S. Brown
Sandra Brown is with the United Stales Environmental Protection Agency's National Health and Environmental Effects Research Laboratory, Western Ecology Division, Corvallis, Oregon, USA.
Note: This article draws heavily on a recent review of this topic for the Intergovernmental Panel on Climate Change Scientific Assessment for 1995 (see Brown et al., in press) and also on valuable additional input from the co-authors of that study.
Forests have the potential to contribute to climate change through their influence on the global carbon cycle. They store large quantities of carbon in vegetation and soil, exchange carbon with the atmosphere through photosynthesis and respiration, release carbon into the atmosphere when they are disturbed, become atmospheric carbon sinks during regrowth after disturbance and can be managed (locally) to alter their role in the carbon cycle. Recent studies suggest that the local management of forests for carbon conservation and sequestration could mitigate emissions of carbon dioxide by an amount equivalent to 11 to 15 percent of fossil fuel emissions over the same period.
The importance of forests for such goods and services as wood products, fuel, conservation of soil and water resources, recreation and biological diversity has been recognized for a long time. Now, forests are also being recognized as playing important roles in global biogeochemical cycles, particularly the global carbon cycle (Dixon et al., 1994). The global carbon cycle is important for its role in regulating the concentration of carbon dioxide (CO2), an important greenhouse gas, in the atmosphere. The increasing concentration of CO2 in the atmosphere contributes to global warming and, thus, to climate change. Major pools of carbon are the atmosphere, fossil fuels, oceans and the terrestrial biota and soils (see Fig. 1). Carbon is exchanged between these pools and the atmosphere as CO2 by: burning fossil fuels; net primary production and respiration of the biota, soils and other forms of dead organic matter; deforestation, forestation (reforestation and afforestation) and regrowth of vegetation after disturbance; and diffusion with the oceans.
Forests are important in the global carbon cycle because they store large quantities of carbon in vegetation and soil, exchange carbon with the atmosphere through photosynthesis and respiration, are sources of atmospheric carbon when they are disturbed by human or natural causes (e.g. wildfires, logging with poor harvesting procedures, clearing and burning for conversion to non-forest uses) and become atmospheric carbon sinks (i.e. with a net absorption of CO2 from the atmosphere) during regrowth after disturbance. Humans have the potential through forest management to change the size of carbon pools and to alter the flow of carbon (or flux) between them, thus modifying their role in the carbon cycle and their potential to affect climate.
The purpose of this article is to present the current state of knowledge about the role of forests in the global carbon cycle, with respect to the magnitude of their carbon pools and flux, and the potential role of forests to mitigate carbon dioxide emissions and thus affect global warming. Forests have the potential to influence global warming in other ways too. For example, through the production of other greenhouse gases such as carbon monoxide, ozone and nitrous oxides (see Brown et al., in press) and changes in albedo or the reflectivity of land as forests are converted to other land cover types. However, the most important way that forests can potentially influence climate change is through their effect on the levels of CO2 in the atmosphere and thus on the global carbon cycle.
Global forests cover about 3.4 Gha (Gha = 109 or 1 000 million ha) (FAO, 1995). Most of the forests are in the low latitudes (0° to 25° N and S) or tropical zone (52 percent), followed by the high latitudes (50° to 75° N and S) or boreal zone (30 percent) and mid latitudes (25° to 50° N and S) or temperate zone (18 percent). In addition, globally there are about 1.7 Gha of other wooded lands, which include open wooded land, scrub, brushland and forest fallows (resulting from shifting cultivation) (FAO, 1995). Even though some of the other wooded lands are unsuitable for forestry, they have considerable potential to mitigate CO2 emissions if managed appropriately (e.g. through fire prevention in savannahs).
Disturbances cause forests to become sources of CO2. These disturbances may be natural or human in origin, including conversion to non-forest uses, particularly agriculture and pastures; overharvesting and degradation; large-scale occurrences of wildfire; fire control; and pest and disease outbreaks. On the other hand, when areas of harvested and degraded forests or agricultural and pasture lands are allowed to revert naturally to forest or are deliberately reforested, they again become carbon sinks for a given period of time. This is the general tendency with many of the forests in temperate and boreal zone countries that were harvested (clear-cut or selectively harvested) or cleared in the past. The area of forests in these zones is undergoing little change (FAO, 1995).
