INTRODUCTION
Since long ago, temples in Korea have been located in places with excellent mountains and waters, and most of the famous temples are included in national parks and cultural heritage protection areas blessed with relatively good natural environments (Yi and Yi, 1995). The Tongdosa is also a typical mountain temple and is located in Yeongchuksan (Mt.) (1,081 m), which has an outstanding natural environment, and is included in the Gajisan Provincial Park together with the Gajisan (Mt.) (1,240 m) and Cheonseongsan (Mt.) (812 m) located in the vicinity. Tongdosa was built by the High Buddhist Priest Jajang and it is a Buddhist temple that enshrines the sarira and kasaya of Buddha. It is one of the three main temples of Korea together with the Dharma temple, Haeinsa (Temple) and the Sangha Temple Songgwangsa(Temple), and it is also called Yeongchukchongnim. Yeongchukchongnim Tongdosa was elevated to a chongnim, or comprehensive temple, in 1972 and contains the Buddha jewel, which is the most valued of the Three Jewels, and is thus called buljijongga and plays a central role in the Korean Buddhist culture known as gukjidaechal.
Gajisan Provincial park is divided into the three districts of the Tongdosa, the Seoknamsa and the Cheonseongsan District. Studies on the Gajisan Provincial Park show differences for research frequency and themes by district, and phytosociological studies were carried out in areas such as the Seoknamsa District and Naewonsa District. In the case of the Gajisan District, there was a wide range of phytosociological studies carried out including research on actual vegetation (Ye, 2002), studies on vegetation structures (Kim et al., 1997; Song, 1998), and research on flora (Kim et al., 2010). For studies on the Seoknamsa District, vegetation studies considering the temple forest features for forests surrounding Seoknamsa (Baek et al., 2000) were made. Studies for the Naewonsa District include vascular plant flora and management plans (Oh and You, 2012) and studies on the vegetation structure (Choi et al., 2005).
In the case of the Tongdosa District, the majority include architectural studies such as research on feng-shui from a humanities and sociological perspective based on the historical and cultural value of temples (Oh, 2009), temple arrangement (Ahn, 1981) and spatial structure changes (Hong, 1993; Kim and Lim, 2002). From a landscape architecture perspective, research on artificial vegetation species found near the temple grounds (Kim and Hong, 1993; You et al., 2010) was conducted. However, there is no research on the vegetation structure of the forest area of Tongdosa's most representative temple forest.
Temple forests have religious, public benefit, and economical functions. These are valuable of vegetation science and sociology in forests in Korea, and have very high important position. Furthermore, as an important forest resource of Korea, it contains value not only for its scenic functions, but also for its historical and cultural value. In addition, temple forests are not only the base of Korean national culture, but also a place of passing down traditional heritage to the next generation, and they are areas with very functional public benefits and roles compared to general private forests. However, most temples, excluding temples within national parks, do not make the temple forests public, and therefore, there are no management plans to maintain and manage the temple forests (Kang et al., 2012).
Studies on temple forests include forest structure analysis for the vegetation and landscape management of the Seoknamsa area (Baek et al., 2000), studies on the Bulguksa temple forest within Gyeongju National Park (Choi et al., 2008; Kang et al., 2012) and research data that identified the ecological value of temple forests by identifying the vegetation structure of forests within the temple grounds of seven temples (Shinheungsa, Beobjusa, Naejangsa, Cheoneunsa, Hwaeomsa, Haeinsa, Bulguksa) located within the national parks around the nation (Yi et al., 2011). While there are few systematic vegetation science studies for other temple forests, the management of the unique landscape of temple forests has been rising (Lee et al., 2008), and as of recently, the importance of temple forests has been recognized by the Buddhist religion and considerations are continuously being made for temple forests (Kang et al., 2012). It is thus expected that vegetation science research will become more active in the future for temple forests.
