Trade-offs and synergies among ecosystem services under different forest management scenarios – The LEcA tool

Ecosystem Services 28 (2017) 67–79

Contents lists available at ScienceDirect

Ecosystem Services journal homepage: www.elsevier.com/locate/ecoser

Trade-offs and synergies among ecosystem services under different forest management scenarios – The LEcA tool Xi Pang a,c,⇑, Eva-Maria Nordström b,c, Hannes Böttcher c,d, Renats Trubins e, Ulla Mörtberg a a

KTH Royal Institute of Technology, Dept. of Sustainable Development, Environmental Science and Engineering, Stockholm, Sweden Swedish University of Agricultural Sciences, Department of Forest Resource Management, Umeå, Sweden c International Institute for Applied Systems Analysis (IIASA), Ecosystems Services and Management Program, Laxenburg, Austria d Öko-Institut e.V., Institute for Applied Ecology, Berlin, Germany e Swedish University of Agricultural Sciences, Southern Swedish Forest Research Centre, Alnarp, Sweden b

a r t i c l e

i n f o

Article history: Received 10 April 2017 Received in revised form 30 August 2017 Accepted 6 October 2017

Keywords: Integrated sustainability assessment Industrial wood Bioenergy Carbon storage Recreation values Habitat networks

a b s t r a c t Forests provide a multitude of ecosystem services. In Sweden, the goal to replace fossil fuels could induce substantial changes in the current management and use of forests. Therefore, methods and tools are needed to assess synergies and trade-offs between ecosystem services for policy and planning alternatives. The aim of this study was to develop methods for integrated sustainability assessment of forest management strategies for long-term provisioning of various ecosystem services. For this purpose, the Landscape simulation and Ecological Assessment (LEcA) tool was developed to analyse synergies and trade-offs among five ecosystem services: bioenergy feedstock and industrial wood production, forest carbon storage, recreation areas and habitat networks. Forest growth and management were simulated for two scenarios; the EAF-tot scenario dominated by even-aged forestry (EAF), and the CCF-int scenario with a combination of continuous-cover forestry (CCF) and intensified EAF. The results showed trade-offs between industrial wood and bioenergy production on one side and habitat, recreation and carbon storage on the other side. The LEcA tool showed great potential for evaluation of impacts of alternative policies for land zoning and forest management on forest ecosystem services. It can be used to assess the consequences of forest management strategies related to renewable energy and conservation policies. Ó 2017 Elsevier B.V. All rights reserved.

1. Introduction Forests play an important role for climate change mitigation by providing bioenergy feedstock to substitute fossil fuels, as well as carbon storage to counteract greenhouse gas emissions. At the same time, they are also important for other ecosystem services and biodiversity. To limit the increase in temperature to well below 2 °C according to the Paris Agreement (UNFCCC, 2015), emissions of greenhouse gases worldwide need to be halved by 2050 and to be close to zero by 2100 (IPCC, 2014). Sweden is a country with relatively good preconditions for both switching to renewable energy sources (water, wind and forest biomass) and for climate change mitigation through carbon sequestration in forests and substitution of fossil-based materials by forest products. In Sweden, the Parliament has declared that the vehicle fleet

⇑ Corresponding author at: KTH Royal Institute of Technology, Dept. of Sustainable Development, Environmental Science and Engineering, Stockholm, Sweden. E-mail address: [email protected] (X. Pang). https://doi.org/10.1016/j.ecoser.2017.10.006 2212-0416/Ó 2017 Elsevier B.V. All rights reserved.

should be independent of fossil fuels by 2030 (IEA, 2014) and adopted a vision for Sweden of zero net emissions of greenhouse gases by 2050 (Ministry of the Environment, 2014). By the year 2013, around 34% of the final energy consumption in Sweden still depended on fossil fuels, which would equal to 131 TWh (SEA, 2015), so, to fulfil the goal, the same amount of renewable energy would be needed to replace the fossil fuels. A similar share (34%) of the domestic energy consumption came from bioenergy in 2013 (SEA, 2015). Other sources that can be seen as carbon neutral and that also have a large share of the energy generation in Sweden are hydropower and nuclear power (SEA, 2015). However, the opportunities for expanding these sources are limited (SEPA, 2009; SOU, 2014), and the latter is even planned to be phased out (SEPA, 2011), which are factors that together with the climate-related goals may lead to an increased demand for forest bioenergy feedstock in Sweden (e.g., Börjesson et al., 2017). This could induce substantial changes in the current management and use of forests. According to the Swedish environmental quality objective ‘‘Reduced climate impact”, climate-related goals should be

68

X. Pang et al. / Ecosystem Services 28 (2017) 67–79

achieved without jeopardizing other goals of sustainable development, such as, biological diversity should be preserved and food production should be assured (Govt. prop. 1997/98:145 and 2004/05:150; SEPA, 2012b). According to another Swedish environmental quality objective, ‘‘Sustainable forests”, the value of forests for biological production must be protected, at the same time as biodiversity, cultural heritage and recreational values are safeguarded (Govt. prop. 1997/98:145 and 2004/05:150). However, according to SEPA (2015), it will not be possible to achieve these objectives by 2020 under current development and policies. On EU level, to combat climate change, the EU Renewable Energy Directive (2009/28/EC) aims to promote renewable energy sources and to reduce greenhouse gas emissions in a sustainable way, ensuring that negative effects on ecosystem services and biodiversity are avoided (EC, 2001, 2010). Thus, it is essential to integrate multiple ecosystem services and biodiversity in assessments of policies and plans for increasing use of forest biomass as a renewable energy source. Today, forest management in Sweden is to a large extent focused on production of industrial wood (sawlogs and pulpwood). Biomass for bioenergy is extracted as harvest residues, mainly tops and branches, and is consequently a by-product of industrial wood production at present. A large share of forest bioenergy also comes from industrial by-products such as black liquor, wood chips, sawdust and bark. In current forest practice, only a part of the technical potential of harvest residues is harvested in Sweden. For instance comparing actually extracted amounts 2015 (SEA, 2016) with the estimated potential 2010–2019 (SFA, 2015) would imply that around 24% is used, and stumps are only harvested on experimental scale (Melin, 2014). To begin with, an increased demand for bioenergy would probably increase the extraction of harvest residues, possibly also stumps. At higher levels of demand, the bioenergy sector might compete with forest industries for raw material, with the wood board industry as well as with pulp and paper producers (Carlsson, 2012; Jonsson, 2012, 2013; Lauri et al., 2014). The sawn wood industry may also be directly affected by competition from bioenergy demand under higher prices for energy wood, and in a transition to green economy where wood is a renewable material used on large scale to substitute, e.g., fossil-based products (EC, 2012), the overall demand for wood can be foreseen to increase. Industrial forestry has been identified as a major cause of depletion of forest biodiversity, mainly due to the simplification of forest structure and the loss of old trees and dead wood (Berg et al., 1994; EEA, 2010; Puettmann et al., 2009; Thompson et al., 2011), even if plans and actions are carried out to protect biodiversity (Eriksson et al., 2015). An increasing demand for wood may call for new forest management practices to increase the supply in Sweden (Larsson et al., 2009; Lidskog et al., 2013). Intensified forestry could increase the biomass production through planting of monocultures of native or introduced tree species, forest fertilization and application of shorter rotation times. Intensified forestry resulting from increasing demand for industrial wood and bioenergy feedstock can be expected to have negative impacts on biodiversity by reducing habitat size and connectivity in forest landscapes (Hanski, 2011; Larsson et al., 2009; Ranius and Roberge, 2011). Further impacts may result from increased extraction of forest residues for bioenergy (e.g., de Jong and Dahlberg, 2017; Hedin et al., 2008). Thus, trade-offs between biodiversity and forest biomass production will be a major challenge for energy and climate policies. Forests provide a multitude of ecosystem services, beside industrial wood, bioenergy and habitat supporting biodiversity, such as cultural ecosystem services including recreation, aesthetics and cultural heritage (Fredman and Tyrväinen, 2010; Milligan and Bingley, 2007; Sonntag-Öström et al., 2014). Forests also play an

