Regional Environmental Change

, Volume 17, Issue 1, pp 3–16 | Cite as

Impacts of business-as-usual management on ecosystem services in European mountain ranges under climate change

  • Harald Bugmann
  • Thomas Cordonnier
  • Heimo Truhetz
  • Manfred J. Lexer
Original Article

Abstract

Mountain forests provide a multitude of services beyond timber production. In a large European project (ARANGE—Advanced multifunctional forest management in European mountain RANGEs), the impacts of climate change and forest management on ecosystem services (ES) were assessed. Here, we provide background information about the concept that was underlying the ARANGE project, and its main objectives, research questions, and methodological approaches are presented. The project focused on synergies and trade-offs among four key ES that are relevant in European mountain ranges: timber production, carbon storage, biodiversity conservation, and protection from gravitational natural hazards. We introduce the concept and selection of case study areas (CSAs) that were used in the project; we describe the concept of representative stand types that were developed to provide a harmonized representation of forest stands and forest management in the CSAs; we explain and discuss the climate data and climate change scenarios that were applied across the seven CSAs; and we introduce the linker functions that were developed to relate stand- and landscape-scale forest features from model simulations to ES provisioning in mountain forests. Finally, we provide a brief overview of the Special Feature, with an attempt to synthesize emerging response patterns across the CSAs.

Keywords

Climate impact assessment Mountain regions Decision support ARANGE project 

Introduction

More than 40% of the European land surface is covered by mountains (Price et al. 2004), and forests cover a disproportionately large fraction of this area (Price et al. 2011). Thus, mountain forests are a very important land cover type, and they provide a multitude of ecosystem services (ES) to the European population, both in the mountains themselves as well as downstream (EEA 2010). Mountain forests provide valuable products such as timber, mushrooms or game. They represent considerable carbon stores, not only because of their large area, but also because of often reduced management intensity compared to the lowlands; they may thus be particularly important for further carbon sequestration in a changing climate, or with continuing abandonment of agricultural land use. Mountains harvest precipitation from the atmosphere, store it in ice, snow, and deep forest soils; they thus contribute to flood mitigation and provide a key freshwater resource for downstream people, plants, and animals. Mountain forests also harbor a significant fraction of the known terrestrial biodiversity of the continent. Further, mountain forests often protect critical human infrastructure such as settlements, roads, and railway lines from gravitational natural hazards such as landslides, avalanches, or rockfall. And lastly, mountains are among the key global destinations for tourism because of their aesthetic beauty and recreational value, which is partly mediated by ecosystems in general and forests in particular.

Often, the demand for ES from a single parcel of mountain forest is manifold, including protection from natural hazards, timber production, and landscape aesthetics (regarding tourism and recreation). Nature conservation values and biodiversity aspects are typically involved as well. The provisioning of these services will be affected and may be threatened considerably by climatic change (e.g., Elkin et al. 2013). It is noteworthy that it is highly likely that warming will be particularly pronounced at higher elevations (Pepin et al. 2015). Thus, the question arises whether the management regimes that have been developed in the past, i.e., under current climatic conditions (hereafter referred to as “business-as-usual management”, BAUM), will still be suitable in the coming decades, or whether changed regimes are required that would need to be planned now and implemented in the coming few decades, due to the slow development and strong lag effects that are characteristic of forests. There is little systematic knowledge on these issues, and although there are case studies that have evaluated some aspects for individual mountain ranges, an assessment is lacking that covers the vastly different environmental, ecological and societal settings that characterize European mountain ranges.

The ARANGE project (acronym for “Advanced multifunctional forest management in European mountain RANGEs”) focused on the impacts of climate change and management on the synergies and trade-offs among four key ES in European mountain ranges: timber production, carbon storage, biodiversity conservation, and protection from gravitational natural hazards. The consortium comprised 16 partner institutions from 11 countries as well as the European Forest Institute (EFI) as an international partner.

Research activities were conducted in seven European mountain ranges that were chosen to represent a broad spectrum of climatic conditions, ranging from mediterranean to boreal, and a wide array of socioeconomic situations, ranging from countries in transition to fully market-based economies where forests still are of high economic value, across boreal countries where commercial forestry and timber production is of pivotal importance, to western European economies where forests are often seen as providers of various ES but not so much as providers of income.

ARANGE was based on a comprehensive and integrative approach with the following key objectives:
  1. 1.

    To produce a consistent spatial and temporal database covering environmental, economic and social information and demands for the seven case study areas (CSAs).

     
  2. 2.

    To adapt forest ecosystem modeling tools to simulate past and future forest development in the CSAs, and to quantify and assess the provisioning of ES according to end user needs.

     
  3. 3.

    To analyze policy and governance conditions in the CSAs.

     
  4. 4.

    To develop new and advanced methods to support multifunctional mountain forest management planning.

     
  5. 5.

    To synthesize the findings from case study analysis.

     
  6. 6.

    To facilitate stakeholder interaction in all CSAs and disseminate information and tools.

     

In the present Special Feature of Regional Environmental Change, we focus on the question whether BAUM can provide the set of ES that is requested by local stakeholders in the seven CSAs under a set of widely differing scenarios of climate change, including the continuation of the current climate as a baseline and reference framework. The work presented here thus builds upon the achievements of several of the general objectives of the ARANGE project (cf. above).

Below, we provide important background information on the selection of the CSAs; the approach we developed for identifying representative stand types (RSTs) in each CSA; the harmonized database on climate and climate change scenarios that we applied across all CSAs; and the methodology that we developed further based on earlier work to relate simulated forest stand properties to ecosystem services, using so-called linker functions. Finally, we provide a brief overview of this Special Feature, with an attempt to synthesize emerging response patterns across the CSAs.

The case study areas

European mountain forests are diverse, with each mountain region featuring a distinct set of environmental conditions, tree species, demands for ES and related management goals of land owners, as well as specific risks and uncertainties. Therefore, multifunctional management must be assessed and optimal strategies developed specifically for the different European mountain regions.

Seven mountain regions across the continent were selected, covering the most important forest types in the main mountain ranges and representing distinct biophysical and governance settings (Fig. 1). Within each mountain range, a CSA was defined that covers a minimum of 50–100 km2 of forested mountain landscape. Several spatial levels were distinguished within each CSA: (1) an administrative district (or similar) as the largest level, representing the general socioeconomic framework, including multiple land owners, (2) a functional entity representing a forest management unit, (3) individual catchments or slopes to consider spatial dependencies between ES, and (4) the stand level, on which the work presented in this Special Feature is focusing. From the slope scale upwards, the forest was abstracted as a mosaic of stands differing in species composition, ecosystem properties and developmental stages.
Fig. 1

Overview of the seven case study areas (CSAs) of the ARANGE project

The multifunctionality of forest management was addressed within a time horizon until the year 2100 and at a range of spatial scales that are relevant in practical mountain forest management, such that practically relevant recommendations for policy makers and resource managers could be derived. Key prerequisites for the selection of the specific CSAs were (1) data availability and (2) willingness of regional stakeholders and small and medium enterprises to participate in the project (cf. Table 1 for an overview of the CSAs).
Table 1

Overview of the seven ARANGE case study areas (CSAs)

 

CSA1

CSA2

CSA3

CSA4

CSA5

CSA6

CSA7

Mountain range

Iberian mountains, Sierra Guardarrama

Western Alps

Eastern Alps

Dinaric mountains

Scandinavian mountains

Western Carpathians

Western Rhodopes

Country

Spain

France

Austria

Slovenia

Sweden

Slovakia

Bulgaria

Name of case study

Montes Valsain, Cabeza de Hierro

Vercors

Montafon

Sneznik, Leskova dolina

Vilhelmia

Kozie chrbty

Shiroka laka

Total area (km2)

