drivers of treeline shift in different european mountains
TRANSCRIPT
CLIMATE RESEARCHClim Res
Vol 73 135ndash150 2017httpsdoiorg103354cr01465
Published August 21
copy The authors 2017 Open Access under Creative Commons byAttribution Licence Use distribution and reproduction are un -restricted Authors and original publication must be credited
Publisher Inter-Research middot wwwint-rescom
Corresponding author cudlinpczechglobecz
Drivers of treeline shift in different European mountains
Pavel Cudliacuten1 Matija Klopltilt2 Roberto Tognetti34 Frantisek Maacutelis56 Concepcioacuten L Alados7 Peter Bebi8 Karsten Grunewald9 Miglena Zhiyanski10
Vlatko Andonowski11 Nicola La Porta12 Svetla Bratanova-Doncheva13 Eli Kachaunova13 Magda Edwards-Jonaacutešovaacute1 Josep Maria Ninot14 Andreas Rigling15
Annika Hofgaard16 Tomaacuteš Hlaacutesny17 Petr Skalaacutek118 Frans Emil Wielgolaski19
1Global Change Research Institute CAS Academy of Sciences of the Czech Republic Ceskeacute Budejovice 370 05 Czech Republic 2University of Ljubljana Biotechnical Faculty Department of Forestry and Renewable Forest Resources Slovenia
3Dipartimento di Bioscienze e Territorio Iniversitagrave degli Studio del Molise Contrada Fonte Lappone 86090 Pesche Italy 4MOUNTFOR Project Centre European Forest Institute 38010 San Michele all Adige (Trento) Italy
5Technical University Zvolen Faculty of Forestry 960 53 Zvolen Slovakia 6National Forest Centre Forest Research Institute Zvolen 960 92 Zvolen Slovakia
7Pyrenean Institute of Ecology (CSIC) Apdo 13034 50080 Zaragoza Spain 8WSL Institute for Snow and Avalanche Research SLF 7260 Davos Dorf Switzerland
9Leibniz Institute of Ecological Urban and Regional Development 01217 Dresden Germany10Forest Research Institute BAS 132 Kl Ohridski Blvd 1756 Sofia Bulgaria
11Faculty of Forestry University Ss Cyril and Methodius Skopje Macedonia 12Research and Innovation Centre Fondazione Edmund Mach (FEM) and MOUNTFOR Project Centre
European Forest Institute 38010 San Michele all Adige (Trento) Italy 13Division of Ecosystem Research IBER-Bulgarian Academy of Sciences 1113 Sofia Bulgaria
14University of Barcelona Department of Plant Biology 08028 Barcelona Spain 15WSL Swiss Federal Institute for Forest Snow and Landscape Research Zuumlrcherstrasse 111 8903 Birmensdorf Switzerland
16Norwegian Institute of Nature Research 7485 Trondheim Norway 17Faculty of Forestry and Wood Sciences Czech University of Life Sciences 165000 Prague Czech Republic
18Czech Hydrometeorological Institute 143 06 Prague Czech Republic19University of Oslo 0316 Oslo Norway
ABSTRACT A growing body of evidence suggests that processes of upward treeline expansionand shifts in vegetation zones may occur in response to climate change However such shifts canbe limited by a variety of non-climatic factors such as nutrient availability soil conditions land-scape fragmentation and some species-specific traits Many changes in species distributions havebeen observed although no evidence of complete community replacement has been registeredyet Climatic signals are often confounded with the effects of human activity for example forestencroachment at the treeline owing to the coupled effect of climate change and highland pastureabandonment Data on the treeline ecotone barriers to the expected treeline or dominant treespecies shifts due to climate and land use change and their possible impacts on biodiversity in 11mountain areas of interest from Italy to Norway and from Spain to Bulgaria are reported Weinvestigated the role of environmental conditions on treeline ecotone features with a focus on tree-line shift The results showed that treeline altitude and the altitudinal width of the treeline eco-tone as well as the significance of climatic and soil parameters as barriers against tree speciesshift significantly decreased with increasing latitude However the largest part of the commonlyobserved variability in mountain vegetation near the treeline in Europe seems to be caused bygeomorphological geological pedological and microclimatic variability in combination with dif-ferent land use history and present socio-economic relations
KEY WORDS Vegetation zone shift middot Climate change middot Climate models middot Treeline ecotone middot European mountains middot Ecosystem services
OPENPEN ACCESSCCESS
Contribution to CR Special 34 lsquoSENSFOR Resilience in SENSitive mountain FORest ecosystemsunder environmental changersquo
Clim Res 73 135ndash150 2017
1 INTRODUCTION
Mountain regions are crucial areas for studying theimpact of climate change on vegetation communitiessteep climatic gradients enable testing of ecologicalecophysiological and evolutionary responses of florato changing geophysical influences related to climatechange In addition most of the species living theregrow in conditions classified as their ecological limits(Koumlrner 2012) Vegetation in European mountainousregions has been subjected to severe changes duringthe remote (eg Kullman 1988) and recent pasts (egVrška et al 2009 Boncina 2011 Bodin et al 2013Elkin et al 2013 Schwoumlrer et al 2014) Changes inspecies distribution and plant community composi-tion have been the most often observed althoughevidence of complete community exchange has notyet been registered
A strong elevational zonation is typical of montanevegetation caused mainly by steep climatic gradi-ents Vegetation zones are characterized by a spe-cific structural-functional type of phytocoenosis con-structed by particular main plant species (edificators)and roughly following latitude (latitudinal vegetationzones) or altitude (altitudinal vegetation zones orvegetation belts) Climatic zones along altitudinalgradients are compressed with large habitat andspecies diversity in successive altitude vegetationzones In these steep gradients species richnessdecreases with increasing altitude although topo-graphic isolation results in increased levels of en -demism (Pedrotti 2013)
The most obvious vegetation boundary at high ele-vations is that of the upper forest limit (Harsch ampBader 2011) Owing to its diffuse character it mightbe better to refer to the treeline ecotone which consists of the belt between the boundary of theclosed forest standminustimberline and the uppermost ornorthern most scattered and stunted individuals ofthe forest-forming tree species regardless of theirgrowth form and height (Holtmeier amp Broll 2005)Within this ecotone the treeline (ie the line connect-ing the tallest patches of forest composed of trees ofge3 m height) is often delimited at this elevation themean temperatures for the growing season are ca64degC (Koumlrner 2012) Climate is one of the mostimportant limiting factors defining the spatial distri-bution of any species (Pearson amp Dawson 2003Wieser et al 2014) Temperature especially summermean temperature and temperature sums is the pri-mary factor causing the formation of treelines at theglobal scale (Grace et al 2002 Moen et al 2004)although it may be substantially affected by several
other non-climatic factors eg mass elevation effect(Ellenberg 1988) past land use or forest management(Gehrig-Fasel et al 2007) The main factors affectingthe expansion andor retreat of tree stands at theirupper limits are scale-dependent (Holtmeier 2009)At the global scale treeline position is determined bygrowing-season temperatures (Koumlrner amp Paulsen2004) whereas at the landscape scale second-orderfactors (ie climatic stress caused by wind or precipi-tation natural disturbances and geomorphologicalfactors) in addition to temperature significantlyaffect treeline elevation and dyna mics (Holtmeier ampBroll 2005 Hagedorn et al 2014 Treml amp Chuman2015) A recent review on the causes producing theupper limits of tree occurrence introduced 6 currentconcepts climatic stress (eg frost damage winterdesiccation) disturbances (eg wind ice blastingavalanches) insufficient carbon supply limitation tocell growth and tissue formation nutritional limita-tion and limited regeneration (Wie ser amp Tausz 2007)Air temperature as the strongest factor influencesthe treeline in 2 different ways temperatures duringthe warm part of the year are the main control oftreeline elevation while temperatures during the coldpart of the year can damage the living tissue of thetrees (evergreen or deciduous species) (Jobbaacutegy ampJackson 2000) Climate change is likely to trigger lat-itudinal and elevational vegetation zone shifts main -ly by altering species mortality and recruitment byexceeding physiological thresholds and changingnatural disturbance regimes (Gonzalez et al 2010)
Vegetation shifts need to be considered as an in -herent adaptation mechanism allowing populationsto track climatically suitable sites Such shifts how-ever can be limited by a variety of non-climatic factors such as nutrient availability soil conditionslandscape fragmentation or species-specific traitsin cluding dispersal capacity competition withground vegetation presence of mycorrhizal fungalsymbionts or increasing virulence of pests andpathogens (Camarero et al 2015b) In particularspecies geographic ranges are expected to shiftdepending on their habitat preferences and theirability to adapt to new conditions A growing bodyof evidence suggests that processes of treeline up -ward expansion drought-induced retraction of spe-cies distributions and shifts of some dominant treespecies may occur in response to global climatechange (Kullman 1999 Kittel et al 2000 Hansen etal 2001 Payette et al 2001 Theurillat amp Guisan2001) Using a meta- analysis treeline advance wasrecorded in 52 of 166 sites around the world(Harsch et al 2009) On the other hand there is
136
Cudliacuten et al Drivers of treeline shift
evidence that treelines have always been dynamicand influenced by climate change and forest devel-opment cycles in the past (Kullman 1988 2007Gehrig-Fasel et al 2007 Vla dovi et al 2014 Tremlamp Chuman 2015)
There are 3 main aspects of environmental changeto which trees are likely to respond increasing tem-perature rising concentrations of CO2 and increasingdeposition of nitrogen (Grace et al 2002 Lindner etal 2014) The trees at the treeline accumulate carbo-hydrates because autotrophic respiration is morelimited by low temperature than photosynthesis Thismeans that factors such as temperature and nitrogenabundance both of which affect the capacity of a treeto use the products of photosynthesis will probablybe more important than factors directly affectingphoto synthesis such as elevated CO2 (Fajardo et al2012)
However climate signals can beconfounded with the effects ofhuman activity Stronger treelinedyna mics due to a coupled effect ofclimate change and highland pastureabandonment frequently occur inEuropean mountain ranges (Ellen-berg 1988) Other human-relatedinfluential factors such as (overly)intensive forest management pas-ture and grazing may be regionallyimportant as well Di rect human-related factors may be crucial inchanging the species compositionand structure of mountain forestsespecially in treeline ecotones whereclimate change does not representthe key factor (Alados et al 2014)
The aims of this study were to eval-uate the effects of geographical posi-tion and different regional climaticand other site conditions on treelineecotone characteristics of selectedEuropean mountain areas to identifythe natural and anthropogenic driv-ers of treeline shift We further dis-cuss the effects of climate and landuse changes on biodiversity in forestsbelow the treeline
2 METHODS
To demonstrate and discuss thepossibilities and limits of treeline
species shift including latitudinal and longitudinalheterogeneity the current situation in 11 mountainareas from Italy to Norway and from Spain to Bul-garia were analysed (Fig 1) To determine the influ-ence of mountain massif size on the mesoclimatic andvegetation conditions including treeline formationwe included large mountain complexes (PyreneesAlps Scandes) as well as relatively small separatedmountains (Pirin and Giant Mountains) in this studyThe characteristics of the treeline ecotone includinghistorical and recent human im pacts on mountainecosystems for the 11 mountain areas of interest aresummarized in Table 1 Extensive mountain massifs(Pyrenees Alps Scandes) were further divided intomore homogeneous units (Central Pyrenees EasternPyrenees Central Alps Eastern Alps Hardan-gervidda Dovre Northern Swedish Lapland InnerFinnmarknorthernmost Finnish Lapland) To esti-
137
Fig 1 Selected European mountains Grey shading differentiates mountainareas 1 Central Pyrenees (CP) 2 Eastern Pyrenees (EP) 3 Apennines (AP) 4Shara Mts (SH) 5 Pirin (PI) 6 Central Balkan Mts (CB) 7 Northern DinaricMts (DI) 8 Central Alps (CA) 9 Eastern Alps (EA) 10 Low Tatra Mts (LT) 11Giant Mts (GM) 12 Hardangervidda (HG) 13 Dovre (DO) 14 NorthernSwedish Lapland (NS) 15 Inner Finnmarknorthernmost Finnish Lapland (FL)
Clim Res 73 135ndash150 2017138
Country Coordi- Elevation Elevation Range of Annual Area VZs near Tr part of nates of range range of Tr tree species altitudinal of study (m asl)mountains central site of T (m asl) limit shift (km2)
(m asl) (m asl) of T or Tr (m yrminus1)
Spain 42deg37rsquoN 1930 No data Pu 2700 049 (T 15 800 Pine with Aa Spanish 0deg27rsquoE (1800minus2050) 1956minus2006) Ps and Fs (1400minus2050) Pyrenees subalpine with Pu
(1800minus2300)
Italy 42deg05rsquoN 1800 No data Fs 2100 1 (Tr [Pm ] 741 Beech forest (800minus1850) Majella Massif 14deg05rsquoE (1700minus1850) Pm 2300 1954minus2007) mixed beechminusdwarf pine part of (1800minus2000) dwarf pine Apennines (2100minus2300)
Macedonia 41deg47rsquorsquoN 1850 1900 Pa 2100 ~1 (T ) 829 Mixed firminusbeechminusShara Mountain 20deg33rsquoE (1800minus1900) (1850minus1950) Pp 2200 1934minus2010 spruce (800minus1900) (Balkan beech (1600minus1900) Peninsula) dwarf pine (1600minus2200)
Bulgaria 41deg42rsquoN 1900 2100 Pp 2500 No data 384 Beech (1000minus1500) Pirin 23deg31rsquoE (1800minus2100) (1900minus2300) coniferous forest (Ps Pp
Ph Pa 1300minus2300) dwarf pine (2000minus2500)
Bulgaria 42deg47rsquoN Beech T 1600 1600 Fs 1700 No data 717 Beech (800minus1700) Central 24deg36rsquoE (1500minus1700) (1500ndash1850) Pa Aa 1850 coniferous (Pa Aa Balkan Mts 1500minus1850) juniper
(1500minus1850)
Slovenia 45deg36rsquoN 1540 1600 Fs 1650 No data ~200 Mixed firminusbeechminusspruce Northern 14deg28rsquoE (1455minus1600) (1550minus1650) Pa 1700 (600minus1400) beech (1300minus1550) Dinaric Mts Aa 1650 dwarf pine with small groups
or individual trees of Fs Pa Aa Ap (1500minus1700)
Switzer- 46deg46rsquoN 1900 2100 Pa 2400 ca13 (T ~250 Spruce with Ld land Italy 9deg52rsquoE (1700minus2150) (1850minus2300) Ld 2450 1954minuspresent) + ~750 (1000minus1900) forest with Pc Central and 46deg17rsquoN Ld and Pa (1700minus2200) Eastern Alps 11deg45rsquoE dwarf shrub vegetation partly
pastures (2200minus2500)
Table 1 Timberline and treeline parameters and characteristics of human activities changes according to climatic models andlimits to climate change induced species shifts in selected European mountains A1B A2 emissions scenarios (IPPC) AaAbies alba Ap Acer pseudoplatanus Fs Fagus sylvatica Ld Larix decidua Pa Picea abies Pc Pinus cembra Ph Pi-nus heldreichii Pm Pinus mugo Pp Pinus peuce Ps Pinus sylvestris Pu Pinus uncinata T timberline Tr treeline TSL
tree species limit VZ vegetation zone C century mng management Elevation range of T and Tr mean (minminusmax)
Cudliacuten et al Drivers of treeline shift 139
Human Annual Limits Referencesinfluence temperature (T ) to climate
in VZs in past and precipitation (P) change and recently differences induced species
between 1961minus1990 shiftand 2021minus2050
mean of CORDEX models
Since 11th C HIRHAM5 model Inadequacy of Batllori et al (2009) clear-cutting and grazing period ΔT2021minus2050 substrata (rocky) Alados et al (2014)
moderate on irregular = 15minus2degC ΔT2051minus2080 climatic constraints Gartzia et a (2014) slopes and ridges since 1930 = 4degC ΔP2021minus2050 (including growing- Camarero et al (2015a)significant release of influence = +5 ΔP2051minus2080 season extreme events
but recovery mainly at = +30 compared to such as late-spring moderate elevations 1960minus1990 freezing summer
drought etc)
Since 1000 BC HadCM3_A2 period Low density of Fs van Gils et al (2008) cutting burning and 2020minus2080 in T dispersal distances Palombo et al (2013 2014)
grazing mng since the mid-20th ΔTmin_Jan2050 = 17degC of Fs seeds upslope C grazing intensity decreased ΔTmin_Jan2080 = 31degC poor soils excessive
now occasionally managed ΔP2050 = minus7 exposure to solar radiation and regeneration of Pm ΔP2080 = minus23 in open areas
Since 14th C clear-cutting Mean of 4 GCMs Differences in soil types Em (1986) and grazing to enlarge (CSIROMk2 HadCM3 between subalpine beech Strid et al (2003)
alpine pastures and to produce ECHAM4OPYC3 and mixed forest (poor Bergant (2006) charcoal since 1970 NCAR minus PCM) ΔT2050 = calcareous soils or rankers on Amidžic et al (2012)
abandonment of traditional 26degC ΔT2100 = 53degC silicate bedrock) cliffs agricultural practices ΔP2050 = minus3 ΔP2100 = minus8 and rocks cattle grazing
Since 680 AD intensive HADCM3_A2 Rocky sites extreme Velchev (1997) fires deforestation and grazing CGGM2 hydrothermal conditions Grunewald amp
since 1962 traditional mng ΔT2050 = 17degC competition of shrubs and tree Scheithauer (2011) has changed last decades tourism ΔT2080 = 28degC seedlings locally intense Raev et al (2011)
impact has increased ΔP2080 = minus15 pasturing
Human impact (mostly burning) HADCM3_A2 Low density of Fs Raev et al (2011) started in 17th C excessive CGCM2 ΔT2050 = 17degC in T dispersal distances Gikov et al (2016) and improper forest mng ΔT2080 = 28degC of Fs seeds upslope pasture Zhiyanski et al
release of influence since 1980 ΔP2050 = minus15 intensity new (2008 2016)ΔP2080 = minus15 expansion of juniper stands
15minus18th C slash burn DMI-HIRHAM5_A1B Impact of wild large Klopltilt et al grazing mng 18minus19th C period 2001minus2100 ΔT2050 ungulates on tree seedlings (2010 2015)
uncontrolled cutting 20th C = 12degC ΔT2100 = 25degC (especially Aa) shallow Mina et al (2017)irregular shelterwood systems ΔP2050 = +15 mm nutrient-poor soils
and over-exploitation mng ΔP2100 = minus18 mm (ie ranker and rendzina)
Since 12th C slash burn grazing mng ENSEMBLES models_ Frequent disturbances Tattoni et al (2010) 13minus19th C intensive forest A1B period 1980minus2100 by snow avalanches Kulakowski et al (2011)
mng at the end of the 19th C ΔT2050 = 12minus18degC and snow gliding ongoing Barbeito et al (2012) decrease in grazing ΔT2100 = 27minus41degC cattle grazing game grazing Bebi et al (2012)
ΔP2050 = 0 mm in the coldest years Dawes et al (2015)ΔP2100 = minus5 mm
Continued on next page
Clim Res 73 135ndash150 2017
mate the geographical position of these mountainunits 1 value was used for the geographical latitudeand longitude of the locus of the studied mountainunits The analysed parameters of these 15 mountainunits are shown in Table 2
In addition we calculated the following climatecharacteristics for all mountain units annual meanair temperature annual sum of precipitation thegrowing season length and the date of the onset ofthe growing season These characteristics were thencompared for 2 distinct periods 1961minus1990 and1991minus 2015 (Table 3) The climate characteristicswere derived from the E-OBS gridded dataset of sta-tion observations version 131 (Haylock et al 2008)We used the E-OBS version on the regular longi-tudeminuslatitude grid with a horizontal resolution of 025degrees The climate data from all grid points withinthe mountain units or near their geographical bor-ders were considered
Data processing for 15 mountain units was made onthe basis of data obtained from the literature and
personal studies by the co-authors these are sum ma -rized in Tables 1 amp 2 and in Tables S1 amp S2 in theSupplement at www int-res com articles suppl c073p135 _ supp pdf In addition a semi-quantitative valu-ation of abiotic biotic and anthropogenic factors lim-iting an upward treeline shift was performed (Fig 2)To answer the question how natural and anthro-pogenic factors have influenced treeline ecotonecharacteristics with a focus on treeline shift redun-dancy analysis (RDA) was used to describe and testthe effect of the explanatory variables (geographicalposition size of the whole mountain massif start ofhuman influence and start of the decrease in humaninfluence) on treeline ecotone characteristics (tim-berline elevation width of the treeline ecotone treespecies forming the treeline and treeline shift peryear) and identify groups of mountain units with sim-ilar variability of the dependent data (ter Braak ampŠmilauer 2012) The whole data set for this analysis isshown in Table S2 We tested both their simple ef -fects which show how much variation every ex plan -
140
Country Coordi- Elevation Elevation Range of Annual Area VZs near Tr part of nates of range range of Tr tree species altitudinal of study (m asl)mountains central site of T (m asl) limit shift (km2)
(m asl) (m asl) of T or Tr (m yrminus1)
Slovakia 48deg55rsquoN 1400 1530 Pa 1730 ~03 (Tr ~400 Spruceminusfirminusbeech Dumbier Tatra 19deg31rsquoE (1350minus1450) (1400minus1630) 1950minuspresent) (1200minus1350) spruce (1300minus(part of Low 1550) Pm with individual Tatra Mts) Pa trees (1400minus1700)
Czech Republic 50deg42rsquoN 1250 1320 Pa 1500 043 (Tr 550 Beechminusspruce (700minus1050) Giant Mts 15deg38rsquoE (1130minus1460) (1250minus1460) 1936minus2005) Pa dominates with increasing
altitude spruce (1000minus1400) Pm with Pa trees (1400minus1560)
Norway 60deg10rsquoN 850 600minus1050 1500 08 (Tr 40 600 Birch forest dominance Southern 7deg40rsquoE 62deg20rsquoN 1915minus2007) at high altitudes but with Scandes 10deg05rsquoE some scattered Ps or Pa
mixed birchminusconifer below 800
Norway 68deg20rsquoN West 650 West 750 West 1000 06 (Tr 35 000 Birch forest dominance Sweden 18deg20rsquoE East 100 East 290 East 0 1958minus2008) at all altitudes but with some Finland 69deg50rsquoN scattered pines or pine stands Northern 27deg00rsquoE on sandy soilsScandes
Table 1 (continued)
Cudliacuten et al Drivers of treeline shift
atory variable can explain separately without usingthe other variables and their conditional ef fectswhich depend on the variables already selected inthe model The statistical significance of the expla -natory variables was tested using a Monte Carlo permutation test and only predictors with p le 005were included in the subsequent RDA (Šmilauer ampLepš 2014)
To find the drivers of biodiversity loss in forestsmeadows and animal communities (dependent vari-ables) the explanatory variables (geographical posi-tion timberline elevation tree species forming thetreeline start of human influence start of the de -crease in human influence and temperature increasebetween the 2 periods 1961minus1990 and 2021minus2050)were tested again by RDA in Canoco 5 as mentionedabove (Figs 3 amp 4) The whole data set for this analy-sis mdash including additional information about tree-lines in the mountain areas of interest with a focuson treeline shift (its rate drivers limits and problemswith its assessment) mdash is presented in Tables S1 amp S2
Selected univariate graphs were constructed to visu-alize the data of relationships between the explana-tory and dependent variables of all analyses whichwere not seen distinctly from the results of the multi-variate statistics or linear regression (Fig 5)
3 RESULTS
31 Effect of environmental variables on treeline ecotone characteristics with a focus
on treeline shift
The variation in the dependent variables (timber-line elevation width of treeline ecotone tree speciesforming the treeline and treeline shift per year) wassignificantly affected only by latitude and size of thewhole mountain massif (Fig 3) The adjusted ex -plained variability by all explanatory variables was342 (F = 46 p = 001) The RDA diagram showsthat in the mountains located more to the north the
141
Human Annual Limits Referencesinfluence temperature (T ) to climate
in VZs in past and precipitation (P) change and recently differences induced species
between 1961minus1990 shiftand 2021minus2050
mean of CORDEX models
From 14th C to 1922 Average of 10 RCM_B1 More rocky and nutrient- Koumlrner (2003) mining forest change ΔT2050 = 18degC poor soils late frosts Fridley et al (2011)
to spruce monocultures ΔT2100 = 37degC steep slopes with avalanches Hlaacutesny et al (2011 2016)from 13th C to 1978 pastures ΔP2050 = +24 mm absence of mature trees
in upper parts of mountain ΔP2100 = minus35 mm
9minus11th C forest clear cutting ALADIN_A1B period 1961minus Different soil types Treml and Banaš (2000) grazing in dwarf pine VZ 2100 ΔT2050 = 13degC for Fs and Pa (cambisol Cudliacuten et al (2013)
since 18th C forest change ΔT2100 = 34degC versus podzol) only TSL of Pa Treml amp Chuman (2015) to spruce monocultures ΔP2050 = +25 mm could elevate on ranker Treml amp Migo (2015)
since 19th C artificially planted ΔP2100 = minus14 mm soils in Pm VZ
Grazed and browsed RegClimMPIHadley Sheep and reindeer Dalen amp Hofgaard (2005) by reindeer and livestock ΔT2050 = 12degC grazing and episodic Wielgolaski (2005)
(sheep and cattle) ΔT2100 = 22degC insect outbreaks (Epirrita) Hofgaard et al (2009) over thousands of years ΔP2050 = +23 mm Kullman Oumlberg (2009)
ΔP2100 = minus12 mm
Grazed and browsed RegClimMPIHadley Continued grazing regime Dalen amp Hofgaard (2005) by semi-domestic reindeer ΔT2050 = 16degC and frequent episodic Wielgolaski (2005)
for hundreds of years ΔT2100 = 29degC insect outbreaks (Epirrita) Toslashmmervik et al (2005) increased grazing from ΔP2050 = +18 mm Hofgaard et al (2009)
year 1960 ΔP2100 = minus10 mm Aune et al (2011) Mathisen et al (2014)
Clim Res 73 135ndash150 2017
treeline ecotone is narrower treeline shiftis smaller the timberline occurs at lowerelevations and the treeline is formed moreby broadleaved species Higher values oftreeline ecotone width and treeline shiftwere found in the mountains with a greatersize of the whole mountain massif
When every explanatory variable wastested separately values of the start of thedecrease of human influence in the moun-tains also had a significant effect on thevariation of the tree line ecotone data(explained variation = 206 pseudo-F =34 p = 0024) Some selected relationshipsof the explanatory and dependent vari-ables even those not showing significanteffects on the variation in the tree line eco-tone parameters are depicted in Fig 5
32 Effect of recent changes in climateand land use on the biodiversity
Biodiversity loss in forests meadows andanimal communities analysed by RDAwas explained by geo graphical positiontreeline species composition and tempera-ture increase between the 2 periods1961minus1990 and 2021minus2050 (Fig 4) Theadjusted ex plained variation by all ex pla -natory variables was 632 (F = 58 p =0001) The RDA results indicated that bio-diversity loss in forest communities in crea -sed with increasing latitude and longitudeas well as in the case where broad leavedspecies formed the treeline In contrast thehighest biodiversity loss in meadows wasfound in the mountains positioned more tothe south and with higher temperatureincrease The biodiversity loss in animalcommunities was negatively correlatedwith longitude and temperature increase
4 DISCUSSION
The comparison of several mountainareas situated across Europe shows a varia-tion in understanding of what constitutesthe treeline across countries The primaryissue is the different approach to the defini-tion of forest when applying a tree heightthreshold This threshold decreases from
142
Mou
nta
in u
nit
Nor
th
Eas
t M
oun
-T
imb
er-
Wid
th
Tre
e T
emp
erat
ure
T
reel
ine
Sta
rt
Sta
rt
lati
tud
elo
ng
itu
de
tain
lin
eof
spec
ies
incr
ease
sh
ift
of H
I of
HI
size
(m a
sl
)tr
eeli
ne
form
ing
b
etw
een
(m
yrminus
1 )(y
r)d
ecre
ase
(km
2 )ec
oton
eth
e 19
61minus
1990
(y
r)(m
)tr
eeli
ne
and
199
1minus20
15 (
degC)
Cen
tral
Pyr
enee
s (C
P)
42deg2
1rsquoN
0deg30
rsquoW55
000
2250
300
C0
81
92 (
1956
minus20
06)
1000
BC
1930
Eas
tern
Pyr
enee
s (E
P)
42deg2
8rsquoN
1deg30
rsquoE55
000
2400
300
C0
80
71 (
1956
minus20
06)
1000
1930
Ap
enn
ines
(A
P)
42deg0
5rsquoN
14deg0
5rsquoE
5000
018
0030
0B
14
1 (1
954minus
2007
)10
00 B
C19
50S
har
a M
ts (
SH
)41
deg47rsquo
N20
deg33rsquo
E16
0018
5035
0C
B0
71
(193
4minus20
10)
1300
1970
Pir
in (
PI)
41deg4
2rsquoN
23deg3
1rsquoE
2585
1900
400
C0
5n
o d
ata
680
1962
Cen
tral
Bal
kan
Mts
(C
B)
42deg4
7primeN
24deg3
6primeE
1160
016
0025
0C
B0
7n
o d
ata
1700
1980
Nor
ther
n D
inar
ic M
ts (
DI)
45deg3
6rsquoN
14deg2
8rsquoE
1000
1540
160
B1
3n
o d
ata
1400
1900
Cen
tral
Alp
s (C
A)
46deg4
6rsquoN
9deg52
rsquoE18
000
021
0055
0C
07
13
(195
4minus20
15)
1300
1880
Eas
tern
Alp
s (E
A)
46deg1
7rsquoN
11deg4
5rsquoE
180
000
1800
400
C0
91
3 (1
954minus
2015
)11
5019
50L
ow T
atra
Mts
(LT
)48
deg55rsquo
N19
deg31rsquo
E12
4014
0033
0C
09
057
(19
75minus
2016
)14
0019
22G
ian
t M
ts (
GM
)50
deg42rsquo
N15
deg38rsquo
E63
012
5025
0C
07
043
(19
36minus
2005
)10
0019
63H
ard
ang
ervi
dd
a (H
G)
60deg0
3rsquoN
7deg47
rsquoE90
0090
015
0C
B0
50
8 (1
915minus
2007
)80
019
70D
ovre
(D
O)
62deg2
1rsquoN
8deg52
rsquoE60
0010
0040
0B
07
08
(191
5minus20
07)
1000
no
dec
reas
eN
orth
ern
Sw
edis
h L
apla
nd
(N
S)
68deg1
2rsquoN
18deg3
3rsquoE
4000
700
300
B1
10
6 (1
958minus
2008
)10
00n
o d
ecre
ase
Inn
er F
inn
mar
kn
orth
ern
mos
t 69
deg10rsquo
N25
deg12rsquo
E60
000
290
90B
10
06
(195
8minus20
08)
1600
no
dec
reas
eF
inn
ish
Lap
lan
d (
FL
)
Tab
le 2
Fac
tors
infl
uen
cin
g t
reel
ine
ecot
one
char
acte
rist
ics
in s
elec
ted
Eu
rop
ean
mou
nta
ins
Th
e m
oun
tain
siz
e is
th
e si
ze o
f th
e w
hol
e m
oun
tain
mas
sive
fu
nct
ion
ing
as a
mez
osca
le c
lim
ate
un
it n
ot o
nly
the
stu
die
d m
oun
tain
are
a W
idth
of t
he
tree
lin
e ec
oton
e w
as d
eter
min
ed b
y su
btr
acti
ng
the
valu
e of
the
mea
n ti
mb
erli
ne
elev
atio
n
from
th
e va
lue
of t
he
up
per
tre
e sp
ecie
s li
mit
C c
onif
ers
B b
road
leav
ed t
rees
HI
hu
man
infl
uen
ce
Cudliacuten et al Drivers of treeline shift
5 m (Jeniacutek amp Lokvenc 1962) to 3 m(Koumlrner 2012) in Northern andCentral Europe to only 2 m (Holt-meier 2009) in Central and South-ern Europe where even shrub -by stands (eg Pinus mugo) areconsi dered as a forest in somecountries eg in Spain Italy andMa ce donia (Batllori et al 2009)An other problem is a different de -signation of forest stands belowthe treeline lsquoupper montane fo -restsrsquo in Central Europe and lsquosub-alpine forestsrsquo in Southern Europe(Ellenberg 1988) Further differen -ces are related to the tradition ofdifferent branches of science (ega more traditional lsquogeobotanicalrsquoapproach in Central Europe ver-sus a more experimental approachin Western Europe) especially inthe rate of applying new pro -gressive methods (eg climaticmodelling or molecular biologicalmethods)
Our comparison of selectedregions (Tables 1 amp 2 Figs 1 amp 2)showed that southern mountainscompared to those located morecentrally and in the north had (1) alonger and more profound exploi -tation by humans in the past (2)greater differences in climatic para -meters (temperature length ofgrowing period) between the 2periods 1961minus1990 and 1991minus2015(Table 3) and (3) more dramaticclimate change scenarios espe-cially concerning temperature in -creases Longitude also had someinfluence on the start and intensityof human influence and on tree-line elevation and rate of treelineshift (Table 2 Fig 5) The occur-rence of broadleaf tree species inthe treeline ecotone in the north-ern (and to some extent the south-ern) countries distinguishes thesefrom the Central European coun-tries where conifers prevail It isinteresting that grazing an im -portant factor shaping treelineecosystems (Dirnboumlck et al 2003
143
Length Beginning Mean Mean of growing of growing annual annual
period season temperature precipitation (d)a (d)b (degC) ()
Central Pyrenees 135 minus89 08 minus107Eastern Pyrenees 117 minus91 08 minus92Apennines 284 minus232 14 57Shara Mts 16 01 07 minus100Pirin 26 minus14 05 27Central Balkan Mts 169 minus136 07 76Northern Dinaric Mts 160 minus95 13 minus56Central Alps 147 minus86 07 62Eastern Alps 125 minus96 09 minus33Low Tatra Mts 24 minus34 09 minus91Giant Mts minus10 13 07 55Hardangervidda Blefjell 63 34 05 56Dovre 138 32 07 64Northern Swedish Lapland 81 05 11 minus27Inner Finnmarknorthern- 17 08 10 77most Finnish Lapland
aPositive numbers indicate a prolongation bNegative numbers indicate an earlier beginning
Table 3 Differences of selected climatic parameters between the 2 study periods (1961minus1990 and 1991minus2015) in selected European mountains
Tein Drou Gefa Nadi Laus Abli Bili Huli
0 1 2 3
Central Pyrenees (CP)
Eastern Pyrenees (EP)
Apennines (AP)
Shara Mts (SH)
Pirin (PI)
Central Balkan Mts (CB)
Northern Dinaric Mts (DI)
Central Alps (CA)
Eastern Alps (EA)
Low Tatra Mts (LT)
Giant Mts (GM)
Hardangervidda (HG)
Dovre (DO)
Northern Swedish Lapland (NS)
Inner FinnmarknorthernmostFinnish Lapland (FL)
Fig 2 Rates of treeline drivers (temperature increase [Tein] land use change[Laus]) and treeline shift limits (drought [Drou] geomorphological factors [Gefa]natural disturbances [Nadi] other abiotic biotic and human factors) in selectedEuropean mountains Abli Bili and Huli abiotic biotic and human treeline shiftlimits respectively) Rate of treeline driver influence mdash 0 no influence 1 weak
influence 2 middle influence 3 strong influence
Clim Res 73 135ndash150 2017
Gehrig-Fasel et al 2007) limits treeline shift in halfof the countries of interest regardless of latitude lon-gitude or recent political developments (Shara PirinCentral Balkans Alps Scandes Fig 2)
Despite the stated differences between studieslooking at climate change impacts on the treelineit is clear that current and future changes in temp -erature will seriously affect treeline ecotones in allmountain ranges of Europe Winter is a period whentreeline stands and individual trees have to survivesevere life-limiting conditions (Wieser amp Tausz2007) Therefore winter warming might increase thechance of young trees surviving and passing the mostcritical period from seedling to sapling and further tothe mature stage (Koumlrner 2003) There is alreadymuch evidence worldwide that winter warming isone of the significant drivers of treeline advance(Harsch et al 2009) Although it is expected that treegrowth at the upper distributional margins (and inthe northern European countries) will increase due toongoing climate change (Pentildeuelas amp Boada 2003Chen et al 2011 Hlaacutesny et al 2011 Lindner et al2014) extremes in temperature (eg black froststemperature reversals in the spring) or winter precip-itation (eg lack of snow drought) might also limit
this process in the future in some regions and localsituations (Holtmeier amp Broll 2005 Hagedorn et al2014) An earlier start to the growing period (espe-cially in the Apennines Central Balkan and SpanishPyrenees see Table 2) can play a negative role in theresistance of trees and seedlings to early springfrosts On the other hand at lower distributional mar-gins (and in Southern European countries) it isexpected that tree growth will decrease or forestswill experience some level of dieback due to drought(Breshears et al 2005 Piovesan et al 2008 Allen etal 2010 Huber et al 2013) A serious implication isthe positive effect that warming can have on root rotfungi and populations of bark beetles resulting inlarge-scale disturbances to temperate and also high-altitude forests (Jankovskyacute et al 2004 Elkin et al2013 Millar amp Stephenson 2015)
Although recent upward advances of treeline eco-tones are widespread in mountain regions (Harsch et
144
Fig 3 Redundancy analysis diagram with variation in tim-berline elevation width of treeline ecotone the tree speciesforming the treeline and treeline shift used as dependentvariables explained by explanatory variables (latitudemountain size) Mountain units (blue) are projected as thecentres of abbreviations (see Table 2) Explanatory variablesaccount for 436 of the total variation in the dependentdata The first canonical axis explained 466 of variationthe second axis explained 30 of variation Dependentvariables (black) are Ecot altitudinal width of treeline eco-tone TLShift treeline shift TrspC (TrspB) treeline formedby conifers (broadleaved trees) Timb timberline elevationExplanatory variables (red) are N north latitude Mtsize
size of the mountain massif
Fig 4 Redundancy analysis diagram with variation in bio-diversity loss in forests meadows and animal communitiesused as dependent variables explained by explanatory vari-ables (geographical position tree species forming the tree-line and temperature increase between the 2 study periods[1961minus1990 and 2021minus2050]) Mountain units (green) areprojected as centres of abbreviations (see Table 2) Explana-tory variables account for 763 of total variation in the de-pendent data The first canonical axis explained 484 ofvariation the second axis explained 183 of variation De-pendent variables (black) are BdfoBdmeBdan biodiversityloss in forestsmeadowsanimal communities Explanatoryvariables (red) are N north latitude E east longitude TrspC(TrspB) treeline formed by conifers (broadleaved trees)Temp temperature increase between the 2 study periods
Cudliacuten et al Drivers of treeline shift
al 2009) only a few published studies have includedquantitative data about treeline shifts (Devi et al2008 Kullman amp Oumlberg 2009 Diaz-Varela et al 2010Van Bogaert et al 2011) Additionally our know -ledge of the spatial patterns in treeline ecotone shiftsat the landscape scale is still surprisingly poor exceptfor some studies from subarctic areas the UralMountains and the Alps (Lloyd et al 2002 Diaz-Varela et al 2010 Hagedorn et al 2014) In the 15studied mountain units the values of treeline shiftranged from 043 to 19 m yrminus1 showing a rather dis-tinct spatial pattern of the dynamics within Europeanmountains The observed treeline shift had a signifi-cant positive relationship only with northern latitudeand a weak positive relationship with mountain-range size (ie the size of the whole mountain massifFig 3) Nevertheless this illustrates the importanceof the mass elevation effect (Koumlrner 2012) and par-tially ex plains why treelines could be at different alti-tudes at the same latitude Small negative regres-sions with east longitude and timberline elevationare shown in Fig 5 There was no apparent depend-ence of the treeline shift rate on climatic parameterchanges between the periods 1961minus1990 and1991minus2015
A whole forest vegetation zone shift is a complexprocess as not only the life strategies of an individualtree need to be considered but plant animal andmicroorganism species and their interrelationshipsas well as the relationships to their specific microhab-
itats (including soil conditions) are also involved(Urban et al 2012) Across all regions we identifiedseveral important obstacles to treeline shifts oftenrelated to site properties such as significant rocki-ness having shallow or low-nutrient soils extremerelief causing disturbances by avalanches snow glid-ing or wind damage (data from the Pyrenees Apen-nines Shara Northern Dinaric Low Tatra and GiantMountains see Fig 2) Soil heterogeneity certainlyhas an important role in plant responses to climatechange and could also maintain the resilience of thecommunity assemblages (Fridley et al 2011) An -other group of obstacles includes extreme climaticparameters especially in winter (reported eg fromthe Pyrenees Apennines Pirin Central BalkanNorthern Dinaric and Giant Mountains Fig 2) Theircombination can result in edaphic andor climaticunsuitability of habitats where species could poten-tially migrate Climatically and edaphically suitablesites will likely decline over the next century partic-ularly in mountain landscapes (Bell et al 2014)There fore trees in the treeline ecotone will colonizepreviously forested habitats The colonization suc-cess of individual forest communities is affected bydifferences in species dispersal and recruitmentbehaviour (Dullin ger et al 2004 Jonaacutešovaacute et al2010) For these reasons colonization of new foresthabitats by the next tree generation may not be suc-cessful and may result in loss of species diversity(Honnay et al 2002 Ibaacutentildeez et al 2009 Dobrowski et
145
Fig 5 Univariate graphs of the relationships of dependent variables viz timberline (Timb) and treeline shift per year (Shift) and their explanatory variables Nola north latitude Ealo east longitude DD decimal degrees
Clim Res 73 135ndash150 2017
al 2015) as reported from Macedonia Slovenia theCzech Republic Slovakia and Norway (Fig 4)
Another limiting factor is the existence of strongadaptation mechanisms of the dominant tree specieswhich allow them to survive in their current distribu-tion areas Species migration is likely to be slow dueto the limited quantity of climatically and edaphicallysuitable sites and the slow velocity of seed dispersaland tree regeneration and will be hampered evenmore by fragmentation of high mountain landscapescaused by human activities including brush invasionin abandoned meadows (Gartzia et al 2014) A verti-cal shift in a vegetation zone involves not only singlespecies of plants animals and microorganisms butalso their interrelationships and links to soil condi-tions The speed at which climatic conditions changewill be different from changes in the soil conditionsdue to the slowness of soil-forming processes Soiltypes and conditions (cambisol versus podzol) arecrucial obstacles for the shift from a beech forest zoneinto the spruce forest zone in the Czech Republic(Vacek amp Mat jka 2010) According to other sourcesthe dominant tree species can influence soil for -mation processes (especially humus forms) Underfavourable climatic and orographic conditions beechis able to change its soil conditions over decades tocenturies (J Macku unpubl data)
Significant changes in biodiversity must be expec -ted in all of the mountain areas of interest We foundthat biodiversity declined particularly in the southernregions of Europe where the timberline is situated athigher elevations and human impact is mostlylonger-lasting (Fig 4) Similarly Pauli et al (2012)reported an increase in species richness of mountaingrasslands across Europe except for Mediterraneanregions and assigned this different response of thesouthern regions to decreased water availability Onthe other hand ecosystems which exhibited bio -diversity increases were not only enriched by mig -rant species but the assemblages underwent thermo -philization ie cold-adapted species declined andmore warm-adapted species spread (Gottfried et al2012) However not all species are able to track cli-mate changes It is expected that around 40 ofhabitats will become climatically unsuitable for manymountain species particularly endemic ones in -creasing their extinction probability during this cen-tury (Dullinger et al 2012) The abandonment of tra-ditional land use forms is another source of this losswhich may be a more important driver than tempera-ture increase (Fig 4) Serious biodiversity losses inmountain meadows were reported from Spain ItalyMacedonia Bulgaria Slovenia and Slovakia There-
fore controlled grazing must occur in order to main-tain alpine grasslands (Dirnboumlck et al 2003) Unfor-tunately grazing is still active only in smaller parts ofEuropean mountains (eg in the Central Alps and theCentral Balkan Mountains but pasturing can alsonegatively affect vegetation diversity) and some-times does not serve to maintain alpine grasslandThe same is true for grassland management whichcan help to maintain mountain species and decreasehabitat loss A serious biodiversity loss in mountainmeadows was reported from the Pyrenees the Apen-nines and the Shara Pirin Central Balkan NorthernDinaric and Low Tatra Mountains (Fig 4)
In forests species responses to climate change maybe equivocal some recent studies indicated contra-dictory shifts in species distributions (eg Lenoir etal 2008 Zhu et al 2012 Rabasa et al 2013) This dis-parity is widely discussed and assigned to the greatvariety of non-climatic factors or even tree ontogeny(Grytnes et al 2014 Lenoir amp Svenning 2015 Maacutelišet al 2016) Disturbances leading to tree mortalitymay also play an important role (Cudliacuten et al 2013)changes in tree canopy cover modify light availabil-ity and microclimate and can induce the thermo -philization of forest vegetation (De Frenne et al2015 Stevens et al 2015) The loss of this microcli-mate buffering effect of forests may induce a biotichomogenization of forests (Savage amp Vellend 2015)or the creation of novel non-analogical communitiessuch as oakminuspine forests (Urban et al 2012) whichmay be a new threat to forest biodiversity
The observed changes in treeline forest ecosystemsare often related to changes in land-use intensity(Theurillat amp Guisan 2001 Alados et al 2014) Ac -cording to climate change predictions and the recentand future exploitation intensity of European moun-tains trees and forest communities will shift upwarddue to land use change or climate change or bothReduced land-use intensity certainly will interactwith climate change by facilitating or inhibiting spe-cies occurrence and will accelerate forest expansionabove the present treeline particularly to previouslyforested habitats (Theurillat amp Guisan 2001) Thesimultaneous action of both main treeline shift driv-ers viz temperature increase and decrease in landuse intensity was recorded from all of our studiedmountain areas The extensive differences in timber-line and treeline elevations in almost all studiedmountains (Table 1) indicate the anthropo-zoogenictreeline type (according to Ellenberg 1998) In mostmountains (eg in the Apennines Shara MountainsCentral Balkans and Alps) land use is the prevailingfactor influencing vegetation drift (Fig 5) Previous
146
Cudliacuten et al Drivers of treeline shift
research showed that the upward shift of the treelinein the Swiss Alps was predominantly attributable toland abandonment and only in some situations to cli-mate change (Gehrig-Fasel et al 2007) In the Apen-nines in the last few decades tree establishment hasbeen mainly controlled by land use while treegrowth has been controlled by climate pointing to aminor role played by climate in shaping the currenttreeline (Palombo et al 2013)
The impact of climate change and the connectedland use change on biodiversity and ecosystem serv-ices provision in several European countries is sum-marized by Wielgolaski et al (2017) and Kyriazopou-los et al (2017) (both this Special) and Sarkki et al(2016) One of the possible adaptive management op-tions in response to climate change assisted migra-tion as human-assisted movement of species hasbeen frequently debated in the last few years (Ste-Marie et al 2011) It is possible to apply it as a type ofassisted colonization ie intentional movement andrelease of an organism outside its indigenous range toavoid extinction of populations of the focal species(eg Macedonian pine species) or as an ecologicalre placement ie the intentional movement and re-lease of an organism outside its indigenous range toperform a specific ecological function (eg planting ofPinus mugo in the Alps IUCNSSC 2013)
5 CONCLUSIONS
The analysis of 11 mountain areas across Europeshowed that with increasing latitude the treelineand altitudinal width of the treeline ecotone signifi-cantly decreases as do the significance of climaticand soil parameters as barriers against tree speciesshift Although temperature is the overwhelmingcontrolling factor of tree growth and establishment intemperate and boreal treeline ecotones late-sea-sonal drought might also play a driving role in Medi-terranean treeline ecotones Longitude was lessinfluential mostly affecting climate and land usechange effects on increased biodiversity loss as wellas the size of the area that forms one mesoclimateunit affecting altitudinal ecotone width The biggestpart of the commonly observed remaining variabilityin mountain vegetation near the treeline in Europeseems to be caused by geomorphological geologicalpedological and microclimatic variability in combina-tion with different land use history and the presentsocio-economic relations The observed differencesin climatic parameters between the mountain areasof interest in comparison with the reference period
1960minus1990 (09degC) have not explained the relativelyhigh differences in the rate of treeline shift per year(149 m) The predicted variability in temperatureincrease due to global warming (12minus26degC in 2050and 22minus53degC in 2100) could lead to a much biggerdifferentiation in treeline ecotone biodiversity andecosystem processes between southern and northernEuropean mountains in the future Therefore thesedifferences must be taken into account by scientistsand EU policy makers when formulating efficientadaptive forest management strategies for treelineecosystems at the European level (eg assistedmigration of adapted genotypes)
Acknowledgements This paper is based firstly upon workfrom the COST Action ES 1203 SENSFOR supported byCOST (European Cooperation in Science and Technologywwwcosteu) This international work was further sup-ported by projects granted by the Ministry of EducationYouth and Sports of the Czech Republic grant NPU ILO1415 and LD 14039 by the agency APVV SR under pro-jects APVV-14-0086 and APVV-15-0270 We acknowledgethe E-OBS dataset from the EU-FP6 project ENSEMBLES(httpensembles-eumetofficecom) and the data providersin the ECAampD project (wwwecadeu)
LITERATURE CITED
Alados CL Errea P Gartzia M Saiz H Escoacutes J (2014) Posi-tive and negative feedbacks and free-scale pattern distri-bution in rural-population dynamics PLOS ONE 9 e114561
Allen CD Macalady AK Chenchouni H Bachelet D andothers (2010) A global overview of drought and heat-induced tree mortality reveals emerging climate changerisks for forests For Ecol Manag 259 660minus684
Amidžic L Bartula M Jaric S (2012) Syntaxonomy overviewof vegetation of Šar Planina in the Balkan PeninsulaVegetos 25 348minus360
Aune S Hofgaard A Soumlderstroumlm L (2011) Contrasting cli-mate and land use driven tree encroachment pattern ofsub-arctic tundra in Northern Norway and Kola Penin-sula Can J For Res 41 437minus449
Barbeito I Dawes M Rixen C Senn J Bebi P (2012) Factorsdriving mortality and growth at treeline a 30-year exper-iment of 92000 conifers Ecology 93 389minus401
Batllori E Camarero JJ Ninot JM Gutieacuterrez E (2009)Seedling recruitment survival and facilitation in alpinePinus uncinata treeline ecotones Implications andpotential responses to climate warming Glob Ecol Bio-geogr 18 460minus472
Bebi P Teich M Hagedorn F Zurbriggen N Brunner SHGreacutet-Regamey A (2012) Veraumlnderung von Wald undWaldleistungen in der Landschaft Davos im Zuge desKlimawandels Schweiz Z Forstwes 163 493minus501
Bell DM Bradford JB Lauenroth WK (2014) Mountain land-scapes offer few opportunities for high elevation treespecies migration Glob Change Biol 20 1441minus1451
Bergant K (2006) Climate change scenarios for MacedoniaUniversity of Nova Gorica Center for Atmospheric
147
Clim Res 73 135ndash150 2017
Research Nova GoricaBodin J Badeau V Bruno E Cluzeau C Moisselin JM
Walther GR Dupouey JL (2013) Shifts of forest speciesalong an elevational gradient in Southeast France cli-mate change or stand maturation J Veg Sci 24 269minus283
Boncina A (2011) History current status and future pros -pects of uneven-aged forest management in the Dinaricregion an overview Forestry 84 467minus478
Breshears DD Cobb NS Rich PM Price KP and others(2005) Regional vegetation die-off in response to global-change-type drought Proc Natl Acad Sci USA 102 15144minus15148
Camarero JJ Garciacutea-Ruiz JM Sanguumlesa-Barreda G GalvaacutenJD and others (2015a) Recent and intense dynamics in aformerly static Pyrenean treeline Arct Antarct Alp Res47 773minus783
Camarero JJ Gazol A Galvaacuten JD Sanguumlesa-Barreda GGutieacuterrez E (2015b) Disparate effects of global-changedrivers on mountain conifer forests warming-inducedgrowth enhancement in young trees vs CO2 fertilizationin old trees from wet sites Glob Change Biol 21 738minus749
Chen IC Hill JK Ohlemuumlller R Roy DB Thomas CD (2011)Rapid range shifts of species associated with high levelsof climate warming Science 333 1024minus1026
Cudliacuten P Sejaacutek J Pokornyacute J Albrechtovaacute J Bastian OMarek M (2013) Forest ecosystem services under climatechange and air pollution In Matyssek R Clarke N Cud-liacuten P Mikkelsen TN Tuovinen JP Wiesner G Paoletti E(eds) Climate change air pollution and global chal-lenges understanding and perspectives from forestresearch Developments in Environmental Science Vol13 Elsevier Oxford p 521minus546
Dalen L Hofgaard A (2005) Differential regional treelinedynamics in the Scandes mountains Arct Antarct AlpRes 37 284minus296
Dawes MA Philipson CD Fonti P Bebi P Haumlttenschwiler SHagedorn F Rixen C (2015) Soil warming and CO2
enrichment induce biomass shifts in alpine tree line veg-etation 2014 Glob Change Biol 21 2005minus2021
De Frenne P Rodriacuteguez-Saacutenchez F De Schrijver A CoomesDA Hermy M Vangansbeke P Verheyen K (2015) Lightaccelerates plant responses to warming Nat Plants 1 15110
Devi N Hagedorn F Moiseev P Bugmann H ShiyatovSG Mazepa V Rigling A (2008) Expanding forests andchanging growth forms of Siberian larch at the treeline ofthe Polar Urals during the 20th century Glob ChangeBiol 14 1581minus1591
Diaz-Varela R Colombo R Meroni M Calvo-Iglesias MSBuffoni A Tagliaferri A (2010) Spatio-temporal analysisof alpine ecotones A spatial explicit model targeting altitudinal vegetation shifts Ecol Model 221 621minus633
Dirnboumlck T Dullinger S Grabherr G (2003) A regionalimpact assessment of climate and land-use change onalpine vegetation J Biogeogr 30 401minus417
Dobrowski SZ Swanson AK Abatzoglou JT Holden ZASafford HD Schwartz MK Gavin DG (2015) Forest struc-ture and species traits mediate projected recruitmentdeclines in western US tree species Glob Ecol Biogeogr24 917minus927
Dullinger S Dirnboumlck S Grabherr G (2004) Modelling cli-mate change-driven treeline shifts relative effects oftemperature increase dispersal and invisibility J Ecol92 241minus252
Dullinger S Gattringer A Thuiller W Moser D and others
(2012) Extinction debt of high-mountain plants undertwenty-first-century climate change Nat Clim Chang 2 619minus622
Elkin C Gutierrez AG Leuzinger S Manusch C TemperliC Rasche L Bugmann H (2013) A 2degC warmer world isnot safe for ecosystem services in the European AlpsGlob Change Biol 19 1827minus1840
Ellenberg H (1988) Vegetation ecology of Central Europe4th edn Cambridge University Press Edinburgh
Em H (1986) Spruce southern distribution range Spruce for-est on Shara Mountain in Macedonia Contrib Macedon-ian Acad Sci Arts Skopje 5 11minus28
Fajardo A Piper FI Pfund L Koumlrner CH Hoch G (2012)Variation of mobile carbon reserves in trees at the alpinetreeline ecotone is under environmental control NewPhytol 195 794minus802
Fridley JD Grime JP Askew AP Moser B Stevens CJ (2011)Soil heterogeneity buffers community response to cli-mate change in species-rich grassland Glob Change Biol17 2002minus2011
Gartzia M Alados CL Peacuterez-Cabello F (2014) Assessment ofthe effects of biophysical and anthropogenic factors onwoody plant encroachment in dense and sparse moun-tain grasslands based on remote sensing data Prog PhysGeogr 38 201minus217
Gehrig-Fasel J Guisan A Zimmermann NE (2007) Tree lineshifts in the Swiss Alps climate change or land abandon-ment J Veg Sci 18 571minus582
Gikov A Dimitrov P Zhiyanski M (2016) Land cover dynam-ics on the northern slope of the Troyan Passage during 30years period Probl Geogr 1-2 78minus92
Gonzalez P Neilson RP Lenihan JM Drapek RJ (2010)Global patterns in the vulnerability of ecosystems to veg-etation shifts due to climate change Glob Ecol Biogeogr19 755minus768
Gottfried M Pauli H Futschik A Akhalkatsi M and others(2012) Continent-wide response of mountain vegetationto climate change Nat Clim Chang 2 111minus115
Grace J Berninger F Nagy L (2002) Impacts of climatechange on the tree line Ann Bot 90 537minus544
Grunewald K Scheithauer J (2011) Landscape developmentand climate change in Southwest Bulgaria (Pirin Mts)Springer Heidelberg
Grytnes JA Kapfer J Jurasinski G Birks HH and others(2014) Identifying the driving factors behind observedelevational range shifts on European mountains GlobEcol Biogeogr 23 876minus884
Hagedorn F Shiyatov FG Mazepa VS Devi NM and others(2014) Treeline advances along the Urals mountainrange mdash driven by improved winter conditions GlobChange Biol 20 3530minus3543
Hansen AJ Neilson RP Dale VH Flather CH and others(2001) Global change in forests responses of speciescommunities and biomes BioScience 51 765minus779
Harsch MA Bader MY (2011) Treeline formminusa potential keyto understanding treeline dynamics Glob Ecol Biogeogr20 582minus596
Harsch MA Hulme PE McGlone MS Duncan RP (2009) Aretreelines advancing A global meta-analysis of treelineresponse to climate warning Ecol Lett 12 1040minus1049
Haylock MR Hofstra N Klein Tank AMG Klok EJ JonesPD New M (2008) A European daily high-resolutiongridded dataset of surface temperature and precipita-tion J Geophys Res Atmos 113 D20119
Hlaacutesny T Barcza Z Fabrika M Balaacutezs B and others (2011)
148
Cudliacuten et al Drivers of treeline shift
Climate change impacts on growth and carbon balanceof forests in Central Europe Clim Res 47 219minus236
Hlaacutesny T Trombik J Dobor L Barcza Z Barka I (2016)Future climate of the Carpathians climate change hot-spots and implications for ecosystems Reg EnvironChange 16 1495minus1506
Hofgaard A Dalen L Hytteborn H (2009) Tree recruitmentabove the treeline and potential for climate driven tree-line change J Veg Sci 20 1133minus1144
Holtmeier FK (2009) Mountain timberlines Ecology patchi-ness and dynamics 2nd edn Advances in GlobalChange Research 36 Springer Science amp ScienceMediaBV Dordrecht
Holtmeier FK Broll G (2005) Sensitivity and response ofnorthern hemisphere altitudinal and polar treelines toenvironmental change at landscape and local scalesGlob Ecol Biogeogr 14 395minus410
Honnay O Verheyen K Butaye J Jacquemyn H Bossuyt BHermy M (2002) Possible effects of habitat fragmentationand climate change on the range of forest plant speciesEcol Lett 5 525minus530
Huber R Rigling A Bebi P Brand F and others (2013) Sus-tainable land use in mountain regions under globalchange synthesis across scales and disciplines Ecol Soc18 36
Ibaacutentildeez I Clark JS Dietze MC (2009) Estimating colonizationpotential of migrant tree species Glob Change Biol 15 1173minus1188
IUCNSSC (International Union for Conservation of NatureSpecies Survival Commission) (2013) Guidelines for re -introductions and other conservation translocations Version 10 IUCNSpecies Survival Commission Gland
Jankovskyacute L Cudliacuten P gtermaacutek P Moravec I (2004) The pre-diction of development of secondary Norway sprucestands under the impact of climatic change in the Dra-hany highlands (The Czech Republic) Ekologia (Bratisl)23(Suppl 2) 101minus112
Jeniacutek J Lokvenc T (1962) Die alpine Waldgrenze im Krko -noše Gebirge Rozpravy SAV Rada matematickyacutech apriacuterodniacutech ved 72 gteskoslovenskaacute akademie ved Praha
Jobbaacutegy EG Jackson RB (2000) Global controls of forest lineelevation in the northern and southern hemispheresGlob Ecol Biogeogr 9 253minus268
Jonaacutešovaacute M Vaacutevrovaacute E Cudliacuten P (2010) Western Carpa -thian mountain spruce forest after a windthrow naturalregeneration in cleared and uncleared areas For EcolManag 259 1127minus1134
Kittel TGF Steffen WL Chapin FS (2000) Global andregional modelling of Arcticminusboreal vegetation distribu-tion and its sensitivity to altered forcing Glob ChangeBiol 6 1minus18
Klopcic M Jerina K Bonltina A (2010) Long-term changes ofstructure and tree species composition in Dinaric un -even- aged forests Are red deer an important factor EurJ For Res 129 277minus288
Klopcic M Simonltilt T Bonltina A (2015) Comparison of re -generation and recruitment of shade-tolerant and light-demanding tree species in mixed uneven-aged forests ex -periences from the Dinaric region Forestry 88 552minus563
Koumlrner C (2003) Alpine plant life functional plant ecology ofhigh mountain ecosystems with 47 tables Springer Sci-ence amp Business Media BV Dordrecht
Koumlrner C (2012) Alpine treelines Functional ecology of theglobal high elevation tree limits Springer Basel
Koumlrner C Paulsen J (2004) A world-wide study of high alti-
tude treeline temperatures J Biogeogr 31 713minus732Kulakowski D Bebi P Rixen C (2011) The interacting effects
of land use change climate change and suppression ofnatural disturbances on landscape forest structure in theSwiss Alps Oikos 120 216minus225
Kullman L (1988) Holocene history of the forestminusalpine tun-dra ecotone in the Scandes Mountains (central Sweden)New Phytol 108 101minus110
Kullman L (1999) Early holocene tree growth at a high elevation site in the northernmost Scandes of Sweden(Lapland) a palaeobiogeographical case study based onmega fossil evidence Geogr Ann 81 63minus74
Kullman L (2007) Tree line population monitoring of Pinussylvestris in the Swedish Scandes 1973minus2005 implica-tions for tree line theory and climate change ecologyJ Ecol 95 41minus52
Kullman L Oumlberg L (2009) Post-Little Ice Age treeline riseand climate warming in the Swedish Scandes a land-scape ecological perspective J Ecol 97 415minus429
Kyriazopoulos AP Skre O Sarkki S Wielgolaski FE Abra-ham EM Ficko A (2017) Humanminusenvironment dynamicsin European treeline ecosystems a synthesis based onthe DPSIR framework Clim Res 73 17ndash29
Lenoir J Svenning JC (2015) Climate-related range shifts mdasha global multidimensional synthesis and new researchdirections Ecography 38 15minus38
Lenoir J Geacutegout JC Marquet PA de Ruffray P Brisse H(2008) A significant upward shift in plant species opti-mum elevation during the 20th century Science 320 1768minus1771
Lindner M Fitzgerald JB Zimmermann NE Reyer C andothers (2014) Climate change and European forests What do we know what are the uncertainties and whatare the implications for forest management J EnvironManag 146 69minus83
Lloyd AH Rupp TS Fastie CL Srafield AM (2002) Patternsand dynamic of treeline advance on the Seward Pen -isula Alaska J Geophys Res 107 8161
Maacuteliš F Kopeckyacute M Petriacutek P Vladovic J Merganic J VidaT (2016) Life-stage not climate change explains ob -served tree range shifts Glob Change Biol 22 1904minus1914
Mathisen IE Mikheeva A Tutubalina OV Aune S and Hof-gaard A (2014) Fifty years of tree line change in theKhibiny Mountains Russia advantages of combinedremote sensing and dendroecological approaches ApplVeg Sci 17 6ndash16
Millar CI Stephenson NL (2015) Temperate forest healthin an era of emerging megadisturbance Science 349 823minus826
Mina M Bugmann H Klopcic M Cailleret M (2017) Accu-rate modeling of harvesting is key for projecting futureforest dynamics a case study in the Slovenian moun-tains Reg Environ Change 17 49minus64
Moen J Aune K Edenius L Angerbjoumlrn A (2004) Potentialeffects of climate change on treeline position in theSwedish mountains Ecol Soc 9 16 wwwecologyandsocietyorgvol9iss1art16
Palombo C Chirici G Marchetti M Tognetti R (2013) Is landabandonment affecting forest dynamics at high elevationin Mediterranean mountains more than climate changePlant Biosyst 147 1minus11
Palombo C Marchetti M Tognetti R (2014) Mountain vege-tation at risk current perspectives and research reedsPlant Biosyst 148 35minus41
Pauli H Gottfried M Dullinger S Abdaladze O Akhalkatsi
149
Clim Res 73 135ndash150 2017
M Alonso JLB Grabherr G (2012) Recent plant diversitychanges on Europersquos mountain summits Science 336 353minus355
Payette S Fortin MJ Gamachet I (2001) The subarctic forestminustundra the structure of a biome in a changing climate BioScience 51 709minus718
Pearson GR Dawson TP (2003) Predicting the impacts of cli-mate change on the distribution of species Are bio -climate envelope models useful Glob Ecol Biogeogr 12 361minus371
Pedrotti F (2013) Plant and vegetation mapping GeobotanyStudies Springer-Verlag Berlin
Pentildeuelas J Boada M (2003) A global change-induced biomeshift in the Montseny mountains (NE Spain) GlobChange Biol 9 131minus140
Piovesan G Biondi F Filippo AD Alessandrini A MaugeriM (2008) Drought driven growth reduction in old beech(Fagus sylvatica L) forests of the central ApenninesItaly Glob Change Biol 14 1265minus1281
Rabasa SG Granda E Benavides R Kunstler G and others(2013) Disparity in elevational shifts of European trees inresponse to recent climate warming Glob Change Biol19 2490minus2499
Raev I Zhelev P Grozeva M Markov I and others (2011)Programme of measures for adaptation of forests in Bul-garia and mitigate the negative impact of climate changeon them Project FUTURE forest helping Europe tackleclimate change INTERREG IVC ERDF Sofia
Sarkki S Ficko A Grunewald K Nijnik M (2016) Benefitsfrom and threats to European treeline ecosystem serv-ices an exploratory study of stakeholders and gover-nance Reg Environ Change 16 2019minus2032
Savage J Vellend M (2015) Elevational shifts biotic homo -genization and time lags in vegetation change during 40years of climate warming Ecography 38 546minus555
Schwoumlrer C Henne PD Tinner W (2014) A model-data com-parison of Holocene timberline changes in the SwissAlps reveals past and future drivers of mountain forestdynamics Glob Change Biol 20 1512minus1526
Šmilauer P Lepš J (2014) Multivariate analysis of ecologicaldata using Canoco 5 Cambridge University Press Cam-bridge
Ste-Marie C Nelson EA Dabros A Bonneau ME (2011)Assisted migration introduction to a multifaceted con-cept For Chron 87 724minus730
Stevens JT Stafford HD Harrison S Latimer AM (2015) For-est disturbance accelerates thermophilization of under-story plant communities J Ecol 103 1253minus1263
Strid A Andonoski A Andonovski V (2003) Alpine biodiver-sity in Europe Ecological Studies 167 Springer-VerlagBerlin
Tattoni C Ciolli M Ferretti F Cantiani MG (2010) Monitor-ing spatial and temporal pattern of Paneveggio forest(Northern Italy) from 1859 to 2006 iForest Biogeosci For3 72minus80
Ter Braak CJF Šmilauer P (2012) Canoco reference manualand userrsquos guide software for ordination (version 50)Microcomputer Power Ithaca NY
Theurillat JP Guisan A (2001) Potential impact of climatechange on vegetation in the European Alps a reviewClim Change 50 77minus109
Toslashmmervik H Wielgolaski FE Neuvonen S Solberg BHoslashgda KA (2005) Biomass and production on a land-scape level in the mountain birch forests In WielgolaskiFE (ed) Plant ecology herbivory and human impact in
Nordic mountain birch forests Ecos Studies 180Springer Berlin p 53minus70
Treml V Banaš M (2000) Alpine timberline in the HighSudetes Acta Univ Carol Geogr 15 83minus99
Treml V Chuman T (2015) Ecotonal dynamics of the altitu-dinal forest limit are affected by terrain and vegetationstructure variables an example from the Sudetes Moun-tains in Central Europe Arct Antarct Alp Res 47 133minus146
Treml V Migo P (2015) controlling factors limiting timber-line position and shifts in the Sudetes a review GeogrPol 88 55minus70
Urban MC Tewksbury JJ Sheldon KS (2012) On a collisioncourse competition and dispersal differences create no-analogue communities and cause extinctions during cli-mate change Proc R Soc Lond B Biol Sci 279 2072minus2080
Vacek S Matejka K (2010) State and development of phyto-cenoses on research plots in the Krkonoše Mts foreststands J For Sci 56 505minus517
Van Bogaert R Haneca K Hoogesteger J Jonasson C deDapper M Callaghan TV (2011) A century of tree linechanges in sub-Arctic Sweden shows local and regionalvariability and only a minor influence of 20th century climate warming J Biogeogr 38 907minus921
Van Gils H Batsukh O Rossiter D Munthali W Liberato -scioli E (2008) Forecasting the pattern and pace of Fagusforest expansion in Majella National Park Italy ApplVeg Sci 11 539minus546
Velchev V (1997) Types of vegetation In Yordanova MDonchev D (eds) Geography of Bulgaria BAN Sofiap 269minus283 (in Bulgarian)
Vladovic J Merganic J Maacuteliš F Križovaacute E and others (2014)The response of forest vegetation diversity to changes inedaphic-climatic conditions in Slovakia Technickaacute uni-verzita vo Zvolene Zvolene (in Slovak)
Vrška T Adam D Hort L Kolaacuter T Janiacutek F (2009) Europeanbeech (Fagus sylvatica L) and silver fir (Abies alba Mill)rotation in the Carpathians mdash a developmental cycle or alinear trend induced by man For Ecol Manag 258 347minus356
Wielgolaski FE (ed) (2005) Plant ecology herbivory andhuman impact in Nordic mountain birch forests Eco -logical Studies 180 Springer Verlag Berlin
Wielgolaski FE Hofgaard A Holtmeier FK (2017) Sensitivityto environmental change of the treeline ecotone and itsassociated biodiversity in European mountains Clim Res73151ndash166
Wieser G Tausz M (2007) Trees at their upper limit Treelifelimitation at the alpine timberline Springer Dordrecht
Wieser G Holtmeier FK Smith WK (2014) Treelines in achanging global environment In Tausz M Grulke N(eds) Trees in a changing environment Plant Ecophys 9Springer Dordrecht p 221minus263
Zhiyanski M Kolev K Sokolovska M Hursthouse A (2008)Tree species effect on soils in Central Stara PlaninaMountains Nauka Gorata 4 65minus82
Zhiyanski M Gikov A Nedkov S Dimitrov P Naydenova L(2016) Mapping carbon storage using land coverlanduse data in the area of Beklemeto Central Balkan In Koulov B Zhelezov G (eds) Sustainable mountain re -gions challenges and perspectives in southeasternEurope Springer Basel p 53minus65
Zhu K Woodall CW Clark JS (2012) Failure to migrate lackof tree range expansion in response to climate changeGlob Change Biol 18 1042minus1052
150
Editorial responsibility Nils Chr Stenseth Oslo Norway
Submitted April 8 2016 Accepted March 8 2017Proofs received from author(s) July 7 2017
Clim Res 73 135ndash150 2017
1 INTRODUCTION
Mountain regions are crucial areas for studying theimpact of climate change on vegetation communitiessteep climatic gradients enable testing of ecologicalecophysiological and evolutionary responses of florato changing geophysical influences related to climatechange In addition most of the species living theregrow in conditions classified as their ecological limits(Koumlrner 2012) Vegetation in European mountainousregions has been subjected to severe changes duringthe remote (eg Kullman 1988) and recent pasts (egVrška et al 2009 Boncina 2011 Bodin et al 2013Elkin et al 2013 Schwoumlrer et al 2014) Changes inspecies distribution and plant community composi-tion have been the most often observed althoughevidence of complete community exchange has notyet been registered
A strong elevational zonation is typical of montanevegetation caused mainly by steep climatic gradi-ents Vegetation zones are characterized by a spe-cific structural-functional type of phytocoenosis con-structed by particular main plant species (edificators)and roughly following latitude (latitudinal vegetationzones) or altitude (altitudinal vegetation zones orvegetation belts) Climatic zones along altitudinalgradients are compressed with large habitat andspecies diversity in successive altitude vegetationzones In these steep gradients species richnessdecreases with increasing altitude although topo-graphic isolation results in increased levels of en -demism (Pedrotti 2013)
The most obvious vegetation boundary at high ele-vations is that of the upper forest limit (Harsch ampBader 2011) Owing to its diffuse character it mightbe better to refer to the treeline ecotone which consists of the belt between the boundary of theclosed forest standminustimberline and the uppermost ornorthern most scattered and stunted individuals ofthe forest-forming tree species regardless of theirgrowth form and height (Holtmeier amp Broll 2005)Within this ecotone the treeline (ie the line connect-ing the tallest patches of forest composed of trees ofge3 m height) is often delimited at this elevation themean temperatures for the growing season are ca64degC (Koumlrner 2012) Climate is one of the mostimportant limiting factors defining the spatial distri-bution of any species (Pearson amp Dawson 2003Wieser et al 2014) Temperature especially summermean temperature and temperature sums is the pri-mary factor causing the formation of treelines at theglobal scale (Grace et al 2002 Moen et al 2004)although it may be substantially affected by several
other non-climatic factors eg mass elevation effect(Ellenberg 1988) past land use or forest management(Gehrig-Fasel et al 2007) The main factors affectingthe expansion andor retreat of tree stands at theirupper limits are scale-dependent (Holtmeier 2009)At the global scale treeline position is determined bygrowing-season temperatures (Koumlrner amp Paulsen2004) whereas at the landscape scale second-orderfactors (ie climatic stress caused by wind or precipi-tation natural disturbances and geomorphologicalfactors) in addition to temperature significantlyaffect treeline elevation and dyna mics (Holtmeier ampBroll 2005 Hagedorn et al 2014 Treml amp Chuman2015) A recent review on the causes producing theupper limits of tree occurrence introduced 6 currentconcepts climatic stress (eg frost damage winterdesiccation) disturbances (eg wind ice blastingavalanches) insufficient carbon supply limitation tocell growth and tissue formation nutritional limita-tion and limited regeneration (Wie ser amp Tausz 2007)Air temperature as the strongest factor influencesthe treeline in 2 different ways temperatures duringthe warm part of the year are the main control oftreeline elevation while temperatures during the coldpart of the year can damage the living tissue of thetrees (evergreen or deciduous species) (Jobbaacutegy ampJackson 2000) Climate change is likely to trigger lat-itudinal and elevational vegetation zone shifts main -ly by altering species mortality and recruitment byexceeding physiological thresholds and changingnatural disturbance regimes (Gonzalez et al 2010)
Vegetation shifts need to be considered as an in -herent adaptation mechanism allowing populationsto track climatically suitable sites Such shifts how-ever can be limited by a variety of non-climatic factors such as nutrient availability soil conditionslandscape fragmentation or species-specific traitsin cluding dispersal capacity competition withground vegetation presence of mycorrhizal fungalsymbionts or increasing virulence of pests andpathogens (Camarero et al 2015b) In particularspecies geographic ranges are expected to shiftdepending on their habitat preferences and theirability to adapt to new conditions A growing bodyof evidence suggests that processes of treeline up -ward expansion drought-induced retraction of spe-cies distributions and shifts of some dominant treespecies may occur in response to global climatechange (Kullman 1999 Kittel et al 2000 Hansen etal 2001 Payette et al 2001 Theurillat amp Guisan2001) Using a meta- analysis treeline advance wasrecorded in 52 of 166 sites around the world(Harsch et al 2009) On the other hand there is
136
Cudliacuten et al Drivers of treeline shift
evidence that treelines have always been dynamicand influenced by climate change and forest devel-opment cycles in the past (Kullman 1988 2007Gehrig-Fasel et al 2007 Vla dovi et al 2014 Tremlamp Chuman 2015)
There are 3 main aspects of environmental changeto which trees are likely to respond increasing tem-perature rising concentrations of CO2 and increasingdeposition of nitrogen (Grace et al 2002 Lindner etal 2014) The trees at the treeline accumulate carbo-hydrates because autotrophic respiration is morelimited by low temperature than photosynthesis Thismeans that factors such as temperature and nitrogenabundance both of which affect the capacity of a treeto use the products of photosynthesis will probablybe more important than factors directly affectingphoto synthesis such as elevated CO2 (Fajardo et al2012)
However climate signals can beconfounded with the effects ofhuman activity Stronger treelinedyna mics due to a coupled effect ofclimate change and highland pastureabandonment frequently occur inEuropean mountain ranges (Ellen-berg 1988) Other human-relatedinfluential factors such as (overly)intensive forest management pas-ture and grazing may be regionallyimportant as well Di rect human-related factors may be crucial inchanging the species compositionand structure of mountain forestsespecially in treeline ecotones whereclimate change does not representthe key factor (Alados et al 2014)
The aims of this study were to eval-uate the effects of geographical posi-tion and different regional climaticand other site conditions on treelineecotone characteristics of selectedEuropean mountain areas to identifythe natural and anthropogenic driv-ers of treeline shift We further dis-cuss the effects of climate and landuse changes on biodiversity in forestsbelow the treeline
2 METHODS
To demonstrate and discuss thepossibilities and limits of treeline
species shift including latitudinal and longitudinalheterogeneity the current situation in 11 mountainareas from Italy to Norway and from Spain to Bul-garia were analysed (Fig 1) To determine the influ-ence of mountain massif size on the mesoclimatic andvegetation conditions including treeline formationwe included large mountain complexes (PyreneesAlps Scandes) as well as relatively small separatedmountains (Pirin and Giant Mountains) in this studyThe characteristics of the treeline ecotone includinghistorical and recent human im pacts on mountainecosystems for the 11 mountain areas of interest aresummarized in Table 1 Extensive mountain massifs(Pyrenees Alps Scandes) were further divided intomore homogeneous units (Central Pyrenees EasternPyrenees Central Alps Eastern Alps Hardan-gervidda Dovre Northern Swedish Lapland InnerFinnmarknorthernmost Finnish Lapland) To esti-
137
Fig 1 Selected European mountains Grey shading differentiates mountainareas 1 Central Pyrenees (CP) 2 Eastern Pyrenees (EP) 3 Apennines (AP) 4Shara Mts (SH) 5 Pirin (PI) 6 Central Balkan Mts (CB) 7 Northern DinaricMts (DI) 8 Central Alps (CA) 9 Eastern Alps (EA) 10 Low Tatra Mts (LT) 11Giant Mts (GM) 12 Hardangervidda (HG) 13 Dovre (DO) 14 NorthernSwedish Lapland (NS) 15 Inner Finnmarknorthernmost Finnish Lapland (FL)
Clim Res 73 135ndash150 2017138
Country Coordi- Elevation Elevation Range of Annual Area VZs near Tr part of nates of range range of Tr tree species altitudinal of study (m asl)mountains central site of T (m asl) limit shift (km2)
(m asl) (m asl) of T or Tr (m yrminus1)
Spain 42deg37rsquoN 1930 No data Pu 2700 049 (T 15 800 Pine with Aa Spanish 0deg27rsquoE (1800minus2050) 1956minus2006) Ps and Fs (1400minus2050) Pyrenees subalpine with Pu
(1800minus2300)
Italy 42deg05rsquoN 1800 No data Fs 2100 1 (Tr [Pm ] 741 Beech forest (800minus1850) Majella Massif 14deg05rsquoE (1700minus1850) Pm 2300 1954minus2007) mixed beechminusdwarf pine part of (1800minus2000) dwarf pine Apennines (2100minus2300)
Macedonia 41deg47rsquorsquoN 1850 1900 Pa 2100 ~1 (T ) 829 Mixed firminusbeechminusShara Mountain 20deg33rsquoE (1800minus1900) (1850minus1950) Pp 2200 1934minus2010 spruce (800minus1900) (Balkan beech (1600minus1900) Peninsula) dwarf pine (1600minus2200)
Bulgaria 41deg42rsquoN 1900 2100 Pp 2500 No data 384 Beech (1000minus1500) Pirin 23deg31rsquoE (1800minus2100) (1900minus2300) coniferous forest (Ps Pp
Ph Pa 1300minus2300) dwarf pine (2000minus2500)
Bulgaria 42deg47rsquoN Beech T 1600 1600 Fs 1700 No data 717 Beech (800minus1700) Central 24deg36rsquoE (1500minus1700) (1500ndash1850) Pa Aa 1850 coniferous (Pa Aa Balkan Mts 1500minus1850) juniper
(1500minus1850)
Slovenia 45deg36rsquoN 1540 1600 Fs 1650 No data ~200 Mixed firminusbeechminusspruce Northern 14deg28rsquoE (1455minus1600) (1550minus1650) Pa 1700 (600minus1400) beech (1300minus1550) Dinaric Mts Aa 1650 dwarf pine with small groups
or individual trees of Fs Pa Aa Ap (1500minus1700)
Switzer- 46deg46rsquoN 1900 2100 Pa 2400 ca13 (T ~250 Spruce with Ld land Italy 9deg52rsquoE (1700minus2150) (1850minus2300) Ld 2450 1954minuspresent) + ~750 (1000minus1900) forest with Pc Central and 46deg17rsquoN Ld and Pa (1700minus2200) Eastern Alps 11deg45rsquoE dwarf shrub vegetation partly
pastures (2200minus2500)
Table 1 Timberline and treeline parameters and characteristics of human activities changes according to climatic models andlimits to climate change induced species shifts in selected European mountains A1B A2 emissions scenarios (IPPC) AaAbies alba Ap Acer pseudoplatanus Fs Fagus sylvatica Ld Larix decidua Pa Picea abies Pc Pinus cembra Ph Pi-nus heldreichii Pm Pinus mugo Pp Pinus peuce Ps Pinus sylvestris Pu Pinus uncinata T timberline Tr treeline TSL
tree species limit VZ vegetation zone C century mng management Elevation range of T and Tr mean (minminusmax)
Cudliacuten et al Drivers of treeline shift 139
Human Annual Limits Referencesinfluence temperature (T ) to climate
in VZs in past and precipitation (P) change and recently differences induced species
between 1961minus1990 shiftand 2021minus2050
mean of CORDEX models
Since 11th C HIRHAM5 model Inadequacy of Batllori et al (2009) clear-cutting and grazing period ΔT2021minus2050 substrata (rocky) Alados et al (2014)
