A Synthesis of the Science on Forests and Carbon for U.S. Forests

IssuesIssues inin EcologyEcology

Published by the Ecological Society of America

Michael G. Ryan, Mark E. Harmon, Richard A. Birdsey, Christian P. Giardina, Linda S. Heath, Richard A. Houghton, Robert B. Jackson, Duncan C. McKinley, James F. Morrison, Brian C. Murray, Diane E. Pataki, and Kenneth E. Skog

Spring 2010 Report Number 13 eessaa ISSUES IN ECOLOGY NUMBER THIRTEEN SPRING 2010

A Synthesis of the Science on Forests and Carbon for U.S. Forests SUMMARY

orests play an important role in the U.S. and global carbon cycle, and carbon sequestered by U.S. forest growth and Fharvested wood products currently offsets 12-19% of U.S. fossil fuel emissions. The cycle of forest growth, death, and regeneration and the use of wood removed from the forest complicate efforts to understand and measure forest carbon pools and flows. Our report explains these processes and examines the science behind mechanisms proposed for increasing the amount of carbon stored in forests and using wood to offset fossil fuel use. We also examine the tradeoffs, costs, and benefits associated with each mechanism and explain how forest carbon is measured.

Current forests are recovering from past land use as agriculture, pasture, or harvest, and because this period of recovery will eventually end, the resulting forest carbon sink will not continue indefinitely. Increased fertilization from atmospheric nitrogen deposition and increased atmospheric carbon dioxide may also be contributing to forest growth. Both the magnitude of this growth and the future of the carbon sink over the next hundred years are uncertain. Several strategies can increase forest carbon storage, prevent its loss, and reduce fossil fuel consumption (listed in order of increasing uncertainty or risk):

Avoiding deforestation retains forest carbon and has many co-benefits and few risks. Afforestation increases forest carbon and has many co-benefits. Afforesting ecosystems that do not naturally support forests can decrease streamflow and biodiversity. Decreasing harvests can increase species and structural diversity, with the risk of products being harvested elsewhere and carbon loss in disturbance. Increasing the growth rate of existing forests through intensive silviculture can increase both forest carbon storage and wood production, but may reduce stream flow and biodiversity. Use of biomass energy from forests can reduce carbon emissions but will require expansion of forest management and will likely reduce carbon stored in forests. Using wood products for construction in place of concrete or steel releases less fossil fuel in manufacturing. Expansion of this use mostly lies in the non-residential building sector and expansion may reduce forest carbon stores. Urban forestry has a small role in sequestering carbon but may improve energy efficiency of structures. Fuel treatments trade current carbon storage for the potential of avoiding larger carbon losses in wildfire. The carbon savings are highly uncertain.

Each strategy has risks, uncertainties, and, importantly, tradeoffs. For example, avoiding deforestation or decreasing harvests in the U.S. may increase wood imports and lower forest carbon elsewhere. Increasing the use of wood or forest biomass energy will likely reduce carbon stores in the forest and require expansion of the area of active forest management. Recognizing these tradeoffs will be vital to any effort to promote forest carbon storage. Climate change may increase disturbance and forest carbon loss, potentially reducing the effectiveness of management intended to increase forest carbon stocks. Finally, most of these strategies currently do not pay enough to make them viable. Forests offer many benefits besides carbon, and these benefits should be considered along with carbon storage potential.

Cover photo credit: Old-growth forest in the Valley of the Giants in Oregon. Photo by Mark E. Harmon, Oregon State University. Inset: Logs harvested at Manitou Experimental Forest in Colorado. Photo by Richard Oakes, USDA Forest Service. © The Ecological Society of America • [email protected] esa 1 ISSUES IN ECOLOGY NUMBER THIRTEEN SPRING 2010 A Synthesis of the Science on Forests and Carbon for U.S. Forests

Introduction increase and the warmer earth will have an impact on the earth’s climate, climate variabil- The movement of carbon between the earth ity, and ecosystems. Rain and snowfall patterns and its atmosphere controls the concentration will shift, and extreme weather events may

of carbon dioxide (CO2) in the air. CO2 is become more common. Some regions that cur- important because it is a greenhouse gas and rently support forests will no longer do so, and traps heat radiation given off when the sun other regions that currently do not support warms the earth. Higher concentrations of forests may become suitable for forest growth. greenhouse gases in the atmosphere cause the Forests store large amounts of carbon in earth to warm. Before the Industrial their live and dead wood and soil and play an

Revolution, the concentration of CO2 in the active role in controlling the concentration of atmosphere was less than 280 parts per million. CO2 in the atmosphere (Figure 1). In the U.S. The burning of fossil fuel for energy and the in 2003, carbon removed from the atmosphere clearing of forests for agriculture, building by forest growth or stored in harvested wood material, and fuel has led to an increase in the products offset 12-19% of U.S. fossil fuel emis-

concentration of atmospheric CO2 to its cur- sions (the 19% includes a very uncertain esti- rent (2010) level of 388 parts per million. This mate of carbon storage rate in forest soil). U.S. current level far exceeds the 180-300 parts per forest growth rates are thought to be higher million found over the last 650,000 years. than those before European settlement Figure 1. Plants and soil play a As a result of rising CO and other green- because of recovery from past land use and dis- large role in the global carbon 2 cycle as shown by global stocks house gases in the atmosphere, global surface turbance, but the current growth rate will not (boxes) and flows (arrows) of temperatures have increased by 0.74˚C (1.3˚F) continue indefinitely. carbon in petagrams (1000 since the late 1800s, with the rate of warming Given the role that U.S. forests play in offset- teragrams). Numbers in light increasing substantially. As more CO2 is added ting CO2 emissions, our report asks: 1) Which blue and green are the historical to the air, temperatures will continue to human actions influence forest carbon sinks fluxes between the oceans and the atmosphere and plants and (storage rates) and can these soil and the atmosphere that Global Stocks and Flows of Carbon sinks be enhanced for a would have occurred without meaningful period of time human influence. The number in ATMOSPHERE 8.7 through management and dark blue is the additional ocean 100 816 (+4.1/year) use of forest products? and 2) absorption of CO2, resulting from 1.4 increased CO in the atmosphere 100 What are some of the major 2 3.0 100 2.3 100 since the Industrial Revolution. risks, uncertainties, tradeoffs, The numbers in black are the PLANTS & SOIL and co-benefits of using fluxes to the atmosphere from 2,000 forests and forest products in OCEANS fossil fuel combustion or proposed carbon emission deforestation. The number in 37,000 brown is the flux from the mitigation strategies? atmosphere to the land, mostly The purpose of our report from forest regrowth. The COAL, OIL & is to answer these ques- measured atmospheric increase NATURAL GAS tions, or, if answers are not of 4.1 petagrams per year is not SEDIMENTS AND SEDIMENTARY ROCKS 10,000 yet available, to present the equal to the sum of the additions 66,000,000 – 100,000,000 and withdrawals because they best current information. are estimated separately and We present the state of

with associated uncertainties. Institute, 2009. Hole Research A. Houghton, Woods Courtesy of Richard knowledge on the role of 2 esa © The Ecological Society of America • [email protected] ISSUES IN ECOLOGY NUMBER THIRTEEN SPRING 2010 forests in the carbon cycle in a straightforward Recent CO2 manner so that it can be understood by forest Photosynthesis CO managers, policymakers, educators, and the 2 Recent and interested public. We begin with a description Older CO2 of the forest carbon cycle and biophysical effects. We then present details on the strate- Dead Wood gies that have been proposed for using forests Plant Carbon Microbe to slow the amount of CO2 entering the air. Respiration in Respiration leaves, Litter These strategies include: wood, roots Microbes • Avoiding deforestation – Keeping forests intact. • Afforestation – The restoration of forest on land that has been without forest cover for Old, Stable New, Labile some time, and the establishment of forest on Dead Roots Soil Carbon Soil Carbon land that has not previously been forested. • Forest management: decreasing carbon loss – Increasing the harvest interval and/or Figure 2. Flows of carbon from decreasing harvest intensity. synthesis, where leaves capture the energy in the atmosphere to the forest • Forest management: increasing forest growth – sunlight and convert CO2 from the atmos- and back. Carbon is stored Use of improved silvicultural practices, phere and water into sugars that are used to mostly in live and dead wood as genetic improvement, and rapid regeneration. build new leaves, wood, and roots as trees forests grow.