100 and 20

500

75

50

8500

132

92

Forest area (%)

90

55

90

97

62

90

97

Ownership

Public, private

Public, private

Cooperative, private

State-owned

Public, private, church, municipality

Church

Public, non-industrial private owners, cooperative

Elevational range

1200–1900 m

560–2270 m

600–2000 m

250–1700 m

300–650 m

600–1800 m

700–2000 m

Tree species

Scots pine, pyrenean oak

Spruce, fir, beech

Beech, maple, fir, spruce

Beech, fir, spruce

Scots pine, spruce, birch

Spruce, fir, beech

Scots pine, black pine, fir, beech, spruce

Ecosystem service

TP, CS, NC, REC

TP, BMdE, PF (rockfall), NC

TP, PF (rockfall, avalanches, flooding, erosion), NC, GM

TP, GM, NC, PF (erosion), water retention

TP, CS, NC, RDH

TP, NC, REC, PF (erosion), water resources

TP, BMfE, CS, NC, PF (flooding/erosion), tourism

Planning issues

Currently segregative approach, climate change adaption, fire disturbance

Profitability of TP, Trade-offs within portfolios of ES: (TP, PF, NC, CS), (TP, REC, NC, CS)

TP and PF versus GM, disturbances (storm, insects) under climate change

Profitablity of TP, trade-offs (TP, GM, NC)

TP versus RDH (spatio-temporal problem)

Trade-offs (TP, REC, NC, GM), disturbances (insects, storm), intensive browsing by game

Trade-offs (TP, NC, PF (erosion), water resources, tourism)

Key to ecosystem services: TP timber production, CS carbon sequestration, NC nature conservation, REC recreation, BMfE biomass for energy, PF protective function, GM game management, RDH reindeer herding

The CSAs served four main purposes: (1) to promote interdisciplinary and transdisciplinary research efforts in the analysis and the development of multifunctional mountain forest management; (2) to provide the data and information base to develop and test forest models and methods to quantify ES, and to promote participatory model–data fusion for management planning; (3) to demonstrate the applicability of the planning tools; and (4) to produce per CSA both stand- and landscape-level results for generic portfolios of ES.

Representing a landscape via stand types

The smallest silvicultural planning and treatment unit is a “stand”. A stand is defined as a forest area with reasonably homogeneous conditions regarding (1) site conditions, (2) tree species composition, (3) stand structure, (4) tree age, (5) management objectives, and (6) silvicultural treatments. The area of such stands ranges from about 0.5 ha to as much as 10–20 ha, depending on local conditions. Ownership structure plays an important role as well. In each CSA, a set of RSTs was defined to represent “typical” forest conditions. RSTs are defined via species composition, forest structure, the management system (i.e., even-aged, uneven-aged), which has shaped stand and tree characteristics, and site conditions. In CSAs that are characterized by age class forests (i.e., even-aged systems), the RSTs had to represent the respective age class sequence.

Depending on data availability in the CSAs, RSTs were derived from (1) inventory plot data, (2) stand data available from management plans, or (3) data specifically gathered within the ARANGE project. The RSTs were the core units of simulation and analysis within the ARANGE project. Data used to characterize the RSTs must allow for initializing the various forest simulation models within the ARANGE consortium (e.g., Bugmann 2001; Cordonnier et al. 2008; Lexer and Hönninger 2001; Seidl et al. 2007; Schumacher and Bugmann 2006; Fabrika and Durský 2005; Lämås and Eriksson 2003). Table 2 shows the site, tree, and stand attributes which were consistently provided for each RST across all seven CSA. The number of RSTs in the CSAs ranged from 19 to 88.
Table 2

Attributes to characterize the representative stand types (RSTs)

Attribute

Description (Unit)

Elevational range

(m a.s.l.)

Aspect

(0–360°)

Slope

[%]

Soil type

FAO classification (–)

Soil depth

From top of soil horizon to bedrock (cm)

pH

Measured in the top soil layer (0–20 cm) (–)

Nutrient supply

[poor, intermediate, rich]

Plant-available nitrogen

Estimated from total N in mineral soil and a mineralization rate (0.5–4%) (kg/ha year)

Water storage capacity

Water column that can be stored in the mineral soil profile (soil depth); ranges between 50 mm (sandy, shallow soil) and 250 mm (high storage capacity in loamy deep soils with low rockiness) (mm)

DBH distribution

Class width 0–10, 10–15, 15–20, 20–25 cm etc., per tree species (ha−1)

Tree height

Representative tree height per DBH class and species. Provided via a height-diameter equation or as mean height per diameter class and species [m]

Mixture form

Qualitative description of the mixture form in mixed stands (e.g., random, patches) (–)

Regeneration

Seedling density per tree species in height classes (10–30 cm, 31–50 cm, 51–80 cm, 81–130 cm) (n/ha)

Mixture form

Qualitative description of the distribution and mixture form of the regeneration, (a) patchy, (b) random, (c) regular (–)

To allow the assessment of forest ecosystem provisioning at spatial scales beyond stand level one or more representative landscapes (RL) were defined as well in each CSA. These RL provide a spatially explicit matrix of stand polygons and cover areas from 216 to more than 10,000 hectares. Each of the stand polygons is represented by one of the RSTs (multiple assignments of RSTs to polygons is possible).

Climatic data and climate change scenarios

When climate change impacts need to be quantified, several issues must be considered. (1) There is often a gap in scales along the modeling chain that needs to be bridged: while global climate models (GCMs) operate on grids with a spacing of ~100 km and more (e.g., Meehl et al. 2007), impact models are usually run on scales many times smaller than this. Therefore, GCM results need to be downscaled (cf. Maraun et al. 2010). (2) Climate models are approximations of the climate systems and hence are affected by errors due to simplifications that induce biases (Stainforth et al. 2007). These biases may become too large for impact models (e.g., Finger et al. 2012; Ravazzani et al. 2014; Smith et al. 2014; Stoffel et al. 2014). In such cases, the output of climate models needs to be bias-corrected first. (3) For overcoming issue (1) and (2), long-term observational data of high quality play an essential role for two reasons: first, without knowing the characteristics of the current climate in the study region of concern, issues (1) and (2) cannot be solved, and second, impact models may need to be calibrated in the study region, which is usually accomplished against long-term observations. (4) Our knowledge about the climate system is incomplete (Stainforth et al. 2007), and natural variability inherently induces uncertainty in climate projections (Ghil 2002). (5) Another source of uncertainty is the unknown future human behavior. (6) Since many impact studies are costly, climate projections need to be carefully pre-selected so as not to underestimate uncertainty in climate change impact assessments.

In the face of these difficulties, the work for creating climate and climate change data in ARANGE was based on the widely accepted assumption (cf. van der Linden and Mitchell 2009; Heinrich et al. 2014) that climate change effects and their uncertainty can be estimated reasonably well by means of a limited number of properly selected (but imperfect) climate projections from GCMs that are scaled appropriately to the level of the impact models, i.e., in the case of ARANGE, forest stands and landscapes.

Calculation of climate change scenarios for the CSAs

In the first step, a baseline climate reflecting current (1961–1990) area-wide climate conditions for each CSA was created (Thurnher 2013) based on in situ meteorological observation stations, or the gridded observational dataset E-OBS (Haylock et al. 2008) if no observation stations were available. This baseline climate consists of stochastic time series on a daily basis for maximum temperature, minimum temperature, daylight temperature, precipitation, global radiation, and vapor pressure deficit. These time series were generated by well-known and robust methods: the weather generator LARS-WG (Racsko et al. 1991; Semenov and Barrow 1997) and the geo-statistical post-processing tool MT-CLIM (Running et al. 1987; Thornton and Running 1999; Thornton et al. 2000) to implement the dependencies on elevational zones, slope, and aspect.

