Stand development patterns of forest cover types in the natural forests of northern Baekdudaegan in South Korea

Stand development patterns of forest cover types in the natural forests of northern Baekdudaegan in South Korea

Ji Hong Kim?Guangze Jin?Sang Hoon Chung

The purpose of this study was to classify current forest cover types,and to investigate stand development patterns for natural forests in six areas in northern Baekdudaegan,South Korea.Twenty-eight independent forest communities were aggregated into eight forest cover types by species composition in the overstory of each forest community.The forest cover types were of mixed mesophytic,‘others’deciduous,Quercus mongolica dominant, Q.mongolica pure,Pinus densiflora–Q.mongolica,P. densiflora,Betula ermanii,and Q.mongolica–P.koraiensis.The ecological information was organized by importance value and species diversity for each forest type. Based on the correlation between species diversity index and the abundance of Q.mongolica plus P.densiflora for corresponding forest cover types,we compared the developmental process and approximate successional pathwaybetween each cover type.The P.densiflora forest cover type changes into the P.densiflora–Q.mongolica cover type,followed by the Q.mongolica dominant cover type through continuous invasion of the oak trees.Furthermore, the Q.mongolica pure cover type would spread toward the Q.mongolica dominant cover type with a mixture of various deciduous tree species.The Q.mongolica dominant cover type progresses through the other deciduous cover types to the mixed mesophytic cover type with diversified composition and structure.On the mid to lower slopes, with loamy soils and good moisture conditions,various deciduous forest types should progress,by ecological succession,toward the mixed mesophytic cover type without any further disturbance.

Baekdudaegan·Forest cover type·Species composition·Species diversity·Stand development patterns

Introduction

Natural forests are composed of a variety of plant species that have evolved by adapting to the climatic and environmental conditions of specific regions,forming a forest community with other biological components such as animals and microbes.Trees in the natural forest will evolve into a group or community to take advantage of certain features such elevation or water source,or they may present pure stands,such as Pinus densiflora in several locations in the Korean peninsula(Kim 2002).

These aggregations and pure stands can be classified into forest cover types based on the species composition of the over-story and these have been widely used by both professional and scientific societies as the standardizeddescription of forests(Kimmins 2004).The definition of forest cover type in this article was adopted from Eyre (1980)as‘a(chǎn) descriptive classification offorestland based on present occupancy of an area by tree species.’It is a category offorestcharacterized by the trees presently occupying the area,which convey no information as to whetherthey are temporary orpermanent.Furthermore,itis a group ofstands with similar composition that develops in the same way in response to given ecologicalfactors(Kimmins 2004).

A certain forest cover type may be either steady or transient in its species composition.Some have occupied the same site for a long period of time,and others have temporarily colonized a disturbed area.Through succession,a given forest type will gradually give way to more stable one and ultimately toward the climax forest.

The cover type approach has long been the major forest vegetation classification system used in the United States. The Society of American Foresters(SAF)first addressed the subject of forest vegetation classification in the early 20th century.SAF emphasized the need for a simple classification system,based on foresttree cover,which was designed to meet the needs of forest management,especially wood utilization(Eyre 1980).

A cover type or dominance is recognized by its overstory trees,which are usually the most economically important species and are emphasized.The dominance approach is based on current tree cover,the way it is presently found on the site.The evaluation and description of contemporary forests is useful to anyone engaged in forest conservation and management practices in different areas.From this point of view,familiarity with forestcover types should be helpful to foresters,land managers,and ecologists.

Patterns of stand development portray forest succession as an orderly progression of ecological changes in an otherwise constant environment.Succession is a natural change that involves the gradual and continuous replacement of one group of tree species by another.In natural forests on the Korean peninsula,the dominant growth form of broad-leaved trees of the deciduous forestexhibita wide variety of morphology and ecophysiology connected with the seral stage(Kim 2002).

The early stages of forest succession are characterized by communities containing a smallnumber of species.The direction of change is toward a more diverse community, and therefore a more complex physiognomy and structure. From this point of view,the analysis of species composition and diversity might help to evaluate the progressive status for forest stand development(Jin and Kim 2006).

