Journal of Climate Change Research 2016, Vol. 7, No. 3, pp DOI: * * * ** *, ** Potential Impact

Transcription

1 Journal of Climate Change Research 2016, Vol. 7, No. 3, pp DOI: * * * ** *, ** Potential Impact of Climate Change on Distribution of Hedera rhombea in the Korean Peninsula Park, Seon Uk *, Koo, Kyung Ah *, Seo, Changwan * and Kong Woo-Seok ** * Division of Ecosystem Services & Research Planning, National Institute of Ecology, Korea ** Department of Geography, Kyung Hee University, Korea ABSTRACT We projected the distribution of Hedera rhombea, an evergreen broad-leaved climbing plant, under current climate conditions and predicted its future distributions under global warming. Inaddition, weexplained model uncertainty by employing 9 single Species Distribution model (SDM)s to model the distribution of Hedera rhombea. 9 single SDMs were constructed with 736 presence/absence data and 3 temperature and 3 precipitation data. Uncertainty of each SDM was assessed with TSS (Ture Skill Statistics) and AUC (the Area under the curve) value of ROC (receiver operating characteristic) analyses. To reduce model uncertainty, we combined 9 single SDMs weighted by TSS and resulted in an ensemble forecast, a TSS weighted ensemble. We predicted future distributions of Hedera rhombea under future climate conditions for the period of 2050 ( ), which were estimated with HadGEM2-AO. RF (Random Forest), GBM (Generalized Boosted Model) and TSS weighted ensemble model showed higher prediction accuracies (AUC > 0.95, TSS > 0.80) than other SDMs. Based on the projections of TSS weighted ensemble, potential habitats under current climate conditions showed a discrepancy with actual habitats, especially in the northern distribution limit. The observed northern boundary of Hedera rhombea is Ulsan in the eastern Korean Peninsula, but the projected limit was eastern coast of Gangwon province. Geomorphological conditions and the dispersal limitations mediated by birds, the lack of bird habitats at eastern coast of Gangwon Province, account for such discrepancy. In general, potential habitats of Hedera rhombea expanded under future climate conditions, but the extent of expansions depend on RCP scenarios. Potential Habitat of Hedera rhombea expanded into Jeolla-inland area under RCP 4.5, and into Chungnam and Wonsan under RCP 8.5. Our results would be fundamental information for understanding the potential effects of climate change on the distribution of Hedera rhombea. Key words: Hedera rhombea, Species Distribution Model(SDM), Potential Habitat, Climate Change 1. 서론 산업화이후인간에의한인위적인이산화탄소배출은지구의기후시스템에영향을미쳤으며, 그로인해지난 100년간지구의지표평균온도는약 0.78 상승한것으로관측되었다. 또한미래기후변화시나리오에따라 21세기말에지구표면온도는현재보다 증가할것으로예측되었다 (IPCC, 2014). 한반도의경우, 연평균기온은지난 30년간 1.2 상승 하였고, 미래기후변화시나리오에따라 년의기온상승폭이전지구평균상승경향의 1.2배, 동아시아지역평균상승경향의 1.4배를상회할것으로전망되었다 (National Institute of Meteorological Sciences, 2012). 기후변화는기온의상승, 강수량과강수패턴의변화, 해수면상승과같은지구물리적시스템의변화가동반되며, 이로인해육상과해양생태계의부정적인영향을초래한다. 해양생태계의경우, 해양산성화로인해산호초, 갑각류및어류가위험에직면할것으로 Corresponding author: kyungah.koo@gmail.com Received July 18, 2016 / Revised August 23, 2016 / Accepted September 9, 2016

