108190. [Google Scholar]
  14. Mu, X.; Guo, X.; Ming, Q.; Hu, C. Dynamic evolution characteristics and driving factors of tourism ecological security in the Yellow River Basin. Acta Geogr. Sin. 2022, 77, 714–735. (In Chinese) [Google Scholar]
  15. Qin, X.; Cheng, Y. Evaluation and Type Classification on Chinese Tourist Urban Ecological Security System: Based on the Network DEA Model with Node Weight. Sci. Geogr. Sin. 2019, 39, 156–163. (In Chinese) [Google Scholar]
  16. Li, Y.; Chen, T.; Hu, J.; Wang, J. Tourism Ecological Security in Wuhan. J. Resour. Ecol. 2013, 4, 149–156. [Google Scholar]
  17. Wolfslehner, B.; Vacik, H. Evaluating sustainable forest management strategies with the Analytic Network Process in a Pressure-State-Response framework. J. Environ. Manage. 2008, 88, 1–10. [Google Scholar] [CrossRef]
  18. Kurniawan, F.; Adrianto, L.; Bengen, D.G.; Prasetyo, L.B. Vulnerability assessment of small islands to tourism: The case of the Marine Tourism Park of the Gili Matra Islands, Indonesia. Glob. Ecol. Conserv. 2016, 6, 308–326. [Google Scholar] [CrossRef]
  19. Qin, X.; Sun, F.; Yuan, W. Research on action mechanism of ecological security system of Chinese tourist cities. China Popul. Resour. Environ. 2019, 29, 31–40. (In Chinese) [Google Scholar]
  20. Tang, C.; Wu, X.; Zheng, Q.; Lyu, N. Ecological security evaluations of the tourism industry in Ecological Conservation Development Areas: A case study of Beijing’s ECDA. J. Clean. Prod. 2018, 197, 999–1010. [Google Scholar] [CrossRef]
  21. Malekmohammadi, B.; Jahanishakib, F. Vulnerability assessment of wetland landscape ecosystem services using driver-pressure-state-impact-response (DPSIR) model. Ecol. Indic. 2017, 82, 293–303. [Google Scholar] [CrossRef]
  22. Yang, X.; Jia, Y.; Liu, L.; Wang, Q.; Hou, Y.; Niu, J. Research trends and prospects of tourism ecological security at home and abroad. Ecol. Sci. 2022, 41, 223–230. (In Chinese) [Google Scholar]
  23. Cao, X.; Chen, T.; Wang, W. Ecological Security Evaluation of Tourism Destination—A Case of Kaifeng City. Res. Soil Water Conserv. 2006, 13, 209–212. (In Chinese) [Google Scholar]
  24. Zhong, L.; Li, P. Ecological risk assessment of tourism development in Awancang Wetland, Gansu Province. Prog. Geogr. 2014, 33, 1444–1451. (In Chinese) [Google Scholar]
  25. Zhou, B.; Zhong, L.; Wang, L.; Zhang, S. Assessment on Tourism Ecological Health of the Nature Reserve—A Case Study of Xinging Hooded Crane Nature Reserve in Heilongjiang Province. For. Resour. Manag. 2015, 5, 145–150. (In Chinese) [Google Scholar]
  26. Liu, X.; Yang, Z.; Di, F.; Chen, X. Evaluation on tourism ecological security in nature heritage sites—Case of Kanas nature reserve of Xinjiang, China. Chinese Geogr. Sci. 2009, 19, 265–273. [Google Scholar] [CrossRef]
  27. Lu, X.; Yao, S.; Fu, G.; Lv, X.; Mao, Y. Dynamic simulation test of a model of ecological system security for a coastal tourist city. J. Destin. Mark. Manag. 2019, 13, 73–82. [Google Scholar] [CrossRef]
  28. Zhao, J.; Guo, H. Spatial and Temporal Evolution of Tourism Ecological Security in the Old Revolutionary Region of the Dabie Mountains from 2001 to 2020. Sustainability 2022, 14, 10762. [Google Scholar] [CrossRef]
  29. Peng, J.; Yang, Y.; Liu, Y.; Hu, Y.; Du, Y.; Meersmans, J.; Qiu, S. Linking ecosystem services and circuit theory to identify ecological security patterns. Sci. Total Environ. 2018, 644, 781–790. [Google Scholar] [CrossRef]