The Baekdudaegan is a region that incorporates six central mountain ranges in the Korean peninsula,which starts from Baekdusan,through Geumgangsan and Seoraksan,to Jirisan,extending approximately 1,400 km.For the 680 km portion of the mountain ranges from Hyangrobong to Jirisan in South Korea,the Korea ForestService enacted the Law of Protection ofthe Baekdudaegan in 2003, and established fundamental projection planning in 2006.

The Baekdudaegan contains essentialhabitats for plants, animals,and microbes,and is the original source of the major river systems that form the watersheds of the Korean peninsula.The northern Baekdudaegan in particular,has strong biodiversity and is important for humanistic,social, and ecological reasons for the whole area is about 130,000 ha accounting for 51%of the South Korean Baekdudaegan.It represents the core of the natural ecosystems of the Baekdudaegan.

This study was performed to classify current forest cover types,evaluate ecological characteristics for each classified forest cover type,and compare the developmental and successional process between types for natural forests in the northern Baekdudaegan in South Korea.

Data and methods

Study area

Fig.1 Location of the six study areas in the northern Baekdudaegan

The vegetation data were collected from six mountains representing the northern Baekdudaegan in South Korea (Fig.1).The mountains are more than 1,000 m above sea level(a.s.l.)and have experienced few exogenous disturbances because of government protection.TheHyangrobong(38°20′N,128°19′E;1,296 m a.s.l.)is located in the northernmost region of Baekdudaegan;was designated as the 247th Natural Monument in 1973,access by non-officialcivilians is restricted because of its close proximity to demilitarized zone(DMZ).The area of Odaesan (37°47′N,128°29′E;1,563 m a.s.l.)—including Birobong, Horyeongbong,Sangwangbong,and Noinbong—were designated as a national park in 1975.The Seokbyeongsan (37°35′N,128°53′E;1,055 m a.s.l.)and Deokhangsan (37°17′N,129°00′E;1,071 m a.s.l.)are at a relatively low altitude in the northern Baekdudaegan.The Dutasan (37°25′N,129°00′E)including Cheongoksan,is 1,353 m a.s.l.Neighboring Daedeoksan,Geumdaebong,and Hambaeksan(37°09′N,128°55′E;1,573 m a.s.l.)were appointed as an ecosystem preservation area in 1993(Kim etal.2003).

Collection and analysis of data

The point-quarter sampling method was adopted to collect vegetation data to reduce labor and time while maintaining the information and accuracy obtained from plot sampling (Brower and Zar 1977).On the randomly selected sample point P,the four compass directions were divided into four quadrants.In each quadrant,all woody plants closest from the point were tallied by three vertical strata(over-story, mid-story,and understory);identified by the species level; and measured from this point(diameter,height,and distance).The taxonomy of woody plant species was described by Lee(2003).The distance between points was approximately 30–50 m and the areas of forest plantation and spaces were excluded to eliminate biased vegetation data.A total of 1,004 sample points was temporarily established by topographic position and aspect in each study area.

The collected vegetation data were subjected to cluster analysis to classify similar sample points and to form a group for each study area according to the species composition in the overstory.The sample points of each area were classified by Ward’s method of hierarchical clustering.Ward’s method,or minimum variance clustering,has been widely used by ecologists for grouping purposes (Everitt 1974;Hartigan 1975;Orloci 1967).This method has great intuitive appeal because it is based on the simple underlying principle that at each stage of clustering the variance within clusters is minimized with respect to the variance between clusters(Ludwig and Reynolds 1988). Furthermore,it is effective at minimizing information loss associated with any interactive step in cluster formation (Lattin etal.2003).The IBM SPSS Statistics 20.0 software was employed to implement the analysis.

The classified forest communities in each study area were named according to the ratio of importance value(IV) of the forest overstory in each region(Lee 2013).The IV was obtained by summation the relative density,relative frequency,and relative coverage of the percentage base (Curtis and McIntosh 1951).Species diversity is measured by species richness(the number of species present in a sample of a particular community)and species evenness (the relative abundance of the different species).The Shannon-Wiener diversity index(Brower and Zar 1977) would be a useful index for communities in which two or three species are dominant,such as the natural deciduous forest of Korea(Kim etal.2011).Considering that species composition and diversity are closely associated with stand development,the pathway for each forest cover type was delineated.