2 326 박선욱 구경아 서창완 공우석 예측되며, 육상생태계는고온, 가뭄, 해충과질병발생과같은영향으로많은지역에서육상생물의위험이증가할것으로전망되었다 (IPCC, 2014). 기후변화에대한다양한전망이예측되면서, 기후변화가생물의서식지에미치는영향을연구하고, 대응대책을수립하는다양한연구가진행되었다 (Collins et al., 2008; Feeley et al., 2011; Feeley et al., 2012; Díaz et al., 2015; Koo et al., 2001; Kong, 1999; Kong, 2005). 많은연구들이기온과강수의변화에따른식물계절의변화, 난대성식물의서식지확대, 한대성식물의서식지유실과멸종의가능성에예측하였다 (Choi and Jung, 2014; Díaz-Varela et al., 2010; Dullinger et al., 2012; Ju and Bang, 2008; Jung et al., 2015; Kim and Yon, 2010; Koo et al., 2015). 고산과극지역에서식하는식물은기후변화에의해봄철식물계절현상이빨라졌으며, 유럽의산지부근에분포하던난대성식물의서식지도기후변화에따라고도가높은지역까지확대되었다 (Zimmermann et al., 2013). 이와반대로, 유럽의고산식물들은기후변화에따라분포지가쇠퇴하고있는것으로분석되었다 (Randin et al., 2009). 한반도의전체면적은 220,847 km 2 이며, 남북으로길게뻗어있는반도형태로면적의 70% 정도가산지로분류된다. 특히남한의난대성상록활엽수림의경우, 북위 35 이남의해안과연평균기온 14 이상인지역에주로분포한다 (National Institute of Forest Science, 2009). 송악 (Hedera rhombea(miq.) Siebold & Zucc. Ex Bean) 은난대성상록활엽수로덩굴성식물로분류된다. 줄기와잎은상춘등이라고하여류마티스관절염, 간염등의치료제로사용되었고, 열매에는 Rutin, Kizutasaponin K 5, Caffeic acid가함유되어있어과거부터약재로활용되었다 (Kim, 2010). 송악에관한연구는대부분이실험을통한함유물질에관한것과조경에사용되는용도로써식재가능범위와인공광원과광도에따른생육변화에관한내용이대부분이며, 종분포모형을사용한연구는아직이루어지지않았다 (Ju and Bang, 2008; Kim, 2010; Kim, 1997). 송악은한반도에서기후변화에따른생태계의영향을확인하고, 지속적으로관찰하기위해선정한기후변화 100종에포함되어있다 (Ministry of Environment, 2010). 하지만기후변화에대한영향을알아볼수있는생태모델링에관한연구는진행되어있지못한상태이다. 종분포모형은 1950년대허치슨의생태적지위 (Ecological niche) 에기반을두고발전된모형으로종의출현은지역의서식지환경에영향을받는것을전제로하여연구지역의종출현가능성을예측한다 (Guisan and Zimmermann, 2000; Hutchinson, 1957). 이러한종분포모형은생태적과정을기반으로 한과정기반모형 (Process based model) (Kleidon and Mooney, 2000; Morin et al., 2007) 과통계기법을이용한상관모형 (Correlative model) (Guisan and Zimmermann, 2000), 그리고종의분포와환경변수와의역학적작용을바탕으로종의분포를예측하는역학모형 (Mechanistic model) (Kearney and Porter, 2009) 으로분류된다. 각각은종의출현자료와환경변수를입력변수로사용하여종의출현가능성을예측하기때문에기후변화에따른영향평가와같은연구에다양하게활용된다 (Choi et al., 2015; Chun and Lee, 2013; Crossman et al., 2012; Koo et al., 2015; Park et al., 2014; Wang et al., 2016). 특히통계기법을사용하는상관모형은 GLM, GAM과같은회귀모형, RF, GBM과같은기계적학습등다양한알고리즘이개발되어있다. 각알고리즘은각기강조하는점이다르기때문에이들의불확실성을확인하고, 보다폭넓은결과를보여주기위하여앙상블기법이사용된다 (Ahn et al., 2015; Buisson et al., 2010; Elith and Graham, 2009; Kwon, 2014; Thuiller, 2003; Thuiller et al., 2009). 송악은약재로인간의생활에활용되는약용식물이며, 기후변화에따른생태계변화를관찰하기위해환경부에서지정한기후변화 100종에속하는지표종이다 (Ministry of Environment, 2010). 하지만기후변화에따른분포역변화예측과같은기후변화영향에관한연구는아직국내에서미비한실정이다. 본연구의목적은송악의현재분포를통해종분포모형에따른불확실성을측정하고, 미래기후변화에따른분포변화를예측하는것이다. 이를위해 9개의종분포모형알고리즘과앙상블모형을활용하여종분포모형을구축하고, 기후변화에따른분포변화를예측하였다. 본연구의결과는기후변화지표종으로알려진송악에관한기후변화영향을평가함으로써기후변화에따른송악의미래서식지변화예측에대한기초자료를제공할것이다. 이와더불어본연구결과는난대성상록활엽수림전체의분포변화에관한연구에기초자료로활용될수있을것이다. 2. 연구방법및자료 2.1 연구종및연구지역 송악 (Hedera rhombea(miq.) Siebold & Zucc. Ex Bean) 은두릅나무과의난대성상록활엽수로주로남해와서해의도서및해안지역저지대에분포한다. 음성식물로그늘진숲속에서잘자라며, 생활형은덩굴성으로내한성은강하지않지만, 중부해안지역에서월동이가능하다 (Lee, 1980). 전북고창의삼 Journal of Climate Change Research 2016, Vol. 7, No. 3

