x-xEPA
United States       Environmental Research   EPA/600/3-90/080
Environmental Protection   Laboratory        September 1990
Agency          Corvallis, OR 97333

Research and Development


RESPONSE AND FEEDBACKS OF

FOREST SYSTEMS TO GLOBAL

CLIMATE CHANGE
             Edited By: George A. King, Jack K. Winjum, Robert K. Dixon,
             and Lynn Y. Arnaut

-------

-------
                                       Table of Contents

Table of Contents	  ii

List of Figures  	  v

List of Tables	  vi

List of Contributors	viii

Acknowledgments  	  ix

Preface  	x

EXECUTIVE SUMMARY	  xi

1  INTRODUCTION	  1
       1.1     Subject  	  2
       1.2     Scope  	  4
       1.3     EPA's 1989 Report to Congress  	  5
               1.3.1  Findings	  5
               1.3.2 Uncertainties	  6
       1.4     Background	  7
               1.4.1  World Forests	  7
               1.4.2 Global Carbon Cycle	  10
               1.4.3 Climate Scenarios	:..  12

2      EFFECTS OF GLOBAL CLIMATE CHANGE ON FORESTS  	'..  15
       2.1  Treeline Response	  15
               2.1.1  Ecophysiological Characteristics of Treeline  	  15
               2.1.2 Treeline Response to Climate Change  	  16
               2.1.3 Research Needs  	  17
       2.2  Forest Stand Effects	  17
               2.2.1  Pacific Northwest Forests  	  18
               2.2.2 Future Modeling of United States Forests	  19
               2.2.3 Boreal Forests	  20
               2.2.4 Research Needs  	  21
       2.3     Carbon Sequestration in  Soils	  21
               2.3.1  Forest Soils and the Global  Carbon Cycle 	  22
               2.3.2 Soil Carbon Dynamics: A Conceptual Model	  23
               2.3.3 Factors Affecting the Accumulation of Soil Carbon  	  26
               2.3.4   Potential Effects of Global Change on Carbon Storage in Soils	  27
               2.3.5 Research Needs  	  28
       2.4     Effects of Global Climate Change on Global Vegetation  	  29
               2.4.1  Global Vegetation Models  	  30
               2.4.2 Equilibrium Simulations of Future Vegetation Patterns	  33
               2.4.3 Discussion of Vegetation Scenario Results	  41
               2.4.4 Research Needs  	  41
       2.5     Impacts  of Climate Change on Biological Diversity	  42

-------
3  FEEDBACKS: TERRESTRIAL CARBON STORAGE	   47
       3.1     Effects of Climate Change on Carbon Storage in Terrestrial Ecosystems:
              Equilibrium Analyses at the Global Level  	   47
              3.1.1  Methods	   47
              3.1.2  Results and Discussion  	   52
              3.1.3  Conclusions  	   56
    '   3.2     Biosphere Feedback During Climate Change	   57
              3.2.1  The Model	   59
              3.2.2  Results	   65
              3.2.3  Sensitivity Analyses	   68
              3.2.4  Discussion  	   72
              3.2.5  Research Needs 	   75

4      DIRECT EFFECTS OF ATMOSPHERIC CO2 ENRICHMENT ON FORESTS	   77
       4.1     Trees	   77
              4.1.1  Effect of C02 Enrichment on Plant Development   	   77
              4.1.2  Water-Use Efficiency	   84
              4.1.3  Research Needs 	   86
       4.2     Ecosystem Perspective	   87

5      MITIGATION OF GLOBAL CHANGE IMPACTS THROUGH FOREST MANAGEMENT  	   89
       5.1  Forest Management Opportunities  	   89
              5.1.1  Potential for Large-Scale Reforestation to  Sequester Atmospheric CO2 ......   92
              5.1.2  Calculation  of the Carbon Sequestering Potential of Temperate
                     Plantations: An Example	   98
              5.1.3  Calculation  of the Carbon Sequestering Potential of Several Forest Stand
                     Treatment Practices 	103
              5.1.4  Research Needs	  109
       5.2     Soil Management to Conserve and Sequester Carbon  	110
              5.2.1  Maintaining the Soil Carbon Pool	110
              5.2.2  Restoring Soil Carbon	114
              5.2.3  Enlarging the Soil Carbon Pool  	114
              5.2.4  Future Considerations	115
              5.2.5  Research Needs	117
       5.3  Forest Mitigation Compared with Other Biosphere Options	118
              5.3.1  Forestry Management Options	118
              5.3.2  Other Forest Management and Terrestrial  Options to Sequester Carbon  ....  118
              5.3.3  Aquatic Options  	120
              5.3.4  Conclusions  	121

6      RESEARCH NEEDS AND PLANS	123
       6.1     Feedback Processes Research (Science Question #1)	123
       6.2     Response Research (Science Question #2)	124
       6.3     Mitigation/Adaptation Research (Science Question #3)  	125
       6.4     Field Assessment  of Forest and Agroforestry Management Opportunities to
              Sequester Atmospheric C02	*•	126
              6.4.1  Goal	126
              6.4.2  Rationale	126
              6.4.3  Approach  . .	126
              6.4.4  Product	127
                                             in

-------
       6.5    Assessment and Validation of Terrestrial Biosphere Models of Global Change Feedbacks
              and Responses	127
              6.5.1  Goal	127
              6.5.2  Rationale	127
              6.5.3  Approach  	128
              6.5.4  Product	128

7      LITERATURE CITED	129

8      APPENDICES	150
       8.1    Appendix A: Extramural Research Projects Sponsored by ERL-C Global Change
              Research Program in FY 90	150
       8.2    Appendix B:   Participants in Carbon Sequestering and Soils
              Workshop, February 27-28, 1990	152
       8.3    Appendix C: Invited Participants, Workshop on Ecological and Operational
              Considerations for Large-Scale Reforestation, Corvallis, Or. May 8-10, 1990  	155
       8.4    Appendix D: Abbreviations	156
                                              IV

-------
                                         List of Figures

Figure 1.1-1.    The global carbon cycle, including major pools and annual flux of carbon             3

Figure 1.1-2.    Latitudinal variations in atmospheric C02 concentrations from 1982-1984 (adapted from
               Tucker et al. 1986)                              •                                11

Figure 2.3-1.    Simplified conceptual model of carbon pools and fluxes in terrestrial
               ecosystems.                                                                     24

Figure 2.4-1.    Holdridge life zone classification system (Holdridge 1967)                            31

Figure 2.4-2.    Histograms showing changes in the total area of six major vegetation types
               after a double-C02 induced climate change.                                        38

Figure 2.4-3.    a) Global and b) North American areas (shaded) in which predicted future
               vegetation is different from current vegetation using the UKMO double-
               COa climate scenario (Smith et al. submitted).                                      40

Figure 2.5-1.    Density/richness maps illustrating one level of biodiversity for a) amphibians (Kiester
               1971), b) reptiles (ibid.), and c) trees (Currie and Paquin 1987) in North America and/or
               the contiguous 48 United States.                                              44/45

Figure 3.1-1.    Relationship of belowground carbon pools to Holdridge life zones (Post et al.
               1982).                                                                          51

Figure 3.1-2.    Potential changes in terrestrial carbon storage based on redistribution of vegetation
               types.                                                                           53

Figure 3.2-1.    Transient carbon pulse with the a) UKMO, b) OSU, c) GFDL, and d) GISS GCMs.      67

Figure 3.2-2.    Sensitivity of extratropical emission pulse to key parameters.                         69

Figure 3.2-3.    Sensitivity of global net carbon pulse to key parameters.                             70

Figure 5.1-1.    Comparison of changes in estimated mean annual growth increment of
               Douglas-fir and loblolly pine.                                                    100

Figure 5.1-2.    Estimated yield curves of total standing crop for Douglas-fir and loblolly pine based
               on the same assumptions as Figure 5.1-1.                                        101

Figure 5.1-3.    Cumulative 10-year increases  in Douglas-fir stem biomass (carbon gain)
               resulting from fertilization with 224 kgN/ha                                       106

Figure 5.2-1.    COaflux from soil and forest floor to the atmosphere at fertilized Douglas-fir and
               western hemlock forested sites in the Oregon Cascades.                          116

-------
                                        List of Tables

Table 1.4-1     World Forested Area by Region, circa 1985   	  8

Table 1.4-2     Forest Fallow and Shrubland, circa 1980	  8

Table 1.4-3     General Circulation Models Used to Generate Double-C02 Climate Scenarios  	  13

Table 2.3-1     Biomass and Soil Carbon in Boreal, Tropical, and Temperate Forest Ecosystems  .  .  23

Table 2.4-1     Changes in Global Vegetation Distribution in a Double-COa Atmosphere using the
              Holdridge Classification Systems (as applied by Emanuel 1985 and Emanuel et
              al. 1985)	                       34

Table 2.4-2     Changes in Global Potential Vegetation Distribution in a Double-C02 Atmosphere
              using the Holdridge Classification System (as applied by Smith et al. submitted)  ..  35

Table 2.4-3     Changes in Global Vegetation Distribution in a Double-C02 Atmosphere using
              the Holdridge Classification System (as modified by Prentice 1990 and applied
              by Prentice and Fung in press)	  36

Table 2.4-4     Comparison of Global Vegetation Distribution in a Double-CO2 Atmosphere
              using the Holdridge Classification System, as applied by Three Teams of
              Investigators	  37

Table 2.4-5     Areal Extent of Land on the Globe Changing Vegetation Cover under Double-
              C02 Conditions as Estimated Using the Holdridge system (Smith et
              al.submitted)  	  39

Table 3.1-1     Aggregation Scheme for Combining Holdridge Life Zones into Biomes (Cramer
              and  Leemans in press)	  48

Table 3.1-2     Above- and  Belowground Carbon Pools for World Biomes	  50

Table 3.1-3     Changes in Areal Extent of Different Vegetation Types as Predicted under Four GCM
              Scenarios (Smith et al. submitted)  	•..  52

Table 3.2-1     Initial Conditions and Parameter Values (all units are for aboveground organic
              carbon	  65

Table 3.2-2     Range of CO2 Cumulative Pulse to the Atmosphere  	  72

Table 4.1-1     Summary of Arctic and  Salt  Marsh  (C3 and CJ Annual Plant Responses to
              Enriched C02 Environments in Field Exposure Chambers  	  78/79

Table 4.1-2     Seedling Morphological and Physiological Responses to Enriched-C02
              Environments Based on Short-term Controlled-Environment Studies (<. 1 yr, with
              an exception as marked)	  80

Table 4.1-3     Early Growth (14 weeks) of Acer macrophyllum Seedlings (n = 48) in Ambient- and
              Enriched-C02 Environments (350, 575, and 700 ppm)	  82
                                              VI

-------
Table 5.1-1     Tropical Lands with Potential for Reforestation (million ha)  	  93

Table 5.1-2     Range of Estimates of Carbon Sequestration for Tropical Forests under Scenario
               of Reforestation or Afforestation	  94

Table 5.1-3     Examples of Potential Wood Volume Growth and Calculated Carbon Fixation
               Rates, and Representative Establishment Costs for Several Species in
               the Tropical and Temperate Zones.  Based on Opinions of Individual Reforestation
               Experts at an International Workshop  	  96

Table 5.1-4     Global Reforestation/Afforestation Rates in the 1980s	  97

Table 5.1-5     The Effect of Stand Thinning on Carbon Sequestration	103

Table 5.2-1     Prioritized Strategies for Sequestering and Storing Carbon in Soils	112
                                                vii

-------
                                      List of Contributors
Section 1



Section 2.1

Section 2.2

Section 2.3

Section 2.4


Section 2,5

Section 3.1

Section 3.2



Section 4



Section 5.1
Section 5.2


Section 5.3


Section 6
Jack K. Winjum
George A. King
Robert K. Oixon

Terry D. Droessler

Terry D. Droessler

Mark G. Johnson

George A. King
Rik Leemans

Sandra Henderson

David P. Turner

Rbnald P. Neilson
Hermann Gucinski
Donald L Phillips

John D. Bailey
Robert K. Dixon
George A. King

Jack K. Winjum
Paul E. Schroeder
Kim G. Mattson
George A. King

Mark G. Johnson
Robert K. Dixon

Robert K. Dixon
Jack K. Winjum

Robert K. Dixon
Jack K. Winjum
                                              viii

-------
Acknowledgments


We would like to thank the peer reviewers for their constructive comments and contributions. We also
are very grateful to Jill Jones for her dedication and invaluable assistance in the production of this
report.
                                               ix

-------
Preface
This report was prepared by the Global Effects and Global Biogeochemistry Teams of the Global Change
Research Program (GCRP) at the US Environmental Protection Agency's Environmental Research
Laboratory in Corvallis, Oregon (ERL-C).  The research was completed in cooperation with the USDA
Forest Service, Oregon State University, the University of Virginia, the National Center for Atmospheric
Research, and the International Institute for Applied Systems Analysis, as part of the ERL-C GCRP
commitment to ORD for fiscal year 1990.  The overall  ORD GCRP plan identified two areas of research
emphasis for ERL-C:  1) assess ecological impacts of climate change on terrestrial systems; and 2)
determine the role of biospheric feedbacks to global climate change.  This report addresses both
ecological effects and biofeedbacks to global change  as manifested through the global carbon cycle.
Portions of this report have been presented at workshops and scientific meetings.  Sections of this
report have been (or will be) submitted to scientific journals or proceedings volumes for publication.

-------
                                     EXECUTIVE SUMMARY

A growing base of evidence suggests that the accumulation of greenhouse gases such as carbon
dioxide (CO^, methane, nitrous oxide,  and chlorofluorocarbons in the atmosphere could significantly
change the earth's climate. If this is true, forest systems throughout the world are projected to change
in extent and composition.  Biospheric feedbacks from forests to the atmosphere will also be altered in a
changing global climate.  Since a third of the biosphere's livable land is forested, such changes will
dramatically affect humankind.

The purpose of this report is to summarize present knowledge in three areas:  forest responses to
climate change; the role of forest feedbacks to the atmosphere; and the potential of forest management
to increase carbon sequestering. The  report discusses these  topics in the context of the global carbon
cycle and the broader concept of managing the biosphere to  offset increases in atmospheric C02.  New
research results are presented on the effects of vegetation redistribution on terrestrial carbon pools.

Alteration of global climate could result in a 20% to 60% change in vegetation cover type of terrestrial
ecosystems.  The responses of world forests to rapid climate  change will vary by latitude.  In the boreal
region, where temperatures are projected to warm the most, forest boundaries are expected to migrate
toward the poles.  Within the temperate latitudes, warmer temperatures are expected to cause drier
conditions for some regions, leading to forest declines.  Migrations are likely to be slower than declines
in both regions, thereby decreasing forest distributions.  Closer to the equator, projections call  for
increases in precipitation, which would favor the productivity of moist tropical forests.

Numerous additional changes in forests are projected.  For instance, forest species other than trees,
both animal  and plant, may experience stress while transitioning to changed climates. Some species are
likely to disappear altogether, thereby reducing biological diversity. Other significant changes are related
to increases in atmospheric CO2.  In addition to contributing to climate change, elevated CO2 could
produce a fertilizer effect in forests. That is, photosynthesis is favored  by increased C02, and forest
growth and productivity might increase where moisture is not  limiting.  At the same time,  high levels of
C02 are known to improve water-use efficiency in plants, and  forests in drier areas could benefit from
this effect.

Changes in global vegetation could cause terrestrial carbon storage to increase (using climate scenarios
generated by the  GISS and OSU Global Circulation Models, or GCMs)  or decrease (UKMO and GFDL
GCMs).  Carbon cycling in changing forest ecosystems may shift and thereby alter the size of above-
                                               xi

-------
and belowground carbon pools.  Where forest regions are decreasing in size and vegetation is in
transition to different environments, carbon could be emitted to the atmosphere, creating an undesirable
positive feedback.  Transient response of vegetation could increase atmospheric carbon by 0-3 gigatons
(106 metric tons) annually. This is an area of great uncertainty, and much more needs to be learned for
a clear understanding of the potential feedbacks.

A possible mitigation technique for slowing the increase of CO2 in the atmosphere is increasing the
amount of carbon stored in the terrestrial biosphere. Forests and other terrestrial components of the
biosphere, such as agroecosystems, can be managed to sequester carbon and contribute to mitigation
efforts.  Of the many possible approaches, reforestation to create new productive forests and expansion
of agroforestry systems offer the greatest potential to fix atmospheric C02. However, sociopolitical
factors, investment capital, and effective  technology currently limit all approaches to managing the
carbon cycle.
                                               XII

-------
      RESPONSE AND FEEDBACKS OF FOREST SYSTEMS TO GLOBAL CLIMATE CHANGE

1  INTRODUCTION

A third of the world's lands are forested and people in all nations benefit by the contributions of forests
to basic necessities (fuel, food, fiber) and to a functional terrestrial biosphere.  Forests are not static,
however. Through geologic time, for instance, natural changes in climate have greatly altered the extent
and composition of forests throughout the world. The reverse is also true: forests can influence climate.
For example, transpired water vapor and natural gas emissions from forests act as feedbacks that affect
climate.

In recent millennia, anthropogenic activities have rivaled climate as an influence on forests. That is,
throughout the world, humankind has used, abused, and, in some cases, sustainably managed forests in
all major terrestrial regions.  At the same time, human activities have grown  in the last century to the
point that they are now suspected of altering the atmosphere through emissions of waste chemicals and
paniculate matter. These changes in the atmosphere are projected to cause dramatic global warming in
the next century (Intergovernmental Panel  on  Climate Change [IPCC] 1990).

Thus, in the climate-forest-human triad of today, each component can influence the condition of the
other two.  How this situation will play out in the coming decades is uncertain.  What is the potential risk
to the terrestrial biosphere, particularly to humankind, and can measures such as forest management
offset the suspected adversities of rapid global change?

Investigating these questions has become a principal research initiative  at the US Environmental
Protection Agency's (EPA) Environmental Research Laboratory-Corvallis (ERL-C). In the last two years,
ERL-C and other research groups have begun to address the response  and feedbacks of the terrestrial
biosphere to global change.  New research approaches have produced insights and projections
regarding global change and forest systems.  This report is a current summary of scientific hypotheses
and evidence regarding projected global change, as well as of new research on potential effects of
climate change on terrestrial carbon pools.

-------
1.1     Subject

Climate simulations using General Circulation Models (GCMs) suggest that the production of greenhouse
gases (e.g., carbon dioxide, methane, chlorofluorocarbons, and nitrogen oxides) from anthropogenic
activities could cause large changes in global climate. This results when atmospheric increases in these
gases in trapping heat that is radiated outward from the earth's surface.  By the year 2020, global mean
temperature is predicted to rise 1.8<€ above preindustrial levels, with a probable increase between 1.3
and 2.5
-------
growth and water-use efficiency in plants growing in a CO^enriched environment (Eamus and Jarvis

1989, Strain and Cure 1985).  Warm temperature, in the absence of moisture stress, has been shown to

increase growth and  maintenance respiration of some plants (Berry and Bjorkman 1980, Drake et al.

submitted). Long-term responses of vegetation, especially forests, to the combined effect of global

warming and  CO2 enrichment are uncertain (Mooney et al. in press).
                                    Atmosphere
                                  740Gt(in 1988)
                                   +3 Gt per year
         1 10GI
      Photosynthesis
  93 Gt
Biological
    5Gt
Fossil Fuel Use
    1-2 Gt
Deforestation
                 55 Gt
               Respiration
Chemical
Processes
                            90 Gt
                          Biological
                                                                         emical
                                                                      Processes
                                       Fossil Fuels
                                     5,000-10,00061
           Soil, Litter, Peat
            1170-1 740Gt
             Ocean
            38.500 Gt
Figure 1.1-1.   The  global carbon  cycle,  including  major pools  and annual flux of carbon  (adapted
               from Schneider  1989b).

-------
Global climate change is predicted to significantly Influence the condition, distribution, and migration of
forests (Smith and Tirpak 1989, Neilson et al. 1989).  Estimates of impacts on future forest condition,
distribution, and productivity are based upon two types of models, GCMs and global vegetation models
(Box 1981, Emanuel 1985, Emanuel et al. 1985, Prentice and Fung in press). GCMs are three-
dimensional representations of the earth's atmosphere, and, although extremely complex, they produce
only very rough estimates of climate change for the next few centuries. About six models are used
widely, and their climate predictions vary substantially.  Global vegetation models also vary widely in
their predictions.

In summary, our knowledge of global change impacts and feedbacks on forest ecosystem-level carbon
cycling is meager but  growing rapidly.  This report presents new results based on recent work at ERL-C
on the potential changes in terrestrial carbon storage caused by vegetation redistribution (see Section 3)
and a summary of the relevant scientific literature. A number of new research activities were initiated in
FY90  (Appendix A). Emphasis is placed on results obtained since the 1989 EPA Report to Congress on
the effects  of climate change on resources of the United States (Smith and Tirpak 1989), and on topics
not discussed in that report.

12.    Scope

This report discusses the potential ecological responses and feedbacks of forest systems to a changing
global climate.  Forests will be considered on a global scale, including the boreal, temperate, and
tropical forest regions.  Specific topics to be discussed are:

1)     The current knowledge of the effects of global climate  change on forests (Section 2).

2)     Potential changes in terrestrial carbon storage as vegetation responds to global climate change
       (Section 3).

3)     Estimates of the effect of increases in atmospheric C02on forest ecophysiological processes
       that influence carbon assimilation and cycling, especially water-use efficiency, photosynthesis,
       and respiration (Section 4).

4)     The role of forest management, especially large-scale reforestation, in sequestering atmospheric
       C02 (Section 5).

-------
5)     Management of the biosphere as an aid in mitigating global climate change (Section 5.3).

6)     Key research needs leading from uncertainties about the role of the terrestrial biosphere
       (Section 6).

13     EPA's 1989 Report to Congress

In the late 1980s, Congress requested that the EPA undertake two studies on 'climate change due to the
greenhouse effect." One study was to be on the potential effects of climate change on United States
agriculture, forests, human health, water systems, and other vital resources (Smith and Tirpak 1989).
The second study was to focus on "policy options to stabilize current levels of atmospheric greenhouse
gas concentrations" (Lashof and Tirpak 1989).

In EPA's  1989 Report to Congress on global  climate change, the Forest Effects chapter presented
results of studies and  projections after two years of analyses conducted by the Global Climate Team at
ERL-C (Winjum and Neilson 1989). Principal findings  and limitations reported at that time are reviewed
in this section.  This document builds on the  earlier information and, in addition, it includes the following
considerations about forests:  1) feedbacks to the atmosphere; 2) C02 mitigation potential; and 3) role in
global ecology.

1.3.1  Findings

1)     Global warming could significantly impact forests of the United States, with the effects apparent
       in 30 to 80 years.

2)     The potential northern range  of forest species in the eastern United States could shift northward
       by as much as 600 to 700 km over the next century.  Actual northward migration could be
       limited to as little as 100 km because of slow  natural rates of species migrations.

3)     The southern ranges of many eastern United States tree species could die back by as much as
       1000 km because of higher temperatures and drier soils. Together with the projections of the
       previous finding, distributions for some tree species would  significantly decrease. Overall, forest
       productivity in the United  States is likely to decline, especially along the southern edges of
       present forest regions.

-------
4)      Additional impacts of climate change on forests could include increases in wildfire, insect and
        disease outbreaks, wind and air pollution damage, and soil erosion. A reduction in biodiversity
        is also  projected (Barker et al. 1990).

5)      Regarding forest policy, institutions such as the United States Department of Agriculture's
        (USDA) Forest Service (USFS), state forest agencies, and private forest companies should
        consider climate change impacts in their long-term planning. A coordinated  public and private
        reforestation effort, together with development of new and adapted silvicultural practices, may
        also be required.

1.3.2  Uncertainties

1)      Under conditions of  rapid climate change, the response of entire forest  ecosystems is very
        difficult to project. The 1989 studies were based upon available knowledge of major tree
        species, largely for eastern United States forests.  Ecosystems, however, are composed of a
        myriad  of both plant and animal species, each likely to have its own characteristic response.
        Ecosystem response, therefore,  remains highly uncertain.

2)      The mathematical simulation models generated in the 1980s to simulate forest response to
        climate change did not include dispersal rates for the tree species studied, and therefore
        projections of forest migrations were only crudely estimated. Furthermore, the potential of
        aiding migration through reforestation practices was not analyzed in detail.

3)      Forest declines are often triggered by periods of high environmental stress. The forest models
        used for the studies  in the 1989  Report to Congress were not run far beyond current conditions,
        such as for extremely dry soils.  Therefore, the model projections may not estimate the timing
        and behavior of forest declines under future climate conditions as closely as  desired.

4)      The response of mature trees or forests to elevated atmospheric CO2was not evaluated in the
        study for  EPA's 1989 Report to Congress.  At that time, few C02 response data were available
        for forest tree species, and the available information was for seedlings exposed to elevated CO2
        for a few growing seasons.  This factor led to a large uncertainty in the early study results
        regarding response by forest stands whether  under the present climate or with rapidly changing
        climate (and the uncertainty remains, as noted in Section 1.1).

-------
5)     The 1989 Report to Congress focused exclusively on the United States.  Effects of global climate
       change on world forests are uncertain.

1.4     Background

Three important  background topics for the report are: 1) a brief description of the extent of world
forests; 2) the role of forests in the global carbon cycle; and 3) the climate change scenarios that will
affect both forests and the global carbon cycle. Though presented as background information,
knowledge about each topic is incomplete primarily because their study on a global  scale is relatively
new.

1.4.1  World Forests

Forests were estimated in 1985 to cover 4139 million ha in the world, or about 31% of the total land area
(Sedjo and Lyon 1990) (see Table 1.4-1). Another  1030 million ha, or 8%, is  classified as shrubland and
forest fallow (i.e., cleared, not fully reforested, but with scattered trees; Postel and Heise 1988) (see
Table 1.4-2).  The total by this classification, therefore, is about 38% of the world's land area as of 1985.
Matthews (1983) estimated that in pre-agricultural history, world forests were about 15% greater in extent
than they are today.

-------
Table 1.4-1    World Forested  Area by Region,  circa 1985 (from Sedjo and Lyon 1990)
Closed Total
Region
North America
Central America
South America
Africa
Europe
Soviet Union
Asia
Pacific Area
(Oceania)
World
Forestland1
734
65
730
800
160
930
530

190
4,139
Closed
forest
(million
459
60
530
190
148
792
400

80
2,659
Open
woodland
ha)
275
2
150
570
12
138
60

105
1,200
Total
land area
1.829
272
1,760
2,970
472
2,240
2,700

842
13,105
forest forest
(percentage
of land area)
25
22
30
6
31
35
15

10
20
40
24
41
27
34
42
20

23
31
       Forestland is not always the sum of closed forest plus open woodland,
       as it includes scrub and brushland areas that are neither forest nor open
       woodland as well as deforested areas where forest regeneration is not taking
       place.  In computation of total land area, Antarctica, Greenland, and Svalbard
       are not included;  19% of all Arctic regions are included.

       Source: Economic Commission for Europe/Food and Agriculture Organization of
       the United Nations (ECE/FAO), The Forest Resources of the ECE Region (Geneva,
       ECE/FAO 1985).
Table 1.4-2   Forest Fallow and Shrubland, circa  1980 (Postel and Heise 1988)


               Region
  Area
(million ha)
       Asia (except China)             107
       Africa                          608
       Central and South America      313
       Oceania                         2
       Total
  1,030

-------
Forests are a significant component of the landscape on every continent and large Island group in the
world except Antarctica (Young 1982). They have historically contributed to human economic and social
progress by providing such resources as shelter, fuel, food, water, recreation, and many commercial
products (Perlin 1989). Some projections for global climate change suggest that the ability of forests to
sequester (i.e., to capture and hold) vast quantities of carbon is yet another role of vital importance to
humankind.  In contrast, predictions from some global vegetation models suggest that forests could
become major sources of carbon that might stimulate global climate change (Leemans 1990, Smith et al.
submitted).  Many forest classification systems appear in the literature (Young 1982). Often forests are
classified as commercial and noncommercial.  For example, the USFS calls forests commercial if they
can grow 1.4 m3/ha/yr of wood that is of sufficient quality to justify timber harvests. Both commercial
and noncommercial forests can be either coniferous, broadleaved, or a mixture (Young 1982).
Noncommercial forests, though less productive in commercial wood  yields than commercial forests,
nevertheless have significant roles in providing resources such as food, water recreation and fuel wood,
and sustaining biological diversity  (Reid and Miller 1989).

Typically, ecological classifications are by tree density or crown cover. Closed forests are those in
which the tree canopy cover is about 20% or more of the land surface (Westoby 1989).  These forests
grow where the annual precipitation is at least 400 mm.  In drier areas, forests have more scattered trees
so the canopy cover is less; where canopy cover is only about 5-20%, the plant communities are called
woodlands (Young 1982).

Closed forests of the world represent about two-thirds of the forest land area and woodlands make up
the other third (28.3 and 13.2 million ha,  respectively; see Table 1.4-1). Closed forests are commonly
classified as follows (Young 1982):

       1)  Mainly evergreen forests:
               a. broadleaved evergreen forests (primarily in warmer climates; e.g., moist tropical
               forests);
               b. coniferous forests (primarily in cooler climates; e.g., boreal and
               temperate forests);

       2)  Mainly deciduous forests:
               a. drought deciduous forests (leaves shed in dry season; e.g., dry tropical forests);
               b. cold deciduous forests (leaves shed in winter; e.g., temperate hardwood forests).

-------
1.4.2 Global Carbon Cycle

The global carbon cycle links significant biogeochemical processes that affect the flux (movement) of
carbon between the atmosphere, terrestrial biosphere, and ocean pools.  Although the general outline of
the cycle is well known, large uncertainties still exist in estimates of the magnitude of carbon pools and
fluxes.

In  general, the oceans store by far the largest fraction of carbon on the globe (38,500 Gt), followed by
fossil fuels in the earth's crust (i.e., coal and oil; 5,000-10,000 Gt)(see Figure 1.1-1, Bolin et al. 1979,
Schneider 1989a). For terrestrial ecosystems, estimates of aboveground carbon storage are highly
uncertain, but they range from 560 to 830 Gt (Whittaker and Likens 1975, Schlesinger 1984, Post et al.
1979, Ajtay et al. 1979,  Olson et al. 1983, Mooney et al. 1987. Schneider 1989a). The earth's
atmosphere is estimated to have 740 Gt of carbon, about the same amount as terrestrial vegetation.
Terrestrial soils, however, contain between 1.5 to 3 times as much carbon (1170 to 1740 Gt) as either
terrestrial vegetation or the atmosphere.

Even though the terrestrial biosphere stores much less carbon than the amount stored in oceans, annual
fluxes between the atmosphere and terrestrial biosphere are approximately equal to those between the
atmosphere and oceans (excluding anthropogenic  influences) (Figure 1.1-1).  On an areal basis,
therefore, terrestrial fluxes are much  greater than oceanic fluxes.  The total annual contribution to the
atmosphere of carbon from the terrestrial biosphere and oceans is about 30% of that stored in the
atmosphere, with  an annual contribution back to these systems of about the same order of magnitude
(Bolin et al. 1979, Schneider 1989b).

These characteristics of terrestrial biosphere fluxes imply that changes in land use and land management
practices that might impact carbon fluxes (e.g., deforestation, reforestation, conservation of soil carbon)
could have significant impacts on atmospheric C02 concentrations (as discussed in Section 5). That the
terrestrial biosphere can affect atmospheric concentrations of C02 is best illustrated by the seasonal
changes in atmospheric C02 (Figure 1.1-2) (Keeling et al. 1989, Tucker et al.  1986), which are caused by
seasonal changes in the relative magnitude of photosynthesis and respiration in the biosphere  (Mooney
etal. 1987, Kingetal. 1987).

Of concern from the perspective of possible global climate change is that the fluxes into and out of the
atmosphere are not balanced, and the atmosphere is gaining about 3 Gt of carbon per year (Schneider
1989b, Tans et al. 1990).  The two principal sources of the additional atmospheric C02 are fossil fuel
                                               10

-------
combustion (Marland et al. 1989, Schneider 1989b) and deforestation (Woodwell et al. 1983, Palm et al.
1986, Houghton et al. 1983. Houghton et al. 1985, Houghton et al. 1987, Detwiler and Hall 1988).  The
contributions of these two carbon sources are about equal according to Keeling (1989), and the
combined total (6-7 Gt) is almost double the 3 Gt of carbon accumulating in the atmosphere each year.
Moreover, the contributions from the terrestrial biosphere (Keeling et al. 1989) and fossil fuel combustion
are increasing with time.

One of the critical unknowns in balancing the carbon budget is determining where the carbon released
to the atmosphere from deforestation and fossil fuel combustion is stored, since only about half
accumulates In the atmosphere.  The only two possibilities are the oceans and the terrestrial biosphere.
A recent study suggests that the terrestrial  biosphere (primarily at temperate latitudes) is a greater sink
for the remaining carbon than previously thought (Tans et al. 1990).  Increased productivity of vegetation
caused  by carbon fertilization (see Section 4.1) could be  affecting the amount of carbon being stored  in
the terrestrial biosphere (Keeling et al. 1989).
Figure 1.1-2.   Latitudinal variations in atmospheric  CO2concentrations  from 1982-1984  (adapted
               from Tucker et al. 1986)
                                                11

-------
The other large uncertainty in balancing the carbon budget is the amount of carbon being released to
the atmosphere through deforestation. Current estimates are about 3 Gt per year, but these estimates
vary by a factor of 2-3 because of uncertainties about rates of deforestation, the amount of carbon in
cleared forests, decomposition rates, and the amount of regrowth after deforestation (Dixon in press, c,
Woodwell et al. 1983. Houghton et al. 1987, Detwiler and Hall 1988).  Clearly, more research Is needed
to measure terrestrial carbon fluxes and to  Improve the understanding of the processes driving
atmosphere-biosphere carbon exchanges (see recent papers by Tans et al. (1990) and Keeling et al.
(1989) for a discussion of uncertainties in balancing the carbon budget and research needs).

To conclude, recent research has clearly indicated that the terrestrial biosphere is an important
component of the global carbon cycle, particularly as it affects fluxes to and from the atmospheric
carbon pool.  Future changes in the terrestrial biosphere caused by climate change and human activities
can be expected to further influence atmospheric carbon pools and thus global climate.

1.4.3  Climate Scenarios

To determine the possible impacts of  global climate change on the biosphere and the  resulting
feedbacks to the climate system, quantitative  estimates are needed of the magnitude of climate change
induced by trace gases. GCMs  of the earth's atmosphere are the only tool available for making
quantitative estimates on a global scale of climate variables such as temperature and precipitation.
GCMs are complex computer models based on fundamental  principles of physics and thermodynamics
(e.g.,  Schlesinger 1988). In the past decade,  several  GCMs have been used to simulate the possible
effect on global climate of a doubling  of CC-2 concentrations.  Results of simulations of four different
GCMs (listed in Table 1.4-3) are  widely available for use by researchers and were used in this report to
obtain the data in Sections 2 and 3.  Some of the overall conclusions about future climate change and
uncertainties  in the GCMs are discussed in this section, as well as the general  methodology for creating
the future climate scenarios used in this report.
                                               12

-------
Table 1.4-3.    General Circulation Models  Used to Generate  Double-CO 2 Climate Scenarios
Model Name                                Reference
Geophysical Fluid Dynamics Model (GFDL)     Manabe and Wetherald
                                           1987
Goddard Institute for Space Studies (GISS)     Hansen et al.
                                           1988
Oregon State University (OSU)                Schlesinger and Zhao
                                           1989
United Kingdom Meteorological Office (UKMO) Mitchell et al.
                                           1989
The most recent results from the GCMs suggest that double-CO2 conditions, or the radiative equivalent,
will increase global temperatures by approximately 1.9 to 4.4
-------
highly uncertain. Research efforts are underway to improve GCMs (the Department of Energy's ARM
and CHAAMP programs) and to determine how well they simulate past climates as a validation
technique (Cooperative Holocene Mapping Project 1988).

Despite the limitations of GCM outputs, especially their poor spatial resolution, they remain the only tool
that can provide spatially distributed estimates of future climates based on the principles of atmospheric
physics. The challenge for researchers estimating the effects of climate change on components of the
biosphere is to use GCM output to generate reasonable scenarios of future climate for input into their
models.  The most common technique is to calculate the difference between (or for precipitation, the
ratio of) the double-C02 estimate for a particular gridpoint and the control estimate (current conditions)
for that  same gridpoint (Parry et al. 1987, Smith and Tirpak 1989, ICF 1989). These differences are then
added to (or multiplied by, in the case of ratios) the corresponding historical weather data, often the
mean value for the 1951-1980 time period. For instance, if a GCM estimates that July temperatures are
2
-------
2      EFFECTS OF GLOBAL CLIMATE CHANGE ON FORESTS

Forests historically have responded to changes in climate (Webb 1987), leaving little doubt that rapid
climate change will affect forests in the future. Research to date has focused on the response of certain
tree species and forest stands to climate change (e.g., Smith and Tirpak 1989). In this section, recent
results concerning the potential effects of climate change on treeline, forest stands, biomes, and
biodiversity are presented.  In addition, a discussion of the effects of climate on belowground processes
in forests from a soil carbon perspective is presented.  Much of the information presented in this section
forms a basis for the later analyses discussed in Sections 3 and 5.

2.1  Treeline Response

Altitudinal and latitudinal treelines are a striking and sensitive example of climate control of species
distributions (Garfinkel and  Brubaker 1980, Goldstein 1981, Goldstein et al. 1985, Edwards et al. 1985,
Cooper 1986, Odasz 1983).  Consequently, movements in treeline could be an early indicator of climate
change impacts.  Studying  and monitoring the northern treeline is especially important since the GCM
simulations of double-C02 climate conditions show greatest warming at high latitudes. Recently, a
global network  of researchers in Scandinavia, the Soviet Union, the United States, and Canada was
organized to evaluate the taiga and tundra boundary response to global change (Solomon in press).
The purpose of this section is to review our current understanding of the factors controlling treeline and
to discuss possible impacts of future climate change on treeline.

2.1.1  Ecophysiological Characteristics of Treeline

Tranquillini (1979), Arno (1984), and Larsen (1989) discuss treeline characteristics ranging from general
features to specific ecophysiological processes of treeline environments.  Tranquillini (1979) emphasized
the synergism of three closely related  factors controlling alpine treeline: 1) limited dry matter production;
2) incomplete tissue maturation; and 3) inadequate climate resistance.  Arno (1984) present physical and
climatic factors that affect treeline position.  Larsen (1989) discusses biotic communities and ecological
relationships of treeline in Canada and Alaska.

Climate at high-latitude and high-altitude treelines is characterized by interactions between a short
growing season, low air temperatures, frozen soils, drought stress, high levels of solar radiation, irregular
snow accumulations and frequent strong winds  (Ives and Hansen-Bristow  1983, Hansen-Bristow and  Ives
                                                15

-------
1984, 1985).  An elevational gradient (low to high) in the forest-alpine tundra ecotone is typified by a
gradual decrease in tree stature, an increase in tree deformation, and a decrease in tree stand densities
(Hansen-Bristow 1986).  Geographically, the transition from subalpine continuous forest to alpine tundra
occurs over a relatively short distance (< 2 km).

