IUCN's Application of Cluster Analysis to Usage Classification Areas in Korean protected areas: Based on Gangwon Province
Article information
Abstract
Background and objective
Protected areas in South Korea are currently designated by different ministries for similar purposes, causing overlapping management and conflicts of interest. Moreover, protected areas after designation have been lacking sufficient management. Thus, it is necessary to classify the types of protected areas in South Korea considering variables related to designation, management, and the natural environment, and to verify the applicability of IUCN categories.
Methods
We selected variables that are easy to apply to spatial data, easy to interpret, and can solve problems within protected areas based on previous studies and literature review. The optimal number of clusters was derived by conducting hierarchical and non-hierarchical cluster analysis on 174 protected areas using the selected variables. The applicability of IUCN categories was verified based on the results.
Results
Cluster centroids represent the characteristics of the variables in each cluster, which were applied to IUCN categories for classification. Based on this standard, we formed clusters of similar protected areas through cluster analysis and classified them according to IUCN categories. The clusters were compared with the protected areas classified by the existing IUCN categories and were further subdivided according to the characteristics of the use districts.
Conclusion
This study classified protected areas located in Gangwon Province as well as overlapping protected areas using criteria such as the natural environment, activity restrictions, protected target and area as set in the research method, and derived the characteristics of each cluster. This study has significance in that it verified the protected areas classified as IUCN categories and suggested appropriate management measures for unclear overlapping protected areas.
Introduction
Since the 1992 Rio Earth Summit in Brazil, there has been an emphasis on the importance of biodiversity and ecosystems (Kil et al., 2014). The importance of biodiversity conservation and management has been highlighted internationally, leading countries to come up with guidelines for designating and managing protected areas based on set criteria. However, each country has different standards, institutions in charge, and management guidelines for designating protected areas, which is why there are difficulties in understanding the global situation and suggesting new management measures (Kim and Kang, 2011). The International Union for Conservation of Nature (IUCN) is striving to strengthen cooperation among nations and establish countermeasures. The IUCN has developed protected areas categories to systematically classify and manage protected areas designated by each country, providing international standards and management measures (Heo et al., 2007; Kim and Kang, 2011).
Recently, Europe declared in the EU Biodiversity Strategy for 2030 that it will expand protected areas up to 30% of the EU’s land areas (Mammola et al., 2020). Moreover, it set a binding target for strictly protecting 10% of the EU’s land areas (Cazzolla et al., 2023). While securing areas that must be strictly protected may not be enough for biodiversity conservation, it can help preserve ecological processes in the long run and support the sustainability of high levels of biodiversity (Primm et al., 2018).
The 7th Conference of the Parties to the Convention on Biological Diversity (CBD) emphasized the importance of protected areas for maintaining biodiversity and ecosystem and officially adopted the IUCN protected areas categories as an international standard (Cho et al., 2010). Currently, the IUCN and the UN Environment Programme World Conservation Monitoring Centre (UNEP WCMC) provide data on protected areas through the jointly created Protected Planet website (Kil et al., 2014). These IUCN categories are used as fundamental data in environmental assessment such as the World Economic Forum (WEF) Environmental Sustainability Index (ESI), Environmental Performance Index (EPI), and OECD Environmental Performance Reviews (EPR) (Cho et al., 2010). As such, IUCN categories have become a crucial standard for evaluating environmental management levels among countries. Therefore, it is necessary to apply IUCN categories to improve the management level of protected areas and obtain various national benefits (Heo et al., 2007). There have been various discussions regarding the application of IUCN categories. Dudley et al. (2011) argued that the competency of IUCN categories must be enhanced, built, and developed. Burgin et al. (2014) pointed out that applying these categories may lead to social and economic issues for owners as the existing use of Bakossi National Park in Cameroon is classified under IUCN Category Ia and may cause economic problems.