By contrast, tropical forests are undergoing high rates of loss, currently estimated to be about 15.4 million ha/yr during the period 1980-1990, although there is much uncertainty about this figure (FAO, 1993). Much of the deforested area is converted to agricultural land, pasture or shifting cultivation. In addition to deforestation, large areas of forests are harvested. Between 1986 and 1990, an estimated 5.9 million ha/yr of tropical forests were logged. While logging reduces the amount of carbon stored on these lands, the forests will regenerate and accumulate carbon (generally at a faster rate than that previous to logging) if they are not severely damaged during harvesting operations and if they remain under sound management and are protected from natural and human forces that would negatively affect their vegetation and soils. However, many of these forests become degraded (see Lanly in FAO, 1982; Brown et al., 1993b), thus affecting their carbon sequestration rate or capacity. Forest degradation, resulting in a loss of carbon from the system either from the vegetation or from the soils, results from damage to residual trees and soil through poor logging practices, illegal logging, excessive fuelwood collection, overgrazing and fire (Goldammer, 1990; Brown et al., 1993b; FAO, 1993; Flint and Richards, 1994). Clearly, human influences have implications for the present role of forests in global carbon cycles and in future carbon sequestration.
Carbon pools in forests
The world's forests contain an estimated 340 Pg C (Pg = 1015 g or 1 gigaton or 1 000 million tonnes) in live and dead above- and below-ground vegetation and 618 Pg C in mineral soil and the O horizon (Table 1). At present, carbon budgets are incomplete for all the world's forests, about 6 percent of which have no carbon budgets. The world's forests contain more than 55 percent of the global carbon stored in vegetation and more than 45 percent of that in soil (see Fig. 1). The carbon content of all forest components was estimated to arrive at these estimates. However, some components are poorly known, such as the carbon content in woody detritus, slash and roots. This contributes to the uncertainty in estimates of the total carbon pool.
Most of the carbon pool in forest vegetation is located in tropical forests (62 percent), whereas most of the carbon pool in forest soils is located in boreal forests (54 percent). Within the tropics, the quantity of carbon in forest vegetation is virtually equal to that in forest soil (Table 1). Soil under temperate forests contains about one-third more carbon than that found in forest vegetation, while soils in the boreal zone contain more than four times as much carbon as that in the forest vegetation.
Carbon flux from forests
Temperate and boreal forests are currently estimated to be a net carbon sink of about 0.7±0.2 Pg/yr (Table 1). That is to say, they already offset some of the emissions from the burning of fossil fuels. Temperate and boreal forests are, on average, composed of relatively young classes with relatively high rates of growth (and therefore of carbon sequestration) following past human and natural disturbances. A larger proportion of these forests are actively managed than are the forests of the tropical zone. Moreover, some areas may be responding to increased levels of atmospheric CO2 and nitrogen (fertilization effect) (Brown et. al., in press). However, there is a finite time period over which net carbon absorption can occur, after which forests reach a steady state. Some scientists believe the current carbon sink in European forests may disappear within 50 to 100 years (Kauppi, Mielikainen and Kuusela, 1992), while others suggest that it may take forests several centuries or even millennia to reach a carbon-steady state in all their components, including soil (Lugo and Brown, 1986).
Tropical forests were estimated to be a relatively large net carbon source of 1.65±0.4 Pg/yr in 1990 (Table 1). This source is equivalent to almost 30 percent of annual emissions of CO2 related to the use of fossil fuel. Although this is the best estimate available in the literature, there are many reasons to believe that the mean is smaller and the uncertainty range larger than shown (Lugo and Brown, 1992). Unlike the temperate and boreal forests, where estimated carbon fluxes are for the most part based on data from periodic national inventories (i.e. field measurements), the estimated carbon flux for tropical forests is based on a theoretical model. The model considers the case of forests that are cleared or harvested and followed by regrowth for up to about 50 to 100 years, during which time carbon is allowed to accumulate. Furthermore, the model assumes that all other forests not reported to be directly affected by humans during the period of model simulation (about 1850-1990) are in a carbon-steady state (Houghton et al., 1987) and that none of the regrowth is influenced by increased levels of atmospheric CO2 and: nitrogen. Recent work questions the steady state assumption because humans have brought about changes in forest cover in previous centuries from which nature may still be recovering (Brown et al., 1993a). Therefore, the net tropical carbon flux could be higher or lower than that reported here, depending on the relative contribution of forest lands that are still gaining carbon through recovery from past human disturbances or are losing carbon through continued human use (Lugo and Brown, 1992; Brown et al., 1993a).
