The Sustainable Island Tourism Evaluation Model Using the FDM-DEMATEL-ANP Method

Abstract

The purpose of this study is first to propose a comprehensive evaluation model for sustainable island tourism, and then to provide guidelines and suggestions for the development thereof. Based on the advantages of using fuzzy set theory, this study’s method included the fuzzy Delphi method (FDM), the decision-making trial and evaluation laboratory (DEMATEL), the analytic network process (ANP), and FDM- DEMATEL-ANP (FDANP). From the literature review results and experts’ surveys, the dimensions of the evaluation criteria for sustainable island tourism are governance, economy and finance, socio-culture, and the environment. Compared with other studies, its major contributions and differences are the governance and finance dimensions, and the evaluation criteria for the marine industry, marine cultures, and marine environments. The findings show that the relative importance of the dimensions from high to low are economy and finance, governance, the environment, and socio-culture. The top five key criteria begin with having an official administration organization, having a tourism industry, and revenue uncertainties based on public health events. These, along with policies and regulations, and local food and drink, are thought to provide the necessary conditions for sustainable island tourism. The implications for theory and practice and future research directions are discussed.

1. Introduction

Before the COVID-19 era, island tourism was the most popular type of tourism, and it accounted for the largest part of most islands’ gross domestic product. However, COVID-19 has had significant negative impacts on the tourism industry, especially for island tourism. COVID-19 has interrupted the flow of island tourists due to restrictions on travel, but because island tourism is crucial for economic growth, more governments and scholars are trying to determine successful policies and strategies to support sustainable island tourism [1,2,3,4,5]. Chand et al. [6] observed that for some islands, such as Singapore, Barbados, and the Seychelles, tourist inflow often accounted for more people than the local population before COVID-19, but it is almost zero now (2021). The Committee for the Coordination of Statistical Activities [7] found that 2020 was the worst year on the tourism record based on experiencing the highest number of losses of international tourist arrivals and international tourism expenditures. In 2020, international tourist arrivals accounted for USD 380 million (1.5 billion), and the estimated losses in international tourism expenditure in 2020 were USD 1.3 trillion.

Lukoseviciute and Pereira [8] found that the rapid increase in coastal tourism negatively affected vital coastal and beach environments, as well as social and cultural sustainability. The equipment, services, and access needed to measure tourists’ perceptions of beach quality are wildlife, vegetation, biodiversity, sand dunes, trails and boardwalks, seawater, garbage facilities, public toilets, showers, cafés/bars, and renewable energy. Ponte et al. [9] believed that the multiple dimensions of sustainable tourism activity include the quality of visitor and local population experiences, the cultural identity of the host community, and the negotiations and cooperation among visitors, the local community, and the destination.

Although the impact of COVID-19 continued in 2021, many countries still relied on island tourism and sustainable tourism as development policies. COVID-19 may have allowed some islands’ tourism to slow down, allowing more free time and resources to prepare for sustainable island tourism [10]. The purpose of this paper is to design and present a comprehensive evaluation model and procedure for achieving sustainable island tourism. By surveying tourists and residents of Taiwan, a case study of island tourism was conducted in order to contribute to the development of guidelines for sustainable island tourism in the tourist and marine industries. Agrusa et al. [10] thought that COVID-19 might allow Hawaii’s tourism to reset or reposition itself within sustainable island tourism, in order to attract the tourists that would benefit from the sustainability of the natural environment and local community. The Committee for the Coordination of Statistical Activities [7] found that the economic performance of Small Island Developing States (SIDS) has deteriorated sharply due to the impact of COVID-19 on tourism, based on SIDS’s international tourist arrivals experiencing a 77% loss in 2020. They found this to be due to SIDS’s economic dependence on tourism, SIDS’s limited scale of local markets, travel restrictions due to COVID-19, and the overall economic impact of COVID-19 on tourists [11].

The differences between island tourism and other tourism are the economic, social, cultural and environmental settings for island- and marine-specific areas, which are more developable, controllable, isolatable, vulnerable, and sensitive. Thus, the sustainability of financal, marine, industry, environmental, and cultures are the key issues of sustainable island tourism, for the following reasons: (1) the resource and environment threshold of marine areas and water; (2) the limited land, resource, and external transportation of islands; (3) the political and socio-economical inability, fragility and instability of islands; and (4) the island-specific and marine environmental and cultural tourism [12,13,14,15].

The players of sustainable island tourism should be its tourists, employees in the tourism and marine industries, and its residents, whose evaluations of the dimensions of sustainable island tourism should be the basis of the comprehensive evaluation model’s calculation. For the advantages of the Fuzzy Delphi method (FDM), the decision-making trial and evaluation laboratory (DEMATEL), and the analytic network process (ANP), this paper used FDM-DEMATEL-ANP (FDANP) to set the comprehensive evaluation model for sustainable island tourism [16,17,18]. Nilashi et al. [19] used a DEMATEL-ANP-based multicriteria decision-making (MCDM) approach to evaluate the factors in construction projects. Hall and Brown [20] believe that the major players of island tourism are tourists, employees, island residents, island animals, and natural and sociocultural environments.

The main purpose of this study is to propose a comprehensive evaluation model for sustainable island tourism, to find out the crucial sustainable island tourism criteria and promote the development of tourist destinations using a combined MCDM model based on the FDM, the decision-making trial and evaluation laboratory (DEMATEL), and an ANP in order to obtain more objective evaluation constructs and weights to evaluate sustainable island tourism. To the best of our knowledge, the sustainable tourism model has very few applications in the island context. The FDM-DEMATE-ANP method has been used to study tourists, employees in the tourism and marine industries, and residents’ viewpoints, accounting for sustainable island tourism constructs and criteria, in order to increase the competitive advantage of tourism destinations.

In addition to the Introduction, sustainable island tourism evaluation criteria are discussed in Section 2. The Methodology is proposed in Section 3. A case study of Taiwan is conducted in Section 4. The conclusion and a discussion on the application of our results is presented in Section 5. Future work is presented in Section 6.

2. Literature Review

2.1. The Initial List of Sustainable Island Tourism Evaluation Criteria

From the surveyed results in Web of Science, we found 765 articles with “island tourism” in the title from between 1990 and 10 August 2021. These articles focused on the social science, environmental science, and business economics of island tourism. There are 68 articles with the title “sustainable island tourism”, such as one by Fauzel and Tandrayen-Ragoobur [21], who found that sustainable economic, social, and environmental development variables influence tourism development. From the surveyed results in the Chinese national knowledge internet, 168 articles with “island tourism” in the title were found from between 1991 and 10 August 2021. These articles focused on the development planning and resource management of island tourism.

Past literature reviews believed that island-specific resources were limited, and that their socio-cultures were fragile, which did not align with the assumptions of an economic development model, and little consideration was given to the diversified development of island-specific and marine resources, industries and culture [22,23,24]. Dłużewska and Giampiccoli [1] thought that the island characteristics related to tourism were isolation (transportation problems) and closed culture/social groups. Guden et al. [25] thought that Cyprus’s characteristics were vulnerable tourism, limited economic independence, biological and cultural diversity, scarce resources, a fragile and sensitive ecosystem, and a high dependence on economic support from Turkey.

From the literature review results and experts’ surveys, the initial list of sustainable island tourism evaluation constructs and criteria in this paper is based on the impacts of tourism on islands, and on the advantages and disadvantages of island tourism. The major impacts of tourism on islands are economic development (as GDP) and industries (as infrastructure and agriculture), which are the primary considerations of island governments [1].