In the second step, transient climate projections reflecting future conditions in each CSA covering the period up to the year 2100 were created (Truhetz 2013). For that purpose, the baseline climate and regional climate model (RCM) simulations from the EU-FP6 project ENSEMBLES (Hewitt and Griggs 2004) were combined. In ENSEMBLES, numerous highly resolved (~25-km grid spacing) RCM simulations were conducted for the European continent by making use of GCM simulations that were based on the greenhouse gas (GHG) emission scenario A1B (Nakicenovic et al. 2000). Note that the baseline climate can be used to calculate the area-averaged climate for each CSA, which is of the same spatial extent as a single grid cell of the ENSEMBLES RCMs. For ARANGE, a subset of 14 ENSEMBLES simulations was selected that are reaching to the end of the twenty-first century.

To take CSA-specific effects into account and to correct biases in the RCM data, a quantile-based approach (QM, Quantile Mapping; cf. Dobler and Ahrens 2008; Piani et al. 2010; Themeßl et al. 2011) was trained with the baseline climate (1961–1990) and applied to the 14 ENSEMBLES simulations. QM is an empirical–statistical downscaling and bias correction method that aims at adjusting the empirical cumulative distribution function of the model data toward a reference dataset. Therefore, it accounts for errors in the shape of the modeled distribution (Themeßl et al. 2011; Déqué 2007). Although QM has some limitations (e.g., Maurer and Pierce 2014), we selected it because it has been found to be robust and flexible even for non-normally distributed variables such as precipitation (e.g., Dobler and Ahrens 2008; Themeßl et al. 2011, 2012; Piani et al. 2010; Finger et al. 2012; Wilcke et al. 2013).

In the third and final step, a subset of five representative climate projections per CSA that span the range of uncertainty in climate projections reasonably well was selected. Uncertainty due to natural variability was implicitly taken into account via the use of different GCMs in ENSEMBLES, but uncertainty due to the GHG emission scenarios was not included (i.e., all simulations were forced by the scenario A1B). However, Prein et al. (2011) found that the largest contribution to uncertainty in climate change signals comes from the formulation of the climate models themselves. While the selection of climate simulations is not straightforward and still inconclusive (cf. Knutti et al. 2010; Déqué and Somot 2010), a pragmatic approach was used in ARANGE: in order to mimic the ENSEMBLES range, four enveloping projections plus one lying next to the ensemble median were selected for each CSA. This selection was based on mean changes of the bias-corrected daylight temperature and precipitation (and global radiation in those CSAs where the precipitation signal dominates) of the summer half year.

Climate change scenarios for the CSAs

While temperatures and temperature-based variables (e.g., vapor pressure deficit) are expected to increase throughout the CSAs, precipitation and solar radiation do not show such a consistent behavior (cf. Table 3). In addition to regional deviations (e.g., increase in Northern Europe, decrease in Southern Europe; also known as the European Climate Change Oscillation; Giorgi and Coppola 2007; Heinrich et al. 2014), precipitation also shows deviations along the annual cycle (shifts to positive changes in the winter half year). For instance, in CSA4 (Dinaric Mountains) precipitation is expected to decrease in summer and to increase in winter. In CSA3 (Eastern Alps) and CSA6 (Carpathians), however, either trend (increase and decrease) seems to be equally possible. This model disagreement indicates high uncertainty in the climate change effects. Highest uncertainty, in general, may be expected for solar radiation: five out of the seven CSAs experience both an increase as well as a decrease, depending on the scenario that is taken into consideration.
Table 3

Ranges of climate change signals (2071–2100 vs. 1961–1990) in the seven ARANGE CSAs for a flat plane on highest elevations

 

Daylight temperature (°C)

Maximum temperature (°C)

Minimum temperature (°C)

Precipitation (%)

Water vapor pressure deficit (%)

Solar radiation (%)

CSA1 Cabeza (2000 m)

3.2–6.4

(3.2–5.2)

3.3–6.6

(3.5–5.6)

2.8–5.9

(2.4–4.3)

−52.0 to −27.5

(−24.4 to 3.0)

22.7–53.3

(55.5–85.0)

2.5–6.4

(2.2–9.6)

CSA1 Valsain (2000 m)

3.5–6.9

(3.5–5.8)

3.5–6.9

(3.8–6.1)

3.4–6.9

(2.7–5.1)

−51.9 to −23.6

(−20.6 to 2.4)

21.6–52.1

(52.2–84.5)

2.5–7.8

(2.4–10.7)

CSA2 (1800 m)

1.7–5.3

(3.4–5.6)

1.7–5.7

(3.6–6.0)

1.7–4.4

(2.9–4.8)

−49.2 to 0.6

(3.8–22.8)

10.4–52.3

(33.6–64.7)

0.1–8.4

(−3.9 to 5.2)

CSA3 (2000 m)

1.7–5.3

(4.0–7.4)

1.7–5.8

(4.4–8.2)

1.7–4.1

(3.0–5.2)

−26.8 to 13.3

(5.4–37.6)

11.1–67.0

(65.4–151.5)

−2.7 to 7.3

(−6.1 to 2.0)

CSA4 (1800 m)

1.4–5.0

(2.8–6.1)

1.4–5.2

(2.8–6.5)

1.4–4.5

(2.7–5.1)

−42.4 to −5.0

(−2.0 to 19.4)

9.4–41.9

(25.7–70.0)

−12.7 to 3.9

(−7.5 to 3.7)

CSA5 (800 m)

2.1–5.4

(4.4–6.4)

2.1–5.7

(4.0–6.1)

2.1–4.7

(5.3–7.2)

2.6–28.0

(22.7–64.8)

9.4–50.3

(24.8–56.5)

−11.4 to  0.2

(−6.2 to 1.3)

CSA6 (1550 m)

1.2–4.5

(3.0–6.3)

1.2–4.8

(3.1–6.6)

1.2–3.7

(3.0–5.4)

−28.5 to 9.6

(12.3–53.8)

7.3–43.6

(33.9–87.9)

−8.4 to 4.5

(−9.4 to 1.1)

CSA7 (2000 m)

2.1–6.4

(3.0–6.7)

2.2–6.8

(3.1–7.2)

1.7–5.3

(2.7–5.3)

−47.9 to −15.8

(−22.3 to −1.9)

18.8–65.0

(33.1–96.6)

−17.9 to 6.6

(−13.3 to 9.4)

Changes of the averaged summer half year (April 16th to September 15th) and winter half year (in parentheses) are shown

The temporal development of the climate projections (cf. Figure 2 for an example in CSA1) reveals further climate change effects: (1) the range of the projections increases in time, and hence, uncertainty at the end of the twenty-first century is larger than for the near future; (2) the variability in single projections increases over time; hence changes in meteorological conditions from one year to the next are expected to progressively increase and exceed current patterns.
Fig. 2

Effect of climate change in CSA1 (Cabeza de Hierro) expressed in terms of the development of the annual mean values of bias-corrected daylight temperature from 1961 to 2100. The graphs depict the five selected representative simulations (colored) and the baseline climate (black)

Linking forest stand properties to ecosystem services

The definition of linker functions, i.e., algorithms that provide an unbiased and accurate quantification of ecosystem service provision from forest simulation model outputs, is challenging because it must be based on three features. First, it calls for models that are able to reproduce key properties of forest dynamics, structure, and functioning (Lafond et al. 2015). Second, it implies that ES can be quantified sufficiently well by the scale of outputs and ecological entities considered in the model. Finally, it necessitates developing a method for assessing ecosystem service provision from specific model outputs (Duncker et al. 2012; Schwenk et al. 2012). Every step has its own scientific and technical challenges, whose discussion is beyond the scope of this paper (cf. Cordonnier et al. 2013).