TABLE 1. Estimated carbon pools and flux in forest vegetation (above- and below-ground living and dead mass, including woody debris) and soils (O horizon and mineral soil to 1 m depth) in world forests
|
Region/country |
C pools (Pg) |
C flux(Pg/yr-1) |
|
|
Vegetation |
Soils |
||
|
BOREAL ZONE |
|||
|
Former USSR1 |
63 |
111 |
+0.30 to + 0.50 |
|
Canada |
12 |
211 |
+ 0.08 |
|
Alaska |
2 |
11 |
|
|
Subtotal |
77 |
333 |
+0.48 ± 0.20 |
|
TEMPERATE ZONE |
|||
|
United States |
15 |
21 |
+0.08 to + 0.25 |
|
Europe² |
10 |
18 |
+0.09 to +0.12 |
|
China |
17 |
16 |
- 0.02 |
|
Australia |
9 |
14 |
trace |
|
Subtotal |
51 |
69 |
+0.26 ± 0.10 |
|
TROPICAL ZONE |
|||
|
Asia |
41-54 |
43 |
-0.50 to - 0.90 |
|
Africa |
52 |
63 |
-0.25 to - 0.45 |
|
America |
119 |
110 |
-0.50 to - 0.70 |
|
Subtotal |
212 |
216 |
-1.65 ± 0.40 |
|
Total |
340 |
618 |
-0.90 ± 0.50 |
Note
Estimate dates vary with country and region but cover the 1980s. Estimates are based on complete carbon budgets in all latitudes using data from original sources and/or from adjustments for completeness (revised version from Brown et al., in press - see footnote 2 for revisions).* Included with United States.
1 Soil pool excludes peat.
² Includes Nordic countries. Total live biomass carbon was assumed to be the product of growing stock in 1990, converted to carbon units, and the mid-point of the expansion factors given in Kauppi. Mielikainen and Kuusela (1992): an additional 40 percent of live biomass was added to account for litter and dead wood. The soil pool is the product of forest area and a soil carbon density of 9 kg/m² (Dixon et al 1994).
The error terms associated with the carbon flux estimates in Table 1 are derived from the range of values resulting from the use of different assumptions in the carbon budgets for a given country or region. They do not represent errors derived from statistical procedures. Errors enter the flux estimation procedure through uncertainties and biases in the primary data, and these compound as the data are combined to draw inferences (Robinson, 1989). Many estimates for components of the forest sector carbon budget are probably no more accurate than ±30 percent of their mean, while others may be as inaccurate as >±50 percent of their mean (Robinson, 1989). These errors are compounded when making global estimates of carbon flux, perhaps to large proportions, but to what extent is at present unknown.
The estimated net carbon flux from the world's forests is a source of 0.9±0.5 Pg/yr, or about 16 percent of the amount currently produced each year by the burning of fossil fuels and cement manufacturing (see Fig. 1, p. 6). Other research on this issue strongly suggests that this missing sink is on the land (Schimel et al., 1995) Because forests have the capacity to store large quantities of carbon at relatively large rates, it is assumed that the "missing sink" is in forests. However, the insertion of this figure of the net carbon flux for forests into the global carbon budget results in an imbalance, often referred to as the "missing sink", of 1.2± 1 Pg/yr. The imbalance is the difference between known sinks and sources of carbon and the amount "needed" to balance the carbon budget (Fig. 1). Because, as noted above, the primary data for carbon budgets for temperate and boreal countries originate from national forest inventories (i.e. hard data), the author concludes that a large part of the imbalance in the global carbon budget must be due to a carbon sink in tropical latitudes, for which the estimates originate from a modelling approach. This conclusion, also suggested by others (Lugo and Brown, 1992; Taylor and Lloyd, 1992; Schimel et al., 1995), could be due to a combination of stimulated regrowth from CO2 and nitrogen fertilization and climate as well as more extensive forest regrowth. To resolve this issue, it is clear that repeated national forest inventories, including permanent sample plots, are needed in tropical latitudes.
In terms of global warming, the present state of understanding as given above suggests that terrestrial ecosystems, and forests in particular, are probably contributing little to the net increase in atmospheric CO2 and are thus contributing little to global warming. However, this may not continue into the future if temperate and boreal forests are allowed to reach maturity, at which point they will contribute less to carbon sequestration (although representing a larger carbon pool), and if current rates of tropical deforestation and degradation continue.