moderate on irregular = 15minus2degC ΔT2051minus2080 climatic constraints Gartzia et a (2014) slopes and ridges since 1930 = 4degC ΔP2021minus2050 (including growing- Camarero et al (2015a)significant release of influence = +5 ΔP2051minus2080 season extreme events
but recovery mainly at = +30 compared to such as late-spring moderate elevations 1960minus1990 freezing summer
drought etc)
Since 1000 BC HadCM3_A2 period Low density of Fs van Gils et al (2008) cutting burning and 2020minus2080 in T dispersal distances Palombo et al (2013 2014)
grazing mng since the mid-20th ΔTmin_Jan2050 = 17degC of Fs seeds upslope C grazing intensity decreased ΔTmin_Jan2080 = 31degC poor soils excessive
now occasionally managed ΔP2050 = minus7 exposure to solar radiation and regeneration of Pm ΔP2080 = minus23 in open areas
Since 14th C clear-cutting Mean of 4 GCMs Differences in soil types Em (1986) and grazing to enlarge (CSIROMk2 HadCM3 between subalpine beech Strid et al (2003)
alpine pastures and to produce ECHAM4OPYC3 and mixed forest (poor Bergant (2006) charcoal since 1970 NCAR minus PCM) ΔT2050 = calcareous soils or rankers on Amidžic et al (2012)
abandonment of traditional 26degC ΔT2100 = 53degC silicate bedrock) cliffs agricultural practices ΔP2050 = minus3 ΔP2100 = minus8 and rocks cattle grazing
Since 680 AD intensive HADCM3_A2 Rocky sites extreme Velchev (1997) fires deforestation and grazing CGGM2 hydrothermal conditions Grunewald amp
since 1962 traditional mng ΔT2050 = 17degC competition of shrubs and tree Scheithauer (2011) has changed last decades tourism ΔT2080 = 28degC seedlings locally intense Raev et al (2011)
impact has increased ΔP2080 = minus15 pasturing
Human impact (mostly burning) HADCM3_A2 Low density of Fs Raev et al (2011) started in 17th C excessive CGCM2 ΔT2050 = 17degC in T dispersal distances Gikov et al (2016) and improper forest mng ΔT2080 = 28degC of Fs seeds upslope pasture Zhiyanski et al
release of influence since 1980 ΔP2050 = minus15 intensity new (2008 2016)ΔP2080 = minus15 expansion of juniper stands
15minus18th C slash burn DMI-HIRHAM5_A1B Impact of wild large Klopltilt et al grazing mng 18minus19th C period 2001minus2100 ΔT2050 ungulates on tree seedlings (2010 2015)
uncontrolled cutting 20th C = 12degC ΔT2100 = 25degC (especially Aa) shallow Mina et al (2017)irregular shelterwood systems ΔP2050 = +15 mm nutrient-poor soils
and over-exploitation mng ΔP2100 = minus18 mm (ie ranker and rendzina)
Since 12th C slash burn grazing mng ENSEMBLES models_ Frequent disturbances Tattoni et al (2010) 13minus19th C intensive forest A1B period 1980minus2100 by snow avalanches Kulakowski et al (2011)
mng at the end of the 19th C ΔT2050 = 12minus18degC and snow gliding ongoing Barbeito et al (2012) decrease in grazing ΔT2100 = 27minus41degC cattle grazing game grazing Bebi et al (2012)
ΔP2050 = 0 mm in the coldest years Dawes et al (2015)ΔP2100 = minus5 mm
Continued on next page
Clim Res 73 135ndash150 2017
mate the geographical position of these mountainunits 1 value was used for the geographical latitudeand longitude of the locus of the studied mountainunits The analysed parameters of these 15 mountainunits are shown in Table 2
In addition we calculated the following climatecharacteristics for all mountain units annual meanair temperature annual sum of precipitation thegrowing season length and the date of the onset ofthe growing season These characteristics were thencompared for 2 distinct periods 1961minus1990 and1991minus 2015 (Table 3) The climate characteristicswere derived from the E-OBS gridded dataset of sta-tion observations version 131 (Haylock et al 2008)We used the E-OBS version on the regular longi-tudeminuslatitude grid with a horizontal resolution of 025degrees The climate data from all grid points withinthe mountain units or near their geographical bor-ders were considered
Data processing for 15 mountain units was made onthe basis of data obtained from the literature and
personal studies by the co-authors these are sum ma -rized in Tables 1 amp 2 and in Tables S1 amp S2 in theSupplement at www int-res com articles suppl c073p135 _ supp pdf In addition a semi-quantitative valu-ation of abiotic biotic and anthropogenic factors lim-iting an upward treeline shift was performed (Fig 2)To answer the question how natural and anthro-pogenic factors have influenced treeline ecotonecharacteristics with a focus on treeline shift redun-dancy analysis (RDA) was used to describe and testthe effect of the explanatory variables (geographicalposition size of the whole mountain massif start ofhuman influence and start of the decrease in humaninfluence) on treeline ecotone characteristics (tim-berline elevation width of the treeline ecotone treespecies forming the treeline and treeline shift peryear) and identify groups of mountain units with sim-ilar variability of the dependent data (ter Braak ampŠmilauer 2012) The whole data set for this analysis isshown in Table S2 We tested both their simple ef -fects which show how much variation every ex plan -
140
Country Coordi- Elevation Elevation Range of Annual Area VZs near Tr part of nates of range range of Tr tree species altitudinal of study (m asl)mountains central site of T (m asl) limit shift (km2)
(m asl) (m asl) of T or Tr (m yrminus1)
Slovakia 48deg55rsquoN 1400 1530 Pa 1730 ~03 (Tr ~400 Spruceminusfirminusbeech Dumbier Tatra 19deg31rsquoE (1350minus1450) (1400minus1630) 1950minuspresent) (1200minus1350) spruce (1300minus(part of Low 1550) Pm with individual Tatra Mts) Pa trees (1400minus1700)
Czech Republic 50deg42rsquoN 1250 1320 Pa 1500 043 (Tr 550 Beechminusspruce (700minus1050) Giant Mts 15deg38rsquoE (1130minus1460) (1250minus1460) 1936minus2005) Pa dominates with increasing
altitude spruce (1000minus1400) Pm with Pa trees (1400minus1560)
Norway 60deg10rsquoN 850 600minus1050 1500 08 (Tr 40 600 Birch forest dominance Southern 7deg40rsquoE 62deg20rsquoN 1915minus2007) at high altitudes but with Scandes 10deg05rsquoE some scattered Ps or Pa
mixed birchminusconifer below 800
Norway 68deg20rsquoN West 650 West 750 West 1000 06 (Tr 35 000 Birch forest dominance Sweden 18deg20rsquoE East 100 East 290 East 0 1958minus2008) at all altitudes but with some Finland 69deg50rsquoN scattered pines or pine stands Northern 27deg00rsquoE on sandy soilsScandes
Table 1 (continued)
Cudliacuten et al Drivers of treeline shift
atory variable can explain separately without usingthe other variables and their conditional ef fectswhich depend on the variables already selected inthe model The statistical significance of the expla -natory variables was tested using a Monte Carlo permutation test and only predictors with p le 005were included in the subsequent RDA (Šmilauer ampLepš 2014)
To find the drivers of biodiversity loss in forestsmeadows and animal communities (dependent vari-ables) the explanatory variables (geographical posi-tion timberline elevation tree species forming thetreeline start of human influence start of the de -crease in human influence and temperature increasebetween the 2 periods 1961minus1990 and 2021minus2050)were tested again by RDA in Canoco 5 as mentionedabove (Figs 3 amp 4) The whole data set for this analy-sis mdash including additional information about tree-lines in the mountain areas of interest with a focuson treeline shift (its rate drivers limits and problemswith its assessment) mdash is presented in Tables S1 amp S2
Selected univariate graphs were constructed to visu-alize the data of relationships between the explana-tory and dependent variables of all analyses whichwere not seen distinctly from the results of the multi-variate statistics or linear regression (Fig 5)
3 RESULTS
31 Effect of environmental variables on treeline ecotone characteristics with a focus
on treeline shift
The variation in the dependent variables (timber-line elevation width of treeline ecotone tree speciesforming the treeline and treeline shift per year) wassignificantly affected only by latitude and size of thewhole mountain massif (Fig 3) The adjusted ex -plained variability by all explanatory variables was342 (F = 46 p = 001) The RDA diagram showsthat in the mountains located more to the north the
141
Human Annual Limits Referencesinfluence temperature (T ) to climate
in VZs in past and precipitation (P) change and recently differences induced species
between 1961minus1990 shiftand 2021minus2050
mean of CORDEX models
From 14th C to 1922 Average of 10 RCM_B1 More rocky and nutrient- Koumlrner (2003) mining forest change ΔT2050 = 18degC poor soils late frosts Fridley et al (2011)
to spruce monocultures ΔT2100 = 37degC steep slopes with avalanches Hlaacutesny et al (2011 2016)from 13th C to 1978 pastures ΔP2050 = +24 mm absence of mature trees
in upper parts of mountain ΔP2100 = minus35 mm
9minus11th C forest clear cutting ALADIN_A1B period 1961minus Different soil types Treml and Banaš (2000) grazing in dwarf pine VZ 2100 ΔT2050 = 13degC for Fs and Pa (cambisol Cudliacuten et al (2013)
since 18th C forest change ΔT2100 = 34degC versus podzol) only TSL of Pa Treml amp Chuman (2015) to spruce monocultures ΔP2050 = +25 mm could elevate on ranker Treml amp Migo (2015)
since 19th C artificially planted ΔP2100 = minus14 mm soils in Pm VZ
Grazed and browsed RegClimMPIHadley Sheep and reindeer Dalen amp Hofgaard (2005) by reindeer and livestock ΔT2050 = 12degC grazing and episodic Wielgolaski (2005)
(sheep and cattle) ΔT2100 = 22degC insect outbreaks (Epirrita) Hofgaard et al (2009) over thousands of years ΔP2050 = +23 mm Kullman Oumlberg (2009)
ΔP2100 = minus12 mm
Grazed and browsed RegClimMPIHadley Continued grazing regime Dalen amp Hofgaard (2005) by semi-domestic reindeer ΔT2050 = 16degC and frequent episodic Wielgolaski (2005)
for hundreds of years ΔT2100 = 29degC insect outbreaks (Epirrita) Toslashmmervik et al (2005) increased grazing from ΔP2050 = +18 mm Hofgaard et al (2009)
year 1960 ΔP2100 = minus10 mm Aune et al (2011) Mathisen et al (2014)
Clim Res 73 135ndash150 2017
treeline ecotone is narrower treeline shiftis smaller the timberline occurs at lowerelevations and the treeline is formed moreby broadleaved species Higher values oftreeline ecotone width and treeline shiftwere found in the mountains with a greatersize of the whole mountain massif
When every explanatory variable wastested separately values of the start of thedecrease of human influence in the moun-tains also had a significant effect on thevariation of the tree line ecotone data(explained variation = 206 pseudo-F =34 p = 0024) Some selected relationshipsof the explanatory and dependent vari-ables even those not showing significanteffects on the variation in the tree line eco-tone parameters are depicted in Fig 5
32 Effect of recent changes in climateand land use on the biodiversity
Biodiversity loss in forests meadows andanimal communities analysed by RDAwas explained by geo graphical positiontreeline species composition and tempera-ture increase between the 2 periods1961minus1990 and 2021minus2050 (Fig 4) Theadjusted ex plained variation by all ex pla -natory variables was 632 (F = 58 p =0001) The RDA results indicated that bio-diversity loss in forest communities in crea -sed with increasing latitude and longitudeas well as in the case where broad leavedspecies formed the treeline In contrast thehighest biodiversity loss in meadows wasfound in the mountains positioned more tothe south and with higher temperatureincrease The biodiversity loss in animalcommunities was negatively correlatedwith longitude and temperature increase
4 DISCUSSION
The comparison of several mountainareas situated across Europe shows a varia-tion in understanding of what constitutesthe treeline across countries The primaryissue is the different approach to the defini-tion of forest when applying a tree heightthreshold This threshold decreases from
142
Mou
nta
in u
nit
Nor
th
Eas
t M
oun
-T
imb
er-
Wid
th
Tre
e T
emp
erat
ure
T
reel
ine
Sta
rt
Sta
rt
lati
tud
elo
ng
itu
de
tain
lin
eof
spec
ies
incr
ease
sh
ift
of H
I of
HI
size
(m a
sl
)tr
eeli
ne
form
ing
b
etw
een
(m
yrminus
1 )(y
r)d
ecre
ase
(km
2 )ec
oton
eth
e 19
61minus
1990
(y
r)(m
)tr
eeli
ne
and
199
1minus20
15 (
degC)
Cen
tral
Pyr
enee
s (C
P)
42deg2
1rsquoN
0deg30
rsquoW55
000
2250
300
C0
81
92 (
1956
minus20
06)
1000
BC
1930
Eas
tern
Pyr
enee
s (E
P)
42deg2
8rsquoN
1deg30
rsquoE55
000
2400
300
C0
80
71 (
1956
minus20
06)
1000
1930
Ap
enn
ines
(A
P)
42deg0
5rsquoN
14deg0
5rsquoE
5000
018
0030
0B
14
1 (1
954minus
2007
)10
00 B
C19
50S
har
a M
ts (
SH
)41
deg47rsquo
N20
deg33rsquo
E16
0018
5035
0C
B0
71
(193
4minus20
10)
1300
1970
Pir
in (
PI)
41deg4
2rsquoN
23deg3
1rsquoE
2585
1900
400
C0
5n
o d
ata
680
1962
Cen
tral
Bal
kan
Mts
(C
B)
42deg4
7primeN
24deg3
6primeE
1160
016
0025
0C
B0
7n
o d
ata
1700
1980
Nor
ther
n D
inar
ic M
ts (
DI)
45deg3
6rsquoN
14deg2
8rsquoE
1000
1540
160
B1
3n
o d
ata
1400
1900
Cen
tral
Alp
s (C
A)
46deg4
6rsquoN
9deg52
rsquoE18
000
021
0055
0C
07
13
(195
4minus20
15)
1300
1880
Eas
tern
Alp
s (E
A)
46deg1
7rsquoN
11deg4
5rsquoE
180
000
1800
400
C0
91
3 (1
954minus
2015
)11
5019
50L
ow T
atra
Mts
(LT
)48
deg55rsquo
N19
deg31rsquo
E12
4014
0033
0C
09
057
(19
75minus
2016
)14
0019
22G
ian
t M
ts (
GM
)50
deg42rsquo
N15
deg38rsquo
E63
012
5025
0C
07
043
(19
36minus
2005
)10
0019
63H
ard
ang
ervi
dd
a (H
G)
60deg0
3rsquoN
7deg47
rsquoE90
0090
015
0C
B0
50
8 (1
915minus
2007
)80
019
70D
ovre
(D
O)
62deg2
1rsquoN
8deg52
rsquoE60
0010
0040
0B
07
08
(191
5minus20
07)
1000
no
dec
reas
eN
orth
ern
Sw
edis
h L
apla
nd
(N
S)
68deg1
2rsquoN
18deg3
3rsquoE
4000
700
300
B1
10
6 (1
958minus
2008
)10
00n
o d
ecre
ase
Inn
er F
inn
mar
kn
orth
ern
mos
t 69
deg10rsquo
N25
deg12rsquo
E60
000
290
90B
10
06
(195
8minus20
08)
1600
no
dec
reas
eF
inn
ish
Lap
lan
d (
FL
)
Tab
le 2
Fac
tors
infl
uen
cin
g t
reel
ine
ecot
one
char
acte
rist
ics
in s
elec
ted
Eu
rop
ean
mou
nta
ins
Th
e m
oun
tain
siz
e is
th
e si
ze o
f th
e w
hol
e m
oun
tain
mas
sive
fu
nct
ion
ing
as a
mez
osca
le c
lim
ate
un
it n
ot o
nly
the
stu
die
d m
oun
tain
are
a W
idth
of t
he
tree
lin
e ec
oton
e w
as d
eter
min
ed b
y su
btr
acti
ng
the
valu
e of
the
mea
n ti
mb
erli
ne
elev
atio
n
from
th
e va
lue
of t
he
up
per
tre
e sp
ecie
s li
mit
C c
onif
ers
B b
road
leav
ed t
rees
HI
hu
man
infl
uen
ce
Cudliacuten et al Drivers of treeline shift
5 m (Jeniacutek amp Lokvenc 1962) to 3 m(Koumlrner 2012) in Northern andCentral Europe to only 2 m (Holt-meier 2009) in Central and South-ern Europe where even shrub -by stands (eg Pinus mugo) areconsi dered as a forest in somecountries eg in Spain Italy andMa ce donia (Batllori et al 2009)An other problem is a different de -signation of forest stands belowthe treeline lsquoupper montane fo -restsrsquo in Central Europe and lsquosub-alpine forestsrsquo in Southern Europe(Ellenberg 1988) Further differen -ces are related to the tradition ofdifferent branches of science (ega more traditional lsquogeobotanicalrsquoapproach in Central Europe ver-sus a more experimental approachin Western Europe) especially inthe rate of applying new pro -gressive methods (eg climaticmodelling or molecular biologicalmethods)
Our comparison of selectedregions (Tables 1 amp 2 Figs 1 amp 2)showed that southern mountainscompared to those located morecentrally and in the north had (1) alonger and more profound exploi -tation by humans in the past (2)greater differences in climatic para -meters (temperature length ofgrowing period) between the 2periods 1961minus1990 and 1991minus2015(Table 3) and (3) more dramaticclimate change scenarios espe-cially concerning temperature in -creases Longitude also had someinfluence on the start and intensityof human influence and on tree-line elevation and rate of treelineshift (Table 2 Fig 5) The occur-rence of broadleaf tree species inthe treeline ecotone in the north-ern (and to some extent the south-ern) countries distinguishes thesefrom the Central European coun-tries where conifers prevail It isinteresting that grazing an im -portant factor shaping treelineecosystems (Dirnboumlck et al 2003
143
Length Beginning Mean Mean of growing of growing annual annual
period season temperature precipitation (d)a (d)b (degC) ()
Central Pyrenees 135 minus89 08 minus107Eastern Pyrenees 117 minus91 08 minus92Apennines 284 minus232 14 57Shara Mts 16 01 07 minus100Pirin 26 minus14 05 27Central Balkan Mts 169 minus136 07 76Northern Dinaric Mts 160 minus95 13 minus56Central Alps 147 minus86 07 62Eastern Alps 125 minus96 09 minus33Low Tatra Mts 24 minus34 09 minus91Giant Mts minus10 13 07 55Hardangervidda Blefjell 63 34 05 56Dovre 138 32 07 64Northern Swedish Lapland 81 05 11 minus27Inner Finnmarknorthern- 17 08 10 77most Finnish Lapland
aPositive numbers indicate a prolongation bNegative numbers indicate an earlier beginning
Table 3 Differences of selected climatic parameters between the 2 study periods (1961minus1990 and 1991minus2015) in selected European mountains
Tein Drou Gefa Nadi Laus Abli Bili Huli
0 1 2 3
Central Pyrenees (CP)
Eastern Pyrenees (EP)
Apennines (AP)
Shara Mts (SH)
Pirin (PI)
Central Balkan Mts (CB)
Northern Dinaric Mts (DI)
Central Alps (CA)
Eastern Alps (EA)
Low Tatra Mts (LT)
Giant Mts (GM)
Hardangervidda (HG)
Dovre (DO)
Northern Swedish Lapland (NS)
Inner FinnmarknorthernmostFinnish Lapland (FL)
Fig 2 Rates of treeline drivers (temperature increase [Tein] land use change[Laus]) and treeline shift limits (drought [Drou] geomorphological factors [Gefa]natural disturbances [Nadi] other abiotic biotic and human factors) in selectedEuropean mountains Abli Bili and Huli abiotic biotic and human treeline shiftlimits respectively) Rate of treeline driver influence mdash 0 no influence 1 weak
influence 2 middle influence 3 strong influence
Clim Res 73 135ndash150 2017
Gehrig-Fasel et al 2007) limits treeline shift in halfof the countries of interest regardless of latitude lon-gitude or recent political developments (Shara PirinCentral Balkans Alps Scandes Fig 2)
Despite the stated differences between studieslooking at climate change impacts on the treelineit is clear that current and future changes in temp -erature will seriously affect treeline ecotones in allmountain ranges of Europe Winter is a period whentreeline stands and individual trees have to survivesevere life-limiting conditions (Wieser amp Tausz2007) Therefore winter warming might increase thechance of young trees surviving and passing the mostcritical period from seedling to sapling and further tothe mature stage (Koumlrner 2003) There is alreadymuch evidence worldwide that winter warming isone of the significant drivers of treeline advance(Harsch et al 2009) Although it is expected that treegrowth at the upper distributional margins (and inthe northern European countries) will increase due toongoing climate change (Pentildeuelas amp Boada 2003Chen et al 2011 Hlaacutesny et al 2011 Lindner et al2014) extremes in temperature (eg black froststemperature reversals in the spring) or winter precip-itation (eg lack of snow drought) might also limit
this process in the future in some regions and localsituations (Holtmeier amp Broll 2005 Hagedorn et al2014) An earlier start to the growing period (espe-cially in the Apennines Central Balkan and SpanishPyrenees see Table 2) can play a negative role in theresistance of trees and seedlings to early springfrosts On the other hand at lower distributional mar-gins (and in Southern European countries) it isexpected that tree growth will decrease or forestswill experience some level of dieback due to drought(Breshears et al 2005 Piovesan et al 2008 Allen etal 2010 Huber et al 2013) A serious implication isthe positive effect that warming can have on root rotfungi and populations of bark beetles resulting inlarge-scale disturbances to temperate and also high-altitude forests (Jankovskyacute et al 2004 Elkin et al2013 Millar amp Stephenson 2015)
Although recent upward advances of treeline eco-tones are widespread in mountain regions (Harsch et
144
Fig 3 Redundancy analysis diagram with variation in tim-berline elevation width of treeline ecotone the tree speciesforming the treeline and treeline shift used as dependentvariables explained by explanatory variables (latitudemountain size) Mountain units (blue) are projected as thecentres of abbreviations (see Table 2) Explanatory variablesaccount for 436 of the total variation in the dependentdata The first canonical axis explained 466 of variationthe second axis explained 30 of variation Dependentvariables (black) are Ecot altitudinal width of treeline eco-tone TLShift treeline shift TrspC (TrspB) treeline formedby conifers (broadleaved trees) Timb timberline elevationExplanatory variables (red) are N north latitude Mtsize
size of the mountain massif
Fig 4 Redundancy analysis diagram with variation in bio-diversity loss in forests meadows and animal communitiesused as dependent variables explained by explanatory vari-ables (geographical position tree species forming the tree-line and temperature increase between the 2 study periods[1961minus1990 and 2021minus2050]) Mountain units (green) areprojected as centres of abbreviations (see Table 2) Explana-tory variables account for 763 of total variation in the de-pendent data The first canonical axis explained 484 ofvariation the second axis explained 183 of variation De-pendent variables (black) are BdfoBdmeBdan biodiversityloss in forestsmeadowsanimal communities Explanatoryvariables (red) are N north latitude E east longitude TrspC(TrspB) treeline formed by conifers (broadleaved trees)Temp temperature increase between the 2 study periods
Cudliacuten et al Drivers of treeline shift
al 2009) only a few published studies have includedquantitative data about treeline shifts (Devi et al2008 Kullman amp Oumlberg 2009 Diaz-Varela et al 2010Van Bogaert et al 2011) Additionally our know -ledge of the spatial patterns in treeline ecotone shiftsat the landscape scale is still surprisingly poor exceptfor some studies from subarctic areas the UralMountains and the Alps (Lloyd et al 2002 Diaz-Varela et al 2010 Hagedorn et al 2014) In the 15studied mountain units the values of treeline shiftranged from 043 to 19 m yrminus1 showing a rather dis-tinct spatial pattern of the dynamics within Europeanmountains The observed treeline shift had a signifi-cant positive relationship only with northern latitudeand a weak positive relationship with mountain-range size (ie the size of the whole mountain massifFig 3) Nevertheless this illustrates the importanceof the mass elevation effect (Koumlrner 2012) and par-tially ex plains why treelines could be at different alti-tudes at the same latitude Small negative regres-sions with east longitude and timberline elevationare shown in Fig 5 There was no apparent depend-ence of the treeline shift rate on climatic parameterchanges between the periods 1961minus1990 and1991minus2015
A whole forest vegetation zone shift is a complexprocess as not only the life strategies of an individualtree need to be considered but plant animal andmicroorganism species and their interrelationshipsas well as the relationships to their specific microhab-
itats (including soil conditions) are also involved(Urban et al 2012) Across all regions we identifiedseveral important obstacles to treeline shifts oftenrelated to site properties such as significant rocki-ness having shallow or low-nutrient soils extremerelief causing disturbances by avalanches snow glid-ing or wind damage (data from the Pyrenees Apen-nines Shara Northern Dinaric Low Tatra and GiantMountains see Fig 2) Soil heterogeneity certainlyhas an important role in plant responses to climatechange and could also maintain the resilience of thecommunity assemblages (Fridley et al 2011) An -other group of obstacles includes extreme climaticparameters especially in winter (reported eg fromthe Pyrenees Apennines Pirin Central BalkanNorthern Dinaric and Giant Mountains Fig 2) Theircombination can result in edaphic andor climaticunsuitability of habitats where species could poten-tially migrate Climatically and edaphically suitablesites will likely decline over the next century partic-ularly in mountain landscapes (Bell et al 2014)There fore trees in the treeline ecotone will colonizepreviously forested habitats The colonization suc-cess of individual forest communities is affected bydifferences in species dispersal and recruitmentbehaviour (Dullin ger et al 2004 Jonaacutešovaacute et al2010) For these reasons colonization of new foresthabitats by the next tree generation may not be suc-cessful and may result in loss of species diversity(Honnay et al 2002 Ibaacutentildeez et al 2009 Dobrowski et
145
Fig 5 Univariate graphs of the relationships of dependent variables viz timberline (Timb) and treeline shift per year (Shift) and their explanatory variables Nola north latitude Ealo east longitude DD decimal degrees
Clim Res 73 135ndash150 2017
al 2015) as reported from Macedonia Slovenia theCzech Republic Slovakia and Norway (Fig 4)
Another limiting factor is the existence of strongadaptation mechanisms of the dominant tree specieswhich allow them to survive in their current distribu-tion areas Species migration is likely to be slow dueto the limited quantity of climatically and edaphicallysuitable sites and the slow velocity of seed dispersaland tree regeneration and will be hampered evenmore by fragmentation of high mountain landscapescaused by human activities including brush invasionin abandoned meadows (Gartzia et al 2014) A verti-cal shift in a vegetation zone involves not only singlespecies of plants animals and microorganisms butalso their interrelationships and links to soil condi-tions The speed at which climatic conditions changewill be different from changes in the soil conditionsdue to the slowness of soil-forming processes Soiltypes and conditions (cambisol versus podzol) arecrucial obstacles for the shift from a beech forest zoneinto the spruce forest zone in the Czech Republic(Vacek amp Mat jka 2010) According to other sourcesthe dominant tree species can influence soil for -mation processes (especially humus forms) Underfavourable climatic and orographic conditions beechis able to change its soil conditions over decades tocenturies (J Macku unpubl data)
Significant changes in biodiversity must be expec -ted in all of the mountain areas of interest We foundthat biodiversity declined particularly in the southernregions of Europe where the timberline is situated athigher elevations and human impact is mostlylonger-lasting (Fig 4) Similarly Pauli et al (2012)reported an increase in species richness of mountaingrasslands across Europe except for Mediterraneanregions and assigned this different response of thesouthern regions to decreased water availability Onthe other hand ecosystems which exhibited bio -diversity increases were not only enriched by mig -rant species but the assemblages underwent thermo -philization ie cold-adapted species declined andmore warm-adapted species spread (Gottfried et al2012) However not all species are able to track cli-mate changes It is expected that around 40 ofhabitats will become climatically unsuitable for manymountain species particularly endemic ones in -creasing their extinction probability during this cen-tury (Dullinger et al 2012) The abandonment of tra-ditional land use forms is another source of this losswhich may be a more important driver than tempera-ture increase (Fig 4) Serious biodiversity losses inmountain meadows were reported from Spain ItalyMacedonia Bulgaria Slovenia and Slovakia There-
fore controlled grazing must occur in order to main-tain alpine grasslands (Dirnboumlck et al 2003) Unfor-tunately grazing is still active only in smaller parts ofEuropean mountains (eg in the Central Alps and theCentral Balkan Mountains but pasturing can alsonegatively affect vegetation diversity) and some-times does not serve to maintain alpine grasslandThe same is true for grassland management whichcan help to maintain mountain species and decreasehabitat loss A serious biodiversity loss in mountainmeadows was reported from the Pyrenees the Apen-nines and the Shara Pirin Central Balkan NorthernDinaric and Low Tatra Mountains (Fig 4)
In forests species responses to climate change maybe equivocal some recent studies indicated contra-dictory shifts in species distributions (eg Lenoir etal 2008 Zhu et al 2012 Rabasa et al 2013) This dis-parity is widely discussed and assigned to the greatvariety of non-climatic factors or even tree ontogeny(Grytnes et al 2014 Lenoir amp Svenning 2015 Maacutelišet al 2016) Disturbances leading to tree mortalitymay also play an important role (Cudliacuten et al 2013)changes in tree canopy cover modify light availabil-ity and microclimate and can induce the thermo -philization of forest vegetation (De Frenne et al2015 Stevens et al 2015) The loss of this microcli-mate buffering effect of forests may induce a biotichomogenization of forests (Savage amp Vellend 2015)or the creation of novel non-analogical communitiessuch as oakminuspine forests (Urban et al 2012) whichmay be a new threat to forest biodiversity
The observed changes in treeline forest ecosystemsare often related to changes in land-use intensity(Theurillat amp Guisan 2001 Alados et al 2014) Ac -cording to climate change predictions and the recentand future exploitation intensity of European moun-tains trees and forest communities will shift upwarddue to land use change or climate change or bothReduced land-use intensity certainly will interactwith climate change by facilitating or inhibiting spe-cies occurrence and will accelerate forest expansionabove the present treeline particularly to previouslyforested habitats (Theurillat amp Guisan 2001) Thesimultaneous action of both main treeline shift driv-ers viz temperature increase and decrease in landuse intensity was recorded from all of our studiedmountain areas The extensive differences in timber-line and treeline elevations in almost all studiedmountains (Table 1) indicate the anthropo-zoogenictreeline type (according to Ellenberg 1998) In mostmountains (eg in the Apennines Shara MountainsCentral Balkans and Alps) land use is the prevailingfactor influencing vegetation drift (Fig 5) Previous
146
Cudliacuten et al Drivers of treeline shift
research showed that the upward shift of the treelinein the Swiss Alps was predominantly attributable toland abandonment and only in some situations to cli-mate change (Gehrig-Fasel et al 2007) In the Apen-nines in the last few decades tree establishment hasbeen mainly controlled by land use while treegrowth has been controlled by climate pointing to aminor role played by climate in shaping the currenttreeline (Palombo et al 2013)
The impact of climate change and the connectedland use change on biodiversity and ecosystem serv-ices provision in several European countries is sum-marized by Wielgolaski et al (2017) and Kyriazopou-los et al (2017) (both this Special) and Sarkki et al(2016) One of the possible adaptive management op-tions in response to climate change assisted migra-tion as human-assisted movement of species hasbeen frequently debated in the last few years (Ste-Marie et al 2011) It is possible to apply it as a type ofassisted colonization ie intentional movement andrelease of an organism outside its indigenous range toavoid extinction of populations of the focal species(eg Macedonian pine species) or as an ecologicalre placement ie the intentional movement and re-lease of an organism outside its indigenous range toperform a specific ecological function (eg planting ofPinus mugo in the Alps IUCNSSC 2013)
5 CONCLUSIONS
The analysis of 11 mountain areas across Europeshowed that with increasing latitude the treelineand altitudinal width of the treeline ecotone signifi-cantly decreases as do the significance of climaticand soil parameters as barriers against tree speciesshift Although temperature is the overwhelmingcontrolling factor of tree growth and establishment intemperate and boreal treeline ecotones late-sea-sonal drought might also play a driving role in Medi-terranean treeline ecotones Longitude was lessinfluential mostly affecting climate and land usechange effects on increased biodiversity loss as wellas the size of the area that forms one mesoclimateunit affecting altitudinal ecotone width The biggestpart of the commonly observed remaining variabilityin mountain vegetation near the treeline in Europeseems to be caused by geomorphological geologicalpedological and microclimatic variability in combina-tion with different land use history and the presentsocio-economic relations The observed differencesin climatic parameters between the mountain areasof interest in comparison with the reference period
1960minus1990 (09degC) have not explained the relativelyhigh differences in the rate of treeline shift per year(149 m) The predicted variability in temperatureincrease due to global warming (12minus26degC in 2050and 22minus53degC in 2100) could lead to a much biggerdifferentiation in treeline ecotone biodiversity andecosystem processes between southern and northernEuropean mountains in the future Therefore thesedifferences must be taken into account by scientistsand EU policy makers when formulating efficientadaptive forest management strategies for treelineecosystems at the European level (eg assistedmigration of adapted genotypes)
Acknowledgements This paper is based firstly upon workfrom the COST Action ES 1203 SENSFOR supported byCOST (European Cooperation in Science and Technologywwwcosteu) This international work was further sup-ported by projects granted by the Ministry of EducationYouth and Sports of the Czech Republic grant NPU ILO1415 and LD 14039 by the agency APVV SR under pro-jects APVV-14-0086 and APVV-15-0270 We acknowledgethe E-OBS dataset from the EU-FP6 project ENSEMBLES(httpensembles-eumetofficecom) and the data providersin the ECAampD project (wwwecadeu)
LITERATURE CITED
Alados CL Errea P Gartzia M Saiz H Escoacutes J (2014) Posi-tive and negative feedbacks and free-scale pattern distri-bution in rural-population dynamics PLOS ONE 9 e114561
Allen CD Macalady AK Chenchouni H Bachelet D andothers (2010) A global overview of drought and heat-induced tree mortality reveals emerging climate changerisks for forests For Ecol Manag 259 660minus684
Amidžic L Bartula M Jaric S (2012) Syntaxonomy overviewof vegetation of Šar Planina in the Balkan PeninsulaVegetos 25 348minus360
Aune S Hofgaard A Soumlderstroumlm L (2011) Contrasting cli-mate and land use driven tree encroachment pattern ofsub-arctic tundra in Northern Norway and Kola Penin-sula Can J For Res 41 437minus449
Barbeito I Dawes M Rixen C Senn J Bebi P (2012) Factorsdriving mortality and growth at treeline a 30-year exper-iment of 92000 conifers Ecology 93 389minus401
Batllori E Camarero JJ Ninot JM Gutieacuterrez E (2009)Seedling recruitment survival and facilitation in alpinePinus uncinata treeline ecotones Implications andpotential responses to climate warming Glob Ecol Bio-geogr 18 460minus472
Bebi P Teich M Hagedorn F Zurbriggen N Brunner SHGreacutet-Regamey A (2012) Veraumlnderung von Wald undWaldleistungen in der Landschaft Davos im Zuge desKlimawandels Schweiz Z Forstwes 163 493minus501
Bell DM Bradford JB Lauenroth WK (2014) Mountain land-scapes offer few opportunities for high elevation treespecies migration Glob Change Biol 20 1441minus1451
Bergant K (2006) Climate change scenarios for MacedoniaUniversity of Nova Gorica Center for Atmospheric
147
Clim Res 73 135ndash150 2017
Research Nova GoricaBodin J Badeau V Bruno E Cluzeau C Moisselin JM
Walther GR Dupouey JL (2013) Shifts of forest speciesalong an elevational gradient in Southeast France cli-mate change or stand maturation J Veg Sci 24 269minus283
Boncina A (2011) History current status and future pros -pects of uneven-aged forest management in the Dinaricregion an overview Forestry 84 467minus478
Breshears DD Cobb NS Rich PM Price KP and others(2005) Regional vegetation die-off in response to global-change-type drought Proc Natl Acad Sci USA 102 15144minus15148
Camarero JJ Garciacutea-Ruiz JM Sanguumlesa-Barreda G GalvaacutenJD and others (2015a) Recent and intense dynamics in aformerly static Pyrenean treeline Arct Antarct Alp Res47 773minus783
Camarero JJ Gazol A Galvaacuten JD Sanguumlesa-Barreda GGutieacuterrez E (2015b) Disparate effects of global-changedrivers on mountain conifer forests warming-inducedgrowth enhancement in young trees vs CO2 fertilizationin old trees from wet sites Glob Change Biol 21 738minus749
Chen IC Hill JK Ohlemuumlller R Roy DB Thomas CD (2011)Rapid range shifts of species associated with high levelsof climate warming Science 333 1024minus1026
Cudliacuten P Sejaacutek J Pokornyacute J Albrechtovaacute J Bastian OMarek M (2013) Forest ecosystem services under climatechange and air pollution In Matyssek R Clarke N Cud-liacuten P Mikkelsen TN Tuovinen JP Wiesner G Paoletti E(eds) Climate change air pollution and global chal-lenges understanding and perspectives from forestresearch Developments in Environmental Science Vol13 Elsevier Oxford p 521minus546
Dalen L Hofgaard A (2005) Differential regional treelinedynamics in the Scandes mountains Arct Antarct AlpRes 37 284minus296
Dawes MA Philipson CD Fonti P Bebi P Haumlttenschwiler SHagedorn F Rixen C (2015) Soil warming and CO2
enrichment induce biomass shifts in alpine tree line veg-etation 2014 Glob Change Biol 21 2005minus2021
De Frenne P Rodriacuteguez-Saacutenchez F De Schrijver A CoomesDA Hermy M Vangansbeke P Verheyen K (2015) Lightaccelerates plant responses to warming Nat Plants 1 15110
Devi N Hagedorn F Moiseev P Bugmann H ShiyatovSG Mazepa V Rigling A (2008) Expanding forests andchanging growth forms of Siberian larch at the treeline ofthe Polar Urals during the 20th century Glob ChangeBiol 14 1581minus1591
Diaz-Varela R Colombo R Meroni M Calvo-Iglesias MSBuffoni A Tagliaferri A (2010) Spatio-temporal analysisof alpine ecotones A spatial explicit model targeting altitudinal vegetation shifts Ecol Model 221 621minus633
Dirnboumlck T Dullinger S Grabherr G (2003) A regionalimpact assessment of climate and land-use change onalpine vegetation J Biogeogr 30 401minus417
Dobrowski SZ Swanson AK Abatzoglou JT Holden ZASafford HD Schwartz MK Gavin DG (2015) Forest struc-ture and species traits mediate projected recruitmentdeclines in western US tree species Glob Ecol Biogeogr24 917minus927
Dullinger S Dirnboumlck S Grabherr G (2004) Modelling cli-mate change-driven treeline shifts relative effects oftemperature increase dispersal and invisibility J Ecol92 241minus252
Dullinger S Gattringer A Thuiller W Moser D and others
(2012) Extinction debt of high-mountain plants undertwenty-first-century climate change Nat Clim Chang 2 619minus622
Elkin C Gutierrez AG Leuzinger S Manusch C TemperliC Rasche L Bugmann H (2013) A 2degC warmer world isnot safe for ecosystem services in the European AlpsGlob Change Biol 19 1827minus1840
Ellenberg H (1988) Vegetation ecology of Central Europe4th edn Cambridge University Press Edinburgh
Em H (1986) Spruce southern distribution range Spruce for-est on Shara Mountain in Macedonia Contrib Macedon-ian Acad Sci Arts Skopje 5 11minus28
Fajardo A Piper FI Pfund L Koumlrner CH Hoch G (2012)Variation of mobile carbon reserves in trees at the alpinetreeline ecotone is under environmental control NewPhytol 195 794minus802
Fridley JD Grime JP Askew AP Moser B Stevens CJ (2011)Soil heterogeneity buffers community response to cli-mate change in species-rich grassland Glob Change Biol17 2002minus2011
Gartzia M Alados CL Peacuterez-Cabello F (2014) Assessment ofthe effects of biophysical and anthropogenic factors onwoody plant encroachment in dense and sparse moun-tain grasslands based on remote sensing data Prog PhysGeogr 38 201minus217
Gehrig-Fasel J Guisan A Zimmermann NE (2007) Tree lineshifts in the Swiss Alps climate change or land abandon-ment J Veg Sci 18 571minus582
Gikov A Dimitrov P Zhiyanski M (2016) Land cover dynam-ics on the northern slope of the Troyan Passage during 30years period Probl Geogr 1-2 78minus92
Gonzalez P Neilson RP Lenihan JM Drapek RJ (2010)Global patterns in the vulnerability of ecosystems to veg-etation shifts due to climate change Glob Ecol Biogeogr19 755minus768
Gottfried M Pauli H Futschik A Akhalkatsi M and others(2012) Continent-wide response of mountain vegetationto climate change Nat Clim Chang 2 111minus115
Grace J Berninger F Nagy L (2002) Impacts of climatechange on the tree line Ann Bot 90 537minus544
Grunewald K Scheithauer J (2011) Landscape developmentand climate change in Southwest Bulgaria (Pirin Mts)Springer Heidelberg
Grytnes JA Kapfer J Jurasinski G Birks HH and others(2014) Identifying the driving factors behind observedelevational range shifts on European mountains GlobEcol Biogeogr 23 876minus884
Hagedorn F Shiyatov FG Mazepa VS Devi NM and others(2014) Treeline advances along the Urals mountainrange mdash driven by improved winter conditions GlobChange Biol 20 3530minus3543
Hansen AJ Neilson RP Dale VH Flather CH and others(2001) Global change in forests responses of speciescommunities and biomes BioScience 51 765minus779
Harsch MA Bader MY (2011) Treeline formminusa potential keyto understanding treeline dynamics Glob Ecol Biogeogr20 582minus596
Harsch MA Hulme PE McGlone MS Duncan RP (2009) Aretreelines advancing A global meta-analysis of treelineresponse to climate warning Ecol Lett 12 1040minus1049
Haylock MR Hofstra N Klein Tank AMG Klok EJ JonesPD New M (2008) A European daily high-resolutiongridded dataset of surface temperature and precipita-tion J Geophys Res Atmos 113 D20119
Hlaacutesny T Barcza Z Fabrika M Balaacutezs B and others (2011)
148
Cudliacuten et al Drivers of treeline shift
Climate change impacts on growth and carbon balanceof forests in Central Europe Clim Res 47 219minus236
Hlaacutesny T Trombik J Dobor L Barcza Z Barka I (2016)Future climate of the Carpathians climate change hot-spots and implications for ecosystems Reg EnvironChange 16 1495minus1506
Hofgaard A Dalen L Hytteborn H (2009) Tree recruitmentabove the treeline and potential for climate driven tree-line change J Veg Sci 20 1133minus1144
Holtmeier FK (2009) Mountain timberlines Ecology patchi-ness and dynamics 2nd edn Advances in GlobalChange Research 36 Springer Science amp ScienceMediaBV Dordrecht
Holtmeier FK Broll G (2005) Sensitivity and response ofnorthern hemisphere altitudinal and polar treelines toenvironmental change at landscape and local scalesGlob Ecol Biogeogr 14 395minus410
Honnay O Verheyen K Butaye J Jacquemyn H Bossuyt BHermy M (2002) Possible effects of habitat fragmentationand climate change on the range of forest plant speciesEcol Lett 5 525minus530
Huber R Rigling A Bebi P Brand F and others (2013) Sus-tainable land use in mountain regions under globalchange synthesis across scales and disciplines Ecol Soc18 36
Ibaacutentildeez I Clark JS Dietze MC (2009) Estimating colonizationpotential of migrant tree species Glob Change Biol 15 1173minus1188
IUCNSSC (International Union for Conservation of NatureSpecies Survival Commission) (2013) Guidelines for re -introductions and other conservation translocations Version 10 IUCNSpecies Survival Commission Gland
Jankovskyacute L Cudliacuten P gtermaacutek P Moravec I (2004) The pre-diction of development of secondary Norway sprucestands under the impact of climatic change in the Dra-hany highlands (The Czech Republic) Ekologia (Bratisl)23(Suppl 2) 101minus112
Jeniacutek J Lokvenc T (1962) Die alpine Waldgrenze im Krko -noše Gebirge Rozpravy SAV Rada matematickyacutech apriacuterodniacutech ved 72 gteskoslovenskaacute akademie ved Praha
Jobbaacutegy EG Jackson RB (2000) Global controls of forest lineelevation in the northern and southern hemispheresGlob Ecol Biogeogr 9 253minus268
Jonaacutešovaacute M Vaacutevrovaacute E Cudliacuten P (2010) Western Carpa -thian mountain spruce forest after a windthrow naturalregeneration in cleared and uncleared areas For EcolManag 259 1127minus1134
Kittel TGF Steffen WL Chapin FS (2000) Global andregional modelling of Arcticminusboreal vegetation distribu-tion and its sensitivity to altered forcing Glob ChangeBiol 6 1minus18
Klopcic M Jerina K Bonltina A (2010) Long-term changes ofstructure and tree species composition in Dinaric un -even- aged forests Are red deer an important factor EurJ For Res 129 277minus288
Klopcic M Simonltilt T Bonltina A (2015) Comparison of re -generation and recruitment of shade-tolerant and light-demanding tree species in mixed uneven-aged forests ex -periences from the Dinaric region Forestry 88 552minus563
Koumlrner C (2003) Alpine plant life functional plant ecology ofhigh mountain ecosystems with 47 tables Springer Sci-ence amp Business Media BV Dordrecht
Koumlrner C (2012) Alpine treelines Functional ecology of theglobal high elevation tree limits Springer Basel
Koumlrner C Paulsen J (2004) A world-wide study of high alti-
tude treeline temperatures J Biogeogr 31 713minus732Kulakowski D Bebi P Rixen C (2011) The interacting effects
of land use change climate change and suppression ofnatural disturbances on landscape forest structure in theSwiss Alps Oikos 120 216minus225
Kullman L (1988) Holocene history of the forestminusalpine tun-dra ecotone in the Scandes Mountains (central Sweden)New Phytol 108 101minus110
Kullman L (1999) Early holocene tree growth at a high elevation site in the northernmost Scandes of Sweden(Lapland) a palaeobiogeographical case study based onmega fossil evidence Geogr Ann 81 63minus74
Kullman L (2007) Tree line population monitoring of Pinussylvestris in the Swedish Scandes 1973minus2005 implica-tions for tree line theory and climate change ecologyJ Ecol 95 41minus52
Kullman L Oumlberg L (2009) Post-Little Ice Age treeline riseand climate warming in the Swedish Scandes a land-scape ecological perspective J Ecol 97 415minus429
Kyriazopoulos AP Skre O Sarkki S Wielgolaski FE Abra-ham EM Ficko A (2017) Humanminusenvironment dynamicsin European treeline ecosystems a synthesis based onthe DPSIR framework Clim Res 73 17ndash29
Lenoir J Svenning JC (2015) Climate-related range shifts mdasha global multidimensional synthesis and new researchdirections Ecography 38 15minus38
Lenoir J Geacutegout JC Marquet PA de Ruffray P Brisse H(2008) A significant upward shift in plant species opti-mum elevation during the 20th century Science 320 1768minus1771
Lindner M Fitzgerald JB Zimmermann NE Reyer C andothers (2014) Climate change and European forests What do we know what are the uncertainties and whatare the implications for forest management J EnvironManag 146 69minus83
Lloyd AH Rupp TS Fastie CL Srafield AM (2002) Patternsand dynamic of treeline advance on the Seward Pen -isula Alaska J Geophys Res 107 8161
Maacuteliš F Kopeckyacute M Petriacutek P Vladovic J Merganic J VidaT (2016) Life-stage not climate change explains ob -served tree range shifts Glob Change Biol 22 1904minus1914
Mathisen IE Mikheeva A Tutubalina OV Aune S and Hof-gaard A (2014) Fifty years of tree line change in theKhibiny Mountains Russia advantages of combinedremote sensing and dendroecological approaches ApplVeg Sci 17 6ndash16
Millar CI Stephenson NL (2015) Temperate forest healthin an era of emerging megadisturbance Science 349 823minus826
Mina M Bugmann H Klopcic M Cailleret M (2017) Accu-rate modeling of harvesting is key for projecting futureforest dynamics a case study in the Slovenian moun-tains Reg Environ Change 17 49minus64
Moen J Aune K Edenius L Angerbjoumlrn A (2004) Potentialeffects of climate change on treeline position in theSwedish mountains Ecol Soc 9 16 wwwecologyandsocietyorgvol9iss1art16
Palombo C Chirici G Marchetti M Tognetti R (2013) Is landabandonment affecting forest dynamics at high elevationin Mediterranean mountains more than climate changePlant Biosyst 147 1minus11
Palombo C Marchetti M Tognetti R (2014) Mountain vege-tation at risk current perspectives and research reedsPlant Biosyst 148 35minus41
Pauli H Gottfried M Dullinger S Abdaladze O Akhalkatsi
149
Clim Res 73 135ndash150 2017
M Alonso JLB Grabherr G (2012) Recent plant diversitychanges on Europersquos mountain summits Science 336 353minus355
Payette S Fortin MJ Gamachet I (2001) The subarctic forestminustundra the structure of a biome in a changing climate BioScience 51 709minus718
Pearson GR Dawson TP (2003) Predicting the impacts of cli-mate change on the distribution of species Are bio -climate envelope models useful Glob Ecol Biogeogr 12 361minus371
Pedrotti F (2013) Plant and vegetation mapping GeobotanyStudies Springer-Verlag Berlin
Pentildeuelas J Boada M (2003) A global change-induced biomeshift in the Montseny mountains (NE Spain) GlobChange Biol 9 131minus140
Piovesan G Biondi F Filippo AD Alessandrini A MaugeriM (2008) Drought driven growth reduction in old beech(Fagus sylvatica L) forests of the central ApenninesItaly Glob Change Biol 14 1265minus1281
Rabasa SG Granda E Benavides R Kunstler G and others(2013) Disparity in elevational shifts of European trees inresponse to recent climate warming Glob Change Biol19 2490minus2499
Raev I Zhelev P Grozeva M Markov I and others (2011)Programme of measures for adaptation of forests in Bul-garia and mitigate the negative impact of climate changeon them Project FUTURE forest helping Europe tackleclimate change INTERREG IVC ERDF Sofia
Sarkki S Ficko A Grunewald K Nijnik M (2016) Benefitsfrom and threats to European treeline ecosystem serv-ices an exploratory study of stakeholders and gover-nance Reg Environ Change 16 2019minus2032
Savage J Vellend M (2015) Elevational shifts biotic homo -genization and time lags in vegetation change during 40years of climate warming Ecography 38 546minus555
Schwoumlrer C Henne PD Tinner W (2014) A model-data com-parison of Holocene timberline changes in the SwissAlps reveals past and future drivers of mountain forestdynamics Glob Change Biol 20 1512minus1526
Šmilauer P Lepš J (2014) Multivariate analysis of ecologicaldata using Canoco 5 Cambridge University Press Cam-bridge
Ste-Marie C Nelson EA Dabros A Bonneau ME (2011)Assisted migration introduction to a multifaceted con-cept For Chron 87 724minus730
Stevens JT Stafford HD Harrison S Latimer AM (2015) For-est disturbance accelerates thermophilization of under-story plant communities J Ecol 103 1253minus1263
Strid A Andonoski A Andonovski V (2003) Alpine biodiver-sity in Europe Ecological Studies 167 Springer-VerlagBerlin
Tattoni C Ciolli M Ferretti F Cantiani MG (2010) Monitor-ing spatial and temporal pattern of Paneveggio forest(Northern Italy) from 1859 to 2006 iForest Biogeosci For3 72minus80
Ter Braak CJF Šmilauer P (2012) Canoco reference manualand userrsquos guide software for ordination (version 50)Microcomputer Power Ithaca NY
Theurillat JP Guisan A (2001) Potential impact of climatechange on vegetation in the European Alps a reviewClim Change 50 77minus109
Toslashmmervik H Wielgolaski FE Neuvonen S Solberg BHoslashgda KA (2005) Biomass and production on a land-scape level in the mountain birch forests In WielgolaskiFE (ed) Plant ecology herbivory and human impact in
Nordic mountain birch forests Ecos Studies 180Springer Berlin p 53minus70
Treml V Banaš M (2000) Alpine timberline in the HighSudetes Acta Univ Carol Geogr 15 83minus99
Treml V Chuman T (2015) Ecotonal dynamics of the altitu-dinal forest limit are affected by terrain and vegetationstructure variables an example from the Sudetes Moun-tains in Central Europe Arct Antarct Alp Res 47 133minus146
Treml V Migo P (2015) controlling factors limiting timber-line position and shifts in the Sudetes a review GeogrPol 88 55minus70
Urban MC Tewksbury JJ Sheldon KS (2012) On a collisioncourse competition and dispersal differences create no-analogue communities and cause extinctions during cli-mate change Proc R Soc Lond B Biol Sci 279 2072minus2080
Vacek S Matejka K (2010) State and development of phyto-cenoses on research plots in the Krkonoše Mts foreststands J For Sci 56 505minus517
Van Bogaert R Haneca K Hoogesteger J Jonasson C deDapper M Callaghan TV (2011) A century of tree linechanges in sub-Arctic Sweden shows local and regionalvariability and only a minor influence of 20th century climate warming J Biogeogr 38 907minus921
Van Gils H Batsukh O Rossiter D Munthali W Liberato -scioli E (2008) Forecasting the pattern and pace of Fagusforest expansion in Majella National Park Italy ApplVeg Sci 11 539minus546
Velchev V (1997) Types of vegetation In Yordanova MDonchev D (eds) Geography of Bulgaria BAN Sofiap 269minus283 (in Bulgarian)
Vladovic J Merganic J Maacuteliš F Križovaacute E and others (2014)The response of forest vegetation diversity to changes inedaphic-climatic conditions in Slovakia Technickaacute uni-verzita vo Zvolene Zvolene (in Slovak)
Vrška T Adam D Hort L Kolaacuter T Janiacutek F (2009) Europeanbeech (Fagus sylvatica L) and silver fir (Abies alba Mill)rotation in the Carpathians mdash a developmental cycle or alinear trend induced by man For Ecol Manag 258 347minus356
Wielgolaski FE (ed) (2005) Plant ecology herbivory andhuman impact in Nordic mountain birch forests Eco -logical Studies 180 Springer Verlag Berlin
Wielgolaski FE Hofgaard A Holtmeier FK (2017) Sensitivityto environmental change of the treeline ecotone and itsassociated biodiversity in European mountains Clim Res73151ndash166
Wieser G Tausz M (2007) Trees at their upper limit Treelifelimitation at the alpine timberline Springer Dordrecht
Wieser G Holtmeier FK Smith WK (2014) Treelines in achanging global environment In Tausz M Grulke N(eds) Trees in a changing environment Plant Ecophys 9Springer Dordrecht p 221minus263
Zhiyanski M Kolev K Sokolovska M Hursthouse A (2008)Tree species effect on soils in Central Stara PlaninaMountains Nauka Gorata 4 65minus82
Zhiyanski M Gikov A Nedkov S Dimitrov P Naydenova L(2016) Mapping carbon storage using land coverlanduse data in the area of Beklemeto Central Balkan In Koulov B Zhelezov G (eds) Sustainable mountain re -gions challenges and perspectives in southeasternEurope Springer Basel p 53minus65
Zhu K Woodall CW Clark JS (2012) Failure to migrate lackof tree range expansion in response to climate changeGlob Change Biol 18 1042minus1052
150
Editorial responsibility Nils Chr Stenseth Oslo Norway
Submitted April 8 2016 Accepted March 8 2017Proofs received from author(s) July 7 2017
Cudliacuten et al Drivers of treeline shift
evidence that treelines have always been dynamicand influenced by climate change and forest devel-opment cycles in the past (Kullman 1988 2007Gehrig-Fasel et al 2007 Vla dovi et al 2014 Tremlamp Chuman 2015)
There are 3 main aspects of environmental changeto which trees are likely to respond increasing tem-perature rising concentrations of CO2 and increasingdeposition of nitrogen (Grace et al 2002 Lindner etal 2014) The trees at the treeline accumulate carbo-hydrates because autotrophic respiration is morelimited by low temperature than photosynthesis Thismeans that factors such as temperature and nitrogenabundance both of which affect the capacity of a treeto use the products of photosynthesis will probablybe more important than factors directly affectingphoto synthesis such as elevated CO2 (Fajardo et al2012)
However climate signals can beconfounded with the effects ofhuman activity Stronger treelinedyna mics due to a coupled effect ofclimate change and highland pastureabandonment frequently occur inEuropean mountain ranges (Ellen-berg 1988) Other human-relatedinfluential factors such as (overly)intensive forest management pas-ture and grazing may be regionallyimportant as well Di rect human-related factors may be crucial inchanging the species compositionand structure of mountain forestsespecially in treeline ecotones whereclimate change does not representthe key factor (Alados et al 2014)
The aims of this study were to eval-uate the effects of geographical posi-tion and different regional climaticand other site conditions on treelineecotone characteristics of selectedEuropean mountain areas to identifythe natural and anthropogenic driv-ers of treeline shift We further dis-cuss the effects of climate and landuse changes on biodiversity in forestsbelow the treeline
2 METHODS
To demonstrate and discuss thepossibilities and limits of treeline
species shift including latitudinal and longitudinalheterogeneity the current situation in 11 mountainareas from Italy to Norway and from Spain to Bul-garia were analysed (Fig 1) To determine the influ-ence of mountain massif size on the mesoclimatic andvegetation conditions including treeline formationwe included large mountain complexes (PyreneesAlps Scandes) as well as relatively small separatedmountains (Pirin and Giant Mountains) in this studyThe characteristics of the treeline ecotone includinghistorical and recent human im pacts on mountainecosystems for the 11 mountain areas of interest aresummarized in Table 1 Extensive mountain massifs(Pyrenees Alps Scandes) were further divided intomore homogeneous units (Central Pyrenees EasternPyrenees Central Alps Eastern Alps Hardan-gervidda Dovre Northern Swedish Lapland InnerFinnmarknorthernmost Finnish Lapland) To esti-
137
Fig 1 Selected European mountains Grey shading differentiates mountainareas 1 Central Pyrenees (CP) 2 Eastern Pyrenees (EP) 3 Apennines (AP) 4Shara Mts (SH) 5 Pirin (PI) 6 Central Balkan Mts (CB) 7 Northern DinaricMts (DI) 8 Central Alps (CA) 9 Eastern Alps (EA) 10 Low Tatra Mts (LT) 11Giant Mts (GM) 12 Hardangervidda (HG) 13 Dovre (DO) 14 NorthernSwedish Lapland (NS) 15 Inner Finnmarknorthernmost Finnish Lapland (FL)
Clim Res 73 135ndash150 2017138
Country Coordi- Elevation Elevation Range of Annual Area VZs near Tr part of nates of range range of Tr tree species altitudinal of study (m asl)mountains central site of T (m asl) limit shift (km2)
(m asl) (m asl) of T or Tr (m yrminus1)
Spain 42deg37rsquoN 1930 No data Pu 2700 049 (T 15 800 Pine with Aa Spanish 0deg27rsquoE (1800minus2050) 1956minus2006) Ps and Fs (1400minus2050) Pyrenees subalpine with Pu
(1800minus2300)
Italy 42deg05rsquoN 1800 No data Fs 2100 1 (Tr [Pm ] 741 Beech forest (800minus1850) Majella Massif 14deg05rsquoE (1700minus1850) Pm 2300 1954minus2007) mixed beechminusdwarf pine part of (1800minus2000) dwarf pine Apennines (2100minus2300)
Macedonia 41deg47rsquorsquoN 1850 1900 Pa 2100 ~1 (T ) 829 Mixed firminusbeechminusShara Mountain 20deg33rsquoE (1800minus1900) (1850minus1950) Pp 2200 1934minus2010 spruce (800minus1900) (Balkan beech (1600minus1900) Peninsula) dwarf pine (1600minus2200)
Bulgaria 41deg42rsquoN 1900 2100 Pp 2500 No data 384 Beech (1000minus1500) Pirin 23deg31rsquoE (1800minus2100) (1900minus2300) coniferous forest (Ps Pp
Ph Pa 1300minus2300) dwarf pine (2000minus2500)
Bulgaria 42deg47rsquoN Beech T 1600 1600 Fs 1700 No data 717 Beech (800minus1700) Central 24deg36rsquoE (1500minus1700) (1500ndash1850) Pa Aa 1850 coniferous (Pa Aa Balkan Mts 1500minus1850) juniper
(1500minus1850)
Slovenia 45deg36rsquoN 1540 1600 Fs 1650 No data ~200 Mixed firminusbeechminusspruce Northern 14deg28rsquoE (1455minus1600) (1550minus1650) Pa 1700 (600minus1400) beech (1300minus1550) Dinaric Mts Aa 1650 dwarf pine with small groups
or individual trees of Fs Pa Aa Ap (1500minus1700)
Switzer- 46deg46rsquoN 1900 2100 Pa 2400 ca13 (T ~250 Spruce with Ld land Italy 9deg52rsquoE (1700minus2150) (1850minus2300) Ld 2450 1954minuspresent) + ~750 (1000minus1900) forest with Pc Central and 46deg17rsquoN Ld and Pa (1700minus2200) Eastern Alps 11deg45rsquoE dwarf shrub vegetation partly
pastures (2200minus2500)
Table 1 Timberline and treeline parameters and characteristics of human activities changes according to climatic models andlimits to climate change induced species shifts in selected European mountains A1B A2 emissions scenarios (IPPC) AaAbies alba Ap Acer pseudoplatanus Fs Fagus sylvatica Ld Larix decidua Pa Picea abies Pc Pinus cembra Ph Pi-nus heldreichii Pm Pinus mugo Pp Pinus peuce Ps Pinus sylvestris Pu Pinus uncinata T timberline Tr treeline TSL
tree species limit VZ vegetation zone C century mng management Elevation range of T and Tr mean (minminusmax)
Cudliacuten et al Drivers of treeline shift 139
Human Annual Limits Referencesinfluence temperature (T ) to climate
in VZs in past and precipitation (P) change and recently differences induced species
between 1961minus1990 shiftand 2021minus2050
mean of CORDEX models
Since 11th C HIRHAM5 model Inadequacy of Batllori et al (2009) clear-cutting and grazing period ΔT2021minus2050 substrata (rocky) Alados et al (2014)
moderate on irregular = 15minus2degC ΔT2051minus2080 climatic constraints Gartzia et a (2014) slopes and ridges since 1930 = 4degC ΔP2021minus2050 (including growing- Camarero et al (2015a)significant release of influence = +5 ΔP2051minus2080 season extreme events
but recovery mainly at = +30 compared to such as late-spring moderate elevations 1960minus1990 freezing summer
drought etc)
Since 1000 BC HadCM3_A2 period Low density of Fs van Gils et al (2008) cutting burning and 2020minus2080 in T dispersal distances Palombo et al (2013 2014)
grazing mng since the mid-20th ΔTmin_Jan2050 = 17degC of Fs seeds upslope C grazing intensity decreased ΔTmin_Jan2080 = 31degC poor soils excessive
now occasionally managed ΔP2050 = minus7 exposure to solar radiation and regeneration of Pm ΔP2080 = minus23 in open areas
Since 14th C clear-cutting Mean of 4 GCMs Differences in soil types Em (1986) and grazing to enlarge (CSIROMk2 HadCM3 between subalpine beech Strid et al (2003)
alpine pastures and to produce ECHAM4OPYC3 and mixed forest (poor Bergant (2006) charcoal since 1970 NCAR minus PCM) ΔT2050 = calcareous soils or rankers on Amidžic et al (2012)
abandonment of traditional 26degC ΔT2100 = 53degC silicate bedrock) cliffs agricultural practices ΔP2050 = minus3 ΔP2100 = minus8 and rocks cattle grazing
Since 680 AD intensive HADCM3_A2 Rocky sites extreme Velchev (1997) fires deforestation and grazing CGGM2 hydrothermal conditions Grunewald amp
since 1962 traditional mng ΔT2050 = 17degC competition of shrubs and tree Scheithauer (2011) has changed last decades tourism ΔT2080 = 28degC seedlings locally intense Raev et al (2011)
impact has increased ΔP2080 = minus15 pasturing
Human impact (mostly burning) HADCM3_A2 Low density of Fs Raev et al (2011) started in 17th C excessive CGCM2 ΔT2050 = 17degC in T dispersal distances Gikov et al (2016) and improper forest mng ΔT2080 = 28degC of Fs seeds upslope pasture Zhiyanski et al
release of influence since 1980 ΔP2050 = minus15 intensity new (2008 2016)ΔP2080 = minus15 expansion of juniper stands
15minus18th C slash burn DMI-HIRHAM5_A1B Impact of wild large Klopltilt et al grazing mng 18minus19th C period 2001minus2100 ΔT2050 ungulates on tree seedlings (2010 2015)
uncontrolled cutting 20th C = 12degC ΔT2100 = 25degC (especially Aa) shallow Mina et al (2017)irregular shelterwood systems ΔP2050 = +15 mm nutrient-poor soils
and over-exploitation mng ΔP2100 = minus18 mm (ie ranker and rendzina)
Since 12th C slash burn grazing mng ENSEMBLES models_ Frequent disturbances Tattoni et al (2010) 13minus19th C intensive forest A1B period 1980minus2100 by snow avalanches Kulakowski et al (2011)
mng at the end of the 19th C ΔT2050 = 12minus18degC and snow gliding ongoing Barbeito et al (2012) decrease in grazing ΔT2100 = 27minus41degC cattle grazing game grazing Bebi et al (2012)
ΔP2050 = 0 mm in the coldest years Dawes et al (2015)ΔP2100 = minus5 mm
Continued on next page
Clim Res 73 135ndash150 2017
mate the geographical position of these mountainunits 1 value was used for the geographical latitudeand longitude of the locus of the studied mountainunits The analysed parameters of these 15 mountainunits are shown in Table 2
In addition we calculated the following climatecharacteristics for all mountain units annual meanair temperature annual sum of precipitation thegrowing season length and the date of the onset ofthe growing season These characteristics were thencompared for 2 distinct periods 1961minus1990 and1991minus 2015 (Table 3) The climate characteristicswere derived from the E-OBS gridded dataset of sta-tion observations version 131 (Haylock et al 2008)We used the E-OBS version on the regular longi-tudeminuslatitude grid with a horizontal resolution of 025degrees The climate data from all grid points withinthe mountain units or near their geographical bor-ders were considered
Data processing for 15 mountain units was made onthe basis of data obtained from the literature and
personal studies by the co-authors these are sum ma -rized in Tables 1 amp 2 and in Tables S1 amp S2 in theSupplement at www int-res com articles suppl c073p135 _ supp pdf In addition a semi-quantitative valu-ation of abiotic biotic and anthropogenic factors lim-iting an upward treeline shift was performed (Fig 2)To answer the question how natural and anthro-pogenic factors have influenced treeline ecotonecharacteristics with a focus on treeline shift redun-dancy analysis (RDA) was used to describe and testthe effect of the explanatory variables (geographicalposition size of the whole mountain massif start ofhuman influence and start of the decrease in humaninfluence) on treeline ecotone characteristics (tim-berline elevation width of the treeline ecotone treespecies forming the treeline and treeline shift peryear) and identify groups of mountain units with sim-ilar variability of the dependent data (ter Braak ampŠmilauer 2012) The whole data set for this analysis isshown in Table S2 We tested both their simple ef -fects which show how much variation every ex plan -
140
Country Coordi- Elevation Elevation Range of Annual Area VZs near Tr part of nates of range range of Tr tree species altitudinal of study (m asl)mountains central site of T (m asl) limit shift (km2)
(m asl) (m asl) of T or Tr (m yrminus1)
Slovakia 48deg55rsquoN 1400 1530 Pa 1730 ~03 (Tr ~400 Spruceminusfirminusbeech Dumbier Tatra 19deg31rsquoE (1350minus1450) (1400minus1630) 1950minuspresent) (1200minus1350) spruce (1300minus(part of Low 1550) Pm with individual Tatra Mts) Pa trees (1400minus1700)
Czech Republic 50deg42rsquoN 1250 1320 Pa 1500 043 (Tr 550 Beechminusspruce (700minus1050) Giant Mts 15deg38rsquoE (1130minus1460) (1250minus1460) 1936minus2005) Pa dominates with increasing
altitude spruce (1000minus1400) Pm with Pa trees (1400minus1560)
Norway 60deg10rsquoN 850 600minus1050 1500 08 (Tr 40 600 Birch forest dominance Southern 7deg40rsquoE 62deg20rsquoN 1915minus2007) at high altitudes but with Scandes 10deg05rsquoE some scattered Ps or Pa
mixed birchminusconifer below 800
Norway 68deg20rsquoN West 650 West 750 West 1000 06 (Tr 35 000 Birch forest dominance Sweden 18deg20rsquoE East 100 East 290 East 0 1958minus2008) at all altitudes but with some Finland 69deg50rsquoN scattered pines or pine stands Northern 27deg00rsquoE on sandy soilsScandes
Table 1 (continued)
Cudliacuten et al Drivers of treeline shift
atory variable can explain separately without usingthe other variables and their conditional ef fectswhich depend on the variables already selected inthe model The statistical significance of the expla -natory variables was tested using a Monte Carlo permutation test and only predictors with p le 005were included in the subsequent RDA (Šmilauer ampLepš 2014)
To find the drivers of biodiversity loss in forestsmeadows and animal communities (dependent vari-ables) the explanatory variables (geographical posi-tion timberline elevation tree species forming thetreeline start of human influence start of the de -crease in human influence and temperature increasebetween the 2 periods 1961minus1990 and 2021minus2050)were tested again by RDA in Canoco 5 as mentionedabove (Figs 3 amp 4) The whole data set for this analy-sis mdash including additional information about tree-lines in the mountain areas of interest with a focuson treeline shift (its rate drivers limits and problemswith its assessment) mdash is presented in Tables S1 amp S2
Selected univariate graphs were constructed to visu-alize the data of relationships between the explana-tory and dependent variables of all analyses whichwere not seen distinctly from the results of the multi-variate statistics or linear regression (Fig 5)
3 RESULTS
31 Effect of environmental variables on treeline ecotone characteristics with a focus
on treeline shift
The variation in the dependent variables (timber-line elevation width of treeline ecotone tree speciesforming the treeline and treeline shift per year) wassignificantly affected only by latitude and size of thewhole mountain massif (Fig 3) The adjusted ex -plained variability by all explanatory variables was342 (F = 46 p = 001) The RDA diagram showsthat in the mountains located more to the north the
141
Human Annual Limits Referencesinfluence temperature (T ) to climate
in VZs in past and precipitation (P) change and recently differences induced species
between 1961minus1990 shiftand 2021minus2050
mean of CORDEX models
From 14th C to 1922 Average of 10 RCM_B1 More rocky and nutrient- Koumlrner (2003) mining forest change ΔT2050 = 18degC poor soils late frosts Fridley et al (2011)
to spruce monocultures ΔT2100 = 37degC steep slopes with avalanches Hlaacutesny et al (2011 2016)from 13th C to 1978 pastures ΔP2050 = +24 mm absence of mature trees
in upper parts of mountain ΔP2100 = minus35 mm
9minus11th C forest clear cutting ALADIN_A1B period 1961minus Different soil types Treml and Banaš (2000) grazing in dwarf pine VZ 2100 ΔT2050 = 13degC for Fs and Pa (cambisol Cudliacuten et al (2013)
since 18th C forest change ΔT2100 = 34degC versus podzol) only TSL of Pa Treml amp Chuman (2015) to spruce monocultures ΔP2050 = +25 mm could elevate on ranker Treml amp Migo (2015)
since 19th C artificially planted ΔP2100 = minus14 mm soils in Pm VZ
Grazed and browsed RegClimMPIHadley Sheep and reindeer Dalen amp Hofgaard (2005) by reindeer and livestock ΔT2050 = 12degC grazing and episodic Wielgolaski (2005)
(sheep and cattle) ΔT2100 = 22degC insect outbreaks (Epirrita) Hofgaard et al (2009) over thousands of years ΔP2050 = +23 mm Kullman Oumlberg (2009)
ΔP2100 = minus12 mm
Grazed and browsed RegClimMPIHadley Continued grazing regime Dalen amp Hofgaard (2005) by semi-domestic reindeer ΔT2050 = 16degC and frequent episodic Wielgolaski (2005)
for hundreds of years ΔT2100 = 29degC insect outbreaks (Epirrita) Toslashmmervik et al (2005) increased grazing from ΔP2050 = +18 mm Hofgaard et al (2009)
year 1960 ΔP2100 = minus10 mm Aune et al (2011) Mathisen et al (2014)
Clim Res 73 135ndash150 2017
treeline ecotone is narrower treeline shiftis smaller the timberline occurs at lowerelevations and the treeline is formed moreby broadleaved species Higher values oftreeline ecotone width and treeline shiftwere found in the mountains with a greatersize of the whole mountain massif
When every explanatory variable wastested separately values of the start of thedecrease of human influence in the moun-tains also had a significant effect on thevariation of the tree line ecotone data(explained variation = 206 pseudo-F =34 p = 0024) Some selected relationshipsof the explanatory and dependent vari-ables even those not showing significanteffects on the variation in the tree line eco-tone parameters are depicted in Fig 5
32 Effect of recent changes in climateand land use on the biodiversity
Biodiversity loss in forests meadows andanimal communities analysed by RDAwas explained by geo graphical positiontreeline species composition and tempera-ture increase between the 2 periods1961minus1990 and 2021minus2050 (Fig 4) Theadjusted ex plained variation by all ex pla -natory variables was 632 (F = 58 p =0001) The RDA results indicated that bio-diversity loss in forest communities in crea -sed with increasing latitude and longitudeas well as in the case where broad leavedspecies formed the treeline In contrast thehighest biodiversity loss in meadows wasfound in the mountains positioned more tothe south and with higher temperatureincrease The biodiversity loss in animalcommunities was negatively correlatedwith longitude and temperature increase
4 DISCUSSION
The comparison of several mountainareas situated across Europe shows a varia-tion in understanding of what constitutesthe treeline across countries The primaryissue is the different approach to the defini-tion of forest when applying a tree heightthreshold This threshold decreases from
142
Mou
nta
in u
nit
Nor
th
Eas
t M
oun
-T
imb
er-
Wid
th
Tre
e T
emp
erat
ure
T
reel
ine
Sta
rt
Sta
rt
lati
tud
elo
ng
itu
de
tain
lin
eof
spec
ies
incr
ease
sh
ift
of H
I of
HI
size
(m a
sl
)tr
eeli
ne
form
ing
b
etw
een
(m
yrminus
1 )(y
r)d
ecre
ase
(km
2 )ec
oton
eth
e 19
61minus
1990
(y
r)(m
)tr
eeli
ne
and
199
1minus20
15 (
degC)
Cen
tral
Pyr
enee
s (C
P)
42deg2
1rsquoN
0deg30
rsquoW55
000
2250
300
C0
81
92 (
1956
minus20
06)
1000
BC
1930
Eas
tern
Pyr
enee
s (E
P)
42deg2
8rsquoN
1deg30
rsquoE55
000
2400
300
C0
80
71 (
1956
minus20
06)
1000
1930
Ap
enn
ines
(A
P)
42deg0
5rsquoN
14deg0
5rsquoE
5000
018
0030
0B
14
1 (1
954minus
2007
)10
00 B
C19
50S
har
a M
ts (
SH
)41
deg47rsquo
N20
deg33rsquo
E16
0018
5035
0C
B0
71
(193
4minus20
10)
1300
1970
Pir
in (
PI)
41deg4
2rsquoN
23deg3
1rsquoE
2585
1900
400
C0
5n
o d
ata
680
1962
Cen
tral
Bal
kan
Mts
(C
B)
42deg4
7primeN
24deg3
6primeE
1160
016
0025
0C
B0
7n
o d
ata
1700
1980
Nor
ther
n D
inar
ic M
ts (
DI)
45deg3
6rsquoN
14deg2
8rsquoE
1000
1540
160
B1
3n
o d
ata
1400
1900
Cen
tral
Alp
s (C
A)
46deg4
6rsquoN
9deg52
rsquoE18
000
021
0055
0C
07
13
(195
4minus20
15)
1300
1880
Eas
tern
Alp
s (E
A)
46deg1
7rsquoN
11deg4
5rsquoE
180
000
1800
400
C0
91
3 (1
954minus
2015
)11
5019
50L
ow T
atra
Mts
(LT
)48
deg55rsquo
N19
deg31rsquo
E12
4014
0033
0C
09
057
(19
75minus
2016
)14
0019
22G
ian
t M
ts (
GM
)50
deg42rsquo
N15
deg38rsquo
E63
012
5025
0C
07
043
(19
36minus
2005
)10
0019
63H
ard
ang
ervi
dd
a (H
G)
60deg0
3rsquoN
7deg47
rsquoE90
0090
015
0C
B0
50
8 (1
915minus
2007
)80
019
70D
ovre
(D
O)
62deg2
1rsquoN
8deg52
rsquoE60
0010
0040
0B
07
08
(191
5minus20
07)
1000
no
dec
reas
eN
orth
ern
Sw
edis
h L
apla
nd
(N
S)
68deg1
2rsquoN
18deg3
3rsquoE
4000
700
300
B1
10
6 (1
958minus
2008
)10
00n
o d
ecre
ase
Inn
er F
inn
mar
kn
orth
ern
mos
t 69
deg10rsquo
N25
deg12rsquo
E60
000
290
90B
10
06
(195
8minus20
08)
1600
no
dec
reas
eF
inn
ish
Lap
lan
d (
FL
)
Tab
le 2
Fac
tors
infl
uen
cin
g t
reel
ine
ecot
one
char
acte
rist
ics
in s
elec
ted
Eu
rop
ean
mou
nta
ins
Th
e m
oun
tain
siz
e is
th
e si
ze o
f th
e w
hol
e m
oun
tain
mas
sive
fu
nct
ion
ing
as a
mez
osca
le c
lim
ate
un
it n
ot o
nly
the
stu
die
d m
oun
tain
are
a W
idth
of t
he
tree
lin
e ec
oton
e w
as d
eter
min
ed b
y su
btr
acti
ng
the
valu
e of
the
mea
n ti
mb
erli
ne
elev
atio
n
from
th
e va
lue
of t
he
up
per
tre
e sp
ecie
s li
mit
C c
onif
ers
B b
road
leav
ed t
rees
HI
hu
man
infl
uen
ce
Cudliacuten et al Drivers of treeline shift
5 m (Jeniacutek amp Lokvenc 1962) to 3 m(Koumlrner 2012) in Northern andCentral Europe to only 2 m (Holt-meier 2009) in Central and South-ern Europe where even shrub -by stands (eg Pinus mugo) areconsi dered as a forest in somecountries eg in Spain Italy andMa ce donia (Batllori et al 2009)An other problem is a different de -signation of forest stands belowthe treeline lsquoupper montane fo -restsrsquo in Central Europe and lsquosub-alpine forestsrsquo in Southern Europe(Ellenberg 1988) Further differen -ces are related to the tradition ofdifferent branches of science (ega more traditional lsquogeobotanicalrsquoapproach in Central Europe ver-sus a more experimental approachin Western Europe) especially inthe rate of applying new pro -gressive methods (eg climaticmodelling or molecular biologicalmethods)
Our comparison of selectedregions (Tables 1 amp 2 Figs 1 amp 2)showed that southern mountainscompared to those located morecentrally and in the north had (1) alonger and more profound exploi -tation by humans in the past (2)greater differences in climatic para -meters (temperature length ofgrowing period) between the 2periods 1961minus1990 and 1991minus2015(Table 3) and (3) more dramaticclimate change scenarios espe-cially concerning temperature in -creases Longitude also had someinfluence on the start and intensityof human influence and on tree-line elevation and rate of treelineshift (Table 2 Fig 5) The occur-rence of broadleaf tree species inthe treeline ecotone in the north-ern (and to some extent the south-ern) countries distinguishes thesefrom the Central European coun-tries where conifers prevail It isinteresting that grazing an im -portant factor shaping treelineecosystems (Dirnboumlck et al 2003
143
Length Beginning Mean Mean of growing of growing annual annual
period season temperature precipitation (d)a (d)b (degC) ()
Central Pyrenees 135 minus89 08 minus107Eastern Pyrenees 117 minus91 08 minus92Apennines 284 minus232 14 57Shara Mts 16 01 07 minus100Pirin 26 minus14 05 27Central Balkan Mts 169 minus136 07 76Northern Dinaric Mts 160 minus95 13 minus56Central Alps 147 minus86 07 62Eastern Alps 125 minus96 09 minus33Low Tatra Mts 24 minus34 09 minus91Giant Mts minus10 13 07 55Hardangervidda Blefjell 63 34 05 56Dovre 138 32 07 64Northern Swedish Lapland 81 05 11 minus27Inner Finnmarknorthern- 17 08 10 77most Finnish Lapland
aPositive numbers indicate a prolongation bNegative numbers indicate an earlier beginning
Table 3 Differences of selected climatic parameters between the 2 study periods (1961minus1990 and 1991minus2015) in selected European mountains
Tein Drou Gefa Nadi Laus Abli Bili Huli
0 1 2 3
Central Pyrenees (CP)
Eastern Pyrenees (EP)
Apennines (AP)
Shara Mts (SH)
Pirin (PI)
Central Balkan Mts (CB)
Northern Dinaric Mts (DI)
Central Alps (CA)
Eastern Alps (EA)
Low Tatra Mts (LT)
Giant Mts (GM)
Hardangervidda (HG)
Dovre (DO)
Northern Swedish Lapland (NS)
Inner FinnmarknorthernmostFinnish Lapland (FL)
Fig 2 Rates of treeline drivers (temperature increase [Tein] land use change[Laus]) and treeline shift limits (drought [Drou] geomorphological factors [Gefa]natural disturbances [Nadi] other abiotic biotic and human factors) in selectedEuropean mountains Abli Bili and Huli abiotic biotic and human treeline shiftlimits respectively) Rate of treeline driver influence mdash 0 no influence 1 weak
influence 2 middle influence 3 strong influence
Clim Res 73 135ndash150 2017
Gehrig-Fasel et al 2007) limits treeline shift in halfof the countries of interest regardless of latitude lon-gitude or recent political developments (Shara PirinCentral Balkans Alps Scandes Fig 2)
Despite the stated differences between studieslooking at climate change impacts on the treelineit is clear that current and future changes in temp -erature will seriously affect treeline ecotones in allmountain ranges of Europe Winter is a period whentreeline stands and individual trees have to survivesevere life-limiting conditions (Wieser amp Tausz2007) Therefore winter warming might increase thechance of young trees surviving and passing the mostcritical period from seedling to sapling and further tothe mature stage (Koumlrner 2003) There is alreadymuch evidence worldwide that winter warming isone of the significant drivers of treeline advance(Harsch et al 2009) Although it is expected that treegrowth at the upper distributional margins (and inthe northern European countries) will increase due toongoing climate change (Pentildeuelas amp Boada 2003Chen et al 2011 Hlaacutesny et al 2011 Lindner et al2014) extremes in temperature (eg black froststemperature reversals in the spring) or winter precip-itation (eg lack of snow drought) might also limit
this process in the future in some regions and localsituations (Holtmeier amp Broll 2005 Hagedorn et al2014) An earlier start to the growing period (espe-cially in the Apennines Central Balkan and SpanishPyrenees see Table 2) can play a negative role in theresistance of trees and seedlings to early springfrosts On the other hand at lower distributional mar-gins (and in Southern European countries) it isexpected that tree growth will decrease or forestswill experience some level of dieback due to drought(Breshears et al 2005 Piovesan et al 2008 Allen etal 2010 Huber et al 2013) A serious implication isthe positive effect that warming can have on root rotfungi and populations of bark beetles resulting inlarge-scale disturbances to temperate and also high-altitude forests (Jankovskyacute et al 2004 Elkin et al2013 Millar amp Stephenson 2015)
Although recent upward advances of treeline eco-tones are widespread in mountain regions (Harsch et
144
Fig 3 Redundancy analysis diagram with variation in tim-berline elevation width of treeline ecotone the tree speciesforming the treeline and treeline shift used as dependentvariables explained by explanatory variables (latitudemountain size) Mountain units (blue) are projected as thecentres of abbreviations (see Table 2) Explanatory variablesaccount for 436 of the total variation in the dependentdata The first canonical axis explained 466 of variationthe second axis explained 30 of variation Dependentvariables (black) are Ecot altitudinal width of treeline eco-tone TLShift treeline shift TrspC (TrspB) treeline formedby conifers (broadleaved trees) Timb timberline elevationExplanatory variables (red) are N north latitude Mtsize
size of the mountain massif
Fig 4 Redundancy analysis diagram with variation in bio-diversity loss in forests meadows and animal communitiesused as dependent variables explained by explanatory vari-ables (geographical position tree species forming the tree-line and temperature increase between the 2 study periods[1961minus1990 and 2021minus2050]) Mountain units (green) areprojected as centres of abbreviations (see Table 2) Explana-tory variables account for 763 of total variation in the de-pendent data The first canonical axis explained 484 ofvariation the second axis explained 183 of variation De-pendent variables (black) are BdfoBdmeBdan biodiversityloss in forestsmeadowsanimal communities Explanatoryvariables (red) are N north latitude E east longitude TrspC(TrspB) treeline formed by conifers (broadleaved trees)Temp temperature increase between the 2 study periods
