d i s c u s s i o n pa p e r n u m B e r 3 august 2010 d e v e l o p m e n t a n d c l i m at e c h a n g e d i s c u s s i o n pa p e r s 1 56661 d e v e l o p m e n t a n d c l i m a t e c h a n g e Adaptation of Forests to Climate Change D I S C u S S I O N PA P E R N u M B E R 3 AuGuST 2010 D E V E L O P M E N T A N D C L I M A T E C H A N G E Adaptation of Forests to Climate Change By Roger A. Sedjo Resources for the Future sedjo@rff.org Papers in this series are not formal publications of the World Bank. They are circulated to encourage thought and discussion. The use and citation of this paper should take this into account. The views expressed are those of the authors and should not be attributed to the World Bank. Copies are available from the Environment Department of the World Bank by calling 202-473-3641. © 2010 The International Bank for Reconstruction and Development / THE WORLD BANK 1818 H Street, NW Washington, DC 20433, U.S.A. Telephone: 202-473-1000 Internet: www.worldbank.org/climatechange E-mail: feedback@worldbank.org All rights reserved. August 2010 This volume is a product of the staff of the International Bank for Reconstruction and Development / The World Bank. The findings, interpretations, and conclusions expressed in this volume do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgement on the part of the World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. RIGHTS AND PERMISSIONS The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applica- ble law. The International Bank for Reconstruction and Development / The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center Inc., 222 Rosewood Drive, Danvers, MA 01923, USA; telephone 978-750-8400; fax 978-750-4470; Internet: www.copyright.com. Cover image: Forest fire, Eastern Europe. © Konstantin Minronov / Shutterstock Images, LLC. All dollars are U.S. dollars unless otherwise indicated. III Table Of COnTenTs 1. Context 1 1.1 What are the Potential Impacts of Climate Change, including Extreme Weather Events, on the Sector? 1 1.1.1 Disturbances and extreme events 3 1.2 Who (across and within Countries) is likely to be Most Affected? 3 1.2.1 Geographically 3 1.2.2 By income or vulnerability class 4 1.3 What Experience is there with Adaptation in the Sector? 4 1.3.1 Autonomous adaptation 5 1.3.2 "Soft" adaptation ­ policies and regulations 5 1.3.3 Reactive adaptation 5 1.4 What is the Nature and Extent of the Adaptation/Development Deficit in this Sector? 5 1.5 How will Emerging Changes in Development and Demographics Influence Adaptation? 7 2. Literature Review 8 2.1 Previous Studies Relevant to the Sector and their Major Conclusions. 8 2.1.1 The results 9 2.2 How our Study Complements Existing Work 12 2.3 Methodology 12 2.3.1 Capturing ecological changes in the economic model 13 2.3.2 The economic model 14 2.4 Demand 16 2.4.1 Possible changes in demand 16 2.5 How We Represent the Future ­ 2010 to 2050 17 2.5.1 The baseline results without climate change 17 2.6 Global Results: 18 2.6.1 With climate change impact on the forest sector 18 2.6.2 Regional impacts 19 2.6.2 Global "wet" and "dry" scenarios 19 2.7 How Climate Change Impacts are Calculated 20 2.8 How Costs of Adaptation are Defined 20 2.9 How Costs of Adaptation are Calculated 20 IV A DAPTATION Of fORESTS TO CLIMATE C H A N G E 3. Results of Damage Offset Investments 21 3.1 Investment Costs (Upfront and Maintenance) in the Baseline (No Climate Change) Scenario 21 3.2 Country Case Studies: Brazil, South Africa, and China 21 3.2.1 Brazil 22 3.2.2 South Africa 24 3.2.3 China 26 4. Limitations 28 4.1 Treatment of Extreme Events 28 4.2 Treatment of Technological Change 28 4.3 Treatment of Inter-Temporal Choice 28 4.4 Treatment of "Soft" Adaptation Measures 28 4.5 Treatment of Cross-Sectoral Measures 28 4.6 Areas for Follow-up Work and Research Advances 29 5. Investments and Compensation: Some Thoughts 31 6. Summary and Conclusions 32 7. References 34 Tables 1 Examples of simulated climate change impacts on forestry 9 2 Summary: Timber market results to date 10 3 Percentage change in forest areas in longer term (by 2145), based on the Hamburg and IIuC climate scenario used for ecological projections 10 4 Percentage change in timber growth rates by 2145. 11 5 Percentage change in regional timber production for 50-year time periods 11 6 Percentage change in regional timber production to the year 2050, based on the climate scenarios used for ecological projections 12 Figures 1 Modeled vegetation distribution 2 2 Changes in the ranges of four tree species since the last ice age 6 3 Adaptation in managed ecosystems 7 4 Presents projections of timber harvests by region for the baseline scenario 17 5 Global timber prices over time 18 6 forest plantation development area (000 ha) 21 7 Change in precipitation, from 2000 to 2050 22 8 Brazil area: 8.5 million km2 23 9 Eucalyptus 3,800,000 ha 23 D E V E L O P M E N T A N D C L I M AT E C H A N G E D I S C u S S I O N PA P E R S V Figures (conTinued) 10 Pinus 1,800,000 ha 23 11 South African provinces 26 12 Map of China 26 13 Ecosystem areas by type 27 14 Land cover of Heilongjiang Province 28 1 1. COnTexT forestry in various regions and countries. Based on the assessment of these projections, adaptation measures are suggested to mitigate damages likely to be This study is part of a World Bank effort intended, first, incurred and identify adaptations that might be made. to help decision makers in developing countries to Preliminary cost estimates are made. The approach better understand and assess the risks posed by climate will not involve a new model or new projections. change and to better design sector strategies to adapt to Rather, the study draws from the literature and the climate change. The second objective is to develop a results of earlier investigations. These are reported and "global" estimate of adaptation costs to inform the the most comprehensive results fused into a single international community's efforts, including UNFCCC report. These do not perfectly fit the precise guide- and the Bali Action Plan, to provide access to adequate, lines--for example, GDP and population--applied to predictable, and sustainable support, and to provide new some of the other World-Bank-sponsored studies. and additional resources to help the most vulnerable However, the models often do not calibrate to GDP developing countries meet adaptation costs. or population. Rather, they make some simplifying assumptions on the demand side with the focus of the To meet these two objectives, the broad World Bank analysis being on the supply side. The results are, to a effort will proceed on two tracks, a case study and an large extent, invariant to demand-side projections that aggregate track. This study is part of the aggregate are only modestly varied. track, which has two objectives. The first is to ensure the availability of developing country/regional adapta- The results of this study are consistent with the general tion cost estimates to contribute to the discussion on findings of the IPCC Fourth Assessment of Climate climate change leading up to the Copenhagen confer- Change, WG II (2007), which states: "The changes on ence in late 2009. The second objective of the aggregate global forest products range from a modest increase to a track to begin to develop procedures that will be slight decrease, although regional and local changes will needed to generate aggregate adaptation cost numbers be large. Production increases will shift from low-lati- once the country case studies are completed. Within tude regions in the short term to high latitude regions that track, the forestry component focuses on the in the long term." This correspondence is not surprising, industrial wood sector. Traditional fuelwood is also of since this study draws in part on the IPCC findings and interest. However, since a significant amount of fuel- on the literature that went into developing those wood is not traded in markets, there is not much data. findings. In general, we would expect that conditions favorable to an expanding forest would also be favorable to the 1 . 1 W h aT a r e Th e pO T e nT i a l i m paC T s creation of fuelwood, and vice versa. O f C l i m aTe C h a n g e , i nC l u d i n g e xTr e m e We aTh e r e v e nTs , O n Th e The approach of this forestry study is to draw from s e C TOr ? the considerable existing literature (see Table 1) to provide perspective, as well as estimates and projec- The ecological literature suggests that warming is tions of the impacts of climate change on forests and likely to result in an expansion of forest in the 2 A DAPTATION Of fORESTS TO CLIMATE C H A N G E high-latitude areas that were previously devoid of large differences in the location of forests and other forest. In the mid-latitudes, some species are likely to vegetative types across models. For example, while some experience dieback of some forest species and types, models--for example, CCCM and UKMO--predict while others migrate to areas with more friendly the forests of the U.S. Southeast to be replaced by grass- climates (Smith and Shugart 1993; Easterling and lands, others-- for example, HADCM2SUL and Aggarwal.). Ecological studies suggest that tree species HADCM2GHG--expect forests to flourish (Figure 1). at the edge of their ecological range may persist even if This is probably due largely to predicted differences in they are not able to regenerate in those conditions moisture. Although models now project on subconti- (Clark 1998). nental scales, it is still well-recognized that GCMs do less well when predicting regional climate effects Figure 1 provides projections of forest configuration (Climatewire 2009). under several alternative GCMs. Note that there are figure 1. mOdeled vegeTaTiOn disTribuTiOn Current HADCM2SUL OSU HADCM2GHG GFDL R30 GISS Tundra CCCM1 UKMO Taiga/Tundra Conifer Forest Northeast Mixed Forest Temperate Deciduous Forest Southeast Mixed Forest Tropical Broadleaf Forest Savanna/Woodland Shrub/Woodland Grassland Arid Lands The above maps were generated by the MAPSS vegetation distribution model (10-km resolution), and depict patterns of major vegetation types in the conterminous United States under current conditions and in response to a doubling of pre-industrial atmospheric CO2 concentrations. The map in the top left corner represents the current distribution of major vegetation types. The remaining seven maps represent the change in distribution of those vegetation types as predicted by different climate models. Source: neilson 1995, reproduced from shugart et al. 2003. D E V E L O P M E N T A N D C L I M AT E C H A N G E D I S C u S S I O N PA P E R S 3 In addition to a changing temperature, the amount and This might be partially explained by the difficulties in pattern of precipitation and moisture is critical to controlling for constant CO2 concentration in a large- forests. In general, getting warmer and wetter will scale experiment. However, economic models often enhance forest growth, while warmer and drier is likely presume high fertilization effects--as did the Sohngen to be detrimental to growth. If drying is significant, et al. (2001) study, which used projections that increased grasses will often replace forests in natural systems NPP by 35 percent under a 2xCO2 scenario. Regardless (Bowes and Sedjo 1992). A number of biogeographical of the contradictory effects of variations in CO2 models demonstrate a polarward shift of potential concentrations, however, empirical evidence indicates vegetation for the 2xCO2 climate by 500 km or more that forest growth rates have been increasing since the for the boreal zone (Solomon and Kirilenko 1997). middle of the 20th century, as noted by Biosvenue and The equilibrium models and some dynamic vegetation Running (2006). models project that this vegetation shift toward newly available areas with favorable climate conditions will 1.1.1 disturbances and extreme events eventually result in forest expansion and replacing of up to 50 percent of current tundra area. There is, Natural disturbances are an integral part of the environ- however, a concern that the lagged forest migration ment in which most forests flourish and evolve. (compare the tree species migration rates after the last Wildfires, outbreaks of insects and pathogens, and glacial period of a few kilometers per decade or less to extreme events such as high winds, are often an integral a projected future climate zones shift rate of 50 kilo- part of the forest environment. These disturbances often meters per decade) could lead to massive loss of natu- precipitate stand-replacing events. Ecological systems ral forests, with increased deforestation at the southern adapt to their climate. Changing climates create new boundary of the boreal forests and a correspondent conditions less consistent with the current ecosystems, large carbon pulse (Malcolm et al. 2002). However, thereby creating increasing stress on the systems. such a result could also lead to an increased rate of Climate change will almost surely change the timing of harvest to allow the capture of the value of the trees the disturbances and will probably increase their sever- before it is lost to mortality. For timber production, ity. Indeed, climate-induced changes in disturbance which typically relies on managed forests with migra- regimes already appear to be occurring (van Mantgem tion facilitated by human actions, this negative effect et al. 2009; Westerling et al. 2006). Modifications of of lagged migration might be of lesser importance than temperature and precipitation, which weaken the forest for natural forests. and can increase the frequency and intensity of infesta- tion and fire, may be as important as the direct impact Carbon Dioxide Fertilization. Increasing concentra- of higher temperatures and elevated CO2. An example tions of atmospheric CO2, aside from modifying the of such a situation may be the extreme beetle outbreak temperature and precipitation pattern, may also increase in the Canadian western forests (Kurz et al. 2008). production through the "carbon fertilization effect" as Many observers believe the beetle population has flour- noted above. Earlier experiments in closed or open-top ished due to the warmer winters, and thus insect chambers demonstrated very high potential for mortality has been dramatically reduced. CO2-induced growth enhancement, such as an 80 percent increase in wood production for orange trees Indeed, some have argued that extreme events in (Ipso et al. 2001). The free-air CO2 enrichment forestry are often the vehicle for facilitating the (FACE) experiments demonstrated a smaller effect of replacement of an established forest with a new, increased CO2 concentrations on tree growth. Long- perhaps more suitable forest, should conditions change term FACE studies suggest an average NPP increase of (Sedjo 1991). Although extreme events could well 23 percent in response to doubling of the CO2 concen- increase due to climate change, few forest production tration in young tree stands, with a range 0­35 percent models include these effects. However, from a timber (Norby 2005.) However, in another FACE study of production perspective, one response would be to mature 100-year-old tree stands, little long-term anticipate the disturbances with shorter and/or more increase in stem growth was found (Korner et al. 2005). targeted harvests. 4 A DAPTATION Of fORESTS TO CLIMATE C H A N G E 1.2 Wh O (a CrOss and WiThin entities. However, there are many exceptions. In South COun Tries) is likely TO be mO sT Africa, for example, although the pulp plantations are affe CTed? privately owned, sawtimber plantations are typically owned by the state. In China, large areas of planta- 1.2.1 geographically tions were established and are managed by the state. However, private international forest companies are In general, climate change is likely to shift natural forests now beginning to establish tree plantations. In Kenya, toward the poles. The same shift is likely for planted government plantations provide wood for both forests, although it would probably be propelled by forest sawmills and pulp operations, while small-scale private management decisions that involved the replacement of tree growing for industrial purposes is also encouraged harvested plantation stands into new areas. Most GCMs (Sedjo 2004). indicate that temperature changes will be least at the equator and increase as the poles are approached. Thus, The income vulnerabilities probably reside mostly with for forests, the locational changes should be greatest in the forestry labor force, which is largely unskilled and the boreal and temperate countries, most of which are low income. Although tree growing is a relatively developed. This suggests the likely migration of boreal modest user of labor, labor is needed both for planting forest into areas that formerly were devoid of trees--for and for harvests. More importantly, wood processing example, parts of the tundra--accompanied by temperate facilities often use substantial amounts of labor. Thus, forests moving into some areas that were formerly occu- any climate-induced disruptions in the industrial forest pied by boreal forests, assuming soils, photoperiod, etc. resource are likely to generate employment losses in the are appropriate. Although not often discussed, tropical processing industries, as well as in the forest. forests may be impacted differently, since the anticipated amount of temperature warming is lower at those lati- 1 . 