Regional Environmental Change

, Volume 17, Issue 1, pp 79–91 | Cite as

Impact of climate change on vulnerability of forests and ecosystem service supply in Western Rhodopes Mountains

  • Tzvetan Zlatanov
  • Che Elkin
  • Florian Irauschek
  • Manfred Josef Lexer
Original Article

Abstract

The vulnerability of forest ecosystem services to climate change is expected to depend on landscape characteristic and management history, but may also be influenced by the proximity to the southern range limit of constituent tree species. In the Western Rhodopes in South Bulgaria, Norway spruce is an important commercial species, but is approaching its current southern limit. Using climate sensitive forest models, we projected the impact of climate change on timber production, carbon storage, biodiversity and soil retention in two representative landscapes in the Western Rhodopes; a lower elevation landscape (1000–1450 m a.s.l) dominated by mixed species forests, and a higher elevation landscape (1550–2100 m a.s.l.) currently dominated by spruce. In both landscapes climate change is projected to induce a shift in forest composition, with drought-sensitive species, such as Norway spruce, being replaced by more drought-tolerant species such as Scots pine and black pine at lower elevations. In the higher elevation landscape a reduction in spruce growth is projected, particularly under the more severe climate change scenarios. Under most climate scenarios a reduction in growing stock is projected to occur, but under some scenarios a moderate increase in higher elevation stands (>1500 m a.s.l.) is expected. Climate change is projected to negatively influence carbon storage potential across landscapes with the magnitude depending on the severity of the climate change scenario. The impact of climate change on forest diversity and habitat availability is projected to differ considerably between the two landscapes, with diversity and habitat quality generally increasing at higher elevations, and being reduced at lower elevations. Our results suggest that if currently management practices are maintained the sensitivity of forests and forest ecosystem services in the Western Rhodopes to climate change will differ between low and higher elevation sites and will depend strongly on current forest composition.

Keywords

Climate change Forest ecosystem services Ecosystem modelling Sustainable forest management 

Introduction

In mountain regions, the impact of climate change on forests and forest ecosystem services (ES) is often expected to be elevation and site dependent (Gonzalez et al. 2010; MEA 2005). Changes in forest species composition along elevation gradients, reflecting elevation dependent temperature, precipitation and edaphic conditions, may promote differences in the sensitivity of certain stands, or local areas, to climate change (EEA 2010). Heterogeneous responses of forest stands to climate change may be particularly prevalent in regions such as the Western Rhodopes in South Bulgaria where important timber species such as Norway spruce (Piceaabies) are approaching the current southern edge of their distribution and may be sensitive to small climatic shifts that exceed the species tolerances (Hanewinkel et al. 2012; Temperli et al. 2012). In regions such as this, where key species are expected to be close to environmental thresholds, relatively small elevation differences and differences in historic forest management have the potential to significantly alter the vulnerability of ES to climate and environmental shifts. These factors highlight the need to improve our understanding of how climate change will potentially impact forest ecosystem services at both the stand and landscape scale, but also illustrate the complexity of managing forests in heterogeneous regions where a uniform response across the landscape to climate change is not projected to occur.

The vulnerability of mountain forest ES is not only expected to vary between and within regions (EEA 2010), but also to differ between services (Elkin et al. 2013). For example, ES that depend on the maintenance of specific species for timber or other provisioning services may be particularly vulnerable. Similarly, rare and endangered animal species that are affiliated with particular forest types and structures, such as birds that have specific nesting and foraging requirements, may also be more vulnerable (Nikolov et al. 2011; Popov 2011). In contrast, ES that mainly depend on the continued presence of forest cover, such as carbon storage or soil protection, will potentially be less vulnerable to climate change. The provisioning of individual ES may also vary across different spatial and temporal scales, especially when environmental changes and management impact species composition and forest structure differently across the landscape. Assessing potential climate change and management impacts on a range of ecosystem services thus also requires a multiscale perspective (Seidl et al. 2013).

In the Western Rhodopes, and mountainous parts of South-Eastern Europe in general, an increase in extreme weather events and shifts in climate have been observed over the last 20 years. These shifts include a decrease in cold temperature extremes, an increase in warm temperature extremes, an increase in the duration of summer dry periods, and an increase in the number of heavy precipitation events (Aleksandrov et al. 2009; IPCC 2014). In conjunction with these environmental changes, there has been a growing societal demand for forest services and functions which extend beyond timber production, with some ES perceived as potentially being incompatible with others. Stakeholder panels established in West Rhodopes Mountain, conducted within the EU FP7 ARANGE project framework, have identified that regional planning authorities face problems finding a consensus for multiple societal needs and demands. These perceived conflicting demands promote tension over using the same piece of land for different purposes and services. For example, the total annual volume of roundwood harvested has been steadily increasing for the last two decades (GFSAF 2000–2014), and there is a concern that the forest sector may not take the necessary steps to comprehensively assess the vulnerability, risks and uncertainties related to increases in harvest intensity. These concerns are increased by an insufficient understanding of potential trade-offs and co-benefits between wood production and biodiversity conservation. An increased focus on multiuse forests, in conjunction the potential impacts of climate change, contributes to the complexity and challenges that must be addressed by regional forest managers (Lexer and Brooks 2005; Rauscher et al. 2005).

The objective of this study is to evaluate the vulnerability of forests and forest ecosystem services to climate change using a multiple-ecosystem service approach. We focus on two spatially proximate but ecologically distinct landscapes in the Western Rhodopes Mountains in Bulgaria; a low elevation landscape that is dominated by mixed forests, and a higher elevation landscape that has a high Norway spruce component. To account for spatial heterogeneity, and scale-dependent processes, two complimentary simulation models are used that evaluate forest dynamics and ecosystem services at separate spatial and temporal scales. This approach allows our findings to be better evaluated at the operational scale at which management decisions are taken (within or beyond the stand). Within this framework, we evaluate the projected impacts of climate change on mountain forests in the Western Rhodopes provided currently implemented management practices are maintained. We focus on four regionally relevant ES: timber production, carbon storage, soil protection and biodiversity conservation. In particular we address four key questions: (1) How is the species composition and timber growing stock in these two forest landscapes projected to change under future climate scenarios? (2) Which ES are projected to be most vulnerable under future conditions? (3) Does the vulnerability of the various ES differ between the two representative landscapes? and (4) Do differences in ES vulnerability reflect the climate sensitivity of the current dominant tree species that comprise each landscape?

