**An Entropy-Perspective Study on the Sustainable Development Potential of Tourism Destination Ecosystem in Dunhuang, China**

#### **Huihui Feng, Xingpeng Chen, Peter Heck and Hong Miao**

**Abstract:** This paper analyzed the characteristic of the tourism destination ecosystem from perspective of entropy in Dunhuang City. Given these circumstances, an evaluation index system that considers the potential of sustainable development was formed based on dissipative structure and entropy change for the tourism destination ecosystem. The sustainable development potential evaluation model for tourism destination ecosystem was built up based on information entropy. Then, we analyzed each indicator impact for the sustainable development potential and proposed some measures for the tourism destination ecosystem. The conclusions include: (a) the requirements of Dunhuang tourism destination ecosystem on the natural ecosystem continuously grew between 2000 and 2012; (b) The sustainable development potential of the Dunhuang tourism destination ecosystem was on an oscillation upward trend during the study period, which is dependent on government attention, and pollution problems were improved.

Reprinted from *Sustainability*. Cite as: Feng, H.; Chen, X.; Heck, P.; Miao, H. An Entropy-Perspective Study on the Sustainable Development Potential of Tourism Destination Ecosystem in Dunhuang, China. *Sustainability* **2014**, *6*, 8980-9006.

#### **1. Introduction**

Sustainable tourism (ST) is an important part of global sustainable development (SD) [1]. The rapid development of tourism has brought great benefits to tourism destinations, while a variety of other problems are emerging, such as, resources and environmental issues and poor management of the tourism industry. Generally, as a reception carrier, the tourism destination of tours concentrates all elements of tourism on an effective framework, which is the most vital part for examining the impact of tourism. Hence, research for sustainable development of tourism destinations could improve the overall efficiency of tourism and optimize the ecological services related to the tourism. Until now, a number of organizations and academics have paid attention to this topic as well as achieved great progress in research methods and practices.