Bulchand-Gidumal [26] discussed the post-COVID-19 recovery of island tourism using a smart tourism destination framework, in which the dimensions of smart sustainability include sustainable infrastructure, sustainable destination planning, and sustainable tourism economies. Li and Lau [27] explored how the perceived sociocultural benefits of festivals articulate island tourism’s production of a sense of place and its respective locality. Zhao et al. [28] found that there are recreational and cultural ecosystem service values for island tourism. Eugenio and Rotarou [29] found that the key issues of island tourism are the governance and management of natural resources and tourism activities, the impacts of climate change, the diversification of economies, and the promotion of innovative and personalized tourist experiences.

The literature review results show that the advantages of island tourism are the developments of its production, employment, small and medium size enterprises, infrastructure, science and technologies, and natural and cultural resources [20,30,31,32,33]. Quevedo et al. [34] thought that the advantages of tourism are life quality, employment opportunities, economic diversity, natural and cultural attractions, and food and hospitality industries. Manwa and Manwa [30] thought that ecotourism should have direct (employment, small and medium sized enterprises), secondary (linkages/partnerships) and dynamic effects (sustainable livelihoods) on residents. Eagles et al. [13] thought that the goals of sustainable tourism in protected areas are economic sustainability, social/cultural sustainability, and the protection of sustainable resources. The impact of tourism on island socioeconomics, politics, and the environment are almost always negative because of infrastructure usage, energy consumption, environmental problems, natural resource deterioration, price appreciation, an increase in cross-border crimes, the degradation of island-specific cultures, and the worsening of life quality [1,35]. Manhas [36] thought that the management systems of island tourism destinations should aim at economically, culturally, and environmentally favorable tourism development for the benefits of its environment and stakeholders. Figueroa and Rotarou [11] thought that sustainable island tourism should consider institutional, environmental, social, and economic imperatives. Lim and Cooper [12] thought that the development dimensions of island tourism included external factors, internal factors, managerial factors, and key factors. The indicators of external factors are transportation, lifestyle, economic level, and the market. The indicators of internal factors are employment, social impact, economic benefit, and the political system. The indicators of managerial factors are policy and planning, and niche markets. The indicators of key factors are sustainability, vulnerability, resilience, tourism identity, and IT.

The literature review results show that the disadvantages of island tourism are: (1) governance, as unclear policies from different departments exist; (2) economics and finance, as revenue uncertainties stem from the emergent events of public health and seasonality, wage–price spirals, infrastructure maintenance costs, and economic leakage; (3) socio-cultural impacts, based on cross-border crimes, community antagonism, and cultural assimilation and exploitation; (4) environmental impacts, as in air, water, marine, and land pollution; noise, wastes, and sewages; and the loss of natural resources and biodiversity [37,38,39,40]. Jordan and Vogt [41] found that the negative impacts of tourism on residents are crowding/congestion, increased living costs, pollution, police harassment, and overused utilities.

Mota et al. [3] thought that the carrying capacity evaluation dimensions of island tourism are its infrastructure, the environment, economics, and society. Sosa et al. [42] thought that the sustainable tourism indicators for community-based food tourism are socio-culture, the environment, tourism, and economics. The dimensions of socio-culture include the local population’s knowledge of the culinary culture, including its recipes, traditions, and methods. The dimensions of its environment are the environmental sustainability of the food production process. The dimensions of tourism are its resources, services, and the infrastructure of food tourism, as well as its restaurants, hotels, and intangible and natural heritages. The dimensions of economics are the tourists’ expenditures on food tourism.

2.2. Sustainable Island Tourism Evaluation Criteria

Increasing numbers of literature reviews show that the key dimensions of sustainable tourism are governance, the environment, and socioeconomics. Fewer pieces of literature focus on finance, marine culture, and industry. Thus, this paper determined the key dimensions of sustainable island tourism to be governance, economy–finance, socio-culture, and the environment, and focused on marine culture, industry, and the environment [43,44,45,46,47].

The final sustainable island tourism evaluation criteria are listed in

Table 1. Sustainable island tourism evaluation criteria.

Dimensions	Sub-Dimensions	Indicators
Governance ($A_1$)	Governance Organizations ($A_{11}$)	Official Governance Organizations ($A_{111}$)
		Semi-Official Governance Organizations ($A_{112}$)
		Unofficial Governance Organizations ($A_{113}$)
	Governance Systems ($A_{12}$)	Policies and Regulations ($A_{121}$)
		Budgets and Subsidies ($A_{122}$)
		Governance Efficiency ($A_{123}$)
Economy and Finance ($A_2$)	Tourist’s Economic Sustainability ($A_{21}$)	Local Food and Drink ($A_{211}$)
		Shopping ($A_{212}$)
		Entertainment Activities ($A_{213}$)
	Industrial Economic Sustainability ($A_{22}$)	Tourism Industry ($A_{221}$)
		Marine Industry ($A_{222}$)
		Competitiveness & Service Quality ($A_{223}$)
	Financial Risks ($A_{23}$)	Emergent Events of Public Health ($A_{231}$)
		Seasonality ($A_{232}$)
		Wage—Price Spirals ($A_{233}$)
Scio-Culture ($A_3$)	Social Stability ($A_{31}$)	Cross-Border Crimes ($A_{311}$)
		Community’s Antagonism ($A_{312}$)
		Lifestyle Changes ($A_{313}$)
	Cultural Sustainability ($A_{32}$)	Island-Specific Cultural Sustainability ($A_{321}$)
		Marine Cultural Sustainability ($A_{322}$)
		Cultural Assimilation and Exploitation ($A_{323}$)
Environment ($A_4$)	Island-Specific Environmental Sustainability ($A_{41}$)	Air, Water, and Land Pollution ($A_{411}$)
		Noise, wastes and Sewages ($A_{412}$)
		Loss of Natural Resources and Biodiversity ($A_{413}$)
	Marine Environmental Sustainability ($A_{42}$)	Pollutions of Marine ($A_{421}$)
		Marine Ecological System ($A_{422}$)
		Coastal Erosion Disasters ($A_{423}$)

. The reasons for the governance dimension of sustainable island tourism are that most papers think that the purpose of sustainable tourism is finding balance among its economic, sociocultural, and environmental dimensions. Fewer papers discuss the matter of who is the balancer. Thus, this paper set the governance dimension ($A_1$) of sustainable island tourism to evaluate its balancer, the sub-dimensions of which are governance organizations ($A_{11}$) and governance systems ($A_{12}$). Quevedo et al. [34] and Dedeke [48] thought the purpose of sustainable tourism is the balance between environmental protection and economic development, and the improvements of the local economies and people’s wellbeing. Singh et al. [49] thought that the governance problems in Fiji island tourism are the governance, community capacity, and resources to implement environment policy, planning, and regulation. Yeniasir and Gokbulut [50] thought that the sustainable cultural tourism policies on Nicosia Island should be determined cooperatively by people, the government, and civil society. Figueroa and Rotarou [17] found that the main developing factor of sustainable island tourism is cooperation among its relevant stakeholders.

The indicators of governing organizations—which include official, semi-official, and unofficial entities ($A_{111}$, $A_{112}$., $A_{113}$)—aim to analyze and balance the costs and benefits of island tourism’s economic, sociocultural, and environmental dimensions. The indicators of governance systems—which include policies and regulations ($A_{121}$), budgets and subsidies ($A_{122}$), and government efficiency ($A_{123}$)—aim to evaluate the performances and efficiency of island tourism’s various governing systems. Aprilani et al. [51] found that the role of official government organizations in island tourism is to plan and develop infrastructure, policies, and regulations. Zgit and Ozturen [52] thought that insufficient and ineffective collaboration and policies are the main obstacles of sustainable island tourism in North Cyprus. Zhang et al. [53] thought that the optimization of island resources and tourism environments should be the foundation for sustainable island tourism.