3 W h aT e x p e r i e nCe i s T h e r e W iT h tudes. However, tropical forests may have less tolerance a d a pTaTi O n i n Th e s eC TOr ? for adaptation. In recent decades industrial forestry has undergone Perhaps more important than temperature are the major changes as planted forests have been established changes in precipitation and moisture. Limits on mois- in an increasing number of countries and regions. ture could result in forestlands being converted to Often, these have not been traditional wood producer grasses. However, climate models are not generally countries, but tropical and subtropical countries in regarded as good predictors of regional precipitation which forest planting has occurred (Bael and Sedjo changes. In general, however, interiors of continents 2006). Indeed, the changes have been so great that an tend to be dry, and this tendency should be exacerbated increasing percentage of the world's industrial wood under climate change and warming. comes from planted forests. The share is expected to be over one-half by 2050, even in the absence of any Over the next fifty years, the forest industry could climate change. Climate change could be expected to probably adapt without major relocation of its process- accelerate this process. ing facilities for reasons discussed below. Over long periods of time, assuming appropriate foresightedness, A host of approaches and tools may be used to adapt to processing facilities could adjust gradually though the changing conditions such as climate change (Sohngen phasing out of obsolete facilities, often with 50-year 2007; Seppala et al. 2009), with a major set of adapta- lives, and adjust the location for new investments, tions associated with the planted forest. A decision to thereby keeping additional climate-inducted costs very plant also involves considerations with respect to loca- modest. tion, choice of species, and quality of the stock to be planted. The planting approach allows regeneration to be 1.2.2 by income or vulnerability class for the species of choice, which is often a rapidly grow- ing species appropriate for intensively managed industrial Forest assets have a variety of ownerships. Most forests. This choice can be desirable for timber produc- industrial forest plantations are owned by private tion and/or for other forest values. Adaptations that may D E V E L O P M E N T A N D C L I M AT E C H A N G E D I S C u S S I O N PA P E R S 5 be useful during climate warming include changing rota- investment could consist of roads and other infrastruc- tion periods, salvage where damage is incurred, replant- ture that allows harvesting to take place, although forest ing of new species if conditions warrant, and adjusting roads are usually the responsibility of the forest harvest- future investment levels, including relocation of selected ing entity. In the context of climate change and forest plantations if warranted. relocation, some major new roads might be required to facilitate the delivery of raw wood to the mills. In addi- In a recent paper on forest adaptation to climate change, tion, forests are often publicly owned or assisted by Roberts (2009) points out those policies that serve multi- public funds. In this context, the public investment ple purposes can be useful in adapting to climate change. could take the form of tree planting to replace or antici- He notes that some forest managers are already begin- pate forest losses. In some cases, the public investment ning to anticipate climate change in their management could take the form of aerial seeding and/or other activ- decisions. He also points out that existing policies tend ities to facilitate the more effective migration and to be reactive rather then proactive. Given the uncertain- regrowth of the forest, although aerial seeding is usually ties of how climate is likely to affect any specific forest, not recommended for commercial forests. however, one might maintain that a reactive policy with a high degree of flexibility is highly appropriate. 1.3.2 "soft" adaptation ­ policies and regulations 1.3.1 autonomous adaptation "Soft" adaptation might include policies and regula- tion to facilitate the "natural" migration and regenera- The evidence indicates that natural forests have been tion of the forest, such as those discussed above. Fire migrating at least since the last glacial period, as the control might also be viewed as a soft adaptation earth warmed and moisture patterns changed. In the policy. However, the broader implications involve absence of very rapid climate change, tree species have short-term emissions releases, and fire control could shown that they are able to migrate and adapt to the be difficult changing environment, in some cases creating forests with a new combination of tree species (Shugart 2003). 1.3.3 reactive adaptation Figure 2 shows the migration of some forest species in North America in the post-glacial period. Figure 2 Reactive adaptation could probably involve activities shows that the forest changes will depend upon the that might be undertaken should damages be occurring specifics of the climatic change. However, climate in the forest. An example might be attempts to control changes have accelerated in recent decades, and some or limit the effects of wild fire. Limiting wildfire may observers anticipate an increase in die-back toward the result in extending the life of the trees, thereby allowing end of this century (IPCC 2007). the harvest of the timber before it is destroyed. However, early wildfire control has often been cited as a For managed and planted forests, human actions may cause for larger firers in the longer term. In addition, facilitate the transition. For short rotation plantations, salvage logging is common, whereby after damage asso- the optimal approach may be simply to replant a site ciated with a nature event, such as fire or infestation, after harvest with a more appropriate provenance. The the remaining merchantable timber in the forest is adjustment problems for mills are generally negligible, harvested and utilized. since the species are likely to be similar to the ones replaced; for example, slash pine replacing loblolly as 1 . 4 W h aT i s Th e n aTu r e a n d e xTe n T the temperature rises. Thus, the adaptation costs are O f Th e a d a pTaT iO n / d e v e lO p m e n T likely to be very small, since artificial regeneration d e f iCi T i n Th i s s eC TOr ? would occur anyway. The only serious question regards replanting with the appropriate species and adjusting The timber producing sector has a high degree of the management regime to that new climate situation. potential for adaptation. In the near term, damaged forests can still be harvested and the usable wood Public sector investment. Forest ownership varies commercially utilized. In the longer term, the forest can considerably across the globe. Relevant public sector usually renew itself through natural regeneration, 6 A DAPTATION Of fORESTS TO CLIMATE C H A N G E figure 2. Changes in The ranges Of fOur Tree speCies sinCe The lasT iCe age 6 10 8 12 12 14 12 14 Picca spp. Pinus strobus Spruce White Pine 0 400 km 0 400 km 6 8 7 8 9 10 10 11 12 12 13 14 14 18 Quercus spp. Acer Maple Oak 0 400 km 0 400 km The lines in the map above mark the boundaries of the species ranges in units of millennia (e.g., 12 indicates the range boundary of the species 12,000 years ago). The changes in the species ranges are in response to climate changes of roughly the same magnitude as that projected over the 21st century climate due to climate change. The species clearly displayed marked differences with respect to their migration patterns and rates. Source: davis 1981, reprinted from shugart et al. 2003. D E V E L O P M E N T A N D C L I M AT E C H A N G E D I S C u S S I O N PA P E R S 7 although not always with the same species. In the very regeneration. The major consideration is that humans long term, the forest can migrate and adapt to new can facilitate an accelerated adjustment. climatic conditions, although not all new conditions will be conducive to forests. 1.5 hOW Will emerging Changes in d e v e lO p m e nT a n d d e mO g r a p h iC s Figure 3 describes how adaptation through harvesting i n f l u e nCe a d a pTaTi O n ? and replanting can substantially reduce losses that would otherwise occur if natural systems were allowed to adapt Forests compete with a variety of other uses for land. on their own. The die-back regime often assumes that Increasing development and growing populations often tree mobility is exceeded by the rate of climate change involve forest clearing, which could involve greater use (Davis and Shaw 2001). Note that in a die-back of previously forested land for agriculture. These alter- scenario, human management plays a large role in both native pressures will continue to compete for land with salvage logging and in promoting rapid regeneration. or without climate change. However, climate change Salvage logging captures some of the timber values that could modify the comparative productivity of the lands might otherwise be lost, and timely artificial regenera- for the various uses. Thus, in some cases forest uses may tion provides for more future commercial timber at an be benefited by climate change, while in others they will earlier time that would be the case relying on natural be disadvantaged. figure 3. adapTaTiOn in managed eCOsysTems A · daptationthroughharvestingandreplantingsubstantiallyreducethelossesthatwouldotherwiseoccurifnatural systemsadaptontheirown, · ResultsbelowarefortheUSonly. MAPSS & B-BGC; UKMO DOLY&TEM; UKMO 1000 800 800 600 400 Tg Carbon per decade Tg Carbon per decade 600 200 400 0 200 ­200 0 ­400 ­200 ­600 ­800 ­400 ­1000 ­600 ­1200 ­800 ­1400 1995 2015 2035 2055 2075 2095 2115 2135 1995 2015 2035 2055 2075 2095 2115 2135 Year Year Human Response Natural Model Source: reprinted from sohngen et al. 1998. 8 2. liTeraTure revieW these models also have been modified to examine the effects of climate change on forestry. This study has a major interest in this last set of models; their projections form the basis of this current study. 2.1 p revi Ous sT udies relevanT TO T he seCTOr and Their majO r One early economic assessment of regional climate COn Clusi Ons. impacts on forests and agriculture was the MINK study (Rosenberg et al. 1991, 1993), which examined the abil- A number of studies have examined the implications of ity of the agricultural and forest areas in a region in the climate change on forests and sometimes on industrial U.S. to adapt to a new and changing climate, with wood production (Table 1). The usual modeling mobility of crops and forests playing a major role. A approach is to combine general circulation models country-focused effort ( Joyce et al. 1995) looked at the (climate models) and ecological models to provide a U.S. forest sector using the terrestrial ecosystems model representation of the climate-modified environment. (TEM) to predict changes in timber growth rates, Economists then treat this as the underlying production timber inventories, and timber supply. An early global function, upon which economic models can and are effort by Binkley (1988), which focused on forestry's imposed to make their assessments. However, since response to climate, used a simple regression approach. different GCMs are used and different ecological Darwin et al. (1995) examined the adjustment of agri- models, the underlying production functions are often culture and forest markets to climate change in the U.S. different, even for the same region. Some have not However, the computable general equilibrium (CGE) allowed for natural and/or human-induced mobility of approach used did not capture the intertemporal adjust- forests and other vegetation. Many of the ecological ment process so critical in forests. More recent global models have focused only on individual countries or efforts include those by Perez-Garcia et al. (1997, 2002). regions. In most cases, the models examined impacts of A quite recent effort was that of Irland et al. (2007). warming on aspects of terrestrial vegetation. These efforts used a global forest economic model to examine the effect of climate change on forest growth The basic approach of any analysis of the economic and its effects on timber markets. However, while the impact of climate change on forests requires the inte- analysis uses the TEM, the approach ignores the grated use of three types of models: economic, climate dynamic migration aspects of tree species. models (GCMs), and ecological models. A number of economic models have been developed to examine long- Finally, the economic study that most directly and term timber supply. Some of these models have been comprehensively examined the effects of climate change modified to estimate the effects of forestry on climate on forests is that of Sohngen et al. (2001). The approach change, as forest activities can sequester and release utilizes the modified timber supply model (Sedjo and carbon. A major focus of recent work has been on the Lyon 1990). This report uses those results and the ability of forests to capture carbon, thereby offsetting or results of its successor models--particularly Sohngen et mitigating to some degree global warming. Some of al. 2001 and Daigneault et al. 2007--to estimate the D E V E L O P M E N T A N D C L I M AT E C H A N G E D I S C u S S I O N PA P E R S 9 Table 1. examples Of simulaTed ClimaTe Change impaCTs On fOresTry Reference; location Scenario and GCM Production impact Economic impact schngen et al., 2001; uiuC and hamburg · 2045: production up by 29­38%; reductions in · 2045: prices reduced, high-latitude loss, low- schngen and sedjo, T-106 for CO2 topping n. america, russia; increases in s. america altitude gain. 2005. 550 ppm in 2060 and Oceania. · 2145: prices increase up by up to 80% (no Global · 2145: production up by 30%, increases in n. climate change), high-latitude gain, low-latitude america, s. america, and russia loss. benefits go to consumers. solberg et al., 2002. baseline, 20­40%, · increased production W. europe, price drop with an increase in welfare to producers Global increase in forest growth · decreased production in e. europe and consumers. increased profits of forest industry by 2020 and forest owners. perez-garcia et al., Tem & CgTm miT gCm, · harvest increase in the us West (+2 to +11%), demand satisfied; prices drop with an increase in 2002. miT eppa emissions new Zealand (+10 to +12%), and s. america welfare to producers and consumers. Global (+10 to +13%). · harvest decrease in Canada. lee and lyon, 2004. eCham-3 (2xCO2 in · 2060s, no climate change: increase of the no climate change: Global 2060), Tsm 2000, biOme industrial timber harvest by 65% (normal · pulpwood price increases 44% 3, hamburg model demand) or 150% (high demand); emerging · solid wood increases 21% regions triple their production. With climate change: · With climate change: increase of the industrial · pulpwood price decreases 25% timber harvest by 25% (normal demand) or · solid wood decreases 34% 56% (high demand). e. siberia & us south · global welfare 4.8% higher than in no climate dominate production. change scenario. nabuurs et al., 2002. hadCm2 under is92a 18% extra increase in annual stemwood incre- both decreases or increases in prices are Europe 1990­2050 ment by 2030, slowing down on a longer term. possible. schroeter, 2004. ipCC a1f1, a2, b1, b2 · increased forest growth (especially in n. in the a1f1 and a2 scenarios, wood demand Europe upto 2100. few manage- europe) and stocks, except for a1f1. exceeds potential felling, particularly in the second ment scenarios · 60­80% of stock change is due to manage- half of the 21st century, while in the b1 and b2 ment, climate explains 10­30% and rest due to scenarios future wood demand can be satisfied. land use change. alig et al., 2002; CgCm1+Tem · increase in the timber inventory by 12% (mid- · reduction in log prices joyce et al., 2001. hadCn2+Tem term); 24% (long-term) and small increase in · producer welfare reduced compared to n cli- USA CgCm1+vemap harvest. major shift in species and an increase mate change scenario hadCm2+vemap in burnt area by 25­50%. · lower prices; consumers will gain and forest is92a · generally, high evaluation and northern forests owners will loose decline, southern forests expand. Source: reprinted from ipCC 2007, easterling et al. Wg 2. base and climate change deviations from that base. As with almost all studies of the effects of climate Related efforts, including some subsequent follow-on change on forests, the results show increased biological efforts using variants of the same model, provide addi- forest productivity, with forest area roughly unchanged tional inputs. These models generate projections of the and a modest increase in timber harvests, which results global forest and associated timber harvests with and in an overall decline in wood prices. All the large devel- without climate change into the middle of the 22nd oping regions show net benefits over the period to 2050 Century. Other studies that are part of the literature and and generally beyond. However, forest stocks cannot are particularly involved in this current study include increase indefinitely, and at some future time stocks Shugart et al. 2003 and Kirilenko and Sedjo 2007. must stabilize or decline. However, this need not imply a decrease in industrial wood supplies. 