• Forest management: thinning to reduce fire threat. grow (Figure 2). About half of the CO2 that is • Urban forestry – Planting trees in urban converted to sugars is respired by living trees areas for carbon storage and shading for to maintain their metabolism, and the other energy savings. half produces new leaves, wood, and roots. As • Biomass energy – Using fuel from wood and they grow, trees shed dead branches, leaves, biomass in place of fossil fuel. and roots and some of the trees die. • Carbon storage in forest products and substitu- Microorganisms decompose this dead material,

tion – Storing carbon in long-lived forest releasing CO2 back to the atmosphere, but products (such as lumber) and substituting some of the carbon remains in the soil. Live forest products for products (such as steel and and dead trees contain about 60% of the car- concrete) whose manufacture releases much bon in a mature forest, and soil and forest lit-

more CO2 than does the processing of wood. ter contain about 40%. The carbon in live and dead trees (50% of their biomass) varies the We then discuss carbon offsets and credits, most with forest age. how forest carbon could be monitored to deter- Carbon can leave the forest in several ways mine whether changes result in the desired besides tree and microorganism respiration. outcomes, and what the costs would need to be Forest fires release stored carbon into the for carbon to encourage changes. We also dis- atmosphere from the combustion of leaves and cuss some of the uncertainties inherent in the small twigs, the litter layer, and some dead use of forests for carbon storage, because trees and logs, leaving behind a great deal of changes in climate, population, and land use stored carbon in dead trees and soil. Storms may lower projected carbon storage. We espe- and insect outbreaks also kill trees and increase cially note the potential loss of carbon that the amount of material available for decompo- might occur with increased disturbance in a sition. Harvesting removes carbon from the warmer climate. Finally, we provide conclu- forest, although some of it is stored in wood sions and recommendations. products (preventing its immediate release to the atmosphere) and some is available for use Forests and carbon as biomass energy (displacing fossil fuel use). Carbon in the forest In addition, water can remove carbon from a forest either by transporting soil and litter Forest carbon storage differs from many other away in streams (especially from erosion after mechanisms that control atmospheric CO2 fire) or by transporting soluble carbon mole- because forests have a life cycle during which cules created during decomposition. After fire, carbon stocks, gains, and losses vary with for- other disturbance, or harvest, regenerated est age. Carbon enters a forest through photo- forests will eventually recover all of the car- © The Ecological Society of America • [email protected] esa 3 ISSUES IN ECOLOGY NUMBER THIRTEEN SPRING 2010

“boom and bust” cycles may not be appar- Fire Total Carbon ent because the landscape is composed of 150 forest stands that are in different stages of Figure 3. If a forest regenerates recovery from disturbance or harvesting after a fire, and the recovery is (Figure 4). Dead Trees long enough, the forest will 100 To determine how quickly carbon recover the carbon lost in the Wood fire and in the decomposition of increases in a forest system, it is impor- trees killed by the fire. This 50 tant to know the starting point or “base-

figure illustrates this concept by Carbon (Mg C/ha) line.” A forest that already stores a sub- Soil showing carbon stored in 0 stantial amount of carbon is likely to lose forests as live trees, dead wood, –20 0 20 40 60 80 100 120 carbon when converted to something and soil and how these pools else, and a system with the potential to change after fire. (Adapted from Year since fire Kashian and others 2006. store carbon but that does not currently BioScience 56(7):598-606.) store much is easier to convert to one that stores more carbon (Figure 5). A for- bon lost so that a complete cycle is carbon est’s timeline for increasing carbon storage is neutral regarding storage if the recovery is important because carbon must be removed

long enough (Figure 3). But if disturbances quickly to reduce CO2 in the atmosphere and increase, as is projected with climate change, thereby slow global warming. a fire, storm, or insect outbreak may occur While the biological processes of photosyn- before the ecosystem recovers the carbon it thesis, respiration, and decomposition are had prior to the disturbance. In that case, the similar for all forests, their relative impor- amount of carbon stored on the landscape will tance differs by forest type and location. Some decrease. forests grow more rapidly, but dead trees in Forests are biological systems that continu- fast-growing forests also decompose more ally gain and lose carbon via processes such as rapidly. In addition, disturbances vary region- photosynthesis, respiration, and combustion; ally: for example, fire disturbance is more whether forests show a net gain or loss of car- common in the western U.S. and hurricanes bon depends on the balance of these processes. more common in the East. Forests are man- The observation that carbon is lost from forests aged in different ways with varying harvest has led to the notion that carbon cannot be intervals and regeneration practices that will permanently stored in forests. However, this influence the optimum strategy for storing view ignores the inevitable increase and even- more carbon. Each forest has a different tual recovery of carbon that follows most dis- potential to store carbon. For example, this turbances. Thus over time, a single forest will potential is particularly high in the Pacific vary dramatically in its ability to store carbon; Northwest where forests are relatively produc- however, when considering many different tive, trees live a long time, decomposition is forests over a large area or landscape, such relatively slow, and fires are infrequent. The differences between forests must Figure 4. Management actions should be examined for large therefore be taken into consider- areas and over long time ation when determining how 1500 they should be managed to store periods. This figure illustrates 1 stand 10 stands how the behavior of carbon carbon. stores changes as the area 100 stands becomes larger and more stands are included in the analysis. As 1000 Carbon from the forest the number of stands increases, the gains in one stand tend to be All forest products eventually offset by losses in another and decompose, but before they do, hence the flatter the carbon 500 they store carbon. Some prod- stores curve becomes. The average carbon store of a large ucts have a short lifespan (such number of stands is controlled as fence posts) and some a longer

by the interval and severity of Carbon Storage (Mg/ha) lifespan (for example, houses) – disturbances, as shown in Figure 0 the longer the lifespan, the more 7. That is, the more frequent and 0 100 200 300 400 carbon is stored. Disposed forest severe the disturbances, the lower the average becomes. Years products in landfills can have a (Courtesy of Mark E. Harmon, very long lifespan; however, the Oregon State University, 2009.) decomposition in landfills

4 esa © The Ecological Society of America • [email protected] ISSUES IN ECOLOGY NUMBER THIRTEEN SPRING 2010

