Scenario analysis for the biomass supply potential and the future development of Finnish forest resources Jari Hynynen, Hannu Salminen, Anssi Ahtikoski, Saija Huuskonen, Risto Ojansuu, Jouni Siipilehto, Mika Lehtonen, Arto Rummukainen, Soili Kojola and Kalle Eerikäinen Working Papers of the Finnish Forest Research Institute 302 http://www.metla.fi/julkaisut/workingpapers/2014/mwp302.htm ISBN 978-951-40-2487-0 (PDF) ISSN 1795-150X www.metla.fi Working Papers of the Finnish Forest Research Institute 302 http://www.metla.fi/julkaisut/workingpapers/2014/mwp302.htm 2 Working Papers of the Finnish Forest Research Institute publishes preliminary research results and conference proceedings. The papers published in the series are not peer-reviewed. http://www.metla.fi/julkaisut/workingpapers/ ISSN 1795-150X Office Post Box 18 FI-01301 Vantaa, Finland tel. +358 29532 2111 e-mail julkaisutoimitus@metla.fi Publisher Finnish Forest Research Institute Post Box 18 FI-01301 Vantaa, Finland tel. +358 29532 2111 e-mail info@metla.fi http://www.metla.fi/ Working Papers of the Finnish Forest Research Institute 302 http://www.metla.fi/julkaisut/workingpapers/2014/mwp302.htm 3 Authors Jari Hynynen, Hannu Salminen, Anssi Ahtikoski, Saija Huuskonen, Risto Ojansuu, Jouni Siipilehto, Mika Lehtonen, Arto Rummukainen, Soili Kojola and Kalle Eerikäinen Title Scenario analysis for the biomass supply potential and the future development of Finnish forest resources Year 2014 Pages 106 ISBN 978-951-40-2487-0 (PDF) ISSN 1795-150X Unit / Research programme / Projects Vantaa Unit / Forest and silviculture in the future / 3559, 3587, 743902 Accepted by Taneli Kolström, tutkimusjohtaja, 30.6.2014 Abstract The potential, cost-efficiency and impacts of intensified management of Finnish forests for next 100 years were assessed using a national-level scenario analysis. This document serves as a technical description of the applied models and methods, but also includes a brief synthesis of the results. Data from the 10th Finnish National Forest Inventory was used to forecast the consequences of alternative management scenarios. Four final scenarios were constructed using MOTTI stand simulator and linear programming package J. Business-as-usual –scenario was compared to three other options that aimed either at high quality raw material, intensive management resulting both quantity and quality of timber, or at low-cost-low-output (extensive) forestry. If the intensity of forest management will remain at the current level, the growing stock will increase. Increasing amount of high quality raw material for forest industry can be produced but it necessitates also increase in annual management practices. For example, treatment areas of young stand management should be doubled compared to current areas in order to maintain or increase cutting removals of high quality wood. It is possible to increase annual removals in a sustainable manner by applying more intensive forest management that also improves profitability nearly 50%. The annual removals can be ca. 40% higher than the current level, and the annual energy wood removal can be over 10 mill m3. Despite increased removals, sustainable wood and biomass production during next 100 years can be achieved. Intensively managed forest are more efficient capturing carbon from atmosphere than extensively managed forests, but the climate impacts depend on the use of removed carbon (end-products made from the removed wood biomass). Keywords forest management, carbon, profitability, silvicultural practices, growth model, simulation Available at http://www.metla.fi/julkaisut/workingpapers/2014/mwp302.htm Replaces Is replaced by Contact information Dr. Jari Hynynen, PO Box 18, FI-01301 VANTAA, FINLAND, jari.hynynen@metla.fi Other information Layout: Anne Siika, Metla Working Papers of the Finnish Forest Research Institute 302 http://www.metla.fi/julkaisut/workingpapers/2014/mwp302.htm 4 Contents Preface.................................................................................................................................... 6 Abbreviations........................................................................................................................ 6 1 Introduction..................................................................................................................... 7 1.1 Changes in operational environment................................................................... 7 1.2 Scope and the aim of the work............................................................................. 8 2 Description of scenarios ............................................................................................... 9 2.1 Definition process.................................................................................................. 9 2.2 Scenario 1: Business as usual (BAU)................................................................. 10 2.3 Scenario 2: Active forest sector and intensive biomass production (INT).... 11 2.4 Scenario 3: High quality raw material production for forest industry and bioenergy (QLTY)............................................................................................... 12 2.5 Extensive forestry due to decreasing activities of forest industry – increasing non-material services........................................................................ 13 3 Calculation of scenarios................................................................................................ 15 3.1 Overview of calculation process.......................................................................... 15 3.1.1 Description of data..................................................................................... 16 3.1.2 Converting NFI data into input data of MOTTI simulator.................... 17 3.2 Forest management assumptions, rules, and calculation parameters............. 17 3.2.1 General forest management principles.................................................... 17 3.2.2 Economical parameters.............................................................................. 19 3.2.3 Logging parameters.................................................................................... 20 3.3 Implementation of scenarios at stand-level ....................................................... 20 3.4 Simulation procedure............................................................................................ 22 3.5 Linear programming.............................................................................................. 22 3.6 Compilation of the final scenarios....................................................................... 24 4 Results and discussion................................................................................................... 25 4.1 Forest management practices .............................................................................. 25 4.3 Growing stock ....................................................................................................... 31 4.4 Carbon stock and removed carbon...................................................................... 33 4.5 Profitability of forest management...................................................................... 34 5 Highlights.......................................................................................................................... 37 References.............................................................................................................................. 37 Appendix................................................................................................................................ 39 Working Papers of the Finnish Forest Research Institute 302 http://www.metla.fi/julkaisut/workingpapers/2014/mwp302.htm 5 Appendix 1: Prediction models for stand dynamics in Motti simulator................. 39 Results by Forestry centres: Appendix 2: Forestry centre Rannikko................................................................. 55 Appendix 3: Forestry centre Lounais-Suomi........................................................ 59 Appendix 4: Forestry centre Häme-Uusimaa........................................................ 63 Appendix 5: Forestry centre Kaakkois-Suomi...................................................... 67 Appendix 6: Forestry centre Pirkanmaa................................................................ 71 Appendix 7: Forestry centre Etelä-Savo............................................................... 75 Appendix 8: Forestry centre Etelä-Pohjanmaa...................................................... 79 Appendix 9: Forestry centre Keski-Suomi............................................................ 83 Appendix 10: Forestry centre Pohjois-Savo.......................................................... 87 Appendix 11: Forestry centre Pohjois-Karjala...................................................... 91 Appendix 12: Forestry centre Kainuu................................................................... 95 Appendix 13: Forestry centre Pohjois-Pohjanmaa................................................ 99 Appendix 14: Forestry centre Lappi...................................................................... 103 Working Papers of the Finnish Forest Research Institute 302 http://www.metla.fi/julkaisut/workingpapers/2014/mwp302.htm 6 Preface This working paper is a technical documentation of a large-scale scenario analysis carried out within the EffFibre “Value through Intensive and Efficient Fibre supply” research program of Finnish Bioeconomy Cluster FIBIC Oy. The goal of the three-year EffFibre program (2010– 2013) was to improve the competitiveness of whole forest cluster. The programme focused on improving the availability and supply of high-quality raw material from Finnish forests and developing new production technologies for chemical pulping. The programme was financed by the Finnish Funding Agency for Technology and Innovation (Tekes), which provided 60% of the financing. The remainder sourced from the participating companies and research institutes. One of the central research themes of EffFibre program was to study sustainable availability and cost-efficient supply of domestic forest-based raw material. It is widely acknowledged that productive domestic forests resources and competitive wood supply are crucial for the vitality of Finnish forest cluster. In the EffFibre program, this topic was addressed in Work Package 2 “Potential and feasibility of intensive wood and biomass production”. This topic was tackled by carrying out an extensive scenario analysis. It was based on comprehensive information on the current forest resources, and analysis of the operational environment of Finnish forest sector. As the result of close cooperation between the industrial and research partners of EffFibre partners, alternative future scenarios were defined, and for each of them, future forecasts on wood supply and the development of forest resources were calculated. This effort resulted in research- based information on the production potential of Finnish forest resources, and comprehensive information on the impacts of forest management of varying intensity on Finnish forests and forestry. Abbreviations –– Scenarios: BAU: Business as usual INT: Active forest sector and intensive biomass production QLTY: High quality raw material production for forest industry and bioenergy EXT: Extensive forestry due to decreasing activities of forest industry – increasing non-material services –– WS1-WS7; seven working scenarios were produced in order to technically allow a flexible compilation of the final scenarios –– Forestry centres = The Finnish Forest Centre is a state-funded organisation covering the whole country and operating on 13 local Forestry Centres (see the map). –– NPV = net present value –– NFI = National Forest Inventory 1 2 3 4 5 6 7 8 9 10 11 12 13 Ra L-S H-U Ka-S Pi E-Sa E-Po Ka-S P-Sa P-Ka Ka P-Po La 13 12 11 1091 1 8 7 65 432 Working Papers of the Finnish Forest Research Institute 302 http://www.metla.fi/julkaisut/workingpapers/2014/mwp302.htm 7 1 Introduction 1.1 Changes in operational environment Globally, the demand for renewable raw material and product will increase in the future. In that context the wood based products are in significant role. The main factors affecting long term demand for wood products globally are the increase in world’s population, economic growth in terms of global GDP, regional shifts because of the rapid growth of developing economies, environmental policies and regulation and energy policies (State of the world’s forests. 