  30. Yang, L.; Cao, K. Tourism Ecological Security Early Warning of Ili River Valley Based on DPSIR Model. Ecol. Econ. 2020, 36, 111–117. (In Chinese) [Google Scholar]
  31. Wu, X.; Liu, S.; Sun, Y.; An, Y.; Dong, S.; Liu, G. Ecological security evaluation based on entropy matter-element model: A case study of Kunming city, southwest China. Ecol. Indic. 2019, 102, 469–478. [Google Scholar] [CrossRef]
  32. He, G.; Yu, B.; Li, S.; Zhu, Y. Comprehensive evaluation of ecological security in mining area based on PSR–ANP–GRAY. Environ. Technol. 2018, 39, 3013–3019. [Google Scholar] [CrossRef] [PubMed]
  33. He, X.; Cai, C.; Shi, J. Evaluation of Tourism Ecological Security and Its Driving Mechanism in the Yellow River Basin, China: Based on Open Systems Theory and DPSIR Model. Systems 2023, 11, 336. [Google Scholar] [CrossRef]
  34. Shi, Y.; Shao, C.; Zhang, Z. Efficiency and Driving Factors of Green Development of Tourist Cities Based on Ecological Footprint. Sustainability 2020, 12, 8589. [Google Scholar] [CrossRef]
  35. Yang, Q.; Liu, G.; Hao, Y.; Coscieme, L.; Zhang, J.; Jiang, N.; Casazza, M.; Giannetti, B.F. Quantitative analysis of the dynamic changes of ecological security in the provinces of China through emergy-ecological footprint hybrid indicators. J. Clean. Prod. 2018, 184, 678–695. [Google Scholar] [CrossRef]
  36. Li, Y.; Wang, K.; Cai, Q.; Wu, Y.; Zhang, D. County Forest Ecological Carrying Capacity Evaluation and Spatial Analysis in Zhejiang Province. Resour. Environ. Yangtze Basin 2019, 28, 554–564. (In Chinese) [Google Scholar]
  37. Yang, X.; Jia, Y.; Wang, Q.; Li, C.; Zhang, S. Space-time evolution of the ecological security of regional urban tourism: The case of Hubei Province, China. Environ. Monit. Assess. 2021, 193, 566. [Google Scholar] [CrossRef]
  38. Lu, F.; Ren, H.; Zhai, X. Dynamic evolution characteristics and driving factors of tourism ecosystem health in China. Front. Public Health 2023, 11, 1127980. [Google Scholar] [CrossRef]
  39. The State Council of the People’s Republic of China. Available online: https://www.gov.cn/ldhd/2012-11/08/content_2260053.htm (accessed on 1 June 2023). (In Chinese)
  40. The State Council of the People’s Republic of China. Available online: https://www.gov.cn/zhengce/2020-06/12/content_5518797.htm (accessed on 1 June 2023). (In Chinese)
  41. Xu, M.; Liu, C. Tourism Ecological Security Evaluation and Obstacle Factors Analysis of Zhangjiajie. Resour. Environ. Yangtze Basin 2018, 27, 605–614. (In Chinese) [Google Scholar]
  42. Wang, Z.; Chen, Q. Spatio-temporal pattern evolution and trend prediction of tourism ecological security in the Yangtze River Economic Belt since 1998. Acta Ecol. Sin. 2021, 41, 320–332. (In Chinese) [Google Scholar]
  43. Friedman, B.D.; Allen, K.N. Systems Theory. In Essentials of Clinical Social Work; Sage: Washington, DC, USA, 2014; pp. 3–20. [Google Scholar]
  44. Robbins, S.; Chatterjee, P.; Canda, E. Contemporary Human Behavior Theory: A Critical Perspective for Social Work Practice, 2nd ed.; Allyn & Bacon: Boston, MA, USA, 2006; ISBN 9780205033126. [Google Scholar]
  45. Bertalanffy, L. Von General System Theory: Foundation, Development, Application; George Braziller: New York, NY, USA, 1968; ISBN 0807604534. [Google Scholar]