In the ARANGE project, the linker functions were partly constrained by the set of forest models and the set of output variables that they were able to produce. Below, we provide a brief overview of the rationale and principles as well as the choices we made (cf. Table 4). Here, we focus on the linker functions for the key ecosystem services that were considered in all CSAs: timber production, carbon storage, protection against gravitational hazards, as well as biodiversity and nature conservation.
Table 4

Overview of the indicators used to link forest properties as simulated by the ARANGE forest models to ecosystem services

For details, see text and Cordonnier et al. (2013)

Rationale and principles

We defined as linker functions a set of indices based either on stand structural properties or on fluxes related to management actions (e.g., wood extraction). Structure encompasses species abundance, diameter and height distributions, density, standing deadwood, and coarse woody debris (CWD). Many studies have shown a strong dependence of ES on these properties, provided that the effect of environmental conditions (soil, climate, topography) is taken into account; this is well recognized for protection against gravitational hazards (Dupire et al. 2016), wood production (Forrester and Bauhus 2016; Pretzsch and Rais 2016), and carbon stocks (Wang et al. 2011). However, the relationship is less straightforward for biodiversity due to its multidimensional character (taxa, species abundances, species richness, composition, etc.). Several studies have found that the quality and quantity of CWD as well as some specific characteristics of living trees strongly influence the richness of forest-dwelling species in a wide range of conditions (Gao et al. 2014; Lassauce et al. 2011). Thus, the use of stand structural attributes to quantify biodiversity effects may be coarse, but it is a pragmatic approach to assess forest habitat quality for biodiversity conservation (e.g., Redon et al. 2014).

Scale represents another key element to consider for assessing ES. As indices related to wood production and carbon stocks display additive properties, the results can be scaled easily from RSTs to the landscape. For protection against gravitational hazards (spatial propagation) and biodiversity conservation (scale dependency of metrics), specific scales and indices have to be chosen. In ARANGE, we defined coherent indices at both the stand and the landscape (or slope) scale, so as to allow for cross-scale comparisons between CSAs (cf. Table 4).

Wood production

Wood production was represented by four metrics: total volume of harvested timber; volume harvested per species and DBH class; forest productivity; and forest stocking. Each metric was reported on a volumetric basis, such that harvested timber and productivity were measured in m3 ha−1 year−1, while forest stocking was measured in m3 ha−1. Of the nine forests models in ARANGE, seven provided forest development on a volumetric basis directly. For the others, their output was converted to volume post hoc using algorithms that are internally consistent with the simulated growth and allometric relationships of the model. For all wood production metrics, 5-cm DBH classes were used, with a minimum diameter of 5 cm. Volume was defined as total bole volume over bark without leaves and branches.

Carbon stocks

Above- and belowground carbon storage in living tree biomass was calculated by all forest models in all CSAs. In addition, carbon in coarse woody debris (CWD; standing and lying) was calculated by those models that included these components (e.g., PICUS), whereas soil carbon content was calculated using the model BIOME-BGC for all CSAs (Pietsch et al. 2005). All carbon storage metrics were calculated in t ha−1. Along with the wood production metrics, all models that did not natively simulate carbon pools in units of t ha−1 converted their output using algorithms that are internally consistent with the simulated growth and allometric relationships of the models. Equations and parameters for calculating forest carbon pools based on forest biomass, wood volume and tree size inputs were adopted from IPCC (2006), Nabuurs et al. (2003) and Vallet et al. (2006).

Protection against gravitational hazards

A rockfall protection index was calculated based on the concept of the probable residual hazard, which corresponds to the percentage of rocks that are able to pass through and exit a forested slope (cf. ROCKFORNET, http://www.ecorisq.org/en/rockfornet.php; Berger and Dorren 2007). This metric was calculated for a slope or for a representative stand with a reference distance along the slope of 250 m, which is the minimal distance for which a stand can provide effective protection. The probable residual hazard is defined as the ratio between the dissipating maximal energy developed by the rock (calculated using the energy line principle) and the energy dissipation ability of the current stand.

For snow avalanches, snow interception represents 70% of the protection effect, and mechanical anchorage of the snow by stems accounts for 30% of the effect (Berger 1997). As for rockfall, an avalanche protection index was calculated per stand based on the ratio of current stand properties to those needed for instantaneous optimum protection. The main assumption is that for a given mean diameter at breast height, basal area is the dendrometric variable that synthesizes both the interception and mechanical effects. Provided that one can specify the reference basal area that is needed to avoid a snow avalanche release, it is possible to calculate a snow protection index via the ratio of current stand basal area to reference basal area; evergreen and deciduous species need to be considered separately.

For landslides, we used the recommendations provided by the Swiss NaiS (Frehner et al. 2005) and French GSM (Gauquelin and Courbaud 2006) projects. Forests can reduce the likelihood and extent of landslides or erosion by mechanically reinforcing the soil through their rooting system and positively influence the water balance (in the sense of reducing water loading to and in the soil, which is conducive to erosion) through interception, transpiration, higher water holding capacity, and enhanced soil permeability (Frehner et al. 2005). Well-developed, multi-layered forests provide the greatest protection, as they provide an extensive rooting system that minimizes landslide potential. Guidelines suggest that in areas where landslides may originate, the minimum requirement is a multi-layered forest with canopy cover ≥30 or ≥40%, depending on the source (cf. above). The guidelines agree that the ideal profile is a multi-layered forest with canopy cover ≥60%. We thus calculated a landslide protection index (LPI) using these three thresholds (low, medium, and high protection effect) and forest cover calculated by the models.

Biodiversity conservation

We defined six complementary indices. The first deals with tree species diversity, which is considered a major forest structural feature (Pommerening 2002) and may influence forest functioning (Nadrowski et al. 2010) as well as other biodiversity components such as floristic diversity (e.g., Zilliox and Gosselin 2014). We used the “equivalent number of tree species,” which is the exponential value of the Shannon entropy index (Jost 2006). It is based on the number of species and their relative abundances as measured by basal area.

The second index captures tree size diversity, which is often considered in studies relating stand structure to biodiversity (McElhinny et al. 2005). The main assumption is that high tree size diversity increases the diversity of habitats for forest-dwelling species (Rouvinen and Kuuluvainen 2005; Bagnaresi et al. 2002). We used the post hoc index defined by Staudhammer and LeMay (2001) without the species diversity component. It corresponds to the Shannon entropy index applied to the basal area of the different size classes (calculated for both height and diameter).

CWD volume is a good surrogate for the diversity of saproxylic species (Martikainen et al. 2000; Grove 2002), as it provides habitats as well as resources for these species (Müller and Butler 2010; Müller et al. 2008). Moreover, it is directly related to tree removal and tree retention practices and, thus, is a key feature to characterize the trade-off between timber production and biodiversity conservation. Standing dead trees (snags) contain more microhabitats for saproxilic species than living trees (Vuidot et al. 2011; Fan et al. 2003) and provide specific habitats for some species compared to lying CWD. The number of large standing dead trees (dbh ≥ 30 cm) was selected as a complementary index to total CWD volume. To this end, the snag submodel of PICUS (Seidl et al. 2007) was implemented in other models.

