Advertisement

Regional Environmental Change

, Volume 18, Issue 5, pp 1555–1567 | Cite as

Future forest landscapes of the Carpathians: vegetation and carbon dynamics under climate change

  • Ivan KruhlovEmail author
  • Dominik Thom
  • Oleh Chaskovskyy
  • William S. Keeton
  • Robert M. Scheller
Original Article

Abstract

Climate change will alter forest ecosystems and their provisioning of services. Forests in the Carpathian Mountains store high amounts of carbon and provide livelihoods to local people; however, no study has yet assessed their future long-term dynamics under climate change. Therefore, we selected a representative area of 1340 km2 to investigate the effects of changing climate and disturbance regimes on (i) the spatial dynamics of the dominant tree species and forest types and (ii) the trajectories of the associated aboveground live carbon (ALC). We simulated 500 years of change under four Representative Concentration Pathway (RCP) scenarios, incorporating wind and bark beetle disturbances using the LANDIS-II forest change model. Our simulations revealed a lagged adaptation of the forest landscape to climate change. While Picea abies dominance declined in all scenarios, Carpinus betulus expanded at low elevations and Acer pseudoplatanus at mid-elevations. We also found a slow but continuous expansion of Quercus petraea and Q. robur at low elevations and of Fagus sylvatica at mid and high elevations. This change in species composition was accompanied by a significant reduction of ALC: on average over the simulation period, unmitigated climate change reduced ALC between − 2.1% (RCP2.6) and − 14.0% (RCP8.5), while disturbances caused an additional reduction of ALC between − 4.5% (RCP2.6) and − 6.6% (RCP8.5). Therefore, foresighted management strategies are needed to facilitate vegetation adaptation to climate change, with the goal of stabilizing carbon storage and maintaining economic value of future Carpathian forests.

Keywords

Carpathian Mountains Forest landscape Climate change Forest disturbance Aboveground carbon LANDIS-II landscape change model Adaptation 

Introduction

The Carpathian Ecoregion, encompassing an area of 213,000 km2 and spanning seven countries (Czechia, Hungary, Poland, Romania, Serbia, Slovakia, and Ukraine), is the second largest mountain range in Europe predominantly covered with forests (Griffiths et al. 2014). Carpathian forests provide goods and services for human communities at multiple scales (Gratzer and Keeton 2017; Keeton et al. 2013), and intensive forest management over the last two centuries has significantly modified the natural forest cover, specifically by expanding even-aged Norway spruce (Picea abies L. [Karst.]) plantations to increase timber production (e.g., Mikoláš et al. 2017). Still, the ecoregion constitutes a refuge of biodiversity (Mráz and Ronikier 2016) and contains the largest areas of old-growth forests in the temperate zone of Europe (e.g., Hobi et al. 2015). Carpathian ecosystems accumulate large amounts of carbon, especially in old spruce- and silver fir- (Abies alba [Mill.]) dominated forests, which have the highest carbon storage in Europe (Erb 2004). Because carbon sequestration is the most important climate regulating function in European temperate forests (Naudts et al. 2016; Schwaab et al. 2015; Thom et al. 2017b), the Carpathians play a key role in climate change mitigation for the region.

Ecosystem services provided by Carpathian forests, including timber provision and carbon accumulation, are endangered by rapid climate change (Hlásny et al. 2016, 2017; Keeton et al. 2013), which modifies site conditions, affects plant physiology, and alters disturbance regimes (Keeton et al. 2007; Liu and Wimberly 2016). In the Carpathians, similarly to other Central European regions, wind and bark beetles are the most common agents of natural forest disturbances (Hobi et al. 2015; Mezei et al. 2014; Svoboda et al. 2014; Thom et al. 2013). Natural disturbances have already intensified during the past decades in European forests as a result of interactions between forest management and climate change (Seidl et al. 2011) and are predicted to increase further (Seidl et al. 2014). In particular, bark beetle infestations will likely increase due to more favorable thermal conditions and higher susceptibility of host trees due to stronger drought stress (Kautz et al. 2017; Netherer et al. 2015). This would cause significant changes in tree species composition and a reduction in the forest carbon sink capacity (Bonan 2008), making the role of Carpathian forests in timber production and climate change mitigation less certain in the future.

To date, no study has investigated the complex interactions among vegetation, climate change, and disturbances and their effect on the climate regulating function of Carpathian forests (see also Hlásny et al. 2014). Our goal was to delineate possible future transformations of the Carpathian forest vegetation and trajectories of associated carbon stocks at a regional scale. We adopted a dynamic simulation approach (i) to estimate the influence of climate change, windthrow, and bark beetle outbreaks on the community composition of trees and (ii) to derive associated changes in carbon stocks. Human activities, such as logging or planting, were not considered as they constitute a driver of forest change by themselves, obscuring the intrinsic processes of forest development.

We hypothesized a significant reduction in Norway spruce dominance in the study area under climate warming as a consequence of changing site conditions and intensified bark beetle disturbances (see also Boden et al. 2014). We expected that changes in tree species composition, specifically the decline of spruce, would decrease productivity (Hanewinkel et al. 2013) and, as a consequence, reduce carbon stocks. Moreover, we hypothesized significant time lags in the response of forest communities to climate change. However, the intensification of natural disturbances was expected to accelerate the pace of vegetation change (Thom et al. 2017a).

Data and methods

Study area

The study area is located in the center of the Carpathian ecoregion and encompasses the main forest types of Central Europe. The limits are defined by an administrative unit—Rakhiv Raion in Western Ukraine (1892 km2)—which coincides with the headwater basin of the Tysa (Tisza) River (Fig. 1). The elevations range from 265 m in the SW to 2061 m in the NE. The landforms are predominantly low (up to ~ 1000 m) and medium (up to ~ 2000 m) elevation mountains and some hills (up to ~ 500 m). They are composed mostly of alternating sandstone and shale strata (flysch), but areas in the south include metamorphic and volcanic rocks, which produce steep slopes. The regolith is generally well developed on all types of available rocks. The climate is moderately continental temperate with prevailing western winds; the annual mean air temperature varies from 8.8 °C at the lowest elevations to 2.7 °C at the highest locations. The annual precipitation ranges from 1000 to 1500 mm, respectively. Most of the precipitation takes place during the vegetation period. The area is dominated by acid brown soils: Eutric Cambisols occur mostly in the hills and low mountains, while Dystric Cambisols are more typical for the medium mountains (Prots and Kagalo 2012).
Fig. 1

Location and key natural landscape features of the study area (according to Prots and Kagalo 2012). a, boundaries of physiographic regions. 0, subalpine-alpine zone and valley bottoms. 1–5, current natural forest altitudinal zones: 1, Oak Zone; 2, Beech-Oak Zone; 3, Beech Zone; 4, Spruce-Beech Zone; 5, Spruce Zone (see Table 1 for definitions)