The Tondgosa temple forest, which is located at Yeongchuksan (Mt.), has received attention for its excellent scenery and natural environment, and it is expected that this interest will continue to grow. This study was carried out in order to examine the value and significance of the Tongdosa temple forest and to provide the basis for its preservation. Accordingly, in order to investigate the value of the Gajisan Provincial Park Tongdosa District's natural environment, this study aimed at identifying land usage, actual vegetation and the vegetation structure to construct preliminary data for the management of the temple forest in the future focusing on the Tongdosa, one of the most representative temples in Korea.
MATERIALS AND METHODS
1.Study Area
Gajisan Provincial Park is located on the southeastern part of Korea and was designated a provincial park in November 1979 for the leisure of the people of Korea and to foster a culture of aesthetic sentiment. Including such temples as Tongdosa, Naewonsa and Seoknamsa, Gajisan Provincial Park is divided into the Tongdosa, Naewonsa and Seoknamsa District. The Gajisan Provincial Park Tongdosa District, which is the target of this research, is located at Jisan-ri, Habuk-myeon, Yangsan-si, Gyeongnam, and is located near Yeongchuksan (Mt., 1,081 m), which is a branch of the Yeongnam Alps to the north, it borders the Ulsan Metropolitan Area and Ulju-gun.
The research target is a part of the south province area according to regional divisions based on the flora of the Korean Peninsula, and by temperature zones, it is part of the temperature broadleaf deciduous forest zone when examining the level of vegetation formation (Kim et al., 2010).
2.Investigation and Analysis Method
1)Land Use and Actual Vegetation
To conduct an investigation on land use, it was first divided into forest and non-forest areas based on satellite images, and after digitizing, the non-forest area properties were given on site between February and June, 2013 to draft a land usage map, and an actual vegetation investigation was carried out to examine the characteristics of the forest areas. For an actual vegetation investigation, the same method was used as for land-use and based on satellite images, the forest type map was referred to for categorizing the vegetational physiognomy of the tree stratum dominant species as conifers and broad-leaved trees. Digitizing was then conducted on these to examine the sub-species on site and properties were granted to them.
Based on the field data surveyed, the AutoCAD 2013 Map and ArcView GIS programs were used at indoors to draw up illustrations, and the rate of the area by land usage type and actual vegetation type was calculated against total area.
2)Plant Community Structure
On-site examinations to analyze the plant community structure of the Gajisan Provincial Park Tongdosa District were carried out for three days from July 10 to 12 after completing pre-investigations in February 2013.
In order to identify the features of the plant community structures of the main types based on the actual vegetation map, 58 quadrats were configured in units of 10 m × 10 m(100 m2) citing Ellenberg (1956) and Westhoff and Maarel (1973). The method of Monk et al. (1969) was taken into reference for the plant community structure investigation and divided into the tree, the understory, and shrub stratum (Park, 1985) to measure and record the names of the tree species and their dimensions. The categorization of the tree species was based on Lee's (1980) Korean illustrated plant book and the academic names were written citing the Korean Plant Name Index (KPNI). Trees that make up the upper crown layer were categorized as the tree stratum, trees with a height of less than 2.0m were categorized as the shrub stratum, and all other trees were categorized as the understory stratum. For the tree and understory stratum, the tree height and breast height were investigated, and for the shrub stratum, the breast height and width of the crown (wide × narrow face). Common items that were investigated were altitude, direction, slope, vegetation rate per crown layer, and number of species.
In order to compare the relative dominance among species per crown layer of the investigation area, the method of Lee et al. (1990) that applied Curtis and McIntosh (1951) was used to analyze the importance percentage (I.P.). I.P. was calculated using (relative density + relative coverage / 2, and taking into consideration the size of the trees, (tree stratum I.P × 3 + understory stratum I.P. × 2 + shrub stratum I.P. × 1) / 6 that were given additional value per crown layer was used to find the average mean importance percentage (M.I.P.). As an indirect expression of the tree age and stand dynamics, distribution per breast height class was analyzed to estimate the aspects by the Korea Forest Service (Harcomb and Marks, 1978).