The reasons for the economic and finance dimensions of sustainable island tourism are that more and more island economies are depending on tourism, but because of COVID-19 and the seasonality of tourism, their finances have faced serious challenges. Fewer papers discussed financial risks. Thus, this paper set the economy and finance dimension ($A_2$) of sustainable island tourism to achieve economic and financial sustainability, the sub-dimensions of which are tourists’ economic sustainability ($A_{21}$), industrial economic sustainability ($A_{22}$), and financial risks ($A_{23}$). Fauzel and Tandrayen-Ragoobur [21] proved that there are bi-directional causality and feedback effects in tourism and economic growth. Cirer-Costa [54] thought that the core elements of island tourism were economic growth, the island environment, and the entrepreneurship of tourism. Zhang et al. [53] thought that the economic environment of island tourism included transportation, accommodation, recreation, and shopping.

The indicators of tourism’s economic sustainability aim to analyze island tourist satisfaction with consumption behavior, including local food and drink ($A_{211}$), shopping ($A_{212}$), and entertainment activities ($A_{213}$). The indicators of industrial economic sustainability aim to evaluate the economic contribution and competition of island tourism’s related industries and activities, which include the tourism industry ($A_{221}$), the marine industry ($A_{222}$), and competitiveness and service quality ($A_{223}$). Island tourism-related marine industries refer to marine sightseeing, marine leisure and entertainment, marine accommodation and casting, and sports and other activities carried out on islands.

Chand et al. [6] found that tourism and marine industries contribute important factors to an island’s economy; for example, the average economic contributions of the tourism sector account for 30% of a given island’s gross domestic product. Quevedo et al. [34] found that island tourism depends highly on coastal and marine resources on Busuanga Island of the Philippines. Tsilimigkas and Rempis [55] found that dive tourism is the key to Rhodes’s socio-economic development. Yu and Spencer [56] thought that island tourism’s benefits for farmers include higher profits, farm resource efficiencies, food education, and rural cultural tradition education.

The indicators of financial risks evaluate revenue uncertainties from the events of island tourism, which are emergent events of public health ($A_{231}$), seasonality ($A_{232}$), and wage–price spirals ($A_{233}$). Chand et al. [6] found that COVID-19 destroyed many island economies, with unhealthy financial structures with high shares of foreign debt (more than 70%) for those whose foreign revenues come mostly from tourism. Lasso and Dahles [57] found that the characteristics of Komodo island tourism are a limited market, fierce competition, short tourist seasons, and dependence on cruise ships.

For the socio-cultural dimension of sustainable island tourism, more reviews found that the uniqueness and vulnerability of an island’s sociocultural environment include factors of sustainable island tourism. Thus, this paper set the socio-culture dimension ($A_3$) of sustainable island tourism to achieve social and culture sustainability, the sub-dimensions of which are social stability ($A_{31}$) and cultural sustainability ($A_{32}$). Zhang et al. [53] thought that the humanistic environment of island tourism included its fork culture, villages, and Buddhist culture.

The indicators of social stability evaluate the social impacts from the events of island tourism, which are cross-border crimes ($A_{311}$), community antagonism ($A_{312}$), and lifestyle changes ($A_{313}$). Andrefouet et al. [58] found that Nusa Lembongan Island tourism has changed its seaweed farming and socio-ecosystems. Grilli et al. [2] found that tourists are interested in tourism experiences with sociocultural and environmental resources. Yamagishi et al. [59] thought that the evaluation criteria of Bantayan Island tourism included the destination characteristics of marketing, the economy, society, and the environment.

The indicators of cultural sustainability evaluate the cultural impacts from the events of island tourism, which are island-specific cultural sustainability ($A_{321}$) and marine cultural sustainability ($A_{322}$). Li et al. [60] found that the cultural sustainability of Jeju Island is affected by its island tourism, which strongly correlates to collaboration with community-based island tourism. Thi [61] thought that the factors of Con Dao Islands’ tourism included tourism commodification, geographic location, and culture and beliefs.

Many reviews have proven that the environmental dimensions of sustainable island tourism include sensitivity and vulnerability to the natural environment. Environmental tourists’ motivations are unique, and the marine environment should be the most important component. Thus, this paper set the environment dimension ($A_4$) of sustainable island tourism to achieve its island-specific and marine sustainability, the sub-dimensions of which are island-specific environmental sustainability ($A_{41}$) and marine environmental sustainability ($A_{42}$). Phong and Tien [62] thought that sustainable island tourism should consider its marine environment and water capacity. Zhang et al. [53] thought that the natural environment of island tourism includes forests, beaches, seas, and coral reefs.

The indicators of island-specific environmental sustainability evaluate the island-specific environmental impacts from the events of island tourism, which include air, water, and land pollution ($A_{411}$); noise, waste, and sewages ($A_{412}$); and the loss of natural resources and biodiversity ($A_{413}$). Martins, and Cro [63] found tourism to be important in Madeira Island for its GDP share (26.6%) and its employment share (16.7%), but its solid-waste generation has been a major environmental problem based on its high share per resident of 41.9–46.6%. Phong and Tien [62] thought that the risk factors of sustainable island tourism included water supply facilities, wastewater treatment, and law enforcement. Brtnicky et al. [64] found that the biggest environmental problem of island tourism is heavy metal pollution, as the soil’s pollution due to Santorini Island tourism is contamination by heavy metals (Cu, Cr, and Pb).

Singh et al. [49] thought that environmental problems in Fiji island tourism include mangrove clearance, coastal degradation, vulnerability to natural disasters, low fish stock, low water quality, pollution, and decreased biodiversity. Akadiri et al. [65] found causal impacts of globalization, economic growth, and carbon emissions on island tourism. Figueroa and Rotarou [11] thought that the environmental issues of island tourism included waste disposal and management, sewage systems, water quality, and decreased biodiversity.

The indicators of marine environmental sustainability evaluate the marine environmental impacts from the events of island tourism, which are marine pollution ($A_{421}$), the marine ecological system ($A_{422}$), and coastal erosion disasters ($A_{423}$). Callejas-Jiménez et al. [66] indicated that Cozumel island’s reefs are a resource for tourism and renewable ocean energy. Cristiano et al. [67] found that Fernando de Noronha Island tourism depends on its beach landscapes. Grelaud and Ziveri [68] thought that island tourism is the main litter gateway of the marine environment, and found that the marine litter accumulation of Mediterranean islands follows a seasonal pattern.

From this discussion and NGT’s results, this paper set the dimensions of sustainable tourism as governance, economy and finance, socio-culture, and the environment, as shown in Table 1.

3. Methodology

From the results of the literature review and experts’ surveys, the comprehensive evaluation model and procedures for sustainable island tourism were set by FDANP, and are shown in Figure 1. As criteria screening tools, FDM, DEMATEL and ANP have the following advantages: they reduce the frequency and cost of investigations, they express experts’ opinions completely, they realize the combination of quantitative calculation and qualitative analysis, they quantify the inevitable fuzziness in the survey, they construct interrelations among dimensions and criteria in the integrated model, and they test the interdependency strengths among the dimensions and criteria [16,17,18].

The FDM incorporates the concept of Fuzzy Theory into the traditional Delphi method in order to integrate opinions and find the best method of solving the uncertainty of expert opinions, and to achieve better consensus convergence [69]. Using the fuzzy Delphi method can reduce the number of back-and-forth questionnaires, improve the efficiency and quality of the questionnaire, and screen out more objective evaluation factors. Therefore, it has the following advantages: (1) it can reduce the number of survey questionnaires; (2) individual expert opinions can be clearly explained and not distorted; (3) it can clarify the unavoidable ambiguity in the expert interview process and receive more rational answers; (4) it can improve the time and cost of the investigation; and (5) it can simplify the calculation process, which can handle multi-level, multi-attribute and multi-objective decision-making problems [70].