Trees with large diameter are known to contain more microhabitats (e.g., cavities, dead, and branches) than small trees (Vuidot et al. 2011; Larrieu and Cabanettes 2012; Nilsson et al. 2002; Michel and Winter 2009; Winter and Möller 2008). Some studies found that the probability of carrying a microhabitat is low for trees with DBH <30 cm (Vuidot et al. 2011; Fan et al. 2003; Schreiber and deCalesta 1992) or DBH <40 cm (Larrieu et al. 2012; Larrieu and Cabanettes, 2012). Several studies suggest that the probability and abundance of microhabitats increases with DBH, with a significant threshold around 60–70 cm, depending on the species. Although a clear threshold is observed at DBH ≈ 70 cm for conifers (Larrieu et al. 2012; Michel and Winter 2009; Schreiber and deCalesta 1992), there seems to be more variability for broadleaves, with values ranging from 50 cm (Lachat and Bütler 2007) to 70 cm (Larrieu et al. 2012) and 90 cm (Larrieu and Cabanettes 2012). We selected the number of large living trees with DBH >50 and >70 cm for broadleaved and conifer species, respectively, as the fifth biodiversity index.

Habitat quality models are complementary to the indices described above as they target specific species or specific groups of species. In the ARANGE project, bird habitat quality indices were defined based on scientific and expert knowledge. The selected group of typical forest bird species consists of all the woodpeckers that are present in the case study areas, Tengmalm’s owl (Aegolius funereus), which is common to almost all case study areas, and the Eurasian tree creeper (Certhia familiaris). Based on thresholds, a score (poor, medium, good) was defined for each of the following five indicators: volume of standing dead wood (dbh > 30 cm), time since last harvest, number of veteran trees (dbh > 50 cm), canopy cover, and share of alien tree species. Then, specific aggregation rules were applied to compute a bird habitat quality score at the stand level (poor, medium, good). For instance, the score is “good” if at least three out of five indicators are classified as “good” (Cordonnier et al. 2013).

Forest responses to changes in climate in European mountains

As mentioned earlier, the current Special Feature provides results quantifying the impacts of climate change on mountain forests from five European regions (cf. Figure 1): Valsain in the Sierrra da Guadarrama in central Spain (CSA1; Pardos et al. 2016), Montafon in the Eastern Alps in Austria (CSA3; Irauschek et al. 2015), Sneznik in the Dinaric Mountains in Slovenia (CSA4; Mina et al. 2015), the Carpathian mountains in Slovakia (CSA6; Hlásny et al. 2015), and the Western Rhodopes (CSA7; Zlatanov et al. 2015) in Bulgaria (cf. Table 1). Several similarities, but also striking differences arise from the comparison of the findings across these climatically, ecologically and societally vastly different regions.

First, in all CSAs there was a strong elevational gradient of forest responses to climate change, i.e., there is no such thing as “the response of forests in the Rhodopes Mountains to climate change.” Uniformly, precipitation levels at high elevation are rather high today and will remain sufficiently high also under the climate change scenario investigated in this project, such that drought effects at high elevations will remain negligible. Instead, increases in temperature typically will result in higher biomass production. Moderate decreases in growth at high elevations were found in exceptional cases only, i.e., when climate changes induced a deficit in water budget, namely in the continental case study regions of ARANGE. On the contrary, at low elevations drought almost always was found to lead to strong negative impacts on tree growth, triggering mortality events in particularly sensitive tree species such as Picea abies, which is a major timber species today also at low elevations. This general response pattern was determined by current species composition, but it was independent of the currently applied BAUM regime.

Second, the variation in current climate across European mountain ranges leads to strongly different sensitivity to climate change. Specifically, precipitation levels under current climate are considerably lower in the more continental CSAs 6 (Carpathians) and 7 (Rhodopes) compared to the other CSAs, which makes forest response and ES vulnerability to climate change particularly drastic in these latter regions.

Third, within the time frame of our investigations, i.e., within the twenty-first century, quite extreme scenarios of climate change are required to produce very substantial impacts on forest growth, composition, and dynamics in most CSAs. The more “mellow” climate scenarios did not often lead to dramatic changes of forest properties. This is at least partly due to the fact that forests, as long-lived systems, feature particularly pronounced lag effects, i.e., the stronger signals of climate change that are becoming evident in the second half of the twenty-first century will have visible consequences for forests well beyond the end of this century, but these were not investigated systematically in the ARANGE project. Thus, the fact that small changes are looming over the coming decades does not imply that larger changes would be absent in the longer term.

Fourth, in CSAs that feature a relatively dry climate already today or that are prone to possibly experiencing a strong decrease in precipitation in the future (as was the case for CSA1, Valsain), uncertainty with regard to suitable future tree species is exceedingly large. This is due to the fact that the extent of drought that will be experienced by future forests depends very strongly on the exact climatic future (scenario) that is chosen. This is due to the fact that responses to drought are highly nonlinear, whereas the temperature response is likely to be more linear. Under severe climate change scenarios, future conditions may be close to the drought tolerance limits of most species. Thus, more research on drought thresholds and tree mortality is highly needed.

Fifth, the studies reported in this Special Feature covered the continuation of business-as-usual management only. While it became evident that this approach would not lead to dramatic deterioration of ecosystem services provisioning except under climatically very harsh future conditions, it may still be feasible to improve the provisioning of ecosystem services under more moderate scenarios of climate change by the use of adaptive forest management practices. These are the subject of further publications from ARANGE and related projects.

Finally, our results should be interpreted with the caveat that due to the partly large uncertainties along the modeling chain, climate change effects in the real world may lie outside the simulated bandwidths. In addition, large-scale disturbances such as windthrow, wildfires, and insect attacks were not considered, at least not in a systematic fashion across CSAs. Even though browsing of regeneration, which is another widespread “disturbance”, was taken into account in most CSAs, additional disturbances that are triggered by extreme weather events may lead to strong effects on forest properties much earlier than the chronic, long-term changes of average climatic conditions that were the focus of the ARANGE work. Large-scale disturbances will very likely be a major driver of changes in forest structure and related ES provisioning. This suggests that considerable emphasis must be placed on including relevant disturbance agents in model-based assessments of forest ES provisioning under climate change.

Notes

Acknowledgements

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 no. 289437.