Sufficient summer heat may be the most important individual factor limiting the elevation or latitude to
which trees can migrate.  Treeline position ultimately depends on the increasingly unfavorable heat
balance associated with rising elevation or latitude.  Germination, growth, photosynthesis, and other
biological processes have varying minimum temperature requirements and optimums (Tranquillini 1979).
The hardiest of conifers require a growing season of two months in which no more than light frosts
occur.  Treelines have been correlated with the location of the 10oC  isotherm for the warmest month,
usually July in the northern hemisphere.  This correlation approximately holds for cold treelines
throughout North America and Eurasia north  of the tropics (Pearson 1931, LaMarche and Mooney 1967,
Arno 1984). Thornthwaite's thermal efficiency index (Hare 1950) and Kira's warmth index (Kira 1965)
have also been correlated with treeline position.

2.1.2 Treeline Response to Climate Change

An  increase in temperature would  shift the IOC isotherm northward  in latitude and higher in elevation.
High-latitude and  high-elevation regions that were previously limited  by growing season might then be
able to support tree establishment and growth.  One mechanism for this shift is a change in reproductive
success.  For example, treeline on the south  slope of the Brooks Range in Alaska and  isolated clusters
of balsam poplar (Populus balsamifera) trees on the north slope commonly reproduce  by vegetative
means only (Edwards and Dunwiddie 1985, Lev 1987).  A temperature increase may allow sufficient time
for flowering to occur and for sexual reproduction to take place.  Migration rates could increase
substantially given that seed  dispersal will occur at greater distances than branch or root vegetative
propagation.

These qualitative considerations are supported  by simulations of the response of global vegetation to
future climate change. The simulations show a significant northward movement of the  deciduous boreal
forest and associated treeline in response to  global warming (Emanuel 1985,  Emanuel et al.  1985,
Leemans 1990, Smith et al. submitted. Prentice and Fung in press).
                                               16

-------
2.1.3  Research Needs

Treeline response to climate depends on unique local site characteristics.  The challenge will be to
identify the most influential and limiting factors across latitude and elevation, to determine how climate
will influence these factors, and to project treeline response.  There is evidence of climate moderation in
the Arctic and this moderation is predicted to continue; thus, studying treeline response at high latitudes
may allow a linkage between tree growth and migration rates and climate.  The following research areas
are suggested as providing the most direct link between treeline and climate:

1)      Detection and linkage  of treeline growth and migration rates to climates at high latitudes and
        altitudes.

2)      Development of mechanistic models of tree growth in response to limiting climate conditions at
        high latitude and altitude treelines.

22  Forest Stand Effects

Stands1 of trees and the ecosystems of which they are a part are important for social and economic
reasons and for their ability to capture and retain carbon for extended periods of time (Harmon et al.
1990). It is at the stand level that effects of climate change will be most strongly felt by society and at
which adaptation and mitigation efforts will be focused (see Section 5).

Considering possible effects of climate change on the  composition of forest stands in the United States,
Smith and Tirpak (1989) summarized modeling work completed in the Great Lakes region and
southeastern United States.  Dixon (in press,c) reviewed forest management planning options to cope
with global change in the southern United States.

Solomon (1986) presented simulation model results for the transient response of forests in eastern North
    1  The term "forest stand" has a clear definition in forest science. Smith (1986) states: "A stand is a
contiguous group of trees sufficiently uniform in species composition, arrangement of age classes, and
condition to be a distinguishable unit.  The internal structure of stands varies mainly with respect to the
degree that different species and age classes are intermingled. The simplest kind of structure is exemplified
by that of the pure, even-aged plantation consisting of trees of a single species.  The range of complexity
can extend to a wide variety of combinations of age classes and species in various vertical and horizontal
arrangements."
                                                17

-------
America to COrinduced climate change. A distinctive dieback of extant trees was noted at most
locations.  The pattern of dieback and recovery varied across the region and by species.  Transient
responses to species composition and biomass continued for centuries after climate had stabilized.

Pastor and Post (1988) presented simulation model results for the response of northern forests to CO2-
induced climate change. The greatest changes were shown at the boreal-cool temperate forest border.
The linkages between species composition, productivity, biomass, carbon and nitrogen cycles, and soil
moisture were discussed. Forest productivity and biomass were directly linked to soil moisture levels.
Species composition altered soil nitrogen availability which then influenced species composition. The
positive feedback of the carbon and nitrogen cycles was bounded by soil moisture availability and
temperature.

In this section, preliminary results are presented from a study designed to estimate the impact of climate
change on forests in the Pacific Northwest. The objectives of newly initiated EPA research on the effects
of climate on United States forests are summarized.  Finally, the geographic scope of EPA research is
broadened to discuss forest models that could be used to quantify the  effects of climate change on
boreal forests.

2.2.1  Pacific  Northwest  Forests

Preliminary modeling work on the potential response of Pacific Northwest forests to climate change was
initiated as a joint research project between the EPA, the USFS Pacific  Northwest Forest and Range
Experiment Station, Oregon State University, and the University of Virginia. The model ZELIG (Urban
and Shugart 1989) has been parameterized for the Pacific Northwest species temperature gradients and
distributions2. The GCM scenarios for a double-CO2 environment predict a warmer and wetter climate
for the Pacific Northwest. Preliminary ZELIG model  results for a period of 500 years show an upward
shift of 500 to 1000 m in species distribution. For example, high-elevation forest sites would warm to
resemble present-day mid- to low-elevation sites, and so on.  If GCM predictions are reasonably
accurate, substantial changes in forest composition can be expected in the Pacific Northwest in the next
century.
    2  ZELIG is an updated version of a class of models (known as forest gap models) that simulate tree
establishment, diameter growth, and mortality on a yearly basis on a plot of defined size (Botkin et al. 1972,
Shugart 1984).
                                               18

-------
2.2.2  Future Modeling of United States Forests

A joint EPA and USFS Rocky Mountain Forest and Range Experiment Station companion modeling
project has been initiated to simulate the effects of climate change on Rocky Mountain tree species.
The work will initially use the model FIRESUM (Keane et al. 1989), an ecological process model for fire
succession In western conifer forests.  The model Is currently being calibrated for elevational and climate
gradients and tree species in the Rocky Mountains. Preliminary model simulations predict that high-
elevation sites will warm and that species composition and distribution will be altered.

The EPA is supporting work at Utah State University-Logan to develop a spatially-explicit regional scale
forest simulation model with which to predict the effects of climate change on Rocky Mountain forest
composition, structure, and function. At  the stand level, the model will simulate the effects of climate
change on reproduction, leaf area dynamics, water use, carbon/nitrogen dynamics j.-^d fire regime
across a range of environments.  The individual stand components will be simulated in a spatial context
that allows predictions of species migration and forest fragmentation for individual regions. A set of
climate change scenarios will be developed based on a regional interpretation of GCM predictions.
These climate scenarios will then serve as input to the regional simulation model to predict a range of
possible regional responses, each specific to a given climate scenario.

The EPA is also supporting work at the University of Florida-Gainesville to model carbon dynamics of
slash pine plantations in response to climate change. The objectives are to incorporate measured  slash
pine physiological responses and climate change scenarios into a labile carbon model (Cropper 1987,
Cropper and Ewel 1983, 1984,  1987).  The  research will result in simulations of the physiological and
stand-level responses of slash pine plantations to potential climate change scenarios.  The relative
importance of respiration responses, leaf area dynamics, stand water balance, and assimilation will be
assessed with the simulations.

The EPA is supporting work at the University of Montana-Missoula to couple the FOREST-BGC
(BioGeoChemical Cycles; Running and Coughlan 1988) and FORET  (Shugart 1984) ecosystem
simulation models for projection of  forest responses to global climate change. The combined  model
simulations will range from daily temporal scale photosynthesis-respiration balance to annual primary
production to species-specific stand development over centuries. Final products are expected to include
the simulation of forest production,  succession, growth, and development for a western coniferous forest
and an eastern deciduous forest over a 200-year period. Simulations will first be done for current
climate conditions over a range of sites and disturbance conditions,  and then with either a transient or a
                                               19

-------
fixed double-CO2 climate change scenario. Shifts in primary production, recruitment-mortality, and
species migration will be tracked.  Relative differences in response of western coniferous and eastern
deciduous forests will also be emphasized.

Prentice et al. (1989) describe the development of a global vegetation dynamics model.  The modeling
approach is to generate vegetation dynamics on a small patch for several climate scenarios. The
scenario projections represent a range of spatial distributions of environmental variables among patches
within landscapes.  Mean values of environmental variables can then be interpolated using existing
global environmental databases.  The modeling approach allows a linkage between small-scale
vegetation process and global scale atmospheric processes, making full use of existing environmental
data.  The range of  scales also allows natural and anthropogenic disturbances to be incorporated into
predictions.

2.2.3  Boreal Forests

Boreal forests are important components of the terrestrial biosphere, and their feedbacks and responses
to global changes will be significant (Dixon in press,a).  Bonan and Shugart (1989) present a unifying
model that links the structure, function, and pattern of boreal forest vegetation with environmental factors
that interact and account for the pattern of boreal forest types and productivity: climate,  solar radiation,
soil moisture and the presence of permafrost, the forest floor organic layer, nutrient availability, insect
outbreaks, and wildfires.  Bonan (1989) discusses a computer model to predict regional  patterns of soil
radiation,  soil moisture, and soil thermal regimes in boreal forests in North America, Scandinavia, and
the Soviet Union. Environmental  factors are calculated on a monthly  time step, a temporal scale suitable
for use in models of forest succession. The ecological consequences of climate change for
environmental factors can then be explored.

Two gap models based on individual trees have been developed for the boreal forest region: LOKI for
North American boreal forests (Bonan 1988), and FORSKA for the Scandinavian boreal forests (Leemans
and Prentice 1989).   A general boreal forest simulator has been developed by combining FORSKA and
LOKI at the University of Virginia in collaboration with Bonan and Leemans.  The EPA is  supporting work
to use this simulator to generate baseline boreal forest dynamics as well as transient departures under
climate change scenarios.
                                               20

-------
2.2.4  Research Needs

Research should link detection of tree and stand response (extensive research sites) to mechanisms of
tree and stand response (intensive research sites) and to climate change disturbance.  Once
mechanisms of forest response are identified and modeled, climate change scenarios from GCMs can
be used to predict vegetation response over transient and equilibrium conditions.  Research projects
currently funded by the EPA and the following suggested research will help make the linkages necessary
to achieve these goals:

o      A pole-to-pole transect of extensive and intensive research sites at ecotonal boundaries,
       especially high-altitude areas, to provide local, regional,  and global data on species interaction
       and on factors limiting species distribution at the stand level for a wide range of species.

23     Carbon Sequestration  in Soils

Globally, soils contain the largest non-fossil, terrestrial  reservoir of carbon (see Figure 1.1-1). The soil
reservoir is estimated to contain between 1170 and 1740 Gt of organic carbon (Bolin 1983, Schlesinger
1984, Houghton and Woodwell 1989, and Schneider 1989b); that may be three times the amount of
carbon stored in vegetation and twice the amount of carbon currently in the atmosphere.

Soils have been often overlooked as a component of the global  carbon cycle. Soils, as a carbon pool,
were not considered in 1989 EPA Report to Congress  (Smith and Tirpak 1989). A number of studies
have shown that the clearing of forests for agriculture has resulted in declines in soil organic matter
(Giddens 1957, Paul  1976, Mann 1986). The amount and rate of decline in soil organic matter are a
function of a number of factors other than forest clearing per se (Allen 1985, Oades 1988).  However,
such declines demonstrate the lability of soil carbon and have raised the question of how losses in
organic matter relate to C02 releases to the atmosphere. The role of soils in relation to global change is
receiving increased attention not only because of the size of the soil carbon pool that could become a
net source of atmospheric CO^ but because soils also represent a  potential sink for carbon (Greenland
and Nye 1959, Jenkinson and Rayner 1977, Armentano and Ralston 1980, Lugo and Brown 1986).

The purpose of this section is to: 1) summarize the importance of forest soils in the global carbon cycle;
2) present a heuristic model of soil carbon  dynamics; 3) identify and discuss factors that affect the size
of the soil carbon pool; and 4) assess the potential effects of increasing CO2 and climate change on this
                                               21

-------
reservoir of carbon. In Section 5.2, the management of soils for the purpose of storing carbon is
reviewed and discussed.

2.3.1  Forest Soils and the Global Carbon Cycle

Terrestrial photosynthesis removes approximately 110 Gt of carbon from the atmosphere annually, only a
portion of which is retained in the annual increment of aboveground biomass (Oades 1988, Schneider
!989b).  Nearly half of this carbon is returned to the atmosphere through direct plant respiration, and the
other half is deposited on or in the soil. While this carbon is being incorporated into the soil, an
equivalent amount is returned to the atmosphere from the soil via oxidation of organic matter to COz At
steady state, the amount of carbon fixed (110 Gt) is equal to the amount released (110 Gt) by respiration
and decomposition (Figures 1.1-1 and 2.3-1).

In an undisturbed world, climate and soils are the principal factors controlling the distribution and
productivity of forests.  These two variables also control the amount of carbon stored in forest
ecosystems.  Boreal, temperate, and tropical forests cover approximately one-third of the earth's land
surface (Table 2.3-1, Waring and Schlesinger 1985). Combined, these three forest biomes contain
approximately 743 Gt  of carbon, or 90% of the aboveground carbon stored in biomass.  Belowground,
they contain approximately 576 Gt of carbon, or about 40% of the carbon sequestered in soils.

Allometric relationships between aboveground  and belowground carbon productivity in forests have
been proposed (Newbould  1967), but are not universally accepted (Bo hm 1979). Rather, the observed
relationships are often a function of environmental conditions. For example, the data in Table 3.1-2
suggest that total belowground carbon pools are at least equal to total aboveground carbon pools and,
in some cases, are 40-fold greater than aboveground carbon pools.

Because of warmer climates (given similar moisture regimes), tropical forests are more productive than
temperate and boreal  forests. At the same time, they have greater decomposition rates and less
belowground relative to aboveground storage (Table 3.1-2). In boreal systems the converse is true.
Temperate forests are intermediate.  Additionally, tropical forest soils have low carbon:nitrogen (C:N)
ratios, whereas boreal forest soils have high C:N ratios (Post et al. 1985).  The C:N ratio is a good
indicator of the extent of decomposition (low ratios indicate more readily decomposable materials).
                                               22

-------
Table 23-1.  Biomass and Soil Carbon in Boreal, Tropical,  and Temperate Forest Ecosystems
       (from Waring and Schlesinger  1985)
Forest
Ecosystem
Tropical
Temperate
Boreal
Total
World Totals
Area
(106km^
24.5
12
12
48.5
147
Total Biomass
(Gt)
460
175
108
743
827
Total Soil Carbon
(Gt)
255
142
179
576
1456
With increased greenhouse gases in the atmosphere and the potential for changes in global climate, as
well as the demands of an increasing world population, the balance between fixed and released carbon
may be altered. This may result in changes in the large stocks of above- and belowground carbon in
forest systems and ultimately in the concentration of CO2 in the atmosphere. These changes in
atmospheric C02 in turn are likely to lead to additional redistribution of forests, with an accompanying
change in the amount of carbon stored in forest ecosystems in both the above- and belowground  pools.

2.3.2  Soil Carbon Dynamics: A Conceptual Model

Soil organic matter can be divided into two principal components: a living component and a nonliving,
or dead, component (Theng et al. 1989).  The living fraction includes plant roots and macro- and
microorganisms. This fraction usually constitutes less than 4% of the total amount of soil organic matter
(Theng et al. 1989). The remaining 96-98% of soil organic matter is nonliving and is divided between
macro-organic matter and humus.  Macro-organic matter includes plant residues that are in various
stages of decomposition but that are larger than 0.25 mm.  Humus is simply the organic material
remaining after the removal of the macro-organic matter.

Figure 2.3-1  depicts a conceptual model of the pools and fluxes of carbon in terrestrial ecosystems.  The
boxes represent pools and the arrows represent the flux of carbon between  pools and the direction of
carbon flow. In this model, carbon is divided into living and nonliving pools. The model distinguishes
                                              23

-------
between aboveground and belowground pools and fluxes. Such a model is broadly applicable across

temporal and spatial scales, and it can be an important heuristic in developing a better understanding of

the terrestrial carbon cycle.  It is also useful for identifying areas of uncertainty that need research and

quantification.


Five carbon pools are  presented in the conceptual model. Pool Ci refers to the aboveground living

portion of vegetation.  Pool C2 is the belowground component of vegetation, including coarse and fine

roots. C2 also includes living macro- and microorganisms.  On the nonliving or detrital side of the model

is the aboveground pool, C3, which  is the pool of organic litter or detritus.  Belowground are pools CA

and Cj.  Pool C4 is the fraction of soil carbon that is readily accessible for decomposition (also called

the "labile" pool).  Pool C5 represents that  portion of the soil that is protected, either chemically or

physically, from decomposition.
                                CO,
                                 F.
                   LIVING
                   CARBON
DETRITAL
 CARBON
                          ABOVE-GROUND
                                                           LITTER
                                  F4
                           BELOW-GROUND
                                              F6l
                                                               F7
                                   UNPROTECTED
                                     SOIL Oft.
Fluxes:
FI  •gross pnotosyntnts Is
F2  • ttove-ground rtsplratlon
F3  • lltttr deposition
F4  • btlow-groutd transport
F5  • mizospntrt rtsplratlon
F6  • rhlzo-dtposltlon • txudatts
F7  • lltttr incorporation • Itacnlng
F8  - lltttr respiration
F9  • sell organic matter rtsplratlon
FIO • Itacning and troslon
Fl I • cntmical and pnysical fixation
FI2 • cnemlcal and onysieal release
                                                                                      ABOVE-GROUND
                                                                                      BELOW-GROUND
                                                               F10
                                                    on. • Organic flatter
                                                    Cn's • Terrestrial Caroon Pools
                                                    Change in pool size:
                                                                 • Z Influxes -Z effluxes
                                                                  - (Cn) t1 - (Cn) 12

                                                      Where: ACp - the change in the size of the nth carbon poc

                                                              t1  - time one, t2 - time two




 Figure 23-1.   Simplified conceptual   model  of carbon pools and fluxes  in terrestrial ecosystems.
                                                   24

-------
The conceptual model identifies 12 distinct pathways for carbon to flow between pools or out of the
system, each of which represents the combination of several processes:  Flux F1 represents the fixation
of C02 (gross photosynthesis). The arrows pointing up and away from boxes or pools (F2, F5, F8, and
F9) represent losses through either autotrophic or heterotrophic respiration.  Flux F3 represents litter
deposition at the soil surface.  Flux F4 represents the shoot-to-root transport of carbon-based materials.
Flux F6 represents root and microbial biomass turnover. Flux F7 represents the incorporation of litter
Into the soil, which occurs either by a physical mixing process or by leaching.  Flux F10 is the loss of
carbon through leaching, including organics and bicarbonate. Flux F11  represents chemical and
physical fixation or protection, and F12 represents release by either chemical or physical protection.

Equation 2.3-1  can be used to evaluate the dynamics of the terrestrial carbon cycle, and, in particular, of
individual carbon pools.  Equation 2.3-2 represents an alternate method for determining ACn.

              aCn    =  Z  influxes  - S effluxes                           (2.3-1)
                      =   (CJt2 -  (Cn)tl                                         (2.3-2)
  where:    ACn    =  the change  in  the  size  of nth  carbon  pool
              tl      =  time  1
              tz      =  time  2

The usefulness of Equation 2.3-1 is based upon the ability to measure or determine the magnitude of all
of the carbon fluxes acting on a particular carbon pool.  It is not necessary to know the actual size of
the pool to determine *Cn if all the fluxes can be quantified.   Direct measurement of *Cn may facilitate
quantification of one or more carbon fluxes.  When ACn is equal to zero, the respective carbon pool (i.e.,
no net change in CJ, is at steady state; carbon inputs to the pool  are therefore equal to carbon outputs.
When ACn is positive, net productivity of the pool is positive and the pool is a sink for carbon.
Conversely, when aCn is negative, net  pool productivity is negative, and the pool is a source of carbon.

It would be useful to have measures of *Cn for each  of the pools of terrestrial carbon to evaluate the
sink/source performance of the whole system and to model  terrestrial carbon dynamics. Obtaining
accurate measures of aboveground carbon pools is difficult due to the heterogeneity of individual
stands.  Likewise, accurate measures of belowground pools  of carbon are difficult to make, as are the
flux measures. Techniques for measuring carbon fluxes and belowground pools must be improved or
new methods developed because current methods lack accuracy and specificity.  When accurate
measures are available, experiments can be conducted to evaluate more conclusively the effects of
                                              25

-------
internal and external factors on the fluxes of carbon, which will lead to a better understanding of the
potential effects of global change on terrestrial carbon dynamics. In the following section, the factors
that affect carbon flux are discussed.

2.3.3  Factors Affecting the Accumulation of Soil Carbon

The accumulation and distribution of soil organic matter (carbon) depend primarily on the quality and
quantity of organic inputs and the rates of microbially mediated decomposition (Oades 1988). They also
depend upon the extent to  which the soil protects some fraction of the organic material from
decomposition  (Oades 1988,  Duxbury et al. 1989).  The sequestration of carbon in soils is complex and
depends on both biotic and abiotic factors, including: 1) edaphic factors; 2) precipitation; 3) biology; 4)
temperature; 5) solar radiation; 6) time; 7)  landscape factors; 8) history; 9) management;  and 10)
atmospheric factors. The rationale for each factor is as follows:

1)      Edaphic factors include soil physical and chemical properties, such as soil nutrient status, pH,
        particle size  distribution,  clay mineralogy, bulk density, and water holding capacity.

2)      Precipitation includes the amount and timing of rainfall, snowfall, etc.

3)      Biology includes the vegetation that fixes carbon through photosynthesis and the belowground
        plant parts that acquire water and  nutrient resources.  It includes the physiology of growth and
        the habit of rooting. It also includes soil biology; that is, the suite of organic matter reducers
        and decomposers that ultimately lead to the oxidization of a portion of the fixed carbon.

4)      Temperature includes both the aboveground and belowground temperature. The aboveground
        temperature  affects primary productivity; the belowground temperature affects the rates of
        decomposition.

5)      Solar radiation refers  to the amount and duration of photosynthetically active radiation (PAR),
        which is the  light required to drive photosynthesis. Annual PAR is determined to  a large extent
        by latitude.  Other factors, such as air pollution, water vapor, and cloud cover, also affect the
        amount of PAR. At the highest latitudes,  insufficient PAR may limit primary production.

6)      Time is important because soil carbon accumulates over long periods of time.  Consequently,
                                               26

-------
       the time since a disturbance (e.g., fire, harvest) will influence the amount of carbon that
       accumulates in soil.

7)     Landscape features such as aspect, slope, landscape position (e.g., toe slope, bottom), and
       shape (convex versus concave) affect the mass movement of soil materials and the hydrologic
       routing of soil water. Consequently, these factors affect the accumulation of soil organic matter.

8)     History includes the effects of past disturbances that affect the accumulation of soil carbon, such
       as fire, pests, flooding, mass wasting, glaciers, etc.

9)     Management could also be called land use, mismanagement, or non-natural disturbance.  It
       refers to effects  due to human influence.  Examples include conversion of forests to agriculture,
       deforestation, urbanization, road building, etc.

10)    Atmospheric chemistry refers to the composition of the gaseous atmosphere in which vegetation
       is growing.  There is evidence that increased levels of ozone (Sharpe et al. 1989) and C02
       (Norby et al. 1987) affect the allocation of carbon from  shoots to roots.

No attempt has been made here to present quantitative relationships between these factors and the
accumulation of carbon  in soils.  In fact, many of the relationships are only qualitative at this point. The
relationships that are reported in the literature are usually for single factors, not for several interacting
factors. However, some quantitative bivariate relationships have been reported, such as the positive
correlation between soil  clay content and the stabilization of so'il organic matter (Oades 1988).  The
primary purpose for taking a heuristic approach is to emphasize the complexity of soils and to identify
the relevant influences on  the storage of carbon in soils.

2.3.4   Potential Effects of Global Change on Carbon Storage in Soils

Because the extent and  magnitude of global  climate change is  uncertain, the effects of such a change
on carbon storage in forest soils are even more uncertain. It Is possible, however, to speculate on the
likely effects based upon potential climate change scenarios and  known or expected relationships
between soil carbon  and the factors that influence it.  Some factors will not be affected by climate
change (e.g., soil clay content), but they may affect the amount of carbon stored in soil.
                                               27

-------
If mean annual air temperatures increase, with all other factors constant, ecosystem productivity will
probably increase and the amount of carbon fixed will increase.  At the same time,  however,
decomposition rates will be amplified. The net result is likely to be a reduction in the amount of carbon
stored in the soil. Tundra and boreal soils are likely to be particularly sensitive to warming.  If
temperatures are elevated and precipitation  is reduced, the negative effect on soil carbon will be
exacerbated to a point.  As long as the soils are moist enough to meet microbial needs, decomposition
will proceed. If the  systems get too dry, decomposition will decrease significantly.  Concomitantly,
primary production will decrease.  If precipitation increases along with soil temperature, the amount of
carbon stored in soils is likely to increase because decomposition of soil organic matter is reduced in
very wet soils (Ino and  Monsi 1969, Oades 1988).
                                               *
One potential cumulative effect of climate change is the shifting of the areal extent and distribution of
biomes.  This will affect the storage of carbon in soils.  Shifts from forests to grassland may lead to more
carbon being stored belowground, but much less stored in aboveground biomass.  Using the data of
Waring and Schlesinger (1985), mean biomass for temperate forests and temperate grasslands are 15
and 1 kg carbon/m2, respectively. Mean soil carbon for these two systems are 11.8 and  19.2 kg
carbon/m2, respectively.  A shift from temperate forests to temperate grasslands would result in a
change in mean ecosystem carbon from 26.8 to 20.2 kg carbon/m2. The result is a net loss of carbon
to the atmosphere, but a net increase in belowground carbon.  A shift from temperate grasslands to
forests would have the equal but opposite effect, and it is a possibility.  A detailed discussion of such
analyses is presented in Section 3.1.

As a result of the direct effect of elevated C02 on vegetation, global change might indirectly affect soil
carbon storage.  As mentioned previously, elevated CO2 has been shown in some circumstances to
increase the allocation of carbon from shoots to roots (Norby et al. 1987), primarily because of a
fertilizer-like effect of C02 on plant growth. In response, plants allocate more carbon to build more
extensive root systems necessary for acquiring additional resources (e.g., water, nutrients). The
allocation of additional carbon belowground could lead to increased levels of carbon sequestering in
soils.

2.3.5 Research Needs

The storage of carbon in soils is a very complex process that is not fully characterized or understood.
To develop strategies to mitigate global change, including strategies to manage the biosphere, the
carbon cycle must be more fully understood.  In addition, a number of other areas related to terrestrial
                                               28

-------
carbon dynamics require more research.  One purpose of current analyses at ERL-C is to characterize
the terrestrial carbon cycle and then to project the long-term effects of increasing C02and changing
climates on terrestrial carbon pools and fluxes.  In that light, work is needed in the following areas:

1)     Quantification of carbon pools and fluxes in specific ecosystems under ambient or steady state
       conditions to develop general principles from the specific examples.

2)     Characterization and  quantification of the factors that control carbon fluxes.

3)     Development of methods to characterize the pools and lability of soil organic matter so that
       these measures can be related to dynamic and biologically relevant processes.

4)     Quantification of the effects of land use  and management on soil carbon.

5)     Characterization and  quantification of the stabilization of soil organic matter by abiotic soil
       factors.

6)     Development of simulation models that accurately project how carbon fluxes (and thus pools
       and feedbacks to the atmosphere) will shift in specific ecosystems under a series of altered
       climate scenarios.

2.4    Effects  of Global Climate Change  on Global Vegetation

As described in Section 1.4.2, changes in fluxes of carbon to the atmosphere from the terrestrial
biosphere can significantly affect atmospheric C02 concentrations. Projected changes in global climate
(Section 1.4.3) could result in a significant redistribution of global vegetation, resulting in large changes
in carbon  pools and fluxes. These changes could provide significant positive or negative feedbacks to
climate change.

To begin an evaluation of the potential  significance of this feedback phenomenon, the current best
estimates  of changes in global vegetation caused by climate change are summarized here. The data
presented in this section form a basis for calculating changes in terrestrial carbon storage in Section 3.
                                               29

-------
2.4.1  Global Vegetation Models

Several models describing the global distribution of vegetation (or at least classification schemes relating
climate to vegetation) are available with which to estimate the redistribution of vegetation in response to
climate change.  These include models of Holdridge (1947,1967), KJ ppen (1900,1918,1936), Box
(1981), and Lashof (1987).  Tuhkanen (1980) and Prentice (1990) have reviewed these models.  The
original Holdridge (1947,1967). modified  Holdridge (Prentice 1990), Box (1981), and Lashof (1987)
approaches have been used to predict the redistribution of vegetation In response to climate change.

The Holdridge life zone classification uses three climate parameters to define the occurrence of major
plant formations: biotemperature, mean annual precipitation, and a potential evapotranspiration (PET)
ratio.  Biotemperature is a temperature sum over the course of a year, with unit values (e.g., daily
values) that are less than 0
-------
G   1.5
o
o

 o>
 u

 •-•
 ra
 k.
 o>
 a


 0>
•^
 o

5
c
c

 3.0





 6.0





12.0




24.0
          T    Boreal
                Latitudinal

                belts
            Polar




            Subpolar
          I    Cool

          -    Temperate
         =Warrn Temperate «c
         =   /Subtropical
                                                                                          Altitudinal

                                                                                          belts
                                                                                          Nival
                                                                                         Alpine     ~
  \  Subalpine  ~

^	


      Montane  H


    <§>
   Q Lower   —~

      montane  r
     Premontane =
                                                     ________
                                  / '/'•/// i /1 illil'l'in) >)»»i>>ni»n/n<
       Critical

       temperature

       line
                                                                                                          1.5
                                                                                                         3.0
                                                                                                        6.0
                                                                                                       12.0
                                                                                                       24.0
    Figure 2.4-1.    Holdridge  life zone classification  system  (Holdridge 1967).
                                                        31

-------
A significant factor in evaluating simulations of future vegetation patterns is understanding how well the
vegetation model simulates current vegetation patterns.  The Holdridge life zone classification
reproduces broad-scale global vegetation patterns (Leemans 1990). but it is inaccurate for many regions
of the world (only vegetation in 40% of l°x 1°gridboxes are correctly simulated by the Holdridge system
(Prentice 1990)).  Prentice (1990) evaluated and refined the Holdridge system to improve its accuracy.
In the initial simulation, in which climate space was divided on a finer scale than in the original Holdridge
classification, vegetation in 58% of the land gridboxes was correctly simulated.  An additional refinement
was to aggregate observed vegetation units based on similarity of vegetation and climate. For the final
analysis, 18 primary vegetation types and 11 transitional zones were defined. Using this scheme,
vegetation in 77% of the gridboxes was  predicted correctly.

A more complicated  but more biologically realistic global vegetation  model was developed by Box
(1981).  Instead of analyzing climate-biome relationships, Box analyzed the distributions of 77 plant life
forms (e.g., summergreen broadleaved trees)  throughout the world.  For each life form, a set of eight
different climate values were correlated with the range limits of the life form. In essence,  Box created a
set of 77 different climate envelopes within which the life form occurs and outside of which the life form
is absent.  The Box model can  predict combinations of  growth forms at any given location, so it is
capable of predicting canopy structure.

Box evaluated his model by simulating life forms present at 74 sites around the world.  The actual
dominant growth forms  were predicted for 92% of the sites, but all dominants and codominants, and no
others, were correctly predicted for only 50%  of the sites.

Lashof (1987) developed a statistical model of climate-vegetation relationships using Olson et al.'s (1983)
vegetation database. Since this database reflects actual rather than potential vegetation,  the Lashof
model is the only currently available approach that incorporates land use.

2.4.1.1  Limitations of Global Vegetation Models

The four global vegetation models just summarized have significant limitations for predicting future
global vegetation patterns.  First and foremost, all are steady-state models and are non-dynamic.  They
give no information on how long it would take the vegetation to return to an equilibrium with climate. It
may take 200-500 years for forest and shrub species to respond to a large climate change (Davis et al.
1986, Webb 1986). Thus, these models cannot be used directly to estimate vegetation patterns that
could exist in the next century.  Second, the Holdridge  and Lashof approaches, and to a  lesser degree
                                               32

-------
the Box approach, are based on empirical relationships between climate and vegetation.  Since future
climate regimes in some regions may be different from any present in the world today, vegetation
models based on the mechanistic response of species or growth forms to climate must be developed.
Third, the models are imperfect predictors of present vegetation, which introduces bias into the
simulations of future vegetation patterns. Fourth, none of the models take into account the effect of soils
(e.g., nutrient availability, texture) or land use (with the exception of the Lashof model) on vegetation
distribution.  Thus, simulating future vegetation using these models assumes that there are appropriate
changes in soils to support the predicted vegetation type.  In regions where soils have a major effect on
vegetation cover,  predictions by the vegetation models will likely be incorrect.  Fifth, the direct effects of
C02 on plants are not incorporated into the models.

2.4.2  Equilibrium Simulations of Future Vegetation Patterns

Despite the limitations discussed above, simulations of the potential response of vegetation to future
climate change are useful for understanding both the magnitude of possible vegetation change and
biospheric feedbacks to climate change. Several global simulations have been recently completed that
form the basis for estimating changes in carbon  pools and fluxes.  Emanuel (1985) and Emanuel et al.
(1985) made the first global projection of future vegetation patterns using the Holdridge approach,
although only changes  in temperature were used to calculate future vegetation patterns.  A more recent
and complete simulation using the  Holdridge approach and five climate scenarios has been completed
by Leemans (1990) and Smith et al. (submitted). Prentice and  Fung (in press) used the refined
Holdridge system (Prentice 1990) to project future  vegetation patterns and carbon storage. Smith et al.
(submitted) also used the Box model to project changes in vegetation in the tropical and boreal regions
of the world.

Major changes in the distribution of the world's vegetation are projected under all the double-CO2
climate scenarios and the different  vegetation models (Emanuel 1985, Emanuel et al. 1985,  Smith et al.
submitted, Prentice and Fung in press). The results of each analysis are briefly summarized here.  It
should be emphasized  that these analyses are for potential natural vegetation and do not take into
account  human land usage (e.g., agriculture).

Emanuel (1985) and Emanuel et al. (1985)  created a future climate scenario using temperature data  from
Manabe and Stouffer's  (1980) simulation of global climate under quadruple-CO2 concentrations.
Temperature changes were divided by two to create a double-CO2 climate scenario.  Under this
scenario, tropical forests, grasslands, subtropical deserts, and boreal deserts showed significant
                                               33

-------
expansion, whereas subtropical forests, boreal forests, warm temperate forests, tundra, and ice
contracted significantly (Table 2.4-1).
Table 2.4-1. Changes in Global Vegetation Distribution in a Double-CO 2 Atmosphere using the
Holdridge Classification System (as applied by Emanuel 1985 and Emanuel et al.
1985)
Vegetation Type
ice
Tundra
Boreal Deserts
Boreal Forests
Cool Temperate Deserts
Grasslands
Cool Temperate Forests
Warm Temperate Deserts
Warm Temperate Forests
Subtropical Deserts
Tropical Deserts
Subtropical Forests
Tropical Forests
Current
Area
(106krrv)
2.22
4.47
1.31
17.26
4.84
22.78
11.29
6.75
15.81
2.91
10.74
11.96
19.03
Double-CO2
Area
(106krrv)
0.57
3.03
2.59
10.88
4.04
28.52
11.63
5.41
14.69
3.66
.12.63
9.38
24.33
Percent
Change
-74
-32
98
-37
-16
25
3
20
-7
26
18
-22
28
Total
131.37
131.36
                                                34

-------
Smith et al. (submitted) used five different climate scenarios generated from output from the four GCMs
described in Section 1.4.2.  The fifth scenario is a second run of the GFDL model using a more realistic
heat flux, and it is referred to here as GFDL-QFIux.  Under all climate scenarios, the tundra, cold
parklands, forest tundra, boreal forest, cool desert, warm temperate forest, and tropical seasonal forest
biomss generally decreased significantly in areal extent (Table 2.4-2). The temperate forest, tropical
semi-arid, tropical dry forest, and tropical rain forest biomes increased in areal extent.  Tropical rain
forests doubled in area under the GISS and OSU climate scenarios. The GFDL-Qflux scenario results
represent the least change from the present, with only the cold parklands and tropical  rain forests
changing in size by more than 20%.
Table 2.4-2. Changes in Global Potential Vegetation Distribution in a Double-CO 2 Atmosphere
using the Holdridge Classification System (as applied by Smith et al. submitted)


Biome
Tundra
Cold Parklands
Forest Tundra
Boreal Forest
Cool Desert
Steppe
Temperate Forest
Hot Desert
Chaparral
Warm Temperate Forest
Tropical Semi-Arid
Tropical Dry Forest
Tropical Seasonal Forest
Tropical Rain Forest
Total
Current
Area

9.30
2.79
8.90
15.03
4.01
7.39
9.94
20.85
5.58
3.17
9.56
14.86
15.13
8.46
134.97



Percentage
i GFDL
-66
1
-56
-36
-24
57
19
- 1
33
-38
46
32
-34
82

GFDL-QFIx
-9
-28
- 1
3
-13
-10
17
-4
-3
7
-10
-5
-5
47

GISS
-54
-15
-34
-10
-42
-6
35
-15
-2
-40
75
30
-48
105


Change
OSU
-49
-4
-33
-6
-21
18
16
-7
-12
-23
27
0
-33
137



UKMO
-69
-39
-62
-32
-48
-3
31
-4
54
-9
74
75
-49
52

Prentice and Fung (in press), using a GISS climate scenario, also reported significant increases in the
areal extent of tropical rain forests, as well as savanna, cold deciduous broadleaved forest and
woodland, and drought deciduous forest (Table 2.4-3).  Drought deciduous woodland, arid grassland,
desert, evergreen needle-leaved forest, tundra,  and temperate evergreen seasonal forest all showed
significant decreases in area.
                                               35

-------
Table 2.4-3.    Changes  in Global Vegetation  Distribution  in a Double-CO 2 Atmosphere  using the
               Holdridge Classification  System  (as modified by Prentice  1990 and  applied by
               Prentice  and Fung in press)
Area
Vegetation Type (106km^
Polar Desert and Ice 3.00
Tundra 7.00
Cold-Deciduous Needleleaved 16.00
Forest and Woodland
Evergreen Needleleaved Forest 6.00
and Woodland
Mesic Grassland 2.00
Drought-Deciduous Woodland 6.00
Arid Grassland and Shrubland 30.00
Cold-Deciduous Broadleaved 12.00
Forest and Woodland
Temperate Evergreen Seasonal 2.00
Broadleaved Forest
Mediterranean Forest 2.00
and Woodland
Desert 14.00
Savanna 5.00
Drought-Deciduous, Drought- 9.00
Seasonal Broadleaved Forest
Tropical Rain Forest 19.00
Total 133.00
Percent
Change
0
-63
5

-66

-17
-42
-22
40

-31

4

-62
35
21

75
129.44
Double-C02
Area
3.00
2.59
16.80

2.04

1.66
3.48
23.40
16.80

1.38

2.08

5.32
6.75
10.89 •

33.25

In contrast to the results using the Holdridge system presented, preliminary estimates using the Box
approach suggest that tropical forests will decrease in area rather than increase (Smith personal
communication).