Figure 5. Projections of carbon generates methane, which is a much more (a) 250 Soil Landfill storage and fossil fuel potent greenhouse gas than CO2, reducing displacement if all biomass is Litter Substitution the carbon storage benefit. In addition, Biomass used shows considerable wood and bark that are burned to run a mill 200 Trees energy storage and offsets for (A) a or heat houses, or made into liquid biofuel, Products–L project that reestablishes forests 150 Products–S with periodic harvests. lower emissions from fossil fuel use. Once Harvesting a high-biomass old the carbon leaves the forest, it becomes growth forest (B) shows carbon more difficult to track and measure than 100 losses, even under the best carbon in the forest, particularly because possible scenario, for several imports and exports must then be tracked. 50 harvests. At each harvest, forest biomass (and thus carbon stock) is removed for use in long- and Biophysical effects may cause 0 short-lived wood products (‘Products-L’ and ‘Products-S’, warming or cooling 250 (b) respectively) substituted for more carbon-intensive products, Forests have other influences on climate and for biomass energy to besides that of carbon; these are known as bio- 200 displace emissions from fossil

physical effects (Figure 6) and include the Cumulative carbon (Mg/ha) fuel use. Because substitution reflection of solar radiation and transpiration 150 generates more fossil fuel of water vapor. Trees are dark and absorb more savings than the carbon it 100 contains, substitution would radiation than other types of land cover, such yield a greater carbon benefit as crops or snow-covered tundra. Therefore, after harvest than that which is converting non-forested land to forest can 50 stored in the biomass. The warm the land and air. Evergreen trees absorb biomass energy and substitution much more energy than deciduous trees in the fossil fuel savings accumulate 0 but represent only hypothetical winter and burned forests absorb more than 20 40 60 80 100 carbon benefits, as currently unburned forests, so species and disturbance Years little biomass energy use and can also alter the energy absorbed by forests. substitution occurs in the U.S. In addition, transpiration from forests may (Adapted from IPCC 2007.) have a cooling effect by contributing to the bon (1012 grams; see Box 1 for units) to the formation of clouds that reflect sunlight. atmosphere, and two-thirds of this release Biophysical effects sometimes act in a direc- occurs in tropical forests. The amount of car- tion opposite to that of the effects of storing or bon released by deforestation equals 17-25% releasing CO2. For instance, whereas convert- of global fossil fuel emissions every year and is ing cropland to forest will sequester more CO2, roughly the amount of U.S. annual fossil fuel which reduces global warming, it will also emissions. If current deforestation rates con- increase solar absorption, which increases tinue, more than 30,000 teragrams of carbon warming. Generally, biophysical effects on cli- could be released to the atmosphere from mate are not as strong as the effects of green- deforestation in the Amazon alone by the house gases. Biophysical effects will be most year 2050. important in evaluating the benefits of In the U.S., forested area increased 0.1% per afforestation because the land use change will year from 2000-2005, and this gain in forested cause large differences. Unfortunately, current area is partially responsible for the current for- estimates of biophysical effects are uncertain est sink of 162 teragrams of carbon per year. because few studies have been done. The net growth in forested area results from both deforestation and afforestation: About Strategies for increasing carbon 6,000 km2 are deforested annually, but more 2 stores in forests than 10,000 km of non-forest are afforested. The net increase in forestlands results from 1. Avoiding deforestation changes in land use and possibly from reduced demand for U.S. timber. Deforestation, or the conversion of forest land Although the U.S. forest carbon sink bene- to other uses, has a significant impact on fits from increased forest area, these carbon global CO2 emissions. Globally, deforestation benefits need to be weighed against the global converts approximately 90,000 km2 (about the consequences of land use change within the size of Indiana) of forests per year (0.2% of all U.S. If afforestation or avoided deforestation forests) to other land uses. Deforestation in the U.S. pushes crop and cattle production annually releases 1,400-2,000 teragrams of car- to other countries, it can lead to deforestation © The Ecological Society of America • [email protected] esa 5 ISSUES IN ECOLOGY NUMBER THIRTEEN SPRING 2010

occurs, substantial carbon is lost to the atmos- Reflected sunlight phere and not recovered by the ecosystem. Evaporation Tree planting would help recover forest carbon Transmitted heat where natural regeneration fails. There are not many risks associated with avoidance of deforestation. Three to note, however, would be risks related to highly fire- prone ecosystems near human settlement, economic consequences for not developing agricultural or pasture land, and an increase in forest products harvested elsewhere. On the other hand, avoiding deforestation has many of the co-benefits identified in Box 2.

2. Afforestation

We define afforestation as both reestablishing forests on land that has been without forest cover for some time and the establishment of forest on land that has not previously been forested (note that some entities involved in Figure 6. Biophysical effects of and loss of forest carbon elsewhere to create carbon markets and reporting use different different land use can have pasture and cropland. Carbon loss associated definitions for this term). Afforestation can important impacts on climate. with such deforestation – especially in the remove substantial CO from the atmosphere. Cropland reflects more sunlight 2 than forest, produces less water tropics – is greater than carbon gain associated Between 1850 and 2000, global land-use vapor, and transmits less heat. with tree growth from afforestation in the change resulted in the release of 156,000 tera- (From Jackson et al. 2008. U.S. grams of carbon to the atmosphere, mostly Environmental Research Letters Forest retention in the western U.S. may be from deforestation. This amount is equivalent 3:article 044006.) even more important in the future as climate to 21.9 years of global fossil fuel CO2 emis- changes. Our warming climate is very likely sions at the 2003 level. causing, at least in part, the current increase in The rate of carbon storage in tree growth forest fire size and intensity, insect outbreaks, varies with species, climate, and management, and storm intensity. If forest regeneration fails ranging widely from about 3-20 megagrams because the disturbances or regeneration con- (Mg, 106 grams) per hectare per year. In the ditions are outside of the ecological norms, dis- continental U.S., the highest potential growth turbances can convert forests to meadows or rates are found in the Pacific Northwest, the shrublands. When this type of deforestation Southeast, and the South Central U.S. Much land currently in pasture and agricultural use Box 1. UNITS FOR CARBON in the eastern U.S. and in the Lake States will naturally revert to forests if left fallow, while When discussing regional, national, or global carbon stores and fluxes, the num- reestablishing forests in many western forests bers get large quickly. We report carbon in teragrams (1012 grams). Other requires tree planting. reports may use other units, so we provide a conversion table below. For stand- The benefits of afforestation (outlined in 6 or forest-level stores and fluxes, we use megagrams (Mg) per hectare (10 Box 2) are enhanced where forests include a grams). Our report uses carbon mass, not CO mass, because carbon is a stan- 2 substantial proportion of native species. dard “currency” and can easily be converted to any other unit. Many reports Planting native species or allowing natural suc- give stocks and fluxes of the mass of CO2, not carbon. To convert carbon mass to CO mass, multiply by 3.67 to account for the mass of the O . cession to recreate the forest that historically 2 2 occupied the site will yield the greatest benefits 1000 teragrams (Tg) 1 petagram (Pg) for species diversity and wildlife habitat and 1000 teragrams 1 billion metric tonnes the lowest risk for unintended consequences. 1000 teragrams 1 gigatonne Because native species often grow more slowly 1 teragram 1 million metric tonnes than exotics or trees selected for improved 1 teragram 1 megatonne growth, restoration of the historical ecosystem 1 megagram (Mg) 1 metric tonne may yield lower carbon accumulation rates 1 metric tonne 0.98 U.S. long ton than other forest reestablishment practices. 1 metric tonne per hectare 0.4 U.S. long tons per acre carbon (C) mass * 3.67 carbon dioxide (CO ) mass Planting monocultures of non-native or native 2 improved-growth species on historical forest 6 esa © The Ecological Society of America • [email protected] ISSUES IN ECOLOGY NUMBER THIRTEEN SPRING 2010 land will likely yield greater carbon accumulation Box 2. CO-BENEFITS OF FORESTS rates but fewer benefits in terms of biodiversity. Afforestation can have negative conse- Our report focuses on forests seen through the lens of carbon, and only carbon. quences, too. Planting forests where they were However, forests are managed for many purposes, and carbon storage and the not present historically can have drawbacks growth of wood for products and fuel to offset fossil fuel use are far from the only such as lower species diversity (if trees are reasons forests are valuable. Forests also provide many other ecosystem ser- planted in native grassland), changes in water vices that are important to the well-being of the U.S. and its inhabitants: protec- table, and a higher energy absorption com- tion of watersheds from erosion, nutrient retention, good water quality, reduction pared to the native ecosystem. In addition, of peak streamflow and an increase in base streamflow, wildlife habitat and afforestation generally reduces streamflow diversity, recreational opportunities and aesthetic and spiritual fulfillment, and regardless of the ecosystem type because trees biodiversity conservation. Americans are strongly attached to their forests. In use more water than grass or crops. some cases, managing strictly for carbon would conflict with other co-benefits Conversion of agricultural or grazing lands to of forests. The option of avoided deforestation retains the co-benefits of forests and the carbon in forest ecosystems, while afforestation adds these co-benefits forest reduces revenue from agricultural prod- in addition to increasing carbon storage. Even simple forests, such as planta- ucts. If afforestation efforts include the additions, generally reduce erosion, regulate streamflow, and increase wildlife habitat tion of nitrogen fertilizer, emissions of nitrous and biodiversity compared to crops or livestock pasture because the frequency oxide (a greenhouse gas roughly 300 times as of harvest or stand–replacing disturbance is much less for forests. powerful as CO2) will increase.