  46. Singh, R.K.; Murty, H.R.; Gupta, S.K.; Dikshit, A.K. An overview of sustainability assessment methodologies. Ecol. Indic. 2012, 15, 281–299. [Google Scholar] [CrossRef]
  47. Asmelash, A.G.; Kumar, S. Assessing progress of tourism sustainability: Developing and validating sustainability indicators. Tour. Manag. 2019, 71, 67–83. [Google Scholar] [CrossRef]
  48. Zha, J.; He, L.; Liu, Y.; Shao, Y. Evaluation on development efficiency of low-carbon tourism economy: A case study of Hubei Province, China. Socioecon. Plan. Sci. 2019, 66, 47–57. [Google Scholar] [CrossRef]
  49. Cecchini, L.; Venanzi, S.; Pierri, A.; Chiorri, M. Environmental efficiency analysis and estimation of CO2 abatement costs in dairy cattle farms in Umbria (Italy): A SBM-DEA model with undesirable output. J. Clean. Prod. 2018, 197, 895–907. [Google Scholar] [CrossRef]
  50. Chen, M.; Zheng, L.; Zhang, D.; Li, J. Spatio-Temporal Evolution and Obstacle Factors Analysis of Tourism Ecological Security in Huanggang Dabieshan UNESCO Global Geopark. Int. J. Environ. Res. Public Health 2022, 19, 8670. [Google Scholar] [CrossRef] [PubMed]
  51. Wang, S.; Sun, C.; Li, X.; Zou, W. Sustainable Development in China’s Coastal Area: Based on the Driver-Pressure-State-Welfare-Response Framework and the Data Envelopment Analysis Model. Sustainability 2016, 8, 958. [Google Scholar] [CrossRef]
  52. Rees, W.E. Ecological footprints and appropriated carrying capacity: What urban economics leaves out. Environ. Urban. 1992, 4, 121–130. [Google Scholar] [CrossRef]
  53. Fatemi, M.; Rezaei-Moghaddam, K.; Karami, E.; Hayati, D.; Wackernagel, M. An integrated approach of Ecological Footprint (EF) and Analytical Hierarchy Process (AHP) in human ecology: A base for planning toward sustainability. PLoS ONE 2021, 16, e0250167. [Google Scholar] [CrossRef]
  54. Xiao, J.; Yu, Q.; Liu, K. Evaluation of the Ecological Security of Island Tourist Destination and Island Tourist Sustainable Development: A Case Study of Zhoushan Islands. Acta Geogr. Sin. 2011, 66, 842–852. (In Chinese) [Google Scholar]
  55. Ali, Q.; Yaseen, M.R.; Anwar, S.; Makhdum, M.S.A.; Khan, M.T.I. The impact of tourism, renewable energy, and economic growth on ecological footprint and natural resources: A panel data analysis. Resour. Policy 2021, 74, 102365. [Google Scholar] [CrossRef]
  56. Anser, M.; Yousaf, Z.; Nassani, A.; Abro, M.M.; Zaman, K.; Kabbani, A. Evaluating Ecological Footprints through Inbound Tourism, Population Density, and Global Trade. Polish J. Environ. Stud. 2020, 30, 555–560. [Google Scholar] [CrossRef] [PubMed]
  57. Liu, C.; Zhang, R.; Wang, M.; Xu, J. Measurement and Prediction of Regional Tourism Sustainability: An Analysis of the Yangtze River Economic Zone, China. Sustainability 2018, 10, 1321. [Google Scholar] [CrossRef]
  58. Zhao, J.; Wang, S.; Li, J. Study on the Spatial–Temporal Pattern and Driving Mechanism of Tourism Eco-Security in the Yellow River Basin. Int. J. Environ. Res. Public Health 2023, 20, 3562. [Google Scholar] [CrossRef] [PubMed]
  59. Bremaud, P. Markov Chains: Gibbs Fields, Monte Carlo Simulation, and Queues; Springer Science & Business Media: New York, NY, USA, 2000; ISBN 0387985093. [Google Scholar]
  60. Anderson, T.W.; Goodman, L.A. Statistical Inference about Markov Chains. Ann. Math. Stat. 1957, 28, 89–110. [Google Scholar] [CrossRef]