The 2008 East Asia Report by the IUCN provides the current status and detailed information on protected areas of South Korea, and also mentions the potential for a peace park in the demilitarized zone (DMZ) (McKinnon et al., 2008). Moreover, the analysis using the IUCN’s management effectiveness evaluation in protected areas shows that South Korea excels in vision and planning, comparable to Europe or Australia. However, despite having a high level of vision and planning for protected areas, the current status of these areas is not well known due to a lack of international discussions and exchanges. Up until 2006, most protected areas were national parks that fell under the lower protection levels of IUCN Categories IV (habitat or species management area) and V (protected landscape or seascape) (Kim and Kang, 2011). In particular, classifying all national parks under IUCN Category II is considered an issue. National parks are currently classified into four use districts (park nature conservation districts, park natural environment districts, park village districts, park cultural heritage districts), with different protection scopes, activity restriction criteria, and social and economic activities within protected areas for each district. Accordingly, it is necessary to apply IUCN categories according to each use district. Nevertheless, national parks are currently classified under IUCN Category II as a whole regardless of use districts.
Although South Korea has not sufficiently publicized its outstanding protected areas to the international community, efforts are being made to approach and engage at an international level. The Korea Database on Protected Areas (KDPA), a system similar to Protected Planet, has been in operation since 2017. KDPA is a website (http://kdpa.kr) that provides data based on geographic information on the status and statistics of protected areas (land and marine) in South Korea, providing and managing all kinds of data such as area, occupancy rate, overlapping area, IUCN category levels, designation year, and current status. Despite these efforts, several complex issues arise, including arithmetic errors in the total protected area due to additional overlapping areas, conflicts of interest and activity restrictions between owners and managing agencies due to the management of protected areas by different ministries, and issues related to overlapping protected areas designated where the importance of the ecological environment appears as a common designation purpose of each protected area (Ministry of Environment, 2011).
Previous studies and reports (Korea Research Institute for Human Settlements, 2007; Kil et al., 2014; Kim et al., 2011; Kim and Kang, 2011; Heo et al., 2007; Choe et al., 2018) have highlighted various issues and solutions regarding protected areas in South Korea, but most of them focused on large-scale protected areas such as Baekdudaegan protection areas, national parks, wildlife protection areas, and ecological landscape conservation areas. Furthermore, while they employed the IUCN categories system or other methods, they lacked consideration of overlapping protected areas.
Therefore, this study seeks to categorize and cluster the protected areas overlapping with the existing ones in Gangwon Province based on designation criteria, guidelines, protected targets, and activity restrictions as specified in relevant laws, thereby forming similar clusters. The goal is to apply these clusters to IUCN categories and examine ways to manage overlapping protected areas within the applicable scope.
Research Method
Scope
The subjects of this study are protected areas located in Gangwon Province. Gangwon Province occupies 16.8% (16,826.47 km2) of South Korea’s land area, 81.7% (13,783.68 km2) of which is forest land that has a natural environment with diverse landscapes compared to flatlands or agricultural areas. In addition, 61.17% of the province’s administrative area is classified as Grade 1 in environmental assessment, which is the highest level in the country (https://ecvam.neins.go.kr/contents/rating.do). Gangwon Province is traversed by the Baekdudaegan mountain range and includes four national parks (Seoraksan, Odaesan, Chiaksan, Taebaeksan) along with various types of protected areas such as forest genetic resource protection areas and water source protection areas, encompassing a total of 17 different types of protected areas (Fig. 1). The protected areas examined in this study consist of 17 types of protected areas: landscape protection areas, urban nature park, water source protection area, waterfront area, fishery resource protection areas, water source nourishment protection areas, wetland protection areas, forest genetic resource protection areas, wildlife protection areas, natural monuments, natural reserves, scenic spots, provincial ecological landscape conservation areas, ecological landscape conservation areas, Baekdudaegan protection areas, military parks, and national parks. Various characteristics of protected areas and overlapping protected areas appear accordingly.
Research sites (protected areas in Gangwon Province).
Method
Procedures
In this study, we overlaid spatial data of protected areas in Gangwon Province as shown in Fig. 2, and conducted a cluster analysis using items such as natural environment (Jeon et al., 2008; Kim et al., 2017), activity restrictions (Kil et al., 2014; Ministry of Environment, 2011), and protected target and area (Korea Research Institute for Human Settlements, 2007). Based on the clustering results, we compared the characteristics of each cluster with those outlined in the IUCN categories, and used this comparison to analyze management measures by applying IUCN categories to the clusters.
Research diagram.