Natural vegetation consists of broad-leaf temperate and coniferous montane forests as well as of subalpine and alpine communities occupying the highest ridgetops and is arranged into several altitudinal zones (Table 1, see Fig. 1). European beech (Fagus sylvatica [L.]) is the prevailing species in natural forests here. It mixes with sessile oak (Quercus petraea [Matt.]) at lower elevations (Oak and Beech-Oak Zones) and with Norway spruce and silver fir at mid-high elevations (Beech and Spruce-Beech Zones). The highest forest locations (Spruce Zone) are dominated by spruce. Pedunculated oak (Q. robur [L.]), European hornbeam (Carpinus betulus [L.]), and sycamore maple (Acer pseudoplatanus [L.]) are also very common for the study region (Prots and Kagalo 2012).
Table 1

Natural forest altitudinal zones (according to Prots and Kagalo 2012)

Id

Zone name

Elevation span (m)

Prevailing natural communities

Phytosociological nomenclature

EU Habitats Directive Annex I name

1

Oak

< 750

Carpinion betuli Issler 1931

9170 Galio-Carpinetum oak-hornbeam forests

2

Beech-Oak

500–1000

Dentario glandulosae-Fagenion Oberd. et Th. Müller 1984;

Carpinion betuli Issler 1931

9130 Asperulo-Fagetum beech forests;

9170 Galio-Carpinetum oak-hornbeam forests

3

Beech

700–1300

Dentario glandulosae-Fagenion Oberd. et Th. Müller 1984

9130 Asperulo-Fagetum beech forests

4

Spruce-Beech

1000–1450

Piceion excelsae Pawlowski in Pawlowski et al. 1928;

Luzulo-Fagenion Lohmeyer et Tx. in Tx. 1954

9410 Acidophilous spruce forests (Vaccinio-Piceetea);

9110 Luzulo-Fagetum beech forests;

5

Spruce

1200–1650

Piceion excelsae Pawlowski in Pawlowski et al. 1928

9410 Acidophilous spruce forests (Vaccinio-Piceetea)

Natural forest cover has been reduced by human settlements and agriculture and modified by forest management, although 20% of the area is dedicated to conservation. We limited our interest to the current forest cover excluding fragmented tree vegetation in valley bottoms and the subalpine zone. Thus, the total forested area (TFA) covered by our study equals to 1340 km2. Spruce-dominated forests occupy 68% TFA. These are mostly middle-aged and old (over 100 years) mixed stands with beech, fir, and maple, as well as younger monodominant even-aged plantations. Beech-dominated forest types constitute 26% TFA, while oak-, hornbeam-, fir-, and maple-dominated forests occur at the rest of the area (Ukrderzhlisproekt 2014).

Study design

We focused our research on the tree species that dominate 99.3% TFA (European beech, Norway spruce, sessile and pedunculated oaks, silver fir, European hornbeam, and sycamore maple) and thus representing the vast majority of forest types. We estimated the effects of climate change in conjunction with wind and European spruce bark beetle (Ips typographus) disturbances on the natural development of tree species’ performance with focus on the spatial changes of forest composition and trajectories of the aboveground live carbon (ALC). ALC is associated with aboveground components of living trees (stems, branches, and foliage) and usually constitutes the largest carbon pool in forest ecosystems beyond the first decades after establishment (e.g., Peichl and Arain 2006). ALC can be derived from the aboveground biomass (AGB) data available for the study area, and it is a more meaningful indicator of climate mitigation than biomass.

For climate change simulations, we used four Representative Concentration Pathway scenarios, RCP2.6, RCP4.5, RCP6.0, and RCP8.5 (Moss et al. 2010), and also projected natural succession under baseline climate conditions as a reference. Although a stabilization of the climate was assumed at the end of the century, we explored the development of the vegetation and associated ALC storage for 500 years, considering that forest ecosystems are dominated by long-living organisms with delayed responses to environmental change (Thom et al. 2017a). We simulated ecosystem development under each of the climate change scenarios as well as the baseline scenario, once including and once excluding natural disturbances. This approach enabled us to distinguish between direct climate and indirect disturbance effects on the forest vegetation and the ALC dynamics.

We employed the LANDIS-II forest landscape model (Scheller et al. 2007) in our study (Fig. 2) (see also www.landis-ii.org). LANDIS-II is an extensible process-based stochastic geographically explicit simulation tool that operates at the level of cohorts, which are defined as age groups of tree species within sites (ecologically homogeneous landscape units). A raster geodataset of initial communities represents spatial combinations of the cohorts at the beginning of the simulation period. LANDIS-II simulates competition, establishment, maturation, seed dispersal, reproduction, and senescence/death of the cohorts using species life history attributes and site parameters. Site parameters are linked to a raster geodataset of ecotopes (i.e., ecoregions or land types). Dynamic inputs, which modify parameters of the ecotopes during the simulation period, account for climate change. We applied the Biomass Succession extension (Scheller and Mladenoff 2004) to simulate the dynamics of cohort AGB and thus to trace changes in forest composition and associated ALC. The dynamic input parameters, which control simulated species AGB development, are maximum annual aboveground net primary production (ANPPMAX), maximum possible AGB (BMAX), and probability of establishment (Pest). Species-specific growth and mortality curve parameters asymptotically slow AGB increase as it approaches BMAX. To model wind disturbances, we used the Base Wind extension (Mladenoff and He 1999), while the Biological Disturbance Agent (BDA) extension (Sturtevant et al. 2004) was applied to simulate bark beetle outbreaks. Wind disturbance parameters are linked to ecotopes and include values characterizing the wind rotation periods; minimum, maximum, and mean areas of disturbed patches; and the severity of wind events (i.e., the mortality rate of the cohorts). The BDA extension parameters are associated with host species and their age, ecotopes as modifiers of insect activity, and values identifying spatial dispersal and regional outbreak patterns. The probability of an outbreak can be linked to wind disturbance events.
Fig. 2

Model structure, data sources, and dataflow of the study

Model parametrization

We derived distributions of the species age groups with a 10-year interval along the forest compartments from the database of the Ukrainian Forest Resources Agency (Ukrderzhlisproekt 2014). The units were represented with an accuracy of a 1:25,000 map scale and had a mean size of 6 ha. As a result, a geodataset of the initial forest communities was prepared containing 22,624 records. The ecotopes, which are associated with small landscape units characterizing relatively homogeneous pedogeomorphic and bioclimatic conditions (e.g., Bastian and Steinhardt 2002), were delineated via interpretation of the Shuttle Radar Topography Mission data (Jarvis et al. 2008) as the overlay of two topography-derived spatial structures: (1) 30 altitudinal bioclimatic zones having 50-m spans and (2) four classes of slope elements characterized by topographic position (lower/upper slopes) and surface inclination (moderate/steep slopes). The zones were delineated considering distribution of current natural forest types and represent general bioclimatic provisions. The slope elements depict additional differentiation of site conditions (e.g., soil depth and moisture) within the zones. More details are provided in Supplement 1. The overlay of the two structures produced 120 classes of ecotopes. Geospatial data were reprojected into UTM coordinates and converted into 30-m raster grids.