Results

Classification of forest cover type

Five distinct forest communities were recognized in Hyangrobong(Fig.2a);five in Odaesan(Fig.2b);four in Seokbyeongsan(Fig.2c);four in Dutasan(Fig.2d);five in Deokhangsan(Fig.2e);and five in Hambaeksan(Fig.2f).

Twenty-eight communities were independently divided by cluster analysis in each study area and were aggregated into eight representative forest cover types based on the over-story species composition.These were the mixed mesophytic cover type,the‘others’deciduous cover type, Quercus mongolica dominant cover type,Q.mongolica pure cover type,Pinus densiflora–Q.mongolica cover type, P.densiflora cover type,Betula ermanii cover type,and Q.mongolica–P.koraiensis cover type(Fig.3).

Ecological information by forest cover types

The ecological information for each classified forest cover type,the species composition by importance values(IV), and species diversity of all trees from three vertical layers and overstory by the Shannon-Wiener index for each cover type were presented in Tables 1 and 2,respectively.The species composition and diversity was characteristically varied by forest cover type.The compositional ratio of species and diversity index of cover types showed peculiar structural attributes and developmental pattern in this forest.

The mixed mesophytic cover type

Fig.2 Dendrograms of forest com munities yielded the minimum variance method(Ward’s method)o f cluster analysis for 204 sampling points in Hyangrobong(a);233 sampling points in Odaes an (b);157 sampling points in Seok byeongsan(c);140 sampling point s in Dutasan(d);110 sampling po ints in Deokhangsan(e);and 160 sampling points in Hambaeksan(f).*Co and**N mean forest community and a number of sampl e point,respectively

Out of 1,004 sample points,125(12.5%)were of the mixed mesophytic forestcover type,where dominance was shared by a number of species.This type was the most diverse of the deciduous forest that lacked prominent dominance.Widespread dominants included Tilia amurensis,Quercus mongolica,Betula ermanii,Acer pictum subsp.mono,and Cornus controversa(Table 1).In addition,another 15–20 species thatachieved dominance in particular stands,presenting the highest species diversity index of 3.541 in three layers,and 3.038 in the overstory among forest cover types(Table 2).Overall,the best indicators of this type were T.amurensis and A.pictum subsp.mono.In this cover type,Q.mongolica,which is the most abundant deciduous species in Korea,had only 11.9%IV(Table 1).

Fig.3 Eightforestcover types were grouped by the similar species composition of the overstory.The Roman numerals are applied to the name of forest cover types in Table 1 and 2

In the narrow valley regions of the Appalachian Mountains in the United States,trees of the same genera— including Acer,Tillia,Betula,Fraxinus,Quercus,Ulmus, and its endemic genera of Fagus,Aesculus,and Liriodendron—show very similar physiognomy and ecological characteristics to those of this cover type(Braun 1950; Vankat 1979;Barbour and Billings 1988).

The‘others’deciduous cover type

Of the eight communities in this cover type,any three of the most abundant species shared less than 50 IV with no obvious dominant species.Therefore,we named this the‘others’deciduous cover type(Fig.3).However,Quercus mongolica maintained dominance with 31.4%of IV,followed by Tilia amurensis,Fraxinus rhynchophylla,and Acer pictum subsp.mono,consisting of 41 canopy species (Table 1).The highestnumber of sample points(309 points out of 1,004)belonged to this cover type,indicating that this type covered extensive mid-slope and valley areas of the study forests.

The remarkable difference of this type from the mixed mesophytic cover type was the difference in the abundance of Q.mongolica and F.rhynchophylla,thereupon the cover type was individually defined.In spite of a larger number of tree species,the diversity was comparatively low,calculated indices of 3.347 of all three layers,and 2.703 of overstory canopy trees(Table 2).This is mainly due to the cover type having lower evenness than that of the mixed mesophytic cover type.