Cudliacuten et al Drivers of treeline shift
al 2009) only a few published studies have includedquantitative data about treeline shifts (Devi et al2008 Kullman amp Oumlberg 2009 Diaz-Varela et al 2010Van Bogaert et al 2011) Additionally our know -ledge of the spatial patterns in treeline ecotone shiftsat the landscape scale is still surprisingly poor exceptfor some studies from subarctic areas the UralMountains and the Alps (Lloyd et al 2002 Diaz-Varela et al 2010 Hagedorn et al 2014) In the 15studied mountain units the values of treeline shiftranged from 043 to 19 m yrminus1 showing a rather dis-tinct spatial pattern of the dynamics within Europeanmountains The observed treeline shift had a signifi-cant positive relationship only with northern latitudeand a weak positive relationship with mountain-range size (ie the size of the whole mountain massifFig 3) Nevertheless this illustrates the importanceof the mass elevation effect (Koumlrner 2012) and par-tially ex plains why treelines could be at different alti-tudes at the same latitude Small negative regres-sions with east longitude and timberline elevationare shown in Fig 5 There was no apparent depend-ence of the treeline shift rate on climatic parameterchanges between the periods 1961minus1990 and1991minus2015
A whole forest vegetation zone shift is a complexprocess as not only the life strategies of an individualtree need to be considered but plant animal andmicroorganism species and their interrelationshipsas well as the relationships to their specific microhab-
itats (including soil conditions) are also involved(Urban et al 2012) Across all regions we identifiedseveral important obstacles to treeline shifts oftenrelated to site properties such as significant rocki-ness having shallow or low-nutrient soils extremerelief causing disturbances by avalanches snow glid-ing or wind damage (data from the Pyrenees Apen-nines Shara Northern Dinaric Low Tatra and GiantMountains see Fig 2) Soil heterogeneity certainlyhas an important role in plant responses to climatechange and could also maintain the resilience of thecommunity assemblages (Fridley et al 2011) An -other group of obstacles includes extreme climaticparameters especially in winter (reported eg fromthe Pyrenees Apennines Pirin Central BalkanNorthern Dinaric and Giant Mountains Fig 2) Theircombination can result in edaphic andor climaticunsuitability of habitats where species could poten-tially migrate Climatically and edaphically suitablesites will likely decline over the next century partic-ularly in mountain landscapes (Bell et al 2014)There fore trees in the treeline ecotone will colonizepreviously forested habitats The colonization suc-cess of individual forest communities is affected bydifferences in species dispersal and recruitmentbehaviour (Dullin ger et al 2004 Jonaacutešovaacute et al2010) For these reasons colonization of new foresthabitats by the next tree generation may not be suc-cessful and may result in loss of species diversity(Honnay et al 2002 Ibaacutentildeez et al 2009 Dobrowski et
145
Fig 5 Univariate graphs of the relationships of dependent variables viz timberline (Timb) and treeline shift per year (Shift) and their explanatory variables Nola north latitude Ealo east longitude DD decimal degrees
Clim Res 73 135ndash150 2017
al 2015) as reported from Macedonia Slovenia theCzech Republic Slovakia and Norway (Fig 4)
Another limiting factor is the existence of strongadaptation mechanisms of the dominant tree specieswhich allow them to survive in their current distribu-tion areas Species migration is likely to be slow dueto the limited quantity of climatically and edaphicallysuitable sites and the slow velocity of seed dispersaland tree regeneration and will be hampered evenmore by fragmentation of high mountain landscapescaused by human activities including brush invasionin abandoned meadows (Gartzia et al 2014) A verti-cal shift in a vegetation zone involves not only singlespecies of plants animals and microorganisms butalso their interrelationships and links to soil condi-tions The speed at which climatic conditions changewill be different from changes in the soil conditionsdue to the slowness of soil-forming processes Soiltypes and conditions (cambisol versus podzol) arecrucial obstacles for the shift from a beech forest zoneinto the spruce forest zone in the Czech Republic(Vacek amp Mat jka 2010) According to other sourcesthe dominant tree species can influence soil for -mation processes (especially humus forms) Underfavourable climatic and orographic conditions beechis able to change its soil conditions over decades tocenturies (J Macku unpubl data)
Significant changes in biodiversity must be expec -ted in all of the mountain areas of interest We foundthat biodiversity declined particularly in the southernregions of Europe where the timberline is situated athigher elevations and human impact is mostlylonger-lasting (Fig 4) Similarly Pauli et al (2012)reported an increase in species richness of mountaingrasslands across Europe except for Mediterraneanregions and assigned this different response of thesouthern regions to decreased water availability Onthe other hand ecosystems which exhibited bio -diversity increases were not only enriched by mig -rant species but the assemblages underwent thermo -philization ie cold-adapted species declined andmore warm-adapted species spread (Gottfried et al2012) However not all species are able to track cli-mate changes It is expected that around 40 ofhabitats will become climatically unsuitable for manymountain species particularly endemic ones in -creasing their extinction probability during this cen-tury (Dullinger et al 2012) The abandonment of tra-ditional land use forms is another source of this losswhich may be a more important driver than tempera-ture increase (Fig 4) Serious biodiversity losses inmountain meadows were reported from Spain ItalyMacedonia Bulgaria Slovenia and Slovakia There-
fore controlled grazing must occur in order to main-tain alpine grasslands (Dirnboumlck et al 2003) Unfor-tunately grazing is still active only in smaller parts ofEuropean mountains (eg in the Central Alps and theCentral Balkan Mountains but pasturing can alsonegatively affect vegetation diversity) and some-times does not serve to maintain alpine grasslandThe same is true for grassland management whichcan help to maintain mountain species and decreasehabitat loss A serious biodiversity loss in mountainmeadows was reported from the Pyrenees the Apen-nines and the Shara Pirin Central Balkan NorthernDinaric and Low Tatra Mountains (Fig 4)
In forests species responses to climate change maybe equivocal some recent studies indicated contra-dictory shifts in species distributions (eg Lenoir etal 2008 Zhu et al 2012 Rabasa et al 2013) This dis-parity is widely discussed and assigned to the greatvariety of non-climatic factors or even tree ontogeny(Grytnes et al 2014 Lenoir amp Svenning 2015 Maacutelišet al 2016) Disturbances leading to tree mortalitymay also play an important role (Cudliacuten et al 2013)changes in tree canopy cover modify light availabil-ity and microclimate and can induce the thermo -philization of forest vegetation (De Frenne et al2015 Stevens et al 2015) The loss of this microcli-mate buffering effect of forests may induce a biotichomogenization of forests (Savage amp Vellend 2015)or the creation of novel non-analogical communitiessuch as oakminuspine forests (Urban et al 2012) whichmay be a new threat to forest biodiversity
The observed changes in treeline forest ecosystemsare often related to changes in land-use intensity(Theurillat amp Guisan 2001 Alados et al 2014) Ac -cording to climate change predictions and the recentand future exploitation intensity of European moun-tains trees and forest communities will shift upwarddue to land use change or climate change or bothReduced land-use intensity certainly will interactwith climate change by facilitating or inhibiting spe-cies occurrence and will accelerate forest expansionabove the present treeline particularly to previouslyforested habitats (Theurillat amp Guisan 2001) Thesimultaneous action of both main treeline shift driv-ers viz temperature increase and decrease in landuse intensity was recorded from all of our studiedmountain areas The extensive differences in timber-line and treeline elevations in almost all studiedmountains (Table 1) indicate the anthropo-zoogenictreeline type (according to Ellenberg 1998) In mostmountains (eg in the Apennines Shara MountainsCentral Balkans and Alps) land use is the prevailingfactor influencing vegetation drift (Fig 5) Previous
146
Cudliacuten et al Drivers of treeline shift
research showed that the upward shift of the treelinein the Swiss Alps was predominantly attributable toland abandonment and only in some situations to cli-mate change (Gehrig-Fasel et al 2007) In the Apen-nines in the last few decades tree establishment hasbeen mainly controlled by land use while treegrowth has been controlled by climate pointing to aminor role played by climate in shaping the currenttreeline (Palombo et al 2013)
The impact of climate change and the connectedland use change on biodiversity and ecosystem serv-ices provision in several European countries is sum-marized by Wielgolaski et al (2017) and Kyriazopou-los et al (2017) (both this Special) and Sarkki et al(2016) One of the possible adaptive management op-tions in response to climate change assisted migra-tion as human-assisted movement of species hasbeen frequently debated in the last few years (Ste-Marie et al 2011) It is possible to apply it as a type ofassisted colonization ie intentional movement andrelease of an organism outside its indigenous range toavoid extinction of populations of the focal species(eg Macedonian pine species) or as an ecologicalre placement ie the intentional movement and re-lease of an organism outside its indigenous range toperform a specific ecological function (eg planting ofPinus mugo in the Alps IUCNSSC 2013)
5 CONCLUSIONS
The analysis of 11 mountain areas across Europeshowed that with increasing latitude the treelineand altitudinal width of the treeline ecotone signifi-cantly decreases as do the significance of climaticand soil parameters as barriers against tree speciesshift Although temperature is the overwhelmingcontrolling factor of tree growth and establishment intemperate and boreal treeline ecotones late-sea-sonal drought might also play a driving role in Medi-terranean treeline ecotones Longitude was lessinfluential mostly affecting climate and land usechange effects on increased biodiversity loss as wellas the size of the area that forms one mesoclimateunit affecting altitudinal ecotone width The biggestpart of the commonly observed remaining variabilityin mountain vegetation near the treeline in Europeseems to be caused by geomorphological geologicalpedological and microclimatic variability in combina-tion with different land use history and the presentsocio-economic relations The observed differencesin climatic parameters between the mountain areasof interest in comparison with the reference period
1960minus1990 (09degC) have not explained the relativelyhigh differences in the rate of treeline shift per year(149 m) The predicted variability in temperatureincrease due to global warming (12minus26degC in 2050and 22minus53degC in 2100) could lead to a much biggerdifferentiation in treeline ecotone biodiversity andecosystem processes between southern and northernEuropean mountains in the future Therefore thesedifferences must be taken into account by scientistsand EU policy makers when formulating efficientadaptive forest management strategies for treelineecosystems at the European level (eg assistedmigration of adapted genotypes)
Acknowledgements This paper is based firstly upon workfrom the COST Action ES 1203 SENSFOR supported byCOST (European Cooperation in Science and Technologywwwcosteu) This international work was further sup-ported by projects granted by the Ministry of EducationYouth and Sports of the Czech Republic grant NPU ILO1415 and LD 14039 by the agency APVV SR under pro-jects APVV-14-0086 and APVV-15-0270 We acknowledgethe E-OBS dataset from the EU-FP6 project ENSEMBLES(httpensembles-eumetofficecom) and the data providersin the ECAampD project (wwwecadeu)
LITERATURE CITED
Alados CL Errea P Gartzia M Saiz H Escoacutes J (2014) Posi-tive and negative feedbacks and free-scale pattern distri-bution in rural-population dynamics PLOS ONE 9 e114561
Allen CD Macalady AK Chenchouni H Bachelet D andothers (2010) A global overview of drought and heat-induced tree mortality reveals emerging climate changerisks for forests For Ecol Manag 259 660minus684
Amidžic L Bartula M Jaric S (2012) Syntaxonomy overviewof vegetation of Šar Planina in the Balkan PeninsulaVegetos 25 348minus360
Aune S Hofgaard A Soumlderstroumlm L (2011) Contrasting cli-mate and land use driven tree encroachment pattern ofsub-arctic tundra in Northern Norway and Kola Penin-sula Can J For Res 41 437minus449
Barbeito I Dawes M Rixen C Senn J Bebi P (2012) Factorsdriving mortality and growth at treeline a 30-year exper-iment of 92000 conifers Ecology 93 389minus401
Batllori E Camarero JJ Ninot JM Gutieacuterrez E (2009)Seedling recruitment survival and facilitation in alpinePinus uncinata treeline ecotones Implications andpotential responses to climate warming Glob Ecol Bio-geogr 18 460minus472
Bebi P Teich M Hagedorn F Zurbriggen N Brunner SHGreacutet-Regamey A (2012) Veraumlnderung von Wald undWaldleistungen in der Landschaft Davos im Zuge desKlimawandels Schweiz Z Forstwes 163 493minus501
Bell DM Bradford JB Lauenroth WK (2014) Mountain land-scapes offer few opportunities for high elevation treespecies migration Glob Change Biol 20 1441minus1451
Bergant K (2006) Climate change scenarios for MacedoniaUniversity of Nova Gorica Center for Atmospheric
147
Clim Res 73 135ndash150 2017
Research Nova GoricaBodin J Badeau V Bruno E Cluzeau C Moisselin JM
Walther GR Dupouey JL (2013) Shifts of forest speciesalong an elevational gradient in Southeast France cli-mate change or stand maturation J Veg Sci 24 269minus283
Boncina A (2011) History current status and future pros -pects of uneven-aged forest management in the Dinaricregion an overview Forestry 84 467minus478
Breshears DD Cobb NS Rich PM Price KP and others(2005) Regional vegetation die-off in response to global-change-type drought Proc Natl Acad Sci USA 102 15144minus15148
Camarero JJ Garciacutea-Ruiz JM Sanguumlesa-Barreda G GalvaacutenJD and others (2015a) Recent and intense dynamics in aformerly static Pyrenean treeline Arct Antarct Alp Res47 773minus783
Camarero JJ Gazol A Galvaacuten JD Sanguumlesa-Barreda GGutieacuterrez E (2015b) Disparate effects of global-changedrivers on mountain conifer forests warming-inducedgrowth enhancement in young trees vs CO2 fertilizationin old trees from wet sites Glob Change Biol 21 738minus749
Chen IC Hill JK Ohlemuumlller R Roy DB Thomas CD (2011)Rapid range shifts of species associated with high levelsof climate warming Science 333 1024minus1026
Cudliacuten P Sejaacutek J Pokornyacute J Albrechtovaacute J Bastian OMarek M (2013) Forest ecosystem services under climatechange and air pollution In Matyssek R Clarke N Cud-liacuten P Mikkelsen TN Tuovinen JP Wiesner G Paoletti E(eds) Climate change air pollution and global chal-lenges understanding and perspectives from forestresearch Developments in Environmental Science Vol13 Elsevier Oxford p 521minus546
Dalen L Hofgaard A (2005) Differential regional treelinedynamics in the Scandes mountains Arct Antarct AlpRes 37 284minus296
Dawes MA Philipson CD Fonti P Bebi P Haumlttenschwiler SHagedorn F Rixen C (2015) Soil warming and CO2
enrichment induce biomass shifts in alpine tree line veg-etation 2014 Glob Change Biol 21 2005minus2021
De Frenne P Rodriacuteguez-Saacutenchez F De Schrijver A CoomesDA Hermy M Vangansbeke P Verheyen K (2015) Lightaccelerates plant responses to warming Nat Plants 1 15110
Devi N Hagedorn F Moiseev P Bugmann H ShiyatovSG Mazepa V Rigling A (2008) Expanding forests andchanging growth forms of Siberian larch at the treeline ofthe Polar Urals during the 20th century Glob ChangeBiol 14 1581minus1591
Diaz-Varela R Colombo R Meroni M Calvo-Iglesias MSBuffoni A Tagliaferri A (2010) Spatio-temporal analysisof alpine ecotones A spatial explicit model targeting altitudinal vegetation shifts Ecol Model 221 621minus633
Dirnboumlck T Dullinger S Grabherr G (2003) A regionalimpact assessment of climate and land-use change onalpine vegetation J Biogeogr 30 401minus417
Dobrowski SZ Swanson AK Abatzoglou JT Holden ZASafford HD Schwartz MK Gavin DG (2015) Forest struc-ture and species traits mediate projected recruitmentdeclines in western US tree species Glob Ecol Biogeogr24 917minus927
Dullinger S Dirnboumlck S Grabherr G (2004) Modelling cli-mate change-driven treeline shifts relative effects oftemperature increase dispersal and invisibility J Ecol92 241minus252
Dullinger S Gattringer A Thuiller W Moser D and others
(2012) Extinction debt of high-mountain plants undertwenty-first-century climate change Nat Clim Chang 2 619minus622
Elkin C Gutierrez AG Leuzinger S Manusch C TemperliC Rasche L Bugmann H (2013) A 2degC warmer world isnot safe for ecosystem services in the European AlpsGlob Change Biol 19 1827minus1840
Ellenberg H (1988) Vegetation ecology of Central Europe4th edn Cambridge University Press Edinburgh
Em H (1986) Spruce southern distribution range Spruce for-est on Shara Mountain in Macedonia Contrib Macedon-ian Acad Sci Arts Skopje 5 11minus28
Fajardo A Piper FI Pfund L Koumlrner CH Hoch G (2012)Variation of mobile carbon reserves in trees at the alpinetreeline ecotone is under environmental control NewPhytol 195 794minus802
Fridley JD Grime JP Askew AP Moser B Stevens CJ (2011)Soil heterogeneity buffers community response to cli-mate change in species-rich grassland Glob Change Biol17 2002minus2011
Gartzia M Alados CL Peacuterez-Cabello F (2014) Assessment ofthe effects of biophysical and anthropogenic factors onwoody plant encroachment in dense and sparse moun-tain grasslands based on remote sensing data Prog PhysGeogr 38 201minus217
Gehrig-Fasel J Guisan A Zimmermann NE (2007) Tree lineshifts in the Swiss Alps climate change or land abandon-ment J Veg Sci 18 571minus582
Gikov A Dimitrov P Zhiyanski M (2016) Land cover dynam-ics on the northern slope of the Troyan Passage during 30years period Probl Geogr 1-2 78minus92
Gonzalez P Neilson RP Lenihan JM Drapek RJ (2010)Global patterns in the vulnerability of ecosystems to veg-etation shifts due to climate change Glob Ecol Biogeogr19 755minus768
Gottfried M Pauli H Futschik A Akhalkatsi M and others(2012) Continent-wide response of mountain vegetationto climate change Nat Clim Chang 2 111minus115
Grace J Berninger F Nagy L (2002) Impacts of climatechange on the tree line Ann Bot 90 537minus544
Grunewald K Scheithauer J (2011) Landscape developmentand climate change in Southwest Bulgaria (Pirin Mts)Springer Heidelberg
Grytnes JA Kapfer J Jurasinski G Birks HH and others(2014) Identifying the driving factors behind observedelevational range shifts on European mountains GlobEcol Biogeogr 23 876minus884
Hagedorn F Shiyatov FG Mazepa VS Devi NM and others(2014) Treeline advances along the Urals mountainrange mdash driven by improved winter conditions GlobChange Biol 20 3530minus3543
Hansen AJ Neilson RP Dale VH Flather CH and others(2001) Global change in forests responses of speciescommunities and biomes BioScience 51 765minus779
Harsch MA Bader MY (2011) Treeline formminusa potential keyto understanding treeline dynamics Glob Ecol Biogeogr20 582minus596
Harsch MA Hulme PE McGlone MS Duncan RP (2009) Aretreelines advancing A global meta-analysis of treelineresponse to climate warning Ecol Lett 12 1040minus1049
Haylock MR Hofstra N Klein Tank AMG Klok EJ JonesPD New M (2008) A European daily high-resolutiongridded dataset of surface temperature and precipita-tion J Geophys Res Atmos 113 D20119
Hlaacutesny T Barcza Z Fabrika M Balaacutezs B and others (2011)
148
Cudliacuten et al Drivers of treeline shift
Climate change impacts on growth and carbon balanceof forests in Central Europe Clim Res 47 219minus236
Hlaacutesny T Trombik J Dobor L Barcza Z Barka I (2016)Future climate of the Carpathians climate change hot-spots and implications for ecosystems Reg EnvironChange 16 1495minus1506
Hofgaard A Dalen L Hytteborn H (2009) Tree recruitmentabove the treeline and potential for climate driven tree-line change J Veg Sci 20 1133minus1144
Holtmeier FK (2009) Mountain timberlines Ecology patchi-ness and dynamics 2nd edn Advances in GlobalChange Research 36 Springer Science amp ScienceMediaBV Dordrecht
Holtmeier FK Broll G (2005) Sensitivity and response ofnorthern hemisphere altitudinal and polar treelines toenvironmental change at landscape and local scalesGlob Ecol Biogeogr 14 395minus410
Honnay O Verheyen K Butaye J Jacquemyn H Bossuyt BHermy M (2002) Possible effects of habitat fragmentationand climate change on the range of forest plant speciesEcol Lett 5 525minus530
Huber R Rigling A Bebi P Brand F and others (2013) Sus-tainable land use in mountain regions under globalchange synthesis across scales and disciplines Ecol Soc18 36
Ibaacutentildeez I Clark JS Dietze MC (2009) Estimating colonizationpotential of migrant tree species Glob Change Biol 15 1173minus1188
IUCNSSC (International Union for Conservation of NatureSpecies Survival Commission) (2013) Guidelines for re -introductions and other conservation translocations Version 10 IUCNSpecies Survival Commission Gland
Jankovskyacute L Cudliacuten P gtermaacutek P Moravec I (2004) The pre-diction of development of secondary Norway sprucestands under the impact of climatic change in the Dra-hany highlands (The Czech Republic) Ekologia (Bratisl)23(Suppl 2) 101minus112
Jeniacutek J Lokvenc T (1962) Die alpine Waldgrenze im Krko -noše Gebirge Rozpravy SAV Rada matematickyacutech apriacuterodniacutech ved 72 gteskoslovenskaacute akademie ved Praha
Jobbaacutegy EG Jackson RB (2000) Global controls of forest lineelevation in the northern and southern hemispheresGlob Ecol Biogeogr 9 253minus268
Jonaacutešovaacute M Vaacutevrovaacute E Cudliacuten P (2010) Western Carpa -thian mountain spruce forest after a windthrow naturalregeneration in cleared and uncleared areas For EcolManag 259 1127minus1134
Kittel TGF Steffen WL Chapin FS (2000) Global andregional modelling of Arcticminusboreal vegetation distribu-tion and its sensitivity to altered forcing Glob ChangeBiol 6 1minus18
Klopcic M Jerina K Bonltina A (2010) Long-term changes ofstructure and tree species composition in Dinaric un -even- aged forests Are red deer an important factor EurJ For Res 129 277minus288
Klopcic M Simonltilt T Bonltina A (2015) Comparison of re -generation and recruitment of shade-tolerant and light-demanding tree species in mixed uneven-aged forests ex -periences from the Dinaric region Forestry 88 552minus563
Koumlrner C (2003) Alpine plant life functional plant ecology ofhigh mountain ecosystems with 47 tables Springer Sci-ence amp Business Media BV Dordrecht
Koumlrner C (2012) Alpine treelines Functional ecology of theglobal high elevation tree limits Springer Basel
Koumlrner C Paulsen J (2004) A world-wide study of high alti-
tude treeline temperatures J Biogeogr 31 713minus732Kulakowski D Bebi P Rixen C (2011) The interacting effects
of land use change climate change and suppression ofnatural disturbances on landscape forest structure in theSwiss Alps Oikos 120 216minus225
Kullman L (1988) Holocene history of the forestminusalpine tun-dra ecotone in the Scandes Mountains (central Sweden)New Phytol 108 101minus110
Kullman L (1999) Early holocene tree growth at a high elevation site in the northernmost Scandes of Sweden(Lapland) a palaeobiogeographical case study based onmega fossil evidence Geogr Ann 81 63minus74
Kullman L (2007) Tree line population monitoring of Pinussylvestris in the Swedish Scandes 1973minus2005 implica-tions for tree line theory and climate change ecologyJ Ecol 95 41minus52
Kullman L Oumlberg L (2009) Post-Little Ice Age treeline riseand climate warming in the Swedish Scandes a land-scape ecological perspective J Ecol 97 415minus429
Kyriazopoulos AP Skre O Sarkki S Wielgolaski FE Abra-ham EM Ficko A (2017) Humanminusenvironment dynamicsin European treeline ecosystems a synthesis based onthe DPSIR framework Clim Res 73 17ndash29
Lenoir J Svenning JC (2015) Climate-related range shifts mdasha global multidimensional synthesis and new researchdirections Ecography 38 15minus38
Lenoir J Geacutegout JC Marquet PA de Ruffray P Brisse H(2008) A significant upward shift in plant species opti-mum elevation during the 20th century Science 320 1768minus1771
Lindner M Fitzgerald JB Zimmermann NE Reyer C andothers (2014) Climate change and European forests What do we know what are the uncertainties and whatare the implications for forest management J EnvironManag 146 69minus83
Lloyd AH Rupp TS Fastie CL Srafield AM (2002) Patternsand dynamic of treeline advance on the Seward Pen -isula Alaska J Geophys Res 107 8161
Maacuteliš F Kopeckyacute M Petriacutek P Vladovic J Merganic J VidaT (2016) Life-stage not climate change explains ob -served tree range shifts Glob Change Biol 22 1904minus1914
Mathisen IE Mikheeva A Tutubalina OV Aune S and Hof-gaard A (2014) Fifty years of tree line change in theKhibiny Mountains Russia advantages of combinedremote sensing and dendroecological approaches ApplVeg Sci 17 6ndash16
Millar CI Stephenson NL (2015) Temperate forest healthin an era of emerging megadisturbance Science 349 823minus826
Mina M Bugmann H Klopcic M Cailleret M (2017) Accu-rate modeling of harvesting is key for projecting futureforest dynamics a case study in the Slovenian moun-tains Reg Environ Change 17 49minus64
Moen J Aune K Edenius L Angerbjoumlrn A (2004) Potentialeffects of climate change on treeline position in theSwedish mountains Ecol Soc 9 16 wwwecologyandsocietyorgvol9iss1art16
Palombo C Chirici G Marchetti M Tognetti R (2013) Is landabandonment affecting forest dynamics at high elevationin Mediterranean mountains more than climate changePlant Biosyst 147 1minus11
Palombo C Marchetti M Tognetti R (2014) Mountain vege-tation at risk current perspectives and research reedsPlant Biosyst 148 35minus41
Pauli H Gottfried M Dullinger S Abdaladze O Akhalkatsi
149
Clim Res 73 135ndash150 2017
M Alonso JLB Grabherr G (2012) Recent plant diversitychanges on Europersquos mountain summits Science 336 353minus355
Payette S Fortin MJ Gamachet I (2001) The subarctic forestminustundra the structure of a biome in a changing climate BioScience 51 709minus718
Pearson GR Dawson TP (2003) Predicting the impacts of cli-mate change on the distribution of species Are bio -climate envelope models useful Glob Ecol Biogeogr 12 361minus371
Pedrotti F (2013) Plant and vegetation mapping GeobotanyStudies Springer-Verlag Berlin
Pentildeuelas J Boada M (2003) A global change-induced biomeshift in the Montseny mountains (NE Spain) GlobChange Biol 9 131minus140
Piovesan G Biondi F Filippo AD Alessandrini A MaugeriM (2008) Drought driven growth reduction in old beech(Fagus sylvatica L) forests of the central ApenninesItaly Glob Change Biol 14 1265minus1281
Rabasa SG Granda E Benavides R Kunstler G and others(2013) Disparity in elevational shifts of European trees inresponse to recent climate warming Glob Change Biol19 2490minus2499
Raev I Zhelev P Grozeva M Markov I and others (2011)Programme of measures for adaptation of forests in Bul-garia and mitigate the negative impact of climate changeon them Project FUTURE forest helping Europe tackleclimate change INTERREG IVC ERDF Sofia
Sarkki S Ficko A Grunewald K Nijnik M (2016) Benefitsfrom and threats to European treeline ecosystem serv-ices an exploratory study of stakeholders and gover-nance Reg Environ Change 16 2019minus2032
Savage J Vellend M (2015) Elevational shifts biotic homo -genization and time lags in vegetation change during 40years of climate warming Ecography 38 546minus555
Schwoumlrer C Henne PD Tinner W (2014) A model-data com-parison of Holocene timberline changes in the SwissAlps reveals past and future drivers of mountain forestdynamics Glob Change Biol 20 1512minus1526
Šmilauer P Lepš J (2014) Multivariate analysis of ecologicaldata using Canoco 5 Cambridge University Press Cam-bridge
Ste-Marie C Nelson EA Dabros A Bonneau ME (2011)Assisted migration introduction to a multifaceted con-cept For Chron 87 724minus730
Stevens JT Stafford HD Harrison S Latimer AM (2015) For-est disturbance accelerates thermophilization of under-story plant communities J Ecol 103 1253minus1263
Strid A Andonoski A Andonovski V (2003) Alpine biodiver-sity in Europe Ecological Studies 167 Springer-VerlagBerlin
Tattoni C Ciolli M Ferretti F Cantiani MG (2010) Monitor-ing spatial and temporal pattern of Paneveggio forest(Northern Italy) from 1859 to 2006 iForest Biogeosci For3 72minus80
Ter Braak CJF Šmilauer P (2012) Canoco reference manualand userrsquos guide software for ordination (version 50)Microcomputer Power Ithaca NY
Theurillat JP Guisan A (2001) Potential impact of climatechange on vegetation in the European Alps a reviewClim Change 50 77minus109
Toslashmmervik H Wielgolaski FE Neuvonen S Solberg BHoslashgda KA (2005) Biomass and production on a land-scape level in the mountain birch forests In WielgolaskiFE (ed) Plant ecology herbivory and human impact in
Nordic mountain birch forests Ecos Studies 180Springer Berlin p 53minus70
Treml V Banaš M (2000) Alpine timberline in the HighSudetes Acta Univ Carol Geogr 15 83minus99
Treml V Chuman T (2015) Ecotonal dynamics of the altitu-dinal forest limit are affected by terrain and vegetationstructure variables an example from the Sudetes Moun-tains in Central Europe Arct Antarct Alp Res 47 133minus146
Treml V Migo P (2015) controlling factors limiting timber-line position and shifts in the Sudetes a review GeogrPol 88 55minus70
Urban MC Tewksbury JJ Sheldon KS (2012) On a collisioncourse competition and dispersal differences create no-analogue communities and cause extinctions during cli-mate change Proc R Soc Lond B Biol Sci 279 2072minus2080
Vacek S Matejka K (2010) State and development of phyto-cenoses on research plots in the Krkonoše Mts foreststands J For Sci 56 505minus517
Van Bogaert R Haneca K Hoogesteger J Jonasson C deDapper M Callaghan TV (2011) A century of tree linechanges in sub-Arctic Sweden shows local and regionalvariability and only a minor influence of 20th century climate warming J Biogeogr 38 907minus921
Van Gils H Batsukh O Rossiter D Munthali W Liberato -scioli E (2008) Forecasting the pattern and pace of Fagusforest expansion in Majella National Park Italy ApplVeg Sci 11 539minus546
Velchev V (1997) Types of vegetation In Yordanova MDonchev D (eds) Geography of Bulgaria BAN Sofiap 269minus283 (in Bulgarian)
Vladovic J Merganic J Maacuteliš F Križovaacute E and others (2014)The response of forest vegetation diversity to changes inedaphic-climatic conditions in Slovakia Technickaacute uni-verzita vo Zvolene Zvolene (in Slovak)
Vrška T Adam D Hort L Kolaacuter T Janiacutek F (2009) Europeanbeech (Fagus sylvatica L) and silver fir (Abies alba Mill)rotation in the Carpathians mdash a developmental cycle or alinear trend induced by man For Ecol Manag 258 347minus356
Wielgolaski FE (ed) (2005) Plant ecology herbivory andhuman impact in Nordic mountain birch forests Eco -logical Studies 180 Springer Verlag Berlin
Wielgolaski FE Hofgaard A Holtmeier FK (2017) Sensitivityto environmental change of the treeline ecotone and itsassociated biodiversity in European mountains Clim Res73151ndash166
Wieser G Tausz M (2007) Trees at their upper limit Treelifelimitation at the alpine timberline Springer Dordrecht
Wieser G Holtmeier FK Smith WK (2014) Treelines in achanging global environment In Tausz M Grulke N(eds) Trees in a changing environment Plant Ecophys 9Springer Dordrecht p 221minus263
Zhiyanski M Kolev K Sokolovska M Hursthouse A (2008)Tree species effect on soils in Central Stara PlaninaMountains Nauka Gorata 4 65minus82
Zhiyanski M Gikov A Nedkov S Dimitrov P Naydenova L(2016) Mapping carbon storage using land coverlanduse data in the area of Beklemeto Central Balkan In Koulov B Zhelezov G (eds) Sustainable mountain re -gions challenges and perspectives in southeasternEurope Springer Basel p 53minus65
Zhu K Woodall CW Clark JS (2012) Failure to migrate lackof tree range expansion in response to climate changeGlob Change Biol 18 1042minus1052
150
Editorial responsibility Nils Chr Stenseth Oslo Norway
Submitted April 8 2016 Accepted March 8 2017Proofs received from author(s) July 7 2017
Clim Res 73 135ndash150 2017138
Country Coordi- Elevation Elevation Range of Annual Area VZs near Tr part of nates of range range of Tr tree species altitudinal of study (m asl)mountains central site of T (m asl) limit shift (km2)
(m asl) (m asl) of T or Tr (m yrminus1)
Spain 42deg37rsquoN 1930 No data Pu 2700 049 (T 15 800 Pine with Aa Spanish 0deg27rsquoE (1800minus2050) 1956minus2006) Ps and Fs (1400minus2050) Pyrenees subalpine with Pu
(1800minus2300)
Italy 42deg05rsquoN 1800 No data Fs 2100 1 (Tr [Pm ] 741 Beech forest (800minus1850) Majella Massif 14deg05rsquoE (1700minus1850) Pm 2300 1954minus2007) mixed beechminusdwarf pine part of (1800minus2000) dwarf pine Apennines (2100minus2300)
Macedonia 41deg47rsquorsquoN 1850 1900 Pa 2100 ~1 (T ) 829 Mixed firminusbeechminusShara Mountain 20deg33rsquoE (1800minus1900) (1850minus1950) Pp 2200 1934minus2010 spruce (800minus1900) (Balkan beech (1600minus1900) Peninsula) dwarf pine (1600minus2200)
Bulgaria 41deg42rsquoN 1900 2100 Pp 2500 No data 384 Beech (1000minus1500) Pirin 23deg31rsquoE (1800minus2100) (1900minus2300) coniferous forest (Ps Pp
Ph Pa 1300minus2300) dwarf pine (2000minus2500)
Bulgaria 42deg47rsquoN Beech T 1600 1600 Fs 1700 No data 717 Beech (800minus1700) Central 24deg36rsquoE (1500minus1700) (1500ndash1850) Pa Aa 1850 coniferous (Pa Aa Balkan Mts 1500minus1850) juniper
(1500minus1850)
Slovenia 45deg36rsquoN 1540 1600 Fs 1650 No data ~200 Mixed firminusbeechminusspruce Northern 14deg28rsquoE (1455minus1600) (1550minus1650) Pa 1700 (600minus1400) beech (1300minus1550) Dinaric Mts Aa 1650 dwarf pine with small groups
or individual trees of Fs Pa Aa Ap (1500minus1700)
Switzer- 46deg46rsquoN 1900 2100 Pa 2400 ca13 (T ~250 Spruce with Ld land Italy 9deg52rsquoE (1700minus2150) (1850minus2300) Ld 2450 1954minuspresent) + ~750 (1000minus1900) forest with Pc Central and 46deg17rsquoN Ld and Pa (1700minus2200) Eastern Alps 11deg45rsquoE dwarf shrub vegetation partly
pastures (2200minus2500)
Table 1 Timberline and treeline parameters and characteristics of human activities changes according to climatic models andlimits to climate change induced species shifts in selected European mountains A1B A2 emissions scenarios (IPPC) AaAbies alba Ap Acer pseudoplatanus Fs Fagus sylvatica Ld Larix decidua Pa Picea abies Pc Pinus cembra Ph Pi-nus heldreichii Pm Pinus mugo Pp Pinus peuce Ps Pinus sylvestris Pu Pinus uncinata T timberline Tr treeline TSL
tree species limit VZ vegetation zone C century mng management Elevation range of T and Tr mean (minminusmax)
Cudliacuten et al Drivers of treeline shift 139
Human Annual Limits Referencesinfluence temperature (T ) to climate
in VZs in past and precipitation (P) change and recently differences induced species
between 1961minus1990 shiftand 2021minus2050
mean of CORDEX models
Since 11th C HIRHAM5 model Inadequacy of Batllori et al (2009) clear-cutting and grazing period ΔT2021minus2050 substrata (rocky) Alados et al (2014)
moderate on irregular = 15minus2degC ΔT2051minus2080 climatic constraints Gartzia et a (2014) slopes and ridges since 1930 = 4degC ΔP2021minus2050 (including growing- Camarero et al (2015a)significant release of influence = +5 ΔP2051minus2080 season extreme events
but recovery mainly at = +30 compared to such as late-spring moderate elevations 1960minus1990 freezing summer
drought etc)
Since 1000 BC HadCM3_A2 period Low density of Fs van Gils et al (2008) cutting burning and 2020minus2080 in T dispersal distances Palombo et al (2013 2014)
grazing mng since the mid-20th ΔTmin_Jan2050 = 17degC of Fs seeds upslope C grazing intensity decreased ΔTmin_Jan2080 = 31degC poor soils excessive
now occasionally managed ΔP2050 = minus7 exposure to solar radiation and regeneration of Pm ΔP2080 = minus23 in open areas
Since 14th C clear-cutting Mean of 4 GCMs Differences in soil types Em (1986) and grazing to enlarge (CSIROMk2 HadCM3 between subalpine beech Strid et al (2003)
alpine pastures and to produce ECHAM4OPYC3 and mixed forest (poor Bergant (2006) charcoal since 1970 NCAR minus PCM) ΔT2050 = calcareous soils or rankers on Amidžic et al (2012)
abandonment of traditional 26degC ΔT2100 = 53degC silicate bedrock) cliffs agricultural practices ΔP2050 = minus3 ΔP2100 = minus8 and rocks cattle grazing
Since 680 AD intensive HADCM3_A2 Rocky sites extreme Velchev (1997) fires deforestation and grazing CGGM2 hydrothermal conditions Grunewald amp
since 1962 traditional mng ΔT2050 = 17degC competition of shrubs and tree Scheithauer (2011) has changed last decades tourism ΔT2080 = 28degC seedlings locally intense Raev et al (2011)
impact has increased ΔP2080 = minus15 pasturing
Human impact (mostly burning) HADCM3_A2 Low density of Fs Raev et al (2011) started in 17th C excessive CGCM2 ΔT2050 = 17degC in T dispersal distances Gikov et al (2016) and improper forest mng ΔT2080 = 28degC of Fs seeds upslope pasture Zhiyanski et al
release of influence since 1980 ΔP2050 = minus15 intensity new (2008 2016)ΔP2080 = minus15 expansion of juniper stands
15minus18th C slash burn DMI-HIRHAM5_A1B Impact of wild large Klopltilt et al grazing mng 18minus19th C period 2001minus2100 ΔT2050 ungulates on tree seedlings (2010 2015)
uncontrolled cutting 20th C = 12degC ΔT2100 = 25degC (especially Aa) shallow Mina et al (2017)irregular shelterwood systems ΔP2050 = +15 mm nutrient-poor soils
and over-exploitation mng ΔP2100 = minus18 mm (ie ranker and rendzina)
Since 12th C slash burn grazing mng ENSEMBLES models_ Frequent disturbances Tattoni et al (2010) 13minus19th C intensive forest A1B period 1980minus2100 by snow avalanches Kulakowski et al (2011)
mng at the end of the 19th C ΔT2050 = 12minus18degC and snow gliding ongoing Barbeito et al (2012) decrease in grazing ΔT2100 = 27minus41degC cattle grazing game grazing Bebi et al (2012)
ΔP2050 = 0 mm in the coldest years Dawes et al (2015)ΔP2100 = minus5 mm
Continued on next page
Clim Res 73 135ndash150 2017
mate the geographical position of these mountainunits 1 value was used for the geographical latitudeand longitude of the locus of the studied mountainunits The analysed parameters of these 15 mountainunits are shown in Table 2
In addition we calculated the following climatecharacteristics for all mountain units annual meanair temperature annual sum of precipitation thegrowing season length and the date of the onset ofthe growing season These characteristics were thencompared for 2 distinct periods 1961minus1990 and1991minus 2015 (Table 3) The climate characteristicswere derived from the E-OBS gridded dataset of sta-tion observations version 131 (Haylock et al 2008)We used the E-OBS version on the regular longi-tudeminuslatitude grid with a horizontal resolution of 025degrees The climate data from all grid points withinthe mountain units or near their geographical bor-ders were considered
Data processing for 15 mountain units was made onthe basis of data obtained from the literature and
personal studies by the co-authors these are sum ma -rized in Tables 1 amp 2 and in Tables S1 amp S2 in theSupplement at www int-res com articles suppl c073p135 _ supp pdf In addition a semi-quantitative valu-ation of abiotic biotic and anthropogenic factors lim-iting an upward treeline shift was performed (Fig 2)To answer the question how natural and anthro-pogenic factors have influenced treeline ecotonecharacteristics with a focus on treeline shift redun-dancy analysis (RDA) was used to describe and testthe effect of the explanatory variables (geographicalposition size of the whole mountain massif start ofhuman influence and start of the decrease in humaninfluence) on treeline ecotone characteristics (tim-berline elevation width of the treeline ecotone treespecies forming the treeline and treeline shift peryear) and identify groups of mountain units with sim-ilar variability of the dependent data (ter Braak ampŠmilauer 2012) The whole data set for this analysis isshown in Table S2 We tested both their simple ef -fects which show how much variation every ex plan -
140
Country Coordi- Elevation Elevation Range of Annual Area VZs near Tr part of nates of range range of Tr tree species altitudinal of study (m asl)mountains central site of T (m asl) limit shift (km2)
(m asl) (m asl) of T or Tr (m yrminus1)
Slovakia 48deg55rsquoN 1400 1530 Pa 1730 ~03 (Tr ~400 Spruceminusfirminusbeech Dumbier Tatra 19deg31rsquoE (1350minus1450) (1400minus1630) 1950minuspresent) (1200minus1350) spruce (1300minus(part of Low 1550) Pm with individual Tatra Mts) Pa trees (1400minus1700)
Czech Republic 50deg42rsquoN 1250 1320 Pa 1500 043 (Tr 550 Beechminusspruce (700minus1050) Giant Mts 15deg38rsquoE (1130minus1460) (1250minus1460) 1936minus2005) Pa dominates with increasing
altitude spruce (1000minus1400) Pm with Pa trees (1400minus1560)
Norway 60deg10rsquoN 850 600minus1050 1500 08 (Tr 40 600 Birch forest dominance Southern 7deg40rsquoE 62deg20rsquoN 1915minus2007) at high altitudes but with Scandes 10deg05rsquoE some scattered Ps or Pa
mixed birchminusconifer below 800
Norway 68deg20rsquoN West 650 West 750 West 1000 06 (Tr 35 000 Birch forest dominance Sweden 18deg20rsquoE East 100 East 290 East 0 1958minus2008) at all altitudes but with some Finland 69deg50rsquoN scattered pines or pine stands Northern 27deg00rsquoE on sandy soilsScandes
Table 1 (continued)
Cudliacuten et al Drivers of treeline shift
atory variable can explain separately without usingthe other variables and their conditional ef fectswhich depend on the variables already selected inthe model The statistical significance of the expla -natory variables was tested using a Monte Carlo permutation test and only predictors with p le 005were included in the subsequent RDA (Šmilauer ampLepš 2014)
To find the drivers of biodiversity loss in forestsmeadows and animal communities (dependent vari-ables) the explanatory variables (geographical posi-tion timberline elevation tree species forming thetreeline start of human influence start of the de -crease in human influence and temperature increasebetween the 2 periods 1961minus1990 and 2021minus2050)were tested again by RDA in Canoco 5 as mentionedabove (Figs 3 amp 4) The whole data set for this analy-sis mdash including additional information about tree-lines in the mountain areas of interest with a focuson treeline shift (its rate drivers limits and problemswith its assessment) mdash is presented in Tables S1 amp S2
Selected univariate graphs were constructed to visu-alize the data of relationships between the explana-tory and dependent variables of all analyses whichwere not seen distinctly from the results of the multi-variate statistics or linear regression (Fig 5)
3 RESULTS
31 Effect of environmental variables on treeline ecotone characteristics with a focus
on treeline shift
The variation in the dependent variables (timber-line elevation width of treeline ecotone tree speciesforming the treeline and treeline shift per year) wassignificantly affected only by latitude and size of thewhole mountain massif (Fig 3) The adjusted ex -plained variability by all explanatory variables was342 (F = 46 p = 001) The RDA diagram showsthat in the mountains located more to the north the
141
Human Annual Limits Referencesinfluence temperature (T ) to climate
in VZs in past and precipitation (P) change and recently differences induced species
between 1961minus1990 shiftand 2021minus2050
mean of CORDEX models
From 14th C to 1922 Average of 10 RCM_B1 More rocky and nutrient- Koumlrner (2003) mining forest change ΔT2050 = 18degC poor soils late frosts Fridley et al (2011)
to spruce monocultures ΔT2100 = 37degC steep slopes with avalanches Hlaacutesny et al (2011 2016)from 13th C to 1978 pastures ΔP2050 = +24 mm absence of mature trees
in upper parts of mountain ΔP2100 = minus35 mm
9minus11th C forest clear cutting ALADIN_A1B period 1961minus Different soil types Treml and Banaš (2000) grazing in dwarf pine VZ 2100 ΔT2050 = 13degC for Fs and Pa (cambisol Cudliacuten et al (2013)
since 18th C forest change ΔT2100 = 34degC versus podzol) only TSL of Pa Treml amp Chuman (2015) to spruce monocultures ΔP2050 = +25 mm could elevate on ranker Treml amp Migo (2015)
since 19th C artificially planted ΔP2100 = minus14 mm soils in Pm VZ
Grazed and browsed RegClimMPIHadley Sheep and reindeer Dalen amp Hofgaard (2005) by reindeer and livestock ΔT2050 = 12degC grazing and episodic Wielgolaski (2005)
(sheep and cattle) ΔT2100 = 22degC insect outbreaks (Epirrita) Hofgaard et al (2009) over thousands of years ΔP2050 = +23 mm Kullman Oumlberg (2009)
ΔP2100 = minus12 mm
Grazed and browsed RegClimMPIHadley Continued grazing regime Dalen amp Hofgaard (2005) by semi-domestic reindeer ΔT2050 = 16degC and frequent episodic Wielgolaski (2005)
for hundreds of years ΔT2100 = 29degC insect outbreaks (Epirrita) Toslashmmervik et al (2005) increased grazing from ΔP2050 = +18 mm Hofgaard et al (2009)
year 1960 ΔP2100 = minus10 mm Aune et al (2011) Mathisen et al (2014)
Clim Res 73 135ndash150 2017
treeline ecotone is narrower treeline shiftis smaller the timberline occurs at lowerelevations and the treeline is formed moreby broadleaved species Higher values oftreeline ecotone width and treeline shiftwere found in the mountains with a greatersize of the whole mountain massif
When every explanatory variable wastested separately values of the start of thedecrease of human influence in the moun-tains also had a significant effect on thevariation of the tree line ecotone data(explained variation = 206 pseudo-F =34 p = 0024) Some selected relationshipsof the explanatory and dependent vari-ables even those not showing significanteffects on the variation in the tree line eco-tone parameters are depicted in Fig 5
32 Effect of recent changes in climateand land use on the biodiversity
Biodiversity loss in forests meadows andanimal communities analysed by RDAwas explained by geo graphical positiontreeline species composition and tempera-ture increase between the 2 periods1961minus1990 and 2021minus2050 (Fig 4) Theadjusted ex plained variation by all ex pla -natory variables was 632 (F = 58 p =0001) The RDA results indicated that bio-diversity loss in forest communities in crea -sed with increasing latitude and longitudeas well as in the case where broad leavedspecies formed the treeline In contrast thehighest biodiversity loss in meadows wasfound in the mountains positioned more tothe south and with higher temperatureincrease The biodiversity loss in animalcommunities was negatively correlatedwith longitude and temperature increase
4 DISCUSSION
The comparison of several mountainareas situated across Europe shows a varia-tion in understanding of what constitutesthe treeline across countries The primaryissue is the different approach to the defini-tion of forest when applying a tree heightthreshold This threshold decreases from
142
Mou
nta
in u
nit
Nor
th
Eas
t M
oun
-T
imb
er-
Wid
th
Tre
e T
emp
erat
ure
T
reel
ine
Sta
rt
Sta
rt
lati
tud
elo
ng
itu
de
tain
lin
eof
spec
ies
incr
ease
sh
ift
of H
I of
HI
size
(m a
sl
)tr
eeli
ne
form
ing
b
etw
een
(m
yrminus
1 )(y
r)d
ecre
ase
(km
2 )ec
oton
eth
e 19
61minus
1990
(y
r)(m
)tr
eeli
ne
and
199
1minus20
15 (
degC)
Cen
tral
Pyr
enee
s (C
P)
42deg2
1rsquoN
0deg30
rsquoW55
000
2250
300
C0
81
92 (
1956
minus20
06)
1000
BC
1930
Eas
tern
Pyr
enee
s (E
P)
42deg2
8rsquoN
1deg30
rsquoE55
000
2400
300
C0
80
71 (
1956
minus20
06)
1000
1930
Ap
enn
ines
(A
P)
42deg0
5rsquoN
14deg0
5rsquoE
5000
018
0030
0B
14
1 (1
954minus
2007
)10
00 B
C19
50S
har
a M
ts (
SH
)41
deg47rsquo
N20
deg33rsquo
E16
0018
5035
0C
B0
71
(193
4minus20
10)
1300
1970
Pir
in (
PI)
41deg4
2rsquoN
23deg3
1rsquoE
2585
1900
400
C0
5n
o d
ata
680
1962
Cen
tral
Bal
kan
Mts
(C
B)
42deg4
7primeN
24deg3
6primeE
1160
016
0025
0C
B0
7n
o d
ata
1700
1980
Nor
ther
n D
inar
ic M
ts (
DI)
45deg3
6rsquoN
14deg2
8rsquoE
1000
1540
160
B1
3n
o d
ata
1400
1900
Cen
tral
Alp
s (C
A)
46deg4
6rsquoN
9deg52
rsquoE18
000
021
0055
0C
07
13
(195
4minus20
15)
1300
1880
Eas
tern
Alp
s (E
A)
46deg1
7rsquoN
11deg4
5rsquoE
180
000
1800
400
C0
91
3 (1
954minus
2015
)11
5019
50L
ow T
atra
Mts
(LT
)48
deg55rsquo
N19
deg31rsquo
E12
4014
0033
0C
09
057
(19
75minus
2016
)14
0019
22G
ian
t M
ts (
GM
)50
deg42rsquo
N15
deg38rsquo
E63
012
5025
0C
07
043
(19
36minus
2005
)10
0019
63H
ard
ang
ervi
dd
a (H
G)
60deg0
3rsquoN
7deg47
rsquoE90
0090
015
0C
B0
50
8 (1
915minus
2007
)80
019
70D
ovre
(D
O)
62deg2
1rsquoN
8deg52
rsquoE60
0010
0040
0B
07
08
(191
5minus20
07)
1000
no
dec
reas
eN
orth
ern
Sw
edis
h L
apla
nd
(N
S)
68deg1
2rsquoN
18deg3
3rsquoE
4000
700
300
B1
10
6 (1
958minus
2008
)10
00n
o d
ecre
ase
Inn
er F
inn
mar
kn
orth
ern
mos
t 69
deg10rsquo
N25
deg12rsquo
E60
000
290
90B
10
06
(195
8minus20
08)
1600
no
dec
reas
eF
inn
ish
Lap
lan
d (
FL
)
Tab
le 2
Fac
tors
infl
uen
cin
g t
reel
ine
ecot
one
char
acte
rist
ics
in s
elec
ted
Eu
rop
ean
mou
nta
ins
Th
e m
oun
tain
siz
e is
th
e si
ze o
f th
e w
hol
e m
oun
tain
mas
sive
fu
nct
ion
ing
as a
mez
osca
le c
lim
ate
un
it n
ot o
nly
the
stu
die
d m
oun
tain
are
a W
idth
of t
he
tree
lin
e ec
oton
e w
as d
eter
min
ed b
y su
btr
acti
ng
the
valu
e of
the
mea
n ti
mb
erli
ne
elev
atio
n
from
th
e va
lue
of t
he
up
per
tre
e sp
ecie
s li
mit
C c
onif
ers
B b
road
leav
ed t
rees
HI
hu
man
infl
uen
ce
Cudliacuten et al Drivers of treeline shift
5 m (Jeniacutek amp Lokvenc 1962) to 3 m(Koumlrner 2012) in Northern andCentral Europe to only 2 m (Holt-meier 2009) in Central and South-ern Europe where even shrub -by stands (eg Pinus mugo) areconsi dered as a forest in somecountries eg in Spain Italy andMa ce donia (Batllori et al 2009)An other problem is a different de -signation of forest stands belowthe treeline lsquoupper montane fo -restsrsquo in Central Europe and lsquosub-alpine forestsrsquo in Southern Europe(Ellenberg 1988) Further differen -ces are related to the tradition ofdifferent branches of science (ega more traditional lsquogeobotanicalrsquoapproach in Central Europe ver-sus a more experimental approachin Western Europe) especially inthe rate of applying new pro -gressive methods (eg climaticmodelling or molecular biologicalmethods)
Our comparison of selectedregions (Tables 1 amp 2 Figs 1 amp 2)showed that southern mountainscompared to those located morecentrally and in the north had (1) alonger and more profound exploi -tation by humans in the past (2)greater differences in climatic para -meters (temperature length ofgrowing period) between the 2periods 1961minus1990 and 1991minus2015(Table 3) and (3) more dramaticclimate change scenarios espe-cially concerning temperature in -creases Longitude also had someinfluence on the start and intensityof human influence and on tree-line elevation and rate of treelineshift (Table 2 Fig 5) The occur-rence of broadleaf tree species inthe treeline ecotone in the north-ern (and to some extent the south-ern) countries distinguishes thesefrom the Central European coun-tries where conifers prevail It isinteresting that grazing an im -portant factor shaping treelineecosystems (Dirnboumlck et al 2003
143
Length Beginning Mean Mean of growing of growing annual annual
period season temperature precipitation (d)a (d)b (degC) ()
Central Pyrenees 135 minus89 08 minus107Eastern Pyrenees 117 minus91 08 minus92Apennines 284 minus232 14 57Shara Mts 16 01 07 minus100Pirin 26 minus14 05 27Central Balkan Mts 169 minus136 07 76Northern Dinaric Mts 160 minus95 13 minus56Central Alps 147 minus86 07 62Eastern Alps 125 minus96 09 minus33Low Tatra Mts 24 minus34 09 minus91Giant Mts minus10 13 07 55Hardangervidda Blefjell 63 34 05 56Dovre 138 32 07 64Northern Swedish Lapland 81 05 11 minus27Inner Finnmarknorthern- 17 08 10 77most Finnish Lapland
aPositive numbers indicate a prolongation bNegative numbers indicate an earlier beginning
Table 3 Differences of selected climatic parameters between the 2 study periods (1961minus1990 and 1991minus2015) in selected European mountains
Tein Drou Gefa Nadi Laus Abli Bili Huli
0 1 2 3
Central Pyrenees (CP)
Eastern Pyrenees (EP)
Apennines (AP)
Shara Mts (SH)
Pirin (PI)
Central Balkan Mts (CB)
Northern Dinaric Mts (DI)
Central Alps (CA)
Eastern Alps (EA)
Low Tatra Mts (LT)
Giant Mts (GM)
Hardangervidda (HG)
Dovre (DO)
Northern Swedish Lapland (NS)
Inner FinnmarknorthernmostFinnish Lapland (FL)
Fig 2 Rates of treeline drivers (temperature increase [Tein] land use change[Laus]) and treeline shift limits (drought [Drou] geomorphological factors [Gefa]natural disturbances [Nadi] other abiotic biotic and human factors) in selectedEuropean mountains Abli Bili and Huli abiotic biotic and human treeline shiftlimits respectively) Rate of treeline driver influence mdash 0 no influence 1 weak
influence 2 middle influence 3 strong influence
Clim Res 73 135ndash150 2017
Gehrig-Fasel et al 2007) limits treeline shift in halfof the countries of interest regardless of latitude lon-gitude or recent political developments (Shara PirinCentral Balkans Alps Scandes Fig 2)
Despite the stated differences between studieslooking at climate change impacts on the treelineit is clear that current and future changes in temp -erature will seriously affect treeline ecotones in allmountain ranges of Europe Winter is a period whentreeline stands and individual trees have to survivesevere life-limiting conditions (Wieser amp Tausz2007) Therefore winter warming might increase thechance of young trees surviving and passing the mostcritical period from seedling to sapling and further tothe mature stage (Koumlrner 2003) There is alreadymuch evidence worldwide that winter warming isone of the significant drivers of treeline advance(Harsch et al 2009) Although it is expected that treegrowth at the upper distributional margins (and inthe northern European countries) will increase due toongoing climate change (Pentildeuelas amp Boada 2003Chen et al 2011 Hlaacutesny et al 2011 Lindner et al2014) extremes in temperature (eg black froststemperature reversals in the spring) or winter precip-itation (eg lack of snow drought) might also limit
this process in the future in some regions and localsituations (Holtmeier amp Broll 2005 Hagedorn et al2014) An earlier start to the growing period (espe-cially in the Apennines Central Balkan and SpanishPyrenees see Table 2) can play a negative role in theresistance of trees and seedlings to early springfrosts On the other hand at lower distributional mar-gins (and in Southern European countries) it isexpected that tree growth will decrease or forestswill experience some level of dieback due to drought(Breshears et al 2005 Piovesan et al 2008 Allen etal 2010 Huber et al 2013) A serious implication isthe positive effect that warming can have on root rotfungi and populations of bark beetles resulting inlarge-scale disturbances to temperate and also high-altitude forests (Jankovskyacute et al 2004 Elkin et al2013 Millar amp Stephenson 2015)
Although recent upward advances of treeline eco-tones are widespread in mountain regions (Harsch et
144
Fig 3 Redundancy analysis diagram with variation in tim-berline elevation width of treeline ecotone the tree speciesforming the treeline and treeline shift used as dependentvariables explained by explanatory variables (latitudemountain size) Mountain units (blue) are projected as thecentres of abbreviations (see Table 2) Explanatory variablesaccount for 436 of the total variation in the dependentdata The first canonical axis explained 466 of variationthe second axis explained 30 of variation Dependentvariables (black) are Ecot altitudinal width of treeline eco-tone TLShift treeline shift TrspC (TrspB) treeline formedby conifers (broadleaved trees) Timb timberline elevationExplanatory variables (red) are N north latitude Mtsize
size of the mountain massif
Fig 4 Redundancy analysis diagram with variation in bio-diversity loss in forests meadows and animal communitiesused as dependent variables explained by explanatory vari-ables (geographical position tree species forming the tree-line and temperature increase between the 2 study periods[1961minus1990 and 2021minus2050]) Mountain units (green) areprojected as centres of abbreviations (see Table 2) Explana-tory variables account for 763 of total variation in the de-pendent data The first canonical axis explained 484 ofvariation the second axis explained 183 of variation De-pendent variables (black) are BdfoBdmeBdan biodiversityloss in forestsmeadowsanimal communities Explanatoryvariables (red) are N north latitude E east longitude TrspC(TrspB) treeline formed by conifers (broadleaved trees)Temp temperature increase between the 2 study periods
Cudliacuten et al Drivers of treeline shift
al 2009) only a few published studies have includedquantitative data about treeline shifts (Devi et al2008 Kullman amp Oumlberg 2009 Diaz-Varela et al 2010Van Bogaert et al 2011) Additionally our know -ledge of the spatial patterns in treeline ecotone shiftsat the landscape scale is still surprisingly poor exceptfor some studies from subarctic areas the UralMountains and the Alps (Lloyd et al 2002 Diaz-Varela et al 2010 Hagedorn et al 2014) In the 15studied mountain units the values of treeline shiftranged from 043 to 19 m yrminus1 showing a rather dis-tinct spatial pattern of the dynamics within Europeanmountains The observed treeline shift had a signifi-cant positive relationship only with northern latitudeand a weak positive relationship with mountain-range size (ie the size of the whole mountain massifFig 3) Nevertheless this illustrates the importanceof the mass elevation effect (Koumlrner 2012) and par-tially ex plains why treelines could be at different alti-tudes at the same latitude Small negative regres-sions with east longitude and timberline elevationare shown in Fig 5 There was no apparent depend-ence of the treeline shift rate on climatic parameterchanges between the periods 1961minus1990 and1991minus2015
A whole forest vegetation zone shift is a complexprocess as not only the life strategies of an individualtree need to be considered but plant animal andmicroorganism species and their interrelationshipsas well as the relationships to their specific microhab-
itats (including soil conditions) are also involved(Urban et al 2012) Across all regions we identifiedseveral important obstacles to treeline shifts oftenrelated to site properties such as significant rocki-ness having shallow or low-nutrient soils extremerelief causing disturbances by avalanches snow glid-ing or wind damage (data from the Pyrenees Apen-nines Shara Northern Dinaric Low Tatra and GiantMountains see Fig 2) Soil heterogeneity certainlyhas an important role in plant responses to climatechange and could also maintain the resilience of thecommunity assemblages (Fridley et al 2011) An -other group of obstacles includes extreme climaticparameters especially in winter (reported eg fromthe Pyrenees Apennines Pirin Central BalkanNorthern Dinaric and Giant Mountains Fig 2) Theircombination can result in edaphic andor climaticunsuitability of habitats where species could poten-tially migrate Climatically and edaphically suitablesites will likely decline over the next century partic-ularly in mountain landscapes (Bell et al 2014)There fore trees in the treeline ecotone will colonizepreviously forested habitats The colonization suc-cess of individual forest communities is affected bydifferences in species dispersal and recruitmentbehaviour (Dullin ger et al 2004 Jonaacutešovaacute et al2010) For these reasons colonization of new foresthabitats by the next tree generation may not be suc-cessful and may result in loss of species diversity(Honnay et al 2002 Ibaacutentildeez et al 2009 Dobrowski et
145
Fig 5 Univariate graphs of the relationships of dependent variables viz timberline (Timb) and treeline shift per year (Shift) and their explanatory variables Nola north latitude Ealo east longitude DD decimal degrees
Clim Res 73 135ndash150 2017
al 2015) as reported from Macedonia Slovenia theCzech Republic Slovakia and Norway (Fig 4)
Another limiting factor is the existence of strongadaptation mechanisms of the dominant tree specieswhich allow them to survive in their current distribu-tion areas Species migration is likely to be slow dueto the limited quantity of climatically and edaphicallysuitable sites and the slow velocity of seed dispersaland tree regeneration and will be hampered evenmore by fragmentation of high mountain landscapescaused by human activities including brush invasionin abandoned meadows (Gartzia et al 2014) A verti-cal shift in a vegetation zone involves not only singlespecies of plants animals and microorganisms butalso their interrelationships and links to soil condi-tions The speed at which climatic conditions changewill be different from changes in the soil conditionsdue to the slowness of soil-forming processes Soiltypes and conditions (cambisol versus podzol) arecrucial obstacles for the shift from a beech forest zoneinto the spruce forest zone in the Czech Republic(Vacek amp Mat jka 2010) According to other sourcesthe dominant tree species can influence soil for -mation processes (especially humus forms) Underfavourable climatic and orographic conditions beechis able to change its soil conditions over decades tocenturies (J Macku unpubl data)
Significant changes in biodiversity must be expec -ted in all of the mountain areas of interest We foundthat biodiversity declined particularly in the southernregions of Europe where the timberline is situated athigher elevations and human impact is mostlylonger-lasting (Fig 4) Similarly Pauli et al (2012)reported an increase in species richness of mountaingrasslands across Europe except for Mediterraneanregions and assigned this different response of thesouthern regions to decreased water availability Onthe other hand ecosystems which exhibited bio -diversity increases were not only enriched by mig -rant species but the assemblages underwent thermo -philization ie cold-adapted species declined andmore warm-adapted species spread (Gottfried et al2012) However not all species are able to track cli-mate changes It is expected that around 40 ofhabitats will become climatically unsuitable for manymountain species particularly endemic ones in -creasing their extinction probability during this cen-tury (Dullinger et al 2012) The abandonment of tra-ditional land use forms is another source of this losswhich may be a more important driver than tempera-ture increase (Fig 4) Serious biodiversity losses inmountain meadows were reported from Spain ItalyMacedonia Bulgaria Slovenia and Slovakia There-
fore controlled grazing must occur in order to main-tain alpine grasslands (Dirnboumlck et al 2003) Unfor-tunately grazing is still active only in smaller parts ofEuropean mountains (eg in the Central Alps and theCentral Balkan Mountains but pasturing can alsonegatively affect vegetation diversity) and some-times does not serve to maintain alpine grasslandThe same is true for grassland management whichcan help to maintain mountain species and decreasehabitat loss A serious biodiversity loss in mountainmeadows was reported from the Pyrenees the Apen-nines and the Shara Pirin Central Balkan NorthernDinaric and Low Tatra Mountains (Fig 4)
In forests species responses to climate change maybe equivocal some recent studies indicated contra-dictory shifts in species distributions (eg Lenoir etal 2008 Zhu et al 2012 Rabasa et al 2013) This dis-parity is widely discussed and assigned to the greatvariety of non-climatic factors or even tree ontogeny(Grytnes et al 2014 Lenoir amp Svenning 2015 Maacutelišet al 2016) Disturbances leading to tree mortalitymay also play an important role (Cudliacuten et al 2013)changes in tree canopy cover modify light availabil-ity and microclimate and can induce the thermo -philization of forest vegetation (De Frenne et al2015 Stevens et al 2015) The loss of this microcli-mate buffering effect of forests may induce a biotichomogenization of forests (Savage amp Vellend 2015)or the creation of novel non-analogical communitiessuch as oakminuspine forests (Urban et al 2012) whichmay be a new threat to forest biodiversity
The observed changes in treeline forest ecosystemsare often related to changes in land-use intensity(Theurillat amp Guisan 2001 Alados et al 2014) Ac -cording to climate change predictions and the recentand future exploitation intensity of European moun-tains trees and forest communities will shift upwarddue to land use change or climate change or bothReduced land-use intensity certainly will interactwith climate change by facilitating or inhibiting spe-cies occurrence and will accelerate forest expansionabove the present treeline particularly to previouslyforested habitats (Theurillat amp Guisan 2001) Thesimultaneous action of both main treeline shift driv-ers viz temperature increase and decrease in landuse intensity was recorded from all of our studiedmountain areas The extensive differences in timber-line and treeline elevations in almost all studiedmountains (Table 1) indicate the anthropo-zoogenictreeline type (according to Ellenberg 1998) In mostmountains (eg in the Apennines Shara MountainsCentral Balkans and Alps) land use is the prevailingfactor influencing vegetation drift (Fig 5) Previous
146
Cudliacuten et al Drivers of treeline shift
research showed that the upward shift of the treelinein the Swiss Alps was predominantly attributable toland abandonment and only in some situations to cli-mate change (Gehrig-Fasel et al 2007) In the Apen-nines in the last few decades tree establishment hasbeen mainly controlled by land use while treegrowth has been controlled by climate pointing to aminor role played by climate in shaping the currenttreeline (Palombo et al 2013)
The impact of climate change and the connectedland use change on biodiversity and ecosystem serv-ices provision in several European countries is sum-marized by Wielgolaski et al (2017) and Kyriazopou-los et al (2017) (both this Special) and Sarkki et al(2016) One of the possible adaptive management op-tions in response to climate change assisted migra-tion as human-assisted movement of species hasbeen frequently debated in the last few years (Ste-Marie et al 2011) It is possible to apply it as a type ofassisted colonization ie intentional movement andrelease of an organism outside its indigenous range toavoid extinction of populations of the focal species(eg Macedonian pine species) or as an ecologicalre placement ie the intentional movement and re-lease of an organism outside its indigenous range toperform a specific ecological function (eg planting ofPinus mugo in the Alps IUCNSSC 2013)
5 CONCLUSIONS
The analysis of 11 mountain areas across Europeshowed that with increasing latitude the treelineand altitudinal width of the treeline ecotone signifi-cantly decreases as do the significance of climaticand soil parameters as barriers against tree speciesshift Although temperature is the overwhelmingcontrolling factor of tree growth and establishment intemperate and boreal treeline ecotones late-sea-sonal drought might also play a driving role in Medi-terranean treeline ecotones Longitude was lessinfluential mostly affecting climate and land usechange effects on increased biodiversity loss as wellas the size of the area that forms one mesoclimateunit affecting altitudinal ecotone width The biggestpart of the commonly observed remaining variabilityin mountain vegetation near the treeline in Europeseems to be caused by geomorphological geologicalpedological and microclimatic variability in combina-tion with different land use history and the presentsocio-economic relations The observed differencesin climatic parameters between the mountain areasof interest in comparison with the reference period
1960minus1990 (09degC) have not explained the relativelyhigh differences in the rate of treeline shift per year(149 m) The predicted variability in temperatureincrease due to global warming (12minus26degC in 2050and 22minus53degC in 2100) could lead to a much biggerdifferentiation in treeline ecotone biodiversity andecosystem processes between southern and northernEuropean mountains in the future Therefore thesedifferences must be taken into account by scientistsand EU policy makers when formulating efficientadaptive forest management strategies for treelineecosystems at the European level (eg assistedmigration of adapted genotypes)
Acknowledgements This paper is based firstly upon workfrom the COST Action ES 1203 SENSFOR supported byCOST (European Cooperation in Science and Technologywwwcosteu) This international work was further sup-ported by projects granted by the Ministry of EducationYouth and Sports of the Czech Republic grant NPU ILO1415 and LD 14039 by the agency APVV SR under pro-jects APVV-14-0086 and APVV-15-0270 We acknowledgethe E-OBS dataset from the EU-FP6 project ENSEMBLES(httpensembles-eumetofficecom) and the data providersin the ECAampD project (wwwecadeu)
LITERATURE CITED
Alados CL Errea P Gartzia M Saiz H Escoacutes J (2014) Posi-tive and negative feedbacks and free-scale pattern distri-bution in rural-population dynamics PLOS ONE 9 e114561
Allen CD Macalady AK Chenchouni H Bachelet D andothers (2010) A global overview of drought and heat-induced tree mortality reveals emerging climate changerisks for forests For Ecol Manag 259 660minus684
Amidžic L Bartula M Jaric S (2012) Syntaxonomy overviewof vegetation of Šar Planina in the Balkan PeninsulaVegetos 25 348minus360
Aune S Hofgaard A Soumlderstroumlm L (2011) Contrasting cli-mate and land use driven tree encroachment pattern ofsub-arctic tundra in Northern Norway and Kola Penin-sula Can J For Res 41 437minus449
Barbeito I Dawes M Rixen C Senn J Bebi P (2012) Factorsdriving mortality and growth at treeline a 30-year exper-iment of 92000 conifers Ecology 93 389minus401
Batllori E Camarero JJ Ninot JM Gutieacuterrez E (2009)Seedling recruitment survival and facilitation in alpinePinus uncinata treeline ecotones Implications andpotential responses to climate warming Glob Ecol Bio-geogr 18 460minus472
Bebi P Teich M Hagedorn F Zurbriggen N Brunner SHGreacutet-Regamey A (2012) Veraumlnderung von Wald undWaldleistungen in der Landschaft Davos im Zuge desKlimawandels Schweiz Z Forstwes 163 493minus501
Bell DM Bradford JB Lauenroth WK (2014) Mountain land-scapes offer few opportunities for high elevation treespecies migration Glob Change Biol 20 1441minus1451
Bergant K (2006) Climate change scenarios for MacedoniaUniversity of Nova Gorica Center for Atmospheric
147
Clim Res 73 135ndash150 2017
Research Nova GoricaBodin J Badeau V Bruno E Cluzeau C Moisselin JM
Walther GR Dupouey JL (2013) Shifts of forest speciesalong an elevational gradient in Southeast France cli-mate change or stand maturation J Veg Sci 24 269minus283
Boncina A (2011) History current status and future pros -pects of uneven-aged forest management in the Dinaricregion an overview Forestry 84 467minus478
Breshears DD Cobb NS Rich PM Price KP and others(2005) Regional vegetation die-off in response to global-change-type drought Proc Natl Acad Sci USA 102 15144minus15148
Camarero JJ Garciacutea-Ruiz JM Sanguumlesa-Barreda G GalvaacutenJD and others (2015a) Recent and intense dynamics in aformerly static Pyrenean treeline Arct Antarct Alp Res47 773minus783
Camarero JJ Gazol A Galvaacuten JD Sanguumlesa-Barreda GGutieacuterrez E (2015b) Disparate effects of global-changedrivers on mountain conifer forests warming-inducedgrowth enhancement in young trees vs CO2 fertilizationin old trees from wet sites Glob Change Biol 21 738minus749
Chen IC Hill JK Ohlemuumlller R Roy DB Thomas CD (2011)Rapid range shifts of species associated with high levelsof climate warming Science 333 1024minus1026
Cudliacuten P Sejaacutek J Pokornyacute J Albrechtovaacute J Bastian OMarek M (2013) Forest ecosystem services under climatechange and air pollution In Matyssek R Clarke N Cud-liacuten P Mikkelsen TN Tuovinen JP Wiesner G Paoletti E(eds) Climate change air pollution and global chal-lenges understanding and perspectives from forestresearch Developments in Environmental Science Vol13 Elsevier Oxford p 521minus546
Dalen L Hofgaard A (2005) Differential regional treelinedynamics in the Scandes mountains Arct Antarct AlpRes 37 284minus296
Dawes MA Philipson CD Fonti P Bebi P Haumlttenschwiler SHagedorn F Rixen C (2015) Soil warming and CO2
enrichment induce biomass shifts in alpine tree line veg-etation 2014 Glob Change Biol 21 2005minus2021
De Frenne P Rodriacuteguez-Saacutenchez F De Schrijver A CoomesDA Hermy M Vangansbeke P Verheyen K (2015) Lightaccelerates plant responses to warming Nat Plants 1 15110
Devi N Hagedorn F Moiseev P Bugmann H ShiyatovSG Mazepa V Rigling A (2008) Expanding forests andchanging growth forms of Siberian larch at the treeline ofthe Polar Urals during the 20th century Glob ChangeBiol 14 1581minus1591
Diaz-Varela R Colombo R Meroni M Calvo-Iglesias MSBuffoni A Tagliaferri A (2010) Spatio-temporal analysisof alpine ecotones A spatial explicit model targeting altitudinal vegetation shifts Ecol Model 221 621minus633
Dirnboumlck T Dullinger S Grabherr G (2003) A regionalimpact assessment of climate and land-use change onalpine vegetation J Biogeogr 30 401minus417
Dobrowski SZ Swanson AK Abatzoglou JT Holden ZASafford HD Schwartz MK Gavin DG (2015) Forest struc-ture and species traits mediate projected recruitmentdeclines in western US tree species Glob Ecol Biogeogr24 917minus927
Dullinger S Dirnboumlck S Grabherr G (2004) Modelling cli-mate change-driven treeline shifts relative effects oftemperature increase dispersal and invisibility J Ecol92 241minus252
Dullinger S Gattringer A Thuiller W Moser D and others
(2012) Extinction debt of high-mountain plants undertwenty-first-century climate change Nat Clim Chang 2 619minus622
Elkin C Gutierrez AG Leuzinger S Manusch C TemperliC Rasche L Bugmann H (2013) A 2degC warmer world isnot safe for ecosystem services in the European AlpsGlob Change Biol 19 1827minus1840
Ellenberg H (1988) Vegetation ecology of Central Europe4th edn Cambridge University Press Edinburgh
Em H (1986) Spruce southern distribution range Spruce for-est on Shara Mountain in Macedonia Contrib Macedon-ian Acad Sci Arts Skopje 5 11minus28
Fajardo A Piper FI Pfund L Koumlrner CH Hoch G (2012)Variation of mobile carbon reserves in trees at the alpinetreeline ecotone is under environmental control NewPhytol 195 794minus802
Fridley JD Grime JP Askew AP Moser B Stevens CJ (2011)Soil heterogeneity buffers community response to cli-mate change in species-rich grassland Glob Change Biol17 2002minus2011
Gartzia M Alados CL Peacuterez-Cabello F (2014) Assessment ofthe effects of biophysical and anthropogenic factors onwoody plant encroachment in dense and sparse moun-tain grasslands based on remote sensing data Prog PhysGeogr 38 201minus217
Gehrig-Fasel J Guisan A Zimmermann NE (2007) Tree lineshifts in the Swiss Alps climate change or land abandon-ment J Veg Sci 18 571minus582
Gikov A Dimitrov P Zhiyanski M (2016) Land cover dynam-ics on the northern slope of the Troyan Passage during 30years period Probl Geogr 1-2 78minus92
Gonzalez P Neilson RP Lenihan JM Drapek RJ (2010)Global patterns in the vulnerability of ecosystems to veg-etation shifts due to climate change Glob Ecol Biogeogr19 755minus768
Gottfried M Pauli H Futschik A Akhalkatsi M and others(2012) Continent-wide response of mountain vegetationto climate change Nat Clim Chang 2 111minus115
Grace J Berninger F Nagy L (2002) Impacts of climatechange on the tree line Ann Bot 90 537minus544
Grunewald K Scheithauer J (2011) Landscape developmentand climate change in Southwest Bulgaria (Pirin Mts)Springer Heidelberg
Grytnes JA Kapfer J Jurasinski G Birks HH and others(2014) Identifying the driving factors behind observedelevational range shifts on European mountains GlobEcol Biogeogr 23 876minus884
Hagedorn F Shiyatov FG Mazepa VS Devi NM and others(2014) Treeline advances along the Urals mountainrange mdash driven by improved winter conditions GlobChange Biol 20 3530minus3543
Hansen AJ Neilson RP Dale VH Flather CH and others(2001) Global change in forests responses of speciescommunities and biomes BioScience 51 765minus779
Harsch MA Bader MY (2011) Treeline formminusa potential keyto understanding treeline dynamics Glob Ecol Biogeogr20 582minus596
Harsch MA Hulme PE McGlone MS Duncan RP (2009) Aretreelines advancing A global meta-analysis of treelineresponse to climate warning Ecol Lett 12 1040minus1049
Haylock MR Hofstra N Klein Tank AMG Klok EJ JonesPD New M (2008) A European daily high-resolutiongridded dataset of surface temperature and precipita-tion J Geophys Res Atmos 113 D20119
Hlaacutesny T Barcza Z Fabrika M Balaacutezs B and others (2011)
148
Cudliacuten et al Drivers of treeline shift
Climate change impacts on growth and carbon balanceof forests in Central Europe Clim Res 47 219minus236
Hlaacutesny T Trombik J Dobor L Barcza Z Barka I (2016)Future climate of the Carpathians climate change hot-spots and implications for ecosystems Reg EnvironChange 16 1495minus1506
Hofgaard A Dalen L Hytteborn H (2009) Tree recruitmentabove the treeline and potential for climate driven tree-line change J Veg Sci 20 1133minus1144
Holtmeier FK (2009) Mountain timberlines Ecology patchi-ness and dynamics 2nd edn Advances in GlobalChange Research 36 Springer Science amp ScienceMediaBV Dordrecht
Holtmeier FK Broll G (2005) Sensitivity and response ofnorthern hemisphere altitudinal and polar treelines toenvironmental change at landscape and local scalesGlob Ecol Biogeogr 14 395minus410
Honnay O Verheyen K Butaye J Jacquemyn H Bossuyt BHermy M (2002) Possible effects of habitat fragmentationand climate change on the range of forest plant speciesEcol Lett 5 525minus530
Huber R Rigling A Bebi P Brand F and others (2013) Sus-tainable land use in mountain regions under globalchange synthesis across scales and disciplines Ecol Soc18 36
Ibaacutentildeez I Clark JS Dietze MC (2009) Estimating colonizationpotential of migrant tree species Glob Change Biol 15 1173minus1188
IUCNSSC (International Union for Conservation of NatureSpecies Survival Commission) (2013) Guidelines for re -introductions and other conservation translocations Version 10 IUCNSpecies Survival Commission Gland
Jankovskyacute L Cudliacuten P gtermaacutek P Moravec I (2004) The pre-diction of development of secondary Norway sprucestands under the impact of climatic change in the Dra-hany highlands (The Czech Republic) Ekologia (Bratisl)23(Suppl 2) 101minus112
Jeniacutek J Lokvenc T (1962) Die alpine Waldgrenze im Krko -noše Gebirge Rozpravy SAV Rada matematickyacutech apriacuterodniacutech ved 72 gteskoslovenskaacute akademie ved Praha
Jobbaacutegy EG Jackson RB (2000) Global controls of forest lineelevation in the northern and southern hemispheresGlob Ecol Biogeogr 9 253minus268
Jonaacutešovaacute M Vaacutevrovaacute E Cudliacuten P (2010) Western Carpa -thian mountain spruce forest after a windthrow naturalregeneration in cleared and uncleared areas For EcolManag 259 1127minus1134
Kittel TGF Steffen WL Chapin FS (2000) Global andregional modelling of Arcticminusboreal vegetation distribu-tion and its sensitivity to altered forcing Glob ChangeBiol 6 1minus18
Klopcic M Jerina K Bonltina A (2010) Long-term changes ofstructure and tree species composition in Dinaric un -even- aged forests Are red deer an important factor EurJ For Res 129 277minus288
Klopcic M Simonltilt T Bonltina A (2015) Comparison of re -generation and recruitment of shade-tolerant and light-demanding tree species in mixed uneven-aged forests ex -periences from the Dinaric region Forestry 88 552minus563
Koumlrner C (2003) Alpine plant life functional plant ecology ofhigh mountain ecosystems with 47 tables Springer Sci-ence amp Business Media BV Dordrecht
Koumlrner C (2012) Alpine treelines Functional ecology of theglobal high elevation tree limits Springer Basel
Koumlrner C Paulsen J (2004) A world-wide study of high alti-
tude treeline temperatures J Biogeogr 31 713minus732Kulakowski D Bebi P Rixen C (2011) The interacting effects
of land use change climate change and suppression ofnatural disturbances on landscape forest structure in theSwiss Alps Oikos 120 216minus225
Kullman L (1988) Holocene history of the forestminusalpine tun-dra ecotone in the Scandes Mountains (central Sweden)New Phytol 108 101minus110
Kullman L (1999) Early holocene tree growth at a high elevation site in the northernmost Scandes of Sweden(Lapland) a palaeobiogeographical case study based onmega fossil evidence Geogr Ann 81 63minus74