The life history attributes of the focal species considered here were longevity and sexual maturity (Keeton et al. 2010; Schütt et al. 2007; Trotsiuk et al. 2012), effective and maximum seed dispersal (Vittoz and Engler 2007), and shade tolerance (Niinemets and Valladares 2006), as well as shape parameters for growth and mortality, which were derived based on yield tables and the decline in production efficiency (Landsberg and Waring 1997) (Table 2). We parametrized dynamic inputs for baseline climate conditions as a preliminary step. The ANPPMAX values were estimated using forest inventory data (Ukrderzhlisproekt 2014), while the BMAX was derived via the equation:
$$ {B}_{\mathrm{MAX}}={\mathrm{ANPP}}_{\mathrm{MAX}}\times \mathrm{Longevity}\times {10}^{-1} $$
Table 2

Tree species and their life history attributes

Focal species

Longevity (years)

Sexual maturity (years)

Seed dispersal distance (m)

Shade tolerance

Shape parameters

Effective

Maximum

Growth

Mortality

Abies alba

400

65

60

460

Very high (5)

0.95

18

Acer pseudoplatanus

300

30

100

460

High (4)

0.80

13

Carpinus betulus

150

30

130

460

High (4)

0.75

15

Fagus sylvatica

350

65

13

150

Very high (5)

0.90

17

Picea abies

350

40

60

460

High (4)

0.75

16

Quercus petraea/robur

550

35

34

150

Moderate (3)

0.80

17

This approach generally follows the suggestion of Scheller and Mladenoff (2004), but considers significant differences between longevity of the species, which influences their ability to accumulate biomass (see also Keeling and Phillips 2007). The Pest values were calculated via several iterations. At first, they were estimated for larger natural forest altitudinal zones based on the natural forest type information (see Table 1, Fig. 1). Second, inside each of the zones, the values were differentiated between the four slope elements based on their ANPPMAX—larger Pest values were assigned to the slope elements with higher ANPPMAX. Third, the obtained Pest values were downscaled (interpolated) to the 50-m altitudinal zones to ensure smoother gradients and higher accuracy during climate change simulations. Each of the steps was accompanied by calibration runs of the model. Finally, we validated the parametrization by comparing a simulation of the potential natural forest types under historic climate conditions against 83 field-estimated sites, resulting in an overall accuracy of 79.7%. The generalized results of the parametrization are summarized in Table 3.
Table 3

Baseline mean values of ANPPMAX (g m−2) and Pest for the tree species aggregated for the natural forest altitudinal zones (see Table 1 for definitions)

Focal species

Oak Zone

Beech-Oak Zone

Beech Zone

Spruce-Beech Zone

Spruce Zone

ANPPMAX

P est

ANPPMAX

P est

ANPPMAX

P est

ANPPMAX

P est

ANPPMAX

P est

Abies alba

786

0.09

775

0.17

738

0.40

665

0.39

0

0

A. pseudoplatanus

450

0.09

450

0.18

434

0.30

405

0.41

157

0.08

Fagus sylvatica

768

0.27

738

0.43

681

0.75

555

0.19

0

0

Carpinus betulus

633

0.75

560

0.44

0

0

0

0

0

0

Picea abies

987

0.01

983

0.05

936

0.13

833

0.41

580

0.59

Q. petraea/robur

517

0.71

487

0.47

0

0

0

0

0

0

To derive wind disturbance parameters, we used data from the forestry database (Ukrderzhlisproekt 2014) as well as from studies dealing with disturbances in similar forest types (Hobi et al. 2015; Petritan et al. 2013; Svoboda et al. 2014; Thom et al. 2017a). Simulations of the spatial dispersal of the bark beetle were based on empirically derived kernel functions (Kautz et al. 2011). Parameters on regional outbreak patterns were adopted from Seidl and Rammer (2017). The ecotope classes, as modifiers for bark beetle activity, were weighted based on disturbance information from the forestry database (Ukrderzhlisproekt 2014). As wind is the main driver of bark beetle disturbance (Kautz et al. 2017), we accounted for increasing bark beetle disturbance probability after wind events following Pasztor et al. (2014). The variation of host susceptibility and mortality over age was estimated based on Altenkirch et al. (2002).

We derived the estimates for temperature change over each of the four RCP scenarios (i.e., averages over 15–26 projections per scenario) until the period of 2071–2095 (baseline period 1980–2005). Predicted precipitation changes in this region are minor and thus considered negligible (Alder and Hostetler 2013). It is expected that the temperature will increase between 1.5 and 4.5 °C—depending on the scenario (Supplement 2). Based on a lapse rate within the study area as provided by Prots and Kagalo (2012), we configured dynamic inputs to change each time the predicted annual temperature of the region increased by 0.3 °C. To capture stochastic model behavior, we run simulations with five replications per climate and disturbance scenario (50 runs in total). The simulations covered a 500-year period with a 10-year time step for the Biomass Succession and Biomass Output extensions and with a 5-year time step for the Wind and BDA extensions.

Analysis

We aggregated the replicated outputs of each scenario combination—species AGB distribution maps and tables—to mean values. The aggregated species AGB maps were reclassified to derive the main forest types according to the dominant tree species at the beginning and at the end of the simulation period. Then, for each scenario, we analyzed changes in the areas of the main forest types within the whole study area and along natural forest altitudinal zones (as described in Table 1) as % TFA. Further, AGB was converted into ALC using a factor of 0.5 (Neumann et al. 2016). We derived the relative average deviation of ALC under climate change compared to baseline climate runs over the 500-year period (i.e., we divided the mean ALC over the simulation period of each run under climate change with the mean ALC under baseline climate conditions in the respective disturbance scenarios) for each altitudinal zone (see Table 1) and tested for significant impacts of climate change in each zone by means of Wilcoxon rank-sum test. We also differentiated climate change and disturbance impacts on spatial distributions of the main forest types as well as on ALC for each climate change scenario by comparing the respective scenario combinations (i.e., climate change scenarios vs. baseline climate and disturbance vs. non-disturbance scenarios). The statistical analysis was performed in the R programming language and software environment (R Development Core Team 2017).