Materials and methods

Case study area

The case study region is located on the northern slopes of Mount Perelik—the highest mountain of Western Rhodopes (South Bulgaria). Due to the altitude (2191 m a.s.l.) and orientation of the main ridge of the mountain, the climate on its northern slopes is less influenced by the Mediterranean Sea compared to surrounding territories, especially those located in the Greek part of the Rhodopes. Average annual precipitation (Shiroka Laka station, 1 km from the study area at 1050 m a.s.l.) for the years 1960–2010 is 850 mm, with precipitation being evenly distributed through most of the year. The driest period is August–October (165 mm precipitation). The mean annual temperature at 1050 m a.s.l. for the same reference period is 6 °C, with a July mean temperature of 15.5 °C and a January mean temperature of −3 °C. The altitudinal lapse rate for precipitation on Mount Perelik averages 20 mm per 100 m. The decrease in mean annual temperature is 0.4 °C per 100 m (Galabov 1973). Representative landscape 1 (coordinates 41°40′N and 24°32′E) includes six representative stand types (RSTs 1–6). The 736 ha landscape has an elevation range from 1000 to 1450 m a.s.l., and is characterized by a variety of site conditions and mixed forests (Table 1). Representative landscape 2 (RSTs 7–10; coordinates 41°37′N and 24°35′E) is larger (1001 ha), is dominated by mountainous and subalpine spruce forests and spans an elevation range from 1550 to 2100 m a.s.l. (Table 1).
Table 1

Characteristic of the representative stand types (RSTs)

RST ID

RST name

Species composition

Stand development stage

Altitude [m a.s.l.]

Soil typea

Soil nutrient supplya

Soil water regimea

1

Beech forests on mesotrophic mesic sites

European Beech (Fagussilvatica L.)

Pole and mature

1000–1150

Cambisol

Intermediate

Moderately moist

2

Black pine forests on oligotrophic xeric sites

Black pine (Pinusnigra Arn.)

Pole and mature

1200–1450

Rendzina

Poor

Dry

3

Black pine dominated forests on submesotrophic subxeric sites

Black pine, Norway spruce and European beech

Thicket, pole and mature

1200–1450

Rendzina

Moderately poor

Moderately dry

4

Mixed forests on Mesotrophic mesic sites

Black pine, Norway spruce and European beech

Pole and mature

1200–1400

Cambisol

Intermediate

Moderately moist

5

Scots pine dominated forests on submesotrophic subxeric sites

Scots pine (Pinussylvestris L.), black pine, Norway spruce and European beech

Pole and mature

1100–1300

Cambisol

Moderately poor

Moderately dry

6

Spruce-fir forests on permesotrophic mesic sites

Norway Spruce and Silver fir (Abiesalba Mill.)

Thicket, pole and mature

1200–1350

Cambisol

Rich

Moist

7

Mountainous spruce forests on permesotrophic mesic sites

Norway Spruce

Thicket, pole and mature

1550–1850

Cambisol

Rich

Moist

8

Mountainous spruce forests on mesotrophic mesic sites

Norway Spruce

Thicket, pole and mature

1550–1850

Cambisol

Intermediate

Moist

9

Subalpine spruce forests on mesotrophic mesic sites

Norway Spruce

Mature

1900–2050

Cambisol

Intermediate

Moist

10

Subalpine spruce forests on former pastures

Norway Spruce

Thicket and pole

1900–2100

Cambisol

Rich

Moist

a Data—SLFEMP 2007

Climate change scenarios

A baseline climate (C0) and five transient climate change scenarios (C1–C5), each consisting of a 100-year time series of daily temperature, precipitation, radiation and vapour pressure deficit, were evaluated. The baseline climate was generated from gridcell data of the E-OBS dataset (Haylock et al. 2008) and bias corrected for temperature and precipitation with empirical data from the weather station Shiroka Laka. The base climate data set was adjusted for representative site types within the case study area regarding altitude, slope and aspect using the algorithms in Thornton and Running (1999). The five climate change scenarios were based on regional climate simulations from the ENSEMBLES project (Hewitt and Griggs 2004; http://www.ensembleseu.org). For details on the definition and description of the climate change scenarios as well as on the downscaling approach, see Lexer et al. (this volume). Projected temperature and precipitation anomalies under the five climate scenarios are shown as the deviation from the baseline climate in Fig. 1.
Fig. 1

Projected temperature and precipitation anomalies under the five climate scenarios DMI-HIRHAM5_BCM (C1), DMI-HIRHAM5_ARPEGE (C2), ETHZ-CLM_HadCM3Q0 (C3), MPI-REMO_ECHAM5-r3 (C4) and HC-HadRM3_HadCM3Q16 (C5). For presentation purposes, monthly temperature and precipitation delta values are shown as a smoothed trace

Models and simulation experiments

Stand-scale simulations

Stand-scale simulations were done using the forest ecosystem model PICUS v1.5 (Lexer and Hönninger 2001; Seidl et al. 2005). It is a hybrid of classical gap model (Lexer and Hönninger 2001) and process-based stand-level NPP model 3PG (Landsberg and Waring 1997), designed to provide spatially explicit site-specific decision support to forest resource managers in heterogeneous mountainous landscapes. PICUS allows the simulation of individual tree within a 3D structure of 10 × 10 m patches extended by crown cells of 5 m height. Forest dynamics are driven by the key processes growth, mortality and regeneration, as described in Lexer and Hönninger (2001) and Seidl et al. (2005). The model follows a modular structure, which includes among others a regeneration module (Woltjer et al. 2008), and a management module (Seidl et al. 2005). The management module enables the user to conduct almost any silvicultural activity in the model environment and encompasses a flexible approach to implement thinning and harvesting regimes. The PICUS version used for this study runs on monthly climate data of temperature, precipitation, radiation and vapour pressure deficit. The model has been used in several application studies focusing on the assessment of climate change impacts. Previous studies found realistic responses to climatic gradients (Lexer and Hönninger 2001; Seidl et al. 2005, 2011). Simulation of BAUM (business-as-usual management) was based on stands of 1 ha size. Initial states for the RSTs were based on diameter at breast height (DBH) distributions and height–diameter relationships for individual tree species. Management operations where implemented using sorting algorithms for the list of trees present in a stand by species and diameter (tending, thinning and harvesting operations) or by choosing trees according to their position at a resolution of the simulated patches (10 m × 10 m raster).