Kullman L (2007) Tree line population monitoring of Pinussylvestris in the Swedish Scandes 1973minus2005 implica-tions for tree line theory and climate change ecologyJ Ecol 95 41minus52
Kullman L Oumlberg L (2009) Post-Little Ice Age treeline riseand climate warming in the Swedish Scandes a land-scape ecological perspective J Ecol 97 415minus429
Kyriazopoulos AP Skre O Sarkki S Wielgolaski FE Abra-ham EM Ficko A (2017) Humanminusenvironment dynamicsin European treeline ecosystems a synthesis based onthe DPSIR framework Clim Res 73 17ndash29
Lenoir J Svenning JC (2015) Climate-related range shifts mdasha global multidimensional synthesis and new researchdirections Ecography 38 15minus38
Lenoir J Geacutegout JC Marquet PA de Ruffray P Brisse H(2008) A significant upward shift in plant species opti-mum elevation during the 20th century Science 320 1768minus1771
Lindner M Fitzgerald JB Zimmermann NE Reyer C andothers (2014) Climate change and European forests What do we know what are the uncertainties and whatare the implications for forest management J EnvironManag 146 69minus83
Lloyd AH Rupp TS Fastie CL Srafield AM (2002) Patternsand dynamic of treeline advance on the Seward Pen -isula Alaska J Geophys Res 107 8161
Maacuteliš F Kopeckyacute M Petriacutek P Vladovic J Merganic J VidaT (2016) Life-stage not climate change explains ob -served tree range shifts Glob Change Biol 22 1904minus1914
Mathisen IE Mikheeva A Tutubalina OV Aune S and Hof-gaard A (2014) Fifty years of tree line change in theKhibiny Mountains Russia advantages of combinedremote sensing and dendroecological approaches ApplVeg Sci 17 6ndash16
Millar CI Stephenson NL (2015) Temperate forest healthin an era of emerging megadisturbance Science 349 823minus826
Mina M Bugmann H Klopcic M Cailleret M (2017) Accu-rate modeling of harvesting is key for projecting futureforest dynamics a case study in the Slovenian moun-tains Reg Environ Change 17 49minus64
Moen J Aune K Edenius L Angerbjoumlrn A (2004) Potentialeffects of climate change on treeline position in theSwedish mountains Ecol Soc 9 16 wwwecologyandsocietyorgvol9iss1art16
Palombo C Chirici G Marchetti M Tognetti R (2013) Is landabandonment affecting forest dynamics at high elevationin Mediterranean mountains more than climate changePlant Biosyst 147 1minus11
Palombo C Marchetti M Tognetti R (2014) Mountain vege-tation at risk current perspectives and research reedsPlant Biosyst 148 35minus41
Pauli H Gottfried M Dullinger S Abdaladze O Akhalkatsi
149
Clim Res 73 135ndash150 2017
M Alonso JLB Grabherr G (2012) Recent plant diversitychanges on Europersquos mountain summits Science 336 353minus355
Payette S Fortin MJ Gamachet I (2001) The subarctic forestminustundra the structure of a biome in a changing climate BioScience 51 709minus718
Pearson GR Dawson TP (2003) Predicting the impacts of cli-mate change on the distribution of species Are bio -climate envelope models useful Glob Ecol Biogeogr 12 361minus371
Pedrotti F (2013) Plant and vegetation mapping GeobotanyStudies Springer-Verlag Berlin
Pentildeuelas J Boada M (2003) A global change-induced biomeshift in the Montseny mountains (NE Spain) GlobChange Biol 9 131minus140
Piovesan G Biondi F Filippo AD Alessandrini A MaugeriM (2008) Drought driven growth reduction in old beech(Fagus sylvatica L) forests of the central ApenninesItaly Glob Change Biol 14 1265minus1281
Rabasa SG Granda E Benavides R Kunstler G and others(2013) Disparity in elevational shifts of European trees inresponse to recent climate warming Glob Change Biol19 2490minus2499
Raev I Zhelev P Grozeva M Markov I and others (2011)Programme of measures for adaptation of forests in Bul-garia and mitigate the negative impact of climate changeon them Project FUTURE forest helping Europe tackleclimate change INTERREG IVC ERDF Sofia
Sarkki S Ficko A Grunewald K Nijnik M (2016) Benefitsfrom and threats to European treeline ecosystem serv-ices an exploratory study of stakeholders and gover-nance Reg Environ Change 16 2019minus2032
Savage J Vellend M (2015) Elevational shifts biotic homo -genization and time lags in vegetation change during 40years of climate warming Ecography 38 546minus555
Schwoumlrer C Henne PD Tinner W (2014) A model-data com-parison of Holocene timberline changes in the SwissAlps reveals past and future drivers of mountain forestdynamics Glob Change Biol 20 1512minus1526
Šmilauer P Lepš J (2014) Multivariate analysis of ecologicaldata using Canoco 5 Cambridge University Press Cam-bridge
Ste-Marie C Nelson EA Dabros A Bonneau ME (2011)Assisted migration introduction to a multifaceted con-cept For Chron 87 724minus730
Stevens JT Stafford HD Harrison S Latimer AM (2015) For-est disturbance accelerates thermophilization of under-story plant communities J Ecol 103 1253minus1263
Strid A Andonoski A Andonovski V (2003) Alpine biodiver-sity in Europe Ecological Studies 167 Springer-VerlagBerlin
Tattoni C Ciolli M Ferretti F Cantiani MG (2010) Monitor-ing spatial and temporal pattern of Paneveggio forest(Northern Italy) from 1859 to 2006 iForest Biogeosci For3 72minus80
Ter Braak CJF Šmilauer P (2012) Canoco reference manualand userrsquos guide software for ordination (version 50)Microcomputer Power Ithaca NY
Theurillat JP Guisan A (2001) Potential impact of climatechange on vegetation in the European Alps a reviewClim Change 50 77minus109
Toslashmmervik H Wielgolaski FE Neuvonen S Solberg BHoslashgda KA (2005) Biomass and production on a land-scape level in the mountain birch forests In WielgolaskiFE (ed) Plant ecology herbivory and human impact in
Nordic mountain birch forests Ecos Studies 180Springer Berlin p 53minus70
Treml V Banaš M (2000) Alpine timberline in the HighSudetes Acta Univ Carol Geogr 15 83minus99
Treml V Chuman T (2015) Ecotonal dynamics of the altitu-dinal forest limit are affected by terrain and vegetationstructure variables an example from the Sudetes Moun-tains in Central Europe Arct Antarct Alp Res 47 133minus146
Treml V Migo P (2015) controlling factors limiting timber-line position and shifts in the Sudetes a review GeogrPol 88 55minus70
Urban MC Tewksbury JJ Sheldon KS (2012) On a collisioncourse competition and dispersal differences create no-analogue communities and cause extinctions during cli-mate change Proc R Soc Lond B Biol Sci 279 2072minus2080
Vacek S Matejka K (2010) State and development of phyto-cenoses on research plots in the Krkonoše Mts foreststands J For Sci 56 505minus517
Van Bogaert R Haneca K Hoogesteger J Jonasson C deDapper M Callaghan TV (2011) A century of tree linechanges in sub-Arctic Sweden shows local and regionalvariability and only a minor influence of 20th century climate warming J Biogeogr 38 907minus921
Van Gils H Batsukh O Rossiter D Munthali W Liberato -scioli E (2008) Forecasting the pattern and pace of Fagusforest expansion in Majella National Park Italy ApplVeg Sci 11 539minus546
Velchev V (1997) Types of vegetation In Yordanova MDonchev D (eds) Geography of Bulgaria BAN Sofiap 269minus283 (in Bulgarian)
Vladovic J Merganic J Maacuteliš F Križovaacute E and others (2014)The response of forest vegetation diversity to changes inedaphic-climatic conditions in Slovakia Technickaacute uni-verzita vo Zvolene Zvolene (in Slovak)
Vrška T Adam D Hort L Kolaacuter T Janiacutek F (2009) Europeanbeech (Fagus sylvatica L) and silver fir (Abies alba Mill)rotation in the Carpathians mdash a developmental cycle or alinear trend induced by man For Ecol Manag 258 347minus356
Wielgolaski FE (ed) (2005) Plant ecology herbivory andhuman impact in Nordic mountain birch forests Eco -logical Studies 180 Springer Verlag Berlin
Wielgolaski FE Hofgaard A Holtmeier FK (2017) Sensitivityto environmental change of the treeline ecotone and itsassociated biodiversity in European mountains Clim Res73151ndash166
Wieser G Tausz M (2007) Trees at their upper limit Treelifelimitation at the alpine timberline Springer Dordrecht
Wieser G Holtmeier FK Smith WK (2014) Treelines in achanging global environment In Tausz M Grulke N(eds) Trees in a changing environment Plant Ecophys 9Springer Dordrecht p 221minus263
Zhiyanski M Kolev K Sokolovska M Hursthouse A (2008)Tree species effect on soils in Central Stara PlaninaMountains Nauka Gorata 4 65minus82
Zhiyanski M Gikov A Nedkov S Dimitrov P Naydenova L(2016) Mapping carbon storage using land coverlanduse data in the area of Beklemeto Central Balkan In Koulov B Zhelezov G (eds) Sustainable mountain re -gions challenges and perspectives in southeasternEurope Springer Basel p 53minus65
Zhu K Woodall CW Clark JS (2012) Failure to migrate lackof tree range expansion in response to climate changeGlob Change Biol 18 1042minus1052
150
Editorial responsibility Nils Chr Stenseth Oslo Norway
Submitted April 8 2016 Accepted March 8 2017Proofs received from author(s) July 7 2017
Cudliacuten et al Drivers of treeline shift 139
Human Annual Limits Referencesinfluence temperature (T ) to climate
in VZs in past and precipitation (P) change and recently differences induced species
between 1961minus1990 shiftand 2021minus2050
mean of CORDEX models
Since 11th C HIRHAM5 model Inadequacy of Batllori et al (2009) clear-cutting and grazing period ΔT2021minus2050 substrata (rocky) Alados et al (2014)
moderate on irregular = 15minus2degC ΔT2051minus2080 climatic constraints Gartzia et a (2014) slopes and ridges since 1930 = 4degC ΔP2021minus2050 (including growing- Camarero et al (2015a)significant release of influence = +5 ΔP2051minus2080 season extreme events
but recovery mainly at = +30 compared to such as late-spring moderate elevations 1960minus1990 freezing summer
drought etc)
Since 1000 BC HadCM3_A2 period Low density of Fs van Gils et al (2008) cutting burning and 2020minus2080 in T dispersal distances Palombo et al (2013 2014)
grazing mng since the mid-20th ΔTmin_Jan2050 = 17degC of Fs seeds upslope C grazing intensity decreased ΔTmin_Jan2080 = 31degC poor soils excessive
now occasionally managed ΔP2050 = minus7 exposure to solar radiation and regeneration of Pm ΔP2080 = minus23 in open areas
Since 14th C clear-cutting Mean of 4 GCMs Differences in soil types Em (1986) and grazing to enlarge (CSIROMk2 HadCM3 between subalpine beech Strid et al (2003)
alpine pastures and to produce ECHAM4OPYC3 and mixed forest (poor Bergant (2006) charcoal since 1970 NCAR minus PCM) ΔT2050 = calcareous soils or rankers on Amidžic et al (2012)
abandonment of traditional 26degC ΔT2100 = 53degC silicate bedrock) cliffs agricultural practices ΔP2050 = minus3 ΔP2100 = minus8 and rocks cattle grazing
Since 680 AD intensive HADCM3_A2 Rocky sites extreme Velchev (1997) fires deforestation and grazing CGGM2 hydrothermal conditions Grunewald amp
since 1962 traditional mng ΔT2050 = 17degC competition of shrubs and tree Scheithauer (2011) has changed last decades tourism ΔT2080 = 28degC seedlings locally intense Raev et al (2011)
impact has increased ΔP2080 = minus15 pasturing
Human impact (mostly burning) HADCM3_A2 Low density of Fs Raev et al (2011) started in 17th C excessive CGCM2 ΔT2050 = 17degC in T dispersal distances Gikov et al (2016) and improper forest mng ΔT2080 = 28degC of Fs seeds upslope pasture Zhiyanski et al
release of influence since 1980 ΔP2050 = minus15 intensity new (2008 2016)ΔP2080 = minus15 expansion of juniper stands
15minus18th C slash burn DMI-HIRHAM5_A1B Impact of wild large Klopltilt et al grazing mng 18minus19th C period 2001minus2100 ΔT2050 ungulates on tree seedlings (2010 2015)
uncontrolled cutting 20th C = 12degC ΔT2100 = 25degC (especially Aa) shallow Mina et al (2017)irregular shelterwood systems ΔP2050 = +15 mm nutrient-poor soils
and over-exploitation mng ΔP2100 = minus18 mm (ie ranker and rendzina)
Since 12th C slash burn grazing mng ENSEMBLES models_ Frequent disturbances Tattoni et al (2010) 13minus19th C intensive forest A1B period 1980minus2100 by snow avalanches Kulakowski et al (2011)
mng at the end of the 19th C ΔT2050 = 12minus18degC and snow gliding ongoing Barbeito et al (2012) decrease in grazing ΔT2100 = 27minus41degC cattle grazing game grazing Bebi et al (2012)
ΔP2050 = 0 mm in the coldest years Dawes et al (2015)ΔP2100 = minus5 mm
Continued on next page
Clim Res 73 135ndash150 2017
mate the geographical position of these mountainunits 1 value was used for the geographical latitudeand longitude of the locus of the studied mountainunits The analysed parameters of these 15 mountainunits are shown in Table 2
In addition we calculated the following climatecharacteristics for all mountain units annual meanair temperature annual sum of precipitation thegrowing season length and the date of the onset ofthe growing season These characteristics were thencompared for 2 distinct periods 1961minus1990 and1991minus 2015 (Table 3) The climate characteristicswere derived from the E-OBS gridded dataset of sta-tion observations version 131 (Haylock et al 2008)We used the E-OBS version on the regular longi-tudeminuslatitude grid with a horizontal resolution of 025degrees The climate data from all grid points withinthe mountain units or near their geographical bor-ders were considered
Data processing for 15 mountain units was made onthe basis of data obtained from the literature and
personal studies by the co-authors these are sum ma -rized in Tables 1 amp 2 and in Tables S1 amp S2 in theSupplement at www int-res com articles suppl c073p135 _ supp pdf In addition a semi-quantitative valu-ation of abiotic biotic and anthropogenic factors lim-iting an upward treeline shift was performed (Fig 2)To answer the question how natural and anthro-pogenic factors have influenced treeline ecotonecharacteristics with a focus on treeline shift redun-dancy analysis (RDA) was used to describe and testthe effect of the explanatory variables (geographicalposition size of the whole mountain massif start ofhuman influence and start of the decrease in humaninfluence) on treeline ecotone characteristics (tim-berline elevation width of the treeline ecotone treespecies forming the treeline and treeline shift peryear) and identify groups of mountain units with sim-ilar variability of the dependent data (ter Braak ampŠmilauer 2012) The whole data set for this analysis isshown in Table S2 We tested both their simple ef -fects which show how much variation every ex plan -
140
Country Coordi- Elevation Elevation Range of Annual Area VZs near Tr part of nates of range range of Tr tree species altitudinal of study (m asl)mountains central site of T (m asl) limit shift (km2)
(m asl) (m asl) of T or Tr (m yrminus1)
Slovakia 48deg55rsquoN 1400 1530 Pa 1730 ~03 (Tr ~400 Spruceminusfirminusbeech Dumbier Tatra 19deg31rsquoE (1350minus1450) (1400minus1630) 1950minuspresent) (1200minus1350) spruce (1300minus(part of Low 1550) Pm with individual Tatra Mts) Pa trees (1400minus1700)
Czech Republic 50deg42rsquoN 1250 1320 Pa 1500 043 (Tr 550 Beechminusspruce (700minus1050) Giant Mts 15deg38rsquoE (1130minus1460) (1250minus1460) 1936minus2005) Pa dominates with increasing
altitude spruce (1000minus1400) Pm with Pa trees (1400minus1560)
Norway 60deg10rsquoN 850 600minus1050 1500 08 (Tr 40 600 Birch forest dominance Southern 7deg40rsquoE 62deg20rsquoN 1915minus2007) at high altitudes but with Scandes 10deg05rsquoE some scattered Ps or Pa
mixed birchminusconifer below 800
Norway 68deg20rsquoN West 650 West 750 West 1000 06 (Tr 35 000 Birch forest dominance Sweden 18deg20rsquoE East 100 East 290 East 0 1958minus2008) at all altitudes but with some Finland 69deg50rsquoN scattered pines or pine stands Northern 27deg00rsquoE on sandy soilsScandes
Table 1 (continued)
Cudliacuten et al Drivers of treeline shift
atory variable can explain separately without usingthe other variables and their conditional ef fectswhich depend on the variables already selected inthe model The statistical significance of the expla -natory variables was tested using a Monte Carlo permutation test and only predictors with p le 005were included in the subsequent RDA (Šmilauer ampLepš 2014)
To find the drivers of biodiversity loss in forestsmeadows and animal communities (dependent vari-ables) the explanatory variables (geographical posi-tion timberline elevation tree species forming thetreeline start of human influence start of the de -crease in human influence and temperature increasebetween the 2 periods 1961minus1990 and 2021minus2050)were tested again by RDA in Canoco 5 as mentionedabove (Figs 3 amp 4) The whole data set for this analy-sis mdash including additional information about tree-lines in the mountain areas of interest with a focuson treeline shift (its rate drivers limits and problemswith its assessment) mdash is presented in Tables S1 amp S2
Selected univariate graphs were constructed to visu-alize the data of relationships between the explana-tory and dependent variables of all analyses whichwere not seen distinctly from the results of the multi-variate statistics or linear regression (Fig 5)
3 RESULTS
31 Effect of environmental variables on treeline ecotone characteristics with a focus
on treeline shift
The variation in the dependent variables (timber-line elevation width of treeline ecotone tree speciesforming the treeline and treeline shift per year) wassignificantly affected only by latitude and size of thewhole mountain massif (Fig 3) The adjusted ex -plained variability by all explanatory variables was342 (F = 46 p = 001) The RDA diagram showsthat in the mountains located more to the north the
141
Human Annual Limits Referencesinfluence temperature (T ) to climate
in VZs in past and precipitation (P) change and recently differences induced species
between 1961minus1990 shiftand 2021minus2050
mean of CORDEX models
From 14th C to 1922 Average of 10 RCM_B1 More rocky and nutrient- Koumlrner (2003) mining forest change ΔT2050 = 18degC poor soils late frosts Fridley et al (2011)
to spruce monocultures ΔT2100 = 37degC steep slopes with avalanches Hlaacutesny et al (2011 2016)from 13th C to 1978 pastures ΔP2050 = +24 mm absence of mature trees
in upper parts of mountain ΔP2100 = minus35 mm
9minus11th C forest clear cutting ALADIN_A1B period 1961minus Different soil types Treml and Banaš (2000) grazing in dwarf pine VZ 2100 ΔT2050 = 13degC for Fs and Pa (cambisol Cudliacuten et al (2013)
since 18th C forest change ΔT2100 = 34degC versus podzol) only TSL of Pa Treml amp Chuman (2015) to spruce monocultures ΔP2050 = +25 mm could elevate on ranker Treml amp Migo (2015)
since 19th C artificially planted ΔP2100 = minus14 mm soils in Pm VZ
Grazed and browsed RegClimMPIHadley Sheep and reindeer Dalen amp Hofgaard (2005) by reindeer and livestock ΔT2050 = 12degC grazing and episodic Wielgolaski (2005)
(sheep and cattle) ΔT2100 = 22degC insect outbreaks (Epirrita) Hofgaard et al (2009) over thousands of years ΔP2050 = +23 mm Kullman Oumlberg (2009)
ΔP2100 = minus12 mm
Grazed and browsed RegClimMPIHadley Continued grazing regime Dalen amp Hofgaard (2005) by semi-domestic reindeer ΔT2050 = 16degC and frequent episodic Wielgolaski (2005)
for hundreds of years ΔT2100 = 29degC insect outbreaks (Epirrita) Toslashmmervik et al (2005) increased grazing from ΔP2050 = +18 mm Hofgaard et al (2009)
year 1960 ΔP2100 = minus10 mm Aune et al (2011) Mathisen et al (2014)
Clim Res 73 135ndash150 2017
treeline ecotone is narrower treeline shiftis smaller the timberline occurs at lowerelevations and the treeline is formed moreby broadleaved species Higher values oftreeline ecotone width and treeline shiftwere found in the mountains with a greatersize of the whole mountain massif
When every explanatory variable wastested separately values of the start of thedecrease of human influence in the moun-tains also had a significant effect on thevariation of the tree line ecotone data(explained variation = 206 pseudo-F =34 p = 0024) Some selected relationshipsof the explanatory and dependent vari-ables even those not showing significanteffects on the variation in the tree line eco-tone parameters are depicted in Fig 5
32 Effect of recent changes in climateand land use on the biodiversity
Biodiversity loss in forests meadows andanimal communities analysed by RDAwas explained by geo graphical positiontreeline species composition and tempera-ture increase between the 2 periods1961minus1990 and 2021minus2050 (Fig 4) Theadjusted ex plained variation by all ex pla -natory variables was 632 (F = 58 p =0001) The RDA results indicated that bio-diversity loss in forest communities in crea -sed with increasing latitude and longitudeas well as in the case where broad leavedspecies formed the treeline In contrast thehighest biodiversity loss in meadows wasfound in the mountains positioned more tothe south and with higher temperatureincrease The biodiversity loss in animalcommunities was negatively correlatedwith longitude and temperature increase
4 DISCUSSION
The comparison of several mountainareas situated across Europe shows a varia-tion in understanding of what constitutesthe treeline across countries The primaryissue is the different approach to the defini-tion of forest when applying a tree heightthreshold This threshold decreases from
142
Mou
nta
in u
nit
Nor
th
Eas
t M
oun
-T
imb
er-
Wid
th
Tre
e T
emp
erat
ure
T
reel
ine
Sta
rt
Sta
rt
lati
tud
elo
ng
itu
de
tain
lin
eof
spec
ies
incr
ease
sh
ift
of H
I of
HI
size
(m a
sl
)tr
eeli
ne
form
ing
b
etw
een
(m
yrminus
1 )(y
r)d
ecre
ase
(km
2 )ec
oton
eth
e 19
61minus
1990
(y
r)(m
)tr
eeli
ne
and
199
1minus20
15 (
degC)
Cen
tral
Pyr
enee
s (C
P)
42deg2
1rsquoN
0deg30
rsquoW55
000
2250
300
C0
81
92 (
1956
minus20
06)
1000
BC
1930
Eas
tern
Pyr
enee
s (E
P)
42deg2
8rsquoN
1deg30
rsquoE55
000
2400
300
C0
80
71 (
1956
minus20
06)
1000
1930
Ap
enn
ines
(A
P)
42deg0
5rsquoN
14deg0
5rsquoE
5000
018
0030
0B
14
1 (1
954minus
2007
)10
00 B
C19
50S
har
a M
ts (
SH
)41
deg47rsquo
N20
deg33rsquo
E16
0018
5035
0C
B0
71
(193
4minus20
10)
1300
1970
Pir
in (
PI)
41deg4
2rsquoN
23deg3
1rsquoE
2585
1900
400
C0
5n
o d
ata
680
1962
Cen
tral
Bal
kan
Mts
(C
B)
42deg4
7primeN
24deg3
6primeE
1160
016
0025
0C
B0
7n
o d
ata
1700
1980
Nor
ther
n D
inar
ic M
ts (
DI)
45deg3
6rsquoN
14deg2
8rsquoE
1000
1540
160
B1
3n
o d
ata
1400
1900
Cen
tral
Alp
s (C
A)
46deg4
6rsquoN
9deg52
rsquoE18
000
021
0055
0C
07
13
(195
4minus20
15)
1300
1880
Eas
tern
Alp
s (E
A)
46deg1
7rsquoN
11deg4
5rsquoE
180
000
1800
400
C0
91
3 (1
954minus
2015
)11
5019
50L
ow T
atra
Mts
(LT
)48
deg55rsquo
N19
deg31rsquo
E12
4014
0033
0C
09
057
(19
75minus
2016
)14
0019
22G
ian
t M
ts (
GM
)50
deg42rsquo
N15
deg38rsquo
E63
012
5025
0C
07
043
(19
36minus
2005
)10
0019
63H
ard
ang
ervi
dd
a (H
G)
60deg0
3rsquoN
7deg47
rsquoE90
0090
015
0C
B0
50
8 (1
915minus
2007
)80
019
70D
ovre
(D
O)
62deg2
1rsquoN
8deg52
rsquoE60
0010
0040
0B
07
08
(191
5minus20
07)
1000
no
dec
reas
eN
orth
ern
Sw
edis
h L
apla
nd
(N
S)
68deg1
2rsquoN
18deg3
3rsquoE
4000
700
300
B1
10
6 (1
958minus
2008
)10
00n
o d
ecre
ase
Inn
er F
inn
mar
kn
orth
ern
mos
t 69
deg10rsquo
N25
deg12rsquo
E60
000
290
90B
10
06
(195
8minus20
08)
1600
no
dec
reas
eF
inn
ish
Lap
lan
d (
FL
)
Tab
le 2
Fac
tors
infl
uen
cin
g t
reel
ine
ecot
one
char
acte
rist
ics
in s
elec
ted
Eu
rop
ean
mou
nta
ins
Th
e m
oun
tain
siz
e is
th
e si
ze o
f th
e w
hol
e m
oun
tain
mas
sive
fu
nct
ion
ing
as a
mez
osca
le c
lim
ate
un
it n
ot o
nly
the
stu
die
d m
oun
tain
are
a W
idth
of t
he
tree
lin
e ec
oton
e w
as d
eter
min
ed b
y su
btr
acti
ng
the
valu
e of
the
mea
n ti
mb
erli
ne
elev
atio
n
from
th
e va
lue
of t
he
up
per
tre
e sp
ecie
s li
mit
C c
onif
ers
B b
road
leav
ed t
rees
HI
hu
man
infl
uen
ce
Cudliacuten et al Drivers of treeline shift
5 m (Jeniacutek amp Lokvenc 1962) to 3 m(Koumlrner 2012) in Northern andCentral Europe to only 2 m (Holt-meier 2009) in Central and South-ern Europe where even shrub -by stands (eg Pinus mugo) areconsi dered as a forest in somecountries eg in Spain Italy andMa ce donia (Batllori et al 2009)An other problem is a different de -signation of forest stands belowthe treeline lsquoupper montane fo -restsrsquo in Central Europe and lsquosub-alpine forestsrsquo in Southern Europe(Ellenberg 1988) Further differen -ces are related to the tradition ofdifferent branches of science (ega more traditional lsquogeobotanicalrsquoapproach in Central Europe ver-sus a more experimental approachin Western Europe) especially inthe rate of applying new pro -gressive methods (eg climaticmodelling or molecular biologicalmethods)
Our comparison of selectedregions (Tables 1 amp 2 Figs 1 amp 2)showed that southern mountainscompared to those located morecentrally and in the north had (1) alonger and more profound exploi -tation by humans in the past (2)greater differences in climatic para -meters (temperature length ofgrowing period) between the 2periods 1961minus1990 and 1991minus2015(Table 3) and (3) more dramaticclimate change scenarios espe-cially concerning temperature in -creases Longitude also had someinfluence on the start and intensityof human influence and on tree-line elevation and rate of treelineshift (Table 2 Fig 5) The occur-rence of broadleaf tree species inthe treeline ecotone in the north-ern (and to some extent the south-ern) countries distinguishes thesefrom the Central European coun-tries where conifers prevail It isinteresting that grazing an im -portant factor shaping treelineecosystems (Dirnboumlck et al 2003
143
Length Beginning Mean Mean of growing of growing annual annual
period season temperature precipitation (d)a (d)b (degC) ()
Central Pyrenees 135 minus89 08 minus107Eastern Pyrenees 117 minus91 08 minus92Apennines 284 minus232 14 57Shara Mts 16 01 07 minus100Pirin 26 minus14 05 27Central Balkan Mts 169 minus136 07 76Northern Dinaric Mts 160 minus95 13 minus56Central Alps 147 minus86 07 62Eastern Alps 125 minus96 09 minus33Low Tatra Mts 24 minus34 09 minus91Giant Mts minus10 13 07 55Hardangervidda Blefjell 63 34 05 56Dovre 138 32 07 64Northern Swedish Lapland 81 05 11 minus27Inner Finnmarknorthern- 17 08 10 77most Finnish Lapland
aPositive numbers indicate a prolongation bNegative numbers indicate an earlier beginning
Table 3 Differences of selected climatic parameters between the 2 study periods (1961minus1990 and 1991minus2015) in selected European mountains
Tein Drou Gefa Nadi Laus Abli Bili Huli
0 1 2 3
Central Pyrenees (CP)
Eastern Pyrenees (EP)
Apennines (AP)
Shara Mts (SH)
Pirin (PI)
Central Balkan Mts (CB)
Northern Dinaric Mts (DI)
Central Alps (CA)
Eastern Alps (EA)
Low Tatra Mts (LT)
Giant Mts (GM)
Hardangervidda (HG)
Dovre (DO)
Northern Swedish Lapland (NS)
Inner FinnmarknorthernmostFinnish Lapland (FL)
Fig 2 Rates of treeline drivers (temperature increase [Tein] land use change[Laus]) and treeline shift limits (drought [Drou] geomorphological factors [Gefa]natural disturbances [Nadi] other abiotic biotic and human factors) in selectedEuropean mountains Abli Bili and Huli abiotic biotic and human treeline shiftlimits respectively) Rate of treeline driver influence mdash 0 no influence 1 weak
influence 2 middle influence 3 strong influence
Clim Res 73 135ndash150 2017
Gehrig-Fasel et al 2007) limits treeline shift in halfof the countries of interest regardless of latitude lon-gitude or recent political developments (Shara PirinCentral Balkans Alps Scandes Fig 2)
Despite the stated differences between studieslooking at climate change impacts on the treelineit is clear that current and future changes in temp -erature will seriously affect treeline ecotones in allmountain ranges of Europe Winter is a period whentreeline stands and individual trees have to survivesevere life-limiting conditions (Wieser amp Tausz2007) Therefore winter warming might increase thechance of young trees surviving and passing the mostcritical period from seedling to sapling and further tothe mature stage (Koumlrner 2003) There is alreadymuch evidence worldwide that winter warming isone of the significant drivers of treeline advance(Harsch et al 2009) Although it is expected that treegrowth at the upper distributional margins (and inthe northern European countries) will increase due toongoing climate change (Pentildeuelas amp Boada 2003Chen et al 2011 Hlaacutesny et al 2011 Lindner et al2014) extremes in temperature (eg black froststemperature reversals in the spring) or winter precip-itation (eg lack of snow drought) might also limit
this process in the future in some regions and localsituations (Holtmeier amp Broll 2005 Hagedorn et al2014) An earlier start to the growing period (espe-cially in the Apennines Central Balkan and SpanishPyrenees see Table 2) can play a negative role in theresistance of trees and seedlings to early springfrosts On the other hand at lower distributional mar-gins (and in Southern European countries) it isexpected that tree growth will decrease or forestswill experience some level of dieback due to drought(Breshears et al 2005 Piovesan et al 2008 Allen etal 2010 Huber et al 2013) A serious implication isthe positive effect that warming can have on root rotfungi and populations of bark beetles resulting inlarge-scale disturbances to temperate and also high-altitude forests (Jankovskyacute et al 2004 Elkin et al2013 Millar amp Stephenson 2015)
Although recent upward advances of treeline eco-tones are widespread in mountain regions (Harsch et
144
Fig 3 Redundancy analysis diagram with variation in tim-berline elevation width of treeline ecotone the tree speciesforming the treeline and treeline shift used as dependentvariables explained by explanatory variables (latitudemountain size) Mountain units (blue) are projected as thecentres of abbreviations (see Table 2) Explanatory variablesaccount for 436 of the total variation in the dependentdata The first canonical axis explained 466 of variationthe second axis explained 30 of variation Dependentvariables (black) are Ecot altitudinal width of treeline eco-tone TLShift treeline shift TrspC (TrspB) treeline formedby conifers (broadleaved trees) Timb timberline elevationExplanatory variables (red) are N north latitude Mtsize
size of the mountain massif
Fig 4 Redundancy analysis diagram with variation in bio-diversity loss in forests meadows and animal communitiesused as dependent variables explained by explanatory vari-ables (geographical position tree species forming the tree-line and temperature increase between the 2 study periods[1961minus1990 and 2021minus2050]) Mountain units (green) areprojected as centres of abbreviations (see Table 2) Explana-tory variables account for 763 of total variation in the de-pendent data The first canonical axis explained 484 ofvariation the second axis explained 183 of variation De-pendent variables (black) are BdfoBdmeBdan biodiversityloss in forestsmeadowsanimal communities Explanatoryvariables (red) are N north latitude E east longitude TrspC(TrspB) treeline formed by conifers (broadleaved trees)Temp temperature increase between the 2 study periods
Cudliacuten et al Drivers of treeline shift
al 2009) only a few published studies have includedquantitative data about treeline shifts (Devi et al2008 Kullman amp Oumlberg 2009 Diaz-Varela et al 2010Van Bogaert et al 2011) Additionally our know -ledge of the spatial patterns in treeline ecotone shiftsat the landscape scale is still surprisingly poor exceptfor some studies from subarctic areas the UralMountains and the Alps (Lloyd et al 2002 Diaz-Varela et al 2010 Hagedorn et al 2014) In the 15studied mountain units the values of treeline shiftranged from 043 to 19 m yrminus1 showing a rather dis-tinct spatial pattern of the dynamics within Europeanmountains The observed treeline shift had a signifi-cant positive relationship only with northern latitudeand a weak positive relationship with mountain-range size (ie the size of the whole mountain massifFig 3) Nevertheless this illustrates the importanceof the mass elevation effect (Koumlrner 2012) and par-tially ex plains why treelines could be at different alti-tudes at the same latitude Small negative regres-sions with east longitude and timberline elevationare shown in Fig 5 There was no apparent depend-ence of the treeline shift rate on climatic parameterchanges between the periods 1961minus1990 and1991minus2015
A whole forest vegetation zone shift is a complexprocess as not only the life strategies of an individualtree need to be considered but plant animal andmicroorganism species and their interrelationshipsas well as the relationships to their specific microhab-
itats (including soil conditions) are also involved(Urban et al 2012) Across all regions we identifiedseveral important obstacles to treeline shifts oftenrelated to site properties such as significant rocki-ness having shallow or low-nutrient soils extremerelief causing disturbances by avalanches snow glid-ing or wind damage (data from the Pyrenees Apen-nines Shara Northern Dinaric Low Tatra and GiantMountains see Fig 2) Soil heterogeneity certainlyhas an important role in plant responses to climatechange and could also maintain the resilience of thecommunity assemblages (Fridley et al 2011) An -other group of obstacles includes extreme climaticparameters especially in winter (reported eg fromthe Pyrenees Apennines Pirin Central BalkanNorthern Dinaric and Giant Mountains Fig 2) Theircombination can result in edaphic andor climaticunsuitability of habitats where species could poten-tially migrate Climatically and edaphically suitablesites will likely decline over the next century partic-ularly in mountain landscapes (Bell et al 2014)There fore trees in the treeline ecotone will colonizepreviously forested habitats The colonization suc-cess of individual forest communities is affected bydifferences in species dispersal and recruitmentbehaviour (Dullin ger et al 2004 Jonaacutešovaacute et al2010) For these reasons colonization of new foresthabitats by the next tree generation may not be suc-cessful and may result in loss of species diversity(Honnay et al 2002 Ibaacutentildeez et al 2009 Dobrowski et
145
Fig 5 Univariate graphs of the relationships of dependent variables viz timberline (Timb) and treeline shift per year (Shift) and their explanatory variables Nola north latitude Ealo east longitude DD decimal degrees
Clim Res 73 135ndash150 2017
al 2015) as reported from Macedonia Slovenia theCzech Republic Slovakia and Norway (Fig 4)
Another limiting factor is the existence of strongadaptation mechanisms of the dominant tree specieswhich allow them to survive in their current distribu-tion areas Species migration is likely to be slow dueto the limited quantity of climatically and edaphicallysuitable sites and the slow velocity of seed dispersaland tree regeneration and will be hampered evenmore by fragmentation of high mountain landscapescaused by human activities including brush invasionin abandoned meadows (Gartzia et al 2014) A verti-cal shift in a vegetation zone involves not only singlespecies of plants animals and microorganisms butalso their interrelationships and links to soil condi-tions The speed at which climatic conditions changewill be different from changes in the soil conditionsdue to the slowness of soil-forming processes Soiltypes and conditions (cambisol versus podzol) arecrucial obstacles for the shift from a beech forest zoneinto the spruce forest zone in the Czech Republic(Vacek amp Mat jka 2010) According to other sourcesthe dominant tree species can influence soil for -mation processes (especially humus forms) Underfavourable climatic and orographic conditions beechis able to change its soil conditions over decades tocenturies (J Macku unpubl data)
Significant changes in biodiversity must be expec -ted in all of the mountain areas of interest We foundthat biodiversity declined particularly in the southernregions of Europe where the timberline is situated athigher elevations and human impact is mostlylonger-lasting (Fig 4) Similarly Pauli et al (2012)reported an increase in species richness of mountaingrasslands across Europe except for Mediterraneanregions and assigned this different response of thesouthern regions to decreased water availability Onthe other hand ecosystems which exhibited bio -diversity increases were not only enriched by mig -rant species but the assemblages underwent thermo -philization ie cold-adapted species declined andmore warm-adapted species spread (Gottfried et al2012) However not all species are able to track cli-mate changes It is expected that around 40 ofhabitats will become climatically unsuitable for manymountain species particularly endemic ones in -creasing their extinction probability during this cen-tury (Dullinger et al 2012) The abandonment of tra-ditional land use forms is another source of this losswhich may be a more important driver than tempera-ture increase (Fig 4) Serious biodiversity losses inmountain meadows were reported from Spain ItalyMacedonia Bulgaria Slovenia and Slovakia There-
fore controlled grazing must occur in order to main-tain alpine grasslands (Dirnboumlck et al 2003) Unfor-tunately grazing is still active only in smaller parts ofEuropean mountains (eg in the Central Alps and theCentral Balkan Mountains but pasturing can alsonegatively affect vegetation diversity) and some-times does not serve to maintain alpine grasslandThe same is true for grassland management whichcan help to maintain mountain species and decreasehabitat loss A serious biodiversity loss in mountainmeadows was reported from the Pyrenees the Apen-nines and the Shara Pirin Central Balkan NorthernDinaric and Low Tatra Mountains (Fig 4)
In forests species responses to climate change maybe equivocal some recent studies indicated contra-dictory shifts in species distributions (eg Lenoir etal 2008 Zhu et al 2012 Rabasa et al 2013) This dis-parity is widely discussed and assigned to the greatvariety of non-climatic factors or even tree ontogeny(Grytnes et al 2014 Lenoir amp Svenning 2015 Maacutelišet al 2016) Disturbances leading to tree mortalitymay also play an important role (Cudliacuten et al 2013)changes in tree canopy cover modify light availabil-ity and microclimate and can induce the thermo -philization of forest vegetation (De Frenne et al2015 Stevens et al 2015) The loss of this microcli-mate buffering effect of forests may induce a biotichomogenization of forests (Savage amp Vellend 2015)or the creation of novel non-analogical communitiessuch as oakminuspine forests (Urban et al 2012) whichmay be a new threat to forest biodiversity
The observed changes in treeline forest ecosystemsare often related to changes in land-use intensity(Theurillat amp Guisan 2001 Alados et al 2014) Ac -cording to climate change predictions and the recentand future exploitation intensity of European moun-tains trees and forest communities will shift upwarddue to land use change or climate change or bothReduced land-use intensity certainly will interactwith climate change by facilitating or inhibiting spe-cies occurrence and will accelerate forest expansionabove the present treeline particularly to previouslyforested habitats (Theurillat amp Guisan 2001) Thesimultaneous action of both main treeline shift driv-ers viz temperature increase and decrease in landuse intensity was recorded from all of our studiedmountain areas The extensive differences in timber-line and treeline elevations in almost all studiedmountains (Table 1) indicate the anthropo-zoogenictreeline type (according to Ellenberg 1998) In mostmountains (eg in the Apennines Shara MountainsCentral Balkans and Alps) land use is the prevailingfactor influencing vegetation drift (Fig 5) Previous
146
Cudliacuten et al Drivers of treeline shift
research showed that the upward shift of the treelinein the Swiss Alps was predominantly attributable toland abandonment and only in some situations to cli-mate change (Gehrig-Fasel et al 2007) In the Apen-nines in the last few decades tree establishment hasbeen mainly controlled by land use while treegrowth has been controlled by climate pointing to aminor role played by climate in shaping the currenttreeline (Palombo et al 2013)
The impact of climate change and the connectedland use change on biodiversity and ecosystem serv-ices provision in several European countries is sum-marized by Wielgolaski et al (2017) and Kyriazopou-los et al (2017) (both this Special) and Sarkki et al(2016) One of the possible adaptive management op-tions in response to climate change assisted migra-tion as human-assisted movement of species hasbeen frequently debated in the last few years (Ste-Marie et al 2011) It is possible to apply it as a type ofassisted colonization ie intentional movement andrelease of an organism outside its indigenous range toavoid extinction of populations of the focal species(eg Macedonian pine species) or as an ecologicalre placement ie the intentional movement and re-lease of an organism outside its indigenous range toperform a specific ecological function (eg planting ofPinus mugo in the Alps IUCNSSC 2013)
5 CONCLUSIONS
The analysis of 11 mountain areas across Europeshowed that with increasing latitude the treelineand altitudinal width of the treeline ecotone signifi-cantly decreases as do the significance of climaticand soil parameters as barriers against tree speciesshift Although temperature is the overwhelmingcontrolling factor of tree growth and establishment intemperate and boreal treeline ecotones late-sea-sonal drought might also play a driving role in Medi-terranean treeline ecotones Longitude was lessinfluential mostly affecting climate and land usechange effects on increased biodiversity loss as wellas the size of the area that forms one mesoclimateunit affecting altitudinal ecotone width The biggestpart of the commonly observed remaining variabilityin mountain vegetation near the treeline in Europeseems to be caused by geomorphological geologicalpedological and microclimatic variability in combina-tion with different land use history and the presentsocio-economic relations The observed differencesin climatic parameters between the mountain areasof interest in comparison with the reference period
1960minus1990 (09degC) have not explained the relativelyhigh differences in the rate of treeline shift per year(149 m) The predicted variability in temperatureincrease due to global warming (12minus26degC in 2050and 22minus53degC in 2100) could lead to a much biggerdifferentiation in treeline ecotone biodiversity andecosystem processes between southern and northernEuropean mountains in the future Therefore thesedifferences must be taken into account by scientistsand EU policy makers when formulating efficientadaptive forest management strategies for treelineecosystems at the European level (eg assistedmigration of adapted genotypes)
Acknowledgements This paper is based firstly upon workfrom the COST Action ES 1203 SENSFOR supported byCOST (European Cooperation in Science and Technologywwwcosteu) This international work was further sup-ported by projects granted by the Ministry of EducationYouth and Sports of the Czech Republic grant NPU ILO1415 and LD 14039 by the agency APVV SR under pro-jects APVV-14-0086 and APVV-15-0270 We acknowledgethe E-OBS dataset from the EU-FP6 project ENSEMBLES(httpensembles-eumetofficecom) and the data providersin the ECAampD project (wwwecadeu)
LITERATURE CITED
Alados CL Errea P Gartzia M Saiz H Escoacutes J (2014) Posi-tive and negative feedbacks and free-scale pattern distri-bution in rural-population dynamics PLOS ONE 9 e114561
Allen CD Macalady AK Chenchouni H Bachelet D andothers (2010) A global overview of drought and heat-induced tree mortality reveals emerging climate changerisks for forests For Ecol Manag 259 660minus684
Amidžic L Bartula M Jaric S (2012) Syntaxonomy overviewof vegetation of Šar Planina in the Balkan PeninsulaVegetos 25 348minus360
Aune S Hofgaard A Soumlderstroumlm L (2011) Contrasting cli-mate and land use driven tree encroachment pattern ofsub-arctic tundra in Northern Norway and Kola Penin-sula Can J For Res 41 437minus449
Barbeito I Dawes M Rixen C Senn J Bebi P (2012) Factorsdriving mortality and growth at treeline a 30-year exper-iment of 92000 conifers Ecology 93 389minus401
Batllori E Camarero JJ Ninot JM Gutieacuterrez E (2009)Seedling recruitment survival and facilitation in alpinePinus uncinata treeline ecotones Implications andpotential responses to climate warming Glob Ecol Bio-geogr 18 460minus472
Bebi P Teich M Hagedorn F Zurbriggen N Brunner SHGreacutet-Regamey A (2012) Veraumlnderung von Wald undWaldleistungen in der Landschaft Davos im Zuge desKlimawandels Schweiz Z Forstwes 163 493minus501
Bell DM Bradford JB Lauenroth WK (2014) Mountain land-scapes offer few opportunities for high elevation treespecies migration Glob Change Biol 20 1441minus1451
Bergant K (2006) Climate change scenarios for MacedoniaUniversity of Nova Gorica Center for Atmospheric
147
Clim Res 73 135ndash150 2017
Research Nova GoricaBodin J Badeau V Bruno E Cluzeau C Moisselin JM
Walther GR Dupouey JL (2013) Shifts of forest speciesalong an elevational gradient in Southeast France cli-mate change or stand maturation J Veg Sci 24 269minus283
Boncina A (2011) History current status and future pros -pects of uneven-aged forest management in the Dinaricregion an overview Forestry 84 467minus478
Breshears DD Cobb NS Rich PM Price KP and others(2005) Regional vegetation die-off in response to global-change-type drought Proc Natl Acad Sci USA 102 15144minus15148
Camarero JJ Garciacutea-Ruiz JM Sanguumlesa-Barreda G GalvaacutenJD and others (2015a) Recent and intense dynamics in aformerly static Pyrenean treeline Arct Antarct Alp Res47 773minus783
Camarero JJ Gazol A Galvaacuten JD Sanguumlesa-Barreda GGutieacuterrez E (2015b) Disparate effects of global-changedrivers on mountain conifer forests warming-inducedgrowth enhancement in young trees vs CO2 fertilizationin old trees from wet sites Glob Change Biol 21 738minus749
Chen IC Hill JK Ohlemuumlller R Roy DB Thomas CD (2011)Rapid range shifts of species associated with high levelsof climate warming Science 333 1024minus1026
Cudliacuten P Sejaacutek J Pokornyacute J Albrechtovaacute J Bastian OMarek M (2013) Forest ecosystem services under climatechange and air pollution In Matyssek R Clarke N Cud-liacuten P Mikkelsen TN Tuovinen JP Wiesner G Paoletti E(eds) Climate change air pollution and global chal-lenges understanding and perspectives from forestresearch Developments in Environmental Science Vol13 Elsevier Oxford p 521minus546
Dalen L Hofgaard A (2005) Differential regional treelinedynamics in the Scandes mountains Arct Antarct AlpRes 37 284minus296
Dawes MA Philipson CD Fonti P Bebi P Haumlttenschwiler SHagedorn F Rixen C (2015) Soil warming and CO2
enrichment induce biomass shifts in alpine tree line veg-etation 2014 Glob Change Biol 21 2005minus2021
De Frenne P Rodriacuteguez-Saacutenchez F De Schrijver A CoomesDA Hermy M Vangansbeke P Verheyen K (2015) Lightaccelerates plant responses to warming Nat Plants 1 15110
Devi N Hagedorn F Moiseev P Bugmann H ShiyatovSG Mazepa V Rigling A (2008) Expanding forests andchanging growth forms of Siberian larch at the treeline ofthe Polar Urals during the 20th century Glob ChangeBiol 14 1581minus1591
Diaz-Varela R Colombo R Meroni M Calvo-Iglesias MSBuffoni A Tagliaferri A (2010) Spatio-temporal analysisof alpine ecotones A spatial explicit model targeting altitudinal vegetation shifts Ecol Model 221 621minus633
Dirnboumlck T Dullinger S Grabherr G (2003) A regionalimpact assessment of climate and land-use change onalpine vegetation J Biogeogr 30 401minus417
Dobrowski SZ Swanson AK Abatzoglou JT Holden ZASafford HD Schwartz MK Gavin DG (2015) Forest struc-ture and species traits mediate projected recruitmentdeclines in western US tree species Glob Ecol Biogeogr24 917minus927
Dullinger S Dirnboumlck S Grabherr G (2004) Modelling cli-mate change-driven treeline shifts relative effects oftemperature increase dispersal and invisibility J Ecol92 241minus252
Dullinger S Gattringer A Thuiller W Moser D and others
(2012) Extinction debt of high-mountain plants undertwenty-first-century climate change Nat Clim Chang 2 619minus622
Elkin C Gutierrez AG Leuzinger S Manusch C TemperliC Rasche L Bugmann H (2013) A 2degC warmer world isnot safe for ecosystem services in the European AlpsGlob Change Biol 19 1827minus1840
Ellenberg H (1988) Vegetation ecology of Central Europe4th edn Cambridge University Press Edinburgh
Em H (1986) Spruce southern distribution range Spruce for-est on Shara Mountain in Macedonia Contrib Macedon-ian Acad Sci Arts Skopje 5 11minus28
Fajardo A Piper FI Pfund L Koumlrner CH Hoch G (2012)Variation of mobile carbon reserves in trees at the alpinetreeline ecotone is under environmental control NewPhytol 195 794minus802
Fridley JD Grime JP Askew AP Moser B Stevens CJ (2011)Soil heterogeneity buffers community response to cli-mate change in species-rich grassland Glob Change Biol17 2002minus2011
Gartzia M Alados CL Peacuterez-Cabello F (2014) Assessment ofthe effects of biophysical and anthropogenic factors onwoody plant encroachment in dense and sparse moun-tain grasslands based on remote sensing data Prog PhysGeogr 38 201minus217
Gehrig-Fasel J Guisan A Zimmermann NE (2007) Tree lineshifts in the Swiss Alps climate change or land abandon-ment J Veg Sci 18 571minus582
Gikov A Dimitrov P Zhiyanski M (2016) Land cover dynam-ics on the northern slope of the Troyan Passage during 30years period Probl Geogr 1-2 78minus92
Gonzalez P Neilson RP Lenihan JM Drapek RJ (2010)Global patterns in the vulnerability of ecosystems to veg-etation shifts due to climate change Glob Ecol Biogeogr19 755minus768
Gottfried M Pauli H Futschik A Akhalkatsi M and others(2012) Continent-wide response of mountain vegetationto climate change Nat Clim Chang 2 111minus115
Grace J Berninger F Nagy L (2002) Impacts of climatechange on the tree line Ann Bot 90 537minus544
Grunewald K Scheithauer J (2011) Landscape developmentand climate change in Southwest Bulgaria (Pirin Mts)Springer Heidelberg
Grytnes JA Kapfer J Jurasinski G Birks HH and others(2014) Identifying the driving factors behind observedelevational range shifts on European mountains GlobEcol Biogeogr 23 876minus884
Hagedorn F Shiyatov FG Mazepa VS Devi NM and others(2014) Treeline advances along the Urals mountainrange mdash driven by improved winter conditions GlobChange Biol 20 3530minus3543
Hansen AJ Neilson RP Dale VH Flather CH and others(2001) Global change in forests responses of speciescommunities and biomes BioScience 51 765minus779
Harsch MA Bader MY (2011) Treeline formminusa potential keyto understanding treeline dynamics Glob Ecol Biogeogr20 582minus596
Harsch MA Hulme PE McGlone MS Duncan RP (2009) Aretreelines advancing A global meta-analysis of treelineresponse to climate warning Ecol Lett 12 1040minus1049
Haylock MR Hofstra N Klein Tank AMG Klok EJ JonesPD New M (2008) A European daily high-resolutiongridded dataset of surface temperature and precipita-tion J Geophys Res Atmos 113 D20119
Hlaacutesny T Barcza Z Fabrika M Balaacutezs B and others (2011)
148
Cudliacuten et al Drivers of treeline shift
Climate change impacts on growth and carbon balanceof forests in Central Europe Clim Res 47 219minus236
Hlaacutesny T Trombik J Dobor L Barcza Z Barka I (2016)Future climate of the Carpathians climate change hot-spots and implications for ecosystems Reg EnvironChange 16 1495minus1506
Hofgaard A Dalen L Hytteborn H (2009) Tree recruitmentabove the treeline and potential for climate driven tree-line change J Veg Sci 20 1133minus1144
Holtmeier FK (2009) Mountain timberlines Ecology patchi-ness and dynamics 2nd edn Advances in GlobalChange Research 36 Springer Science amp ScienceMediaBV Dordrecht
Holtmeier FK Broll G (2005) Sensitivity and response ofnorthern hemisphere altitudinal and polar treelines toenvironmental change at landscape and local scalesGlob Ecol Biogeogr 14 395minus410
Honnay O Verheyen K Butaye J Jacquemyn H Bossuyt BHermy M (2002) Possible effects of habitat fragmentationand climate change on the range of forest plant speciesEcol Lett 5 525minus530
Huber R Rigling A Bebi P Brand F and others (2013) Sus-tainable land use in mountain regions under globalchange synthesis across scales and disciplines Ecol Soc18 36
Ibaacutentildeez I Clark JS Dietze MC (2009) Estimating colonizationpotential of migrant tree species Glob Change Biol 15 1173minus1188
IUCNSSC (International Union for Conservation of NatureSpecies Survival Commission) (2013) Guidelines for re -introductions and other conservation translocations Version 10 IUCNSpecies Survival Commission Gland
Jankovskyacute L Cudliacuten P gtermaacutek P Moravec I (2004) The pre-diction of development of secondary Norway sprucestands under the impact of climatic change in the Dra-hany highlands (The Czech Republic) Ekologia (Bratisl)23(Suppl 2) 101minus112
Jeniacutek J Lokvenc T (1962) Die alpine Waldgrenze im Krko -noše Gebirge Rozpravy SAV Rada matematickyacutech apriacuterodniacutech ved 72 gteskoslovenskaacute akademie ved Praha
Jobbaacutegy EG Jackson RB (2000) Global controls of forest lineelevation in the northern and southern hemispheresGlob Ecol Biogeogr 9 253minus268
Jonaacutešovaacute M Vaacutevrovaacute E Cudliacuten P (2010) Western Carpa -thian mountain spruce forest after a windthrow naturalregeneration in cleared and uncleared areas For EcolManag 259 1127minus1134
Kittel TGF Steffen WL Chapin FS (2000) Global andregional modelling of Arcticminusboreal vegetation distribu-tion and its sensitivity to altered forcing Glob ChangeBiol 6 1minus18
Klopcic M Jerina K Bonltina A (2010) Long-term changes ofstructure and tree species composition in Dinaric un -even- aged forests Are red deer an important factor EurJ For Res 129 277minus288
Klopcic M Simonltilt T Bonltina A (2015) Comparison of re -generation and recruitment of shade-tolerant and light-demanding tree species in mixed uneven-aged forests ex -periences from the Dinaric region Forestry 88 552minus563
Koumlrner C (2003) Alpine plant life functional plant ecology ofhigh mountain ecosystems with 47 tables Springer Sci-ence amp Business Media BV Dordrecht
Koumlrner C (2012) Alpine treelines Functional ecology of theglobal high elevation tree limits Springer Basel
Koumlrner C Paulsen J (2004) A world-wide study of high alti-
tude treeline temperatures J Biogeogr 31 713minus732Kulakowski D Bebi P Rixen C (2011) The interacting effects
of land use change climate change and suppression ofnatural disturbances on landscape forest structure in theSwiss Alps Oikos 120 216minus225
Kullman L (1988) Holocene history of the forestminusalpine tun-dra ecotone in the Scandes Mountains (central Sweden)New Phytol 108 101minus110
Kullman L (1999) Early holocene tree growth at a high elevation site in the northernmost Scandes of Sweden(Lapland) a palaeobiogeographical case study based onmega fossil evidence Geogr Ann 81 63minus74
Kullman L (2007) Tree line population monitoring of Pinussylvestris in the Swedish Scandes 1973minus2005 implica-tions for tree line theory and climate change ecologyJ Ecol 95 41minus52
Kullman L Oumlberg L (2009) Post-Little Ice Age treeline riseand climate warming in the Swedish Scandes a land-scape ecological perspective J Ecol 97 415minus429
Kyriazopoulos AP Skre O Sarkki S Wielgolaski FE Abra-ham EM Ficko A (2017) Humanminusenvironment dynamicsin European treeline ecosystems a synthesis based onthe DPSIR framework Clim Res 73 17ndash29
Lenoir J Svenning JC (2015) Climate-related range shifts mdasha global multidimensional synthesis and new researchdirections Ecography 38 15minus38
Lenoir J Geacutegout JC Marquet PA de Ruffray P Brisse H(2008) A significant upward shift in plant species opti-mum elevation during the 20th century Science 320 1768minus1771
Lindner M Fitzgerald JB Zimmermann NE Reyer C andothers (2014) Climate change and European forests What do we know what are the uncertainties and whatare the implications for forest management J EnvironManag 146 69minus83
Lloyd AH Rupp TS Fastie CL Srafield AM (2002) Patternsand dynamic of treeline advance on the Seward Pen -isula Alaska J Geophys Res 107 8161
Maacuteliš F Kopeckyacute M Petriacutek P Vladovic J Merganic J VidaT (2016) Life-stage not climate change explains ob -served tree range shifts Glob Change Biol 22 1904minus1914
Mathisen IE Mikheeva A Tutubalina OV Aune S and Hof-gaard A (2014) Fifty years of tree line change in theKhibiny Mountains Russia advantages of combinedremote sensing and dendroecological approaches ApplVeg Sci 17 6ndash16
Millar CI Stephenson NL (2015) Temperate forest healthin an era of emerging megadisturbance Science 349 823minus826
Mina M Bugmann H Klopcic M Cailleret M (2017) Accu-rate modeling of harvesting is key for projecting futureforest dynamics a case study in the Slovenian moun-tains Reg Environ Change 17 49minus64
Moen J Aune K Edenius L Angerbjoumlrn A (2004) Potentialeffects of climate change on treeline position in theSwedish mountains Ecol Soc 9 16 wwwecologyandsocietyorgvol9iss1art16
Palombo C Chirici G Marchetti M Tognetti R (2013) Is landabandonment affecting forest dynamics at high elevationin Mediterranean mountains more than climate changePlant Biosyst 147 1minus11
Palombo C Marchetti M Tognetti R (2014) Mountain vege-tation at risk current perspectives and research reedsPlant Biosyst 148 35minus41
Pauli H Gottfried M Dullinger S Abdaladze O Akhalkatsi
149
Clim Res 73 135ndash150 2017
M Alonso JLB Grabherr G (2012) Recent plant diversitychanges on Europersquos mountain summits Science 336 353minus355
Payette S Fortin MJ Gamachet I (2001) The subarctic forestminustundra the structure of a biome in a changing climate BioScience 51 709minus718
Pearson GR Dawson TP (2003) Predicting the impacts of cli-mate change on the distribution of species Are bio -climate envelope models useful Glob Ecol Biogeogr 12 361minus371
Pedrotti F (2013) Plant and vegetation mapping GeobotanyStudies Springer-Verlag Berlin
Pentildeuelas J Boada M (2003) A global change-induced biomeshift in the Montseny mountains (NE Spain) GlobChange Biol 9 131minus140
Piovesan G Biondi F Filippo AD Alessandrini A MaugeriM (2008) Drought driven growth reduction in old beech(Fagus sylvatica L) forests of the central ApenninesItaly Glob Change Biol 14 1265minus1281
Rabasa SG Granda E Benavides R Kunstler G and others(2013) Disparity in elevational shifts of European trees inresponse to recent climate warming Glob Change Biol19 2490minus2499
Raev I Zhelev P Grozeva M Markov I and others (2011)Programme of measures for adaptation of forests in Bul-garia and mitigate the negative impact of climate changeon them Project FUTURE forest helping Europe tackleclimate change INTERREG IVC ERDF Sofia
Sarkki S Ficko A Grunewald K Nijnik M (2016) Benefitsfrom and threats to European treeline ecosystem serv-ices an exploratory study of stakeholders and gover-nance Reg Environ Change 16 2019minus2032
Savage J Vellend M (2015) Elevational shifts biotic homo -genization and time lags in vegetation change during 40years of climate warming Ecography 38 546minus555
Schwoumlrer C Henne PD Tinner W (2014) A model-data com-parison of Holocene timberline changes in the SwissAlps reveals past and future drivers of mountain forestdynamics Glob Change Biol 20 1512minus1526
Šmilauer P Lepš J (2014) Multivariate analysis of ecologicaldata using Canoco 5 Cambridge University Press Cam-bridge
Ste-Marie C Nelson EA Dabros A Bonneau ME (2011)Assisted migration introduction to a multifaceted con-cept For Chron 87 724minus730
Stevens JT Stafford HD Harrison S Latimer AM (2015) For-est disturbance accelerates thermophilization of under-story plant communities J Ecol 103 1253minus1263
Strid A Andonoski A Andonovski V (2003) Alpine biodiver-sity in Europe Ecological Studies 167 Springer-VerlagBerlin
Tattoni C Ciolli M Ferretti F Cantiani MG (2010) Monitor-ing spatial and temporal pattern of Paneveggio forest(Northern Italy) from 1859 to 2006 iForest Biogeosci For3 72minus80
Ter Braak CJF Šmilauer P (2012) Canoco reference manualand userrsquos guide software for ordination (version 50)Microcomputer Power Ithaca NY
Theurillat JP Guisan A (2001) Potential impact of climatechange on vegetation in the European Alps a reviewClim Change 50 77minus109
Toslashmmervik H Wielgolaski FE Neuvonen S Solberg BHoslashgda KA (2005) Biomass and production on a land-scape level in the mountain birch forests In WielgolaskiFE (ed) Plant ecology herbivory and human impact in
Nordic mountain birch forests Ecos Studies 180Springer Berlin p 53minus70
Treml V Banaš M (2000) Alpine timberline in the HighSudetes Acta Univ Carol Geogr 15 83minus99
Treml V Chuman T (2015) Ecotonal dynamics of the altitu-dinal forest limit are affected by terrain and vegetationstructure variables an example from the Sudetes Moun-tains in Central Europe Arct Antarct Alp Res 47 133minus146
Treml V Migo P (2015) controlling factors limiting timber-line position and shifts in the Sudetes a review GeogrPol 88 55minus70
Urban MC Tewksbury JJ Sheldon KS (2012) On a collisioncourse competition and dispersal differences create no-analogue communities and cause extinctions during cli-mate change Proc R Soc Lond B Biol Sci 279 2072minus2080
Vacek S Matejka K (2010) State and development of phyto-cenoses on research plots in the Krkonoše Mts foreststands J For Sci 56 505minus517
Van Bogaert R Haneca K Hoogesteger J Jonasson C deDapper M Callaghan TV (2011) A century of tree linechanges in sub-Arctic Sweden shows local and regionalvariability and only a minor influence of 20th century climate warming J Biogeogr 38 907minus921
Van Gils H Batsukh O Rossiter D Munthali W Liberato -scioli E (2008) Forecasting the pattern and pace of Fagusforest expansion in Majella National Park Italy ApplVeg Sci 11 539minus546
Velchev V (1997) Types of vegetation In Yordanova MDonchev D (eds) Geography of Bulgaria BAN Sofiap 269minus283 (in Bulgarian)
Vladovic J Merganic J Maacuteliš F Križovaacute E and others (2014)The response of forest vegetation diversity to changes inedaphic-climatic conditions in Slovakia Technickaacute uni-verzita vo Zvolene Zvolene (in Slovak)
Vrška T Adam D Hort L Kolaacuter T Janiacutek F (2009) Europeanbeech (Fagus sylvatica L) and silver fir (Abies alba Mill)rotation in the Carpathians mdash a developmental cycle or alinear trend induced by man For Ecol Manag 258 347minus356
Wielgolaski FE (ed) (2005) Plant ecology herbivory andhuman impact in Nordic mountain birch forests Eco -logical Studies 180 Springer Verlag Berlin
Wielgolaski FE Hofgaard A Holtmeier FK (2017) Sensitivityto environmental change of the treeline ecotone and itsassociated biodiversity in European mountains Clim Res73151ndash166
Wieser G Tausz M (2007) Trees at their upper limit Treelifelimitation at the alpine timberline Springer Dordrecht
Wieser G Holtmeier FK Smith WK (2014) Treelines in achanging global environment In Tausz M Grulke N(eds) Trees in a changing environment Plant Ecophys 9Springer Dordrecht p 221minus263
Zhiyanski M Kolev K Sokolovska M Hursthouse A (2008)Tree species effect on soils in Central Stara PlaninaMountains Nauka Gorata 4 65minus82
Zhiyanski M Gikov A Nedkov S Dimitrov P Naydenova L(2016) Mapping carbon storage using land coverlanduse data in the area of Beklemeto Central Balkan In Koulov B Zhelezov G (eds) Sustainable mountain re -gions challenges and perspectives in southeasternEurope Springer Basel p 53minus65
Zhu K Woodall CW Clark JS (2012) Failure to migrate lackof tree range expansion in response to climate changeGlob Change Biol 18 1042minus1052
150
Editorial responsibility Nils Chr Stenseth Oslo Norway
Submitted April 8 2016 Accepted March 8 2017Proofs received from author(s) July 7 2017
Clim Res 73 135ndash150 2017
mate the geographical position of these mountainunits 1 value was used for the geographical latitudeand longitude of the locus of the studied mountainunits The analysed parameters of these 15 mountainunits are shown in Table 2
In addition we calculated the following climatecharacteristics for all mountain units annual meanair temperature annual sum of precipitation thegrowing season length and the date of the onset ofthe growing season These characteristics were thencompared for 2 distinct periods 1961minus1990 and1991minus 2015 (Table 3) The climate characteristicswere derived from the E-OBS gridded dataset of sta-tion observations version 131 (Haylock et al 2008)We used the E-OBS version on the regular longi-tudeminuslatitude grid with a horizontal resolution of 025degrees The climate data from all grid points withinthe mountain units or near their geographical bor-ders were considered
Data processing for 15 mountain units was made onthe basis of data obtained from the literature and
personal studies by the co-authors these are sum ma -rized in Tables 1 amp 2 and in Tables S1 amp S2 in theSupplement at www int-res com articles suppl c073p135 _ supp pdf In addition a semi-quantitative valu-ation of abiotic biotic and anthropogenic factors lim-iting an upward treeline shift was performed (Fig 2)To answer the question how natural and anthro-pogenic factors have influenced treeline ecotonecharacteristics with a focus on treeline shift redun-dancy analysis (RDA) was used to describe and testthe effect of the explanatory variables (geographicalposition size of the whole mountain massif start ofhuman influence and start of the decrease in humaninfluence) on treeline ecotone characteristics (tim-berline elevation width of the treeline ecotone treespecies forming the treeline and treeline shift peryear) and identify groups of mountain units with sim-ilar variability of the dependent data (ter Braak ampŠmilauer 2012) The whole data set for this analysis isshown in Table S2 We tested both their simple ef -fects which show how much variation every ex plan -
140
Country Coordi- Elevation Elevation Range of Annual Area VZs near Tr part of nates of range range of Tr tree species altitudinal of study (m asl)mountains central site of T (m asl) limit shift (km2)
(m asl) (m asl) of T or Tr (m yrminus1)
Slovakia 48deg55rsquoN 1400 1530 Pa 1730 ~03 (Tr ~400 Spruceminusfirminusbeech Dumbier Tatra 19deg31rsquoE (1350minus1450) (1400minus1630) 1950minuspresent) (1200minus1350) spruce (1300minus(part of Low 1550) Pm with individual Tatra Mts) Pa trees (1400minus1700)
Czech Republic 50deg42rsquoN 1250 1320 Pa 1500 043 (Tr 550 Beechminusspruce (700minus1050) Giant Mts 15deg38rsquoE (1130minus1460) (1250minus1460) 1936minus2005) Pa dominates with increasing
altitude spruce (1000minus1400) Pm with Pa trees (1400minus1560)
Norway 60deg10rsquoN 850 600minus1050 1500 08 (Tr 40 600 Birch forest dominance Southern 7deg40rsquoE 62deg20rsquoN 1915minus2007) at high altitudes but with Scandes 10deg05rsquoE some scattered Ps or Pa
mixed birchminusconifer below 800
Norway 68deg20rsquoN West 650 West 750 West 1000 06 (Tr 35 000 Birch forest dominance Sweden 18deg20rsquoE East 100 East 290 East 0 1958minus2008) at all altitudes but with some Finland 69deg50rsquoN scattered pines or pine stands Northern 27deg00rsquoE on sandy soilsScandes
Table 1 (continued)
Cudliacuten et al Drivers of treeline shift
atory variable can explain separately without usingthe other variables and their conditional ef fectswhich depend on the variables already selected inthe model The statistical significance of the expla -natory variables was tested using a Monte Carlo permutation test and only predictors with p le 005were included in the subsequent RDA (Šmilauer ampLepš 2014)
To find the drivers of biodiversity loss in forestsmeadows and animal communities (dependent vari-ables) the explanatory variables (geographical posi-tion timberline elevation tree species forming thetreeline start of human influence start of the de -crease in human influence and temperature increasebetween the 2 periods 1961minus1990 and 2021minus2050)were tested again by RDA in Canoco 5 as mentionedabove (Figs 3 amp 4) The whole data set for this analy-sis mdash including additional information about tree-lines in the mountain areas of interest with a focuson treeline shift (its rate drivers limits and problemswith its assessment) mdash is presented in Tables S1 amp S2
Selected univariate graphs were constructed to visu-alize the data of relationships between the explana-tory and dependent variables of all analyses whichwere not seen distinctly from the results of the multi-variate statistics or linear regression (Fig 5)
3 RESULTS
31 Effect of environmental variables on treeline ecotone characteristics with a focus
on treeline shift
The variation in the dependent variables (timber-line elevation width of treeline ecotone tree speciesforming the treeline and treeline shift per year) wassignificantly affected only by latitude and size of thewhole mountain massif (Fig 3) The adjusted ex -plained variability by all explanatory variables was342 (F = 46 p = 001) The RDA diagram showsthat in the mountains located more to the north the
141
Human Annual Limits Referencesinfluence temperature (T ) to climate
in VZs in past and precipitation (P) change and recently differences induced species
between 1961minus1990 shiftand 2021minus2050
mean of CORDEX models
From 14th C to 1922 Average of 10 RCM_B1 More rocky and nutrient- Koumlrner (2003) mining forest change ΔT2050 = 18degC poor soils late frosts Fridley et al (2011)
to spruce monocultures ΔT2100 = 37degC steep slopes with avalanches Hlaacutesny et al (2011 2016)from 13th C to 1978 pastures ΔP2050 = +24 mm absence of mature trees
in upper parts of mountain ΔP2100 = minus35 mm
9minus11th C forest clear cutting ALADIN_A1B period 1961minus Different soil types Treml and Banaš (2000) grazing in dwarf pine VZ 2100 ΔT2050 = 13degC for Fs and Pa (cambisol Cudliacuten et al (2013)
since 18th C forest change ΔT2100 = 34degC versus podzol) only TSL of Pa Treml amp Chuman (2015) to spruce monocultures ΔP2050 = +25 mm could elevate on ranker Treml amp Migo (2015)
since 19th C artificially planted ΔP2100 = minus14 mm soils in Pm VZ
Grazed and browsed RegClimMPIHadley Sheep and reindeer Dalen amp Hofgaard (2005) by reindeer and livestock ΔT2050 = 12degC grazing and episodic Wielgolaski (2005)
(sheep and cattle) ΔT2100 = 22degC insect outbreaks (Epirrita) Hofgaard et al (2009) over thousands of years ΔP2050 = +23 mm Kullman Oumlberg (2009)
ΔP2100 = minus12 mm
Grazed and browsed RegClimMPIHadley Continued grazing regime Dalen amp Hofgaard (2005) by semi-domestic reindeer ΔT2050 = 16degC and frequent episodic Wielgolaski (2005)
for hundreds of years ΔT2100 = 29degC insect outbreaks (Epirrita) Toslashmmervik et al (2005) increased grazing from ΔP2050 = +18 mm Hofgaard et al (2009)
year 1960 ΔP2100 = minus10 mm Aune et al (2011) Mathisen et al (2014)
Clim Res 73 135ndash150 2017
treeline ecotone is narrower treeline shiftis smaller the timberline occurs at lowerelevations and the treeline is formed moreby broadleaved species Higher values oftreeline ecotone width and treeline shiftwere found in the mountains with a greatersize of the whole mountain massif
When every explanatory variable wastested separately values of the start of thedecrease of human influence in the moun-tains also had a significant effect on thevariation of the tree line ecotone data(explained variation = 206 pseudo-F =34 p = 0024) Some selected relationshipsof the explanatory and dependent vari-ables even those not showing significanteffects on the variation in the tree line eco-tone parameters are depicted in Fig 5
32 Effect of recent changes in climateand land use on the biodiversity
Biodiversity loss in forests meadows andanimal communities analysed by RDAwas explained by geo graphical positiontreeline species composition and tempera-ture increase between the 2 periods1961minus1990 and 2021minus2050 (Fig 4) Theadjusted ex plained variation by all ex pla -natory variables was 632 (F = 58 p =0001) The RDA results indicated that bio-diversity loss in forest communities in crea -sed with increasing latitude and longitudeas well as in the case where broad leavedspecies formed the treeline In contrast thehighest biodiversity loss in meadows wasfound in the mountains positioned more tothe south and with higher temperatureincrease The biodiversity loss in animalcommunities was negatively correlatedwith longitude and temperature increase
4 DISCUSSION
The comparison of several mountainareas situated across Europe shows a varia-tion in understanding of what constitutesthe treeline across countries The primaryissue is the different approach to the defini-tion of forest when applying a tree heightthreshold This threshold decreases from
142
Mou
nta
in u
nit
Nor
th
Eas
t M
oun
-T
imb
er-
Wid
th
Tre
e T
emp
erat
ure
T
reel
ine
Sta
rt
Sta
rt
lati
tud
elo
ng
itu
de
tain
lin
eof
spec
ies
incr
ease
sh
ift
of H
I of
HI
size
(m a
sl
)tr
eeli
ne
form
ing
b
etw
een
(m
yrminus
1 )(y
r)d
ecre
ase
(km
2 )ec
oton
eth
e 19
61minus
1990
(y
r)(m
)tr
eeli
ne
and
199
1minus20
15 (
degC)
Cen
tral
Pyr
enee
s (C
P)
42deg2
1rsquoN
0deg30
rsquoW55
000
2250
300
C0
81
92 (
1956
minus20
06)
1000
BC
1930
Eas
tern
Pyr
enee
s (E
P)
42deg2
8rsquoN
1deg30
rsquoE55
000
2400
300
C0
80
71 (
1956
minus20
06)
1000
1930
Ap
enn
ines
(A
P)
42deg0
5rsquoN
14deg0
5rsquoE
5000
018
0030
0B
14
1 (1
954minus
2007
)10
00 B
C19
50S
har
a M
ts (
SH
)41
deg47rsquo
N20
deg33rsquo
E16
0018
5035
0C
B0
71
(193
4minus20
10)
1300
1970
Pir
in (
PI)
41deg4
2rsquoN
23deg3
1rsquoE
2585
1900
400
C0
5n
o d
ata
680
1962
Cen
tral
Bal
kan
Mts
(C
B)
42deg4
7primeN
24deg3
6primeE
1160
016
0025
0C
B0
7n
o d
ata
1700
1980
Nor
ther
n D
inar
ic M
ts (
DI)
45deg3
6rsquoN
14deg2
8rsquoE
1000
1540
160
B1
3n
o d
ata
1400
1900
Cen
tral
Alp
s (C
A)
46deg4
6rsquoN
9deg52
rsquoE18
000
021
0055
0C
07
13
(195
4minus20
15)
1300
1880
Eas
tern
Alp
s (E
A)
46deg1
7rsquoN
11deg4
5rsquoE
180
000
1800
400
C0
91
3 (1
954minus
2015
)11
5019
50L
ow T
atra
Mts
(LT
)48
deg55rsquo
N19
deg31rsquo
E12
4014
0033
0C
09
057
(19
75minus
2016
)14
0019
22G
ian
t M
ts (
GM
)50
deg42rsquo
N15
deg38rsquo
E63
012
5025
0C
07
043
(19
36minus
2005
)10
0019
63H
ard
ang
ervi
dd
a (H
G)
60deg0
3rsquoN
7deg47
rsquoE90
0090
015
0C
B0
50
8 (1
915minus
2007
)80
019
70D
ovre
(D
O)
62deg2
1rsquoN
8deg52
rsquoE60
0010
0040
0B
07
08
(191
5minus20
07)
1000
no
dec
reas
eN
orth
ern
Sw
edis
h L
apla
nd
(N
S)
68deg1
2rsquoN
18deg3
3rsquoE
4000
700
300
B1
10
6 (1
958minus
2008
)10
00n
o d
ecre
ase
Inn
er F
inn
mar
kn
orth
ern
mos
t 69
deg10rsquo
N25
deg12rsquo
E60
000
290
90B
10
06
(195
8minus20
08)
1600
no
dec
reas
eF
inn
ish
Lap
lan
d (
FL
)
Tab
le 2
Fac
tors
infl
uen
cin
g t
reel
ine
ecot
one
char
acte
rist
ics
in s
elec
ted
Eu
rop
ean
mou
nta
ins
Th
e m
oun
tain
siz
e is
th
e si
ze o
f th
e w
hol
e m
oun
tain
mas
sive
fu
nct
ion
ing
as a
mez
osca
le c
lim
ate
un
it n
ot o
nly
the
stu
die
d m
oun
tain
are
a W
idth
of t
he
tree
lin
e ec
oton
e w
as d
eter
min
ed b
y su
btr
acti
ng
the
valu
e of
the
mea
n ti
mb
erli
ne
elev
atio
n
from
th
e va
lue
of t
he
up
per
tre
e sp
ecie
s li
mit
C c
onif
ers
B b
road
leav
ed t
rees
HI
hu
man
infl
uen
ce
Cudliacuten et al Drivers of treeline shift
5 m (Jeniacutek amp Lokvenc 1962) to 3 m(Koumlrner 2012) in Northern andCentral Europe to only 2 m (Holt-meier 2009) in Central and South-ern Europe where even shrub -by stands (eg Pinus mugo) areconsi dered as a forest in somecountries eg in Spain Italy andMa ce donia (Batllori et al 2009)An other problem is a different de -signation of forest stands belowthe treeline lsquoupper montane fo -restsrsquo in Central Europe and lsquosub-alpine forestsrsquo in Southern Europe(Ellenberg 1988) Further differen -ces are related to the tradition ofdifferent branches of science (ega more traditional lsquogeobotanicalrsquoapproach in Central Europe ver-sus a more experimental approachin Western Europe) especially inthe rate of applying new pro -gressive methods (eg climaticmodelling or molecular biologicalmethods)
Our comparison of selectedregions (Tables 1 amp 2 Figs 1 amp 2)showed that southern mountainscompared to those located morecentrally and in the north had (1) alonger and more profound exploi -tation by humans in the past (2)greater differences in climatic para -meters (temperature length ofgrowing period) between the 2periods 1961minus1990 and 1991minus2015(Table 3) and (3) more dramaticclimate change scenarios espe-cially concerning temperature in -creases Longitude also had someinfluence on the start and intensityof human influence and on tree-line elevation and rate of treelineshift (Table 2 Fig 5) The occur-rence of broadleaf tree species inthe treeline ecotone in the north-ern (and to some extent the south-ern) countries distinguishes thesefrom the Central European coun-tries where conifers prevail It isinteresting that grazing an im -portant factor shaping treelineecosystems (Dirnboumlck et al 2003
143
Length Beginning Mean Mean of growing of growing annual annual
period season temperature precipitation (d)a (d)b (degC) ()
Central Pyrenees 135 minus89 08 minus107Eastern Pyrenees 117 minus91 08 minus92Apennines 284 minus232 14 57Shara Mts 16 01 07 minus100Pirin 26 minus14 05 27Central Balkan Mts 169 minus136 07 76Northern Dinaric Mts 160 minus95 13 minus56Central Alps 147 minus86 07 62Eastern Alps 125 minus96 09 minus33Low Tatra Mts 24 minus34 09 minus91Giant Mts minus10 13 07 55Hardangervidda Blefjell 63 34 05 56Dovre 138 32 07 64Northern Swedish Lapland 81 05 11 minus27Inner Finnmarknorthern- 17 08 10 77most Finnish Lapland
aPositive numbers indicate a prolongation bNegative numbers indicate an earlier beginning
Table 3 Differences of selected climatic parameters between the 2 study periods (1961minus1990 and 1991minus2015) in selected European mountains
Tein Drou Gefa Nadi Laus Abli Bili Huli
0 1 2 3
Central Pyrenees (CP)
Eastern Pyrenees (EP)
Apennines (AP)
Shara Mts (SH)
Pirin (PI)
Central Balkan Mts (CB)
Northern Dinaric Mts (DI)
Central Alps (CA)
Eastern Alps (EA)
Low Tatra Mts (LT)
Giant Mts (GM)
Hardangervidda (HG)
Dovre (DO)
Northern Swedish Lapland (NS)
Inner FinnmarknorthernmostFinnish Lapland (FL)
Fig 2 Rates of treeline drivers (temperature increase [Tein] land use change[Laus]) and treeline shift limits (drought [Drou] geomorphological factors [Gefa]natural disturbances [Nadi] other abiotic biotic and human factors) in selectedEuropean mountains Abli Bili and Huli abiotic biotic and human treeline shiftlimits respectively) Rate of treeline driver influence mdash 0 no influence 1 weak
influence 2 middle influence 3 strong influence
Clim Res 73 135ndash150 2017
Gehrig-Fasel et al 2007) limits treeline shift in halfof the countries of interest regardless of latitude lon-gitude or recent political developments (Shara PirinCentral Balkans Alps Scandes Fig 2)
Despite the stated differences between studieslooking at climate change impacts on the treelineit is clear that current and future changes in temp -erature will seriously affect treeline ecotones in allmountain ranges of Europe Winter is a period whentreeline stands and individual trees have to survivesevere life-limiting conditions (Wieser amp Tausz2007) Therefore winter warming might increase thechance of young trees surviving and passing the mostcritical period from seedling to sapling and further tothe mature stage (Koumlrner 2003) There is alreadymuch evidence worldwide that winter warming isone of the significant drivers of treeline advance(Harsch et al 2009) Although it is expected that treegrowth at the upper distributional margins (and inthe northern European countries) will increase due toongoing climate change (Pentildeuelas amp Boada 2003Chen et al 2011 Hlaacutesny et al 2011 Lindner et al2014) extremes in temperature (eg black froststemperature reversals in the spring) or winter precip-itation (eg lack of snow drought) might also limit
this process in the future in some regions and localsituations (Holtmeier amp Broll 2005 Hagedorn et al2014) An earlier start to the growing period (espe-cially in the Apennines Central Balkan and SpanishPyrenees see Table 2) can play a negative role in theresistance of trees and seedlings to early springfrosts On the other hand at lower distributional mar-gins (and in Southern European countries) it isexpected that tree growth will decrease or forestswill experience some level of dieback due to drought(Breshears et al 2005 Piovesan et al 2008 Allen etal 2010 Huber et al 2013) A serious implication isthe positive effect that warming can have on root rotfungi and populations of bark beetles resulting inlarge-scale disturbances to temperate and also high-altitude forests (Jankovskyacute et al 2004 Elkin et al2013 Millar amp Stephenson 2015)
Although recent upward advances of treeline eco-tones are widespread in mountain regions (Harsch et
144
Fig 3 Redundancy analysis diagram with variation in tim-berline elevation width of treeline ecotone the tree speciesforming the treeline and treeline shift used as dependentvariables explained by explanatory variables (latitudemountain size) Mountain units (blue) are projected as thecentres of abbreviations (see Table 2) Explanatory variablesaccount for 436 of the total variation in the dependentdata The first canonical axis explained 466 of variationthe second axis explained 30 of variation Dependentvariables (black) are Ecot altitudinal width of treeline eco-tone TLShift treeline shift TrspC (TrspB) treeline formedby conifers (broadleaved trees) Timb timberline elevationExplanatory variables (red) are N north latitude Mtsize
size of the mountain massif