Code assigning
To facilitate the classification of each protected area, we assigned unique codes to individual protected areas, with specific digits indicating their status. For example, in Baekdudaegan protection areas, core zones were coded as 10 and buffer zones as 20. In ecological landscape conservation areas, core zones were coded as 100, buffer zones as 200, transition areas as 300, and other areas as 400. Natural reserves were coded as 1000, given they are single protected areas. In this way, codes were assigned according to their differences for each protected area and sub-area. After this coding process, we identified overlapping protected areas using the Field Calculator in ArcGIS.
Variable selection process
In the process of selecting variables, we considered the characteristics of protected areas through literature review (Korea Research Institute for Human Settlements, 2007; Kil et al., 2014; Kim et al., 2011; Kim and Kang, 2011; Heo et al., 2007; Choe et al., 2018) for the classification of single and overlapping protected areas. The items used to evaluate the natural environment included naturalness and cluster stability. There are 7 environmental and ecological items of the Environmental Conservation Value Assessment Map: naturalness, cluster stability, diversity, vulnerability, rareness, abundance, and connectivity. Among these, naturalness and cluster stability using the ecological zoning map and forest type map were selected as items that are considered closely related to the natural environment. Diversity and rareness use the ecological zoning map, and vulnerability, abundance, and connectivity are overlapping with cluster stability and naturalness, and thus were excluded. The assessment criteria proposed by Kim et al. (2017) were applied to assess naturalness and cluster stability.
To select items related to activity restrictions within protected areas, we reviewed data from the Ministry of Environment (2011) and the respective laws for each protected area. We classified the items for activity restrictions with reference to previous studies into 7 items: facility installation, use, development, natural damage, resident support, research (academic), and tourism. Activities within each protected area were divided into three categories: allowed, partially allowed for certain situations, or strictly restricted, as shown in Table 1.
Activity restrictions for each protected area
Classification of protected targets and areas
For classification of protected targets and areas, the characteristics of protected targets were further specified using data from the Ministry of Environment (2011) and the Korea Research Institute for Human Settlements (2007). Table 2 below summarizes the characteristics of protected targets for each protected area. Areas were considered in terms of point, line, and face based on the data from the Korea Research Institute for Human Settlements (2007). Protected areas in the form of faces have unclear protected targets and are large in size, and thus protected areas are wide, thereby including various ecosystems. Protected areas in the form of lines refer to those like water source nourishment protection areas and fishery resource protection areas, which are linear in form and smaller in scale compared to those in the form of faces. Protected areas in the form of points have specific protected targets (such as wildlife protection areas and wetland protection areas), and thus small-scale protected areas were selected.
Protected target by protected area
We used the minimum indicator method for overlapping values when applying variables to protected areas. The minimum indicator method used the maximum function as shown in (Equation 1) when integrating the values of each variable.
In Equation 1, I represents the overall index value, and In represents the n-th overall index value. According to the minimum indicator method, the evaluation of natural environment, activity restrictions, and protected target and area for overlapping protected areas reflects the highest value among the values classified by each variable for the relevant protected areas. This approach allows for relatively simple expression of results when evaluating natural environment, activity restrictions, and protected target and area of overlapping protected areas, while also providing clear and straightforward interpretation (Lee, 2006).
Clustering of protected areas
Cluster analysis was conducted using variables to represent similar groups for each of protected areas and overlapping protected areas. Cluster analysis is a method that groups homogeneous clusters based on the similarity of various attributes that each subject possesses (Park and Lee, 2014). It is used to distinguish homogeneous groups of similar types using selected random criteria (variables) for multiple subjects and to analyze the common characteristics represented by each group.
In this study, we converted categorical variables such as activity restrictions and protected target and area into dummy variables for cluster analysis (Ahn, 2009). Dummy variables are generated for the number of categorical values represented by each variable, assigning ‘1’ to the relevant dummy variable item and ‘0’ otherwise. For example, for protected areas where facility installation is partially allowed, ‘1’ is assigned to facility installation dummy_partially allowed, and ‘0’ is assigned otherwise. Naturalness and cluster stability were used as final variables through standardization of the normal distribution.