Landscape simulations

Landscape scale simulations were done using the forest simulation model LandClim (Schumacher et al. 2004, 2006), a spatially explicit, process-based model that incorporates competition-driven stand dynamics and landscape-level disturbances and management to simulate forest dynamics on a landscape scale. LandClim was designed to examine the impact of climate change and forest management on forest development and structure (Schumacher and Bugmann 2006; Schumacher et al. 2006). The model has been tested in the Central Alps, North American Rocky Mountains, and Mediterranean forests, and has been used to simulate current as well as paleo-ecological (Colombaroli et al. 2010; Henne et al. 2011) and future forest dynamics (Schumacher and Bugmann 2006). Here we provide a brief overview of the structure of the model; for further details see Schumacher et al. (2004). LandClim simulates forest growth in 25 m by 25 m cells using simplified versions of tree recruitment, growth and competition processes derived from those that are commonly used in gap models (Bugmann 2001). Forest growth is determined by climatic variables, soil properties and topography, land use and large-scale disturbances. Individual cells are linked by the spatially explicit processes of seed dispersal and landscape disturbances. Succession processes, including tree growth and intrinsic mortality within each cell, are simulated on a yearly time step, while landscape-level processes, tree establishment and forest management are simulated on a decadal time step. Forest dynamics within each cell are simulated by following tree size cohorts, where cohorts are characterized by the mean biomass of an individual tree and the number of trees in the cohort.

Forest management

Due to many societal changes in the study area during the last century, it is difficult do provide a unified definition of the traditional forest management. Prior to the beginning of the nineteenth century, the area was subjected to heavy grazing. During the 1950s the forests were nationalized and the area partially depopulated. Stock breeding gradually lost its significance in the local economy. Silvicultural systems used were clearcutting and classical shelterwood, with annual allowable harvest usually being less than or close to the annual increment of the forests (SLFEMP 2007). During the post socialistic period (since 1990s), the forests in the study area were again privatized. Annual allowable harvest has increased since privatization, and now it often exceeds the annual increment of the forests. The current business-as-usual management (BAUM) from the period 1990–2010 is used in the simulations. In RSTs 1, 3, 4, 5, 6, 7 and, 8 BAUM is used in the simulations, while in RSTs 2, 9 and 10 no management approach is applied (as it is in reality). The regeneration system used in managed RSTs is a shelterwood system. Rejuvenation of stands is performed in three successive regeneration fellings. The first felling is done at an age of 90 years, while two regeneration fellings follow approximately every 15 years (at 105 and 120 years of stand age). The intensity of the first regeneration felling is approximately 30 % of the total stand volume. Removal percentage of the second felling varies between 45 and 55 % of total stand volume, while in the third and final felling the entire remaining stand is harvested. No weeding operation or other early release treatments are performed in the RSTs. One tending operation is implemented approximately 20 years after the final harvest. During this operation 20 % of the trees that have reached the early pole stage are removed to promote growth and improve crown quality. Thinning is implemented as a combination of thinning from above (e.g. crown thinning) and thinning from below. Usually, three thinning operations are performed. The first thinning operation is done approximately 30 years after the final harvest. Thinning intensity varies between 20 and 25 % of total stand volume. The next two thinning operations follow at approximately 20-year periods. The thinning intensities vary between 20 and 30 % of total stand volume. At the time of the last thinning, the age of trees varies between 70 and 100 years. If trees in the smallest relative DBH class are still left in a stand, they are all removed. As described, BAUM is maintained throughout the simulation period (years 2010–2110).

Ecosystem service indicators

Stand-scale ecosystem indicators

The forest ecosystem services (ES) indicators used in this study represent regional service demands and were evaluated at the scale where management decisions are taken. Wood production was represented by two main indicators: growing stock of live trees (TGS) and annual wood harvest (AWH). Both indicators were derived at stand scale and reported on a volumetric basis, TGS being separated by species and AWH given as a total value. In situ carbon stock (CS) aggregated both the above ground (CSAbove) and below ground (CSBelow) carbon. CSAbove was calculated as a function of TGS, BEF (biomass expansion factor for conversion of growing stock to above ground tree biomass; BEFBroadleaf = 0.48, BEFConifer = 0.51), WD (wood density in t dry matter/m3 fresh volume; WDAbies = 0.4, WDFagus = 0.58, WDPicea = 0.4, WDPinus = 0.42), and CF (carbon fraction of dry matter in t C/t d.m.; CFBroadleaf = 1.4, CFConifer = 1.3) (IPCC 2006)
$${\text{CS}}_{\text{Above}} = {\text{TGS}} * {\text{BEF}} * {\text{WD}} * {\text{CF}}$$
(1)
CSBelow was estimated based on the above ground dry carbon stock using IPCC root-to-shoot ratio (R, Table 2)
Table 2

Root-to-shoot ratios (R) for estimating below ground carbon mass (IPCC 2006)

Forest type

Root-to-shoot

Temperate conifer (above ground biomass <50 t/ha)

0.40

Temperate conifer (above ground biomass 50–150 t/ha)

0.29

Temperate conifer (above ground biomass >150 t/ha)

0.20

Temperate broadleaf (above ground biomass <75 t/ha)

0.46

Temperate broadleaf (above ground biomass 75–150 t/ha)

0.23

Temperate broadleaf (above ground biomass >150 t/ha)

0.24

$${\text{CS}}_{\text{Below}} = {\text{CS}}_{\text{Above}} * R$$
(2)

Landscape scale ecosystem indicators

Four landscape-level ecosystem service indicators were used: forest species composition, Shannon’s diversity index, bird (woodpecker and owl species) habitat quality, and soil stabilization. Each metric was evaluated at a 25 m by 25 m grain (cell) size across the full extent of the simulated landscape.

Projected shifts in tree species composition were analysed by examining how the dominant tree species, based on cumulative basal area, within each cell changed through time.

Impacts on forest diversity were evaluated by calculating Shannon’s tree diversity index based on the biomass of each species (pi):
$$H_{\text{species}} = - \sum\limits_{i = 1}^{S} {p_{i} (\ln (p_{i} ))} .$$
(3)

We also estimated a stand structural diversity index (Staudhammer and LeMay 2001) by calculating Shannon’s diversity index using the height and DBH of each species-specific cohort simulated.

The suitability of forests for bird habitat was evaluated by calculating a bird habitat quality index that was then used to define whether a cell represented good, medium or poor habitat. Within each cell habitat quality was estimated based on a combination of the volume of dead wood, the number of years the forest had been unmanaged, the number of large (>50 cm DBH) trees present, and per cent canopy cover. Increases in the amount of dead wood, years since forest management occurred, and volume of large trees all linearly increased habitat quality. In contrast, a nonlinear relationship between canopy cover and habitat quality was assumed: cells with 60–80 % canopy cover represented good sites, cells with canopy cover between 80–90 % and 40–60 % were classified as medium, and cells with canopy cover >90 or <40 % were classified as poor sites. The selection of these predictors and their interpretation was based on previous research (e.g. Angelstam et al. 2003; Guénette and Villard 2005; Ranius and Fahrig 2006). The four contributing factors to bird habitat quality were normalized and combined to produce a single bird habitat quality estimate. Cells with quality estimates in the highest and lowest quartile were set as good and poor habitat, respectively, while cells that fell between these bounds were defined as medium quality habitat.