Results

Spatial changes in tree species composition

Projections across the 500-year period revealed distinct changes in species dominance in all scenarios (Figs. 3 and 4; Supplement 3). Spruce forests, which are currently most widespread on the landscape occupying 68.3% TFA, contracted following the climate change forcing to 59.7% (RCP2.6)–8.9% (RCP8.5) TFA. In turn, hornbeam- and maple-dominated forests expanded in the area as spruce declined: hornbeam increased its dominance from the initial 2.6 to 16.6% (RCP2.6)–38.3% (RCP8.5) TFA, and maple-dominated forests increased their area from the initial 1.1 to 8.4% (RCP2.6)–24.8% (RCP8.5) TFA. Other “winners” were oak-dominated forests, which are currently barely present on the landscape at 1.6% TFA, but expanded rather equally under all climate change scenarios to 8.2–8.4% TFA. Overall, beech-dominated forests slightly shrunk from the initial 25.8 to 24.5% (RCP2.6)–18.3% (RCP8.5) TFA, and fir-dominated forests remained of similar extent (0.6–1.2% TFA).
Fig. 3

Spatial distributions of simulated forest types (according to dominant species). Shown are initial conditions (a) as well as all future scenarios (including disturbance) assuming baseline climate (b) and projections of climate change (cf) at the end of the 500-year simulation period

Fig. 4

Proportion (%) of the area covered by forest types (according to dominant species) and simulation scenario. Presented are the initial proportion and the proportions at the end of the 500-year simulation period for all scenarios. “nd” indicates non-disturbance scenarios

Disturbances, when compared with non-disturbance simulations (see Fig. 4; Supplement 3), generated additional changes in the species’ dominance on between 5.0% TFA (RCP4.5) and 8.9% TFA (RCP8.5). Spruce-dominated forests additionally shrunk due to disturbances by between 1.9% TFA (RCP4.5) and 3.4% TFA (baseline), while maple-dominated forests benefitted the most from disturbances with an additional expansion in the area of up to 3.3% TFA (baseline). Also, oak-dominated forests gained an additional 0.6–0.7% TFA over all scenarios, and hornbeam-dominated forests expanded by up to 1.2% TFA (RCP4.5). Beech-dominated forests differently reacted on disturbances depending on climate change scenario: they expanded in dominance by 2.4% TFA under the RCP2.6 scenario and shrunk by 2.3% TFA under the RCP8.5 scenario. Moreover, disturbances had a slightly negative influence on fir-dominated forests, decreasing their area by up to 0.3% TFA (RCP4.5).

Changes in forest types varied across altitudinal zones (Supplement 4). Spruce-dominated forests shifted upwards even under baseline climate conditions, while under the RCP8.5 scenario they decreased by almost two thirds even in the highest Spruce Zone (see Figs. 1 and 3). In turn, spruce forests were predominantly substituted by hornbeam-dominated forests at lower locations (mainly in Oak-Beech Zone) as well as by maple- and beech-dominated forests at higher locations (in Spruce-Beech and Spruce Zones). Oak-dominated forests expanded almost exclusively in the lowest altitudinal zone.

Aboveground live carbon trajectories

At the beginning of the simulation period, total ALC stocks were estimated to be 64.3 Mg ha−1, of which spruce had a share of 66.6%. Under baseline climate, total ALC storage increased by 30.8% (non-disturbance scenario) and 25.8% (disturbance scenario) on average over the simulation period (Supplement 5). We found diverging impacts of climate change on ALC trajectories (Fig. 5). Initially, ALC increased slightly more in climate change scenarios compared to baseline scenarios, a difference of between 2.3% (RCP8.5) and 2.8% (RCP4.5). However, this trend reversed after 70–90 years, and ALC stocks strongly decreased with a stabilization of ALC between 300 (RCP8.5) and 370 (RCP2.6) years. This resulted in considerably lower values than under the baseline conditions by between − 8.1% (RCP2.6) and − 35.4% (RCP8.5). The trajectories of total ALC storage were mainly defined by the dynamics of spruce-associated ALC as spruce represents the largest carbon stock (Supplement 6). On average over the simulation period, ALC decreased in response to climate change by between 1.7 Mg ha−1 under the RCP2.6 scenario and 11.8 Mg ha−1 under the RCP8.5 scenario, i.e., between 2.1 and 14.0% (Fig. 6; Supplement 7).
Fig. 5

Aboveground live carbon (ALC) trajectories under climate change scenarios, relative to baseline scenarios. Dashed lines indicate non-disturbance scenarios

Fig. 6

Average change in aboveground live carbon (ALC) over the 500-year simulation period with the relative impacts of climate change and disturbance differentiated

Disturbances smoothed ALC trajectories. They caused additional reductions of ALC during the first 300–370 years, but increased ALC afterwards (see Fig. 5). However, on average during the whole simulation period, disturbances reduced ALC from − 3.2 Mg ha−1 under the baseline scenario to − 4.8 Mg ha−1 under the RCP8.5 scenario (see Supplement 5)—i.e., between − 3.8 and − 6.6% (see Fig. 6; Supplement 7). While Wilcoxon rank-sum tests revealed a significant impact of climate on ALC in all altitudinal zones (p < 0.05), the effect size of ALC changes varied strongly (Fig. 7; Supplement 8). The largest reduction occurred at mid-elevations (Beech Zone); here, average ALC declined across all climate change scenarios by 16.7%. Conversely, ALC increased at the highest locations within the Spruce Zone by an average of 9.6%. The largest variations of ALC under different scenarios were detected in the Spruce-Beech Zone (ranging from − 25.9 to + 3.2%), while the smallest variations occurred at lower elevations in the Oak Zone (ranging from + 0.5 to + 2.2%).
Fig. 7

Relative effect of climate change on aboveground live carbon (ALC) in each altitudinal zone (see Table 1 for definitions). The relative divergence between climate change and baseline climate shown here includes disturbed and undisturbed scenario combinations. Bold lines indicate the median effect size; boxes, the interquartile range; whiskers, the total range; and stars, the significance of the difference for each altitudinal zone

Discussion

Climate change effects on forest vegetation spatial distribution

Our simulations indicated similar patterns in the spatial dynamics of forest communities under all climate change scenarios. These included contraction of spruce-dominated forests; expansion of oak-, hornbeam-, and maple-dominated forests; and upward shift of beech- and fir-dominated forests. The future contraction of spruce forests in the Ukrainian Carpathians has also been revealed by Shvidenko et al. (2017). They identified soil water stress in response to increasing air temperatures as the most detrimental factor for a decrease in spruce dominance under climate change. This finding is consistent with previous studies on vegetation dynamics under climate change in Europe, which suggest a reduction of spruce and its substitution by other late seral species, such as beech and oak (Hanewinkel et al. 2013; Hickler et al. 2012; Thom et al. 2017a). However, according to our results, these species were not able to immediately colonize suitable “vacant” sites owing to slow migration rates (see also Scheller and Mladenoff 2008; Meier et al. 2012). For example, oak, which currently grows at the lowest elevations within the study area, expanded almost equally in all climate change scenarios, because the main limiting factor of its spatial propagation was short seed dispersal distance and relatively low shade tolerance. This provided an opportunity for more readily dispersing hornbeam and maple to temporarily fill in the gaps. Consequently, hornbeam- and maple-dominated forests expanded considerably under all scenarios (see Figs. 3 and 4).

Intensification of disturbances was an additional driver of forest landscape change. While spruce experienced the most significant losses due to disturbances, maple, which grows as a subdominant species in mixed forests together with spruce, gained the most area via fast expansion. Oak also benefited slightly from wind disturbances, which contributed to the reduction of more shade-tolerant beech and hornbeam at lower elevations. Although disturbances accelerated succession, especially during the first 200 years (see Fig. 5), the forest landscape did not reach a dynamic equilibrium with climate conditions at the end of a 500-year simulation period. This was a consequence of the slow migration rates of beech and oak, which were not able to occupy all suitable ecotopes over that time frame. The century-long lead times in forest adaptation to novel environmental conditions correspond to the findings of Thom et al. (2017a).