3. Forest management: decreasing ing dead plant matter, which are common in carbon loss the Pacific Northwest. The carbon benefit of either of these practices will depend on the Lengthening the harvest interval or reducing temporal and spatial scales at which they are the amount removed in a harvest will store administered – applying these practices over more carbon in the forest. The greater the longer timeframes and larger landscapes leads increase in harvest interval over the current to greater carbon benefits. level, the higher the increase in carbon stor- In addition to an increase in carbon storage, age. For example, a five-year increase in the benefits of decreased harvesting also include an harvest interval would lead to a 15% increase increase in structural and species diversity. On in carbon storage if the harvest interval was the other hand, the costs are an increased risk Figure 7. Average carbon changed from 25 to 30 years, but only a 4% of carbon loss due to disturbance and the stored on a landscape will vary increase if the interval was changed from 55 to potential for increased harvesting elsewhere to with the time between harvests (harvest interval) and how much 60 years (Figure 7). A 50-year increase from compensate for the reduction in forest products biomass is removed each 25 to 75 years would increase carbon storage generated. harvest. Lengthening the 92% (Figure 7). harvest interval will have a The carbon impact of reducing the amount 4. Forest management: increasing greater effect for harvests where removals are high (blue arrows of trees removed in a harvest also varies with forest growth show an increase in harvest the harvest interval. For example, reducing the interval from 25 to 75 years). harvest from 100% to 20% of the live trees In addition to afforestation, another strategy Decreasing harvest intensity would increase the average forest carbon stock for increasing carbon storage is to increase the from 100% of trees to 20% of by 97% for a 25-year harvest interval, but only growth rate of existing or new forests. trees (black arrows) will have a greater effect for shorter harvest by 30% for a 100-year harvest interval (Figure Management practices that can increase forest intervals. (Courtesy of Mark E. 7). Some natural forests are dominated by growth include: regenerating harvested or Harmon, Oregon State small disturbances that kill a few trees at a damaged forests, controlling competing vege- University, 2009.) time. Reducing harvest amounts in these sys- tation, fertilizing, planting tems from complete removal of trees to simply genetically improved trees, 200 a percentage, for example, could mimic the and selecting species for natural disturbance regime common to the superior productivity. Yield 150 northeastern and midwestern United States. In gains from these practices addition, reducing harvests could be desirable can be impressive. In pine 100 in public forests that are managed for multiple forests in the southern (Mg/ha) purposes, such as recreation, biodiversity, and U.S., tree breeding has 50 20% Removal water. improved wood growth Carbon Storage 100% Removal These strategies would be most suitable in (and carbon storage rate) 0 forest regions with active management and a by 10-30%, and fertilization 0 50 100 150 200 high potential to store carbon, such as those can show 100% gains for Harvest Interval (Years) with long-lived species and slowly decompos- wood growth. For southern © The Ecological Society of America • [email protected] esa 7 ISSUES IN ECOLOGY NUMBER THIRTEEN SPRING 2010