References

  1. Bagnaresi U, Giannini R, Grassi G, Minotta G, Paffetti D, Pini Prato E et al (2002) Stand structure and biodiversity in mixed, uneven-aged coniferous forests in the eastern Alps. Forestry 75:357–364. doi:10.1093/forestry/75.4.357 CrossRefGoogle Scholar
  2. Berger F (1997) Interaction forêt de montagne-risques naturels. Détermination de Zones d’Interventions Forestières Prioritaires—L’exemple du département de la Savoie, thèse de doctorat, Paris, EngrefGoogle Scholar
  3. Berger F, Dorren L (2007) Principle of the tool Rockfor.net for quantifying the rockfall hazard below a protection forest. Schweiz Z Forstwes 158(6):157–165. doi:10.3188/szf.2007.0157 CrossRefGoogle Scholar
  4. Bugmann H (2001) A review of forest gap models. Clim Change 51:259–305. doi:10.1023/A:1012525626267 CrossRefGoogle Scholar
  5. Cordonnier T, Courbaud B, Berger F, Franc A (2008) Permanence of resilience and protection efficiency in mountain Norway spruce forest stands: A simulation study. For Ecol Manage 256:347–354. doi:10.1016/j.foreco.2008.04.028 CrossRefGoogle Scholar
  6. Cordonnier T, Berger F, Elkin C, Lamas T, Martinez M (2013) Models and linker functions (indicators) for ecosystem services, FP7-289437-ARANGE/D2.2. http://www.arange-project.eu/wp-content/uploads/ARANGE-D2.2_linkerfunctions.pdf
  7. Déqué M, Somot S (2010) Weighted frequency distributions express modelling uncertainties in the ENSEMBLES regional climate experiments. Clim Res 44:195–209. doi:10.3354/cr00866 CrossRefGoogle Scholar
  8. Dobler A, Ahrens B (2008) Precipitation by a regional climate model and bias correction in Europe and South Asia. Meteorol Z 17:499–509. doi:10.1127/0941-2948/2008/0306 CrossRefGoogle Scholar
  9. Duncker PS, Raulund-Rasmussen K, Gundersen P, Katzensteiner K, De Jong J, Ravn HP, Smith M, Eckmüllner O, Spiecker H (2012) How forest management affects ecosystem services, including timber production and economic return: synergies and trade-offs. Ecol Soc 17(4):50. doi:10.5751/ES-05066-170450 Google Scholar
  10. Dupire S, Bourrier F, Monnet J-M, Bigot S, Borgniet L, Berger F, Curt T (2016) Novel quantitative indicators to characterize the protective effect of mountain forests against rockfall. Ecol Indic 67:98–107. doi:10.1016/j.ecolind.2016.02.023 CrossRefGoogle Scholar
  11. EEA (ed) (2010) Europe’s ecological backbone: recognising the true value of our mountains. European Environmental Agency, Technical Report 6/2010. doi:10.2800/43450
  12. Elkin C, Gutierrez AG, 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. Global Change Biol 19:1827–1840. doi:10.1111/gcb.12156 CrossRefGoogle Scholar
  13. Fabrika M, Durský J (2005) Algorithms and software solution of thinning models for SIBYLA growth simulator. J For Sci 51:431–445Google Scholar
  14. Fan ZF, Shifley SR, Spetich MA, Thompson FR, Larsen DR (2003) Distribution of cavity trees in midwestern old-growth and second-growth forests. Can J For Res 33:1481–1494. doi:10.1139/X03-068 CrossRefGoogle Scholar
  15. Finger D, Heinrich G, Gobiet A, Bauder A (2012) Projections of future water resources and their uncertainty in a glacierized catchment in the Swiss Alps and the subsequent effects on hydropower production during the 21st century. Water Resour Res 48:W02521. doi:10.1029/2011WR010733 Google Scholar
  16. Forrester DI, Bauhus J (2016) A review of processes behind diversity—productivity relationships in forests. Curr For Rep 2:45–61. doi:10.1007/s40725-016-0031-2 CrossRefGoogle Scholar
  17. Frehner M, Wasser B, Schwitter R (2005) Nachhaltigkeit und Erfolgskontrolle im Schutzwald. Wegleitung für Pflegemassnahmen in Wäldern mit Schutzfunktion. © OFEV, Bern, SwitzerlandGoogle Scholar
  18. Gao T, Hedblom M, Emilsson T, Nielsen AB (2014) The role of forest stand structure as biodiversity indicator. For Ecol Manag 330:82–93. doi:10.1016/j.foreco.2014.07.007 CrossRefGoogle Scholar
  19. Gauquelin X, Courbaud B (eds) (2006) Guide des sylvicultures de montagne des Alpes du Nord Françaises. French National Forest Office, ParisGoogle Scholar
  20. Ghil M (2002) Natural climate variability. In: Munn RE (ed) Encyclopedia of global environmental change, volume 1, the earth system: physical and chemical dimensions of global environmental change. Wiley, ChichesterGoogle Scholar
  21. Giorgi F, Coppola E (2007) European climate-change oscillation (ECO). Geophys Res Lett 34:L21703. doi:10.1029/2007GL031223 CrossRefGoogle Scholar
  22. Grove SJ (2002) Tree basal area and dead wood as surrogate indicators of saproxylic insect faunal integrity: a case study from the Australian lowland tropics. Ecol Indic 1:171–188. doi:10.1016/S1470-160X(01)00016-4 CrossRefGoogle Scholar
  23. Haylock MR, Hofstra N, Klein-Tank AMG, Klok EJ, Jones PD, New M (2008) A European daily high-resolution gridded dataset of surface temperature and precipitation. JGR (Atmospheres) 113(1–12):D20119. doi:10.1029/2008JD10201 CrossRefGoogle Scholar
  24. Heinrich G, Gobiet A, Mendlik T (2014) Extended regional climate model projections for Europe until the mid-twentyfirst century: combining ENSEMBLES and CMIP3. Clim Dyn 42(1–2):521–535. doi:10.1007/s00382-013-1840-7 CrossRefGoogle Scholar
  25. Hewitt CD, Griggs DJ (2004) Ensembles-based predictions of climate changes and their impacts (Ensembles). EOS Trans AGU 85(52):566. doi:10.1029/2004EO520005 CrossRefGoogle Scholar
  26. IPCC (2006) Guidelines for National Greenhouse Gas Inventories. http://www.ipcc-nggip.iges.or.jp/public/2006gl/vol4.html
  27. Irauschek F, Rammer W, Lexer MJ (2015) Can current management maintain forest landscape multifunctionality in the Eastern Alps in Austria under climate change? Reg Environ Change. doi:10.1007/s10113-015-0908-9
  28. Jost L (2006) Entropy and diversity. Oikos 113:363–375. doi:10.1111/j.2006.0030-1299.14714.x CrossRefGoogle Scholar
  29. Knutti R, Furrer R, Tebaldi C, Cermak J, Meehl GA (2010) Challenges in combining projections from multiple climate models. J. Clim 23:2739–2758. doi:10.1175/2009JCLI3361.1 CrossRefGoogle Scholar
  30. Lachat T, Bütler R (2007) Gestion des vieux arbres et du bois mort: Îlots de sénescence, arbres-habitat et métapopulations saproxyliques. Mandat de l’Office fédéral de l’environnement, OFEV. http://www.wsl.ch/forschung/forschungsunits/walddynamik/diversitaet/totholzmanagement/rapport_bafu_2007.pdf
  31. Lafond V, Cordonnier T, Courbaud B (2015) Reconciling biodiversity conservation and timber production in uneven-aged mountain forests: identification of ecological intensification pathways. Envir Manag 56:1118–1133. doi:10.1007/s00267-015-0557-2 CrossRefGoogle Scholar
  32. Lämås T, Eriksson LO (2003) Analysis and planning systems for multi-resource, sustainable forestry—the Heureka research programme at SLU. Can J For Res 33:500–508. doi:10.1139/cjfr-2016-0068 CrossRefGoogle Scholar