Carbon feedbacks under climate change

Under the baseline climate scenarios, the increase of ALC (see Supplement 5) was mainly caused by the legacy of past forest management, which resulted in vast areas of young and middle-aged, mainly spruce-dominated, forests (see Fig. 3a) that continued to mature and accumulate carbon (see also Loudermilk et al. 2013). Under the climate change scenarios, simulated trajectories of ALC stocks revealed similar patterns relative to baseline conditions: initial increase, which was followed by the significant decrease and stabilization (see Fig. 5). However, the magnitude of ALC variations differed considerably and was positively dependent on the extent of simulated warming and disturbance impact (see Figs. 5 and 6). The relative initial growth of ALC in the climate change scenarios is associated with accelerated productivity of spruce and other species, which compensated for the negative impact of intensifying disturbances (see Supplement 6). This effect echoes ecophysiological studies, which demonstrate initial productivity growth of spruce in optimum locations under climate change (e.g., Ge et al. 2013; Hlásny et al. 2014). The subsequent abrupt ALC decrease is the result of spruce forest decline at low and medium elevations due to intensifying disturbances and increasing dominance of hornbeam and maple. These species could not compensate for previous losses in ALC stocks because of relatively low (compared to spruce) biomass accumulation capacity (see Table 2; Supplement 5). The final stabilization of ALC was mainly associated with the continuous expansion of beech forests at higher elevations and of oak forests at lower elevations, which were gradually substituting hornbeam- and maple-dominated forest types.

Changes of ALC stocks were distributed unevenly across the landscape and were mainly associated with the availability of spruce. ALC changes at lower elevations (Oak and Beech-Oak Zones, see Fig. 7) were smaller because of lower shares of spruce in the initial composition there. However, ALC markedly declined at mid-elevations (Beech and Spruce-Beech Zones) under climate change scenarios, where spruce is currently the dominant species largely because of past forest management (Keeton et al. 2013). Contrastingly, simulated ALC increased at high elevations (Spruce Zone), where site conditions remained suitable for spruce (under all scenarios, except RCP8.5) and productivity improved owing to a moderate temperature increase (see also Ge et al. 2013).

From the perspective of climate regulation, simulated ALC trajectories (see Fig. 5) reveal variable feedbacks of forest ecosystems to climate change. An initial increase of ALC was associated with a negative (self-damping) feedback, which implied an increasing role for forests as a carbon sink and climate mitigation. It was succeeded, however, by a strong positive (self-amplifying) feedback (due to spruce decline) to climate change, which turned these forests into a carbon source thus reinforcing climate warming. Future studies should aim to consider these variable feedbacks in the climate-vegetation system (e.g., Thom et al. 2017b).

Study limitations

Our study is based on the empirical data available for parametrization. For example, contemporary species distribution was estimated from a forest inventory. However, even detailed terrestrial inventories tend to underestimate the presence of less common species (e.g., Didion et al. 2009), such as single seed trees and seedlings of beech and fir in spruce monocultures, which may act as additional sources of species spatial propagation. Therefore, we assume that the expansion of slowly migrating species, especially beech, may be somewhat underestimated in our simulations. Similarly, our estimates of ANPPMAX, BMAX, and Pest were derived from available data that all contain associated uncertainties, particularly under novel climatic conditions and heightened CO2.

Furthermore, we were not able to simulate all processes within the constraints of our study design. Consideration of other carbon pools would likely have changed the results slightly. For example, deadwood generated by disturbances has lagged carbon flux dynamics reflecting species-specific residency periods and decomposition rates. Additional disturbances (e.g., frost damage, fungal pathogens) were not incorporated. These shortcomings, however, do not influence the relative differences between species and areas here, because all scenarios contained the same bias. In particular, human activity has profoundly altered this landscape and will similarly bend future trajectories; subsequent research will focus on this dimension of forest change. Finally, there is additional uncertainty outside the bounds of our research, including economic shifts that can induce sudden changes in land use, abrupt climatic change, or other plausible scenarios.

The need for foresighted forest management

Besides reduced carbon stocks, climate change and increasing disturbances will also negatively affect other ecosystem services, including timber provision. Under past climatic conditions, spruce-dominated forests have been highly productive and economically most important in Europe (Hanewinkel et al. 2013). Thus, the projected future declines in spruce constitute a key challenge for forest managers, who need to find new solutions for sustainable revenue production from timber harvests. This is particularly important for low-income and timber-dependent communities, such as in parts of the Eastern and Southern Carpathians. Moreover, water-regulating functions of the forest should be reassessed in the face of intensifying disturbances. These may increase soil erosion, floods, mudflows, and other natural hazards (Dale et al. 2001; Maroschek et al. 2015) that endanger agriculture, human settlements, and infrastructure in these regions.

To optimize the provisioning of ecosystem services, such as carbon and timber in forest landscapes, we recommend that managers consider (i) fostering highly productive tree species where they are expected to be adaptable in the future and (ii) facilitating the adaptation of forest vegetation to novel environmental conditions where disturbances are expected to increase significantly. As presented here, climate-induced reductions in carbon will likely be most pronounced in the mid-altitudinal zones occupied by mixed spruce-fir-beech forests, which are regarded as the most productive in Europe (Erb 2004). A strategy to maintain the performance of these forests types would be to adapt the shares of the species at landscape scales, considering changing climate and landscape gradients. Thereby the proportion of spruce at lower elevations must be generally low to avoid severe disturbances, but it might be kept as an important species in higher altitudinal zones. Instead of spruce, an increased share of fir might constitute an alternative for maintaining valuable conifers as this species can produce very high timber volumes while being resistant to various disturbance agents (Kerr et al. 2015). Facilitation of compositional changes includes planting and fostering of tree species currently dominating lower altitudinal zones. For instance, it might be beneficial for the future provisioning of ecosystem services to facilitate migration of oak to higher locations before hornbeam, which has a lower economic value and carbon accumulation capacity, occupies these ecotopes. In general, a foresighted management strategy that considers changing environmental conditions and the needs of the local population for forest goods and services is required for the sustainable development of the Carpathians.

Conclusions

We investigated the effects of climate change on the long-term development of a forest landscape and associated carbon stocks of the Carpathians. We projected changes in competition among the main tree species as well as in disturbance regimes of the region over several centuries. Hence, our investigation constitutes the first study estimating the future forest and carbon dynamics of the Carpathians by means of a forest landscape model including interactions between vegetation, climate, and disturbance regimes.

We found that the initial spatial distribution of species and their migration rates drive vegetation dynamics on the regional level. The adaptation of forest landscape and associated carbon stocks may lag several centuries behind climate changes. This climate disequilibrium of forest ecosystems constitutes a challenge for forest management of the region, where people strongly rely on forests for their livelihoods. Active measures, like planting of oak, beech, and fir at higher locations, may facilitate the adjustment process. However, a question remains whether such adaptation should be applied in strictly protected areas, which encompass large portions of Carpathian forests. Our study of natural forest succession under climate change is the first step toward an integrative research initiative investigating the interactions between forests and the local population of the Carpathians in a changing world.