3)Age and Growth Increment Investigation
Trees fitting the average breast height diameter among the dominant species or representative or characteristics trees were selected to investigate the age and growth increments. An increment borer was used at a height of 1.2 m above ground for the selected trees to extract a block of wood, and the extracted block was then analyzed to identify the age and growth increments of the tree.
RESULTS AND DISCUSSION
1.Land Use and Actual Vegetation Map
Upon investigating the actual vegetation in the Gajisan Provincial Park Tongdosa District (Figure 1, Table 1), forest areas accounted for approximately 89.26 % of the entire area, and non-forest areas made up about 10.74 % including farmland, grassland, rock land, and the temple. In the forest area, the vegetation type in which Quercus variabilis was dominant took up the largest area at 32.35 % of the total area. Pinus densiflora forests were distributed centering on Tongdosa and around Banyaam and Jajangam, and areas in which P. densiflora are dominant and compete with Quercus spp. were 12.08 % of the total. In the valley regions that have a relatively lower altitude, there was Carpinus tschonoskii, and in other areas, there was a wide distribution of Q. variabilis and Q. serrata, while in areas with high altitudes, there was a distribution of Q. mongolica. Like the study that suggested Q. mongolica forests as the climax forest (Kim, 1992; Pavel et al., 2006; Kim et al., 2011) that stated that C. laxiflora forests are species that distribute secondarily in areas with relatively low altitude and steep slopes, and it is judged that systematic research that investigates the climax forest of the Tongdosa temple forest will be required.
Based on the actual vegetation investigation results, six main communities were selected that can be said to have representative trees with over a specific area coverage such as Q. variabilis community, C. tschonoskii community, P. densiflora community, P. densiflora-Broadleaf deciduous Forest community, Q. mongolica community, and Broadleaf deciduous Forest community to carry out vegetation structure analysis.
2.Vegetation Structure
1)Study Area Conditions
With reference to the actual vegetation map, the area of investigation was installed around the Geumsuam area which includes brodleaf deciduous trees such as C. tschonoskii and Q. variabilis, the Baekunam area where there are Q. such as Q. mongolica and Q. serrata, and the Banyaam area which includes P. densiflora (Figure 2). Upon examining the altitude, bearings, slope, and crown height and tree coverage per stratum for each investigation area (Table 2), it was found that the altitude range of the area of investigation was between 218 m to 877 m, and the bearing was normally southwest, northwest, and west, with a slope of between 5 and 35°. The number of tree species varied from as little as 4 to as many as 18 and the mean breast height diameter for tree stratum was 16.8-51.3 cm, and 1.3-22.5 cm for the understory stratum.
2)Vegetation Structure
When examining the I.P. analysis per community for the six total communities: Quercus variabilis, Carpinus tschonoskii, Pinus densiflora, P. densiflora-Broadleaf deciduous Forest , Q. mongolica, and Broadleaf deciduous Forest, the results are as follows (Table 3).
For the Q. variabilis community, eight investigation areas were examined. In the tree stratum, Q. variabilis was the dominant species with I.P. 90.8 %, and there were also Q. serrata (I.P.: 5.4 %) and Platycarya strobilacea (I.P.: 1.9 %). In the understory stratum, the I.P. of Q. variabilis was 1.6 %, while the I.P. of Carpinus tschonoskii was 28.8 %, and thus it is expected that in the future, Q. variabilis forests will compete with C. tschonoskii. In the shrub stratum, Lindera erythrocarpa (I.P.: 34.5 %), L. obtusiloba (I.P.: 18.7 %), and Fraxinus sieboldiana (I.P.: 12.2 %) were the major species.