2.1.1 The results Table 3 provides the estimates of Sohngen et al. (2001) Tables 3­6 below provide the projected estimates of of the percentage change in forest areas in the longer Sohngen et al. 2001, which form the basis of this paper. term (by 2145), based on the projections of the 10 A DAPTATION Of fORESTS TO CLIMATE C H A N G E Hamburg and IIUC climate scenarios of the late Table 2. summary: Timber markeT 1990s used for ecological projections. Note that for resulTs TO daTe each GCM, eight of the nine regions experience a net Output area change over this longer period. Additionally, all of Producer Region 2000­2050 2050­2100 returns the regions experiencing a decline are developed north america ­4% to +10% +12% to +16% decreases regions. The next table, Table 4, provides estimates of the percentage change in NPP and timber growth europe ­4% to +5% +2% to +12% decreases rates by 2145 for the two climate models. For all russia +2% to +6% +7% to +18% decreases regions except Oceania, both NPP and timber yield south america +10% to +20% +20% to +50% increases rates are positive. Oceania experiences a decline in aus./new ­3% to +12% ­10% to +30% decr. & incr. NPP for only the Hamburg model. Table 5 presents Zealand the percentage change estimates in regional timber africa +5% to +14% +17% to +31% increases production for the Hamburg and UIUC models for China +10% to +11% +26% to +29% increases three 50-year periods to 2145. For all periods and se asia +4% to +10% +14% to +30% increases regions, the change is positive except for the three Source: alig et al. (2002), irland et al.(2007), joyce et al. (1995, 2001), perez- Hamburg projection for Oceania and the tow projec- garcia et al. (1997, 2002), sohngen et al. (2001), sohngen mendelsohn (1998, tions for North America. Finally, Table 6 draws the 1999), sohngen and sedjo (2005), karajaianen et al. (2003), nabuurs et al. summary results from Table 5, adjusted to the year (2002), perez-garcia et al. (2002), sohngen et al. (2001), lely et al. (1997). 2050. Note that projected timber production in North adaptation to forests and people to Climate Change. 2009. alexander buck, pia America and Oceania has declined modestly under the katila risto seppala, (eds.). iufro World series volume 22. helsinki, 224 p. Hamburg scenario, while only North America has declined under the UIUC scenario. Table 3. perCenTage Change in fOresT areas in lOnger Term (by 2145), based On The hamburg and iiuC ClimaTe sCenariO used fOr eCOlOgiCal prOjeCTiOns Hamburg3 UIUC3 Net Area Access1 Inacess. Net Net Area Access1 Inacess. Net Change Net Change Change Change Net Change Change High-latitude Forests North America 3 (7) 35 4 (2) 24 Europe 16 14 23 7 4 36 former Soviet union 12 14 13 14 15 15 China 41 5 188 20 0 109 Oceania (3) (12) 20 0 6 38 low- to Mid-latitude Forests South America 42 6 44 27 (2) 33 India 10 9 -- (1) (1) -- Asia-Pacific 23 0 282 33 (3) 392 Africa 71 5 74 38 (4) 41 Total 27 5 41 19 5 31 Source: sohngen et al. 2001. 1 accessible forest areas are forests used for industrial purposes. for the low- to mid-latitude forests, accessible includes only industrial plantations or highly man- aged forests. 2 for the asia-pacific region, inaccessible forests are the valuable dipterocarp (tropical hardwood) forests of that region. inaccessible forests also expand in both ecological scenarios for that region, but those changes are suppressed here in order to show changes for the most important market species. 3 hamburg and uiuC refer to the climate scenarios used for the ecological predictions. D E V E L O P M E N T A N D C L I M AT E C H A N G E D I S C u S S I O N PA P E R S 11 Table 4. perCenTage Change in Timber grOWTh raTes by 2145. Hamburg UIUC BIOME 3 Predicted % % Change in Merchantable BIOME 3 Predicted % Change in Merchantable Change in NPP Timber Yield % Change in NPP Timber Yield High-latitude Forests North America 17 34 17 41 Europe 23 4 23 24 former Soviet union 53 44 52 66 China 36 27 38 32 Oceania (16) 10 13 29 low- to Mid-latitude Forests South America 46 42 23 23 India 45 47 28 29 Asia-Pacific 29 28 12 11 Africa 37 37 21 21 Source: sohngen et al. 2001. To summarize, all the developing regions show positive ecological scenario. Note that all the developing country growth in timber production to the year 2050. regions have exhibited timber harvest production Additionally, all the regions with non-negative growth increases both to 2050 and continuing to 2145. to 2050 under the Hamburg scenario also show contin- ued expansion to 2145. Also, all regions show timber For the period under consideration up to the middle of production expansion after 2050 under the UIUC the 21st century, total global forest timber harvests Table 5. perCenTage Change in regiOnal Timber prOduCTiOn fOr 50-year Time periOds Hamburg UIUC Region* 1995­2045 2045­2095 2095­2145 1995­2045 2045­2095 2095­2145 High-latitude Forests North America (1) 12 19 (2) 16 27 Europe 5 2 14 10 13 26 former Soviet union 6 18 71 3 7 95 China 11 29 71 10 26 31 Oceania (3) (5) (10) 12 32 31 low- to Mid-latitude Forests South America 19 47 50 10 22 23 India 22 55 59 14 30 29 Asia-Pacific 10 30 37 4 14 17 Africa 14 31 39 5 17 7 Total All forests 6 21 30 5 18 29 Source: sohngen et al. 2001. * time periods each cover a 50 year period. 12 A DAPTATION Of fORESTS TO CLIMATE C H A N G E analyses being on the supply side. Earlier sensitivity Table 6. perCenTage Change in regiOnal analysis shows that the projections are, to a large extent, Timber prOduCTiOn TO The year 2050, only minimally impacted by modest demand-side based On The ClimaTe sCenariOs used changes (see Sedjo and Lyon 1990). Based on the fOr eCOlOgiCal prOjeCTiOns assessment of these projections, the study suggests some adaptation measures to mitigate for damages likely to be Hamburg UIUC incurred, and makes preliminary estimates of costs. Region* 1995­2050 1995­2050 High-latitude Forests The results of this study are consistent with most of the North America (1) (2) studies of this question, as well as with the general find- Europe 6 11 ings of the IPCC Fourth Assessment of Climate former Soviet union 7 3 Change ( Easterling et al. 2007), which states that the China 12 11 changes globally range from "a modest increase to a Oceania (3) 13 slight decrease, although regional and local changes will low- to Mid-latitude Forests be large." It also notes that "production increases will South America 19 10 shift from low-latitude regions in the short term to India 22 14 high-latitude regions in the long term." The similarity Asia-Pacific 10 4 with the findings of this study is not surprising, since Africa 14 5 this study draws in large part on the IPCC findings and Total all Forests 6 5 on the literature that went into developing those find- ings. On average, most of the studies find forest produc- Source: adapted from sohngen et al. 2001. * time period cover a 55 year period. tivity and area increasing modestly Note: The results of the period 1995­2045 were straight-line extended to 2050. Uncertainties increase over the longer term, which raise concerns about the possibilities over the longer term. increase about 6 percent. The largest percentage The IPCC (2007) anticipates "significant forest dieback increases occur in the developing world, specifically toward the end of the century." However, forests cannot China, South America, India, the Asia-Pacific and expand indefinitely even in the absence of climate Africa. Europe and the Former Soviet Union also expe- effects. Indeed, dieback occurs as mortality overtakes a rience modest gains, with declines being experienced in forest. The dieback exacerbated by climate change is only North American. Oceania has a decline under one likely to be different--and indeed more severe--as the climate model and an increase in another. process of replacing earlier forests by forests more appropriate to the changing climate are established. 2.2 hOW O ur sTudy COmplemen Ts However, dieback need not threaten the adequacy of exis Ting WOrk timber supply, given the ability to salvage a portion of the mortality and the huge surpluses of forest stocks The approach of this study does not involve any new over the requirements of industrial wood demand. model runs. Rather the study draws from the existing literature and the results of earlier investigations, report- 2 . 3 m e Th O d Ol O g y ing the latest comprehensive projections in the litera- ture. The most comprehensive results are fused into this The basic approach of an analysis of the economic single report. These do not fit perfectly the precise impact of climate change on forests requires the inte- guidelines--for example, GDP and population--that grated use of three types of models: economic, climate were applied to some of the other World-Bank- models (GCMs), and ecological models. Specifically, a sponsored studies. However, the basic models used are number of global general circulation models (GCMs) not calibrated to either GDP or population. Rather they exist. These models provide climate change scenarios. make some simplifying assumptions on the demand side Ecological models are also needed to estimate the with the richness of the model and the focus of the response of vegetation to whatever climate change is D E V E L O P M E N T A N D C L I M AT E C H A N G E D I S C u S S I O N PA P E R S 13 anticipated. Together these models estimate changes in level of approximately 550 ppm, which is a doubling of vegetative composition, location, and productivity, which the 1998 atmospheric CO2 level of 340 ppm. are driven by temperature and precipitation change. Specifically, the models used for climate change predict from two equilibrium general circulation models Ecological models project the extent to which a specific (GCMs). Steady-state forecasts from the Hamburg climate change is expected to shift the geographic distri- T-106 model (Claussen; Bengtsson et al.1996) and the bution of plants and particularly tree species (Emanuel et UIUC model (Schlesinger et al. 1997) are used to al. 1985; Shugart et al. 1986; Solomon et al. 1996; predict changes in climate for 0.5 x 0.5 degree grid cells Neilson and Marks 1994). Responses by forests to past across the globe. climate change have consisted of the independent move- ments of the ranges of important tree species (Shugart et Globally, the Hamburg model predicts a 1°C increase in al. 2003). A critical issue in the location of natural forests temperature over land and water, while UIUC predicts a is the rate at which tree species migrate. This issue is less 3.4°C change. The Hamburg scenario predicts relatively important for plantation forests, because people can be larger temperature changes in the high latitudes involved in the replanting of the appropriate species for compared to the UIUC scenario, and the UIUC the new climate conditions. In addition, climate change scenario predicts larger temperature changes in the low is projected to alter tree productivity--in the aggregate in latitudes. These regional differences suggest that the a positive direction--through temperature and precipita- two climate models will have different regional impacts tion changes (Melillo et al. 1993). Although these studies on timber supply. are relatively old, there are no new studies that under- mine these results over the period to 2050. 2.3.1 Capturing ecological changes in the economic model Finally, the carbon dioxide fertilization effect may be an important enhancer of productivity. Although the An ecological model, the global terrestrial biosphere science is still inconclusive and size of the effect appears model BIOME3 (Haxeltine and Prentice 1996; to vary considerably (see Shugart et al. 2003, pp. 19­20, Haxeltine 1996), is used to predict vegetative changes for a more complete discussion of the literature), these that would be expected to be precipitated by the climate effects are usually introduced. Supporting the use of a changes predicted by the GCMs. The climate predic- carbon fertilization factor are the findings of Boisvenue tions are used by a global terrestrial biosphere model and Running (2006), which indicate that tree produc- (BIOME3) to estimate equilibrium changes in the tivity generally has increased in recent periods. distribution of timber species and the productivity of those species across the globe. Biomes are ecological This report, although not developing a new methodol- types that represent accumulations of different species, ogy, uses a consistent methodological approach that referred to as forest types. While some models predict now has a well-established literature. The study draws net primary productivity (Melillo et al. 1993) and some heavily from the results of Sohngen et al. (2001) utiliz- models predict global changes in the distribution of ing a modified version of the timber supply model forest types (Neilson and Marks 1994), most models do (Sedjo and Lyon 1990). This economic model is utilized not capture the two effects simultaneously. together with two climate models and an ecological model. The approach uses the climate change predic- The approach of Sohngen et al. (2001) considers two tions of two general circulation models: the Hamburg types of transition and optimizes over both effects. The T-106 model (Claussen 1996; Bergtsson et al. 1996) first involves forest dieback where some portion of the and the UIUC model (Schlesinger et al. 1997). Again, forest dies due to climate change. The second makes the there are no new studies that undermine these results transition without the dieback as forest regeneration over the period to 2050. more quickly fills in the gaps without a high disruption due to mortality. The dieback scenario also involves The analysis assumes that climate changes linearly until salvage logging, where the timber in the dead forest is 2060, at which time it stabilizes at an atmospheric CO2 salvaged and gradually replaced by regeneration. Under 14 A DAPTATION Of fORESTS TO CLIMATE C H A N G E dieback, the prices are slightly higher as the value of the yield; see Table 2) because Hamburg suggests that salvage is lower than if salvage timber were replaced by species movement in Europe causes mostly an expan- high-value timber from live trees. sion of forests into marginal shrublands in Mediterranean areas. While more productive than The stock of forests also depends on the movement of shrublands, these new forests are less productive than species across the landscape. Two different and likely current forests in Europe, and they lower the long-run extreme scenarios of dynamic processes that govern the average yield of all forests. The change is similar for the movement of species are used to capture this movement: UIUC scenario (23 percent change in NPP and 24 dieback and regeneration. As forest types move because percent change in merchantable timber yield) because of climate change, the dieback scenario predicts the loss UIUC predicts mostly conversions of northern species of a large fraction of the existing stock (King and to southern species and less forest expansion (Table 1). Neilson 1992; Smith and Shugart 1993). By directly affecting stock, dieback can cause net growth in our Note, however, that productivity increases over time are timber types to decline even if NPP is positive. Dieback different from a future loss of biomass. Forests cannot also alters timber harvests because some of the stock expand forever. Thus, even with higher growth, forest that dies back will be salvaged. This salvage enters the stock will inevitability decline for a period after a period market through harvests. The proportion of salvage in of initial increase. Thus the two statements in the IPCC each timber type varies by region. 