Notes

Acknowledgments

This paper is part of a larger project initiated by Garry Sotnik to introduce LANDIS-II to the Ukrainian Carpathians for the study of human adaptation to climate change. We are grateful to Garry Sotnik for organizing the team and raising funding for the Ukrainian team members. We would also like to thank two anonymous reviewers for very helpful remarks, which led to the significant improvement of this paper.

Supplementary material

10113_2018_1296_MOESM1_ESM.pdf (1.3 mb)
ESM 1 (PDF 1371 kb)

References

  1. Alder JR, Hostetler SW (2013) CMIP5 global climate change viewer. US Geological Survey. http://regclim.coas.oregonstate.edu/gccv/index.html. Accessed 8 July 2016
  2. Altenkirch W, Majunke C, Ohnesorge B (2002) Waldschutz auf ökologischer Grundlage. Ulmer, StuttgartGoogle Scholar
  3. Bastian O, Steinhardt U (2002) Development and perspectives of landscape ecology. Springer, Dordrecht.  https://doi.org/10.1007/978-94-017-1237-8 CrossRefGoogle Scholar
  4. Boden S, Kahle H-P, von Wilpert K, Spiecker H (2014) Resilience of Norway spruce (Picea abies (L.) Karst) growth to changing climatic conditions in Southwest Germany. Forest Ecol Manag 315:12–21.  https://doi.org/10.1016/j.foreco.2013.12.015 CrossRefGoogle Scholar
  5. Bonan GB (2008) Forests and climate change: forcings, feedbacks, and the climate benefits of forests. Science 320(5882):1444–1449.  https://doi.org/10.1126/science.1155121 CrossRefGoogle Scholar
  6. Dale VH, Joyce LA, McNnulty S, Neilson RP, Ayres MP, Flannigan MD, Hanson PJ, Irland LC, Lugo AE, Peterson CJ, Simberloff D, Swanson FJ, Stocks BJ, Wotton MB (2001) Climate change and forest disturbances: climate change can affect forests by altering the frequency, intensity, duration, and timing of fire, drought, introduced species, insect and pathogen outbreaks, hurricanes, windstorms, ice storms, or landslides. Bioscience 51(9):723–734. https://doi.org/10.1641/0006-3568(2001)051[0723:CCAFD]2.0.CO;2Google Scholar
  7. Didion M, Kupferschmid AD, Lexer MJ, Rammer W, Seidl R, Bugmann H (2009) Potentials and limitations of using large-scale forest inventory data for evaluating forest succession models. Ecol Model 220(2):133–147.  https://doi.org/10.1016/j.ecolmodel.2008.09.021 CrossRefGoogle Scholar
  8. Erb K-H (2004) Land use related changes in aboveground carbon stocks of Austria’s terrestrial ecosystems. Ecosystems 7(5):563–572.  https://doi.org/10.1007/s10021-004-0234-4 CrossRefGoogle Scholar
  9. Ge Z-M, Kellomäki S, Peltola H, Zhou X, Väisänen H, Strandman H (2013) Impacts of climate change on primary production and carbon sequestration of boreal Norway spruce forests: Finland as a model. Clim Chang 118(2):259–273.  https://doi.org/10.1007/s10584-012-0607-1 CrossRefGoogle Scholar
  10. Gratzer G, Keeton WS (2017) Mountain forests and sustainable development: the potential for achieving the United Nations’ 2030 Agenda. Mt Res Dev 37(3):246–253.  https://doi.org/10.1659/MRD-JOURNAL-D-17-00093.1 CrossRefGoogle Scholar
  11. Griffiths P, Kuemmerle T, Baumann M, Radeloff VC, Abrudan IV, Lieskovsky J, Munteanu C, Ostapowicz K, Hostert P (2014) Forest disturbances, forest recovery, and changes in forest types across the Carpathian ecoregion from 1985 to 2010 based on Landsat image composites. Remote Sens Environ 151:72–88.  https://doi.org/10.1016/j.rse.2013.04.022 CrossRefGoogle Scholar
  12. Hanewinkel M, Cullmann DA, Schelhaas M-J, Nabuurs G-J, Zimmermann NE (2013) Climate change may cause severe loss in the economic value of European forest land. Nature Clim Change 3(3):203–207.  https://doi.org/10.1038/nclimate1687 CrossRefGoogle Scholar
  13. Hickler T, Vohland K, Feehan J, Miller PA, Smith B, Costa L, Giesecke T, Fronzek S, Carter TR, Cramer W, Kühn I, Sykes MT (2012) Projecting the future distribution of European potential natural vegetation zones with a generalized, tree species-based dynamic vegetation model. Glob Ecol Biogeogr 21(1):50–63.  https://doi.org/10.1111/j.1466-8238.2010.00613.x CrossRefGoogle Scholar
  14. Hlásny T, Barcza Z, Barka I, Merganičová K, Sedmák R, Kern A, Pajtík J, Balázs B, Fabrika M, Churkina G (2014) Future carbon cycle in mountain spruce forests of Central Europe: modelling framework and ecological inferences. Forest Ecol Manag 328:55–68.  https://doi.org/10.1016/j.foreco.2014.04.038 CrossRefGoogle Scholar
  15. Hlásny T, Barka I, Kulla L, Bucha T, Sedmák R, Trombik J (2017) Sustainable forest management in a mountain region in the Central Western Carpathians, northeastern Slovakia: the role of climate change. Reg Environ Chang 17(1):65–77.  https://doi.org/10.1007/s10113-015-0894-y CrossRefGoogle Scholar
  16. Hlásny T, Trombik J, Dobor L, Barcza Z, Barka I (2016) Future climate of the Carpathians: climate change hot-spots and implications for ecosystems. Reg Environ Chang 16(5):1495–1506.  https://doi.org/10.1007/s10113-015-0890-2 CrossRefGoogle Scholar
  17. Hobi ML, Commarmot B, Bugmann H (2015) Pattern and process in the largest primeval beech forest of Europe (Ukrainian Carpathians). J Veg Sci 26(2):323–336.  https://doi.org/10.1111/jvs.12234 CrossRefGoogle Scholar
  18. Jarvis A, Reuter HI, Nelson A, Guevara E (2008) Hole-filled SRTM for the globe Version 4, available from the CGIAR-CSI SRTM 90m Database. http://srtm.csi.cgiar.org. Accessed 12 September 2009
  19. Kautz M, Dworschak K, Gruppe A, Schopf R (2011) Quantifying spatio-temporal dispersion of bark beetle infestations in epidemic and non-epidemic conditions. Forest Ecol Manag 262(4):598–608.  https://doi.org/10.1016/j.foreco.2011.04.023 CrossRefGoogle Scholar
  20. Kautz M, Meddens AJH, Hall RJ, Arneth A (2017) Biotic disturbances in Northern Hemisphere forests—a synthesis of recent data, uncertainties and implications for forest monitoring and modelling. Glob Ecol Biogeogr 26(5):533–552.  https://doi.org/10.1111/geb.12558 CrossRefGoogle Scholar