  61. Kullback, S.; Kupperman, M.; Ku, H.H. Tests for Contingency Tables and Marltov Chains. Technometrics 1962, 4, 573–608. [Google Scholar] [CrossRef]
  62. Pu, Y.; Ma, R.; Ge, Y.; Huang, X. Spatial-Temporal Dynamics of Jiangsu Regional Convergencewith Spatial Markov Chains Approach. Acta Geogr. Sin. 2005, 60, 817–826. (In Chinese) [Google Scholar]
  63. Rey, S.J.; Montouri, B.D. US Regional Income Convergence: A Spatial Econometric Perspective. Reg. Stud. 1999, 33, 143–156. [Google Scholar] [CrossRef]
  64. Rey, S.J. Spatial Empirics for Economic Growth and Convergence. Geogr. Anal. 2001, 33, 195–214. [Google Scholar] [CrossRef]
  65. Torres, E.J.; López, E.J.; Huertas, M.R. Spatial Markov Chains Implemented in GIS. In Proceedings of the 2018 International Conference on Computational Science and Computational Intelligence (CSCI), Las Vegas, NV, USA, 12–14 December 2018; pp. 361–367. [Google Scholar]
  66. Zhou, L.; Che, L.; Zhou, C. Spatio-temporal evolution and influencing factors of urban green development efficiency in China. Acta Geogr. Sin. 2019, 74, 2027–2044. (In Chinese) [Google Scholar] [CrossRef]
  67. Hu, S.; Jiao, S.; Zhang, X. Spatio-temporal evolution and influencing factors of China’s tourism development: Based on the non-static spatial Markov chain model. J. Nat. Resour. 2021, 36, 854–865. (In Chinese) [Google Scholar] [CrossRef]
  68. Zhao, Y.; Xie, D.; Zhang, X.; Ma, S. Integrating Spatial Markov Chains and Geographically Weighted Regression-Based Cellular Automata to Simulate Urban Agglomeration Growth: A Case Study of the Guangdong–Hong Kong–Macao Greater Bay Area. Land 2021, 10, 633. [Google Scholar] [CrossRef]
  69. Song, C.; Kwan, M.-P.; Zhu, J. Modeling Fire Occurrence at the City Scale: A Comparison between Geographically Weighted Regression and Global Linear Regression. Int. J. Environ. Res. Public Health 2017, 14, 396. [Google Scholar] [CrossRef] [PubMed]
  70. Yang, L.; Zhang, X. Study on Measurement and Influencing Factors of Household Carbon Emission in Jiangsu Province: An Empirical Analysis Based on GTWR Model. Ecol. Econ. 2020, 36, 31–38. (In Chinese) [Google Scholar]
  71. Huang, B.; Wu, B.; Barry, M. Geographically and temporally weighted regression for modeling spatio-temporal variation in house prices. Int. J. Geogr. Inf. Sci. 2010, 24, 383–401. [Google Scholar] [CrossRef]
  72. Brunsdon, C.; Fotheringham, A.S.; Charlton, M.E. Geographically Weighted Regression: A Method for Exploring Spatial Nonstationarity. Geogr. Anal. 1996, 28, 281–298. [Google Scholar] [CrossRef]
  73. Xiao, H.; Yi, D. Empirical Study of Carbon Emissions Drivers Based on Geographically Time Weighted Regression Model. Stat. Inf. Forum 2014, 29, 83–89. (In Chinese) [Google Scholar]
  74. Li, E.; Lu, Y.; Chen, Y. Geographic pattern evolution of China’s merchandise export and its influencing factors: Based on the analysis of merchandise export distance. Geogr. Res. 2019, 38, 2624–2638. (In Chinese) [Google Scholar]
  75. Sifriyani, S.; Rasjid, M.; Rosadi, D.; Anwar, S.; Wahyuni, R.D.; Jalaluddin, S. Spatial-Temporal Epidemiology of COVID-19 Using a Geographically and Temporally Weighted Regression Model. Symmetry 2022, 14, 742. [Google Scholar] [CrossRef]
Sustainability 15 13680 g001
Figure 1. DPSIR model for the evaluation of tourism ecological security.