Major objectives for managing forest lands generally include: industrial wood production, fuelwood production, production of non-wood forest products, protection of natural resources (e.g. biological diversity, water and soil), wildlife management, recreation, rehabilitation of degraded lands, and the like. Carbon conservation and sequestration resulting from management for these above-mentioned objectives will be an added benefit because they reduce atmospheric concentrations of CO2 and thus mitigate climate change. Forest management practices for mitigating climate change can be grouped into three categories: management for carbon conservation, carbon storage and carbon substitution (Brown et al., in press).
FIGURE 1. The global carbon cycle showing average pools and fluxes during the 1980s.
The total net flux to me atmosphere 7.1 Pg C/yr (fossil fuels cement and changing kind corer or use). Of the net flux to tire atmosphere, 3.2 Pg C/yr stay in the atmosphere, a net 2 Pg C/yr go into the oceans (ocean sink), while it is assumed that 1.9 Pg C/yr are taken up by terrestrial ecosystems to balance the budget
Note: Pg = 1015g.
Source: Adapted from Schimel et al., 1995.
The goal of conservation management is essentially to prevent carbon emissions by conserving existing carbon pools in forests through options such as controlling deforestation, protecting forests in reserves, changing harvesting regimes and controlling other disturbances such as fire and pest outbreaks. The most significant improvement in carbon conservation practices could occur through a reduction of deforestation and degradation in the tropics. It is noteworthy, however, that these practices are caused primarily by the expansion of arable and grazing lands and by subsistence and commodity demand for wood products, which in turn are a response to the underlying pressures of population growth, socio-economic development and political forces. Thus, programmes aimed at carbon conservation through reduced deforestation must be accompanied by measures that increase agricultural productivity and sustainability and deal with overriding socio-economic and political concerns.
In recent years, there has been significant expansion of "protected areas", comprising both mature forests and forests in other development stages, for the conservation of biological diversity and watershed protection, for example. Carbon pools should remain the same or increase in size in these areas, depending on their present age class distribution. It is also likely that the trend towards the management of forests for sustainable timber production in a significantly larger proportion of the world's forests will continue. Using forests this way, including extending rotation cycles, reducing damage to remaining trees, reducing logging waste, implementing soil conservation practices and using wood more efficiently (e.g. reducing postharvest losses in processing), can help to ensure that a larger portion of their total carbon is conserved.
The goal of storage management is to increase the storage of carbon in the vegetation and soil of forest ecosystems by increasing the area and/or carbon density of natural and plantation forests, and also to increase its storage in durable wood products. Increasing the carbon pool in vegetation and soil can be accomplished by protecting secondary forests and other degraded forests whose biomass and soil carbon densities are less than their maximum value, thereby allowing them to sequester carbon by natural or artificial regeneration and soil enrichment. Other approaches are to establish plantations on non-forested lands; promote natural or assisted regeneration in secondary forests, following this up with protection; and increase the tree cover on agricultural or pasture lands through agroforestry. The carbon stored in durable wood products can be increased by expanding demand for wood products at a faster rate than the decay of wood and by extending the lifetime of wood products. Sequestering carbon by storage management is only a short-term option, producing a finite carbon sequestration potential beyond which little additional carbon can be accumulated. The process may take place over a period in the order of several decades to a century or more, depending on the present age class of a forest, the attainable maximum carbon density, forest type, species selection and the wood products produced and retained.
The third category aims at the substitution of forest biomass carbon for non-renewable sources of raw material and fossil fuel-based energy, such as construction materials and biofuels. This approach involves increasing the use of forests for wood products and fuels, obtained either from establishing new forests or plantations, or increasing the growth and subsequent potential fibre production of existing forests through silvicultural treatments (Brown et al., in press). In the case of forests established on previously non-forested lands for energy products such as fuelwood, there is the potential both to increase the amount of carbon stored on the land and, if the wood is burnt as fuel, to increase carbon storage as unburnt fossil fuels (Sampson et al., 1993). Over long periods, the substitution management method is likely to be more effective in reducing carbon emissions than the physical storage of carbon in forests or forest products (Marland and Marland, 1992).