The DEMATEL method uses matrix-based related mathematical theory to calculate the degree of mutual influence between various criteria and elements, and establishes an impact relation map (IRM) in order to understand its causal relationship structure, thereby finding key influencing factors [71,72]. DANP (DEMATEL-based ANP) is a hybrid MCDM model that combines the DEMATEL and the ANP. DANP uses the functions and characteristics of DEMATEL to obtain the mutual causal relationship between all of the criteria by observing the degree of influence between each criterion, and expresses the strength and causal relationship between the criteria through IRM, so as to find the core problems and improvement directions in complex issues [73,74].

This study considers that the interactions between the evaluation criteria of sustainable island tourism are complex and cannot be fully explained by the analytic hierarchy process. Therefore, the ANP is adopted, which solves the problem of internal dependencies and external criteria feedback. In addition, the ANP defines each dimension as having the same weight, whereas, in practice, the pairwise influences between dimensions are often not the same. In other words, the ANP ignores the importance of the weight ratio relationship between dimensions. Therefore, this study uses the DEMATEL method to convert the degree of interaction between complex dimensions into a causal relationship, establish a network relationship structure, and finally calculate the weight of interaction between dimensions. In this way, the empirical results are more scientifically objective, which is the biggest difference between this study and previous research.

This study uses source data provided for the tourists, residents, and employers of the tourism and marine industries to discuss sustainable island tourism evaluation criteria. However, the available information in the sustainable island tourism context is inherently ambiguous, inaccurate, imprecise, and uncertain by nature. The comprehensive FDM-DEMATEL-ANP method may be useful for various MCDM problems. This is a crucial contribution to the paper. Therefore, there are many recent studies using the same method, such as Kiani et al. [75], who used FDANP, which combined the FDEMATEL and FANP methods to determine and prioritize the weight of the criteria and dimensions of blasting operations. Liao et al. [76] presented a hybrid model for the selection of the best online travel agencies (OTAs) using the FDM-DEMATEL-ANP. Hsu and Liou [77] applied the DANP (DEMATEL-based ANP) approach to select suppliers in the airline industry. Mltab et al. [17] used the hybrid method with FDM to evaluate smart product service system functions with high uncertainty and system complexity. Yang et al. [18] thought that the FDM was the most effective method to evaluate the vulnerability of urban wetland ecological environments. Izadpanah et al. [78] used the hybrid approach of FDANP to classify the medication management criteria of hospital accreditation standards. Shakeri et al. [79] used the FDANP technique to prioritize the determinants of maternal health promotion. Aghaee et al. [80] proposed a hybrid-structured MCDM based on FDM-DEMATEL-ANP as a structured methodology to assist strategic maintenance decision makers. Horng et al. [81] adopted the FDM-DEMATEL-ANP method to understand the assessment model of corporate social responsibility practices in the tourism industry.

The comprehensive evaluation model and procedure for sustainable island tourism is divided into three steps. The first step is to prepare an FDM questionnaire and conduct questionnaire surveys for the supervisors of tourists, residents, and the employers of the tourism and marine industries, and to combine those with the triangular fuzzy number to construct the evaluation dimensions and criteria of sustainable island tourism [33]. The second step is to use DEMATEL to evaluate the interdependence between the dimensions from different supervisors, and then explore the weight of each evaluation criterion and derive clear core criteria of sustainable island tourism. [74,82]. The third step is to use ANP to improve the interdependence and feedback between each evaluation criterion of sustainable island tourism. A super matrix is formed through a nonlinear network structure, and the super matrix is multiplied to a steady state in order to obtain relative weights [74,82].

4. Case Study of Taiwan

In the East Asian island arc formed by the Kuril Islands, Japan, the Ryukyu Islands, Taiwan, the Philippines, and many other islands in the western Pacific Ocean, Taiwan occupies a central position at the intersection of Northeast Asia and Southeast Asia. Taiwan is an important hub for Asia-Pacific trade and transportation, and sits in an important strategic location. Taiwan is about 395 km from north to south, and has a maximum width of 144 km from east to west. The coastline around the island is about 1151 km long. The administrative area includes Taiwan and its 22 affiliated islands, as well as 90 islands in the Penghu Islands in the Taiwan Strait.

Taiwan is rich in natural resources. It is surrounded by the sea on all sides. The many different terrains and landforms along the coast are even more special. It is suitable for various outdoor activities. It is very convenient for international tourists to go to the ocean, mountains, hot springs, and other places to experience Taiwan’s natural charm. The fusion of different ethnic groups has created unique and rich cultural assets from its aboriginal, Chinese, Japanese and Western cultures. Taiwan’s convenient transportation construction is complete, which is an important element in the development of tourism. In terms of air transportation, Taiwan currently has three international airports: Taoyuan International Airport, Kaohsiung International Airport, and Taipei Songshan Airport. There are frequent international direct flights every day. In terms of island transportation, in addition to the MRT system in the two major metropolitan areas, there is a dense road and railway network around the island, as well as a high-speed railway in the western area, making it more convenient for tourists to travel. Finally, the government supports and promotes the star rating system for the hotel industry, classifying hotels such that international tourists can quickly choose the type of hotel they need, and there is even the provision of customized services for enterprises. These factors have given Taiwan an excellent advantage as an important tourist destination [83].

4.1. Fuzzy Delphi Method (FDM)

This study reviewed the relevant literature and considered the governance, economic–financial, sociocultural, and environmental backgrounds of Taiwan’s islands, which provided the reference for the FDM questionnaire of sustainable island tourism. Thirty (30) FDM questionnaire surveys were issued to and collected from tourists, residents, and employers of the tourism and marine industries, and the results of the evaluations were summarized.

The statistical results for the individual variables of the first-stage questionnaire respondents are: 1/ Group: one third each for tourists, residents, and employers; 2/ Gender: 46.67% men and 53.33% women; 3/ Age: the average respondent age is 41.00 years old (with a standard deviation of 8.00); 4/ Education: 80.00% of the respondents have a bachelor’s degree or below, and 20.00% of the respondents have a master’s degree or above; 5/ Income: the average monthly income of respondents is TWD 35407.20 (with a standard deviation of 19,121.50).

From the questionnaire results,

Table 2. The triangular fuzzy functions of the various evaluation indicators of sustainable island tourism.

Indicators	Max	GM	Min	Threshold Value	Eliminated (E)/Remain(R)
$A_{111}$	10.00	7.89	5.00	7.00	R
$A_{112}$	10.00	7.26	4.00		R
$A_{113}$	10.00	8.03	5.00		R
$A_{121}$	10.00	8.14	3.00		R
$A_{122}$	10.00	7.06	4.00		R
$A_{123}$	10.00	8.32	6.00		R
$A_{211}$	10.00	7.28	3.00		R
$A_{212}$	10.00	7.98	2.00		R
$A_{213}$	10.00	8.32	4.00		R
$A_{221}$	10.00	7.63	3.00		R
$A_{222}$	10.00	7.35	3.00		R
$A_{223}$	10.00	8.13	5.00		R
$A_{231}$	10.00	7.50	5.00		R
$A_{232}$	10.00	7.86	3.00		R
$A_{233}$	9.00	5.39	2.00		E
$A_{311}$	10.00	7.63	4.00		R
$A_{312}$	10.00	7.74	5.00		R
$A_{313}$	10.00	6.58	2.00		E
$A_{321}$	10.00	7.30	3.00		R
$A_{322}$	10.00	7.03	4.00		R
$A_{323}$	10.00	7.20	4.00		R
$A_{411}$	10.00	7.91	5.00		R
$A_{412}$	10.00	8.31	5.00		R
$A_{413}$	10.00	8.65	4.00		R
$A_{421}$	10.00	7.26	5.00		R
$A_{422}$	10.00	8.03	3.00		R
$A_{423}$	10.00	6.39	2.00		E

shows a summary of the triangular fuzzy number of the various evaluation criteria of sustainable island tourism. From

Table 3. The initial influence matrix on each dimension of sustainable island tourism.