  33. Larrieu L, Cabanettes A (2012) Species, live status, and diameter are important tree features for diversity and abundance of tree microhabitats in subnatural montane beech–fir forests. Can J For Res 42:1433–1445. doi:10.1139/x2012-077 CrossRefGoogle Scholar
  34. Larrieu L, Cabanettes A, Delarue A (2012) Impact of silviculture on dead wood and on the distribution and frequency of tree microhabitats in montane beech-fir forests of the Pyrenees. Eur J For Res 131:773–786. doi:10.1007/s10342-011-0551-z CrossRefGoogle Scholar
  35. Lassauce A, Paillet Y, Jactel H, Bouget C (2011) Deadwood as a surrogate for forest biodiversity: meta-analysis of correlations between deadwood volume and species richness of saproxylic organisms. Ecol Indic 11:1027–1039. doi:10.1016/j.ecolind.2011.02.004 CrossRefGoogle Scholar
  36. Lexer MJ, Hönninger K (2001) A modified 3D-patch model for spatially explicit simulation of vegetation composition in heterogeneous landscapes. For Ecol Manage 144:43–65. doi:10.1016/S0378-1127(00)00386-8 CrossRefGoogle Scholar
  37. Maraun D, Wetterhall F, Ireson AM, Chandler RE, Kendon EJ, Widmann M, Brienen S, Rust HW, Sauter T, Themessl M, Venema VKC, Chun KP, Goodess CM, Jones RG, Onof C, Vrac M, Thiele-Eich I (2010) Precipitation downscaling under climate change: recent developments to bridge the gap between dynamical models and the end user. Rev Geophys 48:3003. doi:10.1029/2009RG000314 CrossRefGoogle Scholar
  38. Martikainen P, Siitonen J, Punttila P, Kaila L, Rauh J (2000) Species richness of Coleoptera in mature managed and old-growth boreal forests in southern Finland. Biol Conserv 94:199–209. doi:10.1016/S0006-3207(99)00175-5 CrossRefGoogle Scholar
  39. Maurer EP, Pierce DW (2014) Bias correction can modify climate model simulated precipitation changes without adverse effect on the ensemble mean. Hydrol Earth Syst Sci 18(3):915–925. doi:10.5194/hess-18-915-2014 CrossRefGoogle Scholar
  40. McElhinny C, Gibbons P, Brack C, Bauhus J (2005) Forest and woodland stand structural complexity: its definition and measurement. For Ecol Manag 218:1–24. doi:10.1016/j.foreco.2005.08.034 CrossRefGoogle Scholar
  41. Meehl GA, Covey C, Delworth T, Latif M, McAvaney B, Mitchell JFB, Stouffer RJ, Taylor KE (2007) The WCRP CMIP3 multimodel dataset: A new era in climate change research. Bull Am Meteorol Soc 88(9):1383–1394. doi:10.1175/BAMS-88-9-1383 CrossRefGoogle Scholar
  42. Michel AK, Winter S (2009) Tree microhabitat structures as indicators of biodiversity in Douglas-fir forests of different stand ages and management histories in the Pacific Northwest, USA. For Ecol Manag 257:1453–1464. doi:10.1016/j.foreco.2008.11.027 CrossRefGoogle Scholar
  43. Mina M, Bugmann H, Klopcic M et al (2015) Accurate modeling of harvesting is key for projecting future forest dynamics: a case study in the Slovenian mountains. Reg Environ Change. doi:10.1007/s10113-015-0902-2
  44. Müller J, Butler R (2010) A review of habitat thresholds for dead wood: a baseline for management recommendations in European forests. Eur J For Res 129:981–992. doi:10.1007/s10342-010-0400-5 CrossRefGoogle Scholar
  45. Müller J, Bussler H, Kneib T (2008) Saproxylic beetle assemblages related to silvicultural management intensity and stand structures in a beech forest in Southern Germany. J Insect Conserv 12:107–124. doi:10.1007/s10841-006-9065-2 CrossRefGoogle Scholar
  46. Nabuurs G-J et al (2003) LUCF sector good practice guidance, Chap 3. In: Penman J et al (eds.), Good practice guidance for land use, land use change and forestry. Special Report of the IPCC, WMO, Geneva. http://www.ipcc-nggip.iges.or.jp/public/gpglulucf/gpglulucf_files/GPG_LULUCF_FULL.pdf
  47. Nadrowski K, Wirth C, Scherer-Lorenzen M (2010) Is forest diversity driving ecosystem function and service? Curr Opin Environ Sust 2:75–79. doi:10.1016/j.cosust.2010.02.003 CrossRefGoogle Scholar
  48. Nakicenovic N, Alcamo J, Davis G, de Vries B, Fenhann J, Gaffin S, Gregory K, Grübler A, Jung TY, Kram T, La Rovere EL, Michaelis L, Mori S, Morita T, Pepper W, Pitcher H, Price L, Raihi K, Roehrl A, Rogner H-H, Sankovski A, Schlesinger M, Shukla P, Smith S, Swart R, van Rooijen S, Victor N, Dadi Z (2000) Special Report on Emissions Scenarios: A Special Report of Working Group III of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge. www.grida.no/climate/ipcc/emission
  49. Nilsson SG, Niklasson M, Hedin J et al (2002) Densities of large living and dead trees in old-growth temperate and boreal forests. For Ecol Manag 161:189–204. doi:10.1016/S0378-1127(01)00480-7 CrossRefGoogle Scholar
  50. Pardos M, Pérez S, Calama R et al (2016) Ecosystem service provision, management systems and climate change in Valsaín forest, central Spain. Reg Environ Change. doi:10.1007/s10113-016-0985-4
  51. Pepin N, Bradley RS, Diaz HF, Baraer M, Caceres EB, Forsythe N, Fowler H, Greenwood G, Hashmi MZ, Liu XD, Miller JR, Ning L, Ohmura A, Palazzi E, Rangwala I, Schöner W, Severskiy I, Shahgedanova M, Wang MB, Williamson SN, Yang DQ (2015) Elevation-dependent warming in mountain regions of the world. Nat Clim Change 5:424–430. doi:10.1038/nclimate2563 CrossRefGoogle Scholar
  52. Piani C, Haerter JO, Coppola E (2010) Statistical bias correction for daily precipitation in regional climate models over Europe. Theoret Appl Climatol 99(1–2):187–192. doi:10.1007/s00704-009-0134-9 CrossRefGoogle Scholar
  53. Pietsch SA, Hasenauer H, Thornton PE (2005) BGC-model parameters for tree species growing in central European forests. For Ecol Manage 211:264–295. doi:10.1016/j.foreco.2005.02.046 CrossRefGoogle Scholar
  54. Pommerening A (2002) Approaches to quantify forest structure. Forestry 75:305–324. doi:10.1093/forestry/75.3.305 CrossRefGoogle Scholar
  55. Prein AF, Gobiet A, Truhetz H (2011) Analysis of uncertainty in large scale climate change projections over Europe. Meteorol Z 20(4):383–395. doi:10.1127/0941-2948/2011/0286 CrossRefGoogle Scholar
  56. Pretzsch H, Rais A (2016) Wood quality in complex forests versus even-aged monocultures: review and perspectives. Wood Sci Technol 50:845–880. doi:10.1007/s00226-016-0827-z CrossRefGoogle Scholar
  57. Price MF, Lysenko I, Gloersen E (2004) Delineating Europe’s mountains. Rev Géogr Alp 92:75–86. doi:10.3406/rga.2004.2293 CrossRefGoogle Scholar
  58. Price MF, Gratzer G, Duguma LA, Thomas K, Maselli D, Romeo R (eds) (2011) Mountain forests in a changing world—realizing values, addressing challenges. FAO/MPS, RomeGoogle Scholar
  59. Racsko P, Szeidl L, Semenov M (1991) A serial approach to local stochastic weather models. Ecol Model 57(1–2):27–41. doi:10.1016/0304-3800(91)90053-4 CrossRefGoogle Scholar
  60. Ravazzani G, Ghilardi M, Mendlik T, Gobiet A, Corbari C, Mancini M (2014) Investigation of climate change impact on water resources for an alpine basin in Northern Italy: implications for evapotranspiration modeling complexity. PLoS ONE 9(10):e109053. doi:10.1371/journal.pone.0109053 CrossRefGoogle Scholar