Fig 4 Redundancy analysis diagram with variation in bio-diversity loss in forests meadows and animal communitiesused as dependent variables explained by explanatory vari-ables (geographical position tree species forming the tree-line and temperature increase between the 2 study periods[1961minus1990 and 2021minus2050]) Mountain units (green) areprojected as centres of abbreviations (see Table 2) Explana-tory variables account for 763 of total variation in the de-pendent data The first canonical axis explained 484 ofvariation the second axis explained 183 of variation De-pendent variables (black) are BdfoBdmeBdan biodiversityloss in forestsmeadowsanimal communities Explanatoryvariables (red) are N north latitude E east longitude TrspC(TrspB) treeline formed by conifers (broadleaved trees)Temp temperature increase between the 2 study periods
Cudliacuten et al Drivers of treeline shift
al 2009) only a few published studies have includedquantitative data about treeline shifts (Devi et al2008 Kullman amp Oumlberg 2009 Diaz-Varela et al 2010Van Bogaert et al 2011) Additionally our know -ledge of the spatial patterns in treeline ecotone shiftsat the landscape scale is still surprisingly poor exceptfor some studies from subarctic areas the UralMountains and the Alps (Lloyd et al 2002 Diaz-Varela et al 2010 Hagedorn et al 2014) In the 15studied mountain units the values of treeline shiftranged from 043 to 19 m yrminus1 showing a rather dis-tinct spatial pattern of the dynamics within Europeanmountains The observed treeline shift had a signifi-cant positive relationship only with northern latitudeand a weak positive relationship with mountain-range size (ie the size of the whole mountain massifFig 3) Nevertheless this illustrates the importanceof the mass elevation effect (Koumlrner 2012) and par-tially ex plains why treelines could be at different alti-tudes at the same latitude Small negative regres-sions with east longitude and timberline elevationare shown in Fig 5 There was no apparent depend-ence of the treeline shift rate on climatic parameterchanges between the periods 1961minus1990 and1991minus2015
A whole forest vegetation zone shift is a complexprocess as not only the life strategies of an individualtree need to be considered but plant animal andmicroorganism species and their interrelationshipsas well as the relationships to their specific microhab-
itats (including soil conditions) are also involved(Urban et al 2012) Across all regions we identifiedseveral important obstacles to treeline shifts oftenrelated to site properties such as significant rocki-ness having shallow or low-nutrient soils extremerelief causing disturbances by avalanches snow glid-ing or wind damage (data from the Pyrenees Apen-nines Shara Northern Dinaric Low Tatra and GiantMountains see Fig 2) Soil heterogeneity certainlyhas an important role in plant responses to climatechange and could also maintain the resilience of thecommunity assemblages (Fridley et al 2011) An -other group of obstacles includes extreme climaticparameters especially in winter (reported eg fromthe Pyrenees Apennines Pirin Central BalkanNorthern Dinaric and Giant Mountains Fig 2) Theircombination can result in edaphic andor climaticunsuitability of habitats where species could poten-tially migrate Climatically and edaphically suitablesites will likely decline over the next century partic-ularly in mountain landscapes (Bell et al 2014)There fore trees in the treeline ecotone will colonizepreviously forested habitats The colonization suc-cess of individual forest communities is affected bydifferences in species dispersal and recruitmentbehaviour (Dullin ger et al 2004 Jonaacutešovaacute et al2010) For these reasons colonization of new foresthabitats by the next tree generation may not be suc-cessful and may result in loss of species diversity(Honnay et al 2002 Ibaacutentildeez et al 2009 Dobrowski et
145
Fig 5 Univariate graphs of the relationships of dependent variables viz timberline (Timb) and treeline shift per year (Shift) and their explanatory variables Nola north latitude Ealo east longitude DD decimal degrees
Clim Res 73 135ndash150 2017
al 2015) as reported from Macedonia Slovenia theCzech Republic Slovakia and Norway (Fig 4)
Another limiting factor is the existence of strongadaptation mechanisms of the dominant tree specieswhich allow them to survive in their current distribu-tion areas Species migration is likely to be slow dueto the limited quantity of climatically and edaphicallysuitable sites and the slow velocity of seed dispersaland tree regeneration and will be hampered evenmore by fragmentation of high mountain landscapescaused by human activities including brush invasionin abandoned meadows (Gartzia et al 2014) A verti-cal shift in a vegetation zone involves not only singlespecies of plants animals and microorganisms butalso their interrelationships and links to soil condi-tions The speed at which climatic conditions changewill be different from changes in the soil conditionsdue to the slowness of soil-forming processes Soiltypes and conditions (cambisol versus podzol) arecrucial obstacles for the shift from a beech forest zoneinto the spruce forest zone in the Czech Republic(Vacek amp Mat jka 2010) According to other sourcesthe dominant tree species can influence soil for -mation processes (especially humus forms) Underfavourable climatic and orographic conditions beechis able to change its soil conditions over decades tocenturies (J Macku unpubl data)
Significant changes in biodiversity must be expec -ted in all of the mountain areas of interest We foundthat biodiversity declined particularly in the southernregions of Europe where the timberline is situated athigher elevations and human impact is mostlylonger-lasting (Fig 4) Similarly Pauli et al (2012)reported an increase in species richness of mountaingrasslands across Europe except for Mediterraneanregions and assigned this different response of thesouthern regions to decreased water availability Onthe other hand ecosystems which exhibited bio -diversity increases were not only enriched by mig -rant species but the assemblages underwent thermo -philization ie cold-adapted species declined andmore warm-adapted species spread (Gottfried et al2012) However not all species are able to track cli-mate changes It is expected that around 40 ofhabitats will become climatically unsuitable for manymountain species particularly endemic ones in -creasing their extinction probability during this cen-tury (Dullinger et al 2012) The abandonment of tra-ditional land use forms is another source of this losswhich may be a more important driver than tempera-ture increase (Fig 4) Serious biodiversity losses inmountain meadows were reported from Spain ItalyMacedonia Bulgaria Slovenia and Slovakia There-
fore controlled grazing must occur in order to main-tain alpine grasslands (Dirnboumlck et al 2003) Unfor-tunately grazing is still active only in smaller parts ofEuropean mountains (eg in the Central Alps and theCentral Balkan Mountains but pasturing can alsonegatively affect vegetation diversity) and some-times does not serve to maintain alpine grasslandThe same is true for grassland management whichcan help to maintain mountain species and decreasehabitat loss A serious biodiversity loss in mountainmeadows was reported from the Pyrenees the Apen-nines and the Shara Pirin Central Balkan NorthernDinaric and Low Tatra Mountains (Fig 4)
In forests species responses to climate change maybe equivocal some recent studies indicated contra-dictory shifts in species distributions (eg Lenoir etal 2008 Zhu et al 2012 Rabasa et al 2013) This dis-parity is widely discussed and assigned to the greatvariety of non-climatic factors or even tree ontogeny(Grytnes et al 2014 Lenoir amp Svenning 2015 Maacutelišet al 2016) Disturbances leading to tree mortalitymay also play an important role (Cudliacuten et al 2013)changes in tree canopy cover modify light availabil-ity and microclimate and can induce the thermo -philization of forest vegetation (De Frenne et al2015 Stevens et al 2015) The loss of this microcli-mate buffering effect of forests may induce a biotichomogenization of forests (Savage amp Vellend 2015)or the creation of novel non-analogical communitiessuch as oakminuspine forests (Urban et al 2012) whichmay be a new threat to forest biodiversity
The observed changes in treeline forest ecosystemsare often related to changes in land-use intensity(Theurillat amp Guisan 2001 Alados et al 2014) Ac -cording to climate change predictions and the recentand future exploitation intensity of European moun-tains trees and forest communities will shift upwarddue to land use change or climate change or bothReduced land-use intensity certainly will interactwith climate change by facilitating or inhibiting spe-cies occurrence and will accelerate forest expansionabove the present treeline particularly to previouslyforested habitats (Theurillat amp Guisan 2001) Thesimultaneous action of both main treeline shift driv-ers viz temperature increase and decrease in landuse intensity was recorded from all of our studiedmountain areas The extensive differences in timber-line and treeline elevations in almost all studiedmountains (Table 1) indicate the anthropo-zoogenictreeline type (according to Ellenberg 1998) In mostmountains (eg in the Apennines Shara MountainsCentral Balkans and Alps) land use is the prevailingfactor influencing vegetation drift (Fig 5) Previous
146
Cudliacuten et al Drivers of treeline shift
research showed that the upward shift of the treelinein the Swiss Alps was predominantly attributable toland abandonment and only in some situations to cli-mate change (Gehrig-Fasel et al 2007) In the Apen-nines in the last few decades tree establishment hasbeen mainly controlled by land use while treegrowth has been controlled by climate pointing to aminor role played by climate in shaping the currenttreeline (Palombo et al 2013)
The impact of climate change and the connectedland use change on biodiversity and ecosystem serv-ices provision in several European countries is sum-marized by Wielgolaski et al (2017) and Kyriazopou-los et al (2017) (both this Special) and Sarkki et al(2016) One of the possible adaptive management op-tions in response to climate change assisted migra-tion as human-assisted movement of species hasbeen frequently debated in the last few years (Ste-Marie et al 2011) It is possible to apply it as a type ofassisted colonization ie intentional movement andrelease of an organism outside its indigenous range toavoid extinction of populations of the focal species(eg Macedonian pine species) or as an ecologicalre placement ie the intentional movement and re-lease of an organism outside its indigenous range toperform a specific ecological function (eg planting ofPinus mugo in the Alps IUCNSSC 2013)
5 CONCLUSIONS
The analysis of 11 mountain areas across Europeshowed that with increasing latitude the treelineand altitudinal width of the treeline ecotone signifi-cantly decreases as do the significance of climaticand soil parameters as barriers against tree speciesshift Although temperature is the overwhelmingcontrolling factor of tree growth and establishment intemperate and boreal treeline ecotones late-sea-sonal drought might also play a driving role in Medi-terranean treeline ecotones Longitude was lessinfluential mostly affecting climate and land usechange effects on increased biodiversity loss as wellas the size of the area that forms one mesoclimateunit affecting altitudinal ecotone width The biggestpart of the commonly observed remaining variabilityin mountain vegetation near the treeline in Europeseems to be caused by geomorphological geologicalpedological and microclimatic variability in combina-tion with different land use history and the presentsocio-economic relations The observed differencesin climatic parameters between the mountain areasof interest in comparison with the reference period
1960minus1990 (09degC) have not explained the relativelyhigh differences in the rate of treeline shift per year(149 m) The predicted variability in temperatureincrease due to global warming (12minus26degC in 2050and 22minus53degC in 2100) could lead to a much biggerdifferentiation in treeline ecotone biodiversity andecosystem processes between southern and northernEuropean mountains in the future Therefore thesedifferences must be taken into account by scientistsand EU policy makers when formulating efficientadaptive forest management strategies for treelineecosystems at the European level (eg assistedmigration of adapted genotypes)
Acknowledgements This paper is based firstly upon workfrom the COST Action ES 1203 SENSFOR supported byCOST (European Cooperation in Science and Technologywwwcosteu) This international work was further sup-ported by projects granted by the Ministry of EducationYouth and Sports of the Czech Republic grant NPU ILO1415 and LD 14039 by the agency APVV SR under pro-jects APVV-14-0086 and APVV-15-0270 We acknowledgethe E-OBS dataset from the EU-FP6 project ENSEMBLES(httpensembles-eumetofficecom) and the data providersin the ECAampD project (wwwecadeu)
LITERATURE CITED
Alados CL Errea P Gartzia M Saiz H Escoacutes J (2014) Posi-tive and negative feedbacks and free-scale pattern distri-bution in rural-population dynamics PLOS ONE 9 e114561
Allen CD Macalady AK Chenchouni H Bachelet D andothers (2010) A global overview of drought and heat-induced tree mortality reveals emerging climate changerisks for forests For Ecol Manag 259 660minus684
Amidžic L Bartula M Jaric S (2012) Syntaxonomy overviewof vegetation of Šar Planina in the Balkan PeninsulaVegetos 25 348minus360
Aune S Hofgaard A Soumlderstroumlm L (2011) Contrasting cli-mate and land use driven tree encroachment pattern ofsub-arctic tundra in Northern Norway and Kola Penin-sula Can J For Res 41 437minus449
Barbeito I Dawes M Rixen C Senn J Bebi P (2012) Factorsdriving mortality and growth at treeline a 30-year exper-iment of 92000 conifers Ecology 93 389minus401
Batllori E Camarero JJ Ninot JM Gutieacuterrez E (2009)Seedling recruitment survival and facilitation in alpinePinus uncinata treeline ecotones Implications andpotential responses to climate warming Glob Ecol Bio-geogr 18 460minus472
Bebi P Teich M Hagedorn F Zurbriggen N Brunner SHGreacutet-Regamey A (2012) Veraumlnderung von Wald undWaldleistungen in der Landschaft Davos im Zuge desKlimawandels Schweiz Z Forstwes 163 493minus501
Bell DM Bradford JB Lauenroth WK (2014) Mountain land-scapes offer few opportunities for high elevation treespecies migration Glob Change Biol 20 1441minus1451
Bergant K (2006) Climate change scenarios for MacedoniaUniversity of Nova Gorica Center for Atmospheric
147
Clim Res 73 135ndash150 2017
Research Nova GoricaBodin J Badeau V Bruno E Cluzeau C Moisselin JM
Walther GR Dupouey JL (2013) Shifts of forest speciesalong an elevational gradient in Southeast France cli-mate change or stand maturation J Veg Sci 24 269minus283
Boncina A (2011) History current status and future pros -pects of uneven-aged forest management in the Dinaricregion an overview Forestry 84 467minus478
Breshears DD Cobb NS Rich PM Price KP and others(2005) Regional vegetation die-off in response to global-change-type drought Proc Natl Acad Sci USA 102 15144minus15148
Camarero JJ Garciacutea-Ruiz JM Sanguumlesa-Barreda G GalvaacutenJD and others (2015a) Recent and intense dynamics in aformerly static Pyrenean treeline Arct Antarct Alp Res47 773minus783
Camarero JJ Gazol A Galvaacuten JD Sanguumlesa-Barreda GGutieacuterrez E (2015b) Disparate effects of global-changedrivers on mountain conifer forests warming-inducedgrowth enhancement in young trees vs CO2 fertilizationin old trees from wet sites Glob Change Biol 21 738minus749
Chen IC Hill JK Ohlemuumlller R Roy DB Thomas CD (2011)Rapid range shifts of species associated with high levelsof climate warming Science 333 1024minus1026
Cudliacuten P Sejaacutek J Pokornyacute J Albrechtovaacute J Bastian OMarek M (2013) Forest ecosystem services under climatechange and air pollution In Matyssek R Clarke N Cud-liacuten P Mikkelsen TN Tuovinen JP Wiesner G Paoletti E(eds) Climate change air pollution and global chal-lenges understanding and perspectives from forestresearch Developments in Environmental Science Vol13 Elsevier Oxford p 521minus546
Dalen L Hofgaard A (2005) Differential regional treelinedynamics in the Scandes mountains Arct Antarct AlpRes 37 284minus296
Dawes MA Philipson CD Fonti P Bebi P Haumlttenschwiler SHagedorn F Rixen C (2015) Soil warming and CO2
enrichment induce biomass shifts in alpine tree line veg-etation 2014 Glob Change Biol 21 2005minus2021
De Frenne P Rodriacuteguez-Saacutenchez F De Schrijver A CoomesDA Hermy M Vangansbeke P Verheyen K (2015) Lightaccelerates plant responses to warming Nat Plants 1 15110
Devi N Hagedorn F Moiseev P Bugmann H ShiyatovSG Mazepa V Rigling A (2008) Expanding forests andchanging growth forms of Siberian larch at the treeline ofthe Polar Urals during the 20th century Glob ChangeBiol 14 1581minus1591
Diaz-Varela R Colombo R Meroni M Calvo-Iglesias MSBuffoni A Tagliaferri A (2010) Spatio-temporal analysisof alpine ecotones A spatial explicit model targeting altitudinal vegetation shifts Ecol Model 221 621minus633
Dirnboumlck T Dullinger S Grabherr G (2003) A regionalimpact assessment of climate and land-use change onalpine vegetation J Biogeogr 30 401minus417
Dobrowski SZ Swanson AK Abatzoglou JT Holden ZASafford HD Schwartz MK Gavin DG (2015) Forest struc-ture and species traits mediate projected recruitmentdeclines in western US tree species Glob Ecol Biogeogr24 917minus927
Dullinger S Dirnboumlck S Grabherr G (2004) Modelling cli-mate change-driven treeline shifts relative effects oftemperature increase dispersal and invisibility J Ecol92 241minus252
Dullinger S Gattringer A Thuiller W Moser D and others
(2012) Extinction debt of high-mountain plants undertwenty-first-century climate change Nat Clim Chang 2 619minus622
Elkin C Gutierrez AG Leuzinger S Manusch C TemperliC Rasche L Bugmann H (2013) A 2degC warmer world isnot safe for ecosystem services in the European AlpsGlob Change Biol 19 1827minus1840
Ellenberg H (1988) Vegetation ecology of Central Europe4th edn Cambridge University Press Edinburgh
Em H (1986) Spruce southern distribution range Spruce for-est on Shara Mountain in Macedonia Contrib Macedon-ian Acad Sci Arts Skopje 5 11minus28
Fajardo A Piper FI Pfund L Koumlrner CH Hoch G (2012)Variation of mobile carbon reserves in trees at the alpinetreeline ecotone is under environmental control NewPhytol 195 794minus802
Fridley JD Grime JP Askew AP Moser B Stevens CJ (2011)Soil heterogeneity buffers community response to cli-mate change in species-rich grassland Glob Change Biol17 2002minus2011
Gartzia M Alados CL Peacuterez-Cabello F (2014) Assessment ofthe effects of biophysical and anthropogenic factors onwoody plant encroachment in dense and sparse moun-tain grasslands based on remote sensing data Prog PhysGeogr 38 201minus217
Gehrig-Fasel J Guisan A Zimmermann NE (2007) Tree lineshifts in the Swiss Alps climate change or land abandon-ment J Veg Sci 18 571minus582
Gikov A Dimitrov P Zhiyanski M (2016) Land cover dynam-ics on the northern slope of the Troyan Passage during 30years period Probl Geogr 1-2 78minus92
Gonzalez P Neilson RP Lenihan JM Drapek RJ (2010)Global patterns in the vulnerability of ecosystems to veg-etation shifts due to climate change Glob Ecol Biogeogr19 755minus768
Gottfried M Pauli H Futschik A Akhalkatsi M and others(2012) Continent-wide response of mountain vegetationto climate change Nat Clim Chang 2 111minus115
Grace J Berninger F Nagy L (2002) Impacts of climatechange on the tree line Ann Bot 90 537minus544
Grunewald K Scheithauer J (2011) Landscape developmentand climate change in Southwest Bulgaria (Pirin Mts)Springer Heidelberg
Grytnes JA Kapfer J Jurasinski G Birks HH and others(2014) Identifying the driving factors behind observedelevational range shifts on European mountains GlobEcol Biogeogr 23 876minus884
Hagedorn F Shiyatov FG Mazepa VS Devi NM and others(2014) Treeline advances along the Urals mountainrange mdash driven by improved winter conditions GlobChange Biol 20 3530minus3543
Hansen AJ Neilson RP Dale VH Flather CH and others(2001) Global change in forests responses of speciescommunities and biomes BioScience 51 765minus779
Harsch MA Bader MY (2011) Treeline formminusa potential keyto understanding treeline dynamics Glob Ecol Biogeogr20 582minus596
Harsch MA Hulme PE McGlone MS Duncan RP (2009) Aretreelines advancing A global meta-analysis of treelineresponse to climate warning Ecol Lett 12 1040minus1049
Haylock MR Hofstra N Klein Tank AMG Klok EJ JonesPD New M (2008) A European daily high-resolutiongridded dataset of surface temperature and precipita-tion J Geophys Res Atmos 113 D20119
Hlaacutesny T Barcza Z Fabrika M Balaacutezs B and others (2011)
148
Cudliacuten et al Drivers of treeline shift
Climate change impacts on growth and carbon balanceof forests in Central Europe Clim Res 47 219minus236
Hlaacutesny T Trombik J Dobor L Barcza Z Barka I (2016)Future climate of the Carpathians climate change hot-spots and implications for ecosystems Reg EnvironChange 16 1495minus1506
Hofgaard A Dalen L Hytteborn H (2009) Tree recruitmentabove the treeline and potential for climate driven tree-line change J Veg Sci 20 1133minus1144
Holtmeier FK (2009) Mountain timberlines Ecology patchi-ness and dynamics 2nd edn Advances in GlobalChange Research 36 Springer Science amp ScienceMediaBV Dordrecht
Holtmeier FK Broll G (2005) Sensitivity and response ofnorthern hemisphere altitudinal and polar treelines toenvironmental change at landscape and local scalesGlob Ecol Biogeogr 14 395minus410
Honnay O Verheyen K Butaye J Jacquemyn H Bossuyt BHermy M (2002) Possible effects of habitat fragmentationand climate change on the range of forest plant speciesEcol Lett 5 525minus530
Huber R Rigling A Bebi P Brand F and others (2013) Sus-tainable land use in mountain regions under globalchange synthesis across scales and disciplines Ecol Soc18 36
Ibaacutentildeez I Clark JS Dietze MC (2009) Estimating colonizationpotential of migrant tree species Glob Change Biol 15 1173minus1188
IUCNSSC (International Union for Conservation of NatureSpecies Survival Commission) (2013) Guidelines for re -introductions and other conservation translocations Version 10 IUCNSpecies Survival Commission Gland
Jankovskyacute L Cudliacuten P gtermaacutek P Moravec I (2004) The pre-diction of development of secondary Norway sprucestands under the impact of climatic change in the Dra-hany highlands (The Czech Republic) Ekologia (Bratisl)23(Suppl 2) 101minus112
Jeniacutek J Lokvenc T (1962) Die alpine Waldgrenze im Krko -noše Gebirge Rozpravy SAV Rada matematickyacutech apriacuterodniacutech ved 72 gteskoslovenskaacute akademie ved Praha
Jobbaacutegy EG Jackson RB (2000) Global controls of forest lineelevation in the northern and southern hemispheresGlob Ecol Biogeogr 9 253minus268
Jonaacutešovaacute M Vaacutevrovaacute E Cudliacuten P (2010) Western Carpa -thian mountain spruce forest after a windthrow naturalregeneration in cleared and uncleared areas For EcolManag 259 1127minus1134
Kittel TGF Steffen WL Chapin FS (2000) Global andregional modelling of Arcticminusboreal vegetation distribu-tion and its sensitivity to altered forcing Glob ChangeBiol 6 1minus18
Klopcic M Jerina K Bonltina A (2010) Long-term changes ofstructure and tree species composition in Dinaric un -even- aged forests Are red deer an important factor EurJ For Res 129 277minus288
Klopcic M Simonltilt T Bonltina A (2015) Comparison of re -generation and recruitment of shade-tolerant and light-demanding tree species in mixed uneven-aged forests ex -periences from the Dinaric region Forestry 88 552minus563
Koumlrner C (2003) Alpine plant life functional plant ecology ofhigh mountain ecosystems with 47 tables Springer Sci-ence amp Business Media BV Dordrecht
Koumlrner C (2012) Alpine treelines Functional ecology of theglobal high elevation tree limits Springer Basel
Koumlrner C Paulsen J (2004) A world-wide study of high alti-
tude treeline temperatures J Biogeogr 31 713minus732Kulakowski D Bebi P Rixen C (2011) The interacting effects
of land use change climate change and suppression ofnatural disturbances on landscape forest structure in theSwiss Alps Oikos 120 216minus225
Kullman L (1988) Holocene history of the forestminusalpine tun-dra ecotone in the Scandes Mountains (central Sweden)New Phytol 108 101minus110
Kullman L (1999) Early holocene tree growth at a high elevation site in the northernmost Scandes of Sweden(Lapland) a palaeobiogeographical case study based onmega fossil evidence Geogr Ann 81 63minus74
Kullman L (2007) Tree line population monitoring of Pinussylvestris in the Swedish Scandes 1973minus2005 implica-tions for tree line theory and climate change ecologyJ Ecol 95 41minus52
Kullman L Oumlberg L (2009) Post-Little Ice Age treeline riseand climate warming in the Swedish Scandes a land-scape ecological perspective J Ecol 97 415minus429
Kyriazopoulos AP Skre O Sarkki S Wielgolaski FE Abra-ham EM Ficko A (2017) Humanminusenvironment dynamicsin European treeline ecosystems a synthesis based onthe DPSIR framework Clim Res 73 17ndash29
Lenoir J Svenning JC (2015) Climate-related range shifts mdasha global multidimensional synthesis and new researchdirections Ecography 38 15minus38
Lenoir J Geacutegout JC Marquet PA de Ruffray P Brisse H(2008) A significant upward shift in plant species opti-mum elevation during the 20th century Science 320 1768minus1771
Lindner M Fitzgerald JB Zimmermann NE Reyer C andothers (2014) Climate change and European forests What do we know what are the uncertainties and whatare the implications for forest management J EnvironManag 146 69minus83
Lloyd AH Rupp TS Fastie CL Srafield AM (2002) Patternsand dynamic of treeline advance on the Seward Pen -isula Alaska J Geophys Res 107 8161
Maacuteliš F Kopeckyacute M Petriacutek P Vladovic J Merganic J VidaT (2016) Life-stage not climate change explains ob -served tree range shifts Glob Change Biol 22 1904minus1914
Mathisen IE Mikheeva A Tutubalina OV Aune S and Hof-gaard A (2014) Fifty years of tree line change in theKhibiny Mountains Russia advantages of combinedremote sensing and dendroecological approaches ApplVeg Sci 17 6ndash16
Millar CI Stephenson NL (2015) Temperate forest healthin an era of emerging megadisturbance Science 349 823minus826
Mina M Bugmann H Klopcic M Cailleret M (2017) Accu-rate modeling of harvesting is key for projecting futureforest dynamics a case study in the Slovenian moun-tains Reg Environ Change 17 49minus64
Moen J Aune K Edenius L Angerbjoumlrn A (2004) Potentialeffects of climate change on treeline position in theSwedish mountains Ecol Soc 9 16 wwwecologyandsocietyorgvol9iss1art16
Palombo C Chirici G Marchetti M Tognetti R (2013) Is landabandonment affecting forest dynamics at high elevationin Mediterranean mountains more than climate changePlant Biosyst 147 1minus11
Palombo C Marchetti M Tognetti R (2014) Mountain vege-tation at risk current perspectives and research reedsPlant Biosyst 148 35minus41
Pauli H Gottfried M Dullinger S Abdaladze O Akhalkatsi
149
Clim Res 73 135ndash150 2017
M Alonso JLB Grabherr G (2012) Recent plant diversitychanges on Europersquos mountain summits Science 336 353minus355
Payette S Fortin MJ Gamachet I (2001) The subarctic forestminustundra the structure of a biome in a changing climate BioScience 51 709minus718
Pearson GR Dawson TP (2003) Predicting the impacts of cli-mate change on the distribution of species Are bio -climate envelope models useful Glob Ecol Biogeogr 12 361minus371
Pedrotti F (2013) Plant and vegetation mapping GeobotanyStudies Springer-Verlag Berlin
Pentildeuelas J Boada M (2003) A global change-induced biomeshift in the Montseny mountains (NE Spain) GlobChange Biol 9 131minus140
Piovesan G Biondi F Filippo AD Alessandrini A MaugeriM (2008) Drought driven growth reduction in old beech(Fagus sylvatica L) forests of the central ApenninesItaly Glob Change Biol 14 1265minus1281
Rabasa SG Granda E Benavides R Kunstler G and others(2013) Disparity in elevational shifts of European trees inresponse to recent climate warming Glob Change Biol19 2490minus2499
Raev I Zhelev P Grozeva M Markov I and others (2011)Programme of measures for adaptation of forests in Bul-garia and mitigate the negative impact of climate changeon them Project FUTURE forest helping Europe tackleclimate change INTERREG IVC ERDF Sofia
Sarkki S Ficko A Grunewald K Nijnik M (2016) Benefitsfrom and threats to European treeline ecosystem serv-ices an exploratory study of stakeholders and gover-nance Reg Environ Change 16 2019minus2032
Savage J Vellend M (2015) Elevational shifts biotic homo -genization and time lags in vegetation change during 40years of climate warming Ecography 38 546minus555
Schwoumlrer C Henne PD Tinner W (2014) A model-data com-parison of Holocene timberline changes in the SwissAlps reveals past and future drivers of mountain forestdynamics Glob Change Biol 20 1512minus1526
Šmilauer P Lepš J (2014) Multivariate analysis of ecologicaldata using Canoco 5 Cambridge University Press Cam-bridge
Ste-Marie C Nelson EA Dabros A Bonneau ME (2011)Assisted migration introduction to a multifaceted con-cept For Chron 87 724minus730
Stevens JT Stafford HD Harrison S Latimer AM (2015) For-est disturbance accelerates thermophilization of under-story plant communities J Ecol 103 1253minus1263
Strid A Andonoski A Andonovski V (2003) Alpine biodiver-sity in Europe Ecological Studies 167 Springer-VerlagBerlin
Tattoni C Ciolli M Ferretti F Cantiani MG (2010) Monitor-ing spatial and temporal pattern of Paneveggio forest(Northern Italy) from 1859 to 2006 iForest Biogeosci For3 72minus80
Ter Braak CJF Šmilauer P (2012) Canoco reference manualand userrsquos guide software for ordination (version 50)Microcomputer Power Ithaca NY
Theurillat JP Guisan A (2001) Potential impact of climatechange on vegetation in the European Alps a reviewClim Change 50 77minus109
Toslashmmervik H Wielgolaski FE Neuvonen S Solberg BHoslashgda KA (2005) Biomass and production on a land-scape level in the mountain birch forests In WielgolaskiFE (ed) Plant ecology herbivory and human impact in
Nordic mountain birch forests Ecos Studies 180Springer Berlin p 53minus70
Treml V Banaš M (2000) Alpine timberline in the HighSudetes Acta Univ Carol Geogr 15 83minus99
Treml V Chuman T (2015) Ecotonal dynamics of the altitu-dinal forest limit are affected by terrain and vegetationstructure variables an example from the Sudetes Moun-tains in Central Europe Arct Antarct Alp Res 47 133minus146
Treml V Migo P (2015) controlling factors limiting timber-line position and shifts in the Sudetes a review GeogrPol 88 55minus70
Urban MC Tewksbury JJ Sheldon KS (2012) On a collisioncourse competition and dispersal differences create no-analogue communities and cause extinctions during cli-mate change Proc R Soc Lond B Biol Sci 279 2072minus2080
Vacek S Matejka K (2010) State and development of phyto-cenoses on research plots in the Krkonoše Mts foreststands J For Sci 56 505minus517
Van Bogaert R Haneca K Hoogesteger J Jonasson C deDapper M Callaghan TV (2011) A century of tree linechanges in sub-Arctic Sweden shows local and regionalvariability and only a minor influence of 20th century climate warming J Biogeogr 38 907minus921
Van Gils H Batsukh O Rossiter D Munthali W Liberato -scioli E (2008) Forecasting the pattern and pace of Fagusforest expansion in Majella National Park Italy ApplVeg Sci 11 539minus546
Velchev V (1997) Types of vegetation In Yordanova MDonchev D (eds) Geography of Bulgaria BAN Sofiap 269minus283 (in Bulgarian)
Vladovic J Merganic J Maacuteliš F Križovaacute E and others (2014)The response of forest vegetation diversity to changes inedaphic-climatic conditions in Slovakia Technickaacute uni-verzita vo Zvolene Zvolene (in Slovak)
Vrška T Adam D Hort L Kolaacuter T Janiacutek F (2009) Europeanbeech (Fagus sylvatica L) and silver fir (Abies alba Mill)rotation in the Carpathians mdash a developmental cycle or alinear trend induced by man For Ecol Manag 258 347minus356
Wielgolaski FE (ed) (2005) Plant ecology herbivory andhuman impact in Nordic mountain birch forests Eco -logical Studies 180 Springer Verlag Berlin
Wielgolaski FE Hofgaard A Holtmeier FK (2017) Sensitivityto environmental change of the treeline ecotone and itsassociated biodiversity in European mountains Clim Res73151ndash166
Wieser G Tausz M (2007) Trees at their upper limit Treelifelimitation at the alpine timberline Springer Dordrecht
Wieser G Holtmeier FK Smith WK (2014) Treelines in achanging global environment In Tausz M Grulke N(eds) Trees in a changing environment Plant Ecophys 9Springer Dordrecht p 221minus263
Zhiyanski M Kolev K Sokolovska M Hursthouse A (2008)Tree species effect on soils in Central Stara PlaninaMountains Nauka Gorata 4 65minus82
Zhiyanski M Gikov A Nedkov S Dimitrov P Naydenova L(2016) Mapping carbon storage using land coverlanduse data in the area of Beklemeto Central Balkan In Koulov B Zhelezov G (eds) Sustainable mountain re -gions challenges and perspectives in southeasternEurope Springer Basel p 53minus65
Zhu K Woodall CW Clark JS (2012) Failure to migrate lackof tree range expansion in response to climate changeGlob Change Biol 18 1042minus1052
150
Editorial responsibility Nils Chr Stenseth Oslo Norway
Submitted April 8 2016 Accepted March 8 2017Proofs received from author(s) July 7 2017
Cudliacuten et al Drivers of treeline shift
atory variable can explain separately without usingthe other variables and their conditional ef fectswhich depend on the variables already selected inthe model The statistical significance of the expla -natory variables was tested using a Monte Carlo permutation test and only predictors with p le 005were included in the subsequent RDA (Šmilauer ampLepš 2014)
To find the drivers of biodiversity loss in forestsmeadows and animal communities (dependent vari-ables) the explanatory variables (geographical posi-tion timberline elevation tree species forming thetreeline start of human influence start of the de -crease in human influence and temperature increasebetween the 2 periods 1961minus1990 and 2021minus2050)were tested again by RDA in Canoco 5 as mentionedabove (Figs 3 amp 4) The whole data set for this analy-sis mdash including additional information about tree-lines in the mountain areas of interest with a focuson treeline shift (its rate drivers limits and problemswith its assessment) mdash is presented in Tables S1 amp S2
Selected univariate graphs were constructed to visu-alize the data of relationships between the explana-tory and dependent variables of all analyses whichwere not seen distinctly from the results of the multi-variate statistics or linear regression (Fig 5)
3 RESULTS
31 Effect of environmental variables on treeline ecotone characteristics with a focus
on treeline shift
The variation in the dependent variables (timber-line elevation width of treeline ecotone tree speciesforming the treeline and treeline shift per year) wassignificantly affected only by latitude and size of thewhole mountain massif (Fig 3) The adjusted ex -plained variability by all explanatory variables was342 (F = 46 p = 001) The RDA diagram showsthat in the mountains located more to the north the
141
Human Annual Limits Referencesinfluence temperature (T ) to climate
in VZs in past and precipitation (P) change and recently differences induced species
between 1961minus1990 shiftand 2021minus2050
mean of CORDEX models
From 14th C to 1922 Average of 10 RCM_B1 More rocky and nutrient- Koumlrner (2003) mining forest change ΔT2050 = 18degC poor soils late frosts Fridley et al (2011)
to spruce monocultures ΔT2100 = 37degC steep slopes with avalanches Hlaacutesny et al (2011 2016)from 13th C to 1978 pastures ΔP2050 = +24 mm absence of mature trees
in upper parts of mountain ΔP2100 = minus35 mm
9minus11th C forest clear cutting ALADIN_A1B period 1961minus Different soil types Treml and Banaš (2000) grazing in dwarf pine VZ 2100 ΔT2050 = 13degC for Fs and Pa (cambisol Cudliacuten et al (2013)
since 18th C forest change ΔT2100 = 34degC versus podzol) only TSL of Pa Treml amp Chuman (2015) to spruce monocultures ΔP2050 = +25 mm could elevate on ranker Treml amp Migo (2015)
since 19th C artificially planted ΔP2100 = minus14 mm soils in Pm VZ
Grazed and browsed RegClimMPIHadley Sheep and reindeer Dalen amp Hofgaard (2005) by reindeer and livestock ΔT2050 = 12degC grazing and episodic Wielgolaski (2005)
(sheep and cattle) ΔT2100 = 22degC insect outbreaks (Epirrita) Hofgaard et al (2009) over thousands of years ΔP2050 = +23 mm Kullman Oumlberg (2009)
ΔP2100 = minus12 mm
Grazed and browsed RegClimMPIHadley Continued grazing regime Dalen amp Hofgaard (2005) by semi-domestic reindeer ΔT2050 = 16degC and frequent episodic Wielgolaski (2005)
for hundreds of years ΔT2100 = 29degC insect outbreaks (Epirrita) Toslashmmervik et al (2005) increased grazing from ΔP2050 = +18 mm Hofgaard et al (2009)
year 1960 ΔP2100 = minus10 mm Aune et al (2011) Mathisen et al (2014)
Clim Res 73 135ndash150 2017
treeline ecotone is narrower treeline shiftis smaller the timberline occurs at lowerelevations and the treeline is formed moreby broadleaved species Higher values oftreeline ecotone width and treeline shiftwere found in the mountains with a greatersize of the whole mountain massif
When every explanatory variable wastested separately values of the start of thedecrease of human influence in the moun-tains also had a significant effect on thevariation of the tree line ecotone data(explained variation = 206 pseudo-F =34 p = 0024) Some selected relationshipsof the explanatory and dependent vari-ables even those not showing significanteffects on the variation in the tree line eco-tone parameters are depicted in Fig 5
32 Effect of recent changes in climateand land use on the biodiversity
Biodiversity loss in forests meadows andanimal communities analysed by RDAwas explained by geo graphical positiontreeline species composition and tempera-ture increase between the 2 periods1961minus1990 and 2021minus2050 (Fig 4) Theadjusted ex plained variation by all ex pla -natory variables was 632 (F = 58 p =0001) The RDA results indicated that bio-diversity loss in forest communities in crea -sed with increasing latitude and longitudeas well as in the case where broad leavedspecies formed the treeline In contrast thehighest biodiversity loss in meadows wasfound in the mountains positioned more tothe south and with higher temperatureincrease The biodiversity loss in animalcommunities was negatively correlatedwith longitude and temperature increase
4 DISCUSSION
The comparison of several mountainareas situated across Europe shows a varia-tion in understanding of what constitutesthe treeline across countries The primaryissue is the different approach to the defini-tion of forest when applying a tree heightthreshold This threshold decreases from
142
Mou
nta
in u
nit
Nor
th
Eas
t M
oun
-T
imb
er-
Wid
th
Tre
e T
emp
erat
ure
T
reel
ine
Sta
rt
Sta
rt
lati
tud
elo
ng
itu
de
tain
lin
eof
spec
ies
incr
ease
sh
ift
of H
I of
HI
size
(m a
sl
)tr
eeli
ne
form
ing
b
etw
een
(m
yrminus
1 )(y
r)d
ecre
ase
(km
2 )ec
oton
eth
e 19
61minus
1990
(y
r)(m
)tr
eeli
ne
and
199
1minus20
15 (
degC)
Cen
tral
Pyr
enee
s (C
P)
42deg2
1rsquoN
0deg30
rsquoW55
000
2250
300
C0
81
92 (
1956
minus20
06)
1000
BC
1930
Eas
tern
Pyr
enee
s (E
P)
42deg2
8rsquoN
1deg30
rsquoE55
000
2400
300
C0
80
71 (
1956
minus20
06)
1000
1930
Ap
enn
ines
(A
P)
42deg0
5rsquoN
14deg0
5rsquoE
5000
018
0030
0B
14
1 (1
954minus
2007
)10
00 B
C19
50S
har
a M
ts (
SH
)41
deg47rsquo
N20
deg33rsquo
E16
0018
5035
0C
B0
71
(193
4minus20
10)
1300
1970
Pir
in (
PI)
41deg4
2rsquoN
23deg3
1rsquoE
2585
1900
400
C0
5n
o d
ata
680
1962
Cen
tral
Bal
kan
Mts
(C
B)
42deg4
7primeN
24deg3
6primeE
1160
016
0025
0C
B0
7n
o d
ata
1700
1980
Nor
ther
n D
inar
ic M
ts (
DI)
45deg3
6rsquoN
14deg2
8rsquoE
1000
1540
160
B1
3n
o d
ata
1400
1900
Cen
tral
Alp
s (C
A)
46deg4
6rsquoN
9deg52
rsquoE18
000
021
0055
0C
07
13
(195
4minus20
15)
1300
1880
Eas
tern
Alp
s (E
A)
46deg1
7rsquoN
11deg4
5rsquoE
180
000
1800
400
C0
91
3 (1
954minus
2015
)11
5019
50L
ow T
atra
Mts
(LT
)48
deg55rsquo
N19
deg31rsquo
E12
4014
0033
0C
09
057
(19
75minus
2016
)14
0019
22G
ian
t M
ts (
GM
)50
deg42rsquo
N15
deg38rsquo
E63
012
5025
0C
07
043
(19
36minus
2005
)10
0019
63H
ard
ang
ervi
dd
a (H
G)
60deg0
3rsquoN
7deg47
rsquoE90
0090
015
0C
B0
50
8 (1
915minus
2007
)80
019
70D
ovre
(D
O)
62deg2
1rsquoN
8deg52
rsquoE60
0010
0040
0B
07
08
(191
5minus20
07)
1000
no
dec
reas
eN
orth
ern
Sw
edis
h L
apla
nd
(N
S)
68deg1
2rsquoN
18deg3
3rsquoE
4000
700
300
B1
10
6 (1
958minus
2008
)10
00n
o d
ecre
ase
Inn
er F
inn
mar
kn
orth
ern
mos
t 69
deg10rsquo
N25
deg12rsquo
E60
000
290
90B
10
06
(195
8minus20
08)
1600
no
dec
reas
eF
inn
ish
Lap
lan
d (
FL
)
Tab
le 2
Fac
tors
infl
uen
cin
g t
reel
ine
ecot
one
char
acte
rist
ics
in s
elec
ted
Eu
rop
ean
mou
nta
ins
Th
e m
oun
tain
siz
e is
th
e si
ze o
f th
e w
hol
e m
oun
tain
mas
sive
fu
nct
ion
ing
as a
mez
osca
le c
lim
ate
un
it n
ot o
nly
the
stu
die
d m
oun
tain
are
a W
idth
of t
he
tree
lin
e ec
oton
e w
as d
eter
min
ed b
y su
btr
acti
ng
the
valu
e of
the
mea
n ti
mb
erli
ne
elev
atio
n
from
th
e va
lue
of t
he
up
per
tre
e sp
ecie
s li
mit
C c
onif
ers
B b
road
leav
ed t
rees
HI
hu
man
infl
uen
ce
Cudliacuten et al Drivers of treeline shift
5 m (Jeniacutek amp Lokvenc 1962) to 3 m(Koumlrner 2012) in Northern andCentral Europe to only 2 m (Holt-meier 2009) in Central and South-ern Europe where even shrub -by stands (eg Pinus mugo) areconsi dered as a forest in somecountries eg in Spain Italy andMa ce donia (Batllori et al 2009)An other problem is a different de -signation of forest stands belowthe treeline lsquoupper montane fo -restsrsquo in Central Europe and lsquosub-alpine forestsrsquo in Southern Europe(Ellenberg 1988) Further differen -ces are related to the tradition ofdifferent branches of science (ega more traditional lsquogeobotanicalrsquoapproach in Central Europe ver-sus a more experimental approachin Western Europe) especially inthe rate of applying new pro -gressive methods (eg climaticmodelling or molecular biologicalmethods)
Our comparison of selectedregions (Tables 1 amp 2 Figs 1 amp 2)showed that southern mountainscompared to those located morecentrally and in the north had (1) alonger and more profound exploi -tation by humans in the past (2)greater differences in climatic para -meters (temperature length ofgrowing period) between the 2periods 1961minus1990 and 1991minus2015(Table 3) and (3) more dramaticclimate change scenarios espe-cially concerning temperature in -creases Longitude also had someinfluence on the start and intensityof human influence and on tree-line elevation and rate of treelineshift (Table 2 Fig 5) The occur-rence of broadleaf tree species inthe treeline ecotone in the north-ern (and to some extent the south-ern) countries distinguishes thesefrom the Central European coun-tries where conifers prevail It isinteresting that grazing an im -portant factor shaping treelineecosystems (Dirnboumlck et al 2003
143
Length Beginning Mean Mean of growing of growing annual annual
period season temperature precipitation (d)a (d)b (degC) ()
Central Pyrenees 135 minus89 08 minus107Eastern Pyrenees 117 minus91 08 minus92Apennines 284 minus232 14 57Shara Mts 16 01 07 minus100Pirin 26 minus14 05 27Central Balkan Mts 169 minus136 07 76Northern Dinaric Mts 160 minus95 13 minus56Central Alps 147 minus86 07 62Eastern Alps 125 minus96 09 minus33Low Tatra Mts 24 minus34 09 minus91Giant Mts minus10 13 07 55Hardangervidda Blefjell 63 34 05 56Dovre 138 32 07 64Northern Swedish Lapland 81 05 11 minus27Inner Finnmarknorthern- 17 08 10 77most Finnish Lapland
aPositive numbers indicate a prolongation bNegative numbers indicate an earlier beginning
Table 3 Differences of selected climatic parameters between the 2 study periods (1961minus1990 and 1991minus2015) in selected European mountains
Tein Drou Gefa Nadi Laus Abli Bili Huli
0 1 2 3
Central Pyrenees (CP)
Eastern Pyrenees (EP)
Apennines (AP)
Shara Mts (SH)
Pirin (PI)
Central Balkan Mts (CB)
Northern Dinaric Mts (DI)
Central Alps (CA)
Eastern Alps (EA)
Low Tatra Mts (LT)
Giant Mts (GM)
Hardangervidda (HG)
Dovre (DO)
Northern Swedish Lapland (NS)
Inner FinnmarknorthernmostFinnish Lapland (FL)
Fig 2 Rates of treeline drivers (temperature increase [Tein] land use change[Laus]) and treeline shift limits (drought [Drou] geomorphological factors [Gefa]natural disturbances [Nadi] other abiotic biotic and human factors) in selectedEuropean mountains Abli Bili and Huli abiotic biotic and human treeline shiftlimits respectively) Rate of treeline driver influence mdash 0 no influence 1 weak
influence 2 middle influence 3 strong influence
Clim Res 73 135ndash150 2017
Gehrig-Fasel et al 2007) limits treeline shift in halfof the countries of interest regardless of latitude lon-gitude or recent political developments (Shara PirinCentral Balkans Alps Scandes Fig 2)
Despite the stated differences between studieslooking at climate change impacts on the treelineit is clear that current and future changes in temp -erature will seriously affect treeline ecotones in allmountain ranges of Europe Winter is a period whentreeline stands and individual trees have to survivesevere life-limiting conditions (Wieser amp Tausz2007) Therefore winter warming might increase thechance of young trees surviving and passing the mostcritical period from seedling to sapling and further tothe mature stage (Koumlrner 2003) There is alreadymuch evidence worldwide that winter warming isone of the significant drivers of treeline advance(Harsch et al 2009) Although it is expected that treegrowth at the upper distributional margins (and inthe northern European countries) will increase due toongoing climate change (Pentildeuelas amp Boada 2003Chen et al 2011 Hlaacutesny et al 2011 Lindner et al2014) extremes in temperature (eg black froststemperature reversals in the spring) or winter precip-itation (eg lack of snow drought) might also limit
this process in the future in some regions and localsituations (Holtmeier amp Broll 2005 Hagedorn et al2014) An earlier start to the growing period (espe-cially in the Apennines Central Balkan and SpanishPyrenees see Table 2) can play a negative role in theresistance of trees and seedlings to early springfrosts On the other hand at lower distributional mar-gins (and in Southern European countries) it isexpected that tree growth will decrease or forestswill experience some level of dieback due to drought(Breshears et al 2005 Piovesan et al 2008 Allen etal 2010 Huber et al 2013) A serious implication isthe positive effect that warming can have on root rotfungi and populations of bark beetles resulting inlarge-scale disturbances to temperate and also high-altitude forests (Jankovskyacute et al 2004 Elkin et al2013 Millar amp Stephenson 2015)
Although recent upward advances of treeline eco-tones are widespread in mountain regions (Harsch et
144
Fig 3 Redundancy analysis diagram with variation in tim-berline elevation width of treeline ecotone the tree speciesforming the treeline and treeline shift used as dependentvariables explained by explanatory variables (latitudemountain size) Mountain units (blue) are projected as thecentres of abbreviations (see Table 2) Explanatory variablesaccount for 436 of the total variation in the dependentdata The first canonical axis explained 466 of variationthe second axis explained 30 of variation Dependentvariables (black) are Ecot altitudinal width of treeline eco-tone TLShift treeline shift TrspC (TrspB) treeline formedby conifers (broadleaved trees) Timb timberline elevationExplanatory variables (red) are N north latitude Mtsize
size of the mountain massif
Fig 4 Redundancy analysis diagram with variation in bio-diversity loss in forests meadows and animal communitiesused as dependent variables explained by explanatory vari-ables (geographical position tree species forming the tree-line and temperature increase between the 2 study periods[1961minus1990 and 2021minus2050]) Mountain units (green) areprojected as centres of abbreviations (see Table 2) Explana-tory variables account for 763 of total variation in the de-pendent data The first canonical axis explained 484 ofvariation the second axis explained 183 of variation De-pendent variables (black) are BdfoBdmeBdan biodiversityloss in forestsmeadowsanimal communities Explanatoryvariables (red) are N north latitude E east longitude TrspC(TrspB) treeline formed by conifers (broadleaved trees)Temp temperature increase between the 2 study periods
Cudliacuten et al Drivers of treeline shift
al 2009) only a few published studies have includedquantitative data about treeline shifts (Devi et al2008 Kullman amp Oumlberg 2009 Diaz-Varela et al 2010Van Bogaert et al 2011) Additionally our know -ledge of the spatial patterns in treeline ecotone shiftsat the landscape scale is still surprisingly poor exceptfor some studies from subarctic areas the UralMountains and the Alps (Lloyd et al 2002 Diaz-Varela et al 2010 Hagedorn et al 2014) In the 15studied mountain units the values of treeline shiftranged from 043 to 19 m yrminus1 showing a rather dis-tinct spatial pattern of the dynamics within Europeanmountains The observed treeline shift had a signifi-cant positive relationship only with northern latitudeand a weak positive relationship with mountain-range size (ie the size of the whole mountain massifFig 3) Nevertheless this illustrates the importanceof the mass elevation effect (Koumlrner 2012) and par-tially ex plains why treelines could be at different alti-tudes at the same latitude Small negative regres-sions with east longitude and timberline elevationare shown in Fig 5 There was no apparent depend-ence of the treeline shift rate on climatic parameterchanges between the periods 1961minus1990 and1991minus2015
A whole forest vegetation zone shift is a complexprocess as not only the life strategies of an individualtree need to be considered but plant animal andmicroorganism species and their interrelationshipsas well as the relationships to their specific microhab-
itats (including soil conditions) are also involved(Urban et al 2012) Across all regions we identifiedseveral important obstacles to treeline shifts oftenrelated to site properties such as significant rocki-ness having shallow or low-nutrient soils extremerelief causing disturbances by avalanches snow glid-ing or wind damage (data from the Pyrenees Apen-nines Shara Northern Dinaric Low Tatra and GiantMountains see Fig 2) Soil heterogeneity certainlyhas an important role in plant responses to climatechange and could also maintain the resilience of thecommunity assemblages (Fridley et al 2011) An -other group of obstacles includes extreme climaticparameters especially in winter (reported eg fromthe Pyrenees Apennines Pirin Central BalkanNorthern Dinaric and Giant Mountains Fig 2) Theircombination can result in edaphic andor climaticunsuitability of habitats where species could poten-tially migrate Climatically and edaphically suitablesites will likely decline over the next century partic-ularly in mountain landscapes (Bell et al 2014)There fore trees in the treeline ecotone will colonizepreviously forested habitats The colonization suc-cess of individual forest communities is affected bydifferences in species dispersal and recruitmentbehaviour (Dullin ger et al 2004 Jonaacutešovaacute et al2010) For these reasons colonization of new foresthabitats by the next tree generation may not be suc-cessful and may result in loss of species diversity(Honnay et al 2002 Ibaacutentildeez et al 2009 Dobrowski et
145
Fig 5 Univariate graphs of the relationships of dependent variables viz timberline (Timb) and treeline shift per year (Shift) and their explanatory variables Nola north latitude Ealo east longitude DD decimal degrees
Clim Res 73 135ndash150 2017
al 2015) as reported from Macedonia Slovenia theCzech Republic Slovakia and Norway (Fig 4)
Another limiting factor is the existence of strongadaptation mechanisms of the dominant tree specieswhich allow them to survive in their current distribu-tion areas Species migration is likely to be slow dueto the limited quantity of climatically and edaphicallysuitable sites and the slow velocity of seed dispersaland tree regeneration and will be hampered evenmore by fragmentation of high mountain landscapescaused by human activities including brush invasionin abandoned meadows (Gartzia et al 2014) A verti-cal shift in a vegetation zone involves not only singlespecies of plants animals and microorganisms butalso their interrelationships and links to soil condi-tions The speed at which climatic conditions changewill be different from changes in the soil conditionsdue to the slowness of soil-forming processes Soiltypes and conditions (cambisol versus podzol) arecrucial obstacles for the shift from a beech forest zoneinto the spruce forest zone in the Czech Republic(Vacek amp Mat jka 2010) According to other sourcesthe dominant tree species can influence soil for -mation processes (especially humus forms) Underfavourable climatic and orographic conditions beechis able to change its soil conditions over decades tocenturies (J Macku unpubl data)
Significant changes in biodiversity must be expec -ted in all of the mountain areas of interest We foundthat biodiversity declined particularly in the southernregions of Europe where the timberline is situated athigher elevations and human impact is mostlylonger-lasting (Fig 4) Similarly Pauli et al (2012)reported an increase in species richness of mountaingrasslands across Europe except for Mediterraneanregions and assigned this different response of thesouthern regions to decreased water availability Onthe other hand ecosystems which exhibited bio -diversity increases were not only enriched by mig -rant species but the assemblages underwent thermo -philization ie cold-adapted species declined andmore warm-adapted species spread (Gottfried et al2012) However not all species are able to track cli-mate changes It is expected that around 40 ofhabitats will become climatically unsuitable for manymountain species particularly endemic ones in -creasing their extinction probability during this cen-tury (Dullinger et al 2012) The abandonment of tra-ditional land use forms is another source of this losswhich may be a more important driver than tempera-ture increase (Fig 4) Serious biodiversity losses inmountain meadows were reported from Spain ItalyMacedonia Bulgaria Slovenia and Slovakia There-
fore controlled grazing must occur in order to main-tain alpine grasslands (Dirnboumlck et al 2003) Unfor-tunately grazing is still active only in smaller parts ofEuropean mountains (eg in the Central Alps and theCentral Balkan Mountains but pasturing can alsonegatively affect vegetation diversity) and some-times does not serve to maintain alpine grasslandThe same is true for grassland management whichcan help to maintain mountain species and decreasehabitat loss A serious biodiversity loss in mountainmeadows was reported from the Pyrenees the Apen-nines and the Shara Pirin Central Balkan NorthernDinaric and Low Tatra Mountains (Fig 4)
In forests species responses to climate change maybe equivocal some recent studies indicated contra-dictory shifts in species distributions (eg Lenoir etal 2008 Zhu et al 2012 Rabasa et al 2013) This dis-parity is widely discussed and assigned to the greatvariety of non-climatic factors or even tree ontogeny(Grytnes et al 2014 Lenoir amp Svenning 2015 Maacutelišet al 2016) Disturbances leading to tree mortalitymay also play an important role (Cudliacuten et al 2013)changes in tree canopy cover modify light availabil-ity and microclimate and can induce the thermo -philization of forest vegetation (De Frenne et al2015 Stevens et al 2015) The loss of this microcli-mate buffering effect of forests may induce a biotichomogenization of forests (Savage amp Vellend 2015)or the creation of novel non-analogical communitiessuch as oakminuspine forests (Urban et al 2012) whichmay be a new threat to forest biodiversity
The observed changes in treeline forest ecosystemsare often related to changes in land-use intensity(Theurillat amp Guisan 2001 Alados et al 2014) Ac -cording to climate change predictions and the recentand future exploitation intensity of European moun-tains trees and forest communities will shift upwarddue to land use change or climate change or bothReduced land-use intensity certainly will interactwith climate change by facilitating or inhibiting spe-cies occurrence and will accelerate forest expansionabove the present treeline particularly to previouslyforested habitats (Theurillat amp Guisan 2001) Thesimultaneous action of both main treeline shift driv-ers viz temperature increase and decrease in landuse intensity was recorded from all of our studiedmountain areas The extensive differences in timber-line and treeline elevations in almost all studiedmountains (Table 1) indicate the anthropo-zoogenictreeline type (according to Ellenberg 1998) In mostmountains (eg in the Apennines Shara MountainsCentral Balkans and Alps) land use is the prevailingfactor influencing vegetation drift (Fig 5) Previous
146
Cudliacuten et al Drivers of treeline shift
research showed that the upward shift of the treelinein the Swiss Alps was predominantly attributable toland abandonment and only in some situations to cli-mate change (Gehrig-Fasel et al 2007) In the Apen-nines in the last few decades tree establishment hasbeen mainly controlled by land use while treegrowth has been controlled by climate pointing to aminor role played by climate in shaping the currenttreeline (Palombo et al 2013)
The impact of climate change and the connectedland use change on biodiversity and ecosystem serv-ices provision in several European countries is sum-marized by Wielgolaski et al (2017) and Kyriazopou-los et al (2017) (both this Special) and Sarkki et al(2016) One of the possible adaptive management op-tions in response to climate change assisted migra-tion as human-assisted movement of species hasbeen frequently debated in the last few years (Ste-Marie et al 2011) It is possible to apply it as a type ofassisted colonization ie intentional movement andrelease of an organism outside its indigenous range toavoid extinction of populations of the focal species(eg Macedonian pine species) or as an ecologicalre placement ie the intentional movement and re-lease of an organism outside its indigenous range toperform a specific ecological function (eg planting ofPinus mugo in the Alps IUCNSSC 2013)
5 CONCLUSIONS
The analysis of 11 mountain areas across Europeshowed that with increasing latitude the treelineand altitudinal width of the treeline ecotone signifi-cantly decreases as do the significance of climaticand soil parameters as barriers against tree speciesshift Although temperature is the overwhelmingcontrolling factor of tree growth and establishment intemperate and boreal treeline ecotones late-sea-sonal drought might also play a driving role in Medi-terranean treeline ecotones Longitude was lessinfluential mostly affecting climate and land usechange effects on increased biodiversity loss as wellas the size of the area that forms one mesoclimateunit affecting altitudinal ecotone width The biggestpart of the commonly observed remaining variabilityin mountain vegetation near the treeline in Europeseems to be caused by geomorphological geologicalpedological and microclimatic variability in combina-tion with different land use history and the presentsocio-economic relations The observed differencesin climatic parameters between the mountain areasof interest in comparison with the reference period
1960minus1990 (09degC) have not explained the relativelyhigh differences in the rate of treeline shift per year(149 m) The predicted variability in temperatureincrease due to global warming (12minus26degC in 2050and 22minus53degC in 2100) could lead to a much biggerdifferentiation in treeline ecotone biodiversity andecosystem processes between southern and northernEuropean mountains in the future Therefore thesedifferences must be taken into account by scientistsand EU policy makers when formulating efficientadaptive forest management strategies for treelineecosystems at the European level (eg assistedmigration of adapted genotypes)
Acknowledgements This paper is based firstly upon workfrom the COST Action ES 1203 SENSFOR supported byCOST (European Cooperation in Science and Technologywwwcosteu) This international work was further sup-ported by projects granted by the Ministry of EducationYouth and Sports of the Czech Republic grant NPU ILO1415 and LD 14039 by the agency APVV SR under pro-jects APVV-14-0086 and APVV-15-0270 We acknowledgethe E-OBS dataset from the EU-FP6 project ENSEMBLES(httpensembles-eumetofficecom) and the data providersin the ECAampD project (wwwecadeu)
LITERATURE CITED
Alados CL Errea P Gartzia M Saiz H Escoacutes J (2014) Posi-tive and negative feedbacks and free-scale pattern distri-bution in rural-population dynamics PLOS ONE 9 e114561
Allen CD Macalady AK Chenchouni H Bachelet D andothers (2010) A global overview of drought and heat-induced tree mortality reveals emerging climate changerisks for forests For Ecol Manag 259 660minus684
Amidžic L Bartula M Jaric S (2012) Syntaxonomy overviewof vegetation of Šar Planina in the Balkan PeninsulaVegetos 25 348minus360
Aune S Hofgaard A Soumlderstroumlm L (2011) Contrasting cli-mate and land use driven tree encroachment pattern ofsub-arctic tundra in Northern Norway and Kola Penin-sula Can J For Res 41 437minus449
Barbeito I Dawes M Rixen C Senn J Bebi P (2012) Factorsdriving mortality and growth at treeline a 30-year exper-iment of 92000 conifers Ecology 93 389minus401
Batllori E Camarero JJ Ninot JM Gutieacuterrez E (2009)Seedling recruitment survival and facilitation in alpinePinus uncinata treeline ecotones Implications andpotential responses to climate warming Glob Ecol Bio-geogr 18 460minus472
Bebi P Teich M Hagedorn F Zurbriggen N Brunner SHGreacutet-Regamey A (2012) Veraumlnderung von Wald undWaldleistungen in der Landschaft Davos im Zuge desKlimawandels Schweiz Z Forstwes 163 493minus501
Bell DM Bradford JB Lauenroth WK (2014) Mountain land-scapes offer few opportunities for high elevation treespecies migration Glob Change Biol 20 1441minus1451
Bergant K (2006) Climate change scenarios for MacedoniaUniversity of Nova Gorica Center for Atmospheric
147
Clim Res 73 135ndash150 2017
Research Nova GoricaBodin J Badeau V Bruno E Cluzeau C Moisselin JM
Walther GR Dupouey JL (2013) Shifts of forest speciesalong an elevational gradient in Southeast France cli-mate change or stand maturation J Veg Sci 24 269minus283
Boncina A (2011) History current status and future pros -pects of uneven-aged forest management in the Dinaricregion an overview Forestry 84 467minus478
Breshears DD Cobb NS Rich PM Price KP and others(2005) Regional vegetation die-off in response to global-change-type drought Proc Natl Acad Sci USA 102 15144minus15148
Camarero JJ Garciacutea-Ruiz JM Sanguumlesa-Barreda G GalvaacutenJD and others (2015a) Recent and intense dynamics in aformerly static Pyrenean treeline Arct Antarct Alp Res47 773minus783
Camarero JJ Gazol A Galvaacuten JD Sanguumlesa-Barreda GGutieacuterrez E (2015b) Disparate effects of global-changedrivers on mountain conifer forests warming-inducedgrowth enhancement in young trees vs CO2 fertilizationin old trees from wet sites Glob Change Biol 21 738minus749
Chen IC Hill JK Ohlemuumlller R Roy DB Thomas CD (2011)Rapid range shifts of species associated with high levelsof climate warming Science 333 1024minus1026
Cudliacuten P Sejaacutek J Pokornyacute J Albrechtovaacute J Bastian OMarek M (2013) Forest ecosystem services under climatechange and air pollution In Matyssek R Clarke N Cud-liacuten P Mikkelsen TN Tuovinen JP Wiesner G Paoletti E(eds) Climate change air pollution and global chal-lenges understanding and perspectives from forestresearch Developments in Environmental Science Vol13 Elsevier Oxford p 521minus546
Dalen L Hofgaard A (2005) Differential regional treelinedynamics in the Scandes mountains Arct Antarct AlpRes 37 284minus296
Dawes MA Philipson CD Fonti P Bebi P Haumlttenschwiler SHagedorn F Rixen C (2015) Soil warming and CO2
enrichment induce biomass shifts in alpine tree line veg-etation 2014 Glob Change Biol 21 2005minus2021
De Frenne P Rodriacuteguez-Saacutenchez F De Schrijver A CoomesDA Hermy M Vangansbeke P Verheyen K (2015) Lightaccelerates plant responses to warming Nat Plants 1 15110
Devi N Hagedorn F Moiseev P Bugmann H ShiyatovSG Mazepa V Rigling A (2008) Expanding forests andchanging growth forms of Siberian larch at the treeline ofthe Polar Urals during the 20th century Glob ChangeBiol 14 1581minus1591
Diaz-Varela R Colombo R Meroni M Calvo-Iglesias MSBuffoni A Tagliaferri A (2010) Spatio-temporal analysisof alpine ecotones A spatial explicit model targeting altitudinal vegetation shifts Ecol Model 221 621minus633
Dirnboumlck T Dullinger S Grabherr G (2003) A regionalimpact assessment of climate and land-use change onalpine vegetation J Biogeogr 30 401minus417
Dobrowski SZ Swanson AK Abatzoglou JT Holden ZASafford HD Schwartz MK Gavin DG (2015) Forest struc-ture and species traits mediate projected recruitmentdeclines in western US tree species Glob Ecol Biogeogr24 917minus927
Dullinger S Dirnboumlck S Grabherr G (2004) Modelling cli-mate change-driven treeline shifts relative effects oftemperature increase dispersal and invisibility J Ecol92 241minus252
Dullinger S Gattringer A Thuiller W Moser D and others
(2012) Extinction debt of high-mountain plants undertwenty-first-century climate change Nat Clim Chang 2 619minus622
Elkin C Gutierrez AG Leuzinger S Manusch C TemperliC Rasche L Bugmann H (2013) A 2degC warmer world isnot safe for ecosystem services in the European AlpsGlob Change Biol 19 1827minus1840
Ellenberg H (1988) Vegetation ecology of Central Europe4th edn Cambridge University Press Edinburgh
Em H (1986) Spruce southern distribution range Spruce for-est on Shara Mountain in Macedonia Contrib Macedon-ian Acad Sci Arts Skopje 5 11minus28
Fajardo A Piper FI Pfund L Koumlrner CH Hoch G (2012)Variation of mobile carbon reserves in trees at the alpinetreeline ecotone is under environmental control NewPhytol 195 794minus802
Fridley JD Grime JP Askew AP Moser B Stevens CJ (2011)Soil heterogeneity buffers community response to cli-mate change in species-rich grassland Glob Change Biol17 2002minus2011
Gartzia M Alados CL Peacuterez-Cabello F (2014) Assessment ofthe effects of biophysical and anthropogenic factors onwoody plant encroachment in dense and sparse moun-tain grasslands based on remote sensing data Prog PhysGeogr 38 201minus217
Gehrig-Fasel J Guisan A Zimmermann NE (2007) Tree lineshifts in the Swiss Alps climate change or land abandon-ment J Veg Sci 18 571minus582
Gikov A Dimitrov P Zhiyanski M (2016) Land cover dynam-ics on the northern slope of the Troyan Passage during 30years period Probl Geogr 1-2 78minus92
Gonzalez P Neilson RP Lenihan JM Drapek RJ (2010)Global patterns in the vulnerability of ecosystems to veg-etation shifts due to climate change Glob Ecol Biogeogr19 755minus768
Gottfried M Pauli H Futschik A Akhalkatsi M and others(2012) Continent-wide response of mountain vegetationto climate change Nat Clim Chang 2 111minus115
Grace J Berninger F Nagy L (2002) Impacts of climatechange on the tree line Ann Bot 90 537minus544
Grunewald K Scheithauer J (2011) Landscape developmentand climate change in Southwest Bulgaria (Pirin Mts)Springer Heidelberg
Grytnes JA Kapfer J Jurasinski G Birks HH and others(2014) Identifying the driving factors behind observedelevational range shifts on European mountains GlobEcol Biogeogr 23 876minus884
Hagedorn F Shiyatov FG Mazepa VS Devi NM and others(2014) Treeline advances along the Urals mountainrange mdash driven by improved winter conditions GlobChange Biol 20 3530minus3543
Hansen AJ Neilson RP Dale VH Flather CH and others(2001) Global change in forests responses of speciescommunities and biomes BioScience 51 765minus779
Harsch MA Bader MY (2011) Treeline formminusa potential keyto understanding treeline dynamics Glob Ecol Biogeogr20 582minus596
Harsch MA Hulme PE McGlone MS Duncan RP (2009) Aretreelines advancing A global meta-analysis of treelineresponse to climate warning Ecol Lett 12 1040minus1049
Haylock MR Hofstra N Klein Tank AMG Klok EJ JonesPD New M (2008) A European daily high-resolutiongridded dataset of surface temperature and precipita-tion J Geophys Res Atmos 113 D20119
Hlaacutesny T Barcza Z Fabrika M Balaacutezs B and others (2011)
148
Cudliacuten et al Drivers of treeline shift
Climate change impacts on growth and carbon balanceof forests in Central Europe Clim Res 47 219minus236
Hlaacutesny T Trombik J Dobor L Barcza Z Barka I (2016)Future climate of the Carpathians climate change hot-spots and implications for ecosystems Reg EnvironChange 16 1495minus1506
Hofgaard A Dalen L Hytteborn H (2009) Tree recruitmentabove the treeline and potential for climate driven tree-line change J Veg Sci 20 1133minus1144
Holtmeier FK (2009) Mountain timberlines Ecology patchi-ness and dynamics 2nd edn Advances in GlobalChange Research 36 Springer Science amp ScienceMediaBV Dordrecht
Holtmeier FK Broll G (2005) Sensitivity and response ofnorthern hemisphere altitudinal and polar treelines toenvironmental change at landscape and local scalesGlob Ecol Biogeogr 14 395minus410
Honnay O Verheyen K Butaye J Jacquemyn H Bossuyt BHermy M (2002) Possible effects of habitat fragmentationand climate change on the range of forest plant speciesEcol Lett 5 525minus530
Huber R Rigling A Bebi P Brand F and others (2013) Sus-tainable land use in mountain regions under globalchange synthesis across scales and disciplines Ecol Soc18 36
Ibaacutentildeez I Clark JS Dietze MC (2009) Estimating colonizationpotential of migrant tree species Glob Change Biol 15 1173minus1188
IUCNSSC (International Union for Conservation of NatureSpecies Survival Commission) (2013) Guidelines for re -introductions and other conservation translocations Version 10 IUCNSpecies Survival Commission Gland
Jankovskyacute L Cudliacuten P gtermaacutek P Moravec I (2004) The pre-diction of development of secondary Norway sprucestands under the impact of climatic change in the Dra-hany highlands (The Czech Republic) Ekologia (Bratisl)23(Suppl 2) 101minus112
Jeniacutek J Lokvenc T (1962) Die alpine Waldgrenze im Krko -noše Gebirge Rozpravy SAV Rada matematickyacutech apriacuterodniacutech ved 72 gteskoslovenskaacute akademie ved Praha
Jobbaacutegy EG Jackson RB (2000) Global controls of forest lineelevation in the northern and southern hemispheresGlob Ecol Biogeogr 9 253minus268
Jonaacutešovaacute M Vaacutevrovaacute E Cudliacuten P (2010) Western Carpa -thian mountain spruce forest after a windthrow naturalregeneration in cleared and uncleared areas For EcolManag 259 1127minus1134
Kittel TGF Steffen WL Chapin FS (2000) Global andregional modelling of Arcticminusboreal vegetation distribu-tion and its sensitivity to altered forcing Glob ChangeBiol 6 1minus18
Klopcic M Jerina K Bonltina A (2010) Long-term changes ofstructure and tree species composition in Dinaric un -even- aged forests Are red deer an important factor EurJ For Res 129 277minus288
Klopcic M Simonltilt T Bonltina A (2015) Comparison of re -generation and recruitment of shade-tolerant and light-demanding tree species in mixed uneven-aged forests ex -periences from the Dinaric region Forestry 88 552minus563
Koumlrner C (2003) Alpine plant life functional plant ecology ofhigh mountain ecosystems with 47 tables Springer Sci-ence amp Business Media BV Dordrecht
Koumlrner C (2012) Alpine treelines Functional ecology of theglobal high elevation tree limits Springer Basel
Koumlrner C Paulsen J (2004) A world-wide study of high alti-
tude treeline temperatures J Biogeogr 31 713minus732Kulakowski D Bebi P Rixen C (2011) The interacting effects
of land use change climate change and suppression ofnatural disturbances on landscape forest structure in theSwiss Alps Oikos 120 216minus225
Kullman L (1988) Holocene history of the forestminusalpine tun-dra ecotone in the Scandes Mountains (central Sweden)New Phytol 108 101minus110
Kullman L (1999) Early holocene tree growth at a high elevation site in the northernmost Scandes of Sweden(Lapland) a palaeobiogeographical case study based onmega fossil evidence Geogr Ann 81 63minus74
Kullman L (2007) Tree line population monitoring of Pinussylvestris in the Swedish Scandes 1973minus2005 implica-tions for tree line theory and climate change ecologyJ Ecol 95 41minus52
Kullman L Oumlberg L (2009) Post-Little Ice Age treeline riseand climate warming in the Swedish Scandes a land-scape ecological perspective J Ecol 97 415minus429
Kyriazopoulos AP Skre O Sarkki S Wielgolaski FE Abra-ham EM Ficko A (2017) Humanminusenvironment dynamicsin European treeline ecosystems a synthesis based onthe DPSIR framework Clim Res 73 17ndash29
Lenoir J Svenning JC (2015) Climate-related range shifts mdasha global multidimensional synthesis and new researchdirections Ecography 38 15minus38
Lenoir J Geacutegout JC Marquet PA de Ruffray P Brisse H(2008) A significant upward shift in plant species opti-mum elevation during the 20th century Science 320 1768minus1771
Lindner M Fitzgerald JB Zimmermann NE Reyer C andothers (2014) Climate change and European forests What do we know what are the uncertainties and whatare the implications for forest management J EnvironManag 146 69minus83
Lloyd AH Rupp TS Fastie CL Srafield AM (2002) Patternsand dynamic of treeline advance on the Seward Pen -isula Alaska J Geophys Res 107 8161
Maacuteliš F Kopeckyacute M Petriacutek P Vladovic J Merganic J VidaT (2016) Life-stage not climate change explains ob -served tree range shifts Glob Change Biol 22 1904minus1914
Mathisen IE Mikheeva A Tutubalina OV Aune S and Hof-gaard A (2014) Fifty years of tree line change in theKhibiny Mountains Russia advantages of combinedremote sensing and dendroecological approaches ApplVeg Sci 17 6ndash16
Millar CI Stephenson NL (2015) Temperate forest healthin an era of emerging megadisturbance Science 349 823minus826
Mina M Bugmann H Klopcic M Cailleret M (2017) Accu-rate modeling of harvesting is key for projecting futureforest dynamics a case study in the Slovenian moun-tains Reg Environ Change 17 49minus64
Moen J Aune K Edenius L Angerbjoumlrn A (2004) Potentialeffects of climate change on treeline position in theSwedish mountains Ecol Soc 9 16 wwwecologyandsocietyorgvol9iss1art16
Palombo C Chirici G Marchetti M Tognetti R (2013) Is landabandonment affecting forest dynamics at high elevationin Mediterranean mountains more than climate changePlant Biosyst 147 1minus11
Palombo C Marchetti M Tognetti R (2014) Mountain vege-tation at risk current perspectives and research reedsPlant Biosyst 148 35minus41
Pauli H Gottfried M Dullinger S Abdaladze O Akhalkatsi
149
Clim Res 73 135ndash150 2017
M Alonso JLB Grabherr G (2012) Recent plant diversitychanges on Europersquos mountain summits Science 336 353minus355
Payette S Fortin MJ Gamachet I (2001) The subarctic forestminustundra the structure of a biome in a changing climate BioScience 51 709minus718
Pearson GR Dawson TP (2003) Predicting the impacts of cli-mate change on the distribution of species Are bio -climate envelope models useful Glob Ecol Biogeogr 12 361minus371
Pedrotti F (2013) Plant and vegetation mapping GeobotanyStudies Springer-Verlag Berlin
Pentildeuelas J Boada M (2003) A global change-induced biomeshift in the Montseny mountains (NE Spain) GlobChange Biol 9 131minus140
Piovesan G Biondi F Filippo AD Alessandrini A MaugeriM (2008) Drought driven growth reduction in old beech(Fagus sylvatica L) forests of the central ApenninesItaly Glob Change Biol 14 1265minus1281
Rabasa SG Granda E Benavides R Kunstler G and others(2013) Disparity in elevational shifts of European trees inresponse to recent climate warming Glob Change Biol19 2490minus2499
Raev I Zhelev P Grozeva M Markov I and others (2011)Programme of measures for adaptation of forests in Bul-garia and mitigate the negative impact of climate changeon them Project FUTURE forest helping Europe tackleclimate change INTERREG IVC ERDF Sofia
Sarkki S Ficko A Grunewald K Nijnik M (2016) Benefitsfrom and threats to European treeline ecosystem serv-ices an exploratory study of stakeholders and gover-nance Reg Environ Change 16 2019minus2032
Savage J Vellend M (2015) Elevational shifts biotic homo -genization and time lags in vegetation change during 40years of climate warming Ecography 38 546minus555
Schwoumlrer C Henne PD Tinner W (2014) A model-data com-parison of Holocene timberline changes in the SwissAlps reveals past and future drivers of mountain forestdynamics Glob Change Biol 20 1512minus1526
Šmilauer P Lepš J (2014) Multivariate analysis of ecologicaldata using Canoco 5 Cambridge University Press Cam-bridge
Ste-Marie C Nelson EA Dabros A Bonneau ME (2011)Assisted migration introduction to a multifaceted con-cept For Chron 87 724minus730
Stevens JT Stafford HD Harrison S Latimer AM (2015) For-est disturbance accelerates thermophilization of under-story plant communities J Ecol 103 1253minus1263
Strid A Andonoski A Andonovski V (2003) Alpine biodiver-sity in Europe Ecological Studies 167 Springer-VerlagBerlin
Tattoni C Ciolli M Ferretti F Cantiani MG (2010) Monitor-ing spatial and temporal pattern of Paneveggio forest(Northern Italy) from 1859 to 2006 iForest Biogeosci For3 72minus80
Ter Braak CJF Šmilauer P (2012) Canoco reference manualand userrsquos guide software for ordination (version 50)Microcomputer Power Ithaca NY
Theurillat JP Guisan A (2001) Potential impact of climatechange on vegetation in the European Alps a reviewClim Change 50 77minus109
Toslashmmervik H Wielgolaski FE Neuvonen S Solberg BHoslashgda KA (2005) Biomass and production on a land-scape level in the mountain birch forests In WielgolaskiFE (ed) Plant ecology herbivory and human impact in
Nordic mountain birch forests Ecos Studies 180Springer Berlin p 53minus70
Treml V Banaš M (2000) Alpine timberline in the HighSudetes Acta Univ Carol Geogr 15 83minus99
Treml V Chuman T (2015) Ecotonal dynamics of the altitu-dinal forest limit are affected by terrain and vegetationstructure variables an example from the Sudetes Moun-tains in Central Europe Arct Antarct Alp Res 47 133minus146
Treml V Migo P (2015) controlling factors limiting timber-line position and shifts in the Sudetes a review GeogrPol 88 55minus70
Urban MC Tewksbury JJ Sheldon KS (2012) On a collisioncourse competition and dispersal differences create no-analogue communities and cause extinctions during cli-mate change Proc R Soc Lond B Biol Sci 279 2072minus2080
Vacek S Matejka K (2010) State and development of phyto-cenoses on research plots in the Krkonoše Mts foreststands J For Sci 56 505minus517
Van Bogaert R Haneca K Hoogesteger J Jonasson C deDapper M Callaghan TV (2011) A century of tree linechanges in sub-Arctic Sweden shows local and regionalvariability and only a minor influence of 20th century climate warming J Biogeogr 38 907minus921
Van Gils H Batsukh O Rossiter D Munthali W Liberato -scioli E (2008) Forecasting the pattern and pace of Fagusforest expansion in Majella National Park Italy ApplVeg Sci 11 539minus546
Velchev V (1997) Types of vegetation In Yordanova MDonchev D (eds) Geography of Bulgaria BAN Sofiap 269minus283 (in Bulgarian)
Vladovic J Merganic J Maacuteliš F Križovaacute E and others (2014)The response of forest vegetation diversity to changes inedaphic-climatic conditions in Slovakia Technickaacute uni-verzita vo Zvolene Zvolene (in Slovak)
Vrška T Adam D Hort L Kolaacuter T Janiacutek F (2009) Europeanbeech (Fagus sylvatica L) and silver fir (Abies alba Mill)rotation in the Carpathians mdash a developmental cycle or alinear trend induced by man For Ecol Manag 258 347minus356
Wielgolaski FE (ed) (2005) Plant ecology herbivory andhuman impact in Nordic mountain birch forests Eco -logical Studies 180 Springer Verlag Berlin
Wielgolaski FE Hofgaard A Holtmeier FK (2017) Sensitivityto environmental change of the treeline ecotone and itsassociated biodiversity in European mountains Clim Res73151ndash166
Wieser G Tausz M (2007) Trees at their upper limit Treelifelimitation at the alpine timberline Springer Dordrecht
Wieser G Holtmeier FK Smith WK (2014) Treelines in achanging global environment In Tausz M Grulke N(eds) Trees in a changing environment Plant Ecophys 9Springer Dordrecht p 221minus263
Zhiyanski M Kolev K Sokolovska M Hursthouse A (2008)Tree species effect on soils in Central Stara PlaninaMountains Nauka Gorata 4 65minus82
Zhiyanski M Gikov A Nedkov S Dimitrov P Naydenova L(2016) Mapping carbon storage using land coverlanduse data in the area of Beklemeto Central Balkan In Koulov B Zhelezov G (eds) Sustainable mountain re -gions challenges and perspectives in southeasternEurope Springer Basel p 53minus65
Zhu K Woodall CW Clark JS (2012) Failure to migrate lackof tree range expansion in response to climate changeGlob Change Biol 18 1042minus1052
150
Editorial responsibility Nils Chr Stenseth Oslo Norway
Submitted April 8 2016 Accepted March 8 2017Proofs received from author(s) July 7 2017
Clim Res 73 135ndash150 2017
treeline ecotone is narrower treeline shiftis smaller the timberline occurs at lowerelevations and the treeline is formed moreby broadleaved species Higher values oftreeline ecotone width and treeline shiftwere found in the mountains with a greatersize of the whole mountain massif