more than 500 million acres and consists of almost all U.S. public and private forestland, excluding remote and reserved areas such as national parks. However, even reserved areas could potentially be managed to restore damaged ecosystems, which could also lead to increased forest growth. Increasing forest growth through management has benefits and costs. The benefits Figure 8. A hydro-axe is used to include increased wood production and the grind up trees to reduce canopy fuel loads and lower the risk of potential for planting species and genotypes crown fire. Photo by Dan Binkley, adapted to future climates. The costs include Colorado State University. reducing the carbon benefit by emissions of nitrous oxide from forest fertilization, reduced U.S. pines, operational plantations using water yield (faster growth uses more water), improved seedlings, control of competing veg- and a loss of biodiversity if faster growth is etation, and fertilization grow wood four times accomplished by replacing multi-species forests faster than naturally regenerated second- with monocultures. growth pine forests without competition control. The potential to increase forest growth 5. Forest management: fuel manage- varies by climate, soil, tree species, and man- ment to reduce fire threat agement. Increases in carbon stocks will generally be Fuel management uses thinning (Box 3) to proportional to increases in growth rates. That lower foliage biomass to reduce the risk of is, a 10% increase in growth will result in a crown fire because crown fires are difficult, if 10% increase in carbon stocks, assuming that not impossible, to control. Fuel management the harvest interval and amount harvested do occurs in forests with a variety of historical fire not change. As shown in Figure 3, the rate of regimes – from forests where historical forest forest growth will naturally slow down as the density was lower and the natural fires were forest ages. Management decisions for increas- mostly surface fires, to forests with stand- ing carbon stocks should take into account for- replacement fire regimes in which crown fires est growth over time, the amount of timber naturally occurred. Fuel management tem- that would end up in wood products if the for- porarily lowers the carbon stored in forest bio- est were harvested, and how long the harvested mass and dead wood because the thinned trees carbon would remain sequestered in the wood are typically piled and burned or mulched and products. Knowledge of these variables will then decompose. help determine when or whether to harvest. If a crown fire burns through a forest that The area of forestland in the U.S. that could was thinned to a low density, the fire may be managed to increase forest growth includes change from a crown to a surface fire in which many of the trees can often survive the fire. In Box 3. THINNING AND CARBON contrast, many or all of the trees in an unthinned stand will be killed by a crown fire. Thinning is an effective forest management technique used to produce larger This contrast in survival has led to the notion stems more quickly, reduce fire risk, and increase tree resistance to insects that fuel treatments offer a carbon benefit: and disease. Thinning increases the growth of the remaining individual trees, removing some carbon from the forest may but generally decreases overall forest wood growth until the remaining trees grow enough to re-occupy the site. The carbon stock in a thinned stand is gen- protect the remaining carbon. erally lower than that in an unthinned stand. If the harvested trees are used for There are two views regarding the science biomass energy or long-lived forest products, these carbon benefits may com- on carbon savings through fuel treatments. pensate for the lower biomass and the wood growth of the thinned stand. Some studies have shown that thinned stands Because of lower overall growth of a thinned stand, even 100% use of the har- have much higher tree survival and lower car- vested trees for products or biomass energy may not produce a total carbon bon losses in a crown fire, or have used mod- benefit greater than that of the higher storage and storage rate in an unthinned eling to estimate lower carbon losses from stand. The net carbon consequences of thinning will depend the most on whether the harvested trees are used for long-lasting wood products or bio- thinned stands if they were to burn. However, mass energy, but also on the change in risk of a crown fire relative to the prob- other stand-level studies have not shown a ability of fire occurring, the species, the site, the thinning regime, and the carbon benefit from fuels treatments, and evi- length of the harvest interval. dence from landscape-level modeling suggests that fuel treatments in most forests will 8 esa © The Ecological Society of America • [email protected] ISSUES IN ECOLOGY NUMBER THIRTEEN SPRING 2010 decrease carbon, even if the thinned trees are used for biomass energy. More research is urgently needed to resolve these different conclusions because thinning to reduce fuel is a widespread forest treatment in the U.S. We recommend that such research focus on the landscape scale because carbon loss in thinning needs to be placed in the context of the expected fire frequency and extent, and the potential for regener- Figure 9. Sycamores lining ation after fire. Regardless of the out- Sycamore Street in Los Angeles, California. Photo by Diane E. come of such research, the carbon Pataki, University of California, benefits of fuel treatments can be Irvine. improved by using the harvested trees for wood or biomass energy. cities located in grasslands and deserts, urban forests require large amounts of irrigation 6. Urban forestry water for maintenance. Because of these many tradeoffs, the fol- Urban forestry offers very limited potential to lowing factors must be taken into account to store carbon, but we address urban forests determine the net climate impact of urban here because of the large interest in using trees: 1) the carbon storage rate of the trees, them to offset carbon emissions and because 2) fossil fuel emissions from energy associ- urban trees provide many co-benefits, includ- ated with planting and maintenance, 3) fos- ing aesthetic benefits and environmental sil fuel emissions resulting from the irriga- advantages in addition to carbon sequestration process, 4) nitrous and nitric oxide tion. The potential for carbon offsets of emissions from fertilizer use, and 5) the net greenhouse gas emissions through urban effect of trees on local air temperature and forestry is very limited for two reasons: 1) its impact on building energy use. These fac- urban areas make up only a small fraction of tors are likely to be highly variable by region the U.S. landscape and 2) urban forests are and by species. intensively managed and may require large energy, water, and fertilizer inputs for planting 7. Biomass energy, carbon storage in and maintenance. products, and substitution Urban forests can have important biophysi- Biomass energy cal effects on climate. Trees have a cooling effect on local temperatures due both to shad- The use of forest biomass energy prevents car- ing effects and to evaporative cooling in tran- bon emissions from fossil fuel use. In 2003, spiration. Shading intercepts incoming radia- biomass energy was 28% of the U.S. renew- tion in the daytime, which can reduce both able energy supply and 2% of the total U.S. day and night surface temperatures. When energy use. Biomass energy is used primarily trees are planted very close to buildings, they for electric power in the forest products indus- cool building temperatures and reduce the fos- try and for residential heating. In the future, sil fuel emissions associated with air condi- biomass may become an important feedstock tioning. When urban forests are planted over for liquid biofuels. very large regions, the climate effects are less If cost were not a constraint and the public certain, as trees can have both warming supported this use of forests, U.S. forests could (absorption) and cooling effects. potentially provide energy production offset- The higher the maintenance required for ting 190 teragrams of fossil fuel carbon emis- urban trees, the less likely they will help miti- sions per year, or the equivalent of 12% of gate climate change. In some regions, cities U.S. fossil fuel emissions in 2003 (as discussed are located in what would naturally be further in Environmental costs below). It has forested areas; thus, urban forests serve to been estimated that by 2022, forest biomass restore forests to land that was previously feedstocks could produce 4 billion gallons of deforested. In such regions, trees may have rel- liquid biofuel per year (offsetting 2.6 teragrams atively low maintenance requirements. In of fossil fuel carbon emissions). © The Ecological Society of America • [email protected] esa 9 ISSUES IN ECOLOGY NUMBER THIRTEEN SPRING 2010

review of studies suggests that if wood products containing one unit of carbon were used in buildings as a substitute for steel or concrete, fossil fuel emissions from manufacturing would be reduced by two units or more. Opportunities for increased substitution in the U.S. will mostly need to be found outside of the housing industry because most housing is already built using wood. Figure 10. Logs harvested at Manitou Experimental Forest in Colorado. Photo by Richard Environmental costs of biomass Oakes, USDA Forest Service. energy and forest products use

Carbon storage in wood and paper The carbon benefits of increasing the use of products wood for biomass energy and for product substitution would require more intensive forest man- In the U.S., forest products are stored in two agement over a much broader area than cur- major “pools”: those that are in use, and those rently occurs. For example, to obtain the held in landfills. Current additions of carbon to aforementioned 190 teragrams per year of bio- these pools from trees harvested in the U.S. are mass energy would involve harvesting all of the greater than decomposition losses from these current annual net forest growth in the U.S. To pools, so carbon stored in these pools is increas- do that would require intensive management ing. In 2007, the net increase in carbon stored on much of the U.S. forest estate and would as products in use and in landfills was 30 tera- reduce the carbon stored in the forest. If grams of carbon (offsetting 1.7% of 2003 U.S. branches and foliage were to be removed for fossil fuel emissions), with about two thirds of biomass energy, fertilization would likely be the 30 teragrams being net carbon additions to needed to replace the nutrients removed to landfills. Recently, additions have been declin- maintain productivity. Additionally, dead wood ing due to decreases in U.S. timber harvests. will decrease and soil carbon may decrease Carbon is also accumulating in “products in under harvesting, creating a carbon debt that use”, primarily in buildings. The total carbon will require time to pay off. held in single and multifamily homes in 2001 was about 700 teragrams of carbon. Annual net Links between strategies carbon accumulation in landfills is larger than that for products “in use” because about 80% of Strategies can be combined to increase the car- wood and 40% of paper decays very slowly bon benefit. For example, Figure 5 shows that under the anaerobic conditions in landfills. the maximum potential benefit from a project However, these same anaerobic conditions that that reestablished forest increases if the stand is slow decomposition also produce methane, a periodically harvested and the wood is used for greenhouse gas with greater than 25 times the substitution and the biomass used for fuel.

warming potential of CO2. Because only 50% of Increased wood use for forest products and bio- methane is captured or oxidized before release, mass energy would be compatible with afforesta- methane release reduces the carbon storage ben- tion, increasing forest growth, and fuel manage- efits in landfills. If we were to use the 30 tera- ment to reduce fire threat. However, increased grams per year of forest products currently going wood use may conflict with increasing carbon into landfills as biomass energy, we would offset stores on the landscape from reducing harvests 1.2% of U.S. fossil fuel use, lower emissions of and avoiding deforestation. Increased forest methane, and extend the life of landfills. growth would be compatible with reducing harvests and avoiding deforestation if the increased Substitution growth frees land for these other uses.