important role in carbon storage for mitigating climate change (Canadell and Raupach, 2008; Pan et al., 2011). Forests are responsible for almost half of the total terrestrial photosynthesis, and improved carbon-focused forest management has been shown to almost always result in net carbon sequestration (Malhi et al., 2002). The supply of ecosystem services and the balance between them will depend on forest management strategy on both stand and landscape scales. However, there are trade-offs to be made between all these ecosystem services since it may not be possible to increase the supply of one ecosystem service without affecting some other ecosystem service negatively. Even if such trade-offs are seldom analysed in energy assessments (Pang et al., 2014), assessment of ecosystem services is currently a rapidly growing area of research. Depending on the ecosystem service in focus and the geographical scale, different models and techniques have been used. Many assessment initiatives have been large scale, e.g., global (the Millennium Ecosystem Assessment; MEA, 2005) or European (the RUBICODE project; Vandewalle et al., 2009), and may provide important information for policy and decision making on international level, but there is a need for studies that can support decision making on national, regional and local level (Burkhard et al., 2010). Although the research on forest ecosystem services have kept growing, the trade-offs between services are still poorly understood (Filyushkina et al., 2016). Currently, most trade-off analysis on forest ecosystem services are focused on comparison of two ecosystem services, such as the conflict between bioenergy extraction and carbon storage (Bottalico et al., 2016; Hoel and Sletten, 2016). Many studies provide biophysical mapping of ecosystem services, i.e., descriptions of the present state, but with no projections of possible future trends or with projections based on historical trends or simplified assumptions on future development. In a few studies, development of multiple ecosystem services and trade-offs among them have been projected over time (Forsius et al., 2015; Verkerk et al., 2014). However, some ecosystem services have a spatial component and have to be considered in a landscape context, such as the spatial distribution of habitat for species and of recreation areas for people. Most policy assessments of ecosystem services conducted so far have not included such spatial aspects in long-term projections. A critical issue is thus to develop models that enable projections of the development of different ecosystem services on landscape level as a function of the forest management. The aim of this study was to develop methods for integrated sustainability assessment of alternative forest management policies, for long-term provisioning of various ecosystem services, considering climate and other environmental and societal goals. Two scenarios based on different land-zoning policies with related forest management strategies were simulated for Kronoberg County, a study area in southern Sweden. This was done using the recently developed LandSim model which is a spatially explicit model for long term projection of forest development (Pang et al., 2017). Building on previous studies, this paper connects existing models for projection of industrial wood production, bioenergy feedstock and carbon storage with spatially explicit methods for recreation area assessment and habitat network assessment. In this way it was possible to analyse trade-offs and synergies among five ecosystem services: provision of industrial wood and bioenergy feedstock, forest carbon storage, recreation areas and habitat networks for selected focal species. We integrated methods and models in a Landscape simulation and Ecological Assessment (LEcA) tool in order to project the corresponding changes of the ecosystem services under alternative forest management scenarios. The LEcA tool thus aims to provide decision support to stakeholders for integrated sustainability assessment of policy and planning alternatives.

X. Pang et al. / Ecosystem Services 28 (2017) 67–79

2. Materials and methods 2.1. Study area The study area embraced Kronoberg County, with an area of 9426 km2 and located in the southern part of Sweden (Fig. 1). Kronoberg County is situated in the hemi-boreal zone and forest covers 66% of the total area and 78% of the land area (County Administrative Boards, CAB, 2006; Statistics Sweden, 2013). The forest is dominated by conifers, mainly Norway spruce (Picea abies) and Scots pine (Pinus sylvestris), often mixed with deciduous trees. Southern broadleaves, mainly oak (Quercus robur) and beech (Fagus sylvatica), can be found around larger lakes and scattered in coniferous forest stands (Swedish Forest Agency, SFA, 2014). The increasingly intensive forest management as well as lack of protected areas have been seen as problematic for reaching regional environmental quality objectives in the County, while old forest, in particular deciduous, is of main importance for biodiversity conservation (CAB, 2006; SEPA, 2014). 2.2. Data Spatial datasets that were used in the study were; forest data on timber volume by dominating tree species, age, and height (SLU, 2013a), which is based on remote sensing data (Landsat TM and Spot) together with field data from the Swedish National Forest Inventory (Reese et al., 2003); land cover data (Lantmäteriet, 2013a); topographic data (Lantmäteriet, 2013b); data on protected areas and areas of national interest for nature conservation, cultural and recreational values (SEPA, 2006; SNBH, 2013), retrieved from CAB (2013); as well as data on areas of high biodiversity value, defined as part of a national green infrastructure but not necessarily protected (SEPA, 2012a). 2.3. Scenarios Two scenarios were defined to provide a basis for the assessment of trade-offs among the ecosystem services. These scenarios represented two different land-zoning policies with related forest management strategies (Pang et al., 2017), which would create and maintain different forest structures on stand and landscape level. Thus they were assumed to affect the potential for industrial

69

wood production, bioenergy feedstock extraction, carbon storage, recreation values, as well as the habitat networks in different ways. The first scenario was the ‘‘EAF-tot” scenario, with even-aged forestry (EAF) management in all (97%) of the productive forest land that was not formally protected (3%) in the County. The related land-zoning policy was to use all forest that is not protected for forest biomass extraction. EAF means that the forest is divided into management units or stands based on the properties of the forest and the site. Each stand is managed according to a cycle of final felling, regeneration, cleaning, thinning, and again final felling and so on. EAF thus results in landscapes where the forest within each stand is relatively homogeneous in terms of age, dimensions and tree species, and the landscape develops into a patchwork of stands with forest in different stages. In Sweden EAF is carried out as retention forestry, which means that some trees, groups of trees and dead wood are retained at final felling (Simonsson et al., 2014). However, in the EAF-tot scenario in this study no retention was applied and trees were harvested from around the age of 70 years. The second scenario was the ‘‘CCF-int” scenario, which involved no forestry in the formally protected areas (3% of the productive forest land), and continuous-cover forestry (CCF) on 27% of the forest land (Fig. 2). CCF management was applied in areas with a weaker protection status, i.e., in areas of national interest for nature conservation, cultural and recreational values (SEPA, 2006; SNBH, 2013), outside protected areas; as well as in areas of high biodiversity value, defined as part of a green infrastructure on national level (SEPA, 2012a), outside protected areas. The related land-zoning policy is to dedicate 30% of the forest landscape to multiple ecosystem services and biodiversity, while 70% to more intensive extraction of forest biomass. CCF involved a selective harvesting method, including bioenergy feedstock extraction but without final felling (Pommerening and Murphy, 2004). The rest of the landscape, 70% of the productive forest area, was managed with intensified EAF, applying 10 years shortened rotation time and including extraction of bioenergy feedstock. 2.4. The LEcA approach for the ecosystem services assessment The LEcA tool comprises of three modules; the landscape simulator LandSim (Pang et al., 2017), a storage and yield calculator and a habitat assessment tool, see Fig. 3. LandSim was used for simulat-

Fig. 1. The study area Kronoberg County in southern Sweden. Spatial data Ó Lantmäteriet i2012/920, coordinate system Sweref 99 TM.

70

X. Pang et al. / Ecosystem Services 28 (2017) 67–79

Fig. 2. Area under CCF management and intensified EAF management within the CCF-int scenario. Spatial data Ó Lantmäteriet i2012/920, coordinate system Sweref 99 TM.

ing forest growth and management under the two different land zoning scenarios (EAF-tot and CCF-int), and their impacts on five different ecosystem services were assessed. These can be grouped following the scheme of e.g., Balmford et al. (2008), as:
Provisioning services: o Industrial wood production, o Bioenergy feedstock production,
Regulating services: carbon storage,
Cultural services: recreational value, and
Habitat services: habitat networks for two focal species representing coniferous forest and southern broadleaved forest, respectively. The consequences for the ecosystem services of the different scenarios were then mapped, quantified and compared (see Subsections 2.4.2 and 2.4.3). Trade-offs among different ecosystem services in the two scenarios could then be compared and assessed. Outside LandSim, all spatial modelling and visualisation were performed in ArcGIS 10.2 (ESRI, 2009). 2.4.1. Simulation of forest growth and management LandSim is designed for projecting the state of a landscape under a specified management strategy (Pang et al., 2017). Input data for the simulations was forest data on tree species, volume, age and height (Reese et al., 2003). The landscape was represented by 25 ⁄ 25 m pixels and the state of each pixel was described by four different variables, each divided into a number of classes: 1. 2. 3. 4.