Particularly, their research focus on the following contents [2–12]: (a) The concept of ST [13–22], for example, Hunter suggested that the concept of ST be redefined in terms of an over-arching paradigm which incorporates a range of approaches to the tourism/environment system within destination areas [13]. Swarbrooke [14] and Aall [17] noted that ST is not just about protecting the environment; it is also concerned with economic viability and social justice, and a suitable balance must be established between these three dimensions to guarantee its long-term sustainability. Sharpley point that there has been significant differences between the concepts of ST and SD, the principles and objectives of SD cannot be transposed onto the specific context of tourism [15]. Hardy argued that ST has traditionally given more focus to aspects related to the environment and economic development, which should be more focused on community involvement [4]. Gianna noted the need to very clearly distinguish between the concept of ST and the idea of tourism as one possible tool to support sustainability at multiple levels [16]. Saarinen concluded that perspectives of the resource-based tradition and the community-based tradition have their advantages in different use contexts and they can complement each other, but in respect to the idea of sustainability and the future challenges of humanity, they all share the same major limitation, which is the strong focus on the local scale [18]. Moyle and McLennan noted that the frequency of occurrence of sustainability as a concept has slightly increased in strategies over the past decade. At the same time, there has been a shift in the conceptualisation of sustainability, with thinking evolving from nature-based, social and triple bottom line concepts toward a focus on climate change, responsibility, adaption and transformation [20]; (b) The indicators of ST [23–41], for example, McElroy constructed a "Composite Tourism Penetration Index" from per capita visitor spending, daily visitor densities per 1000 population and hotel rooms per square kilometer. They tested it on 20 small Caribbean islands and yielded three levels of increasing penetration [26]. McCool and Moisey provided a tourism industry perspective on what items could be sustained and what indicators should be used to monitor for sustainability policies [27]. Wang analyzed the principle of indicators of ST, constructed the indicators of ST, the indicator weight and selected the comprehensive evaluation method [40]. Ward and Butler investigated how to monitor sustainable tourism development (STD) in Samoa. It described some of the methodological considerations and processes involved in the development of STD indicators and particularly highlighted the importance of formulating clear objectives before trying to identify indicators, the value of establishing a multi-disciplinary advisory panel, and the necessity of designing an effective and flexible implementation framework for converting indicator results into management action [31]. Ko proposed that the "Barometer of tourism sustainability" (BTS) model represents the comprehensive level of tourism sustainability in a given destination, combining human and natural indicators into an index of sustainable tourism development, without trading one off against the other. The "AMOEBA of tourism sustainability indicators" (ATSI) model was introduced to complement the BTS analysis and to illustrate individual levels of sustainability of tourism indicators [24]. Chris and Sirakaya employed a modified Delphi technique to constructed indicators from political, social, ecological, economic, technological and cultural dimensions for community tourism development (CTD) [25]. Schianetz and Kavanagh proposed the methodological framework for the selection and evaluation of sustainability indicators for tourism destinations, the systemic indicator system (SIS) this framework takes the interrelatedness of sociocultural, economic and environmental issues into account [33]. Reddy engaged in the identification, selection and evaluation of sustainability indicators for rapid assessment of tourism development in Andaman and Nicobar Islands of India (ANI). These indicators are developed and assessed mainly for developed countries and evaluated a feasible bottom-up approach based on local knowledge [38]. Blancas and Gonzalez introduced an indicator system to evaluate sustainability in established coastal tourism destinations, as well as developing a new synthetic indicator to simplify the measurement of sustainability and facilitate the comparative analysis of destination ranking [30]. Buckley suggested that the indicators of ST should include: population, peace, prosperity, pollution, protection [6]. Oyola and Blancas presented an indicator system to evaluate ST at cultural destinations. Also, they suggested a method based on goal programming to construct composite indicators. Then, they proposed three basic practical uses for these indicators: the formulation of general action plans at a regional level, the definition of short-term strategies for destinations and the establishment of destination benchmarking practices [36]. Delgado and Saarinenc examined the significant role of indicators based on literature review in tourism planning and management. The indicator type (set or index) needs to be carefully selected depending on the situation under analysis and the purpose underpinning the study. However, indicator effectiveness to achieve the ideals of sustainable tourism development is affected by the ambiguity in the definition of the concept of ST and problems associated with data availability and baseline knowledge. The main challenge is to overcome strategic guidelines and political and theoretical proposals of indicators and achieve practical applications for the sustainable development of tourism [23]; (c) Ecological security and environment carrying capacity for tourism [42–54], for instance, Ahn used the limits of acceptable change (LAC) framework as a guide to examine and inform the process of ST on a regional scale. Also, he examined resident attitudes toward tourism development in general, toward desirable types of tourism services, toward local conditions and finally, toward perceptions about if and how conditions might change due to tourism [42]. Gössling provided a methodological framework for the calculation of ecological footprints (EF) related to leisure tourism. Based on the example of the Seychelles, it reveals the statistical obstacles that have to be overcome in the calculation process and discusses the strengths and weaknesses of such an approach [44]. Hunter attempted to connect, conceptually, the realms of ST and EF thinking. It is argued that primary research should focus on calculating the TEF associated with individual tourism products, throughout the product's life-cycle. As well as bringing another dimension to our understanding of tourism's actual ecological demand, it is also argued that the concept of the TEF may be used to clarify theoretical aspects of the sustainable tourism debate, helping to rejuvenate this debate in the process [47]. Cui put out the tourism bearing capacity index and its arithmetic model of operation. He defined the tourism environmental bearing capacity as the bearing intensity of tourism destinations during a period which does not do harm to the present and future people in its current state and which can be accepted by the residents. The bearing intensity of tourism destinations mainly includes the tourist density, the tourism land use intensity and the tourism income value [50]. Xiao constructed the models of general ecological security coefficient (GESC) of island tourism destination and special ecological security coefficient (SESC) of island tourism destinations, and then the assessment framework and judgment criterion were proposed on the ecological security of island tourist destination (ESITD) and island tourist sustainable development (ITSD). Furthermore, the models of island tourist ecological footprints were established based on the idea of EF and an empirical analysis of Zhoushan Islands, China was conducted [51]. Salerno and Viviano describes how the concept of Tourism Carrying Capacity (TCC) has shifted from a uni-dimensional approach to incorporating environmental, social and political aspects. Then, an empirical analysis of internationally popular protected area used by trekkers, the Mt. Everest Region, was conducted [52]. Zhong examined the applicability of the model to China's Zhangjiajie National Forest Park. At the same time, both external and internal factors affecting the park's tourism development as well as the environmental, social, and economic changes of the area are also discussed [53]; (d) The development pattern of ST [55–60] was examined by Rodríguez along with an analysis of the life cycle of Tenerife (Canary Islands, Spain), and two types of strategic decisions are considered: the political– legal decisions of the regional government to regulate tourism activity and the decisions to regrade supply, developed by the administrative institutions related to tourism activity in this destination [55]. Keitumetse devised a Community-Based Cultural Heritage Resources Management (COBACHREM) model that merges the technical and academic approaches to illustrate a symbiosis between cultural and natural resources for sustainable resources conservation at community levels [57]. Rizio explored a forest ecosystem and identified its potential flows of utility, addressing those which best satisfy tourism activities and recreational purposes; to identify the most appropriate tools to manage the flows of utility based on sustainable principles which integrate tourism activities [58]; (e) With regards to perception of residents and visitors [61–75], for example, Xuan studied residents' perceptions of the economic, socio-culture, environment impacts of tourism and residents' attitudes to tourism development in Hainan and Sanya, China. It was concluded that the residents are more aware of positive tourism impacts than negative impacts, and they support tourism development with some reasonable attitudes [61]. Choi and Sirakaya too developed and validated a scale assessing residents' attitudes toward sustainable tourism (SUS-TAS). Then, they administered a 51-item scale of resident attitudes toward sustainable community tourism and 800 households in a small tourism community in Texas. Psychometric properties of SUSTAS along with its practical and theoretical implications are discussed within the framework of sustainable tourism development [65]. Nicholas and Thapa examined visitors' perspectives and support for sustainable tourism development in the Pitons Management Area (PMA) in St. Lucia. Specifically, the focus was on visitors' environmental, economic, and social attitudes based on a sustainable tourism development framework and the effect and best predictive validity of attitudes on support for sustainable tourism development were explored [68]. Bimonte and Punzo analysed how distinct groups of residents, characterised by different levels of involvement in tourism-related activities, perceived the tourism phenomenon, and to check whether there exists a latent or potential ground for conflicts between groups of residents [75]. Cottrell and Vaske examined the relative influence of four sustainability dimensions (environmental, economic, socio-cultural, and institutional) in predicting resident satisfaction with sustainable tourism development in Frankenwald Nature Park, Germany. Structural equation modeling supported the hypothesis that all four dimensions were significant predictors of satisfaction. The economic dimension was the strongest predictor, followed by the institutional, social, and environmental dimensions. Findings indicate that all four dimensions should be included for a holistic approach to planning and monitoring sustainable tourism development [69]. Sörensson and Friedrichs used importance-performance analysis (IPA) to examine the performance of one particular tourist destination with regard to social and environmental sustainability, and to establish whether international tourists and national tourists differ in the sustainability factors they consider important [70]. Dorcheh and Mohamed reviewed existing literature on local perceptions of tourism development and its process. It also discusses influential theories and explains the social exchange theory as an effective framework for sustainable cultural tourism [71]. Miller and Merrilees examined tourists' pro-environmental behaviours in four major categories: recycling; green transport use; sustainable energy/material use (lighting/water usage), and green food consumption. They explored five major antecedents to those categories: habitual behaviour, environmental attitudes, facilities available, a need to take a break from environmental duties, and sense of tourist social responsibility. Also, the poorly understood belief that pro-environmental behaviour weakens when residents become tourists was examined [73]; (f) With regards to research for stakeholders [76–84], Hardy and Beeton argued that the nexus involves an understanding of stakeholder perceptions, and applies this to the Daintree region of Far North Queensland, Australia, to determine whether tourism in the region is operating in a sustainable or maintainable manner. In order to do this, an iterative approach was taken and local people, operators, regulators and tourists were interviewed, and content analysis applied to management and strategic documents for the region. The results illustrate the importance of understanding stakeholder perceptions in facilitating sustainable tourism [83]. Timur and Getz examined the concept of sustainable tourism development in urban destinations. Both qualitative and quantitative data were employed, from interviews and questionnaires undertaken in Victoria and Calgary, Canada, and San Francisco, USA. Respondents representing the three clusters of the tourism industry, local government and the host environment were examined on their interpretation of "sustainable tourism", sustainability goals and barriers to achieving sustainable tourism in urban destinations. Results revealed important similarities and differences among key stakeholders, and particularly a lack of appreciation for a triple bottom line approach among the tourism industry respondents [81]. Vellecco and Mancino demonstrate that in lacking shared responsibility, conflicts and tensions inside the local community paralyse innovative environmental behaviors when they ought really to be turned into opportunities for debate so that shared strategies and solutions may be identified in three Italian areas [78]. Holden found that although stakeholders shared positive perceptions of the economic benefits of tourism, its continued use for sustainable development is uncertain. Key challenges include a lack of confidence in the economic certainty of tourism and its use for out-migration, a maturing tourism market, and challenges to the local control of natural resources with external hegemonic forces [80]. Dabphet and Scott explored the diffusion of the sustainable tourism development concept among stakeholders in the tourism destination of Kret Island, Thailand. It is argued that both interpersonal and media communication and the identification of key actors in the community are needed to effectively diffuse sustainable tourism ideas among destination stakeholders. The results validate the use of diffusion theory as a means to understand the transfer of the sustainable tourism development concept among stakeholders, and they also provide information useful for the design of information dissemination programs [82]; (g) In regard to relevant policy for ST [85–94], for instance, Farsari and Butler explored policies for sustainable tourism development and potential interrelationships between policy considerations. Such policies have been characterized as *ad hoc* and incremental, lacking a clear orientation towards sustainable development, and the complex relationships underpinning them have rarely been considered in decision-making for sustainable tourism [87]. Dodds and Butler found that although respondents were aware of sustainable tourism, the individual advantage from exploiting shared pooled or shared resources is often perceived as being greater than the potential long-term shared losses that result from the deterioration of such resources, which means that there is little motivation for individual actors (whether governments, elected officials, or individual operators), to invest or engage in protection or conservation for more sustainable tourism [86]. Yasarataa and Altinay noted key political actors' interests and how to mitigate personal interests to facilitate and maintain sustainable tourism development in small states North Cyprus, Turkish [92]. Whitford and Ruhanen recommends that there cannot be a "one size fits all" framework for indigenous tourism development to suit all circumstances. Policies need to draw upon indigenous diversity and, in a consistent, collaborative, coordinated and integrated manner, provide the mechanisms and capacity-building to facilitate long-term sustainable indigenous tourism [93]. Solstrand suggest that the environmental and socio-cultural sustainability of marine angling tourism (MAT) in Norway and Iceland requires a complex socio-ecological systems perspective, with interactive governance strategies leading management policies. Sustainability requires that a management strategy not only focuses on the economic aspects; priority must also be given to minimizing multi-stakeholder conflicts and providing sufficient resource data to protect vulnerable fish stocks [89]. Xu and Sofield examined the situation that little guidance is provided to promote sustainability principles in tourism development strategies in China. In the future, a pro-active sustainability approach should be integrated with environmental concerns to allow tourism to participate constructively in the national transformation to a sustainable society [90].