Quercus mongolica dominant cover type

The Q.mongolica dominant cover type ranked next to the‘others’deciduous cover type in terms of the sample area, with 302 sample points out of 1,004.This cover type wasspread over a substantial area,mainly on topographic positions from mid-slopes through upper-slopes onto ridges.Over much of its range,Q.mongolica was probably the most dominant species in natural deciduous forests in Korea.As P.densiflora harvested,there was a fire and therefore oak became the dominant species by persistent sprouting as well as the failure of pine trees to regenerate themselves.

Table 1 IV(%)of overstory by eight forest cover types

Table 2 Species diversity index of eight forest cover types

Q.mongolica was associated with many species(36 species)in the canopy(Table 1),some common to dry upland sites and others more common to moist bottomland sites.The species diversity index was 2.780 in three layers and 1.351 in the overstory.The oak trees were often accompanied by F.rhynchophylla and B.davurica,which were rarely dominant.Commonly favorable sites on northfacing and adjacent valleys in the mountainous part of the study area supported A.pictum subsp.mono and T.amurensis in company with Q.mongolica.

Quercus mongolica pure cover type

Almost 10%(105 sample points)of the study area was solely composed of Q.mongolica trees in the over-story, which was therefore called the Q.mongolica pure cover type.However,we found 50 woody plant species in the mid-story and understory by this sampling procedure.The species diversity index was 3.912 in three layers and zero in the over-story because of the presence of only one species(Table 1,2).

Q.mongolica had some unique characteristics,which permitted itto form 100%pure stands including its ability to sproutreadily and to be competitive.Significantamountsof newly regenerated oak trees in stands were sprouts, either from advance regeneration or from the stumps of cut trees.New sprouts from advance regeneration arose when old stems were damaged during logging or other disturbance.The vertical growth of new sprouts was related to the size of the old damaged stem.The larger the old stem, the faster the new sprout would grow.New sprouts grew rapidly and were usually straight and well formed.

Good seed crops of Q.mongolica are not produced every year and,in the years in which there is a poor yield many of the acorns are ridden with weevils.This variation in acorn production exists not only among isolated stands of oaks,but also among individual trees within stands,as well as annually.When there is a large population of squirrels and other animal populations,a large proportion of the acorn crop may be consumed during the late falland winter.For these and other reasons,seedling reproduction was seldom found to be abundant.

Once established under a forest stand,Q.mongolica seedlings seldom remained true seedlings for more than a few years.Conditions such as fire,poor light,poor moisture conditions,or animal activity killed the tops of seedlings,butnot the roots.One or more dormant buds near the root collar then produced new sprouts.This dieback and resprouting might occur several times.

Pinus densiflora–Quercus mongolica cover type

This forest cover type is a componentof widespread mixed pine-oak forests in which P.densiflora and Q.mongolica and other oak species,dominate the stocking.Other associated species included Betula davurica,Prunus sargentii, Tilia amurensis,Ulmus davidiana var.japonica.This cover type was characterized as a community of low species diversity,and an index of 2.827 in three layers and 1.682 in the over-story was calculated(Table 2).

The structure of the predominant P.densiflora related stands were determined by water deficiency,low fertility, periodic fires,and logging.Owing to their ability to grow on dry,infertile,and coarse-textured soils,and with the aid of surface fire to control more shade tolerant understory hardwoods,P.densiflora were once able to form pure open stands.However,logging and the control of fire had reduced the proportion of pine trees and increased other deciduous tree species,mainly Q.mongolica and other oak species.

This cover type exhibited an uneven-aged condition in which the pine was significantly older than the hardwood component,and was of a relatively narrower age class.The younger and smaller hardwoods were uneven-aged or represented a few very broad age classes.

Even though the combination of P.densiflora and Q. mongolica constituted the majority of density and volume stocking,we named this type P.densiflora forest cover type,distinct to the previous type,due an IV of more than 77%of the pine.In addition to the pine and oak,a small number of other species such as Betula schmidtii,Pinus koraiensis,and Sorbus alnifolia were growing in the canopy layer,presenting the lowest species diversity index of 2.175 in three layers and 0.603 in the over-story among forest cover types(Table 1,2).