To compare the results of Emanuel (1985), Emanuel et al. (1985), Smith et al. (submitted), and Prentice
and Fung (in press), the vegetation classes they used have been grouped into six broad vegetation
types (Table 2.4-4, Figure 2.4-2). Generally, there is agreement in terms of the sign of the predicted
vegetation change.  Deserts, boreal forests, and tundra types decreased in area! extent, whereas
grasslands, temperate forests, and tropical forests generally increased in area under double-CO2
climates.  Across the seven different simulations and six vegetation categories, there are only four
differences in the direction of the predicted change.  The GFDL-Qflux climate scenario caused the
                                               36

-------
smallest changes in global vegetation. Despite the general agreement in the sign of the changes, there
are significant differences in the estimated magnitude of the vegetation changes. For instance, estimates
of the future distribution of boreal forest differ by a factor of two.
Table 2.4-4.
Comparison  of Global Vegetation  Distribution in a Double-CO 2 Atmosphere  using
the Holdridge Classification  System,  as applied  by  Three Teams of Investigators
   Smith et al. (submitted) Predictions
                              Current
                               Area
                                       Percentage Change
Vegetation Type
Deserts
Grasslands/Shrubland
Temperate Forests
Boreal Forests
Tropical Forests
Tundra
Total
Prentice and Fung (in
Vegetation Type
Deserts
Grasslands/Shrubland
Temperate Forests
Boreal Forests
Tropical Forests
Tundra
Total
Emanuel et al. (1985)
Vegetation Type
Desert
Grasslands
Temperate Forests
Boreal Forests
Tropical Forests
Tundra
Total
(106.knv)
20.80
26.60
13.10
26.70
38.50
9.30
135.00
press) Predictions
Current
Area
(106km^
14.00
38.00
16.00
22.00
33.00
10.00
133.00
Predictions
Current
Area
(106knv)
20.40
27.62
27.10
18.57
30.99
6.69
131.37
GFDL GFDL-QFIx GISS
-1 -4-15
35 -9 18
5 14 17
-39 -2 -19
17 6 16
-66 - 9 -54

Percentage
(GISS)
-62
-25
27
- 14
54
-44


Percentage
Change
6
18
- 3
-27
9
-46

OSU UKMO
-7 -4
9 30
7 21
-15 -43
17 21
-49 -69














                                               37

-------
60
UJ
g 40
tx
£ 20
t o
5 -20
« -40
a.
-60

-
1 	 1
1U.


r wr\t J i
-
n^nnn "
i i i i i i i
» „

> i i i i i i







_
n^nrir^n "
_ _
„ _

t I 1 i | ! |
GFDL GISS GFDL OFLX GISS OSU UKMO CFDL GISS CFDL OFLX GISS OSU UKMO
                       tlco      SmllK ol ml. 1MO
                •I ol.  uio
                IMS   '•••.
                     I MO
                                           CIM«I»I <>rmll»       tmln M ol. 1MO
                                            M «L  ••*
                                            Kit   r«i.«
         60

         40

         20

          0

        -20
     o
     S  -40
        -60
             BOREAL  FOREST
                                          GRASSLAND/SHRUBLAND
 CFDL CISS CFDL QFLX  CISS  OSU  UKMO
[mofivol P'ootko      Smith ot ol. IMO
 ol ol.   on*
 1MB  Tunt
      IMO
                                                           CFDL CISS GFOL QFLX  GISS  OSU  UKMO
                                                          Cfn«nw«l
                                                           •I el.
                                                                           Smith « ol. l»«0
                                                                H90
             DESERT
                                          TUNDRA
                                                                                       I
                CFDL CISS GFDL QFLX  CISS  OSU  UKMO
               Cmonuol Pr««tlco
                •I ol.  •««
                                Smith ol ol. llfO
                                            CFDL  CISS  GFDL QFLX CISS  OSU UKMO
                                           CmoiMMl »ro«!l«o      SmltK ol ol. ItIO
                                             o< ol.  ••«
                                             IMS   rim*
                                                  1MO
Figure 2.4-2.    Histograms  showing changes  in the total area of six major vegetation  types
                after a double-CO 2induced climate  change.   Data are also summarized  in
                Table 2.4-4.
                                               38

-------
The results presented here in effect are summaries of global vegetation before and after a climate
change caused by a doubling of C02 concentrations. What is not indicated by these numbers is the
amount of land predicted to change from one vegetation type to another. This could significantly impact
the global carbon cycle because of the transient release of carbon to the atmosphere from vegetation
dieback (see Section 3.2).  In the Prentice and Fung (in press) scenario, 60% of the vegetated
landscape on the globe changed vegetation type.  In the Smith et al. (submitted) scenarios, 16% (GFDL-
QFlux) to 56% (UKMO) of the earth's land surface changed vegetation type (Table 2.4-5, Figure 2.4-3).
Such predicted changes would have significant impacts on biodiversity (discussed in the next section),
water resources, forest resources, agriculture, and land management.
Table 2.4-5.     Area! Extent of Land on the  Globe Changing  Vegetation  Cover under Double-CO 2
               Conditions  as Estimated  Using the  Holdridge system  (Smith et al.submitted)
Amount of Land Changing
Climate Vegetation Cover Percent of Land
Model (106km^ Surface Changing
GFDL
GFDL-QFlux
GISS
OSU
UKMO
(Total Land Area:
65.24
21.51
60.36
53.73
76.21
134.97)
48.34
15.94
44.72
39.81
56.46

                                               39

-------
                                                                    V,
Figure 2.4-3.   a) Global and  b) North American  areas  (shaded)  in which predicted future
               vegetation  is different  from current vegetation  using the UKMO double-CO 2
               climate scenario (Smith et al. submitted).
                                                 40

-------
2.4.3  Discussion of Vegetation Scenario Results

The fact that the vegetation scenarios are similar (in terms of the sign of the change) using the Holdridge
approach is not surprising, given that the GCM scenarios on a global basis predict warmer and wetter
conditions than exist at present (see Section 1.4.3).  In general, a point in a given climate space on the
Holdridge diagram (Figure 2.4-1) can be expected to move down and to the right under double-CO2
climate conditions.  More detailed results from Smith et al. (submitted) not presented here in fact show
that this is often the case. For instance, areas of current tundra vegetation are estimated to change to
forest tundra or boreal forest in the future. Tropical seasonal forest regions that change type tend to
change to tropical  rain forest.

Estimates of current area of the six vegetation types can differ substantially between investigators.  That
the Prentice and Fung (in press) areas are different is not surprising, as they used a different vegetation
classification system from the original Holdridge system. However, Emanuel (1985), Emanuel et al.
(1985), and Smith et al. (submitted) used the same basic vegetation units in their analyses, and they
aggregated the units in the same way to obtain the areas for the vegetation types presented in Table
2.4-4. For instance, estimates of boreal forest area differ by  over 8 million km2.  A possible reason for
this is that Emanuel et al. (1985) used monthly temperature values to calculate annual biotemperature,
whereas  Smith et al. (submitted) used daily values. Consequently, differences in the percent changes
listed in Table 2.4-4 are in part due to differences in how the current vegetation areas were calculated,
as well as to how the individual vegetation units were aggregated to form the vegetation units used in
the table. The estimated changes in vegetation may also differ because of large differences in the
climate scenarios used to drive the vegetation models (see Section 1.4.3).

2.4.4  Research Needs

Predicting the transient response of global vegetation to climate change is essential for estimating
biospheric feedbacks and regional sensitivity to climate change, and for evaluating the effectiveness of
various mitigation strategies (e.g., reforestation).  Current global vegetation models cannot be used to
estimate  transient vegetation dynamics.  Moreover, these models  are based on correlations between
climate and vegetation, correlations that may  not persist under altered climate conditions.  Finally, the
models do  not incorporate soils or the potential effects of C02 fertilization on vegetation.

There is much work to be done in producing a dynamic and realistic global vegetation model.  The
following areas of research need attention in order to develop such a model:
                                               41

-------
1)      Development of a plant life-form classification system that can form the basis of a global
        vegetation model.  Life-forms grouped together should have common autoecological characters,
        such as physiognomy, seed dispersal, and response to macroclimate and soils.

2)      Determination of the ecological mechanisms through which climate controls the distributions of
        these life-forms.

3)      Incorporation of disturbance and migration into global vegetation models.

4)      Production of global vegetation maps based on remotely sensed data to calibrate and validate
        vegetation models.

5)      Generation of digitized databases of soil texture and nutrient availability at a resolution
        appropriate for global modeling.

6)      Calculation of the effects of COa enrichment on ecosystem water-use efficiency.

2.5     Impacts of Climate Change on  Biological  Diversity

Loss of biological diversity (biodiversity) is becoming one of the most critical environmental issues of the
1990s.  Estimates of the annual global rate of species extinction range from 1,000 to 10,000 times the
rate before human intervention (Wilson 1988).  The basic issue driving concerns about biodiversity is the
accelerating and irreplaceable loss of genes, species, populations, and ecosystems. Associated with
this loss are the loss of products presently or potentially obtained from nature, the possible disruption of
essential ecological processes ancfservices, and the forfeit of options for biological and cultural
adaptation to an uncertain future.

Forested ecosystems account for much of the world's biological diversity.  Indeed, the greatest
concentration of species diversity in the world  is in tropical rain forests. These forests account for only
7% of the earth's land cover, yet they contain at least 50% of all species (Reid and Miller 1989).  Current
estimates indicate that 1% of this biome is being deforested each year and that another 1% is being
significantly degraded (Myers 1988). Thus, much of the current concern about loss of biodiversity has
centered on tropical  forests. However, maintenance of biodiversity is also very important in temperate
areas in which human activities have greatly altered the terrestrial landscape. Temperate forest zones
                                                42

-------
have been more uniformly and extensively altered by human activities than any other region of the world.
Most remaining areas of temperate forests are fragmented and highly modified.

In forested ecosystems, climate change is predicted to cause  an intensified loss of biodiversity
(Henderson et al. 1989).  Projected effects of climate change on forested ecosystems Include reductions
in species diversity in low-elevation forests as well as elevational and latitudinal shifts in species ranges
(Leverenz and Lev 1987).  Species are most likely to be stressed at the edge of their ranges. Plants and
animal species with low vagility, or that are prevented from migrating by a lack of habitat corridors, may
become regionally extinct.  Habitat modification and climate change are certain to interact to produce
other synergistic effects that will be difficult to predict or mitigate.

Any impact to vegetation from global climate change will  be extended to animals through food chains
and alteration of habitat structure.  Historically, animals have responded to climate changes by following
the shifting vegetation assemblages.  Gradual change in species composition is a natural response to
climate change.  However, the projected  high rate of climate change over the next few decades is of
concern because species with limited dispersal capacities (i.e., with slow rates  of migration) may not be
able to track  shifting climate zones to stay within their limits of physiological tolerance (Peters and
Darling 1985). Migrating species will have to contend not only with natural geographic barriers to
dispersal, but also with a landscape altered by humans and with barriers such as roads, cities, and
agricultural lands.

We have examined existing species richness maps for various taxa (amphibians, reptiles, trees, birds,
and mammals, see Figure 2.5-1) to determine current patterns of biodiversity.  By overlaying the different
taxa maps, areas of high biodiversity can be identified. For example, the southeastern portion of the
United States is characterized by large numbers of amphibian, tree, and  reptile species, as well as
relatively high levels of bird species diversity. Thus, this area  is of significant interest for maintenance of
biodiversity and further research in understanding species and ecosystem sensitivity to climate change.
Preliminary analyses overlaying the density maps on the vegetative change maps (e.g., Figure 2.4-3)
generated using the Holdridge system and GCM scenarios, indicate that much of the southern United
States could  change vegetative cover under double-CO2 scenarios. Vegetative change will affect animal
habitat. Consequently, the southern United States is at high risk of losing species under all GCM
scenarios.
                                                43

-------
      AMPHIBIA*  SPECIES DEK5ITT
                                                                     IEPTIIE SPECIES OEKSITY
                            TREE  SPECIES  DENSITY
Figure 2.5-1.   Density/richness   maps  illustrating one level of biodiversity  for a) amphibians
               (Kiester 1971), b) reptiles  (ibid.), and c) trees (Currie and Paquin  1987)  in North
               America  and/or the contiguous  48 United States.  Units refer to the number of
               different  species  found  in that area.
                                                44

-------
                                                  BIRD  SPECIES  DENSITY
                     I   I   < 40
                     II 40 - It
                         10 - 120
                      ill!!!! 120 • 110
                         110 - )00
                         100 - 140
                         HO • 210
                         110 - 120
                         110 - 310
                          > 360
                                                 MAMMALS  SPECIES  DENSITY
Figure  2.5-1 (Cont'd).  Density/richness   maps  illustrating one level of biodiversity  for d) birds
               (Cook 1969),  and e) mammals  (Simpson  1964)  in North America  and/or the
               contiguous  48 United  States.  Units refer to the number of different species  found
               in that area.
                                              45

-------
Global climate change could have significant consequences for biodiversity.  Anticipated effects vary
among regions and ecosystem types;  thus, it will be necessary to identify those areas where changes
would  be most catastrophic for biodiversity.  Further research is needed on how biodiversity will be
affected through projected changes in major vegetation zones. Specific needs are to:

1)      Determine climate and land use controls of biodiversity.

2)      Identify high-diversity, high-risk ecosystems.

3)      Identify vulnerable species  and guilds at a regional and continental scale.  Analyze specific
        species at risk, their life history characteristics, and the availability of suitable habitat and or
        corridors for migration and dispersal.
                                                 46

-------
3  FEEDBACKS:  TERRESTRIAL CARBON STORAGE

Changes in global vegetation of the magnitude described in Section 2.4 will have a significant impact on
terrestrial carbon storage and will thus act as a positive or negative feedback to climate change.  In
Section 3.1, the amount of above- and belowground carbon associated with vegetation in a double-C02
atmosphere is estimated.  In Section 3.2, the analysis is expanded to estimate the dynamics in terrestrial
carbon associated with vegetation change from the present to a double-CO2 atmosphere.

3.1     Effects  of Climate Change  on Carbon Storage in Terrestrial Ecosystems:   Equilibrium
       Analyses at the Global Level

The objective of this section is to investigate the potential for a net flux of carbon between the terrestrial
biosphere and the atmosphere due to climate change.  This is an equilibrium analysis (i.e., the
assumption is made that climate and vegetation are always in equilibrium), and a bookkeeping approach
is used in which: 1) vegetation types are distributed across the land surface based on the current
climate and double-CO2 climate scenarios using an existing vegetation-climate correlation system
(Holdridge 1947); 2) the vegetation types are assigned representative above- and belowground carbon
pool values; 3) terrestrial carbon storage is summed by climate scenario; and 4) the differences between
current storage  and future storage are determined.

3.1.1  Methods

3.1.1.1 Vegetation data

The double-C02 atmosphere simulations of global vegetation  generated by Leemans (1990) and Smith et
at. (submitted) and summarized in Section 2.4 were used in this analysis.  Holdridge life zones were
aggregated into biomes following the scheme shown in Table 3.1-1.
                                              47

-------
Table 3.1-1.    Aggregation  Scheme  for Combining Holdridge  Life Zones into Biomes  (Cramer and
              Leemans in press)
                     Biome
                            Holdridge Zone
        2
        3
        4
        6
        7
        9
       10
              Tundra
Cold Parklands
Forest Tundra
Boreal Forest
Cool Desert

Steppe
Temperate Forest
              Hot Desert
Chaparral
Warm Temperate Forest
       11     Tropical Semi-Arid


       12     Tropical Dry Forest

       13     Tropical Seasonal Forest

       14     Tropical Rain Forest
Ice
Polar Desert
Subpolar Dry Tundra
Subpolar Moist Tundra
Subpolar Wet Tundra
Subpolar Rain Tundra
Boreal Desert
Boreal Dry Scrub
Boreal Moist Forest
Boreal Wet Forest
Boreal Rain Forest
Cool Temperate Desert
Cool Temperate Desert Scrub
Cool Temperate Steppe
Cool Temperate Moist Forest
Cool Temperate Wet Forest
Cool Temperate Rain Forest
Warm Temperate Desert
Warm Temperate Desert Scrub
Subtropical Desert
Subtropical Desert Scrub
Tropical Desert
Tropical Desert Scrub
Warm Temperate Thorn Steppe
Warm Temperate Dry Forest
Warm Temperate Moist Forest
Warm Temperate Wet Forest
Warm Temperate Rain Forest
Subtropical Thorn Woodland
Tropical Thorn Woodland
Tropical Very Dry Forest
Subtropical Dry Forest
Tropical Dry Forest
Subtropical Moist Forest
Tropical Moist Forest
Subtropical Wet Forest
Subtropical Rain Forest
Tropical Wet Forest
Tropical Rain Forest
                                             48

-------
3.1.1.2 Carbon pools

Aboveground carbon ranges from 25 kg/2 in dense forests to less than 0.5 kg/m2 in the arctic (Olson et
al. 1983).  The summary of Olson et al. (1983) was used as a basis for assigning aboveground carbon
pool values to the aggregated Holdridge vegetation types (Table 3.1-2).  Belowground carbon pools
were assigned to the Holdridge vegetation types (Table 3.1-2) based on a study by Post et al. (1982) in
which data from 2696 global sites were used to construct isolines for belowground  carbon storage within
the Holdridge climate diagram  (Figure 3.1-1).  Based on the current climate, this analysis estimated
totals for aboveground carbon (852 Gt) and belowground carbon (1456 Gt) that are consistent with other
studies (Schlesinger 1977, Woodwell et al. 1978, Oades 1988).

Certainly there is large spatial heterogeneity in these pools within each vegetation type,  but evaluation of
that heterogeneity will require detailed survey  information that is not currently available.  Mean values
were used in this analysis as a first approximation to account for the spatial heterogeneity.

The absolute value of the representative carbon pools is also in question.  Recent analysis of boreal
forest ecosystems (Botkin and Simpson 1990) suggests that previous  estimates of aboveground carbon
in these and other ecosystems have been uniformly high.  Additional studies, probably coordinated with
remote sensing, are needed to evaluate this question.  In the present analysis, the carbon pools
associated with each vegetation type were held constant, and a difference between current and future
storage was calculated. Thus, the approach is not particularly sensitive to consistent overestimation of
above- or belowground carbon across all vegetation types.
                                               49

-------
Table 3.1-2.     Above-  and Belowground  Carbon Pools  for World Biomes
                                                               Carbon Density1

1
2
3
4
5
6
7
8
9
10
11
12
13
14
Biome
Tundra
Cold Parklands
Forest Tundra
Boreal Forest
Cool Desert
Steppe
Temperate Forest
Hot Desert
Chaparral
Warm Temperate Forest
Tropical Semi-Arid
Tropical Dry Forest
Tropical Seasonal Forest
Tropical Rain Forest
Above-
ground
0.5
0.8
6.0
11.0
0.6
1.5
11.0
0.4
4.0
10.0
5.0
7.0
10.0
15.0
Below-
ground
22.0
10.0
11.0
15.0
9.0
13.0
18.0
1.0
8.0
10.0
4.0
7.0
11.0
19.0
lAboveground data after Olson et al. 1983; belowground data after Post et al. 1982
                                              50

-------
Polar

Subpolar
Boreal
Cool temperate
Warm temperate
Subtropical
Tropical
                                   . o
Nival

Alpine

Subalpine
              1.5" •
                 0
                                                                                         6.
Montane       ^e«
Lower montane   £
prem0ntane
              24*
                                    2       4 6 8    10    1418   22
                                Carbon in mineral soil (kg m~2)
    Figure 3.1-1.   Relationship  of belowground  carbon  pools to Holdridge  life zones  (Post et al.
                 1982).
                                               51

-------
    3.1.2  Results and Discussion


    Differences in the areal extent of the biomes, as predicted by four GCMs, are listed in Table 3.1-3.
    Results for the four GCMs show some basic similarities.  At middle to high latitudes, tundra and boreal
    forests contract, and temperate forest (mostly coniferous) expands. In the tropics, both the semi-arid
    woodland and the tropical rain forest have large increases in areal  extent, mostly at the expense of
    tropical seasonal forests.
    Table 3.1-3.
Biome
Changes  in Areal Extent of Different Vegetation  Types as Predicted  under Four GCM
       Scenarios  (Smith et al. submitted)

                       Area                     Difference in Biome Area (106km^
                     (106krrv)       GFDL           GISS          OSU      UKMO
     1      Tundra
     2      Cold Parklands
     3      Forest Tundra
     4      Boreal Forest
     5      Cool Desert
     6      Steppe
     7      Temperate Forest
     8      Hot Desert
     9      Chaparral
    10     Warm Temperate Forest
    11      Tropical Semi-Arid
    12     Tropical Dry Forest
    13     Tropical Seasonal Forest
    14     Tropical Rain Forest

           Total
9.30
2.79
8.90
15.03
4.01
7.39
9.94
20.85
5.58
3.17
9.56
14.86
15.13
8.46
134.97
-6.11
0.03
-5.02
-5.45
-0.97
4.20
1.92
-0.20
1.83
-1.22
4.43
4.71
-5.11
6.95

                                                   -5.05
                                                   -0.41
                                                   -3.03
                                                   -1.54
                                                   -1.67
                                                   -0.46
                                                    3.49
                                                   -3.22
                                                   •0.13
                                                   -1.25
                                                    7.18
                                                    4.49
                                                   -7.24
                                                    8.85
-4.56
-0.10
-2.90
-0.89
-0.82
 1.30
 1.63
-1.42
-0.69
-0.72
 2.58
-0.00
-4.98
11.57
 -6.43
 -1.09
 -5.50
 -4.85
 -1.93
 -0.21
 3.04
 -0.92
 2.99
 -0.29
 7.07
11.19
 -7.48
 4.40
    Analysis of changes in carbon storage reveals that all climate scenarios predict an uptake of carbon
    from the atmosphere ranging from 37 to 116 Gt for the aboveground component of the biosphere
    (Figure 3.1-2).  There are great discrepancies between the model runs for the change in belowground
    carbon storage. The UKMO model predicts a flux to the atmosphere of 126 Gt, whereas the OSU model
    predicts a belowground accumulation of 37 Gt.  The large fluxes to the atmosphere appear to be
    associated with loss of carbon from reduction in tundra and boreal forest areas, both of which have
    relatively high levels of belowground carbon.
                                                  52

-------
   At) ova ground
      GFDL

        GISS

        osu

     UKMO


    Balow Sround
      GFDL   h
               i
        GISS   !-

        osu   r
               i
     UKMO   ^
        Total
       GFDL  r

        GISS  r

         OSU  h

      UKMO
             -150   -100    -50      0
                Flux to  Atmosphere
   50     100    150
Uptake by Biosphere
200
                                CARBON CHANGE (Gt)
Figure 3.1-2.   Potential changes in terrestrial carbon storage based on redistribution of vegetation
            types.  Values are differences  between  stored carbon  under the current climate
            and under double-CO 2 climates as predicted by four CGMs.
                                       53

-------
The net change of above- and below/ground carbon ranged from a 169 Gt uptake predicted by the OSU
model to a 68 Gt release predicted by the UKMO model. A key question appears to be how great the
gains will be in tropical rain forest. Because both above- and belowground components have high
carbon storage in these forests, the changes from tropical dry forest to tropical wet forest tend to drive
the global trends.  In terms of climate, the question is whether precipitation will substantially Increase in
tropical and subtropical latitudes.

A number of other investigators have estimated potential changes in carbon storage based on the
climate-vegetation correlation systems and other GCM double-CO2 climate scenarios.  For aboveground
carbon,  Sedjo and Solomon (1989) predicted a net flux of 13.9 Gt to the atmosphere. Much of that
change in aboveground carbon was accounted for by loss of boreal forest and increases in savanna.
Lashof (1987, 1989), evaluating potential above- and belowground changes, reported a range from a 64
Gt uptake from the atmosphere to a 26 Gt release, again depending on the GCM used.  A modified
Holdridge classification system (Prentice 1990), which better represented the current vegetation,
suggested a much greater potential flux of carbon (270 Gt) to the surface pools of above- and
belowground  carbon (Prentice and Fung in press).  As in our analyses, the uptake was related primarily
to expansion  of the tropical rain forest biome which has relatively high levels of above- and belowground
carbon.

To evaluate the magnitude of feedbacks associated with these carbon fluxes, it is necessary to know: 1)
how much of the carbon would be retained in the atmosphere (in the case of a positive feedback); 2)
what the radiative  forcing of the CO2 added or removed from the atmosphere would be; and 3) what the
sensitivity of the climate system is to that forcing. Likewise, for a negative feedback, the magnitude can
most readily be evaluated via the impact on atmospheric C02 concentration.

The proportion of  carbon released to the atmosphere due  to fossil fuel combustion and deforestation
that has accumulated in the atmosphere has been approximated at 40%  (IPCC 1990). The remainder
has been taken up by the ocean, in part due to photosynthetic uptake of CO2 by plankton, and by the
terrestrial biosphere via a net imbalance between C02 assimilation during photosynthesis and C02 loss
via plant, animal, and microbial respiration.  The atmospheric retention factor is difficult to determine
because of uncertainties about many terms in the global carbon cycle (Keeling et al. 1989).   It is also
likely to change over the coming decades in response to factors such as C02 fertilization of plants and
warming of the ocean.
                                              54

-------
Assuming a retention factor of 40%, the 68 Gt release of carbon in the case of the maximum positive
feedback discussed earlier would result in a 27 Gt increase in the atmospheric pool.  Given a net change
in the atmospheric carbon pool, a conversion to a change in atmospheric C02 concentration can be
made using an approximation of 2 Gt carbon to 1  ppm C02.  Thus, the increase would be about 13 ppm
CO2.  In the case of a negative feedback, Gt can be converted directly into ppm CO2 extracted from the
atmosphere by dividing by two.

The estimation of a climate forcing associated with a change in atmospheric CO^ independent of
feedbacks in the climate system, is relatively straightforward (Hansen et al. 1988) and the GCMs are in
general agreement (Cess et al. 1989).  Following the approach of Hansen, the climate forcing of 13 ppm
added to a reference C02 concentration of about 450 ppm for a double-CO2 climate (the rest of the
warming  being provided by other trace gases), is about +0.05=0.  The effect of subtracting 84 ppm
(l69Gt/2) would be about -0.28CC. These values compare with the double-CO2 climate forcing without
feedbacks of  +1.25<€.  The GCMs are  in much less agreement in accounting for various feedbacks in
the climate system given an initial forcing (Hansen et al. 1984).  The total climate forcing for the given
potential  changes in C02 is thus much less certain. However, the total effects would probably scale to
the magnitude of the original forcing, so a maximum negative feedback of perhaps 20% of the original
forcing is predicted from this analysis due to changing terrestrial carbon pools.

3.1.2.1 Modifiers and uncertainties

The most troublesome assumption using the equilibrium approach described here is that the
climate/vegetation correlation will remain constant. Physiological responses of plants to CO2 suggests
that water-use efficiency increases with ambient C02 concentration.  If this effect is large, aboveground
carbon associated with a particular water regime might  be expected to increase, and thus the magnitude
of the predicted feedbacks associated with carbon'pool changes would be different.

A second concern is that the analysis assumes that climate and vegetation are now, and will remain, in
equilibrium.  In fact, the predicted rates of climate change (>0.1<€ per decade) are an order of
magnitude faster than what natural systems experienced during the glacial-interglacial cycles of the
Pleistocene period.  Analyses of tree seed dispersal distances and pollen records indicate that the rate
of climate change may exceed rates of vegetation redistribution (Davis 1981).  The predicted
discrepancy between these rates may have significant implications for changes in aboveground carbon
pools.  Large transient carbon fluxes could occur if there is a rapid burnoff of carbon as the vegetation

                                              55

-------
drifts out of equilibrium with climate, and vegetation recovery is slow because of both limited migration
rates for appropriate species and slow regrowth.

There will  also be lags in the equilibrium between belowground carbon pools and climate. Where
carbon gains are expected, as with the transition from tropical seasonal to tropical rain forest, there are
few data to predict how rapidly the accumulation would occur.  Factors such as soil mineralogy and
texture may have strong local influences (Oades 1988).  Land use considerations will likewise influence
carbon storage because of the tendency for cultivation to decrease stored carbon.

Where equilibrium analyses suggest a reduction in belowground carbon, an important consideration will
be the fractionation of the  soil organic matter. Microbial respiration increases with temperature,
assuming  optimal moisture, but belowground carbon is  not all readily oxidizable.  A large proportion of
grassland  soil carbon pools has been classified as recalcitrant to decomposition, with a turnover time of
up to 1000 years (Parton et al. 1988).  Forest soils are also characterized by a large  recalcitrant fraction,
and lags in the  reduction of soil organic matter due to warming may be on the order of hundreds of
years.  Continued research on the mechanisms and modeling of soil organic matter turnover is needed.

These analyses also ignore the large influence that humans will have on rates of vegetation change.
Currently anthropogenic factors use or co-opt approximately 40% of the  earth's potential net primary
production, and over large areas the productive capacity of the land is being reduced (Vitousek et al.
1986).  Land  used for human purposes will probably not be readily converted to the  high carbon-storage
vegetation types predicted in this analysis. However, humanity may be able to promote changes that
would favor carbon sequestration.  Management options are available  to reduce carbon fluxes to the
atmosphere and maximize carbon sinks associated with the terrestrial  biosphere. Their implementation
will require an understanding of current climate-vegetation-soil relationships  and of locations where the
climate will favor specific vegetation types in decades to come.

3.1.3  Conclusions

These equilibrium analyses suggest that the changes in  terrestrial carbon pools induced by climate
warming could  provide a moderate negative feedback to that warming. That is, as global warming
increases, carbon storage  will increase and the overall increase in atmospheric C02 concentration will be
moderated. The greater the increase in precipitation for the low latitudes, and thus in tropical wet forest
area, the stronger that feedback will be.  These analyses do not consider: 1) potential short-term fluxes
of carbon  due to changing disturbance regimes; 2) the relatively slow  response of belowground carbon
                                               56

-------
pools to climate change; or 3) potential enhancements or reductions of carbon storage due to
management practices.

32     Biosphere  Feedback   During Climate Change

The previous section discussed the impact on the terrestrial carbon pool of a redistribution of global
vegetation. That analysis focused exclusively on comparing the carbon stored in the terrestrial
biosphere before and after a climate  change brought about by a doubling of atmospheric COa assuming
vegetation was in equilibrium with climate after the climate change. This section expands on that
analysis by estimating the transient (time-dependent) effects of changing vegetation on the aboveground
terrestrial carbon pool and carbon fluxes to the atmosphere.

Consider first'the vegetation dynamics induced by a double-CO2 atmosphere climate change. According
to Leemans (1990)  and Smith et al. (submitted) (see Section 2.4 for a summary of Leemans' results),
there will be a poleward shift in the extratropical forests and expansion of the tropical forests using the
Holdridge life zone model. The total  area of extratropical forests is not expected to change a great deal
under a double-C02 climate, with estimates ranging from 83% (GFDL)  to 102% (GISS) of the current
area. However, 46% (OSU) to 88% (UKMO) of the southern extent of the extratropical forests would be
displaced toward the poles. Estimates of tropical forest expansion are similar, ranging from 116%
(GISS) to 121% (UKMO) of their current  area. The Box biosphere model (Box 1981) produces similar
results in the extratropics, but it predicts a decline in tropical forests due to an upper thermal limit (Smith
personal communication).

During the simulated vegetation change  from current to double-C02 climate conditions, three processes
will dominate the impact of these changes on global carbon dynamics:

1)     As forests shift poleward, their equatorward extents should go through a decline with concurrent
       carbon release (see following discussion). This is a critical assumption that requires further
       research to confirm it (Solomon 1986, Smith and Tirpak 1989,  Neilson et al. 1989, Bonan et al.
       1990, Overpeck et al.  1990).

2)     New vegetation will advance and regrow in recently opened landscapes (defined here as the
       "change" zone), and the new vegetation will begin sequestering carbon.
                                              57

-------
3)      As the tropical forests expand (according to Smith et al. submitted), they will sequester more
        carbon.  This assumption also requires further research.

The balance of these three processes, one releasing and two sequestering carbon, will determine the net
transient source or sink impact of the biosphere on the atmospheric carbon concentration.  In a study
conducted at ERL-C, a simple model of these three rate processes was developed, and a preliminary
sensitivity analysis of the key parameters was performed with respect to atmospheric loading of CO2
This section presents the results of this analysis for forest systems (as described in Section 1.4.2, forest
systems contain the bulk of the carbon in the terrestrial biosphere and, for simplicity, they were the only
biomes considered in this analysis). The analysis was restricted to aboveground organic matter
dynamics.  Belowground carbon  pools are also expected to change in the same direction as the
aboveground change, but the potential decline rate would be slower than that in the aboveground
component. The conclusions, based solely on aboveground carbon, should thus be conservative with
regard to transient carbon loading of the atmosphere.

The most critical process in the analysis is the removal of the extant forest from the zone of change.  If
this were to occur through competitive displacement, the release of carbon would be slow or
insignificant.  However, as forests are stressed climatically, which is probable under the range of current
GCM  climate projections (Neilson et al. 1989), their decline stage would likely be mediated by
catastrophic disturbance, particularly by fire (Bonan et al. 1990, Neilson et al. 1989, Overpeck et al. 1990,
Winjum and Neilson 1989).  Arguments exist for both processes, i.e., competitive displacement and
catastrophic decline.

The boundary between the boreal and temperate forests is apparently controlled by cold temperature,
the -AO'C isotherm (Burke et al. 1976). If this climate boundary shifts north, one might anticipate a
simple competitive replacement of boreal by temperate species with no increase in disturbance or
significant change in carbon pools.  However, water availability must also be considered. Within their
boundaries, most of the world's biomes may  be transpiring most of their available soil water resources
during each growing season (Woodward 1987, Neilson et al. 1989).

More  water could become available through altered potential evapotranspiration, rainfall, or water-use
efficiency (WUE) of the vegetation.  In theory, if more water were available, more leaf area would be
generated, thus water would be transpired. If a landscape contains too much leaf area, water is
transpired too rapidly and is depleted before the end of the growing season, causing drought-induced
vegetation decline that reduces leaf area.  As a result of these feedback processes, regional biomass
                                               58

-------
levels should be in rough equilibrium with the regional water balance.  A rapid change in that water
balance, through increased potential evapotranspiration (PET), decreased rainfall (P), or increased deficit
(PET - P), should produce drought-induced vegetation decline.

Thus, global increases in temperature may allow some boundaries to shift. However, the concomitant
rise in PET could deplete soil water reserves, causing widespread drought-induced decline (Bonan et al.
1990, Neilson et al. 1989, Solomon 1986, Smith and Tirpak 1989). Marks (submitted) calculated PET
under current and double-CO2 conditions for the continental United States. Large increases in PET are
projected to occur over much of the United States. Anticipated increases in rainfall in northern latitudes
would not likely be sufficient to balance the increased PET (Bonan et al. 1990, Grotch 1988).
Catastrophic fires, induced by the increasing frequency of extreme events such as drought and high
wind (Bonan et al. 1990, Neilson et al. 1989, Overpeck et al. 1990) are assumed here to be the dominant
mechanism removing forests from the change zone.  Drought stress could also increase the intensity of
pest infestation, which also usually results in catastrophic fire. These generalizations may not hold in the
far north where some biomes may be limited more by energy than by water (Woodward 1987).

3.2.1  The Model

The rate of C02 loading  to the atmosphere  during the decline period will be determined by: 1) the
amount of direct combustion; 2) the efficiency of combustion; 3) the amount of residual, dead organic
matter; 4) the rate of decay of that organic  matter; and  5) the amount of carbon lost to surface waters.
The simplest assumptions are that the annual rate of area burned will track the rate of climate change
(likely after some lag period), and that any carbon shunted to surface  waters will be rapidly released as
CO2.  If a doubling of C02 is expected to occur in 50 years, then the entire change zone would burn off
in a 50-year period. This appears to be an  extreme assumption, invoking a massive and continuous
amount of wildfire affecting much of the world's forested area. The alternative, If the assumption of
extreme drought is accepted, is that the rate of forest incineration will  lag behind the rate of climate
change.  Ecosystems would then be subjected to ever-increasing levels of drought, producing
considerable amounts of dead biomass.  It  is thus an equally extreme assumption to presume that vast
amounts of standing dead  biomass would not eventually burn off. The prospect of climate change
forces upon us the examination of almost unimaginable ecological scenarios.  The intent of this exercise
is to attempt to bound the  extent of ecological change and the resulting biospheric feedback to the
atmosphere. Through a focused sensitivity analysis of the important model parameters, including the
rate of decline, we hope to help set priorities for future research.
                                               59

-------
Carbon loss from catastrophic forest dieback was modeled as a combination of two processes:  the
fraction directly incinerated followed by exponential decay of the remaining large fragments.  Imagine
that the current climate over a biome  is a veil. As the climate changes, the veil shifts slowly toward the
pole, revealing a constant fraction of the  eventual change zone each year.  If NR is defined as the total
aboveground organic matter in the change zone that will be Incinerated over n years, then the fractional
organic matter available for combustion each year is as follows:

                                          ND  = —5                               (3.2-1)
                                            v     n

where Np is the aboveground organic matter per land parcel that will eventually be lost.  A new parcel
becomes susceptible each year.  The C02 emissions from a single parcel over time are calculated as
follows:

                                                   l-f)e-*fc                     (3.2-2)
where Ei is the accumulated amount of carbon emitted from parcel / at time t from the combination of:
1) direct incineration of a fraction, f, of aboveground organic matter when the parcel became
susceptible; and 2) subsequent decomposition at rate /(/yr of the remaining aboveground organic matter
on that parcel. The remaining  organic matter is assumed to be large woody debris with 100% mortality
of aboveground vegetation.

The temporal release of C02from each parcel will be identical to all other parcels.  Parcel emissions are
simply initiated sequentially,  one each year.  Thus, one only need calculate the time-dependent
emissions from one parcel.  The total regional emissions at time f can be calculated from the single
parcel by summing emissions from time 1 to time f, for t less than or equal to n, and from (t-n) + 1 lot,
for f greater than n.  The accumulated regional release of C02 over all parcels up to time t, EK(t), is thus
modeled as follows:
                                        c
                                             Ei(j) ,      fox tzn
                                                                                  (3.2-3)
                                             Ei(j) ,      for t>n
Equations 3.2-2 and 3.2-3 form a hierarchical set with Equation 3.2-2 simulating the carbon release from

                                             60

-------
a parcel and 3.2-3 the accumulated release over the region. The parcel-level process rates and the
regional rate of new parcels becoming susceptible are thus combined for the regional emissions
estimate.

Once burned, a parcel becomes available for regrowth and the gradual accumulation of more organic
matter.  Net Ecosystem Productivity (NEP) is the total accumulation of organic.matter per year
accounting for carbon losses due to both autotrophic and heterotrophic respiration (Peet 1981, Sprugel
1985, Waring and Schlesinger 1985, Shugart 1984, Whittaker 1975). In  general, NEP is low after
disturbance, gradually increases to a maximum during mid-succession, and approaches zero as the
ecosystem approaches a carrying capacity in late succession. The sigmoid curve produced by the
logistic equation was used to model this process:
                                     dt       i    K.
                                                                                   0.2-4)
where dN^/dt is the aboveground NEP on parcel /', KJs the carrying capacity of carbon in units of
kg/m2 on parcel /, and A/Js the organic carbon accumulated at time f on parcel /'.  In demography, r is
the intrinsic rate of increase (birthrate - deathrate). In our use r does not carry an exact definition, but
would represent an intrinsic production capacity (Emanuel et al.  1984, Harvey 1989). In the early stages
of succession, actual aboveground NEP is much lower than r, due to a lack of vegetation capable of
fixing carbon, i.e., modulation by A/i(Peet  1981, Waring and Schlesinger 1985).  In the later stages of
succession, aboveground NEP is low because decomposition nearly balances carbon fixation, i.e.,
modulation by K.