The ability of forest to increase soil stabilization was estimated by calculating a landslide protection index. Forests reduce the likelihood and extent of landslides and erosion by mechanically reinforcing the soil, and by positively influence the soil water balance through interception, transpiration and enhanced soil permeability (Frehner et al. 2005). Well-developed forests that are multilayered are therefore expected to provide the greatest protection from both landslides and erosion. We calculated a landslide reduction index (LRI) that incorporates leaf area index (θLAI) as a surrogate for high soil stability in multilayered forests with high canopy coverage, and a metric that evaluates the structural diversity of the forest \((\theta_{\text{structure}} )\). Details on ecosystem service indicators can be found in Bugmann et al. (this volume).
$${\text{LRI}} = \theta_{\text{LAI}} + \theta_{\text{structure}}$$
(4)

Results

Stand-scale simulations

Most scenarios show substantial impacts of climate change on simulated forest ecosystem services at a stand scale. Norway spruce is projected to be the most vulnerable species to climate change, especially at elevations below 1400–1500 m a.s.l in landscape one. The share of spruce in mixed forests on mesotrophic mesic sites and in spruce-fir forests on permesotrophic mesic sites is projected to substantially diminish under climate change (Fig. 2, RST4 and RST6). This will result in a significant decrease in total stand growing stock, especially in spruce-fir forests. The productivity of stands is projected to be substantially reduced by climate change in the mountainous spruce forests of landscape two, as well. Exceptions are under C1 (which projects the smallest alteration in temperature and precipitation amount) and partially C4 (Fig. 2, RST7 an RST8).
Fig. 2

Growing stock dynamics in the representative stand types (RSTs) under baseline climate (C0) and the five climate change scenarios DMI-HIRHAM5_BCM (C1), DMI-HIRHAM5_ARPEGE (C2), ETHZ-CLM_HadCM3Q0 (C3), MPI-REMO_ECHAM5-r3 (C4) and HC-HadRM3_HadCM3Q16 (C5). Projected growing stock dynamics under the five climate scenarios is shown as the deviation from the baseline climate

European beech is also projected to be vulnerable to climate change. This is most pronounced in the lowest elevated beech forests on mesotrophic mesic sites, where the species is projected to undergo abrupt decline during the second half of the simulation period (Fig. 2, RST1) under C2, C3 and C5 climate change scenarios. At higher elevations and more fertile sites beech is less vulnerable to climate change. For example, towards the end of the simulated period, the share of beech in mixed forests on mesotrophic mesic sites is projected increase from 3.8 % (13 m3 ha−1), under baseline climate, to 8.5 % (24 m3 ha−1) and 39.6 % (56 m3 ha−1) under C1 and C2 climate change scenarios, respectively (Fig. 2, RST4). Drought-tolerant species such as Scots pine and black pine are projected to persist in all representative stands where they currently occur, under all climate change scenario.

Black pine is projected to be more robust to climate change and is generally projected to increase relative to other species. For example, towards the end of the simulated period, the share of black pine in Scots pine-dominated forests on submesotrophic subxeric sites is projected to increase from 32 % (78 m3 ha−1), under baseline climate, to 40 % (86 m3 ha−1) and 56 % (78 m3 ha−1) under C4 and C5 climate change scenarios, respectively (Fig. 2, RST5). These per cent changes are mainly due to a projected decrease in total growing stock of 15–44 %. Climate scenario C2, characterized by increased summer temperatures and decreased summer precipitation, is projected to be most unfavourable for black pine.

Silver fir only grows in Spruce-fir forests on permesotrophic mesic sites. Here, Silver fir is projected to gradually increase its relative share in the total stand growing stock under all climate change scenarios (e.g. up to 88 % under C3 towards the end of the simulation period). In absolute terms, however, the fir is projected to increase its growing stock only under the C2 and C5 climate change scenarios (Fig. 2, RST4).

As expected, the mean in situ carbon stock is projected to be most constant through the projected period in the unmanaged stands (RSTs 2, 9 and 10) due to the continuing growth and the accumulation of deadwood pools on site. Intensive management (predominantly regeneration cuts) leads to substantial decrease in carbon stock during the first decades of the simulated period followed by moderate to strong increases under C1 and C4 climate change scenarios (Fig. 3, RSTs 1, 2, 4, 5, 6, 7 and 8). Carbon stock remains relatively constant or starts decreasing again during the middle of the simulated period under C2, C3 and C5 climates, the relative decrease in carbon being greater on mesotrophic mesic (RST 1, 4 and 8) and permesotrophic mesic sites (RST 6 and 7) at elevations below 1800 m a.s.l in both studied landscapes. At higher elevations (above 1800 m a.s.l.) in landscape 2, increased temperature under all climate change scenarios initially leads to increased forest biomass in the unmanaged spruce forests; however, after 2070 drought is projected to decrease forest biomass under climate change scenarios C2, C3 and C5 (Fig. 3).
Fig. 3

Carbon sequestration dynamics in the representative stand types (RSTs) under baseline climate (C0) and the five climate change scenarios DMI-HIRHAM5_BCM (C1), DMI-HIRHAM5_ARPEGE (C2), ETHZ-CLM_HadCM3Q0 (C3), MPI-REMO_ECHAM5-r3 (C4) and HC-HadRM3_HadCM3Q16 (C5)

Under all climate scenarios business-as-usual management results in large fluctuations in the volume of wood harvested annually throughout the simulated period (Fig. 4), with harvested volumes decreasing substantially around the middle of the century. This variation partially masks the impact of climate change harvest volumes. However, climate impacts on harvest are evident in spruce-fir forests on permesotrophic mesic sites, especially in mountainous spruce forests on mesotrophic mesic sites (Fig. 4, RST6 an RST8). In both cases, pronounced decrease in annual wood harvest is observed under climate change scenarios C2, C3 and C5 during the last decades of simulation.
Fig. 4

Annual harvest in the representative stand types (RSTs) under baseline climate (C0) and the five climate change scenarios DMI-HIRHAM5_BCM (C1), DMI-HIRHAM5_ARPEGE (C2), ETHZ-CLM_HadCM3Q0 (C3), MPI-REMO_ECHAM5-r3 (C4) and HC-HadRM3_HadCM3Q16 (C5)

Landscape scale simulations

At a landscape level, divergent responses in forest composition are projected between the lower elevation regions of landscape 1 and higher elevation forests in landscape 2 (Fig. S1). Dieback of drought-intolerant species at lower elevations is projected to reduce species richness, while improved growth conditions at higher elevations are projected to result in a landscape-level increase in species richness. These changes are projected to be the largest under the more severe climate change scenarios, but still observed even under moderate climate change (Fig. S1d, e). Climate change is projected to slow a landscape-level reduction in forest diversity, as estimated by Shannon’s index, at lower elevations (Fig. S2), while increasing diversity at higher elevations.