Due to tourism research involved in geography, ecology, environmental science, sociology and so on, also combined with different scales covering the micro to the macro [2], the research focus on the multi-index comprehensive evaluation method (MICEM), tourism carrying capacity (TCC), tourism environment impacts assess (TEIA), ecological footprint analysis of tourism (EFT), life cycle assessment (LCA), limits of acceptable change (LAC) and Geographic Information Systems (GIS) [95–97]. The MICEM could quantify the level of tourism sustainable development, which employed AHP and Delphi method by the level of sustainable development and other potential targets. However, selection indicator and its weights usually by personal decision-making [33,39,98–101]. TCC could comprehensively measure the carrying capacity of tourism destinations such as ecology, resources, psychological and space, which employed remote sensing (RS), field measurements, questionnaire and Delphi method, *etc.* However, it has a characteristic of randomness and subjectivity when assessing environmental carrying capacity [102–104]. TEIA is an effectively method to analysis the effect of tourism on ecological environments by mathematical statistical analysis methods form microscopic view, which construct assessment index system and select assessment model based on environmental background to monitoring the feedback mechanism for impact of tourism environmental. However, it's usually ignoring the effectiveness of monitoring and feedback mechanism [105,106]. EFT constructs the tourism ecological footprint according to various data of per capita consumption by bottom-up questionnaire and investigates statistics. The consequent was directly comparable based on productive land area; it is a global standards value [56]. It is suitable to research on a small-scale since the ecological burden is likely to be passed on by interregional trade in tourism destination [43,47,107]. LCA could identify the stage of development of the tourism destination and solve some problems [49]. It is difficult to quantify the environmental problems in sustainable development [108,109]. The theoretical framework for LAC include identifying the issues and concerns, and defining and describing the types of tourism opportunities, *etc.* in the planning area. It could solve the contradiction between development of tourism and