Hierarchical cluster analysis was conducted to derive the optimal number of clusters, which were then applied to non-hierarchical cluster analysis to derive classified cluster groups. For the application process of cluster analysis, we referred to the method in Ahn (2009). The first step was selecting variables for cluster division, the second step was calculating the distances between subjects, the third step was performing clustering using the distances between subjects, and the final fourth step was determining the adequate number of clusters.
Application of IUCN categories
IUCN categories are subdivided into 7 items in a total of 6 categories based on the characteristics of protected areas. The 2008 IUCN report provides detailed explanations about the characteristics and criteria for protected areas under each category (Kim and Kang, 2011). The characteristics of the classified clusters were applied to IUCN categories based on the results of hierarchical and non-hierarchical cluster analyses. It was expected that this would improve the ambiguous criteria and the characteristics and management standards of overlapping protected areas. IUCN categories address the characteristics, management standards, and selection criteria of each protected area, and through this information, we referred to the application standard for IUCN categories by Dudley (2008) and Korea Research Institute for Human Settlements (2007) to the clustered protected areas.
Results and Discussions
Results of analyzing overlapping protected areas
As a result of performing initial coding of overlapping protected areas due to protected areas designated in Gangwon Province and analyzing them on GIS, there were a total of 148 significant overlapping protected areas. Among these, there were 79 cases in which 2 protected areas are overlapping. The most frequently overlapping protected areas were forest genetic resource protection areas followed by park nature conservation districts. In particular, forest genetic resource protection areas seemed to be overlapping primarily due to their extensive coverage, while park nature conservation districts were overlapping with the influence of 4 national parks (Seoraksan, Odaesan, Taebaeksan, and Chiaksan).
There were 53 cases in which 3 protected areas are overlapping, with the most frequent overlaps occurring in the core zones and buffer zones of Baekdudaegan protection areas. This frequent overlapping with other protected areas is likely due to the large scale of Baekdudaegan protection areas spanning the length of Gangwon-do from north to south.
There were 13 cases in which 4 protected areas are overlapping, with the most frequently overlapping protected areas being the core zones of Baekdudaegan protection areas, park nature conservation districts, and forest genetic resource protection areas. As mentioned earlier, this is due to the overlapping of protected areas that cover large areas.
There were 2 cases in which 5 protected areas are overlapping. These include the overlapping of landscape protection areas, natural reserves, Baekdudaegan core zones, park nature conservation districts, and water source nourishment protection areas; and the overlapping of landscape protection areas, natural reserves, scenic spots, Baekdudaegan core zones, and park nature conservation districts.
Hierarchical cluster analysis and validation
Hierarchical cluster analysis was conducted on a total of 174 cases, including single protected areas, use districts of protected areas (34 cases), and overlapping protected areas (148 cases). The analysis was effective for 160 (92%) out of 174 protected areas, with 14 (8%) cases being missing data. The results of the hierarchical cluster analysis derived 7 clusters. As a result of analyzing the discriminant analysis results for validation of hierarchical cluster analysis, canonical correlation coefficients ranged from 0.361 to 0.985, indicating high explanatory power for the discriminant functions in classifying clusters. The Wilks’ lambda values ranged from 0 to 0.869, with significance levels of p < .05, revealing that the clusters were well classified by the discriminant functions. Thus, among the 160 protected areas classified as effective through discriminant analysis as well as clusters classified by hierarchical cluster analysis, 11 were identified as incorrectly classified cases. As a result, the effectiveness of the cluster classification outcome from the hierarchical cluster analysis was found to be 93.1%. Therefore, the number of 7 clusters derived from the hierarchical cluster analysis can be considered valid.
Non-hierarchical cluster analysis and validation
As a result of conducting non-hierarchical cluster analysis based on the 7 clusters derived from the hierarchical cluster analysis, it was found that 160 (92%) out of 174 cases including single protected areas, use districts of protected areas, and overlapping protected areas were effective variables, with 14 (8%) cases identified as missing data. The analysis of variance (ANOVA) between clusters indicated that most variables were well classified within a significance level of p < .05, except for “facility installation – allowed” and “protected target – point”.