2007 report cited above earlier (pp. 227 and 275), projecting increased growth and a decline in biomass at This approach assessing two effects is important because some future time, need not be in fundamental conflict. changes in net primary productivity (NPP) can affect species dominance within a forest type, and the species Although initial stocks are not heavily influenced by present can affect NPP. BIOME3 also includes carbon climate change in the regeneration scenario, harvesting fertilization through the physiological effects of increased behavior is affected. For instance, in northern regions carbon dioxide on water use efficiency of plants. where it becomes possible to introduce southern timber types that grow faster, landowners may have an incen- In the long run, the yield of forests is likely to rise tive to harvest even young trees to make way for new because of these two factors. First, BIOME3 predicts species that grow more quickly. that climate change increases the annual growth of merchantable timber by raising NPP (the "BIOME3" The results are reported in Sohngen et al.(2001) for the columns in Table 2). This is the only effect captured by two climate GCMs: Hamburg and UIUC. In the most other climate change studies of forests ( Joyce et al. Hamburg scenario, BIOME predicts fairly large losses 1995; Perez-Garcia et al.1997; and McCarl et al. 1999). of existing timber stands in high-latitude regions, but a Second, BIOME3 predicts that more productive species global forest expansion of 27 percent and a 38 percent move poleward. In the long run, this tends to increase increase in productivity. With the UIUC scenario, the average timber yield for most regions by increasing predicted losses of existing stands are even more wide- the area of more productive species, although the effects spread, overall forests expand less (19 percent), and depend on the climatic predictions. For example, the productivity increases less (29 percent). From this prediction for North America from the NPP increase approach, which includes net primary productivity alone is that long-run timber yield should increase 17 changes and the carbon dioxide fertilization effects, the percent, but with the expansion of southern species into projected changes in the distribution of timber species territory previously occupied by northern species, the and the productivity of those species by location is economic model predicts an average (continental) obtained. increase in merchantable yields of 34 to 41 percent. Alternatively, long-run merchantable timber yield in 2.3.2 The economic model Europe is not predicted to increase as much under Hamburg as would be predicted by the change in NPP The economic model (TSM) of Sohngen et al. (1999) from BIOME3 alone (that is, a 23 percent change in provides the base case (no climate change) results. This NPP and 4 percent change in merchantable timber model is applied to the vegetative changes to project D E V E L O P M E N T A N D C L I M AT E C H A N G E D I S C u S S I O N PA P E R S 15 changes in industrial wood availability and costs that are The model includes consideration of forest manage- reported in Sohngen et al. (2001). The period examined ment and silvicultural practices, alternative species, as is that up to 2060, approximately the same as the 2050 well as various growth rates, harvest costs, and deliv- target called for in the World Bank's terms of reference. ered costs to mills. The model adjusts the level of The results of Sohngen et al. (2001) are adjusted in this management to economically optimum levels, and report to fit the 2050 time frame. The model focuses on provides for the introduction of new lands to establish net primary productivity (NPP) and assumes a carbon new plantation forests through time where economi- fertilization enhancement of 35 percent (Haxeltine cally justified. Since the model includes many different 1996). Although some believe this to be too high land classes and a variety of site and climatic condi- (Norby et al. 2005), the consensus is that fertilization is tions, these give rise to a host of individual regional positive and the direction of forest growth in recent supply curves. Locational considerations and transport decades is empirically borne out by the results of costs are built into the model, given the relationship Boisvenue and Running (2006). between the regional mills and the major market locations. BIOME3 provides more disaggregated results than the economic model can use. The data is aggregated and The model follows each land class through time, noting provides predicted effects for each contiguous forest the age and size of the various trees. An optimal type in BIOME3 for each region in our economic economic rotation is determined endogenously within model. These aggregated effects are used to predict the model. However, that rotation may vary with the changes in average productivity, changes in forest types, market price. Each period, the separate supplies are and the area of land that can be regenerated in each aggregated and, together with demand, a price that timber type, in the economic model. clears the market is determined. The model is forward- looking (rationale expectations) and thus considers The basic economic model utilized was developed as an current demand and supply conditions in the context of optimizing control theory model designed originally to future conditions. It maximizes the sum of producers' focus on industrial timber supply. The model is designed and consumers' surplus for each period and for the to examine carefully the various aspects that go into system. timber supply by region and land class. Supply is provided by a number of regions that have varying loca- Given the supply of various producers and regions tions, species, site conditions, and harvesting and trans- together with global demand, the model determines port costs. Initially, the supply regions consisted of optimal harvest levels and forest management invest- twenty-two homogeneous land classes; substantial detail ments through time. This model has appeared and on the various supply sources of timber can be found in been utilized in a number of published papers and Sedjo and Lyon (1990). Initially, a large nebulous area of reports to address not only timber supply issues land, much unmanaged with limited details, was assumed (Sohngen et al. 1999), but also the questions of forest to autonomously provide a certain portion of the world's carbon sequestration (Sedjo et al. 2000; Sohngen and industrial wood. Subsequently, additional regions have Sedjo 2006) and long-term international trade adjust- been added to the model as greater detail became avail- ments (Daigneault et al. 2007). The version of the able. About fifty regions were used in the 2001 version, model utilized in this study is that which examined which generated the result utilized in this study. The forest modifications in response to climate change model is designed to capture the inter-temporal transi- (Sohngen et al. 2001). That methodology used climate tion nature of the forest inventory, with young trees change estimates of GCM, to which the ecological becoming older and experiencing growth. Both natural literature was applied to create projections of the forest and plantation forests are included, although as different ecosystem around 2050. The underlying economic land classes. Growth is unmanaged in natural stands, but projections model for this period is applied to this subject to modification through forest management. 2050 forest. The approach reports and compares the Plantations are managed intensively. Additional areas of situation under two "climate change scenarios" with plantation also can be added gradually, subject to the the projections of the "base case," i.e., without changes availability of suitable land and economic returns. due to climate change. 16 A DAPTATION Of fORESTS TO CLIMATE C H A N G E 2.4 d emand 2.4.1 possible changes in demand Contrary to earlier FAO predictions of fast-growing Although the demand for industrial wood has been sta- demand for industrial timber to 2.1 billion m3 by 2015 ble and predictable over time, the expansion in the use and 2.7 billion m3 by 2030 (Sedjo and Lyon 1983), of raw wood to energy uses in the form of biofuels, bio- actual demand growth has been much slower. For exam- mass energy, and other energy uses could dramatically ple, current demand for 1.6 billion m3 is just slightly change the trajectory of future demand (Sedjo and above the demand for 1.5 billion m3 in the early 1980s Sohngen 2009). Wood is clearly a potential substitute (FAO 2005a). Additionally, there is little reason to for fossil fuels and, since carbon dioxide can be viewed expect the very modest growth trend in industrial wood as recycled in the biological system, wood energy has use to change in the foreseeable future (Sedjo 2004). substantial appeal. Should wood energy of this type be- Although some markets are growing, others are declin- come important, this would almost surely escalate the ing. For example, major segments of the paper market-- demand for wood, thereby invalidating all of the current for example, newsprint--have decline markedly in some projections regarding the future demand for industrial parts of the world with the advent of the wide spread wood. Although wood energy is not technically an in- use of the Internet. Also, paper recycling is reducing dustrial wood demand, it would draw from essentially demand for virgin fiber. Recent projections of the FAO, the same natural resource base as industrial wood. Wood as well as models of the global forest sector, often is viewed as renewable and as recycling the emitted car- assume the continuation of the more modest demand bon and thus not contributing to the long-term buildup growth to the range of 1.8­1.9 billion m3 by of atmospheric carbon. Although it appears unlikely that 2010­2015. traditional fuelwood will expand significantly, some model-based estimates project an increase in biofuel de- World demand is imposed in this model, but in much mand during the next 50 years by as much as a factor of less detail than supply. The model assumes that ten (Alcamo et al. 2005). In many industrial countries, demand will increase very modestly over the next 100 biofuels, particularly ethanol from grains and other plant years. Demand is initially position to clear the market materials--such as sugarcane--have already become an in the base period. It is then shifted out through time important source of nonconventional transport energy. at a decreasing rate asymptotically approach a steady Biofuels derived from cellulosic biomass--fibrous and level at a period 100 years out. This approach is used wood portions of trees and plants--may offer an even for two reasons. First, projections based on population more attractive opportunity as an alternative to conven- and GDP have provided notoriously inaccurate tional energy sources. In addition, wood cellulose can be projections on the high side (Sedjo and Lyon 1990, used in gasification processes--for example, the integrat- Shugart et al. 2003). Second, since the model is ed gasification combined cycle (IGCC) process--to pro- forward looking with trees growing through multi- duce synthetic gases, including hydrogen. These gases decades periods, mathematical convergence required can be further used to produce energy directly, or as a movement to a long-term steady state. The model feedstock to produce a variety of energy products, in- used assumes demand is shifting at 0.4 percent annu- cluding not only ethanol but also biocrude, using pro- ally initially gradually converging to a stable situation cesses such as Fisher-Tropsch. Wood-fired gasification in 100 years. plants can be constructed as stand-alone projects, as is now under consideration in some locations. An intrigu- Although traditional fuelwood is not model in the ing possibility is that new gasification biorefineries re- model or this analysis, it is unlikely to upset the projec- place aging traditional boilers in existing pulp mills tions. Global fuelwood use appears to have already (Larson et al. 2008). Pulp mills have large energy re- peaked at 1.9 billion m3 and is stable or declining quirements and are designed to facilitate the flow of (Goldammer and Mutch 2002). large amounts of wood. This study, however, assumes that changes in the demand for wood for energy purpos- es will be modest and have a negligible impact on overall industrial wood demand. D E V E L O P M E N T A N D C L I M AT E C H A N G E D I S C u S S I O N PA P E R S 17 2.5 hOW We represenT T he increased timber supply derived from fast-growing fu Ture--2010 TO 2050 industrial wood plantations located in subtropical regions, and including Australia, New Zealand, the 2.5.1 The baseline results without climate change Asia-Pacific, and parts of Asia. Using the dynamic global timber market model results The driving force in global timber production--and the developed by Sohngen et al. (1999) to project future incremental increases in timber production--has been timber supply in the absence of climate change, the the expanding area of managed subtropical plantation model maximizes the net present value of consumer forest. As has been true in recent decades, most of the plus producer surplus in global timber markets. It opti- incremental increases in production are projected to mally manages harvest rotations, timberland area, forest occur in plantations of non-native species--such as management investments, and age-class distributions of southern U.S. pine, Caribbean pine, Monterrey pine, forest types in about 50 land classes, which are reported and eucalyptus--established in subtropical regions of in 10 regions worldwide to 2145. A slightly updated the world, including most importantly South America, version of the model was used by Daugneault et al. but also parts of Africa, Asia and Oceania. (2007) to examine the effects of changes in exchange rates on production and trade flows. The basic run of In Figure 5 higher (real) prices for the no-climate- that model, which did not assume climate change or change scenario (base case) are projected. In the base- exchange rate changes, was used as the updated base; its line scenario, timber prices are projected to rise results are presented in Figure 6. The global model approximately 0.4 percent per year during the period to covers all major timber producing regions of the world. 2050 as increases in demand are anticipated to slightly out-run productivity increases. As noted, most of the In the absence of climate change, the world's overall growth in production in the base case is projected to area of forest is projected to decline over the 21st occur in plantations of non-indigenous species estab- Century. Figure 4 provides historical and projected esti- lished in subtropical regions of South America, Oceania, mates of timber harvests by major global regions in the Asia-Pacific, and Africa. base case from 1960 to 2060. Even in the absence of climate change, the projections show major changes by These areas have been successful in converting marginal region. This includes the historical data that show the agricultural lands and native forestlands to high-value harvests from the former Soviet Union states, which dropped dramatically in the early 1990s. Projections estimate that harvests of those states will not reach figure 4. presenTs prOjeCTiOns Of levels of the late 1980s until the 2030s. The projections Timber harvesTs by regiOn fOr The also anticipate U.S. harvest leveling off in the 1990s and baseline sCenariO declining after 2020. Europe follows essentially the same path to about 2020, but production increases 900 800 thereafter and into the 2030s, after which it declines. 700 Million cubic meters Canadian production continues its rise until about 2015, 600 after which it too declines. Throughout the entire 500 400 period, South American output is projected to increase, 300 reflecting the continuing expansion of planted forests 200 100 and timber production from that region. The projec- 0 tions indicate that production from the "rest of the 1961 1970 1980 1990 2000 2010 2020 2030 2040 2050 world" will not achieve 1990 levels again until after Year 2030, reflecting the full recovery of the forest Soviet Rest of the World United States Europe South America Canada states. The "rest of the world" increases also reflect increased harvests from a host of countries, including Source: daigneault et al. 2007. not only the former Soviet Union states, but also 18 A DAPTATION Of fORESTS TO CLIMATE C H A N G E forest plantations. The model conservatively projects increases only slightly over this time period, from 1.64 subtropical plantations to increase in the baseline by billion m3 to 1.71 billion m3 per year. The regional 273,000 hectares per year on average, with 27 percent of results are reported in Tables 1, 2, 3, and 4 to the year the new plantations predicted to occur in South 2045. For all regions except Oceania, the projected America, 20 percent in Oceania, 8 percent in Asia- changes in the direction are in the same through time, Pacific, and 25 percent in Africa (Daigneault et al. although the magnitude of the change varies somewhat. 