Estimated amount of carbon conserved and sequestered
Two recent studies (Nilsson and Schopfhauser, 1995; Trexler and Haugen, 1994) were combined to arrive at a global estimate of the potential amount of carbon that could be conserved and sequestered by different forested regions of the earth between 1995 and 2050. These studies were chosen because they were the only ones that were global in nature, had done an extensive literature review of the land availability issue and included feasible rates of establishment of management options. Both studies assumed aggressive but unspecified policy and financial interventions in the forestry sectors, with no future change in climate that might interfere with the proposed strategies.
Nilsson and Schopfhauser (1995) estimated the potential for carbon sequestration through a feasible global forestation programme. Using many sources, they estimated the amount of land likely to be available, feasible annual planting rates, likely growth rates and rotation lengths for different countries and regions. They used a growth model to estimate the quantity of carbon fixed in above- and below-ground biomass, litter and soil organic matter for the period 1995-2050. No assumptions about the "life expectancy" of the wood produced were made. Further, the amount of carbon calculated to be sequestered by this programme will only be realized if the forests are harvested at their designated rotation lengths.
Trexler and Haugen (1994) focused on the tropics and included the options of slowing deforestation and natural or assisted regeneration of forests. They made country-level estimates for each decade between 1990 and 2050 for 52 tropical countries accounting for virtually all of the tropical forests. For each country and decade, based on a detailed country-by-country analysis, they estimated current and projected future deforestation rates, the potential reduction in deforestation through the feasible implementation of alternative land uses and the area available at present for natural or assisted native forest regeneration, followed by protection, as well as the likely rates of implementation. They also estimated the change in above-ground biomass carbon associated with each land-use change. To be consistent with the study of Nilsson and Schopfhauser (1995) the author has added estimates of below-ground biomass, soil and litter.
Together, the studies suggest that, globally, 700 million ha of land might be available for carbon conservation and sequestration programmes - 345 million ha on currently non-forested or understocked land for plantations and agroforestry, 138 million ha through slowed tropical deforestation, and 217 million ha through natural and assisted regeneration of tropical forests (Trexler and Haugen, 1994; Nilsson and Schopfhauser, 1995). This amount of land could conserve and sequester 60 to 87 Pg C by 2050 (Table 2). Globally, forestation and agroforestry account for 50 percent of the total (38 Pg C), with about 20 percent of this accumulating in soils, litter and below-ground biomass (Nilsson and Schopfhauser, 1995). The amount of carbon that could be conserved and sequestered by forest sector practices by 2050 under baseline conditions is equivalent to about 11 to 15 percent of total fossil fuel emissions over the same period (the IS92 scenario from Houghton, Calland and Varney, 1992).
The tropics have the potential to conserve and sequester by far the largest quantity of carbon (80 percent) followed by the temperate zone (17 percent) and the boreal zone (3 percent only) (Table 2). More than half of the tropical amount would result from natural and assisted regeneration followed by protection and slowed deforestation. Forestation and agroforestry would contribute less than half of the tropical amount but, without them, regeneration and slowed deforestation would be highly unlikely (Trexler and Haugen, 1994).
Annual rates of carbon conservation and sequestration from all practices would increase over time and reach about 2.2 Pg/yr by 2045 (Fig. 2), with most of the carbon accumulating in the tropical zone and the least in the boreal zone. Carbon savings from slowed deforestation and regeneration would initially be the highest but, from 2025 onwards, when plantations would have reached their maximum carbon accretion, they would sequester practically identical amounts as slowed deforestation and regeneration (Fig. 2b).
The mitigation potential of forests described above is based on a calculation of what is physically possible for increasing carbon storage and sequestration; the calculations do not consider the social or economic feasibility of making such changes in land use and forest management. In addition, the calculations do not take into account the effects of increases in the concentration of CO, and other atmospheric pollutants, the effects of a changing climate, or the effects of future changes in land use caused by increases in human population densities. Each of the promising forest management options for mitigating carbon emissions would be affected differently under a changed climate and atmosphere as well as a changed land use. For natural forest regeneration and slowed deforestation options in the tropics, demand by an increasing human population for more land for agriculture and wood and non-wood forest products at the expense of forest cover is likely to have a major effect on land availability and forest management objectives; direct and indirect effects of climate change on land-use potentials may be less important in comparison (Brown et al., 1993a; Solomon et al., in press). In countries of the middle and high latitudes, land-use patterns are relatively stable at present, and the effects of a change in climate and atmospheric composition would probably be more significant (Kirschbaum et al., in press). For forestation options, the key factors are how a changed climate and atmosphere would affect the suitability and availability of lands for plantation and agroforestry establishment as well as the effects on species selection, rates of tree growth and other pathways of sequestering carbon in, for example, soil, litter, dead wood and roots. However, because plantations are managed, adaptations to changes in climate and atmospheric composition may be feasible, including species substitutions and alternate-length rotations.