3 기후변화에따른송악의잠재서식지분포변화예측 327 인리선운사입구의절벽에위치한자생지는내륙북한계지와가까우며, 노거수인점에서천연기념물로보호및관리되고있다. 종분포모델에사용된분포자료에서북한계지는충남태안군의북격렬비도와우배도로나타났다. 송악은 1933년식생조사를통해서발표된난대성상록활엽수북한계선에포함되어있으며, 2009년에식물표본을통해북한계지가북상한것으로확인되었다 (Ministry of Environment, 2010; Uyeki, 1941; Yun et al., 2014). 연구지역은북위 33 38, 동경 에위치한한반도와부속도서로전체면적의 70% 정도가산림으로분류된다. 남북으로길게뻗은한반도는동고서저의지형적특징과함께해안선이단조로운동해안과복잡한서해안과남해안이비교되며, 3,400여개의부속도서대부분이복잡한해안선을가지는서해안과남해안에위치한다. 동일위도에서서해안과동해안의기온은한반도주변해류의영향으로동해안이상대적으로높게나타난다 (National Geographic Information Institute, 2014). 그렇기때문에동해안에난대성상록활엽수의북한계선이위치할것으로생각되나, 대부분의북한계선은서해안도서및해안에서나타난다. 2.2 환경변수들기온과강수는식물의생장에지배적인영향을미치는기후요소들로식물들의생태적지위모형 (Ecological niche model) 연구에서중요한인자들로고려되어왔다 (Guisan and Zimmermann, 2000). 생물기후적변수들 (Bioclimatic Variables, Bioclim) 은생물의분포와성장에영향을주는기후요소들로이루어진 19개의변수로생물의분포와기후요소와의상관관계를찾는연구에서자주사용되었다. 본연구에서는 worldclim 에서제공하는 30-Arc Second(ca. 1 km 2 ) 해상도의 년기후자료를통해생성된 Bioclim 변수를사용하였다. 19개의 Bioclim 변수들은각각상관관계를가지기때문에 Pearson s r correlation 을통해상관계수가 0.7보다높은변수를제외하고, 식물의생장과서식지적합도를설명하는동시에다른변수들을대표할수있는기온변수 Bioclim 1 3과강수변수인 Bioclim 를사용하였다 (Table 1) (Koo et al., 2015). 미래기후변화예측에서는미래에온실가스절감정책이상당부분실현되는 RCP(Representative Concentration Pathways) 4.5 시나리오와현재온실가스배출추세가계속되는 RCP 8.5 를선택하였다. 전지구기후모델 (Global Climate Model; GCM) 은한반도와유사한환경을가지며, 다양한선행연구에서사용된영국헤들리연구소의 HadGEM2-AO(the Fully-coupled atmosphere-ocean version of the Hadley Centre Global Envi- Table 1. Variables for species distribution model Variable Bio1 Bio2 Annual mean temperature Description Mean diurnal range (Mean of monthly (max temp min temp)) Bio3 Isothermality ( 100) Bio12 Bio13 Bio14 Annual precipitation Precipitation of wettest month Precipitation of driest month ronment Model 2) 를사용하여예측된 2050 년 ( 년 ) 기후자료를사용하였다 (Collins et al., 2008; Korea Meteoro- logical Administration, 2012). 2.3 모형개발및검증 종분포모형의개발에필요한송악의출현 / 비출현자료는 Koo(2000), Lee and Yim(2002), 국립수목원의한반도관속식 물분포도 (2004; 2005; 2006; 2007; 2008; 2009; 2010a; 2010b; 2011) 에수록된분포자료를사용하였다. 식물조사지점이동 일하거나인접한지점은통합하여 736 개의식물분포자료를사 용하였다. 각지점에대한조사자료들은조사지역의서식하는 식물들을모두기록한것으로, 송악이나타나지않은조사지점 은비출현자료로활용하였다. 종분포모형의개발을위해서 736 개의분포지점의 bioclim 1 3, 자료와종의출현 / 비출현자료는 70 대 30 의비율 로훈련자료 (training data) 와시험자료 (test data) 로나누었다. 훈련자료는종분포모형의개발에사용되었으며, 시험자료는 종분포모형의검증에사용되었다. 종분포모형의개발과검증 에사용되는분포자료는각각무작위로분할되었으며, 분할과 정에서발생할수있는편향성을최소화하기위해 50 회반복 수행하였다. 종분포모형개발에는 9 개의종분포모형알고리 즘 (Generalized Linear Models(GLM), Generalized Boosted Mo- dels(gbm), Generalized Additive Models(GAM), Classification Tree Analysis(CTA), Artificial Neural Networks(ANN), Surface Range Envelope (SRE), Flexible Discriminant Analysis(FDA), Multivariate Adaptive Regression Splines(MARS), Random Fo- rest(rf)) 을적용하였고, 각알고리즘의종분포모형결과는앙 상블모형을개발하는데사용되었다 (Thuiller et al., 2009). 다 양한알고리즘이앙상블모형개발에사용되고있으며, 그중 가중치를사용한앙상블방법이모형정확도가가장높은것