For the C. tschonoskii community, 14 investigation areas were examined. In the tree stratum, C. tschonoskii (I.P. 76.7 %) was the dominant species and it was also the dominant species in the understory stratum (I.P.: 39.5 %). Aside from C. tschonoskii, there were tree species in the tree stratum such as C. laxiflora (I.P.: 12.1 %), Q. serrata (I.P.: 4.8%), and Q. variabilis (I.P.: 2.7 %). In the understory stratum, there were C. laxiflora, Styrax japonicus, and L. erythrocarpa. In the shrub stratum, L. obtusiloba (I.P.: 30.5 %) and L. erythrocarpa (I.P.: 30.4 %) were the major species that were discovered. For the C. tschonoskii community, C. tschonoskii evenly made up the tree stratum and understory stratum, and also appeared in the shrub stratum, showing that it will maintain a stable C. tschonoskii community in the future.
For the P. densiflora community, eight investigation areas were examined. P. densiflora was absolutely dominant in the tree stratum with I.P. 90.4 %, but in the understory and shrub stratum, it had little force. On the other hand, in the understory stratum, C. laxiflora displayed I.P. 55.9 %, thus actively expanding its coverage, and it is therefore judged that there will be severe competition in the future between P. densiflora and C. laxiflora.
For the P. densiflora-Broadleaf deciduous Forest community, sixteen plots were examined. Like the P. densiflora community, the P. densiflora-Broadleaf deciduous Forest community in the tree stratum showed that P. densiflora was dominant with I.P. 53.6 %, but in the understory and shrub stratum, they did not have much force. Currently, in the tree stratum, P. densiflora (I.P.: 53.6 %), Q. serrata (I.P.: 13.2 %), and C. tschonoskii (I.P.: 11.3 %) competed and in the understory stratum, C. laxiflora (I.P.: 20.1 %), C. tschonoskii (I.P.: 18.1 %), and Styrax japonicus (I.P.: 12.0 %) appeared, while in the shrub stratum, Stewartia koreana (I.P.: 20.3 %), L. erythrocarpa (I.P.: 14.2 %), and L. obtusiloba (I.P.: 11.1 %) were discovered.
For the Q. mongolica community, three plots were examined. Q. mongolica was the dominant species in the tree stratum with I.P. 86.9 %. In the understory stratum, Q. mongolica (I.P. 13.4 %) competes with Rhododendron schlippenbachii (I.P.: 28.4 %) and Fraxinus sieboldiana (I.P.: 21.2%). In the shrub stratum, Rh. schlippenbachii (I.P.: 36.5 %) and L. obtusiloba (I.P.: 27.9 %) were the major species.
9 investigation areas were examined for the broadleaf deciduous tree mixed forest. For the broadleaf deciduous tree mixed forest, in the tree stratum Q. variabilis (I.P.: 22.3 %), Q. acutissima (I.P.: 21.8 %), Q. serrata (I.P.: 21.1 %), and C. laxiflora (I.P.: 20.7 %) had similar dominance and made their community. Particularly in the understory stratum, C. laxiflora (I.P.: 57.1 %) was dominant, and in the future, it is expected that it will transform to a community where C. laxiflora is in fact dominant.
In the Tongodsa temple forest, P. densiflora, Q. variabilis, C. tschonoskii, and Q. mongolica are dominant, but they are competing. Also, even if P. densiflora is dominant in the tree stratum, they do not have much force in the understory and shrub stratum, and it is expected that C. tschonoskii and Q. spp. will gain more force. Therefore, management plans according to these findings should be established.
3)Number of Species and Individuals
In order to compare the number of species and number of individual trees per community, the number of mean appearance of individuals and species analysis was carried out per unit area (100 m2) for the 58 investigation areas of the six communities by each investigation area stratum (Table 4) and each community stratum (Table 5).
Upon analyzing the mean number of individuals that appeared per stratum, it was found that 5.66 ± 3.08 appeared in the tree stratum, and 9.64 ± 7.97 in the understory stratum, and the overall number of individuals that appeared per investigation area was 161.81 ± 58.87. Upon analyzing the mean number of species that appeared per stratum, it was found that 2.31 ± 1.19 in the tree stratum and 3.74 ± 2.16 in the understory stratum, and the overall number of species that appeared totalled 11.40 ± 3.21 species. This is similar to the 11.60 ± 4.19 species (Baek et al., 2000) that are found in the nearby Seoknamsa temple forest, and slightly higher than the research results of the Gyeongju National Park Bulguksa temple forest that showed mean species appearance of 10.67 ± 3.96 (Kang et al., 2012).