2009). In Finland, the domestic forest resources are important to Finnish forest industry’s ability to produce wood based products to global markets. In the year 2012, 93% of paper and paperboard, 63% sawn goods and 83% of plywood production were exported abroad (Finnish statistical Yearbook of Forestry 2012). Finland’s share of the global export trade of forest products in 2011 was 5.7% (Forest Finland in brief. 2013). Comparing this to the fact that the forest area in Finland is tiny at global scale, the role of forest industry and production of wood based products is significant globally. In addition, the proportion of imported wood is in minor part and even decreased from the average of this millennium (24%), being 17% of wood procurement at year 2011 (Ylitalo 2012). Thus, the role of domestic wood as raw material is highly significant. The forest land area in Finland is 20.3 million ha (Ylitalo 2012). The growing stock volume is 2306 mill m3 and has increased continuously since the 1970s. Majority (91%) of the total growing stock resides in commercially exploitable forests. Nowadays, the annual increment of the growing stock is 104 mill m3. It has steadily increased since the 1970s from the level of 60 mill m3. At the same time, the total drain has continuously remained at lower level than the annual increment. Currently, the total drain amounts only 68% of the annual increment of growing stock (Ylitalo 2012). This unused potential means that the industrial utilization of wood could be increased. Despite the abundant forest resources, there are trends in the operational environment resulting in considerable challenges for domestic wood supply chain. These trends affect forest resources and forest management, wood supply and wood markets, forest industry, forest policy, forest ownership and multiple-use of forests. The current forest resources and forest structure have influenced by the management practices of the past. The structure of Finnish forest has changed significantly during the past 80 years. Forest management has aimed at increasing wood production in commercial forests by emphasising intensive silviculture in even-aged stands of coniferous tree species. Draining of peatlands for wood production especially in the 1960s and 1970s has notably increased the amount of growing stock and affected tree species composition on peatland forests. These investments in forest management have resulted in increased cutting possibilities. However, from 1990’s there has been ongoing trend in the structure of fellings, which is likely to continue, and even strengthen in the future. The increased cutting potential is not in most cost-efficiently harvestable stands, that is final fellings, but instead in intermediate fellings, and fellings in peatland forests, which are not so cost-efficient to harvest. New management practices with increasing interest in un-even aged forest management will possibly increase in future. Over 60% of Finland’s commercial forest are owned by non-industrial private forest owners (Forest Finland in brief 2011).The average size of these altogether about 347 000 small scale Working Papers of the Finnish Forest Research Institute 302 http://www.metla.fi/julkaisut/workingpapers/2014/mwp302.htm 8 family forest holdings is 30.3 ha (Leppänen and Sevola 2013). Roughly 70% of forest owners are over 55 years old (Hänninen et al. 2011), which gives rise to a gradual change in the structure of private forest ownership. This is likely to lead into the situation where the number of forest owners is increasing while the average size of forest holdings is decreasing. Forest incomes are more important to younger that to elder forest owners but, at the same time, more significant to owners of large forest properties than small ones. The small scale forestry has also effect on stand management practices, the small management units lead low efficiency in management and operations and high unit costs. For many future forest owners, forest management has other priorities than wood production, which leads to favouring of less wood production oriented forest management. Overall this may lead to the situation where only some forest owners focus on wood production. Changes in forest ownership altogether may induce volatility and other negative impacts on wood markets and delayed or even neglected silvicultural operations and other stand management practices. The above mentioned trends in operational environment of industrial wood supply require new kind of thinking and novel solutions. The situation can be seen as great potential where the Finn- ish forests meet the different needs of society. Forests are the basis for wood based products, renewable energy and multi-use and protection of forests. All these aims can be met with inten- sified forest management. Overall this increases the income, job opportunities and welfare of the society. On the other hand, these also give room to more professional forest ownership. One solution to combine these different aims is to consider more intensive, cost-efficient and sustain- able management and logging measures in those forest areas, where prerequisites for commercial wood and biomass production are more favourable. 1.2 Scope and the aim of the work The objective of this work was to assess the potential, cost-efficiency and impacts of intensive management of Finnish forest resources in order to provide high quality raw material for forest industry. The aim was to identify the most cost-efficient, ecologically sustainable, and feasible ways to increase the production of domestic biomaterial with high utility value for current and future forest industry. This work comprises large-scale scenarios of the development of forest resources under varying intensity of management in commercially exploitable forests in Finland. The potential and requirements of different end-users of forest-based raw-material (such as pulp and paper, bioenergy, saw and veneer industry, other wood products) and the role of forest management and wood supply was emphasized when constructing and analysing the scenarios. Working Papers of the Finnish Forest Research Institute 302 http://www.metla.fi/julkaisut/workingpapers/2014/mwp302.htm 9 2 Description of scenarios 2.1 Definition process The “scenarios” can be implemented as visions or aspects of possible future. Scenarios are not predictions about the future but rather simulations of some possible futures. Traditionally scenarios are seen as qualitative method to analyse future alternatives. In parallel there have been also quantitative scenario processes. Recently Amer et al. (2013) reviewed the scenario planning literature. They conducted the conclusion that combining qualitative and quantitative scenario methods will provide more robust scenarios. In this work we used scenarios as tools for analysing and understanding the effects of key competitive decisions on forest resource management, and how the forest resources in the future depend on different management strategies. The used scenario definition process is based on action scenario process described by Meristö et al. (2000). At first phase of the scenario process, the key points are to define the current situation and identify important factors in operational environment that should be taken into account in the analyses. Thereafter, the alternative possible future scenarios can be defined. This part of our research was based on qualitative analysis (Fig. 1), in which we examined the development of Finnish Forest resources and wood production on each scenario. The predictions were based on several stand management practices based on implications of the defined scenarios. Quantitative analyses will be defined more detailed in the next chapter. In scenario definition process, the alternative scenarios were implemented by the project group with members from both industrial and research partners of EffFibre program. The group had several meetings during the different stages of planning process. In addition, a one-day workshop addressing the properties of detailed scenarios was organized gathering together project group members, experts of pulp industry and research, forest ownership research, and climate policy and carbon issues . In the beginning of planning process of scenarios, project group discussed the current operational environment of forest sector, and mapped the most significant ongoing trends (Fig. 1). The discussion led to elaboration of SWOT-analysis. The strengths, weaknesses, opportunities and threats of forest sector were analysed. Evaluation was elaborated from three different viewpoints focusing wood and biomass production, wood and biomass supply and availability and utilization of wood and biomass. SWOT-analysis was the starting point for scenario definitions. Based on SWOT analyses, research group of Metla sketched the first draft of the scenarios (Fig. 1). These were further elaborated in a workshop. As an output of the workshop, a detailed and structured list of key factors, properties and assumptions under different scenarios were listed. As a result of the planning process, the following scenarios were agreed to be assessed in the project 1. Business as usual (BAU) 2. Active forest sector and intensive biomass production (INT) 3. High quality raw material production for forest industry and bioenergy (QLTY) 4. Extensive forestry due to decreasing activities of forest industry – increasing non-material services (EXT) Working Papers of the Finnish Forest Research Institute 302 http://www.metla.fi/julkaisut/workingpapers/2014/mwp302.htm 10 Basic assumptions common to all scenarios were that the area of protected forests will remain at least at the current level and the international agreements and commitments for climate change mitigation hold also in the future. The scenarios were defined from different viewpoints: forest industry and production, management of forest resources, timber sales, procurement and logistics, forest policy and forest owners and ownership. The viewpoints are used in the next chapters when presenting the identified key factors of the scenarios. 