Tree species (8 classes), Site index (4 classes), Age (33 classes), Volume (10 classes).

Each pixel was initially assigned to a class for each of the four variables. To simulate the future forest development, the classes for each pixel were projected by using the probability of transition between classes for the pixels based on data from the Swedish

National Forest Inventory (Fridman et al., 2014). The projection of the future states of the pixels was directed by the forest management treatments specified for each individual pixel, i.e., final felling, thinning or no action. Starting from the current situation, the simulations were carried out in 5 year time steps for 100 years and the output from LandSim worked as input data for the assessment of each ecosystem service. 2.4.2. Storage and yield estimation Industrial wood production (sawlogs and pulpwood) in each time step was estimated by using the output from LandSim, which included simulated data on stem volume and information of activities such as final felling or thinning for each pixel. The industrial wood production was estimated using the volume reduction between two periods of time for pixels with final felling or thinning, i.e., the harvest. To estimate the bioenergy feedstock production, we used biomass expansion factors applied to the stem volumes to get the dry weight of the tree component i (tops, branches, foliage and bark), which were then summed up. Eq. (1) was fitted using linear regression and was applied to input data on stand age and tree species (Lehtonen et al., 2004; Pang et al., 2017):

W i ¼ ðai þ bi e0:01t ÞV

ð1Þ

where Wi was the dry weight of tree component i. The parameters ai and bi are shown in Table 1. The variable e0.001t is a timedependent term (related to tree age) and V is the stem volume. These biomass expansion factors were developed using data from pine and spruce dominated stands in Finland (Lehtonen et al., 2004) and were considered to be useful for describing also Swedish conditions. In order to estimate the function of the forest as a carbon sink, the biomass expansion factors were used to estimate the carbon content of a whole tree using stem volume as input data (Penman et al., 2003). Lehtonen et al. (2004) provided biomass expansion factors that are dependent on stand age and tree species as a basis for this quantification. In the current study, we estimated whole tree carbon stocks, C, from stem volume, V, multiplied by the

71

X. Pang et al. / Ecosystem Services 28 (2017) 67–79 Table 1 Parameters for dry weight calculation (Lehtonen et al., 2004). Tree compartment (i)

Pine stands

Foliage Branches Branches (dead) Bark

Spruce stands

Broadleaved stands

a

b

a

b

a

b

0.0177 0.0706 0.0104 0.0254

0.0499 0.0212 0.0059 0.0221

0.0388 0.0905 0.0088 0.0353

0.0849 0.0719 0.0040 0.0125

0.1011 0.0053 0.0588

-0.0180 0.0082 0.0105

Table 2 Biomass expansion factors (Lehtonen et al., 2004). Age

Biomass expansion factors

10–19 20–29 30–39 40–49 50–59 60–69 70–79 80–89 90–99 100–119 120–139 140–

Pine

Spruce

Mixed conifers

Mixed conifers and broadleaves

Broadleaves

0.697 0.705 0.710 0.702 0.701 0.710 0.708 0.707 0.704 0.703 0.698 0.690

0.862 0.860 0.841 0.820 0.816 0.791 0.784 0.777 0.782 0.784 0.782 0.788

0.780 0.783 0.776 0.761 0.759 0.751 0.746 0.742 0.743 0.744 0.740 0.739

0.744 0.750 0.748 0.742 0.738 0.736 0.727 0.725 0.725 0.726 0.724 0.723

0.707 0.716 0.720 0.723 0.718 0.720 0.709 0.709 0.707 0.707 0.707 0.707

biomass expansion factors (Table 2) at age i for species j, and used the IPCC default carbon conversion factor 0.5 (Penman et al., 2003, Eq. (2)):

C¼

n X n X 0:5  BEFi;j  V

ð2Þ

i¼1 j¼1

2.4.3. Habitat assessment for estimation of space-specific ecosystem services The total potential recreation area was used as an indicator to estimate the supply of recreational ecosystem services and its changes with time, quantitatively and spatially. According to several studies, mature forest seems to be linked to high recreation values, both to coniferous forest (Sonntag-Öström et al., 2014) and broadleaves (Mattsson and Li, 1994; Norman et al., 2010). In addition, a survey by Hörnsten and Fredman (2000) showed that when the distance to forest suitable for recreation was longer than 2 km most people would not go by foot but would take the car instead, which may decrease the probability for everyday recreation taking place. In the current study, forest more than 70 years old and within 2000 m from housing areas as well as accessible within 300 m distance along each side of small roads were viewed as valuable recreation areas. Land cover data on residential areas (Lantmäteriet, 2013b) and the road system (Lantmäteriet, 2013a), together with data on tree age from Reese et al. (2003) and from the output of LandSim, were used to find potential recreation areas, A(k), summarized as the total recreation area, A(total) (Eq. (3)).

AðtotalÞ ¼

n X AðkÞ

ð3Þ

k¼0

Changes in the forest structure due to the management strategies of the scenarios would result in changes in the total size of different habitat types across the study area. However, too small and isolated habitat patches may not support species persistence in a landscape, but to assess available habitat, both habitat size and connectivity of habitat networks need to be taken into account (Hanski, 2011; Saura et al., 2010, 2011). The habitat network assessment was applied to two selected focal species, representing

targets for habitat of high biodiversity value in the County. The three-toed woodpecker (Picoides tridactylus) represented targets for mature and old spruce-dominated coniferous forest and the middle spotted woodpecker (Dendrocopos medius) represented targets for mature and old southern broadleaved forest. Both these species are resident with large area requirements, specialized habitat demands and are vulnerable to landscape changes, which are properties that make good indicators in north European forests (Schmiegelow and Mönkkönen, 2002; Roberge and Angelstam, 2006). By targeting habitat networks of focal species with such indicative values, important biodiversity components can be assessed on landscape level (Mörtberg et al., 2012). The habitat network assessment involved a definition of habitat demands (habitat quality, quantity and connectivity) for the threetoed woodpecker (Bütler et al., 2004; D’Eon et al., 2002; Leonard, 2001) and the middle spotted woodpecker (Müller et al., 2009; Pasinelli, 2000; Pettersson, 1984). Areas with suitable habitat types were then extracted based on forest data (Reese et al., 2003) as well as from the output of LandSim, and information on home range and median movement distances were used to derive habitat patches and connectivity parameters (see Table 3). When the habitat patches were found for each of the focal species, the Equivalent Connected Area (ECA) was used for measuring changes in available habitat within the habitat networks. ECA is a probability of connectivity index that relates the connectivity changes to the amount of available habitat. It is defined as the size of a single habitat patch (maximally connected) that would provide the same value of the probability of connectivity as the actual habitat pattern in the landscape (Saura et al., 2010; Saura and Rubio, 2010). It was calculated following Eq. (4):

vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi uX n u n X ECA ¼ t ai aj pij

ð4Þ

i¼1 j¼1

where ai and aj is the area of habitat patches i and j, and pij is the maximum product probability of all the possible paths between patches i and j. ECA has an area unit which makes it easy to interpret, as changes in available habitat area within the habitat net-

72

X. Pang et al. / Ecosystem Services 28 (2017) 67–79

Table 3 Habitat demands of the focal species, for coniferous forest (three-toed woodpecker) and southern broadleaved forest (middle spotted woodpecker). The habitat suitability can take the values 0: not suitable; or 1: suitable. Habitat for focal species

Con.