Upon analyzing the mean number of individuals that appeared per community, in the tree stratum, Q. mongolica community was found have the highest number of individuals at 11.67 ± 3.79, and also in the understory stratum, Q. mongolica community was found to have the highest appearance at 24.33 ± 6.81. The community with the highest overall appearance per unit area was the P. densiflora-Broadleaf deciduous Forest community with 148.38±70.68, and it is evident that the impact of the shrub stratum had a huge effect.
Upon analyzing the mean number of species that appeared per community, in the tree stratum, communities that Quercus variabilis, Carpinus tschonoskii, Pinus densiflora, and Quercus mongolica are dominant did not show much difference, but Pinus densiflora-Broadleaf deciduous Forest communities and Broadleaf deciduous Forest communities that displayed competition among different tree species showed relatively higher number of species compared to other communities.
4)Distribution by Breath Height Diameter
When examining the distribution by breast height diameter (Figure 3, Table 6), in the Q. variabilis community, Q. variabilis DBH distribution range was even at 7-52 cm, and in particular, it was largest between the range of 17 and 42 cm. In the 2-22 cm range, C. tschonoskii and C. laxiflora were the major species, and in the 2-7 cm range, L. obtusiloba were the most widespread. The Q. variabilis community of the Tongdosa temple forest maintained a force of Q. variabilis, but it is expected that there will be competition in the event where the forces of C. tschonoskii and C. laxiflora expand in the future.
In the C. tschonoskii community, C. tschonoskii formed its forces in all ranges from DBH 2 cm to over 52 cm, and L. erythrocarpa and L. obtusiloba were the major species. C. tschonoskii and C. laxiflora were widespread in the 2-17 cm range, and thus, it is judged that in the future, the C. tschonoskii community will maintain its force mainly with C. tschonoskii and C. laxiflora.
In the P. densiflora community, P. densiflora appeared in mid to large diameter trees with a DBH of 12 cm or higher. In particular, 31 P. densiflora 37~52 cm or higher were observed showing the dominance of P. densiflora community at the Tongdosa temple forest. In the DBH of 2-12 cm range, C. tschonoskii, C. laxiflora and Styrax japonicus had high appearance frequency. Kim (2007) stated that P. densiflora community would show a gradual decrease due to vegetation competition with broadleaf deciduous forest. In the case of the Tongdosa temple forest's P. densiflora community, P. densiflora is quite large size in the tree stratum, but is very small DBH in the understory stratum, and therefore, it is expected that it will transform into a broadleaf deciduous forest. Likewise, in the P. densiflora-Broadleaf deciduous Forest community as well, the P. densiflora is prevalent in the tree stratum, but not as so in the understory and shrub stratum. Therefore, it is expected that competition with broadleaf deciduous trees will become fiercer in the future. In conclusion, while P. densiflora community of the Tongdosa is being artificially maintained and managed to fit the characteristics of the temple forest, if natural succession continues without human intervention, it is expected that it will transform into a broadleaf deciduous forest.
In the Q. mongolica community, Q. mongolica is prevalent in the DBH 2-37 cm sector. In the 2-17 cm range, C. laxiflora appeared, and in the 2-12 cm range, Rhodod endron schlippenbachii and Fraxinus sieboldiana appear with relatively high frequency. The Q. mongolica community is expected to be maintained with Q. mongolica as the dominant species.
In the Broadleaf deciduous Forest community in the DBH 2-47 cm range, C. tschonoskii, C. laxiflora, Q. acutissima, Q. variabilis, and Q. serrata appeared. While it has a competitive relationship among species, it is judged that the broadleaf deciduous forest such as C. tschonoskii, C. laxiflora, and Q. spp. will maintain their prevalence (Figure 4).