Figure 1. DPSIR model for the evaluation of tourism ecological security.
Sustainability 15 13680 g001
Sustainability 15 13680 g002
Figure 2. Threshold standard of tourism ecological security evaluation.
Figure 2. Threshold standard of tourism ecological security evaluation.
Sustainability 15 13680 g002
Sustainability 15 13680 g003
Figure 3. Boxplot of tourism ecological security.
Figure 3. Boxplot of tourism ecological security.
Sustainability 15 13680 g003
Sustainability 15 13680 g004
Figure 4. Spatial pattern of tourism ecological security from 2000 to 2021.
Figure 4. Spatial pattern of tourism ecological security from 2000 to 2021.
Sustainability 15 13680 g004
Sustainability 15 13680 g005
Figure 5. Transfer types of tourism ecological security (TES), (a) spatial pattern of the local and neighbor transfer types of TES, (b) the local transfer types of TES from 2000 to 2021 for each province, (c) the neighbor transfer types of TES from 2000 to 2021 for each province. DL, deteriorative level; RL, risky level; SL, sensitive level; CL, critical security level.
Figure 5. Transfer types of tourism ecological security (TES), (a) spatial pattern of the local and neighbor transfer types of TES, (b) the local transfer types of TES from 2000 to 2021 for each province, (c) the neighbor transfer types of TES from 2000 to 2021 for each province. DL, deteriorative level; RL, risky level; SL, sensitive level; CL, critical security level.
Sustainability 15 13680 g005
Sustainability 15 13680 g006
Figure 6. Spatial variation in the regression coefficient of the explanatory variables. ENP, environmental pollution; IDL, industrial developmental level; EL, educational level; ECP, ecological protection; TEL, tourism economic level; TCV, tourism consumption vitality; LI, labor input; GI, government interference.
Figure 6. Spatial variation in the regression coefficient of the explanatory variables. ENP, environmental pollution; IDL, industrial developmental level; EL, educational level; ECP, ecological protection; TEL, tourism economic level; TCV, tourism consumption vitality; LI, labor input; GI, government interference.
Sustainability 15 13680 g006
Table 1. Evaluation system of tourism ecological security.
Table 1. Evaluation system of tourism ecological security.
Criteria LayerElement LayerIndex Layer (Unit)AttributeWeight
Driving forceEconomic developmentD1: Per capita GDP (yuan)−0.0140
D2: Growth rate of GDP (%)−0.0270
Social elementsD3: Natural growth rate of population (%)−0.0270
D4: Urbanization rate (%)−0.0192
D5: Per capita disposable income (yuan)−0.0127
Tourism developmentD6: Growth rate of tourism revenue (%)−0.0255
D7: Growth rate of tourists (%)−0.0227
PressureTourism trafficP1: Tourism traffic pressure (10,000 passengers · km)−0.0179
P2: Tourism traffic density (km)−0.0195
Social elementsP3: Tourism spatial density (10,000 persons/km2)−0.0066
P4: Tourist density (−)−0.0168
Environmental pollutionP5: Sewage discharge (10,000 tons)−0.0109
P6: SO2 emission volume (10,000 tons)−0.0208
P7: Industrial solid waste generated (10,000 tons)−0.0141
StateTourism resourceS1: Scenic spots density (−)−0.0091
S2: Utilization intensity of tourism resources (10,000 persons)−0.0141
Tourism reassuranceS3: Number of star-rated hotels (−)−0.0189
S4: Number of travel agencies (−)−0.0180
Ecological environmentS5: Ratio of nature reserves area (%)+0.0549
S6: Ratio of good ambient air quality (%)+0.0227
S7: Forest coverage rate (%)+0.0412
ImpactIndustrial economyI1: Ratio of tertiary industry in GDP (%)+0.0437
I2: Ratio of tourism revenue in GDP (%)+0.0403
I3: Ratio of accommodation and restaurants in GDP (%)+0.0734
Tourism economyI4: Per capita consumption of tourists (yuan)+0.0954
I5: Tourism economic density (yuan/km2)−0.0067
ResponseSocial responseR1: Number of college students per 100,000 persons (−)+0.0433
R2: Ratio of employees in the tertiary industry (%)+0.0428
Economic adjustmentR3: Ratio of environmental protection expenditure in GDP (%)+0.0769
R4: Ratio of finance expenditures in GDP (%)+0.0790
Environmental governanceR5: Ratio of industrial wastes treated and utilized (%)+0.0208
R6: Sewage treatment rate (%)+0.0219
R7: Rate of harmless disposal of domestic waste (%)+0.0220
+, positive index; −, negative index.