conservation of resources, which is mostly influenced by the decisions of programmer makers and managers [42,110]. GIS are now recognized widely as a valuable tool for which applications for regional tourism planning have not mushroomed as in other fields. This is also reflected in the field of sustainable tourism. Nonetheless, sustainable tourism decision-making and carrying capacity estimation has a lot to benefit from using such technologies. GIS can be used for managing the various information needs, estimating indicators, and generally assisting decision making in the planning phase, as well as, in the monitoring and evaluation phases [111]. Therefore, researchers have used diverse methodologies with more quantitative analysis, as for each method there are certain advantages and disadvantages [112]. Current studies are using more comprehensive approaches. Previous literatures show that a tourist destination is a relatively complete artificial ecosystem; the ecological is a basis for sustainable development [113]. However, relevant research has failed to exhaustively analyze the structures, considering the functions and evolutionary mechanisms of the compound tourism destination ecosystem. This is indeed a shortcoming of those research studies. Entropy, as a measure of system dissipation or disorganization, has been used to analyze social systems in various contexts [114–118]. In relation to tourism [119–123], for example, Bailey noted that Social Entropy Theory (SET) was a very general macro sociological systems theory [114]. Kenneth and Bailey point out that the most recent applications of entropy are in social entropy theory and macro accounting theory [115]. Stepanic jr and Stefancic hold that the established level of analogy between certain characteristics of social systems and part of thermodynamic formalism in the simplified model encourage one to assume that even deeper analogies might be drawn to construct more complete and detailed models of social systems [116]. Wilson describes entropy in urban and regional modelling introducing a new framework for constructing spatial interaction and associated location models [117]. Cabral and Augusto summarized entropy multifaceted character with regard to its implications for urban sprawl, and propose a framework to apply the concept of entropy to urban sprawl for monitoring and management. Hao point out that the phenomenon of the increase of entropy also exists in the tourism destination's ecological system [121]. Zhao proposed the conception and mainly indicates that research can broaden insights on tourism systems' carrying capacities through entropy change analysis from the view of the tourism system's entropy principle under the tourism dissipative structure mechanism [122]. Qian noted that the tourism environment system was an open system. It is unceasingly exchanging material and energy with the outside. It is impossible to achieve the absolute balance through introduction of the negative entropy flow [123]. Relevant research has indicated that tourism destination ecosystems are a typical dissipative structure. Therefore, there is certain feasibility in analyzing its evolution and sustainable development potential from the perspective of entropy. Given the analysis above and based on the relevant former research, this study based on the structure, function and characteristic of tourism destination ecosystems, applied information entropy theory for Dunhuang city which combines analysis entropy with information entropy to establish a tourist destination quantitative ecosystem model for evaluating the potential of sustainable development of Dunhuang tourism. It could offset the disadvantage of indistinct strategies and lack of specificity in some degrees in the sustainable development of tourism destinations.