  61. Redon M, Luque S, Gosselin F, Cordonnier T (2014) Is generalisation of uneven-aged management in mountain forests the key to improve biodiversity conservation within forest landscape mosaics? Ann For Sci 71:751–760. doi:10.1007/s13595-014-0371-7 CrossRefGoogle Scholar
  62. Rouvinen S, Kuuluvainen T (2005) Tree diameter distributions in natural and managed old Pinus sylvestris-dominated forests. For Ecol Manag 208:45–61. doi:10.1016/j.foreco.2004.11.021 CrossRefGoogle Scholar
  63. Running SW, Nemani RR, Hungerford RD (1987) Extrapolation of synoptic meteorological data in mountainous terrain and its use for simulating forest evapotranspiration and photosynthesis. Can J For Res 17:472–483. doi:10.1139/x87-081 CrossRefGoogle Scholar
  64. Schreiber B, deCalesta DS (1992) The relationship between cavity-nesting birds and snags on clearcuts in western Oregon. For Ecol Manag. 50:299–316. doi:10.1016/0378-1127(92)90344-9 CrossRefGoogle Scholar
  65. Schumacher S, Bugmann H (2006) The relative importance of climatic effects, wildfires and management for future forest landscape dynamics in the Swiss Alps. Global Change Biol 12:1435–1450. doi:10.1111/J.1365-2486.2006.01188.X CrossRefGoogle Scholar
  66. Schwenk WS, Donovan TM, Keeton WS, Nunery JS (2012) Carbon storage, timber production, and biodiversity: comparing ecosystem services with multi-criteria decision analysis. Ecol Appl 22:1612–1627. doi:10.1890/11-0864.1 CrossRefGoogle Scholar
  67. Seidl R, Rammer W, Jäger D, Currie WS, Lexer M (2007) Assessing trade-offs between carbon sequestration and timber production within a framework of multi-purpose forestry in Austria. For Ecol Manag 248:64–79. doi:10.1016/j.foreco.2007.02.035 CrossRefGoogle Scholar
  68. Semenov MA, Barrow EM (1997) Use of a stochastic weather generator in the development of climate change scenarios. Clim Change 35(4):397–414. doi:10.1023/A:1005342632279 CrossRefGoogle Scholar
  69. Smith PC, Heinrich G, Suklitsch M, Gobiet A, Stoffel M, Fuhrer J (2014) Station-scale bias correction and uncertainty analysis for the estimation of irrigation water requirements in the Swiss Rhone catchment under climate change. Clim Change 127(3–4):521–534. doi:10.1007/s10584-014-1263-4 CrossRefGoogle Scholar
  70. Stainforth DA, Allen MR, Tredger ER, Smith LA (2007) Confidence, uncertainty and decision-support relevance in climate predictions. Philos Trans R Soc A 365:2145–2161. doi:10.1098/rsta.2007.2074 CrossRefGoogle Scholar
  71. Staudhammer CL, LeMay VM (2001) Introduction and evaluation of possible indices of stand structural diversity. Can J For Res 31:1105–1115. doi:10.1139/cjfr-31-7-1105 CrossRefGoogle Scholar
  72. Stoffel M, Mendlik T, Schneuwly-Bollschweiler M, Gobiet A (2014) Possible impacts of climate change on debris-flow activity in the Swiss Alps. Clim Change 122(1–2):141–155. doi:10.1007/s10584-013-0993-z CrossRefGoogle Scholar
  73. Themeßl MJ, Gobiet A, Leuprecht A (2011) Empirical-statistical downscaling and error correction of daily precipitation from regional climate models. Int J Climatol 31(10):1530–1544. doi:10.1002/joc.2168 CrossRefGoogle Scholar
  74. Thornton PE, Running SW (1999) An improved algorithm for estimating incident daily solar radiation from measurements of temperature, humidity, and precipitation. Agric For Meteorol 93(4):211–228. doi:10.1016/S0168-1923(98)00126-9 CrossRefGoogle Scholar
  75. Thornton PE, Hasenauer H, White MA (2000) Simultaneous estimation of daily solar radiation and humidity from observed temperature and precipitation: an application over complex terrain in Austria. Agric For Meteorol 104(4):255–271. doi:10.1016/S0168-1923(00)00170-2 CrossRefGoogle Scholar
  76. Thurnher, C., 2013. ARANGE Deliverable D1.1Historic Climate, ARANGE Document FP7-289437-ARANGE/D1.1, 13 pp. Institute for Silviculture and Forest Engineering, University of Life Sciences Vienna, Vienna, Austria. http://www.arange-project.eu/?page_id=601
  77. Hlásny, T., Barka, I., Kulla, L. et al (2015) Sustainable forest management in a mountain region in the Central Western Carpathians, northeastern Slovakia: the role of climate change. Reg Environ Change. doi:10.1007/s10113-015-0894-y
  78. Truhetz H (2013) ARANGE deliverable D1.4—climate change scenarios for case study regions, ARANGE document FP7-289437-ARANGE/D1.4. Wegener Center for Climate and Global Change, University of Graz, Graz, Austria. http://www.arange-project.eu/?page_id=601
  79. Vallet P, Dhôte J-F, Moguédec GL, Ravart M, Pignard G (2006) Development of total aboveground volume equations for seven important forest tree species in France. For Ecol Manag 229:98–110. doi:10.1016/j.foreco.2006.03.013 CrossRefGoogle Scholar
  80. van der Linden P, Mitchell JFB (eds) (2009) ENSEMBLES: climate change and its impacts: summary of research and results from the ENSEMBLES project. Met Office Hadley Centre, ExeterGoogle Scholar
  81. Vuidot A, Paillet Y, Archaux F, Gosselin F (2011) Influence of the tree characteristics and forest management on tree microhabitats. Biol Conserv 144:441–450. doi:10.1016/j.biocon.2010.09.030 CrossRefGoogle Scholar
  82. Wang W, Lei X, Ma Z, Kneeshaw DD, Peng C (2011) Positive relationship between aboveground carbon stocks and structural diversity in spruce-dominated forest stands in New Brunswick, Canada. For Sci 57:506–515Google Scholar
  83. Wilcke RAI, Mendlik T, Gobiet A (2013) Multi-variable error correction of regional climate models. Clim Change 120(4):871–887. doi:10.1007/s10584-013-0845-x CrossRefGoogle Scholar
  84. Winter S, Möller GC (2008) Microhabitats in lowland beech forests as monitoring tool for nature conservation. For Ecol Manag 255:1251–1261. doi:10.1016/j.foreco.2007.10.029 CrossRefGoogle Scholar
  85. Zilliox C, Gosselin F (2014) Tree species diversity and abundance as indicators of understory diversity in French mountain forests: variations of the relationship in geographical and ecological space. For Ecol Manag 321:105–116. doi:10.1016/j.foreco.2013.07.049 CrossRefGoogle Scholar
  86. Zlatanov T, Elkin C, Irauschek F, Lexer MJ (2015) Impact of climate change on vulnerability of forests and ecosystem service supply in Western Rhodopes Mountains. Reg Environ Change. doi:10.1007/s10113-015-0869-z

Copyright information

© Springer-Verlag Berlin Heidelberg 2016

Authors and Affiliations

  1. 1.Forest Ecology, Department of Environmental Systems ScienceETH ZürichZurichSwitzerland
  2. 2.Irstea, UR EMGRUniversité Grenoble AlpesSt-Martin-d’HèresFrance
  3. 3.Wegener Center for Climate and Global ChangeUniversity of GrazGrazAustria
  4. 4.Institute of SilvicultureUniversity of Natural Resources and Life SciencesViennaAustria

Personalised recommendations