Projected shifts in forest cover and structure are expected to decrease the ability of the forest to provide soil stabilization across the landscape (Fig. 5). Some of the largest expected decreases in soil stabilization are projected to occur at higher elevations, in regions where forest currently provides higher levels of stabilization and landslip protection.
Fig. 5

Landscape projections of forest soil stabilization (LRI) under baseline climate (C0: ac), a moderate climate change scenario DMI-HIRHAM5_BCM (C1: d, e), and a strong climate change scenario DMI-HIRHAM5_ARPEGE (C2: f, g), in the years 2010, 2050 and 2100. Figure bg illustrate the projected change in soil stabilization compared to forest conditions in 2010 (a)

Shifts in forest species composition and structure are projected to have only a moderately negative impact on bird habitat quality (Fig. S3) under all scenarios. However, the regions that are projected to experience the largest reduction in habitat quality correspond to areas that are currently assessed as being good-quality habitat. Therefore, while climate change is projected to not substantially alter the proportion of good-quality habitat on the landscape, the location of high quality sites may shift.

Discussion

Species-specific reductions in growth change forest community dynamics by altering species competitive abilities and their relative abundance (Huntley 1991; Peñuelas and Boada 2003; Booth and Grime 2003). In mountain regions the impact of climate change on tree species is often elevation and site dependent. Shifts in species composition along elevation gradients, reflecting elevation dependent temperature and precipitation, may influence stand and landscape resilience to climate change. Heterogeneous responses of forest stands to climate change are projected to be particularly prevalent in Western Rhodopes in South Bulgaria where important timber species such as spruce, beech, Scots pine and fir are near the current southern/south-eastern edge of their distribution. Results from our study support observations from other Bulgarian mountain ranges (e.g. Osogovo mountain and Balkan Range, Tzvetan Zlatanov unpublished data) that the European spruce is the most susceptible native species to climate change, specifically at lower altitudes and less fertile sites (Zang et al. 2011). Similar results are reported for the lower altitudes in the Alps, where the proportion of Norway spruce is projected to decrease, and the importance of Spruce as a crop tree is expected to diminish (e.g. Jolly et al. 2005; Seidl et al. 2011; Elkin et al. 2013). Savva et al. (2006) also projected a decrease in Spruce growth at lower elevation sites in the Tatra Mountain. These past findings support our projections of extinction of spruce on oligotrophic xeric sites and mesotrophic submesic sites (Fig. 2, RST 2 and RST 3) over the next 20–50 years. On mesotrophic submesic sites (Fig. 2, RST 6), silver fir is projected to gradually replace spruce as a dominant species. The loss of spruce is considered undesirable by local stakeholder due to higher economic value of spruce timber.

Our results suggest that the decrease in spruce might be influenced not only by climate change but also by the management (shelterwood system) of mixed coniferous forests in the case study area. Current management has led to a loss of uneven-aged forest and decreased the amount of available light for the less shade tolerant spruce seedlings and saplings, compared to the fir. Comparable shifts in species relative competitive ability have also been observed in the Alps (Diaci and Firm 2011), where group selection practices have increased light and enhanced spruce regeneration in mixed spruce-fir stands.

Silver fir has also been reported as being susceptible to climate change at lower elevated sites (Carrer et al. 2010; Peguero-Pina et al. 2007; Tzvetan Zlatanov, unpublished data regarding Eastern Balkan Range and West Rhodopes). However, in our case study region the species is near its ecological optimum (Delkov 1988) and appears to be more robust to climate change than spruce.

Many studies have predicted increased growth of Norway spruce at higher elevations in the mountains of South and Central Europe. For example, Hartl-Meier et al. (2014), using data from a network of spruce forests of the European Alps, found that trees at lower elevations are negatively impacted by drought and high temperatures, while at higher altitudes spruce benefits from warmer climatic conditions. Correspondingly, studies investigating tree growth during the European 2003 heat-wave have reported increased Spruce growth at high elevation sites in the Swiss Alps (Jolly et al. 2005). Similar findings have also been reported from treeline environments in other regions of the European Alps (Rolland et al. 1998; Paulsen et al. 2000). Savva et al. (2006) concluded that rising temperatures will likely cause increased growth of Norway spruce at high elevation sites in the Tatra Mountains. Results from our study do not fully support the findings that spruce will likely increase its growth potential in higher parts of the mountains in South Europe, as result of higher temperatures. In contrast, our results suggest a reduction in spruce productivity, even on fertile sites at higher elevations (above 1500 m). In only few cases (e.g. above 1900 m and under climate change scenarios C1 and C4, Fig. 2, RST 9 and RST 10) spruce is projected to retain its growth potential.

Although European beech is currently a dominant species in less than 5 % of the stands in our study, beech forests are important for the local economy as they are a source of firewood which is critical for most households in the study area. As a result, our findings that beech might be forced outside its present niche at lower elevations (up to 1100 m. a.s.l., Fig. 2, RST 1) towards the end of the simulation period may contribute to regional forest management problems.

Changing the contemporary even-aged forest management into continuous cover forestry, as suggested by other authors (Von Lüpke and Spellmann 1999; Schütz 1999), may have limited immediate positive benefits in beech stands as alternatives to beech are currently not available. Thus local stakeholders should prepare for a future where the sustainability of beech forests at low elevations in the Western Rhodopes is reduced. In contrast, the projected increase in beech share (under all climate change scenarios) in pine-spruce-dominated forests on Mesotrophic mesic sites above 1200 m a.s.l. may favour alternative silvicultural systems, such as irregular shelterwood or group selection, which better utilize the long-term growth and ES provisioning capacity of new mixtures (Schütz 1999; Pretzsch 2009; Sayer and Maginnis 2013). Black pine is considered the most drought-tolerant native coniferous species in Bulgaria having few alternatives on xeric calcareous soils up to 1300–1400 m a.s.l. This species has long been successfully used in afforestation of degraded land in the lower vegetation belt (up to 800–1000 m a.s.l.) of south Bulgaria which was formerly covered by oak-dominated communities (Zlatanov et al. 2010). The projected decrease in black pine volume = in the case study area is relatively minor and only observed during last four decades of simulations (Fig. 2, RST 2, RST 3, RST 4 and RST 5). Collaborating our results, recent data from a dendrochronology study (Wolfgang Beck and Tzvetan Zlatanov, unpublished data) ascertain that black pine in the study area has exhibited relatively uniform growth during last 300 years, with the species being most influenced by increased late summer temperatures and decreased spring and summer precipitations.