#### **2. Study Area**

Dunhuang City is located in the border area of Gansu, Qinghai and Xinjiang Provinces, as the western end of the Hexi Corridor in Gansu Province, China. It belongs to a typical arid oasis region with unique geographical, historical and cultural status. In history, Dunhuang was an important hub on the ancient Silk Road and the point of integration of Eastern with Western civilization. To some extent, it is known as "human Dunhuang" owing to the intersection and coordination of the world's four major cultural systems [124]. It is rich in tourism resources, with Mogao Grottoes called the "Pearl of Oriental Art", Mingsha Mountain known as "desert spectacle", Crescent Lake and other tourism resources. All these places promote the development of tourism resources. In 2012, the number of visitor arrivals was 0.312 million people-times and the total income was 2.687 billion Yuan [125]. However, this explosive growth has brought a series of severe problems for cultural heritage, such as, tourists' periodic overload. Because of those intensive human activities, the weak regional environment, and the global climate, there has been a shortage of water resources, as the core of the regional ecological problems, which continue to worsen [126]. Therefore, it is very necessary to study the potential of sustainable development of the Dunhuang tourism destination ecosystem.

### **3. Methodology**

Entropy, firstly proposed by German physicist R Clausius [127,128], is the unique macroscopic quantity in thermodynamics and statistical physics. In 1948, Shannon introduced this concept into information theory and named it "information entropy" [129]. Generally speaking, information entropy theory is based on probability and statistics to reflect the degree of disorder and quantify the evolution direction of the system. When analyzing the complexity and uncertainty of problems, it can be used as a multi-dimensional method to quantify and determine the comprehensive benefits [130,131]. According to theory of dissipative structures by I. Prigogine [132–134], an open system, which is far from equilibrium during the process of exchanging matter and energy with outside environments, has the tendency of entropy growth; hence, this system, only by constantly introducing negative entropy from the outside flow in order to offset internal positive entropy, will finally have a new and ordered direction for further development. That means that the large entropy of the system corresponds to the low degree of order and *vice versa*.

#### *3.1. Entropy Change and Dissipative Structure of the Tourism Destination Ecosystem*

The tourism destination ecosystem is a special ecosystem of areas with rich tourism resources and occurrences, that is established based on the original nature or artificial ecosystem during tourism development [121]. As the spatial carrier between tourist activities and the ecological environment, tourism definition of ecosystems involves the continuous exchange of materials, energy and information with external environments. It also makes irreversible the non-equilibrium processes inside the system, which is always producing positive entropy, inflowing negative entropy with the characteristics of openness, which is far from equilibrium, nonlinearity and ordered fluctuation [122,123]. Tourism destination ecosystem is a typical dissipative self-organizing system that possesses dissipative characteristics and analyzes the evolutionary process of entropy changes.

The tourism destination ecosystem development and evolution is led mainly by the evolution of its socioeconomic ecosystem under normal conditions. Given this progression, this study analyzes the interactions between the tourism destination socioeconomic ecosystem and its natural ecosystem and other regions by analyzing the entropy change process of the tourism destination socioeconomic ecosystem. This involves analyzing the evolutionary process and developmental trend of the tourism destination ecosystem, as well as evaluating the sustainable development potential of the tourism destination ecosystem.