As a result of conducting discriminant analysis to validate non-hierarchical cluster analysis, canonical correlation coefficients ranged from 0.640 to 0.97 for discriminant functions 1 to 4 (Kim et al., 2006), implying that the explanatory power for these functions is high. However, for discriminant functions 5 and 6, canonical correlation coefficients were 0.174 and 0.310, respectively, indicating low significance. The Wilks’ lambda values ranged from 0.001 to 0.517 for discriminant functions 1 to 4, with significance levels of p < .05, confirming that the clusters were relatively well classified by discriminant functions. However, for discriminant functions 5 and 6, the Wilks’ lambda values were very close to 1, at 0.877 and 0.970, respectively, and their significance levels were also not significant.
The discriminant analysis of the non-hierarchical cluster analysis revealed that 5 out of 160 protected areas classified by non-hierarchical cluster analysis were misclassified. Accordingly, the effectiveness of the cluster classification results by non-hierarchical cluster analysis was 96.9%. We considered methods of applying IUCN categories based on these analysis results.
Application of IUCN categories
Classification of protected areas
We analyzed the characteristics of the clusters classified through the final cluster centroids of non-hierarchical cluster analysis. Naturalness and cluster stability can be interpreted as they can be quantified, but categorical variables cannot be comparatively analyzed at a glance in figures such as means. Thus, cluster analysis was conducted using dummy variables 0 and 1, thereby analyzing the characteristics of the relevant clusters through the composition ratio of the categorical variables found in cluster centroids. The characteristics of each cluster based on cluster centroids are presented in Tables 3 and 4. We applied these characteristics to IUCN categories and finally established the categories of protected areas.
Final cluster centroid
Characteristics by cluster
We classified the clusters to group overlapping protected areas with similar variables among many, which resulted in 7 clusters. After that we identified the characteristics of each cluster (Table 4) through final cluster centroids (Table 3), based on which we derived activity restrictions for the overlapping protected areas within each cluster.
There are conflicts in activity restrictions in overlapping protected areas due to laws that cannot determine the hierarchy among those areas. We established variables that represent the protected areas and conducted cluster analysis on clusters formed by those with similar variable values. Then we derived the characteristics of overlapping protected areas and integrated them with the IUCN categories system, thereby providing management measures that are not influenced by the limitations of existing legal hierarchies and conflicts of activity restrictions.
Naturalness and cluster stability showed higher value when they were closer to 1, and thus higher values indicate negative numbers after undergoing standardization of normal distribution. Since there were diverse variables existing in one cluster centroid for categorical variables such as activity restrictions and protected targets, we displayed the characteristics of clusters while considering the proportions when applying them to IUCN categories.
Cluster 1 showed average values in natural environment among protected areas. For activity restrictions, use, development, and natural damage were restricted, facility installation and ecotourism were partially allowed, and resident support and academic research were fully allowed. Furthermore, regarding protected target and area, a majority of large-scale protected areas were included. As a result of combining these characteristics and applying them to IUCN categories, the areas were classified as Categories III (if unique natural features are included), IV, V, and VI. The common characteristic of Categories IV to VI is that they provide opportunities for the general public to appreciate and learn about the natural environment of protected areas such as preserved natural state and wildlife habitat while also promoting the benefit of residents or revitalization of the regional economy. Among these, protected areas that include unique natural features such as waterfalls, volcanoes, and cliffs could be classified under Category III. The distribution of protected areas in Gangwon Province suggests that they serve as buffers for protected areas with relatively high conservation value.
Cluster 2 was found to have relatively high values for naturalness and cluster stability. For activity restrictions, resident support was 94% restricted and 6% partially allowed, indicating that most protected areas in Cluster 2 restrict resident support. Moreover, academic research is allowed, and all the protected areas included were large scale in terms of protected target and area. As a result of combining these characteristics and applying them to IUCN categories, the characteristics were similar to Categories Ib and II. Category I b is defined as wilderness areas, and Category II as national parks. In particular, this cluster strongly tends to protect and preserve the natural environment rather than focusing on educational and cultural aspects, as they are relatively restricted use districts such as park nature conservation districts of national parks. This cluster also had the highest conservation value among the classified clusters, including overlapping protected areas with highest value in which 5 protected areas are overlapping. Thus, Cluster 2 has the potential to fall under Category Ia as well, which has the highest conservation value, depending on how it is managed in the future. Considering the distribution of Cluster 2 in Gangwon Province, this cluster is located more at the center of protected areas and has higher conservation value than Cluster 1.