4 328 박선욱 구경아 서창완 공우석 으로보고되었다 (Marmion et al., 2009). 이에따라, 앙상블모형은 9개의개별종분포모형결과를 True Skill Statistics (TSS) 값으로가중치를주어구현되었으며, 각개별모형과앙상블모형은 R 패키지인 Biomod2 를사용하여구축되었다 (Thuiller et al., 2009). 잠재서식지는종의출현에적합한환경조건을바탕으로예측된종의분포가능지역을의미한다. 연구지역에서종의출현가능성은확률로나타나기때문에, 종의서식지변화예측을위해서종의출현여부를결정짓는임계점설정은매우중요하다. 이를위해모형에의해예측된종의출현이실제관찰과얼마나일치하는지를보여주는 Sensitivity( 출현율예측정확도 ) 와모형에의해예측된종의비출현이실제관찰과의일치를나타내는 specificity( 비출현율예측정확도 ) 에의해산출되는 TSS 값이최대가되는지점을임계점으로설정하였다 (Ruete and Leynaud, 2015). 모형예측정확도는 ROC 분석 (Receiver operating characteristic analysis) 의 AUC(Area under the curve) 값과 TSS 값을사용하였다 (Landis and Koch, 1977; Pearson, 2010). AUC 값은종의출현과비출현자료에대한비율에영향을받지않으며, 출현과비출현에대한정확도가모두포함되어있기때문에종분포모형의검증에많이사용된다 (Lobo et al., 2008). 하지만대상종이광범위하게분포하는경우, AUC 값이낮게나타나며, 국지적으로한정된곳에분포하는경우, AUC 값이높게계산된다 (Lobo et al., 2008). 이러한단점을보안하기위해 TSS 값을모형검증에사용한다. TSS 값은 AUC 값과마찬가지로출현과비출현에대한정확 도가모두포함되어있으며출현과비출현자료의비율과자료분포의면적또는형태에영향을거의받지않아검증결과의신뢰도가높은것으로보고되었다 (Allouche et al., 2006). 3. 연구결과및고찰 3.1 송악의잠재서식지분포예측과모형불확실성 개별종분포모형의불확실성을분석하기위해 9개의개별모형에따른현재기후에서송악의잠재서식지가예측되었다. 교차검증을위한자료분할과정에서발생할수있는불확실성을최소화하기위하여, 분활과정을 50회반복수행하였고, 이를통해각모형에따라 50개의종분포모형이구축되었다. 각모형의불확실성을살펴보기위해 50번씩반복수행된결과의 AUC 값과 TSS 값의평균, 중앙및최소, 최대값들을산출하였다 (Table 2 and Table 3). 50회반복수행한 AUC 값의평균에서 SRE와 CTA는 0.82로상대적으로낮은값을보였으며, 나머지 7개의개별모형들은모두 0.9 이상의높은정확도를보였다. TSS 값의평균에서도 AUC 값과비슷한양상을보였다. SRE와 CTA를제외한 7개의개별모형들이 0.7 이상의높은모형정확도를보였다. 9개의개별모형의 AUC 값과 TSS 값의평균들을종합해보면 SRE와 CTA는다른모형보다낮은모형예측력을보였으며, GBM과 RF는높은예측력을가지는것으로나타났다 (Fig. 1). 송악의종분포모형에서 RF, GBM과같은기계적학습알고리즘이다른회귀, 의사결정나무등의다른알고리즘에비해 Table 2. AUC values for SDMs GLM SRE GBM GAM CTA ANN FDA MARS RF Average Median Maximum Minimum Table 3. TSS values for SDMs GLM SRE GBM GAM CTA ANN FDA MARS RF Average Median Maximum Minimum Journal of Climate Change Research 2016, Vol. 7, No. 3

5 기후변화에따른송악의잠재서식지분포변화예측 329 (a) AUC values for single models 하였다 (Fig. 2). 앙상블모형은 450개의개별종분포모형에서 TSS 값이 0.7보다높은 335개모형을대상으로 TSS 값가중치를적용하여구축되었다 (Fig. 2). 앙상블모형의정확도를나타내는 AUC 값과 TSS 값은 0.96과 0.84로개별모형의정확도와비교해서 RF를제외한다른모형보다우수한것으로나타났다. RF 모형의 TSS 평균값은앙상블모형의 TSS 값보다 0.01 높게나타났다. 그러나 RF 모형은서해안도서지역과영남내륙지역에서현재잠재서식지를부정확하게예측한반면, 앙상블모형은영동지역을제외한서해안과영남내륙지역에서송악분포를잘설명하였다. 송악의잠재서식지는현재분포자료에서출현지점과차이를보였는데, 특히동해안지역에서그차이가두드러지게나타났다. 동해안은서해안에비해상대적으로높은온도와강수조건을보여북위 38.5 까지송악의분포가가능할것으로예측되었지만, 출현지점은울릉도와울산목도로잠재서식지와출현지점과는많은거리적차이를보였다. (b) TSS values for single models Fig. 1. Boxplots for AUC and TSS values. 예측의정확도가높은것으로나타났으며, 다른연구들에서도유사한결과들이발표되었다. 특산식물인히어리 (Corylopsis coreana) 를대상으로종분포모형의불확실성을연구한다른연구에도 RF, GBM, Maxent(Maximum entropy) 와같은기계적학습알고리즘이다른알고리즘에비해높은정확도를보이는것으로나타났다 (Kwon, 2014). 연구종들의분포특성에따라변동이있을수있지만, 한반도또는남한과같은지역적범위에서종분포모형은기계적학습알고리즘이다른알고리즘보다정확도가높을것으로판단된다. 그러나지역적범위가바뀌거나, 분포형태가특수한종의경우각각의상황에맞는알고리즘을선택되어야할것이다. 종분포모형을통한송악의잠재서식지예측에서개별종분포모형의불확실성을최소화하기위해앙상블모형을구축 Fig. 2. Potential habitat of Hedera rhombea in current climate ( ).

6 330 박선욱 구경아 서창완 공우석 앙상블모형에서동해안지역은출현분포자료가없음에도잠재서식지로예측되었다. 이는동해안지역이동해로흐르는난류의영향으로동일위도의내륙과서해안지역에비해높은기온을보이며, 태백산맥과같은지형적인자에의해다른지역보다강수조건이적합하기때문이다. 하지만실제분포자료에서는서해안이동해안보다북한계지가높게나타나는데, 이는동해안지역의지형적으로협소한공간과난대성상록활엽수의수분매개체의역할을하는조류의서식지부재를원인으로볼수있다 (Choi et al., 2010). 3.2 기후변화에따른송악의잠재서식지분포변화 TSS 가중치앙상블모형을사용하여기후변화에따른미래잠재서식지변화를예측하였다. 이를위해미래기후요소예측은영국헤들리연구소에서개발된 HadGEM2-AO 에서 RCP 4.5와 8.5 미래기후시나리오를사용하였다. 예측결과, 송악의미래잠재서식지는모두증가하는것으로나타났다 (Fig. 3). 이러한미래예측은 RCP 시나리오에따라확장폭과형태가다르게예측되었다. 서해안의경우, 모든시나리오에서북한계선이북위 38.5 인황해도해안지역까지북상하였다. 하지만동해안과남부내륙지역은시나리오에따라확장추세가다르게나타났다. 동해안의경우, RCP 8.5는북위 39 인함경남도원산까지, RCP 4.5는강원도고성까지확장하였다. 남부내륙에서 RCP 4.5의잠재서식지는전라도내륙으로확장하는데그쳤지만, RCP 8.5는충청남도내륙까지잠재서식지가증가하는것으로예측되었다. 한반도남부해안에주로분포하던송악의잠재서식지가기후변화에따라상승하는것으로예측된것은기온과강수량의증가에의한것으로판단된다. 기온에서는연평균기온과더불어최한월기온의상승에따른연교차의감소가송악의잠재서식지북상에영향을주었으며, 강수량에서는연강수량의증가가주요한요인으로분석된다. 미래예측에있어서동해안은현재보다북위 까지북상하는것으로예측되었지만, (a) RCP 4.5 (b) RCP 8.5 Fig. 3. Potential habitat of Hedera rhombea in future climate ( ). Journal of Climate Change Research 2016, Vol. 7, No. 3