Varying TGS and CS dynamics under the five climate scenarios that we examined indicate that uncertainty remains regarding the extend timber production and carbon sequestration in the case study area will be negatively influenced by climate change. Even under moderate climate scenarios, our results support the concern that currently used even-aged silvicultural systems and increasing harvest intensities in Rhodopes Mountain will likely promote inconsistencies in the future roundwood supply (Velichkov et al. 2008). Increasing demand for timber, in combination with the local economic benefits associated with forest management, means that long-term planning and alternative management options are needed to maintain wood supply in this region. While uneven-aged management is expected to result in a decrease in the absolute values of total growing stock and carbon sequestration in situ, it will likely provide higher annual increment (Nyland 1996) and contribute to a more constant supply of harvestable timber, especially large log wood, into the future. According to Velichkov et al. (2008, 2009), a reduction in large log wood provisioning will substantially decrease the profitability of forestry and forest-related industries in Rhodopes and reduce the potential for long-term carbon storage in products pool (e.g. quality furniture). Although not accounted for in the Kyoto Protocol (UNFCCC 1997), we consider C storage in wood products as a viable option to mitigate the superiority of the unmanaged compared to managed stands regarding C sequestration potential in situ.

Forests ability to provide soil stabilization is a key function of the studied landscapes as a result of the steep slopes and extreme weather events observed during recent years. Loss of productive forest land due to soil erosion will likely negatively impact the provisioning of all forest ES across the representative landscapes. Soil erosion can lead to increased pollution and sedimentation in streams and rivers, clogging these waterways and causing declines in fish and other species. Lands with reduced and degraded soils are also often less able to hold onto water, which can worsen flooding (Marinov et al. 2012). Historically, the loss of forest cover caused by unregulated cutting and overgrazing has triggered mass erosion in many regions of South Bulgaria. Erosion risk has been partially mitigated by establishing extensive forest plantations on the eroded slopes during the beginning of nineteenth century (Zahariev et al. 1977). However, recently parts of these artificial forests have been destroyed by fires and pathogen outbreaks (Aleksandrov et al. 2009), which brought back most of the issues observed before the rehabilitation of these lands. Projected decreases in the soil stabilization potential of forests in the studied representative landscapes, especially in regions where forest currently provide higher levels of stabilization and landslip protection, may combine with these other factors to expedite the risks of soil erosion. A key challenge is therefore assessing how forestry/land use can most effectively help in preventing soil degradation and erosion in the future.

The Western Rhodopes Mountains are important for the protection of forest inhabiting owl and woodpecker species at a European and a national scale. Large number of owl and woodpecker species have been reported (Iankov 2007) in the region. Among them, 5 owl species and 10 woodpecker species have been included in the European Natura 2000 species of concern list (under directive 2009/147/EC). The moderate-to-low bird habitat quality appraised in the studied landscapes (with the exception of part of landscape 2) would likely require currently (and traditionally) management practices to be reconsidered in light of their potential negative impact on bird habitat. In particular, the lack of policies directly addressing the need to increase the number of snags and large trees, as well as management practices that can promote the diversification of the stand spatial structure, may need to be considered. The latter is especially relevant in landscape one where our simulations indicate that climate change, in conjunction with BAU management, is projected to further reduce the forest structural diversity as estimated by Shannon’s index (Fig. S2).

The divergent responses of projected forest ES under the various climate scenarios indicate that a high degree of uncertainty remains with regard to the extent that climate change will negatively impact forest ecosystems in the studied representative landscapes. This uncertainty is even more pronounced since our results do not accounting for climate sensitive biotic disturbances (Bentz et al., 2010) and should therefore be viewed as a conservative estimate. Despite the variability of the climate projections, and their impact on the vulnerability of forests and ES supply in Western Rhodopes, even the most moderate climate change scenarios investigated project a negative effect on drought-sensitive forests at the lower elevations of landscape 1. Some forest ES (e.g. forest carbon storage and wood production) are projected to be negatively impacted, especially during the last 40 years of the 100-year simulated period, across both landscapes, with the magnitude of the reduction depending on the severity of the climate change scenario. Alternative management practices may be able to mitigate some of the unfavourable impacts of climate change on regional forest ecosystem services. However, the suitability of alternative management practices, such as longer regeneration periods, continuous cover forestry systems and the retention of small patches of old growth, will likely depend on the stands location on the landscape and its management history.

Notes

Acknowledgments

Support for this study was provided by the project “Advanced Multifunctional Forest Management in European Mountain Ranges (ARANGE)” within the European commission’s 7th framework program, grant agreement number 289437.

Supplementary material

10113_2015_869_MOESM1_ESM.docx (2.6 mb)
Supplementary material 1 (DOCX 2643 kb)