P.densiflora is intolerantof shade and commonly grows in even-aged stands.This species is often succeeded by the more shade-tolerant associated species of deciduous trees, such as Q.mongolica and other hardwoods,which regenerate more readily in the understory.Over time,pine trees gradually decrease in abundance and finally disappear. Other trees increase in density and stocking through natural invasion of common associates such as oaks,maples, ashes,and birches.

This cover type occurs most commonly on dry sandy sites,but shows the best growth on well-drained sandy loam-to loam soils.However,on better sites,the rate of succession to other foresttypes is much more rapid than on poorer sites.On the best sites,the pine trees may be replaced by deciduous tree species or P.koraiensis(Lee 2002),but on poor sites the succession may halt and P.densiflora,if undisturbed,will remain as a persistent subclimax pure stand.Wild fires may disrupt ecological succession by killing all but the thick-barked trees that survive to provide seed for a new stand that may be either pure or P.densiflora with only a small proportion of intolerant species.

Betula ermanii cover type

B.ermanii cover type was characteristically distributed in ridge or peak areas,and at an average 1,251 m of relatively high altitude in limited areas of Odaesan and Hambaeksan.B.ermanii and Q.mongolica predominated, and the birch was usually present in greater numbers and the key species in this forest type.Associated species showed low variability,and included Phellodendron amurense,Abies holophylla,A.nephrolepis,and Cornus controversa(Table 1).The species diversity index was calculated as 2.835 in three layers and 1.116 in the overstory(Table 2).

B.ermanii is thought to be a pioneer species and has adapted to regenerating areas after severe fires or soil disturbance.We observed that this species often preceded its associates by invading disturbed soils that had been logged or cleared by fire.After reaching the mature stand and as forest succession would proceed without additional major disturbances,the density and stocking of birch trees decreased whilst those of other coniferous and deciduous species increased.Eventually,B.ermanii is eliminatedsooner than other species found in this type and this cover type cannot be maintained.

Quercus mongolica–Pinus koraiensis cover type

The dominant species in this forest cover type were by Q. mongolica(54.9%of IV)and P.koraiensis(35.8%of IV);therefore,itwas named as Q.mongolica–P.koraiensis forest cover type(Table 1).This cover type occurred only in the Hyangrobong area with 21 sample points of upper mid-slopes and ridges.It is likely that P.koraiensis grows naturally over extensive areas,which have seldom been discovered in South Korea(Hwang et al.2012).The associated tree species included Tilia amurensis,Abies holophylla,and Kalopanax septemlobus.The species diversity index of this type was rather low,and was calculated at 2.254 in three layers and 1.026 in the over-story (Table 2).

Q.mongolica and P.koraiensis exhibit intermediate shade tolerance.In addition,they are exceedingly site compatible.Seedlings of both species are slow growing but can survive for a number of years under partial sunlight (Lee 2002).However,they grow well under full sunlight, which may enter the forest following disturbances of the canopy.Oak trees sprout vigorously,which gives them an advantage when competing with the regenerating pine. However,when the cover type develops as an even-aged stand,the P.koraiensis has the possibility of becoming the dominant species.

Discussion

The core of forest development has undergone an overall change of the structure and function of forests.Pickettetal. (1987)defined the term‘forest successional pathway’as the temporal pattern of vegetation change.This typically shows change in community types over time,and may describe the decrease or increase of particular species populations.Species composition and diversity change according to foresttype,and this depends on environmental conditions as well as the physiological predispositions of the interacting species.Accordingly,one forest type may readily transform into another following changes in the environment.

Species diversity increases from simple communities during early succession to richer communities in later succession or in the mature community(Odum 1969; Whittaker 1975).The stability of the climax community is related to its species diversity,and the aging phenomenon or succession in an ecosystem is best described as an evolution toward high diversity—a large number of ecological niches and,consequently,a large number of species(Kormondy 1976;Barbour et al.1987).