7 기후변화에따른송악의잠재서식지분포변화예측 331 지형적요인에따른공간적협소함과수분매개체역할을하는철새및조류의서식지부족등생태적요인을고려했을때, 잠재서식지만큼북상하기힘들것으로판단된다. 서해안의경우, 지형적요인과생태적조건이적합하지만, 기후변화에따른해수면상승은난대성상록활엽수의주요서식지인해안지역의감소를초래하여분포지증가에부정적영향을줄것으로예상된다. 그러므로송악의미래분포를좀더정확하게예측하기위해서는수분매개체, 지형적요인, 기후변화에따른해수면상승과같은여러인자들이고려되어야할것이다. 이를위해이러한인자들을설명하는모형들을종분포모형에연결하는연구가진행되어야할것으로판단된다. 이러한한계에도불구하고, 본연구의결과는기후변화에따른송악서식지변화에관한미래연구들의기초자료로활용될수있을것이다. 4. 결론 본연구에서는한반도남부해안지역에주로분포하는송악을대상으로기후변화에따른잠재서식지분포변화를예측하였다. 이를위해, 9개의개별모형을적용한종분포모형들이구축되었고, 이개별모형들의결과는 TSS 가중치를적용하여앙상블하였다. 앙상블모형을적용한결과, 송악은현재기후조건에서서해와남해및동해의해안지역과도서지역에주로분포하는것으로예측되었다. 특히, 동해안의북한계선이서해안보다고위도까지나타나는것으로예측되었는데, 실제분포의북한계지는경남울산으로예측치와차이를보였다. 송악의잠재서식지는기후변화에따라현재보다모두확장하지만, 시나리오에따라확장추세가다르게나타났다. RCP 4.5와 8.5 모두동해안에서 , 서해안에서 북상하였고, 남해내륙으로확장되었다. 동해안의경우, 현재와비교해서지형적요인과생태적조건에의해실제분포지는확장되지못할것으로예측되지만, 서해안의경우다른조건이적합하여잠재서식지가북상할수있을것으로예측된다. 그러나기후변화에따른해수면상승은해안에주로서식하는난대성상록활엽수의서식지에악영향을줄것으로예측된다. 본연구결과는기후변화에따른송악의잠재서식지분포변화를예측함으로써미래송악의서식지변화를이해하는데기초자료로활용될수있을것이다. 5. 사사 본연구는환경부와한국환경산업기술원 ( 기후변화대응환경기술개발사업 ) 의지원으로수행되었습니다. REFERENCES Ahn Y, Lee D, Kim HG, Park C, Kim J, Kim J Estimating Korean pine(pinus koraiensis) habitat distribution considering climate change uncertainty using species distribution models and RCP scenarios (in Korean with English abstract). Journal of the Korea Society of Environmental Restoration Technology 18(3):5164. Allouche O, Tsoar A, Kadmon R Assessing the accuracy of species distribution models: Prevalence, kappa and the true skill statistic (TSS). J Appl Ecol 43(6): Buisson L, Thuiller W, Casajus N, Lek S, Grenouillet G Uncertainty in ensemble forecasting of species distribution. Global Change Biol 16(4): Choi C, Nam H, Chae H Exotic seeds on the feathers of migratory birds on a stopover island in Korea. Journal of Ecology and Environment 33(1):1922. Choi CH, Jung SG Analysis of the MODISbased vegetation phenology using the HANTS algorithm (in Korean with English abstract). Journal of the Korean Association of Geographic Information Studies 17(3):2038. Choi J, Lee S, Lee S Anticipation of the future suitable cultivation areas for Korean pines in Korean peninsula with climate change (in Korean with English abstract). Journal of the Korea Society of Environmental Restoration Technology 18(1): Chun JH, Lee C Assessing the effects of climate change on the geographic distribution of Pinus densiflora in Korea using ecological niche model (in Korean with English abstract). Korean Journal of Agricultural and Forest Meteorology 15(4): Collins W, Bellouin N, Doutriaux Boucher M, 2008, Evaluation of the hadgem2 model. Hadley Cent Tech Note 74, Met Office, Exeter, UK, available gov.uk/publications/hctn/index. html Crossman ND, Bryan BA, Summers DM Identifying priority areas for reducing species vulnerability to climate change. Divers Distrib 18(1):6072. Diaz S, Demissew S, Carabias J, Joly C, Lonsdale M, Ash