When every explanatory variable wastested separately values of the start of thedecrease of human influence in the moun-tains also had a significant effect on thevariation of the tree line ecotone data(explained variation = 206 pseudo-F =34 p = 0024) Some selected relationshipsof the explanatory and dependent vari-ables even those not showing significanteffects on the variation in the tree line eco-tone parameters are depicted in Fig 5
32 Effect of recent changes in climateand land use on the biodiversity
Biodiversity loss in forests meadows andanimal communities analysed by RDAwas explained by geo graphical positiontreeline species composition and tempera-ture increase between the 2 periods1961minus1990 and 2021minus2050 (Fig 4) Theadjusted ex plained variation by all ex pla -natory variables was 632 (F = 58 p =0001) The RDA results indicated that bio-diversity loss in forest communities in crea -sed with increasing latitude and longitudeas well as in the case where broad leavedspecies formed the treeline In contrast thehighest biodiversity loss in meadows wasfound in the mountains positioned more tothe south and with higher temperatureincrease The biodiversity loss in animalcommunities was negatively correlatedwith longitude and temperature increase
4 DISCUSSION
The comparison of several mountainareas situated across Europe shows a varia-tion in understanding of what constitutesthe treeline across countries The primaryissue is the different approach to the defini-tion of forest when applying a tree heightthreshold This threshold decreases from
142
Mou
nta
in u
nit
Nor
th
Eas
t M
oun
-T
imb
er-
Wid
th
Tre
e T
emp
erat
ure
T
reel
ine
Sta
rt
Sta
rt
lati
tud
elo
ng
itu
de
tain
lin
eof
spec
ies
incr
ease
sh
ift
of H
I of
HI
size
(m a
sl
)tr
eeli
ne
form
ing
b
etw
een
(m
yrminus
1 )(y
r)d
ecre
ase
(km
2 )ec
oton
eth
e 19
61minus
1990
(y
r)(m
)tr
eeli
ne
and
199
1minus20
15 (
degC)
Cen
tral
Pyr
enee
s (C
P)
42deg2
1rsquoN
0deg30
rsquoW55
000
2250
300
C0
81
92 (
1956
minus20
06)
1000
BC
1930
Eas
tern
Pyr
enee
s (E
P)
42deg2
8rsquoN
1deg30
rsquoE55
000
2400
300
C0
80
71 (
1956
minus20
06)
1000
1930
Ap
enn
ines
(A
P)
42deg0
5rsquoN
14deg0
5rsquoE
5000
018
0030
0B
14
1 (1
954minus
2007
)10
00 B
C19
50S
har
a M
ts (
SH
)41
deg47rsquo
N20
deg33rsquo
E16
0018
5035
0C
B0
71
(193
4minus20
10)
1300
1970
Pir
in (
PI)
41deg4
2rsquoN
23deg3
1rsquoE
2585
1900
400
C0
5n
o d
ata
680
1962
Cen
tral
Bal
kan
Mts
(C
B)
42deg4
7primeN
24deg3
6primeE
1160
016
0025
0C
B0
7n
o d
ata
1700
1980
Nor
ther
n D
inar
ic M
ts (
DI)
45deg3
6rsquoN
14deg2
8rsquoE
1000
1540
160
B1
3n
o d
ata
1400
1900
Cen
tral
Alp
s (C
A)
46deg4
6rsquoN
9deg52
rsquoE18
000
021
0055
0C
07
13
(195
4minus20
15)
1300
1880
Eas
tern
Alp
s (E
A)
46deg1
7rsquoN
11deg4
5rsquoE
180
000
1800
400
C0
91
3 (1
954minus
2015
)11
5019
50L
ow T
atra
Mts
(LT
)48
deg55rsquo
N19
deg31rsquo
E12
4014
0033
0C
09
057
(19
75minus
2016
)14
0019
22G
ian
t M
ts (
GM
)50
deg42rsquo
N15
deg38rsquo
E63
012
5025
0C
07
043
(19
36minus
2005
)10
0019
63H
ard
ang
ervi
dd
a (H
G)
60deg0
3rsquoN
7deg47
rsquoE90
0090
015
0C
B0
50
8 (1
915minus
2007
)80
019
70D
ovre
(D
O)
62deg2
1rsquoN
8deg52
rsquoE60
0010
0040
0B
07
08
(191
5minus20
07)
1000
no
dec
reas
eN
orth
ern
Sw
edis
h L
apla
nd
(N
S)
68deg1
2rsquoN
18deg3
3rsquoE
4000
700
300
B1
10
6 (1
958minus
2008
)10
00n
o d
ecre
ase
Inn
er F
inn
mar
kn
orth
ern
mos
t 69
deg10rsquo
N25
deg12rsquo
E60
000
290
90B
10
06
(195
8minus20
08)
1600
no
dec
reas
eF
inn
ish
Lap
lan
d (
FL
)
Tab
le 2
Fac
tors
infl
uen
cin
g t
reel
ine
ecot
one
char
acte
rist
ics
in s
elec
ted
Eu
rop
ean
mou
nta
ins
Th
e m
oun
tain
siz
e is
th
e si
ze o
f th
e w
hol
e m
oun
tain
mas
sive
fu
nct
ion
ing
as a
mez
osca
le c
lim
ate
un
it n
ot o
nly
the
stu
die
d m
oun
tain
are
a W
idth
of t
he
tree
lin
e ec
oton
e w
as d
eter
min
ed b
y su
btr
acti
ng
the
valu
e of
the
mea
n ti
mb
erli
ne
elev
atio
n
from
th
e va
lue
of t
he
up
per
tre
e sp
ecie
s li
mit
C c
onif
ers
B b
road
leav
ed t
rees
HI
hu
man
infl
uen
ce
Cudliacuten et al Drivers of treeline shift
5 m (Jeniacutek amp Lokvenc 1962) to 3 m(Koumlrner 2012) in Northern andCentral Europe to only 2 m (Holt-meier 2009) in Central and South-ern Europe where even shrub -by stands (eg Pinus mugo) areconsi dered as a forest in somecountries eg in Spain Italy andMa ce donia (Batllori et al 2009)An other problem is a different de -signation of forest stands belowthe treeline lsquoupper montane fo -restsrsquo in Central Europe and lsquosub-alpine forestsrsquo in Southern Europe(Ellenberg 1988) Further differen -ces are related to the tradition ofdifferent branches of science (ega more traditional lsquogeobotanicalrsquoapproach in Central Europe ver-sus a more experimental approachin Western Europe) especially inthe rate of applying new pro -gressive methods (eg climaticmodelling or molecular biologicalmethods)
Our comparison of selectedregions (Tables 1 amp 2 Figs 1 amp 2)showed that southern mountainscompared to those located morecentrally and in the north had (1) alonger and more profound exploi -tation by humans in the past (2)greater differences in climatic para -meters (temperature length ofgrowing period) between the 2periods 1961minus1990 and 1991minus2015(Table 3) and (3) more dramaticclimate change scenarios espe-cially concerning temperature in -creases Longitude also had someinfluence on the start and intensityof human influence and on tree-line elevation and rate of treelineshift (Table 2 Fig 5) The occur-rence of broadleaf tree species inthe treeline ecotone in the north-ern (and to some extent the south-ern) countries distinguishes thesefrom the Central European coun-tries where conifers prevail It isinteresting that grazing an im -portant factor shaping treelineecosystems (Dirnboumlck et al 2003
143
Length Beginning Mean Mean of growing of growing annual annual
period season temperature precipitation (d)a (d)b (degC) ()
Central Pyrenees 135 minus89 08 minus107Eastern Pyrenees 117 minus91 08 minus92Apennines 284 minus232 14 57Shara Mts 16 01 07 minus100Pirin 26 minus14 05 27Central Balkan Mts 169 minus136 07 76Northern Dinaric Mts 160 minus95 13 minus56Central Alps 147 minus86 07 62Eastern Alps 125 minus96 09 minus33Low Tatra Mts 24 minus34 09 minus91Giant Mts minus10 13 07 55Hardangervidda Blefjell 63 34 05 56Dovre 138 32 07 64Northern Swedish Lapland 81 05 11 minus27Inner Finnmarknorthern- 17 08 10 77most Finnish Lapland
aPositive numbers indicate a prolongation bNegative numbers indicate an earlier beginning
Table 3 Differences of selected climatic parameters between the 2 study periods (1961minus1990 and 1991minus2015) in selected European mountains
Tein Drou Gefa Nadi Laus Abli Bili Huli
0 1 2 3
Central Pyrenees (CP)
Eastern Pyrenees (EP)
Apennines (AP)
Shara Mts (SH)
Pirin (PI)
Central Balkan Mts (CB)
Northern Dinaric Mts (DI)
Central Alps (CA)
Eastern Alps (EA)
Low Tatra Mts (LT)
Giant Mts (GM)
Hardangervidda (HG)
Dovre (DO)
Northern Swedish Lapland (NS)
Inner FinnmarknorthernmostFinnish Lapland (FL)
Fig 2 Rates of treeline drivers (temperature increase [Tein] land use change[Laus]) and treeline shift limits (drought [Drou] geomorphological factors [Gefa]natural disturbances [Nadi] other abiotic biotic and human factors) in selectedEuropean mountains Abli Bili and Huli abiotic biotic and human treeline shiftlimits respectively) Rate of treeline driver influence mdash 0 no influence 1 weak
influence 2 middle influence 3 strong influence
Clim Res 73 135ndash150 2017
Gehrig-Fasel et al 2007) limits treeline shift in halfof the countries of interest regardless of latitude lon-gitude or recent political developments (Shara PirinCentral Balkans Alps Scandes Fig 2)
Despite the stated differences between studieslooking at climate change impacts on the treelineit is clear that current and future changes in temp -erature will seriously affect treeline ecotones in allmountain ranges of Europe Winter is a period whentreeline stands and individual trees have to survivesevere life-limiting conditions (Wieser amp Tausz2007) Therefore winter warming might increase thechance of young trees surviving and passing the mostcritical period from seedling to sapling and further tothe mature stage (Koumlrner 2003) There is alreadymuch evidence worldwide that winter warming isone of the significant drivers of treeline advance(Harsch et al 2009) Although it is expected that treegrowth at the upper distributional margins (and inthe northern European countries) will increase due toongoing climate change (Pentildeuelas amp Boada 2003Chen et al 2011 Hlaacutesny et al 2011 Lindner et al2014) extremes in temperature (eg black froststemperature reversals in the spring) or winter precip-itation (eg lack of snow drought) might also limit
this process in the future in some regions and localsituations (Holtmeier amp Broll 2005 Hagedorn et al2014) An earlier start to the growing period (espe-cially in the Apennines Central Balkan and SpanishPyrenees see Table 2) can play a negative role in theresistance of trees and seedlings to early springfrosts On the other hand at lower distributional mar-gins (and in Southern European countries) it isexpected that tree growth will decrease or forestswill experience some level of dieback due to drought(Breshears et al 2005 Piovesan et al 2008 Allen etal 2010 Huber et al 2013) A serious implication isthe positive effect that warming can have on root rotfungi and populations of bark beetles resulting inlarge-scale disturbances to temperate and also high-altitude forests (Jankovskyacute et al 2004 Elkin et al2013 Millar amp Stephenson 2015)
Although recent upward advances of treeline eco-tones are widespread in mountain regions (Harsch et
144
Fig 3 Redundancy analysis diagram with variation in tim-berline elevation width of treeline ecotone the tree speciesforming the treeline and treeline shift used as dependentvariables explained by explanatory variables (latitudemountain size) Mountain units (blue) are projected as thecentres of abbreviations (see Table 2) Explanatory variablesaccount for 436 of the total variation in the dependentdata The first canonical axis explained 466 of variationthe second axis explained 30 of variation Dependentvariables (black) are Ecot altitudinal width of treeline eco-tone TLShift treeline shift TrspC (TrspB) treeline formedby conifers (broadleaved trees) Timb timberline elevationExplanatory variables (red) are N north latitude Mtsize
size of the mountain massif
Fig 4 Redundancy analysis diagram with variation in bio-diversity loss in forests meadows and animal communitiesused as dependent variables explained by explanatory vari-ables (geographical position tree species forming the tree-line and temperature increase between the 2 study periods[1961minus1990 and 2021minus2050]) Mountain units (green) areprojected as centres of abbreviations (see Table 2) Explana-tory variables account for 763 of total variation in the de-pendent data The first canonical axis explained 484 ofvariation the second axis explained 183 of variation De-pendent variables (black) are BdfoBdmeBdan biodiversityloss in forestsmeadowsanimal communities Explanatoryvariables (red) are N north latitude E east longitude TrspC(TrspB) treeline formed by conifers (broadleaved trees)Temp temperature increase between the 2 study periods
Cudliacuten et al Drivers of treeline shift
al 2009) only a few published studies have includedquantitative data about treeline shifts (Devi et al2008 Kullman amp Oumlberg 2009 Diaz-Varela et al 2010Van Bogaert et al 2011) Additionally our know -ledge of the spatial patterns in treeline ecotone shiftsat the landscape scale is still surprisingly poor exceptfor some studies from subarctic areas the UralMountains and the Alps (Lloyd et al 2002 Diaz-Varela et al 2010 Hagedorn et al 2014) In the 15studied mountain units the values of treeline shiftranged from 043 to 19 m yrminus1 showing a rather dis-tinct spatial pattern of the dynamics within Europeanmountains The observed treeline shift had a signifi-cant positive relationship only with northern latitudeand a weak positive relationship with mountain-range size (ie the size of the whole mountain massifFig 3) Nevertheless this illustrates the importanceof the mass elevation effect (Koumlrner 2012) and par-tially ex plains why treelines could be at different alti-tudes at the same latitude Small negative regres-sions with east longitude and timberline elevationare shown in Fig 5 There was no apparent depend-ence of the treeline shift rate on climatic parameterchanges between the periods 1961minus1990 and1991minus2015
A whole forest vegetation zone shift is a complexprocess as not only the life strategies of an individualtree need to be considered but plant animal andmicroorganism species and their interrelationshipsas well as the relationships to their specific microhab-
itats (including soil conditions) are also involved(Urban et al 2012) Across all regions we identifiedseveral important obstacles to treeline shifts oftenrelated to site properties such as significant rocki-ness having shallow or low-nutrient soils extremerelief causing disturbances by avalanches snow glid-ing or wind damage (data from the Pyrenees Apen-nines Shara Northern Dinaric Low Tatra and GiantMountains see Fig 2) Soil heterogeneity certainlyhas an important role in plant responses to climatechange and could also maintain the resilience of thecommunity assemblages (Fridley et al 2011) An -other group of obstacles includes extreme climaticparameters especially in winter (reported eg fromthe Pyrenees Apennines Pirin Central BalkanNorthern Dinaric and Giant Mountains Fig 2) Theircombination can result in edaphic andor climaticunsuitability of habitats where species could poten-tially migrate Climatically and edaphically suitablesites will likely decline over the next century partic-ularly in mountain landscapes (Bell et al 2014)There fore trees in the treeline ecotone will colonizepreviously forested habitats The colonization suc-cess of individual forest communities is affected bydifferences in species dispersal and recruitmentbehaviour (Dullin ger et al 2004 Jonaacutešovaacute et al2010) For these reasons colonization of new foresthabitats by the next tree generation may not be suc-cessful and may result in loss of species diversity(Honnay et al 2002 Ibaacutentildeez et al 2009 Dobrowski et
145
Fig 5 Univariate graphs of the relationships of dependent variables viz timberline (Timb) and treeline shift per year (Shift) and their explanatory variables Nola north latitude Ealo east longitude DD decimal degrees
Clim Res 73 135ndash150 2017
al 2015) as reported from Macedonia Slovenia theCzech Republic Slovakia and Norway (Fig 4)
Another limiting factor is the existence of strongadaptation mechanisms of the dominant tree specieswhich allow them to survive in their current distribu-tion areas Species migration is likely to be slow dueto the limited quantity of climatically and edaphicallysuitable sites and the slow velocity of seed dispersaland tree regeneration and will be hampered evenmore by fragmentation of high mountain landscapescaused by human activities including brush invasionin abandoned meadows (Gartzia et al 2014) A verti-cal shift in a vegetation zone involves not only singlespecies of plants animals and microorganisms butalso their interrelationships and links to soil condi-tions The speed at which climatic conditions changewill be different from changes in the soil conditionsdue to the slowness of soil-forming processes Soiltypes and conditions (cambisol versus podzol) arecrucial obstacles for the shift from a beech forest zoneinto the spruce forest zone in the Czech Republic(Vacek amp Mat jka 2010) According to other sourcesthe dominant tree species can influence soil for -mation processes (especially humus forms) Underfavourable climatic and orographic conditions beechis able to change its soil conditions over decades tocenturies (J Macku unpubl data)
Significant changes in biodiversity must be expec -ted in all of the mountain areas of interest We foundthat biodiversity declined particularly in the southernregions of Europe where the timberline is situated athigher elevations and human impact is mostlylonger-lasting (Fig 4) Similarly Pauli et al (2012)reported an increase in species richness of mountaingrasslands across Europe except for Mediterraneanregions and assigned this different response of thesouthern regions to decreased water availability Onthe other hand ecosystems which exhibited bio -diversity increases were not only enriched by mig -rant species but the assemblages underwent thermo -philization ie cold-adapted species declined andmore warm-adapted species spread (Gottfried et al2012) However not all species are able to track cli-mate changes It is expected that around 40 ofhabitats will become climatically unsuitable for manymountain species particularly endemic ones in -creasing their extinction probability during this cen-tury (Dullinger et al 2012) The abandonment of tra-ditional land use forms is another source of this losswhich may be a more important driver than tempera-ture increase (Fig 4) Serious biodiversity losses inmountain meadows were reported from Spain ItalyMacedonia Bulgaria Slovenia and Slovakia There-
fore controlled grazing must occur in order to main-tain alpine grasslands (Dirnboumlck et al 2003) Unfor-tunately grazing is still active only in smaller parts ofEuropean mountains (eg in the Central Alps and theCentral Balkan Mountains but pasturing can alsonegatively affect vegetation diversity) and some-times does not serve to maintain alpine grasslandThe same is true for grassland management whichcan help to maintain mountain species and decreasehabitat loss A serious biodiversity loss in mountainmeadows was reported from the Pyrenees the Apen-nines and the Shara Pirin Central Balkan NorthernDinaric and Low Tatra Mountains (Fig 4)
In forests species responses to climate change maybe equivocal some recent studies indicated contra-dictory shifts in species distributions (eg Lenoir etal 2008 Zhu et al 2012 Rabasa et al 2013) This dis-parity is widely discussed and assigned to the greatvariety of non-climatic factors or even tree ontogeny(Grytnes et al 2014 Lenoir amp Svenning 2015 Maacutelišet al 2016) Disturbances leading to tree mortalitymay also play an important role (Cudliacuten et al 2013)changes in tree canopy cover modify light availabil-ity and microclimate and can induce the thermo -philization of forest vegetation (De Frenne et al2015 Stevens et al 2015) The loss of this microcli-mate buffering effect of forests may induce a biotichomogenization of forests (Savage amp Vellend 2015)or the creation of novel non-analogical communitiessuch as oakminuspine forests (Urban et al 2012) whichmay be a new threat to forest biodiversity
The observed changes in treeline forest ecosystemsare often related to changes in land-use intensity(Theurillat amp Guisan 2001 Alados et al 2014) Ac -cording to climate change predictions and the recentand future exploitation intensity of European moun-tains trees and forest communities will shift upwarddue to land use change or climate change or bothReduced land-use intensity certainly will interactwith climate change by facilitating or inhibiting spe-cies occurrence and will accelerate forest expansionabove the present treeline particularly to previouslyforested habitats (Theurillat amp Guisan 2001) Thesimultaneous action of both main treeline shift driv-ers viz temperature increase and decrease in landuse intensity was recorded from all of our studiedmountain areas The extensive differences in timber-line and treeline elevations in almost all studiedmountains (Table 1) indicate the anthropo-zoogenictreeline type (according to Ellenberg 1998) In mostmountains (eg in the Apennines Shara MountainsCentral Balkans and Alps) land use is the prevailingfactor influencing vegetation drift (Fig 5) Previous
146
Cudliacuten et al Drivers of treeline shift
research showed that the upward shift of the treelinein the Swiss Alps was predominantly attributable toland abandonment and only in some situations to cli-mate change (Gehrig-Fasel et al 2007) In the Apen-nines in the last few decades tree establishment hasbeen mainly controlled by land use while treegrowth has been controlled by climate pointing to aminor role played by climate in shaping the currenttreeline (Palombo et al 2013)
The impact of climate change and the connectedland use change on biodiversity and ecosystem serv-ices provision in several European countries is sum-marized by Wielgolaski et al (2017) and Kyriazopou-los et al (2017) (both this Special) and Sarkki et al(2016) One of the possible adaptive management op-tions in response to climate change assisted migra-tion as human-assisted movement of species hasbeen frequently debated in the last few years (Ste-Marie et al 2011) It is possible to apply it as a type ofassisted colonization ie intentional movement andrelease of an organism outside its indigenous range toavoid extinction of populations of the focal species(eg Macedonian pine species) or as an ecologicalre placement ie the intentional movement and re-lease of an organism outside its indigenous range toperform a specific ecological function (eg planting ofPinus mugo in the Alps IUCNSSC 2013)
5 CONCLUSIONS
The analysis of 11 mountain areas across Europeshowed that with increasing latitude the treelineand altitudinal width of the treeline ecotone signifi-cantly decreases as do the significance of climaticand soil parameters as barriers against tree speciesshift Although temperature is the overwhelmingcontrolling factor of tree growth and establishment intemperate and boreal treeline ecotones late-sea-sonal drought might also play a driving role in Medi-terranean treeline ecotones Longitude was lessinfluential mostly affecting climate and land usechange effects on increased biodiversity loss as wellas the size of the area that forms one mesoclimateunit affecting altitudinal ecotone width The biggestpart of the commonly observed remaining variabilityin mountain vegetation near the treeline in Europeseems to be caused by geomorphological geologicalpedological and microclimatic variability in combina-tion with different land use history and the presentsocio-economic relations The observed differencesin climatic parameters between the mountain areasof interest in comparison with the reference period
1960minus1990 (09degC) have not explained the relativelyhigh differences in the rate of treeline shift per year(149 m) The predicted variability in temperatureincrease due to global warming (12minus26degC in 2050and 22minus53degC in 2100) could lead to a much biggerdifferentiation in treeline ecotone biodiversity andecosystem processes between southern and northernEuropean mountains in the future Therefore thesedifferences must be taken into account by scientistsand EU policy makers when formulating efficientadaptive forest management strategies for treelineecosystems at the European level (eg assistedmigration of adapted genotypes)
Acknowledgements This paper is based firstly upon workfrom the COST Action ES 1203 SENSFOR supported byCOST (European Cooperation in Science and Technologywwwcosteu) This international work was further sup-ported by projects granted by the Ministry of EducationYouth and Sports of the Czech Republic grant NPU ILO1415 and LD 14039 by the agency APVV SR under pro-jects APVV-14-0086 and APVV-15-0270 We acknowledgethe E-OBS dataset from the EU-FP6 project ENSEMBLES(httpensembles-eumetofficecom) and the data providersin the ECAampD project (wwwecadeu)
LITERATURE CITED
Alados CL Errea P Gartzia M Saiz H Escoacutes J (2014) Posi-tive and negative feedbacks and free-scale pattern distri-bution in rural-population dynamics PLOS ONE 9 e114561
Allen CD Macalady AK Chenchouni H Bachelet D andothers (2010) A global overview of drought and heat-induced tree mortality reveals emerging climate changerisks for forests For Ecol Manag 259 660minus684
Amidžic L Bartula M Jaric S (2012) Syntaxonomy overviewof vegetation of Šar Planina in the Balkan PeninsulaVegetos 25 348minus360
Aune S Hofgaard A Soumlderstroumlm L (2011) Contrasting cli-mate and land use driven tree encroachment pattern ofsub-arctic tundra in Northern Norway and Kola Penin-sula Can J For Res 41 437minus449
Barbeito I Dawes M Rixen C Senn J Bebi P (2012) Factorsdriving mortality and growth at treeline a 30-year exper-iment of 92000 conifers Ecology 93 389minus401
Batllori E Camarero JJ Ninot JM Gutieacuterrez E (2009)Seedling recruitment survival and facilitation in alpinePinus uncinata treeline ecotones Implications andpotential responses to climate warming Glob Ecol Bio-geogr 18 460minus472
Bebi P Teich M Hagedorn F Zurbriggen N Brunner SHGreacutet-Regamey A (2012) Veraumlnderung von Wald undWaldleistungen in der Landschaft Davos im Zuge desKlimawandels Schweiz Z Forstwes 163 493minus501
Bell DM Bradford JB Lauenroth WK (2014) Mountain land-scapes offer few opportunities for high elevation treespecies migration Glob Change Biol 20 1441minus1451
Bergant K (2006) Climate change scenarios for MacedoniaUniversity of Nova Gorica Center for Atmospheric
147
Clim Res 73 135ndash150 2017
Research Nova GoricaBodin J Badeau V Bruno E Cluzeau C Moisselin JM
Walther GR Dupouey JL (2013) Shifts of forest speciesalong an elevational gradient in Southeast France cli-mate change or stand maturation J Veg Sci 24 269minus283
Boncina A (2011) History current status and future pros -pects of uneven-aged forest management in the Dinaricregion an overview Forestry 84 467minus478
Breshears DD Cobb NS Rich PM Price KP and others(2005) Regional vegetation die-off in response to global-change-type drought Proc Natl Acad Sci USA 102 15144minus15148
Camarero JJ Garciacutea-Ruiz JM Sanguumlesa-Barreda G GalvaacutenJD and others (2015a) Recent and intense dynamics in aformerly static Pyrenean treeline Arct Antarct Alp Res47 773minus783
Camarero JJ Gazol A Galvaacuten JD Sanguumlesa-Barreda GGutieacuterrez E (2015b) Disparate effects of global-changedrivers on mountain conifer forests warming-inducedgrowth enhancement in young trees vs CO2 fertilizationin old trees from wet sites Glob Change Biol 21 738minus749
Chen IC Hill JK Ohlemuumlller R Roy DB Thomas CD (2011)Rapid range shifts of species associated with high levelsof climate warming Science 333 1024minus1026
Cudliacuten P Sejaacutek J Pokornyacute J Albrechtovaacute J Bastian OMarek M (2013) Forest ecosystem services under climatechange and air pollution In Matyssek R Clarke N Cud-liacuten P Mikkelsen TN Tuovinen JP Wiesner G Paoletti E(eds) Climate change air pollution and global chal-lenges understanding and perspectives from forestresearch Developments in Environmental Science Vol13 Elsevier Oxford p 521minus546
Dalen L Hofgaard A (2005) Differential regional treelinedynamics in the Scandes mountains Arct Antarct AlpRes 37 284minus296
Dawes MA Philipson CD Fonti P Bebi P Haumlttenschwiler SHagedorn F Rixen C (2015) Soil warming and CO2
enrichment induce biomass shifts in alpine tree line veg-etation 2014 Glob Change Biol 21 2005minus2021
De Frenne P Rodriacuteguez-Saacutenchez F De Schrijver A CoomesDA Hermy M Vangansbeke P Verheyen K (2015) Lightaccelerates plant responses to warming Nat Plants 1 15110
Devi N Hagedorn F Moiseev P Bugmann H ShiyatovSG Mazepa V Rigling A (2008) Expanding forests andchanging growth forms of Siberian larch at the treeline ofthe Polar Urals during the 20th century Glob ChangeBiol 14 1581minus1591
Diaz-Varela R Colombo R Meroni M Calvo-Iglesias MSBuffoni A Tagliaferri A (2010) Spatio-temporal analysisof alpine ecotones A spatial explicit model targeting altitudinal vegetation shifts Ecol Model 221 621minus633
Dirnboumlck T Dullinger S Grabherr G (2003) A regionalimpact assessment of climate and land-use change onalpine vegetation J Biogeogr 30 401minus417
Dobrowski SZ Swanson AK Abatzoglou JT Holden ZASafford HD Schwartz MK Gavin DG (2015) Forest struc-ture and species traits mediate projected recruitmentdeclines in western US tree species Glob Ecol Biogeogr24 917minus927
Dullinger S Dirnboumlck S Grabherr G (2004) Modelling cli-mate change-driven treeline shifts relative effects oftemperature increase dispersal and invisibility J Ecol92 241minus252
Dullinger S Gattringer A Thuiller W Moser D and others
(2012) Extinction debt of high-mountain plants undertwenty-first-century climate change Nat Clim Chang 2 619minus622
Elkin C Gutierrez AG Leuzinger S Manusch C TemperliC Rasche L Bugmann H (2013) A 2degC warmer world isnot safe for ecosystem services in the European AlpsGlob Change Biol 19 1827minus1840
Ellenberg H (1988) Vegetation ecology of Central Europe4th edn Cambridge University Press Edinburgh
Em H (1986) Spruce southern distribution range Spruce for-est on Shara Mountain in Macedonia Contrib Macedon-ian Acad Sci Arts Skopje 5 11minus28
Fajardo A Piper FI Pfund L Koumlrner CH Hoch G (2012)Variation of mobile carbon reserves in trees at the alpinetreeline ecotone is under environmental control NewPhytol 195 794minus802
Fridley JD Grime JP Askew AP Moser B Stevens CJ (2011)Soil heterogeneity buffers community response to cli-mate change in species-rich grassland Glob Change Biol17 2002minus2011
Gartzia M Alados CL Peacuterez-Cabello F (2014) Assessment ofthe effects of biophysical and anthropogenic factors onwoody plant encroachment in dense and sparse moun-tain grasslands based on remote sensing data Prog PhysGeogr 38 201minus217
Gehrig-Fasel J Guisan A Zimmermann NE (2007) Tree lineshifts in the Swiss Alps climate change or land abandon-ment J Veg Sci 18 571minus582
Gikov A Dimitrov P Zhiyanski M (2016) Land cover dynam-ics on the northern slope of the Troyan Passage during 30years period Probl Geogr 1-2 78minus92
Gonzalez P Neilson RP Lenihan JM Drapek RJ (2010)Global patterns in the vulnerability of ecosystems to veg-etation shifts due to climate change Glob Ecol Biogeogr19 755minus768
Gottfried M Pauli H Futschik A Akhalkatsi M and others(2012) Continent-wide response of mountain vegetationto climate change Nat Clim Chang 2 111minus115
Grace J Berninger F Nagy L (2002) Impacts of climatechange on the tree line Ann Bot 90 537minus544
Grunewald K Scheithauer J (2011) Landscape developmentand climate change in Southwest Bulgaria (Pirin Mts)Springer Heidelberg
Grytnes JA Kapfer J Jurasinski G Birks HH and others(2014) Identifying the driving factors behind observedelevational range shifts on European mountains GlobEcol Biogeogr 23 876minus884
Hagedorn F Shiyatov FG Mazepa VS Devi NM and others(2014) Treeline advances along the Urals mountainrange mdash driven by improved winter conditions GlobChange Biol 20 3530minus3543
Hansen AJ Neilson RP Dale VH Flather CH and others(2001) Global change in forests responses of speciescommunities and biomes BioScience 51 765minus779
Harsch MA Bader MY (2011) Treeline formminusa potential keyto understanding treeline dynamics Glob Ecol Biogeogr20 582minus596
Harsch MA Hulme PE McGlone MS Duncan RP (2009) Aretreelines advancing A global meta-analysis of treelineresponse to climate warning Ecol Lett 12 1040minus1049
Haylock MR Hofstra N Klein Tank AMG Klok EJ JonesPD New M (2008) A European daily high-resolutiongridded dataset of surface temperature and precipita-tion J Geophys Res Atmos 113 D20119
Hlaacutesny T Barcza Z Fabrika M Balaacutezs B and others (2011)
148
Cudliacuten et al Drivers of treeline shift
Climate change impacts on growth and carbon balanceof forests in Central Europe Clim Res 47 219minus236
Hlaacutesny T Trombik J Dobor L Barcza Z Barka I (2016)Future climate of the Carpathians climate change hot-spots and implications for ecosystems Reg EnvironChange 16 1495minus1506
Hofgaard A Dalen L Hytteborn H (2009) Tree recruitmentabove the treeline and potential for climate driven tree-line change J Veg Sci 20 1133minus1144
Holtmeier FK (2009) Mountain timberlines Ecology patchi-ness and dynamics 2nd edn Advances in GlobalChange Research 36 Springer Science amp ScienceMediaBV Dordrecht
Holtmeier FK Broll G (2005) Sensitivity and response ofnorthern hemisphere altitudinal and polar treelines toenvironmental change at landscape and local scalesGlob Ecol Biogeogr 14 395minus410
Honnay O Verheyen K Butaye J Jacquemyn H Bossuyt BHermy M (2002) Possible effects of habitat fragmentationand climate change on the range of forest plant speciesEcol Lett 5 525minus530
Huber R Rigling A Bebi P Brand F and others (2013) Sus-tainable land use in mountain regions under globalchange synthesis across scales and disciplines Ecol Soc18 36
Ibaacutentildeez I Clark JS Dietze MC (2009) Estimating colonizationpotential of migrant tree species Glob Change Biol 15 1173minus1188
IUCNSSC (International Union for Conservation of NatureSpecies Survival Commission) (2013) Guidelines for re -introductions and other conservation translocations Version 10 IUCNSpecies Survival Commission Gland
Jankovskyacute L Cudliacuten P gtermaacutek P Moravec I (2004) The pre-diction of development of secondary Norway sprucestands under the impact of climatic change in the Dra-hany highlands (The Czech Republic) Ekologia (Bratisl)23(Suppl 2) 101minus112
Jeniacutek J Lokvenc T (1962) Die alpine Waldgrenze im Krko -noše Gebirge Rozpravy SAV Rada matematickyacutech apriacuterodniacutech ved 72 gteskoslovenskaacute akademie ved Praha
Jobbaacutegy EG Jackson RB (2000) Global controls of forest lineelevation in the northern and southern hemispheresGlob Ecol Biogeogr 9 253minus268
Jonaacutešovaacute M Vaacutevrovaacute E Cudliacuten P (2010) Western Carpa -thian mountain spruce forest after a windthrow naturalregeneration in cleared and uncleared areas For EcolManag 259 1127minus1134
Kittel TGF Steffen WL Chapin FS (2000) Global andregional modelling of Arcticminusboreal vegetation distribu-tion and its sensitivity to altered forcing Glob ChangeBiol 6 1minus18
Klopcic M Jerina K Bonltina A (2010) Long-term changes ofstructure and tree species composition in Dinaric un -even- aged forests Are red deer an important factor EurJ For Res 129 277minus288
Klopcic M Simonltilt T Bonltina A (2015) Comparison of re -generation and recruitment of shade-tolerant and light-demanding tree species in mixed uneven-aged forests ex -periences from the Dinaric region Forestry 88 552minus563
Koumlrner C (2003) Alpine plant life functional plant ecology ofhigh mountain ecosystems with 47 tables Springer Sci-ence amp Business Media BV Dordrecht
Koumlrner C (2012) Alpine treelines Functional ecology of theglobal high elevation tree limits Springer Basel
Koumlrner C Paulsen J (2004) A world-wide study of high alti-
tude treeline temperatures J Biogeogr 31 713minus732Kulakowski D Bebi P Rixen C (2011) The interacting effects
of land use change climate change and suppression ofnatural disturbances on landscape forest structure in theSwiss Alps Oikos 120 216minus225
Kullman L (1988) Holocene history of the forestminusalpine tun-dra ecotone in the Scandes Mountains (central Sweden)New Phytol 108 101minus110
Kullman L (1999) Early holocene tree growth at a high elevation site in the northernmost Scandes of Sweden(Lapland) a palaeobiogeographical case study based onmega fossil evidence Geogr Ann 81 63minus74
Kullman L (2007) Tree line population monitoring of Pinussylvestris in the Swedish Scandes 1973minus2005 implica-tions for tree line theory and climate change ecologyJ Ecol 95 41minus52
Kullman L Oumlberg L (2009) Post-Little Ice Age treeline riseand climate warming in the Swedish Scandes a land-scape ecological perspective J Ecol 97 415minus429
Kyriazopoulos AP Skre O Sarkki S Wielgolaski FE Abra-ham EM Ficko A (2017) Humanminusenvironment dynamicsin European treeline ecosystems a synthesis based onthe DPSIR framework Clim Res 73 17ndash29
Lenoir J Svenning JC (2015) Climate-related range shifts mdasha global multidimensional synthesis and new researchdirections Ecography 38 15minus38
Lenoir J Geacutegout JC Marquet PA de Ruffray P Brisse H(2008) A significant upward shift in plant species opti-mum elevation during the 20th century Science 320 1768minus1771
Lindner M Fitzgerald JB Zimmermann NE Reyer C andothers (2014) Climate change and European forests What do we know what are the uncertainties and whatare the implications for forest management J EnvironManag 146 69minus83
Lloyd AH Rupp TS Fastie CL Srafield AM (2002) Patternsand dynamic of treeline advance on the Seward Pen -isula Alaska J Geophys Res 107 8161
Maacuteliš F Kopeckyacute M Petriacutek P Vladovic J Merganic J VidaT (2016) Life-stage not climate change explains ob -served tree range shifts Glob Change Biol 22 1904minus1914
Mathisen IE Mikheeva A Tutubalina OV Aune S and Hof-gaard A (2014) Fifty years of tree line change in theKhibiny Mountains Russia advantages of combinedremote sensing and dendroecological approaches ApplVeg Sci 17 6ndash16
Millar CI Stephenson NL (2015) Temperate forest healthin an era of emerging megadisturbance Science 349 823minus826
Mina M Bugmann H Klopcic M Cailleret M (2017) Accu-rate modeling of harvesting is key for projecting futureforest dynamics a case study in the Slovenian moun-tains Reg Environ Change 17 49minus64
Moen J Aune K Edenius L Angerbjoumlrn A (2004) Potentialeffects of climate change on treeline position in theSwedish mountains Ecol Soc 9 16 wwwecologyandsocietyorgvol9iss1art16
Palombo C Chirici G Marchetti M Tognetti R (2013) Is landabandonment affecting forest dynamics at high elevationin Mediterranean mountains more than climate changePlant Biosyst 147 1minus11
Palombo C Marchetti M Tognetti R (2014) Mountain vege-tation at risk current perspectives and research reedsPlant Biosyst 148 35minus41
Pauli H Gottfried M Dullinger S Abdaladze O Akhalkatsi
149
Clim Res 73 135ndash150 2017
M Alonso JLB Grabherr G (2012) Recent plant diversitychanges on Europersquos mountain summits Science 336 353minus355
Payette S Fortin MJ Gamachet I (2001) The subarctic forestminustundra the structure of a biome in a changing climate BioScience 51 709minus718
Pearson GR Dawson TP (2003) Predicting the impacts of cli-mate change on the distribution of species Are bio -climate envelope models useful Glob Ecol Biogeogr 12 361minus371
Pedrotti F (2013) Plant and vegetation mapping GeobotanyStudies Springer-Verlag Berlin
Pentildeuelas J Boada M (2003) A global change-induced biomeshift in the Montseny mountains (NE Spain) GlobChange Biol 9 131minus140
Piovesan G Biondi F Filippo AD Alessandrini A MaugeriM (2008) Drought driven growth reduction in old beech(Fagus sylvatica L) forests of the central ApenninesItaly Glob Change Biol 14 1265minus1281
Rabasa SG Granda E Benavides R Kunstler G and others(2013) Disparity in elevational shifts of European trees inresponse to recent climate warming Glob Change Biol19 2490minus2499
Raev I Zhelev P Grozeva M Markov I and others (2011)Programme of measures for adaptation of forests in Bul-garia and mitigate the negative impact of climate changeon them Project FUTURE forest helping Europe tackleclimate change INTERREG IVC ERDF Sofia
Sarkki S Ficko A Grunewald K Nijnik M (2016) Benefitsfrom and threats to European treeline ecosystem serv-ices an exploratory study of stakeholders and gover-nance Reg Environ Change 16 2019minus2032
Savage J Vellend M (2015) Elevational shifts biotic homo -genization and time lags in vegetation change during 40years of climate warming Ecography 38 546minus555
Schwoumlrer C Henne PD Tinner W (2014) A model-data com-parison of Holocene timberline changes in the SwissAlps reveals past and future drivers of mountain forestdynamics Glob Change Biol 20 1512minus1526
Šmilauer P Lepš J (2014) Multivariate analysis of ecologicaldata using Canoco 5 Cambridge University Press Cam-bridge
Ste-Marie C Nelson EA Dabros A Bonneau ME (2011)Assisted migration introduction to a multifaceted con-cept For Chron 87 724minus730
Stevens JT Stafford HD Harrison S Latimer AM (2015) For-est disturbance accelerates thermophilization of under-story plant communities J Ecol 103 1253minus1263
Strid A Andonoski A Andonovski V (2003) Alpine biodiver-sity in Europe Ecological Studies 167 Springer-VerlagBerlin
Tattoni C Ciolli M Ferretti F Cantiani MG (2010) Monitor-ing spatial and temporal pattern of Paneveggio forest(Northern Italy) from 1859 to 2006 iForest Biogeosci For3 72minus80
Ter Braak CJF Šmilauer P (2012) Canoco reference manualand userrsquos guide software for ordination (version 50)Microcomputer Power Ithaca NY
Theurillat JP Guisan A (2001) Potential impact of climatechange on vegetation in the European Alps a reviewClim Change 50 77minus109
Toslashmmervik H Wielgolaski FE Neuvonen S Solberg BHoslashgda KA (2005) Biomass and production on a land-scape level in the mountain birch forests In WielgolaskiFE (ed) Plant ecology herbivory and human impact in
Nordic mountain birch forests Ecos Studies 180Springer Berlin p 53minus70
Treml V Banaš M (2000) Alpine timberline in the HighSudetes Acta Univ Carol Geogr 15 83minus99
Treml V Chuman T (2015) Ecotonal dynamics of the altitu-dinal forest limit are affected by terrain and vegetationstructure variables an example from the Sudetes Moun-tains in Central Europe Arct Antarct Alp Res 47 133minus146
Treml V Migo P (2015) controlling factors limiting timber-line position and shifts in the Sudetes a review GeogrPol 88 55minus70
Urban MC Tewksbury JJ Sheldon KS (2012) On a collisioncourse competition and dispersal differences create no-analogue communities and cause extinctions during cli-mate change Proc R Soc Lond B Biol Sci 279 2072minus2080
Vacek S Matejka K (2010) State and development of phyto-cenoses on research plots in the Krkonoše Mts foreststands J For Sci 56 505minus517
Van Bogaert R Haneca K Hoogesteger J Jonasson C deDapper M Callaghan TV (2011) A century of tree linechanges in sub-Arctic Sweden shows local and regionalvariability and only a minor influence of 20th century climate warming J Biogeogr 38 907minus921
Van Gils H Batsukh O Rossiter D Munthali W Liberato -scioli E (2008) Forecasting the pattern and pace of Fagusforest expansion in Majella National Park Italy ApplVeg Sci 11 539minus546
Velchev V (1997) Types of vegetation In Yordanova MDonchev D (eds) Geography of Bulgaria BAN Sofiap 269minus283 (in Bulgarian)
Vladovic J Merganic J Maacuteliš F Križovaacute E and others (2014)The response of forest vegetation diversity to changes inedaphic-climatic conditions in Slovakia Technickaacute uni-verzita vo Zvolene Zvolene (in Slovak)
Vrška T Adam D Hort L Kolaacuter T Janiacutek F (2009) Europeanbeech (Fagus sylvatica L) and silver fir (Abies alba Mill)rotation in the Carpathians mdash a developmental cycle or alinear trend induced by man For Ecol Manag 258 347minus356
Wielgolaski FE (ed) (2005) Plant ecology herbivory andhuman impact in Nordic mountain birch forests Eco -logical Studies 180 Springer Verlag Berlin
Wielgolaski FE Hofgaard A Holtmeier FK (2017) Sensitivityto environmental change of the treeline ecotone and itsassociated biodiversity in European mountains Clim Res73151ndash166
Wieser G Tausz M (2007) Trees at their upper limit Treelifelimitation at the alpine timberline Springer Dordrecht
Wieser G Holtmeier FK Smith WK (2014) Treelines in achanging global environment In Tausz M Grulke N(eds) Trees in a changing environment Plant Ecophys 9Springer Dordrecht p 221minus263
Zhiyanski M Kolev K Sokolovska M Hursthouse A (2008)Tree species effect on soils in Central Stara PlaninaMountains Nauka Gorata 4 65minus82
Zhiyanski M Gikov A Nedkov S Dimitrov P Naydenova L(2016) Mapping carbon storage using land coverlanduse data in the area of Beklemeto Central Balkan In Koulov B Zhelezov G (eds) Sustainable mountain re -gions challenges and perspectives in southeasternEurope Springer Basel p 53minus65
Zhu K Woodall CW Clark JS (2012) Failure to migrate lackof tree range expansion in response to climate changeGlob Change Biol 18 1042minus1052
150
Editorial responsibility Nils Chr Stenseth Oslo Norway
Submitted April 8 2016 Accepted March 8 2017Proofs received from author(s) July 7 2017
Cudliacuten et al Drivers of treeline shift
5 m (Jeniacutek amp Lokvenc 1962) to 3 m(Koumlrner 2012) in Northern andCentral Europe to only 2 m (Holt-meier 2009) in Central and South-ern Europe where even shrub -by stands (eg Pinus mugo) areconsi dered as a forest in somecountries eg in Spain Italy andMa ce donia (Batllori et al 2009)An other problem is a different de -signation of forest stands belowthe treeline lsquoupper montane fo -restsrsquo in Central Europe and lsquosub-alpine forestsrsquo in Southern Europe(Ellenberg 1988) Further differen -ces are related to the tradition ofdifferent branches of science (ega more traditional lsquogeobotanicalrsquoapproach in Central Europe ver-sus a more experimental approachin Western Europe) especially inthe rate of applying new pro -gressive methods (eg climaticmodelling or molecular biologicalmethods)
Our comparison of selectedregions (Tables 1 amp 2 Figs 1 amp 2)showed that southern mountainscompared to those located morecentrally and in the north had (1) alonger and more profound exploi -tation by humans in the past (2)greater differences in climatic para -meters (temperature length ofgrowing period) between the 2periods 1961minus1990 and 1991minus2015(Table 3) and (3) more dramaticclimate change scenarios espe-cially concerning temperature in -creases Longitude also had someinfluence on the start and intensityof human influence and on tree-line elevation and rate of treelineshift (Table 2 Fig 5) The occur-rence of broadleaf tree species inthe treeline ecotone in the north-ern (and to some extent the south-ern) countries distinguishes thesefrom the Central European coun-tries where conifers prevail It isinteresting that grazing an im -portant factor shaping treelineecosystems (Dirnboumlck et al 2003
143
Length Beginning Mean Mean of growing of growing annual annual
period season temperature precipitation (d)a (d)b (degC) ()
Central Pyrenees 135 minus89 08 minus107Eastern Pyrenees 117 minus91 08 minus92Apennines 284 minus232 14 57Shara Mts 16 01 07 minus100Pirin 26 minus14 05 27Central Balkan Mts 169 minus136 07 76Northern Dinaric Mts 160 minus95 13 minus56Central Alps 147 minus86 07 62Eastern Alps 125 minus96 09 minus33Low Tatra Mts 24 minus34 09 minus91Giant Mts minus10 13 07 55Hardangervidda Blefjell 63 34 05 56Dovre 138 32 07 64Northern Swedish Lapland 81 05 11 minus27Inner Finnmarknorthern- 17 08 10 77most Finnish Lapland
aPositive numbers indicate a prolongation bNegative numbers indicate an earlier beginning
Table 3 Differences of selected climatic parameters between the 2 study periods (1961minus1990 and 1991minus2015) in selected European mountains
Tein Drou Gefa Nadi Laus Abli Bili Huli
0 1 2 3
Central Pyrenees (CP)
Eastern Pyrenees (EP)
Apennines (AP)
Shara Mts (SH)
Pirin (PI)
Central Balkan Mts (CB)
Northern Dinaric Mts (DI)
Central Alps (CA)
Eastern Alps (EA)
Low Tatra Mts (LT)
Giant Mts (GM)
Hardangervidda (HG)
Dovre (DO)
Northern Swedish Lapland (NS)
Inner FinnmarknorthernmostFinnish Lapland (FL)
Fig 2 Rates of treeline drivers (temperature increase [Tein] land use change[Laus]) and treeline shift limits (drought [Drou] geomorphological factors [Gefa]natural disturbances [Nadi] other abiotic biotic and human factors) in selectedEuropean mountains Abli Bili and Huli abiotic biotic and human treeline shiftlimits respectively) Rate of treeline driver influence mdash 0 no influence 1 weak
influence 2 middle influence 3 strong influence
Clim Res 73 135ndash150 2017
Gehrig-Fasel et al 2007) limits treeline shift in halfof the countries of interest regardless of latitude lon-gitude or recent political developments (Shara PirinCentral Balkans Alps Scandes Fig 2)
Despite the stated differences between studieslooking at climate change impacts on the treelineit is clear that current and future changes in temp -erature will seriously affect treeline ecotones in allmountain ranges of Europe Winter is a period whentreeline stands and individual trees have to survivesevere life-limiting conditions (Wieser amp Tausz2007) Therefore winter warming might increase thechance of young trees surviving and passing the mostcritical period from seedling to sapling and further tothe mature stage (Koumlrner 2003) There is alreadymuch evidence worldwide that winter warming isone of the significant drivers of treeline advance(Harsch et al 2009) Although it is expected that treegrowth at the upper distributional margins (and inthe northern European countries) will increase due toongoing climate change (Pentildeuelas amp Boada 2003Chen et al 2011 Hlaacutesny et al 2011 Lindner et al2014) extremes in temperature (eg black froststemperature reversals in the spring) or winter precip-itation (eg lack of snow drought) might also limit
this process in the future in some regions and localsituations (Holtmeier amp Broll 2005 Hagedorn et al2014) An earlier start to the growing period (espe-cially in the Apennines Central Balkan and SpanishPyrenees see Table 2) can play a negative role in theresistance of trees and seedlings to early springfrosts On the other hand at lower distributional mar-gins (and in Southern European countries) it isexpected that tree growth will decrease or forestswill experience some level of dieback due to drought(Breshears et al 2005 Piovesan et al 2008 Allen etal 2010 Huber et al 2013) A serious implication isthe positive effect that warming can have on root rotfungi and populations of bark beetles resulting inlarge-scale disturbances to temperate and also high-altitude forests (Jankovskyacute et al 2004 Elkin et al2013 Millar amp Stephenson 2015)
Although recent upward advances of treeline eco-tones are widespread in mountain regions (Harsch et
144
Fig 3 Redundancy analysis diagram with variation in tim-berline elevation width of treeline ecotone the tree speciesforming the treeline and treeline shift used as dependentvariables explained by explanatory variables (latitudemountain size) Mountain units (blue) are projected as thecentres of abbreviations (see Table 2) Explanatory variablesaccount for 436 of the total variation in the dependentdata The first canonical axis explained 466 of variationthe second axis explained 30 of variation Dependentvariables (black) are Ecot altitudinal width of treeline eco-tone TLShift treeline shift TrspC (TrspB) treeline formedby conifers (broadleaved trees) Timb timberline elevationExplanatory variables (red) are N north latitude Mtsize
size of the mountain massif
Fig 4 Redundancy analysis diagram with variation in bio-diversity loss in forests meadows and animal communitiesused as dependent variables explained by explanatory vari-ables (geographical position tree species forming the tree-line and temperature increase between the 2 study periods[1961minus1990 and 2021minus2050]) Mountain units (green) areprojected as centres of abbreviations (see Table 2) Explana-tory variables account for 763 of total variation in the de-pendent data The first canonical axis explained 484 ofvariation the second axis explained 183 of variation De-pendent variables (black) are BdfoBdmeBdan biodiversityloss in forestsmeadowsanimal communities Explanatoryvariables (red) are N north latitude E east longitude TrspC(TrspB) treeline formed by conifers (broadleaved trees)Temp temperature increase between the 2 study periods
Cudliacuten et al Drivers of treeline shift
al 2009) only a few published studies have includedquantitative data about treeline shifts (Devi et al2008 Kullman amp Oumlberg 2009 Diaz-Varela et al 2010Van Bogaert et al 2011) Additionally our know -ledge of the spatial patterns in treeline ecotone shiftsat the landscape scale is still surprisingly poor exceptfor some studies from subarctic areas the UralMountains and the Alps (Lloyd et al 2002 Diaz-Varela et al 2010 Hagedorn et al 2014) In the 15studied mountain units the values of treeline shiftranged from 043 to 19 m yrminus1 showing a rather dis-tinct spatial pattern of the dynamics within Europeanmountains The observed treeline shift had a signifi-cant positive relationship only with northern latitudeand a weak positive relationship with mountain-range size (ie the size of the whole mountain massifFig 3) Nevertheless this illustrates the importanceof the mass elevation effect (Koumlrner 2012) and par-tially ex plains why treelines could be at different alti-tudes at the same latitude Small negative regres-sions with east longitude and timberline elevationare shown in Fig 5 There was no apparent depend-ence of the treeline shift rate on climatic parameterchanges between the periods 1961minus1990 and1991minus2015
A whole forest vegetation zone shift is a complexprocess as not only the life strategies of an individualtree need to be considered but plant animal andmicroorganism species and their interrelationshipsas well as the relationships to their specific microhab-
itats (including soil conditions) are also involved(Urban et al 2012) Across all regions we identifiedseveral important obstacles to treeline shifts oftenrelated to site properties such as significant rocki-ness having shallow or low-nutrient soils extremerelief causing disturbances by avalanches snow glid-ing or wind damage (data from the Pyrenees Apen-nines Shara Northern Dinaric Low Tatra and GiantMountains see Fig 2) Soil heterogeneity certainlyhas an important role in plant responses to climatechange and could also maintain the resilience of thecommunity assemblages (Fridley et al 2011) An -other group of obstacles includes extreme climaticparameters especially in winter (reported eg fromthe Pyrenees Apennines Pirin Central BalkanNorthern Dinaric and Giant Mountains Fig 2) Theircombination can result in edaphic andor climaticunsuitability of habitats where species could poten-tially migrate Climatically and edaphically suitablesites will likely decline over the next century partic-ularly in mountain landscapes (Bell et al 2014)There fore trees in the treeline ecotone will colonizepreviously forested habitats The colonization suc-cess of individual forest communities is affected bydifferences in species dispersal and recruitmentbehaviour (Dullin ger et al 2004 Jonaacutešovaacute et al2010) For these reasons colonization of new foresthabitats by the next tree generation may not be suc-cessful and may result in loss of species diversity(Honnay et al 2002 Ibaacutentildeez et al 2009 Dobrowski et
145
Fig 5 Univariate graphs of the relationships of dependent variables viz timberline (Timb) and treeline shift per year (Shift) and their explanatory variables Nola north latitude Ealo east longitude DD decimal degrees
Clim Res 73 135ndash150 2017
al 2015) as reported from Macedonia Slovenia theCzech Republic Slovakia and Norway (Fig 4)
Another limiting factor is the existence of strongadaptation mechanisms of the dominant tree specieswhich allow them to survive in their current distribu-tion areas Species migration is likely to be slow dueto the limited quantity of climatically and edaphicallysuitable sites and the slow velocity of seed dispersaland tree regeneration and will be hampered evenmore by fragmentation of high mountain landscapescaused by human activities including brush invasionin abandoned meadows (Gartzia et al 2014) A verti-cal shift in a vegetation zone involves not only singlespecies of plants animals and microorganisms butalso their interrelationships and links to soil condi-tions The speed at which climatic conditions changewill be different from changes in the soil conditionsdue to the slowness of soil-forming processes Soiltypes and conditions (cambisol versus podzol) arecrucial obstacles for the shift from a beech forest zoneinto the spruce forest zone in the Czech Republic(Vacek amp Mat jka 2010) According to other sourcesthe dominant tree species can influence soil for -mation processes (especially humus forms) Underfavourable climatic and orographic conditions beechis able to change its soil conditions over decades tocenturies (J Macku unpubl data)
Significant changes in biodiversity must be expec -ted in all of the mountain areas of interest We foundthat biodiversity declined particularly in the southernregions of Europe where the timberline is situated athigher elevations and human impact is mostlylonger-lasting (Fig 4) Similarly Pauli et al (2012)reported an increase in species richness of mountaingrasslands across Europe except for Mediterraneanregions and assigned this different response of thesouthern regions to decreased water availability Onthe other hand ecosystems which exhibited bio -diversity increases were not only enriched by mig -rant species but the assemblages underwent thermo -philization ie cold-adapted species declined andmore warm-adapted species spread (Gottfried et al2012) However not all species are able to track cli-mate changes It is expected that around 40 ofhabitats will become climatically unsuitable for manymountain species particularly endemic ones in -creasing their extinction probability during this cen-tury (Dullinger et al 2012) The abandonment of tra-ditional land use forms is another source of this losswhich may be a more important driver than tempera-ture increase (Fig 4) Serious biodiversity losses inmountain meadows were reported from Spain ItalyMacedonia Bulgaria Slovenia and Slovakia There-
fore controlled grazing must occur in order to main-tain alpine grasslands (Dirnboumlck et al 2003) Unfor-tunately grazing is still active only in smaller parts ofEuropean mountains (eg in the Central Alps and theCentral Balkan Mountains but pasturing can alsonegatively affect vegetation diversity) and some-times does not serve to maintain alpine grasslandThe same is true for grassland management whichcan help to maintain mountain species and decreasehabitat loss A serious biodiversity loss in mountainmeadows was reported from the Pyrenees the Apen-nines and the Shara Pirin Central Balkan NorthernDinaric and Low Tatra Mountains (Fig 4)
In forests species responses to climate change maybe equivocal some recent studies indicated contra-dictory shifts in species distributions (eg Lenoir etal 2008 Zhu et al 2012 Rabasa et al 2013) This dis-parity is widely discussed and assigned to the greatvariety of non-climatic factors or even tree ontogeny(Grytnes et al 2014 Lenoir amp Svenning 2015 Maacutelišet al 2016) Disturbances leading to tree mortalitymay also play an important role (Cudliacuten et al 2013)changes in tree canopy cover modify light availabil-ity and microclimate and can induce the thermo -philization of forest vegetation (De Frenne et al2015 Stevens et al 2015) The loss of this microcli-mate buffering effect of forests may induce a biotichomogenization of forests (Savage amp Vellend 2015)or the creation of novel non-analogical communitiessuch as oakminuspine forests (Urban et al 2012) whichmay be a new threat to forest biodiversity
The observed changes in treeline forest ecosystemsare often related to changes in land-use intensity(Theurillat amp Guisan 2001 Alados et al 2014) Ac -cording to climate change predictions and the recentand future exploitation intensity of European moun-tains trees and forest communities will shift upwarddue to land use change or climate change or bothReduced land-use intensity certainly will interactwith climate change by facilitating or inhibiting spe-cies occurrence and will accelerate forest expansionabove the present treeline particularly to previouslyforested habitats (Theurillat amp Guisan 2001) Thesimultaneous action of both main treeline shift driv-ers viz temperature increase and decrease in landuse intensity was recorded from all of our studiedmountain areas The extensive differences in timber-line and treeline elevations in almost all studiedmountains (Table 1) indicate the anthropo-zoogenictreeline type (according to Ellenberg 1998) In mostmountains (eg in the Apennines Shara MountainsCentral Balkans and Alps) land use is the prevailingfactor influencing vegetation drift (Fig 5) Previous
146
Cudliacuten et al Drivers of treeline shift
research showed that the upward shift of the treelinein the Swiss Alps was predominantly attributable toland abandonment and only in some situations to cli-mate change (Gehrig-Fasel et al 2007) In the Apen-nines in the last few decades tree establishment hasbeen mainly controlled by land use while treegrowth has been controlled by climate pointing to aminor role played by climate in shaping the currenttreeline (Palombo et al 2013)
The impact of climate change and the connectedland use change on biodiversity and ecosystem serv-ices provision in several European countries is sum-marized by Wielgolaski et al (2017) and Kyriazopou-los et al (2017) (both this Special) and Sarkki et al(2016) One of the possible adaptive management op-tions in response to climate change assisted migra-tion as human-assisted movement of species hasbeen frequently debated in the last few years (Ste-Marie et al 2011) It is possible to apply it as a type ofassisted colonization ie intentional movement andrelease of an organism outside its indigenous range toavoid extinction of populations of the focal species(eg Macedonian pine species) or as an ecologicalre placement ie the intentional movement and re-lease of an organism outside its indigenous range toperform a specific ecological function (eg planting ofPinus mugo in the Alps IUCNSSC 2013)
5 CONCLUSIONS
The analysis of 11 mountain areas across Europeshowed that with increasing latitude the treelineand altitudinal width of the treeline ecotone signifi-cantly decreases as do the significance of climaticand soil parameters as barriers against tree speciesshift Although temperature is the overwhelmingcontrolling factor of tree growth and establishment intemperate and boreal treeline ecotones late-sea-sonal drought might also play a driving role in Medi-terranean treeline ecotones Longitude was lessinfluential mostly affecting climate and land usechange effects on increased biodiversity loss as wellas the size of the area that forms one mesoclimateunit affecting altitudinal ecotone width The biggestpart of the commonly observed remaining variabilityin mountain vegetation near the treeline in Europeseems to be caused by geomorphological geologicalpedological and microclimatic variability in combina-tion with different land use history and the presentsocio-economic relations The observed differencesin climatic parameters between the mountain areasof interest in comparison with the reference period
1960minus1990 (09degC) have not explained the relativelyhigh differences in the rate of treeline shift per year(149 m) The predicted variability in temperatureincrease due to global warming (12minus26degC in 2050and 22minus53degC in 2100) could lead to a much biggerdifferentiation in treeline ecotone biodiversity andecosystem processes between southern and northernEuropean mountains in the future Therefore thesedifferences must be taken into account by scientistsand EU policy makers when formulating efficientadaptive forest management strategies for treelineecosystems at the European level (eg assistedmigration of adapted genotypes)
Acknowledgements This paper is based firstly upon workfrom the COST Action ES 1203 SENSFOR supported byCOST (European Cooperation in Science and Technologywwwcosteu) This international work was further sup-ported by projects granted by the Ministry of EducationYouth and Sports of the Czech Republic grant NPU ILO1415 and LD 14039 by the agency APVV SR under pro-jects APVV-14-0086 and APVV-15-0270 We acknowledgethe E-OBS dataset from the EU-FP6 project ENSEMBLES(httpensembles-eumetofficecom) and the data providersin the ECAampD project (wwwecadeu)
LITERATURE CITED
Alados CL Errea P Gartzia M Saiz H Escoacutes J (2014) Posi-tive and negative feedbacks and free-scale pattern distri-bution in rural-population dynamics PLOS ONE 9 e114561
Allen CD Macalady AK Chenchouni H Bachelet D andothers (2010) A global overview of drought and heat-induced tree mortality reveals emerging climate changerisks for forests For Ecol Manag 259 660minus684
Amidžic L Bartula M Jaric S (2012) Syntaxonomy overviewof vegetation of Šar Planina in the Balkan PeninsulaVegetos 25 348minus360
Aune S Hofgaard A Soumlderstroumlm L (2011) Contrasting cli-mate and land use driven tree encroachment pattern ofsub-arctic tundra in Northern Norway and Kola Penin-sula Can J For Res 41 437minus449
Barbeito I Dawes M Rixen C Senn J Bebi P (2012) Factorsdriving mortality and growth at treeline a 30-year exper-iment of 92000 conifers Ecology 93 389minus401
Batllori E Camarero JJ Ninot JM Gutieacuterrez E (2009)Seedling recruitment survival and facilitation in alpinePinus uncinata treeline ecotones Implications andpotential responses to climate warming Glob Ecol Bio-geogr 18 460minus472
Bebi P Teich M Hagedorn F Zurbriggen N Brunner SHGreacutet-Regamey A (2012) Veraumlnderung von Wald undWaldleistungen in der Landschaft Davos im Zuge desKlimawandels Schweiz Z Forstwes 163 493minus501
Bell DM Bradford JB Lauenroth WK (2014) Mountain land-scapes offer few opportunities for high elevation treespecies migration Glob Change Biol 20 1441minus1451
Bergant K (2006) Climate change scenarios for MacedoniaUniversity of Nova Gorica Center for Atmospheric
147
Clim Res 73 135ndash150 2017
Research Nova GoricaBodin J Badeau V Bruno E Cluzeau C Moisselin JM
Walther GR Dupouey JL (2013) Shifts of forest speciesalong an elevational gradient in Southeast France cli-mate change or stand maturation J Veg Sci 24 269minus283
Boncina A (2011) History current status and future pros -pects of uneven-aged forest management in the Dinaricregion an overview Forestry 84 467minus478
Breshears DD Cobb NS Rich PM Price KP and others(2005) Regional vegetation die-off in response to global-change-type drought Proc Natl Acad Sci USA 102 15144minus15148
Camarero JJ Garciacutea-Ruiz JM Sanguumlesa-Barreda G GalvaacutenJD and others (2015a) Recent and intense dynamics in aformerly static Pyrenean treeline Arct Antarct Alp Res47 773minus783
Camarero JJ Gazol A Galvaacuten JD Sanguumlesa-Barreda GGutieacuterrez E (2015b) Disparate effects of global-changedrivers on mountain conifer forests warming-inducedgrowth enhancement in young trees vs CO2 fertilizationin old trees from wet sites Glob Change Biol 21 738minus749
Chen IC Hill JK Ohlemuumlller R Roy DB Thomas CD (2011)Rapid range shifts of species associated with high levelsof climate warming Science 333 1024minus1026
Cudliacuten P Sejaacutek J Pokornyacute J Albrechtovaacute J Bastian OMarek M (2013) Forest ecosystem services under climatechange and air pollution In Matyssek R Clarke N Cud-liacuten P Mikkelsen TN Tuovinen JP Wiesner G Paoletti E(eds) Climate change air pollution and global chal-lenges understanding and perspectives from forestresearch Developments in Environmental Science Vol13 Elsevier Oxford p 521minus546
Dalen L Hofgaard A (2005) Differential regional treelinedynamics in the Scandes mountains Arct Antarct AlpRes 37 284minus296
Dawes MA Philipson CD Fonti P Bebi P Haumlttenschwiler SHagedorn F Rixen C (2015) Soil warming and CO2
enrichment induce biomass shifts in alpine tree line veg-etation 2014 Glob Change Biol 21 2005minus2021
De Frenne P Rodriacuteguez-Saacutenchez F De Schrijver A CoomesDA Hermy M Vangansbeke P Verheyen K (2015) Lightaccelerates plant responses to warming Nat Plants 1 15110
Devi N Hagedorn F Moiseev P Bugmann H ShiyatovSG Mazepa V Rigling A (2008) Expanding forests andchanging growth forms of Siberian larch at the treeline ofthe Polar Urals during the 20th century Glob ChangeBiol 14 1581minus1591
Diaz-Varela R Colombo R Meroni M Calvo-Iglesias MSBuffoni A Tagliaferri A (2010) Spatio-temporal analysisof alpine ecotones A spatial explicit model targeting altitudinal vegetation shifts Ecol Model 221 621minus633
Dirnboumlck T Dullinger S Grabherr G (2003) A regionalimpact assessment of climate and land-use change onalpine vegetation J Biogeogr 30 401minus417
Dobrowski SZ Swanson AK Abatzoglou JT Holden ZASafford HD Schwartz MK Gavin DG (2015) Forest struc-ture and species traits mediate projected recruitmentdeclines in western US tree species Glob Ecol Biogeogr24 917minus927
Dullinger S Dirnboumlck S Grabherr G (2004) Modelling cli-mate change-driven treeline shifts relative effects oftemperature increase dispersal and invisibility J Ecol92 241minus252
Dullinger S Gattringer A Thuiller W Moser D and others
(2012) Extinction debt of high-mountain plants undertwenty-first-century climate change Nat Clim Chang 2 619minus622
Elkin C Gutierrez AG Leuzinger S Manusch C TemperliC Rasche L Bugmann H (2013) A 2degC warmer world isnot safe for ecosystem services in the European AlpsGlob Change Biol 19 1827minus1840
Ellenberg H (1988) Vegetation ecology of Central Europe4th edn Cambridge University Press Edinburgh
Em H (1986) Spruce southern distribution range Spruce for-est on Shara Mountain in Macedonia Contrib Macedon-ian Acad Sci Arts Skopje 5 11minus28
Fajardo A Piper FI Pfund L Koumlrner CH Hoch G (2012)Variation of mobile carbon reserves in trees at the alpinetreeline ecotone is under environmental control NewPhytol 195 794minus802
Fridley JD Grime JP Askew AP Moser B Stevens CJ (2011)Soil heterogeneity buffers community response to cli-mate change in species-rich grassland Glob Change Biol17 2002minus2011
Gartzia M Alados CL Peacuterez-Cabello F (2014) Assessment ofthe effects of biophysical and anthropogenic factors onwoody plant encroachment in dense and sparse moun-tain grasslands based on remote sensing data Prog PhysGeogr 38 201minus217
Gehrig-Fasel J Guisan A Zimmermann NE (2007) Tree lineshifts in the Swiss Alps climate change or land abandon-ment J Veg Sci 18 571minus582
Gikov A Dimitrov P Zhiyanski M (2016) Land cover dynam-ics on the northern slope of the Troyan Passage during 30years period Probl Geogr 1-2 78minus92
Gonzalez P Neilson RP Lenihan JM Drapek RJ (2010)Global patterns in the vulnerability of ecosystems to veg-etation shifts due to climate change Glob Ecol Biogeogr19 755minus768
Gottfried M Pauli H Futschik A Akhalkatsi M and others(2012) Continent-wide response of mountain vegetationto climate change Nat Clim Chang 2 111minus115
Grace J Berninger F Nagy L (2002) Impacts of climatechange on the tree line Ann Bot 90 537minus544
Grunewald K Scheithauer J (2011) Landscape developmentand climate change in Southwest Bulgaria (Pirin Mts)Springer Heidelberg
Grytnes JA Kapfer J Jurasinski G Birks HH and others(2014) Identifying the driving factors behind observedelevational range shifts on European mountains GlobEcol Biogeogr 23 876minus884
Hagedorn F Shiyatov FG Mazepa VS Devi NM and others(2014) Treeline advances along the Urals mountainrange mdash driven by improved winter conditions GlobChange Biol 20 3530minus3543
Hansen AJ Neilson RP Dale VH Flather CH and others(2001) Global change in forests responses of speciescommunities and biomes BioScience 51 765minus779
Harsch MA Bader MY (2011) Treeline formminusa potential keyto understanding treeline dynamics Glob Ecol Biogeogr20 582minus596
Harsch MA Hulme PE McGlone MS Duncan RP (2009) Aretreelines advancing A global meta-analysis of treelineresponse to climate warning Ecol Lett 12 1040minus1049
Haylock MR Hofstra N Klein Tank AMG Klok EJ JonesPD New M (2008) A European daily high-resolutiongridded dataset of surface temperature and precipita-tion J Geophys Res Atmos 113 D20119
Hlaacutesny T Barcza Z Fabrika M Balaacutezs B and others (2011)
148
Cudliacuten et al Drivers of treeline shift
Climate change impacts on growth and carbon balanceof forests in Central Europe Clim Res 47 219minus236
Hlaacutesny T Trombik J Dobor L Barcza Z Barka I (2016)Future climate of the Carpathians climate change hot-spots and implications for ecosystems Reg EnvironChange 16 1495minus1506
Hofgaard A Dalen L Hytteborn H (2009) Tree recruitmentabove the treeline and potential for climate driven tree-line change J Veg Sci 20 1133minus1144
Holtmeier FK (2009) Mountain timberlines Ecology patchi-ness and dynamics 2nd edn Advances in GlobalChange Research 36 Springer Science amp ScienceMediaBV Dordrecht
Holtmeier FK Broll G (2005) Sensitivity and response ofnorthern hemisphere altitudinal and polar treelines toenvironmental change at landscape and local scalesGlob Ecol Biogeogr 14 395minus410
Honnay O Verheyen K Butaye J Jacquemyn H Bossuyt BHermy M (2002) Possible effects of habitat fragmentationand climate change on the range of forest plant speciesEcol Lett 5 525minus530
Huber R Rigling A Bebi P Brand F and others (2013) Sus-tainable land use in mountain regions under globalchange synthesis across scales and disciplines Ecol Soc18 36
Ibaacutentildeez I Clark JS Dietze MC (2009) Estimating colonizationpotential of migrant tree species Glob Change Biol 15 1173minus1188
IUCNSSC (International Union for Conservation of NatureSpecies Survival Commission) (2013) Guidelines for re -introductions and other conservation translocations Version 10 IUCNSpecies Survival Commission Gland
Jankovskyacute L Cudliacuten P gtermaacutek P Moravec I (2004) The pre-diction of development of secondary Norway sprucestands under the impact of climatic change in the Dra-hany highlands (The Czech Republic) Ekologia (Bratisl)23(Suppl 2) 101minus112
Jeniacutek J Lokvenc T (1962) Die alpine Waldgrenze im Krko -noše Gebirge Rozpravy SAV Rada matematickyacutech apriacuterodniacutech ved 72 gteskoslovenskaacute akademie ved Praha
Jobbaacutegy EG Jackson RB (2000) Global controls of forest lineelevation in the northern and southern hemispheresGlob Ecol Biogeogr 9 253minus268
Jonaacutešovaacute M Vaacutevrovaacute E Cudliacuten P (2010) Western Carpa -thian mountain spruce forest after a windthrow naturalregeneration in cleared and uncleared areas For EcolManag 259 1127minus1134
Kittel TGF Steffen WL Chapin FS (2000) Global andregional modelling of Arcticminusboreal vegetation distribu-tion and its sensitivity to altered forcing Glob ChangeBiol 6 1minus18
Klopcic M Jerina K Bonltina A (2010) Long-term changes ofstructure and tree species composition in Dinaric un -even- aged forests Are red deer an important factor EurJ For Res 129 277minus288
Klopcic M Simonltilt T Bonltina A (2015) Comparison of re -generation and recruitment of shade-tolerant and light-demanding tree species in mixed uneven-aged forests ex -periences from the Dinaric region Forestry 88 552minus563
Koumlrner C (2003) Alpine plant life functional plant ecology ofhigh mountain ecosystems with 47 tables Springer Sci-ence amp Business Media BV Dordrecht
Koumlrner C (2012) Alpine treelines Functional ecology of theglobal high elevation tree limits Springer Basel
Koumlrner C Paulsen J (2004) A world-wide study of high alti-
tude treeline temperatures J Biogeogr 31 713minus732Kulakowski D Bebi P Rixen C (2011) The interacting effects
of land use change climate change and suppression ofnatural disturbances on landscape forest structure in theSwiss Alps Oikos 120 216minus225
Kullman L (1988) Holocene history of the forestminusalpine tun-dra ecotone in the Scandes Mountains (central Sweden)New Phytol 108 101minus110
Kullman L (1999) Early holocene tree growth at a high elevation site in the northernmost Scandes of Sweden(Lapland) a palaeobiogeographical case study based onmega fossil evidence Geogr Ann 81 63minus74
Kullman L (2007) Tree line population monitoring of Pinussylvestris in the Swedish Scandes 1973minus2005 implica-tions for tree line theory and climate change ecologyJ Ecol 95 41minus52
Kullman L Oumlberg L (2009) Post-Little Ice Age treeline riseand climate warming in the Swedish Scandes a land-scape ecological perspective J Ecol 97 415minus429
Kyriazopoulos AP Skre O Sarkki S Wielgolaski FE Abra-ham EM Ficko A (2017) Humanminusenvironment dynamicsin European treeline ecosystems a synthesis based onthe DPSIR framework Clim Res 73 17ndash29
Lenoir J Svenning JC (2015) Climate-related range shifts mdasha global multidimensional synthesis and new researchdirections Ecography 38 15minus38
Lenoir J Geacutegout JC Marquet PA de Ruffray P Brisse H(2008) A significant upward shift in plant species opti-mum elevation during the 20th century Science 320 1768minus1771
Lindner M Fitzgerald JB Zimmermann NE Reyer C andothers (2014) Climate change and European forests What do we know what are the uncertainties and whatare the implications for forest management J EnvironManag 146 69minus83
Lloyd AH Rupp TS Fastie CL Srafield AM (2002) Patternsand dynamic of treeline advance on the Seward Pen -isula Alaska J Geophys Res 107 8161
Maacuteliš F Kopeckyacute M Petriacutek P Vladovic J Merganic J VidaT (2016) Life-stage not climate change explains ob -served tree range shifts Glob Change Biol 22 1904minus1914
Mathisen IE Mikheeva A Tutubalina OV Aune S and Hof-gaard A (2014) Fifty years of tree line change in theKhibiny Mountains Russia advantages of combinedremote sensing and dendroecological approaches ApplVeg Sci 17 6ndash16
Millar CI Stephenson NL (2015) Temperate forest healthin an era of emerging megadisturbance Science 349 823minus826
Mina M Bugmann H Klopcic M Cailleret M (2017) Accu-rate modeling of harvesting is key for projecting futureforest dynamics a case study in the Slovenian moun-tains Reg Environ Change 17 49minus64
Moen J Aune K Edenius L Angerbjoumlrn A (2004) Potentialeffects of climate change on treeline position in theSwedish mountains Ecol Soc 9 16 wwwecologyandsocietyorgvol9iss1art16
Palombo C Chirici G Marchetti M Tognetti R (2013) Is landabandonment affecting forest dynamics at high elevationin Mediterranean mountains more than climate changePlant Biosyst 147 1minus11
Palombo C Marchetti M Tognetti R (2014) Mountain vege-tation at risk current perspectives and research reedsPlant Biosyst 148 35minus41
Pauli H Gottfried M Dullinger S Abdaladze O Akhalkatsi
149
Clim Res 73 135ndash150 2017
M Alonso JLB Grabherr G (2012) Recent plant diversitychanges on Europersquos mountain summits Science 336 353minus355
Payette S Fortin MJ Gamachet I (2001) The subarctic forestminustundra the structure of a biome in a changing climate BioScience 51 709minus718
Pearson GR Dawson TP (2003) Predicting the impacts of cli-mate change on the distribution of species Are bio -climate envelope models useful Glob Ecol Biogeogr 12 361minus371
Pedrotti F (2013) Plant and vegetation mapping GeobotanyStudies Springer-Verlag Berlin
Pentildeuelas J Boada M (2003) A global change-induced biomeshift in the Montseny mountains (NE Spain) GlobChange Biol 9 131minus140
Piovesan G Biondi F Filippo AD Alessandrini A MaugeriM (2008) Drought driven growth reduction in old beech(Fagus sylvatica L) forests of the central ApenninesItaly Glob Change Biol 14 1265minus1281
Rabasa SG Granda E Benavides R Kunstler G and others(2013) Disparity in elevational shifts of European trees inresponse to recent climate warming Glob Change Biol19 2490minus2499
Raev I Zhelev P Grozeva M Markov I and others (2011)Programme of measures for adaptation of forests in Bul-garia and mitigate the negative impact of climate changeon them Project FUTURE forest helping Europe tackleclimate change INTERREG IVC ERDF Sofia
Sarkki S Ficko A Grunewald K Nijnik M (2016) Benefitsfrom and threats to European treeline ecosystem serv-ices an exploratory study of stakeholders and gover-nance Reg Environ Change 16 2019minus2032
Savage J Vellend M (2015) Elevational shifts biotic homo -genization and time lags in vegetation change during 40years of climate warming Ecography 38 546minus555
Schwoumlrer C Henne PD Tinner W (2014) A model-data com-parison of Holocene timberline changes in the SwissAlps reveals past and future drivers of mountain forestdynamics Glob Change Biol 20 1512minus1526
Šmilauer P Lepš J (2014) Multivariate analysis of ecologicaldata using Canoco 5 Cambridge University Press Cam-bridge
Ste-Marie C Nelson EA Dabros A Bonneau ME (2011)Assisted migration introduction to a multifaceted con-cept For Chron 87 724minus730
Stevens JT Stafford HD Harrison S Latimer AM (2015) For-est disturbance accelerates thermophilization of under-story plant communities J Ecol 103 1253minus1263
Strid A Andonoski A Andonovski V (2003) Alpine biodiver-sity in Europe Ecological Studies 167 Springer-VerlagBerlin
Tattoni C Ciolli M Ferretti F Cantiani MG (2010) Monitor-ing spatial and temporal pattern of Paneveggio forest(Northern Italy) from 1859 to 2006 iForest Biogeosci For3 72minus80
Ter Braak CJF Šmilauer P (2012) Canoco reference manualand userrsquos guide software for ordination (version 50)Microcomputer Power Ithaca NY
Theurillat JP Guisan A (2001) Potential impact of climatechange on vegetation in the European Alps a reviewClim Change 50 77minus109
Toslashmmervik H Wielgolaski FE Neuvonen S Solberg BHoslashgda KA (2005) Biomass and production on a land-scape level in the mountain birch forests In WielgolaskiFE (ed) Plant ecology herbivory and human impact in
Nordic mountain birch forests Ecos Studies 180Springer Berlin p 53minus70
Treml V Banaš M (2000) Alpine timberline in the HighSudetes Acta Univ Carol Geogr 15 83minus99
Treml V Chuman T (2015) Ecotonal dynamics of the altitu-dinal forest limit are affected by terrain and vegetationstructure variables an example from the Sudetes Moun-tains in Central Europe Arct Antarct Alp Res 47 133minus146
Treml V Migo P (2015) controlling factors limiting timber-line position and shifts in the Sudetes a review GeogrPol 88 55minus70
Urban MC Tewksbury JJ Sheldon KS (2012) On a collisioncourse competition and dispersal differences create no-analogue communities and cause extinctions during cli-mate change Proc R Soc Lond B Biol Sci 279 2072minus2080