Correlation analysis showed that the species diversity index was negatively associated with IVs of Q.mongolica plus P.densiflora for corresponding cover types in this study(Fig.4).This shows that as the density and stocking of Q.mongolica plus P.densiflora decreased and the abundance other species such as Acer pictum subsp.mono, Betula ermanii,Cornus controversa,Fraxinus mandshurica,Fraxinus rhynchophylla,and Tilia amurensis,increased (Table 1),the species diversity increased accordingly. Therefore,we suspectthatthe mixed mesophyic cover type was the late successional stage in the study forests,followed by the‘others’deciduous cover type,Q.mongolica dominant cover type,P.densiflora–Q.mongolica cover type,and P.densiflora cover type.

Previous studies of forestcommunities predicted thatthe abundance of P.densiflora would decrease;if not,the abundance of Q.mongolica would remain constant or increase the composition ratio during ecologicalsuccession in naturalforests(Kim etal.1990;Oh etal.1998;Byun etal. 1998;Cho and Choi2002;Kim and Choo 2003;Choiand Oh 2003;Oh et al.2005;Choo and Kim 2005;Kim 2010;An and Choo 2010).Some studies reported that Q.mongolica was a late successionalspecies,a climax species(Kang and Oh 1982),or a climate-specific climax community in ridge (Jang and Yim 1985).Other studies reported that Quercus species and P.densiflora forests developed from secondary forest and was estimated as a sub-climax forest(Yun et al. 1987;Kim 1992;Kim et al.1993).

Fig.4 The correlation between species diversity index and abundance of Q.mongolica plus P.densiflora for corresponding forest cover types

Nevertheless,it is well documented that once opened, upland,dry,sandy,and textured sites usually come to be dominated by pure stands of P.densiflora,with allthe trees the same age,forming P.densiflora cover type.However,pines reproduce poorly in their own shade due to a requirement for relatively high light intensity,or because they are poor competitors for soil moisture.Thereafter, pine stands that are several decades old have a welldeveloped understory of hardwoods.

Out of the many hardwood species that are future overstory trees accompanied with the pine,Q.mongolica plays a leading role in the invasion of mature pure pine stands, finally able to replace the pines(Lee et al.2010).This process presumably led to the growth of the P.densiflora–Q.mongolica cover type.Under these conditions,the frequency of P.densiflora occurrence in the overstory decreases steadily because of their greater age and thus shorter life span,and the balance shifts more and more towards hardwood over time.Eventually a classification other than P.densiflora–Q.mongolica forest cover type must be applied,forming a Q.mongolica dominant cover type where few scattered pine trees barely stay in existence (Table 1;Fig.5).

Q.mongolica is an outstanding tree and is widespread across the Korean peninsula.The oak has the ability to grow on all upland aspects and slope positions within its range except extremely dry,shallow-soil ridges,poorly drained flats,and wet bottomland.Although Q.mongolica was once a component of mixed,uneven-aged stands,most current species are in pure to mixed second growth stands of sprout origin(Kim et al.2001).

Q.mongolica is generally classified as having intermediate tolerance to shade.Young trees are the most tolerant but become less tolerant as they grow larger.Q.mongolica seedlings,saplings,and even pole-size trees are nevertheless able to persist under a forest canopy for several decades.Q.mongolica usually becomes dominant in the stand because of its ability to persist for long periods of time in the understory,its ability to respond well after release,and its great longevity,forming pure or absolute dominant stands.The successful regeneration of Q.mongolica typically depends on their existence as saplings in the understory.Although there are many factors associated with oak regeneration,fire is important for successful regeneration for two reasons—fire kills competitive trees and it promotes resprouting of young Q.mongolica(Kim et al. 2001).

Fig.5 Approximate successional pathway through seres of forest cover types

Having intermediate tolerance to shade,Q.mongolica might be at a competitive disadvantage against early successional species in the open understory and against late successional species in the shaded understory.Likewise, mesophytic species regularly invade the understories of Q. mongolica forests via wind and animal dispersal of seeds. The understory-tolerant mesophytic species replacing Q. mongolica in the understory on all but the driest sites are primarily A.pictum subsp.mono,Cornus controversa,and T.amurensis.However,where the oak advance regeneration fails,and logging or wind disturbance of Q.mongolica over-story provide sufficient light,opportunistic species, such as Prunus spp.,Betula spp.,Fraxinus spp.,and Kalopanax septemlobus can become established and develop in the absence of fire(Kim 1992).