References

  1. Advanced Multifunctional Forest Management in European Mountain Ranges (ARANGE). Project within the European commission’s 7th framework program, grant agreement number 289437. http://www.arange-project.eu/. Accessed 28 June 2014
  2. Aleksandrov A, Kostov G, Zlatanov T (2009) COST action FP0703—ECHOES. Expected climate change and options for European silviculture, country report Bulgaria, 14pp. http://www.gip-ecofor.org/echoes. Accessed 28 June 2014
  3. Angelstam P, Bütler R, Lazdinis M, Mikusinski G, Roberge J (2003) Habitat thresholds for focal species at multiple scales and forest biodiversity conservation—dead wood as an example. Ann Zool Fennici 40:473–482Google Scholar
  4. Bentz B, Régnière J, Fettig C, Hansen E, Hicke J, Hayes J, Kelsey Y, Negrón J, Seybold S (2010) Climate change and bark beetles of the Western United States and Canada: direct and indirect effects. Bioscience 60:602–613CrossRefGoogle Scholar
  5. Booth R, Grime J (2003) Effects of genetic impoverishment on plant community diversity. J Ecol 91:721–730CrossRefGoogle Scholar
  6. Bugmann H (2001) A review of forest gap models. Clim Change 51:259–305CrossRefGoogle Scholar
  7. Carrer M, Nola P, Motta R, Urbinati C (2010) Contrasting tree-ring growth to climate responses of Abies alba toward the southern limit of its distribution area. Oikos 119:1515–1525. doi:10.1111/j.1600-0706.2010.18293.x CrossRefGoogle Scholar
  8. Colombaroli D, Henne PD, Kaltenrieder P, Gobet E, Tinner W (2010) Species responses to fire, climate and human impact at tree line in the Alps as evidenced by palaeo-environmental records and a dynamic simulation model. J Ecol 98:1346–1357CrossRefGoogle Scholar
  9. Council of the European Union (1992) Document 31992L0043: Council Directive 92/43/EEC of 21 May 1992 on the conservation of natural habitats and of wild fauna and flora. Official J L 206:7–50Google Scholar
  10. Delkov N (1988) Dendrology. Sofia, ZemizdatGoogle Scholar
  11. Diaci J, Firm D (2011) Long-term dynamics of a mixed conifer stand in Slovenia managed with a farmer selection system. For Ecol Manag 262:931–939. doi:10.1016/j.foreco.2011.05.024 CrossRefGoogle Scholar
  12. EEA (2010) Eurpoe’s ecological backbone: recognising the true value of our mountains. EEA Report, CopenhagenGoogle Scholar
  13. Elkin C, Gutiérrez A, Leuzinger S, Manusch C, Temperli C, Rasche L, Bugmann H (2013) A 2 C warmer world is not safe for ecosystem services in the European Alps. Glob Change Biol 19:1827–1840. doi:10.1111/gcb.12156 CrossRefGoogle Scholar
  14. Frehner M, Waser B, Schwitter R (2005) Sustainability and controlling in protection forests. Guidelines for tending forests with protective function. Bundesamt für Umwelt, Wald und Landschaft (BUWAL), BernGoogle Scholar
  15. Galabov Z (ed) (1973) Atlas of the People’s Republic of Bulgaria. Sofia, Main Bureau of Geodesy and CartographyGoogle Scholar
  16. GFSAF (2000–2014) Report forms 5 GF (Forest Fund). Bulgarian State Forest AgencyGoogle Scholar
  17. Gonzalez P, Neilson R, Lenihan J, Drapek R (2010) Global patterns in the vulnerability of ecosystems to vegetation shifts due to climate change. Glob Ecol Biogeogr 19:755–768. doi:10.1111/j.1466-8238.2010.00558.x CrossRefGoogle Scholar
  18. Guénette J, Villard M (2005) Thresholds in forest bird response to habitat alteration as quantitative targets for conservation. Conserv Biol 19(4):1168–1180CrossRefGoogle Scholar
  19. Hanewinkel M, Cullmann D, Schelhaas M-J, Nabuurs G-J, Zimmermann N (2012) Climate change may cause severe loss in the economic value of European forest land. Nat Clim Change 3:203–207. doi:10.1038/nclimate1687 CrossRefGoogle Scholar
  20. Hartl-Meier C, Zang C, Dittmar C, Esper J, Göttlein A, Rothe A (2014) Vulnerability of Norway spruce to climate change in mountain forests of the European Alps. Clim Res 60:119–132. doi:10.3354/cr01226 CrossRefGoogle Scholar
  21. Haylock M, Hofstra N, Klein Tank A, Klok E, Jones P, New M (2008) A European daily high-resolution gridded dataset of surface temperature and precipitation. J Geophys Res 113:D20119. doi:10.1029/2008JD10201 CrossRefGoogle Scholar
  22. Henne PD, Elkin CM, Reineking B, Bugmann H, Tinner W (2011) Did soil development limit spruce (Picea abies) expansion in the Central Alps during the Holocene? Testing a palaeobotanical hypothesis with a dynamic landscape model. J Biogeogr 38:933–949CrossRefGoogle Scholar
  23. Hewitt C, Griggs D (2004) Ensembles-based predictions of climate changes and their impacts. EOS Trans AGU 85(52):566. doi:10.1029/2004EO520005 CrossRefGoogle Scholar
  24. Huntley B (1991) How plants respond to climate change—migration rates, individualism and the consequences for plant communities. Ann Bot 67:15–22Google Scholar
  25. Iankov P (ed.) (2007) Atlas of breeding birds in Bulgaria. BSPB conservation series, Book 10, SofiaGoogle Scholar
  26. IPCC (2006) Guidelines for National Greenhouse Gas Inventories. http://www.ipcc-nggip.iges.or.jp/public/2006gl/vol4.html. Accessed 20 Feb 2015
  27. IPCC (2014) Climate change 2014: synthesis report. Contribution of Working Groups I, II and III to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, GenevaGoogle Scholar
  28. Jolly M, Dobbertin M, Zimmermann N, Reichstein M (2005) Divergent vegetation growth responses to the 2003 heat wave in the Swiss Alps. Geophys Res Let 32:L18409CrossRefGoogle Scholar
  29. Landsberg J, Waring R (1997) A generalised model of forest productivity using simplified concepts of radiation-use efficiency, carbon balance and partitioning. For Ecol Manag 95:209–228CrossRefGoogle Scholar
  30. Lexer MJ, Brooks R (2005) Decision support for multiple purpose forestry. For Ecol Manag 207:1–3. doi:10.1016/j.foreco.2004.11.002 CrossRefGoogle Scholar
  31. Lexer MJ, Hönninger K (2001) A modified 3D-patch model for spatially explicit simulation of vegetation composition in heterogeneous landscapes. For Ecol Manag 144:43–65. doi:10.1016/S0378-1127(00)00386-8 CrossRefGoogle Scholar
  32. Marinov I, Velizarova E, Nyagolov I et al (2012) Climate change and their effect on the forest ecosystems and water resources in the Struma basin. BON editorial house, BlagoevgradGoogle Scholar
  33. MEA (2005) Ecosystems and human well-being: current state and trends. Island Press, WashingtonGoogle Scholar
  34. Nikolov B, Shurlinkov P, Hristova-Nikolova I (2011) Bird density and species composition in Sweet Chestnut (Castanea sativa) dominated forests in the Bulgarian part of Belasitsa Mountain. In: State and prospects of the Castanea sativa population in Belasitsa Mountain: climate change adaptation; maintenance of biodiversity and sustainable ecosystem management (Zlatanov T, I Velichkov, B Nikolov eds). Project BG 0031 EEA report. http://www.castbelbg.com/. Accessed 21 February 2015