Traditionally, carbon fixation and decomposition processes would be modeled explicitly to calculate
dNj/dt (Emanuel et al. 1984, Harvey 1989).  However, observations of aboveground  organic matter
changes are direct observations of net carbon gain (NEP, aboveground), rather than observations of r.
Therefore, observed literature values of aboveground NEP were substituted in Equation 3.2-4 for
dNi/dt, and were assumed to be the maximum values. The term r is then calculated using Equation
3.2-4 from the maximum observed  dN±/dt (NEP), where NL  =K^2 at  maximum NEP  (Ricklefs 1973)
                                              61

-------
and is substituted into Equation 3.2-5 below.3 Again, this can be viewed from the theory of hierarchies
(Allen and Starr 1982, O'Neill et al. 1986), whereby NEP is an ecosystem property constrained by the
external environment and following a simple sigmoid curve (Sprugel 1985).  Thus, for our purposes,
there  is no need to know or simulate directly the production and respiration of carbon, only the net
process of organic carbon accumulation. The accumulation of organic carbon over time on parcel / can
then be calculated by integrating Equation 3.2-4 as follows:
                                                                                    (3.2-5)
Integration of Equation 3.2-4 introduces a new parameter, Ni0, the initial post-disturbance organic
carbon on parcel /.  As with regional emissions (Equation 3.2-3), the regional accumulation of organic
carbon, WR, is modeled hierarchically using an initial 'growing window' and then a 'moving window' as an
accumulator over a  single parcel through time, as follows:
                                         t
                                              Ni(j) ,      for  tzn
                                                                                    (3.2-6)
                                                            for  t>n
In summary, parameter estimates must be supplied for dN^/dt and K and substituted into Equation
3.2-4 which is solved for r. This completes the input parameters for Equation 3.2-5 which is iterated
through time (f) for a single parcel.  The resulting sigmoid curve is subjected to an 'accumulator window'
to estimate the regional time series of carbon accumulation NR

Tropical expansion was modeled as an increase in the carrying capacity, K, over the area into which the
forests expanded according to Equations 3.2-4 through 3.2-6.  The current total organic carbon of
tropical forests (including moist, dry, and seasonal tropical forests) was parceled by n years.  This
produced an initial  starting organic carbon (Ni0) per parcel of about 10 kg/m2, which was less than
    3  Observed values of NEP are more likely an average over a bell-shaped trajectory of the true NEP
through time.  Therefore, actual maximum values for NEP are likely somewhat higher than the  average
observed values.
                                              62

-------
moist tropical forests (15 kg/m2), about the same as tropical seasonal forests, and greater than that of
dry tropical forests (7 kg/m2, Table 3.1-2). The total increase in aboveground organic carbon in tropical
forest under a double-C02 climate was parceled uniformly over the duration of change and added to the
initial organic carbon per parcel. A/i0. That is,  each year a new parcel of tropical forest became available
for additional growth, its carrying capacity was raised, and growth ensued until the new carrying
capacity was reached.  Since the tropical A/i0 is rather high, the initial calculated r was also high. Thus,
new tropical growth was always begun to the right of the inflection point on the sigmoid curve produced
by Equation 3.2-5.  Although dN^/dt was set artificially high for the tropical forests so that the
calculated r value would be in a realistic maximum range, it declined rapidly as the new carrying
capacity was approached.

Expansion of tropical forests is assumed to occur through competitive displacement or augmented
growth of existing tropical forests. The expansion is predicted  by the Holdridge model primarily because
of large increases in rainfall that are  not offset  by increased PET.  Thus, as discussed earlier, more
available water should result in increased biomass and leaf area of the existing ecosystem (Woodward
1987).  We infer no catastrophic declines or fires in the tropics. Therefore, Equations 3.2-2 and  3.2-3 for
carbon loss only pertain to the extratropical forest dynamics.

Parameter values and initial conditions were determined from the literature. All parameter values were
subjected to a sensitivity analysis with regard to the magnitude of the extratropical emission  pulse, both
with and without tropical expansion.

Emissions of CO2from declining forests through drought and fire (Equation 3.2-2) require two parameter
estimations. The percentage of organic carbon released through direct incineration, /, was set at 10%
(AuClair 1985, Fahnestock and Agee 1983).  The decomposition rate, k, was set at 0.05/yr (AuClair 1985,
Edmonds 1987).

Both the tropical and extratropical growth rates (NEP) were set somewhat higher than the highest
literature values. Recall that these growth rates are used as the maximum that could occur in
calculating  r. The  observed  NEP values are usually linear averages of organic carbon gained over some
period of time.  That is, if a system gained x amount of organic carbon over 100 years, annual NEP
would be calculated as x/100.  Of course, if the true trajectory over 100 years were a sigmoid curve, the
early and late succession NEP rates would be lower than average, whereas NEP during  mid-succession
would be higher than average. Thus, the choice for the maximum values of NEP was set higher than
                                               63

-------
observed average values. The NEP values were also set high to ensure that our estimates of a C02
emission pulse were conservative.

All temperate and boreal forests have been aggregated, and average parameter values for the aggregate
have been imposed (Table 3.2-1).  Extratropical maximum aboveground NEP was set at 2 t/ha/yr of
organic carbon (AuClair 1985, Bormann and  Likens 1979, Waring and Schlesinger 1985). Aboveground
extratropical forest carbon averaged about 10 kg/m2 (Section 3.1), and the area was set at 28.14 million
km2 (Section 2.4, Table 2.4-2) for a total aboveground carbon storage of about 306 Gt, compared with
an estimated 283 Gt (Whittaker 1975, Olson et al. 1983).  The initial organic carbon from which regrowth
begins was set to 0.4 kg/m2, equivalent to two years of maximum NEP. There are very few data with
which to estimate initial post-fire organic carbon, but this number appears to agree with current
information (AuClair 1985, Johnson and van Cleve personal communication).  Sensitivity analyses were
performed on all  key parameters.

Several Holdridge tropical categories were aggregated, including dry, seasonal, and rain forests (see
Section 2.4, Table 2.4-2).  These different types range from relatively low to quite high standing organic
carbon, with an average value of about 10 kg/m2.  So as not to impose a low  initial growth rate, A/0 was
set to about 10 kg/m2, upon which was grown approximately 1.5 kg/m2, depending on the scenario.
Maximum tropical NEP was set to 16 t/ha/yr of organic carbon. Our preferred value for tropical NEP  is
about 5 t/ha/yr (Jordan 1989, Uhl and  Jordan 1984). However, since A/i0 was set quite high (above the
inflection point on the logistic curve), the  calculated NEP (dWi/dt from Equation 3.2-4) was much less
than the potential maximum (at the inflection point).  An artificially high value of 16 provided an initial
NEP that ranged from about 8 to 12 t/ha/yr, depending on trie  scenario.  Although this is still quite high
compared with the preferred value of 5 t/ha/yr, tropical NEP declined rapidly to less than 5 t/ha/yr after
about 10 years, as the forest organic carbon approached carrying capacity. The sensitivity analysis for
this parameter was, therefore, extended to an unrealistically broad range of values from high to low.  It is
important to enforce some duration of high aboveground NEP on tropical forests. Equation 3.2-5
produces a rapid decline in  dNi/dt due to modulation by the  carrying capacity, K,  under an
equilibrium double-CO2climate.  We do not expect climate to equilibrate at double CO2 but  rather to
continue warming. Thus,  tropical expansion  could continue at a high rate for some time.
                                              64

-------
3.2.2  Results
The terrestrial carbon emission trajectories under the four climate scenarios with baseline initial
conditions and parameters (Table 3.2-1) are presented in Figure 3.2-1. The UKMO scenario is the most
extreme in extratropical change and the least extreme in tropical change.  The OSU model exhibits the
least extratropical change and the greatest tropical change.  The GFDL and GISS results'are
intermediate, with GFDL being the more extreme of the two in the extratropics.  The accumulated
emission pulse from the extratropics ranged from about 60 to 130 Gt of carbon released, or about 1 to 2
Gt/yr.  The maximum accumulation of C02 in the atmosphere occurs at about 60 years, after which the
pulse declines.  If the Box model is correct and tropical forests decline under double-C02 conditions, the
tropics would become an additional source and would increase the size of this carbon pulse.
Table 32-1.     Initial Conditions and  Parameter Values  (all units are for aboveground   organic
               carbon)
Fraction of organic C burned
Decay Rate, post-burn organic C
Duration of change
NEP, Extratropical
Initial Extratropical organic C
NEP, Tropical1
Initial Tropical organic C
1
k
n
max dN/dt
NiO
max dN/dt
NiO
10%
0.05 /yr
50 years
2 t/ha/yr
0.4 kg/m2
16 t/ha/yr
9.9 kg/m2
JThe preferred value is about 5 t/ha/yr (Jordan 1989, Uhl and Jordan 1984).

                             UKMO,  Transient Carbon  Pulse
                         o
                        
-------
            OSU, Transient Carbon Pulse


y-x
o
c
o
J2
i»
O
o
V
>
"5
^
£
3
O







JUU
250
200
150

100

50

0

-50


-100
1 4' . - - ' ^>Sv^^l' •-..
~
^S** ''••.
^^ '"••-...
Global Net
Carbon Pulse
"
i.i.i.

0 50 100
Years
i i
•
Extratropical
Forest
Regeneration

Tropical
Expansion
-



^



i . i

150 200

Figure 3.2-1 (b)
            GFDL,  Transient  Carbon  Pulse

        o
        o
        _o
        3
        E
        D
        O
 300

 250

 200

 150

 100

  50

   0

 -50

-100
      r_       r
Extratropical
 Forest
 loss

  Extratropical ,'
    Emission  /
     Pulse  /
                                         Extratropical
                                         Forest
                                         Regeneration
                                          Tropical
                                          Expansion
                      Global Net
                      Carbon Pulse
                         50
                      100

                     Years
                    150
200
Figure 32-1 (c)
                               66

-------
                GISS,   Transient   Carbon  Pulse
.JUU
250
x-, 2°°
~ 150
o
1 100
o
I 50
E °
5 -50
-100

—\ 	 i 	 1 	 1 	 1 	
. Extratropical .Extratropical
_ Forest Extratropical Forest
loss . Emission Regeneration
\. / ,'' Tropical
\ / /' Expansion
: ' ^^^^
Global Net "--...
Carbon Pulse """ 	
1.1.1,1,1
0 50 100 150 200
Years
Figure 3.2-1.    Transient carbon pulse with the a) UKMO, b) OSU, c) GFDL, and d) GISS GCMs.
              The trajectory of forest loss  (solid line) is the total aboveground  organic carbon
              remaining in the original extratropical  forests after annual fire and subsequent
              decay of remaining biomass according  to Equations 32-2 and 32-3. Forest
              Regeneration  is the cumulative  gain of extratropical aboveground  organic carbon,
              post-fire, according to Equations 32-5 and 32-6.  The Emission  Pulse is the
              extratropical pulse of carbon produced  by the imbalance between forest loss and
              forest regrowth.  Tropical Expansion is the cumulative  growth  of tropical forests
              above current levels  of aboveground  organic carbon according to Equations  32-5
              and 32-6.  The  Global Net Carbon Pulse is the Emission Pulse diminished by the
              Tropical Expansion.  All values  on the vertical axis are cumulative over time.
                                           67

-------
According to the Holdridge model,  which predicts tropical expansion, approximately 60 Gt (UKMO) to
125 Gt (OSU) carbon are sequestered over time (Figure 3.2-1, indicated by the negative global net
carbon equilibrium). The large expansion of the tropics and the small extratropical change combined in
the OSU scenario to completely offset the extratropical emissions of C02  However, under the UKMO
scenario, the combination still produces a net pulse of about 75 Gt, that is, more than 1 Gt/yr over
about 50 years.

3.2.3  Sensitivity Analyses

The magnitude of the extratropical carbon pulse, both in isolation (Figure 3.2-2) and as diminished by
tropical  expansion (Figure 3.2-3), was examined over a range of parameter values to determine
parameter sensitivities.  Each parameter was tested individually with the remaining parameters set to the
values in Table 3.2-1. The magnitude of the carbon pulse was the only variable recorded. The duration
of the pulse near peak values varied considerably under the different parameter values, ranging from
about a decade (Figure 3.2-2) to a century or more, and it should be more thoroughly examined. The
extratropical pulse was analyzed separately from the global net pulse (which includes a tropical
expansion) in order to estimate the sensitivity of the parameters under a tropical decline as projected by
the Box model.  We did not examine the carbon dynamics of a tropical decline, which presumably would
exacerbate the extratropical carbon pulse.  Recall that in the  extratropics the Box and Holdridge models
perform about the same.
                                               68

-------
c- 25° r
o
~ 200 -
0)
tn
"a 150 -
QL.
1 1°° '
'55
w RO -
E -
u 0 -
CM
o
o -50 L
1 ' 1 ' 1 ' 1
-
UKMO J
T-x-^^^31---^ •
o-o-o UGFDL
9 :9z«^«=« -
GISS osu
1 I 1 1 1 1 1
0 10 20 30
250 -
200 •
150
100
50
0
50
1 I • • ' i ' • ' i • ' • i
UKMO T
J^GFDL:
— / ^^^9
fc'^1 '
ft GISS o^Tj -
i*y uou
, i , , , i , , , i . , , i






w w
0 4 8 12
Percent Burned Decay Rate (*/yr)
O C rt
C? 250 9sn
o
""" 200
0)
CO
"5 150
0-

I 10°
'In
- 50
.- OU

u o
CM
^^
o -50

i • i • i
UKMO
i /
/
5.
\ HFDT
s\ *2?UL
Jk ^W^"^
• /v\OC H
-/ ^t^5^^
.GISS \^f^S^a .
\ **=w
xosu
.
1,1,1
50 150 250
^ou
200

150


1 00
1 W W
50

0
C O
i ' i • i ' i
" Jx UKMO
: ^X^-GFDL :
^\\
- * NX
• r^NA XX GIQS
v^* O\~>~^
- v^. y^
^0^5^
osu ^-f ^5 -
w

" "
t , i , i , i






•




-50 • 	
01234
Decline Duration (years) N^p Extra-trop., t/Ha/yr
<-*- 250 • i -i -i—r '-r i-'-i •!—- r- 1 — i
^/ ^.JW v ' 1 ' 1 ' 1 ' 1 ' 1
O
~ 200
0)
(0
"3 150
CL
1 1°°
"35
.2 50
* 0
CM
ri _RA
Y ' i ' i ' i ' i ' i
-% 	 UKMO
• /^
-€D /GFDL
•S^
rv^ ^-GISS
^NOv^^
-«%
T^B
\ *^5.
osu •• 	 «

I , 1 , 1 , 1 , 1 . 1
•

•

-
-
•

                           2345

                 Initial  Biomass,  Kg/m2

Figure 32-2.    Sensitivity of extratropical  emission  pulse to key parameters.  Baseline parameter
               values  are in Table 32-1.
                                               69

-------
O ^3U
| 200
3
a- 150
CN
8 10°
° 50
o
o 0
_w
(U
i ' i • i • i
- -
-
1 GFDL ™M°j
1 Jio-o— ^o-"""0 -
IjloO _____0
— y~y~y-~ ~y 	 v -
OSU
i , i , i , i '
ZJU
200

150
100
50

.0

^n
1 i • ' • i ' ' ' i ' ' • i
-
UKMO -
'_ GFDL V J
- /Q-^GISS -
X _ — «^— •• " " "
m^9— v 	 "^
OSU -
, i , . , i , . , i , , . i
0 10 20 30 0 4 8 12
Percent Burned Decay Rate (%/yr)
s 25°
S" 200
i3 150
CM
8 10°
N_X
"5 t;n
_Q ^U
_o
o 0

"S
z -

i ' i • .1
-
_ _
.
UKMO
'
: °^s^|^_ :
— /^fi —^— ^r ^^^ ^P - i- )H| IHI^ )•( __
• GISS XOSU
1,1,1
50 150 250
250
200
150

100

50
0

50

i i i i
-
T UKMO
\
• o T

•GISS°\>
: *~*-^5=8 :

OSU
1 1 II
01234
Decline Duration (years) N^p Extra-trop., t/Ha/y
5 250
£ 20°
2
Q- 150
CN
8 10°

1 50
o
o 0
"S3

i ' i ' i ' i ' i • i
• -
• T 	 UKMO
"i
-T GFDL
•$*/ GISS

. ^^ir
-^'•^• — • 	 •-
i°?M , i , i , i , i
250
200

150
100

50

0
50

1 ' 1 ' 1 ' 1 ' 1
-
* "
— —
lj^KM° GFDL -
' ' O " ^ — ^/"^ ^ — ^
-•^ "~"° — o--g — o~
• ^^^* — •• 	 .JSISS
~ osu~~ ~ ~~ ~*"
1,1,1,1,1
 012345
Initial  Biomass,  Kg/m2
                                                           5    10   15  20   25
                                                         NEP  Tropical. t/Ha/yr
Figure 3.2-3.   Sensitivity  of global net carbon  pulse to key parameters.  Baseline  parameter
              values are in Table 32-1.
                                            70

-------
The size of the global carbon pulse was rather insensitive to variation in the percentage of aboveground
organic carbon initially burned (Figures 3.2-2a and 3.2-3a).  However, the decay rate of the remaining
organic carbon is a critical parameter. The range of literature values is about 0.03 to 0.05/yr.  In that
range the carbon pulse can vary by about 40 Gt (Figures 3.2-2b and 3.2-3b).

The duration, or rapidity, of change is a very important parameter.  The slower the change (i.e., the
greater the duration), the more capable is the biosphere of keeping pace with the change.  The size of
the carbon pulse is very sensitive to small changes in duration over the range of 50 to 100 years.  If the
tropics expand  (Figure 3.2-2c), this parameter is less critical than if the tropics decline  (Figure 3.2-3c).
Under the UKMO scenario, the variation in the size of the carbon pulse with a shift from a 50- to 100-
year duration is about 35 Gt with tropical expansion and over 50 Gt with a tropical decline.

A 50% increase in maximum NEP, from 2 to 3 t/ha/yr of carbon, is about equivalent to a 50-year
increase  in the duration of forest decline; that is, about a 35-Gt decrease in the pulse with tropical
expansion (Figure 3.2-2d) and an over 50-Gt decrease without tropical expansion (Figure 3.2-3d).

The magnitude of the carbon pulse was very sensitive to the amount of post-fire organic carbon that
represents the new flush of aboveground vegetation on which  all future growth depends (Figures 3.2-2e
and 3.2-3e). The initial value of 0.4 kg/m2of carbon,  representing about two years of maximum
productivity, is in an extremely sensitive range for this parameter, but this value  has been observed
(AuClair 1985).  An increase in this parameter to 1  kg/m2, about 10% of the pre-fire organic  carbon,
would drop the carbon pulse by about 45 Gt with tropical expansion and by over 50 Gt without tropical
expansion. This parameter is very sensitive at low values because it places the  starting point  on the
logistic curve further out on an asymptote, effectively producing a large lag before rapid growth  ensues.
AuClair (1985) examined a sequence of boreal forest stands of varying age post-fire and found an  initial
organic carbon quite close to our parameter value. He also observed that significant tree growth and
organic carbon accumulation required about three decades to begin.  Our model produces  very similar
behavior, with regrowth organic carbon increasing rapidly after about three decades (Figure 3.2-1).

If the tropics expand, the rate of tropical growth (NEP) is sufficiently high for further increases in growth
to have little effect on the size of the  carbon pulse (Figure 3.2-2f).

If all  parameter values were to combine for the most benign yet still reasonable  combination, the carbon
pulse for the most extreme scenario (UKMO) is much reduced  (Table 3.2-2). The parameters were set
to 100 years duration, 3 t/ha/yr for extratropical NEP, 1 kg/m2 initial, post-fire organic carbon, 5% initial
                                               71

-------
incineration, and 0.05/yr decay rate. A high NEP was used for the tropical forests.  Under this
combination, the global net carbon pulse with tropical expansion was reduced from 68 Gt to 1 Gt of
carbon, and with tropical decline (i.e, extratropical pulse only) from 131 Gt to 16 Gt carbon (Table 3.2-2).
These values could be considered a conservative boundary on the transient behavior of these forests.
The actual transient response would not likely be this conservative. The same parameters  with the most
benign model (OSU) produced similar  results. Under severe values for the parameters (duration 50 yrs,
1.5 t/ha/yr extratropical NEP, 0.4 kg/m2 initial, post-fire organic carbon, 20%  incineration, 0.05/yr decay
rate, 8 t/ha/yr tropical NEP), the carbon pulse could be as high as 166 Gt (almost 3 Gt/yr) or as low as
15 Gt, depending on the scenario (OSU or UKMO) and on whether the tropics expand or contract.
Table 3.2-2.    Range  of CO2 Cumulative  Pulse  to the Atmosphere  under Benign  and Severe Model
               Parameter Values,  Assuming  Tropical Expansion  (Global) and No Change  in
               Tropical Sequestering   (Extratropical)
Parameter
Values
Benign
Severe
Carbon Pulse (Gt)
OSU UKMO
Global Extratropical Global Extratropical
0 8 1 16
15 86 111 166
3.2.4  Discussion

Depending upon the vegetation scenario (i.e., either expanding or contracting tropics) and the climate
scenario (i.e., either more like OSU or more like UKMO), the range of the carbon pulse from
aboveground organic carbon only could be anywhere from 0 Gt to about 130 Gt, under the standard
parameter values (Table 3.2-1). These estimates are conservative in that consideration of the forest floor
and soil organic carbon pools would likely raise the estimate considerably.  Fires that penetrate the duff
in the boreal forest can release as much as 2 kg/m2 of carbon (Johnson and van Cleve personal
communication). Additional decay from higher soil temperatures and drier sites could elevate the
numbers even more.

The carbon pulse would potentially peak in about 50 to 60 years under the 50-year scenario.  In general,
there is a tradeoff between the size of the peak and its duration. We did not examine this tradeoff in
detail.  However, under parameter values that extended the time frame, the pulse was also extended,
                                               72

-------
albeit with a correspondingly lower amplitude.  The residence time of the excess CO2in the atmosphere
is likely to be of importance to global climate dynamics.

An attribute of the model that is clearly artificial is the implication of an equilibrium endpoint.  Double-
C02 scenarios are only convenient mileposts with which to gauge relative magnitudes and rates of
change.  The double-CO2 point should actually be viewed as a point on a trajectory that may continue to
increase.  More important, the biome response to double-C02 is potentially extreme enough to produce
an almost 90% spatial displacement of extratropical forests under some scenarios.  If the biogeographic
response exceeds that discussed here, the change zone would extend through the entire current
distribution and beyond into the change zone of the neighboring biome.  This would bring on a whole
new set of dynamics.  Solomon (1986) predicted just such shifts with successional  communities being
repeatedly reset to the early succession of a different forest type. Since the model presented here
terminates with an equilibrium condition, interpretation of the transient dynamics should be restricted to
the period encompassing the ascending phase of CO2 emission from the extratropics. After that the
dynamics become increasingly dominated by the ecosystems  approaching carrying capacity.

The direct influence of higher C02 levels on plant processes could affect these ecological scenarios.  As
will be discussed in  Section 4, the direct effects of C02 include accelerated growth and increased WUE
(Strain and Cure 1985).  Faster growth rates have been examined as a part of the sensitivity analysis.
Due to currently high productivity rates, virtually nothing would be gained, with regard to minimizing the
carbon pulse, by increased  growth rates in the tropical forests (Figure 3.2-2f). Increased growth rates in
the extratropics could, however, offset the impacts of climate change to some degree (Figures 3.2-2c
and 3.2-3c).

The most important direct effect of increased C02 could be increased WUE. The phenomenon has been
observed at the level of the  whole plant, but it  has not been well explored at landscape scales  under
limiting water conditions (see Section 4.2). At  large spatial scales, radiation may be more important than
stomatal processes  in the control of landscape water balance  (Cowan 1965, Federer  1982, Jarvis and
McNaughton 1986). These  uncertainties require further research.  However, should increased WUE
prove important at large scales, it could mitigate the decline process directly and potentially preclude
any significant dieback and  disturbance regime.  For these benefits to accrue, the physiological
response must precede the gradually increasing climate stress.  Since both increasing stress and
increasing WUE are consequences  of a gradual increase in CO2, it is uncertain which effect, biotic or
climatic, would manifest itself most rapidly. If increased WUE  lags behind climate-induced stress, it may
occur too late for significant benefit.
                                               73

-------
If a maximum pulse of carbon is considered to be about 130 Gt over about 50 years, annual emissions
of carbon could increase by about 2.6 Gt per year.  This would be about a 37% to 43% increase in
annual emissions of carbon, some of which would be sequestered by the oceans and the terrestrial
biosphere (Keeling et al. 1989, Harvey 1989, Tans et al. 1990).  The potential effect of this positive
feedback to the climate system would be to increase the rate of change of climate, which in turn could
further stress natural ecosystems.

The process of CO2 release from the biosphere through fire is,  of course, not completely efficient.
Organic carbon will be oxidized to C02, carbon monoxide, and other organic compounds, depending on
the conditions of combustion (Gofer et al. 1989).  Incomplete combustion could reduce the size of the
C02 pulse, but it could also produce other consequences. Carbon monoxide is important in
tropospheric chemistry relative to ozone, methane, and hydroxyl radical (Crutzen  1988).  Injections of
particulates from fires over a prolonged forest decline could also produce feedbacks to the climate
(Harvey 1988). Nitrous oxide released from burning green leaves  could act as an additional positive
feedback to climate warming (Cofer et al. 1989, Westberg et al. 1981, Stith  et al. 1981).

In summary, the transient response of the biosphere to climate change could produce a positive
feedback to the changing climate by mobilizing terrestrial carbon at a faster rate than it can be
withdrawn by emerging vegetation. A significant uncertainty in this scenario is the response of the
tropics, which could expand or contract. These uncertainties exist for both the biosphere modeling and
the climate scenarios.

Our results suggest that the transient dynamics of the biosphere could be quite different from those
inferred from equilibrium considerations.  Recall that under equilibrium comparisons, a warmer, wetter
world tends to support more aboveground biomass and is therefore considered to be a net sink for
carbon (Section 3.1).  Apparently the only circumstances under which the terrestrial biosphere might
behave as a carbon sink under transient conditions  are when either the rate of climate change is slow or
the magnitude of the change is relatively small, or both.  If either the rate or magnitude of change is
sufficiently extreme, ecosystem rate processes are likely to be exceeded, resulting in catastrophic forest
decline in the extratropics. Double-C02 equilibrium  scenarios are  only considered as a convenient
milepost  on a  continuous trajectory of CO2 increase. Actual atmospheric equilibrium conditions for CO2
are not expected to occur for some considerable time after doubling. Therefore, even our transient
analyses that carry the biosphere to an equilibrium are conservative with regard to eventual changes that
could occur. The occurrence of the likely driver of the transient decline of forests, drought followed by
fire, must be a high priority for future  research.
                                               74

-------
Bonan et al. (1990) completed a sensitivity analysis of a detailed boreal forest computer model.  They
found that the forests were very sensitive to increases in PET, which produced soil drying and greatly
increased the probability of fire. The stress due to increased PET could be offset if precipitation
increased sufficiently. However, projected increases in precipitation by the GCMs in boreal regions are
not likely to sufficiently offset the rather larger increases in PET due to large increases in temperature
(Bonan et al. 1990, Grotch 1988).  These results are consistent with the dynamics  presumed here and
elsewhere (Solomon 1986).

The most sensitive parameters in the model presented here were the rates of aboveground NEP in
extratropical forests, the decay rate of post-fire debris, and the amount of initial live organic carbon post-
fire. The form of the logistic equation  may not be appropriate for the earliest stages of ecosystem
recovery.  Measuring early successional rates of NEP should be a  priority research area to refine the
form of the equation. The potential mitigating influence of direct CO2 effects and the timing of those
effects relative to the timing of potential drought should also receive attention. It is apparent from these
analyses that the potentially rapid rate of climate change is perhaps the overriding factor producing
catastrophic ecosystem responses with ensuing positive feedbacks to global warming.  This work
underscores the importance of slowing the rate of climate change  as recommended by Lashof and
Tirpak (1989). Most approaches to limiting the rate of climate  change focus on limiting anthropogenic
emissions of C02 and other radiatively important trace gases.  However, large-scale ecosystem
manipulation to sequester carbon,  reduce drought effects, establish appropriate species, mitigate
impacts on surface and near-coastal waters,  and  alleviate atmospheric pollution from large, persistent
forest fires  could be important in limiting positive feedbacks  from the biosphere and in optimizing the
biospheric  responses to climate change.

3.2.5 Research Needs

Future research priorities ensue from these analyses and can be grouped  into several categories.  The
most critical assumption is that forests will decline due to drought  stress and will eventually burn.  This
assumption should  receive considerable scrutiny through focused  research efforts on regional water
balance, C02-induced changes in WUE, and climate-fire interactions.  Given this assumption, the most
critical processes are decomposition rates of coarse, woody debris, and post-fire  biomass and regrowth
rates as determined by successional processes.  Mitigation options to reduce forest decline, sequester
carbon, and facilitate regrowth should  receive attention. All  three of these areas, i.e., decline processes,
regrowth processes, and mitigation processes, must be examined  in the context of current and
anticipated land use practices.
                                                75

-------
The research needs may be summarized as follows:

1)     Processes of forest decline and disturbance regimes; in particular:
               •      Interaction of C02 concentration, WUE and landscape water-balance processes
               •      Climate, drought stress, pest, fire dependencies

2}     Post-disturbance regrowth and succession, specifically:
               •      Post-fire biomass and regrowth rates
               •      Dependencies of regrowth dynamics on species and life-form composition
                      during succession

3)     Mitigation options for:
               •      Stress detection (monitoring), stress amelioration (e.g., thinning)
               •      Regrowth  enhancement (e.g., selective 'weeding',  introductions)
                                               76

-------
4      DIRECT EFFECTS OF ATMOSPHERIC CO2 ENRICHMENT ON FORESTS

Despite the prominent role of the terrestrial biosphere in the global carbon cycle, limited information is
available on long-term response of vegetation to C02 enrichment.  This lack of knowledge is
incongruous with the readily quantified and widely accepted increase in atmospheric CO2 that has
occurred since the turn  of the century.  Preliminary results from studies of two ecosystem types, arctic
tundra and a coastal marsh, reveal that long-term system responses to increases in C02 vary depending
on site conditions and plant genotype (Oechel et al. 1984, Drake et al. 1987, Tissue and Oechel 1987,
Curtis et al. 1989b, Mooney et al. in press).  Long-term field evaluations of forest ecosystem response to
CO2 enrichment have been initiated only in recent years.

4.1     Trees

The ability of individual trees to sequester carbon will change with increasing atmospheric CO2
concentrations and subsequent climate change (moisture and temperature  stress).  A rise in atmospheric
CO2 concentration is the most certain of projected global changes and thus it will be considered first.

4.1.1  Effect of C02 Enrichment on Plant Development

To understand direct and indirect effects of C02 enrichment on individual trees requires knowledge of
how juvenile and mature trees assimilate and allocate carbon during ontogeny. This entails both: 1) a
summation of seasonal  patterns in growth and development; and 2) an annual adjustment to these
seasonal patterns due to inherent changes in plant physiology  with age. The current understanding of
C02 effects on trees and forests is based almost exclusively on studies using tree seedlings measured
during one growing season.  Preliminary studies on the response of arctic and salt marsh (C3 and C4)
annual plants to C02 enrichment have also been completed (Table 4.1-1).  The seedling response
literature contains little information on seasonal growth patterns or changes in those patterns over time.
Furthermore, seedling research has been conducted almost solely in controlled-environment chambers
where growth conditions, and therefore responses, can be artificial. Long-term measurements of forest
ecosystem response to C02 enrichment under realistic conditions are sorely needed.

Despite the limitations imposed by artificial environments and short study durations, these studies have
consistently demonstrated many positive changes in growth, morphology, and physiology in woody
plant due to increased atmospheric C02 concentrations (Table 4.1-2). Indeed, commercial seedling
production facilities have utilized enriched-C02 environments for decades to enhance growth rates (Tinus
                                              77

-------
 and McDonald 1979).  The following summary briefly reviews these potential direct effects of
 seedling growth, morphology, and physiology.  Additional reviews are available in Strain and Cure
 (1985), Shugart et al. (1986), Kramer and Sionit (1987), and Allen (1990).
 Table 4.1-1     Summary  of Arctic and Salt Marsh (C3and C4) Annual Plant Responses  to
               Enriched CO2 Environments  in Field Exposure  Chambers (from Mooney et al. in
               press)
                                            Response to C02 Enrichment1
                                         Arctic             Saltmarsh
I. Plant Effects                                            C3             C<,
 Carbon exchange

 Photosynthesis                             0             +             0
 Photosynthetic acclimation                   +00
 Plant respiration                             0
 Shoot decomposition                        n/a

 Growth

 Shoot expansive growth                     000
 Root biomass                               -/O            +             0
 Number of shoots                           +             +             0
 Size of shoots                              000
 Root/shoot ratio                            -/O            +             °

 Tissue Composition

 Nitrogen concentration                       -              -              0
 Carbon/nitrogen                            +             +             0
 Starch content                              +/0           n/a           n/a
 Tissue density/specific wt.                   +00
 Salt content                                n/a           -              n/a

 Development/Reproduction

 Senescence
 Tillering                                    +             +             0
 Number of flowers                           -              0             0
 Number of seeds/stem                      -              00
 Sexual/asexual reproduction                 -              n/a           n/a

 Water Use

 Transpiration                               0             -              -
 Water use efficiency                         0             +             +
 Leaf temperature                            0             +             +
 Water potential                             n/a           +             +
                                              78

-------
Fable  4.1-1  (cont'd)                                    Response to C02 enrichment
                                                   Arctic              Saltmarsh
                                                                  C3
II. Ecosystem effects
Evapotranspiration
Net carbon storage
Carbon exchange acclimation
Net ecosystem respiration
Species composition
Water use
Nitrogen content of canopy
Soil enzyme activity
Soil solution nitrogen
Nitrogen content of below-
ground biomass

0
+ /0
+
-/O
+
0
n/a
+ /-
-/o
n/a


.
+
0
-
n/a
-
0
n/a
n/a
+


.
+
0
-
n/a
-
0
n/a
n/a
0

1 Relative to ambient concentrations; +,  increase; -, decrease; 0, no change; n/a no data available
                                                      79

-------
Table 4.1-2     Seedling  Morphological and Physiological  Responses  to Enriched-CO 2
               Environments  Based  on Short-term  Controlled-Environment  Studies (<. 1 yr, with an
               exception  as marked)
                                                           Response1
Morphology                                      Conifer                     Broadleaf

        stem diameter                                +                             +
        stem height                                  +                             +
        relative growth rate                           +                             +
        root and shoot weight                        +                             +
        weight: height ratio                            +                             n/a
        root:shoot ratio                              +                             +/-
        fine:coarse root ratio                          +                             n/a
        N-fixing nodule mass                          n/a                           +
        mycorrhizal mass                             +                             +

        number of branches                          +                             +
        secondary leaf production                     +                             n/a
        crown width                                  +                             n/a
        leaf area                                     +                             +
        leaf area:total biomass ratio                   n/a                           +
        leaf thickness                                 +                             +
        leaf area duration                             +                             +
        stomatal density                              n/a                           +

        number of buds                              n/a                           +
        weight per bud                               n/a                           +
        wood density                                 n/a                           +

Physiology

        net photosynthesis                            +                             +
        starch in needles                             +                             n/a
        allocation of carbon to roots                   +                             n/a
        pigment concentrations
        chlorosis)2                                   +                             n/a

        stomatal conductance
        water use efficiency                           +                             +
        water content/potential                       -                             n/a
        leaf temperature2                             +                             n/a
        heat tolerance2                               -                             n/a

        nitrogen concentrations                       n/a
        nutrient uptake                               +                             +
        nutrient use efficiency                         +                             +
        ion leaching from soil                         -                             n/a
1 +, increase; -, decrease; n/a, no data available
2 based solely on a multi-year study
                                               80

-------
4.1.1.1 Seedling growth responses to CO2 enrichment

Table 4.1-2 shows that higher carbon assimilation rates resulting from a reduced C02 gradient between
the leaf and atmosphere in an enriched-C02 environment (all other things being equal)  lead to:

1)     increased growth rates of roots to capture and maintain soil contact for water and nutrient
       uptake (Tolley and Strain 1984b, Sionit et al. 1985, Brown and Higginbotham 1986, Norby et al.
       1986a, Hollinger 1987, Norby 1987, Norby et al. 1987, Campagna and Margolis 1989);

2)     increased stem and foliar biomass, surface area, and foliage duration for photosynthetic surface
       and competition for light (Funsch et al. 1970, Tinus 1972,  Canham and McCavish 1981,  Rogers
       et al.  1983a,b, Tolley and Strain 1984a,b, Higginbotham et al. 1985, Oberbauer et al. 1985,
       Brown and Higginbotham 1986, Norby et al. 1986b);

3)     increased bud size and weight for ensuing years' growth (Norby et al. !986a);  and

4)     more available carbon for metabolite synthesis (e.g., starch and sugar) for stress resistance and
       symbiotic associations (Norby 1987, Norby et al. 1987,  O'Neill et al. I987b, Arnone and Gordon
       in press).

In addition, the optimal temperature for photosynthesis increases  (Kramer and Sionit 1987) and  the light
compensation point decreases with increasing C02 concentration, suggesting that plants can maintain
photosynthetic rates at higher temperatures and lower light levels in higher C02 environments.  Plants do
not respond above a threshold CO2 level  of about two- to three-times ambient CO2 concentrations
(Higginbotham et al. 1985, Rogers et al. 1983a).  Loblolly pine (Pinus taeda) is a notable exception to
these generalities, showing no short-term increase in biomass accumulation due to increased C02
(Tolley and Strain 1984b).

Seedlings respond physiologically to C02 enrichment typically within hours to days of their introduction
to a high-C02 environment, with some acclimation to higher C02 concentrations within weeks to months.
For example, arctic plants under field conditions (Tissue and Oechel 1987) and temperate plants in
growth chambers (Tinus 1972, Tolley and Strain 1984a) show the loss of growth responses within a
growing season. This latter "acclimation" (in temperate plants) may be attributed to saturation of the
carbon sink in a  limited environment (e.g., pot-bound plants). Examination of CO2 acclimation in situ
with coastal marsh grasses suggests that responses will vary widely by site and plant genotype (Curtis
                                              81

-------
et al. 1989a). The response of plants that are first acclimated to a high-C02 environment and then
placed in an environment that is further enhanced with CO2 is unknown.