	${\mathit{A}}_1$	${\mathit{A}}_2$	${\mathit{A}}_3$	${\mathit{A}}_4$
$A_1$	0	3.00	2.30	2.30
$A_2$	2.00	0	2.70	3.00
$A_3$	2.00	1.30	0	1.70
$A_4$	2.30	1.70	1.00	0

, MAX and MIN represent the smallest and largest evaluations of the criterion regarding questionnaire respondents; the range of MAX is 9.00–10.00, and the range of MIN is 2.00–5.00. GM represents the geometric mean of the respondents’ evaluation of each criterion; the range of GM is 5.40–8.70. The threshold value of 7 was used to filter and select the appropriate evaluation criteria, which were decided by the experts.

From Table 3, the criteria for which the geometric mean failed to meet the threshold value are “Revenue Uncertainties from the Wage–Price Spiral” ($A_{233}$, 5.4), “Lifestyle Changes” ($A_{313}$, 6.6), and “Coastal Erosion Disasters” ($A_{423}$, 6.4). After these criteria were eliminated, subsequent analysis was performed on the remaining 24 criteria.

4.2. Decision-Making Trial and Evaluation Laboratory (DEMATEL)

The calculation steps of DEMATEL are described as follows:

(1) Establish the initial average matrix on each dimension of sustainable island tourism. The evaluation scale goes from 0 (no influence) to 4 (very high influence), and is designed to establish the interviewee’s subjective judgments on the elements, and then establish the initial influence matrix on each dimension of sustainable island tourism.

From Table 3, we can see that the influence of governance on the economic and financial, and sociocultural dimensions is between middle and high (2.3–3.0); the influence of economy and finance on governance, socio-culture, and the environment is between middle and high (2.0–3.0); the influence of socio-culture on governance, economy and finance, and the environment is low (1.3–2.0); and the influence of the environment on governance, economy and finance, and socio-culture is low (1.0–1.7).

(2) Establish a full (direct/indirect) influence matrix (see

Table 4. The full influence matrix on each dimension of sustainable island tourism.

	${\mathit{A}}_1$	${\mathit{A}}_2$	${\mathit{A}}_3$	${\mathit{A}}_4$
$A_1$	2.50	2.70	2.62	2.95
$A_2$	2.66	2.33	2.57	2.92
$A_3$	2.01	1.87	1.68	2.11
$A_4$	2.09	1.96	1.87	1.98

). After normalizing Table 4, the sum of the values of each row ($D$) and column ($R$) in the matrix is calculated, and then the full (direct/indirect) influence matrix is created.

(3) Establish the threshold value and generate an impact relation map (IRM). In order to discover the important influence relationship among the various dimensions and criteria, a threshold value is set for the secondary influence factors. If the value of an element in the full influence matrix is higher than this threshold value, it means that there is a greater relevance. If it is less than the threshold value, indicating that the relevance is small, it is deleted and the value in the matrix is set to 0. In order to maintain important relationships, a threshold value of 2 was used after a consensus was arrived at between the experts. In

Table 5. The full influence matrix after setting the threshold value.

	${\mathit{A}}_1$	${\mathit{A}}_2$	${\mathit{A}}_3$	${\mathit{A}}_4$	$\mathit{D}$	$\mathit{R}$	$\mathit{D}+\mathit{R}$	$\mathit{D}-\mathit{R}$
$A_1$	2.50	2.70	2.62	2.95	10.78	9.26	20.03	1.52
$A_2$	2.66	2.33	2.57	2.92	10.48	5.03	15.51	5.45
$A_3$	2.01	0	0	2.11	4.11	5.19	9.30	−1.08
$A_4$	2.09	0	0	0	2.09	7.98	10.07	−5.89

, the full influence after setting the threshold value is shown.

The $D+R$ values represent the total influence levels. The $D-R$ values represent the net influence levels where positive values indicate that one dimension will influence other dimensions more than other ones influence it. The influence level of the four dimensions can be prioritized as Governance ($A_1$) > Economic and Financial ($A_2$) > Environment ($A_4$) > Socio-cultural ($A_3$) based on the $D+R$ values. Based on the $D-R$ values, Governance ($A_1$), and Economic and Financial ($A_2$) are net causes, whereas the others are net receivers.

This study uses Table 5 to calculate the normalized full influence matrix after setting the threshold value, as shown in

Table 6. The normalized full influence matrix after setting the threshold value.

	${\mathit{A}}_1$	${\mathit{A}}_2$	${\mathit{A}}_3$	${\mathit{A}}_4$
$A_1$	0.27	0.54	0.51	0.37
$A_2$	0.29	0.46	0.50	0.37
$A_3$	0.22	0	0	0.26
$A_4$	0.23	0	0	0

. According to Table 6, the network evaluation structure of the ANP is generated in Figure 2. IRM (the network evaluation structure of the ANP) was further developed to accurately reflect the complex causal relationship between measurement dimensions, and IRM might provide a clear picture for the loop of interdependence and feedback between criteria. Therefore, it seems that the IRM provides an overview of this complicated problem.

4.3. Analytic Network Process (ANP)

The ANP calculation method on the super matrix on each dimension of sustainable island tourism is described as follows:

(1) Perform a pairwise comparison, calculate eigenvectors, and derive a super matrix. This paper used pairwise comparisons to estimate the weights of sustainable island tourism evaluation criteria. Because the consistency index (CI) and ratio (CR) of all of the samples in this paper were less than 0.10, they all passed the consistency test.

According to each criterion, the unweighted super matrix can be obtained by comparing the matrix in pairs, as shown in

Table A1. The non-weighted super matrix of each dimension of sustainable island tourism.