Vacek S Matejka K (2010) State and development of phyto-cenoses on research plots in the Krkonoše Mts foreststands J For Sci 56 505minus517
Van Bogaert R Haneca K Hoogesteger J Jonasson C deDapper M Callaghan TV (2011) A century of tree linechanges in sub-Arctic Sweden shows local and regionalvariability and only a minor influence of 20th century climate warming J Biogeogr 38 907minus921
Van Gils H Batsukh O Rossiter D Munthali W Liberato -scioli E (2008) Forecasting the pattern and pace of Fagusforest expansion in Majella National Park Italy ApplVeg Sci 11 539minus546
Velchev V (1997) Types of vegetation In Yordanova MDonchev D (eds) Geography of Bulgaria BAN Sofiap 269minus283 (in Bulgarian)
Vladovic J Merganic J Maacuteliš F Križovaacute E and others (2014)The response of forest vegetation diversity to changes inedaphic-climatic conditions in Slovakia Technickaacute uni-verzita vo Zvolene Zvolene (in Slovak)
Vrška T Adam D Hort L Kolaacuter T Janiacutek F (2009) Europeanbeech (Fagus sylvatica L) and silver fir (Abies alba Mill)rotation in the Carpathians mdash a developmental cycle or alinear trend induced by man For Ecol Manag 258 347minus356
Wielgolaski FE (ed) (2005) Plant ecology herbivory andhuman impact in Nordic mountain birch forests Eco -logical Studies 180 Springer Verlag Berlin
Wielgolaski FE Hofgaard A Holtmeier FK (2017) Sensitivityto environmental change of the treeline ecotone and itsassociated biodiversity in European mountains Clim Res73151ndash166
Wieser G Tausz M (2007) Trees at their upper limit Treelifelimitation at the alpine timberline Springer Dordrecht
Wieser G Holtmeier FK Smith WK (2014) Treelines in achanging global environment In Tausz M Grulke N(eds) Trees in a changing environment Plant Ecophys 9Springer Dordrecht p 221minus263
Zhiyanski M Kolev K Sokolovska M Hursthouse A (2008)Tree species effect on soils in Central Stara PlaninaMountains Nauka Gorata 4 65minus82
Zhiyanski M Gikov A Nedkov S Dimitrov P Naydenova L(2016) Mapping carbon storage using land coverlanduse data in the area of Beklemeto Central Balkan In Koulov B Zhelezov G (eds) Sustainable mountain re -gions challenges and perspectives in southeasternEurope Springer Basel p 53minus65
Zhu K Woodall CW Clark JS (2012) Failure to migrate lackof tree range expansion in response to climate changeGlob Change Biol 18 1042minus1052
150
Editorial responsibility Nils Chr Stenseth Oslo Norway
Submitted April 8 2016 Accepted March 8 2017Proofs received from author(s) July 7 2017
Clim Res 73 135ndash150 2017
Gehrig-Fasel et al 2007) limits treeline shift in halfof the countries of interest regardless of latitude lon-gitude or recent political developments (Shara PirinCentral Balkans Alps Scandes Fig 2)
Despite the stated differences between studieslooking at climate change impacts on the treelineit is clear that current and future changes in temp -erature will seriously affect treeline ecotones in allmountain ranges of Europe Winter is a period whentreeline stands and individual trees have to survivesevere life-limiting conditions (Wieser amp Tausz2007) Therefore winter warming might increase thechance of young trees surviving and passing the mostcritical period from seedling to sapling and further tothe mature stage (Koumlrner 2003) There is alreadymuch evidence worldwide that winter warming isone of the significant drivers of treeline advance(Harsch et al 2009) Although it is expected that treegrowth at the upper distributional margins (and inthe northern European countries) will increase due toongoing climate change (Pentildeuelas amp Boada 2003Chen et al 2011 Hlaacutesny et al 2011 Lindner et al2014) extremes in temperature (eg black froststemperature reversals in the spring) or winter precip-itation (eg lack of snow drought) might also limit
this process in the future in some regions and localsituations (Holtmeier amp Broll 2005 Hagedorn et al2014) An earlier start to the growing period (espe-cially in the Apennines Central Balkan and SpanishPyrenees see Table 2) can play a negative role in theresistance of trees and seedlings to early springfrosts On the other hand at lower distributional mar-gins (and in Southern European countries) it isexpected that tree growth will decrease or forestswill experience some level of dieback due to drought(Breshears et al 2005 Piovesan et al 2008 Allen etal 2010 Huber et al 2013) A serious implication isthe positive effect that warming can have on root rotfungi and populations of bark beetles resulting inlarge-scale disturbances to temperate and also high-altitude forests (Jankovskyacute et al 2004 Elkin et al2013 Millar amp Stephenson 2015)
Although recent upward advances of treeline eco-tones are widespread in mountain regions (Harsch et
144
Fig 3 Redundancy analysis diagram with variation in tim-berline elevation width of treeline ecotone the tree speciesforming the treeline and treeline shift used as dependentvariables explained by explanatory variables (latitudemountain size) Mountain units (blue) are projected as thecentres of abbreviations (see Table 2) Explanatory variablesaccount for 436 of the total variation in the dependentdata The first canonical axis explained 466 of variationthe second axis explained 30 of variation Dependentvariables (black) are Ecot altitudinal width of treeline eco-tone TLShift treeline shift TrspC (TrspB) treeline formedby conifers (broadleaved trees) Timb timberline elevationExplanatory variables (red) are N north latitude Mtsize
size of the mountain massif
Fig 4 Redundancy analysis diagram with variation in bio-diversity loss in forests meadows and animal communitiesused as dependent variables explained by explanatory vari-ables (geographical position tree species forming the tree-line and temperature increase between the 2 study periods[1961minus1990 and 2021minus2050]) Mountain units (green) areprojected as centres of abbreviations (see Table 2) Explana-tory variables account for 763 of total variation in the de-pendent data The first canonical axis explained 484 ofvariation the second axis explained 183 of variation De-pendent variables (black) are BdfoBdmeBdan biodiversityloss in forestsmeadowsanimal communities Explanatoryvariables (red) are N north latitude E east longitude TrspC(TrspB) treeline formed by conifers (broadleaved trees)Temp temperature increase between the 2 study periods
Cudliacuten et al Drivers of treeline shift
al 2009) only a few published studies have includedquantitative data about treeline shifts (Devi et al2008 Kullman amp Oumlberg 2009 Diaz-Varela et al 2010Van Bogaert et al 2011) Additionally our know -ledge of the spatial patterns in treeline ecotone shiftsat the landscape scale is still surprisingly poor exceptfor some studies from subarctic areas the UralMountains and the Alps (Lloyd et al 2002 Diaz-Varela et al 2010 Hagedorn et al 2014) In the 15studied mountain units the values of treeline shiftranged from 043 to 19 m yrminus1 showing a rather dis-tinct spatial pattern of the dynamics within Europeanmountains The observed treeline shift had a signifi-cant positive relationship only with northern latitudeand a weak positive relationship with mountain-range size (ie the size of the whole mountain massifFig 3) Nevertheless this illustrates the importanceof the mass elevation effect (Koumlrner 2012) and par-tially ex plains why treelines could be at different alti-tudes at the same latitude Small negative regres-sions with east longitude and timberline elevationare shown in Fig 5 There was no apparent depend-ence of the treeline shift rate on climatic parameterchanges between the periods 1961minus1990 and1991minus2015
A whole forest vegetation zone shift is a complexprocess as not only the life strategies of an individualtree need to be considered but plant animal andmicroorganism species and their interrelationshipsas well as the relationships to their specific microhab-
itats (including soil conditions) are also involved(Urban et al 2012) Across all regions we identifiedseveral important obstacles to treeline shifts oftenrelated to site properties such as significant rocki-ness having shallow or low-nutrient soils extremerelief causing disturbances by avalanches snow glid-ing or wind damage (data from the Pyrenees Apen-nines Shara Northern Dinaric Low Tatra and GiantMountains see Fig 2) Soil heterogeneity certainlyhas an important role in plant responses to climatechange and could also maintain the resilience of thecommunity assemblages (Fridley et al 2011) An -other group of obstacles includes extreme climaticparameters especially in winter (reported eg fromthe Pyrenees Apennines Pirin Central BalkanNorthern Dinaric and Giant Mountains Fig 2) Theircombination can result in edaphic andor climaticunsuitability of habitats where species could poten-tially migrate Climatically and edaphically suitablesites will likely decline over the next century partic-ularly in mountain landscapes (Bell et al 2014)There fore trees in the treeline ecotone will colonizepreviously forested habitats The colonization suc-cess of individual forest communities is affected bydifferences in species dispersal and recruitmentbehaviour (Dullin ger et al 2004 Jonaacutešovaacute et al2010) For these reasons colonization of new foresthabitats by the next tree generation may not be suc-cessful and may result in loss of species diversity(Honnay et al 2002 Ibaacutentildeez et al 2009 Dobrowski et
145
Fig 5 Univariate graphs of the relationships of dependent variables viz timberline (Timb) and treeline shift per year (Shift) and their explanatory variables Nola north latitude Ealo east longitude DD decimal degrees
Clim Res 73 135ndash150 2017
al 2015) as reported from Macedonia Slovenia theCzech Republic Slovakia and Norway (Fig 4)
Another limiting factor is the existence of strongadaptation mechanisms of the dominant tree specieswhich allow them to survive in their current distribu-tion areas Species migration is likely to be slow dueto the limited quantity of climatically and edaphicallysuitable sites and the slow velocity of seed dispersaland tree regeneration and will be hampered evenmore by fragmentation of high mountain landscapescaused by human activities including brush invasionin abandoned meadows (Gartzia et al 2014) A verti-cal shift in a vegetation zone involves not only singlespecies of plants animals and microorganisms butalso their interrelationships and links to soil condi-tions The speed at which climatic conditions changewill be different from changes in the soil conditionsdue to the slowness of soil-forming processes Soiltypes and conditions (cambisol versus podzol) arecrucial obstacles for the shift from a beech forest zoneinto the spruce forest zone in the Czech Republic(Vacek amp Mat jka 2010) According to other sourcesthe dominant tree species can influence soil for -mation processes (especially humus forms) Underfavourable climatic and orographic conditions beechis able to change its soil conditions over decades tocenturies (J Macku unpubl data)
Significant changes in biodiversity must be expec -ted in all of the mountain areas of interest We foundthat biodiversity declined particularly in the southernregions of Europe where the timberline is situated athigher elevations and human impact is mostlylonger-lasting (Fig 4) Similarly Pauli et al (2012)reported an increase in species richness of mountaingrasslands across Europe except for Mediterraneanregions and assigned this different response of thesouthern regions to decreased water availability Onthe other hand ecosystems which exhibited bio -diversity increases were not only enriched by mig -rant species but the assemblages underwent thermo -philization ie cold-adapted species declined andmore warm-adapted species spread (Gottfried et al2012) However not all species are able to track cli-mate changes It is expected that around 40 ofhabitats will become climatically unsuitable for manymountain species particularly endemic ones in -creasing their extinction probability during this cen-tury (Dullinger et al 2012) The abandonment of tra-ditional land use forms is another source of this losswhich may be a more important driver than tempera-ture increase (Fig 4) Serious biodiversity losses inmountain meadows were reported from Spain ItalyMacedonia Bulgaria Slovenia and Slovakia There-
fore controlled grazing must occur in order to main-tain alpine grasslands (Dirnboumlck et al 2003) Unfor-tunately grazing is still active only in smaller parts ofEuropean mountains (eg in the Central Alps and theCentral Balkan Mountains but pasturing can alsonegatively affect vegetation diversity) and some-times does not serve to maintain alpine grasslandThe same is true for grassland management whichcan help to maintain mountain species and decreasehabitat loss A serious biodiversity loss in mountainmeadows was reported from the Pyrenees the Apen-nines and the Shara Pirin Central Balkan NorthernDinaric and Low Tatra Mountains (Fig 4)
In forests species responses to climate change maybe equivocal some recent studies indicated contra-dictory shifts in species distributions (eg Lenoir etal 2008 Zhu et al 2012 Rabasa et al 2013) This dis-parity is widely discussed and assigned to the greatvariety of non-climatic factors or even tree ontogeny(Grytnes et al 2014 Lenoir amp Svenning 2015 Maacutelišet al 2016) Disturbances leading to tree mortalitymay also play an important role (Cudliacuten et al 2013)changes in tree canopy cover modify light availabil-ity and microclimate and can induce the thermo -philization of forest vegetation (De Frenne et al2015 Stevens et al 2015) The loss of this microcli-mate buffering effect of forests may induce a biotichomogenization of forests (Savage amp Vellend 2015)or the creation of novel non-analogical communitiessuch as oakminuspine forests (Urban et al 2012) whichmay be a new threat to forest biodiversity
The observed changes in treeline forest ecosystemsare often related to changes in land-use intensity(Theurillat amp Guisan 2001 Alados et al 2014) Ac -cording to climate change predictions and the recentand future exploitation intensity of European moun-tains trees and forest communities will shift upwarddue to land use change or climate change or bothReduced land-use intensity certainly will interactwith climate change by facilitating or inhibiting spe-cies occurrence and will accelerate forest expansionabove the present treeline particularly to previouslyforested habitats (Theurillat amp Guisan 2001) Thesimultaneous action of both main treeline shift driv-ers viz temperature increase and decrease in landuse intensity was recorded from all of our studiedmountain areas The extensive differences in timber-line and treeline elevations in almost all studiedmountains (Table 1) indicate the anthropo-zoogenictreeline type (according to Ellenberg 1998) In mostmountains (eg in the Apennines Shara MountainsCentral Balkans and Alps) land use is the prevailingfactor influencing vegetation drift (Fig 5) Previous
146
Cudliacuten et al Drivers of treeline shift
research showed that the upward shift of the treelinein the Swiss Alps was predominantly attributable toland abandonment and only in some situations to cli-mate change (Gehrig-Fasel et al 2007) In the Apen-nines in the last few decades tree establishment hasbeen mainly controlled by land use while treegrowth has been controlled by climate pointing to aminor role played by climate in shaping the currenttreeline (Palombo et al 2013)
The impact of climate change and the connectedland use change on biodiversity and ecosystem serv-ices provision in several European countries is sum-marized by Wielgolaski et al (2017) and Kyriazopou-los et al (2017) (both this Special) and Sarkki et al(2016) One of the possible adaptive management op-tions in response to climate change assisted migra-tion as human-assisted movement of species hasbeen frequently debated in the last few years (Ste-Marie et al 2011) It is possible to apply it as a type ofassisted colonization ie intentional movement andrelease of an organism outside its indigenous range toavoid extinction of populations of the focal species(eg Macedonian pine species) or as an ecologicalre placement ie the intentional movement and re-lease of an organism outside its indigenous range toperform a specific ecological function (eg planting ofPinus mugo in the Alps IUCNSSC 2013)
5 CONCLUSIONS
The analysis of 11 mountain areas across Europeshowed that with increasing latitude the treelineand altitudinal width of the treeline ecotone signifi-cantly decreases as do the significance of climaticand soil parameters as barriers against tree speciesshift Although temperature is the overwhelmingcontrolling factor of tree growth and establishment intemperate and boreal treeline ecotones late-sea-sonal drought might also play a driving role in Medi-terranean treeline ecotones Longitude was lessinfluential mostly affecting climate and land usechange effects on increased biodiversity loss as wellas the size of the area that forms one mesoclimateunit affecting altitudinal ecotone width The biggestpart of the commonly observed remaining variabilityin mountain vegetation near the treeline in Europeseems to be caused by geomorphological geologicalpedological and microclimatic variability in combina-tion with different land use history and the presentsocio-economic relations The observed differencesin climatic parameters between the mountain areasof interest in comparison with the reference period
1960minus1990 (09degC) have not explained the relativelyhigh differences in the rate of treeline shift per year(149 m) The predicted variability in temperatureincrease due to global warming (12minus26degC in 2050and 22minus53degC in 2100) could lead to a much biggerdifferentiation in treeline ecotone biodiversity andecosystem processes between southern and northernEuropean mountains in the future Therefore thesedifferences must be taken into account by scientistsand EU policy makers when formulating efficientadaptive forest management strategies for treelineecosystems at the European level (eg assistedmigration of adapted genotypes)
Acknowledgements This paper is based firstly upon workfrom the COST Action ES 1203 SENSFOR supported byCOST (European Cooperation in Science and Technologywwwcosteu) This international work was further sup-ported by projects granted by the Ministry of EducationYouth and Sports of the Czech Republic grant NPU ILO1415 and LD 14039 by the agency APVV SR under pro-jects APVV-14-0086 and APVV-15-0270 We acknowledgethe E-OBS dataset from the EU-FP6 project ENSEMBLES(httpensembles-eumetofficecom) and the data providersin the ECAampD project (wwwecadeu)
LITERATURE CITED
Alados CL Errea P Gartzia M Saiz H Escoacutes J (2014) Posi-tive and negative feedbacks and free-scale pattern distri-bution in rural-population dynamics PLOS ONE 9 e114561
Allen CD Macalady AK Chenchouni H Bachelet D andothers (2010) A global overview of drought and heat-induced tree mortality reveals emerging climate changerisks for forests For Ecol Manag 259 660minus684
Amidžic L Bartula M Jaric S (2012) Syntaxonomy overviewof vegetation of Šar Planina in the Balkan PeninsulaVegetos 25 348minus360
Aune S Hofgaard A Soumlderstroumlm L (2011) Contrasting cli-mate and land use driven tree encroachment pattern ofsub-arctic tundra in Northern Norway and Kola Penin-sula Can J For Res 41 437minus449
Barbeito I Dawes M Rixen C Senn J Bebi P (2012) Factorsdriving mortality and growth at treeline a 30-year exper-iment of 92000 conifers Ecology 93 389minus401
Batllori E Camarero JJ Ninot JM Gutieacuterrez E (2009)Seedling recruitment survival and facilitation in alpinePinus uncinata treeline ecotones Implications andpotential responses to climate warming Glob Ecol Bio-geogr 18 460minus472
Bebi P Teich M Hagedorn F Zurbriggen N Brunner SHGreacutet-Regamey A (2012) Veraumlnderung von Wald undWaldleistungen in der Landschaft Davos im Zuge desKlimawandels Schweiz Z Forstwes 163 493minus501
Bell DM Bradford JB Lauenroth WK (2014) Mountain land-scapes offer few opportunities for high elevation treespecies migration Glob Change Biol 20 1441minus1451
Bergant K (2006) Climate change scenarios for MacedoniaUniversity of Nova Gorica Center for Atmospheric
147
Clim Res 73 135ndash150 2017
Research Nova GoricaBodin J Badeau V Bruno E Cluzeau C Moisselin JM
Walther GR Dupouey JL (2013) Shifts of forest speciesalong an elevational gradient in Southeast France cli-mate change or stand maturation J Veg Sci 24 269minus283
Boncina A (2011) History current status and future pros -pects of uneven-aged forest management in the Dinaricregion an overview Forestry 84 467minus478
Breshears DD Cobb NS Rich PM Price KP and others(2005) Regional vegetation die-off in response to global-change-type drought Proc Natl Acad Sci USA 102 15144minus15148
Camarero JJ Garciacutea-Ruiz JM Sanguumlesa-Barreda G GalvaacutenJD and others (2015a) Recent and intense dynamics in aformerly static Pyrenean treeline Arct Antarct Alp Res47 773minus783
Camarero JJ Gazol A Galvaacuten JD Sanguumlesa-Barreda GGutieacuterrez E (2015b) Disparate effects of global-changedrivers on mountain conifer forests warming-inducedgrowth enhancement in young trees vs CO2 fertilizationin old trees from wet sites Glob Change Biol 21 738minus749
Chen IC Hill JK Ohlemuumlller R Roy DB Thomas CD (2011)Rapid range shifts of species associated with high levelsof climate warming Science 333 1024minus1026
Cudliacuten P Sejaacutek J Pokornyacute J Albrechtovaacute J Bastian OMarek M (2013) Forest ecosystem services under climatechange and air pollution In Matyssek R Clarke N Cud-liacuten P Mikkelsen TN Tuovinen JP Wiesner G Paoletti E(eds) Climate change air pollution and global chal-lenges understanding and perspectives from forestresearch Developments in Environmental Science Vol13 Elsevier Oxford p 521minus546
Dalen L Hofgaard A (2005) Differential regional treelinedynamics in the Scandes mountains Arct Antarct AlpRes 37 284minus296
Dawes MA Philipson CD Fonti P Bebi P Haumlttenschwiler SHagedorn F Rixen C (2015) Soil warming and CO2
enrichment induce biomass shifts in alpine tree line veg-etation 2014 Glob Change Biol 21 2005minus2021
De Frenne P Rodriacuteguez-Saacutenchez F De Schrijver A CoomesDA Hermy M Vangansbeke P Verheyen K (2015) Lightaccelerates plant responses to warming Nat Plants 1 15110
Devi N Hagedorn F Moiseev P Bugmann H ShiyatovSG Mazepa V Rigling A (2008) Expanding forests andchanging growth forms of Siberian larch at the treeline ofthe Polar Urals during the 20th century Glob ChangeBiol 14 1581minus1591
Diaz-Varela R Colombo R Meroni M Calvo-Iglesias MSBuffoni A Tagliaferri A (2010) Spatio-temporal analysisof alpine ecotones A spatial explicit model targeting altitudinal vegetation shifts Ecol Model 221 621minus633
Dirnboumlck T Dullinger S Grabherr G (2003) A regionalimpact assessment of climate and land-use change onalpine vegetation J Biogeogr 30 401minus417
Dobrowski SZ Swanson AK Abatzoglou JT Holden ZASafford HD Schwartz MK Gavin DG (2015) Forest struc-ture and species traits mediate projected recruitmentdeclines in western US tree species Glob Ecol Biogeogr24 917minus927
Dullinger S Dirnboumlck S Grabherr G (2004) Modelling cli-mate change-driven treeline shifts relative effects oftemperature increase dispersal and invisibility J Ecol92 241minus252
Dullinger S Gattringer A Thuiller W Moser D and others
(2012) Extinction debt of high-mountain plants undertwenty-first-century climate change Nat Clim Chang 2 619minus622
Elkin C Gutierrez AG Leuzinger S Manusch C TemperliC Rasche L Bugmann H (2013) A 2degC warmer world isnot safe for ecosystem services in the European AlpsGlob Change Biol 19 1827minus1840
Ellenberg H (1988) Vegetation ecology of Central Europe4th edn Cambridge University Press Edinburgh
Em H (1986) Spruce southern distribution range Spruce for-est on Shara Mountain in Macedonia Contrib Macedon-ian Acad Sci Arts Skopje 5 11minus28
Fajardo A Piper FI Pfund L Koumlrner CH Hoch G (2012)Variation of mobile carbon reserves in trees at the alpinetreeline ecotone is under environmental control NewPhytol 195 794minus802
Fridley JD Grime JP Askew AP Moser B Stevens CJ (2011)Soil heterogeneity buffers community response to cli-mate change in species-rich grassland Glob Change Biol17 2002minus2011
Gartzia M Alados CL Peacuterez-Cabello F (2014) Assessment ofthe effects of biophysical and anthropogenic factors onwoody plant encroachment in dense and sparse moun-tain grasslands based on remote sensing data Prog PhysGeogr 38 201minus217
Gehrig-Fasel J Guisan A Zimmermann NE (2007) Tree lineshifts in the Swiss Alps climate change or land abandon-ment J Veg Sci 18 571minus582
Gikov A Dimitrov P Zhiyanski M (2016) Land cover dynam-ics on the northern slope of the Troyan Passage during 30years period Probl Geogr 1-2 78minus92
Gonzalez P Neilson RP Lenihan JM Drapek RJ (2010)Global patterns in the vulnerability of ecosystems to veg-etation shifts due to climate change Glob Ecol Biogeogr19 755minus768
Gottfried M Pauli H Futschik A Akhalkatsi M and others(2012) Continent-wide response of mountain vegetationto climate change Nat Clim Chang 2 111minus115
Grace J Berninger F Nagy L (2002) Impacts of climatechange on the tree line Ann Bot 90 537minus544
Grunewald K Scheithauer J (2011) Landscape developmentand climate change in Southwest Bulgaria (Pirin Mts)Springer Heidelberg
Grytnes JA Kapfer J Jurasinski G Birks HH and others(2014) Identifying the driving factors behind observedelevational range shifts on European mountains GlobEcol Biogeogr 23 876minus884
Hagedorn F Shiyatov FG Mazepa VS Devi NM and others(2014) Treeline advances along the Urals mountainrange mdash driven by improved winter conditions GlobChange Biol 20 3530minus3543
Hansen AJ Neilson RP Dale VH Flather CH and others(2001) Global change in forests responses of speciescommunities and biomes BioScience 51 765minus779
Harsch MA Bader MY (2011) Treeline formminusa potential keyto understanding treeline dynamics Glob Ecol Biogeogr20 582minus596
Harsch MA Hulme PE McGlone MS Duncan RP (2009) Aretreelines advancing A global meta-analysis of treelineresponse to climate warning Ecol Lett 12 1040minus1049
Haylock MR Hofstra N Klein Tank AMG Klok EJ JonesPD New M (2008) A European daily high-resolutiongridded dataset of surface temperature and precipita-tion J Geophys Res Atmos 113 D20119
Hlaacutesny T Barcza Z Fabrika M Balaacutezs B and others (2011)
148
Cudliacuten et al Drivers of treeline shift
Climate change impacts on growth and carbon balanceof forests in Central Europe Clim Res 47 219minus236
Hlaacutesny T Trombik J Dobor L Barcza Z Barka I (2016)Future climate of the Carpathians climate change hot-spots and implications for ecosystems Reg EnvironChange 16 1495minus1506
Hofgaard A Dalen L Hytteborn H (2009) Tree recruitmentabove the treeline and potential for climate driven tree-line change J Veg Sci 20 1133minus1144
Holtmeier FK (2009) Mountain timberlines Ecology patchi-ness and dynamics 2nd edn Advances in GlobalChange Research 36 Springer Science amp ScienceMediaBV Dordrecht
Holtmeier FK Broll G (2005) Sensitivity and response ofnorthern hemisphere altitudinal and polar treelines toenvironmental change at landscape and local scalesGlob Ecol Biogeogr 14 395minus410
Honnay O Verheyen K Butaye J Jacquemyn H Bossuyt BHermy M (2002) Possible effects of habitat fragmentationand climate change on the range of forest plant speciesEcol Lett 5 525minus530
Huber R Rigling A Bebi P Brand F and others (2013) Sus-tainable land use in mountain regions under globalchange synthesis across scales and disciplines Ecol Soc18 36
Ibaacutentildeez I Clark JS Dietze MC (2009) Estimating colonizationpotential of migrant tree species Glob Change Biol 15 1173minus1188
IUCNSSC (International Union for Conservation of NatureSpecies Survival Commission) (2013) Guidelines for re -introductions and other conservation translocations Version 10 IUCNSpecies Survival Commission Gland
Jankovskyacute L Cudliacuten P gtermaacutek P Moravec I (2004) The pre-diction of development of secondary Norway sprucestands under the impact of climatic change in the Dra-hany highlands (The Czech Republic) Ekologia (Bratisl)23(Suppl 2) 101minus112
Jeniacutek J Lokvenc T (1962) Die alpine Waldgrenze im Krko -noše Gebirge Rozpravy SAV Rada matematickyacutech apriacuterodniacutech ved 72 gteskoslovenskaacute akademie ved Praha
Jobbaacutegy EG Jackson RB (2000) Global controls of forest lineelevation in the northern and southern hemispheresGlob Ecol Biogeogr 9 253minus268
Jonaacutešovaacute M Vaacutevrovaacute E Cudliacuten P (2010) Western Carpa -thian mountain spruce forest after a windthrow naturalregeneration in cleared and uncleared areas For EcolManag 259 1127minus1134
Kittel TGF Steffen WL Chapin FS (2000) Global andregional modelling of Arcticminusboreal vegetation distribu-tion and its sensitivity to altered forcing Glob ChangeBiol 6 1minus18
Klopcic M Jerina K Bonltina A (2010) Long-term changes ofstructure and tree species composition in Dinaric un -even- aged forests Are red deer an important factor EurJ For Res 129 277minus288
Klopcic M Simonltilt T Bonltina A (2015) Comparison of re -generation and recruitment of shade-tolerant and light-demanding tree species in mixed uneven-aged forests ex -periences from the Dinaric region Forestry 88 552minus563
Koumlrner C (2003) Alpine plant life functional plant ecology ofhigh mountain ecosystems with 47 tables Springer Sci-ence amp Business Media BV Dordrecht
Koumlrner C (2012) Alpine treelines Functional ecology of theglobal high elevation tree limits Springer Basel
Koumlrner C Paulsen J (2004) A world-wide study of high alti-
tude treeline temperatures J Biogeogr 31 713minus732Kulakowski D Bebi P Rixen C (2011) The interacting effects
of land use change climate change and suppression ofnatural disturbances on landscape forest structure in theSwiss Alps Oikos 120 216minus225
Kullman L (1988) Holocene history of the forestminusalpine tun-dra ecotone in the Scandes Mountains (central Sweden)New Phytol 108 101minus110
Kullman L (1999) Early holocene tree growth at a high elevation site in the northernmost Scandes of Sweden(Lapland) a palaeobiogeographical case study based onmega fossil evidence Geogr Ann 81 63minus74
Kullman L (2007) Tree line population monitoring of Pinussylvestris in the Swedish Scandes 1973minus2005 implica-tions for tree line theory and climate change ecologyJ Ecol 95 41minus52
Kullman L Oumlberg L (2009) Post-Little Ice Age treeline riseand climate warming in the Swedish Scandes a land-scape ecological perspective J Ecol 97 415minus429
Kyriazopoulos AP Skre O Sarkki S Wielgolaski FE Abra-ham EM Ficko A (2017) Humanminusenvironment dynamicsin European treeline ecosystems a synthesis based onthe DPSIR framework Clim Res 73 17ndash29
Lenoir J Svenning JC (2015) Climate-related range shifts mdasha global multidimensional synthesis and new researchdirections Ecography 38 15minus38
Lenoir J Geacutegout JC Marquet PA de Ruffray P Brisse H(2008) A significant upward shift in plant species opti-mum elevation during the 20th century Science 320 1768minus1771
Lindner M Fitzgerald JB Zimmermann NE Reyer C andothers (2014) Climate change and European forests What do we know what are the uncertainties and whatare the implications for forest management J EnvironManag 146 69minus83
Lloyd AH Rupp TS Fastie CL Srafield AM (2002) Patternsand dynamic of treeline advance on the Seward Pen -isula Alaska J Geophys Res 107 8161
Maacuteliš F Kopeckyacute M Petriacutek P Vladovic J Merganic J VidaT (2016) Life-stage not climate change explains ob -served tree range shifts Glob Change Biol 22 1904minus1914
Mathisen IE Mikheeva A Tutubalina OV Aune S and Hof-gaard A (2014) Fifty years of tree line change in theKhibiny Mountains Russia advantages of combinedremote sensing and dendroecological approaches ApplVeg Sci 17 6ndash16
Millar CI Stephenson NL (2015) Temperate forest healthin an era of emerging megadisturbance Science 349 823minus826
Mina M Bugmann H Klopcic M Cailleret M (2017) Accu-rate modeling of harvesting is key for projecting futureforest dynamics a case study in the Slovenian moun-tains Reg Environ Change 17 49minus64
Moen J Aune K Edenius L Angerbjoumlrn A (2004) Potentialeffects of climate change on treeline position in theSwedish mountains Ecol Soc 9 16 wwwecologyandsocietyorgvol9iss1art16
Palombo C Chirici G Marchetti M Tognetti R (2013) Is landabandonment affecting forest dynamics at high elevationin Mediterranean mountains more than climate changePlant Biosyst 147 1minus11
Palombo C Marchetti M Tognetti R (2014) Mountain vege-tation at risk current perspectives and research reedsPlant Biosyst 148 35minus41
Pauli H Gottfried M Dullinger S Abdaladze O Akhalkatsi
149
Clim Res 73 135ndash150 2017
M Alonso JLB Grabherr G (2012) Recent plant diversitychanges on Europersquos mountain summits Science 336 353minus355
Payette S Fortin MJ Gamachet I (2001) The subarctic forestminustundra the structure of a biome in a changing climate BioScience 51 709minus718
Pearson GR Dawson TP (2003) Predicting the impacts of cli-mate change on the distribution of species Are bio -climate envelope models useful Glob Ecol Biogeogr 12 361minus371
Pedrotti F (2013) Plant and vegetation mapping GeobotanyStudies Springer-Verlag Berlin
Pentildeuelas J Boada M (2003) A global change-induced biomeshift in the Montseny mountains (NE Spain) GlobChange Biol 9 131minus140
Piovesan G Biondi F Filippo AD Alessandrini A MaugeriM (2008) Drought driven growth reduction in old beech(Fagus sylvatica L) forests of the central ApenninesItaly Glob Change Biol 14 1265minus1281
Rabasa SG Granda E Benavides R Kunstler G and others(2013) Disparity in elevational shifts of European trees inresponse to recent climate warming Glob Change Biol19 2490minus2499
Raev I Zhelev P Grozeva M Markov I and others (2011)Programme of measures for adaptation of forests in Bul-garia and mitigate the negative impact of climate changeon them Project FUTURE forest helping Europe tackleclimate change INTERREG IVC ERDF Sofia
Sarkki S Ficko A Grunewald K Nijnik M (2016) Benefitsfrom and threats to European treeline ecosystem serv-ices an exploratory study of stakeholders and gover-nance Reg Environ Change 16 2019minus2032
Savage J Vellend M (2015) Elevational shifts biotic homo -genization and time lags in vegetation change during 40years of climate warming Ecography 38 546minus555
Schwoumlrer C Henne PD Tinner W (2014) A model-data com-parison of Holocene timberline changes in the SwissAlps reveals past and future drivers of mountain forestdynamics Glob Change Biol 20 1512minus1526
Šmilauer P Lepš J (2014) Multivariate analysis of ecologicaldata using Canoco 5 Cambridge University Press Cam-bridge
Ste-Marie C Nelson EA Dabros A Bonneau ME (2011)Assisted migration introduction to a multifaceted con-cept For Chron 87 724minus730
Stevens JT Stafford HD Harrison S Latimer AM (2015) For-est disturbance accelerates thermophilization of under-story plant communities J Ecol 103 1253minus1263
Strid A Andonoski A Andonovski V (2003) Alpine biodiver-sity in Europe Ecological Studies 167 Springer-VerlagBerlin
Tattoni C Ciolli M Ferretti F Cantiani MG (2010) Monitor-ing spatial and temporal pattern of Paneveggio forest(Northern Italy) from 1859 to 2006 iForest Biogeosci For3 72minus80
Ter Braak CJF Šmilauer P (2012) Canoco reference manualand userrsquos guide software for ordination (version 50)Microcomputer Power Ithaca NY
Theurillat JP Guisan A (2001) Potential impact of climatechange on vegetation in the European Alps a reviewClim Change 50 77minus109
Toslashmmervik H Wielgolaski FE Neuvonen S Solberg BHoslashgda KA (2005) Biomass and production on a land-scape level in the mountain birch forests In WielgolaskiFE (ed) Plant ecology herbivory and human impact in
Nordic mountain birch forests Ecos Studies 180Springer Berlin p 53minus70
Treml V Banaš M (2000) Alpine timberline in the HighSudetes Acta Univ Carol Geogr 15 83minus99
Treml V Chuman T (2015) Ecotonal dynamics of the altitu-dinal forest limit are affected by terrain and vegetationstructure variables an example from the Sudetes Moun-tains in Central Europe Arct Antarct Alp Res 47 133minus146
Treml V Migo P (2015) controlling factors limiting timber-line position and shifts in the Sudetes a review GeogrPol 88 55minus70
Urban MC Tewksbury JJ Sheldon KS (2012) On a collisioncourse competition and dispersal differences create no-analogue communities and cause extinctions during cli-mate change Proc R Soc Lond B Biol Sci 279 2072minus2080
Vacek S Matejka K (2010) State and development of phyto-cenoses on research plots in the Krkonoše Mts foreststands J For Sci 56 505minus517
Van Bogaert R Haneca K Hoogesteger J Jonasson C deDapper M Callaghan TV (2011) A century of tree linechanges in sub-Arctic Sweden shows local and regionalvariability and only a minor influence of 20th century climate warming J Biogeogr 38 907minus921
Van Gils H Batsukh O Rossiter D Munthali W Liberato -scioli E (2008) Forecasting the pattern and pace of Fagusforest expansion in Majella National Park Italy ApplVeg Sci 11 539minus546
Velchev V (1997) Types of vegetation In Yordanova MDonchev D (eds) Geography of Bulgaria BAN Sofiap 269minus283 (in Bulgarian)
Vladovic J Merganic J Maacuteliš F Križovaacute E and others (2014)The response of forest vegetation diversity to changes inedaphic-climatic conditions in Slovakia Technickaacute uni-verzita vo Zvolene Zvolene (in Slovak)
Vrška T Adam D Hort L Kolaacuter T Janiacutek F (2009) Europeanbeech (Fagus sylvatica L) and silver fir (Abies alba Mill)rotation in the Carpathians mdash a developmental cycle or alinear trend induced by man For Ecol Manag 258 347minus356
Wielgolaski FE (ed) (2005) Plant ecology herbivory andhuman impact in Nordic mountain birch forests Eco -logical Studies 180 Springer Verlag Berlin
Wielgolaski FE Hofgaard A Holtmeier FK (2017) Sensitivityto environmental change of the treeline ecotone and itsassociated biodiversity in European mountains Clim Res73151ndash166
Wieser G Tausz M (2007) Trees at their upper limit Treelifelimitation at the alpine timberline Springer Dordrecht
Wieser G Holtmeier FK Smith WK (2014) Treelines in achanging global environment In Tausz M Grulke N(eds) Trees in a changing environment Plant Ecophys 9Springer Dordrecht p 221minus263
Zhiyanski M Kolev K Sokolovska M Hursthouse A (2008)Tree species effect on soils in Central Stara PlaninaMountains Nauka Gorata 4 65minus82
Zhiyanski M Gikov A Nedkov S Dimitrov P Naydenova L(2016) Mapping carbon storage using land coverlanduse data in the area of Beklemeto Central Balkan In Koulov B Zhelezov G (eds) Sustainable mountain re -gions challenges and perspectives in southeasternEurope Springer Basel p 53minus65
Zhu K Woodall CW Clark JS (2012) Failure to migrate lackof tree range expansion in response to climate changeGlob Change Biol 18 1042minus1052
150
Editorial responsibility Nils Chr Stenseth Oslo Norway
Submitted April 8 2016 Accepted March 8 2017Proofs received from author(s) July 7 2017
Cudliacuten et al Drivers of treeline shift
al 2009) only a few published studies have includedquantitative data about treeline shifts (Devi et al2008 Kullman amp Oumlberg 2009 Diaz-Varela et al 2010Van Bogaert et al 2011) Additionally our know -ledge of the spatial patterns in treeline ecotone shiftsat the landscape scale is still surprisingly poor exceptfor some studies from subarctic areas the UralMountains and the Alps (Lloyd et al 2002 Diaz-Varela et al 2010 Hagedorn et al 2014) In the 15studied mountain units the values of treeline shiftranged from 043 to 19 m yrminus1 showing a rather dis-tinct spatial pattern of the dynamics within Europeanmountains The observed treeline shift had a signifi-cant positive relationship only with northern latitudeand a weak positive relationship with mountain-range size (ie the size of the whole mountain massifFig 3) Nevertheless this illustrates the importanceof the mass elevation effect (Koumlrner 2012) and par-tially ex plains why treelines could be at different alti-tudes at the same latitude Small negative regres-sions with east longitude and timberline elevationare shown in Fig 5 There was no apparent depend-ence of the treeline shift rate on climatic parameterchanges between the periods 1961minus1990 and1991minus2015
A whole forest vegetation zone shift is a complexprocess as not only the life strategies of an individualtree need to be considered but plant animal andmicroorganism species and their interrelationshipsas well as the relationships to their specific microhab-
itats (including soil conditions) are also involved(Urban et al 2012) Across all regions we identifiedseveral important obstacles to treeline shifts oftenrelated to site properties such as significant rocki-ness having shallow or low-nutrient soils extremerelief causing disturbances by avalanches snow glid-ing or wind damage (data from the Pyrenees Apen-nines Shara Northern Dinaric Low Tatra and GiantMountains see Fig 2) Soil heterogeneity certainlyhas an important role in plant responses to climatechange and could also maintain the resilience of thecommunity assemblages (Fridley et al 2011) An -other group of obstacles includes extreme climaticparameters especially in winter (reported eg fromthe Pyrenees Apennines Pirin Central BalkanNorthern Dinaric and Giant Mountains Fig 2) Theircombination can result in edaphic andor climaticunsuitability of habitats where species could poten-tially migrate Climatically and edaphically suitablesites will likely decline over the next century partic-ularly in mountain landscapes (Bell et al 2014)There fore trees in the treeline ecotone will colonizepreviously forested habitats The colonization suc-cess of individual forest communities is affected bydifferences in species dispersal and recruitmentbehaviour (Dullin ger et al 2004 Jonaacutešovaacute et al2010) For these reasons colonization of new foresthabitats by the next tree generation may not be suc-cessful and may result in loss of species diversity(Honnay et al 2002 Ibaacutentildeez et al 2009 Dobrowski et
145
Fig 5 Univariate graphs of the relationships of dependent variables viz timberline (Timb) and treeline shift per year (Shift) and their explanatory variables Nola north latitude Ealo east longitude DD decimal degrees
Clim Res 73 135ndash150 2017
al 2015) as reported from Macedonia Slovenia theCzech Republic Slovakia and Norway (Fig 4)
Another limiting factor is the existence of strongadaptation mechanisms of the dominant tree specieswhich allow them to survive in their current distribu-tion areas Species migration is likely to be slow dueto the limited quantity of climatically and edaphicallysuitable sites and the slow velocity of seed dispersaland tree regeneration and will be hampered evenmore by fragmentation of high mountain landscapescaused by human activities including brush invasionin abandoned meadows (Gartzia et al 2014) A verti-cal shift in a vegetation zone involves not only singlespecies of plants animals and microorganisms butalso their interrelationships and links to soil condi-tions The speed at which climatic conditions changewill be different from changes in the soil conditionsdue to the slowness of soil-forming processes Soiltypes and conditions (cambisol versus podzol) arecrucial obstacles for the shift from a beech forest zoneinto the spruce forest zone in the Czech Republic(Vacek amp Mat jka 2010) According to other sourcesthe dominant tree species can influence soil for -mation processes (especially humus forms) Underfavourable climatic and orographic conditions beechis able to change its soil conditions over decades tocenturies (J Macku unpubl data)
Significant changes in biodiversity must be expec -ted in all of the mountain areas of interest We foundthat biodiversity declined particularly in the southernregions of Europe where the timberline is situated athigher elevations and human impact is mostlylonger-lasting (Fig 4) Similarly Pauli et al (2012)reported an increase in species richness of mountaingrasslands across Europe except for Mediterraneanregions and assigned this different response of thesouthern regions to decreased water availability Onthe other hand ecosystems which exhibited bio -diversity increases were not only enriched by mig -rant species but the assemblages underwent thermo -philization ie cold-adapted species declined andmore warm-adapted species spread (Gottfried et al2012) However not all species are able to track cli-mate changes It is expected that around 40 ofhabitats will become climatically unsuitable for manymountain species particularly endemic ones in -creasing their extinction probability during this cen-tury (Dullinger et al 2012) The abandonment of tra-ditional land use forms is another source of this losswhich may be a more important driver than tempera-ture increase (Fig 4) Serious biodiversity losses inmountain meadows were reported from Spain ItalyMacedonia Bulgaria Slovenia and Slovakia There-
fore controlled grazing must occur in order to main-tain alpine grasslands (Dirnboumlck et al 2003) Unfor-tunately grazing is still active only in smaller parts ofEuropean mountains (eg in the Central Alps and theCentral Balkan Mountains but pasturing can alsonegatively affect vegetation diversity) and some-times does not serve to maintain alpine grasslandThe same is true for grassland management whichcan help to maintain mountain species and decreasehabitat loss A serious biodiversity loss in mountainmeadows was reported from the Pyrenees the Apen-nines and the Shara Pirin Central Balkan NorthernDinaric and Low Tatra Mountains (Fig 4)
In forests species responses to climate change maybe equivocal some recent studies indicated contra-dictory shifts in species distributions (eg Lenoir etal 2008 Zhu et al 2012 Rabasa et al 2013) This dis-parity is widely discussed and assigned to the greatvariety of non-climatic factors or even tree ontogeny(Grytnes et al 2014 Lenoir amp Svenning 2015 Maacutelišet al 2016) Disturbances leading to tree mortalitymay also play an important role (Cudliacuten et al 2013)changes in tree canopy cover modify light availabil-ity and microclimate and can induce the thermo -philization of forest vegetation (De Frenne et al2015 Stevens et al 2015) The loss of this microcli-mate buffering effect of forests may induce a biotichomogenization of forests (Savage amp Vellend 2015)or the creation of novel non-analogical communitiessuch as oakminuspine forests (Urban et al 2012) whichmay be a new threat to forest biodiversity
The observed changes in treeline forest ecosystemsare often related to changes in land-use intensity(Theurillat amp Guisan 2001 Alados et al 2014) Ac -cording to climate change predictions and the recentand future exploitation intensity of European moun-tains trees and forest communities will shift upwarddue to land use change or climate change or bothReduced land-use intensity certainly will interactwith climate change by facilitating or inhibiting spe-cies occurrence and will accelerate forest expansionabove the present treeline particularly to previouslyforested habitats (Theurillat amp Guisan 2001) Thesimultaneous action of both main treeline shift driv-ers viz temperature increase and decrease in landuse intensity was recorded from all of our studiedmountain areas The extensive differences in timber-line and treeline elevations in almost all studiedmountains (Table 1) indicate the anthropo-zoogenictreeline type (according to Ellenberg 1998) In mostmountains (eg in the Apennines Shara MountainsCentral Balkans and Alps) land use is the prevailingfactor influencing vegetation drift (Fig 5) Previous
146
Cudliacuten et al Drivers of treeline shift
research showed that the upward shift of the treelinein the Swiss Alps was predominantly attributable toland abandonment and only in some situations to cli-mate change (Gehrig-Fasel et al 2007) In the Apen-nines in the last few decades tree establishment hasbeen mainly controlled by land use while treegrowth has been controlled by climate pointing to aminor role played by climate in shaping the currenttreeline (Palombo et al 2013)
The impact of climate change and the connectedland use change on biodiversity and ecosystem serv-ices provision in several European countries is sum-marized by Wielgolaski et al (2017) and Kyriazopou-los et al (2017) (both this Special) and Sarkki et al(2016) One of the possible adaptive management op-tions in response to climate change assisted migra-tion as human-assisted movement of species hasbeen frequently debated in the last few years (Ste-Marie et al 2011) It is possible to apply it as a type ofassisted colonization ie intentional movement andrelease of an organism outside its indigenous range toavoid extinction of populations of the focal species(eg Macedonian pine species) or as an ecologicalre placement ie the intentional movement and re-lease of an organism outside its indigenous range toperform a specific ecological function (eg planting ofPinus mugo in the Alps IUCNSSC 2013)
5 CONCLUSIONS
The analysis of 11 mountain areas across Europeshowed that with increasing latitude the treelineand altitudinal width of the treeline ecotone signifi-cantly decreases as do the significance of climaticand soil parameters as barriers against tree speciesshift Although temperature is the overwhelmingcontrolling factor of tree growth and establishment intemperate and boreal treeline ecotones late-sea-sonal drought might also play a driving role in Medi-terranean treeline ecotones Longitude was lessinfluential mostly affecting climate and land usechange effects on increased biodiversity loss as wellas the size of the area that forms one mesoclimateunit affecting altitudinal ecotone width The biggestpart of the commonly observed remaining variabilityin mountain vegetation near the treeline in Europeseems to be caused by geomorphological geologicalpedological and microclimatic variability in combina-tion with different land use history and the presentsocio-economic relations The observed differencesin climatic parameters between the mountain areasof interest in comparison with the reference period
1960minus1990 (09degC) have not explained the relativelyhigh differences in the rate of treeline shift per year(149 m) The predicted variability in temperatureincrease due to global warming (12minus26degC in 2050and 22minus53degC in 2100) could lead to a much biggerdifferentiation in treeline ecotone biodiversity andecosystem processes between southern and northernEuropean mountains in the future Therefore thesedifferences must be taken into account by scientistsand EU policy makers when formulating efficientadaptive forest management strategies for treelineecosystems at the European level (eg assistedmigration of adapted genotypes)
Acknowledgements This paper is based firstly upon workfrom the COST Action ES 1203 SENSFOR supported byCOST (European Cooperation in Science and Technologywwwcosteu) This international work was further sup-ported by projects granted by the Ministry of EducationYouth and Sports of the Czech Republic grant NPU ILO1415 and LD 14039 by the agency APVV SR under pro-jects APVV-14-0086 and APVV-15-0270 We acknowledgethe E-OBS dataset from the EU-FP6 project ENSEMBLES(httpensembles-eumetofficecom) and the data providersin the ECAampD project (wwwecadeu)
LITERATURE CITED
Alados CL Errea P Gartzia M Saiz H Escoacutes J (2014) Posi-tive and negative feedbacks and free-scale pattern distri-bution in rural-population dynamics PLOS ONE 9 e114561
Allen CD Macalady AK Chenchouni H Bachelet D andothers (2010) A global overview of drought and heat-induced tree mortality reveals emerging climate changerisks for forests For Ecol Manag 259 660minus684
Amidžic L Bartula M Jaric S (2012) Syntaxonomy overviewof vegetation of Šar Planina in the Balkan PeninsulaVegetos 25 348minus360
Aune S Hofgaard A Soumlderstroumlm L (2011) Contrasting cli-mate and land use driven tree encroachment pattern ofsub-arctic tundra in Northern Norway and Kola Penin-sula Can J For Res 41 437minus449
Barbeito I Dawes M Rixen C Senn J Bebi P (2012) Factorsdriving mortality and growth at treeline a 30-year exper-iment of 92000 conifers Ecology 93 389minus401
Batllori E Camarero JJ Ninot JM Gutieacuterrez E (2009)Seedling recruitment survival and facilitation in alpinePinus uncinata treeline ecotones Implications andpotential responses to climate warming Glob Ecol Bio-geogr 18 460minus472
Bebi P Teich M Hagedorn F Zurbriggen N Brunner SHGreacutet-Regamey A (2012) Veraumlnderung von Wald undWaldleistungen in der Landschaft Davos im Zuge desKlimawandels Schweiz Z Forstwes 163 493minus501
Bell DM Bradford JB Lauenroth WK (2014) Mountain land-scapes offer few opportunities for high elevation treespecies migration Glob Change Biol 20 1441minus1451
Bergant K (2006) Climate change scenarios for MacedoniaUniversity of Nova Gorica Center for Atmospheric
147
Clim Res 73 135ndash150 2017
Research Nova GoricaBodin J Badeau V Bruno E Cluzeau C Moisselin JM
Walther GR Dupouey JL (2013) Shifts of forest speciesalong an elevational gradient in Southeast France cli-mate change or stand maturation J Veg Sci 24 269minus283
Boncina A (2011) History current status and future pros -pects of uneven-aged forest management in the Dinaricregion an overview Forestry 84 467minus478
Breshears DD Cobb NS Rich PM Price KP and others(2005) Regional vegetation die-off in response to global-change-type drought Proc Natl Acad Sci USA 102 15144minus15148
Camarero JJ Garciacutea-Ruiz JM Sanguumlesa-Barreda G GalvaacutenJD and others (2015a) Recent and intense dynamics in aformerly static Pyrenean treeline Arct Antarct Alp Res47 773minus783
Camarero JJ Gazol A Galvaacuten JD Sanguumlesa-Barreda GGutieacuterrez E (2015b) Disparate effects of global-changedrivers on mountain conifer forests warming-inducedgrowth enhancement in young trees vs CO2 fertilizationin old trees from wet sites Glob Change Biol 21 738minus749
Chen IC Hill JK Ohlemuumlller R Roy DB Thomas CD (2011)Rapid range shifts of species associated with high levelsof climate warming Science 333 1024minus1026
Cudliacuten P Sejaacutek J Pokornyacute J Albrechtovaacute J Bastian OMarek M (2013) Forest ecosystem services under climatechange and air pollution In Matyssek R Clarke N Cud-liacuten P Mikkelsen TN Tuovinen JP Wiesner G Paoletti E(eds) Climate change air pollution and global chal-lenges understanding and perspectives from forestresearch Developments in Environmental Science Vol13 Elsevier Oxford p 521minus546
Dalen L Hofgaard A (2005) Differential regional treelinedynamics in the Scandes mountains Arct Antarct AlpRes 37 284minus296
Dawes MA Philipson CD Fonti P Bebi P Haumlttenschwiler SHagedorn F Rixen C (2015) Soil warming and CO2
enrichment induce biomass shifts in alpine tree line veg-etation 2014 Glob Change Biol 21 2005minus2021
De Frenne P Rodriacuteguez-Saacutenchez F De Schrijver A CoomesDA Hermy M Vangansbeke P Verheyen K (2015) Lightaccelerates plant responses to warming Nat Plants 1 15110
Devi N Hagedorn F Moiseev P Bugmann H ShiyatovSG Mazepa V Rigling A (2008) Expanding forests andchanging growth forms of Siberian larch at the treeline ofthe Polar Urals during the 20th century Glob ChangeBiol 14 1581minus1591
Diaz-Varela R Colombo R Meroni M Calvo-Iglesias MSBuffoni A Tagliaferri A (2010) Spatio-temporal analysisof alpine ecotones A spatial explicit model targeting altitudinal vegetation shifts Ecol Model 221 621minus633
Dirnboumlck T Dullinger S Grabherr G (2003) A regionalimpact assessment of climate and land-use change onalpine vegetation J Biogeogr 30 401minus417
Dobrowski SZ Swanson AK Abatzoglou JT Holden ZASafford HD Schwartz MK Gavin DG (2015) Forest struc-ture and species traits mediate projected recruitmentdeclines in western US tree species Glob Ecol Biogeogr24 917minus927
Dullinger S Dirnboumlck S Grabherr G (2004) Modelling cli-mate change-driven treeline shifts relative effects oftemperature increase dispersal and invisibility J Ecol92 241minus252
Dullinger S Gattringer A Thuiller W Moser D and others
(2012) Extinction debt of high-mountain plants undertwenty-first-century climate change Nat Clim Chang 2 619minus622
Elkin C Gutierrez AG Leuzinger S Manusch C TemperliC Rasche L Bugmann H (2013) A 2degC warmer world isnot safe for ecosystem services in the European AlpsGlob Change Biol 19 1827minus1840
Ellenberg H (1988) Vegetation ecology of Central Europe4th edn Cambridge University Press Edinburgh
Em H (1986) Spruce southern distribution range Spruce for-est on Shara Mountain in Macedonia Contrib Macedon-ian Acad Sci Arts Skopje 5 11minus28
Fajardo A Piper FI Pfund L Koumlrner CH Hoch G (2012)Variation of mobile carbon reserves in trees at the alpinetreeline ecotone is under environmental control NewPhytol 195 794minus802
Fridley JD Grime JP Askew AP Moser B Stevens CJ (2011)Soil heterogeneity buffers community response to cli-mate change in species-rich grassland Glob Change Biol17 2002minus2011
Gartzia M Alados CL Peacuterez-Cabello F (2014) Assessment ofthe effects of biophysical and anthropogenic factors onwoody plant encroachment in dense and sparse moun-tain grasslands based on remote sensing data Prog PhysGeogr 38 201minus217
Gehrig-Fasel J Guisan A Zimmermann NE (2007) Tree lineshifts in the Swiss Alps climate change or land abandon-ment J Veg Sci 18 571minus582
Gikov A Dimitrov P Zhiyanski M (2016) Land cover dynam-ics on the northern slope of the Troyan Passage during 30years period Probl Geogr 1-2 78minus92
Gonzalez P Neilson RP Lenihan JM Drapek RJ (2010)Global patterns in the vulnerability of ecosystems to veg-etation shifts due to climate change Glob Ecol Biogeogr19 755minus768
Gottfried M Pauli H Futschik A Akhalkatsi M and others(2012) Continent-wide response of mountain vegetationto climate change Nat Clim Chang 2 111minus115
Grace J Berninger F Nagy L (2002) Impacts of climatechange on the tree line Ann Bot 90 537minus544
Grunewald K Scheithauer J (2011) Landscape developmentand climate change in Southwest Bulgaria (Pirin Mts)Springer Heidelberg
Grytnes JA Kapfer J Jurasinski G Birks HH and others(2014) Identifying the driving factors behind observedelevational range shifts on European mountains GlobEcol Biogeogr 23 876minus884
Hagedorn F Shiyatov FG Mazepa VS Devi NM and others(2014) Treeline advances along the Urals mountainrange mdash driven by improved winter conditions GlobChange Biol 20 3530minus3543
Hansen AJ Neilson RP Dale VH Flather CH and others(2001) Global change in forests responses of speciescommunities and biomes BioScience 51 765minus779
Harsch MA Bader MY (2011) Treeline formminusa potential keyto understanding treeline dynamics Glob Ecol Biogeogr20 582minus596
Harsch MA Hulme PE McGlone MS Duncan RP (2009) Aretreelines advancing A global meta-analysis of treelineresponse to climate warning Ecol Lett 12 1040minus1049
Haylock MR Hofstra N Klein Tank AMG Klok EJ JonesPD New M (2008) A European daily high-resolutiongridded dataset of surface temperature and precipita-tion J Geophys Res Atmos 113 D20119
Hlaacutesny T Barcza Z Fabrika M Balaacutezs B and others (2011)
148
Cudliacuten et al Drivers of treeline shift
Climate change impacts on growth and carbon balanceof forests in Central Europe Clim Res 47 219minus236
Hlaacutesny T Trombik J Dobor L Barcza Z Barka I (2016)Future climate of the Carpathians climate change hot-spots and implications for ecosystems Reg EnvironChange 16 1495minus1506
Hofgaard A Dalen L Hytteborn H (2009) Tree recruitmentabove the treeline and potential for climate driven tree-line change J Veg Sci 20 1133minus1144
Holtmeier FK (2009) Mountain timberlines Ecology patchi-ness and dynamics 2nd edn Advances in GlobalChange Research 36 Springer Science amp ScienceMediaBV Dordrecht
Holtmeier FK Broll G (2005) Sensitivity and response ofnorthern hemisphere altitudinal and polar treelines toenvironmental change at landscape and local scalesGlob Ecol Biogeogr 14 395minus410
Honnay O Verheyen K Butaye J Jacquemyn H Bossuyt BHermy M (2002) Possible effects of habitat fragmentationand climate change on the range of forest plant speciesEcol Lett 5 525minus530
Huber R Rigling A Bebi P Brand F and others (2013) Sus-tainable land use in mountain regions under globalchange synthesis across scales and disciplines Ecol Soc18 36
Ibaacutentildeez I Clark JS Dietze MC (2009) Estimating colonizationpotential of migrant tree species Glob Change Biol 15 1173minus1188
IUCNSSC (International Union for Conservation of NatureSpecies Survival Commission) (2013) Guidelines for re -introductions and other conservation translocations Version 10 IUCNSpecies Survival Commission Gland
Jankovskyacute L Cudliacuten P gtermaacutek P Moravec I (2004) The pre-diction of development of secondary Norway sprucestands under the impact of climatic change in the Dra-hany highlands (The Czech Republic) Ekologia (Bratisl)23(Suppl 2) 101minus112
Jeniacutek J Lokvenc T (1962) Die alpine Waldgrenze im Krko -noše Gebirge Rozpravy SAV Rada matematickyacutech apriacuterodniacutech ved 72 gteskoslovenskaacute akademie ved Praha
Jobbaacutegy EG Jackson RB (2000) Global controls of forest lineelevation in the northern and southern hemispheresGlob Ecol Biogeogr 9 253minus268
Jonaacutešovaacute M Vaacutevrovaacute E Cudliacuten P (2010) Western Carpa -thian mountain spruce forest after a windthrow naturalregeneration in cleared and uncleared areas For EcolManag 259 1127minus1134
Kittel TGF Steffen WL Chapin FS (2000) Global andregional modelling of Arcticminusboreal vegetation distribu-tion and its sensitivity to altered forcing Glob ChangeBiol 6 1minus18
Klopcic M Jerina K Bonltina A (2010) Long-term changes ofstructure and tree species composition in Dinaric un -even- aged forests Are red deer an important factor EurJ For Res 129 277minus288
Klopcic M Simonltilt T Bonltina A (2015) Comparison of re -generation and recruitment of shade-tolerant and light-demanding tree species in mixed uneven-aged forests ex -periences from the Dinaric region Forestry 88 552minus563
Koumlrner C (2003) Alpine plant life functional plant ecology ofhigh mountain ecosystems with 47 tables Springer Sci-ence amp Business Media BV Dordrecht
Koumlrner C (2012) Alpine treelines Functional ecology of theglobal high elevation tree limits Springer Basel
Koumlrner C Paulsen J (2004) A world-wide study of high alti-
tude treeline temperatures J Biogeogr 31 713minus732Kulakowski D Bebi P Rixen C (2011) The interacting effects
of land use change climate change and suppression ofnatural disturbances on landscape forest structure in theSwiss Alps Oikos 120 216minus225
Kullman L (1988) Holocene history of the forestminusalpine tun-dra ecotone in the Scandes Mountains (central Sweden)New Phytol 108 101minus110
Kullman L (1999) Early holocene tree growth at a high elevation site in the northernmost Scandes of Sweden(Lapland) a palaeobiogeographical case study based onmega fossil evidence Geogr Ann 81 63minus74
Kullman L (2007) Tree line population monitoring of Pinussylvestris in the Swedish Scandes 1973minus2005 implica-tions for tree line theory and climate change ecologyJ Ecol 95 41minus52
Kullman L Oumlberg L (2009) Post-Little Ice Age treeline riseand climate warming in the Swedish Scandes a land-scape ecological perspective J Ecol 97 415minus429
Kyriazopoulos AP Skre O Sarkki S Wielgolaski FE Abra-ham EM Ficko A (2017) Humanminusenvironment dynamicsin European treeline ecosystems a synthesis based onthe DPSIR framework Clim Res 73 17ndash29
Lenoir J Svenning JC (2015) Climate-related range shifts mdasha global multidimensional synthesis and new researchdirections Ecography 38 15minus38
Lenoir J Geacutegout JC Marquet PA de Ruffray P Brisse H(2008) A significant upward shift in plant species opti-mum elevation during the 20th century Science 320 1768minus1771
Lindner M Fitzgerald JB Zimmermann NE Reyer C andothers (2014) Climate change and European forests What do we know what are the uncertainties and whatare the implications for forest management J EnvironManag 146 69minus83
Lloyd AH Rupp TS Fastie CL Srafield AM (2002) Patternsand dynamic of treeline advance on the Seward Pen -isula Alaska J Geophys Res 107 8161
Maacuteliš F Kopeckyacute M Petriacutek P Vladovic J Merganic J VidaT (2016) Life-stage not climate change explains ob -served tree range shifts Glob Change Biol 22 1904minus1914
Mathisen IE Mikheeva A Tutubalina OV Aune S and Hof-gaard A (2014) Fifty years of tree line change in theKhibiny Mountains Russia advantages of combinedremote sensing and dendroecological approaches ApplVeg Sci 17 6ndash16
Millar CI Stephenson NL (2015) Temperate forest healthin an era of emerging megadisturbance Science 349 823minus826
Mina M Bugmann H Klopcic M Cailleret M (2017) Accu-rate modeling of harvesting is key for projecting futureforest dynamics a case study in the Slovenian moun-tains Reg Environ Change 17 49minus64
Moen J Aune K Edenius L Angerbjoumlrn A (2004) Potentialeffects of climate change on treeline position in theSwedish mountains Ecol Soc 9 16 wwwecologyandsocietyorgvol9iss1art16
Palombo C Chirici G Marchetti M Tognetti R (2013) Is landabandonment affecting forest dynamics at high elevationin Mediterranean mountains more than climate changePlant Biosyst 147 1minus11
Palombo C Marchetti M Tognetti R (2014) Mountain vege-tation at risk current perspectives and research reedsPlant Biosyst 148 35minus41
Pauli H Gottfried M Dullinger S Abdaladze O Akhalkatsi
149
Clim Res 73 135ndash150 2017
M Alonso JLB Grabherr G (2012) Recent plant diversitychanges on Europersquos mountain summits Science 336 353minus355
Payette S Fortin MJ Gamachet I (2001) The subarctic forestminustundra the structure of a biome in a changing climate BioScience 51 709minus718
Pearson GR Dawson TP (2003) Predicting the impacts of cli-mate change on the distribution of species Are bio -climate envelope models useful Glob Ecol Biogeogr 12 361minus371
Pedrotti F (2013) Plant and vegetation mapping GeobotanyStudies Springer-Verlag Berlin
Pentildeuelas J Boada M (2003) A global change-induced biomeshift in the Montseny mountains (NE Spain) GlobChange Biol 9 131minus140
Piovesan G Biondi F Filippo AD Alessandrini A MaugeriM (2008) Drought driven growth reduction in old beech(Fagus sylvatica L) forests of the central ApenninesItaly Glob Change Biol 14 1265minus1281
Rabasa SG Granda E Benavides R Kunstler G and others(2013) Disparity in elevational shifts of European trees inresponse to recent climate warming Glob Change Biol19 2490minus2499
Raev I Zhelev P Grozeva M Markov I and others (2011)Programme of measures for adaptation of forests in Bul-garia and mitigate the negative impact of climate changeon them Project FUTURE forest helping Europe tackleclimate change INTERREG IVC ERDF Sofia
Sarkki S Ficko A Grunewald K Nijnik M (2016) Benefitsfrom and threats to European treeline ecosystem serv-ices an exploratory study of stakeholders and gover-nance Reg Environ Change 16 2019minus2032
Savage J Vellend M (2015) Elevational shifts biotic homo -genization and time lags in vegetation change during 40years of climate warming Ecography 38 546minus555
Schwoumlrer C Henne PD Tinner W (2014) A model-data com-parison of Holocene timberline changes in the SwissAlps reveals past and future drivers of mountain forestdynamics Glob Change Biol 20 1512minus1526
Šmilauer P Lepš J (2014) Multivariate analysis of ecologicaldata using Canoco 5 Cambridge University Press Cam-bridge
Ste-Marie C Nelson EA Dabros A Bonneau ME (2011)Assisted migration introduction to a multifaceted con-cept For Chron 87 724minus730
Stevens JT Stafford HD Harrison S Latimer AM (2015) For-est disturbance accelerates thermophilization of under-story plant communities J Ecol 103 1253minus1263
Strid A Andonoski A Andonovski V (2003) Alpine biodiver-sity in Europe Ecological Studies 167 Springer-VerlagBerlin
Tattoni C Ciolli M Ferretti F Cantiani MG (2010) Monitor-ing spatial and temporal pattern of Paneveggio forest(Northern Italy) from 1859 to 2006 iForest Biogeosci For3 72minus80
Ter Braak CJF Šmilauer P (2012) Canoco reference manualand userrsquos guide software for ordination (version 50)Microcomputer Power Ithaca NY
Theurillat JP Guisan A (2001) Potential impact of climatechange on vegetation in the European Alps a reviewClim Change 50 77minus109
Toslashmmervik H Wielgolaski FE Neuvonen S Solberg BHoslashgda KA (2005) Biomass and production on a land-scape level in the mountain birch forests In WielgolaskiFE (ed) Plant ecology herbivory and human impact in
Nordic mountain birch forests Ecos Studies 180Springer Berlin p 53minus70
Treml V Banaš M (2000) Alpine timberline in the HighSudetes Acta Univ Carol Geogr 15 83minus99
Treml V Chuman T (2015) Ecotonal dynamics of the altitu-dinal forest limit are affected by terrain and vegetationstructure variables an example from the Sudetes Moun-tains in Central Europe Arct Antarct Alp Res 47 133minus146
Treml V Migo P (2015) controlling factors limiting timber-line position and shifts in the Sudetes a review GeogrPol 88 55minus70
Urban MC Tewksbury JJ Sheldon KS (2012) On a collisioncourse competition and dispersal differences create no-analogue communities and cause extinctions during cli-mate change Proc R Soc Lond B Biol Sci 279 2072minus2080
Vacek S Matejka K (2010) State and development of phyto-cenoses on research plots in the Krkonoše Mts foreststands J For Sci 56 505minus517
Van Bogaert R Haneca K Hoogesteger J Jonasson C deDapper M Callaghan TV (2011) A century of tree linechanges in sub-Arctic Sweden shows local and regionalvariability and only a minor influence of 20th century climate warming J Biogeogr 38 907minus921
Van Gils H Batsukh O Rossiter D Munthali W Liberato -scioli E (2008) Forecasting the pattern and pace of Fagusforest expansion in Majella National Park Italy ApplVeg Sci 11 539minus546
Velchev V (1997) Types of vegetation In Yordanova MDonchev D (eds) Geography of Bulgaria BAN Sofiap 269minus283 (in Bulgarian)
Vladovic J Merganic J Maacuteliš F Križovaacute E and others (2014)The response of forest vegetation diversity to changes inedaphic-climatic conditions in Slovakia Technickaacute uni-verzita vo Zvolene Zvolene (in Slovak)
Vrška T Adam D Hort L Kolaacuter T Janiacutek F (2009) Europeanbeech (Fagus sylvatica L) and silver fir (Abies alba Mill)rotation in the Carpathians mdash a developmental cycle or alinear trend induced by man For Ecol Manag 258 347minus356
Wielgolaski FE (ed) (2005) Plant ecology herbivory andhuman impact in Nordic mountain birch forests Eco -logical Studies 180 Springer Verlag Berlin
Wielgolaski FE Hofgaard A Holtmeier FK (2017) Sensitivityto environmental change of the treeline ecotone and itsassociated biodiversity in European mountains Clim Res73151ndash166
Wieser G Tausz M (2007) Trees at their upper limit Treelifelimitation at the alpine timberline Springer Dordrecht
Wieser G Holtmeier FK Smith WK (2014) Treelines in achanging global environment In Tausz M Grulke N(eds) Trees in a changing environment Plant Ecophys 9Springer Dordrecht p 221minus263
Zhiyanski M Kolev K Sokolovska M Hursthouse A (2008)Tree species effect on soils in Central Stara PlaninaMountains Nauka Gorata 4 65minus82
Zhiyanski M Gikov A Nedkov S Dimitrov P Naydenova L(2016) Mapping carbon storage using land coverlanduse data in the area of Beklemeto Central Balkan In Koulov B Zhelezov G (eds) Sustainable mountain re -gions challenges and perspectives in southeasternEurope Springer Basel p 53minus65
Zhu K Woodall CW Clark JS (2012) Failure to migrate lackof tree range expansion in response to climate changeGlob Change Biol 18 1042minus1052
150
Editorial responsibility Nils Chr Stenseth Oslo Norway
Submitted April 8 2016 Accepted March 8 2017Proofs received from author(s) July 7 2017
Clim Res 73 135ndash150 2017
al 2015) as reported from Macedonia Slovenia theCzech Republic Slovakia and Norway (Fig 4)
Another limiting factor is the existence of strongadaptation mechanisms of the dominant tree specieswhich allow them to survive in their current distribu-tion areas Species migration is likely to be slow dueto the limited quantity of climatically and edaphicallysuitable sites and the slow velocity of seed dispersaland tree regeneration and will be hampered evenmore by fragmentation of high mountain landscapescaused by human activities including brush invasionin abandoned meadows (Gartzia et al 2014) A verti-cal shift in a vegetation zone involves not only singlespecies of plants animals and microorganisms butalso their interrelationships and links to soil condi-tions The speed at which climatic conditions changewill be different from changes in the soil conditionsdue to the slowness of soil-forming processes Soiltypes and conditions (cambisol versus podzol) arecrucial obstacles for the shift from a beech forest zoneinto the spruce forest zone in the Czech Republic(Vacek amp Mat jka 2010) According to other sourcesthe dominant tree species can influence soil for -mation processes (especially humus forms) Underfavourable climatic and orographic conditions beechis able to change its soil conditions over decades tocenturies (J Macku unpubl data)
Significant changes in biodiversity must be expec -ted in all of the mountain areas of interest We foundthat biodiversity declined particularly in the southernregions of Europe where the timberline is situated athigher elevations and human impact is mostlylonger-lasting (Fig 4) Similarly Pauli et al (2012)reported an increase in species richness of mountaingrasslands across Europe except for Mediterraneanregions and assigned this different response of thesouthern regions to decreased water availability Onthe other hand ecosystems which exhibited bio -diversity increases were not only enriched by mig -rant species but the assemblages underwent thermo -philization ie cold-adapted species declined andmore warm-adapted species spread (Gottfried et al2012) However not all species are able to track cli-mate changes It is expected that around 40 ofhabitats will become climatically unsuitable for manymountain species particularly endemic ones in -creasing their extinction probability during this cen-tury (Dullinger et al 2012) The abandonment of tra-ditional land use forms is another source of this losswhich may be a more important driver than tempera-ture increase (Fig 4) Serious biodiversity losses inmountain meadows were reported from Spain ItalyMacedonia Bulgaria Slovenia and Slovakia There-
fore controlled grazing must occur in order to main-tain alpine grasslands (Dirnboumlck et al 2003) Unfor-tunately grazing is still active only in smaller parts ofEuropean mountains (eg in the Central Alps and theCentral Balkan Mountains but pasturing can alsonegatively affect vegetation diversity) and some-times does not serve to maintain alpine grasslandThe same is true for grassland management whichcan help to maintain mountain species and decreasehabitat loss A serious biodiversity loss in mountainmeadows was reported from the Pyrenees the Apen-nines and the Shara Pirin Central Balkan NorthernDinaric and Low Tatra Mountains (Fig 4)
In forests species responses to climate change maybe equivocal some recent studies indicated contra-dictory shifts in species distributions (eg Lenoir etal 2008 Zhu et al 2012 Rabasa et al 2013) This dis-parity is widely discussed and assigned to the greatvariety of non-climatic factors or even tree ontogeny(Grytnes et al 2014 Lenoir amp Svenning 2015 Maacutelišet al 2016) Disturbances leading to tree mortalitymay also play an important role (Cudliacuten et al 2013)changes in tree canopy cover modify light availabil-ity and microclimate and can induce the thermo -philization of forest vegetation (De Frenne et al2015 Stevens et al 2015) The loss of this microcli-mate buffering effect of forests may induce a biotichomogenization of forests (Savage amp Vellend 2015)or the creation of novel non-analogical communitiessuch as oakminuspine forests (Urban et al 2012) whichmay be a new threat to forest biodiversity
The observed changes in treeline forest ecosystemsare often related to changes in land-use intensity(Theurillat amp Guisan 2001 Alados et al 2014) Ac -cording to climate change predictions and the recentand future exploitation intensity of European moun-tains trees and forest communities will shift upwarddue to land use change or climate change or bothReduced land-use intensity certainly will interactwith climate change by facilitating or inhibiting spe-cies occurrence and will accelerate forest expansionabove the present treeline particularly to previouslyforested habitats (Theurillat amp Guisan 2001) Thesimultaneous action of both main treeline shift driv-ers viz temperature increase and decrease in landuse intensity was recorded from all of our studiedmountain areas The extensive differences in timber-line and treeline elevations in almost all studiedmountains (Table 1) indicate the anthropo-zoogenictreeline type (according to Ellenberg 1998) In mostmountains (eg in the Apennines Shara MountainsCentral Balkans and Alps) land use is the prevailingfactor influencing vegetation drift (Fig 5) Previous
146
Cudliacuten et al Drivers of treeline shift
research showed that the upward shift of the treelinein the Swiss Alps was predominantly attributable toland abandonment and only in some situations to cli-mate change (Gehrig-Fasel et al 2007) In the Apen-nines in the last few decades tree establishment hasbeen mainly controlled by land use while treegrowth has been controlled by climate pointing to aminor role played by climate in shaping the currenttreeline (Palombo et al 2013)
The impact of climate change and the connectedland use change on biodiversity and ecosystem serv-ices provision in several European countries is sum-marized by Wielgolaski et al (2017) and Kyriazopou-los et al (2017) (both this Special) and Sarkki et al(2016) One of the possible adaptive management op-tions in response to climate change assisted migra-tion as human-assisted movement of species hasbeen frequently debated in the last few years (Ste-Marie et al 2011) It is possible to apply it as a type ofassisted colonization ie intentional movement andrelease of an organism outside its indigenous range toavoid extinction of populations of the focal species(eg Macedonian pine species) or as an ecologicalre placement ie the intentional movement and re-lease of an organism outside its indigenous range toperform a specific ecological function (eg planting ofPinus mugo in the Alps IUCNSSC 2013)
5 CONCLUSIONS
The analysis of 11 mountain areas across Europeshowed that with increasing latitude the treelineand altitudinal width of the treeline ecotone signifi-cantly decreases as do the significance of climaticand soil parameters as barriers against tree speciesshift Although temperature is the overwhelmingcontrolling factor of tree growth and establishment intemperate and boreal treeline ecotones late-sea-sonal drought might also play a driving role in Medi-terranean treeline ecotones Longitude was lessinfluential mostly affecting climate and land usechange effects on increased biodiversity loss as wellas the size of the area that forms one mesoclimateunit affecting altitudinal ecotone width The biggestpart of the commonly observed remaining variabilityin mountain vegetation near the treeline in Europeseems to be caused by geomorphological geologicalpedological and microclimatic variability in combina-tion with different land use history and the presentsocio-economic relations The observed differencesin climatic parameters between the mountain areasof interest in comparison with the reference period
1960minus1990 (09degC) have not explained the relativelyhigh differences in the rate of treeline shift per year(149 m) The predicted variability in temperatureincrease due to global warming (12minus26degC in 2050and 22minus53degC in 2100) could lead to a much biggerdifferentiation in treeline ecotone biodiversity andecosystem processes between southern and northernEuropean mountains in the future Therefore thesedifferences must be taken into account by scientistsand EU policy makers when formulating efficientadaptive forest management strategies for treelineecosystems at the European level (eg assistedmigration of adapted genotypes)
Acknowledgements This paper is based firstly upon workfrom the COST Action ES 1203 SENSFOR supported byCOST (European Cooperation in Science and Technologywwwcosteu) This international work was further sup-ported by projects granted by the Ministry of EducationYouth and Sports of the Czech Republic grant NPU ILO1415 and LD 14039 by the agency APVV SR under pro-jects APVV-14-0086 and APVV-15-0270 We acknowledgethe E-OBS dataset from the EU-FP6 project ENSEMBLES(httpensembles-eumetofficecom) and the data providersin the ECAampD project (wwwecadeu)
LITERATURE CITED
Alados CL Errea P Gartzia M Saiz H Escoacutes J (2014) Posi-tive and negative feedbacks and free-scale pattern distri-bution in rural-population dynamics PLOS ONE 9 e114561
Allen CD Macalady AK Chenchouni H Bachelet D andothers (2010) A global overview of drought and heat-induced tree mortality reveals emerging climate changerisks for forests For Ecol Manag 259 660minus684
Amidžic L Bartula M Jaric S (2012) Syntaxonomy overviewof vegetation of Šar Planina in the Balkan PeninsulaVegetos 25 348minus360
Aune S Hofgaard A Soumlderstroumlm L (2011) Contrasting cli-mate and land use driven tree encroachment pattern ofsub-arctic tundra in Northern Norway and Kola Penin-sula Can J For Res 41 437minus449