TABLE 2. Global estimates of the potential amount of carbon that could be sequestered and conserved by forest management practices between 1995 and 2050
|
Latitudinal belt |
Country/region |
Practice |
Carbon sequestered and conserved (Pg)1 |
|
Boreal |
Canada | Forestation | 0.68 |
| Nordic Europe | 0.03 | ||
| Former USSR | 1.76 | ||
| Subtotal | 2.40 | ||
|
Temperate
|
Canada² |
Forestation |
0.43 |
|
United States |
|
3.07 |
|
|
Europe |
|
0.96 |
|
|
China |
|
1.70 |
|
|
Asia |
|
2.19 |
|
|
South Africa |
|
0.44 |
|
|
South America |
|
1.02 |
|
|
Australia |
|
0.31 |
|
|
New Zealand |
|
1.70 |
|
|
Subtotal |
|
11.80 |
|
|
United states |
Agroforestry |
0.29 |
|
|
Australia |
|
0.36 |
|
|
Subtotal |
|
0.70 |
|
|
Tropical
|
Tropical |
Forestation |
8.02 |
|
Trop. Africa |
|
0.90 |
|
|
Trop. Asia |
|
7.50 |
|
|
Subtotal |
|
16.40 |
|
|
Trop. America |
Agroforestry |
1.66 |
|
|
Trop. Africa |
|
2.63 |
|
|
Trop. Asia |
|
2.03 |
|
|
Subtotal |
|
6.30 |
|
|
Trop. America |
Regeneration ³ |
4.80-14.30 |
|
|
Trop. Africa |
3.00-6.70 |
|
|
|
Trop. Asia |
3.80-7.70 |
|
|
|
Subtotal |
|
11.50-28.70 |
|
|
Trop America |
Slow deforestation ³ |
5.00-10.70 |
|
|
Trop Africa |
|
2.50-4.40 |
|
|
Trop Asia |
|
3.30-5.80 |
|
|
Subtotal |
|
10.80-20.80 |
|
|
Total |
|
60.00-87.00 |
1 Pg = 1015 g.² Canada has both boreal and temperate forests.
³ Includes an additional 25 percent of above-ground carbon to account for below-ground carbon in roots, litter and (based on data in Nilsson and Schopfhauser 1995 and Brown et al. 1993b). The range in values is based on the use of low and high estimates of biomass carbon density resulting from the uncertainty in these estimates.
Source: Brown et al., in press.
Source: Based on Brown et al., in press.
Although calculations indicate that the world's forests are at present a net source of CO2 for the atmosphere, a logical attempt to balance the global carbon budget suggests that net CO2 emissions from forests and other terrestrial ecosystems must be close to zero. That is, although forests are an important component of the global carbon cycle through their regulation of CO2 fluxes and storage, at present they are likely to be making a minimal contribution to global warming. This could change in the future for many reasons, including the continued and increasing clearing and degradation of tropical forests, the maturing of mid- and high-latitude forests and increased mortality rates and wildfires of mid- and high-latitude forests as they succumb to climate change. However, through the implementation of forest management options that are compatible with traditional objectives of forestry, over the next 50 years or so there is a potential to conserve and sequester an amount of carbon equivalent to about 11 to 15 percent of total fossil fuel emissions over the same period. The adoption of forest management options that conserve and sequester carbon would help to prevent forests from becoming a significant net source of CO2 for the atmosphere in the future and thereby help to offset other factors that contribute to accelerated global warming.
Brown, S., Lugo, A.E. & Wisniewski, J. 1992 Missing carbon dioxide. Science, 257: 11.
Brown, S., Hall, C.A.S., Knabe, W., Raich, J., Trexler, M.C. & Woomer, P. 1993a. Tropical forests: their past, present and potential future role in the terrestrial carbon budget Water Air Soil Pollut., 70: 71-94.