  35. Nyland R (1996) Silviculture concepts and applications. Mc Graw-Hill, New YorkGoogle Scholar
  36. Paulsen J, Weber M, Körner C (2000) Tree growth near treeline: abrupt or gradual reduction with altitude? Arct Antarct Alp Res 32:14–20CrossRefGoogle Scholar
  37. Peguero-Pina J, Camarero J, Abadía A, Martín E, González-Cascón R, Morales F, Gil-Pelegrín E (2007) Physiological performance of silver-fir (Abies alba Mill.) populations under contrasting climates near the south-western distribution limit of the species. Flora 202(3):226–236. doi:10.1016/j.flora.2006.06.004 CrossRefGoogle Scholar
  38. Peñuelas J, Boada M (2003) A global change-induced biome shift in the Montseny Mountains (NE Spain). Glob Change Biol 9:131–140. doi:10.1046/j.1365-2486.2003.00566.x CrossRefGoogle Scholar
  39. Popov V (2011) Animal (spiders, beetles, butterflies and birds) assemblage patterns and environmental gradients in sweet chestnut dominated forests in the Bulgarian part of Belasitsa Mountain. In: State and prospects of the Castanea sativa population in Belasitsa Mountain: climate change adaptation; maintenance of biodiversity and sustainable ecosystem management (Zlatanov T, I Velichkov, B Nikolov eds). Project BG 0031 EEA report. http://www.castbelbg.com/. Accessed 25 February 2015
  40. Pretzsch H (2009) Forest dynamics, growth and yield: from measurement to model. Springer, BerlinCrossRefGoogle Scholar
  41. Ranius T, Fahrig L (2006) Targets for maintenance of dead wood for biodiversity conservation based on extinction thresholds. Scand J For Res 21:201–208. doi:10.1080/02827580600688269 CrossRefGoogle Scholar
  42. Rauscher H, Reynolds K, Vacik H (2005) Decision support systems for forest management. Comput Electron Agric 49(1):1–5. doi:10.1016/j.compag.2005.02.001 CrossRefGoogle Scholar
  43. Rolland C, Petitcolas V, Michalet R (1998) Changes in radial tree growth for Picea abies, Larix decidua, Pinus cembra and Pinus uncinata near the alpine timberline since 1750. Trees Struct Funct 13:40–53. doi:10.1007/PL00009736 Google Scholar
  44. Savva Y, Oleksyn J, Reich P, Tjoelker M, Vaganov E, Modrzynski J (2006) Interannual growth response of Norway spruce to climate along an altitudinal gradient in the Tatra Mountains, Poland. Trees 20:735–746. doi:10.1007/s00468-006-0088-9 CrossRefGoogle Scholar
  45. Sayer J, Maginnis S (2013) Forests in landscapes: ecosystem approaches to sustainability. Taylor & Francis, New YorkGoogle Scholar
  46. Schumacher S, Bugmann H (2006) The relative importance of climatic effects, wildfires and management for future forest landscape dynamics in the Swiss Alps. Glob Change Biol 12:1435–1450CrossRefGoogle Scholar
  47. Schumacher S, Bugmann H, Mladenoff DJ (2004) Improving the formulation of tree growth and succession in a spatially explicit landscape model. Ecol Model 180:175–194CrossRefGoogle Scholar
  48. Schumacher S, Reineking B, Sibold J, Bugmann H (2006) Modeling the impact of climate and vegetation on fire regimes in mountain landscapes. Landsc Ecol 21:539–554CrossRefGoogle Scholar
  49. Schütz J-P (1999) Close-to-nature silviculture: is this concept compatible with species diversity? Forestry 72(4):359–366CrossRefGoogle Scholar
  50. Seidl R, Lexer MJ, Jäger D, Hönninger K (2005) Evaluating the accuracy and generality of a hybrid patch model. Tree Physiol 25(7):939–951CrossRefGoogle Scholar
  51. Seidl R, Rammer W, Lexer MJ (2011) Climate change vulnerability of sustainable forest management in the Eastern Alps. Clim Change 106:225–254. doi:10.1007/s10584-010-9899-1 CrossRefGoogle Scholar
  52. Seidl R, Eastaugh C, Kramer K, Maroschek M, Reyer C, Socha J, Vacchiano G, Zlatanov T, Hasenauer H (2013) Scaling issues in forest ecosystem management and how to address them with models. Eur J For Res 132:653–666. doi:10.1007/s10342-013-0725-y CrossRefGoogle Scholar
  53. SLFEMP (2007) Shiroka Laka. Forest Enterprise Management Plan, AgrolesproektGoogle Scholar
  54. Temperli C, Bugmann H, Elkin C (2012) Adaptive management for competing forest goods and services under climate change. Ecol Appl 22:2065–2077CrossRefGoogle Scholar
  55. Thornton P, Running S (1999) An improved algorithm for estimating incident daily solar radiation from measurements of temperature, humidity, and precipitation. Agr For Meteorol 93:211–228CrossRefGoogle Scholar
  56. UNFCCC (1997) Kyoto protocol to the united nations framework convention on climate change. http://unfccc.int/resource/docs/convkp/kpeng.pdf. Accessed 25 Feb 2015
  57. Velichkov V, Krijan-Velichkova E, Cenov C (2008) Possibilities for valuable timber production in the Rodopes. Manag Sustain Dev 1(19):68–77Google Scholar
  58. Velichkov I, G Popov, G Hinkov, T Zlatanov (2009) Resources, issues, and management perspectives of beech forests in Rodopi mountains. In: G Rafailov (ed.), National conference on beech forests management in Bulgaria, Barziya, 16–17 June 2009, “RUTA–HB”, pp 87–100Google Scholar
  59. Von Lüpke B, Spellmann H (1999) Aspects of stability, growth and natural regeneration in mixed Norway spruce-beech stands as a basis of silvicultural decisions. In: Management of mixed-species forest: silviculture and economics (eds. Olsthoorn A, Bartelink H, Gardiner J). IBN Scientific Contributions 15, Institute of Forestry and Nature Research, Wageningen, the Netherlands, pp 245–267Google Scholar
  60. Woltjer M et al (2008) Coupling a 3D patch model and a rockfall module to assess rockfall protection in mountain forests. J Environ Manag 87:373–388CrossRefGoogle Scholar
  61. Zahariev B, Duhovnikov Y, Iliev A et al (1977) The forestation in Bulgaria. Zemizdat, SofiaGoogle Scholar
  62. Zang C, Rothe A, Weis W, Pretzsch H (2011) Zur Baumarten-eignung bei Klimawandel: ableitung der Trockenstress-Anfälligkeit wichtiger Waldbaumarten aus Jahrring-breiten. Allg For J Ztg 182:98–112Google Scholar
  63. Zlatanov T, Velichkov I, Lexer M, Dubravac T (2010) Regeneration dynamics in aging black pine (Pinus nigra Arn.) plantations on the south slopes of the middle balkan range in Bulgaria. New For 40:289–303. doi:10.1007/s11056-010-9200-5 CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2015

Authors and Affiliations

  1. 1.Forest Research Institute SofiaSofiaBulgaria
  2. 2.Forest EcologyETH Zürich Switzerland & University of Northern British ColumbiaPrince GeorgeCanada
  3. 3.University of Natural Resources and Life SciencesViennaAustria

Personalised recommendations