2007). The baseline plantation establishment prediction used in the base case is somewhat lower than the recent With climate change, the ecological model BIOME3 historical average annual increase in non-indigenous predicts large conversions from one forest type to plantations in subtropical regions of 6 million hectares another, large conversions of non-forest land to forest- per year for the period 1980 to 1990 (FAO 1995). land, and higher NPP. Using the Hamburg climate scenario, BIOME3 predicts fairly large losses of existing The effect of subtropical plantations is understood timber stands in high-latitude regions, but an overall when it is recognized that they commonly grow at rates global forest area expansion of 27 percent and a 38 in excess of 10­15 m3 per hectare per year compared to percent increase in productivity. With the UIUC many temperate forests, which grow at only 2­5 m3 per scenario, predicted losses of existing stands are even more hectare per year (Bazett 1993). The total area of these widespread, overall forests expand less (19 percent), and fast-growing industrial wood plantations is projected to productivity increases less (29 percent). Although the expand from around 70 million hectares currently to results are limited by reliance on only one ecological around 130 million hectares in 2050. Total wood model, these ecological results are broadly consistent production from these subtropical fast-growing planta- with the literature (Watson et al. 1998; Gitay et al. 2001). tions is projected to increase from about 200 million m3 per year, or about 13 percent of total wood supply, to Four transient ecological change scenarios are developed about 700 million m3 [per year, or about 41 percent of to provide decadal predictions of the ecological variables total wood supply by 2050. Total production from all described above. These include a dieback and regenera- planted forests is forecast by some to reach 75 percent tion scenario for both the Hamburg and UIUC climate of total global production by 2050 (Irland et al. 2007). scenarios. The dynamic economic model takes these decadal predictions as exogenous, and predicts how 2.6 gl Obal r esulT s: timber markets may react. The economic model uses dynamic optimization techniques to predict how a risk- 2.6.1 With climate change impact on the forest neutral supplier would change planting, management, sector Prices are a signal of relative scarcity or abundance. Figure 5 presents wood price projections until 2140 for figure 5. glObal Timber priCes Over Time both the baseline case and for the two global warming scenarios (Sohngen et al. 2001). Note the baseline with 150 no climate change has the highest prices, reflecting 1990 US$ per Cubic Meter 130 greatest relative scarcity. The two climate-ecological scenarios give lower prices, with the dieback price some- 110 what higher than that of the regeneration scenario. In 90 either case, the implication of the study is the timber 70 supplies will be enhanced by anticipate climate warming. 50 2000 2020 2040 2060 2080 2100 2120 2140 These projections suggest that global timber prices Year (denominated in 2000 real southern U.S. softwood log Baseline Case Hamburg Regeneration prices) rise from $114 per m3 to $132 per m3 from Hamburg Dieback UIUC Regeneration UIUC Dieback 2000 to 2050, an increase of nearly 0.4 percent per year. Source: sohngen et al. 2001. However, the total quantity of timber produced globally D E V E L O P M E N T A N D C L I M AT E C H A N G E D I S C u S S I O N PA P E R S 19 and harvesting decisions. Aggregating these changes timber stocks initially, and it reduces harvests initially. across the global market, the model predicts how harvest The baseline projections predict most of the increase in quantities and therefore prices will change. The model timber harvests will occur in these subtropical regions, does not capture feedback effects from the market back and climate change appears to strengthen this trend as onto climate itself, because these feedbacks are expected managers adapt quickly with fast growing, non-indige- to be small. However, the market does affect ecosystem nous plantation species. dynamics, as market forces can facilitate change by harvesting slower growing trees or trees destined for Early forest losses are offset by moving more productive dieback and planting trees designed for the new climate. southern species further north. "Net Area Change" in Table 1 is the prediction of the relative area of forests 2.6.2 regional impacts after climate change by BIOME3. BIOME3 predicts relatively large increases in forest area. However, given The economic model predicts that global timber supply the low productivity of these polar forests even with increases and prices decline relative to the base under all climate change, the newly established forest stocks will scenarios (Figure 5). As expected, the regional and be small in 2050. In any event, they are unlikely to be temporal effects on timber production for the two climate major harvested forests for the reasons below. In addi- scenarios are different (Table 3). In the Hamburg tion, model's assumption--that forests do not shift into scenario, production increases most heavily in low- to high-quality agricultural land--limits most of the mid-latitude regions because climate changes are expansion to conversions of one forest type for another predicted to be mild and the trees respond well to the or to shifts of low-value grasslands and tundra to higher levels of carbon dioxide. In the near term (1995 to forests. Accessible forests in the economic model conse- 2045), the Hamburg Model projects the largest relative quently increase by only 5 percent. Most of the increase production losses will centered in mid- to high-latitude in forestland is predicted to occur in inaccessible boreal regions of North America, the Former Soviet Union, and tropical regions (31 percent to 41 percent) that are China, Oceania, and Europe--regions that currently never used for timber harvests. supply 77 percent of the world's industrial wood (FAO 1996). These relative declines reflect the large productiv- In summary, for the most part, the changes in forest ity increases in the low- to mid-latitude regions, includ- areas are consistent with recent experiences in markets. ing South America, India, Asia-Pacific, and Africa. In the To the year 2050, most of the losses occur in high-lati- long run, productive species replace the lost forests so that tude regions, with the lower latitude developing world long-run productivity increases. Initially, prices are rela- generally benefiting. There are slight losses in North tively lower in the regeneration scenario. In the long run, America in accessible forest area. Europe and the however, the period of conversion ends and the same Former Soviet Union gain forestland. productive forests take over, causing long-run prices to converge in both scenarios. The difference in prices 2.6.2 global "wet" and "dry" scenarios between the dieback and regeneration scenarios declines before the conversion process ends because it takes longer In general, a warmer and wetter climate is likely to for more productive species to take hold in the regenera- promote forest growth under many real world condi- tion scenario. In the UIUC scenario, production increases tions (Bowes and Sedjo 1993). In both of the GCM are similar for all regions, but larger tropical warming models used, the warming scenarios showed an increase reduces productivity gains in low- to mid-latitude regions. in average NNP over the base, and forest growth in the aggregate benefited. The Hamburg results might be Although the Former Soviet Union is predicted to gain viewed as the "wet" results, with this model giving significant production relative to the baseline in either generally higher productivity (NPP) outcomes, while scenario, these increases take many years to affect the UIUC results are modestly less productive and can markets because species grow slowly there. Europe be viewed as the "dry" outcomes. The Hamburg scenario harvests heavily during early periods to avoid economic generates an average increase in forest NNP above the losses from dieback in its generally older stock of trees. base of 38 percent, while the UIUC generates an NNP In contrast, North America has relatively younger increase of 29 percent above the base. 20 A DAPTATION Of fORESTS TO CLIMATE C H A N G E Of the two GCMs used, the Hamburg scenario consis- plantations in more suitable locations as replacements tently generated more favorable timber grow results for the old. Finally, tree breeding could be undertaken than the UIUC. Although they vary a bit by region, the to develop more resilient trees to better adapt to the results probably reflected more substantial precipitation changing or new climate. and a more favorable distribution of moisture. The overall results of both models suggest that precipitation Fire, disease, and infestation could be part of the adap- and moisture may have been slightly improved in the tation process by clearing away the old forest as part of aggregate. However, carbon dioxide fertilization also the process of bringing in the new (Sedjo 1991). was a major contributor to the positive results. One Although these may be part of the adaptation process, effect of carbon dioxide fertilization is that it allows the control of these forces is probably desirable, both to plant to use water more efficiently, thereby providing allow for increased salvage and to minimize damage to the potential to offset some declines in moisture. development in the forest. The relevant types of costs could include programs and training in fire, pest, and 2.7 hOW C limaT e C hange impaCTs a r e disease control. Also, the costs of the relocation of a Cal CulaT ed plantation are likely to be higher that the costs of replanting at an existing site. Finally, there are losses · Events that damage forests could include fire, infes- associated with tree damage, even if salvage is successful. tation, disease, and wind-throw. All of these might Fewer trees are harvestable and trees exposed to fire be expected to be exacerbate in a major climate have more limited uses than harvested growing trees. warming. In many cases the event need not be Obviously, these are mitigating and adapting activities unusually extreme, but might simply represent a sit- and real losses will result. Good management, however, uation where the forests, under stress due to the can reduce these costs and losses. changing climate, are increasingly susceptible to the various events above. Under stress from climate 2 . 9 h O W C O sT s Of a d a pTaT i O n a r e change, the forest could experience dieback due the C a lC u l aTe d changes, thereby increasing the probably of wild- fire, wind throw, etc. The costs of establishing a new tree plantation depend · Additionally, extreme events generated by climate upon the site and general economic conditions within change could put healthy forests at greater risk. This a country. Establishment costs, including land, could would be particularly true for windstorms and adja- run about $1,000/ha for a new site. Replanting a stand cent wildfire. These concerns are considered below. after harvest is approximately one-half of that (Sedjo 1983, 2004). Thus, the incremental costs of relocating 2.8 hOW CO sT s Of adapTaT iOn a r e plantations is roughly $500/ha. Rehabilitation of an defined existing forest is likely to be a different type of project. A 1998 World Bank project in India (#49477) put the Adaptation for forests to climate change could occur costs of the rehabilitation of 27,000 ha of forest at naturally, though natural regeneration and tree migra- about $18.8 million, or about $666/ha. A World Bank tion. However, for timber forests, adaptation to main- fire suppression project in Brazil (PO7882) was put at tain continuous industrial wood production may require $1.4 million in the southern Amazon. Obviously, salvage logging of disturbed forest. Additionally, anticipation of climate change induced events and disturbed forests could be replanted in species more mitigating actions will not always prevent damages, suitable to the changed climate, and the plantation and the extent of the climate induced damages can still forests could be relocated by establishing new be substantial. 21 3. resulTs Of damage OffseT 3 . 2 C O u nTry C a s e s Tu d i e s : br a Zi l , s O uTh a f r iC a , a n d C h i n a invesTmenTs 3.1 i nvesTmen T COsT s (upfrOnT a n d The climate effects on forests and industrial wood main T enanCe) in T he baseli n e ( nO production in Brazil, South Africa, and China, countries C limaTe Change) sCenariO chosen by the World Bank, are discussed below. The focus in these countries is on planted forests. All three countries have substantial volumes of timber produced About 0.5 million ha are harvested each year in devel- from their planted forests, and all three have expanded oping countries (assumes 200 M3/ha), of which about their planted forest estates in recent years. 200,000 ha or 40 percent are in tree plantations. If 10 percent of the plantations, 20,000 ha, need to be relo- Figure 6 shows that China and Brazil are both among cated each year, at $1,000/ha, the replanting investments the leading countries in forest plantation establishment, costs would be about $20 million worldwide. However, with China leading the world and Brazil ranked the incremental costs associated with relocation are esti- number seven. Brazil has been concentrating on wood- mated at about one-half the replanting costs, since producing forest; and unlike China, which has a large replanting would occur in any event and the incremen- portion of its planted forest dedicated to environmental tal costs would be those for accessing and preparing the and protection objectives, Brazil has been rapidly new site. Thus total global replanting costs would be about $10 million annually. Incremental fire control costs plus funds for rehabilitation of natural forest could be about $20 million annually. Rehabilitation area could be about $20 million; that is, 40,000 ha @$500/ha. This figure 6. fOresT planTaTiOn develOpmenT might have only a minimal effect on harvest level, since area (000 ha) the rehabilitated areas may not be an important part of the timber base. The total global incremental cost for Thailand Ukraine 4425 Iran relocation and rehabilitation could be approximately 4920 2284 $50 million per year for the developing countries. Brazil 4982 However, the amount related to timber and fire control Indonesia is about $30 million, since the replanted costs could be 9871 China 45083 viewed as the responsibility of the plantation ownership. Japan 10682 Although fire suppression costs can be very high, the USA relevant cost estimates for this report are incremental 16238 India costs related to climate change. In the U.S., much of the Russia 17340 32578 current fire suppression activity is unrelated to timber harvests and relates to protecting development in and Source: reprinted from seppala et al. 2009. adjacent to forests 22 A DAPTATION Of fORESTS TO CLIMATE C H A N G E increasing its production of industrial wood. South available for concessions. However, actual forest conces- Africa, the third case study country, has had a much sions appear to be only about 300,000 ha. Sustainable more modest expansion of planted forest, but it has systems involve low-intensity selective logging, with only provided the base for a domestic pulp and paper indus- a very few trees harvested per ha. The goal involves try, which is very activity in international trade. harvesting 30 m3/ha in large trees each 30 years. This intensity would involve the harvesting of only 1 m3 ha/ Figure 7 provides a global overview of precipitation yr on average. The major environmental effects of using the Hadley GCM. Hadley has one of the largest harvests in these areas probably involves the creation of increases in maximum temperatures and also has severe roads and the possibility of spontaneous migration that precipitation limitations for some regions. Note that for could lead to land-use changes. the regions examined in this study--Brazil, South Africa, and China--precipitation is positive for forestry Although important, natural forests continue their production in southern Brazil and