Information on the growth characteristics of woody perennials that are exposed to enriched-C02
environments from the time of seed germination is also limited. A study at ERL-C with three native tree
species has shown little morphological difference in plants at 14 weeks from germination (growth
responses of one of three species, Acer macrophyllum, are shown in Table 4.1-3).  Research will
proceed through consecutive growth and dormancy cycles to determine if anatomical, morphological or
physiological differences develop that prepare the seedlings for overwintering and ensuing years' growth.
This study is the first to examine multiple-season ontogeny of seedlings and to expose the Acer genus
to enriched C02.
Table 4.1-3.    Early Growth (14 weeks)  or Acer macrophyllum  Seedlings  (n = 48)  in Ambient- and
               Enriched-CO 2 Environments (350,  575, and 700  ppm)
C02
Concentration
(ppm)
350
575
700
Collar diameter Seedling Height
(mm) (cm)
Mean Std Range Mean Std Range
5.7 1.0 3.1-7.5 67.0 14.2 27.0-89.7
6.4 0.9 3.6-7.8 78.4 14.7 30.7-101.9
6.7 0.7 4.9-7.9 76.7 12.7 49.3-99.4
Long-term studies with juvenile trees reported by Surano et al. (1986) and Houpis et al. (1988) have only
partially evaluated issues regarding the duration of short-term responses.  These reports are based on a
2.5-year experiment with tree saplings grown in individual field chambers.  This study demonstrated the
loss of many of the positive short-term responses noted in Table 4.1-2, and it also revealed negative
long-term plant responses.  For example, seedlings tended to accumulate heat energy in leaves due to
partial stomatal closure.  Surano et al. (1986) reported a general decrease in heat tolerance (i.e., in the
upper thermal  limits of the photosynthetic apparatus) with increased C02 concentrations even though
the optimum temperature for photosynthesis is increased.  Foliage showed classic visual stress
symptoms (e.g., chlorosis), and saplings showed eventual decreases in growth rates (relative to control
treatments) after early accelerated height and diameter growth advantages.
                                               82

-------
Based on these results, Surano et al. (1986) and Houpis et al. (1988) proposed that plants reached a
new homeostasis under higher C02 and lost the short-term benefits as environmental stresses
developed. However, these results are from one study conducted under one set of conditions on one
plant each. Additional multiple-year exposure studies are clearly needed to evaluate tree responses for a
range of biomes and environmental conditions.

4.1.1.2  Ecophysiological responses to C02 enrichment

The literature provides a preliminary picture of potential  nongrowth-related effects of CO2 on woody
plants, including: modification of water- and nutrient-use efficiencies, stress tolerance, and plant
dormancy requirements. Water-use efficiency (WUE) is defined as a ratio, the number of CO2 molecules
fixed per water molecules transpired.  As C02 concentration increases, fewer water molecules exit the
stomata for each CO2 molecule that enters. Thus, plants may either put on more biomass while using a
fixed amount of water, or they may use less water to acquire a fixed amount of biomass.  As with growth
responses, the largest changes in WUE are seen early in the enrichment period (Norby and O'Neill
1989), a potentially important time given the synchrony of water availability and plant growth cycles.

An  increase in nutrient-use efficiency appears to stem almost solely from increased biomass production
with little to no increase in nutrient uptake,  interpreted as a nondamaging foliar dilution of elements.
Normally deficient elements such as phosphorus, potassium, magnesium,  and boron may be recycled
internally to new meristems in enriched-C02 environments (Luxmoore et al. 1986).  Increased fine root
growth in a higher CO2 atmosphere enhances the nutrient availability in a given soil profile.  Finally,
improved water- and nutrient-use efficiencies of plants in a C02-enriched environment have  the potential
to significantly enhance seedling  performance on nutrient- and water-limited  lands.

The influences of C02 enrichment on biotic and anthropogenic plant stresses other than water or
nutrient deficits could be significant, but they have not been fully considered (Kramer and Sionit 1987).
For example, insect damage to plants is reduced and parasites grow more slowly and suffer greater
mortality in higher C02 atmospheres (Fajer et al. 1989).  These effects are linked with reduced nutrient
content (nitrogen content or carbon:nitrogen ratio) and water content in host material.  Fajer et al. (1989)
found no change in the levels of defensive  compounds produced by plants.  Air pollution injury is likely
to be reduced with increasing CO2 due to partial stomatal closure (Green and Wright 1977). However,
as with  growth data, only short-term changes in stress responses due to increased C02 concentrations
have been tested.  Long-term, multiple growing season research is required.
                                               83

-------
Carbon dioxide enrichment may also alter the natural dormancy cycle of trees (Bailey et al. 1990),
creating potential reproduction and growth problems (e.g., frost damage to unhardened plants). A short-
term study by Rogers et al. (1983a) revealed that pine trees in high-CO2 environments may continue
active growth after trees in ambient chambers become dormant, an effect similar to that found with over-
irrigation or over-fertilization. Enriched CO2 has been shown to alter carbohydrate storage patterns
(particularly starch).  Starch accumulation (the storage of energy for overwintering and  bud break) is
related directly to plant dormancy and stress  resistance.  Oberbauer et al.  (1985) demonstrated excess
starch accumulation in chloroplast in foliage of tropical tree species, and Campagna and Margolis (1989)
showed increased starch in the needles of black spruce with increasing CO2 concentrations. Finally,
higher leaf temperatures, which may interfere  with temperature control of dormancy initiation, have been
observed in high-C02 environments. The fate of growing season responses (e.g., starch accumulation
and extension of the  season) later in the plant's growth cycle (during dormancy) is unknown, as is their
effect on subsequent years' growth.

The interaction of direct C02 enrichment effects and indirect climate change effects must also be
considered. Section  1.4.3  of this report demonstrated the potential for increased atmospheric CO2
concentrations to induce changes in climate,  primarily in moisture and temperature. Ecophysiological
plant responses to C02 may either offset or amplify these effects.  For example, decreased water
availability for trees could be offset by increased WUE, but changes in water movement through plant
stomata might interfere with temperature regulation. A temperature increase would complement C02
increases in terms of increased carbon assimilation rate (given an increase in the temperature optimum
for photosynthesis), but it may also interfere with bud dormancy regulation and winter cold-hardiness.
The potential  interactions are numerous and must be addressed regionally as projections of climate
change  are refined.

4.1.2 Water-Use Efficiency

Increases in WUE due to increasing atmospheric CO2 concentrations represent the potential for
substantial impacts on forest ecosystem condition, composition, and migration. Unfortunately, our
knowledge of this process Is largely limited to growth chamber studies under non-limiting conditions of
water availability.  It is not clear what benefits could accrue from increased WUE if soil water availability
is decreased or if the demand for water (PET) is increased (Neilson et al. 1989).  The response of forests
to changes in  regional distribution of precipitation in combination with temperature and  PET extremes
has been considered. For example, Neilson et al. (1989)  and Urban and Shugart  (1989) concluded that
forest ecosystems in  the southern United States could experience long-term negative impacts due to
                                               84

-------
extended drought in a double-C02 climate. The following section addresses predictions regarding
changes in seedling and tree water relations in an enriched-CO2 environment, changes that may modify
plant response to an altered climate.

General water relations in plants and adaptation to or tolerance of drought stress in natural and
managed ecosystems have been reviewed by other authors (e.g., Kramer 1983, Jarvis 1987).

4.1.2.1 Short-term water-use responses to C02 enrichment

Several short-term studies of ecophysiological processes have determined that C02 enrichment can alter
the water relations of C3 and C4 plants. Stomatal conductance is a primary plant process influenced by
C02 enrichment, and reduced  stomatal conductance (lower transpiration per unit leaf area) has been
widely observed in plants in enriched-C02 environments (Houpis et al. 1987, Williams et al.  1986,
Hollinger 1987, Eamus and Jarvis 1989,  Norby and O'Neill 1989).  The ratio of carbon gain to water
loss, one measure of WUE, is consistently increased with elevated C02 (Rogers et al. 1983a,b,
Oberbauer et al. 1985, Norby et al. I986a, Conroy et al. 1988, Norby and O'Neill  1989). This
phenomenon has been demonstrated with instantaneous as well as cumulative measurements of water
use (Morison 1987, Mooney et al. in press).  However, instantaneous WUE response can be reduced at
the whole-plant level or over long periods of adjustment to CO2 enrichment (Morison  1987).

An examination of arctic and estuarine plants subjected to C02 enrichment over the long term suggests
that changes in anatomy and other ecophysiological processes  contribute to changes in plant water
relations, as reviewed by Mooney et al. (in press). For tree seedlings, Norby and O'Neill (submitted)
observed that the ratio of leaf area to dry weight decreased and root biomass increased in yellow-poplar
(Liriodendron tulipifera), enhancing WUE under C02 enrichment. Research  has demonstrated that
mycorrhizal development and root biomass of tree seedlings are significantly increased in a high-C02
environment (Norby et al.  1987, O'Neill et al.  1987a). These mycorrhizal relationships can play a
prominent role in drought avoidance or tolerance (Dixon et al. 1980).

Changes in WUE in plants observed in short-term controlled- environment studies may not always be
observed in the field due to the host of interacting factors that regulate plant development.  For example,
Tolley and Strain (1985) observed that C02 enrichment ameliorated the impact of drought in sweetgum
but not in loblolly pine.  Plants  resist drought through tolerance and avoidance mechanisms (Eamus and
Jarvis 1989). Tolerance mechanisms include maintenance of  cell turgor, and thereby of leaf expansion
and photosynthesis, under low water potential.  Avoidance includes low stomatal conductance, leaf
                                              85

-------
abscission and increases in root system biomass.  However, plants that avoid drought through high
WUE can be at a competitive disadvantage in the field if a drought-tolerant competitor can exploit the
more available water for its metabolic processes (DeLucia and Heckathorn 1989).

Plants growing in an enriched-C02 environment may efficiently tolerate drought through osmotic
adjustment of tissue (Eamus and Jarvis 1989), depending partially on carbon metabolism and source-
sink relationships of leaves.  Elevated C02 increased the concentration of soluble sugars in the feeder
roots of loblolly pine under drought stress (Norby personal communication). The accumulation of
osmotic solutes is linked to the degree of water stress in many conifers.  This and other drought
tolerance and avoidance mechanisms in plants growing in an enriched-CO2 environment merit further
investigation.

4.1.2.2 Long-term water use responses to CO2 enrichment

Long-term responses of woody plants to CO2 enrichment have not been evaluated.  Therefore, many
research issues focus on understanding long-term tree responses to changes in CO2 concentration,
moisture, and temperature.  This will require multi-year exposures and measurements on a diverse
spectrum of life stages under field conditions. Research should be designed to explain and model tree
responses mechanistically since empirically testing every possible combination of conditions is
impossible (Mooney et al. in press).

4.1.3  Research Needs

Section 4.1.2.2 noted the almost complete lack of research in the area of long-term seedling and tree
responses to CO2,  particularly in the area of whole plant development and WUE. The response of
terrestrial vegetation to  CO2 enrichment with  or without climate change will have tremendous effects on
the global carbon cycle. The following research tasks are needed:

1)     Long-term exposures of tree species from the time of germination until mature tissue is
       generated.

2)     Assessment of whole-plant development and changes in anatomy, morphology, and physiology,
       to allow the formulation of  process-based models of responses to CO2 enrichment.
                                              86

-------
3)     Characterization of C02 effects on WUE for a wide range of genera (e.g., tropical to boreal) and
       life stages.

4)     Quantification of atmospheric C02 enrichment effects on terrestrial carbon pools and fluxes.

42     Ecosystem   Perspective

The above sections discuss the possible impacts of increased C02 concentrations on WUE of individual
plants. Increased WUE could counteract the potential negative effects on plant growth of increased PET
caused by climate warming and/or declines in precipitation (Mooney et al. in press).  Whether increased
WUE will mitigate the effects of increased PET in forested ecosystems is a key research question
debated in the literature (Oechel and Strain 1985, Strain 1987, Norby and O'Neill 1989, Mooney et al. in
press).

Availability of water appears to be the primary  determinant of the distribution of forests in the continental
United States (e.g., Neilson et al. 1989, Stephenson 1990).  Current forests appear to use nearly all the
water available to them during the growing season for growth or simply for maintenance (Neilson et al.
1989). The amount of regional evapotranspiration in forests is a function of the volume of deep soil
water, rooting depth, leaf area over the region, and atmospheric vapor pressure deficit. The primary
regulator of evaporation from vegetation appears to be leaf area, which increases or decreases over time
as water supplies change (Woodward 1987).  If regional PET were to increase due to increased
temperature (ignoring the direct effects of C02 and assuming no change in water supply), soil water
would be depleted, causing plants to die or shed leaves.  Leaf area and hence evapotranspiration would
decrease  in subsequent years, until a new biomass equilibrium is reached. If this process were carried to
the extreme, either through increased PET or decreased water supply, trees could no longer be
supported (Neilson et al. 1989, Stephenson 1990).

What  could happen under warmer conditions with increased WUE caused by higher CO2
concentrations?  At least two scenarios are plausible.  First, since  more water is effectively available for
growth,  regional leaf area could increase during the growing season in comparison with the unchanged
WUE  scenario.  However, Increased leaf area would increase the amount of evapo-transpiration from the
ecosystem, with the water available for growth being used before the end of the growing season, as in
the unchanged WUE scenario. In fact, increased leaf area could result in faster evapotranspiration, with
water stress beginning  earlier in the growing season.  Net primary productivity need not be any higher
                                               87

-------
and could be less than in the unchanged WUE scenario. Changes in species composition would occur
as in the unchanged scenario.

A second scenario takes into consideration the possibility that root/shoot allocations could change
under higher C02 concentrations with proportionately more growth allocated to roots and mycorrhizal
associates (Norby et al. 1987, O'Neill et al. 1987b).  Root biomass and area can increase up to two-fold
under C02-enriched environments (Rogers et al. 1983a), with roots potentially reaching deeper Into the
water table.  Under this scenario, leaf area and thus evapotranspiration would not increase as much as
in the first scenario.  Net primary productivity could increase relative to the unchanged WUE scenario as
long as the increase in evapotranspiration caused by the limited increase in leaf area did not offset gains
in effective moisture caused by the higher WUE and  root biomass.  Alternatively, increased root area
could simply facilitate the rapid extraction of soil water, again limiting the system (Federer 1982).

If regional ecosystems are to realize a benefit from increased WUE of vegetation, the most effective
strategy would be to use less water and  not accumulate additional  biomass in either root or shoot.
However, this seems unlikely, given that  plants clearly do not know that they will be water limited until
they actually are water limited. Thus, one might expect plants to spend the water and  add the extra
biomass  (in  either root or shoot), which could result  in drought stress as the water  is depleted.

Another issue concerning the direct effects of C02 on ecosystem productivity is nutrient availability. If
plant growth is limited by nutrients  rather than by water, increasing WUE will have minimal impact on
plant growth. Experimental data suggest that plants have mechanisms that use soil nutrients more
efficiently under elevated-CO2 conditions, effectively increasing nutrient-use efficiency (Luxmoore et al.
1986, Norby et al. I986a, O'Neill et al. I987a,b). Longer term experimental studies  with representative
forest soils in combination with water stress are needed to determine the effects of  elevated CO2 on
natural communities.
                                               88

-------
5      MITIGATION OF GLOBAL CHANGE IMPACTS THROUGH FOREST MANAGEMENT

Through its role in the carbon cycle, the terrestrial biosphere is one of the major components in global
climate dynamics along with the oceans of the world (Schneider 1989b, Houghton et at. 1983). The size
of the carbon pool is much greater for oceans than for continents (as shown in Figure 1.1-1, oceans
have about 38,500 Gt of carbon versus approximately 2,000 Gt for the terrestrial biosphere). However,
the flux rates of carbon in the terrestrial biosphere are approximately equal to those of oceans
(Houghton 1987). The terrestrial biosphere, therefore, covering only 30% of the earth's surface, has a
much greater carbon flux rate per  unit area than the oceans have.  This is largely the  result of
photosynthesis by vegetation in the terrestrial ecosystems (Figure 1.1-2).  Forests, in turn, are a
substantial component of the terrestrial biosphere (34% of the area), and have a significant influence
over terrestrial carbon fluxes  into and  from the atmosphere (Harmon et al. 1990, Woodwell et al.  1983).

World forests capture and store significant amounts of atmospheric C02 through photosynthesis  (Shands
and Hoffman 1987).  Waring and Schlesinger (1985) stated that:  1) "the equivalent of the entire CO2
content of the atmosphere passes through the terrestrial biota every seven years, with about 70% of the
exchange occurring through the forest ecosystems"; and 2) the vegetation and soils of forests combined
contain about  60% of the organic carbon in terrestrial ecosystems of the world using the estimates of
765 Gt vegetative carbon and 635  Gt  carbon in  forest soils (Waring and Schlesinger 1985).

Consequently, forest management to  foster forest carbon sequestering is one possible option for
mitigating the  increase in atmospheric C02 concentrations.  The purpose of this section
is to review management options for increasing carbon sequestering in forested regions, specifically,
reforestation and afforestation, forest stand treatment, and management of soil carbon.  In addition to
summarizing the available literature, new analyses are presented on:  1) the potential amount of carbon
that could be sequestered in tropical regions each year through reforestation and afforestation, 2) the
amount of land that would have to be available  for reforestation in the United States to offset its yearly
emission of C02, and 3) the effects of thinning and fertilization on forest growth.

5.1  Forest Management Opportunities

Forest management refers to the combination of business methods and forestry principles to forest
areas to achieve the land owners'  objectives (Barrett 1980, Davis 1966).  Components of forest
management,  while directed at maximizing a specific yield from forest stands (e.g., timber pulp,
                                               89

-------
recreation) can also alter carbon storage and release from forests.  Forest management practices
relevant to the mitigation of atmospheric C02 increases are reviewed here.

1) Silviculture

Silviculture refers to the biological and physical manipulations of forests based upon ecological
principles for the purpose of enhancing yields of wood and other forest outputs (Smith
1986).

Typical silvicultural treatments are:

       a)  Reforestation:  the artificial establishment of forests on land that previously carried forests.
Reforestation frequently involves replacing the previous crop by a new and essentially different crop
(Wiersum 1984).  When essentially the same crop as before is established, the process is sometimes
referred to as artificial regeneration  (as opposed to natural
regeneration or the establishment of new forests by natural means).

       b)  Afforestation: the artificial establishment of forests on land that previously has not carried
forests within anyone's memory or during the last 50 years (Wiersum 1984).

       c)  Stand treatments:  manipulations applied to established  plantations or existing forests usually
to improve growth and yield of wood or other forest resources.  Manipulations include weed control,
thinning (i.e., stocking control), and fertilization.  This practice is also called timber stand improvement
(TSI) or forest tending (Smith 1986, Barrett 1980).

2) Mensuration

Mensuration is the subject of forest stand measurements,  commonly conducted to quantify forest
inventory (i.e., standing  stock), growth, and yield (Dixon et al. 1990, Husch et al. 1972).  Mensuration
includes assembling data and maps of forest land area. Worldwide, definitive data on the amount of
land now occupied by forests or available for reforestation are difficult to obtain (Section 5.1.1).
Currently, development  of more reliable data bases on this subject is a high research priority.
                                                90

-------
3) Harvesting Operations

Harvesting refers to the activities required to extract the wood resource from forest lands (Barrett 1980).
Activities include road construction, logging methods (e.g., selective or clearcutting), and equipment
deployment.  This is also the process needed to convert a forest area from one or a mix of species to a
forest with a more desirable composition.  Unfortunately, harvesting without attempting any reforestation
often leads to less productive land uses, i.e., deforestation (Section 5.3), and in some places,
desertification.

4) Agroforestrv

A current definition for agroforestry is "... a land use that involves deliberate retention, introduction, or
mixture of trees or other woody perennials in crop/animal production fields to benefit from the resultant
ecological and  economic interactions" (MacDicken and Vergara 1990). Humans have practiced
agroforestry for thousands  of years, but only since the mid-1970s has an effort been made to collect
information on the extent and principles of agroforestry management.  Vast areas of land are devoted to
agroforestry worldwide. Wood et at.  (1984) estimate that as much as 2 billion ha of land are available in
the tropical regions for growing tree crops. It likely that agroforestry could be practiced on about a third
of this land base (Buriey and Stewart 1985).

5) Fuel Woodlot Forestry

In the tropical forest regions, many rural people depend  upon wood for cooking fuel, animal fodder,
building  materials, and other domestic uses (Gregersen et al. 1989). Historically, this wood has been
harvested from natural tropical forests.  In some locations where the forest resource has been depleted,
trees are planted especially for fuel wood. Thus, fuel woodlot forestry is the deliberate cultivation of
trees for firewood (Barrett-Lennard et al. 1986).  Rotation lengths are often just 5 to 10 years and as
many as 60 species of tropical hardwoods have been used (National Academy of Sciences 1980). Fuel
woodlots, while sequestering only modest amounts of carbon themselves, can significantly offset
deforestation of natural tropical forests, thereby preventing significant losses of existing carbon to the
atmosphere.  An important research  need is to quantify through local case studies how much fuel
woodlot  forestry actually offsets tropical deforestation (MacDicken and Vergara 1990).
                                                91

-------
6) Urban Forestry

A significant portion of the world's forests that can contribute to CO2 mitigation are those maintained
within urban and  suburban areas. Examples are community parks, greenbelts, roadside forests, and
wooded residential and industrial zones.  In the United States, the urban/suburban forests total
approximately 28 million ha (Grey and Deneke 1978).

5.1.1  Potential for Large-Scale Reforestation to Sequester Atmospheric CO2

Several researchers have calculated the amount of new forest plantations needed to offset all or part of
annual C02 emissions and buildup.  Dyson (1977) estimated that approximately 700  million ha of new
forest would be required to offset the 5 billion tons of total annual global anthropogenic carbon
emissions,  and that the United States could provide about 10% of this area. Marland (1988, 1989) put
the estimate at 500 to 700 million ha depending on the proportions of new plantations in the tropics or
temperate zone.  Sedjo (1989) concluded that 465 million ha would be required to offset the 2.9 billion
ton annual  increase (as opposed to total emissions) in atmospheric carbon. Sedjo also cited an
unpublished paper by Woodwell  (1987) suggesting that 200-400 million ha of additional forest would be
required to sequester 1-2 billion tons of carbon annually.

5.1.1.1 Potential  rates of carbon sequestration in the tropics:   Preliminary analysis

A different approach  for evaluating the potential for sequestering  carbon through reforestation or
afforestation/revegetation is presented in this section.  Estimates of land availability and productivity in
the tropics  are used to calculate the amount of carbon that  could be sequestered each year. The focus
here is on the tropics because estimates of tropical land availability have been  published, and the
potential rates of  sequestration are higher than most temperate locations (Odum 1983, Waring and
Schlesinger 1985).

The term "afforestation/revegetation" is used to indicate a range of options for increasing the productivity
of desertified land. This could be achieved by reestablishment of the pre-desertification vegetation cover
on a site (i.e. desert scrub, shrubland, or woodland), which  is called revegetation here.  Or,
establishment of plantations of drought-tolerant tree species, here termed afforestation, could be used to
increase productivity. In the following analysis, productivity values for natural arid land communities are
used (e.g.,  desert scrub).  Determining how to reestablish vegetation or afforest desertified regions is an
important research need.
                                                92

-------
Land availability in the tropics is taken from Wood et al. (1934, Table 5.1-1). Low and high estimates of
productivity for tropical regions are used to indicate the range of potential sequestration rates (Table 5.1-
2).  The method used to calculate sequestration rates is shown in Table 5.1-2.

Results of the analysis show an order of magnitude difference in the range of carbon sequestration rates.
Both reforestation and afforestation/revegetation have similar ranges. At the high end of the -ange,
sequestration rates approach over half of the rate of fossil fuel estimates.  However, the hie . ^nd is
probably unattainable, because each hectare of land would have to sequester carbon at the ninhest rate
listed in Table 5.1-2.  Differences  in site condition within a region would prevent this from occurring.  It
should also be emphasized that the land availability values are highly uncertain because there is no
standard approach for estimating available land on a global or continental scale.
Table 5.1-1.    Tropical Lands  with Potential  for Reforestation  (million ha) (Wood et al. 1984)

                       Logged        Forest  Deforested     Desert'rfied           All
Region                 Forests1       Fallow2 Watersheds3   Arid Lands*         Lands

Latin America           53.5            65.7      27.5          701.8           848.5
Africa                   42.8            58.7       3.0          685.0           789.6
Asia                    59.8            56.6      56.5          170.0           342.9

Total                  156.2          181.0      87.0         1556.8           1981.0
1 Almost 90% of these are tropical moist forests (tropical rain forest and tropical moist dec..  , js forest)
2 All are In tropical moist forest areas (tropical rain forest and tropical moist deciduous forest)
3 Montane forest. The area of deforested watersheds is only a rough estimate. It has been included in
 'All Lands'.
* Savanna and arid lands
                                                93

-------
Table 5.1-2.    Range  of Estimates  of Carbon Sequestration
               Reforestation  or Afforestation.
                                   for Tropical Forests  under Scenario of



Region

Available
Land1
(million ha)

Range of
Productivity2
t C/ha/yr
Range of
Total C
Accumulated3
(Gt C/yr)
Reforestation:

Latin America
Africa
Asia
 146.6
 104.5
 172.9
1.95-10.0
1.95-10.0
1.95 - 10.0
Total                    424.0

Afforestation/Reveaetation:
Latin America
Africa
Asia

Total
 701.8
 685.0
 170.0

1556.8
0.35 - 3.0
0.35 - 3.0
0.35 - 3.0
0.29-1.47
0.20-1.05
0.34-1.73

0.83 - 4.25
0.25-2.11
0.24 - 2.06
0.06 - 0.51

0.55 - 4.68
Available land estimates are taken from Table 5.1-1; land available for reforestation is the sum of
Columns 1,2, and 3 (logged, forest fallow, and deforested watersheds); land available for afforestation is
desertified and arid lands.

2 Low estimate of productivity for reforestation was derived from the estimated minimum total biomass
accumulation over 40 years in tropical plantations from Brown et al.(1986).  (155.7 t/haw 40 yrs x 0.5
[carbon-to-biomass conversion factor] = 1.95 tC/ha/yr.
High estimate of productivity was derived from an estimate of gross primary production for wet tropical
forests (Odum 1983) (2000 kCal gross primary production/m2/yr = 1 t C net primary production/ha/hr;
assuming 10 kCal equals 1 g C and net primary production equals half of gross primary production).

Range of productivity for afforestation is from estimated net primary productivities by.ecosystems
(Waring and Schlesinger 1985). The "desert scrub" ecosystem  is used for the low estimate and
"woodland and shrubland" ecosystem for the high estimate.

3 Low and high estimates of total carbon accumulated are the product of land area and low and high
productivity, respectively.
                                               94

-------
5.1.1.2 Reforestation workshop

A recent workshop on international reforestation was sponsored by the EPA at ERL-C.  The workshop
explored  ecological, operational, and sociopolitical considerations for successful large-scale reforestation
projects.  Workshop participants included about 50 forest managers and scientists from Brazil, British
Columbia, Congo, Guatemala, India, New Zealand, the United Kingdom, and United States Pacific
Northwest and Southeast regions.  The workshop was designed to gather knowledge from these
participants to guide decision makers in implementing future large-scale reforestation projects. A list of
participants is given in Appendix C.

Participants agreed on two important points:

1)     Current technical and ecological knowledge, though needing research in a few key areas, is
       adequate to undertake large-scale projects for some initial increases in reforestation on a global
       scale (Smith  1986, OTA 1984).

2)     Of all the requirements needed for successful reforestation, the social and political limitations,
       especially in the developing nations, were emphasized repeatedly.  These limitations take many
       forms, but most important  are:  a) the differential awareness many people have of the value of
       forests; b) the need for financial incentives or alternative incomes, particularly to encourage
       small landowners or migrant people at subsistence living standards to plant and grow trees; and
       c) the lack of governmental infrastructure in many nations to undertake large-scale  reforestation
       projects (Gregersen et al. 1989).

Participants also used their combined knowledge to estimate carbon sequestration rates for various
forest systems (Table 5.1-2).

3).  For the tropics, the rates estimated in Table 5.1-3 are at the high end (Brazil and Congo)  of
published values in Table 5.1-2.  The greatest uncertainty in Table 5.1-3 is whether these high rates
could be implemented over millions of hectares.

To place the amount of effort needed to reforest the land listed In Table 5.1-1 in context, consider that
during the 1980s, about 16.4 million ha in the world were planted to trees annually (Table 5.1-4).  Most
forestation occurred in temperate latitudes. To plant 750 million to 1 billion ha of land in the world over
25 years (a period accepted as reasonable at the above workshop), an annual rate of 30 to 40 million ha
                                                95

-------
is required.  Socioeconomic barriers must be overcome before significant forestation will occur in
tropical latitudes.  Establishment of sustainable agroforestry systems with multiple purpose trees offers
one attractive alternative (Jain et al. 1989).

Another factor affecting the feasibility of reforestation as a management option is cost.  Experts at the
workshop estimated establishment costs ranging from $200 to $1700/ha depending upon the species
and geographic location.  Comprehensive economic analyses are needed to evaluate the cost of forest
planting. Evaluation of the societal and environmental impacts  of reforestation and afforestation are also
required.
Table 5.1-3.    Examples  of Potential Wood Volume Growth and Calculated  Carbon Fixation Rates,
               and Representative  Establishment  Costs for Several Species  in the Tropical and
               Temperate  Zones.   Based on Opinions  of Individual Reforestation  Experts at an
               International Workshop  (Winjum and Schroeder  in preparation)

       Species                      Wood Volume Carbon*       Establishment
                                      m3/ha/yr     tC/ha/yr       Cost ($/ha)
Tropics:
Tropical hardwoods (Brazil)              35            15              800
Eucalyptus spp.   (Congo)              25            10             1400
Casuarina spp.    (India)                 9             4              200

Temperate:
Pinus radiata (New Zealand)             25            10              300
Picea sitchensis   (UK)                   14            6             1100
Pinus taeda     (US)                    9             4              300
Pseudotsuga menziesii  (US)             14            6             1700
Pseudotsuga menziesii  (Canada)          6             2.5           1000
1  Carbon sequestration rates are developed by multiplying wood volume by 1.6 to get total biomass
volume/ha/yr and by 0.26 to get t C/ha/yr (Sedjo and Solomon 1989). These rates are just for the first
rotation of the plantations and do not account for carbon lost by harvesting and  growing plantations in
perpetuity.    •
                                               96

-------
Table 5.1-4.     Global  Reforestation/Afforestation   Rates  in the 1980s
Region/Area
  Area
ha/yrx 1000
Reference
Canada
Scandinavia
USSR

Temperate:

Argentina
China
India
Japan
South Africa
United  States
Western Europe1

Tropical:

Africa
Asia
Latin America

Total
 1122
 6202
 5502
  55
7.0002
5.0002
 240
 110
1,200
 475
  127
  439
  534

16,462
Scarratt et al. 1982
Scarratt et al. 1982
Postel and Heise 1988
Postel and Heise 1988
Lonqjun circa 1986
Sharma et al. 1989
Postel and Heise 1988
Postel and Heise 1988
Foreward et al. 1990
Postel and Heise 1988
Postel and Heise 1988
Postel and Heise 1988
Postel and Heise 1988
        Estimate is for 1970s
        Land area estimated by dividing the reported annual numbers of seedlings planted by 1,200
        seedlings/ha.
                                               97

-------
5.1.2 Calculation of the Carbon Sequestering Potential of Temperate Plantations: An Example

The following sections report the results of a study conducted at ERL-C (Schroeder and Ladd in press),
the goal of which was to calculate the area of forests that would have to be planted in the United States
to store the total anthropogenic carbon emissions of the United States over a 50-year period.

5.1.2.1 Approach

The United States emits about 1 billion tons of carbon per year, which  comes to 50 billion tons over a
50-year period If we also assume that emission rates do not change (Manne and Richels in press [a,b]).
Fifty  years is an arbitrary assumption. What is important here is the approach used to calculate the area
of plantations that would be required, given a 50-year time period.  Furthermore, the analysis only
accounted for the amount of carbon that could be stored during the 50-year period and did not consider
its fate at the end of that time.  Also, the analysis did not consider potential climate change effects on
forests.  As discussed in Section 3.2, vegetation redistribution could create a significant pulse of carbon
to the atmosphere, further increasing atmospheric CC^ concentration and making reforestation  efforts
more complicated and potentially less effective. Decomposition has already been accounted for in the
following estimates.

The forestation/mitigation analysis considered two tree species important in the South and Northwest,
respectively:  loblolly pine and Douglas-fir (Barrett 1980).  Calculations  of stand growth and carbon
fixation rates were based on output from standard forestry growth and  yield models. These are
empirical models commonly used to predict the volume growth of stemwood that can be  converted into
usable products.  For loblolly pine, the analysis used the model PTAEDA2 (Burkhart et al. 1987), and for
Douglas-fir the model DFSIM (Curtis et al. 1982) was used.  Because they are empirical models,
PTAEDA2 and DFSIM assume that future stand development will be similar to past  stand development
(i.e.,  similar to the data on which the models are based).  They therefore cannot account for the
potentially complex influence of a changing climate.

As Sedjo (1989)  assumed, 1 m3of stemwood was associated with 1.6 m43of whole-tree biomass (roots,
branches, leaves, etc.). The analysis also followed both Marland's (1988) and Sedjo's (1989) analyses
by assuming that 1 m3of biomass contains 0.26 tons of carbon (tC). As with previous studies,  the
present one did not account for carbon storage in forest floor detritus or soil, nor was the potential
impact of atmospheric C02 enrichment itself considered.
                                               98

-------
5.1.2.2 Results

The carbon fixation rate for loblolly pine was 4.0 tC/ha/yr (Figure 5.1-1), or a total carbon storage
capacity of 200 tons C/ha over 50 years. For Douglas-fir the annual fixation rate was 5.2 tons C/ha/yr
(Figure 5.1-1), and storage was 260 tons C/ha for the 50-year period.  It would thus take 250 million ha
of loblolly pine or 192 million ha of Douglas-fir to capture and  store the 50 billion tons of carbon that
would be emitted over the 50-year time period.  For perspective, consider that 200 million ha Is about
26% of the total area of the 48 contiguous United States, and the total worldwide area of commercial
forest plantations in the mid-1980s was about 92 million  ha (Postel and Heise 1988).

The reason that 30% more area is required to store the same amount of carbon  with loblolly pine than
with  Douglas-fir over a 50-year period is illustrated in part by Figure 5.1-2 and is  explained by Figure 5.1-
1. Although the two species may have similar rates of carbon fixation at certain  points in their life cycle,
Douglas-fir is structurally capable of higher levels of carbon storage.  The point of maximum growth for
loblolly pine occurs relatively early, before much carbon has accumulated (e.g., 20-30 yrs). Indeed,
throughout much of the assumed 50-year period, mean annual increment (MAI) for loblolly pine actually
declined, and total yield or carbon storage was nearly constant. The pine stand essentially reached its
maximum carbon storage capacity at approximately age 30. For the next 20 years little additional
carbon was accumulated. Over a period greater than 50 years, therefore, the disparity in land
requirements for the two species would be even greater. Continuing high rates of fixation while
maintaining  carbon in storage may be difficult with loblolly  pine.
                                                99

-------
 E
 0)
 J-  X-K
 o  u
 c
~»  CQ
 2~3
 c
 c
 C
 CO
          0
                                                                          Douglas-fir
                                                    Loblolly
                                                      pine
             0
10
20
30
40
50
                                           Age
   Figure 5.1-1.   Comparison of changes  in estimated  mean annual growth  increment  of Douglas-fir
                 and loblolly pine.  Estimates assume  Douglas-fir site index 125 (base age 50 yrs)
                 and initial planting density  of 1000 trees/ha, loblolly  pine  site  Index 70 (base age 25
                 yrs), and initial planting density of 1500 trees/ha.  (Site index  is a forestry
                 productivity classification  concept.  It is simply the mean  height of the dominant
                 (tallest) trees at some particular age.  The concept  is based on the understanding
                 that height  growth is very heavily  influenced by, and therefore integrates  the effects
                 of, the physical  and environmental factors that determine  site  quality.)
                                                100

-------
        300
        250
        200
0)
                                                                Douglas-fir
        100
         50
          0
                                                                Loblollv
                                                                          w
                                                                   pine
                       10        20        30        40        50

                                      Age
 Figure 5.1-2.   Estimated yield curves of total standing crop for Douglas-fir and loblolly pine based
             on the same assumptions  as Figure 5.1-1.  Lower early yield for Douglas-fir (i.e.,
             7.5 tons C/ha at age  15) reflects the lower juvenile growth rate illustrated in Figure
             5.1-1.
                                        101

-------
One possible solution is to grow two 25-year crops of a species like loblolly pine during the 50-year
period, harvest trees at age 25 when they have achieved nearly maximum storage, and replace them
with a vigorous young stand.  For this approach to be successful, the harvested carbon must not find its
way back into the atmosphere (or must be used to replace fossil fuel).  Unfortunately, the amount of
carbon that can actually be stored in the form of durable products (e.g., lumber or plywood) is fairly
small.  Harmon et at. (1990) estimated that only about 42% of harvested stemwood carbon goes into
durable products. The  remainder is lost to paper production, fuel consumption, and decomposition.
(Unaccounted for, however, is carbon in an unknown amount of paper that can lie buried but
undecomposed in landfills for several decades.) This analysis accounted for branches, leaves, and roots
in addition to stems.  Since these components comprise about 40% of total carbon and since none of
them go  into durable products, only about 25% of the total carbon accumulation estimated here could
be converted to durable products. This illustrates the transitory nature of carbon storage in forests.

It was calculated that total yield for a 25-year crop of loblolly pine would be 133 tC/ha (Figure 5.1-2).
Because  75% of this amount returns to the atmosphere, however, only about 33 tC/ha would go into
long-term storage.  This 33 tons combined with an additional 133 tons from a second 25-year crop
(unharvested) would succeed in storing 166 tC/ha over a 50-year period. This is 17% less  than a single
50-year crop.

5.1.2.3 Conclusions

This analysis demonstrates the significant influence of forest growth patterns over time on achieving
carbon storage goals.  If the policy objective is to slow the increase in atmospheric CO2 concentrations,
then the emphasis should be on carbon storage over several decades. Concentration on short-term
average annual rates of carbon fixation without consideration of the growth dynamics of forest stands
over time can be misleading.  For periods of 50 or more years,  it is especially important to  be  mindful of
a species' carbon storage potential rather than of its potential maximum growth rate at some point
during its life cycle.

The analysis also shows that large areas  of land are required to offset the CO2 emissions of a developed
economy like that of the United States. Forestation alone may not solve the problem of increasing
atmospheric CO2 concentration, but it can result in the storage of large amounts of carbon. However,
establishment of new forests can be a valuable component in a comprehensive solution that also
includes  reductions  in C02 emissions from fossil fuel combustion and deforestation.  It is unlikely that
the developed countries, the largest sources of anthropogenic CO^ have sufficient land available to
                                              102

-------
offset even their own C02 emissions through forestation (Shands and Hoffman 1987).  Consequently, any
strategy including forestation as a component must be implemented globally.
                                                           *
5.1.3  Calculation of the Carbon Sequestering Potential of Several Forest Stand Treatment Practices

This section analyzes the effects of thinning, fertilization, and vegetation control on carbon storage
based on output from temperate forest growth and yield models, and values derived from the literature.
The species, models, and methods used were the same as those described in the previous section.
Different assumptions, models, etc., are probably applicable to boreal and tropical species (Burley and
Stewart 1985).