Table 2. The variables of the geographically and temporally weighted regression.
Table 2. The variables of the geographically and temporally weighted regression.
Variable Type Variable NameVariable Interpretation
Response variable Tourism ecological security (TES)The value calculated by the TOPSIS method that represents the level of tourism ecological security.
Explanatory variableEnvironmental elementsEnvironmental pollution (ENP)SO2 emission volume
Ecological protection (ECP)Ratio of nature reserves area
Economic elementsIndustrial developmental level (IDL)Ratio of tertiary industry in GDP
Tourism economic level (TEL)Ratio of tourism revenue in GDP
Tourism consumption vitality (TCV)Per capita consumption of tourists
Social elementsEducational level (EL)Number of college students per 100,000 persons
Labor input (LI)Ratio of employees in the tertiary industry
Government interference (GI)Ratio of finance expenditures in GDP
Table 3. Global Moran’s Index of tourism ecological security from 2000 to 2021.
Table 3. Global Moran’s Index of tourism ecological security from 2000 to 2021.
YearMoran’s Indexz-Scorep-ValueYearMoran’s Indexz-Scorep-Value
20000.0416 1.6320 0.0896 20110.0634 2.1217 0.0339
20010.0413 1.6248 0.0909 20120.0666 2.2029 0.0276
20020.0427 1.6564 0.0853 20130.0436 1.6819 0.0926
20030.0413 1.6263 0.0906 20140.0664 2.1789 0.0293
20040.0500 1.8157 0.0621 20150.0550 1.9374 0.0527
20050.0441 1.6867 0.0803 20160.0627 2.1003 0.0357
20060.0415 1.6291 0.0901 20170.0626 2.1059 0.0352
20070.0439 1.6822 0.0810 20180.0665 2.1789 0.0293
20080.0435 1.6736 0.0824 20190.0524 1.8683 0.0559
20090.0541 1.9172 0.0552 20200.0700 2.2604 0.0238
20100.0708 2.2841 0.0224 20210.0415 1.6304 0.0898
Table 4. Markov transition probability matrix of tourism ecological security types from 2000 to 2021.
Table 4. Markov transition probability matrix of tourism ecological security types from 2000 to 2021.
t/t + 1DLRLSLCLn
DL0.5500 0.4500 0020
RL0.0193 0.9130 0.0604 0.0072 414
SL00.1293 0.7959 0.0748 147
CL00.0143 0.1429 0.8429 70
Diagnostic information
LLR48.8003 p-Value0.0000
DL, deteriorative level; RL, risky level; SL, sensitive level; CL, critical security level; LLR, likelihood ratio.
Table 5. Spatial Markov transition probability matrix of tourism ecological security types from 2000 to 2021.
Table 5. Spatial Markov transition probability matrix of tourism ecological security types from 2000 to 2021.
Spatial Lagt/t + 1DLRLSLCLn
DLDL10001
RL00000
SL00000
CL00000
RLDL0.5882 0.4118 0017
RL0.0219 0.9281 0.0438 0.0063 320
SL00.1493 0.7612 0.0896 67
CL00.0370 0.2222 0.7407 27
SLDL01002
RL0.0106 0.8617 0.1170 0.0106 94
SL00.1250 0.8194 0.0556 72
CL000.0930 0.9070 43
CLDL00000
RL00000
SL000.8750 0.1250 8
CL00000
Diagnostic information
LLR88.8852 p-Value0.0000
DL, deteriorative level; RL, risky level; SL, sensitive level; CL, critical security level; LLR, likelihood ratio.