	${\mathit{A}}_{111}$	${\mathit{A}}_{112}$	${\mathit{A}}_{113}$	${\mathit{A}}_{121}$	${\mathit{A}}_{122}$	${\mathit{A}}_{123}$	${\mathit{A}}_{211}$	${\mathit{A}}_{212}$	${\mathit{A}}_{213}$	${\mathit{A}}_{221}$	${\mathit{A}}_{222}$	${\mathit{A}}_{223}$	${\mathit{A}}_{231}$	${\mathit{A}}_{232}$	${\mathit{A}}_{311}$	${\mathit{A}}_{312}$	${\mathit{A}}_{321}$	${\mathit{A}}_{322}$	${\mathit{A}}_{323}$	${\mathit{A}}_{411}$	${\mathit{A}}_{412}$	${\mathit{A}}_{413}$	${\mathit{A}}_{421}$	${\mathit{A}}_{422}$
$A_{111}$	0.09	0.12	0.12	0.12	0.11	0.11	0.13	0.13	0.13	0.17	0.18	0.17	0.13	0.13	0.12	0.08	0.08	0.08	0.08	0.16	0.16	0.15	0.16	0.12
$A_{112}$	0.08	0.04	0.10	0.07	0.06	0.06	0.10	0.10	0.10	0.12	0.12	0.11	0.08	0.08	0.06	0.05	0.05	0.05	0.05	0.08	0.08	0.08	0.08	0.08
$A_{113}$	0.07	0.06	0.04	0.06	0.05	0.06	0.12	0.11	0.12	0.11	0.11	0.11	0.07	0.06	0.06	0.05	0.05	0.05	0.05	0.08	0.08	0.07	0.08	0.07
$A_{121}$	0.13	0.11	0.12	0.06	0.09	0.10	0.12	0.12	0.13	0.16	0.17	0.16	0.13	0.10	0.12	0.07	0.07	0.07	0.07	0.15	0.15	0.12	0.15	0.11
$A_{122}$	0.09	0.08	0.09	0.08	0.04	0.08	0.08	0.08	0.08	0.11	0.11	0.10	0.11	0.08	0.07	0.05	0.05	0.05	0.05	0.11	0.11	0.08	0.11	0.08
$A_{123}$	0.09	0.08	0.10	0.09	0.07	0.05	0.09	0.09	0.09	0.13	0.13	0.11	0.10	0.09	0.10	0.05	0.06	0.06	0.06	0.12	0.12	0.09	0.12	0.08
$A_{211}$	0.09	0.07	0.08	0.07	0.07	0.07	0.09	0.15	0.15	0.17	0.17	0.16	0.12	0.12	0.11	0.08	0.08	0.08	0.08	0.14	0.14	0.14	0.14	0.14
$A_{212}$	0.09	0.07	0.07	0.07	0.07	0.07	0.15	0.09	0.15	0.16	0.17	0.16	0.11	0.11	0.11	0.08	0.08	0.08	0.08	0.14	0.14	0.14	0.14	0.14
$A_{213}$	0.08	0.07	0.07	0.07	0.06	0.07	0.14	0.14	0.09	0.15	0.16	0.15	0.11	0.11	0.11	0.07	0.08	0.08	0.08	0.13	0.13	0.13	0.13	0.13
$A_{221}$	0.11	0.09	0.09	0.09	0.08	0.08	0.14	0.14	0.14	0.12	0.17	0.16	0.12	0.12	0.12	0.10	0.11	0.11	0.11	0.16	0.16	0.15	0.16	0.15
$A_{222}$	0.10	0.08	0.09	0.08	0.08	0.08	0.11	0.11	0.12	0.15	0.11	0.14	0.10	0.09	0.09	0.08	0.08	0.08	0.08	0.14	0.14	0.14	0.14	0.13
$A_{223}$	0.06	0.05	0.06	0.05	0.05	0.05	0.11	0.11	0.11	0.12	0.12	0.07	0.07	0.07	0.06	0.05	0.05	0.05	0.05	0.08	0.08	0.07	0.08	0.07
$A_{231}$	0.13	0.12	0.13	0.12	0.12	0.12	0.15	0.15	0.16	0.17	0.18	0.17	0.08	0.14	0.09	0.07	0.07	0.07	0.07	0.12	0.12	0.11	0.12	0.11
$A_{232}$	0.05	0.05	0.05	0.05	0.04	0.04	0.06	0.06	0.07	0.11	0.11	0.11	0.05	0.04	0.05	0.04	0.04	0.04	0.04	0.07	0.07	0.06	0.07	0.06
$A_{311}$	0.08	0.07	0.07	0.07	0.05	0.05	0.11	0.11	0.12	0.13	0.13	0.12	0.07	0.07	0.05	0.05	0.05	0.05	0.06	0.08	0.08	0.08	0.08	0.08
$A_{312}$	0.06	0.05	0.05	0.05	0.05	0.05	0.09	0.09	0.10	0.11	0.11	0.10	0.06	0.06	0.06	0.03	0.06	0.06	0.06	0.07	0.07	0.07	0.07	0.07
$A_{321}$	0.06	0.05	0.05	0.05	0.05	0.05	0.09	0.09	0.09	0.10	0.10	0.09	0.06	0.06	0.05	0.04	0.03	0.07	0.07	0.07	0.07	0.07	0.07	0.06
$A_{322}$	0.06	0.05	0.05	0.05	0.04	0.05	0.09	0.09	0.09	0.10	0.10	0.09	0.06	0.06	0.05	0.04	0.06	0.03	0.07	0.07	0.07	0.06	0.07	0.06
$A_{323}$	0.06	0.05	0.05	0.05	0.05	0.05	0.09	0.09	0.10	0.10	0.11	0.10	0.06	0.06	0.06	0.05	0.06	0.06	0.03	0.07	0.07	0.07	0.07	0.07
$A_{411}$	0.11	0.06	0.06	0.06	0.06	0.06	0.11	0.11	0.11	0.12	0.13	0.11	0.07	0.07	0.07	0.06	0.06	0.06	0.06	0.08	0.14	0.13	0.14	0.13
$A_{412}$	0.10	0.06	0.06	0.06	0.05	0.05	0.09	0.09	0.10	0.11	0.11	0.10	0.07	0.07	0.06	0.05	0.05	0.05	0.05	0.13	0.07	0.12	0.13	0.12
$A_{413}$	0.11	0.06	0.06	0.06	0.06	0.06	0.10	0.10	0.11	0.12	0.12	0.11	0.07	0.07	0.07	0.05	0.06	0.06	0.06	0.13	0.13	0.07	0.13	0.13
$A_{421}$	0.09	0.05	0.06	0.05	0.05	0.05	0.07	0.07	0.09	0.10	0.12	0.08	0.06	0.06	0.06	0.05	0.05	0.05	0.05	0.12	0.12	0.11	0.07	0.12
$A_{422}$	0.08	0.05	0.05	0.05	0.05	0.05	0.07	0.07	0.09	0.10	0.11	0.08	0.06	0.06	0.06	0.05	0.05	0.05	0.05	0.12	0.12	0.11	0.12	0.06

in Appendix A. In order to normalize all the row vectors, the sum of the rows is 1, and the unweighted super matrix is standardized in order to obtain the standardized super matrix, as shown in

Table A2. The weighted super matrix of each dimension of sustainable island tourism.

	${\mathit{A}}_{111}$	${\mathit{A}}_{112}$	${\mathit{A}}_{113}$	${\mathit{A}}_{121}$	${\mathit{A}}_{122}$	${\mathit{A}}_{123}$	${\mathit{A}}_{211}$	${\mathit{A}}_{212}$	${\mathit{A}}_{213}$	${\mathit{A}}_{221}$	${\mathit{A}}_{222}$	${\mathit{A}}_{223}$	${\mathit{A}}_{231}$	${\mathit{A}}_{232}$	${\mathit{A}}_{311}$	${\mathit{A}}_{312}$	${\mathit{A}}_{321}$	${\mathit{A}}_{322}$	${\mathit{A}}_{323}$	${\mathit{A}}_{411}$	${\mathit{A}}_{412}$	${\mathit{A}}_{413}$	${\mathit{A}}_{421}$	${\mathit{A}}_{422}$
$A_{111}$	0.04	0.07	0.07	0.07	0.08	0.07	0.05	0.05	0.05	0.06	0.06	0.06	0.07	0.07	0.06	0.05	0.05	0.05	0.05	0.06	0.06	0.06	0.06	0.05
$A_{112}$	0.04	0.03	0.05	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.03	0.04	0.04	0.04	0.04	0.03	0.03	0.03	0.03	0.03
$A_{113}$	0.03	0.04	0.02	0.04	0.04	0.04	0.05	0.05	0.04	0.04	0.04	0.04	0.03	0.03	0.03	0.04	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03
$A_{121}$	0.06	0.07	0.07	0.04	0.06	0.06	0.05	0.05	0.05	0.05	0.05	0.06	0.06	0.05	0.07	0.05	0.05	0.05	0.05	0.06	0.06	0.05	0.06	0.05
$A_{122}$	0.05	0.05	0.05	0.05	0.03	0.05	0.03	0.03	0.03	0.04	0.04	0.04	0.05	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.03	0.04	0.03
$A_{123}$	0.04	0.05	0.06	0.06	0.05	0.03	0.04	0.04	0.04	0.04	0.04	0.04	0.05	0.05	0.05	0.04	0.04	0.04	0.04	0.05	0.05	0.04	0.05	0.04
$A_{211}$	0.04	0.04	0.04	0.04	0.05	0.05	0.04	0.06	0.06	0.06	0.05	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06
$A_{212}$	0.04	0.04	0.04	0.04	0.05	0.04	0.06	0.04	0.06	0.05	0.05	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.05	0.05	0.06	0.05	0.06
$A_{213}$	0.04	0.04	0.04	0.04	0.04	0.04	0.05	0.05	0.04	0.05	0.05	0.05	0.06	0.06	0.06	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05
$A_{221}$	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.06	0.05	0.04	0.05	0.06	0.06	0.06	0.06	0.07	0.07	0.07	0.07	0.06	0.06	0.06	0.06	0.06
$A_{222}$	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.03	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.06	0.05	0.06
$A_{223}$	0.03	0.03	0.03	0.03	0.03	0.03	0.04	0.04	0.04	0.04	0.04	0.03	0.04	0.04	0.03	0.04	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03
$A_{231}$	0.07	0.07	0.07	0.07	0.08	0.08	0.06	0.06	0.06	0.06	0.06	0.06	0.04	0.07	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.04
$A_{232}$	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.04	0.04	0.04	0.03	0.02	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03
$A_{311}$	0.04	0.04	0.04	0.04	0.04	0.04	0.05	0.05	0.04	0.04	0.04	0.04	0.03	0.03	0.03	0.04	0.04	0.04	0.04	0.03	0.03	0.03	0.03	0.03
$A_{312}$	0.03	0.03	0.03	0.03	0.03	0.03	0.04	0.04	0.04	0.04	0.03	0.04	0.03	0.03	0.03	0.02	0.04	0.04	0.04	0.03	0.03	0.03	0.03	0.03
$A_{321}$	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.02	0.04	0.04	0.03	0.03	0.03	0.03	0.03
$A_{322}$	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.04	0.02	0.04	0.03	0.03	0.03	0.03	0.03
$A_{323}$	0.03	0.03	0.03	0.03	0.03	0.03	0.04	0.04	0.04	0.03	0.03	0.04	0.03	0.03	0.03	0.03	0.04	0.04	0.02	0.03	0.03	0.03	0.03	0.03
$A_{411}$	0.05	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.03	0.05	0.05	0.05	0.05
$A_{412}$	0.05	0.04	0.03	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.03	0.03	0.03	0.04	0.04	0.04	0.04	0.05	0.03	0.05	0.05	0.05
$A_{413}$	0.05	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.05	0.05	0.03	0.05	0.05
$A_{421}$	0.05	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.04	0.03	0.04	0.03	0.03	0.03	0.03	0.04	0.03	0.03	0.03	0.05	0.05	0.05	0.03	0.05
$A_{422}$	0.04	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.04	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.04	0.04	0.05	0.04	0.03