  21. Keeling HC, Phillips OL (2007) The global relationship between forest productivity and biomass. Glob Ecol Biogeogr 16(5):618–631.  https://doi.org/10.1111/j.1466-8238.2007.00314.x CrossRefGoogle Scholar
  22. Keeton WS, Franklin JF, Mote PW (2007) Climate variability, climate change, and western wildfire with implications for the suburban-wildland interface. In: Troy A, Kennedy R (eds) Living on the edge: economic, institutional and management perspectives on wildfire hazard in the urban interface. Elsevier, New York, pp 223–255.  https://doi.org/10.1016/S1569-3740(06)06013-5 CrossRefGoogle Scholar
  23. Keeton WS, Chernyavskyy M, Gratzer G, Main-Knorn M, Shpylchak M, Bihun Y (2010) Structural characteristics and aboveground biomass of old-growth spruce–fir stands in the eastern Carpathian mountains, Ukraine. Plant Biosyst 144(1):148–159.  https://doi.org/10.1080/11263500903560512 CrossRefGoogle Scholar
  24. Keeton WS, Angelstam PK, Bihun Y, Chernyavdkyy M, Crow SM, Deyneka A, Elbakidze M, Farley J, Kovalyshyn V, Kruhlov I, Mahura B, Myklush S, Nunery JS, Soloviy I, Zahvoyska L (2013) Sustainable forest management alternatives for the Carpathian Mountains with a focus on Ukraine. In: Kozak J, Ostapowicz K, Bytnerowicz A, Wyżga B (eds) The Carpathians: integrating nature and society towards sustainability. Springer, Berlin/Heidelberg, pp 331–352.  https://doi.org/10.1007/978-3-642-12725-0_24 CrossRefGoogle Scholar
  25. Kerr G, Stokes V, Peace A, Jinks R (2015) Effects of provenance on the survival, growth and stem form of European silver fir (Abies alba Mill.) in Britain. Eur J For Res 134(2):349–363.  https://doi.org/10.1007/s10342-014-0856-9 CrossRefGoogle Scholar
  26. Landsberg JJ, Waring RH (1997) A generalised model of forest productivity using simplified concepts of radiation-use efficiency, carbon balance and partitioning. Forest Ecol Manag 95(3):209–228.  https://doi.org/10.1016/S0378-1127(97)00026-1 CrossRefGoogle Scholar
  27. Liu Z, Wimberly MC (2016) Direct and indirect effects of climate change on projected future fire regimes in the western United States. Sci Total Environ 542(Pt A):65–75.  https://doi.org/10.1016/j.scitotenv.2015.10.093 CrossRefGoogle Scholar
  28. Loudermilk EL, Scheller RM, Weisberg PJ, Yang J, Dilts TE, Karam SL, Skinner C (2013) Carbon dynamics in the future forest: the importance of long-term successional legacy and climate–fire interactions. Glob Change Biol 19:3502–3515.  https://doi.org/10.1111/gcb.12310 CrossRefGoogle Scholar
  29. Maroschek M, Rammer W, Lexer MJ (2015) Using a novel assessment framework to evaluate protective functions and timber production in Austrian mountain forests under climate change. Reg Environ Chang 15(8):1543–1555.  https://doi.org/10.1007/s10113-014-0691-z CrossRefGoogle Scholar
  30. Mikoláš M, Tejkal M, Kuemmerle T, Griffiths P, Svoboda M, Hlásny T, Leitão PJ, Morrissey RC (2017) Forest management impacts on capercaillie (Tetrao urogallus) habitat distribution and connectivity in the Carpathians. Landsc Ecol 32(1):163–179.  https://doi.org/10.1007/s10980-016-0433-3 CrossRefGoogle Scholar
  31. Mezei P, Grodzki W, Blaženec M, Jakuš R (2014) Factors influencing the wind–bark beetles’ disturbance system in the course of an Ips typographus outbreak in the Tatra Mountains. Forest Ecol Manag 312:67–77.  https://doi.org/10.1016/j.foreco.2013.10.020 CrossRefGoogle Scholar
  32. Moss R, Edmonds J, Hibbard K, Manning M, Rose S, Vuuren van D, Carter T, Emori S, Kainuma M, Kram T, Meehl G, Mitchell J, Nakicenovic N, Riahi K, Smith S, Stouffer R, Thomson A, Weyant J, Wilbanks T (2010) The next generation of scenarios for climate change research and assessment. Nature 463(7282):747–756.  https://doi.org/10.1038/nature08823 CrossRefGoogle Scholar
  33. Meier ES, Lischke H, Schmatz DR, Zimmermann NE (2012) Climate, competition and connectivity affect future migration and ranges of European trees. Glob Ecol Biogeogr 21(2):164–178.  https://doi.org/10.1111/j.1466-8238.2011.00669.x CrossRefGoogle Scholar
  34. Mladenoff DJ, He HS (1999) Design, behavior and applications of LANDIS, an object-oriented model of forest landscape disturbance and succession. In: Mladenoff DJ, Baker WL (eds) Spatial modeling of forest landscape change: approaches and applications. Cambridge University Press, Cambridge, pp 125–162Google Scholar
  35. Mráz P, Ronikier M (2016) Biogeography of the Carpathians: evolutionary and spatial facets of biodiversity. Biol J Linn Soc 119(3):528–559.  https://doi.org/10.1111/bij.12918 CrossRefGoogle Scholar
  36. Naudts K, Chen Y, MJ MG, Ryder J, Valade A, Otto J, Luyssaert S (2016) Europe’s forest management did not mitigate climate warming. Science 351(6273):597–600.  https://doi.org/10.1126/science.aad7270 CrossRefGoogle Scholar
  37. Netherer S, Matthews B, Katzensteiner K, Blackwell E, Henschke P, Hietz P, Pennerstorfer J, Rosner S, Kikuta S, Schume H, Schopf A (2015) Do water-limiting conditions predispose Norway spruce to bark beetle attack? New Phytol 205(3):1128–1141.  https://doi.org/10.1111/nph.13166 CrossRefGoogle Scholar
  38. Neumann M, Moreno A, Mues V, Härkönen S, Mura M, Bouriaud O, Lang M, Achten WMJ, Thivolle-Cazat A, Bronisz K, Merganič J, Decuyper M, Alberdi I, Astrup R, Mohren F, Hasenauer H (2016) Comparison of carbon estimation methods for European forests. Forest Ecol Manag 361:397–420.  https://doi.org/10.1016/j.foreco.2015.11.016 CrossRefGoogle Scholar
  39. Niinemets Ü, Valladares F (2006) Tolerance to shade, drought, and waterlogging of temperate Northern Hemisphere trees and shrubs. Ecol Monogr 76(4):521–547. https://doi.org/10.1890/0012-9615(2006)076[0521:TTSDAW]2.0.CO;2Google Scholar