southeastern China, decline in significance as sources of industrial wood, as but not as promising for forestry in South Africa. Brazil plans to continue to establish an additional 500,000 ha of plantation forest annually. About 1 3.2.1 brazil percent of Brazil's land area--or about 6 million ha (Seixas 2009)--is planted forest, but this is the core of The forest resource Brazil's forest industry. In recent years, it has planted or Brazil's tropical forests make up 42 percent of its total replanted about 600,000 ha annually, about 40 percent land area, compared to 1 percent for plantations (Figure of which are newly established plantations. Eucalyptus 8). It is estimated that about one-half of the total value and pine constitute 5.6 million ha, or about 93 percent of industrial wood (about $22 billion)--but only about of the total planted forest. Eucalyptus is found predom- 20 percent of the over $10 billion in wood product inantly in the southeast and pine in the south. exports--comes from the natural forest sector. Brazil Currently, eucalyptus is found in warmer regions than also has harvests from its tropical forests. There are 210 pine, in part because it is frost-sensitive (Figures 9 and million ha of federal forest, of which 12 million ha is 10). Tree breeding is currently under way toward the figure 7. Change in preCipiTaTiOn, frOm 2000 TO 2050 Source: provided by gerald nelson. D E V E L O P M E N T A N D C L I M AT E C H A N G E D I S C u S S I O N PA P E R S 23 figure 8. braZil area: 8.5 milliOn km 2 figure 9. euCalypTus 3,800,000 ha Others 9% Regenerating forest 18% Agriculture 10% Pasture 20% Natural Forest 42% Planted Forest 1% = 100,000 ha Source: prof. fernando seixas, esalQ/usp, piracicaba, sp brazil (2009). Source: prof. fernando seixas, esalQ/usp, piracicaba, sp brazil (2009). development of frost-resistant eucalyptus trees. This could expand the area suitable for plantations. Brazil estimates the 2007 sustainable harvest of its pine and Should a warming occur that moved forest, or forest suit- eucalyptus plantation at 191 million m3 annually, with ability toward the poles, the movement of the planted eucalyptus production being more than twice that of forest would be toward the south of Brazil. Thus, both pine (Seixas 2009). forest types could likely be relatively easily shifted to the area of Brazil toward the south. In a country the size of As a result of increasing establishment of fast-growing Brazil, there appears to be adequate room for movement industrial wood plantations, South America generally to the south while still remaining within Brazil. and Brazil in particular is projected to continue expand- ing its market share, experiencing an annual increase in In general, eucalyptus is the preferred species due to its production of approximately 0.8 percent per year over very rapid biological growth. Global warming could be the next 50 years. Under the baseline, most of these met by adjusting species or, if necessary, relocating increases are derived from harvests in industrial wood plantations. The area of land devoted to plantations in South America is projected to more than double during the coming half century, from 10.7 million hectares in figure 10. pinus 1,800,000 ha 2008 to 26.7 million hectares in 2050. Although total harvests are expected to increase in the region, baseline harvests from natural tropical and subtropical forests are projected to decline over the next 50 years. Industrial wood plantations are projected to account for as much as 71 percent of the timber harvested from all of South America by 2050 (Daigneault et al. 2007). Figure 9 identifies areas of major eucalyptus plantation activity, while Figure 10 identifies the areas with major pine plantations. Most of the plantations are in the area = 100,000 ha that was formerly the coastal forest, savannah forest, or caatinga (dry forest vegetation). The north-south range is somewhat greater for the eucalyptus than the pine. Source: prof. fernando seixas, esalQ/usp, piracicaba, sp brazil (2009). 24 A DAPTATION Of fORESTS TO CLIMATE C H A N G E plantations. Warming would probably allow continued, The alterations that climate change will bring to the perhaps greater, expansion of the planted area of forest, tropical forest area of Brazil, largely in the Amazon since few plantation areas would need to be abandoned region, remain to be seen. A few climate models suggest and cooler areas should warm. The existence of a large major vegetative changes. However, most suggest that number of eucalyptus species would allow, in principle, the tropical forest will persist. In any event, the changes the substitution of a more suitable variety. A word of anticipated between now and 2050 appear unlikely to caution, however, since knowledge of the behavior and dramatically disturb the overall forest or, for this report, likely wood-producing performance of the various euca- the timber production drawn from it. Over the longer lyptus and pine species is currently limited to a rela- period, should climate change be such that forests tively few species. Additional research in this area could persist, changes in tree species are to be expected in be important. general (Shugart et al. 2003) and also for tropical forests (Sedjo 2003). Should forest land change fundamentally, Brazil's recent planning and performance indicate they such as to grasslands, attempts to maintain land in plan to establish far more industrial plantation forests forest would probably be futile, and alternative land uses than envisioned in the projections model of Sohngen et would probably be both low-cost and wise. In summary, al. (1999). The government goal is to plant about it is likely that climate change would generate more 500,000 ha annually. Brazil has very rapid biological benefits than damages for Brazil's wood-producing growth of planted forest trees and views itself with a industry, and little public investment is warranted. substantial competitive advantage over most other indus- trial-country forests. Tree improvement has furthered this Offsetting investments advantage, as biological growth rates have continued to Although the relocation of planted forest might best be rise. Investment in the forest and wood-processing left to the private sector investors in those forests, there sectors has been substantial and is expected to continue are some sensible types of public investments to miti- at a relatively high level. Short rotations, continuing gate the impacts of clime change on Brazilian forests. A improvement and adaptation of genetically improved system of forest fire control is probably desirable both in stock, and large areas for expansion suggest the Brazilian the plantation regions and for natural forests. Fire is a forest industry is ideally positioned to adapt to climate continuing problem in parts of the Brazilian forest change. independent of climate change, and the World Bank has a history of supporting fire control capacities. (See Global change the project appraisal document entitled "Brazil-- Global change in Brazil is expected to involve warming Amazon Fire Prevention and Mobilization Project in the plantation areas of the southeast and south. March 2001"). Although natural forests could provide a However, the cost of warming to Brazil's planted forest useful agent to facilitate adaptation to the new climate industry is likely to be minimal. Warming would expand (Sedjo 1993), a fire control capacity is desirable to limit the frost-free areas suitable for eucalyptus, thereby damage to infrastructure and development around the allowing them to be established further to the south. forest. In addition, projects to promote forest rehabilita- With warming, pine could continue to be planted and tion on a selected basis, especially in the natural forest, producing where it is currently located. Adjustments may be desirable. Since wildfires in subtropical Brazilian could be made to the warming either by continuing to forests are common, a program with an annual addi- use the appropriate species of southern (yellow) pine. tional budget of perhaps $2 million, based on earlier Slash pine might be substituted for loblolly pine should World Bank fire projects, might be appropriate. the warming be excessive. Also, tropical pine--for example, Caribbean pine--could be introduced should 3.2.2 south africa temperatures rise substantially. In general, the array of pine and eucalyptus species currently in use and avail- South Africa has a very small area of natural forests able are well-suited to be relocated within these regions that are largely in small scattered patches. The total area to address a regional warming. The same sets of species of 327,600 ha constitutes only about 0.2 percent of the also offer the ability to adjust within limits to changing land area of the country. Open natural savanna wood- precipitation and moisture conditions. lands occupy another 28 million ha (DWAF 1996a). D E V E L O P M E N T A N D C L I M AT E C H A N G E D I S C u S S I O N PA P E R S 25 Forest plantations in South Africa were first initiated in moves eastward. Plantations are concentrated in an area the late 19th century Exotic tree species were tried, and of the country in provinces where rainfall exceeds 800 the area planted increased rapidly after 1920. Plantation mm per annum, specifically in a swath running from species consist largely of eucalyptus and pine. Trees were West Cape in the south to the northeast and parallel to planted on high-lying grassland areas with acceptable the southeast coast of South Africa to Limkpopo in the precipitation and other conditions suitable for forest northeast. Other provinces with substantial tree planta- plantations. Afforestation expanded more rapidly after tions are Mpumalanga, Kwazulu Natal, and Eastern the middle of the 20th century, and a domestic pulp and Cape (Figure 10). Worsening the uncertainty, South paper industry emerged. The annual rate of planting Africa is subject to drought (Vogel 2003). Under a situ- during 1981­90 was about 18,000 ha, and the total ation of global warming, the question of the viability of planted area was 1,487,000 ha in 1995. The pace of South Africa's forest plantations would likely depend planting has varied, decreasing from its peak of 45,000 more on the overall effects on precipitation and mois- hectares in 1991. In recent years, new afforestation has ture, rather than temperature. A number of climate proceeded at a level of around 11,000 hectares per year, studies suggest that South Africa is likely to have drier constrained largely by a limited availability of suitable winters. In addition, some of the land-limit constraints land, either in terms of water use regulations or in terms could be relieved with more secure tenure rights. of insecure land tenure. Furthermore, the Haley GCM (Figure 7) projections suggest moisture difficulties as climate changes to the Several large private companies together own about year 2050. one-half the plantation area, with a large portion owned by the state and some smaller private companies. South Given that a constraint to forest plantations is water, Africa's forest plantation area has continued to increase. should the climate turn toward greater dryness, State-owned plantations have primarily been geared to forestry and timber production would likely suffer. the production of sawlogs, whereas the privately owned Should dryness increase significantly, the likely plantations are mainly used for the production of pulp- outcome is a substantial decline in South African wood. The South Africa pulp and paper industry is timber production, with little possibility of invest- easily the largest in Africa, and it is an important inter- ments to offset the decline (irrigated planted forests national producer and exporter. rarely make financial or economic sense). The lands would likely revert to grasses, with grazing being Biological growth rates of some species--such as perhaps its most economically attractive use. pine--are about 16 m3/ha/year, with harvest rotations Alternatively, should moisture increase, the area suit- varying from 15 to 25 years depending on the able for plantation forests would likely increase, even intended use (Sedjo 1983). Eucalyptus growth is more independently of temperature, resulting in increased rapid and rotations shorter. The country's plantation industrial wood and wood products from South estate consists of 52 percent in pine, 39 percent in Africa. Moreover, increased moisture could potentially eucalyptus, and 7 percent in wattle, with the balance open the savannah lands of South Africa to planted comprising other species such as poplar. Water forestry, since moisture often determines whether the concerns have resulted in regulations--the forestry vegetation is forest or grasses. sector has been put under tighter control and the cost of planting has increased (SH 1999). The current Should the tree plantation value be lost, the financial strategy is to enhance the annual production of round- cost would be the value of the plantation as a lost asset. wood from the existing plantation areas by applying The present value of 1 ha of South African plantation genetic improvement and better silviculture to all forest was estimated to be about $3,700 in 1983 plantation areas. (Sedjo1983). Adjusting for inflation, the current value of a wooded ha could be in the neighborhood of Climate change $10,000 per ha. The cost of establishing additional As noted, a major constraint on planted forest in South plantations is estimated to be quite low in South Africa Africa is water. South Africa climate configures into due to the relative ease of site preparation costs and low arid and semi-arid in the west, becoming wetter as one labor costs (Sedjo 1983). 26 A DAPTATION Of fORESTS TO CLIMATE C H A N G E The role of public investment to offset climate impacts figure 11. sOuTh afriCan prOvinCes on industrial forests appears quite limited. Many forests are public, although the large paper industry is private. If climate were to undermine forest plantations, a sensi- ble approach might be to focus investments on retrain- ing of the displaced labor force. 3.2.3 China China is a large country with a variety of geographic, ecological, and climatic conditions (Figures 12, 13). Since the late 1970s, China's forests have made a remarkable recovery, in large part due to the govern- ment-sponsored program to establish large areas of planted forest. Indeed, China has been the world's lead- ing country in the planting of new and restored forests (Figure 7). Some of these are aimed at increasing industrial wood production, but large areas are also Source: department of Water affairs and forestry, south africa. dedicated to other reforested and afforested purposes figure 12. map Of China D E V E L O P M E N T A N D C L I M AT E C H A N G E D I S C u S S I O N PA P E R S 27 exception. This view is consistent with the estimates of figure 13. eCOsysTem areas by Type Sohngen et al. 2001 (Table 6). Wetlands and water bodies 1% Forests Figure 14 provides a focused look at the land cover of Sparse or barren vegetation; 11% snow and ice Heilongjiang province in the northeast. This assessment 15% shows forest decline through the 1990s, but a modest Shrublands, savanna and grasslands recovery since then. Urban and 44% built-up areas 0% Costs of adaptation Although China is an important producer and exporter Cropland and crop/natural vegetation mosaic of industrial wood products, it is a relatively modest 29% producer of raw industrial wood. Much of its wood used for processing is imported from a variety of suppliers, including Russia, the Asia Pacific, and North America (http://www.woodmarkets.com). As noted, its forest planting programs have at least two purposes: environ- (Figures 12, 13). An FAO report (2005) indicates that mental protection and the production of more industrial China's man-made forests have increased from 28 wood. Thus, while the reforestation program is directed million ha in 1986 to 48 million in 2001, or an average at adding more domestic wood to domestic processing, of about 1.33 million ha annually. Figure 6.3 of that this consideration is not critical for continued wood report presents an estimate that about 45 million ha of processing provided that wood imports are allowed to China's forest is planted. The FAO report (2005) also continue in a relatively unobstructed manner. indicates that China's forested land area has increased from 107.2 million ha to 158.5 million ha between For China, the challenge of climate change to its indus- 1986 and 2005. In a separate study that draws on the trial wood producing forests appears modest. The FAO data, Kauppi et al. (2006) estimate China's forest exception could be infestations, which have tended to area increased by about 1.5 percent annually in recent plague largely non-timber-producing poplar forests in years, among the most rapid worldwide. These numbers the interior. China is responding to this threat with suggest that about 11 percent of China's area was genetically engineered poplar trees, which are resistant forested in the mid-1980s, and that the portion in to the infestation. Most timber trees have not been seri- forests today is about 16 percent, given the recent forest ously adversely affected by the insects. However, infesta- area estimates. tions and genetic adaptations could increase the cost of adaptation. The overall effects of climate change on It is anticipated that China will continue to expand its forestry anticipated by 2050, as reflected in Sohngen et forest even in the absence of climate change. FAO data al. (2001) and in the IPCC map, suggest an overall reveal that China estimated about 86 million ha of improving situation for forestry and industrial wood timber forest and 62 million of protection forest for production in China. This situation should further be 2005, with protection forest increasing rapidly while improved due to the active policies of forest establish- timber forest expanded only modestly. A declining ment, management, and protection being undertaken by portion of the forest was being dedicated to firewood. the Chinese government. Adaptation costs that might be required by climate change may be modest. China's forests are located largely in the northeast and Productivity in the relevant regions is anticipated to southeast sections of the country. Fortuitously, the increase so that regions currently in forest appear likely Haldey map (Figure 7) suggests that both those to benefit from climate change. Additionally, China is regions--the Northeast temperate region and Southeast continuing to establish planted forests for both environ- subtropical region--will be modestly helped by climate mental and industrial wood purposes. Thus, should change to middle of the 21st century. More generally, climate-related problems occur in forest production, IPCC (2006) projects that most of China will experi- modest changes in the choice of new tree planting stock ence increased precipitation, the west being the should be sufficient to adjust to the modified climate. 