Carbon emissions can be offset by substitut- Carbon offsets and credits ing wood products for products such as steel and concrete, which generate more green- A carbon offset is a reduction in greenhouse house gas emissions in their production. A gas emissions (or an increase in carbon seques- 10 esa © The Ecological Society of America • [email protected] ISSUES IN ECOLOGY NUMBER THIRTEEN SPRING 2010 tration) by one entity, which can compensate for – or “offset” – emissions by another entity. The latter can thus continue with business as usual and avoid directly reducing its own emissions. Offsets are typically traded (bought and sold) as “carbon credits.” Typically, offset projects are certified, which instills confidence that the offsets are real and enables the associated carbon credits to be sold or traded to those who voluntar- Figure 11. Tree harvesting at ily wish to reduce their reductions or Manitou Experimental Forest in are regulated to do so. In the U.S., Colorado. Photo by Richard carbon credits are traded as part of a Oakes, USDA Forest Service. voluntary market, and the certifica- tion process varies widely. Europe, which rati- as the climate remains suitable because the fied the Kyoto Protocol, has a regulated car- landscape will maintain a level of carbon bon market. Some of the forest management determined by the disturbance or harvest strategies discussed in this paper could “earn” interval. carbon credits, such as afforestation, decreas- The most serious concern in any effort ing harvest intensity, increasing forest growth, where forest management is changed for car- use of biomass energy, and substitution. bon benefits is leakage – changes outside of Carbon offsets require additionality, meaning the project boundary that reduce or eliminate that the carbon benefits occur directly as a the carbon benefit. For example, afforesting result of an action deliberately taken to agricultural land in the U.S. may increase increase carbon sequestration. Additionality is deforestation elsewhere to meet the demand required because reducing greenhouse gas emis- for food. Or, subsidizing forest carbon in the sions over business as usual is a goal and U.S. could decrease harvests, increase imports because no one wishes to pay for something of wood and wood products, and lead to that would happen anyway. Demonstrating increased forest harvest – and thus reduced additionality for forest activities requires that forest carbon – elsewhere. Leakage occurs, but the activity be compared against a baseline sce- is very difficult to measure because of its global nario without activity. Demonstrating addition- nature and the difficulty of identifying cause ality is relatively straightforward for afforesta- and effect. tion, urban forestry, and biomass energy use Although carbon offsets and credits feature because the “starting point” can be quantified. prominently in comprehensive climate-and- It is much more complex for management that energy legislation and may be critical to a reduces carbon outputs or increases forest society-wide effort to address climate change, growth because larger areas need to be moni- other systems for increasing forest carbon tored for a longer time to validate increased sequestration may be simpler than carbon off- carbon storage. It is also difficult to show addi- sets. For example, direct payments to tionality for the strategy of avoiding deforesta- landowners for a particular land use (as in the tion because carbon storage does not necessar- current Conservation Reserve Program) could ily increase if forests are simply retained. ensure desired management, and could reward Many traders of forest carbon credits are also avoided deforestation. Land-use regulation concerned with permanence, because carbon could also be used to force behavior that credits associated with the offset are sold sequesters carbon (for example, minimum before the management is fully implemented. harvest intervals or requirements to plant Some forest carbon can be temporarily lost in trees on agricultural lands). a disturbance or harvest. It can also be lost with land use changes, some of which can pre- Measuring, monitoring and serve the option of forest reestablishment verifying carbon offsets (such as change to agriculture or pasture) and some of which do not (urban development). As the U.S. does not have a regulated carbon For land maintained as forest, forest carbon market, this discussion of monitoring and storage can be considered permanent as long verifying carbon offsets is based on processes © The Ecological Society of America • [email protected] esa 11 ISSUES IN ECOLOGY NUMBER THIRTEEN SPRING 2010

carbon management. The Forest Inventory and Analysis National Program conducts a national-level strategic forest inventory based on a combination of on-the-ground measurements of all forest carbon pools and remotely- sensed observations. The inventory produces estimates of forest age, cover types, and disturbance and uses modeling for components that Figure 12. Regeneration in are difficult to measure. Yellowstone National Park Carbon stored in wood products is more dif- 19 years after the 1988 fires, ficult to monitor than carbon in the forest. with Dan Kashian (Wayne State University). Carbon in solid wood products in structures Photo by Mike Ryan, could be estimated using current census data USDA Forest Service. with deductions for the fraction of products that are imported. Rates of accumulation for outlined for voluntary markets. Carbon man- all forest products could also be monitored agement begins with a project design that has using data on production rates, recycling rates, been validated by scientific study to increase and discard rates (to landfills). Biomass energy carbon storage rates compared to baseline rates. use could be tracked through surveys of bio- Once additional carbon accumulates, credible mass energy facilities. and accepted measurement and monitoring methods must be used to document carbon How should carbon stores be gains. Next, many offset projects and activities measured? demonstrate that they do not cause leakage, but not all voluntary markets require this important Since carbon-storage projects take place across but difficult step. Finally, an independent verifi- many different scales (stand, landscape, cation confirms that the project was installed regional, and national) and jurisdictions, mul- correctly, is performing as projected, and that tiple methods of measurement are needed. A the carbon reporting is valid. list of approved methods for measuring carbon pools should include the minimum number of Measurement of carbon at various pools to be measured with methods having scales minimal bias (that do not lead to frequent over- or under-counts of carbon) as well as the At the scale of individual forest stands, ade- minimum frequency of measurements. There quate measurements (accurate to about 20%) is an inherent level of uncertainty associated can be made to estimate the carbon stored in with any method for measuring carbon, and trees, plants, dead wood, and in litter on the there is a practical need to decide how to treat forest floor using standard inventory methods. this uncertainty in decision-making. If we use Improvements to these methods would likely high-end estimates for forest carbon storage, involve increased monitoring costs. Stand- we may over-promise what forests can do and level measurements of belowground stocks are obscure the need for mitigation actions in more difficult because of the large cost of sam- other sectors. Given the urgent need to meet pling soil carbon and fewer equations for esti- climate change mitigation objectives and the mating belowground biomass. Soil and below- high risks to society associated with failing to ground carbon monitoring should receive meet them, we recommend discounting car- attention in accounting for forest carbon bon estimates where they are uncertain. As because forest harvest may cause an average sampling frequency and specificity increase, loss of 8% of soil carbon stocks and 30% of the uncertainty should decrease, but costs will also organic layer (forest floor) carbon. rise. Individual groups or entities can decide At the landscape level, projects can be mon- which approved method should be used for itored and verified using remote sensing. each project based on a cost-benefit ratio, Remote-sensing methods enable direct moni- weighing cost against gaining potential carbon toring of forest age, cover types, and distur- benefit. The potential for leakage and bance. Changes in carbon stores can be esti- accounting for and underestimating distur- mated with this information using ecosystem bance losses can be reduced by implementing or accounting models. Monitoring at the a national-level accounting system that vali- regional level assesses the large-scale impact of dates the carbon storage at a national scale. 12 esa © The Ecological Society of America • [email protected] ISSUES IN ECOLOGY NUMBER THIRTEEN SPRING 2010