Based on observations from species diversity and the replacementmechanism of tree species,we believe thatthe P.densiflora cover type changes into the P.densiflora–Q. mongolica cover type following the invasion of the dominant oak species Q.mongolica cover type,.Thus,the Q. mongolica pure cover type would advance toward the Q. mongolica dominant cover type with a mixture of deciduous tree species.The Q.mongolica dominant cover type would progress through the others deciduous cover type to the mixed mesophytic cover type,having a diversified composition and structure(Fig.5).

We do not consider the present mixed mesophytic cover type being the climax forestin the Korean peninsula.Based upon the species composition,species diversity,and stand structure of DBH and height distribution(Hwang et al. 2012;Lee 2013),however,we recognize that the cover type had been moderately disturbed in the pastby humans, fires,and winds,and it had had reverted so that it reached stability and ceased changing so rapidly.On a site ofmid to lower slopes that have been undisturbed,with loamy soils and good moisture conditions,various deciduous forest types should make progress by succession toward the mixed mesophytic forests,which might be representative of the climax forest in Korea.

An HC,Choo GC(2010)Vegetation structure ofthe Woongseokbong in the Jirisan(Mt.).Korea J Environ Ecol 24:547–555

Barbour MG,Billings WD(eds).(1988).North American terrestrial vegetation.Cambridge University Press,New York,p 434

Barbour MG,Burk JH,Pitts WD(1987)Terrestrialplantecology,2nd edn.The Benjamin/Cummings Publishing Company Inc.,Menlo Park,p 634

Braun EL(1950)Deciduous forest of eastern North America. Blakiston,Philadelphia,p 596

Brower JE,Zar JH(1977)Field and laboratory methods for general ecology.WM.C.Brown Company Publishers,Dubuqu,p 136

Byun DW,Lee HJ,Kim CH(1998)Vegetation pattern and successional sere in the forest of Mt.Odae.J Ecol Field Biol 21:283–290

Cho HS,Choi SH(2002)Plantcommunity structure ofthe Baekcheon valley in Taebaeksan area,the Baekdudaegan.Korea J Environ Ecol 15:368–377

Choi SH,Oh KK(2003)Vegetation structure of Mountain Ridge from Jeongryeongchi to Bokseongijae in the Baekdudaegan. Korea J Environ Ecol 16:421–432

Choo GC,Kim GT(2005)Vegetation structure of mountain ridge from Bubong to Poamsan in Baekdudaegan,Korea.Korea J Environ Ecol 19:83–89

Curtis JT,McIntosh RP(1951)An upland forest continuum in the prairie-forest border region of Wisconsin.Ecology 32:476–496

EverittB(1974)Clusteranalysis.John Wiley&Sons Inc,New York, p 122

Eyre FH(ed)(1980)Forest cover types of the United States and Canada.Society of American Foresters,Washington DC,p 148

Hartigan JA(1975)Clustering algorithms.John Wiley&Sons Inc., New York,p 351

Hwang KM,Lee JM,Kim JH(2012)Community classification and successional trends in the natural forest of Baekdudaegan in Gangwon Province.J Agric Life Sci 46:41–55

Jang YS,Yim YJ(1985)Vegetation types and their structures of the Piagol,Mt.Chiri.J Plant Biol 28:165–175

Jin GZ,Kim JH(2006)The estimation of succession index by community types in the naturaldeciduous forestof Mt.Jumbong. J Korean For Soc 95:723–728

Kang YS,Oh KC(1982)An application of ordinations to Kwangnung Forest.J Plant Biol 25:83–99

Kim JH(1992)Analysis of successional trend by transition matrix model in the mixed broadleaved-Abies forest of Mt.Odae. J Korean For Soc 82:325–336

Kim JH(2002)Community ecological view of the natural deciduous forest in Korea.Ecology of Korea.Bumwoo Publishing Company,Seoul,pp 93–104

Kim HD(2010).Studies on community dynamics of vegetation and successive dynamics in Bukhansan National park.Ph.D.dissertation,Changwon national university,Korea,p 161