5)Age and Growth Increment Analysis
In 39 of the 58 total investigation areas, trees that correspond to the mean DBH among dominant species or representative or specific trees within the investigation areas were selected to examine and analyze the age and growth increment. The dominant species and representative tree species were C. tschonoskii, Q. variabilis, P. densiflora, Q. mongolica, Q. serrata, and Q. acutissima. In the age and growth increment analysis results, the oldest tree was P. densiflora of investigation area of 42 in the P. densiflora-Broadleaf deciduous Forest community investigation area, which is approximately 113 years old with an annual mean growth increment of 2.55 mm/year.
Age analysis results showed that the overall forest age of the Tongdosa temple forest is 50 to 100 years. When examining the results of analyzing age and growth increments by tree species, C. tschonoskii had a mean estimated age of approximately 55 years and an annual mean growth increment of 2.65 mm, while Q. variabilis also had mean estimated age of approximately 55 years and an annual mean growth increment of 2.38 mm. The age of P. densiflora was approximately 26 to 113 years, and the mean annual average growth was 2.97 mm. In the case of P. densiflora, the further away from the temple, the less artificial landscaping, and it thus maintained P. densiflora forests in its natural state with trees between the age of 26 to 45 years. It was found that ages were less compared to the P. densiflora community around the temple that had an artificial climax status.
3.Considerations
The Buddhist Temple Tongdosa, which plays a central role in Korean Buddhist culture, received attention mainly from its historical and cultural values, but because the Tongdosa temple forest has, in addition to its value as a temple forest, the functions of having public benefit, high value as a sustainable cultural, scenic and tourist resource, and therefore, this study was carried out with the goal of accumulating data for its preservation and management. This study examined the features of the community structure and major vegetation distribution of the Gajisan Provincial Park Tongdosa District that has a well preserved forest ecosystem focusing on the Tongdosa.
In the Gajisan Provincial Park Tongdosa District, Quercus variabilis community were the most widespread, covering 32.5 % of the total area, and it was found that 89.26 % were forest areas, and 10.74 % non-forest areas. In areas of high altitudes in the Tongdosa temple forest, there was a distribution of Q. mongolica and in areas of relatively low altitudes, there was a distribution of Carpinus tschonoskii, which could be called the climax forest in the southern area in the valley areas for Q. variabilis. This study revealed the vegetation structural features and distribution of Q. mongolica, Q. variabilis, and C. tschonoskii within the Tongdosa temple forest, but it did not investigate plant community distribution and environmental factors. As suggested in the research results, the representative tree species with succession climax in Korea's temperate forest zones are C. laxiflora forests (Ryou et al., 1996; 1991; Park et al., Lee et al., 1996; Choi et al., 1997;) and Q. mongolica forests (Kim, 1992; Pavel et al., 2006; Kim et al., 2011). In order to domonstrate the climax forest of the Tongdosa temple forest in the future, more research on the correlation with environmental factors such as altitude, slope direction and soil must be carried out.
In case of the Tongdosa temple forest's P. densiflora forest, there is a distribution of P. densiflora community around the temple, but further out away from the temple, there is a growing natural expansion of Q. variabilis, Q. serrata, Q. mongolica, or C. tschonoskii, therefore weakening the forces of the P. densiflora community. When examining the cases that maintained disclimax for the management of P. densiflora forests as traditional cultural scenery (Baek and Kim, 1999; Yi and Choi, 2000), it is judged that it would be advisable to consider the history and landscape of the Tongdosa temple forest during its management. Accordingly, detailed investigations focusing on the P. densiflora surrounding the Tongdosa was carried out to differentiate areas with values sufficient enough to maintain the P. densiflora forest, and areas to allow for the natural success of Q. spp. and C. tschonoskii, and it is necessary to establish a suitable management plan. Furthermore, it is judged that research to induce the natural succession of artificial forests such as Larix kaempferim, which is also distributed around the Tongdosa temple forest, will also be necessary.