8 332 박선욱 구경아 서창완 공우석 N, Larigauderie A, Adhikari JR, Arico S, ldi AB. 2015, The IPBES conceptual framework-connecting nature and people. Current Opinion in Eenvironmental Sustainability 14:116. Diaz Varela RA, Colombo R, Meroni M, CalvoIglesias MS, Buffoni A, Tagliaferri A Spatiotemporal analysis of alpine ecotones: A spatial explicit model targeting altitudinal vegetation shifts. Ecol Model 221(4): Dullinger S, Gattringer A, Thuiller W, Moser D, Zimmermann NE, Guisan A, Willner W, Plutzar C, Leitner M, Mang T Extinction debt of highmountain plants under twentyfirstcentury climate change. Nature Climate Change 2(8): Elith J, Graham CH Do they? How do they? WHY do they differ? On finding reasons for differing performances of species distribution models. Ecography 32(1): Feeley KJ, Malhi Y, Zelazowski P, Silman MR The relative importance of deforestation, precipitation change, and temperature sensitivity in determining the future distributions and diversity of Amazonian plant species. Global Change Biol 18(8): Feeley KJ, Silman MR, Bush MB, Farfan W, Cabrera KG, Malhi Y, Meir P, Revilla NS, Quisiyupanqui MNR, Saatchi S Upslope migration of Andean trees. J Biogeogr 38(4): Guisan A, Zimmermann NE Predictive habitat distribution models in ecology. Ecol Model 135(2): Hutchinson GE Cold spring harbor symposium on quantitative biology. Concluding Remarks 22: IPCC Climate change 2014: Synthesis report. Contribution of Working Groups I, II and III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Core Writing Team, R.K. Pachauri and L.A. Meyer (eds.)]. IPCC, Geneva, Switzerland, pp 151. Ju JH, Bang KJ Effects of artificial light sources and light intensities in subway stations on the growth of Hedera rhombea and Saxifraga stolonifera (in Korean with English abstract). Journal of the Korea Society of Environmental Restoration Technology 11(6): Jung M, Shim K, Kim Y, Choi I Change of climatic growing season in Korea (in Korean with English abst- ract). Korean Journal of Environmental Agriculture 34(3): Kearney M, Porter W Mechanistic niche modelling: Combining physiological and spatial data to predict species ranges. Ecol Lett 12(4): Kim HR, Yon YH Effects of elevated CO 2 concentration and temperature on the response of seed germination, phenology and leaf morphology of Phytolacca insularis (Endemic species) and Phytolacca americana(alien species) (in Korean with English abstract). Korean Journal of Environment and Ecology 24(1): Kim JH A studies on the distribution of Hedera rhombea (Miq.) bean in the Korean peninsula and its biological activities. Dissertation(in Korean with English abstract), Jeju National University. Kim K Studies on components and biological activities of Hedera rhombea Bean. Dissertation(in Korean with English