According to the dissipative structure theory [135], there are two parts of entropy changes in tourism destination ecosystems. The first one is the entropy flow caused by the tourism destination socioeconomic ecosystem's exchange of materials, energy and information with external environments *etc.*; it reflects the carrying capacity of the tourism destination nature ecosystem for its socioeconomic ecosystem. Another part is the entropy production caused by irreversible non-equilibrium processes inside the system, which reflects its regeneration potential and could indicate the vitality of the tourism destination ecosystem. The total entropy change of the system is the summation of entropy production and entropy flow, reflecting the overall level of development in the tourism destination ecosystem [136]. The environment is affected by the development of tourism and associated activities as part of the evolution in becoming a tourism destination ecosystem. This is a result of the increase in disorder of entropy production, which is caused by de-vegetation, water pollution, soil fertility, air quality degradation, biodiversity decline and the assimilation features of traditional culture, *etc.* inside of the tourism destination ecosystem. The total entropy changes of the system and disordered parameters will increase if the tourism destination ecosystem does not exchange moderate amounts of material, energy and information with external environments, so the entropy flow does not offset entropy production. This will result in some negative effects for the tourism destination ecosystem, such as an increase in disorder within the system, lack of power, regulatory failure and weaker functioning.

The analysis of entropy changes in tourism destination ecosystems describes the state of tourism system and changes during the exchange of recourses with an external system. The amount or size of entropy not only expresses the level of internal system resources' effective utilization, but also reflects the elastic changes of system affordability. The change in size of the system entropy refers to higher or lower effective utilization rates in the evolutionary process of the tourism destination ecosystem's exchange of materials and energy with external environments [122].

#### *3.2. Indicator System Establishment*

The indicator system is an effective tool for measuring and evaluating the tourism sustainable development level. There are lots of indicators that play a more important role for tourism sustainable development. Depending on the principles of scientific city, comprehensiveness, dynamics, hierarchy, maneuverability and perceptiveness [137,138], using the references from Indicators of Sustainable Development for Tourism Destinations: A Guidebook (WTO, 2004) [139], European Tourism Indicator System For Sustainable Destinations (EU, 2013) [140], ecological civilization city construction indicator system of Dunhuang city and related research results [23–41,141], the article establishes tourism destination ecosystem sustainable development analysis and indicators

system evaluation according to three aspects. They are the structure, function and characteristic of tourism destination ecosystems; the entropy production and entropy flow in the process of system operation; and the ecological environment pollution and destruction during the tourism industry development within the system. The article selected two parts of the entropy production and entropy flow; four aspects that are the supportive entropy input index, the stressful entropy output index, the consumption metabolism index of entropy and the regenerate metabolism index of entropy; 29 representative index (Table 1).


**Table 1.** Index system hierarchy of sustainable development potential evaluation.

Supportive entropy input index: Mainly embodies attractiveness and bearing capacity of the tourism destination ecosystem. The tourism destination satisfied the tourists' demands by supporting resources, infrastructures and services, support foundation of tourism development, and promotes communication with the outside and for internal operations. Therefore, here we select the related index that can reflect tourism destination development status quo and potential of its development (basic infrastructure, available water resources, *etc.*).

Stressful entropy output index: Expresses tourism activities putting pressure on the tourism destination ecosystem. During the evolution process of tourism destination ecosystem and tourist industry development, tourists and their consumption (transport, hotels, sightseeing, catering) and tourist industry energy consumption have direct or indirect influence on the tourism local ecological environment, and puts some pressure on system development, as well as slowing down the positive evolution speed of the system.

Consumption metabolism index of entropy: During the evolution process of the tourism destination ecosystem, the discharge of wastes, pollutants produced and a series of ecological problems from tourism activities to some extent weaken the sustainable development potential of the system.

Regenerate metabolism index of entropy: Mainly shows human being's governance capacity of the tourism destination ecosystem. The waste discharge by tourists is above the system bearing capacity, so that some pollution cannot be purified by the system itself; therefore, the system must rely on artificial management policies and scientific technologies. That is why human beings invest into environmental pollution management as a recovery function of tourism destination ecosystem's sustainable development.

According to the established evolution indexes of the tourism destination ecosystem, there is the calculation formula for entropy flow, entropy production and total entropy changes (Table 2).


**Table 2.** Symbols and formulae of entropy flow, entropy production and total entropy change.

#### *3.3. Depending on Entropy Information Evaluation Model Establishment*

Based on information entropy's benefits, information entropy's evaluation model is widely used in many scientific fields [136,142]. The tourism research focuses on the following: measuring the weight of an indicator based on the entropy method [143]; analyzing the characteristics and development measures based on information entropy, theory of entropy and dissipative structure [121–123,126,144]. More qualitative analysis was be used and a few calculation methods for entropy of tourism systems. According to the information entropy of Shannon, if we used random variables *X* represents uncertainty in the system, the discrete random variable could be supposed as *x* and its value is *X* = {*x*1, *x*2, …, *xn*} (*n* 2), the probability for each value is *P* = {*p*1, *p*2, …, *pn*} (0 *pi* 1, *i* = 1, 2, …, *n*). . The information entropy can be described as <sup>1</sup> ¦*Pi*

follows [131,145,146]:

$$S = -\sum\_{i=1}^{n} P\_i \ln \left(\mathbf{P}\_i\right) \tag{1}$$

where *S* the is information entropy of an uncertain system, *Pi* is the probability of the random state variable *X* in the uncertain system.