S. BL.

Home range (hectares) Median movement distance (km) Habitat suitability value

36 1.27 0 1 0 1 1 0 0 0 0

12.5 0.75 0 1 0 0 0 1 0 1 1

Forest age Dominating tree species

works. The habitat networks were spatially modelled across the study area for the baseline situation and for different time steps in the two scenarios. The outputs were maps of habitat networks for the focal species together with related ECAs. 3. Results 3.1. Assessment of ecosystem services The development in terms of industrial wood harvest for the scenarios is illustrated in Fig. 4a. Scenario EAF-tot featured a substantially higher harvest level compared to scenario CCF-int. Forest stands in the EAF-tot scenario were all managed with EAF and had a rotation period around 70 years in average. The harvest would decrease from the start of the simulation period in 2010 to reach the lowest level around 2070, during which period the trees were relatively young and not ready for harvest. The industrial wood harvest was much lower in the CCF-int scenario, which to a large extent relied on the 70% of the forest area that was managed with intensified EAF. The extraction of biomass for bioenergy included all harvest residues, i.e., tops, branches and foliage, from both thinning and final felling, and bark as by-product of industrial wood. The results show that the yields of biomass for bioenergy were higher in the EAF-tot scenario than in the CCF-int scenario (Fig. 4b). The reason for this is that in the CCF-int scenario fewer residues were available from final felling which is the form of management that produces the largest volumes of residues. According to the scenarios, the

<70 years >70 years Pine Spruce Mixed Con. Mixed (Con. with BL.) Trivial Southern BL. Mixed (S.BL. with other BL.)

bioenergy feedstock production was projected to be lowest around 2070. As can be seen when comparing Fig. 4a and b, the extraction of harvest residues was strongly correlated with the overall harvest levels. Moreover, due to the lower harvested volumes in 30% of the forest in the CCF-int scenario, the carbon stock was higher than in the EAF-tot scenario, and stayed at a high level since the forest canopy was maintained. In the EAF-tot scenario, the extimated carbon stock in the forest was lower due to the higher harvest removals (Fig. 4c). The differences in forest structure in the two scenarios resulted in different overall size of the total potential recreation area. Due to the larger areas of final felling and thus less mature and old trees in the EAF-tot scenario, the potential recreation area was much smaller than that of the CCF-int scenario (Fig. 4d). Furthermore, the results show that in the CCF-int scenario, the amount of habitat assumed to contain mature and old trees, which is habitat for both focal species, was projected to increase (Fig. 4e). Meanwhile, the mature and old forest in the CCF-int scenario also worked as links to other suitable habitat which may otherwise be situated beyond the focal species’ median movement distances and therefore may lose the function as habitat through decreased connectivity and thereby size of available habitat. That is why the ECA was very much higher in the CCF-int scenario than in the EAF-tot scenario. A zoomed in example for coniferous forest is shown in Fig. 5 to exemplify the differences in the area of available habitat (ECA of habitat networks) between the EAF-tot and CCF-int scenarios. There is a high risk that, in the example of the EAF-tot scenario, species dependent on this forest habitat may lose much of the

Fig. 3. Overview of the LEcA tool, see Sub-section 2.4 for explanations.

73

a) Industrial wood

Milll. m3

X. Pang et al. / Ecosystem Services 28 (2017) 67–79

5 4 3 EAF-tot

2

CCF-int

1

b) Bioenergy feedstock

Milll. to on

0

c) Carbon stock

1.2 1.0 0.8 0.6 0.4 0.2 0.0

EAF-tot CCF-int

100 Mill. to on C

80 60 EAF-tot

40

CCF-int

20 0

d)Potenal recreaon area

4000

Km2

3000 2000

EAF-tot

1000

CCF-int

0

e) ECA of habitat networks

2500

Km2

2000 1500 1000 500

EAF_tot CCF_int

0

Fig. 4. Supply of ecosystem services during 2015–2110 in the EAF-tot and CCF-int scenarios: (a) industrial wood production (m3), (b) bioenergy feedstock (million tons), (c) carbon stock (Mt C), (d) potential recreation area (km2), and (e) available habitat of the habitat networks measured as ECA (km2).

74

X. Pang et al. / Ecosystem Services 28 (2017) 67–79

Fig. 5a. Habitat changes in the EAF-tot scenario (a zoomed-in area). The blank or almost blank boxes from 2020 to 2110 show that there is almost no available habitat within the habitat networks in this area during this time period. Spatial data Ó Lantmäteriet [i2012/920], coordinate system Sweref 99 TM.

available habitat in the County during certain time steps, such as around year 2080 (Fig. 5a). By contrast the situation was projected to be much better in the CCF-int scenario (Fig. 5b).

3.2. Trade-offs and synergies The projections for the five ecosystem services were compared with the situation for year 2010, the base year for the assessment, and the increase or decrease in percentages are shown in Fig. 6. According to SFA (2013), the gross felling of Kronoberg in year 2010 was 3,984,000 m3. They calculated that under current forest management, 448,200 tons of tops, branches, foliage and bark could be extracted theoretically for year 2010 in Kronoberg. The carbon stock in 2010 in Kronoberg was 37 million ton based on our results, and the total area of forest suitable for recreation was 735 km2 in year 2010, according to our definition. Furthermore, the total ECA, considering both coniferous and southern broadleaved forest habitat, was according to our estimations 326 km2 in 2010. Trade-offs and synergies among the five ecosystem services for each scenario are thus illustrated in Fig. 6 which display the increase or decrease in percentage over time in relation to the initial state in 2010. In the EAF-tot scenario, the output from the for-

est resource was dominated by industrial wood, bioenergy feedstock production and, to some extent, carbon storage. Available habitat of the habitat networks, as measured by ECA, was predicted to follow a negative trend and would most probably be very low following this development (Fig. 6a). In the CCF-int scenario, the production of timber, pulpwood and bioenergy feedstock were not as high as in the EAF-tot scenario, but the other ecosystem services assessed showed positive development over time and there was a synergy between these services. Especially for habitat networks, the increase was large from 2050 and onwards compared with the base year 2010 (Fig. 6b).

4. Discussion As expected, there was a clear synergy between industrial wood production and bioenergy in this study, since the tops, branches, foliage and bark used for bioenergy were either residuals from harvesting or a by-product from the industrial wood. The scenarios further showed that there were trade-offs between industrial wood and bioenergy production on one side and habitat, recreation and, to some extent, carbon storage on the other side. These tradeoffs became more pronounced over time, as the available habitat and recreation areas increased very much in the CCF-int scenario

Fig. 5b. Habitat changes in the CCF-int scenario (the same zoomed-in area as in Fig. 5a). The green areas in the boxes from 2020 to 2110 show suitable and available habitat within the habitat networks in this area. Spatial data Ó Lantmäteriet [i2012/920], coordinate system Sweref 99 TM.