Kim GT,Choo GC(2003)Vegetation structure of mountain ridge from Nogodan to Goribong in Baekdudaegan.Korea J Environ Ecol 16:441–448

Kim JS,Kim GT,Choo GC(1990)The actual vegetation in Mt. Sokri.Korea J Environ Ecol 4:1–15

Kim JH,Lee DK,Kim ZS,Lee KJ,Hyun JO,Hwang JW,Kwon KW (1993)The interpretation for stand structure in natural oak forests.J Korean For Soc 82:235–245

Kim JH,Yang HM,Jin GZ,Lee WS,Kang SK(2001)The aspectof natural regeneration for major tree species in the natural deciduous forest.J For Environ Sci 17:1–17

Kim GT,Choo GC,Baek GJ(2003)Structure offorestcommunity at Daedeoksan-Geumdaebong nature Ecosystem Preservation Area in Baekdudaegan.Korea J Environ Ecol 17:9–17

Kim JH,Lee HS,Hwang GM(2011)Ecological attributes of species composition by topographical position in the natural deciduous forest.J For Environ Sci 27:17–22

Kimmins JP(2004)Forestecology:a foundation forsustainable forest management and environmental ethics in forestry,3rd edn. Prentice Hall,New Jersey,p 611

Kormondy EJ(1976)Concept of ecology,2nd edn.Prentice-Hall, Inc.,New Jersey,p 238

Lattin J,Carroll JD,Green PE(2003)Analyzing multivariate data. Brooks/Cole-Thomson Learning,Pacific Grove,p 580

Lee WS(2002)Abundance and growth ofnaturally regenerated Pinus koraiensis wildings in four different forest types.Ph.D Dissertation.Department of Forest Management,Kangwon National University,Korea,p 140

Lee TB(2003)Coloured flora of Korea.Hyangmoonsa,Seoul,p 232 1928 pp

Lee JM(2013)The Classification of forest communities and the evaluation of species diversity in natural forests of six Mts.in Gangwon-do Baekdudaegan.MS Thesis.Department of Forest Management,College of Forest&Environmental Sciences, Kangwon National University,Korea,p 85

Lee DK,Kwon KW,Kim JH,Kim GT(2010)Silviculture. Hyangmoonsa,Seoul,p 334

Ludwig JA,Reynolds JF(1988)Statistical ecology.John Wiley& Sons Inc.,New York,pp 125–144

Odum EP(1969)The strategy of ecosystem development.Science 164:262–270

Oh SH,Yun CW,Bae KH,Hong SC(1998)A study on forest vegetation in Mt.Cheongok,Kyungsangpuk-do.J Korean For Soc 87:27–39

Oh KK,Choi SH,Kim SH,Choi WK(2005)Analysis of the forest community structure in the Woraksan National Park.Korea J Environ Ecol 19:90–98

Orloci L(1967)An agglomerative methods for classification of plant communities.J Ecol 55:193–206

Pickett STA,Collins SL,Armesto JJ(1987)Models,mechanism,and pathways of succession.Bot Rev 53:335–371

Vankat JL(1979)The Natural vegetation of North America.John Wiley&Sons,New York,p 261

Whittaker RH(1975)Communities and ecosystem,2nd edn.Macmillan,New York,p 385

Yun JH,Kim JH,Han SS(1987)Structure and environmental conditions in a Virgin forest.Res Bull Exp For 7:3–26

22 May 2014/Accepted:10 October 2014/Published online:29 January 2015

?Northeast Forestry University and Springer-Verlag Berlin Heidelberg 2015

Project funding:This study was conducted with the supportof‘Forest Science&Technology Projects(Project No.S211012L030110)’provided by Korea Forest Service.

The online version is available at http://www.springerlink.com

Corresponding editor:Hu Yanbo

J.H.Kim

Department of Forest Management,College of Forest and Environmental Sciences,Kangwon National University, Chuncheon 200701,Korea

G.Jin

School of Forestry,Northeast Forestry University, Harbin 150040,China

S.H.Chung(?)

Forest Ecology Division,Forest Conservation Department, Korea Forest Research Institute,Seoul 130712,Korea e-mail:chsh@kangwon.ac.kr