28 A DAPTATION Of fORESTS TO CLIMATE C H A N G E figure 14. land COver Of heilOngjiang prOvinCe Late 1970s Around 1990 Paddy land Dry land Dense forestland Sparse forestland Dense grassland Sparse grassland Water body Built-up land Unused land Around 2000 Around 2004 N 0 37.5 75 150 Kilometers Source: shen, yin, and Qi 2009. In recent years, the World Bank has provided financial appear to be minimal. There appears to be little assistance to China for at lease two forestry develop- concrete reason to anticipate any serious investments in ment projects, both of which involved planting trees as offsetting the impacts of climate change on China's a component. However, the impacts of climate change industrial forests. In summary, it is difficult to see on China's industrial forestry sector through 2050 China's industrial wood situation deteriorating signifi- cantly over the next 50 years due to climate change. 29 4. limiTaTiOns 4 . 3 T r e aTm e nT O f i nT e r -Te m pO r a l Ch O i C e 4.1 TreaTmenT Of exT reme eve nTs Inter-temporal choice with forests and forest planta- Damages to wood-producing forests associated with tions could be associated with harvesting before the extreme events seem to be largely limited to the type optimal rotation age, where conditions suggest that of threats common to forests, such as fire and wind- climate change threatens to reduce or destroy the throw. These in turn could become more serious either timber crop. An early harvest may avoid most of this due directly to the extreme events or due to reduced loss (Shugart et al. 2003). forest health, such as infestation or disease related to the changing climate. These threats could be manifest 4 . 4 T r e aTm e nT O f " sO fT " a d a pTaTiO n as forest dieback. As discussed above, these types of measures problems can be addressed in part through salvage logging, which reduces the financial losses. In addition, I would take "soft" adaptation measures in the context regeneration, either natural or artificial, could be of the forest industry to refer to reliance on the natural promoted to facilitate the recovery of a forest. Note, resilience, mobility and reproductive capacity of the however, that for forests extreme events generally are forest. This natural resilience may need to be facilitated not independent, but rather associated with forest through, for example, efforts to ensure the absence of system biological weakness. This weakness can reflect obstructions to natural mobility. In addition, mobility either the age and/or health of the forest, and may be can be facilitated through more active human activities associated with the unsuitability of the forest types to promote mobility--such as aerial seeding--which that became established under the earlier climate although probably inappropriate for industrial forests, regime. New types may need to accompany climate could facilitate mobility among "natural" forests. change. 4 . 5 T r e aTm e nT O f Cr O s s - s eC TOr a l 4.2 TreaTmenT Of Te Chn Ol OgiC a l measures Change No serious cross-sectoral measures were identified. The Modest technological change is built into the basic obvious one would be the question of alternative land model and is not addressed separately for the industrial use among forestry and agricultural uses such as pasture forest industry. Technical change could also be part of and cropland uses. The Sohngen et al. approach does the adaptation process, such as tree breeding designed not allow for the automatic conversion of useful agricul- to facilitate adaptation to drought conditions or to resist tural land to forest uses as climate changes, unless those infestations associated with climate change. lands are not actively being managed or though a 30 A DAPTATION Of fORESTS TO CLIMATE C H A N G E conscious human decision to promote the land use of the regions or countries examined could have differ- change to forests. Indeed, much of the newly developed ent results with a different GCM. For forests, precipita- plantation area of the world reflects land use changes, tion is probably as important as temperature, at least in typically from abandoned and marginal agricultural use the temperature ranges under consideration. to intensive forest plantation management. Climate change, in the form of changing temperature and/or Regarding forest and industrial wood, useful research precipitation, could shift the comparative productivity advances may be found in the development of trees of an unmanaged natural site from some uses to differ- that have the ability to flourish under changing ent uses, such as from grasses to forest. climatic conditions. In addition, for industrial forestry, short rotations facilitate adaptation. It is likely that 4.6 a reas fOr fOll OW-up WOrk a n d future breeding will develop trees customized to the researC h advanCes site and that the genetic features of each new rotation will be adapted to the anticipated changing conditions. A major limitation of this study is the range of possible Short rotations are likely to be an element of the climate changes generated by the various GCMs. Any customized tree. 31 5. invesTmenTs and GHGs into the atmosphere, would have liabilities for these earlier as well as current emissions. More recent COmpensaTiOn: sOme ThOughTs transition countries, such as emerging countries like China and India, also are now major generators of The question arises as to what the proper public sector GHGs and so also have liabilities. The larger a country role should be in addressing climate-change-induced and the longer the country has been industrialized, the adaptation and/or compensation. This question focuses larger share of the GHG emissions are probably its particularly on the appropriateness of external support to responsibility. The developed vs. developing country the country--such as from the World Bank--for invest- dichotomy is an approximation of this reality. Thus, in ments in adaptation and/or compensation. Let me share concept, compensation should flow from developed to a few thoughts. Public sector support is often viewed as developing countries in recognition of the source and appropriate in the cast of severe catastrophic or near- size of the damages. catastrophic disasters. Climate change, natural or human induced, would probably fit. However, the nature of the How should the transfer be allocated between the event is such that substantial time is probably available public and private sector? Using the common law to anticipate and undertake adaptive responses to many paradigm, both private and public entities alike are of the shocks so as to mitigate the size of the direct eligible for compensation for damages from externali- damages. Obviously, something like the Kyoto Protocol ties. For forestry, natural forest restoration and/or is a move in the direction of mitigating warming and its compensation would seem appropriate regardless of consequences. Some of the activities suggested in this ownership if the source of the damages were identified. report are intended to reduce the damages by adaptation, Investments to reduce damages from fires, infestations, perhaps involving investments, from the warming that is wind-throw, storms, etc., should in principle address not successfully mitigated. these problems regardless of forest ownerships for the same common law reasons. For plantation owners, One can think of warming as an externality associated public or private, the damages are likely to be modest with the free or low-cost disposal of a "bad," in this case for the reasons articulated in this report. However, the GHGs, into the atmosphere. This approach to disposal loss of market values of the former forest plantation has been viewed as costless when, in fact, there are real lands could be large if those lands have few alternatives costs in the form of damages associated with GHG uses in the new climate; for example, with permanent build-up. In common law, the generator of a negative moisture reduction, as could occur in South Africa. externality is typically held liable for damages associated Finally, however, the rationale developed above may be with the externality. Thus, the countries of the devel- overwhelmed by real world economic and political oped world, which have a long history of releasing realities. 32 6. summary and COnClusiOns expected to migrate to areas more conducive to their needs. Carbon fertilization will probably increase growth rates at least marginally for most forests, The main components of the study are as follows: although this issue is scientifically less certain. However, forest damage will occur as existing trees become less · Establish a baseline and projected timber and fuel suitable for the new climates. The anticipated trends are harvests for the period 2010 to 2050. These esti- captured in this report by utilizing the projections of mates will be drawn from the existing literature. Sohngen et al. 2001. Although these projections were · Describe the nature of the different climate risks done several years ago, no new comprehensive projec- faced by the forestry sector. tions of this detail are available. Also, there are no new · By major timber producing regions, establish the scientific findings that would lead us to expect these quantitative and economic impact of climate change projections would change appreciably if updated. on the forestry sector. · Develop a data base of adaptation measures, with The general finding is that the future overall availability estimates of their costs and benefits, allowing for of industrial wood is likely to be more than adequate variations in costs by country, or by region. Draw on despite climate change, although the location of some information from relevant World Bank projects and/ forests and some supply sources could change. Forest or other sources. stocks and anticipated growth are more than adequate · By developing region, use the data base to estimate to meet anticipated future industrial wood demand. costs of adaptation as the minimum level of invest- Plantation forests are projected to increasingly supply ment required to partially offset or restore the tim- industrial wood requirements. Plantation forests have ber and fuel productivity of forests to their short rotations and can be planted in the species of without-climate change levels, recognizing that in choice, which can change to fit changing conditions, many cases one country or region's comparative thereby allowing maximum flexibility to adapt to advantage in forestry activities may be reduced or climate change. lost while another's is enhanced. · In addition, the TOR called for the utilization of Three countries are examined in detail: Brazil, South existing methodology and literature to a major coun- Africa, and China. They generally show different capac- try in each of the three regions. The major countries ities to adapt. Brazil has a large and growing forest agreed to were Brazil, South Africa, and China. plantation sector. With short rotations, a relatively large number of species to draw from, and large land areas The report extensively reviewed the literature. available for new or replacement sites, Brazil is in a Overwhelmingly, the literature suggested that the strong position to adapt its timber producing forests. world's overall forest area would probably change little However, should it have widespread moisture problems, most, likely expanding modestly. The tables show that its future supply potential could be compromised. forest productivity (NPP) is expected to increase in However, most GCMs suggest moisture will be most regions. As climate changes, tree species are adequate. Although most of China's industrial wood is D E V E L O P M E N T A N D C L I M AT E C H A N G E D I S C u S S I O N PA P E R S 33 imported, China appears to be in a strong position to more broadly. However, plantations offer many dimen- maintain and expand its forests and increase future sions for flexibility and adaptability. Even with infesta- domestic wood harvests, even in the face of climate tions, for example, plantations provide potential change. China has a very aggressive tree planting adaptability by allowing the replanting of species not program. Most of the industrial forest planting is antici- favored by the pests. Furthermore, planting allows for pated to occur in the southeast, a region that is gener- the introduction of genetically altered trees resistant to ally expecting adequate precipitation with climate the pests, either through traditional breeding or through change. genetic engineering. Thus, plantations, the growing source of industrial wood, provide more options in The third country, South Africa, is more problematic. addressing an infestation problem than would be avail- Plantation forestry has done well in South Africa, as the able in most natural forests. country has built a successful pulp and paper industry oriented toward export markets. However, moisture is One of the larger uncertainties relates to new sources of inadequate for trees in much of South Africa. The areas demand for wood. Although wood was once a major of tree plantations are near the edge of an adequate source of energy, most harvested wood today is used as moisture range. A number of GCMs project decreased industrial wood; that is, lumber and solid wood mate- precipitation, which suggests problems for the existing rial, and pulp and paper. However, there are growing plantations. The opportunities to relocate are limited demands for alternatives to fossil fuels, and wood is and could be further reduced by precipitation problems. commonly mentioned. Wood can be combusted directly or converted into various forms of energy, including Despite the generally optimistic assessment found in biofuels. The potential demand from energy sources is this report, uncertainly persists. Unanticipated problems huge and could dramatically alter the balance between related to climate change could take the form of wide- wood production and demand. This issue is beyond the spread infestation of forests in any of these countries or scope of this report, but cannot be dismissed. 34 7. referenCes Southern Forests." In S. Fox and R. Mickler, eds. Ecological Studies 128: The Productivity and Sustainability of Southern Forest Ecosystems in a Changing Environment. Bael, initial, and R. Sedjo. 2006. "Toward Globalization New York: Springer-Verlag. of the Forest Products Industry: Some Trends." RFF Discussion paper 06­35. Washington, DC: Resources Caspersen, J.P., S.W. Pacala, J.C. Jenkins, G.C. Hurtt, for the Future. http://www.rff.org/rff/Documents/ P.R. Moorcroft, and R.A. Birdsey. 