Brown, S., Iverson, L.R., Prasad, A. & Liu, D. 1993b. Geographical distribution of carbon in biomass and soils of tropical Asian forests. Geocarto Int., 8: 45-60
Brown, S., Sathaye, J., Cannell, M. & Kauppi, P.E. Management of forests for mitigation of greenhouse gas emissions. In Working Group II, Second Assessment Report, Intergovernmental Panel on Climate Change Chap. 24. Cambridge, UK, Cambridge University Press. (in press)
Dixon, R.K., Brown, S., Houghton, R.A., Solomon, A.M., Trexler, M.C. & Wisniewski, J. 1994. Carbon pools and flux of global forest ecosystems. Science, 263: 185-190.
FAO. 1982. Tropical forest resources. FAO Forestry Paper No 30. Rome.
FAO. 1993. Forest resources assessment 1990 - Tropical countries. FAO Forestry Paper No. 112. Rome.
FAO. 1995. Forest resources assessment. 1990 Global synthesis. FAO Forestry Paper No. 124. Rome.
Flint, E.P. & Richards, J.F. 1994. Trends in carbon content of vegetation in South and Southeast Asia associated with changes in land use. In V.H. Dale, ed. Effects of land use change on atmospheric CO2 concentrations: Southeast Asia as a case study, Chap. 6. New York, Springer Verlag.
Goldammer, J.G., ed. 1990. Fire in the tropical biota: ecosystem processes and global challenges. Ecological Studies 84. Berlin, Germany. Springer Verlag.
Houghton, J.T., Callander, B.A. & Varney, S.K., eds. 1992 Climate Change 1992. The supplementary report to the IPCC scientific assessment. Cambridge UK Cambridge University Press.
Houghton, R.A., Boone, R.D., Fruci, J.R., Hobbie, J.E., Melillo, J.M., Palm, C.A., Peterson, B.J., Shaver, G.R., Woodwell, G.M., Moore, B., Skole, D.L. & Myers, N. 1987. The flux of carbon from terrestrial ecosystems to the atmosphere in 1980 due to changes in land use: geographic distribution of the global flux. Tellus, 39B: 122-139.
Kauppi, P.E., Mielikainen, K. & Kuusela, K, 1992. Biomass and carbon budget of European forests, 1971 to 1990. Science, 256: 70-74.
Kirschbaum, M., Fischlin, A., Cannell, M.G.R., Cruz, R.V.O., Cramer, W.P. & Galinski, W. The impacts of climate change on forest ecosystems. In Working Group II, Second Assessment Report, Intergovernmental Panel on Climate Change. Chap. 1. Cambridge. UK, Cambridge University Press (in press)
Lugo, A.E. & Brown, S. 1986. Steady state terrestrial ecosystems and the global carbon cycle. Vegetatio, 68: 83-90.
Lugo, A.E. & Brown, S.1992. Tropical forests as sinks of atmospheric carbon. Forest Ecol. Manage 48: 69-88.
Marland, G. & Marland, S.1992. Should we store C in trees? Water Air Soil Pollut., 64: 181-195
Nilsson, S. & Schopfhauser, W. 1995. The carbon-sequestration potential of a global afforestation program. Climatic Change, 30: 267-293.
Robinson, J.M. 1989. On uncertainty in the computation of global emissions from biomass burning. Climatic Change. 14: 243-262.
Sampson, R. N., Wright, L.L., Winjum, J.K., Kinsman, J.D., Benneman, J., Kursten, E. & Scurlock, J.M.O. 1993. Biomass management and energy. Water Air Soil Pollut., 70: 139-159.
Schimel, D., Enting, I.G., Heimann, M., Wigley, T.M.L., Rayneud, D., Alves, D. & Seigenthaler, U. 1995. CO2 and the carbon cycle. In J. T. Houghton, L.G. Meira Filho, J. Bruce, H. Lee B.A. Callander, E. Haites, N. Harris & K. Maskell, eds. Climate change 1994, radiative forcing of climate change and an evaluation of the IPCC IS92 emission scenarios, p. 35-71 Published for the IPCC. Cambridge. UK, Cambridge University Press.
Solomon, A.M., Ravindranath, N., Stewart, R., Weber M. & Nilsson, S. Wood production under changing climate and land use. In Working Group II, Second Assessment Report, Intergovernmental Panel on Climate Change, Chap. 15 Cambridge, UK, Cambridge University Press. (in press)
Taylor, J.A. & Lloyd, J. 1992. Sources and sinks of atmospheric CO2. Aust. J. Boll. 40: 407-418.
Trexler, M.C. & Haugen, C. 1994. Keeping it green: evaluating tropical forestry strategies to mitigate global warming. Washington, DC, World Resources Institute.