5.1.3.1 Thinning

Thinning of forest stands is a common management practice despite the fact that, paradoxically, it may
actually reduce total yield (in terms of timber or carbon) over the length of a rotation (Smith 1986).
Simulated examples of this for Douglas-fir and loblolly pine are shown in Table 5.1-5.
Table 5.1-5.  The Effect of Stand Thinning on Carbon Sequestration
Species
Douglas-fir1
Loblolly Pine2
Unthinned
tC/ha
260
170
Thinned
tC/ha
210
145
%Reduction
19
15
Including
Thinnings
tC/ha
240
151
%Reduction
8
11
1 Simulation assumptions:  Reasonably good quality site (site index 125 at age 50), planted at 750
trees/ha, 50 yr rotation thinned once at age 35.
2 Simulation assumptions:  Reasonably good quality site (site index 70 at age 25), planted at 1500
trees/ha, 35 yr rotation thinned once at age 25.
By redistributing stand growth to a smaller number of larger trees, thinning can be used to increase the
value of merchantable stemwood volume even if total production of wood, or storage of carbon, is
unaltered or even somewhat reduced.  In his classic silviculture text, Smith (1986) explains that for a
given age, composition, and site, total stand volume production is essentially constant for a wide range
                                              103

-------
of stocking density.  It can be decreased, but not increased, by altering the amount of growing stock to
levels outside of this range. Farnum et al. (1983) allowed that small increases in productivity may result
from early thinnings of dense stands that leave only well-spaced stems on the best microsites, but they
also concluded that overall productivity is not changed substantially by thinning.

5.1.3.2  Nutrient fertilization

The objective of nutrient fertilization is to increase tree growth by supplying the nutrient elements that
are limiting growth (Pritchett and Fisher 1987). A comprehensive review of literature on forest fertilization
is beyond the scope of this report (Dangerfield and Brix 1981). Bather, this section will demonstrate
some concepts and principles regarding the effectiveness of fertilization in enhancing  carbon storage in
forests (Shands and Hoffman 1987).  The results cited below were reported in the literature in terms of
growth or volume responses to fertilization, and they have been converted to estimates of carbon
storage for this report.

General estimates of fertilizer response

Douglas-fir generally responds positively to nitrogen fertilization.  However, the magnitude of the
response can vary greatly, which makes it difficult to predict quantitatively as a function of soils (Ballard
and Shumway 1986).  The estimates that follow should therefore be considered as general and subject
to fairly wide variation. Farnum et al. (1983) estimated that repeated periodic application of  nitrogen
fertilizer would increase the maximum mean annual growth increment of Douglas-fir by 20% over a
rotation. Miller et al. (1986) reported results from an extensive set of nitrogen fertilization trials for
Douglas-fir.  They observed increases in basal area growth (roughly comparable to volume growth and,
therefore, to carbon fixation rate) of about 20% over a six-year period for a variety of application rates,
site conditions, and stand conditions in the Pacific Northwest.  Peterson and  Gessel (1982) reported that
the fertilized stands in the same study were outgrowing controls by about 18% or 1.5  tC/ha/yr after
eight years.

The response of loblolly pine to fertilization is also variable.  Farnum et al. (1983) estimated increases in
mean annual volume growth of  loblolly pine of 20% as a result of phosphorus fertilization. They
estimated an increase of an additional 12% for nitrogen fertilization. Comerford et al.  (1982) reported
that phosphorus fertilization increased the annual growth of a young pine stand by about 64% or just
over 3 tC/ha/yr. Ballard (1981) concluded that nitrogen fertilization can produce a volume increase on a
wide variety of sites used for loblolly pine plantations.  He reported gains of 1-2 tC/ha/yr over a period
                                               104

-------
of five years following nitrogen application. Ballard et al. (1981) reported increases of about 1 tC/ha/yr
for the four-year period following nitrogen fertilization for 12 loblolly pine plantations.

Fertilizer interactions

Fertilization can Interact with site quality,  stand density, and age.  Figure 5.1-3 is derived from data
contained in Miller et al. (1986), and it shows a typical inverse relationship between site quality and
nitrogen fertilizer response. Lower quality sites generally respond more positively to fertilizer application
than better quality sites, both in relative and absolute terms.  Indeed, over the 10-year period presented,
the fertilizer response of the lowest quality sites was about fourfold that of the highest quality sites.  The
same general  relationship was also observed by Peterson and Gessel (1982).
                                                 105

-------

•o
.2
>»
.E
c
*s
O
S
o

V
I-

ou
40

C
to
« - 30
N «
*^™ ^^^
••• ^n^
*c w
^^1 ^BB
1 E »
0
**"
10
n
•



.



•


•
>\
                       20          30          40
                                  Site Index

                       Height (meters) at age  50
50
Figure 5.1-3.   Cumulative 10-year increases in Douglas-fir stem biomass  (carbon gain) resulting
            from fertilization with 224 kgN/ha (from Miller et al. 1986).
                                       106

-------
Regarding interactions with stand density, Peterson (cited in Miller et al. 1986 and personal
communication) has pointed out that by increasing stand growth, and therefore increasing competition
for space and other resources, fertilization may increase mortality in unthinned stands. When thinned
stands are fertilized, all or most of the growth increase stays in the living forest, and is not returned to
the atmosphere via decomposition and respiration. Although thinning by itself does not increase total
carbon storage, fertilizing stands that have already been thinned, or that are soon going to be thinned,
appears to offer a good opportunity for gains in carbon storage.  Douglas-fir and loblolly pine have both
been shown to respond positively to fertilization over a wide range of age.  Positive fertilizer response for
Douglas-fir has been observed in stands from 9 to 120 years of age (Miller and Webster 1979, Miller et
al. 1986, Peterson and Gessel 1982).  Loblolly pine is not as long-lived as Douglas-fir, but Ballard's
(1981) data referenced above include  stands that are 4 to 21 years old.

Carbon cost of fertilizer production

The production of commercial fertilizers must also be considered with respect  to carbon storage.
Ammonia, the base component for the nitrogen fertilizer industry, is produced from atmospheric nitrogen
and hydrogen from natural gas or petroleum.  Carbon dioxide is produced as a byproduct when the
carbon atom is stripped from the natural  gas hydrocarbon to leave the hydrogen atoms.  Approximately
4 tons of COaare produced for every ton of ammonia manufactured  (Sittig 1979). These  reactions are
also carried  out at high temperature and  pressure, which requires energy inputs and subsequent
additional CO2 release. A common nitrogen (N) fertilization application rate is  about 250 kgN/ha, or
0.25 tons N. Manufacturing this amount of fertilizer would, therefore, entail the production of at least a
ton of C02 containing 270 kg carbon. This means that the net carbon storage would be reduced by 2-
5% over a period  of 5-8 years,  based on data presented above.  Other C02 costs of forest management,
primarily from energy usage, should also be noted.

Effects of fertilization on other ecosystem  properties

Besides increasing the rate of plant growth, additions of nutrients may affect a number of other
processes in ecosystems.  Some examples of these processes include decomposition rates (Melillo et al.
1982), rates  of grazing activity (Mattson and Addy 1975), competitive abilities of plants (Ruess et al.
1983), carbon allocation to roots  (Hermann 1977), or even soil chemical properties (Ruess and Johnson
1986).  These effects in turn may act to further increase the potential to sequester carbon if added in
moderate amounts, but they also could reduce carbon sequestration if added  in excess.  An example of
                                              107

-------
excess amounts of nutrient inputs is the chronic deposition of acidic compounds into certain sensitive
forests (Schulze 1989, Materna 1989).

5.1.3.3 Control of competing vegetation

The control of competing vegetation in young plantations is another common practice (Gjerstad et al.
1984) important for enhancing carbon storage capability. A few examples for loblolly pine simulated with
the PTAEDA2 model illustrate the relative sizes of potential effects of competition on growth carbon
storage.  A low level of competition (5% of the total stand stocking) would result in a 13% growth loss or
about 23 tC/ha over a 35-year rotation. If this competition were eliminated at an early age, such as
within the first five to eight years, there would  be no growth loss.  A high level of competition (15% of
stand stocking) would result in a 35% growth loss over 35 years, which is equivalent to nearly 60 tC/ha.
Controlling this level of competition early in stand development would prevent nearly all of the growth
loss.  However, if controlling this level of competition is delayed, until age 20 for example, final carbon
storage would be reduced by about 31 tC/ha  or 18%.

5.1.3.4 Discussion and conclusions

Reforestation or afforestation of  lands currently denuded of trees will ultimately result in larger amounts
of carbon storage than can be attained by  intensifying management of existing forests. Some of the
growth data presented  here, in the previous section, and elsewhere (Schroeder and Ladd in press) show
potential  carbon accumulation rates of 4-6  tC/ha/yr for reforested areas, as compared with the 1-2
tC/ha/yr increase in carbon accumulation for  other forest management practices.  On a unit area basis,
this increase is still notable.  On a global basis, however, the impact of intensified forest management on
atmospheric CO2will depend on the area of forest that can be successfully utilized.

This discussion of forestry practices and the dynamics of forest growth and  development leads to
several conclusions about carbon storage in temperate Douglas-fir and loblolly pine forests:

1)     Thinning does not generally increase carbon storage, although thinning dense young stands
       may increase carbon storage over  the life of the stand (Section 5.1.3.1).

2)     Fertilization generally increases carbon storage,  although the actual level of response can be
       difficult to predict (Section 5.1.3.2).
                                               108

-------
3)      Fertilization may result in greater carbon storage on low-quality sites than on high-quality sites
        (Section 5.1.3.2).

4)      Forest stands over a wide range of ages respond positively to fertilization, which may offer
        opportunities for increasing carbon storage (Section 5.1.3.2).

5)      Fertilizing unthinned or dense stands increases net carbon storage, but fertilizing thinned or less
        dense stands results in an even greater net removal of carbon from the atmosphere (Section
        5.1.3.2).

6)      Control of competing vegetation maximizes carbon storage in trees over the life of a stand
        (Section 5.1.3.3).

5.1.4  Research Needs

The analyses discussed above in Sections 5.1.1 to 5.1.3 suggest that forest management, particularly
reforestation/afforestation, can play a role in reducing the rate of increase of atmospheric CO2  These
analyses, however, are admittedly preliminary. Uncertainties in the estimates presented here can be
reduced by completing the following research tasks:

1)      Inventory and classify land areas that are suitable and available for reforestation/afforestation.

2)      Determine rates of carbon fixation and storage for other tree species, particularly tropical
        species.

3)      Estimate the carbon storage potential of agroforestry systems, including  carbon accumulated as
        soil organic matter.

4)      Determine the effectiveness of agroforestry to reduce rates of deforestation  by reducing the area
        of land required to meet agricultural and wood needs.
5)      Develop methods to reforest land that has been degraded by overuse or misuse.
                                                109

-------
52     Soil Management  to Conserve and Sequester  Carbon

Soil carbon in tropical, temperate, and boreal forests accounts for 40% of the 1500 Gt of carbon
estimated to be in world soils (Waring and Schlesinger 1985).  Land use practices such as harvesting
and permanent clearing can affect these soils and can lead to emissions of carbon to the atmosphere
(Jain et al. 1989). Evaluation of strategies to manage and use world forests to sequester and store
carbon should consider the role that soils play in carbon storage as well as their role in reforestation and
afforestation (Dixon in press, c).

Recently, at a workshop held by EPA at ERL-C, policy and science experts on agricultural and forest
soils considered the potential role of soils in mitigating increases in atmospheric C02 (see Appendix B
for list of attendees). The principal workshop question was, "Can soils be used to store sufficient carbon
to aid in mitigating the extent of global climate change?" The participants concluded that data are
currently insufficient to answer this  question definitively.

There was, however, a general consensus that three strategies could promote carbon storage in soils.
The first is to implement soil management techniques that maintain current pools  of soil carbon and that
minimize the loss of carbon from soils to the atmosphere (as either C02 or methane), such as
minimizing soil disturbance during forest harvesting (Parton et al. 1988, Miller and Sirois 1986, Mattson
and Swank 1989).  The second approach  is to implement management practices that would  restore soil
carbon in carbon-depleted soils (Oades 1988, Pritchett and Fisher 1987); for example, reforesting
marginal lands that were once cleared for agriculture. The third strategy is to promote management
techniques to enlarge  the size of the existing soil carbon pool. For instance, soil carbon sequestering
could be enhanced  by retaining slash on site instead of  removing it by burning (Benson 1982, Jain et al.
1989, Yonker et al. 1988).  Workshop participants produced a list of specific management tools for both
agricultural and forested systems that could be implemented under each of these three approaches, and
ranked their relative priority.  These tools are shown in Table 5.2-1 and a subset is discussed in some
detail below.

5.2.1  Maintaining the  Soil Carbon Pool

As described in Section 2.3, the carbon content of soils  is a function of a variety of factors including past
and present management.  Historically, forest soils have not been managed to conserve carbon
(Pritchett and Fisher 1987). In fact, some forest management practices are purposefully designed to
remove slash and detrital carbon to facilitate or expedite replanting.  Other practices such as the
                                              110

-------
conversion of forested lands to agriculture can lead to the net loss of soil carbon (Giddens 1957,
Delcourt and Harris 1980, Allen 1985, Schlesinger et al. 1990).  In all likelihood, such land use changes
will mean loss of soil carbon to the atmosphere.

The world demand for forest products is not likely to decline in the immediate future, so forests will
continue to be a source of raw materials (USDA 1988). Changing the way forest soils are currently
managed, however, can minimize the loss of soil carbon and promote longer term sustainability of these
systems. Five potential approaches are: 1) maintaining forest soil fertility; 2) retaining forest slash and
residues on site; 3)  using prescribed burning to maximize carbon storage; 4) minimizing site
disturbance; and 5) controlling erosion.
                                               111

-------
Table 52-1.  Prioritized Strategies  for Sequestering  and Storing Carbon in Soils 1  -

                                                                   Approach2

Management Practice                                Maintain       Restore        Enlarge

Minimum Tillage

Reforestation
Maintain or Improve Soil
        Fertility

Retain Forest Slash on Site

Leave Crop Residues

Incorporate Crop Residues

Intensify Tropical Agriculture

Increase Efficiency of  Forest
        Product Use

Minimize Site Disturbance

Use Prescribed Burning to
        Maximize Carbon Storage

Take Marginal Lands out of
        Intensive Agricultural Production

Control Erosion

Preserve Natural Wetlands

Use Municipal, Animal, Industrial
        and Food Processing Wastes

Urban Forestry

Minimize Dryland Fallowing

Use Mulching
M3
H


M

M

L

H

H


M

M
M

H
M
H

H
M
               M
M
               H
               H
H


M

M

M

H
               M
               M
               H

               M
i Recommendations/suggestions made by general consensus of scientific panel at ERL-C Workshop,
  February 1990.  List of participants is given in Appendix B.

2 Maintain, Restore, and Enlarge are defined in the text (Sections 5.2.1, 5.2.2, 5.2.3)

3 Priority: H, high; M, moderate; L, low.
                                               112

-------
5.2.1.1 Maintaining forest soil fertility

The principal objective of this management strategy is to keep the land in production (Barrett-Lennard et
al. 1986).  Maintaining soil fertility simply means the implementation of management practices that
prevent the degradation or loss of the capacity of the soil to supply nutrients. Practices such as
interplanting with nitrogen-fixing plants (Binkley et al.  1982) may lead to more fertile systems and in all
likelihood  more productive systems on a sustained basis (Burley and Stewart 1985).

5.2.1.2 Retaining forest slash and residues on site

Forest harvest residues  are an important source of organic matter and  nutrients for the next rotation as
well as an important substrate for heterotrophic organisms (Swift  1977, Harmon et al. 1986, Pritchett and
Fisher 1987). Quite often these residues are either removed or burned following forest harvesting.
Burning and heterotrophic decomposition of these materials both result in the emission of CO2  Burning
also results in the loss of nitrogen to the atmosphere and may damage the soil and the soil biology
(Office of Technology Assessment [OTA] 1984).  Burning may also leave the soil bare, making it more
susceptible to erosion, particularly in humid regions on steep slopes.  If the organic matter is retained on
site and incorporated  into the soil, net carbon emissions to the atmosphere could be reduced (Yonker et
al. 1988) and site productivity increased (Jain et al.  1989).

5.2.1.3  Using prescribed burning to maximize carbon storage

Prescribed burning has  been used for centuries as an effective method for removing slash following a
forest harvest or in shifting to agriculture. Changing the  way slash is burned  may, in some situations,
increase rather  than decrease, the potential for a site to sequester and store carbon (Pritchett and Fisher
1987). Very hot burns (i.e., hot dry and windy weather and dry slash) may severely damage the soil.
Such  burns can be so hot that they combust organic matter in the forest floor and soil, destroying seed
beds, still-living root systems, mycorrhizae, and free-living soil organisms (Waring and Schlesinger 1985,
Perry et al. 1989).

5.2.1.4  Minimizing site disturbance

Some logging systems during forest  harvesting may compact or disrupt a large portion of the soil in the
harvest areas (Miller and Sirois 1986 Smith 1986). This could reduce the productivity of the site and
                                               113

-------
promote the loss of soil carbon. Harvest and management practices that minimize or eliminate site
disturbance will help maintain the productivity of the forest system (Barrett 1980, Perry et at. 1989).

5.2.1.5  Erosion control

Erosion results in the loss of the soil resource and with it the potential to sequester and store carbon
(Parton et al. 1988).  The purpose of controlling soil erosion is to maintain the soil resource, the
productivity of the system, and  the potential for carbon to be stored belowground.

5.2.2  Restoring Soil  Carbon

Worldwide, millions of hectares of once-forested land may be  denuded of vegetation (Grainger 1988).
Accompanying the loss of vegetative cover is the loss of soil carbon (Schlesinger et al. 1990).
Management practices that restore these lands into production will lead to  increased carbon
sequestration and storage both above- and belowground (Dixon in press.c).

Strategies for restoring soil carbon in forested (or once forested)  systems include: 1) reforestation
(Wiersum 1984); 2) intensifying  tropical agriculture (MacDicken and Vergara 1990); 3)  improving soil
fertility (Pritchett and Fisher 1987); 4)  removing marginal lands from intensive agricultural production
(Barrett-Lennard et al. 1986); and 5) urban forestry (Grey and  Deneke  1978).

5.2.3  Enlarging the Soil Carbon Poo!

Increasing the productivity of ecosystems by enhancing or enlarging the size of the soil carbon pool is a
third strategy for storing more carbon belowground (Dixon in press.c). Most of the opportunities to
enlarge this pool may be in agriculture because agricultural systems are generally more intensively
managed than forested systems (Smit et al. 1988). Of the three strategies for storing  carbon in soils,
this approach has the lowest priority.  Preliminary consideration suggests that the marginal return,
measured in stored carbon, would be greater if the first two strategies are implemented.  However, some
opportunities exist to enlarge soil carbon pools in forested systems, including improving soil fertility and
retaining forest slash on site (Pritchett and Fisher 1987, Shands and Hoffman 1987).

To expand  on one of these strategies, the addition of fertilizers to soil may  Increase soil carbon pools
above their normal, steady-state levels.  Fertilizers can increase carbon flow in the soils via increasing
the rate of carbon fixation into plants and thus also increasing detrital inputs into soils (Attiwill 1986).
                                               114

-------
A second and less understood way fertilizers may increase soil carbon pools is by reducing
decomposition and CO2 efflux back to the atmosphere. For example, following nitrogen fertilizer
treatments in forests in Sweden, reduced decomposition and reduced soil respiration have been
observed (Nohrstedt et at.  1989). Similarly, nitrogen additions have caused a substantial reduction in
C02 efflux to the atmosphere at a forest site in the Oregon Cascades (Figure 5.2-1). The causes for the
reduction are unknown at the  present.  Hypotheses include altered soil solution chemistry, shifts in
microbial population, suppression of lignolytic activity of fungi, and reduced root  respiration.  The large
depression of CO2 efflux suggests that fertilization is an important factor controlling carbon flow.

5.2.4  Future Considerations

The strategies discussed here are aimed at storing carbon in soils of forest ecosystems. The
approaches appear to be applicable to both temperate and tropical forest systems (Jain et al. 1990,
Pritchett and Fisher 1987).  Uncertainties exist in the implementation of these practices because negative
effects associated with a given practice may outweigh the positive benefits at particular sites. Under
certain circumstances, for  instance, some of the recommended carbon sequestering techniques could
lead to the emission of gases  that are more radiatively efficient than CO2 (e.g., reducing environments
caused  by intermittent flooding may enhance the efflux of methane and nitrous oxide,  while reducing the
efflux of COa), and therefore, in terms of global warming,  would be counterproductive  (Tans et al. 1990,
Post et  al. 1982).  The exact nature and extent of the uncertainties cannot be fully evaluated  without
more research (IPCC 1990). The implementation of these management practices should be  considered
on a site or regional basis and should be judged in terms of their effect on the carbon pools and fluxes
(Dixon in press, b).
                                               115

-------
  CO
 T3
CM
  £
3-
  III
  o
          2-
                                                  single   fertilization
                           double  fertilization
                           -i-
                           4
T"
 6
                                                             -r
                                                              8
10
12
                                      Time  (week  #)
                                   (June  18 to August 27, 1990)
Figure 52-1.   COzflux from soil and forest floor to the atmosphere  at fertilized Douglas-fir and
             western  hemlock forested sites  in the Oregon Cascades.  Reductions  of 35% in
             CO2 efflux occur as a result of added  nitrogen.  Nitrogen  additions (urea, 300
             kgN/ha) were added several years ago to both sites  and  again this spring to one
             site; i.e., the single and double  fertilized lines, respectively.
                                         116

-------
In addition to storing carbon assimilated from the atmosphere, managing forested soils to conserve
carbon has additional benefits, including: 1) increased soil water holding capacity; 2) increased nutrient
availability; 3) improved soil physical properties; and 4) decreased soil erosion by wind and water.
Costs for implementing various soil carbon conservation practices need to be determined.

5.2.5  Research Needs

The storage of carbon in soils is a complex process that is not fully characterized or understood as yet
(Post et al.  1982). To implement biospheric management strategies for mitigating global change, the
effects of specific management practices on soil carbon cycling must be more fully  understood (IPCC
1990). If the purpose is to capture atmospheric C02for storage in the biosphere, including storage in
soils, then the following areas require investigation:

1)      Identification and characterization of the specific pools of soil carbon that can be managed (Post
        et al. 1982).

2)      Development of methods for quantifying the pools and fluxes of soil carbon (Parton et al. 1988).

3)      Quantification of carbon pools and fluxes in specific ecosystems under ambient or  steady state
        conditions to develop general principles from the specific  examples (Tans et al. 1990).

4)      Characterization and quantification of the factors that control  carbon fluxes  (Schlesinger et al.
        1990).

5)      Development of simulation models that accurately project how carbon fluxes (and thus pools
        and feedbacks to the atmosphere) will shift in specific ecosystems under different management
        regimes and under a series of altered climate  scenarios (Tans et al. 1990, Dickinson 1986) .

6)      Quantification of the effects of land use and management on  soil carbon in  tropical, temperate,
        and forest ecosystems (Burley and Stewart 1985).

7)      Quantification of the economics of implementing soil management practices that sequester and
        store carbon (Smit et al. 1988, Pritchett and Fisher 1987).
                                               117

-------
53 Forest Mitigation Compared with Other Biosphere  Options

Establishment and management of forests to sequester carbon is one of several options to influence the
global carbon cycle (Dixon in press.c). Examples of other options are: 1) for terrestrial systems,
agroforestry,  agriculture, urban forestry, wood lot forestry, and rehabilitation of degraded lands; and 2)
for aquatic systems, wetland reconstruction and ocean fertilization (Charles 1989, IPCC 1990, Fulkerson
et al.  1989).  In most cases, these options would complement, not compete with, forest management for
atmospheric carbon mitigation.

5.3.1  Forestry Management Options

The two broad forest management options to sequester carbon are reforestation and improving growth
of existing stands (i.e., timber stand improvement or TSI)(Dixon in press.c).  In Sections 5.1.2 and 5.1.3,
the potential of these options  was discussed  using example analyses for two temperate forest regions in
the United States, Douglas-fir  in the Pacific Northwest and loblolly pine in the South. The conclusion
reached in Section 5.1.3.4 was that for each land unit in these forest regions reforestation can sequester
much more carbon than does intensifying management of existing forests.

Assuming this relationship holds for forests in other regions of the world, the potential on a global scale
would depend upon the amount of land area available to implement each practice.  Estimates of land
available for reforestation in the tropics vary (Grainger 1988).  A value of 424 million ha was used here in
Section 5.1.1.1 to calculate potential carbon sequestration in the tropics. For TSI the figure is unknown,
though it is likely to be only a small percentage of the world's closed forests, which encompass about
2700 million ha (Table 1.4-1).  Therefore, based upon  even these preliminary estimates, reforestation still
appears to have substantially  more global carbon sequestering potential than TSI. Other factors such as
cost and the  sociopolitical considerations would also be most important to any comprehensive
determination.

5.3.2  Other Forest Management  and Terrestrial Options to Sequester Carbon

Other forest management approaches  with potential to contribute to global carbon sequestering  include:
urban tree planting, alternative wood harvesting and utilization methods to retain stored carbon, and
plantation establishment for biomass energy through various conversion technologies.  Definitive
evaluations of these approaches for carbon sequestering are a few years away (Trexler 1990), but
preliminary analyses lead to the following tentative conclusions.
                                              118

-------
Planting trees in urban areas around houses and low buildings can reduce year-round demands for
energy from fossil fuels (Grey and Deneke 1978).  In the summer in temperate regions, shade trees near
buildings reduce air conditioning demands during hot weather.  In the winter, trees around houses and
other structures used by people act as windbreaks, thereby decreasing heating energy needs (Trexler
1990).  Ongoing studies by the United States Department of Energy (DOE) and others are attempting to
clarify cost/benefit numbers for urban tree planting on a national basis (Wright in press).

Whether started naturally or by planting, forests ultimately mature, and,  if they are not harvested, trees
senesce and eventually die like all living organisms (Harmon et al. 1990). The wood then oxidizes,
releasing carbon back to the atmosphere as C02 Durable wood products can continue to store the
fixed carbon, but it is estimated that less than half of the wood from harvested trees goes Into durable
wood products (Harmon et al. 1990), and that the life of most durable wood products is usually a few
decades at best.  Research and development designed to improve this  carbon sink is continuing;
examples are improving forest utilization to reduce wood waste, increasing the life of wood products,
and expanding wood recycling opportunities (USDA 1990).

The concept of tree plantations as a source of energy has been under study for some time by DOE
(Wright in press).  Of interest are short-rotation woody crops using fast-growing poplar species
regenerated from stump sprouts (coppicing) about every eight years. The concept is to harvest wood
for burning to generate electricity or to be converted by thermal gasification to ethanol, methanol,
biocrude gasoline, and other synthetic liquid or gaseous fuels (Trexler 1990).  Though these wood uses
release C02 to the atmosphere, this would  reduce fossil fuel burning, which emits higher amounts of
C02.  Using wood for the above bioenergy purposes is still two or more times more expensive than
using fossil fuels,  but technological development continues to improve them from the standpoint of
economic attractiveness.

Another important tree-growing option is agroforestry or the combination of tree and agricultural crops
on the same land unit (MacDicken and Vergara 1990). Estimates of carbon sequestering rates for
agroforestry range from 10 to 30 tC/ha/yr (Gregersen et al. 1989). The high end of this range results
from mixtures of trees and agronomic crops in a few locations in the tropics; their productivity
represents some of the highest carbon sequestering values reported for managed ecosystems (Burley
and Stewart 1985). The values include both above- and belowground sequestering and are applicable to
the peak years in a crop rotation (i.e., they are not long-term averages). The land available for
agroforestry is mostly in the tropics, but definitive estimates of available land are still being developed.
Thus, in addition to conventional forest management, there are other options to use trees to mitigate the
                                              119

-------
buildup of CO2 in the atmosphere.  Each option has multiple benefits similar to forest management and
needs much more study and quantification.  Eventually comparisons should be made based upon a
comprehensive analysis of data that quantifies: 1) the tons of carbon stored annually per ha; 2) the cost
per ton of carbon fixed; and 3) the land or water area available for each option  (IPCC 1990). Social and
environmental impacts must also be evaluated.

Managing lands for agricultural crops can favor carbon in soils. The  strategies discussed in Section 5.2
for sequestering  carbon in soils (maintaining, restoring, and enlarging soil carbon) also apply here
(Parton et al.  1988). In the past, agricultural practices have often depleted soil carbon, resulting in
reduced production.  For example, temperate grassland soils have about 10 kg C/m2(100 tC/ha), but
they can lose 30-40%  through cultivation in about 20 years (Post et al. 1982, World Resources Institute
1988, Burke et al. 1989).  Tropical grasslands start with only 40% of the amount of carbon in temperate
soils, but they can be  depleted in the same manner (Dregne 1983).

World agricultural lands total 850 million ha,  which is generally higher than estimates of land available for
reforestation (i.e., 424  million ha, Table 5.1-2). Analyses are needed  to define how much of the
agricultural lands can, through improved management,  attain net gains in carbon while producing
profitable crops (Parton et al.  1988, Smit et al. 1988).

Rehabilitating lands with degraded soils offers carbon sequestration opportunities, but fixation rates may
be low and costs high compared with forests and agriculture (Barrett-Lennard et al.  1986).  However,
past land abuses in the world have created an amount  of land requiring rehabilitation (1400 million ha)
that is greater than the amount of land available for any of the other  terrestrial options (Grainger 1988).
Ultimately, society will have to confront this option if ever-increasing  populations are to be fed and
supported (Crosson and Rosenberg 1989). The potential for carbon sequestration by this  means may
help to accelerate investments in this option (Jain et al. 1989).

5.3.3  Aquatic Options

Much less is known about aquatic options to sequester C02. For instance, ocean fertilization is a
strategy whereby iron  compounds would be added to marine ecosystems to promote phytoplankton
populations (Martin and Fitzwater 1988).  These organisms would take up CO2 by photosynthesis and
eventually contribute to carbon storage in seafloor sediments. Although potentially effective and
economical by first estimates, this method carries unknown ecological risks such as disruption of marine
food chains.
                                              120

-------
Wetlands have, on average, much higher rates of carbon sequestration and storage than do most forest
ecosystems.  Swamps, marshes, and estuaries have a mean net primary production estimated to be
1000 g C/m2/yr, which is the same level as tropical forests, the most productive of all the world's
ecosystem types (Waring and Schlesinger 1985).  In addition, swamps and marshes have the highest
levels of soil carbon storage in the terrestrial biosphere.  Mean soil organic levels are about 69 kg C/m2,
which is more than three times greater than the second highest levels of about 20 kg C/m2 for
grasslands, tundra, and alpine areas (Waring and Schlesinger 1985). These ecosystems are  potentially
major sources of greenhouse gases (e.g., methane).

Wetlands have greatly diminished  in size due to human impacts.  For example, Heimlich et al. (1989)
stated that"... since European settlement, more than half of the original 215 million acres (87 million ha)
have been drained and converted to other land uses," largely for agriculture.  The most effective way to
manage wetlands for carbon sequestration is to preserve what is remaining.  Preservation is expensive
because much of the remaining wetlands are in private ownership, so that, in most cases, the landowner
must be compensated.  In the United States, for example, wetland purchases by federal agencies
average about $2000/ha; permanent easements are $700/ha; and rental rates average $40/ha/yr.  On
former wetlands, restoration is a second approach to managing wetlands, but the costs for restoration
are even higher than costs for preservation.  Heimlich et al. (1989) report that restoration costs in the
United States have ranged from $l20/ha to $3000/ha.  Research on more efficient methods  to manage
and/or restore wetlands would likely improve the opportunity to utilize their carbon sequestering
capability.

5.3.4  Conclusions

Reforestation or afforestation approaches have a greater potential for sequestering carbon than
intensification of forest management. However, land availability is critical for evaluating the potential
magnitude of carbon sequestration by reforestation/afforestation.

The largest potential probably lies in the tropics, both  in terms of area and potential productivity.
Grainger (1988) has estimated that over 700 million ha of land in the tropics could potentially be
forested.  Although some of this area is semi-arid,  much of it is in humid zones and once supported
forest cover.  Indeed, tropical forests are capable of storing very large amounts of carbon (Evans 1982).
Some of this land, however, may have suffered various forms of degradation, and its potential
productivity must be evaluated (Jain et al. 1989). To further assess the feasibility of global forestation, a
more detailed inventory and classification of available lands is needed.
                                               121

-------
Any global strategy for reducing atmospheric C02 must also address the issue of deforestation
(Woodwell et al. 1983).  The importance of carbon inputs to the atmosphere from tropical deforestation
and burning are well known, although the actual magnitude is uncertain (Woodwell et al. 1983, Detweiler
and Hall  1988, Houghton 1990a).  It can be said that annual rates of deforestation are in the millions of
hectares.  Houghton (1990) reported that"... in 1980, the area of tropical forests cleared annually for
other uses, largely agricultural, was the size of Pennsylvania," or 11.3 million ha. The World Resources
Institute (1990)  now estimates that this level of deforestation has almost doubled.

Conserving existing forests is a much more immediate and direct contribution to solving the C02
problem than is planting new trees (Harmon et al.  1990, IPCC 1990). This is because existing forests
contain levels of biomass and stored carbon not likely to be achieved by plantations (at least not over
decade-long periods) and  because new plantings can take many years to achieve high growth rates and
hence high sequestration rates.

In addition to plantation forestry, variations on forestation practices can be implemented to relieve some
of the pressure and demand on the world's forests. These practices include integrated land uses such
as agroforestry that use the land more intensively than, for example, shifting cultivation, and they provide
a variety  of products such as food, fuel, fiber, and fodder (Gregersen et al. 1989). Instituting these
practices to increase the productivity of currently underutilized lands will also directly increase the
amount of carbon stored in terrestrial vegetation and soils.

A broad-based  research program  is needed to quantitatively evaluate the potential for sequestering
carbon through reforestation and other forest management options. Specific recommendations are listed
in Sections 5.1.4 and 5.2.5. In brief, research is  needed to determine: 1) land availability, 2) techniques
and methods for reforestation and afforestation, 3) actual carbon sequestration rates based on field
experiments, and 4) the socioeconomic and environmental effects of a large-scale forestation program.
                                               122

-------
6      RESEARCH NEEDS AND PLANS

Sections 2 through 5 of this report presented specific research needs for the topics relevant to forest
ecosystems.  Future research at ERL-C will not be limited to forest ecosystems but will include all
dimensions of the terrestrial biosphere.  Presented here are three over-arching scientific questions that
encompass the entire ERL-C global research program for the next five years.  The questions are:

1)     How will  feedbacks from the terrestrial biosphere amplify or reduce climate change during the
       next 25-50 years?  What are the current and future fluxes of greenhouse gases from the
       terrestrial biosphere?

2)     What are the expected long-term responses of the terrestrial biosphere to global climate
       change?  Who will be the ecological winners and losers?

3)     Can the terrestrial biosphere be managed to lessen the impacts of global climate change by
       increasing carbon sequestration, reducing trace gas fluxes, and promoting a more favorable
       surface energy balance? What is the potential for wider use of forest management, particularly
       reforestation, afforestation, and agroforestry?

These three scientific questions emerged through the reviews  and research by ERL-C that have been
summarized in this report as preliminary results.  Over the next five years, the ERL-C effort will continue
as a key portion  of the broader Global Change Research Program (GCRP) within EPA's Office of
Research and Development (ORD).  Indeed, these questions and ORD's program will play an  important
role in the national research program on climate change as outlined  by the Committee on Earth and
Environmental Sciences (CEES) and in turn by the  IPCC (1990).

Full resolution, therefore, of the issues raised by these questions calls for coordination across many
organizations and fields  of expertise.  The process  has already begun and will need ongoing attention.
The research elements significant to EPA's contribution are outlined as follows.

6.1     Feedback Processes  Research  (Science  Question  #1)

Feedbacks to global change from the terrestrial biosphere, and particularly from forests,  are significant.
Several research priorities have emerged:
                                              123

-------
1)      Quantify the global carbon cycle, including pools and annual flux.  Basically, where is the carbon
        in the terrestrial biosphere?

2)      Evaluate the role of C02 enrichment and its interaction with stress agents on vegetation
        condition, distribution, and migration.

3)      Assess the current flux of radiatively important gases (COj, methane, carbon monoxide, and
        nonmethane hydrocarbons) from soils and vegetation.

4)      Evaluate the direction and magnitude of physical feedbacks (e.g., reflecting radiation by changes
        in albedo) to global change.

5)      Determine the emissions of greenhouse gases from forests as positive feedbacks to the
        atmosphere. Information  is needed about forests ranging from vigorous to declining and from
        natural  to managed, within the boreal, temperate, and tropical forest regions.

62      Response  Research  (Science  Question  #2)

Ecological impacts due to global change are projected to be significant. Relevant research issues
include:

1)      Determining the transient and long-term responses of vegetation (condition, migration,
        distribution) to global change.

2)      Defining limits of rate of change for various ecosystems (agroecosystems, forests, deserts).

3)      Characterizing the response of  hydrologic and energy cycles to global change.

4)      Assessing the impact of global  change on terrestrial biodiversity (OTA 1987).
                                               124

-------
63     Mitigation/Adaptation  Research  (Science  Question  #3)

Is It reasonable to expect that increasing ecosystem management throughout the terrestrial biosphere
can contribute to the mitigation of global change?   The preliminary results of analyses presented in
earlier sections make It appear promising, but research aimed at reducing the uncertainty is the next
step. Within EPA/ORD, ERL-C plans call for research beginning in FY91 to provide the following
information:

1)      Improved data  on the land area in the world that is available and suitable for more intensive
       ecosystem management to sequester atmospheric carbon;

2)      Carbon sequestering rates and the potential for storage of carbon above- and belowground for
       management options, including afforestation, agroforestry, and fuel wood forestry programs in
       the boreal, temperate, and tropical regions;

3)      A clear understanding of the global carbon cycle relative to natural and managed forest
       ecosystems throughout the world;

4)      More effective technology for the  propagation and silvicultural management of a wider array of
       forest species to  increase the cropping options across the mix of available lands;

5)      Refined projections about global warming, including climate scenarios to determine how the
       health and productivity of future forests will vary by location;

6)      Data about the response of forests to elevated levels of atmospheric C02, with emphasis on
       photosynthesis, WUE, and net ecosystem production;

7)      Effective techniques to manage forest soils to maintain, restore, and enlarge their carbon
       sequestration rates and storage capacity;

8)      Improved knowledge of biodiversity trends associated with natural  and managed forests as the
       forests change in response to future global climates;
                                               125

-------
9)      Case studies showing how forest management, particularly reforestation and agroforestry, can
        aid in offsetting both  deforestation and large-scale forest decline resulting from global climate
        change or changing land use that reduces the amount of productive land;

10)     Conceptual models based on a Geographic Information System (GIS) analysis to integrate
        results from the above research and to provide projections of the contributions of forest
        management on a world scale as an aid to mitigation of increases in atmospheric COa and

11)     An assessment of the contribution of forest management to carbon sequestering  In the broader
        concept of managing the biosphere for a sustainable living environment.

6.4     Field Assessment  of Forest and Agroforestry  Management  Opportunities to Sequester
        Atmospheric CO2

6.4.1  Goal

The goal of the assessment is to field test the feasibility of increasing carbon storage in the terrestrial
biosphere through managing the carbon cycle; i.e., increasing both above- and belowground carbon
storage through forestry, agroforestry,  and agricultural practices.

6.4.2  Rationale

This study would accelerate implementation of field demonstration trials for terrestrial management
options upon completion of the feasibility study. The aim is to be ready for early on-the-ground testing
as soon as clear guidance is available.