2.2 Scenario 1: Business as usual (BAU) Overview Scenario was based on the assumption that the current situation in forest sector and in wood supply will not change, except for the well-known ongoing trends and changes in forest sector. These changes are assumed occur as widely expected today. In this scenario, it was assumed that the intensity of forest management practices, as well as the current levels of annual commercial fellings will remain at current level. Forest industry and forest-based production There will be no major changes in the selection of forest-based products. The share of paperboard production will strongly increase, and chemical pulp and tissues will slightly increase while the proportion of mechanical pulp and paper production will decrease. The share of saw timber will remain at the current level. The number of large scale enterprises will remain more or less constant, but the number of small and medium size enterprises will slightly decrease. The degree of integration within forest industry will increase. Number of bio-refineries will increase as well as the amount of bioenergy production. Management of forest resources Prevailing forest management and utilization will extend also in the future. Hence, the intensity of silvicultural practices, and the volumes commercial fellings will remain at current level. This implies that the annual areas of completed silvicultural practices, especially tending of young stands, are much lower than recommended areas. In forest regeneration, artificial regeneration maintains its dominant position as a regeneration method. The use of improved regeneration material in artificial regeneration increases slightly. In commercial fellings, the proportion of commercial thinnings and fellings on peatland forests will increase. Recovery of biomass for bioenergy, that is logging residues and stumps in final fellings, and especially small-sized trees in the first thinnings, will continuously increase. Timber trade, procurement and logistics In timber trade, selling by timber assortments will remain as the prevailing method, but timber assortments will be defined in a more flexible manner than today. There will be more buyers in the energy wood markets. Domestic wood supply will be supplemented with imported wood. In timber procurement, the share of thinnings and loggings on peatland forests will increase. The role of extended entrepreneurship will strengthen in wood procurement, and the average company size will increase. In addition, supply of harvesting, regeneration and stand management services Working Papers of the Finnish Forest Research Institute 302 http://www.metla.fi/julkaisut/workingpapers/2014/mwp302.htm 11 for forest owners will increase. Exchange of timber assortments between companies will increase. In road network, the proportion of unsound forest and local roads will increase. At the same time, other forms of wood and biomass transport will be developed. Forest policy There will be no significant changes in forest legislation, neither in the principles and amount of state subsidies to forestry. Taxation practices will remain more or less unchanged. International agreements and commitments do not require any major changes to forest management. Price of emission allowances will remain at current level. Forest owners and ownership Current structure of forest ownership will prevail in near future. No significant changes will occur in forest owners’ objectives of forest management. Current attitude to multi-use of forests will remain at current level; proportion of multi-objective owners will remain at one third of forest owners and one half of forest holdings land area. Joint ownership of forests will slowly become more common. Wood supply from forests owned by forest industries and state will remain at current level. On one hand, the structure of forests will generate increasing wood supply, but on the other hand, a part of forest owners will become less active in selling timber. 2.3 Scenario 2: Active forest sector and intensive biomass production (INT) Overview Scenario is based on the assumption that vitality of forest sector will markedly improve. Increasing business leads to increasing demand for domestic wood and biomass. Measures of intensive wood and biomass production are widely applied. Forest industry and forest-based production Overall, the future for forest industry seems optimistic; brand of business is competitive, and profitability is high, which contributes increase of investments. Wide diverse of wood based products have large demand. Vital forest and energy industry generates increasing demand for domestic raw material. In forest industries, degree of integration will markedly increase. The number and capacity of bio-refineries will rapidly increase. The production of paperboard will strongly increase, as well as the production of wood-based panels, such as veneer sheets and plywood. The production of tissues, chemical pulp, and sawn goods increases only slightly. In addition the production of chemicals, other materials such as plastics and composites increase significantly. The role of forest-based bioenergy in energy production will increase notably. Management of forest resources Common trend in forest management is differentiation of forest management according to goals of forest owners. In commercial forests, measures of more intensive wood and biomass production will take place. In forest regeneration, artificial regeneration, especially planting with genetically improved material will be prevailing method. Cost-efficiency of silvicultural practices will improve through mechanization of silvicultural operations leading to increased areas of, for example, tending of sapling stands and pre-commercial thinnings. Correspondingly, intensity of intermediate thinnings will increase, and forest fertilization becomes a common measure to enhance wood and biomass production in commercial forests. The early and intensive thinnings Working Papers of the Finnish Forest Research Institute 302 http://www.metla.fi/julkaisut/workingpapers/2014/mwp302.htm 12 and increased fertilization lead to shorter rotations. Integrated recovery of pulp wood and energy wood in first commercial thinnings, and recovery of biomass of logging residues and stumps in final fellings for energy will increase. Thus, the recovery of energy wood will markedly increase. Timber trade, procurement and logistics In timber trade, selling by timber assortments will be, at least partly, replaced by pricing systems based on value yield of the respective end-products. Domestic wood supply will be supplemented with imported wood. The average size of cutting areas and cutting removals increases. The role of extended entrepreneurship will strengthen in wood procurement, and company size will increase. More efforts will be put to maintain the road network. Accordingly, other forms of wood and biomass transport will be developed as well. Forest policy Forest legislation and forest policy will be developed to be more supportive to sustainable and cost-efficient wood and biomass production. State subsidies are directed to promote young stand management and energy wood recovery and production. Taxation will be converted to stimulate wood production and to create possibilities to use forest resources enabling economies of scale. Price of emission allowances will remain approximately at the current level. EU climate policy targets to increasing bioenergy production. Forest owners and ownership Concentration will be prevailing trend in private forest ownership structure. The number of forest owners will decrease, and average size of forest holdings will increase. Further, new forms of forest ownership will become more common. Willingness to active forest management and timber sales will increase vigorously among private forest owners leading to increased supply of wood and biomass. The forest management strategy of private forest owners will diversify according to diverse management goals. The proportion of multi-objective owners increases vigorously and, at the same time, the proportion of recreationists and indifferent owners decrease significantly. 2.4 Scenario 3: High quality raw material production for forest industry and bioenergy (QLTY) Overview Scenario is based on the assumption that volume and vitality of forest industry in Finland will remain at least at the current level, but structure of forest industry will change. Wood products and energy industries will strengthen at the expense of mechanical pulp and paper industries. There will be an increasing demand for high quality raw material of forest industry, especially wood products. Thus, the management of forest resources is based on combined energy wood and timber production, and overall activity of management increases. Wood quality aspects are emphasized in forest management, which promotes thinnings for quality and longer rotations. Forest industry and forest-based production The structure of forest industry will change. The number of large scale plants will decrease but, on the other hand, the small and medium size enterprises in wood products industry will increase significantly. In forest industries, degree of integration will decrease. The number and capacity of bio-refineries will rapidly increase. The production of paper and mechanical pulp will decrease. On the other hand, the production of saw timber and wood-based panels (veneer sheets and Working Papers of the Finnish Forest Research Institute 302 http://www.metla.fi/julkaisut/workingpapers/2014/mwp302.htm 13 plywood) will strongly increase. However, chipboard production is going to decrease. The role of forest-based bioenergy will strongly increase in energy production. Management of forest resources Forest regeneration areas will decrease due to forest structure and longer rotations. In forest regeneration, planting of spruce will remain at high level, as well as sowing of pine stands. The usage of improved and vegetatively propagated regeneration material increases slightly. In young stand management, silviculture aiming at combined production of timber and energy wood will become more common. It results in increased mean density of young stands. Therefore, the first commercial thinnings will be either energy wood thinnings or integrated pulpwood and energy wood thinnings. In commercial thinnings, more emphasis will be paid on stem quality in tree selection. Thinnings from above will become more common leading to longer rotations. Advanced thinning stands are fertilized. Usage of nitrogen fertilization will increase in order to compensate nutrient loss caused by energy wood thinnings. Timber trade, procurement and logistics In timber trade, the number of buyers’ of logs and energy wood increase. The proportion of energy wood increases and the proportion of pulp wood decreases with respect to the total removal of thinnings during rotation. Internet-based wood markets gain space substantially. Further, exchange of timber assortments between companies will increase. Number of round wood assortments will increase, including also more energy wood assortments. Pricing systems of timber assortments will be based more strictly on the value yield of the end products. Average transport distances of wood will decrease. Climate and forest policy will increasingly encourage to carbon sequestration, increased use of renewable raw materials and bioenergy, and maintaining of forest biodiversity. State subsidies to forestry will be targeted more precisely to promote desired activities. Price of emission allowances will remain at current level. Forest owners and ownership Concentration will be a prevailing trend in private forest ownership structure. The number of forest owners will decrease, and average size of forest holdings will increase. New forms of forest ownership will become more common. Willingness to active forest management for quality timber and sell timber will increase among private forest owners leading to increased supply of timber and energy wood. The forest management strategy of private forest owners will diversify according to diverse management goals. 2.5 Scenario 4: Decreasing activities of forest industry – increasing non- material services (EXT) Overview Scenario is based on an assumption that the global trends affecting forest and energy industries will decimate the profitability of forest industry in Finland. Especially the volume of pulp and paper industry decreases significantly. Consequently, there will be less demand for domestic industrial wood. In forest management, more emphasis will be put on protection and forest externalities. Working Papers of the Finnish Forest Research Institute 302 http://www.metla.fi/julkaisut/workingpapers/2014/mwp302.htm 14 Forest industry and forest-based production In forest industries, both the number of plants and the average size of companies will decrease. The degree of integration will strongly decrease. Throughout the forest industry, production volumes will decrease. Especially, the production of mechanical pulp will decrease to marginal level, and paper and chemical pulp production will decrease as well. Within wood product industries, especially chipboard production is going to strongly decrease or cease completely. The role of forest-based bioenergy will stay on the current level. The brand of the business will be focused on domestic small-scale industry. The relative importance of ecosystem services will increase. Management of forest resources More extensive forest management will take place. Areas of forest regeneration will decrease due to forest structure, longer rotations, and increased popularity of low-impact forestry. Along with the regeneration areas, the areas of artificial regeneration will decrease as well. Natural regeneration gains space over planting and sowing. Extensive, low-cost forest management practices will be more popular. Areas of intermediate thinning will decrease due to poor demand and low price of pulp-size wood. Rotations will be longer. The popularity of uneven-aged forestry will strongly increase among private forest owners. Permanent and temporary protection of private forests will increase. Timber trade, procurement and logistics In timber trade, the number of local buyers of logs and energy wood increases. Pricing according to the work site will increase. More robust pricing systems will replace the current system based on timber assortments. In wood procurement, part-time entrepreneurship will be more common. Logging conditions will get worse due to extensive forest management. Wood transport conditions are affected by worsening condition of road network. Average transport distances of wood will decrease due to increased local use of timber and biomass. Forest policy Climate and forest policy will be more rigid with respect to utilization of forest resources leading to new restrictions on forest management. Climate policy emphasizes importance of carbon sequestration management. On the other hand, forest protection will get more emphasis, and will be encouraged by state subsidies. As a result, extensive management becomes more common. Forest owners and ownership The number of forest owners will increase, and the average size of forest holdings will decrease. Due to increasing emphasis on immaterial goods on one hand, and poor timber markets on the other hand, forest owners have no motivation to expand their forestry business and increase the size of their forest properties. For increasing number of forest owners, wood and biomass production is not anymore on the top of the list of their forest management priorities. Working Papers of the Finnish Forest Research Institute 302 http://www.metla.fi/julkaisut/workingpapers/2014/mwp302.htm 15 3 Calculation of scenarios 3.1 Overview of calculation process In a nutshell, the following procedure was applied in the calculation of the scenarios ( ). Data from the 10th Finnish National Forest Inventory were applied as initial data and starting point of the simulations. Measurement data from sample plots of NFI10 located on forest land in commercial forests were used to represent current forest resources. For each plot, a set of pre-defined management regimes reflecting the management principles of a given scenario was simulated with MOTTI-stand simulator (Hynynen et al. 2005, Salminen et al. 2005) over 100-year time period. For simulations, seven alternative working scenarios (see chapter 3.4.) were defined. After the simulation stage, linear programming package J (Lappi and Lempinen 2013) was applied as a tool to select a management program for each scenario that met the given constraints. For a detailed description of the optimization procedure, see chapter 3.6 below. The final four scenarios were compilations of the one or more working scenarios (see chapter 3.7 below). The calculation was completed by Forestry Centres, which was also the lowest level at which the final results of scenarios are presented. Qualitative analysis Situation report: the current operational environment of the Finnish forest sector SWOT analysis Defining scenarios Background information for each scenario (scenario table) Guidelines for forest management regimes in different scenarios Detailed definitions for stand-level simulations Quantitative analysis Forest management schedules (with variations) for each scenario Plot 1 Plot 2 Plot 3 Plot n Predictions Linear programming Forestry statisticsSustainabilityNational Forest Inventory (NFI) Scaling Compiling working scenarios Simulation of pre-defined, alternative management regimes ( n ~ 50) for each plot with MOTTI-simulator Forest resources represented by sample plots (n > 40 000) of the NFI Combining working scenarios into final scenarios 765 4 Working scenario 1 BAU INT QLTY EXT 3 2 (a priori) Figure 1. Scenario process: a qualitative phase followed by a quantitative phase. Working Papers of the Finnish Forest Research Institute 302 http://www.metla.fi/julkaisut/workingpapers/2014/mwp302.htm 16 3.1.1 Description of data Data of this study were collected from altogether 46 297 sample plots of the 9th and 10th national forest inventory (NFI9, NFI10) of Finland that was carried out during the period from 2004 to 2008. Majority of the inventory data comprised of a systematic grid of NFI10 sample plots laid over the whole country, excluding the northernmost Lapland where the two-phased stratified sampling technique of NFI9 was applied (see Korhonen et al. 2006, Tomppo et al. 2009). In the NFI10, more than 100 variables were observed to describe the site, growing stock, damages and need for silvicultural operations. The sample plots were distributed over the Forestry centres and covered wide range of site fertility classes (Table 1). Table 1. Number of sample plots with respect to site fertility classes in Forestry Centres. Fertility class Center Soil type 1 (OMaT) 2 (OMT) 3 (MT) 4 (VT) 5 (CT) > 5 (Clt) Total 1 Heath 83 350 995 282 26 57 2233 Peat 14 116 177 86 44 3 2 Heath 43 470 1359 519 118 57 3162 Peat 10 126 197 163 97 3 3 Heath 184 996 995 212 12 28 2816 Peat 12 100 149 64 62 2 4 Heath 128 471 912 459 22 13 2375 Peat 20 54 130 111 54 1 5 Heath 117 628 1119 360 65 22 2817 Peat 9 62 165 158 108 4 6 Heath 127 1013 1458 465 22 15 3844 Peat 22 149 303 208 62 0 7 Heath 15 166 1126 792 129 20 3744 Peat 19 120 369 588 391 9 8 Heath 62 814 1527 660 57 19 3960 Peat 16 107 252 294 151 1 9 Heath 130 1022 1306 388 19 2 3808 Peat 33 156 339 285 125 3 10 Heath 81 604 1357 812 125 5 4143 Peat 24 119 335 397 279 5 11 Heath 19 110 1600 878 127 7 4102 Peat 34 125 295 643 263 1 12 Heath 50 136 1749 970 166 22 5304 Peat 59 264 566 999 322 1 13 Heath 6 57 1793 1101 156 10 3989 Peat 35 232 174 379 46 0 Total 1352 8567 20747 12273 3048 310 46297 Working Papers of the Finnish Forest Research Institute 302 http://www.metla.fi/julkaisut/workingpapers/2014/mwp302.htm 17 3.1.2 Converting NFI data into input data of MOTTI simulator NFI sample plot data were converted into the format compatible with MOTTI simulator,i.e., tree lists representing the stand data. The stand-level characteristics initially assessed in the field in young stands (D < 8 cm, Ddom < 10 cm) were the number of stems (N) and mean height (H). Other stand characteristics that were assessed only in advanced stands were basal area (G) and basal area-median diameter (DGM) and the corresponding height (HGM). In addition, dominant diameter (Ddom) and dominant height (Hdom) were calculated from measured or predicted characteristics of tally trees. Diameters at breast height were obtained for all relascope-sampled tally trees, whereas only every 7th tree was treated as a sample tree and measured for age, height, crown height, diameter and height increment (see Korhonen et al. 2006). Dominant tree characteristics were regarded practical input variables for models if at least four trees were measured. Otherwise, they were predicted. The NFI10 inventory data were used to convert stand characteristics into tree list for MOTTI simulator. Weibull distribution model (unpublished) that was fitted especially for NFI data, utilized G, DGM and Ddom as input variables for predicting diameter distribution. Tree heights and crown ratios of tally trees were obtained with the linear mixed-effects models by Eerikäinen (2009). Sapling stands having the dominant height of less than 8 m were not converted into tree lists. Instead, the vector of stand characteristics was completed and saved for their further development, respectively. The development predictions were based on the family of species-specific models for stand characteristics (see Siipilehto 2006) taking into account tending and thinning-effects of young stands (see Siipilehto et al. 2014). The NFI10 stand characteristics for young stands (N, H) were used as calibrating variables for the aforementioned family of models. In simulations, open areas were immediately regenerated according to guidelines presented in 3.2.1. The stand development after regeneration was simulated using the stand dynamic models of MOTTI (see Appendix 1). 3.2 Forest management assumptions, rules, and calculation parameters 3.2.1 General forest management principles Forest management schedule, i.e. a set of practices applied in a given stand, was derived from scenarios. Therefore, the intensity, timing and type of management practices varied according to scenario. However, in MOTTI-simulator, there are some general pre-defined principles and rules for a given silvicultural or logging operations. Silvicultural guidelines for private forests of Finland (Hyvän metsänhoidon suositukset 2006) were applied as standard management schedule in managed scenarios (BAU, INT and QLTY). However, when management schedules for a given stand were elaborated these guidelines were modified in order to be in line with management principles of a given scenario. Regeneration was simulated on initially open areas, and during the simulation after regeneration felling. Regenerated species and regeneration type were selected according to silvicultural guidelines for private forestry (Hyvän metsänhoidon suositukset. 2006) based on site fertility. Working Papers of the Finnish Forest Research Institute 302 http://www.metla.fi/julkaisut/workingpapers/2014/mwp302.htm 18 In general, the most fertile sites from OMT to MT were planted with Norway spruce. Sub-xeric, VT sites were seeded for pine whereas poorer sites CT and ClT were naturally regenerated for pine. During a simulation period of 100-years, most stands were regenerated once. MT sites are suitable for both Scots pine and Norway spruce as well. Therefore, when the given MT site was regenerated during simulation, it was regenerated with pine if pine was the main tree species in the former generation. As a result, drastic changes in the species structure were avoided. If there were other exceptions in species selection, they were due to different scenarios. Artificial regeneration: Soil preparation was always preceding planting and sowing. In intensive scenarios (INT, QLTY), improved regeneration material was assumed to have been applied in planting and seeding of Scots pine. Based on empirical evidence form progeny trials of Metla, it was assumed that genetic gain in growth was 7% in tree diameter and height growth (Haapanen and Mikola 2008). The implementation of genetic gain into the growth models is described by Ahtikoski et al. (2010). Natural regeneration: The number of seed trees left standing was fixed to 80 per hectare. Their volume and value was assessed at the time of the seed-tree cutting thereby neglecting their growth up to their final removal but at the same time increasing their net present value with any interest rate higher than 0%. This simplification was due to practical reasons; it enabled the removals and incomes to be technically allocated to the right rotation and not to the new tree generation. The volume of seed trees was underestimated but, as a compensating error, their net present value overestimated thus decreasing the total effect of the simplification. A light soil preparation was usually included in the operational chain of natural regeneration. Cleaning of a sapling stand: all the seedlings that are supposed to eventually develop into crop trees were left: remaining stem number was usually 3000–4000 seedlings per hectare Pre-commercial thinning: Aim was to provide adequate growing space to the best crop trees by removing all competitors. The stem number after pre-commercial thinning varied according to dominant tree species and geographical location, site and management scenario. Commercial thinnings: Timing, intensity and type of commercial thinnings varied according to management scenario, dominant tree species, site type, and geographical location. The possible thinning types were thinning from above, thinning from below and systematic thinning. Thinning guidelines for private forests of Finland (Hyvän metsänhoidon suositukset. 