in Appendix A.

From Table A1 in Appendix A, the mean and coefficient of variance (CV) on the weights of sustainable island tourism evaluation criteria are summarized in

Table 7. The mean and CV on the weights of sustainable island tourism evaluation criteria.

Dimensions		Indicators	Mean	C.V.	Dimensions		Indicators	Mean	C.V.
$A_1$	$A_{11}$	$A_{111}$	0.13	24.38%	$A_3$	$A_{31}$	$A_{311}$	0.08	32.10%
		$A_{112}$	0.08	28.27%			$A_{312}$	0.07	29.92%
		$A_{113}$	0.07	33.11%		$A_{32}$	$A_{321}$	0.06	28.36%
	$A_{12}$	$A_{121}$	0.11	27.65%			$A_{322}$	0.06	28.32%
		$A_{122}$	0.08	25.44%			$A_{323}$	0.07	30.88%
		$A_{123}$	0.09	27.37%	$A_4$	$A_{41}$	$A_{411}$	0.09	34.60%
$A_2$	$A_{21}$	$A_{211}$	0.11	31.00%			$A_{412}$	0.08	34.37%
		$A_{212}$	0.11	31.24%			$A_{413}$	0.09	34.96%
		$A_{213}$	0.10	30.52%		$A_{42}$	$A_{421}$	0.08	34.76%
	$A_{22}$	$A_{221}$	0.12	23.10%			$A_{422}$	0.07	35.21%
		$A_{222}$	0.10	24.60%
		$A_{223}$	0.07	33.35%
	$A_{23}$	$A_{231}$	0.12	27.33%
		$A_{232}$	0.06	35.22%

. In Table 7, the top five evaluation criteria by their CV are “Seasonality”, “Marine Ecological System”, “Losses of Natural Resources and Biodiversity”, “Pollutions of Marine”, and “Pollutions of Air, Water, and Land”.

(2) Calculate the overall criterion weight and ranking (Criterion Ranking). By limiting the weighted super matrix, the limiting super matrix that converges to a stable state can be obtained, as shown in

Table A3. The limiting super matrix of each dimension of sustainable island tourism.

	${\mathit{A}}_{111}$	${\mathit{A}}_{112}$	${\mathit{A}}_{113}$	${\mathit{A}}_{121}$	${\mathit{A}}_{122}$	${\mathit{A}}_{123}$	${\mathit{A}}_{211}$	${\mathit{A}}_{212}$	${\mathit{A}}_{213}$	${\mathit{A}}_{221}$	${\mathit{A}}_{222}$	${\mathit{A}}_{223}$	${\mathit{A}}_{231}$	${\mathit{A}}_{232}$	${\mathit{A}}_{311}$	${\mathit{A}}_{312}$	${\mathit{A}}_{321}$	${\mathit{A}}_{322}$	${\mathit{A}}_{323}$	${\mathit{A}}_{411}$	${\mathit{A}}_{412}$	${\mathit{A}}_{413}$	${\mathit{A}}_{421}$	${\mathit{A}}_{422}$
$A_{111}$	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06
$A_{112}$	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04
$A_{113}$	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04
$A_{121}$	0.05	0.06	0.06	0.06	0.06	0.06	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.06	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05
$A_{122}$	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04
$A_{123}$	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04
$A_{211}$	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05
$A_{212}$	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05
$A_{213}$	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05
$A_{221}$	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06
$A_{222}$	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05	0.05
$A_{223}$	0.03	0.03	0.03	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04
$A_{231}$	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06	0.06
$A_{232}$	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03
$A_{311}$	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04
$A_{312}$	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03
$A_{321}$	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03
$A_{322}$	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03
$A_{323}$	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03
$A_{411}$	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04
$A_{412}$	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04
$A_{413}$	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04
$A_{421}$	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04	0.04
$A_{422}$	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.03	0.04

in Appendix A.

From Table A3 in Appendix A, the relative weight ranking of each dimension and criterion can be summarized, as shown in

Table 8. Rankings for the sustainable island tourism dimensions and indicators.

Dimensions		Indicators	Relative Weight	Ranking	Ratio of Each Dimension	Ranking
$A_1$	$A_{11}$	$A_{111}$	0.06	1	0.269	2
		$A_{112}$	0.04	15
		$A_{113}$	0.04	17
	$A_{12}$	$A_{121}$	0.05	4
		$A_{122}$	0.04	12
		$A_{123}$	0.04	9
$A_2$	$A_{21}$	$A_{211}$	0.05	5	0.381	1
		$A_{212}$	0.05	6
		$A_{213}$	0.05	8
	$A_{22}$	$A_{221}$	0.06	2
		$A_{222}$	0.05	7
		$A_{223}$	0.03	18
	$A_{23}$	$A_{231}$	0.06	3
		$A_{232}$	0.03	24
$A_3$	$A_{31}$	$A_{311}$	0.04	14	0.161	4
		$A_{312}$	0.03	20
	$A_{32}$	$A_{321}$	0.03	22
		$A_{322}$	0.03	23
		$A_{323}$	0.03	21
$A_4$	$A_{41}$	$A_{411}$	0.04	10	0.190	3
		$A_{412}$	0.04	13
		$A_{413}$	0.04	11
	$A_{42}$	$A_{421}$	0.04	16
		$A_{422}$	0.03	19

. In Table 8, the results are accompanied by the ranking of each measurement criterion. Based on the results, the study found out that the weight sequence of dimensions from high to low are economy and finance ($A_2$), governance ($A_1$), environment ($A_3$), and socio-culture ($A_4$). “Official Administer Organizations ($A_{111}$)”, “Tourism Industry ($A_{221}$)”, “Revenue Uncertainties from the Emergent Events of Public Health ($A_{231}$)”, “Policies and Regulations ($A_{121}$)”, and “Local Food and Drink ($A_{211}$)” are found to be the five critical sustainable island tourism indicators that should prioritized when considering how to promote the local economy and tourism.