  40. Pasztor F, Matulla C, Rammer W, Lexer MJ (2014) Drivers of the bark beetle disturbance regime in Alpine forests in Austria. Forest Ecol Manag 318:349–358.  https://doi.org/10.1016/j.foreco.2014.01.044 CrossRefGoogle Scholar
  41. Peichl M, Arain MA (2006) Above- and belowground ecosystem biomass and carbon pools in an age-sequence of temperate pine plantation forests. Agric For Meteorol 140(1-4):51–63.  https://doi.org/10.1016/j.agrformet.2006.08.004 CrossRefGoogle Scholar
  42. Petritan AM, Nuske RS, Petritan IC, Tudose NC (2013) Gap disturbance patterns in an old-growth sessile oak (Quercus petraea L.)–European beech (Fagus sylvatica L.) forest remnant in the Carpathian Mountains, Romania. Forest Ecol Manag 308:67–75.  https://doi.org/10.1016/j.foreco.2013.07.045 CrossRefGoogle Scholar
  43. Prots B, Kagalo O (eds) (2012) Kataloh typiv oselyshch Ukrainskykh Karpat ta Zakarpatskoyi nyzovyny. Merkator, Lviv (In Ukrainian)Google Scholar
  44. R Development Core Team (2017) R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna. http://R-project.org. Accessed 7 November 2017
  45. Scheller RM, Mladenoff DJ (2004) A forest growth and biomass module for a landscape simulation model, LANDIS: design, validation, and application. Ecol Model 180(1):211–229.  https://doi.org/10.1016/j.ecolmodel.2004.01.022 CrossRefGoogle Scholar
  46. Scheller RM, Mladenoff DJ (2008) Simulated effects of climate change, fragmentation, and inter-specific competition on tree species migration in northern Wisconsin, USA. Clim Res 36:191–202.  https://doi.org/10.3354/cr00745 CrossRefGoogle Scholar
  47. Scheller RM, Domingo JB, Sturtevant BR, Williams JS, Rudy A, Gustafson EJ, Mladenoff DJ (2007) Design, development, and application of LANDIS-II, a spatial landscape simulation model with flexible temporal and spatial resolution. Ecol Model 201(3-4):409–419.  https://doi.org/10.1016/j.ecolmodel.2006.10.009 CrossRefGoogle Scholar
  48. Schütt P, Schuck J, Stimm B (2007) Lexikon der Baum- und Straucharten. Nikol, HamburgGoogle Scholar
  49. Schwaab J, Bavay M, Davin E, Hagedorn F, Hüsler F, Lehning M, Schneebeli M, Thürig E, Bebi P (2015) Carbon storage versus albedo change: radiative forcing of forest expansion in temperate mountainous regions of Switzerland. Biogeosciences 12(2):467–487.  https://doi.org/10.5194/bg-12-467-2015 CrossRefGoogle Scholar
  50. Shvidenko A, Buksha I, Krakovska S, Lakyda P (2017) Vulnerability of Ukrainian forests to climate change. Sustainability 9(7):1152.  https://doi.org/10.3390/su9071152 CrossRefGoogle Scholar
  51. Seidl R, Schelhaas M-J, Lexer MJ (2011) Unraveling the drivers of intensifying forest disturbance regimes in Europe. Glob Change Biology 17(9):2842–2852.  https://doi.org/10.1111/j.1365-2486.2011.02452.x CrossRefGoogle Scholar
  52. Seidl R, Schelhaas M-J, Rammer W, Verkerk PJ (2014) Increasing forest disturbances in Europe and their impact on carbon storage. Nat Clim Chang 4(9):806–810.  https://doi.org/10.1038/nclimate2318 CrossRefGoogle Scholar
  53. Seidl R, Rammer W (2017) Climate change amplifies the interactions between wind and bark beetle disturbances in forest landscapes. Landsc Ecol 32(7):1485–1498.  https://doi.org/10.1007/s10980-016-0396-4 CrossRefGoogle Scholar
  54. Sturtevant BR, Gustafson EJ, Li W, He HS (2004) Modeling biological disturbances in LANDIS: a module description and demonstration using spruce budworm. Ecol Model 180(1):153–174.  https://doi.org/10.1016/j.ecolmodel.2004.01.021 CrossRefGoogle Scholar
  55. Svoboda M, Janda P, Bače R, Fraver S, Nagel TA, Rejzek J, Mikoláš M, Douda J, Boublík K, Šamonil P, Čada V, Trotsiuk V, Teodosiu M, Bouriaud O, Biriş AI, Sýkora O, Uzel P, Zelenka J, Sedlák V, Lehejček J (2014) Landscape-level variability in historical disturbance in primary Picea abies mountain forests of the Eastern Carpathians, Romania. J Veg Sci 25(2):386–401.  https://doi.org/10.1111/jvs.12109 CrossRefGoogle Scholar
  56. Thom D, Seidl R, Steyrer G, Krehan H, Formayer H (2013) Slow and fast drivers of the natural disturbance regime in Central European forest ecosystems. Forest Ecol Manag 307:293–302.  https://doi.org/10.1016/j.foreco.2013.07.017 CrossRefGoogle Scholar
  57. Thom D, Rammer W, Seidl R (2017a) Disturbances catalyze the adaptation of forest ecosystems to changing climate conditions. Glob Change Biol 23(1):269–282.  https://doi.org/10.1111/gcb.13506 CrossRefGoogle Scholar
  58. Thom D, Rammer W, Seidl R (2017b) The impact of future forest dynamics on climate: interactive effects of changing vegetation and disturbance regimes. Ecol Monogr 87(4):665–684.  https://doi.org/10.1002/ecm.1272 CrossRefGoogle Scholar
  59. Trotsiuk V, Hobi ML, Commarmot B (2012) Age structure and disturbance dynamics of the relic virgin beech forest Uholka (Ukrainian Carpathians). Forest Ecol Manag 265:181–190.  https://doi.org/10.1016/j.foreco.2011.10.042 CrossRefGoogle Scholar
  60. Ukrderzhlisproekt (2014) Ukrainske derzhavne proektne lisovporiadne vyrobnyche obyednannia VO “Ukrderzhlisproekt” (In Ukrainian). http://webulr.lisproekt.gov.ua. Accessed 14 June 2014
  61. Vittoz P, Engler R (2007) Seed dispersal distances: a typology based on dispersal modes and plant traits. Bot Helv 117(2):109–124.  https://doi.org/10.1007/s00035-007-0797-8 CrossRefGoogle Scholar

Copyright information

© Springer-Verlag GmbH Germany, part of Springer Nature 2018

Authors and Affiliations

  1. 1.Department of Physical GeographyFranko National University of LvivLvivUkraine
  2. 2.University of Natural Resources and Life SciencesWienAustria
  3. 3.University of VermontBurlingtonUSA
  4. 4.National Forestry University of UkraineLvivUkraine
  5. 5.North Carolina State UniversityRaleighUSA

Personalised recommendations