Barbeito I Dawes M Rixen C Senn J Bebi P (2012) Factorsdriving mortality and growth at treeline a 30-year exper-iment of 92000 conifers Ecology 93 389minus401
Batllori E Camarero JJ Ninot JM Gutieacuterrez E (2009)Seedling recruitment survival and facilitation in alpinePinus uncinata treeline ecotones Implications andpotential responses to climate warming Glob Ecol Bio-geogr 18 460minus472
Bebi P Teich M Hagedorn F Zurbriggen N Brunner SHGreacutet-Regamey A (2012) Veraumlnderung von Wald undWaldleistungen in der Landschaft Davos im Zuge desKlimawandels Schweiz Z Forstwes 163 493minus501
Bell DM Bradford JB Lauenroth WK (2014) Mountain land-scapes offer few opportunities for high elevation treespecies migration Glob Change Biol 20 1441minus1451
Bergant K (2006) Climate change scenarios for MacedoniaUniversity of Nova Gorica Center for Atmospheric
147
Clim Res 73 135ndash150 2017
Research Nova GoricaBodin J Badeau V Bruno E Cluzeau C Moisselin JM
Walther GR Dupouey JL (2013) Shifts of forest speciesalong an elevational gradient in Southeast France cli-mate change or stand maturation J Veg Sci 24 269minus283
Boncina A (2011) History current status and future pros -pects of uneven-aged forest management in the Dinaricregion an overview Forestry 84 467minus478
Breshears DD Cobb NS Rich PM Price KP and others(2005) Regional vegetation die-off in response to global-change-type drought Proc Natl Acad Sci USA 102 15144minus15148
Camarero JJ Garciacutea-Ruiz JM Sanguumlesa-Barreda G GalvaacutenJD and others (2015a) Recent and intense dynamics in aformerly static Pyrenean treeline Arct Antarct Alp Res47 773minus783
Camarero JJ Gazol A Galvaacuten JD Sanguumlesa-Barreda GGutieacuterrez E (2015b) Disparate effects of global-changedrivers on mountain conifer forests warming-inducedgrowth enhancement in young trees vs CO2 fertilizationin old trees from wet sites Glob Change Biol 21 738minus749
Chen IC Hill JK Ohlemuumlller R Roy DB Thomas CD (2011)Rapid range shifts of species associated with high levelsof climate warming Science 333 1024minus1026
Cudliacuten P Sejaacutek J Pokornyacute J Albrechtovaacute J Bastian OMarek M (2013) Forest ecosystem services under climatechange and air pollution In Matyssek R Clarke N Cud-liacuten P Mikkelsen TN Tuovinen JP Wiesner G Paoletti E(eds) Climate change air pollution and global chal-lenges understanding and perspectives from forestresearch Developments in Environmental Science Vol13 Elsevier Oxford p 521minus546
Dalen L Hofgaard A (2005) Differential regional treelinedynamics in the Scandes mountains Arct Antarct AlpRes 37 284minus296
Dawes MA Philipson CD Fonti P Bebi P Haumlttenschwiler SHagedorn F Rixen C (2015) Soil warming and CO2
enrichment induce biomass shifts in alpine tree line veg-etation 2014 Glob Change Biol 21 2005minus2021
De Frenne P Rodriacuteguez-Saacutenchez F De Schrijver A CoomesDA Hermy M Vangansbeke P Verheyen K (2015) Lightaccelerates plant responses to warming Nat Plants 1 15110
Devi N Hagedorn F Moiseev P Bugmann H ShiyatovSG Mazepa V Rigling A (2008) Expanding forests andchanging growth forms of Siberian larch at the treeline ofthe Polar Urals during the 20th century Glob ChangeBiol 14 1581minus1591
Diaz-Varela R Colombo R Meroni M Calvo-Iglesias MSBuffoni A Tagliaferri A (2010) Spatio-temporal analysisof alpine ecotones A spatial explicit model targeting altitudinal vegetation shifts Ecol Model 221 621minus633
Dirnboumlck T Dullinger S Grabherr G (2003) A regionalimpact assessment of climate and land-use change onalpine vegetation J Biogeogr 30 401minus417
Dobrowski SZ Swanson AK Abatzoglou JT Holden ZASafford HD Schwartz MK Gavin DG (2015) Forest struc-ture and species traits mediate projected recruitmentdeclines in western US tree species Glob Ecol Biogeogr24 917minus927
Dullinger S Dirnboumlck S Grabherr G (2004) Modelling cli-mate change-driven treeline shifts relative effects oftemperature increase dispersal and invisibility J Ecol92 241minus252
Dullinger S Gattringer A Thuiller W Moser D and others
(2012) Extinction debt of high-mountain plants undertwenty-first-century climate change Nat Clim Chang 2 619minus622
Elkin C Gutierrez AG Leuzinger S Manusch C TemperliC Rasche L Bugmann H (2013) A 2degC warmer world isnot safe for ecosystem services in the European AlpsGlob Change Biol 19 1827minus1840
Ellenberg H (1988) Vegetation ecology of Central Europe4th edn Cambridge University Press Edinburgh
Em H (1986) Spruce southern distribution range Spruce for-est on Shara Mountain in Macedonia Contrib Macedon-ian Acad Sci Arts Skopje 5 11minus28
Fajardo A Piper FI Pfund L Koumlrner CH Hoch G (2012)Variation of mobile carbon reserves in trees at the alpinetreeline ecotone is under environmental control NewPhytol 195 794minus802
Fridley JD Grime JP Askew AP Moser B Stevens CJ (2011)Soil heterogeneity buffers community response to cli-mate change in species-rich grassland Glob Change Biol17 2002minus2011
Gartzia M Alados CL Peacuterez-Cabello F (2014) Assessment ofthe effects of biophysical and anthropogenic factors onwoody plant encroachment in dense and sparse moun-tain grasslands based on remote sensing data Prog PhysGeogr 38 201minus217
Gehrig-Fasel J Guisan A Zimmermann NE (2007) Tree lineshifts in the Swiss Alps climate change or land abandon-ment J Veg Sci 18 571minus582
Gikov A Dimitrov P Zhiyanski M (2016) Land cover dynam-ics on the northern slope of the Troyan Passage during 30years period Probl Geogr 1-2 78minus92
Gonzalez P Neilson RP Lenihan JM Drapek RJ (2010)Global patterns in the vulnerability of ecosystems to veg-etation shifts due to climate change Glob Ecol Biogeogr19 755minus768
Gottfried M Pauli H Futschik A Akhalkatsi M and others(2012) Continent-wide response of mountain vegetationto climate change Nat Clim Chang 2 111minus115
Grace J Berninger F Nagy L (2002) Impacts of climatechange on the tree line Ann Bot 90 537minus544
Grunewald K Scheithauer J (2011) Landscape developmentand climate change in Southwest Bulgaria (Pirin Mts)Springer Heidelberg
Grytnes JA Kapfer J Jurasinski G Birks HH and others(2014) Identifying the driving factors behind observedelevational range shifts on European mountains GlobEcol Biogeogr 23 876minus884
Hagedorn F Shiyatov FG Mazepa VS Devi NM and others(2014) Treeline advances along the Urals mountainrange mdash driven by improved winter conditions GlobChange Biol 20 3530minus3543
Hansen AJ Neilson RP Dale VH Flather CH and others(2001) Global change in forests responses of speciescommunities and biomes BioScience 51 765minus779
Harsch MA Bader MY (2011) Treeline formminusa potential keyto understanding treeline dynamics Glob Ecol Biogeogr20 582minus596
Harsch MA Hulme PE McGlone MS Duncan RP (2009) Aretreelines advancing A global meta-analysis of treelineresponse to climate warning Ecol Lett 12 1040minus1049
Haylock MR Hofstra N Klein Tank AMG Klok EJ JonesPD New M (2008) A European daily high-resolutiongridded dataset of surface temperature and precipita-tion J Geophys Res Atmos 113 D20119
Hlaacutesny T Barcza Z Fabrika M Balaacutezs B and others (2011)
148
Cudliacuten et al Drivers of treeline shift
Climate change impacts on growth and carbon balanceof forests in Central Europe Clim Res 47 219minus236
Hlaacutesny T Trombik J Dobor L Barcza Z Barka I (2016)Future climate of the Carpathians climate change hot-spots and implications for ecosystems Reg EnvironChange 16 1495minus1506
Hofgaard A Dalen L Hytteborn H (2009) Tree recruitmentabove the treeline and potential for climate driven tree-line change J Veg Sci 20 1133minus1144
Holtmeier FK (2009) Mountain timberlines Ecology patchi-ness and dynamics 2nd edn Advances in GlobalChange Research 36 Springer Science amp ScienceMediaBV Dordrecht
Holtmeier FK Broll G (2005) Sensitivity and response ofnorthern hemisphere altitudinal and polar treelines toenvironmental change at landscape and local scalesGlob Ecol Biogeogr 14 395minus410
Honnay O Verheyen K Butaye J Jacquemyn H Bossuyt BHermy M (2002) Possible effects of habitat fragmentationand climate change on the range of forest plant speciesEcol Lett 5 525minus530
Huber R Rigling A Bebi P Brand F and others (2013) Sus-tainable land use in mountain regions under globalchange synthesis across scales and disciplines Ecol Soc18 36
Ibaacutentildeez I Clark JS Dietze MC (2009) Estimating colonizationpotential of migrant tree species Glob Change Biol 15 1173minus1188
IUCNSSC (International Union for Conservation of NatureSpecies Survival Commission) (2013) Guidelines for re -introductions and other conservation translocations Version 10 IUCNSpecies Survival Commission Gland
Jankovskyacute L Cudliacuten P gtermaacutek P Moravec I (2004) The pre-diction of development of secondary Norway sprucestands under the impact of climatic change in the Dra-hany highlands (The Czech Republic) Ekologia (Bratisl)23(Suppl 2) 101minus112
Jeniacutek J Lokvenc T (1962) Die alpine Waldgrenze im Krko -noše Gebirge Rozpravy SAV Rada matematickyacutech apriacuterodniacutech ved 72 gteskoslovenskaacute akademie ved Praha
Jobbaacutegy EG Jackson RB (2000) Global controls of forest lineelevation in the northern and southern hemispheresGlob Ecol Biogeogr 9 253minus268
Jonaacutešovaacute M Vaacutevrovaacute E Cudliacuten P (2010) Western Carpa -thian mountain spruce forest after a windthrow naturalregeneration in cleared and uncleared areas For EcolManag 259 1127minus1134
Kittel TGF Steffen WL Chapin FS (2000) Global andregional modelling of Arcticminusboreal vegetation distribu-tion and its sensitivity to altered forcing Glob ChangeBiol 6 1minus18
Klopcic M Jerina K Bonltina A (2010) Long-term changes ofstructure and tree species composition in Dinaric un -even- aged forests Are red deer an important factor EurJ For Res 129 277minus288
Klopcic M Simonltilt T Bonltina A (2015) Comparison of re -generation and recruitment of shade-tolerant and light-demanding tree species in mixed uneven-aged forests ex -periences from the Dinaric region Forestry 88 552minus563
Koumlrner C (2003) Alpine plant life functional plant ecology ofhigh mountain ecosystems with 47 tables Springer Sci-ence amp Business Media BV Dordrecht
Koumlrner C (2012) Alpine treelines Functional ecology of theglobal high elevation tree limits Springer Basel
Koumlrner C Paulsen J (2004) A world-wide study of high alti-
tude treeline temperatures J Biogeogr 31 713minus732Kulakowski D Bebi P Rixen C (2011) The interacting effects
of land use change climate change and suppression ofnatural disturbances on landscape forest structure in theSwiss Alps Oikos 120 216minus225
Kullman L (1988) Holocene history of the forestminusalpine tun-dra ecotone in the Scandes Mountains (central Sweden)New Phytol 108 101minus110
Kullman L (1999) Early holocene tree growth at a high elevation site in the northernmost Scandes of Sweden(Lapland) a palaeobiogeographical case study based onmega fossil evidence Geogr Ann 81 63minus74
Kullman L (2007) Tree line population monitoring of Pinussylvestris in the Swedish Scandes 1973minus2005 implica-tions for tree line theory and climate change ecologyJ Ecol 95 41minus52
Kullman L Oumlberg L (2009) Post-Little Ice Age treeline riseand climate warming in the Swedish Scandes a land-scape ecological perspective J Ecol 97 415minus429
Kyriazopoulos AP Skre O Sarkki S Wielgolaski FE Abra-ham EM Ficko A (2017) Humanminusenvironment dynamicsin European treeline ecosystems a synthesis based onthe DPSIR framework Clim Res 73 17ndash29
Lenoir J Svenning JC (2015) Climate-related range shifts mdasha global multidimensional synthesis and new researchdirections Ecography 38 15minus38
Lenoir J Geacutegout JC Marquet PA de Ruffray P Brisse H(2008) A significant upward shift in plant species opti-mum elevation during the 20th century Science 320 1768minus1771
Lindner M Fitzgerald JB Zimmermann NE Reyer C andothers (2014) Climate change and European forests What do we know what are the uncertainties and whatare the implications for forest management J EnvironManag 146 69minus83
Lloyd AH Rupp TS Fastie CL Srafield AM (2002) Patternsand dynamic of treeline advance on the Seward Pen -isula Alaska J Geophys Res 107 8161
Maacuteliš F Kopeckyacute M Petriacutek P Vladovic J Merganic J VidaT (2016) Life-stage not climate change explains ob -served tree range shifts Glob Change Biol 22 1904minus1914
Mathisen IE Mikheeva A Tutubalina OV Aune S and Hof-gaard A (2014) Fifty years of tree line change in theKhibiny Mountains Russia advantages of combinedremote sensing and dendroecological approaches ApplVeg Sci 17 6ndash16
Millar CI Stephenson NL (2015) Temperate forest healthin an era of emerging megadisturbance Science 349 823minus826
Mina M Bugmann H Klopcic M Cailleret M (2017) Accu-rate modeling of harvesting is key for projecting futureforest dynamics a case study in the Slovenian moun-tains Reg Environ Change 17 49minus64
Moen J Aune K Edenius L Angerbjoumlrn A (2004) Potentialeffects of climate change on treeline position in theSwedish mountains Ecol Soc 9 16 wwwecologyandsocietyorgvol9iss1art16
Palombo C Chirici G Marchetti M Tognetti R (2013) Is landabandonment affecting forest dynamics at high elevationin Mediterranean mountains more than climate changePlant Biosyst 147 1minus11
Palombo C Marchetti M Tognetti R (2014) Mountain vege-tation at risk current perspectives and research reedsPlant Biosyst 148 35minus41
Pauli H Gottfried M Dullinger S Abdaladze O Akhalkatsi
149
Clim Res 73 135ndash150 2017
M Alonso JLB Grabherr G (2012) Recent plant diversitychanges on Europersquos mountain summits Science 336 353minus355
Payette S Fortin MJ Gamachet I (2001) The subarctic forestminustundra the structure of a biome in a changing climate BioScience 51 709minus718
Pearson GR Dawson TP (2003) Predicting the impacts of cli-mate change on the distribution of species Are bio -climate envelope models useful Glob Ecol Biogeogr 12 361minus371
Pedrotti F (2013) Plant and vegetation mapping GeobotanyStudies Springer-Verlag Berlin
Pentildeuelas J Boada M (2003) A global change-induced biomeshift in the Montseny mountains (NE Spain) GlobChange Biol 9 131minus140
Piovesan G Biondi F Filippo AD Alessandrini A MaugeriM (2008) Drought driven growth reduction in old beech(Fagus sylvatica L) forests of the central ApenninesItaly Glob Change Biol 14 1265minus1281
Rabasa SG Granda E Benavides R Kunstler G and others(2013) Disparity in elevational shifts of European trees inresponse to recent climate warming Glob Change Biol19 2490minus2499
Raev I Zhelev P Grozeva M Markov I and others (2011)Programme of measures for adaptation of forests in Bul-garia and mitigate the negative impact of climate changeon them Project FUTURE forest helping Europe tackleclimate change INTERREG IVC ERDF Sofia
Sarkki S Ficko A Grunewald K Nijnik M (2016) Benefitsfrom and threats to European treeline ecosystem serv-ices an exploratory study of stakeholders and gover-nance Reg Environ Change 16 2019minus2032
Savage J Vellend M (2015) Elevational shifts biotic homo -genization and time lags in vegetation change during 40years of climate warming Ecography 38 546minus555
Schwoumlrer C Henne PD Tinner W (2014) A model-data com-parison of Holocene timberline changes in the SwissAlps reveals past and future drivers of mountain forestdynamics Glob Change Biol 20 1512minus1526
Šmilauer P Lepš J (2014) Multivariate analysis of ecologicaldata using Canoco 5 Cambridge University Press Cam-bridge
Ste-Marie C Nelson EA Dabros A Bonneau ME (2011)Assisted migration introduction to a multifaceted con-cept For Chron 87 724minus730
Stevens JT Stafford HD Harrison S Latimer AM (2015) For-est disturbance accelerates thermophilization of under-story plant communities J Ecol 103 1253minus1263
Strid A Andonoski A Andonovski V (2003) Alpine biodiver-sity in Europe Ecological Studies 167 Springer-VerlagBerlin
Tattoni C Ciolli M Ferretti F Cantiani MG (2010) Monitor-ing spatial and temporal pattern of Paneveggio forest(Northern Italy) from 1859 to 2006 iForest Biogeosci For3 72minus80
Ter Braak CJF Šmilauer P (2012) Canoco reference manualand userrsquos guide software for ordination (version 50)Microcomputer Power Ithaca NY
Theurillat JP Guisan A (2001) Potential impact of climatechange on vegetation in the European Alps a reviewClim Change 50 77minus109
Toslashmmervik H Wielgolaski FE Neuvonen S Solberg BHoslashgda KA (2005) Biomass and production on a land-scape level in the mountain birch forests In WielgolaskiFE (ed) Plant ecology herbivory and human impact in
Nordic mountain birch forests Ecos Studies 180Springer Berlin p 53minus70
Treml V Banaš M (2000) Alpine timberline in the HighSudetes Acta Univ Carol Geogr 15 83minus99
Treml V Chuman T (2015) Ecotonal dynamics of the altitu-dinal forest limit are affected by terrain and vegetationstructure variables an example from the Sudetes Moun-tains in Central Europe Arct Antarct Alp Res 47 133minus146
Treml V Migo P (2015) controlling factors limiting timber-line position and shifts in the Sudetes a review GeogrPol 88 55minus70
Urban MC Tewksbury JJ Sheldon KS (2012) On a collisioncourse competition and dispersal differences create no-analogue communities and cause extinctions during cli-mate change Proc R Soc Lond B Biol Sci 279 2072minus2080
Vacek S Matejka K (2010) State and development of phyto-cenoses on research plots in the Krkonoše Mts foreststands J For Sci 56 505minus517
Van Bogaert R Haneca K Hoogesteger J Jonasson C deDapper M Callaghan TV (2011) A century of tree linechanges in sub-Arctic Sweden shows local and regionalvariability and only a minor influence of 20th century climate warming J Biogeogr 38 907minus921
Van Gils H Batsukh O Rossiter D Munthali W Liberato -scioli E (2008) Forecasting the pattern and pace of Fagusforest expansion in Majella National Park Italy ApplVeg Sci 11 539minus546
Velchev V (1997) Types of vegetation In Yordanova MDonchev D (eds) Geography of Bulgaria BAN Sofiap 269minus283 (in Bulgarian)
Vladovic J Merganic J Maacuteliš F Križovaacute E and others (2014)The response of forest vegetation diversity to changes inedaphic-climatic conditions in Slovakia Technickaacute uni-verzita vo Zvolene Zvolene (in Slovak)
Vrška T Adam D Hort L Kolaacuter T Janiacutek F (2009) Europeanbeech (Fagus sylvatica L) and silver fir (Abies alba Mill)rotation in the Carpathians mdash a developmental cycle or alinear trend induced by man For Ecol Manag 258 347minus356
Wielgolaski FE (ed) (2005) Plant ecology herbivory andhuman impact in Nordic mountain birch forests Eco -logical Studies 180 Springer Verlag Berlin
Wielgolaski FE Hofgaard A Holtmeier FK (2017) Sensitivityto environmental change of the treeline ecotone and itsassociated biodiversity in European mountains Clim Res73151ndash166
Wieser G Tausz M (2007) Trees at their upper limit Treelifelimitation at the alpine timberline Springer Dordrecht
Wieser G Holtmeier FK Smith WK (2014) Treelines in achanging global environment In Tausz M Grulke N(eds) Trees in a changing environment Plant Ecophys 9Springer Dordrecht p 221minus263
Zhiyanski M Kolev K Sokolovska M Hursthouse A (2008)Tree species effect on soils in Central Stara PlaninaMountains Nauka Gorata 4 65minus82
Zhiyanski M Gikov A Nedkov S Dimitrov P Naydenova L(2016) Mapping carbon storage using land coverlanduse data in the area of Beklemeto Central Balkan In Koulov B Zhelezov G (eds) Sustainable mountain re -gions challenges and perspectives in southeasternEurope Springer Basel p 53minus65
Zhu K Woodall CW Clark JS (2012) Failure to migrate lackof tree range expansion in response to climate changeGlob Change Biol 18 1042minus1052
150
Editorial responsibility Nils Chr Stenseth Oslo Norway
Submitted April 8 2016 Accepted March 8 2017Proofs received from author(s) July 7 2017
Cudliacuten et al Drivers of treeline shift
research showed that the upward shift of the treelinein the Swiss Alps was predominantly attributable toland abandonment and only in some situations to cli-mate change (Gehrig-Fasel et al 2007) In the Apen-nines in the last few decades tree establishment hasbeen mainly controlled by land use while treegrowth has been controlled by climate pointing to aminor role played by climate in shaping the currenttreeline (Palombo et al 2013)
The impact of climate change and the connectedland use change on biodiversity and ecosystem serv-ices provision in several European countries is sum-marized by Wielgolaski et al (2017) and Kyriazopou-los et al (2017) (both this Special) and Sarkki et al(2016) One of the possible adaptive management op-tions in response to climate change assisted migra-tion as human-assisted movement of species hasbeen frequently debated in the last few years (Ste-Marie et al 2011) It is possible to apply it as a type ofassisted colonization ie intentional movement andrelease of an organism outside its indigenous range toavoid extinction of populations of the focal species(eg Macedonian pine species) or as an ecologicalre placement ie the intentional movement and re-lease of an organism outside its indigenous range toperform a specific ecological function (eg planting ofPinus mugo in the Alps IUCNSSC 2013)
5 CONCLUSIONS
The analysis of 11 mountain areas across Europeshowed that with increasing latitude the treelineand altitudinal width of the treeline ecotone signifi-cantly decreases as do the significance of climaticand soil parameters as barriers against tree speciesshift Although temperature is the overwhelmingcontrolling factor of tree growth and establishment intemperate and boreal treeline ecotones late-sea-sonal drought might also play a driving role in Medi-terranean treeline ecotones Longitude was lessinfluential mostly affecting climate and land usechange effects on increased biodiversity loss as wellas the size of the area that forms one mesoclimateunit affecting altitudinal ecotone width The biggestpart of the commonly observed remaining variabilityin mountain vegetation near the treeline in Europeseems to be caused by geomorphological geologicalpedological and microclimatic variability in combina-tion with different land use history and the presentsocio-economic relations The observed differencesin climatic parameters between the mountain areasof interest in comparison with the reference period
1960minus1990 (09degC) have not explained the relativelyhigh differences in the rate of treeline shift per year(149 m) The predicted variability in temperatureincrease due to global warming (12minus26degC in 2050and 22minus53degC in 2100) could lead to a much biggerdifferentiation in treeline ecotone biodiversity andecosystem processes between southern and northernEuropean mountains in the future Therefore thesedifferences must be taken into account by scientistsand EU policy makers when formulating efficientadaptive forest management strategies for treelineecosystems at the European level (eg assistedmigration of adapted genotypes)
Acknowledgements This paper is based firstly upon workfrom the COST Action ES 1203 SENSFOR supported byCOST (European Cooperation in Science and Technologywwwcosteu) This international work was further sup-ported by projects granted by the Ministry of EducationYouth and Sports of the Czech Republic grant NPU ILO1415 and LD 14039 by the agency APVV SR under pro-jects APVV-14-0086 and APVV-15-0270 We acknowledgethe E-OBS dataset from the EU-FP6 project ENSEMBLES(httpensembles-eumetofficecom) and the data providersin the ECAampD project (wwwecadeu)
LITERATURE CITED
Alados CL Errea P Gartzia M Saiz H Escoacutes J (2014) Posi-tive and negative feedbacks and free-scale pattern distri-bution in rural-population dynamics PLOS ONE 9 e114561
Allen CD Macalady AK Chenchouni H Bachelet D andothers (2010) A global overview of drought and heat-induced tree mortality reveals emerging climate changerisks for forests For Ecol Manag 259 660minus684
Amidžic L Bartula M Jaric S (2012) Syntaxonomy overviewof vegetation of Šar Planina in the Balkan PeninsulaVegetos 25 348minus360
Aune S Hofgaard A Soumlderstroumlm L (2011) Contrasting cli-mate and land use driven tree encroachment pattern ofsub-arctic tundra in Northern Norway and Kola Penin-sula Can J For Res 41 437minus449
Barbeito I Dawes M Rixen C Senn J Bebi P (2012) Factorsdriving mortality and growth at treeline a 30-year exper-iment of 92000 conifers Ecology 93 389minus401
Batllori E Camarero JJ Ninot JM Gutieacuterrez E (2009)Seedling recruitment survival and facilitation in alpinePinus uncinata treeline ecotones Implications andpotential responses to climate warming Glob Ecol Bio-geogr 18 460minus472
Bebi P Teich M Hagedorn F Zurbriggen N Brunner SHGreacutet-Regamey A (2012) Veraumlnderung von Wald undWaldleistungen in der Landschaft Davos im Zuge desKlimawandels Schweiz Z Forstwes 163 493minus501
Bell DM Bradford JB Lauenroth WK (2014) Mountain land-scapes offer few opportunities for high elevation treespecies migration Glob Change Biol 20 1441minus1451
Bergant K (2006) Climate change scenarios for MacedoniaUniversity of Nova Gorica Center for Atmospheric
147
Clim Res 73 135ndash150 2017
Research Nova GoricaBodin J Badeau V Bruno E Cluzeau C Moisselin JM
Walther GR Dupouey JL (2013) Shifts of forest speciesalong an elevational gradient in Southeast France cli-mate change or stand maturation J Veg Sci 24 269minus283
Boncina A (2011) History current status and future pros -pects of uneven-aged forest management in the Dinaricregion an overview Forestry 84 467minus478
Breshears DD Cobb NS Rich PM Price KP and others(2005) Regional vegetation die-off in response to global-change-type drought Proc Natl Acad Sci USA 102 15144minus15148
Camarero JJ Garciacutea-Ruiz JM Sanguumlesa-Barreda G GalvaacutenJD and others (2015a) Recent and intense dynamics in aformerly static Pyrenean treeline Arct Antarct Alp Res47 773minus783
Camarero JJ Gazol A Galvaacuten JD Sanguumlesa-Barreda GGutieacuterrez E (2015b) Disparate effects of global-changedrivers on mountain conifer forests warming-inducedgrowth enhancement in young trees vs CO2 fertilizationin old trees from wet sites Glob Change Biol 21 738minus749
Chen IC Hill JK Ohlemuumlller R Roy DB Thomas CD (2011)Rapid range shifts of species associated with high levelsof climate warming Science 333 1024minus1026
Cudliacuten P Sejaacutek J Pokornyacute J Albrechtovaacute J Bastian OMarek M (2013) Forest ecosystem services under climatechange and air pollution In Matyssek R Clarke N Cud-liacuten P Mikkelsen TN Tuovinen JP Wiesner G Paoletti E(eds) Climate change air pollution and global chal-lenges understanding and perspectives from forestresearch Developments in Environmental Science Vol13 Elsevier Oxford p 521minus546
Dalen L Hofgaard A (2005) Differential regional treelinedynamics in the Scandes mountains Arct Antarct AlpRes 37 284minus296
Dawes MA Philipson CD Fonti P Bebi P Haumlttenschwiler SHagedorn F Rixen C (2015) Soil warming and CO2
enrichment induce biomass shifts in alpine tree line veg-etation 2014 Glob Change Biol 21 2005minus2021
De Frenne P Rodriacuteguez-Saacutenchez F De Schrijver A CoomesDA Hermy M Vangansbeke P Verheyen K (2015) Lightaccelerates plant responses to warming Nat Plants 1 15110
Devi N Hagedorn F Moiseev P Bugmann H ShiyatovSG Mazepa V Rigling A (2008) Expanding forests andchanging growth forms of Siberian larch at the treeline ofthe Polar Urals during the 20th century Glob ChangeBiol 14 1581minus1591
Diaz-Varela R Colombo R Meroni M Calvo-Iglesias MSBuffoni A Tagliaferri A (2010) Spatio-temporal analysisof alpine ecotones A spatial explicit model targeting altitudinal vegetation shifts Ecol Model 221 621minus633
Dirnboumlck T Dullinger S Grabherr G (2003) A regionalimpact assessment of climate and land-use change onalpine vegetation J Biogeogr 30 401minus417
Dobrowski SZ Swanson AK Abatzoglou JT Holden ZASafford HD Schwartz MK Gavin DG (2015) Forest struc-ture and species traits mediate projected recruitmentdeclines in western US tree species Glob Ecol Biogeogr24 917minus927
Dullinger S Dirnboumlck S Grabherr G (2004) Modelling cli-mate change-driven treeline shifts relative effects oftemperature increase dispersal and invisibility J Ecol92 241minus252
Dullinger S Gattringer A Thuiller W Moser D and others
(2012) Extinction debt of high-mountain plants undertwenty-first-century climate change Nat Clim Chang 2 619minus622
Elkin C Gutierrez AG Leuzinger S Manusch C TemperliC Rasche L Bugmann H (2013) A 2degC warmer world isnot safe for ecosystem services in the European AlpsGlob Change Biol 19 1827minus1840
Ellenberg H (1988) Vegetation ecology of Central Europe4th edn Cambridge University Press Edinburgh
Em H (1986) Spruce southern distribution range Spruce for-est on Shara Mountain in Macedonia Contrib Macedon-ian Acad Sci Arts Skopje 5 11minus28
Fajardo A Piper FI Pfund L Koumlrner CH Hoch G (2012)Variation of mobile carbon reserves in trees at the alpinetreeline ecotone is under environmental control NewPhytol 195 794minus802
Fridley JD Grime JP Askew AP Moser B Stevens CJ (2011)Soil heterogeneity buffers community response to cli-mate change in species-rich grassland Glob Change Biol17 2002minus2011
Gartzia M Alados CL Peacuterez-Cabello F (2014) Assessment ofthe effects of biophysical and anthropogenic factors onwoody plant encroachment in dense and sparse moun-tain grasslands based on remote sensing data Prog PhysGeogr 38 201minus217
Gehrig-Fasel J Guisan A Zimmermann NE (2007) Tree lineshifts in the Swiss Alps climate change or land abandon-ment J Veg Sci 18 571minus582
Gikov A Dimitrov P Zhiyanski M (2016) Land cover dynam-ics on the northern slope of the Troyan Passage during 30years period Probl Geogr 1-2 78minus92
Gonzalez P Neilson RP Lenihan JM Drapek RJ (2010)Global patterns in the vulnerability of ecosystems to veg-etation shifts due to climate change Glob Ecol Biogeogr19 755minus768
Gottfried M Pauli H Futschik A Akhalkatsi M and others(2012) Continent-wide response of mountain vegetationto climate change Nat Clim Chang 2 111minus115
Grace J Berninger F Nagy L (2002) Impacts of climatechange on the tree line Ann Bot 90 537minus544
Grunewald K Scheithauer J (2011) Landscape developmentand climate change in Southwest Bulgaria (Pirin Mts)Springer Heidelberg
Grytnes JA Kapfer J Jurasinski G Birks HH and others(2014) Identifying the driving factors behind observedelevational range shifts on European mountains GlobEcol Biogeogr 23 876minus884
Hagedorn F Shiyatov FG Mazepa VS Devi NM and others(2014) Treeline advances along the Urals mountainrange mdash driven by improved winter conditions GlobChange Biol 20 3530minus3543
Hansen AJ Neilson RP Dale VH Flather CH and others(2001) Global change in forests responses of speciescommunities and biomes BioScience 51 765minus779
Harsch MA Bader MY (2011) Treeline formminusa potential keyto understanding treeline dynamics Glob Ecol Biogeogr20 582minus596
Harsch MA Hulme PE McGlone MS Duncan RP (2009) Aretreelines advancing A global meta-analysis of treelineresponse to climate warning Ecol Lett 12 1040minus1049
Haylock MR Hofstra N Klein Tank AMG Klok EJ JonesPD New M (2008) A European daily high-resolutiongridded dataset of surface temperature and precipita-tion J Geophys Res Atmos 113 D20119
Hlaacutesny T Barcza Z Fabrika M Balaacutezs B and others (2011)
148
Cudliacuten et al Drivers of treeline shift
Climate change impacts on growth and carbon balanceof forests in Central Europe Clim Res 47 219minus236
Hlaacutesny T Trombik J Dobor L Barcza Z Barka I (2016)Future climate of the Carpathians climate change hot-spots and implications for ecosystems Reg EnvironChange 16 1495minus1506
Hofgaard A Dalen L Hytteborn H (2009) Tree recruitmentabove the treeline and potential for climate driven tree-line change J Veg Sci 20 1133minus1144
Holtmeier FK (2009) Mountain timberlines Ecology patchi-ness and dynamics 2nd edn Advances in GlobalChange Research 36 Springer Science amp ScienceMediaBV Dordrecht
Holtmeier FK Broll G (2005) Sensitivity and response ofnorthern hemisphere altitudinal and polar treelines toenvironmental change at landscape and local scalesGlob Ecol Biogeogr 14 395minus410
Honnay O Verheyen K Butaye J Jacquemyn H Bossuyt BHermy M (2002) Possible effects of habitat fragmentationand climate change on the range of forest plant speciesEcol Lett 5 525minus530
Huber R Rigling A Bebi P Brand F and others (2013) Sus-tainable land use in mountain regions under globalchange synthesis across scales and disciplines Ecol Soc18 36
Ibaacutentildeez I Clark JS Dietze MC (2009) Estimating colonizationpotential of migrant tree species Glob Change Biol 15 1173minus1188
IUCNSSC (International Union for Conservation of NatureSpecies Survival Commission) (2013) Guidelines for re -introductions and other conservation translocations Version 10 IUCNSpecies Survival Commission Gland
Jankovskyacute L Cudliacuten P gtermaacutek P Moravec I (2004) The pre-diction of development of secondary Norway sprucestands under the impact of climatic change in the Dra-hany highlands (The Czech Republic) Ekologia (Bratisl)23(Suppl 2) 101minus112
Jeniacutek J Lokvenc T (1962) Die alpine Waldgrenze im Krko -noše Gebirge Rozpravy SAV Rada matematickyacutech apriacuterodniacutech ved 72 gteskoslovenskaacute akademie ved Praha
Jobbaacutegy EG Jackson RB (2000) Global controls of forest lineelevation in the northern and southern hemispheresGlob Ecol Biogeogr 9 253minus268
Jonaacutešovaacute M Vaacutevrovaacute E Cudliacuten P (2010) Western Carpa -thian mountain spruce forest after a windthrow naturalregeneration in cleared and uncleared areas For EcolManag 259 1127minus1134
Kittel TGF Steffen WL Chapin FS (2000) Global andregional modelling of Arcticminusboreal vegetation distribu-tion and its sensitivity to altered forcing Glob ChangeBiol 6 1minus18
Klopcic M Jerina K Bonltina A (2010) Long-term changes ofstructure and tree species composition in Dinaric un -even- aged forests Are red deer an important factor EurJ For Res 129 277minus288
Klopcic M Simonltilt T Bonltina A (2015) Comparison of re -generation and recruitment of shade-tolerant and light-demanding tree species in mixed uneven-aged forests ex -periences from the Dinaric region Forestry 88 552minus563
Koumlrner C (2003) Alpine plant life functional plant ecology ofhigh mountain ecosystems with 47 tables Springer Sci-ence amp Business Media BV Dordrecht
Koumlrner C (2012) Alpine treelines Functional ecology of theglobal high elevation tree limits Springer Basel
Koumlrner C Paulsen J (2004) A world-wide study of high alti-
tude treeline temperatures J Biogeogr 31 713minus732Kulakowski D Bebi P Rixen C (2011) The interacting effects
of land use change climate change and suppression ofnatural disturbances on landscape forest structure in theSwiss Alps Oikos 120 216minus225
Kullman L (1988) Holocene history of the forestminusalpine tun-dra ecotone in the Scandes Mountains (central Sweden)New Phytol 108 101minus110
Kullman L (1999) Early holocene tree growth at a high elevation site in the northernmost Scandes of Sweden(Lapland) a palaeobiogeographical case study based onmega fossil evidence Geogr Ann 81 63minus74
Kullman L (2007) Tree line population monitoring of Pinussylvestris in the Swedish Scandes 1973minus2005 implica-tions for tree line theory and climate change ecologyJ Ecol 95 41minus52
Kullman L Oumlberg L (2009) Post-Little Ice Age treeline riseand climate warming in the Swedish Scandes a land-scape ecological perspective J Ecol 97 415minus429
Kyriazopoulos AP Skre O Sarkki S Wielgolaski FE Abra-ham EM Ficko A (2017) Humanminusenvironment dynamicsin European treeline ecosystems a synthesis based onthe DPSIR framework Clim Res 73 17ndash29
Lenoir J Svenning JC (2015) Climate-related range shifts mdasha global multidimensional synthesis and new researchdirections Ecography 38 15minus38
Lenoir J Geacutegout JC Marquet PA de Ruffray P Brisse H(2008) A significant upward shift in plant species opti-mum elevation during the 20th century Science 320 1768minus1771
Lindner M Fitzgerald JB Zimmermann NE Reyer C andothers (2014) Climate change and European forests What do we know what are the uncertainties and whatare the implications for forest management J EnvironManag 146 69minus83
Lloyd AH Rupp TS Fastie CL Srafield AM (2002) Patternsand dynamic of treeline advance on the Seward Pen -isula Alaska J Geophys Res 107 8161
Maacuteliš F Kopeckyacute M Petriacutek P Vladovic J Merganic J VidaT (2016) Life-stage not climate change explains ob -served tree range shifts Glob Change Biol 22 1904minus1914
Mathisen IE Mikheeva A Tutubalina OV Aune S and Hof-gaard A (2014) Fifty years of tree line change in theKhibiny Mountains Russia advantages of combinedremote sensing and dendroecological approaches ApplVeg Sci 17 6ndash16
Millar CI Stephenson NL (2015) Temperate forest healthin an era of emerging megadisturbance Science 349 823minus826
Mina M Bugmann H Klopcic M Cailleret M (2017) Accu-rate modeling of harvesting is key for projecting futureforest dynamics a case study in the Slovenian moun-tains Reg Environ Change 17 49minus64
Moen J Aune K Edenius L Angerbjoumlrn A (2004) Potentialeffects of climate change on treeline position in theSwedish mountains Ecol Soc 9 16 wwwecologyandsocietyorgvol9iss1art16
Palombo C Chirici G Marchetti M Tognetti R (2013) Is landabandonment affecting forest dynamics at high elevationin Mediterranean mountains more than climate changePlant Biosyst 147 1minus11
Palombo C Marchetti M Tognetti R (2014) Mountain vege-tation at risk current perspectives and research reedsPlant Biosyst 148 35minus41
Pauli H Gottfried M Dullinger S Abdaladze O Akhalkatsi
149
Clim Res 73 135ndash150 2017
M Alonso JLB Grabherr G (2012) Recent plant diversitychanges on Europersquos mountain summits Science 336 353minus355
Payette S Fortin MJ Gamachet I (2001) The subarctic forestminustundra the structure of a biome in a changing climate BioScience 51 709minus718
Pearson GR Dawson TP (2003) Predicting the impacts of cli-mate change on the distribution of species Are bio -climate envelope models useful Glob Ecol Biogeogr 12 361minus371
Pedrotti F (2013) Plant and vegetation mapping GeobotanyStudies Springer-Verlag Berlin
Pentildeuelas J Boada M (2003) A global change-induced biomeshift in the Montseny mountains (NE Spain) GlobChange Biol 9 131minus140
Piovesan G Biondi F Filippo AD Alessandrini A MaugeriM (2008) Drought driven growth reduction in old beech(Fagus sylvatica L) forests of the central ApenninesItaly Glob Change Biol 14 1265minus1281
Rabasa SG Granda E Benavides R Kunstler G and others(2013) Disparity in elevational shifts of European trees inresponse to recent climate warming Glob Change Biol19 2490minus2499
Raev I Zhelev P Grozeva M Markov I and others (2011)Programme of measures for adaptation of forests in Bul-garia and mitigate the negative impact of climate changeon them Project FUTURE forest helping Europe tackleclimate change INTERREG IVC ERDF Sofia
Sarkki S Ficko A Grunewald K Nijnik M (2016) Benefitsfrom and threats to European treeline ecosystem serv-ices an exploratory study of stakeholders and gover-nance Reg Environ Change 16 2019minus2032
Savage J Vellend M (2015) Elevational shifts biotic homo -genization and time lags in vegetation change during 40years of climate warming Ecography 38 546minus555
Schwoumlrer C Henne PD Tinner W (2014) A model-data com-parison of Holocene timberline changes in the SwissAlps reveals past and future drivers of mountain forestdynamics Glob Change Biol 20 1512minus1526
Šmilauer P Lepš J (2014) Multivariate analysis of ecologicaldata using Canoco 5 Cambridge University Press Cam-bridge
Ste-Marie C Nelson EA Dabros A Bonneau ME (2011)Assisted migration introduction to a multifaceted con-cept For Chron 87 724minus730
Stevens JT Stafford HD Harrison S Latimer AM (2015) For-est disturbance accelerates thermophilization of under-story plant communities J Ecol 103 1253minus1263
Strid A Andonoski A Andonovski V (2003) Alpine biodiver-sity in Europe Ecological Studies 167 Springer-VerlagBerlin
Tattoni C Ciolli M Ferretti F Cantiani MG (2010) Monitor-ing spatial and temporal pattern of Paneveggio forest(Northern Italy) from 1859 to 2006 iForest Biogeosci For3 72minus80
Ter Braak CJF Šmilauer P (2012) Canoco reference manualand userrsquos guide software for ordination (version 50)Microcomputer Power Ithaca NY
Theurillat JP Guisan A (2001) Potential impact of climatechange on vegetation in the European Alps a reviewClim Change 50 77minus109
Toslashmmervik H Wielgolaski FE Neuvonen S Solberg BHoslashgda KA (2005) Biomass and production on a land-scape level in the mountain birch forests In WielgolaskiFE (ed) Plant ecology herbivory and human impact in
Nordic mountain birch forests Ecos Studies 180Springer Berlin p 53minus70
Treml V Banaš M (2000) Alpine timberline in the HighSudetes Acta Univ Carol Geogr 15 83minus99
Treml V Chuman T (2015) Ecotonal dynamics of the altitu-dinal forest limit are affected by terrain and vegetationstructure variables an example from the Sudetes Moun-tains in Central Europe Arct Antarct Alp Res 47 133minus146
Treml V Migo P (2015) controlling factors limiting timber-line position and shifts in the Sudetes a review GeogrPol 88 55minus70
Urban MC Tewksbury JJ Sheldon KS (2012) On a collisioncourse competition and dispersal differences create no-analogue communities and cause extinctions during cli-mate change Proc R Soc Lond B Biol Sci 279 2072minus2080
Vacek S Matejka K (2010) State and development of phyto-cenoses on research plots in the Krkonoše Mts foreststands J For Sci 56 505minus517
Van Bogaert R Haneca K Hoogesteger J Jonasson C deDapper M Callaghan TV (2011) A century of tree linechanges in sub-Arctic Sweden shows local and regionalvariability and only a minor influence of 20th century climate warming J Biogeogr 38 907minus921
Van Gils H Batsukh O Rossiter D Munthali W Liberato -scioli E (2008) Forecasting the pattern and pace of Fagusforest expansion in Majella National Park Italy ApplVeg Sci 11 539minus546
Velchev V (1997) Types of vegetation In Yordanova MDonchev D (eds) Geography of Bulgaria BAN Sofiap 269minus283 (in Bulgarian)
Vladovic J Merganic J Maacuteliš F Križovaacute E and others (2014)The response of forest vegetation diversity to changes inedaphic-climatic conditions in Slovakia Technickaacute uni-verzita vo Zvolene Zvolene (in Slovak)
Vrška T Adam D Hort L Kolaacuter T Janiacutek F (2009) Europeanbeech (Fagus sylvatica L) and silver fir (Abies alba Mill)rotation in the Carpathians mdash a developmental cycle or alinear trend induced by man For Ecol Manag 258 347minus356
Wielgolaski FE (ed) (2005) Plant ecology herbivory andhuman impact in Nordic mountain birch forests Eco -logical Studies 180 Springer Verlag Berlin
Wielgolaski FE Hofgaard A Holtmeier FK (2017) Sensitivityto environmental change of the treeline ecotone and itsassociated biodiversity in European mountains Clim Res73151ndash166
Wieser G Tausz M (2007) Trees at their upper limit Treelifelimitation at the alpine timberline Springer Dordrecht
Wieser G Holtmeier FK Smith WK (2014) Treelines in achanging global environment In Tausz M Grulke N(eds) Trees in a changing environment Plant Ecophys 9Springer Dordrecht p 221minus263
Zhiyanski M Kolev K Sokolovska M Hursthouse A (2008)Tree species effect on soils in Central Stara PlaninaMountains Nauka Gorata 4 65minus82
Zhiyanski M Gikov A Nedkov S Dimitrov P Naydenova L(2016) Mapping carbon storage using land coverlanduse data in the area of Beklemeto Central Balkan In Koulov B Zhelezov G (eds) Sustainable mountain re -gions challenges and perspectives in southeasternEurope Springer Basel p 53minus65
Zhu K Woodall CW Clark JS (2012) Failure to migrate lackof tree range expansion in response to climate changeGlob Change Biol 18 1042minus1052
150
Editorial responsibility Nils Chr Stenseth Oslo Norway
Submitted April 8 2016 Accepted March 8 2017Proofs received from author(s) July 7 2017
Clim Res 73 135ndash150 2017
Research Nova GoricaBodin J Badeau V Bruno E Cluzeau C Moisselin JM
Walther GR Dupouey JL (2013) Shifts of forest speciesalong an elevational gradient in Southeast France cli-mate change or stand maturation J Veg Sci 24 269minus283
Boncina A (2011) History current status and future pros -pects of uneven-aged forest management in the Dinaricregion an overview Forestry 84 467minus478
Breshears DD Cobb NS Rich PM Price KP and others(2005) Regional vegetation die-off in response to global-change-type drought Proc Natl Acad Sci USA 102 15144minus15148
Camarero JJ Garciacutea-Ruiz JM Sanguumlesa-Barreda G GalvaacutenJD and others (2015a) Recent and intense dynamics in aformerly static Pyrenean treeline Arct Antarct Alp Res47 773minus783
Camarero JJ Gazol A Galvaacuten JD Sanguumlesa-Barreda GGutieacuterrez E (2015b) Disparate effects of global-changedrivers on mountain conifer forests warming-inducedgrowth enhancement in young trees vs CO2 fertilizationin old trees from wet sites Glob Change Biol 21 738minus749
Chen IC Hill JK Ohlemuumlller R Roy DB Thomas CD (2011)Rapid range shifts of species associated with high levelsof climate warming Science 333 1024minus1026
Cudliacuten P Sejaacutek J Pokornyacute J Albrechtovaacute J Bastian OMarek M (2013) Forest ecosystem services under climatechange and air pollution In Matyssek R Clarke N Cud-liacuten P Mikkelsen TN Tuovinen JP Wiesner G Paoletti E(eds) Climate change air pollution and global chal-lenges understanding and perspectives from forestresearch Developments in Environmental Science Vol13 Elsevier Oxford p 521minus546
Dalen L Hofgaard A (2005) Differential regional treelinedynamics in the Scandes mountains Arct Antarct AlpRes 37 284minus296
Dawes MA Philipson CD Fonti P Bebi P Haumlttenschwiler SHagedorn F Rixen C (2015) Soil warming and CO2
enrichment induce biomass shifts in alpine tree line veg-etation 2014 Glob Change Biol 21 2005minus2021
De Frenne P Rodriacuteguez-Saacutenchez F De Schrijver A CoomesDA Hermy M Vangansbeke P Verheyen K (2015) Lightaccelerates plant responses to warming Nat Plants 1 15110
Devi N Hagedorn F Moiseev P Bugmann H ShiyatovSG Mazepa V Rigling A (2008) Expanding forests andchanging growth forms of Siberian larch at the treeline ofthe Polar Urals during the 20th century Glob ChangeBiol 14 1581minus1591
Diaz-Varela R Colombo R Meroni M Calvo-Iglesias MSBuffoni A Tagliaferri A (2010) Spatio-temporal analysisof alpine ecotones A spatial explicit model targeting altitudinal vegetation shifts Ecol Model 221 621minus633
Dirnboumlck T Dullinger S Grabherr G (2003) A regionalimpact assessment of climate and land-use change onalpine vegetation J Biogeogr 30 401minus417
Dobrowski SZ Swanson AK Abatzoglou JT Holden ZASafford HD Schwartz MK Gavin DG (2015) Forest struc-ture and species traits mediate projected recruitmentdeclines in western US tree species Glob Ecol Biogeogr24 917minus927
Dullinger S Dirnboumlck S Grabherr G (2004) Modelling cli-mate change-driven treeline shifts relative effects oftemperature increase dispersal and invisibility J Ecol92 241minus252
Dullinger S Gattringer A Thuiller W Moser D and others
(2012) Extinction debt of high-mountain plants undertwenty-first-century climate change Nat Clim Chang 2 619minus622
Elkin C Gutierrez AG Leuzinger S Manusch C TemperliC Rasche L Bugmann H (2013) A 2degC warmer world isnot safe for ecosystem services in the European AlpsGlob Change Biol 19 1827minus1840
Ellenberg H (1988) Vegetation ecology of Central Europe4th edn Cambridge University Press Edinburgh
Em H (1986) Spruce southern distribution range Spruce for-est on Shara Mountain in Macedonia Contrib Macedon-ian Acad Sci Arts Skopje 5 11minus28
Fajardo A Piper FI Pfund L Koumlrner CH Hoch G (2012)Variation of mobile carbon reserves in trees at the alpinetreeline ecotone is under environmental control NewPhytol 195 794minus802
Fridley JD Grime JP Askew AP Moser B Stevens CJ (2011)Soil heterogeneity buffers community response to cli-mate change in species-rich grassland Glob Change Biol17 2002minus2011
Gartzia M Alados CL Peacuterez-Cabello F (2014) Assessment ofthe effects of biophysical and anthropogenic factors onwoody plant encroachment in dense and sparse moun-tain grasslands based on remote sensing data Prog PhysGeogr 38 201minus217
Gehrig-Fasel J Guisan A Zimmermann NE (2007) Tree lineshifts in the Swiss Alps climate change or land abandon-ment J Veg Sci 18 571minus582
Gikov A Dimitrov P Zhiyanski M (2016) Land cover dynam-ics on the northern slope of the Troyan Passage during 30years period Probl Geogr 1-2 78minus92
Gonzalez P Neilson RP Lenihan JM Drapek RJ (2010)Global patterns in the vulnerability of ecosystems to veg-etation shifts due to climate change Glob Ecol Biogeogr19 755minus768
Gottfried M Pauli H Futschik A Akhalkatsi M and others(2012) Continent-wide response of mountain vegetationto climate change Nat Clim Chang 2 111minus115
Grace J Berninger F Nagy L (2002) Impacts of climatechange on the tree line Ann Bot 90 537minus544
Grunewald K Scheithauer J (2011) Landscape developmentand climate change in Southwest Bulgaria (Pirin Mts)Springer Heidelberg
Grytnes JA Kapfer J Jurasinski G Birks HH and others(2014) Identifying the driving factors behind observedelevational range shifts on European mountains GlobEcol Biogeogr 23 876minus884
Hagedorn F Shiyatov FG Mazepa VS Devi NM and others(2014) Treeline advances along the Urals mountainrange mdash driven by improved winter conditions GlobChange Biol 20 3530minus3543
Hansen AJ Neilson RP Dale VH Flather CH and others(2001) Global change in forests responses of speciescommunities and biomes BioScience 51 765minus779
Harsch MA Bader MY (2011) Treeline formminusa potential keyto understanding treeline dynamics Glob Ecol Biogeogr20 582minus596
Harsch MA Hulme PE McGlone MS Duncan RP (2009) Aretreelines advancing A global meta-analysis of treelineresponse to climate warning Ecol Lett 12 1040minus1049
Haylock MR Hofstra N Klein Tank AMG Klok EJ JonesPD New M (2008) A European daily high-resolutiongridded dataset of surface temperature and precipita-tion J Geophys Res Atmos 113 D20119
Hlaacutesny T Barcza Z Fabrika M Balaacutezs B and others (2011)
148
Cudliacuten et al Drivers of treeline shift
Climate change impacts on growth and carbon balanceof forests in Central Europe Clim Res 47 219minus236
Hlaacutesny T Trombik J Dobor L Barcza Z Barka I (2016)Future climate of the Carpathians climate change hot-spots and implications for ecosystems Reg EnvironChange 16 1495minus1506
Hofgaard A Dalen L Hytteborn H (2009) Tree recruitmentabove the treeline and potential for climate driven tree-line change J Veg Sci 20 1133minus1144
Holtmeier FK (2009) Mountain timberlines Ecology patchi-ness and dynamics 2nd edn Advances in GlobalChange Research 36 Springer Science amp ScienceMediaBV Dordrecht
Holtmeier FK Broll G (2005) Sensitivity and response ofnorthern hemisphere altitudinal and polar treelines toenvironmental change at landscape and local scalesGlob Ecol Biogeogr 14 395minus410
Honnay O Verheyen K Butaye J Jacquemyn H Bossuyt BHermy M (2002) Possible effects of habitat fragmentationand climate change on the range of forest plant speciesEcol Lett 5 525minus530
Huber R Rigling A Bebi P Brand F and others (2013) Sus-tainable land use in mountain regions under globalchange synthesis across scales and disciplines Ecol Soc18 36
Ibaacutentildeez I Clark JS Dietze MC (2009) Estimating colonizationpotential of migrant tree species Glob Change Biol 15 1173minus1188
IUCNSSC (International Union for Conservation of NatureSpecies Survival Commission) (2013) Guidelines for re -introductions and other conservation translocations Version 10 IUCNSpecies Survival Commission Gland
Jankovskyacute L Cudliacuten P gtermaacutek P Moravec I (2004) The pre-diction of development of secondary Norway sprucestands under the impact of climatic change in the Dra-hany highlands (The Czech Republic) Ekologia (Bratisl)23(Suppl 2) 101minus112
Jeniacutek J Lokvenc T (1962) Die alpine Waldgrenze im Krko -noše Gebirge Rozpravy SAV Rada matematickyacutech apriacuterodniacutech ved 72 gteskoslovenskaacute akademie ved Praha
Jobbaacutegy EG Jackson RB (2000) Global controls of forest lineelevation in the northern and southern hemispheresGlob Ecol Biogeogr 9 253minus268
Jonaacutešovaacute M Vaacutevrovaacute E Cudliacuten P (2010) Western Carpa -thian mountain spruce forest after a windthrow naturalregeneration in cleared and uncleared areas For EcolManag 259 1127minus1134
Kittel TGF Steffen WL Chapin FS (2000) Global andregional modelling of Arcticminusboreal vegetation distribu-tion and its sensitivity to altered forcing Glob ChangeBiol 6 1minus18
Klopcic M Jerina K Bonltina A (2010) Long-term changes ofstructure and tree species composition in Dinaric un -even- aged forests Are red deer an important factor EurJ For Res 129 277minus288
Klopcic M Simonltilt T Bonltina A (2015) Comparison of re -generation and recruitment of shade-tolerant and light-demanding tree species in mixed uneven-aged forests ex -periences from the Dinaric region Forestry 88 552minus563
Koumlrner C (2003) Alpine plant life functional plant ecology ofhigh mountain ecosystems with 47 tables Springer Sci-ence amp Business Media BV Dordrecht
Koumlrner C (2012) Alpine treelines Functional ecology of theglobal high elevation tree limits Springer Basel
Koumlrner C Paulsen J (2004) A world-wide study of high alti-
tude treeline temperatures J Biogeogr 31 713minus732Kulakowski D Bebi P Rixen C (2011) The interacting effects
of land use change climate change and suppression ofnatural disturbances on landscape forest structure in theSwiss Alps Oikos 120 216minus225
Kullman L (1988) Holocene history of the forestminusalpine tun-dra ecotone in the Scandes Mountains (central Sweden)New Phytol 108 101minus110
Kullman L (1999) Early holocene tree growth at a high elevation site in the northernmost Scandes of Sweden(Lapland) a palaeobiogeographical case study based onmega fossil evidence Geogr Ann 81 63minus74
Kullman L (2007) Tree line population monitoring of Pinussylvestris in the Swedish Scandes 1973minus2005 implica-tions for tree line theory and climate change ecologyJ Ecol 95 41minus52
Kullman L Oumlberg L (2009) Post-Little Ice Age treeline riseand climate warming in the Swedish Scandes a land-scape ecological perspective J Ecol 97 415minus429
Kyriazopoulos AP Skre O Sarkki S Wielgolaski FE Abra-ham EM Ficko A (2017) Humanminusenvironment dynamicsin European treeline ecosystems a synthesis based onthe DPSIR framework Clim Res 73 17ndash29
Lenoir J Svenning JC (2015) Climate-related range shifts mdasha global multidimensional synthesis and new researchdirections Ecography 38 15minus38
Lenoir J Geacutegout JC Marquet PA de Ruffray P Brisse H(2008) A significant upward shift in plant species opti-mum elevation during the 20th century Science 320 1768minus1771
Lindner M Fitzgerald JB Zimmermann NE Reyer C andothers (2014) Climate change and European forests What do we know what are the uncertainties and whatare the implications for forest management J EnvironManag 146 69minus83
Lloyd AH Rupp TS Fastie CL Srafield AM (2002) Patternsand dynamic of treeline advance on the Seward Pen -isula Alaska J Geophys Res 107 8161
Maacuteliš F Kopeckyacute M Petriacutek P Vladovic J Merganic J VidaT (2016) Life-stage not climate change explains ob -served tree range shifts Glob Change Biol 22 1904minus1914
Mathisen IE Mikheeva A Tutubalina OV Aune S and Hof-gaard A (2014) Fifty years of tree line change in theKhibiny Mountains Russia advantages of combinedremote sensing and dendroecological approaches ApplVeg Sci 17 6ndash16
Millar CI Stephenson NL (2015) Temperate forest healthin an era of emerging megadisturbance Science 349 823minus826
Mina M Bugmann H Klopcic M Cailleret M (2017) Accu-rate modeling of harvesting is key for projecting futureforest dynamics a case study in the Slovenian moun-tains Reg Environ Change 17 49minus64
Moen J Aune K Edenius L Angerbjoumlrn A (2004) Potentialeffects of climate change on treeline position in theSwedish mountains Ecol Soc 9 16 wwwecologyandsocietyorgvol9iss1art16
Palombo C Chirici G Marchetti M Tognetti R (2013) Is landabandonment affecting forest dynamics at high elevationin Mediterranean mountains more than climate changePlant Biosyst 147 1minus11
Palombo C Marchetti M Tognetti R (2014) Mountain vege-tation at risk current perspectives and research reedsPlant Biosyst 148 35minus41
Pauli H Gottfried M Dullinger S Abdaladze O Akhalkatsi
149
Clim Res 73 135ndash150 2017
M Alonso JLB Grabherr G (2012) Recent plant diversitychanges on Europersquos mountain summits Science 336 353minus355
Payette S Fortin MJ Gamachet I (2001) The subarctic forestminustundra the structure of a biome in a changing climate BioScience 51 709minus718
Pearson GR Dawson TP (2003) Predicting the impacts of cli-mate change on the distribution of species Are bio -climate envelope models useful Glob Ecol Biogeogr 12 361minus371
Pedrotti F (2013) Plant and vegetation mapping GeobotanyStudies Springer-Verlag Berlin
Pentildeuelas J Boada M (2003) A global change-induced biomeshift in the Montseny mountains (NE Spain) GlobChange Biol 9 131minus140
Piovesan G Biondi F Filippo AD Alessandrini A MaugeriM (2008) Drought driven growth reduction in old beech(Fagus sylvatica L) forests of the central ApenninesItaly Glob Change Biol 14 1265minus1281
Rabasa SG Granda E Benavides R Kunstler G and others(2013) Disparity in elevational shifts of European trees inresponse to recent climate warming Glob Change Biol19 2490minus2499
Raev I Zhelev P Grozeva M Markov I and others (2011)Programme of measures for adaptation of forests in Bul-garia and mitigate the negative impact of climate changeon them Project FUTURE forest helping Europe tackleclimate change INTERREG IVC ERDF Sofia
Sarkki S Ficko A Grunewald K Nijnik M (2016) Benefitsfrom and threats to European treeline ecosystem serv-ices an exploratory study of stakeholders and gover-nance Reg Environ Change 16 2019minus2032
Savage J Vellend M (2015) Elevational shifts biotic homo -genization and time lags in vegetation change during 40years of climate warming Ecography 38 546minus555
Schwoumlrer C Henne PD Tinner W (2014) A model-data com-parison of Holocene timberline changes in the SwissAlps reveals past and future drivers of mountain forestdynamics Glob Change Biol 20 1512minus1526
Šmilauer P Lepš J (2014) Multivariate analysis of ecologicaldata using Canoco 5 Cambridge University Press Cam-bridge
Ste-Marie C Nelson EA Dabros A Bonneau ME (2011)Assisted migration introduction to a multifaceted con-cept For Chron 87 724minus730
Stevens JT Stafford HD Harrison S Latimer AM (2015) For-est disturbance accelerates thermophilization of under-story plant communities J Ecol 103 1253minus1263
Strid A Andonoski A Andonovski V (2003) Alpine biodiver-sity in Europe Ecological Studies 167 Springer-VerlagBerlin
Tattoni C Ciolli M Ferretti F Cantiani MG (2010) Monitor-ing spatial and temporal pattern of Paneveggio forest(Northern Italy) from 1859 to 2006 iForest Biogeosci For3 72minus80
Ter Braak CJF Šmilauer P (2012) Canoco reference manualand userrsquos guide software for ordination (version 50)Microcomputer Power Ithaca NY
Theurillat JP Guisan A (2001) Potential impact of climatechange on vegetation in the European Alps a reviewClim Change 50 77minus109
Toslashmmervik H Wielgolaski FE Neuvonen S Solberg BHoslashgda KA (2005) Biomass and production on a land-scape level in the mountain birch forests In WielgolaskiFE (ed) Plant ecology herbivory and human impact in
Nordic mountain birch forests Ecos Studies 180Springer Berlin p 53minus70
Treml V Banaš M (2000) Alpine timberline in the HighSudetes Acta Univ Carol Geogr 15 83minus99
Treml V Chuman T (2015) Ecotonal dynamics of the altitu-dinal forest limit are affected by terrain and vegetationstructure variables an example from the Sudetes Moun-tains in Central Europe Arct Antarct Alp Res 47 133minus146
Treml V Migo P (2015) controlling factors limiting timber-line position and shifts in the Sudetes a review GeogrPol 88 55minus70
Urban MC Tewksbury JJ Sheldon KS (2012) On a collisioncourse competition and dispersal differences create no-analogue communities and cause extinctions during cli-mate change Proc R Soc Lond B Biol Sci 279 2072minus2080
Vacek S Matejka K (2010) State and development of phyto-cenoses on research plots in the Krkonoše Mts foreststands J For Sci 56 505minus517
Van Bogaert R Haneca K Hoogesteger J Jonasson C deDapper M Callaghan TV (2011) A century of tree linechanges in sub-Arctic Sweden shows local and regionalvariability and only a minor influence of 20th century climate warming J Biogeogr 38 907minus921
Van Gils H Batsukh O Rossiter D Munthali W Liberato -scioli E (2008) Forecasting the pattern and pace of Fagusforest expansion in Majella National Park Italy ApplVeg Sci 11 539minus546
Velchev V (1997) Types of vegetation In Yordanova MDonchev D (eds) Geography of Bulgaria BAN Sofiap 269minus283 (in Bulgarian)
Vladovic J Merganic J Maacuteliš F Križovaacute E and others (2014)The response of forest vegetation diversity to changes inedaphic-climatic conditions in Slovakia Technickaacute uni-verzita vo Zvolene Zvolene (in Slovak)
Vrška T Adam D Hort L Kolaacuter T Janiacutek F (2009) Europeanbeech (Fagus sylvatica L) and silver fir (Abies alba Mill)rotation in the Carpathians mdash a developmental cycle or alinear trend induced by man For Ecol Manag 258 347minus356
Wielgolaski FE (ed) (2005) Plant ecology herbivory andhuman impact in Nordic mountain birch forests Eco -logical Studies 180 Springer Verlag Berlin
Wielgolaski FE Hofgaard A Holtmeier FK (2017) Sensitivityto environmental change of the treeline ecotone and itsassociated biodiversity in European mountains Clim Res73151ndash166
Wieser G Tausz M (2007) Trees at their upper limit Treelifelimitation at the alpine timberline Springer Dordrecht
Wieser G Holtmeier FK Smith WK (2014) Treelines in achanging global environment In Tausz M Grulke N(eds) Trees in a changing environment Plant Ecophys 9Springer Dordrecht p 221minus263
Zhiyanski M Kolev K Sokolovska M Hursthouse A (2008)Tree species effect on soils in Central Stara PlaninaMountains Nauka Gorata 4 65minus82
Zhiyanski M Gikov A Nedkov S Dimitrov P Naydenova L(2016) Mapping carbon storage using land coverlanduse data in the area of Beklemeto Central Balkan In Koulov B Zhelezov G (eds) Sustainable mountain re -gions challenges and perspectives in southeasternEurope Springer Basel p 53minus65
Zhu K Woodall CW Clark JS (2012) Failure to migrate lackof tree range expansion in response to climate changeGlob Change Biol 18 1042minus1052
150
Editorial responsibility Nils Chr Stenseth Oslo Norway
Submitted April 8 2016 Accepted March 8 2017Proofs received from author(s) July 7 2017
Cudliacuten et al Drivers of treeline shift
Climate change impacts on growth and carbon balanceof forests in Central Europe Clim Res 47 219minus236
Hlaacutesny T Trombik J Dobor L Barcza Z Barka I (2016)Future climate of the Carpathians climate change hot-spots and implications for ecosystems Reg EnvironChange 16 1495minus1506
Hofgaard A Dalen L Hytteborn H (2009) Tree recruitmentabove the treeline and potential for climate driven tree-line change J Veg Sci 20 1133minus1144
Holtmeier FK (2009) Mountain timberlines Ecology patchi-ness and dynamics 2nd edn Advances in GlobalChange Research 36 Springer Science amp ScienceMediaBV Dordrecht
Holtmeier FK Broll G (2005) Sensitivity and response ofnorthern hemisphere altitudinal and polar treelines toenvironmental change at landscape and local scalesGlob Ecol Biogeogr 14 395minus410
Honnay O Verheyen K Butaye J Jacquemyn H Bossuyt BHermy M (2002) Possible effects of habitat fragmentationand climate change on the range of forest plant speciesEcol Lett 5 525minus530
Huber R Rigling A Bebi P Brand F and others (2013) Sus-tainable land use in mountain regions under globalchange synthesis across scales and disciplines Ecol Soc18 36
Ibaacutentildeez I Clark JS Dietze MC (2009) Estimating colonizationpotential of migrant tree species Glob Change Biol 15 1173minus1188
IUCNSSC (International Union for Conservation of NatureSpecies Survival Commission) (2013) Guidelines for re -introductions and other conservation translocations Version 10 IUCNSpecies Survival Commission Gland
Jankovskyacute L Cudliacuten P gtermaacutek P Moravec I (2004) The pre-diction of development of secondary Norway sprucestands under the impact of climatic change in the Dra-hany highlands (The Czech Republic) Ekologia (Bratisl)23(Suppl 2) 101minus112
Jeniacutek J Lokvenc T (1962) Die alpine Waldgrenze im Krko -noše Gebirge Rozpravy SAV Rada matematickyacutech apriacuterodniacutech ved 72 gteskoslovenskaacute akademie ved Praha
Jobbaacutegy EG Jackson RB (2000) Global controls of forest lineelevation in the northern and southern hemispheresGlob Ecol Biogeogr 9 253minus268
Jonaacutešovaacute M Vaacutevrovaacute E Cudliacuten P (2010) Western Carpa -thian mountain spruce forest after a windthrow naturalregeneration in cleared and uncleared areas For EcolManag 259 1127minus1134
Kittel TGF Steffen WL Chapin FS (2000) Global andregional modelling of Arcticminusboreal vegetation distribu-tion and its sensitivity to altered forcing Glob ChangeBiol 6 1minus18
Klopcic M Jerina K Bonltina A (2010) Long-term changes ofstructure and tree species composition in Dinaric un -even- aged forests Are red deer an important factor EurJ For Res 129 277minus288
Klopcic M Simonltilt T Bonltina A (2015) Comparison of re -generation and recruitment of shade-tolerant and light-demanding tree species in mixed uneven-aged forests ex -periences from the Dinaric region Forestry 88 552minus563
Koumlrner C (2003) Alpine plant life functional plant ecology ofhigh mountain ecosystems with 47 tables Springer Sci-ence amp Business Media BV Dordrecht
Koumlrner C (2012) Alpine treelines Functional ecology of theglobal high elevation tree limits Springer Basel
Koumlrner C Paulsen J (2004) A world-wide study of high alti-
tude treeline temperatures J Biogeogr 31 713minus732Kulakowski D Bebi P Rixen C (2011) The interacting effects
of land use change climate change and suppression ofnatural disturbances on landscape forest structure in theSwiss Alps Oikos 120 216minus225
Kullman L (1988) Holocene history of the forestminusalpine tun-dra ecotone in the Scandes Mountains (central Sweden)New Phytol 108 101minus110
Kullman L (1999) Early holocene tree growth at a high elevation site in the northernmost Scandes of Sweden(Lapland) a palaeobiogeographical case study based onmega fossil evidence Geogr Ann 81 63minus74
Kullman L (2007) Tree line population monitoring of Pinussylvestris in the Swedish Scandes 1973minus2005 implica-tions for tree line theory and climate change ecologyJ Ecol 95 41minus52
Kullman L Oumlberg L (2009) Post-Little Ice Age treeline riseand climate warming in the Swedish Scandes a land-scape ecological perspective J Ecol 97 415minus429
Kyriazopoulos AP Skre O Sarkki S Wielgolaski FE Abra-ham EM Ficko A (2017) Humanminusenvironment dynamicsin European treeline ecosystems a synthesis based onthe DPSIR framework Clim Res 73 17ndash29
Lenoir J Svenning JC (2015) Climate-related range shifts mdasha global multidimensional synthesis and new researchdirections Ecography 38 15minus38
Lenoir J Geacutegout JC Marquet PA de Ruffray P Brisse H(2008) A significant upward shift in plant species opti-mum elevation during the 20th century Science 320 1768minus1771
Lindner M Fitzgerald JB Zimmermann NE Reyer C andothers (2014) Climate change and European forests What do we know what are the uncertainties and whatare the implications for forest management J EnvironManag 146 69minus83
Lloyd AH Rupp TS Fastie CL Srafield AM (2002) Patternsand dynamic of treeline advance on the Seward Pen -isula Alaska J Geophys Res 107 8161
Maacuteliš F Kopeckyacute M Petriacutek P Vladovic J Merganic J VidaT (2016) Life-stage not climate change explains ob -served tree range shifts Glob Change Biol 22 1904minus1914
Mathisen IE Mikheeva A Tutubalina OV Aune S and Hof-gaard A (2014) Fifty years of tree line change in theKhibiny Mountains Russia advantages of combinedremote sensing and dendroecological approaches ApplVeg Sci 17 6ndash16
Millar CI Stephenson NL (2015) Temperate forest healthin an era of emerging megadisturbance Science 349 823minus826
Mina M Bugmann H Klopcic M Cailleret M (2017) Accu-rate modeling of harvesting is key for projecting futureforest dynamics a case study in the Slovenian moun-tains Reg Environ Change 17 49minus64
Moen J Aune K Edenius L Angerbjoumlrn A (2004) Potentialeffects of climate change on treeline position in theSwedish mountains Ecol Soc 9 16 wwwecologyandsocietyorgvol9iss1art16
Palombo C Chirici G Marchetti M Tognetti R (2013) Is landabandonment affecting forest dynamics at high elevationin Mediterranean mountains more than climate changePlant Biosyst 147 1minus11
Palombo C Marchetti M Tognetti R (2014) Mountain vege-tation at risk current perspectives and research reedsPlant Biosyst 148 35minus41
Pauli H Gottfried M Dullinger S Abdaladze O Akhalkatsi
149
Clim Res 73 135ndash150 2017
M Alonso JLB Grabherr G (2012) Recent plant diversitychanges on Europersquos mountain summits Science 336 353minus355
Payette S Fortin MJ Gamachet I (2001) The subarctic forestminustundra the structure of a biome in a changing climate BioScience 51 709minus718
Pearson GR Dawson TP (2003) Predicting the impacts of cli-mate change on the distribution of species Are bio -climate envelope models useful Glob Ecol Biogeogr 12 361minus371
Pedrotti F (2013) Plant and vegetation mapping GeobotanyStudies Springer-Verlag Berlin
Pentildeuelas J Boada M (2003) A global change-induced biomeshift in the Montseny mountains (NE Spain) GlobChange Biol 9 131minus140
Piovesan G Biondi F Filippo AD Alessandrini A MaugeriM (2008) Drought driven growth reduction in old beech(Fagus sylvatica L) forests of the central ApenninesItaly Glob Change Biol 14 1265minus1281
Rabasa SG Granda E Benavides R Kunstler G and others(2013) Disparity in elevational shifts of European trees inresponse to recent climate warming Glob Change Biol19 2490minus2499
Raev I Zhelev P Grozeva M Markov I and others (2011)Programme of measures for adaptation of forests in Bul-garia and mitigate the negative impact of climate changeon them Project FUTURE forest helping Europe tackleclimate change INTERREG IVC ERDF Sofia
Sarkki S Ficko A Grunewald K Nijnik M (2016) Benefitsfrom and threats to European treeline ecosystem serv-ices an exploratory study of stakeholders and gover-nance Reg Environ Change 16 2019minus2032
Savage J Vellend M (2015) Elevational shifts biotic homo -genization and time lags in vegetation change during 40years of climate warming Ecography 38 546minus555
Schwoumlrer C Henne PD Tinner W (2014) A model-data com-parison of Holocene timberline changes in the SwissAlps reveals past and future drivers of mountain forestdynamics Glob Change Biol 20 1512minus1526
Šmilauer P Lepš J (2014) Multivariate analysis of ecologicaldata using Canoco 5 Cambridge University Press Cam-bridge
Ste-Marie C Nelson EA Dabros A Bonneau ME (2011)Assisted migration introduction to a multifaceted con-cept For Chron 87 724minus730
Stevens JT Stafford HD Harrison S Latimer AM (2015) For-est disturbance accelerates thermophilization of under-story plant communities J Ecol 103 1253minus1263
Strid A Andonoski A Andonovski V (2003) Alpine biodiver-sity in Europe Ecological Studies 167 Springer-VerlagBerlin
Tattoni C Ciolli M Ferretti F Cantiani MG (2010) Monitor-ing spatial and temporal pattern of Paneveggio forest(Northern Italy) from 1859 to 2006 iForest Biogeosci For3 72minus80
Ter Braak CJF Šmilauer P (2012) Canoco reference manualand userrsquos guide software for ordination (version 50)Microcomputer Power Ithaca NY
Theurillat JP Guisan A (2001) Potential impact of climatechange on vegetation in the European Alps a reviewClim Change 50 77minus109
Toslashmmervik H Wielgolaski FE Neuvonen S Solberg BHoslashgda KA (2005) Biomass and production on a land-scape level in the mountain birch forests In WielgolaskiFE (ed) Plant ecology herbivory and human impact in
Nordic mountain birch forests Ecos Studies 180Springer Berlin p 53minus70
Treml V Banaš M (2000) Alpine timberline in the HighSudetes Acta Univ Carol Geogr 15 83minus99
Treml V Chuman T (2015) Ecotonal dynamics of the altitu-dinal forest limit are affected by terrain and vegetationstructure variables an example from the Sudetes Moun-tains in Central Europe Arct Antarct Alp Res 47 133minus146
Treml V Migo P (2015) controlling factors limiting timber-line position and shifts in the Sudetes a review GeogrPol 88 55minus70
Urban MC Tewksbury JJ Sheldon KS (2012) On a collisioncourse competition and dispersal differences create no-analogue communities and cause extinctions during cli-mate change Proc R Soc Lond B Biol Sci 279 2072minus2080
Vacek S Matejka K (2010) State and development of phyto-cenoses on research plots in the Krkonoše Mts foreststands J For Sci 56 505minus517
Van Bogaert R Haneca K Hoogesteger J Jonasson C deDapper M Callaghan TV (2011) A century of tree linechanges in sub-Arctic Sweden shows local and regionalvariability and only a minor influence of 20th century climate warming J Biogeogr 38 907minus921
Van Gils H Batsukh O Rossiter D Munthali W Liberato -scioli E (2008) Forecasting the pattern and pace of Fagusforest expansion in Majella National Park Italy ApplVeg Sci 11 539minus546
Velchev V (1997) Types of vegetation In Yordanova MDonchev D (eds) Geography of Bulgaria BAN Sofiap 269minus283 (in Bulgarian)
Vladovic J Merganic J Maacuteliš F Križovaacute E and others (2014)The response of forest vegetation diversity to changes inedaphic-climatic conditions in Slovakia Technickaacute uni-verzita vo Zvolene Zvolene (in Slovak)
Vrška T Adam D Hort L Kolaacuter T Janiacutek F (2009) Europeanbeech (Fagus sylvatica L) and silver fir (Abies alba Mill)rotation in the Carpathians mdash a developmental cycle or alinear trend induced by man For Ecol Manag 258 347minus356
Wielgolaski FE (ed) (2005) Plant ecology herbivory andhuman impact in Nordic mountain birch forests Eco -logical Studies 180 Springer Verlag Berlin
Wielgolaski FE Hofgaard A Holtmeier FK (2017) Sensitivityto environmental change of the treeline ecotone and itsassociated biodiversity in European mountains Clim Res73151ndash166
Wieser G Tausz M (2007) Trees at their upper limit Treelifelimitation at the alpine timberline Springer Dordrecht
Wieser G Holtmeier FK Smith WK (2014) Treelines in achanging global environment In Tausz M Grulke N(eds) Trees in a changing environment Plant Ecophys 9Springer Dordrecht p 221minus263
Zhiyanski M Kolev K Sokolovska M Hursthouse A (2008)Tree species effect on soils in Central Stara PlaninaMountains Nauka Gorata 4 65minus82
Zhiyanski M Gikov A Nedkov S Dimitrov P Naydenova L(2016) Mapping carbon storage using land coverlanduse data in the area of Beklemeto Central Balkan In Koulov B Zhelezov G (eds) Sustainable mountain re -gions challenges and perspectives in southeasternEurope Springer Basel p 53minus65
Zhu K Woodall CW Clark JS (2012) Failure to migrate lackof tree range expansion in response to climate changeGlob Change Biol 18 1042minus1052
150
Editorial responsibility Nils Chr Stenseth Oslo Norway
Submitted April 8 2016 Accepted March 8 2017Proofs received from author(s) July 7 2017
Clim Res 73 135ndash150 2017
M Alonso JLB Grabherr G (2012) Recent plant diversitychanges on Europersquos mountain summits Science 336 353minus355
Payette S Fortin MJ Gamachet I (2001) The subarctic forestminustundra the structure of a biome in a changing climate BioScience 51 709minus718
Pearson GR Dawson TP (2003) Predicting the impacts of cli-mate change on the distribution of species Are bio -climate envelope models useful Glob Ecol Biogeogr 12 361minus371
Pedrotti F (2013) Plant and vegetation mapping GeobotanyStudies Springer-Verlag Berlin
Pentildeuelas J Boada M (2003) A global change-induced biomeshift in the Montseny mountains (NE Spain) GlobChange Biol 9 131minus140
Piovesan G Biondi F Filippo AD Alessandrini A MaugeriM (2008) Drought driven growth reduction in old beech(Fagus sylvatica L) forests of the central ApenninesItaly Glob Change Biol 14 1265minus1281
Rabasa SG Granda E Benavides R Kunstler G and others(2013) Disparity in elevational shifts of European trees inresponse to recent climate warming Glob Change Biol19 2490minus2499
Raev I Zhelev P Grozeva M Markov I and others (2011)Programme of measures for adaptation of forests in Bul-garia and mitigate the negative impact of climate changeon them Project FUTURE forest helping Europe tackleclimate change INTERREG IVC ERDF Sofia
Sarkki S Ficko A Grunewald K Nijnik M (2016) Benefitsfrom and threats to European treeline ecosystem serv-ices an exploratory study of stakeholders and gover-nance Reg Environ Change 16 2019minus2032
Savage J Vellend M (2015) Elevational shifts biotic homo -genization and time lags in vegetation change during 40years of climate warming Ecography 38 546minus555
Schwoumlrer C Henne PD Tinner W (2014) A model-data com-parison of Holocene timberline changes in the SwissAlps reveals past and future drivers of mountain forestdynamics Glob Change Biol 20 1512minus1526
Šmilauer P Lepš J (2014) Multivariate analysis of ecologicaldata using Canoco 5 Cambridge University Press Cam-bridge
Ste-Marie C Nelson EA Dabros A Bonneau ME (2011)Assisted migration introduction to a multifaceted con-cept For Chron 87 724minus730
Stevens JT Stafford HD Harrison S Latimer AM (2015) For-est disturbance accelerates thermophilization of under-story plant communities J Ecol 103 1253minus1263
Strid A Andonoski A Andonovski V (2003) Alpine biodiver-sity in Europe Ecological Studies 167 Springer-VerlagBerlin
Tattoni C Ciolli M Ferretti F Cantiani MG (2010) Monitor-ing spatial and temporal pattern of Paneveggio forest(Northern Italy) from 1859 to 2006 iForest Biogeosci For3 72minus80
Ter Braak CJF Šmilauer P (2012) Canoco reference manualand userrsquos guide software for ordination (version 50)Microcomputer Power Ithaca NY
Theurillat JP Guisan A (2001) Potential impact of climatechange on vegetation in the European Alps a reviewClim Change 50 77minus109
Toslashmmervik H Wielgolaski FE Neuvonen S Solberg BHoslashgda KA (2005) Biomass and production on a land-scape level in the mountain birch forests In WielgolaskiFE (ed) Plant ecology herbivory and human impact in
Nordic mountain birch forests Ecos Studies 180Springer Berlin p 53minus70
Treml V Banaš M (2000) Alpine timberline in the HighSudetes Acta Univ Carol Geogr 15 83minus99
Treml V Chuman T (2015) Ecotonal dynamics of the altitu-dinal forest limit are affected by terrain and vegetationstructure variables an example from the Sudetes Moun-tains in Central Europe Arct Antarct Alp Res 47 133minus146
Treml V Migo P (2015) controlling factors limiting timber-line position and shifts in the Sudetes a review GeogrPol 88 55minus70
Urban MC Tewksbury JJ Sheldon KS (2012) On a collisioncourse competition and dispersal differences create no-analogue communities and cause extinctions during cli-mate change Proc R Soc Lond B Biol Sci 279 2072minus2080
Vacek S Matejka K (2010) State and development of phyto-cenoses on research plots in the Krkonoše Mts foreststands J For Sci 56 505minus517
Van Bogaert R Haneca K Hoogesteger J Jonasson C deDapper M Callaghan TV (2011) A century of tree linechanges in sub-Arctic Sweden shows local and regionalvariability and only a minor influence of 20th century climate warming J Biogeogr 38 907minus921
Van Gils H Batsukh O Rossiter D Munthali W Liberato -scioli E (2008) Forecasting the pattern and pace of Fagusforest expansion in Majella National Park Italy ApplVeg Sci 11 539minus546
Velchev V (1997) Types of vegetation In Yordanova MDonchev D (eds) Geography of Bulgaria BAN Sofiap 269minus283 (in Bulgarian)
Vladovic J Merganic J Maacuteliš F Križovaacute E and others (2014)The response of forest vegetation diversity to changes inedaphic-climatic conditions in Slovakia Technickaacute uni-verzita vo Zvolene Zvolene (in Slovak)
Vrška T Adam D Hort L Kolaacuter T Janiacutek F (2009) Europeanbeech (Fagus sylvatica L) and silver fir (Abies alba Mill)rotation in the Carpathians mdash a developmental cycle or alinear trend induced by man For Ecol Manag 258 347minus356
Wielgolaski FE (ed) (2005) Plant ecology herbivory andhuman impact in Nordic mountain birch forests Eco -logical Studies 180 Springer Verlag Berlin
Wielgolaski FE Hofgaard A Holtmeier FK (2017) Sensitivityto environmental change of the treeline ecotone and itsassociated biodiversity in European mountains Clim Res73151ndash166
Wieser G Tausz M (2007) Trees at their upper limit Treelifelimitation at the alpine timberline Springer Dordrecht
Wieser G Holtmeier FK Smith WK (2014) Treelines in achanging global environment In Tausz M Grulke N(eds) Trees in a changing environment Plant Ecophys 9Springer Dordrecht p 221minus263
Zhiyanski M Kolev K Sokolovska M Hursthouse A (2008)Tree species effect on soils in Central Stara PlaninaMountains Nauka Gorata 4 65minus82
Zhiyanski M Gikov A Nedkov S Dimitrov P Naydenova L(2016) Mapping carbon storage using land coverlanduse data in the area of Beklemeto Central Balkan In Koulov B Zhelezov G (eds) Sustainable mountain re -gions challenges and perspectives in southeasternEurope Springer Basel p 53minus65
Zhu K Woodall CW Clark JS (2012) Failure to migrate lackof tree range expansion in response to climate changeGlob Change Biol 18 1042minus1052
150
Editorial responsibility Nils Chr Stenseth Oslo Norway
Submitted April 8 2016 Accepted March 8 2017Proofs received from author(s) July 7 2017