5. Conclusions and Discussion

5.1. Conclusions

This paper proposed an approach that combines the use of FDM, DEMATEL and ANP in the evaluation of key sustainable island tourism criteria. The study first established the sustainable island tourism evaluation framework through decisions by tourists, residents, and tourism and marine industry employers obtained through an FDM. A DEMATEL was used to obtain the clear core criteria based on the subjective judgments of the tourists, residents, and employers. Finally, ANP was used to obtain the weights of sustainable island tourism evaluation criteria and rank their priority. The main contribution of this paper is to propose an efficient and simple method to assist tourism management authorities in choosing the key sustainable island tourism criteria that will drive business opportunities and promote the development of tourist destinations.

First, the study filtered and selected four main dimensions and 24 criteria of sustainable island tourism based on the analytic findings of FDM expert questionnaires. The “revenue uncertainties from the wage–price spirals” ($A_{233}$), “lifestyle changes” ($A_{313}$), and “coastal erosion disasters” ($A_{423}$) were eliminated, as they did not meet the filtering threshold. Secondly, the DEMATEL method was used to develop the interrelationships among the evaluation dimensions in order to form an IRM. The influence level of the four dimensions can be prioritized as Governance ($A_1$) > Economic and Financial ($A_2$) > Environment ($A_4$) > Sociocultural ($A_3$). In the final step, the ANP replaced hierarchies with networks, emphasizing the interdependent relationships among various decision criteria, which were used to calculate the relative weights of the sustainable island tourism evaluation criteria. The weight sequence of the four dimensions from high to low is economy and Finance ($A_2$), Governance ($A_1$), Environment ($A_3$), and Socio-culture ($A_4$). The top five criteria of sustainable island tourism are administering organization criteria, tourism industry criteria, revenue uncertainties from the emergent events of public health criteria, policies and regulations criteria, and local food and drink criteria. Furthermore, based on the IRM results, the “economic and financial” dimension of sustainable island tourism should influence the other dimensions more than it is influenced by them. Furthermore, the “Governance” dimension of sustainable island tourism should affect other criteria and be significantly affected by them.

For the economic impact of COVID-19 on island tourism, most respondents thought the necessary condition of sustainable island tourism is its economic and financial sustainability. The costs and benefits of industries related to island tourism would be significantly affected by the governance dimension. Tourists’ willingness to travel is affected by all dimensions, but the most important criteria are the tourism industry and local food and drink. Lastly, local residents cared about all of the dimensions, but regarding the impact of COVID-19, they hope that the economic and financial problems can be solved as soon as possible.

5.2. Discussion

From the empirical results shown, this paper specifically proposes theoretical and practical implications, which are explained as follows.

In the theoretical implications area, this paper’s contributions are:

(1) As a supplement to the literature of sustainable island tourism, which is evaluated by the dimensions of governance, economy–finance, socio-culture, and the environment. Although many papers have discussed the issues of sustainable tourism, fewer focus on sustainable island tourism based on the uniqueness of island-specific and marine resources, industries, and culture.

(2) As a supplement to the evaluation criteria of sustainable island tourism which focuses on island tourism management organizations and systems. In most papers, sustainable tourism means that there are gains from tourism in all dimensions and for all groups, but this might not be the optimal solution. Sometimes, the optimal solution is that there is less loss in one dimension and many more gains in other dimensions. Thus, this paper used sustainable island tourism management organizations and systems as the criteria to evaluate sustainable island tourism.

(3) As a supplement to the evaluation criteria of sustainable island tourism which focuses on marine industry, culture, and the environment.

(4) As a supplement on the empirical literature of sustainable island tourism with their unique evaluation dimensions and criteria [14,84,85,86,87].

In the practical implications area, this study proposes several directions, which are explained as follows:

(1) Local food and drink are the most important attractions of sustainable island tourism. The reasons why this is important to sustainable island tourism are based on the multiplier effects of tourists’ food consumption, the carbon footprint reduction without food transportation, and tourism reflexivity, with its uniqueness and comparative advantage [42,88,89,90,91,92]. Sims [88], Clark and Chabrel [86], and Ilbery et al. [87] thought that local food and drink could improve the economic and environmental sustainability of tourism.

(2) Protecting marine environments is an important part of sustainable island tourism. This is because the blue carbon ecosystem is a part of the marine ecological system, which provides a species habitat, coastal protection, cultural services, and carbon sequestration [34,93,94]. Zhang et al. [53] thought that the development topics of island tourism include ecotourism concepts and ecological protection, tourism and brand activities, tourism propaganda, and infrastructure and outlets.

(3) More attention should be paid to the resource and environmental issues of sustainable island tourism. As Ara et al. [95] found, island tourism impacts the coral reef and local microclimate, as the reduction in the coral reef footprint was 38% in St. Martin’s Island. Tsilimigkas and Rempis [55] found significant human-induced island tourism activities that underline the conflicts of sea use. Quevedo et al. [40] highlighted the environmental disadvantages of island tourism as the scarcity of water and resources; land, water, and air pollution; and noise and solid waste. Grydehøj and Kelman [22] thought that sustainable island tourism should forbid the eco-island trap of inefficient or ineffective investments into renewable energy and sustainability initiatives for the benefits of ecotourism. Peeters and Landré [37] thought that the main growth for tourism emissions is global time–space expansion. Carbone and Yunis [39] thought that the environmental issues of dynamic tourism growth are the environments of the coast, marine environments, and nature, and the sites of history and cultural heritage.

(4) Society and community issues on sustainable island tourism should be considered. Zhang et al. [53] found that the homestay is tourists’ favorite accommodation option, and beach leisure is their favorite activity. Pham et al. [35] thought that residents’ life quality is the influencing factor of sustainable tourism, and the content of their life quality included the tourism economy, environment, and social and cultural life. Cole [33] thought that the information and empowerment of community participation are the keys to sustainable tourism.

More precisely, this paper’s contribution and the managerial insights are as follows: (1) It provides a supplement to the literature of sustainable island tourism. This paper first discovered that the existing literature lacks details in governance and finance dimensions, and in the marine industry, culture, and environment criteria. (2) The empirical practice of the FDANP method is used to explain the relative weight ranking of each dimension and criterion of sustainable island tourism. The MCDM model provides a simple decision support system to solve complicated problems that are involved with many different departments or units in the real word, such as developing a sustainable island tourism policy in the case of Taiwan. (3) For the stakeholders (such as destination managers and local government administrators) of island tourism, the optimal actions and strategies for sustainable island tourism should be solved by the dynamic multi-actor process, which compares to the FDANP process in this paper. (4) For destination managers of island tourism, financial planning should be the most important consideration, as in the planning of cash flow and closure funding. (5) For local government administrators of island tourism, policies could focus on the financial issues of sustainable island tourism, such as adjustable tax policies and labor right protection policies.

6. Future Works

There are a few limitations in this study. First, although we incorporated the perceptions of tourists, residents, and employers of the tourism and marine industries for evaluation, the current sample size was insufficient. Thus, it is suggested that future research collect more samples. In addition to retesting the research model of this study, it is possible to establish a rigorous statistical analysis model to specifically identify the important influencing factors affecting sustainable island tourism. Second, this study used Taiwan as an example. Therefore, the results of this study may not be generalizable to other countries. It is suggested that future research collect data from other countries in order to retest the model to compare changes in the ranking of dimensions and criteria. Finally, future research could consider or combine a Fuzzy TOPSIS and MCGP model [96], the WIRN-biased COPRAS model [97], and the preference ranking organization method for enrichment evaluation (PROMETHEE) model [98] into the FDANP model, in order to further compare the methods described above with the proposed integrated model.