#### 3.3.1. Measurement for Entropy Flow and Entropy Production of Tourism Destination Ecosystem

According to the measurement models for information entropy theory, we compute a formula for the entropy flow and entropy production based on information entropy theory and models for each year, then analyzed the complexity, coordination order and health for tourism destination ecosystems. Measurement of indictors in *m* years, ¨*S* represents the four types of entropy based on information entropy [130,131,145], *i.e.*, the input supportive type of entropy (¨*eS*1), the output pressure type of entropy (¨*eS*2), the consumption metabolic type of entropy (¨*iS*2) and the regeneration metabolic type of entropy (¨*iS*1). *n*

$$
\Delta S = -\frac{1}{\ln m} \sum\_{i=1}^{n} \frac{q\_{ij}}{q\_j} \ln \frac{q\_{ij}}{q\_j} \tag{2}
$$

where ¨*S* represents the four types of entropy, *qij* is the standardized value of calculated from the raw data, *qj* is sum for standardized value of index in *j* year, *m* is sum for the number of appraisal events and *n* is the number of indicators, *i* is each index, . 1 1,2,..., ; 1,2,..., *n i ij i q q i nj m* ¦ ˄ ˅

If the number of index is *n*, and the number of appraisal events is *m*, then *Ei* denotes the indicator-based information of indicator and can be derived thus: *i*

$$E\_i = -\frac{1}{\ln m} \sum\_{j=1}^{m} \frac{q\_{ij}}{q\_i} \ln \frac{q\_{ij}}{q\_i} \tag{3}$$

where *Ei* is the information entropy of indicator, *qij* is the standardized value calculated from the raw data and *qi* is sum for standardized value all appraisal events in *i* index, . 1 1,2,..., ; 1,2,..., *m i ij j q q i nj m* ¦ ˄ ˅

According to the entropy weighting method [142], the entropy weight of *i* indicator is defined as:

$$\mathbf{Q} = (1 - E\_i) \left\langle \left( n - \sum\_{i=1}^{n} E\_i \right) \right\rangle \tag{4}$$

where Q*i* is the entropy weight of *i*, *Ei* is the indicator-based information entropy of indicator *i*, is the number of indicator and , 0 *Qi* 1, *n* 2. *n* 1 1 1 *n i Q* ¦

The entropy weight of an indicator is not the most important coefficient of the indicator in regard to decision-making issues. It is instead the relative degree of competition with other indicators when a set of evaluation objects is given and the evaluation indicators are determined, the entropy weighting value is closely related to the evaluation objects. From the information perspective, the entropy weight of an evaluation indicator represents how much useful information an indicator can provide [131,145]. When the entropy weighting of an indicator is larger than other indicators in the evaluation index system for the sustainable development of tourism destination ecosystems, the useful information provided by the indicator could have a greater impact on the system than the other indicators [131].

#### 3.3.2. Sustainable Development Evaluation Model of the Tourism Destination Ecosystem

The index weight was calculated by information entropy, and then integrated to the value of normalization calculated a comprehensive score of values [130,131,146]:

$$G = \sum Q\_i X\_i \tag{5}$$

where: *G* is an appraisal score, *Qi* is the weighting factor derived from information entropy (described below), *Xi* is the standardized value between 0 and 1 generated from raw data for each indicator. A larger value of *G* indicates a better state of the tourism destination ecosystem and a better potential of the tourism destination ecosystem for sustainable development.

#### **4. Data Sources and Processing Method**

#### *4.1. Data Sources*

Related data applied in this study were extracted from the 10th Five-Year Statistical Yearbook of Dunhuang City [147], 11th Five-Year Statistical Yearbook of Dunhuang City [148], Statistical Yearbook of Dunhuang City between 2011 and 2013 [149], Environmental Quality Bulletin of Dunhuang City between 2006 and 2012 [150]. Some data are obtained from the interview and questionnaire.

#### *4.2. Data Processing Methods*

This study adopted the standardize deviation to processing data and the score between [0, 1] when analyzing the evolution and development of the tourism destination socioeconomic ecosystem. The following aspects should be noted in processing data [131,146]: (a) As the entropy change model has used the four types of entropy for vector quantization, there is no need to distinguish between positive and negative indicators to standardize the data processing; (b) The assessment model for the sustainable development potential of the tourism destination ecosystem, which is based on information entropy, has not used vector quantization on different types of indicators, the data processing must distinguish between positive and negative indicators.