Cluster 3 showed relatively low value in natural environment among the classified clusters. For activity restrictions, the proportion of partially allowed activities was the highest. Natural damage showed similar proportions of partially allowed and restricted at 53% and 47%, respectively. Ecotourism was 47% allowed, 40% partially allowed, and 13% restricted, indicating low discriminative power for natural damage and ecotourism. As a result of applying the characteristics of Cluster 3 to IUCN categories, Cluster 3 was classified as Categories V and VI. Categories V and VI are complex protected areas where there are natural environments and social and economic activities are possible within protected areas. Therefore, protected areas in Cluster 3 can consider economic activities of residents and actively utilize natural environments as tourism, educational, and cultural elements without damaging them, thereby achieving sustainable development. In Gangwon Province, protected areas in Cluster 3 are scattered across various locations, mainly in peripheral regions.
Cluster 4 showed relatively low values in natural environment, but like Cluster 2, it included protected areas that strictly restrict activities. Thus, Cluster 4 consists of protected areas with relatively low values in natural environment among those with strict activity restrictions due to high conservation value. When applying IUCN categories, Cluster 4 can be classified as Category II. Categories I a and I b require somewhat high values in natural environment when applied to other clusters or the IUCN criteria. Therefore, Cluster 4 has the potential to fall under Categories I a and I b as well if the values in natural environment increase depending on how it is managed. Considering the distribution in Gangwon Province, these protected areas have high conservation value and are located at the central part of protected areas like Cluster 2.
Cluster 5 showed the highest values in natural environment among the clusters. It had low discriminative power for facility installation and ecotourism, as these were not classified as single activities. Other activity restrictions, such as use, development, and natural damage, were restricted, while resident support was allowed. Thus, Cluster 5 can be considered protected areas with excellent natural environments that allow resident activities. Analysis of Cluster 5 showed that it mostly consisted of overlapping protected areas, such as wildlife protection areas, national parks, and Baekdudaegan protection areas. It exhibited characteristics similar to the park natural environment districts of national parks. Hence, when applying the IUCN categories, Cluster 5 can be classified with characteristics in between Categories II and III, restricting access while allowing productive activities of residents within the protected areas.
Cluster 6 showed the lowest values in natural environment among the clusters, and its activity restrictions were similar to those of Cluster 1. Facility installation and ecotourism were partially allowed, while resident support and academic research were fully allowed. However, unlike Cluster 1, the protected target and area for Cluster 6 had a composition of 67% faces and 33% lines, indicating that it included some medium-scale protected areas. Thus, when applying the IUCN categories, Clusters IV to VI with relatively small protected areas would be suitable for this cluster. Considering the distribution in Gangwon Province, the scale was inadequate compared to other clusters.
Cluster 7 displayed average values in natural environment among the clusters, with activities both partially allowed and allowed in terms of restrictions. For protected target and area, protected areas of various scales were included. Thus, Cluster 7 consists of protected areas with relatively fewer restrictions on activities, allowing economic, cultural, and educational activities. Accordingly, Cluster 7 can be classified as Category VI under IUCN categories, allowing diverse activities as long as they do not harm the value of the protected areas. In Gangwon Province, Cluster 7 was distributed widely across the region (Table 5, Fig 3).
Results of applying IUCN categories by cluster
Classification of protected areas through cluster analysis
Furthermore, we comparatively analyzed the IUCN categories system that classified single protected areas with the IUCN categories system applied in this study (Table 6). National parks were previously classified as Category II without distinguishing use districts, but the new classification provided more detailed results based on each use district. Baekdudaegan protection areas, which were previously categorized as IV, are subdivided into core zones and buffer zones to distinguish activity restrictions. These are classified by dividing IUCN categories for core zones and buffer zones. Ecological landscape conservation areas were also previously classified as Category IV regardless of use districts, but here they are classified by core, buffer, and transition zones. Protected areas without distinction of use districts were similarly classified without significant errors. However, natural reserves, which are classified as Category I a under the KDPA, are classified as VI in this new method. This discrepancy is likely due to the lack of detailed classification of activity restrictions in the Cultural Heritage Protection Act for natural reserves compared to scenic spots and natural monuments. Moreover, wetland protection areas were not classified due to the omission of forest type maps, which made it impossible to determine the cluster stability. Military parks were also not classified, as they were not classified by use district in the current spatial data, thereby not following the activity restrictions of each use district in the Natural Parks Act.