Economics of forest carbon a potential for greater loss of carbon stores from forest fires, insect outbreaks, hurricanes, Most of the strategies for increasing forest car- windstorms, and ice storms. Climate change bon storage or the use of forest products would threatens to amplify these risks by increasing require carbon to have a substantial value the frequency of these disturbances. If climate through credits for offsets or through some change increases the frequency of disturbance, other mechanism to compensate those that as observational and modeling studies for the have an economic interest for additional costs U.S. suggest, many forests could release signif- or foregone profit. To sequester an additional icant amounts of carbon to the atmosphere 200-330 teragrams of carbon in forests (the over the next 50-100 years – simultaneous equivalent of offsetting 13-21% of 2003 U.S. with efforts to harness CO2 emissions. It is fossil fuel emissions) would require payments important to remember that, at the landscape of between $110-$183 per metric tonne of car- level over the long term, disturbance does not bon, or 23-60 billion dollars per year. The per- cause a net loss of forest carbon…as long as tonne payment requirement reflects the eco- the forest regenerates. But if the frequency nomic value of the current use. For example, and/or severity of disturbance increase sub- for afforestation, landowners would expect stantially, long-term carbon storage at the compensation for both their lost agricultural landscape scale will be reduced because the revenues and for the cost of planting trees. For fraction of the landscape with large, older lengthening the harvest interval, landowners trees (that have high carbon stores) will would need compensation for the reduced decline. Climate change could also increase product flow. soil decomposition, leading to carbon losses Although the total costs of such an under- from a part of the ecosystem that we consider taking are large, the costs of implementing to be relatively stable and that contains about these forest activities to sequester carbon are 40% of the total carbon in U.S. forests. often far less than the cost of reducing the The largest risk to carbon storage from dis- same amount of greenhouse gas emissions by turbance is that the forest may not regenerate other means, such as through the transporta- and instead be replaced by a meadow or shrub- tion or electric power sectors. Therefore, land ecosystem, losing much carbon in the forests can play a key role in reducing the process. As a result of past fire suppression, we overall cost of achieving greenhouse gas emis- see this happening currently in the western sion reduction targets. Economic modeling of U.S. as high-severity fires occur in ecosystems U.S. climate policy proposals consistently that are adapted to low-severity fire regimes. shows that forest carbon sequestration and Although actions are being taken to reduce other “offset” activities can significantly lower the fire risk, the carbon-related effects are cur- the cost of complying with the proposed regu- rently unknown. Climate change may also lations. increase the likelihood that forests will not regenerate sufficiently since highly adapted Climate change and other risks species and genotypes may have a difficult time to forest carbon storage growing under altered climatic conditions. The potential to increase carbon storage in Conclusions and forests needs to be weighed against the pro- Recommendations jected increases in disturbances promoted by a changing climate that will lower carbon stor- U.S. forests and forest products currently offset age. Climate change may also make regenera- 12-19% of U.S. fossil fuel emissions, largely tion after disturbance more difficult or render owing to recovery from past deforestation and the current tree populations genetically extensive harvesting. Increased nitrogen depo- unsuitable. Finally, population increase and sition and atmospheric CO2 compared to his- exurban development will decrease the gen- torical levels may also be contributing to eral amount of forested area. Because distur- increased forest growth, but the science sup- bances are likely to increase in the future, we porting their contribution is uncertain because recommend conservative estimates of poten- of a limited number of experiments and the tial gains from forest carbon management. difficulty in assessing change over the diverse A potential negative effect of forest manage- forests of the U.S. ment strategies to enhance carbon storage is How long will U.S. forests remain a carbon that, as forest carbon storage increases, there is sink? Since 1940, forest regrowth in the U.S. © The Ecological Society of America • [email protected] esa 13 ISSUES IN ECOLOGY NUMBER THIRTEEN SPRING 2010

has recovered about a third of the carbon lost Because forest carbon loss poses a significant to the atmosphere through the deforestation climate risk and because climate change may and harvesting that occurred from 1700-1935 impede regeneration following disturbance, (Figure 13). To recover the remaining two- avoiding forest loss and promoting regenera- thirds of the carbon that was lost would tion after disturbance should receive high pri- require reestablishing forests in a significant ority as policy considerations. Forest loss portion of what is now agriculture and pasture moves a large portion of the carbon land. However, reforesting this part of the sequestered in forests into the atmosphere, U.S. (almost all land east of the Mississippi) is particularly where the loss includes not only not feasible from an economic and food-secu- trees but also the decomposition of soil car- rity perspective. Today’s recovery from the for- bon. Because of climate change, increasing est clearing and wood-based economy of the threats from disturbance, and continued popu- 1800s and early 1900s will likely sustain car- lation growth and resulting exurban develop- bon storage rates at the current rate for ment, we cannot assume that all existing decades, but not indefinitely. forests will remain. Because there is a high But, forest carbon storage only gets us part of likelihood that climatic patterns will shift and the way. Even under the best scenarios, the the frequency of disturbances will increase – amount of carbon storage potential is finite. potentially making existing tree species less Strategies that combine increased use of forest suited to their environment – it would be pru- products to offset fossil fuel use (such as use of dent to focus on regeneration after disturbance biomass energy and substitution), in conjunc- to help ensure maintenance of forests. tion with increasing carbon storage on The various strategies for storing carbon in forested landscapes, are likely to produce the forests have different associated risks and levels most sustainable forest carbon benefits. of uncertainty. Retaining forests (which also Every strategy we examined has tradeoffs. includes regenerating after disturbance) and Avoiding deforestation and increasing the har- afforestation both involve low levels of uncer- vest interval in the U.S. may move timber pro- tainty regarding carbon consequences and duction elsewhere, resulting in no net benefit therefore low risk to carbon storage – aside for carbon in the atmosphere. Reestablishing from the risks of carbon loss in disturbance or forests has great potential but will also displace that the deforestation will simply happen else- current land uses such as farming and pasture. where. The carbon benefits of using biomass Increasing forest product use and forest bio- energy and long-lived forest products are also mass energy will require more active forest fairly certain, as long as forests regenerate. management over larger areas than currently Lengthening harvest intervals involves a bit occurs and may lower forest carbon stores. more risk because disturbance would occur in Intensive silviculture can increase growth, but forests with higher carbon stores and because decrease streamflow and biodiversity. Forest decision-makers can change harvest intensity products in landfills increase carbon storage, quickly relative to forest growth. but the resulting methane emissions pose a Regardless of the risks and uncertainties, problem. A better use for waste material, there- any policy to encourage forest carbon storage fore, is energy production. Recognizing these should: 1) promote the retention of existing tradeoffs will be vital to any effort to promote forests; 2) account for other greenhouse gas forest carbon. effects, such as methane and nitrous oxide emissions and biophysical changes; 3) account for harvest moving elsewhere indirectly caused by changes in manage- 800 ment with the project boundary; 4) rec- 600 ognize other environmental benefits of Figure 13. The carbon balance Carbon Release forests, such as biodiversity, nutrient of the U.S. forest sector shows 400 that clearing for agriculture, management, and watershed protection; pasture, development, and 200 5) focus on the most robust and certain wood use released ~42,000 Tg 0 carbon storage benefits in any compen- of carbon from 1700 to 1935, Carbon Storage sation scheme; 6) recognize the difficulty