For the four indicators, the input supportive type of entropy (¨*eS*1) *and* regeneration metabolic type of entropy (¨*iS*1) are positive indicators, the bigger value means more coordination of the system. The output pressure type of entropy (¨*eS*2) and the consumption metabolic type of entropy (¨*iS*2) are negative indicators, the bigger value means less coordination of the system. The normalization methods for the positive and negative indicators are listed below:

Normalization method for positive indicators:

$$X\_y^\cdot = X\_y \Big/\text{Max}(X\_\iota) \tag{6}$$

Normalization method for negative indicators:

$$X\_y^\prime = \operatorname{Min}(X\_\iota) / X\_y \tag{7}$$

where is the normalized value of *Xij*, *Xij* is the raw data for indicator *i* in *j* year, *Xi* represents all of the original data for indicator *i*, and *Max*(*Xi*) obtains the maximum of indicator by function during the study period, and *Min*(*Xi*) obtains the minimum of indicator *i* by function during the study period. *Xij i*

#### **5. Results and Analysis**

'

#### *5.1. Entropy Change Analysis*

The supportive entropy input showed a trend to remain stable during study period. The stressful entropy output was fluctuated within a slow upward trend (Table 3, Figure 1). For the stressful entropy output with the turning point in 2003 and 2008, the smallest value was in 2003, because the tourism industry was in a state of depression influenced by SARS. In addition, the global financial crisis and snow disaster of South China led to a smaller value in 2008. The burden on tourism destination ecosystems was decreased during those two years. However, the pressure of tourist destinations was increased with the recovery of the tourism industry. The burden of the Dunhuang tourism destination natural ecosystem was increased under the rapid development of tourism and increasing utilization of tourism resources, while the supportive entropy input experienced relatively slow growth. This shows that the pressure was increasing on the tourism destination socioeconomic ecosystem in some degree from 2008.

The consumption metabolism of entropy showed a slowed trend down between 2000 and 2012, the regenerate metabolism of entropy fluctuated with a sharply upward trend. The turning point of regenerate metabolism of entropy was in 2007, 2008 and 2011, which first increased and then decreased with the turning point in 2007 and slowed down sharply. The minimum value was in 2008 and then showed a slowed upward trend. The turning point was a sharp upward trend in 2011 (Table 3, Figure 2). This indicates that the ecological security was improved, the potential of metabolism was better and the vitality was improved gradually in Dunhuang tourism destination from 2008.


**Table 3.** Entropy production and total entropy change of the tourism destination socio-economic ecosystem in Dunhuang city on information entropy.

**Figure 1.** Score trends of the two entropy exchange types of the tourism destination.

The entropy flow showed fluctuation within a slow upward trend between 2000 and 2012, while the entropy production and total entropy change both sharply fluctuated. The turning points of entropy flow, entropy production and total entropy change were during the period of 2003 and 2008 because the tourism industry was mostly effected by the external environment, with the influences of SARS in 2003 and the global financial crisis and snow disaster of South China in 2008. The entropy flow fluctuated with a slow upward trend, which first decreased and then increased with the turning points in 2003 and 2008. The entropy production and total entropy change sharply fluctuated, first increasing and then decreasing with the turning point in 2003. Also, it first increased and then decreased with the turning point in 2008 (Table 3). This indicates that the Dunhuang tourism destination ecosystem was orderly and healthy during the study period.

**Figure 2.** Score trends of the two entropy production types of tourism destination.

#### *5.2. Analysis of Tourism Destination Ecosystem Sustainable Development Potential in Dunhuang*

The values of supportive entropy input remained stable between 2000 and 2012. These values indicate the carrying capacity was relatively stable as a socio-economic ecosystem in Dunhuang. The values of stressful entropy output were decreased in 2003 and 2008, which showed that the pressure was increased on the nature ecosystem with the development of tourism (Figure 1). The values of consumption metabolism of entropy fluctuated with a slow downwards trend, and the values of regenerate metabolism of entropy fluctuated with a sharp upward trend. That indicated the metabolism of function was strengthened, which indicates some success in protecting the ecological environment and also its quality was improved (Figure 2). The values of sustainable development potential fluctuated with an upward trend (Figure 3). The lowest value was in 2008 and the highest was in 2012. This phenomenon may be related to external features of the tourism industry, which was in a status of trough influenced by the global financial crisis and snow disaster of Southern China in 2008. The stressful entropy output fluctuated with a downward trend; also the regenerate metabolism of entropy had the lowest value in 2008. The highest value is attributed to government attention and an improvement in ecological security, an increase in investment in anti-pollution projects as percentage of GDP and improved proportion of industrial solid waste treated and utilized.

**Figure 3.** Score trends of tourism destination ecosystem sustainable development potential.