75

X. Pang et al. / Ecosystem Services 28 (2017) 67–79

Industrial wood 500 400 300 200 Habitat

Bioenergy 100

2010 2020 2050

0

2080 2110

Recreaon

Carbon stock

Fig. 6a. Trade-offs and synergies in the EAF-tot scenario 2010–2110 (% of the values of year 2010). For example, in year 2110 bioenergy is projected to increase till about 200% of that in year 2010, while the habitat area of the habitat networks will shrink to 22%.

while they decreased in the EAF-tot scenario compared to the base year 2010. Especially available habitat area of the habitat networks (measured as ECA) decreased dramatically in the EAF-tot scenario. In the CCF-int scenario, industrial wood production would be reduced by around 30% compared with the EAF-tot scenario, and thus the income from wood production would be negatively affected. However, the harvested volumes were more evenly distributed over time and the standing volume as well as carbon storage in the forest was increasing over time in the CCF-int scenario (see Fig. 4), and could potentially be harvested for industrial wood production and bioenergy in the future. In our simulations, EAF meant extracting all available harvest residues for bioenergy purposes, both in the EAF-tot and the CCFint scenario. However, as mentioned in the Introduction, this is not the case in forest practice today, and therefore there is such a large increase in bioenergy yield as illustrated in Fig. 6, where the base year was the real situation in 2010. Increased extraction of harvest residues can be expected to have environmental impacts. de Jong et al. (2017) estimated thresholds for how much extraction of harvest residues could maximally increase in Sweden in order to avoid conflicts between different national environmental objectives. According to these authors, the extraction of branches and tops should not reach over 80% of the total final felling area in order to keep forest productivity and avoid leakage of toxic substances, and it should not reach over 50% of the total final felling area to avoid acidification and conflicts with biodiversity targets. Stumps should only be harvested on less than 30% of the total final felling area to keep forest productivity and to avoid leakage of toxic substances, and only on up to 20% of the final felling area to avoid conflicts with biodiversity targets. The residues contribute to the supply of dead wood, which is an important biodiversity component in the forest, and the extraction of harvest residues should be restricted to coniferous forest, mainly avoiding deciduous trees (de Jong et al., 2017; de Jong and Dahlberg, 2017). The LEcA tool could in future steps be used to spatially model and assess implications of applying these thresholds and restrictions. Forest bioenergy is seen as renewable energy and is further generally accounted for as carbon–neutral (EU Renewable Energy Directive 2009/28/EC). This means that forest bioenergy could con-

tribute to climate change mitigation due to the neutrality of GHG emissions in the whole life cycle of the harvest and regeneration process. Thus, it is considered to be an alternative to replace fossil fuels. However, the carbon neutrality of forest bioenergy can be disputed (Zanchi et al., 2012). Until the carbon stock grows back to the level before the forest stand was harvested, the overall GHG impact from forest bioenergy may still be negative. For sawn wood, carbon is stored in wood products, e.g., construction material or furniture, until it is released at burning or decomposition. The two scenarios in this study represent two different policies for climate change mitigation: substitution and storage in the forest. In the EAF-tot scenario, the production of industrial wood and bioenergy was high and could contribute to the substitution of fossil-based fuel and materials and, in addition, carbon would be stored in wood products. In the CCF-int scenario, CCF resulted in lower harvest levels and less industrial wood and bioenergy would be produced but climate change mitigation would be achieved by storing carbon in the forest. In order to measure the function of the forest as a carbon sink, biomass expansion factors were used to estimate the carbon content of a whole tree using stem volume as input data. The biomass expansion factors developed by Lehtonen et al. (2004) are dependent on stand age and tree species, which provide a relatively firm basis for quantifying forest carbon stock. Thus, in this study we assessed only the carbon stock in the tree layer, and the total carbon stock in the forest would be considerably higher if soil carbon and carbon in litter and ground vegetation were included (Cronan, 2003; Peltoniemi et al., 2006). Moreover, if the substitution effects (such as forest regeneration) from using forest biomass as raw material instead of fossil fuels and if carbon stored in wood products would have been taken into account, this would also have affected the carbon balance (Gustavsson and Sathre, 2006). The climate mitigation effects are dependent on the system boundaries, e.g., the temporal and spatial scales considered as well as the extent of the life cycle for forest products included in the analysis. For instance, some recent studies indicate that substitution effects and storage of carbon in harvested wood products may result in higher overall carbon sequestration in a long-term perspective than the strategy to store carbon in the forest (Cintas et al.,

76

X. Pang et al. / Ecosystem Services 28 (2017) 67–79

Industrial wood 500 400 300 200

Habitat

Bioenergy

100

2010 2020 2050

0

2080 2110

Recreation

Carbon stock

Fig. 6b. Trade-offs and synergies in the CCF-int scenario 2010–2110 (% of the values of year 2010). For example, bioenergy is projected to be around 137% in year 2110 compared with year 2010, while the habitat area of the habitat networks would increase by over 5 times given the modelling assumptions.

2017; Gustavsson et al., 2017; Lundmark et al., 2014). These issues should be considered in future studies to widen the scope on effects of forests and forestry on the carbon balance. In this study, the age of the forest was used as an indicator of forest recreation value on stand level; as soon as it was old enough, it was assumed to be good for recreation. Old forest is generally found to have a high recreational value in many preference studies (e.g., Gundersen and Frivold, 2008), but the assumption is still a simplification since there are other factors that can influence the recreational value of a forest stand, such as tree species composition and openness (Gundersen and Frivold, op. cit.), as well as biodiversity (Horne et al., 2005). Since the distance to the forest also affects preferences for recreational forest and how frequently it is used for recreation (Hörnsten and Fredman, 2000; Barbosa et al., 2007), we assumed that suitable forests close to housing areas and easily accessible from roads were most likely to be used for everyday recreation and this was added as a landscape level variable. For future use, recreation indicators in the LEcA tool can be further developed, combining forest age with for instance tree species, stem volume, extraction of harvest residuals, and different spatial parameters such as stand size, proximity to water bodies, trails, viewsheds, and more. In this way, different recreation values can be assessed on landscape level. The habitat network assessment enabled the analysis of habitat size and connectivity simultaneously, which is essential for the possibility for species to persist in a landscape. There are however uncertainties in the parameters used for the modelling, for instance concerning home range size, and the habitat suitability of forests treated with CCF, both addressed in Pang et al. (2017), and other parameters. More knowledge is in general needed concerning species’ demands on their landscapes, including thresholds in available habitat for long-term persistence. Moreover, the habitat demands of the selected focal species can be considered to have high indicator values (Schmiegelow and Mönkkönen, 2002), but of course more habitat types and focal species could be included, representing e.g., fungi, insects, large vertebrates and mammals. Policies for extraction of forest residues, affecting the supply of dead wood, could also be assessed using the spatial dimension of habitat networks. Thus, a wider array of habitat network indicators could

be assessed by the LEcA tool in order to be representative of main biodiversity components affected by the intended forest management. The EAF-tot scenario focussed on forest biomass extraction, but the on-going large-scale application of EAF management is seen as a major cause of negative environmental impacts, even though measures such as forest retention are undertaken (Eriksson et al., 2015; Puettmann et al., 2009; Thompson et al., 2011). While policies for further land zoning and retention measures are suggested, the increasing demand for forest biomass for bioenergy and other purposes may still imply a further intensification of the forestry, with related risks for increasing these impacts. In Sweden, forest management practices and harvesting technology have been developed to make EAF into a very efficient management system from an economic point of view, with a high and even production of industrial wood as well as bioenergy feedstock. However, in order to avoid detrimental impacts, EAF in Sweden is carried out as retention forestry, which means that some trees, groups of trees and dead wood are retained at final felling (Simonsson et al., 2014). By contrast, in the EAF-tot scenario in this study no retention was applied and trees were harvested from around the age of 70 years. Thus, the EAF-tot scenario is more intensive than the current management practices with retention forestry, in which the habitat networks are potentially larger, depending on the amount and spatial distribution of mature and old trees being left. Moreover, final felling at the age of around 70 years is economically rational but it may not be the current practice in reality, since many other factors affect harvesting decisions and the rotation time. Consequently, the EAF-tot scenario is rather an ‘‘intensive management” scenario compared with current management practices, which could be mitigated by different retention measures using the LEcA tool (see Pang et al., 2017). In the CCF-int scenario, the idea was to dedicate 30% of the forest landscape to multiple ecosystem services. In view of the increasing demands on forest biomass for multiple purposes, the current attraction of CCF lies in the selective harvesting method to maintain the forest canopy that reduces the impact of clearcutting (Pommerening and Murphy, 2004), but it is uncertain to what extent old and mature trees are kept. Still, from an ecological