6.4.3  Approach

The strategy will be to establish a network of field trials in major global  biomes to determine the biologic
and economic feasibility of managing terrestrial carbon pools and fluxes.  The trials are expected to  be
established in representative forestry, agroforestry, and agricultural areas of the boreal, temperate, and
tropical regions. Research will be coordinated with existing groups such as the Food and Agriculture
Organization (FAO) of the United Nations, United States Agency for International Development (USAID),
and others.  Key components of this study are:
                                               126

-------
1)     Institute a system of research field trials on unmanaged or poorly managed lands in forested,
       agroforested, and agricultural regions to quantify the potential for expanding terrestrial carbon
       sequestration.  Degraded lands will be emphasized.

2)     Initiate a worldwide system of field surveys In aR major biomes to determine the carbon
       sequestering potential of representative forested, agroforested, and agricultural areas presently
       under management.

6.4.4  Product

In this study, demonstration sites and field trials will be established u:. 1) quantify the carbon
sequestering capability of managed forestry,  agroforestry, and agricultural areas in the world;  and 2)
estimate the potential to optimize that capability through more intensive management practices.

6.5     Assessment  and Validation of Terrestrial Biosphere  Models of Global Change Feedbacks
       and Responses

6.5.1  Goal

The goal of this study is to design and  implement  process research projects to allow better evaluations
and simulations of terrestrial biosphere  management options.  Ecophysiological factors and processes
i   regulate pools of carbon and  exchange  of greenhouse gases via vegetation and soils will L
studied.  Research in a minimum of five sites in the each of the following biomes will be establish^
wet tropical forest; temperate deciduous forest; xeromorphic shrubland;  estuarine wetland; mixed sub-
boreal forest; and boreal forest.

6.5.2  Rationale

Long-term assessments are needed to validate and calibrate large-scale models of globa. :   nge and
terrestrial biospheric responses and feedbacks.  A wide range of biomes should be evaluated because
ecological responses and feedbacks of climate change are expected to vary. Present mechanistic
models do not incorporate the effects of global change  on ecophysiological  processes such as
photosynthesis, transpiration, respiration, and growth, and large-scale models such as GCMs do not
Incorporate any biological process.
                                               127

-------
6.5.3  Approach

Research will be established In representative biomes to assess carbon cycling processes.  A range of
instruments and field exposure facilities will be used to assess above- and belowground processes
(meteorological, plant gas-exchange, soil biogeochemistry). This approach allows a determination of
system-level responses to abiotic and biotic stresses under present and enriched COa regimes and
under management scenarios.   Data collected will be used to calibrate and validate large-scale models
(e.g., GCMs).

6.5.4  Product

The outcome of this research will be General Circulation Model/Global Vegetation Model predictions of
biome specific responses and feedbacks to climate change. The GCMs will be calibrated and validated
from long-term data collected  in the field.
                                              128

-------
       LITERATURE CITED
Ajtay, G.L., P. Ketner, and P. Duvigneaud.   1979.  Terrestrial primary production and phytomass, pp.
        129-182. In: The Global Carbon Cycle. SCOPE Report No. 13. John Wiley, Chichester.

Allen, J.C.  1985.  Soil response to forest clearing in the United States and the tropics: Geological and
        biological factors.  Biotropics 17:15-27.

Allen, L.H. Jr.  1990.  Plant responses to rising carbon dioxide and potential interactions with air
        pollutants.  J. Environ. Qual.  19(1 ):15-34.

Allen, T.FJ.,  and T.B. Starr. 1982. Hierarchy: Perspectives for Ecological Complexity.  University of
        Chicago Press. Chicago.

Armentano,  T.V., and C.W.  Ralston.  1980.  The role of temperate zone forests in the global carbon
        cycle. Can. J. For. Res.  10:53-60.

Arno, S.F.  1984. Timberline: Mountain and Arctic Forest Frontiers.  The Mountaineers, Seattle, WA.

Arnone, JA., and  J.C. Gordon.  In press. Effect of nodulation, nitrogen fixation and C02 enrichment
        on the physiology, growth and dry mass allocation of Alnus rubra seedlings. New Phytol.

Attiwill, P.M. 1986.  Interactions between carbon and nutrients in the forest ecosystem.  Tree Physiol.
        2:389-399.

AuClair, A.N.D.  1985.  Postfire regeneration of plant and soil organic pools in a Picea mariana-Cladonia
        stellaris ecosystems. Can. J. For. Res.  15:279-291.

Bailey,  J.D.,  R.K, Dixon, and R.W.  Rose.  1990. Dormancy and cold-hardiness of Pseudotsuga
        menziesii, Alnus rubra, and Acer macrophyllum in an enriched-carbon dioxide environment.
        11th North American Forest Biology Workshop, June 13-15, Athens, GA.

Ballard, R.  1981.  Optimum nitrogen rates for fertilization of loblolly pine plantations. Southern J. Appl.
        For.  5:212-216.

Ballard, R., and J.S. Shumway.  1986. Soil and site information for  optimum management of
        Douglas-fir, pp. 168-176. In: C.O. Oliver,  D.P. Hanley, and J.A. Johnson, eds.  Douglas-fir: Stand
        Management for the Future. College of Forest Resources, University of Washington, Seattle,
        WA.

Ballard, R., H.W. Duzan, and M.B.  Kane.  1981. Thinning and fertilization of loblolly pine plantations,
        pp.100-104. In: J.P. Barnett, ed. Proceedings of the First Biennial Southern Silviculture
        Research Conference.  Genera! Technical Report SO-34, USDA Forest Service, Southern Forest
        Experiment Station, New Orleans, LA.

Barker, J.R.,  S. Henderson,  R.F. Noss, and D.T. Tingey.  In  press.  Biodiversity and human impacts.
        In: W.A. Nierenberg, ed. Encyclopedia of System Science, Academic Press, San Diego, CA.

Barrett, J.W., ed.  1980.  Regional Silviculture of the United States. John Wiley, New York.  551pp.
                                               129

-------
Barrett-Lennard, E.G.,  C.V. Malcolm, W.R. Stern, and S.M. Wilkins, eds.  1986. Forage and Fuel
       Production from Salt-Affected Wasteland.  Elsevier Publishers, New York. 459 pp.

Benson,  R.E.  1982.  Management consequences of alternative harvesting and residue treatment
       practices - lodgepole pine. General Technical  Report INT-132, USDA Forest Service.

Berry, J.  and O. Bjorkman.  1980. Photosynthetic response and adaptation to temperature in higher
       plants.  Ann. Rev. Plant Physiol. 31:491-543.

Binkley, D., K.  Cromack  Jr., and  R. Friedriksen.   1982. Nitrogen accretion and availability in some
       snowbrush ecosystems. For. Sci.  28:720-724.

Bo hm, W.  1979.  Methods of Studying Root Systems. Springer-Vertag, Berlin.  188 pp.

Bolin, B.  1983. The carbon cycle, pp. 41-45.  In: The Major Biogeochemical Cycles and Their
       Interactions.  SCOPE Report No. 21. John Wiley, Chichester.

Bolin, B.,  E.T. Degens,  S. Kempe, and P. Ketner.  1979.  The Global Carbon Cycle. John Wiley &
       Sons, New York.

Bonan, G.B.  1988. Environmental Processes and Vegetation Patterns in Boreal Forests. Ph.D.
       Dissertation. University of Virginia,  Charlottesville, VA.

Bonan, G.B.  1989. A computer model of the solar radiation, soil moisture,  and soil thermal regimes in
       boreal forests.  Ecol. Modelling 45:275-306.

Bonan, G.B. and  H.H. Shugart.  1989.  Environmental factors and ecological  processes in boreal
       forests.  Ann. Rev. Ecol. Sys.  10:1-28.

Bonan, G.B., H.H. Shugart, and D.L. Urban. 1990. The sensitivity of some high-latitude boreal forests
       to climatic parameters. Climatic Change 16:9-30.

Bormann,  F.H., and G.E. Likens.  1979.  Pattern and Process in a Forested Ecosystem. Springer-
       Verlag, New York.  256 pp.

Botkin, D.B., and L.G.  Simpson.   1990.  Biomass of the North American boreal forest: A step toward
       accurate global measures. Biogeochemistry 9:161-174.

Botkin, D.B., J.F. Janak,  and J.R. Wallis.  1972.  Some ecological consequences of a computer model
       of forest growth. J. Ecol. 60:849-873.

Box, E.G.   1981.  Macroclimate and Plant Forms: An Introduction to Predictive Modeling in
       Phytogeography.  Dr. W. Junk Publishers, The Hague.

Brown, K., and  K.O. Higginbotham.   1986. Effects of carbon dioxide enrichment and nitrogen supply
       on growth of boreal tree seedlings. Tree Physiol.  2:223-232.

Brown, S., A.E. Lugo, and J.  Chapman.   1986. Biomass of tropical tree plantations and its implications
       for the global carbon budget. Can. J. For. Res. 16:390-394.
                                              130

-------
Burke, I.C., C.M. Yonker,  WJ.  Parton, C.V. Cole,  K. Flach,  and D.S. Schimel.  1989.  Texture,
       climate, and cultivation effects on soil organic matter content in U.S. grassland soils.  Soil Sci.
       Soc. Amer. J.  53:800-805.

Burke, MJ., L.V. Gusta, HA. Quamme, CJ. Weiser, and P.H. LI 1976. Freezing injury in plants.
       Ann. Rev. Plant Physiol. 27:507-528.

Burkhart, H.E., K.D. Farrar, ILL. Amateis,  and R.F. Daniels.  1987. Simulation of individual tree
       growth and stand development in loblolly pine plantations on cutover, site-prepared areas.
       Publication FWS-1-87, School of Forest and Wildlife Resources, Virginia Polytechnic Institute and
       State University, Blacksburg, VA. 25 pp.

Burley, J., and J. Stewart. 1985.  Increasing Productivity of Multipurpose Trees.  International Union of
       Forest Research Organizations, Vienna, Austria.  560 pp.

Campagna,  MA., and HA. Margolis.  1989.  Influence of short-term atmospheric CO2 enrichment on
       growth, allocation  patterns, and biochemistry of black spruce seedlings at different stages of
       development.  Can. J. For.  Res.  19:773-782.

Canham,  A.E., and WJ. McCavish.   1981. Some effects of C02, daylength and nutrition on the growth
       of young forest tree plants: In the seedling stage.  Forestry  54(2):169-182.

Cess, R.D., G.L. Potter, J.P. Blanchet, GJ. Boer,  SJ.  Ghan,  J.T. Kiehl, H. Le Treut, Z.-X Li, X.-Z
       Liang, J.F.B.  Mitchell, J.-J Morcrette, DA. Randall, M.R. Riches,  E. Roeckner,  U. Schlese,
       A. Slingo, K.E. Taylor, W.M. Washington,  R.T. Wetherald, and I.Yagai.   1989.  Interpretation
       of cloud-climate feedback as produced by 14 atmospheric general circulation models. Science
       245:513-516.

Charles,  D.  1989.  EPA's  plan for cooling the global greenhouse.  Science 243:1544-1545.

Cofer, W.R., J.S. Levine,  D.I. Sebacher, E.L. Winstead, PJ. Riggan,  BJ. Stocks,  JA. Brass, V.G.
       Ambrosia, and PJ. Boston.   1989. Trace Gas Emissions from Chaparral and Boreal  Forest
       Fires.  J. Geophys. Res.  94:2255-2259.

Comerford, N.B., R.F. Fisher,  and W.L. Pritchett.  1982.  Advances in forest fertilization on the
       southeastern coastal plain, pp. 370-378. In: R. Ballard and S.P. Gessel, eds.  IUFRO Symposium
       on Forest Site and Continuous Productivity. General Technical Report PNW-163, USDA Forest
       Service, Pacific Northwest Forest and  Range Experiment Station, Portland, OR.

Conroy,  J.P.,  M. Ri ppers, B. Kippers, J. Virgona, and E.W.R. Barlow.  1988.  The influence of C02
       enrichment, phosphorus deficiency and water stress on the growth, conductance and  water use
       of Pinus radiata D. Don.  Plant Cell Environ. 1:91-98.

Cook,  R.E.  1969. Variation in species density of North American birds. Syst. Zool.  18:63-84.

Cooper,  DJ.  1986.  Trees above and beyond tree line in the Arrigetch Peaks region, Brooks Range,
       Alaska.  Arctic 39(3):247-252.

Cooperative  Holocene Mapping Project (COHMAP).  1988. Climatic changes of the last 18,000 years:
       Observations and  model simulations.  Science  241:1043-1052.
                                               131

-------
Cowan,  I.R.  1965.  Transport of water in the soil-plant-atmosphere system. J. Appl. Ecol. 2:221-239.

Cramer, W., and R. Leemans.  In press. Assessing impacts of climate change on vegetation using
       global vegetation models.  In: Shugart, H.H. and A.M. Solomon, eds.  Vegetation Dynamics
       Modelling and Global Change.  Chapman-Hall, New York.

Cropper, W.P. Jr.  1987. Labile carbon dynamics in a Florida slash pine plantation, pp. 278-284.  In:
       A.R. Ek, S.R. Shrfley, and I.E. Burk, eds.  Forest Growth  Modelling and Prediction Proceedings,
       August, 1987. General Technical Report NC-120. USDA Forest Service, North Central Forest
       Experiment Station.

Cropper, W.P. Jr.,  and K.C. Ewel.  1983. Computer simulation of long-term carbon storage patterns in
       Florida slash pine plantations. For.  Ecol.  Manage.  6:101-114.

Cropper, W.P. Jr.,  and K.C. Ewel.  1984. Carbon storage patterns  in Douglas-fir ecosystems. Can. J.
       For. Res.  14:855-859.

Cropper, W.P. Jr.,  and K.C. Ewel.  1987. A regional carbon storage simulation for large-scale biomass
       plantations.  Ecol. Model.  36:171-180.

Crosson, P.R., and  NJ. Rosenburg.  1989. Strategies for agriculture. Sci. Amer.  261:128-135.

Crutzen,  PJ.  1988.  Variability in atmospheric-chemical systems. In: T. Rosswall, R.G. Woodmansee
       and P.G. Risser, eds.  Scales and Global  Change.  John Wiley & Sons Ltd., New York, NY.

Currie, DJ., and V. Paquin.  1987.  Large-scale  biogeographical patterns of species richness of trees.
       Nature 329:326-329.

Curtis, P.S., B.C. Drake, P.W. Leadley,  WJ. Arp, and D.F. Whigham.  1989a.  Growth and
       senescence in plant communities exposed to elevated C02 concentrations on an estuarine
       marsh.  Oecologia 78:20-26.

Curtis, P.S., B.C. Drake, and  D.F. Whigham.  1989b. Nitrogen and carbon dynamics in C3and C^
       estuarine marsh plants grown under elevated  CO2 in situ.  Oecologia 78:297-301.

Curtis, R.O., G.W.  Clendenen,  D.L. Reukema, and  DJ. DeMars.   1982. Yield tables for managed
       stands of coast Douglas-Fir.  General Technical Report PNW-135, USDA Forest Service, Pacific
       Northwest Forest and  Range Experiment  Station, Portland, OR.

Dangerfield, J., and  H. Brix.  1981.  Comparative effects of ammonium nitrate and urea fertilizers on
       tree growth and soil processes, pp.  133-139.  In: S.P. Gessel, R.M. Kenady, and W.A.J. Atkinson,
       eds.  Proceedings, Forest Fertilization Conference.  University of Washington.  Canadian
       Forestry Service, Victoria BC.

Davis, K.P.  1966.  Forest  Management: Regulation and Valuation.  McGraw-Hill, New York. 519pp.

Davis, M.B.  1981. Quaternary history and  the stability of forest communities. In: D.C. West,  H.H.
       Shugart, and D.B. Botkin, eds.  Forest Succession: Concepts and Application. Springer-Verlag,
       NY. 517pp.
                                              132

-------
Davis, M.B., K.D. Woods, S.L. Webb, and  R.P. Futyma .  1986.  Dispersal versus climate: expansion of
       Fagus and Tsuga into the Upper Great Lakes region.  Vegetatio 67:93-103.

DeLucia,  E.H., and SA. Heckathorn.   1989.  The effect of soil drought on water-use efficiency in a
       contrasting Great Basin desert and Sierran montane species.  Plant Cell Environ.  12:935-940.

Derwiler,  HP., and CAS, Hall. 1988. Tropical forests and the global carbon cycle.  Science 239:42-
       46.

Dickinson,  HE.  1986.  How will climate change: The climate system and modelling of future climate,
       pp. 221-231.  In: B. Bolin, B.R. Dbo s, J. Ja ger, and R.A. Warrick, eds. Scope 29: The
       Greenhouse Effect, Climatic Change and Ecosystems.  John Wiley, New York.

Dickinson,  R.E.  1989.  Uncertainties of estimates of climatic change: A review.  Climatic Change  15:5-
13.

Dixon,  R.K. In press,a.  Feedbacks and response of boreal forests to global change.  In: A. Shivendko,
       ed.  Proceedings of Boreal Forest Symposium. Arkhanglesk, USSR.

Dixon,  R.K. In press.b.  Regional forest management planning in the southern U.S. In: Proceedings of
       North American Conference on Forestry Response to Climate Change, Climate Institute,
       Washington, DC.

Dixon,  R.K. In press,c.  Response and feedbacks of global forests to climate change.  In: Management
       and Productivity of Western-Montane Forest Soils, General Technical Report, USDA Forest
       Service, IRS.  In press.

Dixon,  R.K., R.S.  Meldahl, GA. Ruark, and W.G. Warren.  1990.  Process Modeling of Forest Growth
       Responses to Environmental Stress. Timber Press, Portland, OR  441 p.

Dixon,  R.K., G.M. Wright, G.T. Behrns, R.D. Teskey,  and T.M. Hinckley.  1980. Water deficits and
       root growth of ectomycorrhizal white oak seedlings.  Can. J. For. Res. 10:545-548.

Drake, E.G., P.W. Leadley,  W. Arp, D. Nassiry, and P.S. Curtis. Submitted. Testing the  effects of
       elevated COaon natural vegetation. I. Control of C02 and microclimate inside an open top
       chamber.  Functional Ecology.

Drake, B.C., P.S.  Curtis, WJ. Arp, P.W. Leadly, J. Johnson,  and D. Whigham.  1987.  Effects of
       elevated carbon dioxide on Chesapeake Bay wetlands. III. Ecosystem and whole plant
       responses in the first year of exposure.  U.S. Department of Energy, Carbon Dioxide Research
       Division Report 044, Office of Energy Research, Washington, DC, 20545.

Dregne, H.E. 1983.  Desertification of Arid Lands.  Harwood Academic Publishers,  New York, NY.  242
       PP-

Duxbury, J.M., M.S.  Smith, J.W. Doran, C. Jordan, L. Scott, and E. Vance.  1989.  Soil organic
       matter as a source and a sink of plant nutrients.  In: D.C. Coleman, J.M. Oades and G. Uehara,
       eds. Dynamics of Soil Organic Matter in Tropical  Ecosystems.  Niftal Project, University of
       Hawaii Press, Honolulu, HI.

Dyson, FJ. 1977. Can we control the carbon dioxide in the atmosphere? Energy 2:287-291.
                                              133

-------
Eamus, D., and P. Jarvis.  1989.  The direct effects of increase in the global atmospheric C02
        concentration on natural and commercial temperate trees and forests.  Adv. Ecol. Res.  19:2-56.

Edmonds, R.L.  1987.  Decomposition rates and nutrient dynamics in small-diameter woody litter in four
        forest ecosystems in Washington, U.S.A. Can. J. For. Res.  17:499-509.

Edwards,  M.E. and  P.W. Dunwiddie.   1985.  Dendrochronological and palynological observations on
        Populus balsamifera in northern Alaska, U.S.A.  Arctic and Alpine Research 17:271-278.

Edwards,  M.E., P.M. Anderson, H.L. Garflnkel, and L.B. Brubaker.  1985. Late Wisconsin and
        Holocene vegetational history of the Upper Koyukuk region, Brooks Range, Alaska.  Can. J. Bot.
        63:616-626.

Emanuel,  W.R.  1985.  Response to comment: Climatic change and the broad-scale distribution of
        terrestrial ecosystem complexes.  Climatic Change 7:457-460.

Emanuel,  W.R.,  G.G.  Killough, W.M. Post, and H.H. Shugart.  1984.  Modeling terrestrial ecosystems
        in the global carbon cycle with shifts in carbon storage capacity by land-use change. Ecology
        65:970-983.

Emanuel,  W.R.,  H.H. Shugart,  and M.P.Stevenson.   1985.  Climatic change and the broad-scale
        distribution of terrestrial ecosystem complexes.  Climate Change  7:29-43.

Evans, J. 1982.  Plantation Forestry in the Tropics. Clarendon Press, Oxford.  472  pp.

Fahnestock,  G.R., and J.K. Agee.  1983.  Biomass consumption and smoke production by prehistoric
        and modern forest fires in western Washington.  J. For.  653-657.

Fajer, E.D., M.D. Bowers,  and FA. Bazzaz.  1989. The effects of enriched carbon dioxide
        atmospheres on plant-insect herbivore interactions.  Science 243:1198-1200.

Farnum, P.,  R. Timmis, and J.L. Kulp.  1983.  Biotechnology of forest yield. Science  219:694-702.

Federer,  C.A.  1982.  Transpirations! supply and demand: plant, soil, and atmospheric effects evaluated
        by simulation. Water Resour.  Res. 18(2):355-362.

Food  and Agriculture Organization (FAO). 1985. Topical forestry action plan.  Food and Agriculture
        Organization of the United Nations, Rome.  275 pp.

Foreward, P.W., R.S. Moulton, and J.D. Snellgrove.   1990.   FY89 U.S. Forest Service Planting Report.
        USDA Forest Service, Coop. Forestry, Washington, DC.  15 pp.

Fulkerson, W.,  D.B. Reister, A.M. Perry,  A.T. Crane, D.E. Kash, and S.I.  Auerbach.   1989. Global
        warming: An energy technology R&D challenge.  Science 246:868-869.

Funsch,  R.W., R.H. Mattson, and G.R. Mowry.  1970.  C02-supplemented atmosphere increases
        growth of Pinus strobus seedlings. For. Sci.  16:459-460.

Garfinkel, H.L. and L.B. Brubaker. 1980. Modern climate-tree-growth relationships and climatic
               reconstruction in sub-Arctic Alaska.  Nature  286:872-874.
                                              134

-------
Gates,  W.L. 1985. Modeling as a means of studying the climate system.  In: M.C. MacCracken and
       P.M. Luther, eds.  Projecting the Climatic Effects of Increasing Carbon Dioxide.  DOE/ER-0237,
       US Department of Energy, Washington, DC.

Giddens,  J. 1957. Rate loss of carbon from Georgia soils. Soil Sci. Soc. Amer. Proc.  21:513-515.
Gjerstad, D.H., L.R. Nelson,  J.H. Dukes, and WjL Retzlaff.  1984.  Vegetation management, pp. 247-
       258. In: M.L Duryea, and G.N. Brown, eds. Seedling Physiology and Reforestation Success.
       Proceedings of the Physiology Working Group Technical Session.  Soc. Amer. Foresters.
       Nijhoff/Junk Publishers, Boston. MA.

Goldstein, G.H. 1981. Ecophysiological and demographic studies of w^  : spruce (Picea glauca
       (Moench) Voss) at treeline in the central Brooks Range of Alaska   ; h.D. Thesis, University of
       Washington.  193 pp.

Goldstein, G.H., Brubaker,  L.B. and T.M. Hinckley.   1985. Water relations of white spruce (P/'cea
       glauca  (Moench) Voss) at tree line in north central Alaska. Can. J. For. Res.  15:1080-1087.

Grainger, A.  1988.  Estimating areas of degraded tropical lands requiring replenishment of forest cover.
       Int. Tree Crops J. 5:31-61.

Green, K., and  R. Wright. 1977. Field  response of photosynthesis to COa enhancement in ponderosa
       pine.  Ecology 58:687-692.

Greenland,  DJ., and  P.M. Nye. 1959.  Increases in the carbon and nitrogen contents of tropical soils
       under natural fallows. J. Soil Sci.  10:284-299.

Gregersen,  H., S. Draper, and D.  Elz.  1989.  People and trees:  The role of social forestry in
       sustainable development. EDI Seminar Series.  World Bank,  Washington, DC.  273 pp.

Grey, G.W., and FJ.  Deneke.  1978. Urban Forestry.  John Wiley & Sons, New York, NY. 175pp.

Crotch, S.L., ed.  1988.  Regional Intercomparisons of General Circulation Model Predictions and
       Historical Climate Data.  U.S. Dept. of Energy Report TR041 , Washington, DC. 291 pp.

Hansen, J., A.  Lacis,  D. Rind, G.  Russel, P. Stone,  I.Fung,  and J. Lerner.  1984. Climate sensitivity:
       Analysis of feedback mechanism.  In: Climate Processes and Climate Sensitivity,  Geophys.
       Monogr. Ser.,  AGU, Washington, DC. 29:130-163.

Hansen, J., I.Fung, A. Lacis,  D. Rind,  S. LebedefT,  R. Ruedy,  and G. Russell.   1988.  Global climate
       changes as forecast  by Goddard Institute for Space Studies three-dimensional model. J.
       Geophys. Res. 93:9341-9364.

Hansen-Bristow, KJ. 19*86.  Influence  of increasing elevation on growth characteristics at timberline.
       Can. J. Bot. 64:2517-2523.

Hansen-Bristow, KJ., and J.D. Ives. 1984.  Changes in the forest-alpine tundra ecotone: Colorado
       Front Range.  Phys. Geog.  5:186-197. '
                                              135

-------
Hansen-Bristow,  KJ. and J.D. Ives.  1985.  Composition, form and distribution of the forest-alpine
       tundra ecotone, Indian Peaks, Colorado, U.S.A.  Erdkunde 39:286-295.

Hare, F.K.  1950. Climate and zonal divisions of the boreal forest formation in eastern Canada.  Geog.
       Rev.  40:615-635.

Harmon, M.E.,  W.K. Ferrell, and J.F. Franklin.  1990. Effects on carbon storage of conversion of
       old-growth forests to young forests.  Science  247:699-702.

Harmon, M.E.,  J.F. Franklin, FJ. Swanson,  P. Sollins, S.V. Gregory, D. Lattin, N.H. Anderson,  S.P.
       Ciine, N.G. Aumen, J.R. Sedell, G.W. Uenkaemper,  K. Cromack  Jr., and R.W. Cummins.
       1986. Ecology of coarse woody debris in temperate ecosystems. Adv. Ecol. Res. 15:133-302.

Harvey, L.D.D.  1988.  Climatic impact of ice-age aerosols.  Nature 334:333-336.

Harvey, L.D.D.  1989.  Transient climatic response to an increase of greenhouse gases.  Climatic
       Change 15:15-30.

Heimlich, R.E.,  M.B. Carey,  and RJ. Brazee.  1989.  Beyond swampbuster: A permanent wetland
       preserve.  J. Soil Water Conserv. 9:445-450.

Henderson,  S.,  R.K. Olson, and  R.F. Noss.  1989. Current and potential threats to biodiversity in
       forests of the lower Pacific Coast states, pp. 325-336.  In: R.K. Olson and A.S. Lefohn, eds.,
       Effects of Air Pollution on Western Forests, Air and Waste Management Association, Anaheim,
       CA.

Hermann, R.K.  1977.  Growth and productivity of tree roots: A review, pp. 7-28. In: J.K. Marshall, ed.
       The Below-Ground Ecosystem: A Synthesis of Plant-Associated Processes.  Colorado State
       University, Fort Collins, CO.

Higginbotham,  K.O., J.M. Mayo, S. L'Hirondelle,  and O.K. Krystofiak.  1985. Physiological ecology
       of lodgepole pine (Pinus contorta) in an enriched C02 environment. Can. J. For. Res. 15:417-
       421.

Holdridge, L.R.  1947.  Determination of world formulations from simple climatic data. Science
       105:367-368.

Holdridge, L.R.  1967.  Life Zone Ecology.  Tropical Science Center, San Jose, CA.

Hollinger, D.Y.  1987.  Gas exchange and dry matter allocation responses to elevation of atmospheric
       C02 concentration in seedlings of three tree species. Tree Physiol. 3:193-202.

Houghton, RA. 1987. Terrestrial metabolism and atmospheric C02 concentrations.  BioScience
       37:672-678.

Houghton, RA. 1990a. The global effects of tropical deforestation.  Environ. Sci. Technol. 24(4) :414-
       422.

Houghton, RA. 1990b. The future role of tropical forests in affecting the carbon dioxide concentration
       of the atmosphere. Ambio. 19:204-209.

Houghlon, RA., and G.M. Woodwell.   1989.  Global climatic change. Sci. Amer.  260:36-44.


                                              136

-------
Houghton,  RA., J.E. Hobble, J.M. Melillo, B. Moore, BJ. Peterson,  G.R. Shaver,  and G.M.
       Woodwell.  1983. Changes in the carbon content of terrestrial biota and soils between 1860
       and 1980: A net release of C02to the atmosphere. Ecol. Monogr.  53:235-262.

Houghton,  RA., R.D. Boone,  J.M. Melillo, CA. Palm, GJV1. Woodwell,  N. Myers, B. Moore HI,and
       O.K. Skole.  1985. Net flux of carbon dioxide from tropical forests in 1980. Nature 316:617-620

Houghton,  RA., R.D. Boone,  J.R. Fruci, J.E. Hobble,  J.M. Melillo, CA. Palm, BJ. Peterson,  G.R.
       Shaver, and G.M. Woodwell.  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  398:122-139.

Houpis, J.LJ.,  KA. Surano,  and J. Shinn.  1987. Seasonal and diurnal water use patterns of Pinus
       ponderosa seedlings exposed for two years with elevated carbon dioxide.  Plant Physio!, (suppl)
       83:45.

Houpis, J.LJ.,  KA. Surano,  S.  Cowles,  and J.H. Shinn.  1988. Chlorophyll and carotenoid
       concentrations in two varieties of Pinus ponderosa seedlings subjected to  long-term elevated
       carbon dioxide.  Tree Physio!. 4:187-193.

Husch, B., C.I. Miller, and T.W. Beers.   1972.  Forest Mensuration.  Ronald Press, New York, 410 pp.

ICF, Inc.  1989.  Summary Report: Scenarios Advisory Meeting Report, 31 Aug-1 Sept, Boulder, CO.
       Report for U.S. Environmental Protection Agency, Office of Policy, Planning and Evaluation,
       Washington, DC,  14 pp.

Ino, Y., and M. Monsi.   1969. An experimental approach to the calculation of CO2 amount evolved
       from several soils. Jap. J. Bot.  2:153-188.

Intergovernmental  Panel on Climate Change (IPCC). 1990. Scientific Assessment of Climate Change.
       Report for WGI Plenary Meeting.

Ives, J.D.,  and  KJ. Hansen-Bristow.   1983.  Stability and instability of natural and modified upper
       timberline landscapes in the Colorado Rocky  Mountains, U.S.A.  Mountain Research Division
       3:149-155.

Jain, R.K., K. Paliwal, R.K. Dixon, and  D.H. Gjerstad.  1989. Improving productivity of multipurpose
       trees growing on substandard soils in India.  J.  For. 87:38-42.

Jarvis, P.G. 1987. Water and carbon fluxes  in ecosystems, pp. 50-67.  In:  E.D. Shulze and H. Zwolfer,
       eds. Potentials and Limitations of Ecosystem  Analysis, Springer-Veriag, Berlin.

Jarvis, P.G., and K.G. McNaughton.  1986.  Stomatal control of transpiration:  scaling up from leaf to
       region.  Adv. Ecol. Res.  15:1-49.

Jenkinson, D.S., and J.H. Rayner.  1977. The turnover of soil organic matter in some of the
       Rothamsted classical experiments. Soil Sci.  123:298-305.

Johnson,  EA.  Personal communication.  Department of Biology, University of Calgary, Calgary,
       Alberta.
                                              137

-------
Jordan, C.F., ed. 1989. Man and the Biosphere Series.  Volume 2: An Amazonian Rain Forest: The
       Structure and Function of a Nutrient Stressed Ecosystem and the Impact of Slash-and-Burn
       Agriculture.  UNESCO and the Parthenon Publishing Group.

Keane, R.E., S.F. Arno, and J.K. Brown.  1989. FIRESUM: An ecological  process model for fire
       succession in western conifer forests.  General Technical Report, INT-266, USDA Forest Service,
       Intermountain Research Station.

Keeling, C.D.,  R.B. Bacastow,  A.F. Carter, S.C. Piper, T.P. Whorf, M. Heimann,  W.G. Mook, and H.
       RoelofTzen.  1989.  A three-dimensional model of atmospheric COatransport based on observed
       winds:  Analysis of observational data.  Geophys. Monogr.  55:165-236.

Keyfitz, N. 1989. The growing human population.  Sci. Amer. 261:119-126.

Kiester, A.R.  1971. Species density of North American amphibians and reptiles.  Syst. Zool.  20:49-67.

King, A.W, D.L. DeAngelis,  and W.M.  Post.  1987.  The seasonal exchange of carbon dioxide between
       the atmosphere and the terrestrial biosphere: Extrapolation from site-specific models to regional
       models.  Publication 2988, Environmental Sciences Division, Oak Ridge National Laboratory, Oak
       Ridge,  TN.

Kira, T. 1965.  Primary production in forests, pp. 5-40. In: J.P. Cooper, ed. Photosynthesis and
       Productivity in Different Environments.  Cambridge University Press, England.

K> ppen, W. 1900. Versuch einer Klass'rfication der Klimate, vorzugsweise  nach ihren Beziehungen zur
       pflanzenwelt. Geographische Zeitschrift. 6:593-611, 657-679.

Koppen, W. 1918. Wassification der Klimate nach Temperatur, Neidenschlag  und Jahreslauf.
       Petermanns Geogr.  Mitt. 64:193-203, 243-248.

K> ppen, W. 1936. Das Geographische System der Klimate. In: W. KJ ppen, and R. Geiger, Handbuch
       der Klimatologie, 1C, Gebr Borntraeger, Berlin, 46 pp.

Kramer, PJ.  1983. Water Relations of Plants. Academic Press, New York, NY. 489 pp.

Kramer, PJ.,  and N. Sionit. 1987. Effects of increasing carbon dioxide concentration on the
       physiology and growth of forest trees, pp. 219-246. In: W.E. Shands and J.S. Hoffman, eds.
       The Greenhouse Effect, Climate Change, and U.S. Forests. The Conservation Foundation,
       Thomson-Shore, Inc., Dexter, Ml.

LaMarche,  V.C.,  Jr., and HA. Mooney.  1967.  Altithermal timbertine advance in the western United
       States. Nature  13:980-982.

Larsen, JA. 1989. The Northern Forest Border in Canada and Alaska.  Springer-Verlag, New York.
       255 pp.

Lashof, D.A.  1987. The role of the biosphere in the global carbon cycle: evaluation through biospheric
       modeling and atmospheric measurement.  Ph.D. Dissertation, Energy and  Resources Group,
       University of California, Berkeley, CA.
                                              138

-------
Lashof,  DA.  1989. The dynamic greenhouse: Feedback processes that may influence future
       concentrations of atmospheric trace gases and climatic change. Climatic Change 14:213-242.

Lashof,  DA., and DA. Tirpak.  1989.  Policy Options for Stabilizing Global Climate. Draft Report to
       Congress, U.S.  Environmental Protection Agency, Office of Policy, Planning and Evaluation,
       Washington, DC.

Leemans,  R. 1990. Possible changes in natural vegetation patterns due to a global warming.
       Publication No. 108, International Institute for Applied Systems Analysis A-2361, Laxenburg,
       Austria. 13 pp.

Leemans,  R., and I.C. Prentice.  1989. FORSKA: A general forest succession model. Institute of
       Ecological Botany, Uppsala, Sweden.

Lev, DJ.  1987.  Balsam poplar (Populus balsamifera) in Alaska: Ecology and growth response to
       climate. Master's Thesis, University of Washington, Seattle, WA. 69 pp.

Leverenz,  J.W., and DJ. Lev.  1987.  Effects of carbon dioxide-induced climate changes on the natural
       ranges of six major commercial tree species in the western United States, pp. 123-155.  In: W.E.
       Shands and J.S. Hoffman, eds. The Greenhouse Effect, Climate Change, and U.S. Forests. The
       Conservation Foundation, Thomson-Shore, Inc., Dexter, Ml.

Lonqjun, G.  circa 1986.  Green China. Office of Central Afforestation Committee.  The Great Wall
       Publishing House.  31 pp.

Lugo, A.E., and S. Brown.  1986. Steady  state terrestrial ecosystems and the global carbon cycle.
       Vegetatio  68:83-90.

Luxmoore,  RJ., E.G. O'Neill, J.M. Ells, and H.H. Rogers.  1986.  Nutrient uptake and growth
       responses of Virginia pine to elevated atmospheric carbon dioxide.  J. Environ. Qual. 15:244-
       251.

MacDicken,  K.G., and  N.T. Vergara.  1990. Agroforestry: Classification and Management.  John Wiley
       & Sons, New York, NY.  382 pp.

Manabe,  S., and RJ.  Stoufler.  1980. Sensitivity of a global climate to an increase of CO2
       concentration in the atmosphere. J. Geophys. Res.  85:5529-5554.

Manabe,  S., and R.T. Wetherald.  1987. Large-scale changes in soil wetness induced by an increase in
       carbon dioxide. J. Atmos. Sci. 44:1211-1235.

Mann, L.K.  1986.  Changes in soil carbon  storage after cultivation. Soil Sci.  142:279-288.

Manne,  A.S., and R.G. Richels.  In press.a.  CO2 emission  limits: an economic cost analysis for the
       USA. Energy J.

Manne,  A.S., and R.G. Richels.  In press.b.  Global C02 emission reductions:  The impacts of rising
       energy costs.  Energy J.

Marks, D.  Submitted.  The sensitivity of potential evapotranspiration to  climate change over the
       continental United States.


                                              139

-------
Marland, G.  1988.  The prospect for solving the C02 problem through reforestation. DOE/NBB-0082,
        U.S. Department of Energy, Office of Energy Research, Washington, DC.  66 pp.

Marland, G.  1989.  The role of forests in addressing the C02 greenhouse. In: J.C. White, ed.  Global
        Climate Change Linkages: Acid Rain, Air Quality and Stratospheric Ozone.  Elsevier Publishing,
        New York.

Martin, J.H., and S.E. Fitzwater.  1988.  Iron deficiency limits phytoplankton growth in the northeast
        Pacific subarctic. Nature.  331:341-343.

Materna, J.  1989. Air pollution and forestry in Czechoslovakia. Environ Monitor. Assess.  12:227-235.

Matthews, E.  1983.  Global vegetation and land use: New high-resolution data base for climate studies.
        J. Climate Appl. Meteorol. 22:474-487.

Mattson, K.G., and  W.T. Swank.  1989.  Soil and detrital carbon dynamics following forest cutting in the
        southern Appalachians.  Biol. Pert. Soils 7:247-253.

Mattson, W.S., and A. Addy.  1975.  Phytophagous insects as regulators of forest primary productivity.
        Science  190:515-521.

Meehl, G A., and Washington, W.M.  1990.  C02 climate sensitivity and snow-sea-ice albedo
        parameterization in an atmospheric GCM coupled to a mixed-layer ocean model. Climatic
        Change  16:283-306.

Melillo, J.M., J.D. Aber, and J.F. Muratore. 1982.  Nitrogen and lignin control Of hardwood leaf litter
        decomposition dynamics. Ecology  63:621-626.

Miller, J.H., and  D.L. Sirois.  1986.  Soil disturbance by skyline yarding vs. skidding in a loamy hill
        forest. Soil Sci. Soc. Amer. J. 50:1579-1583.