2006) were applied as a standard procedure. However, depending on the management scenario, these guidelines were modified into more intensive or extensive direction. In the first commercial thinnings, opening of strip roads were mimicked by removing 18% of stand basal area by systematic thinning. Recovery of energy wood: In thinnings, energy wood was recovered only in the first commercial thinning, either as integrated pulpwood and energy wood thinning or as energy wood thinning only. In combined pulp- and energy wood thinning, recovered energy wood included tops of harvested pulpwood stems and stems of those small trees that were undersized as a pulpwood. The minimum diameter of trees harvested as energy wood was 4 cm. In pure energy wood thinning, all the thinning removal was regarded as energy wood. In all thinnings, only stems were recovered for bioenergy meaning that all branches and foliage were left on site. When energy wood recovery was simulated in final fellings, the applied recovery rates were 100% for stem wood, 80% for branches, 60% for needles, and 70% for stumps. Working Papers of the Finnish Forest Research Institute 302 http://www.metla.fi/julkaisut/workingpapers/2014/mwp302.htm 19 Fertilization: On mineral soils, nitrogen fertilization was applied. The assumed type of fertilizer was ammonium nitrate. The fertilization dose varied from 160 to 180 kg N ha-1 depending on the scenario. In spruce stands, fertilizer was assumed to include also phosphorus in addition to nitrogen. On peatland forests, ash was used as a fertilizer. The assumed dose of ash was constant, equal to 4000 kg ha-1. Ditch maintenance: The need for ditch network maintenance (DNM) at peatland forests were predicted using model by Hökkä et al. (2000, model 3b, p. 4). Model predicts a stand level probability for ditch network condition being poor using the time from the previous ditching and site properties as driving variables. A need for DNM was alarmed if ditch network condition was predicted to be poor. It also resulted in a lower level of basal area growth. The actual DNM was scheduled together with the next thinnings or the final cutting. Growth response to DNM comprised both the growth shift back to normal level and, when DNM was applied for the first time, an additional growth reaction (Hökkä et al. 2000, Hökkä and Kojola 2002). 3.2.2 Economical parameters Forest management costs and stumpage prices Stumpage prizes and those silvicultural costs, which were presented in euros/ha, were based on nominal time series covering the years 1995–2010. Nominal costs and prices were then deflated according to cost-of-living index (base 1951:10=100, and year 2010 reflecting index value of 1751). Since statistics on separate stumpage prices for different thinnings (first thinning, other thinning, and final felling) have not been kept until during the recent few years, an additional comparison was carried out. First we set stumpage prices of final felling to 100% (reference level). Then we compared the real stumpage prices (i.e. deflated) of the first thinning and other thinnings to that of final felling, creating two ratios. Finally according to the two ratios the deflated original time series data were converted to the real stumpage prices for first thinning, other thinnings and for final felling (Table 2). Some of the silvicultural costs were expressed as a function of time consumption (see Table 2). Unit costs were based on averages of the year 2010 (obtained from several sources: e.g. private companies and public organizations). In regeneration, planting density (the number of planted seedlings) was based on the silvicultural guidelines for private forestry (Hyvän metsänhoidon suositukset. 2006). Working Papers of the Finnish Forest Research Institute 302 http://www.metla.fi/julkaisut/workingpapers/2014/mwp302.htm 20 3.2.3 Logging parameters The merchantable stem volume for logs and pulpwood was calculated using the assortment rules that are widely applied in Finland. The minimum length applied for pulpwood was 3.0 m, and the minimum top diameter over bark for Scots pine and broadleaves trees was 6.0 cm, and for Norway spruce 7.0 cm. The minimum log length was 3.1 m for Scots pine and broadleaved trees, and 3.7 m for Norway spruce. The maximum log length was 6.1 m for Scots pine and Norway spruce and for broadleaved trees 7.3 m. The minimum top diameter for log over bark was 20.5 cm for Scots pine, 21.5 cm for Norway spruce and 16.5 cm for birch. For birch this minimum value was constant. However, the minimum top diameter decreased progressively with increasing log length, being 14.5 cm for Scots pine and 16.5 cm for Norway spruce when the log length was 4.3 m or more. 3.3 Implementation of scenarios at stand-level / Simulation of management alternatives The basic principles and general properties of four future scenarios (BAU, INT, QLTY and EXT) are presented in Chapter 2. For calculations, detailed management descriptions were elaborated to be applied as simulation rules (Table 3) for each sample plot of NFI10 data (see Chapter 3.2). The following working scenarios (WS) were calculated for each stand (sample plot) WS1. Management according to silvicultural guidelines of Tapio WS2. Management without silvicultural practices in young stands WS3. Intensive management for effective wood and biomass production (INT) Table 2. Real (i.e. deflated) stumpage prices and silvicultural costs. STUMPAGE PRICES €/m3 Logs Pulp wood Energy wood Pine Spruce Birch Pine Spruce Birch First commercial thinning 46.3 43.9 43.1 15.5 20.2 14.8 3.0 Other commercial thinning 49.8 45.9 47.1 16.6 21.7 15.5 3.0 Final felling 57.5 53.7 52.8 19.3 26.7 19.1 3.0 REGENERATION COSTS SILVICULTURAL COSTS Labour cost of planting pine €/plant 0.147 Cleaning of sapling stand €/h 30.0 spruce 0.163 Pre-commercial thinning €/h 30.0 birch 0.183 Initial clearing of thinning area €/ha 200.0 Material costs of planting pine €/plant 0.179 Fertilization €/ha 204.5 spruce 0.199 Ditch network maintenance €/ha 156.5 birch 0.224 Seeding €/ha 193.1 Soil preparation mounding €/ha 283.0 disc trenching 174.6 scarification 174.6 Working Papers of the Finnish Forest Research Institute 302 http://www.metla.fi/julkaisut/workingpapers/2014/mwp302.htm 21 WS4. Intensive management for producing high quality raw material (QLTY) WS5. Low-cost management with one intermediate thinning WS6. Low-cost management without thinnings WS7. Unmanaged - no activities at all Within each scenario, several management alternatives were simulated for a given stand. These alternatives within a scenario were needed in order to allow flexibility for LP analysis (see chapter 3.6). Scenarios 1 2 3 4 5 6 7 BAU 1 BAU 2 INT QLTY EXT 1 EXT 2 EXT 3 Regeneration Tree species Norway spruce Norway spruce Norway spruce Norway spruce Norway spruce Norway spruce Norway spruce Method planting planting planting planting natural regeneration natural regeneration Density (N ha-1) 1800 1600 2000 2000 Soil preparation spot mounding scarification spot mounding spot mounding Early cleaning Timing, m 1.5 1.5 1.5 Growing density, (N ha-1) ca. 3000 ca. 3000 ca. 3000 Precommercial thinning Tree species selection (10% of growing birch mixture) (10% of growing birch mixture) Timing, (dominant height, m) 4 3 5 5 Growing density, (N ha-1) 1700 1600 2000 2000 First commercial thinning Method below integrated energy and pulp wood thinning below integrated energy and pulp wood thinning integrated energy and pulp wood thinning energy wood thinning Timing, (dominant height, m) 13 16 12 14 16 18 12 14 12 15 Growing density, (N ha-1) 1000 900 1000 900 800 700 700 1000 1100 900 Other thinnings Method below above below below above below above Timing, (dominant height, m) thinning guide lines thinning guide lines thinning guide lines thinning guide lines thinning guide lines thinning guide lines thinning guide lines Growing density, (N ha-1) thinning guide lines thinning guide lines thinning guide lines 350 thinning guide lines thinning guide lines thinning guide lines Final felling Mean diameter, cm 26/28/30 28/30/32 26/28/30 25/27/29 25/27/29 28/30/32 28/30/32 25/27/29 25/27/29 Stand age (yrs) 70/80/90 80/90/100 70/80/90 50/60/70 50/60/70 80/90/100 80/90/100 90/100/110 90/100/110 Recovery of logging residues and stumps YES/NO YES/NO YES/NO YES YES YES YES YES Fertilization Time of 1st fertilization (amount of Nitrogen) 5 years after first commercial thinning (180 kg ha-1 (N+P)) 5 years after first commercial thinning (180 kg ha-1 (N+P)) Time of 2nd fertilization (amount of Nitrogen) 5 years after second commercial thinning (180 kg ha-1 (N+P)) 5 years after second commercial thinning (180 kg ha-1 (N+P)) 5 years after second commercial thinning (160 kg ha-1 (N)) 5 years after second commercial thinning (160 kg ha-1 (N)) 6 6 24 3 3 3 3 3 3 1 No m an ag em en t Table 3. An example of management chains of each working scenarios in Southern Finland at fresh site type (MT) for Norway spruce dominated stands. Working Papers of the Finnish Forest Research Institute 302 http://www.metla.fi/julkaisut/workingpapers/2014/mwp302.htm 22 3.4 Simulation procedure The 100 year-long development of NFI-stands under defined scenarios was predicted by 5-year- periods using the models of forest dynamics (Appendix 1) and obeying the general forest management principles and parameters. Every stand had several optional management regimes resulting in as many alternative predictions within each scenario. The alternative stand-wise simulations were engineered in to batch processes that were computed by one forest centre at a time. The results were saved into six data sets; site properties, stocking, removals, incomes, costs and events. The amount of output data was reduced by aggregating individual observations whenever possible. For example, all incomes and costs were discounted to present values, and thereafter time-dimension of the original results could be reduced. Stocking, removals and events were saved by 5-year periods, which were later on summed up into 10-year results. The simulation results were pre-processed with the statistical software (SAS Institute 2011) that served as a main data-base of this study. The main tasks of pre-processing were to cross-check the data and to retrieve the required variables from the six output sets and compose a data set to be used in the next step, which was linear programming. In addition, some statistical measures were produced, mainly mean and cumulative values. 3.5 Linear programming After having simulated alternative management regimes within each scenario for each stand, they were congregated as a variable space. This variable space consisted of a total of over 14 million individual management regimes. However, for each forestry centre the variable space was smaller, depending on the initial number of NFI stands in that region. In general, linear programming is designed to solve efficiently planning problems (Lappi 1992). Principally, in linear programming alternative schedules are usually simulated (here the management regimes), and each schedule is associated with a vector of input and output variables over time (Lappi 1992). It is assumed that the goal(s) of the decision maker can be described as a linear programming optimization problem. For instance, decision maker may want to maximize the net present value of future incomes, subject to constraints such as constant annual incomes or minimum drainage per 10-year time horizons (Lappi 1992). In this study, we chose to maximize random number (instead of e.g. the net present value), emphasizing the constraints of the linear programming problem. These constraints played a crucial role in this study since through the constraints we could steer the aggregate outcome of numerous (simulated) management regimes to follow the principles of each working scenario (1–7). For optimization algorithm we constructed a specific control file in which e.g. annual cutting removals were restricted to follow a particular pattern in accordance with original working scenario. In this connection it should be emphasized that to some extent this procedure was ad hoc, depending on the initial structure of forests in each forestry centre. The first optimization task was carried out without any constraints. By this we could find out the underlying growth potential (which in turn indicates the potential cutting removals) of each forestry centres. After that the minimum removal for the first ten-year period was searched by an iterative process, leaving other 10-year periods intact. Having found the minimum cutting removal of the first 10 years, further constraints were formulated according to specific constraints such as similarity of annual cutting removals between all 10-year periods or minimum growing stock at the end of time horizon. Working Papers of the Finnish Forest Research Institute 302 http://www.metla.fi/julkaisut/workingpapers/2014/mwp302.htm 23 The underlying idea was to create such a control file that under the formulated constraints the optimum solution would resemble the initial working scenario as much as possible. The overall linear programming procedure is presented in the flowchart (Figure 2). Similar optimization package based on linear programming has been applied in earlier studies covering e.g. carbon sequestration issues (Matala et al. 2009) and peatland wood production (Nuutinen et al. 2000). The results of linear programming set the management regime for each stand in each working scenario. The results were imported back to SAS, and the final forest-centre results for each scenario were produced by cross-tabulating stand-wise figures. Simulated management regimes in forestry center i (i.e. variable space) Unconstrained optimization (blue bars) ...etc 765432Working scenario 1 Optimization under the third set of constraints Control file, version 3 Optimization under the second set of constraints Control file, version 2 Optimization under the first set of constraints (grey bars) Control file, version 1 Figure 2. Flowchart on the linear optimization procedure. The graph in the background is an illustration representing working scenario 1 of a forestry centres in which the unconstrained optimization resulted in a too large cutting removal (the leftmost blue bar for years 0-9). After the first constraint set the cutting removals were balanced (grey bars). Working Papers of the Finnish Forest Research Institute 302 http://www.metla.fi/julkaisut/workingpapers/2014/mwp302.htm 24 3.6 Compilation of the final scenarios The seven working scenarios WS1–WS7 were calculated using the simulation and optimization procedure described above. Working scenarios WS3 (Intensive management for effective wood and biomass production), and WS4 (Intensive management for producing high quality raw material) were used as such for final scenarios INT and QLTY (see chapter 2.4 and 2.5.). However, Business as Usual (BAU) scenario and scenario of extensive management (EXT) were compiled by mixing the working scenarios. The mixture of working scenarios was carried out for aggregated results presented in six data sets at forestry centre-level (see chapter 3.5.2). Business as usual scenario (BAU) was designed to reflect the situation, in which the most important silvicultural activities and annual cuttings will retain at current level also in the future. Young stand management practices, i.e. cleaning of sapling stands and pre-commercial thinnings, have crucial impact on the future development and the production potential of forests. According to assessments made in 10th National Forest Inventory (Korhonen et al. 2013), 53% of the area in the need of young stand management practices has been completed annually during 2001– 2010 on the average. Information available by forestry centres on completed vs. recommended treatment areas (Korhonen et al. 2013) was applied in compiling the BAU scenario with the help of the working scenarios. Further, the ratio between the annual cutting removals and maximum allowable cutting removal by forestry centres (see Table 4.xx in Statistical Yearbook of Forestry… 2011) was used as a measure of current degree of utilization of wood production potential. The applied percentages by forestry centres referring to treatment and cutting rates described above are presented in Table 4. BAU scenario was compiled by mixing the working scenarios in the following manner. The restriction related to current activity in young stand management was taken into account by Table 4. Statistics by Forestry Centres on percentages of the performed areas of young stand management practices with respect to recommended areas of 10th National Forest Inventory (Korhonen et al. 2013), and percentage of removed volumes with respect to maximum sustainable removals (Ylitalo 2013 Forestry Centre Percentage of completed young stand treatment areas of the recommended areas, % Percentage of actual cutting volumes of the maximum sustainable removals, % 1 41.5 65.9 2 53.5 71.1 3 51.7 87.5 4 64.8 85.1 5 49.3 71.1 6 66.2 88.3 7 53.0 78.6 8 61.6 84.2 9 58.2 80.4 10 61.3 77.9 11 67.9 74.9 12 41.9 84.3 13 41.9 67.4 Average 54.8 78.2 ). Working Papers of the Finnish Forest Research Institute 302 http://www.metla.fi/julkaisut/workingpapers/2014/mwp302.htm 25 setting the proportion of WS1 (management according to silvicultural guidelines of Tapio) in BAU scenario equal to the percentage of completed vs. recommended treatment areas by forestry centre (Table 4). The remaining proportion of BAU scenario was composed as the mixture of WS2, WS5 and WS7 working scenarios, in which young stand management practices were not applied. Restriction related to cutting removals was taken into account with the help of calculated mean annual removals of the working scenarios. It was assumed that annual removals of WS1 will reflect the maximum allowable volume of harvest removals (100% cutting level). The removals of BAU scenario by forestry centres were set to the percentages presented in Table 4 with respect to the removal of WS1. The mixture of working scenarios in BAU scenario, agreeing with the constraints presented above was obtained by solving optimization problem using Solver of MS Excel. The analysis resulted in the optimal mixture of WS1, WS2, WS5 and WS7. Scenario of Extensive Management (EXT) was based on the assumption that industrial utilization of forests radically decreases in the future. This decrease was assumed to result in forest management intensity in the following manner: –– 25% of forest area will be managed according to BAU scenario –– on 25% of forest area only one intermediate thinning and regeneration felling will be carried out during the rotation –– on 25% of forest area, only natural regeneration and final fellings will be completed –– 25% of forest area will be left unmanaged and outside commercial wood production EXT scenario was compiled by mixing of BAU scenario (25%) with WS5 (25%), WS6 (25%), and WS7 (25%) working scenarios. 4 Results and discussion 4.1 Forest management practices The forest management practices in the early years of rotation consist of silvicultural operations such as regeneration and precommercial thinning. In more advanced stands, the silvicultural operations, such as fertilization, and ditch maintenance and the commercial cutting operations are carried out. The intensity of forest management varied notably between the scenarios (Figure 3). The annual areas of forest regeneration were significantly higher in INT scenario compared to other scenarios or completed regeneration areas between 2001–2010. In INT scenario the rotations were shorter than in QLTY, BAU or current situation. Areas of young stand management showed the largest contrasts between the scenarios. In intensively managed scenarios (INT and QLTY) annual areas of pre-commercial thinnings were at much higher level than in BAU or EXT scenarios. In INT scenario, the treatment areas were double compared to those of BAU scenario, and triple compared to the actually completed areas. Accordingly, forest fertilization areas were highest in INT and QLTY scenarios. Working Papers of the Finnish Forest Research Institute 302 http://www.metla.fi/julkaisut/workingpapers/2014/mwp302.htm 26 In addition to differences in the treatment areas between the scenarios, there were also some temporal trends within scenarios, especially in the regeneration areas (Figure 4). Due to the current structure of forest resources, which was characterized by great proportion of mature forests, regeneration areas increased markedly during the first decade of the simulation period. The increase was greatest in INT scenario, in which shorter rotations were favoured more than in the other scenarios. 0 50 100 150 200 250 300 350 400 450 500 Forest regeneration Pre-commercial thinnings Fertilization Ditch maintenance BAU INT QLTY EXT Completed 2001–2010 1000 ha 0 300 600 900 1200 1500 Th ou sa nd s BAU 1000 ha INT 0 300 600 900 1200 1500 1–10 11–20 21–30 31–40 41–50 51–60 61–70 71–80 81–90 91–100 Th ou sa nd s QLTY EXT 1–10 11–20 21–30 31–40 41–50 51–60 61–70 71–80 81–90 91–100 Ditch maintenance Fertilization Pre-commercial thinnings Forest regeneration Figure 3. Annual area of different forest management practices 2010–2110 for different scenarios. Completed 2001-2010 refers to treatment areas based on Finnish forest statistics (Ylitalo 2013). Figure 4. Temporal variation of annual area of different forest management practices 2010–2110 for different scenarios. Working Papers of the Finnish Forest Research Institute 302 http://www.metla.fi/julkaisut/workingpapers/2014/mwp302.htm 27 Commercial cuttings include first commercial thinning, other commercial thinnings and regeneration fellings (Figure 5). In INT and QLTY scenario, the annual area of first commercial thinnings was clearly larger than in BAU and EXT scenario, as expected. The share of “other commercial thinnings” was greatest in QLTY scenario. This was driven by the management goal of QLTY scenario aiming at growing high quality large-size timber. In order to meet the goal, longer rotation and successive intermediate thinnings were applied. The temporal trend in cutting areas during the first decades of 100-year period was similar to that of regeneration areas, and arose from the age class structure of forests (Figure 6). After the sharp decrease in cutting areas after the first 10-year period, areas started to steadily increase in intensively managed scenarios, and especially in QLTY scenario due to increase in the areas of intermediate thinnings. Figure 5. Mean annual area of cuttings during 2010–2110 for different scenarios. Completed 2001-2010 refers to cutting areas based on Finnish forest statistics (Ylitalo 2013). Figure 6. Temporal variation of annual area of cuttings during 2010–2110 for different scenarios. 0 100 200 300 400 500 600 700 800 BAU INT QLTY EXT Completed 2001–2010 Regeneration fellings Other commercial thinnings First commercial thinning 1000 ha EXT 0 20 40 60 80 100 120 140 BAU Energy wood Pulpwood Logs mill m3 a-1 INT 0 20 40 60 80 100 120 140 QLTY 1–10 11–20 21–30 31–40 41–50 51–60 61–70 71–80 81–90 91–100 1–10 11–20 21–30 31–40 41–50 51–60 61–70 71–80 81–90 91–100 Working Papers of the Finnish Forest Research Institute 302 http://www.metla.fi/julkaisut/workingpapers/2014/mwp302.htm 28 Intensive management requires considerable investments in silviculture in order to maintain or increase cutting removals of high quality wood. Results strongly suggest that due to current structure of forest resources, the main focus in Finnish forestry today and in the near future should be in increased regeneration and silvicultural practices of young stands, if the goal is to maintain or increase wood and biomass production in the long run. Young stand management is one of the most obvious bottlenecks of forestry that needs to be tackled. 4.2 Harvesting removals Intensive forest management enables a significant potential to increase annual removals in a sustainable manner. The annual removals increased ca. 40% in intensive management scenarios (INT and QLTY) (Figure 7) compared to the current level of removals. As expected, the EXT scenario led to the lowest removals. BAU scenario resulted in only slightly greater removals compared with EXT scenario. The result indicated that the intensity of current forest management in commercial forests is actually at rather low level. The annual removals of logs were highest in QLTY scenario, but the difference was small compared to INT scenario. The energy wood removals consisted only of small-size stem wood below the pulpwood dimensions and logging residues and stumps from final fellings. Despite this rather strict definition set to energy wood, the annual energy wood removal was nearly 13 mill m3 in INT scenario and 10 mill m3 in QLTY scenario (Figure 7). The temporal variation of removals showed the similar pattern to cutting areas, as expected. There was only small variation in the cutting removals of BAU and EXT scenarios throughout the 100- year simulation period, but temporal trends occurred in the removals of intensive scenarios, INT and QLTY (Figure 8). In the beginning of calculation period, removals were at higher level in these scenarios compared to following 10-year periods due to high harvesting reserves. Further, in INT and QLTY scenarios, harvesting removals started gradually increase during the second half of 100-year period. The increasing trend in harvesting removals towards the end of 100-year period was the result of increased intensity of forest management activities (Figure 8). Figure 7. Mean annual harvesting removals 2010–2110 by timber assortments for different scenarios. Removals 2001-2010 refers to completed removals based on Finnish forest statistics (Ylitalo 2013). 0 10 20 30 40 50 60 70 80 90 100 BAU INT QLTY EXT Removals 2001–2010 Energy wood Pulpwood Logs mill m3 Working Papers of the Finnish Forest Research Institute 302 http://www.metla.fi/julkaisut/workingpapers/2014/mwp302.htm 29 In order to assess the economic potential of the scenarios at national level, the gross stumpage earnings of each scenario were calculated applying the stumpage prices, which were presented in chapter 3.2.2. Annual cutting removals in BAU scenario with 51 mill m3 refers to 1807 mill. euros in gross stumpage earnings (Figure 9). Applying INT scenario, the annual gross stumpage earnings could be increased by 57% compared to BAU, being 2847 mill. euros on the average. In all the scenarios the temporal variation of gross stumpage earnings was quite similar to that of annual removals. (Figure 9). EXT 0 20 40 60 80 100 120 140 BAU Energy wood Pulpwood Logs mill m3 a-1 INT 0 20 40 60 80 100 120 140 QLTY 1–10 11–20 21–30 31–40 41–50 51–60 61–70 71–80 81–90 91–100 1–10 11–20 21–30 31–40 41–50 51–60 61–70 71–80 81–90 91–100 Figure 8. Temporal variation of mean annual removals (mill m3a-1) 2010-2110 by timber assortments for different scenarios. Figure 9. Temporal variation and average of gross stumpage earnings during 2010–2110 for different scenarios. 2010 present the completed level of gross stumpage earnings at year 2010. 500 1000 1500 2000 2500 3000 3500 4000 BAU mill € a-1 INT QLTY EXT 0 500 1000 1500 2000 2500 3000 3500 4000 0 1–10 11–20 21–30 31–40 41–50 51–60 61–70 71–80 81–90 91–100 Average2010 1–10 11–20 21–30 31–40 41–50 51–60 61–70 71–80 81–90 91–100 Average2010 Working Papers of the Finnish Forest Research Institute 302 http://www.metla.fi/julkaisut/workingpapers/2014/mwp302.htm 30 In a country with large geographical and climatic variation, there are notable regional differences in wood production potential, and conditions for forestry and forest management. The annual harvesting removals varied significantly by Forestry centres because of differences in forest area, structure of forests and climatic conditions (Figure 10, Figure 11). However, throughout the country, the comparison of scenarios showed the great potential to increase the annual removals with more intensive forest management (INT and QLTY scenario). In some regions the annual harvesting removals are at nearly same level in BAU scenario as in EXT scenario. In Northern Finland (forestry centres 11–13), the difference between INT and BAU scenarios in terms of the mean annual removal per hectare was 1.2 m3ha-1a-1, whereas the difference in Southern and Central Finland (forestry centres 1–10) was 2.7 m3ha-1a-1, respectively. Despite the fact that intensive management in Northern Finland resulted in much smaller gain in removals per hectare (Figure 11), the increase in terms of the total removal of the whole forestry centre was at the same level as in southern parts of the country due to large areas of commercial forests in Northern Finland (see Figure 10). Figure 10. Mean annual harvesting removals (mill m3) by Forestry centres on the average during the years 2010–2110 by timber assortments for different scenarios. 0 1 2 3 4 5 6 7 8 9 BAU Energy wood Pulpwood Logs mill m3 INT EXT 0 1 2 3 4 5 6 7 8 9 QLTY Ra L-S H-U Ka-S Pi E-Sa E-Po Ke-S P-Sa P-Ka Ka P-Po La Ra L-S H-U Ka-S Pi E-Sa E-Po Ke-S P-Sa P-Ka Ka P-Po La Working Papers of the Finnish Forest Research Institute 302 http://www.metla.fi/julkaisut/workingpapers/2014/mwp302.htm 31 4.3 Growing stock During the 100-year simulation period the volume of growing stock increased in BAU and EXT scenarios compared to the current level of growing stock (Figure 12). Intensive management scenarios resulted in lower stocking levels. INT scenario led to the lowest standing volumes of the growing stock due to shorter rotations and intensive intermediate thinnings. In Finland, like in many other European countries, the growing stock has increased during decades because the annual drain has been below the annual increment of growing stock. During the years 2001–2010, ca. 78% (Table 4) of the maximum sustainable removal was actually removed in cuttings resulting in constant increase of growing stock volumes. In BAU scenario it was assumed that the ratio between annual removals and maximum sustainable removals will remain at the current level also in the future. Thus, the results showed that the volume of growing stock will increase from the current volume of 2 billion m3 gradually up to the level above 3 billion m3 by the end of the 100-year calculation period in BAU scenario (Figure 12). The EXT scenario followed nearly same temporal pattern in growing stock than BAU scenario. In EXT scenario, low level of annual removals promoted the increase of accumulation of wood in forests more than in BAU scenario. By the end of calculation period, growing stock in EXT scenario is expected to be only 7% greater than in BAU scenario (Figure 12). The slight difference was due to extensive forest management resulting in lower growth, especially in young stands, compared to that of BAU scenario. Scenarios of intensive management (INT and QLTY) resulted in lower stocking levels than BAU and EXT scenarios. The most significant reduction occurred during the first decades of the simulation period due to intensive cuttings (Figure 12). In INT scenario, there were 20% lower 0 1 2 3 4 5 6 7 8 9 BAU m3 ha-1 INT 0 1 2 3 4 5 6 7 8 9 QLTY Ra L-S H-U Ka-S Pi E-Sa E-Po Ke-S P-Sa P-Ka Ka P-Po La EXT Ra L-S H-U Ka-S Pi E-Sa E-Po Ke-S P-Sa P-Ka Ka P-Po La Energy wood Pulpwood Logs Figure 11. Mean annual harvesting removals per hectare (m3 ha-1) by Forestry centres on the average during the years 2010–2110 by timber assortments for different scenarios. Working Papers of the Finnish Forest Research Institute 302 http://www.metla.fi/julkaisut/workingpapers/2014/mwp302.htm 32 volumes of growing stock compared to QLTY scenario because of shorter rotations and intensive commercial thinnings in INT scenario. However, the intensified silviculture resulted in increasing stocking levels after 20 to 30 years in both INT and QLTY scenarios. Results showed that INT and QLTY scenarios ensured sustainable long-term wood and biomass production despite increased removals during next 100 years (Figure 12). The regional variation in growing stock showed that in Southern Finland the total volume of growing stock, in terms of mill. m3, was lower compared to Northern Finland but on the other hand, the growing stock per hectare, in terms of m3 per hectare, was higher in Southern Finland (Figure 13). The differences in growing stock per hectare with BAU and INT scenario were greatest in Southern Finland (Figure 13). Figure 12. Volume of the growing stock by 10th Finnish National Forest inventory (NFI10), and temporal variation of growing stock 2010-2110 mill m3 by scenarios. QLTY EXT INT 0 500 1000 1500 2000 2500 3000 3500 BAU NFI10 1–10 11–20 21–30 31–40 41–50 51–60 61–70 71–80 81–90 91–100 NFI10 1–10 11–20 21–30 31–40 41–50 51–60 61–70 71–80 81–90 91–100 mill m3 0 500 1000 1500 2000 2500 3000 3500 Working Papers of the Finnish Forest Research Institute 302 http://www.metla.fi/julkaisut/workingpapers/2014/mwp302.htm 33 4.4 Carbon stock and removed carbon Forest management has well-known impact on the magnitude of carbon sequestration to forests. With extensive management, more carbon will be stored in forest biomass, as the results of this scenario analysis also confirmed (Figure 14). The carbon stock increased in business as usual (BAU) and extensive management (EXT) scenarios during next 100 years by 61% and 76%, respectively (Figure 14). In intensive management scenario (INT), carbon storage in living and dead biomass remained at the level of 55% from that of EXT scenario. The aim in QLTY scenario was to produce large size logs by applying extended rotations. Thus, amount of carbon in forest was, on the average, 23% greater than in INT scenario. Compared to the carbon storage in the beginning of calculation period, no significant loss of carbon sequestration occurred in QLTY scenario, although the amount of recovered carbon in cuttings was 53% greater compared to current removals. The results also indicated that in the long run, intensively managed forests were capable to capture more carbon from the atmosphere than extensively managed forests. Already within 40 years the amount of captured carbon in QLTY BAU EXTQLTYINT m3 ha-1 60–80 81–100 101–120 121–140 141–160 161–180 181–200 201–220 A 70–110 111–150 151–190 191–230 231–270 271–310 311–350 351–390 mill m3BAU EXTQLTYINT B Figure 13. Mean growing stock per hectare (A) m3 ha -1 and (B) mill m3 2010–2110 by Forestry Centres for different scenarios. Working Papers of the Finnish Forest Research Institute 302 http://www.metla.fi/julkaisut/workingpapers/2014/mwp302.htm 34 scenario exceeded that of EXT scenario. Because of intensive management, majority of captured carbon was bound to biomass removed from forests in logging operations. Thus, climate impacts finally depend on how harvested wood and biomass will be utilized, i.e. for how long time carbon will be stored in the products, and to what extent forest-based raw materials and energy products will substitute non-renewable materials and fossil-based energy. Finally, in the assessment of climate impacts, the question of time span of the analysis is crucial. 4.5 Profitability of forest management Profitability of scenarios was analysed by calculating net present values of future incomes during 100-year calculation period. In the analysis, current stumpage prices and management costs (see chapter 3.2.2.), and discount rates from 1% to 5% were applied. Results showed clearly that intensive forest management revealed superiority of in terms of profitability compared to business as usual management (BAU), or extensive management (EXT) (Figure 15). In INT scenario, present values of net incomes were highest being ca. 1.5-times greater than in BAU scenario in spite of the applied discount rate. Profitability of INT and QLTY scenarios were equal with 1% discount rate. With higher discount rates, profitability of INT scenario was higher than in QLTY. The explanation for the difference is that the higher discount rate promotes the effect of short rotations and intensive thinnings. Thus, INT scenario resulted in the highest profitability from 2% to 5% discount rate. 0 500 1000 1500 2000 2500 3000 BAU mill Mg INT 0 500 1000 1500 2000 2500 3000 QLTY EXT Carbon in logs and pulpwood Carbon in energy Carbon in logging residues and mortality Carbon in living biomass 1–10 11–20 21–30 31–40 41–50 51–60 61–70 71–80 81–90 91–100 1–10 11–20 21–30 31–40 41–50 51–60 61–70 71–80 81–90 91–100 Figure 14. Temporal variation of carbon stock 2010-2110 mill Mg by scenarios. Working Papers of the Finnish Forest Research Institute 302 http://www.metla.fi/julkaisut/workingpapers/2014/mwp302.htm 35 Results showed the large variation in profitability between geographical regions (Figure 16). The regional analysis revealed a clear decreasing trend in profitability from south to north. Further, between the scenarios the absolute differences in profitability at 3% discount rate was much greater in Southern and Central Finland compared to Northern Finland. This was due to differences in climate, production potential of forest sites and the structure of forests. The differences between Forestry centres were also considerable in management costs, incomes and net incomes (Figure 17). The management costs were lower in Northern Finland due to the greater proportion of infertile sites, where less intensive and more affordable silvicultural practices (e.g. regeneration methods) could be applied. The highest management costs were found in intensive forest management scenarios, INT and QLTY, but similarly the incomes were also highest. On the average, INT scenario resulted in 144% greater net incomes per hectare compared to BAU scenario. Profitability of scenarios was calculated applying current stumpage values and management costs. Applying today’s prices and costs in 100-year scenarios was general assumption and may not predict well the situation in future. For example, management costs may differ from the current situation in the future because of the increased mechanization of management practices. Thus, the results on profitability solely reflected financial potential of alternative management strategies in current market situation and operational environment. Based on results, in current situation intensive forest management is profitable strategy. In the long run, it is obvious that increased wood supply can only be realized with favourable market conditions and increased demand for wood. However, this viewpoint was not analysed in this study, but the results will serve as basis for further analysis from the viewpoint of wood supply potential. Figure 15. Profitability (net present values) at 3% and 4% discount rates by scenarios. 0 1000 2000 3000 4000 5000 6000 BAU INT QLTY EXT 3% 4% € ha-1 Working Papers of the Finnish Forest Research Institute 302 http://www.metla.fi/julkaisut/workingpapers/2014/mwp302.htm 36 Figure 16. Profitability (net present values) by Forestry centres at 3% discount rate by scenarios. Figure 17. Incomes, costs and net incomes by Forestry centres and by scenarios. BAU EXTQLTYINT 700–1000 1001–2000 2001–3000 3001–4000 4001–5000 5001–6000 6001–7000 7001–8000 8001–9000 NPV 3%, € ha-1 9001–10000 -50 0 50 100 150 200 250 300 350 BAU Incomes Costs Net incomes INT -50 0 50 100 150 200 250 300 350 QLTY EXT Ra L-S H-U Ka-S Pi E-Sa E-Po Ke-S P-Sa P-Ka Ka P-Po LaRa L-S H-U Ka-S Pi E-Sa E-Po Ke-S P-Sa P-Ka Ka P-Po La € ha-1 Working Papers of the Finnish Forest Research Institute 302 http://www.metla.fi/julkaisut/workingpapers/2014/mwp302.htm 37 5 Highlights • This study assessed the potential, cost-efficiency and impacts of intensified management of Finnish forests for next 100 years. • If the aim is to produce high quality raw material for forest industry, higher inputs in annual management practices are required. • Young stand management is in the main focus in the forest management. Treatment areas of young stand management should be doubled compared to current areas in order to maintain or increase cutting removals of high quality wood. • Intensive management allows increase of annual removals ca. 40%. • The annual energy wood removal can nearly met the target supply level of 13 mill m3 with intensive forest management by recovering only small-size stem wood, logging residues and stumps. • If the intensity of forest management will remain at current level, , the growing stock will increase. Intensive management does not decrease the amount of growing stock in the long run, despite increasing removals. • Intensively managed forest are more efficient capturing carbon from atmosphere than extensively managed forests, but the climate impacts depend on the use of removed carbon i.e. end-products made from the removed wood biomass. • Intensive forest management improves profitability nearly 50%. Especially the positive effect of intensive management on profitability is notable in Southern Finland References Ahtikoski, A., Alenius, V. & Mäkitalo, K. 2010. Scots pine stand establishment with special emphasis on uncertainty and cost-effectiveness, the case of northern Finland. New Forests 40: 69–84. Amer, M., Daim, T.U. & Jetter, A. 2013. A review of scenario planning. Futures 46: 23-40. Eerikäinen, K. 2009. A Multivariate Linear Mixed-Effects Model for the Generalization of Sample Tree Heights and Crown Ratios in the Finnish National Forest Inventory. Forest Science 55: 480–493. Finland’s national forest programme 2015. Turning the Finnish forest sector into a responsible pioneer in bioeconomy. 2010. Finnish Ministry of Agriculture and Forestry, Helsinki. 52 p. Finnish Government. 2008. Government resolution on the forest biodiversity programme for Southern Finland 2008–2016 (METSO) 15 pp. Available at: http://www.mmm.fi/attachments/metsat/5xO83061Y/ METSO_Resolution_2008-2016_ENGL.pdf. Forest Finland in brief. 2011. Finnish Forest Research Institute, Available at: http://www.metla.fi/metinfo/ tilasto/julkaisut/muut/metlaff2011.pdf. Haapanen, M. & Mikola, J. 2008. Metsänjalostus 2050 – Pitkän aikavälin metsänjalostusohjelma. Finnish Forest Research Institute, 50 p. Hänninen, H., Karppinen, H. & Leppänen, J. 2011. Suomalainen metsänomistaja 2010. Finnish Forest Research Institute, Vantaa. 94 p. Hökkä, H. & Kojola, S. 2002. Kunnostusojituksen kasvureaktioon vaikuttavat tekijät. In: Hiltunen, A. & Kaunisto, S. (eds.). Suometsien kasvatuksen ja käytön teemapäivät 26.–27.9.2001 Joensuu. Metsäntutkimuslaitoksen tiedonantoja 832. p. 30–36. Hökkä, H., Alenius, V. & Salminen, H. 2000. Predicting the need for ditch network maintenance in drained peatland sites in Finland. Suo 51: 1–10. http://www.mmm.fi/attachments/metsat/5xO83061Y/METSO_Resolution_2008-2016_ENGL.pdf http://www.mmm.fi/attachments/metsat/5xO83061Y/METSO_Resolution_2008-2016_ENGL.pdf http://www.metla.fi/metinfo/tilasto/julkaisut/muut/metlaff2011.pdf http://www.metla.fi/metinfo/tilasto/julkaisut/muut/metlaff2011.pdf Working Papers of the Finnish Forest Research Institute 302 http://www.metla.fi/julkaisut/workingpapers/2014/mwp302.htm 38 Hynynen, J., Ahtikoski, A., Siitonen, J., Sievänen, R. & Liski, J. 2005. Applying the MOTTI simulator to analyse the effects of alternative management schedules on timber and non-timber production. Forest Ecology and Management 207: 5–18. Hyvän metsänhoidon suositukset. 2006. Metsätalouden Kehittämiskeskus Tapio, Helsinki. 101 p. Korhonen, K. T., Heikkinen, J., Henttonen, H., Ihalainen, A., Pitkänen, J. & Tuomainen, T. 2006. Suomen metsävarat metsäkeskuksittain 2004–2005. Metsätieteen aikakauskirja 1B/2006: 183–221. Lappi, J. 1992. JLP, A linear programming package for management planning. 134 p. Lappi, J. & Lempinen, R. 2013. J -users’ guide 2.0. Finnish Forest Research Institute, Vantaa. 118 p. Leppänen, J. & Sevola, Y. 2013. Metsämaan omistus 2011. Metsätilastotiedote (SVT Maa-, metsä- ja kalatalous) 16/2013. Finnish Forest Research Institute, Vantaa. 9 p. Matala, J., Kärkkäinen, L., Härkönen, K., Kellomäki, S. & Nuutinen, T. 2009. Carbon sequestration in the growing stock of trees in Finland under different cutting and climate scenarios. European Journal of Forest Research 128: 493–504. Meristö, T., Kettunen, J. & Hagström-Näsi, C. 2000. Metsäklusterin tulevaisuusskenaariot. Teknologiakatsaus 95/2000, 69 p. Nuutinen, T., Hirvelä, H., Hynynen, J., Härkönen, K., Hökkä, H., Korhonen, K. & Salminen, O. 2000. The role of peatlands in Finnish wood production – an analysis based on large-scale forest scenario modelling. Silva Fennica 34 (2):131–153. Salminen, H., Lehtonen, M. & Hynynen, J. 2005. Reusing legacy FORTRAN in the MOTTI growth and yield simulator. Computers and Electronics in Agriculture 49: 103–113. SAS Institute. 2011. The SAS system for Windows. Release 9.2. Siipilehto, J. 2006. Linear prediction application for modelling the relationships between a large number of stand characteristics of Norway spruce stands. Silva Fennica 40(3):517–530. – 2009. Modelling stand structure in young Scots pine dominated stands. Forest Ecology and Management 257: 223–232. Siipilehto, J., Sarkkola, S. & Mehtätalo, L. 2007. Comparing regression estimation techniques when predicting diameter distributions of Scots pine on drained peatlands. Silva Fennica 41(2):333–349. Siipilehto, J., Valkonen, S., Ojansuu, R., Hynynen J., Miina J. & Saksa, T. 2014. Metsikön varhaiskehityksen kuvaus MOTTI-ohjelmistossa. Metlan työraportteja 286. 43 p. State of the world’s forests. 2009. Food and Agriculture Organization of the United Nations, Rome. 152 p. Tomppo, E., Haakana, M., Katila, M., Mäkisara, K. & Peräsaari, J. 2009. The Multi-source National Forest Inventory of Finland – methods and results 2005. Working Papers of the Finnish Forest Research Institute 111. 277 p. Ylitalo, E. 2012. Finnish statistical yearbook of forestry. Finnish Forest Research Institute, Vantaa. 454 p. – 2013. Finnish statistical yearbook of forestry. Finnish Forest Research Institute, Vantaa. 416 p. Working Papers of the Finnish Forest Research Institute 302 http://www.metla.fi/julkaisut/workingpapers/2014/mwp302.htm 39 Appendix 1: Prediction models for stand dynamics in Motti simulator 1 Natural regeneration and early growth The early growth of stands in MOTTI simulator comprises stand establishment and the development of the stand characteristics from establishment to dominant height (Hdom) of height meters. Thereafter, the stand-level predictions are replaced by individual-based models for tree growth. The general approach to early growth prediction applied in MOTTI is described by Siipilehto et al. (2014). Species composition in natural regeneration follows species of seed trees. In artificial regeneration, trees are established according to user-defined rules covering tree species, origin and number of plants or seeding points. In addition to seeded or planted trees, naturally regenerated mixture is predicted using models that are based on data from the Finnish National Forest Inventory (NFI7). Species mixture and species specific stem number is affected by soil type (mineral or organic), site fertility, regeneration method and dominant tree species (see Hynynen et al. 2002, Tables 15 and 17). Corrections to numbers given by Hynynen et al. (2002, Table 15) are calculated according to site preparation method and stand location. After early cleaning of sapling stand (Hdom < 2 m), a new seedling storey is established and its density is scaled according to the time of cleaning. The models for stand characteristics are assumed to be multiplicative, and they were fitted using linear regression after logarithmic transformations. The following dependent variables (Y) were fitted simultaneously: 1) total basal area (G, m2ha-1), 2) total stem number (N, ha-1), 3) arithmetic mean diameter (D, cm), 4) basal area median diameter (dgM, cm), 5) dominant diameter (Ddom, cm), 6) mean height (H, m), 7) basal area median height (hgM, m), 8) dominant height (Hdom, m). The basic models represented the average development of each stand variable over the stand total age. The common structure of the candidate model was as follows: lnY = a0 + a1 ln(T) + a2 Tk + a3 lnDDY + a3 Origin×Tk + aj Sitej + e Eq. 1 where T = total age (yrs), and the candidate power k was either -0.5 or -1, DDY = degree days (i.e. average annual sum of the mean temperatures above +5 °C), “Origin” is a dummy (value either 0 or 1) for artificial regeneration methods and “Site” consists of dummy variables associated with a certain site (j) defined as forest type by Cajander, and the supplementary site characteristic such as stoniness and paludification, a0–ai, are estimated parameters of the model and ε is the random error. The models were fitted simultaneously in order to estimate the cross-model error variance- covariance