2000. Contributions RFF-DP-06-35.pdf of land-use history to carbon accumulation in US forests, Science 290: 1148­1151. Bazett, M. 1993. Industrial Wood. Shell/WWF Tree Plantation Review Study No. 3. WWF (United Clark, J.S. 1998. Why trees migrate so fast: Confronting Kingdom), Panda House, Weyside Park, Godalming, theory with dispersal biology and the paleorecord. Am Surrey. Nat 152: 204­224. Bengtsson, L., M. Botzet, and M. Esch. 1996. "Will Claussen, M. 1996. "Variability of Global Biome Patterns Greenhouse Gas-Induced Warming Over the next 50 as a Function of Initial and Boundary Conditions in a Years Lead to Higher Frequency and Great Intensity of Climate Model." Climate Dynamics 12: 371­379. Hurricanes." Tellus 48A: 57­73. Climatewire. 2009. "Science: Experts struggle to cure Binkley, C.S. 1988. "A Case Study of the Effects of the big flaw in climate models." 2009/04/10/3. http:// CO2-induced Climatic Warming on Forest Growth and www.eenews.net/cw/ the Forest Sector: B. Economic Effects on the World's Forest Sector." In M.L. Parry, T.R. Carter, and N.T. Cramer, W., A. Bondeau, F.I. Woodward, I.C. Prentice, Konijn, eds. The Impact of Climatic Variations on R.A. Betts, V. Brovkin, P.M. Cox, V. Fisher, J. Foley, and Agriculture. Dordrecht, NL: Kluwer. A.D. Friend. 2001. title. Global Change Biology 7: 357­373. Boisvenue, C., and S.W. Running. 2006. Impacts of climate change on natural forest productivity ­ Daigneault, A., B. Sohngen, and R. Sedjo. 2007. evidence since the middle of the 20th century. Global "Exchange rates and the competitiveness of the United Change Biology 12: 1­21. States timber sector in a global economy." In Forest Policy and Economics. Elsevier, the Netherlands Bowes, M., and R. Sedjo. 1993. "Impacts and Responses to Climate Change in Forests of the MINK Region." Darwin, R.F., M. Tsigas, J. Lewandrowski, and A. Climatic Change 24: 63­82. Raneses. 1996. "Land Use and Cover in Ecological Economics." Ecological Economics 17(3): 157­181. Burton, D.M., B.A. McCarl, N.M. de Sousa, D.M. Adams, R.A. Alig, , and S.M. Winnett. 1998. Darwin, R. F., M. Tsigas, J. Lewandrowski, and A. "Economic Dimensions of Climate Change Impacts on Raneses. 1995. "World agriculture and Climate Change: D E V E L O P M E N T A N D C L I M AT E C H A N G E D I S C u S S I O N PA P E R S 35 Economic Adaptations." Agricultural Economic Report Foley, J.A., S. Levis, I.C. Prentice, D. Pollard, and Number 703. Washington, DC: U.S. Department of S.L.Thompson. 1998. Global Change Biology 4: 561­579. Agriculture, Economic Research Service. Gitay, H., S. Brown, W. Easterling, and B. Jallow. 2001. Davis, M. B., and R. G. Shaw. 2001. "Range Shifts and "Ecosystems and Their Goods and Services." In J. Adaptive Responses to Quaternary Climate Change." McCarthy, O. Canziani, N. Leary, D. Dokken, and K. Science 292: 673­679. White, eds. Climate Change 2001: Impacts, Adaptation, and Vulnerability. Cambridge, UK: Cambridge Department of Water Affairs and Forestry, South Africa University Press. (DWAF). 1996. "White paper on sustainable forest development in South Africa: The policy of the govern- Goldammer, J.G., and R.W. Mutch, eds. 2001. Global ment of national unity." Pretoria: Department of Water forest fire assessment 1990­2000. Rome: Food and Affairs and Forestry. Agricultural Organization of the United Nations. Easterling, W. and P. Aggarwal. 2007. "Food, Fibre, and Haxeltine, A. 1996. "Modelling the Vegetation of the Forest Products." In IPCC WGII Fourth Assessment Earth." Unpublished Ph.D Thesis. Lund, Sweden: Report. On Climate Change. Cambridge Faculty of Science, Plant Ecology, Lund University. Emanuel, W.R., H.H. Shugart, and M.P. Stevenson. Haxeltine, A., and I.C. Prentice. 1996. "BIOME3: An 1985. "Climate Change and the Broad-Scale equilibrium perrestrial biosphere modle based on Distribution of Terrestrial Ecosystem Complexes." ecophysicological constraints, resource availability, and Climatic Change 7: 29­43. competition among plant functional types." Global Biogeochemical Cycle 10(4): 693­709. FAO. 1995. "Forest Resources Assessment 1990: Global Synthesis." FAO Forestry Paper 124. Rome: Food and Houghton, J.T., L.G. Meira Filho, B.A. Callander, N. Agricultural Organization of the United Nations. Harris, A. Kattenberg, and K. Maskell. 1996. Climate Change 1995: The Science of Climate Change. Cambridge, FAO. 2004. Trade and Sustainable Forest Management - UK: Cambridge University Press. Impacts and Interactions. (FAO, Rome). IEA. 2004. World Energy Outlook 2004. Paris: Interna- FAO. 1996. "Forest Products." Rome: Food and tional Energy Agency, OECD Publication Service. Agricultural Organization of the United Nations. Idso, S. B., and B.A. Kimball. 2001. CO, enrichment of FAO. 1997. State of the World's Forests, 1997. Rome: sour orange trees: 13 years and counting. Environ Exp Food and Agricultural Organization of the United Bot 46: 147­153. Nations. Irland, L.C., D. Adams, R. Alig, C.J. Betz, C.C. Chen, FAO, FRAP. 2005. "Global Forest5 Resource M. Hutchins, B.A. McCarl, K. Skog, and B.L. Sohngen. Assessment Update 2005: China. 2004." Country 2001. Assessing socioeconomic impacts of climate Report on Forest Resources, Food and Agricultural change on US forests, wood-product markets, and forest Organization of the United Nations; Rome. recreation. BioScience 51: 753­764. FAO. 2005. "FAOSTAT Database, 2005." Rome: Joyce, L.A., J.R. Mills, L.S. Heath, A.D. McGuire, R.W. Food and Agricultural Organization of the United Haynes, and R.A. Birdsey. 1995. "Forest Sector Impacts Nations. from Changes in Forest Productivity Under Climate Change." Journal of Biogeography 22: 703­713. FAO. 2005. Global Forest Resources Assessment 2005. FAO Forestry Paper 147 Rome: Food and Agricultural Kauppi, P.E., J.H. Assubel, J. Fang, A. Mather, R.A. Organization of the United Nations. Sedjo, and P.E. Waggoner. 2006. "Returning Forests 36 A DAPTATION Of fORESTS TO CLIMATE C H A N G E Analyzed with the Forest Identity." in the Proceedings of global climate change. Critical Reviews in Environmental the National Academy of Science, Vol. 103, no.46. Science and Technology 27(Special):S245­S257 CRC pp.17574­17579. Press. Boca Rotan. Kirilenko, A. and R.A.Sedjo. 2007. "Climate Change Neilson, R.P. and D. Marks. 1994. "A Global Impacts on Forestry. " PNAS Vol. 104, no. 50, Perspective of Regional Vegetation and Hydrologic 19697­19702. Sensitivities from Climate Change." Journal of Vegetation Science 5: 715­730. King, G.A. and R.P. Neilson. 1992. "The Transient Response of Vegetation to Climate Change: A Potential Norby, R. J., DeLucia, E. H., Gielen, B., Calfapietra, C., Source of CO2 to the Atmosphere." Water, Air, and Soil Giardina, C. P., King, J. S., Ledford, J.,McCarthy, H. R., Pollution 94: 365­383. Moore, D. J. P., Ceulemans, R., De Angelis, P., Finzi, A. C., Karnosky,D. F., Kubiske, M. E., Lukac, M., Korner, C., R. Asshoff, O. Bignucolo, S. Hottenschwiler, Pregitzer, K. S., Scarascia-Mugnozza, G. E.,Schlesinger, S.G. Keel, S. Pelaez-Riedl, S. Pepin, R.R.T.W. Siegwolf, W. H., & Oren, R. (2005). Forest response to elevated and G. Zotz, G. 2005. Science 309: 1360­1362. CO2 is conserved across a broad range of productivity. Proceedings of the National Academy of Sciences ofthe Kurz, W. A., C. C. Dymond, G. Stinson, G. J. United States of America, 102, 18052­18056. Rampley, E. T. Neilson, A. L. Carroll, T. Ebata, and L. Safranyik. 2008. "Mountain pine beetle and forest Olson, J.S., J.A.Watts, and L.J. Allison. 1983. "Carbon carbon feedback to climate change." Nature 452: in Live Vegetation of Major World Ecosystems." Report 987­990. ORNL-5862. Oak Ridge, TN: Oak Ridge National Laboratory. Larson, E.D., S. Consonni, R.E. Katofsky, K. Iisa, and J. Frederick, "An Assessment of Gasification-Based Perez-Garcia, J., L.A. Joyce, A.D. McGuire, and C.S. Biorefining at Kraft Pulp and Paper Mills in the United Binkley. 1997. "Economic Impact of Climatic Change States, Part A: Background and Assumptions," TAPPI on the Global Forest Sector." In R.A. Sedjo, R.N. Journal, 7(11): 8­14, November 2008. Sampson, and J.Wisniewski, eds. Economics of Carbon Sequestration in Forestry. Boca Raton, FLA: Lewis Malcolm, J.R., A. Markham, R.P. Neilson, and M. Garaci. Publishers. 2002. Estimated migration rates under scenarios of global climate change. Journal of Biogeography 29: 835­849. Perez-Garcia, J., L.A. Joyce, A.D. McGuire, and Xiangming Xiao. 2002. "Impacts of Climatic Change on McCarl, B.A., D.M. Adams, R.D. Alig, D. Burton, and the Global Forest Sector." Climatic Change 54: 439­461. C. Chen. 1999. "The Effects of Global Climate Change on the U.S. Forest Sector: Response Functions Derived Roberts, Geoff. 2009. "Current Adaptation Measures from a Dynamic Resource and Market Simulator." and Policies," In Adaptation of Forest and People to Climate Research 16: 255­259. Climate Change--A Global Assessment Report. Seppala, R., A. Buck, and P. Katila, eds. 2009. Adaptation of Forest Melillo, J. M., A.D. McGuire, D.W. Kicklighter, B and People to Climate Change--A Global Assessment Moore, III, C.J. Vorosmarty, and A.L. Schloss. 1993. Report. IUFRO World Series Vol. 22. Tydall, UK. "Global Climate Change and Terrestrial Net Primary Production." Nature 363: 234­240. Rosenberg, N., et al. 1991. "Report III - Forest Resources." In Processes for Identifying Regional Influences Moulton, R.J., and J.F. Kelly. 1997. "The Physical Risks of and Responses to Increasing Atmospheric CO2 and of Reforestation as a Strategy to Offset Global Climate Climate Change--The MINK Project. Prepared by RFF Change." In Moulton, R.J. and Kelly, J.F. 1997. The for the U.S. Department of Energy, DOE/ physical risks of reforestation as a strategy to offset RL/01830T-H9. Washington, DC: D E V E L O P M E N T A N D C L I M AT E C H A N G E D I S C u S S I O N PA P E R S 37 Rosenberg, N. et al. 1993. "The MINK Methodology: Seppala, R., A. Buck, and P. Katila, eds. 2009. Background and Baseline." Climatic Change 24: 7­22. Adaptation of Forest and People to Climate Change--A Global Assessment Report. IUFRO World Series Vol. 22. Schlesinger, M.E., N. Andronnova, A. Ghanem, S. Tydall, UK. Malyshev, T. Reichler, E. Rosano, W. Wang, and R. Yang. 1997. "Geographic scenarios of green-house gas Shen, Z.Q., R.S.Yin, and J.G. Qi. 2009. "Land cover and anthropogenic-sulfate-aersol induced climate changes in northeast China from the late 1970s to changes." Un-numbered Publication. Urbana, Illinois: 2004." In R. Yin, ed. An Integrative Assessment of China's Climate Research Group, Department of Atmospheric Ecological Restoration Programs. New York: Springer. Sciences, University of Illinois. Shugart, H.H., M. Antonovsky, M. Ya, P.G. Jarvis, and Sedjo, R. 1983. The Comparative Economics of Plantation A.P. Sandford. 1986. "CO2, Climatic Change, and Forestry: A Global Assessment. Washington, DC: Forest Ecosystems." In B. Bolin, B.R. Doos, J. Jager, and Resources for the Future. R.A. Warrick, eds. The Greenhouse Effect, Climatic Change and Ecosystem. Chichester, UK: Wiley. Sedjo, R. 1991. "Economic Aspects of Climate, Forests, and Fire: A North American Perspective." Environment Shugart, H.H., R.A. Sedjo, and B. Sohngen. 2003. International 17 (2/3). "Forest and Climate Change: Potential Impacts on the global U.S. forest industry." Paper prepared for the Pew Sedjo, R.A. 1993. "The Carbon Cycle and Global Center on Climate Change. Arlington, VA Forest Ecosystems." In J. Wisniewski and R. N. Sampson, eds. Terrestrial Biospheric Carbon Fluxes: Smith, T.M., and H.H. Shugart. 1993. "The Transient Quantification of Sinks and Sources of CO2. Boston, MA: Response of Terrestrial Carbon Storage to a Perturbed Kluwer Academic Publishers. Climate." Nature 361: 523­526. Sedjo, R.A. 1999. "Potential for Carbon Forest Sohngen, B. 2008. "Climate Change, Agriculture, Plantation in Marginal Timber Forest: The Case of Forests, and Biofuels." Paper presented to Intern Patagonia, Argentina." RFF Discussion paper 99­27. Agricultural Trade Research Consortium, December Washington, DC: RFF. 2008, Scottsdale, AZ. Sedjo, R. A. 2003. "Tropical Deforestation and Sohngen B., and R. Mendelsohn. 1998. "Valuing the Climate." In T. Potter and B. Colman, eds. Handbook of Market Impact of Large Scale Ecological Change in a Weather, Climate and Water. Hoboken, NJ: Market: The Effect of Climate Change on U.S. Willey-Interscience. Timber." American Economic Review 88: 686­710. Sedjo, R. A. 2004. "Kenya Forestry: Economic and Sohngen, B., and R. Mendelsohn. 2001. "Timber: Financial Viability." Report to the World Bank and ecological-economic analysis." In R. Mendelsohn, ed. PROFOR. Washington, D.C. Global warming and the American economy; a regional assessment of climate change impacts. Northhampton , Sedjo, R.A., and K.S. Lyon. 1990. The Long Term MA: Edward Elgar. Adequacy of the World Timber Supply. Washington, DC: Resources for the Future. Sohngen, B., and R. Mendelsohn. 2003. "An Optimal Control Model for Forest Sequestration." American Sedjo, R.A., and B. Sohngen. 2009. "The Implication of Journal of Agricultural Economics 85 (2): 448­457. increased Use of Wood for Biofuel Production." RFF Issues Brief #09-04. Accessible at: http://www.rff.org/ Sohngen, B., and R. Mendelsohn. 2007. "A Sensitivity rff/documents/RFF-IB-09-04.pdf. Analysis of Carbon Sequestration." In M. Schlesinger, H.S. Kheshgi, J. Smith, F.C. de la Chesnaye, J.M. Reilly, 38 A DAPTATION Of fORESTS TO CLIMATE C H A N G E T. Wilson, and C. Kolstad, eds. Human Induced Climate migrate? Global Ecology and Biogeography Letters 6: Change: An Interdisciplinary Assessment. Cambridge, UK: 139­148. Cambridge University Press. Solomon, A., N.H. Ravindranath, R.B. Steward, M. Sohngen, B., R. Mendelsohn, and R. Sedjo. 1999. Weber, and S. Nilsson. 1996. "Wood Production Under "Forest Conservation, Management, and Global Timber Changing Climate and Land Use." In R.T. Wilson, ed. Markets." American Journal of Agricultural Economics Climate Change 1995: Impacts, Adaptation & 81(1): 1­13. Vulnerability. IPCC Special Reports on Climate Change Sohngen, B., R. Mendelsohn, and R.A. Sedjo. 2001."A Southern Hemisphere Forest Industry Journal H. (SH). Global Model of Climate Change Impacts on Timber 1999. "Planting increasing after a mixed year." The Markets." Journal of Agriculture and Resources Economics Southern Hemisphere Forest Industry Journal, 5 (2): 26(2): 326­343. Accessible at: http://www.fao.org/forestry/18316/en/ zaf/. Sohngen, B., and R.A. Sedjo. 2000. "Potential Carbon Flux from Timber Harvests and Management in the VEMAP Members. 1995. "Vegetation/Ecological Context of a Global Timber Market." Climatic Change modeling and analysis project (VEMAP): comparing 44: 151­172. biogeography and biogeochemistry models in continen- tal scale study of terrestrial ecosystems responses to Sohngen, B., and R. Sedjo. 2000. "Potential Carbon climate change and CO2 doubling." Global Flux From Timber Harvests and Management on the Biogeochemical Cycles 9: 407­437. Context of a Global Carbon Flux." Climatic Change 44: 151­172. van Mantgem, P. J., N. L. Stephenson, J. C. Byrne, L. D. Daniels, J. F. Franklin, P. Z. Fulé, M. E. Harmon, A. J. Sohngen, B., and R. Sedjo. 2006. "Carbon Sequestration Larson, J. M. Smith, A. H. Taylor, and T. T. Veblen. in Global Forests Under Different Carbon Price 2009. "Widespread Increase of Tree Mortality Rates in Regimes." The Energy Journal (Special issue, Multi- the Western United States." Science 323: 521­524. Greenhouse Gas Mitigation and Climate Policy): 109­162. Vogel, C. 2003. "Drought in South Africa." In T. Potter and B. Colman, eds. Handbook of Weather, Climate and Sohngen, B., and R. Sedjo. 2006. "Impacts of climate Water. Hoboken, NJ: Willey-Interscience. change on forest products markets: Implications for North American producers." The Forestry Chronicle Watson, R.T., M.C. Zinyowera, and R.H. Moss. 1998. 81(5): 669­675. The Regional Impacts of Climate Change: An Assessment of Vulnerability. Cambridge, UK: Cambridge University Sohngen, B. 2007. "Adapting Forest and Ecosystems to Press. Climate Change." Accessible at: http://www.iccgov.org/ files/ADAPTATION/Sohngen_ccforest_4_2009.pdf. Westerling, A. L., H. G. Hidalgo, D. R. Cayan, and T. W. Swetnam. 2006. "Warming and Earlier Spring Solomon, A.M. 1986. "Transient Response of Forest to Increases Western U.S. Forest Wildfire Activity." Science. CO2-induced Climate Change: Simulation Modeling 313: 940­943. Experiments in Eastern North America." Oecologia 68: 567­579. Yin, Runsheng, ed. 2009. An Integrative Assessment of China's Ecological Restoration Programs. New York: Solomon, A.M., and A.P. Kirilenko. 1997. Climate Springer. change and terrestrial biomass: What if trees do not The World Bank Group 1818 H Street, NW Washington, D.C. 20433 USA Tel: 202-473-1000 Fax: 202-477-6391 Internet: www.worldbank.org/climatechange