Comparison with existing IUCN categories for single protected areas
Conclusion
The purpose of this study is to establish variables that can represent protected areas and to reclassify the clusters derived through cluster analysis into IUCN categories. Currently in South Korea, protected areas are designated in the same region for similar purposes by different government departments, which resulted in overlapping protected areas, without clear management authorities or guidelines for them, thereby causing conflicts among protected areas as well as various problems. Therefore, this study aimed to classify the protected areas in Gangwon Province as well as the overlapping protected areas using the international standard of IUCN categories and to identify the characteristics and management measures for these protected areas.
More than 80% of the land in Gangwon Province is made up of forests, with a particularly beautiful natural environment compared to other regions; and the province is traversed by the Baekdudaegan mountain range that serves as the backbone of the Korean Peninsula, with four national parks distributed throughout the area. These protected areas were selected for the study due to the large number of protected areas designated by various government ministries. To prevent confusion from overlapping protected areas subdivided by use district with existing spatial data and to facilitate data interpretation, we proceeded with overlapping after coding. We derived variables that are easy to apply to the spatial data of protected areas and that can represent and explain the characteristics based on previous studies and literature review. For variables that represent the overall natural environment of protected areas, we selected naturalness and cluster stability among items of the Environmental Conservation Value Assessment Map. The most frequently occurring issue within overlapping protected areas is conflict of activities among protected areas, which is why we summarized the activities in each protected areas into 7 items based on laws and previous studies related to protected areas. In South Korea, it is rare for protected areas to have clearly designated protection targets, leading to management issues due to ambiguities in protected targets. This is closely related to the size of the protected areas, and thus we set the relationship between protection area and target as a variable. Based on that, we formed clusters of similar protected areas through cluster analysis and classified them according to the IUCN categories system. When comparing these newly formed clusters with the previously classified protected areas based on IUCN categories, the results were found to be either similar or more subdivided according to the characteristics of the use districts.
There is a need for management measures tailored to the characteristics of each protected area for conservation and sustainable use of the classified protected areas. To this end, it is necessary to consider the gap between the high value of these areas as natural resources and the stance of local residents seeking to utilize these areas (Park et al., 2012). Those seeking to utilize protected areas intend to allow more activities within the areas, whereas those seeking to acknowledge and protect the value of these areas intend to restrict many activities. Instead of applying uniform plans and policies based on law to protected areas, management measures that account for these different stances are needed. Therefore, managing protected areas should not be limited to institutions alone by law, but should involve local community organizations comprised of experts, residents, and government (agencies), with management measures to autonomously use and develop protected areas according to their characteristics through localized approach (Park et al., 2012). Efforts from various levels of society are required for sustainable development of protected areas; thus, this study provides management measures using community organizations for clusters derived based on the study results.
This study classified protected areas and overlapping protected areas in Gangwon Province through the criteria set in the research methods such as natural environment, activity restrictions, and protected target and area and provided the characteristics of each cluster. It has significance in verifying protected areas classified by IUCN categories and proposing adequate management measures for unclear guidelines regarding management of overlapping protected areas.
However, this study has limitations, such as restrictions on activity restrictions within protected areas, missing spatial data, and failure to clearly apply the IUCN categories system to the classification of all protected areas and their clusters. Moreover, it is necessary to exclude multicollinearity among items through scientific investigation and analysis and comprehensively consider and analyze representative variables. Currently, South Korea possesses numerous excellent protected areas as they are designated and managed based on legal standards and various data. However, for overlapping protected areas in the data of protected areas provided by IUCN and KDPA, there may be budget overlap in management of protected areas by each government department or ministry. The area of overlapping protected areas may cause errors to the arithmetic statistics of protected area sizes reported to IUCN. There are also issues of activity restrictions. To address these issues, we considered making suggestions using cluster analysis, but it seems further research and analysis must be conducted. Therefore, future research should address the entire network of protected areas and overlapping protected areas in South Korea, and ensure high accuracy in classifying protected areas and applying IUCN categories by diversifying the variables, thereby providing suitable management measures.