Miller, R.E.,  and S.R. Webster.   1979. Fertilizer response in mature stands of Douglas-fir, pp.  126-132.
        In: S.P. Gessel, R.M. Kenady, and W.A. Atkinson, eds. Proceedings: Forest  Fertilization
        Conference.  College of Forest Resources,  University of Washington, Seattle, WA.

Miller, R.E.,  P.R.  Barker,  C.E. Peterson,  and S.R. Webster.   1986. Using nitrogen fertilizers in
        management of coast Douglas-fir: Regional trends of response, pp. 290-303. In: C.O. Oliver,
        D.P. Hanley, and J.A. Johnson, eds.  Douglas-fir: Stand Management for the Future.  College of
        Forest Resources,  University of Washington, Seattle, WA.

Mitchell, J.F.B.,  and DA.  Warrilow.  1987.  Summer dryness in northern  mid-latitude due to increased
        C02.  Nature  330:238-240.

Mitchell, J.F.B.,  CA.  Senior and WJ. Ingram.  1989.  C02 and climate: A missing feedback?  Nature
        341:132-134.

Mooney, HA., P.M. Vitousek, and PA.  Matson.  1987.  Exchange of materials between terrestrial
        ecosystems and the atmosphere.  Science 238:926-932.
                                              140

-------
Mooney,  HA., B.C. Drake, RJ. Luxmoore,  W.C. Oechel and L. Pitelka.  In press. How will terrestrial
       ecosystems interact with the changing C02 concentration of the atmosphere and anticipated
       climate change? BioScience.

Morison, J.I.L. 1987. Intercellular CO2 concentration and stomatal responses to CO^ pp. 229-251. In:
       E. Zeiger, G.D. Farquhar, and I.R. Cowan, eds.  Stomatal Function.  Stanford University Press,
       Stanford, CA.

Munslow,  B., Y. Katerere,  A. Ferf, and P. O'Keefe . 1988. The Fuelwood Trap: A Study of the SADCC
       Region.  Earthscan Publ. Ltd., London.  181 pp.

Myers, N.  1988.  Tropical forests and their species: Going, going ....? In: E.O. Wilson, ed. Biodiversity,
       National Academy Press, Washington, DC.

National Academy  of Sciences.   1980. Firewood Crops. National Academy Press, Washington, DC.
       236 pp.

Neilson, R.P.,  GA.  King, R.L. DeVelice,  J.  Lenihan, D. Marks, J.  Dolph, B. Campbell, and G. Click.
       1989.  Sensitivity of ecological landscapes and regions to global climate change. U.S.
       Environmental Protection Agency, Environmental Research Laboratory, Corvallis, OR.

Neilson, R.P.,  GA.  King, J. Lenihan, and  R.L. DeVelice.  1990. The annual course of precipitation
       over much of the United States: Observed versus GCM simulation, pp.  19-26. In: J.L
       Betancourt and A.M. MacKay, eds.  Proceedings of the Sixth Annual Pacific Climate (PACLIM)
       Workshop, CL/PACLIM-6ATR/90-23, California Department of Water Resources, Sacramento,
       CA.

Newbould,  P.S.  1967.  Methods for estimating  the primary production of forests.  International
       Biological Program Handbook No. 2. Blackwell Scientific, Oxford.

Nohrstedt,  H.-O ., K. Arnebrant, E. Baa th, and  B. So derstro m.  1989. Changes in carbon content,
       respiration rate, ATP content, and microbial biomass in nitrogen-fertilized  pine forest soils in
       Sweden.  Can. J. For. Res.  19:323-328.

Norby, RJ. Personal communication. Oak Ridge National Laboratory, Environmental Sciences
       Division, Oak Ridge, TN.

Norby, RJ. 1987.  Nodulation and nitrogenase activity in nitrogen-fixing woody plants stimulated by
       C02 enrichment of the atmosphere.  Physiol. Plant 71:77-82.

Norby, RJ., and E.G. O'Neill.  1989.  Growth dynamics and water use of seedlings of Quercus alba L
       in C02-enriched atmospheres.  New Phytol.  111:491 -500.

Norby, RJ., and E.G. O'Neill.  Submitted.   Leaf area compensation and nutrient Interactions in C02-
       enriched yellow-poplar (Liriodendron tulipifera  L.) seedlings.

Norby, RJ., E.G. O'Neill, and RJ. Luxmoore.   1986a.  Effects  of atmospheric C02 enrichment on the
       growth and mineral nutrition of Quercus alba seedlings in nutrient-poor soil.  Plant Physiol.
       82:83-89.
                                              141

-------
Norby, RJ., J. Pastor, and J.M. Melillo.  1986b.  Carbon-nitrogen interactions in CO^enriched white
       oak: Physiological and long-term perspectives. Tree Physiol.  2:233-241.

Norby, RJ., E.G. O'Neill, W.G. Hood, and RJ. Luxmoore.  1987. Carbon allocation, root exudation
       and mycorrhizal colonization of Pinus echinata seedlings grown under CO2 enrichment. Tree
       Physiol.  3:203-210.

Oades, J.M.  1988. The retention of organic matter in soils. Biogeochemistry  5:35-70.

Oberbauer, S.F.,  B.R. Strain, and N. Fetcher.  1985. Effect of CO2-enrichment on seedling physiology
       and growth of two tropical tree species.  Physiol. Plant.  65:352-356.

Odasz, A.M.  1983. Vegetational patterns at the tree limit ecotone in the upper Alatna River Drainage of
       the Central Brooks Range, Alaska.  Ph.D. Thesis, University of Colorado, Boulder.  224 pp.

Odum, P.  1983. Basic Ecology.  Saunders College Publishing, Philadelphia, PA. 613 pp.

Oechel,  W., and B.R.  Strain.   1985. Native species responses to increased carbon dioxide
       concentration.  In: B.R. Strain and J.D. Cure,  eds.  Direct Effects of Increasing Carbon Dioxide
       on Vegetation.  U.S. Department of Energy, Washington, DC.

Oechel,  YV.C, S.  Hastings, D. Hilbert, W. Lawrence,  T. Prudhomme,  G. Riechers,  and D. Tissue.
       1984.  The response of arctic ecosystems to elevated CO2 regimes.  U.S. Department of Energy,
       Carbon Dioxide Research Division Report 019, Office of Energy Research, Washington, DC.

Office of Technology   Assessment.   1984. Technologies to Sustain Tropical Forest Resources.  U.S.
       Congress, Washington, DC, 344 pp.

Office of Technology   Assessment.   1987. Technologies to Maintain Biological Diversity.  U.S.
       Congress, Washington, DC.  344 pp.

Olson, J.S., J.A. Watts, and LJ. Allison.  1983.  Carbon in live vegetation of major world ecosystems.
       ORNL-5862, Oak Ridge National Laboratory, Oak Ridge, TN. 180 pp.

O'Neill, R.V., D.L. DeAngelis,  J.B. Waide, and T.F.H. Allen. 1986. A Hierarchical Concept of
       Ecosystems.  Monographs in Population Biology #23.  Princeton University Press, Princeton,
       NJ.  253 pp.

O'Neill, E.G., RJ. Luxmoore  and RJ. Norby.  1987a.  Elevated atmospheric CO2 effects on seedling
       growth, nutrient uptake, and rhizosphere bacterial populations of Liriodendron tuliplfera L  Plant
       and Soil 104:3-11.

O'Neill, E.G., RJ. Luxmoore,  and RJ. Norby.   1987b.  Increases in mycorrhizal colonization and
       seedling growth in Pinus echinata and Quercus alba in an enriched C02 atmosphere. Can. J.
       For. Res.  17:878-883.

Overpeck,  J.T., D. Rind, and  R. Goldberg.  1990. Climate-induced changes in forest disturbance and
       vegetation. Nature 343:51-51.

Palm, CA., RA. Houghton, J.M. Melillo, and D.L. Skole.   1986. Atmospheric carbon dioxide from
       deforestation in southeast Asia.  Biotropica 18:177-188.
                                              142

-------
Parry, M., T. Carter, N. Konijin, and J. Lockwood.  1987. The impact of climatic variations on
       agriculture: Introduction to the  IIASA/UNEP Case Studies in Semi-Arid Regions. International
       Institute for Applied Systems Analysis, Laxenburg, Austria.

Parton, WJ.,  J.W.B.  Stewart, and C.V. Cole.  1988. Dynamics of C, N, P and S in grassland soils:  A
       model.  Biogeochemistry 5:109-131.

Pastor, J., and W.M. Post  1988. Response of northern forests to CO?lnduced climate change.
       Nature 334:55-58.

Paul, EA.  1976.  Nitrogen cycling in terrestrial ecosystems, pp. 225-243.  In: 0. Nriager, ed.
       Environmental Biogeochemistry, Vol. 1.  Ann Arbor Science Publishers, Ann Arbor, Ml.

Pearson,  GA. 1931.  Forest types of the southwest as determined by climate and soil. Technical
       Bulletin 247, U.S.  Department of Agriculture, Washington, DC.

Peet, R.K.  1981.  Changes in biomass and production during secondary forest succession, pp.  324-338.
       In: D.C. West, H.H. Shugart,  and D.B. Botkin, eds.  Forest Succession:  Concepts and
       Application. Springer-Vertag, New York.

Perlin, J.  1989.  A Forest Journey, the Role of Wood in the Development of Civilization. Norton, New
       York. 445 pp.

Perry, DA., M.P.  Amaranthus, J.G.  Borchers, S.L. Borchers,  and R.E. Brainerd.   1989.
       Bootstrapping in ecosystems.  Bioscience 39:230-237.

Peters, R.L., and  J.D.S. Darling.  1985.  The greenhouse effect and nature reserves.  BioScience
       35:707-717.

Peterson,  C.E.  Personal communication. NSI Technology Services, Corp., U.S. Environmental
       Protection Agency, Environmental Research Laboratory, Corvallis, OR.

Peterson,  C.E., and S.P.  Gessel.  1982. Forest fertilization in the Pacific Northwest: Results of the
       Regional Forest Nutrition Research Project, pp.  365-369.  In:  R. Ballard and S.P. Gessel,  eds.
       IUFRO Symposium on Forest Site and Continuous Productivity.  General Technical Report
       PNW-163, USDA Forest Service,  Pacific Northwest Forest and Range Experiment Station,
       Portland, OR.

Post,  W.M., W.R. Emanuel,  PJ. Zinke,  and A.G. Stangenberger  .  1982. Soil carbon pools and world
       life zones.  Nature 298:156-159.

Post,  W.M., J. Pastor,  PJ. Zinke, and A.G. Stangenberger.   1985. Global patterns of soil nitrogen
       storage.  Nature  317:613-616.

Postel, S., and L. Heise.   1988.  Reforesting the earth.   WorldWatch Paper 83, WorldWatch Institute,
       Washington, DC. 66 pp.


Prentice,  K.C.  1990.  Bioclimatic distribution of vegetation for General Circulation Model studies.  J.
       Geophys. Res. 95:11811-11630.
                                              143

-------
Prentice, K.C., and I.Y. Fung.  In press. Bioclimatic simulations test the sensitivity of terrestrial carbon
        storage to perturbed climates.  Nature.

Prentice, I.C., R.S. Webb, M.T. Ter-Mikhaelian, AM. Solomon, T.M. Smith, S.E.  Pitovranov,  N.T.
        Nikolov, AA. Minin, R. Leemans, S. Lavorel, M.D. Korzukhin, J.P. Hrabovszky,  H.O.
        Helmisaari, S.P. Harrison, W.R. Emanuel, and G.B. Bonan.  1989. Developing a global
        vegetation dynamics model: Results of an NASA summer workshop.  Publication RR-89-7,
        International Institute for Applied Systems Analysis, A-2361, Laxenburg, Austria.

Pritchett, W.L., and R.F. Fisher.   1987.  Properties and Management of Forest Soils. John Wiley, New
        York.  494 pp.

Reid,  W.V., and K.R. Miller. 1989. Keeping options alive: The scientific basis for conserving
        biodiversity. World Resource Institute, Washington, DC. 128 pp.

Ricklefs, R.E. 1973. Ecology. Chiron Press Inc., Newton, MA. 861 pp.

Rogers,  H.H., G.E. Bingham,  J.D. Cure, J.M. Smith, and FLA. Surano.  1983a. Responses of selected
        plant species to elevated carbon dioxide in the field. J. Environ. QuaL  12(4):569-574.

Rogers,  H.H., J.F. Thomas, and  G.E. Bingham.  1983b.  Response of agronomic and forest species to
        elevated atmospheric carbon dioxide.  Science 220:428-429.

Ruess, J.O.,  and D.VV.  Johnson.   1986.  Acid Deposition and the Acidification of Soils and Waters.
        Springer-Verlag, New York. 119 pp.

Ruess,  R.W., SJ. McNaughton, and  M.B. Coughenour.   1983. The effects of clipping, nitrogen
        source and nitrogen concentration on growth responses and nitrogen uptake of an East African
        sedge.  Oecologia (Berlin) 59:253-261.

Running, S.W., and J.C. Coughlan.   1988.  A general model of forest ecosystem processes for regional
        applications. I. Hydrologic balance, canopy gas exchange and primary production processes.
        Ecol. Model.  42:125-154.

Scarratt, J.B., C. Glerum, and CA. Plexman,  eds.  1982. Proceedings of the Canadian Containerized
        Tree Seedling Symposium, September 1981, Toronto, Canada.  COJFRC Symposium
        Proceedings O-P-10. 460 pp.

Schlesinger,   M.E., ed.  1988. Physically-Based Modelling and  Simulation of Climate and Climatic
        Change (Part 1).  Proceedings of the NATO Advanced Study Institute, Erice, Italy, May 11-23,
        1986.

Schlesinger,   M.E., and J.F.B. Mitchell.  1985.  Model projections of the equilibrium climatic response
        to increased carbon dioxide. In: M.D. MacCracken and P.M. Luther, eds. Projecting the
        Climatic Effects of Increasing Carbon Dioxide. DOE/ER-0237, US Department of Energy,
        Washington, DC.


Schlesinger,   M.E., and Z.C. Zhao. 1989.  Seasonal climatic change introduced by doubled C02as
        simulated by the OSU atmospheric GCM/mixed-layer ocean model. J. Climate 2:429-495.
                                              144

-------
Schlesinger,  W.H. 1977. Carbon balance in terrestrial detritus.  Ann. Rev. Ecol. Syst.  8:51-81.

Schlesinger,  W.H. 1984. Soil organic matter: A source of atmospheric C02 In: G.M. Woodwell, ed.
       The Role of Terrestrial Vegetation in the  Global Carbon Cycle: Measurement by Remote Sensing.
       John Wiley & Sons, New York, NY.

Schlesinger,  W.H., J.F. Reynolds, G.L. Cunningham,  LF. Keunneke,  VfM. Jamil, RA. Virginia, and
       W.G. Whitford. 1990. Biological feedbacks in global desertification. Science 247:1043-1047.

Schneider,  S.H.  I989a. Global Warming: Are We Entering the Greenhouse Century?  Sierra Club, San
       Francisco.  317 pp.

Schneider,  S.H.  1989b. The changing climate. Sci. Amer. 261(3):70-79.

Schroeder,  P.E., and L. Ladd.  In press. Slowing the increase of atmospheric carbon  dioxide: A
       biological approach. Climate Change.

Schulze,  E.-D.  1989.  Air pollution and forest decline in a spruce (P/cea ab/es) forest.  Science
       244:776-783.

Sedjo,  R. 1989.  Forests to offset the greenhouse effect.  J. For.  87(7):12-15.

Sedjo,  R.A., and K.S. Lyon. 1990. The Long-Term Adequacy of World Timber Supply. Resources for
       the Future, Washington, DC.  230 pp.

Sedjo,  R.A., and A.M. Solomon.   1989.  Climate and forests, pp. 105-120..  In: N. Rosenburg, W.E.
       Easterling III,  P.R. Crosson, and J. Darmstadter, eds.  Greenhouse Warming: Abatement and
       Adaptation.  Resources for the Future, Washington, DC.

Shands,  W.E., and J.S. Hoffman. 1987.  The  Greenhouse Effect, Climate  Change, and U.S. Forests.
       The Conservation Foundation, Washington, DC.

Sharma,  R.N., O.P. Vimal,  H.L. Shanna, and K.S. Rao, eds.  1989.  Proceedings of Bio-Energy Society
       Fifth Convention & Symposium '88.  Bio-Energy Society of India.  New Delhi, India.  486 pp.

Sharpe, J.H., R.D. Spence,  and  EJ.  Rykiel, Jr. 1989.  Diagnosis of sequential ozone effects on
       carbon assimilation, translocation, and allocation in cottonwood and loblolly pine. Technical
       Bulletin No. 565, National Council of the Paper Industry for Air and Stream Improvement, Inc.,
       New York, NY. 76 pp.

Shugart,  H.H. 1984. A Theory of Forest Dynamics. Springer-Verlag, New York, NY. 278pp.

Shugart,  H.H., M.Y. Antonovsky,   P.G. Jarvis, and A.P.  Sandford.  1986.  CO^ climatic change and
       forest ecosystems - Assessing the response of global forest to the direct effects of Increasing
       C02and climatic change, pp 475-521.  In: B. Bolin, B.R. Dob s, J. Jsger, and R.A. Warrick, eds.
       The Greenhouse Effect. Climatic Change, and Ecosystems. SCOPE 29.  John Wiley & Sons,
       Chichester.

Simpson,  G.G.  1964.  Species density of North American recent mammals. Syst. Zool. 13:57-73.
                                              145

-------
Sionit, N., B.R. Strain, H. Hellmers, G.H. Riecbers, and C.H. Jaeger.   1985.  Long-term atmospheric
        C02 enrichment affects the growth and development of Liquidamber styraciflua and Pinus taeda
        seedlings.  Can. J. For. Res. 15:468-471.

Sittig, M. 1979. Fertilizer Industry Processes, Pollution Control, and Energy Conservation. Noyes Data
        Corp., Park Ridge, NJ.

Smit, B., L. Ludlow, and  M. Brklacich.   1988.  Implications of a global climatic warming for agriculture:
        A review and appraisal.  J. Environ. Qual. 17:519-527.

Smith, D.M.  1986. The Practice of Silviculture. John Wiley & Sons, New York, NY. 527  pp.

Smith, J.B.,  and D. Tirpak, eds. 1989.  The Potential Effects  of Global Climate Change on the United
        States.  Report to Congress. EPA-230-05-89-050, U.S. Environmental Protection Agency,
        Washington, DC.  413 pp.

Smith, T. Personal communication.  Department of Environmental Sciences, University of Virginia,
        Charlottesville, VA.

Smith, "TM., R. Leemans,  W. Cramer, I.C. Prentice, A.M. Solomon,  and H.H. Shugart.   Submitted.
        Simulating  changes in the distribution of global terrestrial vegetation under C02-induced climate
        change: Comparison among scenarios based on general circulation models. Climate Change.

Solomon,  A.M.  1986.  Transient response of forests to CCyinduced climate change: Simulation
        modeling experiments in eastern  North America. Oecologia  68:567-579.

Solomon,  A.M.  In press. The taiga-tundra border in first detection of biospheric response to global
        environmental change. In: A. Shivendko, ed. Proceedings of Boreal Forest Symposium,
        Arkhanglesk, USSR.

Sprugel,  D.G.   1985.   Natural Disturbance and Ecosystem Energetics. In: S.T.A. Pickett and P.S. White,
        eds. The Ecology of Natural Disturbance and  Patch Dynamics.  Academic Press,  Inc., London.
        472 pp.

Stephenson,   N.L.  1990.  Climatic control of vegetation distribution: The role of the water  balance.
        Amer. Naturalist.  135(5):649-670.

Stith, J.L., L.F. Radke, and P.V. Hobbs . 1981.  Particle emissions and  the production of ozone and
        nitrogen oxides from burning of forest slash.  Atmos. Environ. 15:73-82.

Strain, B.R.  1987.  Direct effects of increasing atmospheric C02 on plants and ecosystems. Tree  2:18-
        21

Strain, B.R., and J.D.  Cure.  1985.  Direct effects of increasing carbon dioxide on vegetation.   DOE/ER-
        0238, U.S. Department of Energy, Carbon Dioxide Research Division, Washington, DC.

Surano, KJL, P.F.  Daley, J LJ. Houpis,  J.H. Shinn, J.A. Helms,  RJ. Palassou,  and M.P. Costella.
        1986. Growth and physiological responses of Pinus ponderosa Dougl. ex P. Laws, to long-term
        elevated C02 concentrations.  Tree Physiol.  2:243-259.

Swift, M.S.  1977.  The ecology of wood  decomposition.  Sci.    Prog., Oxford  64:179-203.
                                              146

-------
Tans, P.P., I.Y. Fung,  and T. Takahashi.  1990.  Observational constraints on the global atmospheric
       C02 budget. Science 247:1431-1438.

Tbeng, B.K.G.,  K.R. Tate, P. Sollins,  N. Moris, N. Nadkarni, and  RX. Tate, III  1989.  Constituents of
       organic matter in temperate and tropical soils.  In: D.C. Coleman, J.M. Oades and G. Uehara,
       eds. Dynamics of Soil Organic Matter in Tropical Ecosystems.  N'rftal Project, University of
       Hawaii Press, Honolulu, HI.

Tinus, R.W. 1972. C02 enriched atmosphere speeds growth of ponderosa pine and blue spruce
       seedlings. Tree Planters' Notes 23:12-18.

Tinus, R.W., and S.E.  McDonald.   1979.  How to grow tree seedlings in containers in greenhouses, pp.
       128-129. General Technical Report RM-60, USDA Forest Service, Rocky Mountain Forest &
       Range Experiment Station, Ft. Collins, CO.

Tissue, D.T., and W.C. Oechel.  1987.  Response of Eriophorum vaginatum to elevated C02and
       temperature in the Alaskan tussock tundra. Ecology 68(2):401-410.

Tolley, L.C., and B.R. Strain. 1984a.  Effects of C02 enrichment and water stress on growth of
       Liquidamber styraciflua and P/nus  faeda seedlings. Can. J. Bot. 62:2135-2139.

Tolley, L.C., and B.R. Strain. 1984b.  Effects of C02 enrichment on growth of Liquidamber styraciflua
       and Pini   • ^da seedlings under different irradiance levels. Can. J.  For. Res.  14:343-350.

Tolley, L.C., and             1985.  Effects of CO2 enrichment and  water stress on gas exchange of
       Liquidamber .•>.     . -. and Pinus  faeda seedlings grown under different irradiance levels.
       Oecologia  65:166-172.

Tranquillini, U. 1979. Physiological Ecology of the Alpine Timberline.  Springer-Verlag, Berlin.  137pp.

Trexler, M.C.  In press.  Reforesting the United States to Combat Global Warming.  World Resources
       Institute, Washington, DC.

Tuhkanen,  S.  1980.  Climatic parameters and indices in plant geography.  Acta Phytogogr. Suec.  67:1-
       110.

Tucker, CJ.,  I.Y. Fung,  CD. Keeling, and R.H. Gammon.  1986. Relationship between atmospheric
       C02 variations and a satellite-derived vegetation index.  Nature  319:195-199.

Uhl, C., and C.F.  Jordan.  1984.  Succession and nutrient dynamics following forest cutting and burning
       in Amazonia. Ecology  65(5):1476-1490.

United States  Department of Agriculture  (USDA).  1988. The South's Fourth Forest:  Alternatives for
       the Future. USDA Forest Resource Report #24. USDA, Washington, DC.  512 pp.

United States  Department of Agriculture  (USDA).  1990. Global Change Research Program Plan.
       USDA Forest Service. Science and Policy Associates, Washington, DC. 75 pp.
                                              147

-------
Urban, D.L., and  H.H. Shugart  1989.  Forest response to climatic change: a simulation study for
       southeastern forests.  In: J.B. Smith and D. Tirpak, eds.  The Potential Effects of Global Climate
       Change on the United States. EPA-230-05-89-050, U.S. Environmental Protection Agency,
       Washington, DC.

Van Cleve, K. Personal communication.  Forest Soils Laboratory, University of Alaska, Fairbanks, AK.

Vitousek,  P.M., P.R. Ehrlich, A.H. Ehrlich, and PA. Matson.  1986. Human appropriation of the
       products of photosynthesis.  BioScience  36:368.

Waring, R.H., and W.H. Schlesinger.   1985.  Forest Ecosystems: Concepts and Management.
       Academic Press, Inc., New York, NY. 340 pp.

Washington, W.M., and GA, Meehl. 1989. Climate sensitivity due to increased CO* Experiments with
       a coupled atmosphere and ocean general circulation model. Climate Dynamics.  4:1-38

Webb,  T.  1986.  Is vegetation in equilibrium with climate? How to interpret late-quaternary pollen data.
       Vegetatio  67:75-91.

Webb,  T.  1987.  The appearance and disappearance of major vegetational assemblages: Long-term
       vegetational dynamics in eastern North America. Vegetation 69:177-187.

Westberg,  H.,  K. Sexton,  and D. Flyckt.  1981.  Hydrocarbon production and photochemical ozone
       formation  in forest burn  plumes.  J. Air Pollut. Control Assoc. 3:661-664.

Westoby,   J. 1989.  Introduction to World Forestry.  Blackwell, New York, NY. 228pp.

Wetherald, R.T., and S. Manabe. 1988.  Cloud feedback processes in a general circulation model.  J.
       Atmos. Sci.  45:1397-1415.

Whittaker, R.H. 1975.  Communities and Ecosystems, 2nd Edition.  MacMillan Publishing Co., Inc., New
       York. NY.  385 pp.

Whittaker, R.H., and G.E. Likens.  1975. The biosphere and man, pp. 305-328.  In: Primary
       Productivity of the Biosphere. Ecol. Study No. 14, Springer-Verlag, Berlin.

Wlersum,  K.F., ed.  1984. Strategies and Designs for Afforestation, Reforestation and Tree Planting.
       Centre for Agricultural Publishing and Documentation (Pudoc), Wageningen. The Netherlands.
       432pp.

Williams, W.E., K. Garbutt, FA. Bazzaz,  and P.M.  Vitousek.  1986. The response Of plants to
       elevated C02: IV.  Two deciduous-forest tree communities.  Oecologia  69:454-459.

Wilson, E.G.  1988. The  current state of biological diversity,  pp. 3-18.  In: E.O. Wilson, ed., Biodiversity,
       National Academy Press, Washington, DC.

Wiiijum, J.K., and R.P. Nellson. 1989.  Forests, pp. 71-92.  In:  J.B. Smith and D. Tirpak, eds. The
       Potential Effects of Global Climate Change on the United States.  Report to Congress. EPA-230-
       5-89-050, U.S. Environmental Agency, Washington, DC.
                                              148

-------
Winjum, J.K., and  P.E. Schroeder , eds. In preparation.  Proceedings of International Workshop on
       Large-Scale Reforestation, May 9-10,1990, U.S. Environmental Protection Agency,
       Environmental Research Laboratory, Corvallis, OR.

Wood,  PJ.,  J. Burley, and A. Grainger. 1984. Technologies and Technology for Reforestation of
       Degraded Tropical Lands. Office of Technology Assessment, U.S. Congress, Washington, DC.
       114pp.

Woodward,  F.I.  1987.  Climate and Plant Distribution.  Cambridge University Press, London. 174 pp.

Woodwell, G.M. 1987. The warming of the industrialized middle latitudes. Unpublished manuscript,
       Workshop on Developing Policies for Responding to Future Climate Change, Villach, Austria.  15
       pp.

Woodwell, G.M., R.H. Whittaker, WA.  Reiners, G.E. Likens,  C.C. Delwiche, and D.T.  Botkin. 1978.
       The  biota and the world carbon budget. Science 199:141-146.

Woodwell, G.M., J.E.  Hobbie,  RA. Houghton,  J.M. Melillo, B. Moore, BJ. Peterson,  and G.R.
       Shaver.  1983.  Global deforestation:  contribution to atmospheric carbon dioxide.  Science
       222:1081-1086.

World  Resources  Institute.  1988.  World Resources 1988-89.  Basic  Books, Inc., New York, NY. 372
       PP.

World  Resources  Institute.  1990.  World Resources 1990-91.  Oxford University Press,  New York, NY.
       383  pp.

Wright, L.L. In press.  Short rotation woody crop opportunities to mitigate carbon dioxide buildup.  In:
       Proceedings, North American Conference on Forestry Responses to Climate Change. May 15-17,
       1990.  Climate Institute.  Washington, DC.

Yonker, CM., Schimel,  D.S. Paroussis, E. and R.D. Heil. 1988. Patterns of organic carbon
       accumulation in a semi-arid shortgrass steppe,  Colorado.  Soil Sci. Soc. Amer. J. 52:478-483.

Young, R.A. 1982. Introduction to Forest  Science.  John Wiley & Sons, New York, NY.  554pp.
                                              149

-------
8      APPENDICES

8.1     Appendix A; Extramural Research  Projects Sponsored  by ERL-C Global  Change Research

       Program in FY 90.
PRINCIPAL
INVESTIGATOR

L H. Allen
USDA/Agricultural Research
Service

Julio Betancourt
US Geological Survey

William Chang
University of Michigan

Wendall P. Cropper, Jr.
University of Florida
TITLE OF PROJECT

Temperature and C02 Interactions on Rates of
Development, Growth, and Yield of Rice
 FUNDING
MECHANISMS

IAG
Assist in the Organization of the 1990 Pacific Climate  IAG
(PACLIM) Workshop

The Effects of Long-Term Climate Changes on Large  Coop
Lake Basin Ecosystems in China

Modelling Carbon Dynamics of Slash Pine Plantations Coop
in Response to Climate Change Scenarios
James D. Hall
Oregon State University

James M. Hoell, Jr.
National Aeronautics
and Space Administration,
Langley Research Center

John Kineman
National Geophysical Data
Center,
World Data Center

Brian Lamb
Washington State University

Dennis P. Lettenmaier
University of Washington

Robert McKelvey
University of Montana

Ronald P. Neilson
Oregon State University

Richard J. Norby
Oak Ridge National Lab
Effect of Climate Change on Inland Fisheries
Coop
Emissions of CH,, and NMHC from Canadian Wetlands IAG
and Tundra Ecosystems
Co-develop Data, Tools, and Methods for            IAG
Characterization and Analysis of Environmental
System Patterns to Support the EPA's Global Climate
Research and Modeling

Biogenic Hydrocarbon Emissions and Global Climate  Coop
Change

Identification of Long-Term Water Resources Effects of Coop
Global Climate Change

Implication of the Loss  of Biological and Genetic      Coop
Diversity in a Regulatory Conte>ct

Mechanisms of Biome Response to Climate Change:  Coop
North American and Global
Interactions Between Elevated C02 and Drought
Stress in Tree Seedlings
IAG
                                            150

-------
John T. Rani                  Potential Impacts of Climatological Change on the    Coop
University of Idaho             Distribution of Endemic Plants and Animals

David W. Roberts              Assessing the Effects of Climate Change on Forest    Coop
Utah State University           Ecosystem  Dynamics

Steve Running                Coupling of Forest-BGC FORET Ecosystem Simulation Coop
University of Montana          Models for Projection of Forest Responses to Global
                             Climate Change

Herman Shugart               Unking Physiological Process Models to Forest Gap  Coop
University of Virginia           Models

Fredrick Swanson             Technology Transfer and Cooperative Research on    IAG
USDA/Paclfic NW Station       Western Forests

John M. Thomas              Technology Transfer and Cooperative Research in    IAG
Battelle/Pacific NW Lab        Support of EPA's Global Climate Change Program

Starley Thompson             Modeling the Effects of Global Climate Change on    Coop
National Center for Atmospheric Vegetation: Non-Interactive and Interactive Responses
Research

Ted Vinson                   Carbon Cycling in Arctic Tundra and Boreal Forest    Coop
Oregon State University        Ecosystems:  Responses and Feedbacks to Global
                             Climate Change
                                            151

-------
       Appendix
       1990.
B:   Participants in Carbon Sequestering  and Soils Workshop, February 27-28,
Dr. Chris Anderson
US EPA
Environmental Research Lab.
200 SW 35th Street
Corvallis, OR  97333

Dr. Dominique Bachelet
NSI Technology Services
US EPA
Environmental Research Lab.
200 SW 35th Street
Corvallis, OR  97333

Dr. Michael Beare
Institute of Ecology
University of Georgia
Athens, GA  30602

Dr. Stan Buol
Department of Soil Science
North Carolina State University
Raleigh, NC   29650

Dr. Bruce Caldwell
Dept. of Microbiology
Oregon State  University
Corvaliis, OR  97331

Dr. Dale Cole
College of Forest Resources
AR-10
University of Washington
Seattle, WA   98195

Dr. Kermit Cromack
Department of Forest Science
Forest Sciences Laboratory
Oregon State  University
Corvallis, OR  97331

Dr. Richard Dick
Department of Soil Science
Oregon State  University
Corvallis, OR  97331
                               Dr. John Duxbury
                               Dept. of Agronomy
                               Bradfield Hall
                               Cornell University
                               Ithaca,  NY

                               Dr. William Ferrell
                               Forest Science
                               Oregon State University
                               Corvallis. OR 97331

                               Dr. Cheryl Gay
                               NSI Technology Services
                               US EPA
                               Environmental Research Lab
                               200 SW 35th Street
                               Corvallis, OR 97333

                               Dr. Robert Griffiths
                               Dept. of Microbiology
                               Oregon State University
                               Corvallis, OR 97331

                               Dr. Hermann Gucinski
                               NSI Technology Services
                               US EPA
                               Environmental Research Lab
                               200 SW 35th Street
                               Corvallis, OR 97333

                               Ms. Kate Heaton
                               US EPA
                               Office of Policy Analysis
                               PM-221
                               Climate Change Division
                               401 M Street, SW
                               Washington, DC 20460

                               Dr. Richard  Houghton
                               The Woods Hole Research Center
                               P.O. Box 296
                               Woods Hole, MA  02543
                                             152

-------
Dr. Elaine Ingham
Dept. of Botany & Plant Path.
Oregon State University
Corvallis, OR 97331

Dr. Dale Johnson
Biological Sciences Center
Desert Research Institute
P.O. Box 60220
Reno, Nevada  89506

Mr. Jeffrey Kern
US EPA
Environmental Research Lab.
200 SW 35th Street
Corvallis, OR 97333

Dr. Ellis Knox
USDA/Soil Conserv.  Service
100 Centennial Mall N., Rm.345
Lincoln, NE  68508-3866

Dr. Robert Lackey
US EPA
Environmental Research Lab.
200 SW 35th Street
Corvallis, OR 97333

Dr. Rattan Lai
Department of Agronomy
The Ohio State University
2021 Coffey  Road
Columbus, OH  43210-1086

Dr. Duane Lammers
USDA/Forest Science
US EPA
Environmental Research Lab.
200 SW 35th Street
Corvallis, OR 97333

Dr. Leon Liegel
USDA Forest Service
US EPA
Environmental Research Lab
200 SW 35th Street
Corvallis, OR 97333
Dr. Jeff Lee
US EPA
Environmental Research Lab
200 SW 35th StreM
Corvallis, OR  97333

Dr. Ariel Lugo
Institute of Tropical Forestry
USDA Forest Service
So. Forest Experiment Station
Call Box 2500
Rio Piedras, PR  00928-2500

Dr. Kim Mattson
US EPA
Environmental Research Lab
200 SW 35th Street
Corvallis. OR  97333

Dr. David Myrold
Department of Soil Science
Oregon State University
Corvallis, OR  97331

Dr. William Parton
Natural Resource Ecology Laboratory
Colorado State University
Fort Collins, CO  80523

Dr.  Dave Perry
Department of Forest Science
Forest Sciences Laboratory
Oregon State University
Corvallis, OR  97331

Dr. Charles Peterson
NSI Technology Services
US EPA
Environmental Research  Lab
200 SW 35th Street
Corvallis, OR  97333

Dr. W. Mac Post
Environmental Sciences Division
P.O. Box 2008, Bldg.  1000
Oak Ridge National Laboratory
Oak Ridge, TN  37831-6335
                                             153

-------
Dr. Paul Rasmussen                             Dr. Richard Waring
USDA-ARS                                     Department of Forest Science
Columbia Basin Agricultural Research Center        Forest Sciences Laboratory
P.O.  Box 370                                   Oregon State University
Pendleton, OR  97801                           Corvallis, OR  97331

Dr. Paul Rygiewicz                               Dr. Jack Winjum
US EPA                                        Global Climate Team
Environmental Research Lab                      US EPA Envir. Res. Lab.
200 SW 35th Street                              200 SW 35th Street
Corvallis, OR  97333                             Corvallis, OR  97333

Dr. William Schlesinger
Department of Botany
Duke University
Durham, NC  27706

Mr. Paul Schroeder
NSI Technology Services
US EPA
Environmental Research Lab
200 SW 35th Street
Corvallis, OR  97333

Mr. Paul Shaffer
NSI Technology Services
US EPA
Environmental Research Lab.
200 SW 35th Street
Corvallis, OR  97333

Dr. Philip Sollins
Department of Forest Science
Forest Sciences Laboratory
Oregon State University
Corvallis, OR  97331

Dr. David Turner
NSI Technology Services
US EPA
Environmental Research Lab.
200 SW 35th Street
Corvallis, OR  97333

Dr. James  Trappe
Department of Forest Science
Forest Sciences Laboratory
Oregon State University
Corvallis, OR  97331
                                             154

-------
83     Appendix  C: Invited Participants, Workshop  on Ecological and Operational Considerations
       for Large-Scale  Reforestation,  Corvallis, Or. May 8-10,  1990


Dr. Francis Cailliez
Centre Technique Forestier Tropicale
Nogent sur Marne, France

Dr. A.N. Chaturvedi
TATA Energy Research Institute
New Delhi, India

Mr. Renato Moraes De Jesus
Florestas Rio Doce
Espirito Santo, Brazil

Mr. Joe Hughes
Weyerhaeuser Company
New Bern, North Carolina

Dr. Ian Hunter
Forest Research Institute
Rotorua,  New Zealand

Dr. Denis Lavender
University of British Columbia
Vancouver, British Columbia

Dr. Douglas Malcolm
University of Edinburgh
Edinburgh, Scotland

Dr. Peyton Owston/ Mr. Tom Tirpin
U.S. Forest Service
Corvallis,  Oregon

Dr. Mark Trexler
World Resources  Insitute
Washington, D.C.

Dr. Jack  Walstad
Oregon State University
Corvallis,  Oregon
                                               155

-------
8.4     Appendix D:  Abbreviations

CEES         Committee on Earth and Environmental Sciences
DOE          Department of Energy
EPA          Environmental Protection Agency
ERI.-C         Environmental Research Laboratory-Corvallis
FAO          Food and Agriculture Organization
FY            Fiscal year
GCM          General Circulation Model
GIS           Geographic Information System
GCRP         Global Change Research Program
GFDL         Geophysical Data Center
GISS          Goddard Institute for Space Science
IPCC          Intergovernmental Panel on Climate Change
MAI           Mean  annual increment
NEP          Net ecosystem productivity
ORD          Office of Research and Development
OSU          Oregon State University
OTA          Office of Technology and Assessment
PAR          Photosynthetically active radiation
PET           Potential evapotranspiration
TSI            Timber stand improvement
UKMO         United Kingdom Meteorological Organization
USAID         United States Agency for International Development
USDA         United States Department of Agriculture
USFS         USDA Forest Service
WUE          Water-use efficiency
                                            156

-------