abstract), Ewha Womans University. Kleidon A, Mooney HA A global distribution of biodiversity inferred from climatic constraints: Results from a process based modelling study. Global Change Biol 6 (5): Kong W Selection of velnerable indicator plants by global warming (in Korean with English abstract). Asia Pacific Journal of Atmospheric Sciences 41(2): Kong W The vertical distribution on air temperature and thermal amplitude of alpine plants on Mt. Halla, Cheju Island, Korea (in Korean with English abstract). Journal of Korean Geographical Society 34(4): Koo KA Distribution of evergreen broadleaved trees and climatic factors. Dissertation(in Korean with English abstract), KyungHee University. Koo KA, Kong W, Nibbelink NP, Hopkinson CS, Lee JH Potential effects of climate change on the distribution of cold tolerant evergreen broadleaved woody plants in the Korean peninsula. PloS one 10(8). Koo KA, Park WK, Kong W Dendrochronological analysis of Abies koreana W. at Mt. Halla, Korea: Effects of climate change on the growths (in Korean with English abstract). Korean J Ecol 24(5): Korea Meteorological Administration The Korean peninsula climate change outlook report. Journal of Climate Change Research 2016, Vol. 7, No. 3

9 기후변화에따른송악의잠재서식지분포변화예측 333 Korea National Arboretum Distribution maps of vascular plants of Korean peninsula I. South coast province. Korea National Arboretum, Pocheon. Korea National Arboretum Distribution maps of vascular plants of Korean peninsula II. South province (Jeollado & Jirisan). Korea National Arboretum, Pocheon. Korea National Arboretum Distribution maps of vascular plants of Korean peninsula III. Central & South province (Chungcheong-do). Korea National Arboretum, Pocheon. Korea National Arboretum Distribution maps of vascular plants of Korean peninsula IV. Central & south province (Gyeongsangbuk-do). Korea National Arboretum, Pocheon. Korea National Arboretum Distribution maps of vascular plants of Korean peninsula V. Central province (Geonggi-do). Korea National Arboretum, Pocheon. Korea National Arboretum Distribution maps of vascular plants of Korean peninsula VI. Central province (Gangwon-do). Korea National Arboretum, Pocheon. Korea National Arboretum. 2010a. Distribution maps of vascular plants of Korean peninsula VII. South province (Gyeongsangnam-do) and Ulleung-do province. Korea National Arboretum, Pocheon. Korea National Arboretum. 2010b. Distribution maps of vascular plants of Korean peninsula VIII. Jeju-do province. Korea National Arboretum, Pocheon. Korea National Arboretum Distribution maps of vascular plants of Korean peninsula IX. West & South coast province. Korea National Arboretum, Pocheon. Kwon HS Applying ensemble model for identifying uncertainty in the species distribution models (in Korean with English abstract). Journal of the Korean Society for Geospatial Information System 22(4): Landis JR, Koch GG The measurement of observer agreement for categorical data. Biometrics pp Lee TB