(Tg Carbon/year) –200 and recovered about 15,000 Tg and expense of tracking forest carbon, of carbon from 1935-2010. (Used with permission, from 1700 1800 1900 2000 the cyclical nature of forest growth and U.S. Forest Carbon Balance U.S. Forest Journal of Environmental Quality Year regrowth, and the extensive movement 35:1461-1469 (2006)) of forest products globally; 7) recognize 14 esa © The Ecological Society of America • [email protected] ISSUES IN ECOLOGY NUMBER THIRTEEN SPRING 2010 that the value of any carbon credit will Roy, D.J. Barrett, C.W. Cook, K.A. Farley, depend on how well the carbon can be mea- D.C. le Maitre, B.A. McCarl, and B.C. sured and verified; 8) acknowledge that cli- Murray. 2005. Trading water for carbon mate change and population growth will with biological sequestration. Science. 310: 1944-1947. increase the potential for forest loss and may Mitchell, S.R., M.E. Harmon, and K.E.B. keep large-scale projects from reaching their O'Connell. 2009. Forest fuel reduction full potential; 9) recognize the tradeoffs; and alters fire severity and long-term carbon 10) understand that the success of any car- storage in three Pacific Northwest ecosys- bon mitigation strategy depends on human tems. Ecological Applications. 9: 643-655. behavior and technological advances in Murray, B.C., B.L. Sohngen, A.J. Sommer, addition to forest biology. Finally, because B.M. Depro, K.M. Jones, B.A. McCarl, D. CO2 remains in the atmosphere for more Gillig, B. DeAngelo, and K. Andrasko. than 100 years, any action to avoid further 2005. Greenhouse gas mitigation potential emissions should be undertaken as soon as in U.S. forestry and agriculture, EPA-R-05- possible. 006. U.S. Environmental Protection Few forests are managed solely for carbon Agency, Office of Atmospheric Programs, – rather, carbon storage serves as a co-bene- Washington, D.C. fit that accompanies or perhaps helps pay Nabuurs, G.J., O. Masera, K. Andrasko, P. for other ecosystem services provided by Benitez-Ponce, R. Boer, M. Dutschke, E. forests (Box 2). As we have discussed Elsiddig, J. Ford-Robertson, P. Frumhoff, T. above, elevating carbon storage to the pri- Karjalainen, O. Krankina, W.A. Kurz, M. mary focus of management could poten- Matsumoto, W. Oyhantcabal, N.H. Ravin- tially impede the other co-benefits of dranath, M.J. Sanz Sanchez, and X. Zhang. forests. A focus on carbon storage to the 2007: Forestry. In Climate Change 2007. detriment of other ecosystem services would Mitigation. Contribution of Working be short-sighted. Group III to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change [B. Metz, O.R. Davidson, For Further Reading P.R. Bosch, R. Dave, L.A. Meyer (eds)], Cambridge University Press, Cambridge, Birdsey, R.A., K.S. Pregitzer, and A. Lucier. United Kingdom and New York, NY, 2006. Forest carbon management in the USA. United States: 1600-2100. Journal of Nowak, D.J. and D.E. Crane. 2002. Carbon Environmental Quality. 35: 1461-1469. storage and sequestration by urban trees in CCSP. 2007. The First State of the Carbon the USA. Environmental Pollution. 116: Cycle Report (SOCCR): The North 381-389. American Carbon Budget and Implica- Skog, K.E. 2008. Carbon storage in forest tions for the Global Carbon Cycle. A products for the United States. Forest Report by the U.S. Climate Change Products Journal. 58: 56-72. Science Program and the Subcommittee on U.S. Environmental Protection Agency. 2008. Global Change Research [King, A.W., L. Inventory of U.S. Greenhouse Gas Emis- Dilling, G.P. Zimmerman, D.M. Fairman, R.A. Houghton, G. Marland, A.Z. Rose, sions and Sinks: 1990-2006. 394 pp. and T.J. Wilbanks (eds.)]. National Woodbury, P.B., J.E. Smith, and L.S. Heath. Oceanic and Atmospheric Administration, 2007. Carbon sequestration in the US for- National Climatic Data Center, Asheville, est sector from 1990 to 2010. Forest Ecology NC, USA, 242 pp. Management. 241: 14-27. Harmon, M.E., A. Moreno, and J.B. Domingo. 2009. Effects of partial harvest on the carbon stores in Douglas-fir/Western hemlock Acknowledgements forests: A simulation study. Ecosystems. 12: 777-791. We thank Noel Gurwick, special editor, for his Hurteau, M.D., G.W. Koch, and B.A. many helpful suggestions. Funding for this Hungate. 2008. Carbon protection and fire project was provided by Joint Venture risk reduction: toward a full accounting of Agreement 08-JV-11221633-248 between the forest carbon offsets. Frontiers in Ecology and USDA Forest Service Rocky Mountain the Environment. 6: 493-498. Research Station and the Ecological Society Jackson, R.B., E.G. Jobbagy, R. Avissar, S.B. of America.

About the Scientists expressed by the authors in ESA publications is assumed by the editors or the publisher. Michael G. Ryan, USDA Forest Service, Rocky Mountain Research Station, Fort Jill S. Baron, Editor-in-Chief, Collins, CO 80526 and Graduate Degree US Geological Survey and Colorado State Program in Ecology, Colorado State University, [email protected]. University, Fort Collins, CO 80523 Mark E. Harmon, Department of Forest Advisory Board of Issues in Ecosystems and Society, Oregon State Ecology University, Corvallis, OR 97331 Richard A. Birdsey, USDA Forest Service, Charlene D'Avanzo, Hampshire College Northern Research Station, Newtown Square, Robert A. Goldstein, Electric Power Research PA 19073 Institute Christian P. Giardina, USDA Forest Service, Noel P. Gurwick, Union of Concerned Institute of Pacific Islands Forestry, Hilo, HI Scientists 96720 Rachel Muir, US Geological Survey Linda S. Heath, USDA Forest Service, Christine Negra, the H. John Heinz III Northern Research Station, Durham, NH 03824 Center for Science, Economics, and the Richard A. Houghton, Woods Hole Research Environment Center, Woods Hole, MA 02540 Louis Pitelka, University of Maryland - Robert B. Jackson, Nicholas School of the Center for Environmental Science Environment and Earth Sciences and Sandy Tartowski, USDA - Agricultural Department of Biology, Duke University, Research Service, Jornada Experimental Range Durham, NC 27708 David S. Wilcove, Princeton University Duncan C. McKinley, American Association Kerry Woods, Bennington College for the Advancement of Science, Washington, DC 20005 Ex-Officio Advisors James F. Morrison, USDA Forest Service, Northern Region, Missoula, MT 59807 Jayne Belnap, US Geological Survey Brian C. Murray, Nicholas Institute for Clifford S. Duke, Ecological Society of America Environmental Policy Solutions and Nicholas Robert Jackson, Duke University School of the Environment, Duke University, Richard Pouyat, USDA - Forest Service Durham, NC 27708 Diane E. Pataki, Department of Earth System ESA Staff Science and Department of Ecology & Evolutionary Biology, University of California, Clifford S. Duke, Director of Science Programs Irvine, CA 92697 Aleta Wiley, Science Programs Assistant Kenneth E. Skog, USDA Forest Service, Forest Products Laboratory, Madison, WI 53726 Additional Copies Science Writing and Layout This report and all previous Issues in Ecology are available electronically for free at Christine Frame, Science writer http://esa.org/science_resources/issues.php Amanda Mascarelli, Science writer Print copies may be ordered for $5 each online Bernie Taylor, Design and layout or by contacting ESA: About Issues in Ecology Ecological Society of America 1990 M Street NW, Suite 700 Washington, DC 20036 Issues in Ecology uses commonly-understood lan- (202) 833-8773, [email protected] guage to report the consensus of a panel of scientific experts on issues related to the environment. The text for Issues in Ecology is reviewed for technical content by external expert review- ers, and all reports must be approved by the Editor-in-Chief before publication. This report is a publication of the Ecological Society of America. No responsibility for the view 16 esa © The Ecological Society of America • [email protected]