Table 6. The Pearson correlation coefficients between the variables.
Table 6. The Pearson correlation coefficients between the variables.
VariableTESENPECPIDLTELTCVELLI
ENP−0.572 **1
ECP0.546 **−0.327 **1
IDL0.645 **−0.485 **0.247 **1
TEL0.353 **−0.044−0.0130.197 **1
TCV0.610 **−0.0620.0330.286 **0.287 **1
EL0.059−0.125 **−0.173 **0.209 **0.410 **0.0101
LI0.545 **−0.411 **0.240 **0.577 **0.142 **0.288 **0.243 **1
GI0.533 **−0.379 **0.738 **0.191 **0.025−0.011−0.338 **0.050
** indicate a significance level of 0.01.
Table 7. Parameter estimate summaries of the OLS linear regression.
Table 7. Parameter estimate summaries of the OLS linear regression.
VariableRegression Coefficient
(β)		Estimate
(t-Test)		Significance
(p-Value)		Collinearity Statistics
(VIF)
ENP−0.0584 −14.1972 0.0000 1.6388
ECP0.0521 9.1345 0.0000 2.4719
IDL0.0716 12.9245 0.0000 1.7965
TEL0.0415 9.6086 0.0000 1.4444
TCV0.1348 34.0115 0.0000 1.2661
EL0.0076 1.5499 0.1216 1.6890
LI0.0432 7.6281 0.0000 1.7742
GI0.1104 13.7179 0.0000 3.0500
Diagnostic information
R 2 0.9031 R a d j 2 0.9019
AICc−3263.6RSS0.3248
R 2 , determination coefficient; R a d j 2 , adjusted determination coefficient; AICc, corrected Akaike’s information criterion; RSS, residual sum of squares; VIF, variance inflation factor.
Table 8. Parameter estimate summaries of geographically and temporally weighted regression.
Table 8. Parameter estimate summaries of geographically and temporally weighted regression.
VariableMinimumLower QuartileMedianMeanUpper QuartileMaximum
ENP−0.1505 −0.0626 −0.0348 −0.0409 −0.0156 0.0363
ECP−0.0753 0.0015 0.0385 0.0382 0.0705 0.2909
IDL−0.1397 0.0235 0.0429 0.0490 0.0792 0.1815
TEL−0.0640 0.0012 0.0157 0.0215 0.0401 0.1565
TCV−0.5089 0.1291 0.1556 0.1432 0.1712 0.2279
EL−0.1645 0.0070 0.0484 0.0456 0.0814 0.1742
LI−0.0660 0.0416 0.0838 0.0825 0.1230 0.4403
GI−0.7357 0.1177 0.2013 0.1932 0.2973 0.5220
Diagnostic information
R 2 0.9782 R a d j 2 0.9780
AICc−3990.6 RSS0.0731
R 2 , determination coefficient; R a d j 2 , adjusted determination coefficient; AICc, corrected Akaike’s information criterion; RSS, residual sum of squares; VIF, variance inflation factor.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Li, Y.; Liu, Z.; Liu, G. Evaluation of Tourism Ecological Security Based on Driving Force–Pressure–State–Influence–Response Framework and Analysis of Its Dynamic Evolution Characteristics and Driving Factors in Chinese Province Territory. Sustainability 2023, 15, 13680. https://doi.org/10.3390/su151813680

AMA Style

Li Y, Liu Z, Liu G. Evaluation of Tourism Ecological Security Based on Driving Force–Pressure–State–Influence–Response Framework and Analysis of Its Dynamic Evolution Characteristics and Driving Factors in Chinese Province Territory. Sustainability. 2023; 15(18):13680. https://doi.org/10.3390/su151813680

Chicago/Turabian Style

Li, Yingchang, Zhenzhen Liu, and Gaifang Liu. 2023. "Evaluation of Tourism Ecological Security Based on Driving Force–Pressure–State–Influence–Response Framework and Analysis of Its Dynamic Evolution Characteristics and Driving Factors in Chinese Province Territory" Sustainability 15, no. 18: 13680. https://doi.org/10.3390/su151813680

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop