Article F. Stuart Chapin III, Kjell Danell, Thomas Elmqvist, Carl Folke and Nancy Fresco Managing Climate Change Impacts to Enhance the Resilience and Sustainability of Fennoscandian Forests Projected warming in Sweden and other Fennoscandian countries will probably increase growth rates of forest trees near their northern limits, increase the probability of new pest outbreaks, and foster northerly migration of both native and exotic species. The greatest challenges for sustainable forestry are to restore and enhance the ecological and socioeconomic diversity of intensively managed forested landscapes. With appropriate man- agement, climate warming may facilitate the regeneration of this diversity. Experimental transplant gardens along latitudinal or altitudinal gradients and high-resolution maps of expected future climate could provide a scientific basis for predicting the climate response of potential migrant species. Management of corridors and assisted migration could speed the movement of appropriate species. INTRODUCTION The world is experiencing rapid changes in many of the factors that influence both ecosystems and society. For example, human activities have substantially altered climate, the hydro- logic cycle, biodiversity, land cover, the use of biological productivity, the cycling of nitrogen, and the mobility of exotic species at global scales (1). In addition, the globalization of trade and culture and increasing energy costs make local economies sensitive to events that occur elsewhere in the world. These changes challenge society���s capacity to sustain the desirable features of ecosystems on which we depend, especially for forests because the trees that are planted today will likely encounter quite different physical, biological, and economic environments by the time they are harvested. How can plans for sustainable management of long-lived species such as trees be developed when conditions are changing so rapidly? We address this question with respect to Fennoscandian forestry, which is an important part of the ecological, economic, and social fabric of the region. The trends described here are based largely on information from Sweden, but many of these patterns are similar to those occurring elsewhere in Fenno- scandia. Beginning in the Middle Ages, Swedish lands gradually became formally owned by either families or the government. Commercial timber harvesting began in the 17th and 18th centuries, and by 1900 large land areas were contracted to pulp and paper companies. Initially these contracts were demand- based rather than supply-based, so little attention was given to whether forestry practices were sustainable or to factors that might change the nature of forests (2, 3). By the 1920s, fire had been virtually excluded as a major disturbance. Fertilization, pesticide application, replanting, and other intensive silvicul- tural methods were implemented to increase yields. The intensification of management led to an increase in the standing biomass of forests that had been ������mined������ in the 19th century. There was, however, a reduction in average stand age, smaller trees, less structural complexity and coarse woody debris, and drainage of wetlands and peatlands to increase the area available for commercial forestry. Species diversity of plants, insects, and birds typically increases with forest age (4), so the combination of habitat loss from fire exclusion and wetland drainage, the reduction in average stand age, and the intensification of forest management led to substantial biodi- versity loss (5). In recent decades, the idea of maximum sustainable yield of forest products has given way to a broader concept of multiple use of forests in response to changing attitudes and an awareness of the services provided to society by natural ecosystems. The Forestry Act of 1993 codified management for biodiversity and other ecological and social factors, but by this time much of the original complexity and biodiversity of these forests had been lost (5). Most Swedish forestland is privately owned, so decisions by private landowners strongly influence future sustainability. Over the years, forest companies have merged into larger, more international units. Small forest owners have increasingly moved to towns and cities and have contracted forest management to forest owners associations or forest companies. Mechanization has led to a progressive decline in the number of people employed by forest operations. As a consequence of the globalization of the Swedish forest industry, decisions on forest issues are no longer only a Swedish matter. All these changes have reduced the forest owners��� sense of connection to the land. However, forests generally provide a minor part of the income to small-scale private forest owners, so their decisions about forest management can be strongly influenced by other services provided by forests, including the supply of moose (Alces alces) and berries and the aesthetic nature of the forests. A FRAMEWORK FOR SUSTAINABILITY PLANNING Given the rapidly changing global environment and the long history of intensive forest management in Sweden and other Fennoscandian countries, there are both challenges and opportunities for sustainable forest management. What should be sustained? For whom and for how long? In what ways can we use rapid climate change as an opportunity rather than simply as a liability? We define sustainability as the use of environment and resources to meet the needs of the present without compromising the ability of future generations to meet their own needs (6). This well-accepted definition of sustainability does not require that future ecosystems and their use by society be identical to those of today. Sustainable forestry is inherently challenging because of the large magnitude (but uncertain nature) of the ecological and social change that will inevitably occur between planting and harvesting trees. Therefore, for long time intervals it is important to manage for resilience���the capacity of forestry systems to deal with unexpected as well as expected changes in conditions without degrading the founda- tion for long-term sustainability. The science of sustainability (7���10) is still in its early stages, so there are no well accepted principles that provide a formula for sustainability planning. The overall aim is to sustain earth���s 528 Ambio Vol. 36, No. 7, November 2007 �� Royal Swedish Academy of Sciences 2007 http://www.ambio.kva.se

life support systems���the services that ecosystems provide to society (8). Sustainability is a human-oriented concept focused on the linkages between ecological and social processes in an integrated social-ecological system (Fig. 1). Especially impor- tant are ecosystem services, i.e., the benefits that society receives from ecosystems (10, 11). The Millennium Ecosystem Assess- ment described several general categories of ecosystem services that are useful for understanding the linkages between ecosystems and society (10). The ecosystem services that are most actively managed are the goods (provisioning services) that are directly harvested and used by society (e.g., timber products, food, and water). In addition, there are supporting services (basic ecological processes that shape the structure and dynamics of ecosystems) regulating services such as climate and disturbance regulation that affect areas well beyond a managed stand and cultural services that provide a sense of place and identity, aesthetic or spiritual benefits, and opportu- nities for recreation and tourism. Human impacts on ecosystems form a second general category of human-environment interactions. We describe these in terms of institutions, i.e., the repeating patterns of human action in situations structured by rules, norms, and shared strategies (12���14) (Fig. 1). Human actions, mediated by these institutions, affect both ecological and social processes (15). Institutions are a useful way to analyze societal responses to environmental change because they affect both individual decisions and the politics of larger groups (16). EXPECTED CHANGES Through the Arctic Climate Impact Assessment and other studies, there is considerable information about the changes that Fennoscandia will probably experience in the next 50���100 years (17). It is more reasonable to plan for the future based on these expected changes than to assume that the future will be like the past. During the next 50 years, the air temperature in Fennoscan- dia is expected to increase by 0.2���0.58C per decade, with the increases being greatest in the north and in the interior (18���20). Warming will be especially pronounced in winter and spring. Weaker inversions will make warming most pronounced in valleys. Thus, both geographic and seasonal temperature variations will become smaller in the future. Warmer winters will increase the frequency of rain-on-snow or wet-snow events, winter thaws, and snow-free winter conditions, probably leading to more frequent and extreme soil thaw-freeze events. Annual precipitation is expected to increase, especially in the north (21). During summer, however, precipitation is expected to increase only slightly in northern Fennoscandia and to decrease in the south. The combination of warmer temperatures and small changes in summer precipitation will probably increase the frequency of summer drought and risk of wildfire, especially in southern Fennoscandia. All across the boreal forest, the frequency of extreme events, including storms, floods, and droughts, is expected to increase (22). Unless greenhouse gas emissions are reduced in the near future, the changes in the second half of the 21st century are expected to be similar in direction but more extreme than those described above. There is also speculation that current trends in the discharge of high-latitude rivers could weaken thermohaline circulation by the end of the 21st century (23) if this occurred, it might cause a radical cooling of the climate throughout Fennoscandia, but the likelihood and timing of such an extreme climate change can only be speculated. It is also possible that all the global models are incorrect and that current warming trends will not persist. These are examples of unexpected events that cannot be easily planned for. In general, boreal forests are expected to respond to climate warming by increasing productivity, moving northward and upward in elevation, and being displaced in the south by temperate forests or agriculture (24���26). The individual processes leading to these predictions are relatively well understood, but the rates of change and interactions among these processes are not easily predicted, so the overall outcome is relatively uncertain. Here we summarize what is known about the climate sensitivity of individual processes and speculate about their interactions. In southern Fennoscandia, the dominant tree species will probably show little growth response to climatic warming because temperature increases are expected to be small, and wood production by Norway spruce (Picea abies) and Scots pine (Pinus sylvestris) in provenance trials shows little variation in response to regional variation in summer climate (27, 28). In northernmost Fennoscandia, where warming is expected to be greater, the resident tundra vegetation shows little response to experimental warming (29���31) (just like tree growth in the south), but provenance trials suggest that expected warming could lead to a 300% increase in wood production of Norway spruce and a 60% increase in production in Scots pine (28, 32). Nitrogen plays a strong role in these growth responses. Warmer temperatures and longer frost-free periods can increase nitrogen mineralization and availability, stimulating tree growth (33) in both the north and the south (34). Because most species show greater growth response to increased temperature at their northern range limit, as described above for Fennoscandia���s major timber species, there will probably be a gradual migration of southerly taxa, including both native and exotic species, to the north and to higher elevations. However, migration rates and pathways will differ dramatically among species. Species that disperse easily (e.g., birds, winged insects, and plants with wind-dispersed seeds) and that have short generation times (particularly weedy species) will disperse most rapidly. In addition, more continental species are expected to migrate westward from Finland and Russia as warming proceeds in northern Sweden. Foresters may also introduce tree species or genotypes north of their current distribution, especially if resident species respond negatively to the direct or indirect effects of climate change. Species migration Figure 1. Diagram of a social-ecological system composed of an ecological subsystem (left) and a social subsystem (right), each with a spectrum of controls that operate across a range of temporal and spatial scales. Environmental impacts, ecosystem services, and social impacts govern the well being of human actors, who affect ecological and social systems through a variety of institutions. Modified from (15). Ambio Vol. 36, No. 7, November 2007 529 �� Royal Swedish Academy of Sciences 2007 http://www.ambio.kva.se