X. Pang et al. / Ecosystem Services 28 (2017) 67–79

perspective, forests with a CCF management strategy are expected to support species that depend on forest continuity, thus may better promote ecological objectives with the trade-off of lower biomass yields (Kuuluvainen et al., 2012). Moreover, the question of whether CCF is a viable alternative to EAF and under what conditions is one of the major issues currently under debate in Swedish forestry (SLU, 2013b). There is no large-scale practical experience of CCF management of Swedish forestry and growth and yield functions as well as forest decision support systems and models are all mainly based on EAF (e.g., Wikström et al., 2011). Further research on CCF will reduce the uncertainties of its impacts, and in addition, alternative development pathways of CCF can be simulated. In order to assess land zoning policies for forest management and their effect on ecosystem services, the spatial component is essential. Other models that can simulate forest growth and management and ecosystem services on a landscape level are of course available. One such model is LANDIS (Gustafson et al., 2000), a raster-based model addressing forest ecosystem development including a wider range of ecological processes than the LEcA tool. However, the LEcA tool including LandSim as forest development simulator (Pang et al., 2017) has the advantage of using an empirical model for volume growth of the forest (based on data from the National Forest Inventory of Sweden) while in LANDIS forest growth is reflected only in age. Thus the LEcA tool is well suited for modelling forest management, in this case adapted to Sweden. Another decision support model for forest management in Sweden is HEUREKA (Wikström et al., 2011), which can address a wide range of forest management activities and ecosystem services in more detail. However, this model is data-intensive and not so well suited for treating spatially explicit data over larger regions, which is a strength of the LEcA tool. To reduce climate impacts and simultaneously preserve forest environments for multiple ecosystem services and biodiversity conservation, several measures are needed. The quality and scope of measures to counter habitat loss and fragmentation need to increase, so that biodiversity values are sustained on a landscape scale. Nature reserves and other forms of protection are essential, combined with voluntary set-asides of forest by land-owners (SEPA, 2012b). How to apply land zoning policies between protected forest areas, more or less intensified EAF and alternative forestry methods such as CCF is debated (SLU, 2013b), while the LEcA tool can be used to address these issues. At the same time, maintaining high wood production is most likely needed to meet climate goals and for the transition into a green economy. Since some ecosystem services have a commercial aspect and other not, it would be interesting to integrate economic aspects in the policy assessments. All in all, to achieve a balance between conserving and developing biodiversity and a range of ecosystem services, while still remaining competitive as timber and energy resource providers, is going to be a challenge for forest and energy policies in the future. The LEcA tool presented in this study could be used to assess the consequences of various forest management strategies related to both renewable energy and conservation policies. In this way, synergies and trade-offs between forest ecosystem services can be localised, quantified and assessed for different land use related policies and plans, thus providing decision support for a sustainable forest management.

5. Conclusions In this paper we present the LEcA tool for applying a landscape approach to the assessment of land zoning policies with related forest management strategies, linking forest development simula-

77

tion with the change of forest ecosystem services. These strategies were explored through two land zoning scenarios, one dominated by EAF (EAF-tot) and one combining CCF with intensified EAF (CCFint). The results showed that different land zoning policies would have different impacts on long-term potential biomass extraction and other ecosystem services; and that the impacts would change over time. The scenarios showed trade-offs between industrial wood and bioenergy production on one side and habitat, recreation and, to some extent, carbon storage on the other side. These tradeoffs became more pronounced over time, as the available habitat and recreation areas increased very much in the CCF-int scenario while they decreased in the EAF-tot scenario compared to the base year 2010. In the CCF-int scenario, industrial wood production would be reduced by around 30% compared with the EAF-tot scenario, and thus the income from wood production could be negatively affected. The scenarios could be adjusted and re-run in order to find more sustainable solutions. In addition, the LEcA tool is prepared for applying more comprehensive scenarios, with for instance retention of forest in EAF and different applications of CCF. In addition, indicators for recreation and biodiversity can be more detailed, and uncertainties can be expressed by applying multiple-range parameters. In the current study, the LEcA tool showed great potential for evaluation of impacts of alternative policies for land zoning and forest management on forest ecosystem services. It can be used to assess the consequences of various land zoning policies related to renewable energy as well as conservation policies. In this way, synergies and trade-offs between forest ecosystem services can be localised, quantified and assessed for different land use related policies and plans, thus providing decision support for a sustainable forest management. Acknowledgements This study was implemented in collaboration with the Environmental Management and Assessment Research Group, KTH Royal Institute of Technology, and the Swedish University for Agricultural Sciences, within the Young Scientist Summer Program at the International Institute for Applied Systems Analysis (IIASA), Austria, funded by the Swedish Research Council for Environment, Agricultural Sciences and Spatial Planning (FORMAS). The study was also supported by the Swedish strategic research programme STandUP for Energy. References Balmford, A., Rodrigues, A.S.L., Walpole, M., ten Brink, P., Kattunen, M., Braat, L., de Groot, R., 2008. The Economics of Biodiversity and Ecosystems: Scoping the Science. European_Commission, Cambridge. Barbosa, O., Tratalos, J.A., Armsworth, P.R., Davies, R.G., Fuller, R.A., Johnson, P., Gaston, K.J., 2007. Who benefits from access to green space? A case study from Sheffield, UK. Landscape and Urban Planning 83, 187–195. Berg, Å., Ehnström, B., Gustafsson, L., Hallingbäck, T., Jonsell, M., Weslien, J., 1994. Threatened plant, animal, and fungus species in Swedish forests: distribution and habitat associations. Conserv. Biol. 8, 718–731. Börjesson, P., Hansson, J., Berndes, G., 2017. Future demand for forest-based biomass for energy purposes in Sweden. For. Ecol. Manage. 383, 17–26. Bottalico, F., Pesola, L., Vizzarri, M., Antonello, L., Barbati, A., Chirici, G., Corona, P., Cullotta, S., Garfì, V., Giannico, V., Lafortezza, R., Lombardi, F., Marchetti, M., Nocentini, S., Riccioli, F., Travaglini, D., Sallustio, L., 2016. Modeling the influence of alternative forest management scenarios on wood production and carbon storage: a case study in the Mediterranean region. Environ. Res. 144, 72–87. Burkhard, B., Kroll, F., Müller, F., 2010. Landscapes‘ capacities to provide ecosystem services – A concept for land-cover based assessments. Landscape Online, 1–22. Bütler, R., Angelstam, P., Ekelund, P., Schlaepfer, R., 2004. Dead wood threshold values for the three-toed woodpecker presence in boreal and sub-Alpine forest. Biol. Conserv. 119, 305–318. CAB, 2006. Vår miljö i Kronobergs län [In Swedish]. Meddelande nr. 2006:4, Länsstyrelsen i Kronobergs län [County Administrative Board of Kronoberg]. CAB, 2013. Länsstyrelsernas GIS-service [County Administrative Boards GIS data service, in Swedish].2013-01-10.

78

X. Pang et al. / Ecosystem Services 28 (2017) 67–79

Canadell, J.G., Raupach, M.R., 2008. Managing forests for climate change mitigation. Science 320, 1456–1457. Carlsson, M., 2012. Bioenergy from the Swedish Forest Sector – A Partial Equilibrium Analysis of Supply Costs and Implications for the Forest Product Markets. Department of Economics. Swedish University of Agricultural Sciences, Uppsala. Cintas, O., Berndes, G., Hansson, J., Poudel, B.C., Bergh, J., Börjesson, P., Egnell, G., Lundmark, T., Nordin, A., 2017. The potential role of forest management in Swedish scenarios towards climate neutrality by mid century. For. Ecol. Manage. 383, 73–84. Cronan, C.S., 2003. Belowground biomass, production, and carbon cycling in mature Norway spruce, Maine, U.S.A. Can. J. For. Res. 33, 339–350. de Jong, J., Dahlberg, A., 2017. Impact on species of conservation interest of forest harvesting for bioenergy purposes. For. Ecol. Manage. 383, 37–48. de Jong, J., Akselsson, C., Egnell, G., Löfgren, S., Olsson, B.A., 2017. Realizing the energy potential of forest biomass in Sweden – How much is environmentally sustainable? For. Ecol. Manage. 383, 3–16. D’Eon, R.G., Glenn, S.M., Parfitt, I., Fortin, M.-J., 2002. Landscape connectivity as a function of scale and organism vagility in a real fore sted landscape. Conserv. Ecol. 6, 10. EC, 2001. A Sustainable Europe for a Better World: A European Union Strategy for Sustainable Development. European Commission, Brussels. EC, 2010. Report from the Commission to the Council and the European Parliament on sustainability requirements for the use of solid and gaseous biomass sources in electricity, Heating and Cooling. SEC 65 and SEC (2010) 66. European Commission, Brussels. EC, 2012. Innovating for Sustainable Growth: A Bioeconomy for Europe. COM(2012) 60 final. European Commission. EEA, 2010. EU 2010 biodiversity baseline. EEA Technical report No 12/2010, European Environment Agency. Eriksson, A., Snäll, T., Harrison, P.J., 2015. Analys av miljöförhållanden – SKA 15. Rapport nr 11, Swedish Forest Agency. Jönköping. ESRI, 2009. ArcGIS Version 10.2 [GIS Application]. Environmental Systems Research Institute Inc., Redlands, CA. Filyushkina, A., Strange, N., Löf, M., Ezebilo, E.E., Boman, M., 2016. Non-market forest ecosystem services and decision support in Nordic countries. Scand. J. For. Res. 31, 99–110. Forsius, M., Akujärvi, A., Mattsson, T., Holmberg, M., Punttila, P., Posch, M., Liski, J., Repo, A., Virkkala, R., Vihervaara, P., 2015. Modelling impacts of forest bioenergy use on ecosystem sustainability: Lammi LTER region, southern Finland. Ecol. Ind. Fredman, P., Tyrväinen, L., 2010. Frontiers in Nature-Based Tourism. Scand. J. Hospit. Tourism 10, 177–189. Fridman, J., Holm, S., Nilsson, M., Nilsson, P., Hedström Ringvall, A., Ståhl, G., 2014. Adapting National Forest Inventories to changing requirements – the case of the Swedish National Forest Inventory at the turn of the 20th century. Silva Fennica 48: article id 1095, 1029 p. Gundersen, V.S., Frivold, L.H., 2008. Public preferences for forest structures: a review of quantitative surveys from Finland, Norway and Sweden. Urban Forest. Urban Greening 7, 241–258. Gustafson, E.J., Shifley, S.R., Mladenoff, D., Nimerfro, K.K., He, H.S., 2000. Spatial simulation of forest succession and timber harvesting using LANDIS. Can. J. For. Res. 30, 32–42. Gustavsson, L., Sathre, R., 2006. Variability in energy and carbon dioxide balances of wood and concrete building materials. Build. Environ. 41, 940–951. Gustavsson, L., Haus, S., Lundblad, M., Lundström, A., Ortiz, C.A., Sathre, R., Truong, N.L., Wikberg, P.E., 2017. Climate change effects of forestry and substitution of carbon-intensive materials and fossil fuels. Renew. Sustain. Energy Rev. 67, 612–624. Hanski, I., 2011. Habitat loss, the dynamics of biodiversity, and a perspective on conservation. Ambio 40, 248–255. Hedin, J., Isacsson, G., Jonsell, M., Komonen, A., 2008. Forest fuel piles as ecological traps for saproxylic beetles in oak. Scand. J. For. Res. 23, 348–357. Hoel, M., Sletten, T.M., 2016. Climate and forests: The tradeoff between forests as a source for producing bioenergy and as a carbon sink. Resour. Energy Econ. 43, 112–129. Horne, P., Boxall, P.C., Adamowicz, W.L., 2005. Multiple-use management of forest recreation sites: a spatially explicit choice experiment. For. Ecol. Manage. 207, 189–199. Hörnsten, L., Fredman, P., 2000. On the distance to recreational forests in Sweden. Landscape Urban Plan. 51, 1–10. IEA, 2014. Consequences of an increased extraction of forest biofuel in Sweden. International Energy Agency. IPCC, 2014. Climate Change 2014 Synthesis Report Summary for Policymakers. IPCC, Switzerland, Geneva, p. 151. Jonsson, R., 2012. Econometric modelling and projections of wood products demand, supply and trade in Europe. Geneva Timber and Forest Discussion Papers 59, UNECE/FAO Forestry and Timber Section. United_Nations. Geneva, Switzerland. Jonsson, R., 2013. How to cope with changing demand conditions – The Swedish forest sector as a case study: an analysis of major drivers of change in the use of wood resources. Can. J. For. Res. 43, 405–418. Kuuluvainen, T., Tahvonen, O., Aakala, T., 2012. Even-aged and uneven-aged forest management in Boreal Fennoscandia: a review. Ambio 41, 720–737. Lantmäteriet, 2013a. GSD Landcover Data. Ó Lantmäteriet [i2012/920]. Lantmäteriet, 2013b. GSD Topographic Map. Ó Lantmäteriet [i2012/920].

Larsson, S., Lundmark, T., Ståhl, G., 2009. Möjligheter till intensivodling av skog. Slutrapport. [Possibilities for intensive forestry. Final report]. Governmental Commission 2008, No. 1885. Stockholm. Lauri, P., Havlik, P., Kindermann, G., Forsell, N., Böttcher, H., Obersteiner, M., 2014. Woody biomass energy potential in 2050. Energy Policy 66, 19–31. Lehtonen, A., Mäkipää, R., Heikkinen, J., Sievänen, R., Liski, J., 2004. Biomass expansion factors (BEFs) for Scots pine, Norway spruce and birch according to stand age for boreal forests. For. Ecol. Manage. 188, 211–224. Leonard, D.L., 2001. Three-toed woodpecker (Picoides tridactylus). In: Poole, A., Gill, F. (Eds.), The Birds of North America, No. 588. The Birds of North America Inc., Philadelphia, PA. Lidskog, R., Sundqvist, G., Kall, A.-S., Sandin, P., Larsson, S., 2013. Intensive forestry in Sweden: stakeholders’ evaluation of benefits and risk. J. Integr. Environ. Sci. 10, 145–160. Lundmark, T., Bergh, J., Hofer, P., Lundström, A., Nordin, A., Poudel, B.C., Sathre, R., Taverna, R., Werner, F., 2014. Potential roles of Swedish forestry in the context of climate change mitigation. Forests 5, 557–578. Malhi, Y., Meir, P., Brown, S., 2002. Forests, carbon and global climate. Philos. Trans. R. Soc. A: Math. Phys. Eng. Sci. 360, 1567–1591. Mattsson, L., Li, C., 1994. How do different forest management practices affect the non-timber value of forests? – An economic analysis. J. Environ. Manage 41, 78– 88. MEA, 2005. Ecosystems and Human Well-being: Synthesis. Island Press, Washington, DC. Melin, Y., 2014. Impacts of Stumps and Roots on Carbon Storage and Bioenergy Use in a Climate Change Context (Doctoral thesis). Swedish University of Agricultural Sciences, Umeå. Milligan, C., Bingley, A., 2007. Restorative places or scary spaces? The impact of woodland on the mental well-being of young adults. Health Place 13, 799–811. Ministry_of_the_Environment_Sweden, 2014. Sweden’s sixth national communication on climate change – Under the United Nations framework convention on climate change. Report Ds 2014:11, Regeringskansliet, Stockholm. Mörtberg, U., Zetterberg, A., Brokking Balfors, B., 2012. Urban landscapes in transition: lessons from integrating biodiversity and habitat modelling in planning. J. Environ. Assess. Policy Manage. 14, 1250002. Müller, J., Pöllath, J., Moshammer, R., Schröder, B., 2009. Predicting the occurrence of Middle Spotted Woodpecker Dendrocopos medius on a regional scale, using forest inventory data. For. Ecol. Manage. 257, 502–509. Norman, J., Ellingson, L., Boman, M., Mattsson, L., 2010. The value of forests for outdoor recreation in southern Sweden: are broadleaved trees important? Ecol. Bull. 53, 21–31. Pan, Y., Birdsey, R.A., Fang, J., Houghton, R., Kauppi, P.E., Kurz, W.A., Phillips, O.L., Shvidenko, A., Lewis, S.L., Canadell, J.G., Ciais, P., Jackson, R.B., Pacala, S.W., McGuire, A.D., Piao, S., Rautiainen, A., Sitch, S., Hayers, D., 2011. A large and persistent carbon sink in the world’s forests. Science 333, 988–993. Pang, X., Mörtberg, U., Brown, N., 2014. Energy models from a strategic environmental assessment perspective – what is missing concerning renewables? Renew. Sustain. Energy Rev. 33, 353–362. Pang, X., Mörtberg, U., Sallnäs, O., Trubins, R., Nordström, E.-M., Böttcher, H., 2017. Habitat network assessment of forest bioenergy options using the landscape simulator LandSim – a case study of Kronoberg, southern Sweden. Ecol. Model. 345, 99–112. Pasinelli, G., 2000. Oaks (Quercus sp.) and only oaks? Relations between habitat structure and home range size of the middle spotted woodpecker (Dendrocopos medius). Biol. Conserv. 93, 227–235. Peltoniemi, M., Palosuo, T., Monni, S., Mäkipää, R., 2006. Factors affecting the uncertainty of sinks and stocks of carbon in Finnish forests soils and vegetation. For. Ecol. Manage. 232, 75–85. Penman, J., Gytarsky, M., Hiraishi, T., et al, 2003. Good Practice Guidance for Land Use, Land Use Change, and Forestry. Institute for Global Environmental Strategies for the Intergovernmental Panel on Climate Change Hayama, Kanagawa, Japan. Pettersson, B., 1984. Territory size and habitat characters of the middle spotted woodpecker Dendrocopos medius (L.) in Sweden. PM 1813, Statens naturvårdsverk [Swedish Environmental Protection Agency]. Pommerening, A., Murphy, S., 2004. A review of the history, definitions and methods of continuous cover forestry with special attention to afforestation and restocking. Forestry 77, 27–44. Puettmann, K.J., Coates, K.D., Messier, C., 2009. A Critique of Silviculture: Managing for Complexity. Island Press, Washington, DC. Ranius, T., Roberge, J.-M., 2011. Effects of intensified forestry on the landscape-scale extinction risk of dead wood dependent species. Biodivers. Conserv. 20, 2867– 2882. Reese, H., Nilsson, M., Granqvist Pahlén, T., Hagner, O., Joyce, S., Tingelöf, U., Egberth, M., Olsson, H., 2003. Countrywide estimates of forest variables using satellite data and field data from the National Forest Inventory. Ambio 32, 542–548. Roberge, J.-M., Angelstam, P., 2006. Indicator species among resident forest birds – A cross-regional evaluation in northern Europe. Biol. Conserv. 130, 134–147. Saura, S., Rubio, L., 2010. A common currency for the different ways in which patches and links can contribute to habitat availability and connectivity in the landscape. Ecography 33, 523–537. Saura, S., Estreguil, C., Mouton, C., Rodríguez-Freire, M., 2010. Network analysis to assess landscape connectivity trends: Application to European forests (1990– 2000). Ecol. Ind. 11, 407–416.

X. Pang et al. / Ecosystem Services 28 (2017) 67–79 Saura, S., Vogt, P., Velázquez, J., Hernando, A., Tejera, R., 2011. Key structural forest connectors can be identified by combining landscape spatial pattern and network analyses. For. Ecol. Manage. 262, 150–160. Schmiegelow, F.K.A., Mönkkönen, M., 2002. Habitat loss and fragmentation in dynamic landscapes: avian perspectives from the boreal forest. Ecol. Appl. 12, 375–389. SEA, 2015. Energy in Sweden 2015. (access date: 2016-05-26). SEA, 2016. Production of unprocessed wood fuels 2015. ES 2016:05, Swedish Energy Agency. SEPA, 2006. Riksintresse för naturvård och friluftsliv [Areas of national interest for nature conservation and recreation] Swedish Environmental Protection Agency, Bromma. SEPA, 2009. Swedish Nature Conservation 100 years. The Swedish Environmental Protection Agency, Stockholm. SEPA, 2011. Sweden’s Environment Problems and Protection, 1960–2010. The Swedish Environmental Protection Agency, Stockholm. SEPA, 2012a. Grön infrastruktur – Redovisning av regeringsuppdrag [Green infrastructure – report on Governmental commission] Swedish Environmental Protection Agency. Bromma. SEPA, 2012b. Sweden’s Environmental Objectives. Swedish Environmental Protection Agency, Stockholm. SEPA, 2014. Regional uppföljning av miljökvalitetsmålen 2013 [in Swedish]. PM Dnr NV-05447-13, Swedish Environmental Protection Agency. SEPA, 2015. Mål i sikte. Analys och bedömning av de 16 miljökvalitetsmålen i fördjupad utvärdering. Volym 2. Rapport 6662. Swedish Environmental Protection Agency, Bromma. SFA, 2013. Skogsstatistisk årsbok 2013 [Swedish Statistical Yearbook of Forestry 2013]. Swedish Forest Agency. SFA, 2014. Skogsstatistisk årsbok 2014 [Swedish Statistical Yearbook of Forestry 2014]. Swedish Forest Agency. SFA, 2015. Skogliga konsekvensanalyser 2015 – SKA 15 [in Swedish]. Rapport 10, Swedish Forest Agency. Simonsson, P., Gustafsson, L., Östlund, L., 2014. Retention forestry in Sweden: driving forces, debate and implementation 1968–2003. Scand. J. For. Res., 1–20

79

SLU, 2013a. Forest map. Department of Forest Resource Management, S.U.o.A. S.2013-01-10. SLU, 2013b. Future Forests. Swedish University of Agricultural Sciences, Uppsala. SNBH, 2013. Riksintressen – nationella värden och möjligheter. Boverket [The Swedish National Board of Housing Building and Planning]. Stockholm. Sonntag-Öström, E., Nordin, M., Lundell, Y., Dolling, A., Wiklund, U., Karlsson, M., Carlberg, B., Slunga Järvholm, L., 2014. Restorative effects of visits to urban and forest environments in patients with exhaustion disorder. Urban For. Urban Green. 13, 344–354. SOU, 2014. I vått och tort – förslag till ändrade vattenrättsliga regler. Slutbetänkande av Vattenverksamhetsutredningen [in Swedish]. SOU 2014:35. Statens Offentliga Utredningar, Stockholm. Statistics_Sweden, 2013. Markanvändningen i Sverige [Land use in Sweden]. Statistiska centralbyrån [Statistics Sweden]. Örebro. Thompson, I.D., Okabe, K., Tylianakis, J.M., Kumar, P., Brockerhoff, E.G., Schellhorn, N.A., Parrotta, J.A., Nasi, R., 2011. Forest biodiversity and the delivery of ecosystem goods and services: translating science into policy. Bioscience 61, 972–981. UNFCCC, 2015. Historic Paris Agreement on Climate Change. United Nations Framework Convention on Climate Change, (access date: 2016-09-09). Vandewalle, M., Sykes, M., Harrison, P., 2009. Review paper on concepts of dynamic ecosystems and their services. Rationalising Biodiversity Conservation in Dynamic Ecosystems. The RUBICODE project, EC 6th Framework Programme. Verkerk, P.J., Mavsar, R., Giergiczny, M., Lindner, M., Edwards, D., Schelhaas, M.J., 2014. Assessing impacts of intensified biomass production and biodiversity protection on ecosystem services provided by European forests. Ecosyst. Serv. 9, 155–165. Wikström, P., Edenius, L., Elfving, B., Eriksson, L.O., Lämås, T., Sonesson, J., Öhman, K., Wallerman, J., Waller, C., Klintebäck, F., 2011. The Heureka forestry decision support system: an overview. Math. Comput. For. Nat.-Resour. Sci. 3, 87–95. Zanchi, G., Pena, N., Bird, N., 2012. Is woody bioenergy carbon neutral? A comparative assessment of emissions from consumption of woody bioenergy and fossil fuel. GCB Bioenergy 4, 761–772.