Flora of Korea. Hyangmunsa, Seoul. Lee WC, Yim Y Plant geography with special reference to Korea. Kangwon National University Press, Chuncheon. Lobo JM, Nez Valverde AJ, Real R AUC: A misleading measure of the performance of predictive distribution models. Global Ecol Biogeogr 17(2): Marmion M, Parviainen M, Luoto M, Heikkinen RK, Thuiller W Evaluation of consensus methods in predictive species distribution modelling. Divers Distrib 15(1): Ministry of Environment Climatesensitive biological indicator species. Morin X, Augspurger C, Chuine I Processnased modeling of species distributions: What limits temperate tree species range boundaries? Ecology 88(9): National Geographic Information Institute The national atlas of Korea 1st Edition. Human Culture Arirang, Suwon. National Institute of Forest Science The Korea report for sustatinable forest management. National Institute of Meteorological Sciences Global climate change report for response the fifth assessment report of the intergovernmental panel on climate change 2012 predict according to RCP 2.6/4.5/6.0/8.5. Park HC, Lee JH, Lee GG Predicting the suitable habitat of the Pinus pumila under climate change (in Korean with English abstract). Journal of Environmental Impact Assessment 23(5): Pearson RG Species distribution modeling for conservation educators and practitioners. Lessons in Conservation 3: Randin CF, Engler R, Normand S, Zappa M, Zimmermann NE, Pearman PB, Vittoz P, Thuiller W, Guisan A Climate change and plant distribution: Local models predict high elevation persistence. Global Change Biol 15(6): Ruete A, Leynaud GC Goaloriented evaluation of species distribution models accuracy and precision: True skill statistic profile and uncertainty maps. Peer J Pre Prints (3). Thuiller W Biomod-optimizing predictions of species distributions and projecting potential future shifts under global change. Global Change Biol 9(10): Thuiller W, Lafourcade B, Engler R, Ara Jo MB Biomod-a platform for ensemble forecasting of species distributions. Ecography 32(3): Uyeki H Northern distribution limit of Korean evergreen broadleaved trees. Acta Phytotax Geobot 10: Wang T, Wang G, Innes J, Nitschke C, Kang H Climatic niche models and their consensus projections for

10 334 박선욱 구경아 서창완 공우석 future climates for four major forest tree species in the Asia-Pacific region. For Ecol Manage 360: Yun J, Nakao K, Tsuyama I, Higa M, Matsui T, Park C, Lee B, Tanaka N Does future climate change facilitate expansion of evergreen broadleaved tree species in the humandisturbed lasndscape of the Korean peninsula? Implication for monitoring design of the impact assessment. Journal of Forest Research 19(1): Zimmermann NE, Jandl R, Hanewinkel M, Kunstler G, Klling C, Gasparini P, Breznikar A, Meier ES, Normand S, Ulmer U Potential future ranges of tree species in the Alps. in Management Strategies to Adapt Alpine Space Forests to Climate Change Risks, In Tech pp Journal of Climate Change Research 2016, Vol. 7, No. 3